Public Disclosure Authorized

Public Disclosure Authorized

Public Disclosure Authorized

Public Disclosure Authorized

G
___

__ Woi

-, 1012 Report~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~:-

D =-- = -F_e a _

*~~~~~~~~--

an u&ai

10123

-- Z_wL~

LIII~~~~

/ U

,-

ElectricityPricing
Theory and Case Studies

Mohan Munasinghe and Jeremy J. Warford

Published for the World Bank The Johns Hopkins University Press Baltimore and London

Copyright C 1982 by The International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, N.W., Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America The Johns Hopkins University Press Baltimore, Maryland 21218, U.S.A. The views and interpretations in this book are the authors' and should not be attributed to the World Bank, to its affiliated organizations, or to any individual acting in their behalf. The five maps in this book have been prepared exclusively for the convenience of readers; the denominations used and the boundaries shown do not imply, on the part of the World Bank and its affiliates, any judgment on the legal status of any territory or any endorsement or acceptance of such boundaries.

Library of Congress Cataloging in Publication Data Munasinghe, Mohan, 1945Electricity pricing. Bibliography: p. 368 Includes index. 1. Electric utilities-Asia, HD9685.A77S66 ISBN 0-8018-2703-5

Southeastern-Rates. I. Warford, Jeremy J. [1. Title. 338.4'336362 8147613 AACR2

Contents
Preface Definitions Part One. Theory Chapter 1. Objectives and.Pricing Framework Background 3 Purpose of the Study 7 Structure of the Study 8 Objectives of an Electric Power Tariff 10 Tariffs Based on Long-Run Marginal Costs Developing Practical Tariff Structures 12 Summary 13 Chapter 2. Economics of Marginal Cost Pricing Basic Marginal Cost Theory 15 Capital Indivisibilities and Peak Load Pricing Extensions of Simple Models 19 Chapter 3. Prerequisites for Marginal Costing Load Forecasting 27 System Planning 34 Shadow Pricing 42 Chapter 4. Strict Long-Run Marginal Cost Cost Categories and Pricing Periods 52 Marginal Capacity Costs 53 Marginal Energy Costs and Losses 57 Customer Costs 58 Hydroelectric Systems 59 Chapter 5. Adjusting the Strict Long-Run Marginal Cost Second-Best Considerations 66 Subsidized or Lifeline Prices 67 Optimal Electricity Pricing in a Distorted Economy 69 Financial Viability 73 Other Objectives and Constraints 76 Metering, Billing, ancdCustomer Comprehension 79 Chapter 6. Recent Experience with Modem Pricing Structures Types of Tariff Structures 83 Recent Empirical Evidence 87
v

xiii xvii I 3

11

15 17 25

52

66

83

Vi

CONTENTS

Part Two. Case Studies Chapter 7. Overview of the Case Studies Institutional, Technical, and Physical Characteristics System Losses 101 System Planning 102 Financial Viability 102 Divergences from the Strict LRMC 103 Special Metering 104 Market Distortions 104 105 Second-Best and Political Considerations Governmental Attitudes toward Reforms 105 Chapter 8. Indonesia Organization of the Sector 109 Computation of the Strict LRMC for Java Adjustment of the Strict LRMC 128 Policy Issues and Conclusions 138 Chapter 9. Pakistan Organization of the Sector 140 Computation of the Strict LRMC 145 Adjustment of the Strict LRMC 158 165 Policy Issues-and Conclusions Chapter 10. Philippines 167 Organization of the Sector Computation of the Strict LRMC 174 188 Adjustment of the Strict LRMC 198 Policy Issues and Conclusions Chapter 11. Sri Lanka 205 Organization of the Sector Computation of the Strict LRMC 211 217 Adjustment of the Strict LRMC Chapter 12. Thailand Organization of the Sector 223 231 Computation of the Strict LRMC 236 Adjustment of the Strict LRMC 240 Policy Issues and Conclusions Appendixes Appendix A. Seminar Participants Appendix B. Power Sector Statistics for Developing Countries 99

97 99

107 116

140

167

205

223

249 251 259

CONTENTS

vii

Appendix C. Allocation of Capacity and Energy Costs between Peak and Off-peak Consumers Appendix D. STLRMC: A Simple Computer Program for Estimating Strict Long-Run Marginal Costs Calculation of Loss Multipliers 331 Calculation of Marginal T&D Capacity Cost 333 Calculation of Marginal Energy Cost 334 Calculation of Marginal Generation Capacity Cost 335 Shadow Pricing 336 Listing of the STLRMC Program 338 Sample Outputs from the STLRMC Program for the WAPDA System in Pakistan 357 References Index
TABLES

327
331

368 375

1-1. 1-2. 4-1. 6-1. 8-1. 8-2. 8-3. 8-4. 8-5. 8-6. 8-7A. 8-7B. 8-8. 8-9. 8-10. 8-11.

Worldwide Electricity Statistics, 1978 4 Power Sector Capacity, Investment, and Commercial Energy Requirements in the Developing Countries, 1981 to 1990 Marginal Generating Costs in an All-Hydro System Examples of Peak Load Tariffs 88 62

Forecasts of Future Generation Requirements 111 Demand and Loss Forecast at Time of System Peak for Java 112 Incremental Demand Forecast at Time of System Peakfor Java 112 Average and Peak Loss Factors for Energy and Power at Time of System Peak for .ava 114 Generation Capacity Costfor Java (Saguling Hydro Plant First Unit of 175 Megawatts) 118 Calculation of Generation Capacity Costs for Java 120 Transmission Capacity Costs Based on the EHV/HV Investment Plan 121 Distribution Capacity Costs Based on the MV and LV Investment Plan 122 Summary of Incremental Capacity Costs for Java 124 Summary of Peak Capacity Costs for Java 124 Calculation of Operation and Maintenance Costs for Java 126 Summary of Operation and Maintenance Costs for Java 127

viii

CONTENTS

8-12. 8-13. 8-14. 8-15. 8-16. 9-1. 9-2. 9-3. 9-4. 9-5. 9-6. 9-7. 9-8. 9-9. 9-10. 9-11. 9-12. 9-13. 9-14.
10-1.

Energy Generation Cost Calculation for Table 8-13 Energy and Fuel Cost 129 Strict LRMC for Java 130 Equivalent Strict LRMC for Java 132 Typical New Tariff 136

128

Maximum Demand, Generation, and Sales, 1970-71 to 1976-77 141 1 Forecast of Maximum Demand, Generation, and Sales, 1977-78 to 1982-83 143 WAPDA Generation Expansion Program 144 Costs of Generation Capacity at the Busbar 148 WAPDA Investment Program, 1980 to 1990 149 Projected System Peak Demand at EHV/HV, MV, and LV, 1979 to 1990 150 Incremental Peak Demand at EHV/HV, MV, and LV, 1979 to 1990 152 Marginal Energy Costs 153 WAPDA System Loss Factors, 1979 to 1990 154 Customer Costs 156 Capacity Costs 156 Summary of Strict Long-Run Marginal Capacity and Energy Costs (DP) 157 Summary of Main Features of the Existing WAPDA Tariffs 160 LRMC and Proposed New WAPDA Power Tarifffor 1980-81 162
NPC

System Operations for 1977

172

10-2. Physical Targets in NPC'S Power Expansion Program, 1978 to 1987 172 10-3. Annual Incremental Capacity Cost at the HV Levelfor NPC 176 10-4. Incremental Energy-Related Costs at the HV Levelfor NPC 177 10-5. Marginal Cost Structure for NPC 178 10-6. Annual Incremental Capacity Cost for Private Electric Utilities 181 10-7. Annual Incremental Energy-Related Costs for Private Electric Utilities 181 10-8. Annual Incremental Customer-Related Costs at the LV Level for Private Electric Utilities 182 10-9. Marginal Cost Structure at the LV Levelfor Private Electric Utilities, from the Utilities' Viewpoint 182

CONTENTS

ix

10-10. Marginal Cost Structure Cost Allocation for Private Electric Utilities, from the Economy's Viewpoint 183 10-11. Annual Incremental Capacity Cost at the LV Level for Electric Cooperatives 184 10-12. Incremental Energy-Related Costs at the LV Levelfor Electric Cooperatives 185 10-13. Annual Incremental Customer-Related Costs for Electric Cooperatives 186 10-14. Private Marginal Cost Structure at the LV Levelfor Electric Cooperatives, from the Cooperatives' Viewpoint 187 10-15. Private Marginal Cost Structure at the LV Level for Electric Cooperatives, from the Economy's Viewpoint 187 10-16. Average ElectricityTariffs in 1978 189 10-17. Financial Cost Structure of Energy Sales for NPC, 1978 191 10-18. Financial Cost Structure of Energy Sales for Private Electric Utilities, 1978 192 10-19. Financial Cost Structure of Energy Salesfor Electric Cooperatives, 1978 193 10-20. Average Tariffsfor NPC, 1978 194 10-21. Allocation of Chargesfor NPC 195 10-22. Average Tariffs fbr Private Electric Utilities, 1978 196 10-23. Allocation of Charges for Private Utilities 196 10-24. Average Tariffsfor Electric Cooperatives, 1978 197 10-25. Allocation of Charges for Electric Cooperatives 197
11-1. CEB Energy and Power Balance 210 11-2. Peaking Hydro Capacity Costs 213 11-3. Load, Demand, and Loss Forecasts 214 11-4. Incremental Load or Demand Forecast 214 11-5. HV Transmission Investment Costs 216 11-6. MV Distribution Investment Costs 217 11-7. LV Distribution Investment Costs 218 11-8. Incremental Costs Summary (at Source) 219 11-9. Marginal Energy Costs 219 11-10. Summary of Strict LRMC of Capacity and Energy (BP) 11-11. Standard Costs for Service Connections 220 11-12. Summary of Existing Tariff 222

220

12-1. Number of Customers of the Electric Power Sector, by Type of Customer 227 12-2. Energy Consumption, by Type of Customer 227 12-3. Sales of Power and Energy by EGAT to Consumers 230 12-4. Marginal Capacity Costs for EGAT 233 12-5. Marginal Capacity Costsfor MEA 235 12-6. Marginal Capacity Costs for PEA 235

CONTENTS

Consumer-Related Costs for MEA and PEA (with Shadow Prices) 236 12-8. Total Marginal Cost of EGAT, MEA, and PEA (without Shadow Prices) 237 12-9. Total Marginal Cost of EGAT, MEA, and PEA 238 12-10. Adjusted Marginal Cost of EGAT, MEA, and PEA 243 12-11. Average Upper and Lower Limit of Poverty Band of Representative 244 Households, Adjusted by 1977 Consumer Price Index 12-7.
FIGURES

2-1. 2-2. 2-3. 2-4. 2-5. 3-1. 3-2. 3-3. 3-4. 4-1. 4-2. 4-3. 4-4. 4-5. 5-1. 5-2. 5-3. 6-1. 8-1. 8-2. 8-3.

Supply and Demand for Electricity Consumption 16 The Effect of Capital Indivisibilities on Price 18 Peak Load Pricing Model 19 Use of Price Feedback in Estimating Tariffs Based on LRMC 20 Relation between Outage Costs, Supply Costs, and Total Costs at the Optimum Reliability Level 22 Schematic of a Simple Electric Power System 28 Power and Energy Flows, Consumption, and Losses in a Typical Power System 30 Size and Timing of Additional Generating Plants Needed to Meet the Peak Demand for Power 38 Types of Generating Plants Needed to Meet the Annual Load Duration Curve 39 Typical Annual Load Duration Curve 54 Forecast of Demandfor Peak Power 54 Annual Variation of New Water Flows into the Reservoir in an All-Hydroelectric System 60 Typical Daily Load Duration Curves for a Mixed Hydro-Thermal System 63 Imputed Value of Water in a Mixed Hydro-Thermal System 64 Economic Basis for the Social or Lifeline Rate 68 Supply and Demand for Electricity 70 The Metering Decision to Implement a Two-Period Time-of-Day Tariff 80 Structure of Increasing and Decreasing Block Tariffs Typical Daily Load Curves for Java 115 Typical Daily Load Duration Curves for Java 134 Daily Load Profile for a LV Consumer 116 84

CONTENTS

Xi

9-1. 9-2.

Daily Load Curve for December 20, 1978 146 Annual Load Duration Curve for the WAPDA Grid System, 1979 147

11-1. Daily Load Curves for CEB, September 1978 208 11-2. Synthesized Annual Load Duration Curve, 1978 209 11-3. CEB Energy and P'ower Balance of Supply and Demand 12-1. 12-2. 12-3. C- 1.
MAPS

212

Characteristics of the EGAT System 228 Diagram of the E(3AT System 231 Flowchart for Calculation of the Strict LRMC Plant Costs and Annual Load Duration Curve

242 328

1. Indonesia (Java): Principal Power Stations and Transmission Lines in the PLN System 108 2. Pakistan: Principal Power Stations and Transmission Lines in the WAPDA System A142 3. Philippines (Luzon Grid): Principal Power Stations and 230 KV. Tranmission Lines in the NPC System 168 4. Sri Lanka: Principal Power Stations and Tranmission Lines in the CEB System 206 5. Thailand: Principal Power Stations and Transmission Lines in the EGAT System 224

Preface

INCREASING COST OF ELECTRIC POWER in recent years has brought about a growing awareness of the importance of pricing policies in maximizing the net economic benefits of consumption and avoiding waste. Although the theoretical lprinciplesgoverning optimal pricing strategies have long been understood, their practical application to the energy sector in general, and to electric power in particular, has not been seriously pursued until recently. Increasing unit costs, however, have provided the stimulus for a change in approach, and there is now a good deal of evidence that marginal cost pricing-hitherto thought to be a somewhat academic concept-is becoming accepted as an important criterion that should be considered in determining electric power tariffs. A previous World Bank publication, Electricity Economics, by Ralph Turvey and Dennis Anderson (1977), has helped to encourage this change in attitude. The theory and practice of marginal cost pricing covers far too much ground to be dealt with in detail within a volume of this size. Accordingly, we have had to make two principal compromises. First, in the attempt to maintain the balance between theory and practice, the methodology is developed with the emphasis on its operational usefulness. The basic theory of marginal cost pricing is explained, but readers are referred to other sources for the finer points that are more likely to be of interest to purists. Indeed, practitioners of marginal cost pricing and policymakers recognize that lack of data and other constraints will invariably limit the applicability of the more esoteric theoretical aspects, especially in developing countries. The second and related compromise concerns the relative attention paid to the two stages of tariff setting-first, the estimation of the strict long-run marginal costs (LRMC) to meet the objective of economically efficient pricing, and next, the adjustments to the strict LRMC required to derive a realistic tariff schedule that satisfies other constraints, such as financial requirements, social-subsidy considerations, fairness, and metering and billing difficulties. We feel that the adjustment process is at least as important as the calculation of the strict LRMC in setting practical tariffs in developing countries. Regrettably, the past literature has concentrated mainly on the strict LRMC. Therefore, this book seeks only to explain and interpret clearly the existing theoretical basis for estimating the strict LRMC and to outline a practical methodology for applying it. The principles and practice of adjusting the THE RAPIDLY

xiii

XiV

PREFACE

strict LRMC are, however, treated in greater detail, including the development of a new analytical framework for making economic second-best adjustments and estimating subsidized lifeline prices for poor consumers. This tradeoff between analytical rigor and practical applicability, on the one hand, and between calculating and adjusting the strict LRMC, on the other hand, are reflected in both the methodology and case study sections. In the final analysis, we have deliberately given more weight to practical issues that are of greater interest to tariff analysts and policymakers, who are likely to be our principal audience. This book consists of two parts. The first part contains a summary of the economic principles underlying marginal cost pricing for electric power, and emphasizes the importance of the adjustments that need to be made to the strict LRMC to reflect the various economic, social, and engineering objectives and constraints that are actually faced by policymakers in the energy sector. The second part consists of case studies prepared in collaboration with the staff of power and energy authorities in five developing countries in Asia. These studies were associated with the first two seminars in a series that is being conducted by World Bank staff in various parts of the world. The case studies all follow a two-stage procedure in which the LRMC of electric power is used to weigh the costs and benefits of other policy objectives that might be addressed through the medium of power pricing. LRMC is therefore treated largely as a benchmark by which other economic and social objectives may be consciously judged. It is our opinion that local understanding of these factors is absolutely crucial for successful tariff setting. The pervasive effects of electricity pricing policies clearly indicate that the traditional narrowly focused public utility approach, which concentrates almost exclusively upon financial and engineering analysis to the detriment of economic and social criteria, is inadequate. The concerns expressed in the case studies are ample evidence that this is so, and it is to be hoped that the approach they follow, which can readily be adapted to energy pricing studies in general, will be of value to power and energy authorities in other countries in addressing the complex and sensitive issues relating to electricity pricing policy. The lists of persons to whom the authors are indebted for assistance in the completion of this book is long. Most important, we would like to thank all the participants in the seminars- their names are listed in Appendix A. Among them we wish to give special mention to major contributors to the case studies: Soejitno Sujanto, Djaya Santoso, Muhammed Jahangir, K.K.Y.W. Perera, Nelson Wijemanne, S. M. Christy, Pratin Pathanaporn, and Viset Choopiban. Other contributors who could not attend the seminars, but to whom we are equally indebted, are Malaine Manzo Valenzuela and Iqbal Kahn, as well as Mark Gellerson, who helped to prepare one of the

PREFACE

XV

case studies. The assistance of Rajesh Pradhan, Karl Jechoutek, Kam Kalaji, and Cecile Rivera in preiparing two of the appendixes is acknowledged with thanks. We are also grateful for the encouragement and technical advice received from many colleagues at the World Bank, including Yves Rovani, Richard Sheehan, James Fish, John Davis, Edwin Moore, Dennis Anderson, Anandarup Ray, Jack Beach, Frank Lamson-Scribner, Rafael Moscote, Bemard Montfort, Esref Erkmen, Karl Stichenwirth, John Sneddon, Cheruvari Chandran, Vatsal Thakor, John Vance, Colin Warren, Ibrahim Elwan, Karl Schmedtje, Joe Gilling, and Bahman Abadian. We wish to thank Judy Barry for editorial assistance and Sibo Kong, Noriko Clark, Andie Dufigan, and Connie Villacorte for secretarial help. Among those outside the Bank, Ralph Turvey, Fred McCoy, and Rene Males offered helpful comments while Jan Paul Acton and Allen Miedema provided useful information on recent tariff studies in the United States. Virginia deHaven Hitchcock edited the manuscript for publication. Brian J. Svikhart directed design and production. Chris Jerome read and corrected proofs. Raphael Blow prepared the figures. The maps were compiled by Yung Koo and Josephine Cullen Dugan and drawn by Larry A. Bowring under the supervision of the World Bank's Cartography Division. Ralph Ward and James Silvan indexed the text. Joyce C. Eisen designed the cover. The responsibility for all errors and omissions is, of course, ours. Finally, we owe much to our families-our respective wives Sria and Beryl, and children, Anusha and Ranjiva, and Susan and Ian-for their patience and understanding while this book was being written. The volume is dedicated to our children.
MOHAN MUNASINGHE JEREMY WARFORD

Definitions
A&G AIC ARI B BP
BTU

CCF
CEB

CES CF c.i.f. D DP ECF

EGAT EHV

ESWR f.o.b. FTER GT HSD

HV IAEA IEE IEEE KESC

LDC LF LOLP LRMC

LV

MC MCB
MEA MECO

MOC MSB MSC MU

Administrative and general costs Average incremental cost Accounting rate of interest Baht Border price British thermal unit Consumption conversion factor Ceylon Electricity Board (Sri Lanka) Constant elasticity of substitution Conversion factor Cost, insurance, freight Demand Domestic price Electricity conversion factor Electricity Generating Authority of Thailand Extra high voltage Efficiency shladow wage rate Free on board Free trade exchange rate Gas turbine High speed diesel High voltage International Atomic Energy Authority Institution of Electrical Engineers (U.K.) Institute of Electrical and Electronic Engineers (U.S.) Karachi Electric Supply Corporation (Pakistan) Load duration curve Load factor Loss of load probability Long-run marginal cost Low voltage Marginal cost Border-priced marginal cost Metropolitan Electricity Authority (Thailand) Manila Electric Company (Philippines) Marginal opportunity cost Marginal social benefit Marginal social cost Marginal utility
xvii

xviii

DEFINITIONS

MV

MW NB
NEA NPC

OC OCC OER O&M p

PEA

PF
PLN

R RM Rp Rs SC SCF SER SRMC SWR TB TC T&D TOU

WAPDA

Medium voltage Megawatt Net benefit National Electrification Administration (Philippines) National Power Corporation (Philippines) Outage cost Opportunity cost of capital Official exchange rate Operation and maintenance costs Peso Provincial Electricity Authority (Thailand) Power factor Perusahaan Umum Listrik Negara (Indonesia) Reliability Reserve margin Rupiah Rupee Supply cost Standard conversion factor Shadow exchange rate Short-run marginal cost Shadow wage rate Total benefit Total cost Transmission and distribution Time of use Water and Power Development Authority (Pakistan)

PartOne

Theory

Chapter1

Objectives and Pricing Framework

ELECTRIC POWER IS A CRUCIAL FORM OF ENERGY in the world today. In 1980

about 950 terawatt-hours of electricity were generated in the developing countries. Production of electrical energy in the third world will continue to grow at an average rate of 8.5 percent annually, and will require investments of approximately $414 billion during 1980-89.' In comparison, about 2,350 terawatt-hours of electricity were generated in the United States in 1980. During 1980-89 consumption in the United States will grow at an average rate of a little more than 4 percent, requiring about $430 billion to meet the investment requirements.2

Background
A summary of selectedlstatistics for the electric power sector in the developing countries and other regions of the world during 1978 is given in Table 1-1. The third world contains more than 2 billion inhabitants, or about half the world's population, having an average per capita income of approximately $650-less than a third of the world average. Access to electricity in the developing countries is poor because more than two-thirds of the population live in rural areas that are sparsely electrified. In 1978 the developed countries produced more than seven times as much electrical energy as the developing countries although their population was only about one-third as large. The per capita consumption of electricity in the third world, therefore, is about one-twentieth the corresponding value for the developed countries and less than a fifth of the world average. Similarly, the installed generating capacity and the total commercial energy consumption in the developing countries lags far behind that for the rest of the world. This underlines the well-known correlation between per capita income and energy use.
1. Investment figures in this chapter are given in constant 1980 U.S. dollars. See Appendix B for details of electricity statistics in developing countries. 2. U.S. Department of Energy, "Report No. DOE/RG-0036" (Washington, D.C. June 1980; processed); and "Thirtieth Annual Electrical Energy Forecast," Electrical World (September 15, 1979), pp. 75-84. In this book "billion" is used to mean "thousand million." 3

THEORY

Table 1-1. Worldwide Electricity Statistics, 1978 Developed Developing market countries economies
30 70 2.08 650 29 696 377 307 12 331 177 105 70 2 0.085 75 25 0.78 8,000 >90 5,070 3,616 982 472 6,509 1,265 900 270 95 1.624

Item
Population Urban (percentage of total) Rural (percentage of total) Total (billions) Gross national product (GNP) per capita (dollars) Access to electricity (percent), Electricity generation Total (terawatt-hours) Thermal (terawatt-hours) Hydro (terawatt-hours) Nuclear (terawatt-hours) Per capita (kilowatt-hours) Installed capacity Total (gigawatts) Thermal (gigawatts) Hydro (gigawatts) Nuclear (gigawatts) Per capita (kilowatts) Annual commercial energy useb Total (million metric tons of oil equivalent) Per capita (metric tons of oil equivalent) Tons of oil equivalent per thousand dollars of GNP Electricity generated/commercial energyb Electricity generated per dollar of GNP (kilowatt-hours) Average annual growth of electricity consumption, 1973-78 (percent) Average electricity price (cents per kilowatt-hour), n.a. Not available.

Centrally planned economies

35 65 1.36 1,150 n.a. 1,851 1,534 269 48 1,365 387 308 69 10 0.285

World
40 60 4.24 2,100 n.a. 7,617 5,527 1,558 532 1,803 1,829 1,313 409 107 0.433

689 0.33 0.51 0.253 0.51 8.12 3.4

3,613 4.63 0.58 0.351 0.81 3.49 n.a.

2,004 1.47 1.28 0.231 1.19 6.44 n.a.

6.304 1.49 0.71 0.302 0.86 4.55 n.a.

Note: See Appendix B for a list of countries in each region and more disaggregate data for the
developing countries. See list of definitions for unit equivalencies. I Access to electricity defined by fraction of total population living in areas where electricity services are provided. I Commercial energy includes oil, coal, gas, hydroelectric, nuclear, and geothermal sources. I million metric tons of oil equivalent = 101 metric tons of oil equivalent = 1.47 million metric tons of coal equivalent = 41 x 10"2 BTU. I terawatt-hour of electricity a year is assumed to be equivalent to 0.25 million metric tons of oil equivalent of fossil fuel, based on the conversion efficiency of an average thermal power plant. c Average price = sales revenue/electrical energy consumed. Source: U.N. and World Bank data.

OBJECTIVES

AND PRICING

FRAMEWORK

In the developing countries electricity consumption grew at an average of more than 8 percent a year during 1973-78-more than twice the rate for the developed countries. As development continues, the future demand for electricity in the third world will tend to grow faster than the world average. In the past, the convenience and relatively low price of electricity in most developing countries helped to promote electricity use. As electricity and other energy prices rise during the next decade, however, the growth of demand is unlikely to accelerate further. The third world's investment in the power sector and its fossil fuel requirements for the next decade are summarized in Table 1-2. The table shows how capital intensive the power sector is. During the next decade, the installed capacity in the developing countries will more than double, requiring investments averaging more than $40 billion a year. The real costs of a kilowatt will continue to rise for several reasons. These include the shift to more costly coal and nuclear plants following the oil crisis, the reduced availability of cheaply exploitable hydroelectric sources, and the inability to realize further significant economies of scale, particularly as systems expand into regions of lower consumer density, such as small towns and rural areas. Commercial energy requirements for producing electricity, including fossil fuels used in thermal generation, will more than double in the next ten years, growing at an average rate of about 8.5 percent a year. The share of oil-fired generation in total thermal generation is expected to decrease significantly, however, from 57 percent in 1980 to 39 percent in 1990. Although nuclear and geothermal generation will grow dramatically, their contribution in 1990 will still be relatively small because of their small starting base in 980. The large investments in, and rising costs of, power have highlighted the need for increased economic efficiency in the electricity sector. Traditionally, the greatest emphasis has been placed on improving technical and financial efficiency through least-cost, long-range planning for system expansion, optimizing short- and medium-term system operation, and providing better management of utilities. These considerations on "the supply side" of the supply-demand equation have been tackled with a remarkable degree of success, usually by engineers using technically and financially oriented solutions. More recently, power economists have focused attention on the objectives of economic efficiency in the national context. These developments have had a growing and significant effect on "the demand side," chiefly through the application of appropriate tariff policies based on the principles of marginal cost pricing. On the supply side as well, the acceptance of economic arguments is reflected in the use of economic opportunity costs. These costs are represented by shadow prices that are the true costs of economic resources and not purely financial or accounting costs.

THEORY

Table 1-2. Power Sector Capacity, Investment, and Commercial Energy

Requirements in the Developing Countries, 1981 to 1990 Plant type Item
Power sector requirements Installed capacity, 1980 (gigawatts) Unit cost (dollars per kilowatt)a Capacity increments (gigawatts) 1981-85 1986-90 1981-90 Investment required (billions of dollars) 1981-85 1986-90 1981-90

Thermal

Hydro Nuclear Geothermal Total

150 1,160 62.4 81.7 144.1

101 1,730 47.4 54.3 101.7

4 1,920 6.8 27.9 34.7

2,100 1.0 0.9 1.9

255
-

117.6 164.8 282.1

72 95 167

82 94 176

13 54 67

1 2 4

169 245 414

Commercial energy required to produce electricity (million barrels a day of oil equivalent)b 1980 2.54 (56.7;1l.4;31.9)c 1985 3.90 (47.2;17.9;34.9), 1990 5.34 (38.6;21 .2;40.2)c Growth rate 1981-90 (percent) 7.7 (3.7;14.6;10.3)d

2.10 3.13 4.11 6.9

0.08 0.27 1.20 31.1

0.01 0.05 0.08 23.1

4.73 7.35 10.73 8.5

Note: See Appendix B for a list of countries amd more disaggregate data. All monetary values are given in 1980 constant U.S. dollars. - Not applicable. I Includes incremental cost of transmission and distribution facilities. bCommercial energy includes oil, coal, gas, hydroelectric, nuclear, and geothermal sources. I million barrels a day of oil equivalent = 50 million tons of oil equivalent a year. All nonthermal electricity generated transformed into millions of barrels a day of oil equivalent using the conversion efficiency of an average thermal power plant: I terawatt-hour a year = 109 kilowatthours a year = 0.25 million metric tons of oil equivalent = 0.005 million barrels a day of oil equivalent. c Percentage shares of oil, gas, and coal-lignite are given in parentheses. I Growth rates of oil, gas, and coal-lignite are given in parentheses. Source: World Bank data.

OBJECTIVES AND PRICING FRAMEWORK

Purpose of the Study

This book focuses on the importance of adopting correct power pricing policies to maximize the net economic benefits of electricity consumption to society. It reports on the progress of the pricing reform program in the electric power sector that is underway in the developing world with the active encouragement of the World Bank. A valuable contribution to this effort was the book Electricity Economics by Turvey and Anderson, which helped stimulate an awareness of the importance of reflecting long-run marginal 3 economic costs in pricing power in developing countries. The further stimulus of rapidly increasing electricity costs has caused this approach to be better understood by power authorities in the third world than in many industrialized countries. More specifically, the Bank's involvement in electric power project and sector work in the developing countries and the continuing exchange of information on pricing issues with utility companies have resulted in the synthesis and practical application of the pricing principles described in this book. Convincing national authorities that pricing based on marginal cost is not an academic exercise also has been important. Pricing policy is a valuable "soft" tool for managing demand and reducing electricity consumption. The effects of correct pricing policies are enhanced by coordinating their use with the other soft techniques of demand management, such as tax and financial incentives and education and propaganda, as well as with the hard methods of demand management, including direct control of loads, curtailment, and so forth. The integrated use of soft and hard techniques of demand management is most important because the former are more effective in the long run, whereas ihe latter have greater effect in the short run. Because of the lack of data and other constraints, the search for excessively precise cost estimates and pricing policies is likely to be difficult and often even counterproductive. Utilities have been encouraged to make "back of the envelope" cost estimates based on relatively simple considerations. These, however, may be a large improvement over traditional costing methods, because they at least indicate trends in economic cost. It is also necessary to instill confidence in power sector institutions in the developing countries to conduct such analyses and to persuade them to refrain from excessive reliance on foreign experts. External consultants may lack the understanding or "feel" for the developmental issues that are so critical in creating truly effective power pricing policies. All too frequently, the reports are
3. Ralph Turvey and Dennis Anderson, Electricity Economics (Baltimore, Md.: Johns Hopkins University Press, 1977).

THEORY

not applicable and end up on dusty bookshelves-a costly experience for developing countries. The approach of "learning by doing" has been pursued by the power pricing authorities of developing countries with considerable success. This process is considered absolutely essential, given the potentially important role that power pricing policy can play. The effects of the self-help approach are pervasive. Clearly, understanding domestic economic and social objectives and constraints is an indispensable part of the successful tariff analysis. A team of local engineers, economists, and financial analysts is more likely to have this understanding than a foreign consultant. This is particularly true in the third world, in which the range of economic and social policy tools at the disposal of governments is so limited and in which public revenues are at a premium. Also, in view of the rapid changes that may take place in a country's official economic and social objectives, the value of the self-help approach is clear. The government machinery must be able to update and revise tariff analysis constantly to accommodate such changes. In brief, tariff revision and policy should not be treated as a once-and-for-all, ad hoc effort to be carried out by external consultants, but rather as an important, continuous in-house activity. Therefore, the "learning by doing" or "institutionbuilding" approach that the Bank has actively encouraged is of critical importance to all developing countries.

Structure of the Study

The first part of the book begins with a discussion of the many and often conflicting national and sectoral objectives of electricity pricing. Next, a two-stage procedure for setting tariffs is outlined that meets these objectives. First the strict long-run marginal costs (LRMC) of electricity supply are computed to satisfy the criterion for economic efficiency. Then, the strict LRMC is adjusted to arrive at an appropriately realistic tariff structure that meets various other goals and constraints, including social-subsidy considerations for poor consumers, the financial viability of the electric power sector, simplicity of metering and billing, and so on. Part One continues with an overview of the economic principles underlying marginal cost pricing, followed by a discussion of how these basic models may be modified to satisfy more complex situations in the real world. Next, problems involved in the second stage of making tariffs are discussed, that is, adjusting the strict marginal cost price to reflect systematically the various economic, social, and engineering constraints that must actually be faced. Relatively greater emphasis is placed on this stage of tariff setting. Thus a systematic procedure based on shadow prices is proposed, which makes economic second-best adjustments to compensate for distor-

OBJECTIVES

AND PRICING

FRAMEWORK

tions in the prices of other goods and services in the economy. In particular, given the importance of integrating the pricing of alternative forms of energy, such as oil, gas, coal, and traditional fuels, in a consistent manner, power prices need to be modified because of distortions in the prices of other fuels. Ideally, reform of electric power pricing policy should be considered as only one aspect of a comprehensive analysis and revision of all energy prices.4 Indeed, all aspects of planning within the energy sector should be coordinated, as discussed in Chapter 3. In practice, however, many constraints, such as the inefficient pricing of other energy sources, may have to be accepted as given when electricity tariffs are set. In this context, the methodology discussed in this book is equally applicable to energy pricing in general. The first task of energy pricing should be to estimate an economically efficient pricing st-ucture corresponding to the strict marginal (opportunity) cost of various fuels, and then to adjust that structure as required to achieve other financial, economic, or social goals. To broaden the Bank's experience in the practical application of the principles of marginal cost pricing in the electric power field, two power tariff seminars were conducted in Asia. The seminars were designed to increase awareness and understanding of the roles and limitations of these techniques among utility companies in the developing countries. Part Two of the book includes five case studies prepared mainly in response to these seminars by participants representing several Asian power utilities, with some guidance and editing by the authors. An introductory chapter reviews, compares, and analyzes the principal issues and lessons to be learned from the tariff studies. Specific contributors to Part Two include: Soejitno Sujanto and Djaya Santoso (Indonesia); Muhammed Jahangir and Iqbal Khan (Pakistan); Malaine Valenzuela and coworkers (Philippines); K. K. Y. W. Perera, Nelson Wijemanne, and S. M. Christy (Sri Lanka), and the Electricity Tariff Conimittee (Thailand). The seminars were held in Thailand in June 1978 and in Indonesia in January 1979 and were attended by representatives from twelve Asian countries.5 The countries represented were Burma, Fiji, Indonesia, Korea, Malaysia, Nepal, Pakistan, Papua New Guinea, Philippines, Singapore, Sri

4. For details, see Mohan Munasinghe, "An Integrated Framework for Energy Pricing in Developing Countries," Energy Journal, vol. I (July 1980), pp. 1-30; also available as Reprint no. 148 from the World Bank, Washington, D.C. 5. More recently, similar seminars have been successfully conducted in Colombia (November 1979) and Brazil (July 1980) for twenty-six Latin American and Caribbean countries, and in Kenya (June 1980) for twelve African countries. See Mohan Munasinghe and Colin J. Warren (eds.), Power Tariffs: Case Studies in Latin America and the Caribbean (Washington, D.C.: World Bank, 1981).

10

THEORY

6 Lanka, and Thailand. Some of the participants had several years' experience working with the World Bank in this field; others had none. The seminars were successful, resulting in a high level of interaction among the participants. They helped both the authors and those participants that were further along in their thinking to resolve several conceptual and practical problems. In turn, the more experienced delegates were able to help other participants realize that the seminar was critical to their own countries' efforts to ensure the rational use of increasingly scarce energy resources. The frank exchange of information on the common practical and sociopolitical difficulties faced by electric power sector authorities was as valuable to the participants as was the knowledge gained regarding the principles and methodology of the LRMC approach. These pricing studies have been used subsequently as the bases for revising both the level and structure of existing power tariffs in many of the countries concerned. This is convincing evidence of the increasing importance of the marginal cost methodology. The case studies illustrate the practical problems of applying marginal cost pricing techniques. As might be expected, they handle several conceptual issues in different ways that may not fully satisfy the theoretical standards of purists in the marginal cost field. Without exception, these studies indicate an awareness of the fact that the real costs to the economy are the marginal costs. Even when tariff makers are constrained by other economic and social objectives in achieving a long-run marginal tariff structure that is based on cost, such studies indicate to policymakers the economic cost of meeting these other objectives. Marginal cost is, therefore, the benchmark by which the achievement of those other socioeconomic objectives can be consciously judged. Whatever pricing policy eventually emerges, it seems to be crucial to know at least the magnitude and structure of the LRMC of supply. Such an inquiry may be conducted along the lines presented in this book.

Objectives of an Electric Power Tariff

The modern approach to electric power pricing recognizes the existence of several objectives or criteria, not all of which are mutually consistent. First, national economic resources must be allocated efficiently, not only among different sectors of the economy, but also within the electric power sector. This implies that prices that reflect cost must be used to indicate to the electricity consumers the true economic cost of supplying their specific needs, so that supply and demand can be matched efficiently. Second, certain principles relating to fairness and equity must be satisfied,
6. The full list of seminar participants is given in Appendix A.

OBJECTIVES

AND PRICING

FRAMEWORK

11

including: (a) allocating costs among consumers according to the burdens they impose on the system; (b) assuring a reasonable degree of price stability and avoiding large price fluctuations from year to year; and (c) providing a minimum level of service to persons who may not be able to afford the full cost. Third, the power prices should raise sufficient revenues to meet the financial requirements of the sector, as described earlier. Fourth, the structure of electric power tariffs must be simple enough to facilitate the metering and billing of consumers. Fifth, and finally, other economic and political requirements must also be considered. These might include, for example, subsidized electricity supply to certain sectors to enhance growth or to certain geographic areas for regional development. Since the above criteria often conflict with one another, it is necessary to accept certain tradeoffs between them. The LRMC approach to setting prices is both rigorous and flexible enough to provide a tariff structure that is responsive to these basic objectives.

Tariffs Based on Long-Run Marginal Costs

A tariff based on LRMC is consistent with the first objective of efficiently allocating resources. The traditional accounting approach is concerned with recovering historical, or stnk, costs. In calculating the LRMC the important consideration is the amount of future resources used or saved by consumer decisions. Since electricity prices are the amounts paid for increments of consumption, they should generally reflect the incremental cost incurred. Supply costs increase if existing consumers increase their demand, or if new consumers are connected to the system. Therefore, prices that act as a signal to consumers should be related to the economic value of future resources required to meet consumption changes. The accounting approach that uses historical assets and embedded costs implies that future economic resources will be as cheap or as expensive as in the past. This could lead to overinvestment and waste, or underinvestment and the additional costs of unnecessary scarcity. To promote better utilization of capacity, and to avoid unnecessary investments to meet peak demands, which tend to grow very rapidly, the LRMC approach structures prices so that they vary according to the marginal costs of serving demands: by diifferent consumer categories, in different seasons, at different hours of the (Jay, by different voltage levels, in different geographic areas, and so on. In particular, with an arppropriate choice of the peak period, structuring the tariffs based on LRMC by time of day generally leads to the conclusion that peak consumers should pay both capacity and energy costs, whereas

12

THEORY

off-peak consumers should pay only the energy costs. Similarly, analysis of LRMC by voltage level usually indicates that the lower the service voltage, the greater the cost that consumers impose on the system. The structuring of tariffs based on LRMC also meets requirements (a) and (b) of the second, or fairness, objective mentioned above. The economic resource costs of future consumption are allocated as much as possible among the customers according to the incremental costs they impose on the power system. In the traditional approach, fairness was often defined rather narrowly and led to the arbitrary allocation of accounting costs to various consumers. Because the LRMC method deals with future costs over a long period-for example, at least five to ten years-the resulting prices in constant terms tend to be quite stable over time. This smoothing out of costs during a long period is especially important given the large size or "lumpiness" of power system investments. Using economic opportunity costs (or shadow prices, especially for capital, labor, and fuel) instead of purely financial costs and taking externalities into consideration whenever possible also link the LRMC method to efficient resource allocation. Developing Practical Tariff Structures

The first stage of the LRMC approach is the calculation of the pure or strict LRMC that reflects the criterion of economic efficiency. If price was set strictly equal to LRMC, consumers could indicate their willingness to pay for more consumption, thus signaling the justification of further investment to expand capacity. In the second stage of setting tariffs, ways are sought to adjust the strict LRMC to meet the other objectives, among which the financial requirement is most important. If prices were set equal to the strict LRMC, a financial surplus would be likely. This is because marginal costs are higher than average costs when the unit costs of supply are increasing. In principle, financial surpluses of the utility may be taxed away by the state, but in practice the use of power pricing as a tool for raising government revenues is usually politically unpopular and is rarely applied. Such surplus revenues also can be utilized in a way that is consistent with the other objectives. For example, the connection charges can be subsidized without violating the LRMC price, or low-income consumers could be provided with a subsidized block of electricity to meet their basic requirement, thus satisfying sociopolitical objectives. Conversely, if marginal costs are below average costs-typically as a result of economies of scale-then pricing at the strict LRMC will lead to a financial deficit. This will have to be made up, for example, by higher lump-sum connection charges, by flat rate charges, or even by government subsidies.

OBJECTIVES AND PRICING FRAMEWORK

13

Another reason for deviating from the strict LRMC arises because of second-best considerations, When prices elsewhere in the economy do not reflect marginal costs, especially for electric power substitutes and complethe strict marginal cost pricing rule for electricments, then departures fromn ity services would be justified. Clearly the first-best solution would be to remove the price distortions in the other parts of the economy. The second-best adjustments to electricity prices arise only because the first-best option is not practically feasible. For example, in rural areas inexpensive alternative energy may be available in the form of subsidized kerosene or firewood. In this case, pricing electricity below the LRMC may be justified to prevent excessive use of the alternative forms of energy. Similarly, if incentives are provided to import private generators and their fuel is also subsidized, then charging the full marginal cost to industrial consumers may encourage them to purchase their own or captive power plant. This is economically less efficient from a national perspective. Since the computation of the strict LRMC is based on the power utilities' least-cost expansion program, LRMC may also need to be modified by short-term considerations if previously unforeseen events make the long-run power systemrplan suboptimal in the short run. Typical examples of such situations include a sudden reduction in the growth of demand and a large excess of installed capacity, which may justify slightly reduced capacity charges, or a rapid increase in fuel prices, which could warrant a short-term fuel surcharge. As discussed earlier, the LRMC approach permits a high degree of tariff structuring. Practical data constraints and the need to simplify metering and billing procedures, however, usually require that tariffs be differentiated only by: (a) principal customer categories-residential, industrial, commercial, special, rural, and so on; (b) voltage levels (high, medium, and low); (c) time of day (peak, off-peak); and (d) geographic region. Finally, various other constraints also may be incorporated into tariffs based on the LRMC, such as the political requirement of having a uniform national tariff, subsidizing rural electrification, and so on. In each case, however, such deviations from the LRMC will impose an efficiency cost on the economy.

Summary

In the first stage of calculating the LRMC, the objective of economic efficiency in setting tariffs is satisfied, because the method of calculation is based on future economic resource costs rather than on sunk costs, and also incorporates economic considerations, such as shadow prices and externalities. The structuring of marginal costs permits the tariffs to be efficiently and fairly allocated among consumers. In the second stage of developing a

14

THEORY

tariff based on LRMC, deviations from the strict LRMC are considered to meet important financial, social, economic (second-best), and political criteria. This second step of adjusting the strict LRMC is generally as important as the first calculation, especially in developing countries. The LRMC approach provides an explicit framework for analyzing system costs and setting tariffs. If departures from the strict LRMC are required for noneconomic reasons, then the economic efficiency cost of these deviations may be estimated roughly by comparing the effect of the modified tariff relative to the (benchmark) strict LRMC. Since the cost structure may be studied in considerable detail during the LRMC calculations, this analysis also helps to pinpoint weaknesses and inefficiencies in the various parts of the power system-for example, overinvestment; unbalanced investment; or excessive losses during generation, transmission, and distribution in different geographic areas; and so on. This aspect is particularly useful in improving system expansion planning. Finally, any tariff based on the LRMC is a compromise between many different objectives. Therefore, no ideal tariff exists. By using the LRMC approach, it is possible to revise and improve the tariff on a consistent and continual basis. Thus, the optimal price is reached gradually over several years, without subjecting long-standing consumers to unfair shocks because of large abrupt price changes.

Chapter2

Economics ofMarginal Cost Pricing

dates to the pathbreaking efforts of Dupuit and, subsequently, Hotel'ling. Ruggles provided a comprehensive review of work on the subject up to the 1940s.' The development of the theory, especially for application in the electric power sector, received a strong impetus 2 from the work of Boiteux, Steiner, and others in the 1950s. Recent work has led to more sophisticated investment models that enable analysts to structure marginal costs rnore accurately; to detennine developments in peak load pricing; and to consider the effects of uncertainty, the costs of power 3 shortages, and so on.
MARGINAL COST PRICING THEORY

Basic Marginal Cost Theory

The rationale for setting price equal to marginal cost may be clarified with
1. P. Dupuit, "De l'Utilite et de Sa Mesure," La Reforma Soziale (Turin, 1932); H. Hotelling, "The General Welfare in Relation to Problems of Railway and Utility Rates," Econometrica, vol. 6 (July 1938), pp. 242-69; N. Ruggles, "The Welfare Basis of the Marginal Cost Pricing Principle," Review of Economic Studies, vol. 17 (1949-50), pp. 29-46; and N. Ruggles, "Recent Developments in the Theory of Marginal Cost Pricing," Review of Economic Studies, vol. 17 (1949-50), pp. 107-26. 2. See M. Boiteux, ''La Tarification des Demandes en Pointe," Revue Generale de l'Electricite, vol. 58 (1949); P. Steiner, "Peak Loads and Efficient Pricing," Quarterly Journal of Economics (November 1957); M. Boiteux and P. Stasi, "The Determination of Costs of Expansion of an Interconnected System of Production and Distribution of Electricity," in Marginal Cost Pricing in Practice, ed. James Nelson (Englewood Cliffs, N.J.: Prentice-Hall, 1964); 0. E. Williamson, "Peak Load Pricing and Optimal Capacity under Indivisibility Constraints," American Economic Review, vol. 56, no. 4 (September 1966), pp. 810-27; and Ralph Turvey, Optimal Pricing and Investment in Electricity Supply (Cambridge, Mass.: MIT Press, 1968). 3. See "Symposium on Peak Load Pricing," Bell Journal of Economics, vol. 7 (Spring 1976), pp. 197-250; Ralph T'urvey and Dennis Anderson, Electricity Economics (Baltimore, Md.: Johns Hopkins University Press, 1977); Michael A. Crew and Paul R. Kleindorfer, Public Utility Economics (New York' St. Martins Press, 1979); Canadian Electrical Association, Marginal Costing and Pricing of Electrical Energy. Proceedings of the State of the Art Conference, Montreal, May 1978; and Mohan Munasinghe, The Economics of Power System Reliability and Planning (Baltimore, Md.: Johns Hopkins University Press, 1979). 15

16

THEORY

the simple supply-demand diagram shown in Figure 2-1. Let EFGDObe the demand curve, which determines the kilowatt-hours of electricity demanded in a given year at any given average price level. AGS is the supply curve, and is represented by the marginal cost (MC) of supplying additional units of output. At price p and demand Q, the total benefit of consumption is represented by the consumers' willingness to pay, that is, the area under the demand curve OEFJ. The cost of supplying the output is the area under supply curve OAHJ. Therefore, the net benefit, or total benefit minus supply cost, is given by the area AEFH. The maximum net benefit AEG is achieved when price is set equal to marginal cost at the optimal market clearing point G, that is,
(pO, QO).

In mathematical terms, the net benefit (NB) is given by NB = f p(q)dq -

fJQMC

(q)dq

where p (Q) and MC(Q) are the equations of the demand and supply curves, respectively. Maximizing NB yields d(NB)ldQ = p(Q) - MC(Q) 0 the O,
point at which the demand and marginal cost curves intersect (po, QO).

The analysis so far has been static. Now consider the dynamic effect of demand growth from year zero to the first, which leads to an outward shift in the demand curve from Do to D,. Assuming that the correct market clear-

Figure 2-1. Supply and Demand for Electricity Consumption

E
\F

S
\ '

/~~~~~~~~~S (MC)

PiL
pO, ------

,
F------------i-

-----

l----

P~~~~~~~~~~~~~~~P

~~It IIIDD
QC Q' Qt

Kilowatt-hours

ECONOMICS

OF MARGINAL

COST PRICING

17

ing price p, was prevailing in year zero, excess demand equal to GK will occur in the first year. Ideally, the supply should be increased to Q, and the new optimal market clearing price established at p,. The available information conceming the demand curve D, may be incomplete, however, making it difficult to locate point L. Fortunately, the technical-economic relations underlying the production function usually enable the analyst to determine the marginal cost curve more accurately. Therefore, as a first step, the supply may be increased to an intermediate level Q' at price p'. The existence of the excess demand MN indicates that both the supply and the marginal cost price should be further increased. Conversely, if L is overshot and there is excess supply, then it may be necessary to wait until demand growth catches up with the overcapacity. In this iterative manner, it is possible to move along the marginal cost curve toward the optimal market clearing point. As the optimum is approached, it is also shifting with demand growth. Therefore, this moving target may never be reached. The basic rule of setting price equal to the marginal cost and expanding supply until the market clears, however, is still 4 valid.

Capital Indivisibilil:ies and Peak Load Pricing

The analysis of the effect of capital indivisibilities, or "lumpiness" of investments recognizes the fact that, owing to economies of scale, capacity additions to power systems (especially generation) tend to be large and longlived (Figure 2-2). Suppose that in year zero, the maximum supply capacity is Q, while the optimal price and output combination (pa, Q) prevails, corresponding to the demand curve Do and the short-run marginal cost curve SRMC. The SRMC is based on fuel, operating, and maintenance costs, that is, supply costs with fixed capacity. As demand grows from Do to D, over time, the price must be increased to p, to clear the market in the short run, because capacity is fixed and the sup4. As explained later, this simple rule has to be modified when there are constraints in the economy: in particular, the consequences of shadow pricing of inputs, second-best considerations, and subsidized social prices for poor consumers. Also there has been a recent resurgence of interest in location-theoretical models of power supply for systems that have already exploited most of the economies of scale. Typically the bulk transmission network is treated as a common carrier, which purchases power from independent generating stations and then sells power to separate distribution utilities. The possibilities for allocating costs according to the spatial location of generating and load centers and for the existence of different modes of market behavior, such as competition or collusion among generation companies with or without regulation, can give rise to a vwiderange of pricing possibilities. For a more detailed discussion, see Richard E. Schuler and Benjamin F. Hobbs, "Spatial Competition: Applications in the Generation of Electricity" (paper presented at the 1981 Annual Meeting of the Southem Regional Science Association, Washington, D.C., April 1981; processed).

18

THEORY

Figure 2-2. The Effect of Capital Indivisibilities on Price

P2

--

--

SRMC

*_Po

RP14 4 ------

__

__,

Q,

Q Kilowatt-hours

ply curve is extremely steep at Q. When the demand curve has shifted to D 2 and the price is P2, plant is added to increase the capacity to Q. As soon as the capacity increment is completed and becomes a sunk cost, however, SRMC falls to its old trend line. Therefore, p3 is the optimal price corresponding to demand D3 and the SRMC curve. Generally, the large price fluctuations during this process will be unacceptable to consumers. This practical problem may be avoided by adopting a long-run marginal cost (LRMC) approach and by recognizing the need for peak load pricing, as described below. The basic peak load pricing model shown in Figure 2-3 has two demand curves. For example, D,k could represent the peak demand during the daylight and evening hours when electric loads are large, and Do, would indicate the off-peak demand during the remaining hours when loads are light. The marginal cost curve is simplified by assuming a single type of plant with the SRMC of fuel, operating, and maintenance costs given by the constant a, and the LRMC of adding to capacity given by the constant b. Adding to capacity could involve, for example, investment costs suitably annuitized and distributed over the lifetime output of the plant. The static diagram has been drawn to indicate that the pressure on capacity arises because of the peak demand D 0,, and that the off-peak demand D,Ddoes not infringe on the capacity Q. The optimal pricing rule now has two parts corresponding to two distinct pricing periods (differentiated by the time of day): peak period price (pk = a + b) and off-peak period price (p, = a).

ECONOMICS

OF MARGINAL

COST PRICING

19

Figure 2-3. Peak Load Pricing Model

D,,
Q
Kilowatt-hours per hour Note: Dpk = peak demand; D, = off-peal demand; a nance costs; and b = LRMC of adding to capacity. SRMC of fuel, operating, and mainte-

The logic of this simple result is that users during peak periods, who are the cause of capacity additions, should bear full responsibility for the capacity costs, as well as fuel, operating, and maintenance costs. Off-peak consumers need not pay the capacity costs. As explained in Appendix C, the analysis becomes more complicated when there is more than one type of generating plant and there are more than two pricing periods.

Extensions of Simple Models

The models presented so far have been deliberately idealized and simplified to clarify the basic principles involved. Extensions of these models incorporate and analyze the economics of real-world power systems more realistically. Three broad types of practical complications are discussed below. First, the usual procedure adopted in marginal cost pricing studies may require some iteration as shown in Figure 2-4. Typically, a deterministic longrange demand forecast is made assuming that prices will evolve in the future. Then, using power system models and data, several plans are proposed to meet this demand at some fixed target reliability level (see Chapter 3). The cheapest or least-cost system expansion plan is chosen from these alternatives. Finally, the strict LRMC is computed on the basis of this least-cost

Figure 2-4. Use of Price Feedback in Estimating Tariffs Based on LRMC

Iterative feedback loop for price

[Systenm models| and data K. Revised price forecast m alternative system / plans and costs Original price forecast Load-demand pmodels, data, and forecast Fixed target reliability level (R) level (R) Economic efficiency objective Financial viability, social subsidy, fairness, other objectives and ~~~~~~~~~~~other Least-cost system plan

lRMC

Tariff dsed

L0l

ECONOMICS

OF MARGINAL

COST PRICING

21

plan, and a tariff structure based on the adjusted LRMC is prepared. If the newly estimated tariff that is to be imposed on consumers is significantly different from the originally assumed evolution of prices, however, then this first-round tariff must be fed back into the model to revise the demand forecast and repeat the LRMC' calculation. In theory, this iterative procedure could be repeated until future demand, prices, and LRMC-based tariff estimates become mutually self-consistent. In practice, uncertainties in price elasticities of demand and other data may dictate a more pragmatic approach in which the LRMC results would be used after only one iteration to devise new power tariffs and to implement them. The demand is then observed over some time period, the LRMC is reestimated, and tariffs are revised to move closer to the optimum, which may itself have shifted, as described previously. An extreme form of price feedback could result in a shift of the peak outside the original peak period, especially if this period was too narrowly defined. That is, peak load pricing may shift the demand peak from one pricing period to another. If sufficient data on the price elasticity of demand were available, theory indicates that each potential or secondary peak should be priced to keep it just below the available capacity level. Since the necessary information would rarely be available in practice, a combination of techniques-including use of a sufficiently wide peak period, redefining the peak period to include both the actual and potential peaks, direct switching of certain consumer loads, and so on-may be used to avoid the shifting peak problem. Second, the interrelated issues of supply and demand, uncertainty, reserve margins, and costs of shortages raise certain problems. As mentioned before, the least-cost systerm expansion plan to meet the demand forecast is generally determined assuming some arbitrary target level of system reliability. Reliability measures include loss-of-load-probability (LOLP), reserve margin, and so on.5 Therefore, marginal costs depend on the target reliability level. Economic theory suggests that reliability should also be treated as a variable to be optimized and that both price and capacity-or, equivalently, reliability-levels should be optimized simultaneously. The optimal price is the marginal cost price as described earlier. The optimal reliability level is achieved when the marginal cost of adding capacity to improve the reliability are equal to the expected value of the cost savings to consumers resulting from the electricity supply shortages averted by those capacity increments. These considerations lead to a more generalized approach to system expansion planning as shown below.6
5. For a more detailed review of measures of reliability and their relation to shortage costs, see Munasinghe, The Economics of Power System Reliability and Planning. 6. For details, see Mohan Munasinghe, "A New Approach to Power System Planning," IEEE Transactions on Power Apparatus and Systems, vol. PAS-99 (May-June 1980), pp. 1198-1209; also available as Reprint no. 147 from the World Bank, Washington, D.C.

22

THEORY

Consider a simple static expression for the net benefits (NB) of electricity consumption, which is to be maximized:
NB (D,R)
-

TB (D) -

SC (D,R) - OC (D,R)

where TB = total benefits of consumption if there were no outages, SC = supply costs (system costs), OC = outage costs (costs to consumers of supply shortages), D = demand, and R = reliability. In the traditional approach to system planning, both D and R are exogenously fixed, and therefore NB is maximized when SC is minimized. This is conventional system expansion planning on a least-cost basis. If R is treated as a variable, however, then
d(NB)ldR
= -[d(SC

+ OC)/dR]

[a(TB- SC

OC)IaD] (ODIlR)= 0

is the necessary first-order maximization condition. Next, assuming aDlaR = 0, the result is

aSCIaR=

dOCldR.

Therefore, as described earlier, reliability should be increased by adding to capacity until the above condition is satisfied. In Figure 2-5, as reliability Figure 2-5. Relation betwveenOutage Costs, Supply Costs, and Total Costs at the Optimal Reliability Level

~Jslope
I '

C Reliability Note: OC = outage costs, SC mal reliability level.

=

RI

100 percent
=

supply costs, TC = total costs = OC + SC, and RI

opti-

ECONOMICS

OF MARGINAL

COST PRICING

23

increases, outage costs fall, whereas supply costs rise more and more rapidly. Reliability is optimized at RI when the slope of the SC curve is equal to the negative slope of the OC curve. An altemative way of expressing this result is that since TB is independent of R, NB is maximized when total costs TC = (SC + OC) are minimized. The above criterion is one that effectively subsumes the traditional rule for least-cost system planning of minimizing only the system costs.' Third, some practical problems may be raised by the dichotomy of having to choose between SRMC and LRMC. A simplified, intuitive explanation will clarify this issue. SRMC may be defined in economic terms as the cost of meeting additional electricity consumption with fixed capacity. LRMC is the cost of meeting an increase in consumption, sustained indefinitely into the future, when needed capacity adjustments are possible. If there is an incremental increase in consumption, in the short run both the system operating costs and the outage costs (especially during the peak period) will also rise at the margin. Similarly, in the long run, an increase in demand will result in a corresponding increase in the operating costs as well as in the capacity costs. Thus in both the short and long run an equivalent increase in operating costs will occur. But the optimal reliability rule ensures that the marginal outage and capacity costs are also equal. Therefore, when the system is optimally planned and operated-that is, capacity and reliability are 8 optimal-SRMC and LRMC coincide. Therefore, the estimation and use of the strict LRMC is simplest when the system is near the ideal operating point. If the system plan is suboptimal, however, significant deviations between SRMC and LRMC will have to be resolved within the pricing policy framework. For example, after 1973 many utilities began to replace oil-fired plants with coal-fired units to save fuel costs. Such a situation could result in significant excess capacity and low marginal capacity costs in the short to medium run, thus justifying reduced demand charges below the LRMC level. In this situation, however, as peak demand grows and the system approaches its ideal operating point again, the capacity charges should rise
7. The emphasis on outage costs requires greater effort to measure these costs; see Mohan Munasinghe and Mark Gellerson, "Economic Criteria for Optimizing Power System Reliability Levels," Bell Journal of Economics, vol. 10 (Spring 1979), pp. 353-65, also available as Reprint no. 112 from the World Bank, Washington, D.C.; and Mohan Munasinghe, "The Costs of Electric Power Shortages to Residential Consumers," Journal of Consumer Research, vol. 6 (March 1980), pp. 361-69, also available as Repaint no. 128 from the World Bank, Washington, D.C. An even more generalized system planning criterion in which the assumption aDlaR = 0 is relaxed, thus allowing D to vary with R, is presented in Mohan Munasinghe, "System Planning to Achieve Optimal Reliability Levels," Proceedings of the Third International Conference on Analysis, Forecasting and Planning for Utilities, Paris, June 1980 (processed). 8. For a more rigorous proof of this result, see Crew and Kleindorfer, Public Utility Economics, chap. 7.

24

THEORY

smoothly toward LRMC. If demand charges are suppressed too far below LRMC or for too long, then electricity demand may be overstimulated. This could lead to costly shortages or uneconomic advancement of investments in future capacity. Furthermore, when future prices have to be increased to the LRMC level, the transition may be undesirably abrupt, resulting in consumer discontent. Conversely, when there are significant shortages, the economically efficient short-run solution would be to raise prices to ration the limited supplies according to the consumer's willingness to pay. Such a price increase would generally be impractical, however, because consumers would rarely accept higher prices when the quality of service is poor (see Chapter
5).

Finally, if there are substantial outage costs outside the peak period, then the optimal marginal capacity costs may be allocated among the pricing periods in proportion to the corresponding marginal outage costs. Another suggestion is to allocate capacity costs to different pricing periods in inverse proportion to LOLP, but this would be only a rough approximation, because aggregate reliability indexes such as LOLP generally are poor proxies for
prorating marginal outage costs."

An alternate approach to this problem has been proposed recently and uses spot or instantaneous prices for electricity.'" Advances in solid state switching, metering, and communications technology have made it possible to vary prices continuously instead of relying on a predetermined schedule. Very briefly, prices at any moment are set to reflect marginal supply costs. As demand rises toward the capacity limit, price is allowed to rise as high as necessary to choke off demand before a shortage occurs. This type of pricing scheme is most appropriate for large electricity users, especially in industrialized countries, where the increased costs of communications and metering hardware are more than compensated for by the more economic and efficient use of electricity (see Chapter 5).
9. See, for example, Joseph Vardi, Jacob Zahavi, and Benjamin Avi-ltzhak, "Variable Load Pricing in the Face of Loss of Load Probability," Bell Journal of Econiomics, vol. 8 (Spring 1977), pp. 270-88. 10. One of the earliest articles is William Vickrey, "Responsive Pricing of Public Utility Services," Bell Journal of Economics, vol. 2 (Spring 1971), pp. 337-46. For a recent review of the topic, see Roger E. Bohn, Michael C. Caramanis, and Fred C. Schweppe, "Optimal Spot Pricing of Electricity," Working Paper no. MIT-EL-81-008WP (Cambridge, Mass., MIT Energy Laboratory, March 1981; processed).

Chapter 3

Prerequisites forMarginalCosting
SEVERAL PRELIMINARY STEPS must be

takenbefore the basic theorypresented

earlier for estimating- long-run marginal costs (LRMC) can be successfully applied. The intimate theoretical link between the optimal investment and pricing rules was stressed in the previous chapter. If the load or demand forecast and the plan for least-cost system investment are not prepared accurately, the marginal costs that are derived from those earlier calculations are also likely to be incorrect. Therefore, the rationale underlying the investment decision and the elements of demand forecasting and system planning, which are essential prerequisites for a pricing exercise, and the principles of shadow pricing need to be understood. Long before the decision to invest in electric power is made, even broader issues must be faced. They involve the total future energy needs, availability of supply, and the optimal mix of different sources to be developed. Ideally, the energy sector investrnent and pricing policies for the entire nation should be analyzed and determined within an explicit integrated framework.' In practice, such decisions are often made by policymakers case by case. Modem societies require increasing amounts of energy for domestic, industrial, commercial, agricultural, and transport uses. Arrayed against these energy needs are the short-term, depletable fossil fuel supplies-petroleum, coal, and natural gas-as well as the longer-run, renewable energy sources. The latter include water, nuclear, solar, geothermal, wind, tidal, and biomass, along with traditional or noncommercial fuels such as wood and animal waste. At this stage, three basic energy policy decisions are required.2 First, the appropriate level of demand for energy that must be served to achieve social
1. See Mohan Munasinghe, "Integrated National Energy Planning (INEP) for the Developing Countries," Natural Resources Forum, vol. 4 (October 1980), pp. 359-73; also available as Reprint no. 165 from the World Bank, Washington, D.C. Traditional or noncommercial fuels such as firewood are particularly important in developing countries, amounting to about 40 percent of total energy consumption in 1975. See Mohan Munasinghe and Colin J. Warren, "Rural Electrification, Energy Economics, and National Policy in the Developing Countries," in Future Energy Concepts, publication no. 171 (London: Institution of Electrical Engineers, 1979), pp. 44-47. 2. For a nontechnical discussion of these three aspects, that is, load forecast, investment 25

26

THEORY

goals, such as economic development, growth, and basic human needs, should be determined. Second, the optimal mix of energy sources must be established that will meet the desired demand, based on several national objectives, such as minimum cost, independence from foreign sources, conservation of resources, environmental considerations, and price stability. The analysis is complicated by uncertainties regarding the future evolution of demand and supply, relative costs and prices, and incompatibility of the different energy sources with all the various energy uses. Third, closely associated with and following the investment decision is the pricing policy, which will be based on criteria such as economic efficiency in resource allocation, economic second-best considerations, sector financial requirements, social 3 equity considerations, and other political constraints. Energy pricing decisions may also have feedback effects on investment decisions through demand (see Chapter 2). The electric power sector is usually one of the most important elements within the broader energy framework. Once the important decisions regarding energy policy have been made at the national level, the electric power sector authorities must perform a similar but more detailed analysis. More specifically, the demand for electricity must be forecast, preferably disaggregated by geographic region and customer category. Then alternative long-range investment programs are compared that will meet this demand forecast, subject to various constraints, as described below. The cheapest or least-cost program is usually selected as the optimal one. The appropriateness of such a least-cost program for power expansion must be further verified in the sense that, broadly defined, the social benefits of this expansion should exceed the social costs by the maximum amount, that is, until the net benefits of consumption are maximized. A pricing policy based on LRMC effectively permits the burden for the cost-benefit analysis to be placed on the electricity consumer, because he signals the justifi-

plan, and output price, relating specifically to electric power, see Mohan Munasinghe, "Planning for Electrical Power: Costs and Technologies." National Development (April 1980), pp. 75-84; also available as Reprint no. 149 from the World Bank, Washington, D.C. 3. For details of energy pricing, see Mohan Munasinghe, "An Integrated Framework for Energy Pricing in Developing Countries," Energy Journal. vol. I (July 1980), pp. 1-30; also available as Reprint no. 148 from the World Bank, Washington, D.C. Interactions between different energy sources must not be neglected; see Gunter Schramm and Mohan Munasinghe, "Interrelationships in Energy Planning: The Case of the Tobacco Industry in Thailand," Energy Systems and Policy vol. 5 (March 1981), pp. 117-39; also available as Reprint no. 187 from the World Bank, Washington, D.C.; and Oli Havrylyshyn and Mohan Munasinghe, "Interactions among Alternative Modes of Energy Production: Empirical Case Study for Sri Lanka," Economics Discussion Paper (Washington, D.C.: George Washington University, April 1980).

PREREQUISITES

FOR MARGINAL

COSTING

27

cation of further investment by his willingness to pay the marginal cost of electricity supply.

Load Forecasting
Many elements of load forecasting are relevant to the process of system planning. Further details may be found in the references cited in this chap4 ter. The terms demand and load forecast are used interchangeably to indicate the magnitude as well as the structure of the requirements of both electrical power (kilowatts) and energy (kilowatt-hours), unless an appropriate distinction is necessary. The structure of demand includes disaggregation by geographic area, consumer category, and time period, as well as characteristics such as load and diversity factors and system losses. Since demand characteristics vary by type of consumer, geographic area, and time period, a knowledge of loads at the disaggregate level is required. The properties of the aggregate demand at the system level may be quite different from the characteristics of the individual loads. Disaggregate loads are important in the system planning process. In generation planning, the system may be modeled as one source feeding a single lumped load; in the design of transmission networks, the characteristics of demand by region and by principal load center, such as a city, become important. Ultimately, for planning distribution grids, a detailed knowledge of demand at each load center is required, as discussed below. Furthermore, the sum of many disaggregate demand forecasts, estimated by using various different techniques, can often serve as a useful check on an independently made global forecast for the same region. Power, energy, and the loadfactor Before discussing the other characteristics of demand, it would be helpful to examine the basic relation between power and energy. The commonly used unit of energy in heavy current electricity applications is the kilowatthour. The rate of flow of energy per unit of time is called power, which is measured in kilowatts. The distinction between power and energy is important because the same amount of energy may be delivered within a small interval of time at a relatively high rate of power flow, or over a longer period at a lower power level.
4. See Federal Power Commission, The Methodology of Load Forecasting, 1970 National Power Survey (Washington, D.C.: Department of Energy, 1970); Robert L. Sullivan, Power System Planning (New York: McGraw-Hill, 1977), chapter 2; and G. E. Huck, "Load Forecast Bibliography: Phase 1," Proceedings of the IEEE PES Summer Meeting, Vancouver, July 1979, paper no. F7950-3.

Figure 3-1. Schematic of a Simple Electric Power System

G, TF 1
TT2

TT 1,

D dt L---_----J

Ts

T3

r- -I
7',T 2 v

-_-_,

= HVtanmssolns 5T

T~~~~
281 1 F2

S~ , ~ 52 1 seonay

=ins

M2

~~DT,

Note: G,, Ga generators; TF,, TF2 = voltage transformers; T,, T,2,T., T6 = HVtransmission lines; MS,TS2 = transmission substations; T. = BHV transmission; DS1,DS&= distribution substations; Fl, F2 = primary feeders (or circuits); DT1, DT2 = distribution transformers; S,, 52 = secondary lines; M,, m = customer service inlets and meters;

28

PREREQUISITES

FOR MARGINAL

COSTING

29

The load factor (LF), which is the ratio of average to maximum or peak kilowatts over a given interval of time, may be specified for a single customer, the whole systemn,on a daily or annual basis, and so on. The LF is important, because the size or capacity and, therefore, the cost of power system components are determined to a great extent by their ability to handle peak power flows. Since practically all customers use maximum power only during a short peak period during the day, the LF is also a measure of the intensity of capacity use. (See the discussion on the load duration curve below.) A load forecast may be made either in terms of peak power or of total energy consumed during a given period. The LF is used to convert from one unit to the other. Kilowatt peaks for different types of customers do not occur simultaneously. The diversity factor for a group of consumers measures the divergence over time of the individual peak loads and permits the combined peak load for the group to be computed starting from disaggregate peak values. It is convenient at this point to examine the load forecast requirements that will facilitate disaggregation of the LRMC estimates. As explained in detail in the next chapter, the LRMC of supply is estimated at various points in the system which usually are distinguished by the voltage level, for example, with respect to generation, transmission, and distribution. Therefore, the load-demand forecast must be sufficiently disaggregate to permit this type of structuring. Components of typical electric power systems The basic component and voltage levels in the simple power system are shown in Figure 3-1. lThe output from two power generating plants G, and G, (hydroelectric or thermal) is increased to the high voltage (HV) level at source by the respective transformers TF, and TF2 and is fed into the busbars in transmission substation TS, through the HV transmission links T, and T2. At TS, the voltage is further increased to the extra high voltage (EHv) line T, to substation TS 2 , after which the electric 2 . A voltage reduction occurs at TS power is fed, through HV tranmission lines T, and T5, into the distribution substations DS, and DS, located close to the load centers. Another voltage reduction occurs at DS, and DS2 , and then power flows out of the distribution substation busbars, through primary feeders such as F,, to various distribution transformers (for example, DT,), where the voltage is decreased further. Finally, the power is delivered to a representative consumer through the secondary distribution line S, and service inlet and meter M,. In transporting electric power the general principle is that voltage level depends on distance and power flow. Transmission lines are used to carry large amounts of power over long distances, whereas distribution lines involve smaller power flows over shorter distances. The technical distinction

30

THEORY

Figure 3-2. Power and Energy Flows, Consumption, and Losses in a Typical Power System
System peak power and energy at generators , Generation Station use + 2 EHV and HV transmission losses + nontechnical losses
EHV

and

HV

System peak (3) power and energy at HV

(4)

Power and energy demands of HV- consumers

(5 1_

MVdistribution losses

+ nontechnical losses

System peak (6) power and

energy at MV

Power and energy (7) demands of

MV consumers

(8) System peak power and

energy at Note:
LV

distribution losses and nontechnical losses

LV

LV

(9) Power and energy demands

of
LV

consumers

power and energy flows losses generation

'3D transformation

PREREQUISITES

FOR MARGINAL

COSTING

31

between transmission and distribution facilities is usally made on the basis of their operating voltages. Although voltage standards vary greatly from country to country, broadly accepted definitions for voltage levels are as follows: EHV transmission, more than 220 kilovolts; HV transmission, 45 to 220 kilovolts; primary distribution, 6 to 25 kilovolts; and secondary distribution, 110 to 380 volts. In practice, considerable overlap is likely, and facilities operating at voltages in the range of 25 to 45 kilovolts (often called subtransmission) may be included within the distribution or transmission category, depending on their function. An operating power system is much more complex than the one described above, and in general has many generating sources linked to many load centers through an interconnected transmission network. Transmission interconnections may be used to tie the power systems of several different utilities together to form an even larger power pool. The distribution grid at a single load center would serve thousands of consumers. Furthermore, many other components, including protective and relaying devices as well as load control and dispatching equipment, would play important roles in a large interconnected system. Figure 3-2 shows the flow of power and energy from generators to consumers. The total amount of both kilowatts and kilowatt-hours generated at the source will be greater than the corresponding values consumed because of losses in the system. Generation losses usually involve station use, such as driving various auxiliary equipment. Transmission and distribution network losses occur in the lines, substations, and transformers. Nontechnical losses, including theft and unmetered consumption, arise frequently in some countries and must be allowed for at each voltage level, where necessary. To establish LRMC at the generators and at the high- medium- and lowvoltage levels, the system peak power and energy forecast should be known at points (1), (3), (6), and (9) in the figure. An estimate of the consumption represented by (4), (7), and (9)-as well as the losses at (2), (5), and (8)would be required to reconstruct the complete power and energy flow balance.5 Similarly, if a need exists for structuring of LRMC by geographic region, the load forecast must also be disaggregated by region and combined with the spatially disaggregate system costs (see Chapter 4). 6 Load forecasts involve time periods of varying duration. Daily or weekly
5. As explained in Chapter 4, the energy and peak power loss factors will, in general, be different. Also, the peak power demand of consumers at different voltage levels may not coincide in time with the system peak. T'herefore, when the energy and power flow balances are drawn up, energy flows can be added arithmetically, but peak power flows have to be aggregated after considering the time of day when these flows occur. These points are brought out in greater numerical detail in the case studies (see Chapter 8, Indonesia). 6. K. N. Stanton, "Medium-Range, Weekly, and Seasonal Peak Demand Forecasting by Probability Methods," IEEE Transactions on Power Apparatus and Systems, vol. PAS-71 (May 1971), pp. 1183-89.

32

THEORY

demand projections are made to optimize system operation and to schedule hydro units. Short-run forecasts ranging between one and three years are used in hydro reservoir management, distribution system planning, and so on. Medium-term demand projections cover about four to eight years, corresponding to the lead times required for large transmission and generation projects. These forecasts, therefore, are especially useful to determine the next steps in building facilities. Long-range demand projections, usually encompassing ten to thirty-five years, are extremely important in long-run system expansion planning. Since they are the most relevant to the topic addressed in this book, methods of making long-range forecasts are discussed below. Forecasting techniques At first, long-term load forecasting basically consisted of extrapolating trend lines. A time series of past demand for electricity was analyzed as a function of time, using a variety of techniques ranging from simple graphing to sophisticated mathematical curve fitting. The resulting trend was extended into the future, sometimes expressed in terms of compound growth rates. Because this method implicitly assumes that present growth patterns will remain essentially unchanged in the future, it is often acceptable when stable socioeconomic conditions encourage electricity demand to increase steadily over many years. Adjustments may be made to the trend line estimates if significant changes are expected in the overall environment, such as economic and demographic shifts and weather changes. Another simple technique that has been used with large consumers is the systematic enumeration and summation of such individual loads to develop the overall forecast. Correlation methods of forecasting are more sophisticated than the techniques described above. The basic rationale underlying the use of correlation methods is that the growth of the electricity load could be indirectly forecast by linking it to several other variables whose future evolution would be easier to predict than the demand for electricity. The simplest techniques include correlating demand for electricity to some measure of economic activity, such as per capita gross national product, personal disposable income, industrial or commercial value-added, and so on. In addition, demand could be correlated with the price of electricity through the price elasticity of demand. The elasticity of the dependent variable y, with respect to the independent variable x, is given by: e, = (dyvdx)/(ylx). Projected values of income and price then could be used to estimate future growth of the load. Both extrapolation and correlation approaches require some knowledge of historical data on electricity demand for the area or country under consideration. If this information is unavailable, as is often the case in developing countries where there are large amounts of unsatisfied demand, techniques

PREREQUISITES

FOR MARGINAL

COSTING

33

that can use appropriate data from other comparable regions may be required.'

More complex multiple correlation techniques consider several additional variables, such as prices of competing fuels and energy sources, temperature and weather, land use, and population shifts. They often involve fitting nonlinear demand functions. Such econometric methods, which explicitly consider the price of electricilLy as an explanatory variable, are particularly relevant here. One recent review of the topic notes that problems encountered to 9 date have been mainly conceptual. The difficulties include (a) modeling tariffs that involve fixed and variable charges, prices for different consumption blocks, as well as discontinuities in the demand function; (b) distinguishing between short-run versus long-run or dynamic adjustments; and (c) incorporating the effects caused by changes in the prices of competing fuels. The recent resurgence of interest:in marginal cost pricing is resulting in the continued development of econometric models that emphasize isolating the effects of electricity price on the level of demand.' Past empirical research on price elasticities of demand for electric power carried out mainly for industrialized countries shows massive variations in the results. The confidence that can be placed in econometric demand analysis, even in developed countries, is therefore not considerable. The problems are magnified, however, when similar analysis is attempted in a developing country. For exarnple, time series analysis may be frustrated by histories of power shortages, so that, at the margin, consumption at any one point in time might be (letermined by supply rather than demand. This would give rise to formidable identification problems in econometric estimation. For the general case it might be expected that long-term elasticities are greater than short-term elasticities, as existing capital stock gets replaced and fuel substitutes become available, but the effects tend to be blurred because of all the other events taking place over the period. This problem may
7. See H. Aoki, "New Method of Long Range or Very Long Range Demand Forecast of Energy, Including Electricity, Viewed from a Worldwide Standpoint" (Tokyo: Electric Power Development Company Ltd., 1974; processed), in which an attempt is made to correlate the growth rate and the level of per capita gross national product and per capita electricity consumption in kilowatt-hours for 111countries. 8. See Lester D. Taylor, "The Demand for Electricity: A Survey," Bell Journal of Economics, vol. 6 (Spring 1975), pp. 74-110. For a valuable survey of the economic principles and issues underlying econometric analysis of residential electricity consumption, see Charles J. Cicchetti, William J. Gillen, and Paul Smolensky, The Marginal Cost and Pricing of Electricity (Cambridge, Mass.: Ballinger, 1977), appendix B (by Denis Aigner). For a more general review of energy demand models, see Robert S. Pindyck, The Structure of World Energy Demand (Cambridge, Mass.: MIT Press, 1979). 9. See several papers that recently appeared in Canadian Electrical Association, Marginal Costing and Pricing of Electrical Energy, Proceedings of the State of the Art Conference (Montreal, 1978).

34

THEORY

assume special importance in developing countries in which significant structural economic changes may occur in a relatively short time period, thus statistically swamping the effect of price changes, even though the price elasticity may in fact be substantial. The models discussed so far have used deterministic load projections; uncertainty is usually accounted for by parametric shifts in the forecasts to test sensitivity. Recently, more sophisticated techniques became available. They use matrix methods, such as input-output matrixes, to capture the relations between productive sectors, as well as stochastic processes that represent probabilities for sectoral growth in terms of a Markov model. These techniques encourage explicit consideration of probabilistic effects." The resulting long-range demand forecasts are still deterministic, however. Again, for sensitivity testing, the basic assumptions may be varied to provide low-, medium-, and high-demand forecasts, covering a range of relevant values of load growth. In summary, no single, universally superior method of load forecasting exists. The more sophisticated techniques tend to obscure the basic methodological assumptions as well as weaknesses in the data. Therefore, the simpler techniques may be more appropriate when the available information is not reliable. The determinants of demand must be carefully selected because they vary from case to case, and the level of disaggregation chosen should depend on the data, as well as on the ultimate purpose of the projection. Demand forecasting is a dynamic process in the sense that the procedure must be repeated often, with improved data and analytical techniques. Creating a reliable data bank is an invaluable tool in this process of constant revision. Finally, in some cases, institutionally imposed constraints that suppress demand-such as peak suppression through load shedding or not connecting customers-and the existence of privately owned generation may lead to a substantial gap between the potential and the realistic demand.

System Planning
The costs used in system planning and, subsequently, in pricing should be based on a good physical model of the power system under review, using shadow prices where appropriate. Some analysts have recently taken the al10. See, for example, Nuclear Power Planning Study for Indonesia (Vienna: IAEA, 1976); N. R. Prasad, J. M. Perkins, and G. Nesgos, "A Markov Process Applied to Forecasting: Part 1, Economic Development," Proceedings of the IEEE Power Engineering Society Summer Meeting, Vancouver, July 1973; and N. R. Prasad, J. M. Perkins, and G. Nesgos. "A Markov Process Applied to Forecasting: Part 11,The Demand for Electricity," Proceedings of the IEEE Power Engineering Society Winter Meeting, New York, January 1974.

PREREQUISITES FOR MARGINAL COSTING

35

ternative approach of estimating neoclassical long-run production functions of electric power systems to determine the cost structure, economies of scale, technical progress, factor substitution possibilities, and so on." The Cobb-Douglas, constant elasticity of substitution (CES) and trans-logarithmic formulations have been used. No generalized production model derived from neoclassical economic theory could hope to capture the detailed information of the physical system that is included in an engineering planning model. Sector-wide production fianctions based on cross section (or pooled, cross section, and time series) data are appropriate to determine historical trends in the economics of power sector evolution, rather than to represent a given system accurately. Production functions estimated for a single system, using time series data, tend to be more specific, but they are still based on pastrather than future-oriented data. In summary, an engineering model used in the design and planning of a system, as described below, is far more appropriate than an economic production function model for estimating future marginal supply costs for a power system and an appropriate price structure. The principal economic criteria for optimizing price is to maximize the net benefit accruing to the whole society, rather than to the power utility alone. Therefore, the strict LRMC to be used to set tariffs should be based on goods and services used as physical inputs to the electrical power system. These include labor, land, physical capital, and materials, which are shadow priced as scarce economic resources that could be used in altemative production. The basic principles of shadow pricing appropriate to the electricity sector are described in the concluding part of this chapter. Shadow pricing is particularly useful for publicly owned power utilities, which are common in developing countries. The alternative, which is the financial viewpoint of private electric power utilities similar to those in the United States, is not ap2 propriate.' In this case, the items to be included in the costing exercise are based primarily on accounting concepts-physical assets, services, depreciation, interest, and taxes--that may be affected by regulatory requirements. The values placed on them are the private financial costs incurred by the utility, which may be quite inadequate as indicators of economic costs. Ideally, the planning of the whole system should be integrated. To make the practical problems of system design workable, however, it is convenient to break the procedure down into parts associated with the main subsystems:
11. See Phoebus J. Dhrymes and M. Kurtz, "Technology and Scale in Electricity Generation," Econometrica, vol. 32 (July 1965), pp. 287-315; M. Galatin, Economies of Scale and Technical Change in Thermal Power Generation (Amsterdam: North-Holland, 1968); Fred McCoy, "The Production Function, Method 1," in Marginal Costing and Pricing of Electrical Energy; and Laurits Christensen, "The Production Function, Method I," in Marginal Costing and Pricing of Electrical Energy. 12. See James Suelflow, Public Utility Accounting: Theory and Application (East Lansing, Mich.: Michigan State University, 1973).

36

THEORY

3 This type of hierarchical reducgeneration, transmission, and distribution.' tion is logical from several points of view. As was noted in the previous section on load forecasting, various completion times are associated with the discrete projects in a long-range plan for system expansion. These range from up to ten years for the design and construction of a large nuclear generating plant to one or two years for distribution schemes. Also, the problems and design philosophies associated with the different parts of the total power system are inherently dissimilar. For example, distribution system planners require detailed knowledge of localized geographic areas and must deal with many small components such as distribution transformers, switches, and reclosers. At the other extreme, generation planning is carried out on a global level and is concerned with large components such as turbines, and these generally have interaction modes that are more complex to analyze. Consequently, the models and criteria now used to design the various subsystems are also different. The following nonmathematical discussion is cast in the context of traditional system expansion planning, that is, choosing the design that minimizes system cost while meeting given requirements for load and reliability. The basic ideas involved, however, may be readily adapted to the new approach in which the reliability level is treated as a variable, and the total cost to soci4 ety, or sum of system and shortage costs, is minimized.' Early approaches to system design and planning were rather intuitive and relied heavily on the designers' past experience. In contrast, the mathematical methods currently used rely on optimization models. The goal is to minimize an objective function, usually the present discounted value of system cost subject to various constraints, such as meeting targets for load and reli5 ability, and maintaining technical and operating requirements.' In general, if the problem is deterministic, with known objective function and constraints-that is, best estimates of plant costs, load, and demand in generation planning-then some definitive nominal design may be found. With uncertainty, flexible design strategies are required. In the worst-case strategy, upper and lower design limits are determined that correspond to

13. For details of power system design and planning, see Sullivan, Power System Planning; and Upton G. Knight, Power System Engineering and Mathematics (New York: Pergamon Press, 1972). For a discussion of hierarchical systems see D. A. Wismer, Optimization Methods for Large Scale Systems (New York: McGraw-Hill Book Company, 1971). 14. See Chapter 2 and Mohan Munasinghe, The Economics of Power System Reliability and Planning (Baltimore, Md.: Johns Hopkins University Press, 1979). 15. For a more detailed and mathematical treatment. see Application of Optimization Methods in Power System Engineering, IEEE tutorial course, text no. 76CH1 107-2-PWR (New York: IEEE, 1976); and R. Rischl, "Optimal System Expansion Planning: A Critical Review," Systems Engineering for Power: Status and Prospects, Proceedings of the USERDA Conference, Henniker, N.H., August 1975, NTIS no. CONF-750867, pp. 233-60.

PREREQUISITES FOR MARGINAL COSTING

37

maximum and minimum values of the objective function and constraints. Given a probabilistic distribution of values of the objective function and constraints, the statistical technique results in the average design configuration emerging from among the design variations. The specific techniques for solving the optimization problem may be considered under three categories: programming, search, and other special procedures. The mathematical programming approaches, which include linear, nonlinear, integer, mixed-integer, and dynamic programming versions, are most useful when the objective function is simple and well defined. They rely on explicit (computerized) algorithms that are used to find the solution after a finite number of steps. Search techniques, which may be used when programming methods are difficult to apply, are generally easier to implement, but they do not guarantee convergence to the solution within a reasonable time. Special procedures are often elegant, but they are usually applica6 ble only to particular problems." Generation The basic questions in the design of generation systems involve the size,
timing, type (or mix), and location of generating plants. The economic as-

pects of each of the choices involved are clarified below by using a simple approach. Figure 3-3 depicts the problem of size and timing resulting from the growth of peak megawatt (MW) demand over time in a typical system with limited capacity. The installed capacity of existing generating plant MW will not be sufficient to meet the demand in two years, allowing for a suitable reserve margin. Two expansion altematives are available that meet all the constraints. In th,e first, two medium-size generating plants would come on stream in the second and fourth years. In the second option, a small plant would be commissioned in the second year, followed by a large power station in the sixth year. The system planner must choose the cheaper alternative. In a real system, the expansion paths over a longer period of time would be compared. In this example the two alternative plans have been made roughly equivalent after the sixth year to highlight the nature of the choice of size and timing of the next two generation additions. Moreover, in predominantly hydroelectric systems, energy shortages caused by limited storage space may precede any capacity constraints. The selection of an appropriate plant mix is best discussed in the context of the annual load duration curve (LDC) for a system. The LDC for a given year is shown in Figure 3-4, in which the hourly megawatt demand of the system is plotted against the number of hours of the year during which this

16. See, for example, L. R. Ford and D. R. Fulkerson, Flows in Networks (Princeton, N.J.: Princeton University Press, 1962).

38

THEORY

Figure 3-3. Size and Timing of Additional Generating Plants Needed to Meet the Peak Demandfor Power

Alternative 2

Alternative]

Peak megawatt

demand

11f1+

1. 4 Years

level of demand is equaled or exceeded. The peak demand MW occurs only over a short period of time. Since the average level of megawatt demandthat is, the total energy in megawatt-hours or area under the LDC curve divided by 8,760 hours-is less than the peak demand, the load factor will be less than unity. For simplicity, only two kinds of generating units are considered: gas turbines and base-load steam plants. The relevant characteristics of both these types of machines are shown by the solid lines in the upper diagram of Figure 3-4, which is a linearized graph of the total average cost per megawatt of installed capacity, that is, capital cost plus operating cost, plotted as a function of the number of hours of operation. Gas turbines have lower capital costs, represented by the intercept a, which is the investment cost of a megawatt, annuitized over the lifetime of the machine. They have higher fuel costs than steam plants, however, as indicated by the slope b, of the operating cost curve. First, consider the problem of determining the amounts of new gas turbine and base-load steam capacity that must be built to meet the given load, if the

PREREQUISITES

FOR MARGINAL

COSTING

39

Figure 3-4. Types of Generating Plants Needed to Meet the Annual Load Duration Curve
New gas turbines(slopeb,) Existingplant (slopeB3)

0.

aZ, t

~~~~~ New

base-load steamn plant (slope b2)

~;

I H

Hours per year of plant operation

,=DX X,.S

_
_ _

0 H 8,760 Hours per year during whichthe givenlevel of demandis equalledor exceeded

existing generating plant is ignored. From the economic point of view, gas turbines are more expensive than base-load units if they are to be used more than H hours a year. This essentially static picture is only illustrative, because the relevant streams of present discounted investment and the operating costs of the machines would have to be compared over many years. To pursue this further, the lower diagram indicates that it would be economical to serve the bottom X 2 megawatts (the base load) with steam units and the

40

THEORY

upper X, megawatts (the peak load) with gas turbines. A reserve margin may be included if necessary. Next, the modifications required in generation expansion planning because of old or existing plants are illustrated using a simple example. Suppose that Y3 megawatts of generating capacity exist already. This plant has an operating cost given by the slope B, in Figure 3-4. To clarify the representation, assume b, > B3 > b2. The capacity costs of the existing plant have already been incurred, and the planner must determine the optimal point of operation for these Y 3 megawatts on the LDC. This is done geometrically by moving the line parallel to itself up or down on the diagram until the vertical distance between the appropriate points F and G on the LDC is equal to Y3 megawatts. The intercept A, is the shadow value of capacity of the existing plant, that is, the equivalent saving in capacity costs for the system resulting from use of the existing generating plant. The new plan requires full use of the Y3 megawatts of existing capacity and X', and Y', megawatts, respectively, of new gas turbines and base-load steam units to be constructed to meet the LDC. Since X', i X, and X' 2 i X, the existence of old capacity has altered the optimal mix of new plants required. In more complicated cases, if some of the existing capacity is inefficient and has very high operating costs, the optimal plant mix may include only part or none of this old plant. In practice, many kinds of generating machines of various sizes would have to be taken into account including nuclear; coal and oil-fired steam; gas turbine; conventional and pumped-storage hydro; and, more recently, geothermal, tidal, and wind-powered units. When generation additions were being considered to meet future load growth, however, the basic principle of selecting units with lower capital but higher operating costs for peaking purposes-and the opposite for nonpeaking operation-would generally hold true. Transmission Usually, the problem of locating the generators is closely linked with the design of the transmission system; therefore, they should be discussed in the same context. Transmission planning typically involves solving a dynamic network-type problem. The basic objective is to select the type, timing, and location of additional lines that would connect the various generating stations to the different load centers in the least expensive way. Requirements involving the load, reliability level, and other constraints must be satisfied in addition. A commonly used network model is the dc load flow model, which may be used in either a linear or nonlinear framework. In the single-step approaches that are essentially static, the design optimization is repeated at

PREREQUISITES

FOR MARGINAL

COSTING

41

various points. When generation and load levels are specified-for example, in the last year of the planning horizon-the repetition provides instantaneous snapshots of the system's evolution. Long computing times are required in dynamic methods to handle transmission line variations in both space and time, as in the analysis of a large interconnected area during each year of the planning period. The problem of time is simplified by preselecting several expansion alternatives according to certain performance criteria. None of these techniques can guarantee a definitive solution, however, given the size of the problem and limits on calculating times.' There are two general types of criteria for transmission system performance: steady state and transient.1 Steady state performance of the system consists basically of verifying that line and transformer design limits are not exceeded under normal loading and operating conditions or with single contingencies, such as loss of one line, transformer, or generator. Transient performance criteria are used to test the model for stability in various situations. One group of tests involves examining how well an interconnected system would maintain steady state stability under both normal and contingency conditions, when excited by a small disturbance such as the switching on of a minor load. Verifying the system's dynamic stability is also important. This involves the degree of damping of high amplitude oscillations induced by a large fault, such as failure of a major line because of a three-phase fault to ground or the loss of a block of generators. A return to the question of generator location is appropriate at this point. The analysis of combined generation and transmission system models raises formidable problems. Therefore, although the generation and transmission subsystems are intimately related, separate models are generally used for the design of each subsystem. More specifically, a typical large computerized generation planning model would yield the appropriate plant mix, size, and timing. Sometimes a limited and often inaccurate representation of the transmission network is included.'9 The optimal locations to supply the specified load centers would be chosen from among several potential sites within the context of a more detailed transmission planning model. In practice, it would be necessary to iterate between the models to derive the best mutually consistent solution.
17. See Brian Stott, "Review of Load-Flow Calculation Methods," Proceedingsof the IEEE, vol. 62 (July 1974), pp. 916-29. 18. For a more detailed discussion of criteria for transmission design, see Canadian International Development Agency, "Proceedings of the Regional Power Systems Engineefing Seminar, Kathmandu, Nepal, February-March 1977"; processed. 19. See, for example, D. Anderson, "Models for Determining Least Cost Investments in Electricity Supply," Bell Journal of Economics and Management Science, vol. 3 (Spring 1972), pp. 267-301.

42

THEORY

Distribution Although planning generation and transmission systems is an interrelated process, the distribution grid serving a localized geographic area or load center can be designed independently. Recently, computerized methods have been developed to determine the optimal size, timing, and location of distri20 bution substations and primary feeders. Sophisticated models and computerized algorithms, however, are not as widely used in the design of distribution systems as in generation and transmission work. In the more conventional approach, forecast loads are imposed on the existing distribution system at various times in the future, and the network is systematically strengthened to meet these loads adequately. The location of new substations and primary feeders and the upgrading of existing ones would be determined to meet consumer demand within acceptable voltage limits under both normal and emergency conditions, as well as during peak and off-peak periods. Flexibility in the design to facilitate isolating faulted sections of feeders and to permit customers to be switched from faulted to functioning feeders would be an important consideration. Often, because of the large numbers of customers and components such as line sections, distribution transformers, and so on, only certain representative feeders would be analyzed in detail to establish general criteria such as maximum kilowatt-amps for a circuit. Such criteria could then be applied to other feeders without further detailed study. In general, the larger and more complex the service area, the greater would be the advantage of using even simple computerized techniques for circuit-by-circuit analysis.

Shadow Pricing
Usually, when market imperfections occur, shadow prices may be used instead of distorted market prices to represent the true economic resource costs of various goods and services in an economy. These shadow or social accounting prices stem from welfare economic theory and from cost-benefit analysis. If perfect competition existed, a Pareto-optimal situation would result as many individual profit-maximizing producers and utility-maximizing consumers interacted. In this ideal state, prices reflect the true marginal social costs, scarce resources are efficiently allocated, and for a given income dis20. See S. B. Holt, Jr., and D. M. Crawford, "Distribution Substation Planning Using Optimization Methods," in Application of Optimization Methods in Power System Engineering,
IEEE tutorial

course, pp. 69-76.

PREREQUISITES

FOR MARGINAL

COSTING

43

tribution no individual can be made better off without making someone else 2 worse off. " Conditions in the real world are different, however. Distortions caused by monopoly practices; external economies and diseconomies; interventions in the market process through taxes, duties, and subsidies; and so on result in market prices for goods and services that may be substantially different from their shadow prices or true economic values. If large income disparities exist, the passive acceptance of the existing skewed income distribution, which is implied by the reliance on strict efficiency criteria for determining economic welfare, may be socially and politically unacceptable. These considerations require appropriate shadow prices instead of market prices for economic analysis, especially in the developing countries where market distortions are more prevalent. Consider a general equilibrium model of the economy in which the national goal is embodied in an acceptable objective function such as aggregate consumption. This consumption is to be maximized subject to constraints that might include limits on resource availabilities, distortions in the economy, and so on. Then, the shadow price of a given scarce economic resource represents the change in the value of the objective function caused by 22 a marginal change in the availability of that resource. In the more specific context of a mathematical programming macroeconomic model, it is possible to formulate the so-called dual problem. The dual variables correspond to the mathematical constraints on the availability of resources in the original or primal problem. The optimal values of these dual variables have di23 mensions of price and could be interpreted as shadow prices. Two basic types of shadow prices exist. These involve whether or not society is indifferent to considerations of income distribution. To illustrate this point, consider the simple national goal of maximizing the present value of aggregate consumption for a given time period. If the consumption of different individuals is added (lirectly without regard to income level, then the shadow prices derived from such a model are termed efficiency prices because they reflect the pure efficiency of resource allocation. When the con21. See Francis J. Bator, "General Equilibrium, Welfare, and Allocation," American Economic Review (March 1957), pp. 22-59; and Ezra J. Mishan, Cost-Benefit Analysis (New York: Praeger, 1976), part 7. 22. See Partha Dasgupta and Joseph Stiglitz, "Benefit-Cost Analysis and Trade Policies," Journal of Political Economy, vol. 82 (January-February 1974), pp. 1-33; and Peter Warr, "On the Shadow Pricing of Traded Commodities," Journal of Political Economy, vol. 85 (1977), pp. 865-72. 23. See David G. Luenberger, Introduction to Linear and Non-Linear Programming (Reading, Mass.: Addison-Wesley, 1973), chap. 4; and Peter G. Sassone, "Shadow Pricing in CBA: Mathematical Programming and Economic Theory," Engineering Economist, vol. 22 (Spring 1977), pp. 219-33.

44

THEORY

sumption of the lower income groups is of particular interest, this consideration is given a greater weight in evaluating aggregate consumption, and the 24 resultant shadow prices are called social prices. There are two practical difficulties. First, some analysts object to using social pricing in project work on the grounds that weighting consumption benefits by income category is necessarily arbitrary and that income is redistributed more efficiently through fiscal and other policy instruments rather than by biasing investment decisions. But acceptance of efficiency prices implies that equal weights are arbitrarily attached to benefits accruing to all individuals. Also, in many countries the government may have few practical 25 policy tools available to influence the existing income distribution. In practice, the use of social versus efficiency prices depends on factors such as availability of data and the commitment of the government to equalizing income distribution. Second, estimating shadow prices with the full-fledged general equilibrium model referred to in the theory is difficult, because many countries do not have the requisite data. Reanalyzing such a large model whenever new investment decisions are required would be wasteful and time consuming. Instead, partial equilibrium techniques may be used that evaluate the effect of the change in the availability of a given resource by its effect on a few key areas, rather than throughout the economy. The partial equilibrium method has been criticized on the basis of nonmarginality, for its lack of simultaneity, and because of the element of judgment involved in choosing which linkages to analyze and which to neglect when the effect of some disturbance is to be evaluated. These objections can be overcome by using an iterative process of successive (convergent) approximations and by considering enough linkages that the partial equilibrium shadow prices closely approximate the corresponding general equilibrium values. Numeraire To derive a consistent set of economic prices for goods and services, a common yardstick or numeraire to measure value is necessary, as illustrated by a simple example. If one wishes to compare avocados with coconuts, the equivalent units might be either one avocado for one coconut, or one kilo of
24. See Lyn Squire and Herman G. van der Tak, Economic Analysis of Projects (Baltimore, Md.: Johns Hopkins University Press, 1975), chap. 6. 25. See Partha Dasgupta. Amartya Sen, and Stephen Marglin, Guidelines for Project Evaluation (UNIDO) (New York: United Nations, 1972), chap. 7; Ian M. D. Little and James A. Mirrlees, Project Appraisal and Planning for Developing Countries (New York: Basic Books, 1974), chap. 4.

PREREQUISITESFOR MARGINAL COSTING

45

avocados for one kilo of coconuts. In the first instance, the common yardstick is one fruit; in the second, it is the unit of weight. Clearly, if the weights of the two types of fruits are different, the result of the comparison will depend on the numeraire used. With a numeraire of economic value the situation is more complicated, because the same nominal unit of currency may have a different value depending on the economic circumstances in which it is used. For example, one rupee's worth of a certain good purchased in a duty-free shop is likely to be more than the physical quantity of the same good obtained for one rupee from a retail store, after import duties and taxes have been levied. Therefore, it is possible to distinguish intuitively between the border-priced rupee, which is used in intemational markets free of import tariffs, and a domesticpriced rupee, which is used in the domestic market subject to various distortions. A more sophisticated example of the differences in value of a currency unit in various uses arises in countries where the aggregate investment for future economic growth is considered inadequate. In these instances, a rupee of savings that could be invested to increase the level of future consumption may be considered more valuable than a rupee devoted to current consumption. The choice of the numeraire, like the choice of a currency unit, should not influence the economic criteria for decisionmaking except in relation to magnitude, provided the same consistent framework and assumptions are used in the analysis. For example, only one difference exists between a study using cents as units and one using rupees (where the rupee is defined as 100 cents). In the study using cents all monetary quantities will be numerically 100 times larger than in the one using rupees. Therefore, a numeraire may be selected purely on the basis of convenience of application. A most appropriate numeraire in many instances is a unit of uncommitted public income at border prices. Essentially, this unit is the same as freely disposable foreign exchange available to the govemment, but expressed in
terms of units of local currency converted at the official exchange rate.26 The

discussion in the next section is developed in relation to this particular yardstick of value. The border-priced numeraire is particularly relevant for developing countries, which have scarce foreign exchange. It represents the set
26. Little and Mirrlees, Project Appraisal and Planning for Developing Countries, chap. 9; and Squire and van der Tak, Economic Analysis of Projects, chap. 3. A numeraire based on private consumption is advocated in Dasgupta, Sen, and Marglin, Guidelines for Project Evaluation. For a discussion of the different assumptions used in various approaches to shadow pricing, see Harry Schwartz and Richard Berney (eds.), Social and Economic Dimensions of Project Evaluation (Washington., D.C.: Inter-American Development Bank, 1977); and Deepak Lal, Methods of Project Analysis, World Bank Staff Occasional Papers no. 16 (Baltimore, Md.: Johns Hopkins University Press, 1974).

46

THEORY

of opportunities available to a country to purchase goods and services on the international market. Applying shadow prices The estimation and use of shadow prices is facilitated by dividing economic resources into tradable and nontradable items. The values of directly imported or exported goods and services are already known in border prices, that is, their foreign exchange costs converted at the official exchange rate. Locally purchased items whose values are known only in terms of domestic market prices, however, must be converted to border prices, by multiplying the former prices by appropriate conversion factors (CF). Therefore, tradables and nontradables are treated differently.
BORDER PRICES. For tradables with infinite elasticities-of world supply for imports and of world demand for exports-the cost, insurance, and freight (c.i.f.) border price for imports and the free-on-board (f.o.b.) border price for exports may be used with a suitable adjustment for the marketing margin. If the relevant elasticities are finite, then the change in import costs or export revenues, as well as any shifts in other domestic consumption or production levels or in income transfers should be considered (see Appendix D). The free trade assumption is not required to justify the use of border prices since domestic price distortions are, in effect, adjusted by netting out all taxes, duties, and subsidies. To clarify this point, imagine a household in which a child is given an allowance of 20 pesos a month as pocket money. The youngster may purchase a bag of lollipops from the grocery store for 2 pesos. Since the parents want to discourage consumption of sweets, however, they impose a fine of I peso on each bag of lollipops. The fine is exactly like an import duty, and the child must surrender 3 pesos for every bag of candy (valued at its domestic price, inside the household). From the family's perspective, however, the total external payment for the item is only 2 pesos, because the 1 peso fine is a net transfer within the household. Therefore, the true economic cost, or shadow price, of the bag of lollipops (valued at its border price) is 2 pesos, when the effect of the fine on the distribution of income between parents and child is ignored. There are two types of nontradable economic resources. A nontradable is conventionally defined as a commodity whose domestic supply price lies between the f.o.b. and c.i.f. prices for export and import, respectively. Items that are not traded at the margin because of prohibitive trade barriers, such as bans or rigid quotas, are also included within this category. If the increased demand for a given nontradable good or service causes the domestic

PREREQUISITES FOR MARGINAL COSTING

47

supply or imports to expand, the associated border-priced marginal social cost (MSC) of this increased supply is the relevant resource cost. If consumption of other domestic or foreign users decreases, the border-priced marginal social benefit (MSB) of this forgone domestic consumption or of reduced export earnings would be a more appropriate measure of social costs. The socially optimal level of total consumption for the given input (Qp,) would lie at the point where the curves of MSC and MSB intersect. The differences between MSB and MSC arise from price and other distortions that lead to levels of consumption below the optimal level Q i Qp,. More generally, if both effects are present, a weighted average of MSC and MSB should be used. The MSB would tend to dominate in a short-run, supplyconstrained situation; the MSC would be more important in the longer run, 27 when expansion of output is possible. Clearly the MSC of nontradable goods and services from many sectors can be determined through appropriate decomposition. For example, suppose 1 rupee's worth, in domestic prices, of the output of the domestic construction sector may be broken down successively into n components. This would include capital, labor, and materials, which are valued at rupees a,, a2, * , a_, in border prices. Since the conversion factor of any good is defined as the ratio of the border price to the domestic price, the construction conversion factor equals i=,, a,. The standard conversion factor (SCF) may be used with nontradables that are not important enough to merit individual attention or lack sufficient data. The SCF is equal to the official exchange rate (OER) divided by the more familiar shadow exchange rate (SER), appropriately defined. Converting domestic prices into border price equivalents by applying the SCF to the value is conceptually the inverse of the traditional practice of multiplying foreign currency costs by the SER (instead of the OER) to convert to the equivalent domestic price. The SCF- may be approximated by the ratio of the OER to the free trade exchange rate (FTER), when the country is moving toward a freer trade regime: SCF = OERIFTER = (eX + nM)l[eX(l - t) + nM(l + t,j] where X = f.o.b. value of exports; M = c.i.f. value of imports; e = elasticity of export supply; n = elasticity of import demand; t, = average
27. For details relevant to the electricity sector, see Mohan Munasinghe and Jeremy Warford, "Shadow Pricing and Power Tariff Policy," in Marginal Costing and Pricing of Electrical Energy; and Munasinghe, The Economics of Power System Reliability and Planning, chap. 9 and app. B.

48

THEORY

tax rate on exports (is negative for subsidy); and t. = average tax rate on
imports.2 "
TRADABLE AND NONTRADABLE INPUTS. The most important tradable inputs used in the electric power sector are capital goods and petroleum-based fuels. Some countries may have other domestic fuels available, such as natural gas or coal deposits. If no clear-cut export market exists for these indigenous energy resources, they cannot be treated as tradables. In addition, if there is no alternative use for the fuels, an appropriate economic value is the MSC of production, that is, of extracting gas or coal plus a markup for the 29 discounted value of future consumption forgone, or "user cost." If another high value use exists for this fuel, the opportunity cost of not using the resource in the alternative use should be considered as the economic cost of the fuel. The most important nontradable primary factor inputs are labor and land. Consider a typical case of unskilled labor in a country with surplus laborfor example, rural workers employed for dam construction. The forgone output of workers used in the electric power sector is the dominant component of the shadow wage rate (SWR). Complications arise because the original rural income earned may not reflect the marginal product of agricultural labor, and, furthermore, for every new job created, more than one rural worker 30 may give up former employment. Allowance must also be made for seasonal activities, such as harvesting. In theory, if the laborer has to work harder in a new job than he did before, then the disutility of forgone leisure should be included in the SWR, but in practice this component is ignored. Overhead costs incurred, such as transport expenses, should also be considered. The foregoing may be summarized by the following basic equation for the efficiency shadow wage rate (ESWR):

ESWR = a m + c u

where m and u are the forgone marginal output and overhead costs of labor in domestic prices, and a and c are corresponding conversion factors to con28. For a more detailed discussion of alternative interpretations of the SCF and SER, see Squire and van der Tak, Economic Analysis of Projects; and Bela Balassa, "Estimating the Shadow Price of Foreign Exchange in Project Appraisal," Oxford Economic Papers, vol. 26 (July 1974), pp. 147-68. 29. For a review of practical energy pricing rules, especially with respect to nonrenewable energy sources, see Mohan Munasinghe, "An Integrated Framework for Energy Pricing in Developing Countries," Energy Journal, vol. I (July 1980), pp. 1-30; also available as Reprint no. 148 from the World Bank, Washington, D.C. 30. See John R. Harris and Michael P. Todaro, "Migration, Unemployment, and Development," American Economic Review, vol. 60 (March 1970), pp. 126-42.

PREREQUISITES FOR MARGINAL COSTING

49

vert these values into border prices.

Consider the effect of these changes on consumption patterns. Suppose a worker receives a wage W, in a new job and that the income forgone is W,, both in domestic prices; W, may not necessarily be equal to the marginal product forgone m. It could be assumed, quite plausibly, that low-income workers consume the entire increase in income (W, - W). Then this increase in consumption will result in a resource cost to the economy of b(W1 - W.). The increased consumption also provides a benefit given by w(W, - W), where w represents the MSB, in border prices, of increasing domestic-priced private sector consumption by one unit. Therefore,
SWR = a m + c u + (b w) (W. WO).

The letter b represents the MSC to the economy resulting from the use of the increased income. For example, if all the new income is consumed, then b is the relevant consumption conversion factor or resource cost (in units of the numeraire) of making available to consumers one unit's worth (in domestic prices) of the marginal basket of goods they would purchase. In this case b = gj CF,, where gi is the proportion or share of the ith good in the marginal consumption basket, and CFi is the appropriate conversion factor. The corresponding MSB of increased consumption may be decomposed further:
w = dlv

where l/v is the value (in units of the numeraire) of a one-unit increase in domestic-priced consumption accruing to someone at the average level of consumption e. Therefore, v may be roughly thought of as the premium attached to public savings, compared with "average" private consumption. Under certain simplifying assumptions, b = 1/v. If MU(c) denotes the marginal utility of consumption at some level c, then
d = MU(c)/MU(c).

Assuming that the marginal utility of consumption is diminishing, d would be greater than unity for poor consumers with c < e, and vice versa. A simple form of marginal utility function which could be assumed is: MU(c) = c-'. Therefore, d = MU(c)IMU(e) = (elc)'. Making the further assumption that the distribution parameter n = 1, gives d
= elc =

Ill

50

THEORY

where 111is the ratio of net incomes, which may be used as a proxy for the corresponding consumption ratio. The consumption term (b - w) in the expression for SWR disappears if, at the margin (a) society is indifferent to the distribution of income-or consumption-so that everyone's consumption has equivalent value; and (b) private consumption is considered to be as socially valuable as uncommitted public savings, that is, the numeraire. The appropriate shadow value placed on land depends on its location. Usually, the market price of urban land is a good indicator of its economic value in domestic prices, and the application of an appropriate conversion factor, such as the SCF, to this domestic price will yield the border-priced cost of urban land inputs. Rural land that can be used in agriculture may be valued at its opportunity cost, the net benefit of forgone agricultural output. The MSC of other rural land is usually assumed to be negligible, unless there is a specific reason to the contrary. Examples might be the-flooding of virgin jungle because of a hydroelectric dam that would involve the loss of valuable timber, or spoilage of a recreational area that has some commercial potential. The shadow price of capital is usually reflected in the discount rate or accounting rate of interest (ARI), which is defined as the rate of decline in the value of the numeraire over time. Although there has been much discussion concerning the choice of an appropriate discount rate, in practice the opportunity cost of capital (OCC) may be used as a proxy for the ARI in the pure 3 efficiency price regime. " The OCC is defined as the expected value of the annual stream of consumption, in border prices net of replacement, which is yielded by the investment of one unit of public income at the margin. A simple formula for ARI, which also includes consumption effects, is given by:
ARI = OCC [s + (I - s)wlb]

where s is the fraction of the yield from the original investment that will be saved and reinvested. If the LRMC is calculated by using border prices to shadow price the inputs, a further adjustment is necessary to convert this shadow-priced LRMC into an equivalent optimal electricity price that will be perceived and interpreted by consumers like any other domestic market price. Thus, if only economic efficiency is considered, a typical expression for the optimal market price based on LRMC would be
31. See, for example, Stephen Marglin, "The Social Rate of Discount and the Optimal Rate of Investment," Quarterly Journal of Economics (1963); Mishan, Cost-Benefit Analysis, chaps. 31-34; and Arnold C. Harberger, Project Evaluation (London: Macmillan and Company, 1972).

PREREQUISITES

FOR MARGINAL

COSTING

51

Pe* = (MCB,lb) where MCB, is the LRMC of electricity in border prices. The value of b for a given consumer group depends on the expenditure pattern of these electricity users, as described earlier. Therefore, it is possible to have different prices pe* for various categories of electricity consumers, such as residential and industrial users, based on the same value of MCB,. It may sometimes be simpler to use an economy-wide average value of b, such as the SCF, for all electricity consumers. This is especially true when detailed information on different consumer categories is unavailable. Since the SCF usually is less than 1, the optimal market price is greater than the border-priced LRMC, pe* > MCB. 32 Usually it is a long and complex task to estimate shadow prices on a rigorous basis. Therefore, the analyst in the electric power sector is best advised to use whatever shadow prices have already been calculated. Alternatively, the analyst would estimate a few important items such as the SCF, OCC, and SWR. When the data are not precise enough, sensitivity studies may be made over a range of values of such key national parameters.
32. A simple numerical example illustrates this point. Suppose that, based on the world import price for high-speed diesel fuel and the running costs of peaking gas turbines, the borderpriced LRMC of peak period energy is MCB, 1.6 pesos a kilowatt-hour. That is, 8 cents a kilowatt-hour, times the OER of 20 pesos = I dollar. Let the appropriate SER, which reflects the average level of import duties and export subsidies (for example, FTER), be 25 pesos = I dollar. Therefore SCF = OERISER = 0.8, and the optimal market price for peak period energy: p,* = 1.6/0.8 = 2 pesos a kilowatt-hour.

Chapter4

StrictLong-Run Marginal Cost

MARGINAL COST (LRMC) may be defined broadly as the incremental cost of all adjustments in the system expansion plan and system operations attributable to an incremental increase in demand that is sustained into the future. Usually the term "increment" conveys the concept of a small but discrete amount of cost or demand, whereas the term "marginal" is often interpreted in the strictly mathematical sense of an infinitesimal change. In the present context, incremental and marginal may be used interchangeably to denote a discrete change. LRMC must be evaluated within a disaggregated framework, however. This structuring of LRMC is based chiefly on technical grounds and may include breaking down marginal cost by time of day, voltage level, geographic area, season of the year, and so on. The degree of structuring and sophistication of the LRMC calculation depends on data constraints and the usefulness of the results, given the practical problems of computing and applying a complex tariff. In theory, the LRMC of each individual consumer at each moment of time can be estimated. In the following calculation of the strict LRMC, the structuring framework is limited to one that has operational value in a typical developing country. Points at which the computation may be pursued at a more sophisticated level are indicated in the text. The methodology of computing the strict LRMC is summarized in this chapter; practical problems and more detailed numerical examples are presented in the case studies in Part Two.' THE STRICT LONG-RUN

Cost Categories and Pricing Periods

Three broad categories of marginal costs may be identified for the LRMC calculations: capacity costs, energy costs, and consumer costs. Marginal capacity costs are basically the costs of investment in generation, transmission, and distribution facilities to supply additional kilowatts. Marginal energy
1. For other case studies that emphasize the calculation of the strict LRMC in various situations in developing countries, see Ralph Turvey and Dennis Anderson, Electricity Economics (Baltimore, Md.: Johns Hopkins University Press. 1977).

52

STRICT LONG-RUN

MARGINAL

COST

53

costs are the fuel and operating costs needed to provide additional kilowatthours from a thermal plant, whereas in a hydroelectric system a part of the investment cost associated with storage may be related to energy. Marginal customer costs are the incremental costs directly attributable to consumers, including costs of hook-up, metering, and billing. Wherever appropriate, these elements of LRMC must be broken down by time of day, voltage level, and so on. The first step in structuring is to select appropriate pricing or rating periods. The system load duration curves (LDC) and generation schedules should be examined to determine periods during which demand presses on capacity and supply costs are highest. These cyclical critical periods may be caused by daily demand variations (for example, evening lighting load), or seasonal variations in both demand (for example, summer air-conditioning peak load) and supply (for example, dry season for hydro systems). The illustration of the principles of structuring and calculating strict LRMC begins with a thermal system, in which demand is not greatly affected by the seasons. Only two pricing periods by time of day-peak and off-peak-are chosen. Seasonal variations in LRMC and the analysis of hydroelectric systems are discussed later.

Marginal Capacity Costs

Consider, in Figure 4-1, the typical system annual LDC, ABEF, for the starting year 0 divided into two rating periods, peak and off-peak. As demand grows over time, the LDC increases in size, and the resultant forecast of peak demand is given by the curve D in Figure 4-2, starting from the initial value MW,. The LRMC of capacity may be determined by asking the following question: what is the change in system capacity costs AC associated with a sustained increment AD in the long-run peak demand? This is shown by the shaded area of Figure 4-1 and the broken line D + AD in Figure 4-2. Consequently, the LRMC of generation would be AC/AD, where the increment of demand AD is marginal both in time and in megawatts. In theory AD can be either positive or negative, that is, both increments and decrements should be considered symmetrically. Generally the ratio ACIAD will vary with the sign as well as the magnitude of AD. If many such values of ACIAD are computed, they can be averaged to obtain LRMC. The easiest procedure, however, would be to consider only a single representative positive increment of demand. In an optimal system the change in the program for expanding generation to meet the new incremental load would normally consist of advancing the commissioning date of a future plant or inserting new units, such as gas turbines or peaking hydro plants (see Figure 4-2). If system planning involves a

54

THEORY

Figure 4-1. Typical Annual Load Duration Curve

0)

0
Peak

B~~~~~

.4

Off-peak

8,760

Hours per year

Figure 4-2. Forecast of Demand for Peak Power

Advanced capaci'ty /D

+ AD

Originally planned
,

capacity

MWo-

-----l l

Years

STRICT LONG-RUN

MARGINAL

COST

55

computerized model, it is relatively easy to determine the change in capacity costs AC by simulating the expansion path and system operation with and without the demand increment (or decrement) AD. If a more sophisticated tariff structure having more rating periods is used, then the LRMC in any rating period may be estimated by running the computerized model for system expansion with a sustained load increment added to the LDC during that period. This method that simulates the optimal planning process is based on the dynamic LRMC concept. When constraints relating to time, data, and computer facilities preclude this ideal procedure, it is usually possible to use simple considerations to derive marginal generation capacity costs based on a more static interpretation of LRMC.2 Suppose gas turbines are used for peaking. Then the required LRMC of generating capacity (LRMCOc) may be approximated by the cost of advancing or by the cost saving from delaying 1 kilowatt of gas turbine. This may be estimated by the cost of a kilowatt installed, annuitized over the expected lifetime, and adjusted for the reserve margin (RM) and appropriate percentage loss (LG) typically caused by station use.3 Therefore,
LRMCOC = (annuitized cost per kilowatt)

+ (1 + RM/100)/(1 - LI,100). If expected outage costs are significant in a rating period outside the peak period, then it is possible to allocate marginal capacity costs over several pricing periods, as explained in Chapter 2. In this simple model, however, all capacity costs are charged to peak period consumers. Therefore, if the capacity costs of base-load generating units are to be included in the calculations, it is important to net out potential fuel savings resulting from displacement of less efficient plant by these new base-load units. The theoretical rationale underlying this conclusion is clarified in Appendix C, using a simplified system model. From a more practical point of view, it is clear that if peak period consumers are incorrectly charged the high capacity costs of expensive base-load units, such as nuclear plants, this may encourage them to install their own gas turbine plant. Even intuitively, such a pricing policy would not appear to be sensible.
2. For a more detailed discussion of dynamic versus static LRMC and different practical techniques of estimating the marginal costs of generation, see NARUC,Electricity Rate Design Study, Report nos. 66 and 67 (Palo Alto, Calif.: EPRI,June-July 1980). 3. The annuitized value of a lump-sum investment is defined as the fixed annual payment over the lifetime of the investment, whose present value at the given discount rate is exactly equal to the original expenditure. Annuitizing the investment is conceptually equivalent to advancing the capital outlay by one year. For example, if the investment cost of gas turbines is $400 per kilowatt installed, the corresponding annuitized value is $400 x 0.1175 = $47 per kilowatt per year, at a discount rate of 10 percent, over twenty years. The annuity factor of 0.1175 may be read off a set of standard tables. See Richard S. Burington, Handbook of MathematicalTables and Formulas, 5th ed. (New York: McGraw-Hill, 1973), p. 445.

56

THEORY

Next, the LRMC of transmission and distribution is calculated. Generally, all costs of investment in transmission and distribution (T&D)-except customer costs discussed later-are allocated to incremental capacity, because the designs of these facilities are determined principally by the peak kilowatts that they carry rather than the kilowatt-hours. Particularly at the distribution level, however, the size of a given feeder may depend on the local demand peak, which may not occur within or coincide with the system peak period. This could complicate the problem of allocating distribution capacity costs among the various pricing periods.4 The concept of structuring by voltage level may be introduced at this stage. Consider several supply voltage categories: extra high, high, medium, and low (EHV, HV, Mv, and LV, respectively). Since consumers at each voltage level are charged only upstream costs, capacity costs at each voltage level must be identified. The simplest approach is to use the average incremental cost (AIC) method to estimate the LRMC of T&D. Suppose that in year i, AMW, and I, are the increase in demand served (relative to the previous year) and the investment cost, respectively. Then, the AIC of capacity is given by AIC
=

, I,/(I + r)i/[EiLT

AMWIll1 + r)']

where r is the discount rate (for example, the opportunity cost of capital), T is the planning horizon (for example, ten years), and L is the average time delay between the investment and commissioning dates for new facilities. In the AIC method the actual additional increments of demand are considered as they occur, rather than the hypothetical fixed demand increment AD used (more rigorously) in calculating generation LRMC. Because there is no problem of plant mix with investments in T&D, however, AIC and the hypothetical increment method will yield similar results. AIC is also usually much easier to calculate using readily available planning data.5 An alternative method of determining marginal T&D costs at several different voltage levels would be to use historical data to fit regression equations such as: (transmission costs) = a + b (peak demand). There is no guarantee, however, that such past relations would hold true in the future as the system expands. Assume that the AIC of EHV and HV transmission has been computed and annuitized over the lifetime of the plant-for example, thirty years-to yield the marginal costs ALRMCHV. Then, the total LRMC of capacity during
4. See, for example, M. Boiteux and P. Stasi, "The Determination of Costs of Expansion of an Interconnected System of Production and Distribution of Electricity," in Marginal Cost Pricing in Practice, ed. James Nelson (Englewood Cliffs, N.J.: Prentice-Hall, 1964). 5. AIC is a good approximation to LRMC in most cases. For numerical examples, see Robert J. Saunders, Jeremy J. Warford, and Patrick C. Mann, Alternative Concepts of Marginal Cost Pricing for Public Utility Pricing: Problems of Application in the Water Supply Sector, World Bank Staff Working Paper no. 259 (Washington, D.C.: World Bank, May 1977).

STRICT LONG-RUN

MARGINAL

COST

57

the peak period at the HV level would be

LRMC.VC = LRMC3.I/(l - LHv/100) + ALRMCHV

where LHV is the percentage of incoming peak power that is lost in the EHV and HV network. (See also the power and energy flow balance shown in Figure 3-2.) This procedure may be repeated at the MV and LV levels. Thus the LRMC of capacity to MV consunmersis given by
LRMCMVC
= LRMCHVC

/(1

LMv/100) + ALRMCMV

where ALRMCMV is the element of incremental MV capacity costs, for example, the AIC of distribution substations and primary feeders; and LMVis the percentage of incoming peak power that is lost at the MV level. The LRMC of T&D calculated in this way is based on actual growth of future demand and is averaged over many consumers. Some exceptions should be noted, however. First, certain transmission lines may be specifically associated with particular generating sites and could be considered a generation cost rather than a transmission cost. Second, some transmission may be associated with specific areas or loads and, therefore, the costs of such facilities should be allocated accordingly. For example, suppose that a particular system has two geographically distinct load areas: a central region consisting of a group of cities and a remote area that is predominantly rural. Then the concept of a generation power pool can be used to calculate a common generating cost for serving both areas, but different transmission costs may be determined for each area by allocating the costs of transmission links appropriately. Third, it may be possible to identify facilities that are specific to certain consumers. These could be allocated to consumer costs, for example, a special subtransmission link and substation for a large industry. This last point is also important in dividing the distribution costs between LV and consumer costs. A given customer, for example, may have a long service drop line that should be specifically allocated to this user, rather than being included in the LV calculation.

Marginal Energy Costs and Losses

The system lambda concept is useful in calculating marginal energy costs. Thus, it may be deduced from Figure 4-1 that the LRMC of energy during the peak period will be the running costs of the machines to be used last in the merit order to meet the incremental peak kilowatt-hours corresponding to the demand increment AD. In the simple model presented, this would be the fuel and operating costs of gas turbines. These costs have to be adjusted by the appropriate peak loss factors at each voltage level in the same way as applied to marginal capacity costs.

58

THEORY

The LRMC of off-peak energy corresponding to a load increment during the off-peak period would usually be the running costs of the least efficient base-load or cycling plant used during this period. Exceptions to this generalization would occur when the marginal plant used during a pricing period was not necessarily the least efficient machine that could have been used. For example, less efficient plants that have long start-up times and are kept running because they are required in the next pricing period may be operated earlier in the loading order than more efficient plants. This would correspond to minimization of operating costs over several pricing periods rather than on an hourly basis. Again since the heat rate of the plants could vary with output level, the simple linear relation usually assumed between generation costs and kilowatt-hours may need to be replaced by a more realistic nonlinear model. The loss factors for adjusting off-peak costs will be smaller than the loss factors for the peak period. For example, resistive losses are a function of the square of the current flows and are greatest during the peak period. The treatment of losses generally raises several important issues. Total normal technical losses, including station use, vary from system to system. If these are significantly greater than about 15 percent of gross generation, then reduction of the losses should have a high priority. When engineering losses in excess of acceptable levels are routinely passed on to the customer, this may act as a disincentive to improve technical or administrative efficiency. Losses caused by theft and unpaid bills are also often charged to paying customers. Again, the issue is whether these nontechnical losses could be reduced by appropriate measures, or if incremental consumption always has an unavoidable component of such losses associated with it. Theft in U.S. systems has been estimated to average about 2 percent of gross generation, but norms in developing countries may have to be set somewhat 6 higher. The LRMC analysis at the generation, transmission, and distribution levels helps to establish whether these incremental costs are excessive because of overinvestment, high losses, or both.

Customer

Costs

Allocating a part of the investment and operating costs of the distribution system to customer costs has proved difficult. Thus, unsuccessful attempts have been made to identify a pure component of customer costs: on the basis of a hypothetical skeleton distribution network required to serve a minimum load or by using regression analysis of past data to fit equations such as:
6. See Alan Donziger, "The Underground Economy and the Theft of Utility Services." Public Utilities Fortnightly (November 22, 1979), pp. 23-27.

STRICT LONG-RUN

MARGINAL

COST

59

Distribution costs = a + b (peak demand) + c (number of customers). The latter is particularly difficult because the number of consumers and peak demand are usually highly correlated. It is simpler to consider the general costs of the distribution systenn as capacity costs, whereas fixed and recurrent customer costs are defined as those that can be readily allocated to users. Initial customer costs consist of nonrecurrent expenses attributable to items such as service drop lines, meters, and labor for installation. They may be charged to the customer as a lump sum or may involve payments over several years. Recurrent customer costs stem from meter reading, billing, administrative, and other expenses. These could be imposed as a flat charge on a repeated basis, in addition to the usual kilowatt and kilowatt-hour charges. In general, systems vary in allocating incremental (nonfuel) operating, maintenance, and administrative costs among the three basic cost categories: capacity, energy, and customer. This requires an analysis of specific systems, but such incremental costs are usually small and do not greatly affect the results.

Hydroelectric

Systems

The analysis of the LRMC of generation has been discussed so far chiefly in terms of an all-thermal power system. There are several special issues that arise in computing LRAMCfor all-hydro and mixed hydro-thermal systems, where seasonal variations in LRMC are particularly important.' Again, the most accurate method of calculating LRMC would be to use a computerized model for long-range system planning and to simulate the effect of different load increments on system capacity and energy costs, as described at the beginning of this chapter. In the absence of such a sophisticated planning model, simpler methods of analysis described below must be used to derive the LRMC. All-hydro systems To present the analysis of an all-hydroelectric system clearly, it is necessary to make the following convenient assumptions: a. The operation of an entire multi-reservoir, multi-plant system is modeled in terms of an "equivalent" single composite reservoir. b. The system is analyzed essentially for one year, for example, with long-run capacity costs represented by annuitized investment costs. c. Deterministic or expected values of water inflows and outflows are used, and the effects of uncertainty are ignored. The "equivalent" composite reservoir concept is a representation often used
7. For a more mathematical treatment of the effects of seasonal water flows on the structure of LRMC, see Turvey and Anderson, Electricity Economics, chap. 15.

60

THEORY

8 Also, although a onein the engineering analysis of hydroelectric systems. period or annual analysis is no substitute for a complete long-range system model, it illustrates the essential features to be analyzed in the LRMC estimations. Finally, the complications raised by uncertainty may be introduced later, as described in Chapter 2. Thus, these simplifying assumptions do not alter the validity of the basic conclusions to be drawn from the model. Figure 4-3 depicts the typical annual flow of water into and out of the lumped equivalent reservoir. From month 0 to 5, the reservoir is being discharged or drawn down, because this is the dry season, and the water required for generation exceeds the inflow. Therefore, from point A to B the reservoir is being constantly depleted, until at point B the reservoir will be at its lowest level. For a well-planned and well-operated system in a normal year, the reservoir at this time of year will be almost empty, containing only a small safety reserve margin of water. Months 5 to 10 represent the wet

Figure 4-3. Annual Variation of New Water Flows into the Reservoir in an All-Hydroelectric System
C

Filling

Spilling

or sluicing

Discharging

Discharging B I D I

. O

~~A'

B'

j1

I 1

5 6

8 9 10 Months

12

15

8. See, for example, N. V. Arvanitidis and J. Rosing. "Composite Representation of a Multireservoir Hydroelectric Power System," and "Optimal Operation of Multireservoir System Using Composite Representation," both papers in IEEE Transactions on Power Apparatus

and Systems, vol. PAS-90 (February 1970). For more sophisticated modeling of hydroelectric power system operation, see S. Soares, C. Lyra, and H. Tavares, "Optimal Generation Scheduling of Hydrothermal Power Systems," IEEE Transactions on Power Apparatus and Systems,

vol. PAS-99 (May-June 1980). pp. 1107-18.

STRICT

LONG-RUN

MARGINAL

COST

61

season during which water inflow from the catchment area exceeds the desired outflow. The reservoir fills up, consequently, between points B and C. In month 8 the reservoir is full, and excess water must be spilled, since the net inflow is still positive even after the daily generation requirements have been fully satisfied. At point D spilling ceases, and the onset of the dry season again requires a steady draw-down of the reservoir between months 10 and 12. After point A, the annual cycle described above will be repeated. If significant spilling occurs over several months and this pattern is likely to continue for many years, then long-run marginal energy costs during this period will be essentially zero. Assuming that there is sufficient plant capacity to meet the peak demand, additional energy may be generated by running through the turbines the water that would otherwise have been spilled. The only marginal costs involved might be a small increase in operating and maintenance costs. In contrast, reservoir capacity is required when the net water inflow is negative, and the corresponding costs represent marginal energy costs during the dry season. Long-run marginal costs of power are incurred, as in the case of the all-thermal system discussed earlier, when additional generating capacity is needed to meet a sustained increment in peak period demand. In summary, therefore, in an all-hydro system the LRMC of generating capacity incurred during the peak period would be based on the cost of increasing peaking capability, that is, additional turbines, penstocks, expansion of the power-house, and so on. Incremental energy costs chiefly attributable to the dry season would be the costs of expanding reservoir storage. In certain instances it may be justifiable to allocate a fraction of the dam costs to the capacity costs. This depends, however, on the specific case involved, for example, whether or not more storage is needed to firm up the additional capacity. When significant spilling of water occurs-for example, during the wet season--incremental energy costs would be small, usually involving operation and maintenance costs only. Correspondingly, at times when demand does not press on capacity, incremental generating capacity costs may be ignored. If the system is likely to be energy constrained and all incremental capacity is needed primarily to generate more energy because the energy shortage precedes the capacity constraint for many future years, however, then the distinction between peak and off-peak costs and between capacity and energy costs tends to blur. In the extreme case of a system constrained for energy, it may be sufficient only to levy a simple kilowatt-hour charge at all times, because hydro energy consumed during any period, except when spilling, usually leads to an equivalent draw-down of the reservoir. This charge could be determined by applying the AIC method to total incremental system costs. These results are summarized in Table 4-1.

62

THEORY

Table 4-1. Marginal Generating Costs in an All-Hydro System

Marginal costs Peak period Type Capacityconstrained Energyconstrained Wet season Capacity Dry season Off-peak period Wet season Dry season

Capacity None Energy and energy Total incremental costs requiredto supply additionalkilowatthours

Mixed hydro-thermal systems

The estimation of marginal costs in mixed hydro-thermal generating systems depends critically on the mix of generating plants used at different times. Several basic configurations illustrate the principles governing the calculation of LRMC. First, consider a predominantly thermal system in which all available water is used and thermal generation is required throughout the year. The typical daily pattern of generation for such a system is shown in Figure 4-4a. Base-load steam plants represented by the shaded area I would be run continuously during the twenty-four-hour period; gas turbines depicted by the shaded area III would be required to meet the peak period demand. All the water collected during the year would be used to supplement the thermal generation on a daily basis by filling up the unshaded area II of the LDC. The LRMC analysis for this system would closely follow the procedure described earlier in the all-thermal system. A general principle for determining the value of water used in a mixed hydrothermal system, for example, in area II of Figure 4-4a, is analyzed geometrically below. Consider the pattern of plant operation discussed earlier, which is analyzed in Figure 4-5, on the basis of minimizing cost. As described in Chapter 3, the gas turbines with relatively low capital and high operating costs, represented by the steepest solid line, are used for peaking duty. The steam turbines having higher capital and lower operating costs, as shown by the flatter solid line, are operated as base-load plants. Since the water is used between these two types of plants, an imputed value for water would be given by the broken line. The position of this line would usually be determined by conditions of operation that would seek to ensure that all available water in storage would be used up during the year to supply the energy represented by the shaded area of the annual LDC. This area also corresponds to the unshaded portion II of Figure 4-4a. The intermediate slope

STRICT

LONG-RUN

MARGINAL

COST

63

of the broken line indicaLtesthat the cost for each kilowatt-hour of watergenerated energy lies between the energy costs of peaking plants and baseload thermal plants. In brief, one method of determining the value of water in a mixed hydrothermal system is based on the equivalent costs of the thermal generation it displaces. If pumped storage is involved, the value of the water is the cost of pumping, net of appropriate losses.

Figure 4-4. Typical Daily Load Duration Curves for a Mixed Hydro-Thermal System
PREDOMINANTLY THERMAL SYSTEM

Hoursof the day

PREDOMINANTLY HIYDRO SYSTEM

12

16

20

24

16

Hours of the day

64

THEORY

Figure 4-5. Imputed Value of Water in a Mixed Hydro-Thermal System

Gas turbines (peaking) E \/lmp~~~~~~~~uted curvefor hydro

Steam turbines (base-load)

Hours per year of plant operation

7,l

Annualload durationcurve

0 Hours per year

8,760

The second configuration is one in which hydro makes a bigger contribution than in the previous case. Here, steam turbines or diesels are operated continuously to supplement limited hydro resources during the dry season. This mode of operation is shown in Figure 4-4b, where the shaded area I represents thermal plants, and the unshaded area II is hydro generation. To minimize the present value of system costs in the long run, the smallest pos-

STRICT LONG-RUN MARGINAL COST

65

sible amount of thermal capacity must be installed and base-loaded or run twenty-four hours a day in the dry season, with stored water meeting peak demands. In the rainy season, the situation is reversed. To avoid spilling or sluicing water the maximum amount of installed hydro capacity is used continuously as shown by area I. Thermal generation provides the residual energy required to supply area II. This type of plant may also be used for following the load during the peak period. In this situation, incremental energy must always be supplied by operating the thermal plants, that is, marginal fuel costs must be incurred whenever additional kilowatt-hours are needed, including the wet and dry seasons as well as peak and off-peak periods. Even if more stored water can be run off to supply additional energy during a given period, this implies that there must be greater thermal generation at another time to compensate for the water loss. In addition, since thermal capacity is fully utilized, the marginal costs of the additional capacity divided by the number of hours of operation at full capacity must be added to the marginal fuel costs to obtain marginal energy costs. An increment of demand shown by the broken line enclosing area IV could be met by increasing installed hydro capacity. But the additional energy in area IV must be supplied by augmenting thermal capacity, and this cost has already been charged to marginal energy costs. Therefore, the marginal capacity cost during the peak period would be the incremental cost for a kilowatt of hydro capacity minus the cost of the appropriate kilowatt fraction of thermal capacity used to generate energy. Other hydrothermal generation configurations are possible, but usually the LRMC for these cases can also be analyzed in the same way, using fairly simple and straightforward considerations. Complications may arise, particularly if the pattern of plant operation is likely to change rapidly in the future. This might occur, for example, if the marginal plant used in the peak period shifts from gas turbines to peaking hydro, or if the reverse situation occurs. Then the value of LRMC would have to be calculated as a weighted average. The weights would depend on the share of future generation by the different types of plants used and on the timing of the transitions.

Chapter 5

Adjusting the StrictLong-Run Marginal Cost

ONCE THE STRICT LONG-RUN MARGINAL COST (LRMC) has been calculated,

the first stage of setting tariffs is complete. In the second stage, the actual tariff structure that meets economic second-best, social, financial, political, and other constraints must be derived by modifying the strict LRMC. This process of adjusting LRMC will, in general, result in deviations in both the magnitude and structure of the strict LRMC. Changes in tariff structure at this stage will be based mainly on sociopolitical factors. These would include differentiation by type of consumer, such as residential, commercial, industrial, and so on, or by income levels for residential consumers. Practical considerations such as the difficulties of metering and billing also affect the final tariff structure. The constraints that produce deviations in the final tariffs relative to the strict LRMC fall into two categories.' The first consists of distortions that may be analyzed basically within an economic framework, such as secondbest consideration and subsidized (or lifeline) tariffs for low-income consumers. In these cases, it is possible to quantify the extent of the deviation from the strict LRMC by using an appropriate pricing model and explicit system of shadow prices. The second group includes all other considerations, such as financial viability, sociopolitical constraints, and problems of metering and billing, where strict economic analysis is difficult to apply. These two groups of constraints may be related. For example, subsidized tariffs can simultaneously have~ economic welfare, financial, and sociopolitical implications. Second-Best Considerations

Where prices elsewhere in the economy-especially electric power substitutes and complements-do not reflect marginal costs, a second-best depar1. The strict (shadow-priced) LRMC also deviates from the LRMC calculated on the basis of financial costs, because shadow prices are used instead of the market prices of inputs to the electric power sector. This preliminary adjustment corrects distortions in the economy. Therefore, the constraints that force further departures from the strict LRMC in the second stage of setting tariffs may also be considered consequently as distortions that impose their own shadow

66

ADJUSTING

THE STRICT

LRMC

67

ture from a strict marginal cost pricing policy for electricity services may be required.2 For example, the subsidies for imported generators or diesel fuel, which exist in many developing countries, may make it advantageous for users to establish their own plant, even though to the economy as a whole this is not the least expensive way to meet the demand. The first-best solution in this case might be for the government to revoke such subsidies, or to restrict imports of private generating plants and to price electricity at the strict LRMC. If transport policy dictates the need to maintain subsidies for diesel fuel, however, or if strong pressure groups make such changes politically infeasible, the low cost of (subsidized) private generation may require that a second-best price lfor grid-supplied electricity be set below the strict LRMC. The extent of the deviation from the strict LRMC is determined by the size of the subsidy and the degree of substitutability of the alternative energy source, as explainedL later. A related question concerns the availability of subsidized kerosene that meets the basic energy requirements for low-income consumers in many developing countries at prices they can afford. In part, the subsidy may also prevent low-income households, especially in rural areas, from shifting to noncommercial fuels. Such a shift can cause environmental damage. For example, the over-use of wood leads to deforestation and other associated environmental consequences. At the same time, animal dung has a high opportunity cost in that it can be used as a fertilizer. Several unfortunate side effects may follow, however, including the practice of mixing kerosene with gasoline, since the latter is usually sold at a much higher price. Such a policy could also have negative effects on income distribution. For example, in some countries the relatively wealthy have benefited from the subsidy by converting motor car engines to kerosene use. It is also common in several countries to mix kerosene with more expensive diesel fuel for industrial use. Finally, however, when there is a kerosene subsidy, an argument exists for reducing the price of electricity proportionately.

Subsidized or Lifeline Prices

In addition to the second-best economic arguments-for example, those associated with subsidized kerosene-sociopolitical or equity arguments are
values on the calculation. For details, see Mohan Munasinghe and Jeremy J. Warford, "Shadow Pricing and Power Tariff Policy," in CanadianElectrical Association,Marginal
Costing and Pricing of Electrical Energy (Montreal, May 1978).

2. Pricedistortions affectinginputsintothe production of electricpowerandoutputsof other sectorsthat are electricityintensive,such as aluminum,shouldalso be considered.Production inputsmay be dealt withby directshadowpricingof inputs.The outputsof electricity-intensive sectorsrequiremore specificanalysis,althoughsuch casesare rare.

68

THEORY

often advanced in favor of lifeline tariffs for electric power to supply the basic electricity needs of poor consumers. These arguments are heard most forcefully where the cost of electricity consumption is high compared with the relevant income levels. In theory, lump sum cash transfers to lowincome households to meet their overall basic needs would be preferable to subsidies on specific items of consumption such as electricity (Chapter 3). The ineffectiveness of administration and policy tools for directly redistributing income and the persistence of large disparities of income, however, often force authorities to use subsidized prices as a redistributive measure. Since electric power utilities can act as discriminating monopolists, they have an advantage in addressing these issues. Clearly the appropriateness of the lifeline pricing policy and the size of the tariff blocks involved requires detailed analysis. This would involve examining a chain of related energy and other effects, which are generally more complex in developing countries than in developed countries. The concept of a subsidized social block, or lifeline rate, for low-income consumers has another important economic rationale, based on the argument of income redistribution. Figure 5-1 shows the respective demand curves AB and GH of low-income (1,) and average-income (12) domestic users, the social tariff P, over the minimum consumption block 0 to Qi,, and the price level based on marginal cost p. All tariff levels are in domestic market prices. If the actual tariff p = p,, then the average household will be conFigure 5-1. Economic Basis for the Social or Lifeline Rate
G

'2

>

1I

'.r.

P_______________.

PI~~~~~

Q 1

Q..
Quantity

Q2

ADJUSTING

THE STRICr

LRMC

69

suming at the optimal level Q , but the poor household will not be able to 2 afford the service. If increased benefits accruing to the poor have a high social value or weight, the consumer surplus segment ABF must be multiplied by an appropriate social weight Wj greater than unity. Then, although in nominal market prices the point A lies below pe, the socially weighted distance OA could be greater than the marginal cost of supply. The adoption of the increasing block tariff shown in Figure 5-1, consisting of the lifeline rate p, followed by the full tariff p,, helps to capture the consumer surplus of the poor user. But it does not affect the optimal consumption pattern of the average consumer, if the income effect caused by reduced expenditure of the average consumer for the first block of consumption is ignored, that is, up to Qm.i In practice, the size of Q,m, must be carefully determined to avoid subsidizing relatively well-off consumers. It should be based on acceptable criteria for identifying low-income groups and reasonable estimates of their minimum levels of electricity consumption. These might be established as the amount that is sufficient to supply basic requirements for lighting and minor appliances. In most developing countries Q,in would be about 50 kilowatt-hours and certainly would be less than 100 kilowatt-hours a month. In contrast, minimum consumption levels may typically amount to several hundred kilowatt-hours a month in industrialized countries such as the United States, where the average household used about 700 kilowatt-hours monthly in 1979. The level of p5in relation to the theoretical LRMC might be determined on the basis of the poor consumer's income relative to some critical consumption level and also his overall ability to pay, as discussed later. The utility's revenue constraints would also be considered in determining p, and Qmi.This lifeline tariff policy may be reiniorced by an appropriate connections program, such as subsidized house connections.

Optimal Electricity Pricing in a Distorted Economy

A general expression for the socially optimal electricity price can be developed that is based on border prices and that compensates for distortions in the economy. From the general equation, results for optimal electricity pricing can be derived for cases that reflect: a. a perfectly competitive economy (classical results) b. economic second-best considerations c. subsidized social prices or lifeline rates for poor consumers. The supply and demanid for electricity is shown in Figure 5-2, where S is the supply curve represented by the marginal cost of supply at domestic market prices of inputs, and D is the corresponding market demand curve for a specific consumer. Starting with the initial combination of price and consumption (p, Q), consider the effect of a small price reduction (dp) and the

70

THEORY

Figure 5-2. Supply and Demand for Electricity

dp-,B

_L

a
--

0~~~~~~~~~~~~~~~~~

F-____ >s
---

>

~~~~~~~~(MC)

-MC, ___

I Q

Q + dQ

Q,

Quantity

resulting increase in demand (dQ) on the net social benefits of electricity consumption. Before evaluating the net social benefit of this price change, several parameters will be defined. First, suppose the marginal cost of electricity supply (MC) is calculated without shadow pricing the inputs, that is, considering them, in market prices. Then a, is defined as the electricity conversion factor (ECF), which transforms the market-priced MC into the corresponding real economic resource cost. When all the inputs to the electricity sector are correctly shadow priced, the border-priced MC is MCB, = (a, MC). Second, a specific social weight W, is assigned to each marginal unit of consumption (valued in market prices) of a given individual i in the economy. For example, if this electricity user is poor, the corresponding social weight may be much larger than for a rich customer, reflecting society's emphasis on the increased consumption of low-income groups. Third, if the given individual's consumption of goods and services other than electricity valued in market prices increases by one unit, then the border-priced marginal social cost of economic resources used or the shadow cost to the economy is bi. As a result of the price reduction, the consumer is using dQ units more of electricity, area IFGH, which has a market value of (pdQ).3 The consumer's
3. The little triangle LFG may be neglected throughoutthis analysis because its area is (dp dQ)12,whereboth dp and dQ are small increments.

ADJUSTING THE STRICT LRMC

71

income net of expenditure on electricity has increased by the amount pQ (p - dp) (Q + dQ), hovvever. Assuming none of the income is saved, this individual's consumption of other goods and services will increase by the amount (Qdp - pdQ)-that is, area BEFG minus area IFGH, also valued in market prices. Therefore, the consumer's total consumption of electricity plus other goods will increase by the net amount (Qdp) in market prices. This is the traditional increase in consumer surplus benefits. The shadow value of this increased consumption is W, (Qdp), where W, is the social weight appropriate to this consumer's income/consumption level. Next, consider the resource costs of these changes in consumption. The shadow cost of increasing electricity supply is (a, MCdQ), that is, a, times area IJKH. The resources used to provide the other additional goods consumed are b, (Q dp - p dQ), where a, is the conversion factor for electricity production, and b, is the conversion factor for other goods consumed by this individual. Finally, the income change of the electricity producer must also be considered, but this effect may be ignored if, as is likely, the producer is the government. The total increase in net social benefits caused by the decrease in the electricity price is given by dNB = W, (p dQ) + ap (MCdQ) - b, (Qdp - pdQ). Therefore,
(dNBldp) = Q [(W, - b,) + n b, - nap (MCIp)]

where n = (p dQIQ dp) is the elasticity of demand (magnitude). The necessary first-order condition for maximizing net social benefits in the limit is d(NB)Idp = 0. This yields the optimal price level (5.1) p* = a, MCI[b, + (W
-

bc)ln].

This expression may be reduced to a more familiar form by making some simplifying assumptions. Case 1. Classical case
Assume a perfectly competitive economy where market prices and

shadow prices are the same, and income transfer effects are ignored, that is,
there is no social weighting. Therefore, a, = W, = bc = 1, and equation

(5.1) reduces to
(5.2) p* = MC.

This is the classical result (discussed in Chapter 2), where net social benefits are maximized when price is set equal to marginal cost at the market clearing point (p,, Q) in Figure 5-1.

72

THEORY

Case 2. Efficiency pricing case Assume that income transfer effects are ignored, because the marginal social benefit of consumption is equal to the marginal social cost to the economy of providing this consumption. Therefore, W, = b,, and equation (5.1) becomes (5.3) PI* = (a, MC)lb, = MCB lb,.

This is the optimal marginal cost price for electricity, when efficiency (shadow) prices are used that emphasize the efficient allocation of resources and neglect income distribution considerations. As mentioned earlier, the marginal social cost of electricity supply (MCB,) may be evaluated directly by applying the appropriate efficiency shadow prices to the least-cost mix of technically determined inputs used to produce electricity. The conversion factor b, depends on the type of consumer involved, however. For residential electricity consumers, b, represents the consumption conversion factor (CCF), which reflects the resource cost or shadow value of one (market priced) unit of the household's marginal consumption basket. If the CCF < 1, then PI* < (MCB). Another interesting case illustrates the application of equation (5.3) to correct for economic second-best consideration arising from possibilities of energy substitution. As an extreme case, suppose all expenditures diverted from grid-supplied electricity will be used to purchase alternative energy that is subsidized by the government. Examples could be kerosene for lighting or diesel for autogeneration. In this case, b, is the ratio of the border-priced marginal cost of alternative energy to its market price and may be written
b, = MCBae'Pac.

Thus, from equation (5.3),

(5.4)

pe = MCB, (p.,IMCB-).

The logic of this expression may be clarified by considering the case when p, > pe*. Then, the shadow cost of one unit of electricity is MCB,lp,. If this sum were used to purchase alternative energy, the shadow cost would be MCB_. Since pe > pp*,MCB,1p, > MCBa,lP,,, and the country is better off if more electricity is used instead of the alternative energy. Therefore, p, should be reduced to pe. Similar reasoning can be used to show that if p. < pe, then p should be increased to the value pe*. If the alternative energy is priced below its border-priced marginal cost, that is, b, > 1, then from equation (5.4), pe* < MCB, also. Therefore, the subsidization of substitute energy prices will result in an optimal electricity price that is below its shadow supply cost.

ADJUSTING

THE STRICT

LRMC

73

If it is not possible to determine the consumption patterns of specific consumer groups, then b, could be defined broadly as the average conversion factor for all electricity users, as discussed in Chapter 3. Case 3. General case Assume equation (5.1) represents the optimal electricity price calculated using social (shadow) prices, which incorporate concerns about income distribution. In this case MCB, should be computed using the social shadow prices of inputs to the electricity sector. In practice the difference in MCB, calculated in efficiency and social prices, however, is likely to be small and may be ignored. That is, the efficiency-priced value of MCB, may be used. Consider the case of a group of poor consumers for whom W>> b, (n 1). Therefore, equation (5.1) may be written p* n MCB,IW,. An even greater simplification is possible if it is assumed that n = 1; thus,
P* = MCB,IW,.

For illustration, suppose that the income-consumption level of these poor consumers (c) is one-third the critical income-consumption level (c), which is similar to a poverty line. Then a simple expression for the social weight is
W, = c/c = 3.

Therefore, p5 = MCB/3, which is the lifeline rate or subsidized tariff appropriate to this group oi low-income consumers.

Financial Viability
To introduce more rigor into the analysis of tariff policy and financial performance criteria for public enterprises, it is necessary to distinguish between the economic and social objectives of tariff policy, and the minimum level of financial performance necessary for the continued efficient operation of the enterprise concerned. Such a distinction should not be interpreted as downgrading the role of financial analysis; it simply reflects the fact that although a minimum level of financial performance is almost by definition a necessary condition, it is rarely a sufficient condition for achieving a range of social and economic objectives that might be assisted by the pricing and investment policies of the enterprise. As described earlier, preeminent among those objectives that transcend the financial interests of the enterprise itself is the goal of using marginal cost pricing to influence consumer behavior in order to improve resource allocation. Other nonfinancial policy objectives are discussed later. The principal financial objectives most often encountered relate to the revenue requirements of the sector. They are often embodied in criteria such as

74

THEORY

some target financial rate of return on assets or an acceptable rate of contribution toward the future investment program. In principle, for state-owned power utilities, the most efficient solution would be to set price equal to marginal cost and to rely on government subsidies or taxes to meet the utilities' financial needs. In practice, some measure of financial autonomy and self-sufficiency is an important goal for the sector. Because of the premium that is placed on public funds, a marginal cost pricing policy that fails to achieve minimum financial targets for continued operation of the power sector would rarely be acceptable. The converse and more typical case, where marginal cost pricing would result in financial surpluses well in excess of traditional revenue targets, usually attracts the criticism of consumers. Therefore, in either case changes in revenues have to be achieved by adjusting the strict marginal cost. A widely used criterion of financial viability is the utility's potential to earn an acceptable rate of return on assets, for example, the net operating income after taxes given as a fraction of net fixed assets in operation plus, in some cases, adequate working capital. In the case of private utilities-for example, in the United States-the regulatory authorities have traditionally imposed a fair rate of return as an upper limit on earnings (and therefore, on average price per unit sold).4 Where utilities are owned by the government, as in most developing countries, the target rate of return is usually considered a minimum requirement to help resist sociopolitical pressures that tend to keep prices too low. If the asset base is defined in revalued terms, then this requirement is more consistent with the forward-looking approach of LRMC. Another future-oriented financial criterion that is especially useful when the system expands rapidly requires the utility to make a reasonable contribution to its future investment program from its own revenues. This self-financing ratio is often expressed by the amount of internally generated funds available after operating expenses and debt service, as a fraction of capital expenditures. The application of the financial criteria often raises serious conceptual and practical problems. Thus, if a rate of return test is to be used, then the question of asset revaluation arises. The use of historical costs for working assets-typically original cost less depreciation-would tend to understate their value when capital costs are rising rapidly. If assets are to be revalued,
4. See Peter J. Garfield and William F. Lovejoy, Public Utility Economics (Englewood Cliffs, N.J.: Prentice-Hall, 1974); or James Suelflow, Public Utility Accounting: Theory and Applications (East Lansing: Michigan State University, 1973). It has been suggested that such regulation of the rate of return might cause utilities to overcapitalize, thus increasing their rate base and permitting greater earnings, that is, substitute capital for other units like labor, beyond the levels consistent with economically efficient cost-minimizing production. For a discussion of this phenomenon, referred to as the Averch-Johnson (or A-J) effect, see, for example, Michael A. Crew and Paul R. Kleindorfer, Public Utility Economics (New York: St. Martin's Press, 1979), chap. 9.

ADJUSTING

THE STRICT LRMC

75

the costs of either exactly reproducing the power system at today's prices or replacing it with an equivalent system, also at today's prices, might be used after netting out depreciation to allow for the loss of value corresponding to the economic and functional obsolescence of existing equipment. Whichever approach is adapted, significant difficulties of interpretation clearly will oc5 cur in its practical application. Furthermore, in certain countries public utilities or corporations are not legally allowed to revalue their assets. Whichever criterion or combination of criteria is used, it is important that the initial tariffs based on the strict LRMC be included in the utility's financial forecast. Then these first-round tariffs may be adjusted through an iterative process until the chosen parameters of financial viability fall within the acceptable range. Although in practice this process is quite ad hoc, some practical guidelines may be effectively used to reconcile the strict LRMC and the revenue requirement. The relative adjustments to the strict LRMC between principal consumer categories (like residential and industrial), as well as among the different pricing periods (like peak and off-peak) within a given consumer category, will determine the share of the revenue burden to be borne by each user group in a given pricing period. The simplest practical method of adjustment, which also appears to be the most equitable, is to retain the relative structure of LRMC and to vary the average tariff level by equi-proportional changes. In general this procedure will not be economically efficient. The application of the Baumol-Bradford inverse elasticity rule, whereby the greatest (least) divergence from the strict LRMC occurs for the consumer group and pricing period where the price elasticity is lowest (highest), is the most satisfactory adjustment procedure from the viewpoint of economic efficiency.6 In other words, price adjustments should be smallest for consumers whose electricity usage is most sensitive to price changes and largest for those who are least sensitive. It is intuitively clear that this guideline would result in the smallest deviations from the optimal levels of consumption consistent with the strict application of the marginal cost pricing rule. In the case of two goods, the following expression applies:
(1 - LRMC, 1p,)/.(l -- LRMC2 ip2) = (lie + Ile,2)I(1le2 + lie2,).

LRMCj and pi are the strict LRMC and price, respectively, of good i; whereas

e; = (aQi1api)I(Q4,9i) and eii = (Qilapo)I(Q/lp)

(i, j = 1, 2)

are the own and cross-price elasticities, respectively, of demand Q with re5. See, for example, G. Robert Faust and H. Gary Larson, "International Valuation Procedures for Electric Utilities," Public Utilities Fortnightly (August 14, 1980), pp. 19-25. 6. See William J. Baumol and David F. Bradford, "Optimal Departures from Marginal Cost Pricing," American Economic Review (June 1970), pp. 265-83.

76

THEORY

spect to price p. The two goods 1 and 2 may be interpreted as either the electricity consumption of two different consumer groups in the same pricing period or the consumption of the same consumer group in two distinct pricing periods. In practice, a larger number of consumer types and rating periods must be considered, and application of the rule will be limited by lack of data on price elasticities and the need to use subjective estimates. This technique may appear to penalize some customers more than others, thus violating the fairness objective. Adjustments involving lump sum payments or rebates, or changes in customer and connection charges are also consistent with economic efficiency provided consumers' electricity usage is relatively unaffected by these procedures, that is, consumption depends mainly on the variable charges. The magnitude of the adjustments that can be made may be insufficient, however. Another related approach for reducing revenues is to charge the strict LRMC only for marginal consumption and to reduce the price for an initial block of electricity use. These subsidies on customer charges or on the initial consumption block can also be tailored to satisfy the lifeline rate requirement for poor consumers, but such measures tend to complicate the price structure. In practice, an eclectic approach involving a combination of all these methods is most likely to be successful.

Other Objectives and Constraints

Aside from second-best adjustments and the satisfaction of basic needs of low-income consumers by means of an increasing block rate, several other objectives and constraints face the policymaker in determining an appropriate level and structure of electricity tariffs. In practice, several legitimate economic reasons, quite apart from political or social concerns, could justify a departure from a tariff policy based strictly on marginal cost. Thus, the decision to promote productive electricity in a particularly deprived region in which large external benefits from development might be reaped may well provide legitimate economic reasons for departing from the strict LRMC ideal. It is rarely possible to completely disentangle the economic from the noneconomic implications or justification of such policies. Thus, maintaining a viable regional development program or reducing rural to urban migration clearly incorporates notions of income redistribution, general welfare, and considerations of economic growth and efficiency. For example, the decision to electrify a remote rural area may require subsidized tariffs because the beneficiaries are unable to pay a price that fully reflects the relatively high unit costs of rural electrification. Such a decision might be made on grounds that are not completely economic, and the principal objective might, in the simplest terms, be to alleviate local political discontent.

ADJUSTING

THE STRICT LRMC

77

Whether or not the primary rationale for subsidization lies in economic or social considerations, however, the decision to subsidize should in all cases be made with a clear understanding of the economic efficiency-that is, the marginal costs that are entailed. Arguments advanced in favor of subsidized electricity prices often may not survive careful scrutirny. Consider the possible use of low power tariffs to promote industrial development or to attract multinational enterprises. Except in certain specific electricity-intensive industries, such as aluminum smelting or chemicals, however, the viability of existing industrial consumers or the decision to set up a new enterprise will depend much less on electricity prices than on other factors, including the availability of a skilled work force, wage rates, incentives such as tax concessions offered by the government, and the provision of an adequate infrastructure. Direct cash subsidies to specific industries from the government are likely to be much more effective than blanket indirect subsidies through electricity tariffs, which probably could be taken advantage of by a much broader group of industrial consumers. In general, decisions to subsidize electric power should be approached with care in developing countries. Subsidization implies either a corresponding drain on public revenues or, at least, if cross subsidization from other electricity consumers is the source, frustration of certain legitimate demands for electric power. Should the burden of subsidization be placed on public revenues, particular costs are implied in the typical developing country in which a premium should be placed on public revenues because of a low tax base and inadequate fiscal machinery. In other words, at the margin a rupee or a peso of investment in the public sector is of greater value than a rupee or a peso of private consumption. Indeed, one might go further and argue that because of this situation, an electric power system could become an appropriate vehicle for generating public revenue. Under current cost conditions in which the unit costs of energy are rising fairly dramatically, an LRMC pricing policy will typically tend to yield considerable financial surpluses to the utility. In this case, efficient use of resources is fully consistent with the generation of public revenues, and a pricing policy that helps to satisfy both objectives would be advantageous. Moreover, as noted earlier, it is reasonably straightforward to build into the tariff structure an element to redistribute income or at least to satisfy the basic electricity needs of low-income groups. It is particularly difficult to reform pricing policy where low incomes and a tradition of subsidized power prevail. In practice, price changes have to be gradual, in view of the costs that may be imposed on those who have already incurred expenditures on electrical equipment and made other decisions expecting little or no changes in traditional policies of power pricing. The efficiency costs of gradualism arising from the need to avoid abrupt implementation of tariffs based on marginal cost can be seen as an additional

78

THEORY

implicit shadow value placed on the full social benefits that result from this policy. Although the more extensive application of peak load pricing based on LRMC is clearly justified, the transition period may take many years. First, the truths and myths regarding modem tariff structures must be well understood by customers, utility managers, and government regulators. Any means of disseminating information will play a key role in this respect. Second, the application of new tariffs could begin with the larger customers at HV and MV levels because they are less numerous and may be sensitized more easily. Over the years, more and more consumers-including at least the larger LV users-would become subject to the new tariffs. Third, the changes in tariff structure should be initially small, to avoid customer resistance resulting from unfamiliarity or from hardship caused by large increases in the total electricity bill. Later, tariffs could be altered more rapidly to approximate the LRMC better. Finally, alternative rate structures could be offered simultaneously-that is, both the conventional and new tariffs-and customers could gradually be won over to the new tariff. Another frequent constraint arises when the government has a policy of uniform national tariffs for each consumer category throughout the country. This would deviate most from the strict LRMC when there are many isolated grids or when disparities between urban and rural electrification costs are large, especially for distribution. Again, there may be pressing sociopolitical and economic reasons for such a policy, including reducing public discontent, promoting development of remote areas, satisfying basic energy needs, and so on. The benefits of a uniform national tariff policy, however, must be weighed once again against the efficiency costs of deviating from the strict LRMC (see Chapter 7). A macroeconomic argument that is often used appears to be a valid policy constraint, but on closer analysis is not such an important issue. The argument involves the possible inflationary effect of increases in power tariffs implied by charging prices based on LRMC. For domestic consumers, moderate electricity price increases will not significantly affect the consumer price index since the value of electricity they consume usually amounts to a relatively small percentage of total household expenditures. Similarly, for industrial and commercial users, the cost of electricity is only a small fraction of total production costs. In general, unless the increase in electricity prices was manyfold, its effect on total costs would be quite small, except in the case of a few power intensive consumers such as metal or chemical plants. Conversely, if demand for electricity was unduly stimulated by low prices, then the additional expenditures on capacity required to meet this demand may have a much more inflationary effect in the long run; ignoring for the moment the losses in economic efficiency inherent in this situation. For example, a consumer's decision to add 1 kilowatt of demand for cooking or air-conditioning at a private cost of about $100 to $200 will usually imply

ADJUSTING

THE STRICT LRMC

79

investments about ten times as great by the utility to meet this increased demand. In certain cases, however, raising electricity tariffs in conjunction with the prices of other government-supplied essential items may have an adverse psychological effect on the inflationary expectations of the public. Also, producers could use the increase as an excuse to raise the prices of their outputs.

Metering, Billing, and Customer Comprehension

Because of both the practical difficulties and the economics of metering and billing, the tariff structure may have to be simplified. Another crucial factor is that the tariff structure must be comprehensible to the average customer. Otherwise, individuals will not be able to adjust their consumption according to the price signal. Therefore, the number of customer categories, rating periods, consumption blocks, voltage levels, and so on will have to be limited. The degree of sophistication of metering-for example, by time of dayshould be determined by the net benefit of metering, the practical problems of installation and maintenance, ease of billing, and so on. The economic rationale for making the metering decision depends on the net benefit of metering based on a cost-benefit analysis that compares the lower supply costs of reduced consumption with the cost of metering and billing plus the decrease in net consumption benefits. This procedure is illustrated in the following discussion, using a two-period tariff.' Figure 5-3 depicts the analysis underlying the metering decision to implement a two-period, time-of-day tariff. D,k and Do are the demand curves for an average hour during the peak and off-peak periods of the day, respectively, each lasting H hours and (24 - H) hours, respectively. If a simple uniform price pavis levied throughout the day, the daily consumption will be Q, H and Q2 (24 - H) kilowatt-hours, respectively, during the peak and offpeak periods. Instead, if a two-period tariff (ppk, pO,) is charged during the peak and off-peak hours, the corresponding levels of consumption will be Q 3 H and Q, (24 - H). Suppose that the LRMC of supplying a kilowatt-hour during the peak and off-peak periods are MCp, and MC,, respectively, where the former consists of both capacity and energy costs and the latter includes only the energy cost. The net consumption benefit of the uniform tariff is given by
NB, = B, C,

tariff involvingseveralgenera7. For an economicanalysisof the meritsof a multi-period Electriction technologies,see EdwardF. Renshaw,"ExpectedWelfareGainsfrom Peak-Load ity Charges," Energy Econornics (January1980),pp. 37-45.

80

THEORY

Figure 5-3. The Metering Decision to Implement a Two-Period Time-of-Day Tariff

G
C~~~~~~~~~~~~~~~~~

$~

T T

T'

1;L;-----t vpa * -----------Pr,~~~~~~~~~~~~~~~,

pv-Pl d

r~ -~ -~---~----r----------------

M.'

.L
Q3

J
,

Q2

Q4

(Q2+ dQ:) Average kilowatt-hours per hour

(Q,+ dQ,)

NB,

[(area OGIJ) H + (area OFRM) (24 - H)]

-

[Q H MCPk + Q2(24

-H)

MCJP].

Similarly, the net consumption benefit of the two-period tariff is given by

NB,, = B,, - C,,

=

[(area OGHL) H + (area OFNK) (24

H)]

- [Q 3 H MC,, + Q. (24 - H) MCoJI. The change in net benefit, therefore, is ANB = NB,,
-

NBI
-

= [(areaMRNK)(24
+

H) - (area LHIJ) H]

[(Q2 - Q 4 )(24 - H) MCr, + (Q, - Q 3) H MCPk].

A further simplification is possible if it is assumed that the two-period tariff reflects the theoretical LRMC, that is, p,p= MC,,k and pt, = MC,. Then, (Q 4 - Q2)MCI, = area MSNK, and (Q, - Q 3) MC,,, = area LHTJ. Therefore, the following result is obtained:

ANB = (area HTI) H + (area SRN) (24

H).

Assuming no change in this average daily cycle during the first year, the annual increase in net consumption benefits is given by

ADJUSTING

THE STRICT

LRMC

81

AANB, = 365 ANB.

Let the difference between the two streams of metering and billing costs (installation, maintenance, administration, and so forth) associated with the
two-period and uniform tariffs be AC,, AC2 , be estimated as AANB,, ST
* * *, AANB,.
*

*, ACT, over the lifetime T

of the meter. Suppose that the corresponding stream of net benefits can also
If the condition (AANB, -

AC,)I(l + r) '- > 0

is satisfied, implementation of the two-period price structure would be economically justified. Even if the decision to install more complex metering is negative, further analysis is helpful. Thus, it is clear that applying the uniform price pa. instead of the strict LRMC prices Pp' = MCPk and p, = MCo, will result in a loss of net efficiency benefits:
NB' = (area HTI) H + (area SRN) (24
-

H).

The level of pa. may be adjusted, however, to minimize this loss of net benefits. Suppose that a reduction in price from pa, to (pa, - dp,,) results in increasedconsumption given by (Q, + dQ,) and (Q 2 + dQ2) in periods I and 2, respectively. The new net benefits may be written NB2 = (area HT'I') H + (area S'R'N) (24 - H). The change in net efficiency benefits is d(ANB) = (NB 2
= -NB')
--

(MCpk

pav)dQ. H

(pa, -

MC,o) dQ 2 (24

H).

Using the familiar first-order condition d(ANB)/dp,, = 0 yields the optimal uniform price that will minimize the loss of net benefits:
*

(H dQ,ldpaj) MC,, + [(24 - H) dQ21dP.J MC,, (11 dQ,idp_v)+ [(24 - H) dQJdp_,J

This result may be generalized. Thus, suppose the tariff analyst wishes to consolidate n pricing periods of the day, where Hi is the duration, Q, is the average hourly consumption, and MC; is the LRMC during the ith period. The optimal uniforn price would be

PI = [Ei-= (Hi dQ,Idp,,) MCI/[E,n

(Hi dQildp,J].

Derivatives of the form H, (dQildp,,) refer to a change in consumption during a given pricing period i caused by a uniform change in price across all the pricing periods that need to be consolidated under the single price regime. Generally, in practically all developing countries information on demand would be inadequate to determine the magnitude of such derivatives

82

THEORY

accurately. In this case, a practical approximation might be p*v = [E,-, (Hi Q,) MCi]1[Ej=i (Hi Q2i)].

Here the uniform price is set equal to the weighted average LRMC using the 8 consumption levels in the different pricing periods as the weights. For poor consumers receiving a subsidized rate, a simple device that limits current may suffice, because the cost of even simple kilowatt-hour metering may exceed its net benefit. In general, various forms of peak load pricing (that is, maximum demand or time-of-day metering) would be more 9 applicable to large mv and HV industrial and commercial consumers. Most LV consumption, especially for households, is metered only on a kilowatthour basis, with the kilowatt-hour price based on a combined energy and rolled in capacity charge-for example, using appropriate coincidence and load factors. More sophisticated meters, such as time-of-day meters that incorporate synchronous clocks, may be affected by power outages. At the other end of the scale, current-limiting devices are easier to tamper with. Some commonly used types of electricity pricing structures are discussed in the next chapter. Recently, the concept of homeostatic utility control has been proposed in the United States, using advanced solid-state technology (including use of microprocessors) to implement sophisticated metering, automatic meter 10 In contrast, reading, load management techniques, and pricing structures. some developing countries may lack technically skilled labor to install and maintain sophisticated meters, or may lack even reliable meter readers. Therefore, choice of appropriate metering is usually specific to each country and is likely to involve many practical considerations.
8. For a more detailed discussion, see for example: Ralph Turvey and Dennis Anderson, Electricity Economics (Baltimore, Md.: Johns Hopkins University Press, 1977), chapter 16. 9. For practical details of metering, see NARUC, Electricity Rate Design Study (Palo Alto, Calif.: EPRI, November 1977), chapter 5; and CIRED, International Conference on Electricity Distribution, Conference Publication no. 151 (London: Institution of Electrical Engineers, 1977). 10. MIT Energy Laboratory, New Utility Management and Control Systems (Cambridge, Mass., 1979).

Chapter 6

RecentExperience with Modern Pricing Structures

SEVERAL TYPESOF TARIFF:STRUCTURES are commonly used to implement the

long-run marginal cost (LRMC) approach, and there is some recent empirical evidence regarding the effectiveness of such modem pricing structures.

Types of Tariff Structures

Many different types of electricity tariffs have been used in various countries over the past fifty years. Price structures have become increasingly complex as both the tec]hniques for analyzing the structure of supply costs and the metering hardware available to apply these tariffs have become progressively more sophisticated. Since the quantity, quality, and price of electricity supplied to each consumer can be, if necessary, individually controlled or at least monitored, a high degree of discrimination and structuring is possible with electricity prices. In theory, a separate tariff could be devised for each customer. In practice, however, as discussed in the previous chapter, the complexity of the tariff would be limited by the metering capabilities, the problems of billing, and the ability of electricity users to comprehend and react to the price signals provided by the power utility. The structuring of LRMC with respect to voltage level, geographic area, and customer type has been discussed earlier. This section focuses on devising and implementing tariffs that vary in relation to the following principal aspects: (a) energy or kilowatt-hour consumption; (b) power demand based on kilowatt or kilovolt-amp consumption; and (c) fixed charges, including both nonrecurrent and recurrent charges. Aspects (a) and (b) can be structured by time of use and usage level. Interruptible rates, the use of tariff adjustment clauses to correct for power factor, fuel surcharges, and so on must also be considered. These basic building blocks may be combined in many various ways to yield literally hundreds of tariff structures differing in their finer details. The most common form of tariff is the energy charge based on the customer's kilowatt-hour consumption over a given period of time, typically 83

84

THEORY

one month. Kilowatt-hour meters that record consumption continuously over shorter periods-for example, at fifteen-minute intervals or during two different periods of the day-may be used to implement electricity prices that vary by time of use (TOU). During the peak period, typically, the capacity charge is converted into an equivalent kilowatt-hour charge and then added to the energy charge. Unit charges may also be varied according to the number of kilowatthours consumed, yielding the two basic types of block tariff structures shown in Figure 6-1. Block structures may also be used with kilowatt or capacity charges, but this is not a common practice. In the increasing block tariff, the kilowatt price increases as consumption rises. This type of pricing structure is also called an inverted tariff, compared with the more traditional decreasing block tariff. Incorporation of the increasing block structure in applying the LRMC-based methodology has already been discussed, particularly in Chapter 5. A two-block tariff consisting of a subsidized-social block followed by normally priced consumption would be adequate in most cases. A larger number of consumption blocks would tend to complicate billing and may confuse the customer. Decreasing block tariffs have been important historically and are still prevalent in many countries. The decreasing block tariff, in which the initial slab of consumption has the highest price followed by successively cheaper blocks (see Figure 6-1), has been widely used, especially for households and small consumers with only kilowatt-hour metering, where more complex metering would not be economically justified (see Chapter 5). The rationale for this policy included arguments that: (a) the utility could recover some of the fixed customer costs through the high-priced initial block even though kilowatt-hour consumption was low; (b) the first block corresponded to the high cost of supplying the customers' peak period load, whereas additional consumption was mainly caused by off-peak appliance use that could be supplied at relatively low

Figure 6-1. Structure of Increasing and Decreasing Block Tariffs

Increasing block tariff E {

Decreasing

block tariff

Units consumed per month

EXPERIENCE

WITH PRICING

STRUCTURES

85

cost; (c) the utility shoulcd encourage increased consumption to realize economies of scale in production; (d) price discrimination could be used to extract the maximum revenue from smaller users who had low price elasticities of demand while also encouraging consumption of larger users who were more sensitive to high prices; and (e) if temporary excess capacity existedfor example, when a new hydro site was developed-the new energy could be supplied without extra cost and, therefore, higher consumption should be encouraged to collect the maximum potential revenues. All of these arguments ignore the fact that if any slab of the decreasing block tariff is significantly below LRMC, it signals the consumer that electricity is much cheaper thtanit really is, thus encouraging wasteful consumption. First, if customer costs must be recovered, then single or recurring fixed charges should be used. Second, unless there is clear evidence that customers with greater consumption have higher user load factors and consume relatively more off-peak energy at the margin, any additional kilowatthours consumed by all users will be equally costly to supply. Therefore, there would be little basis for price discrimination according to consumption level. Third, even if economies of scale exist at the aggregate level of the utility, they do not apply in the case of the variable costs to individual customers. In fact, few utilities currently exhibit any economies of scale, and real unit costs of supply in the long run are rising.' Fourth, it cannot be generally assumed that the consumption of larger users would be more sensitive to price. Fifth, using up any short-run excess capacity is not costless in the long run, because if demand growth is unduly stimulated, future investments must be advanced. Finally, the decreasing block rate is highly regressive and unfair, because it penalizes poorer consumers who generally use less electricity but must pay higher prices on average for each unit purchased (see the discussion of lifeline tariffs in Chapter 5). As explained earlier, the purpose of structuring tariffs by TOU, voltage level, geographic area, and so on, is to convey the LRMC of supply to consumers as accurately as possible. Although peak load or TOU tariffs may also be determined on the basis of accounting costs, the allocation of capacity costs to different pricing periods in this case is usually arbitrary. For example, one method attempts to identify peaking, intermediate, and base-load generation plant and then allocates the costs of these units to the peak, shoulder, and off-peak periods. Another procedure uses the probability of contribution to the peak based on the number of hours in each rating period in which demand exceeds some arbitrary threshold level, divided by the total number of hours in the year, as the allocation criterion. None of these methods satisfy the economic efficiency objective, and, therefore, references
1. For evidence that LRMC is rising for U.S. systems, see Gregory C. Krohm, "Growth and the Cost of Electric Power," Public Utilities Fortnightly (December 18, 1980), pp. 32-35.

86

THEORY

to peak-load and TOU rates in the subsequent discussions imply that these are based on LRMC. TOU metering (when this is justified) is the best way to apply pricing structures based on LRMC. The Hopkinson or two-part tariff with separate energy and peak demand charges is used widely, but if the consumer's maximum kilowatt or kilowatt-amp demand is not measured at the time of the system peak, then he or she could be unfairly penalized. If only kilowatthour metering is available, the capacity charge is sometimes levied on the customer's connected kilovolt-amps-for example, with a current limiting breaker or fuse to limit the maximum load. But this is even more questionable, since it requires that the relation between the consumer's peak demand and connected kilovolt-amps be accurately known. Interruptible tariffs are an extreme form of peak load pricing in which the customer agrees to be disconnected or shed at short notice when there is a power shortage. These prices have to be low because there is no burden on system capacity. Sometimes the interruptible customer is offered the option of remaining on the system at a time of shortage provided he pays a much higher price. In either case, when demand presses on supply the interruptible tariff increases rapidly either to a high value or to infinity if the customer is automatically shed. Fixed charges are most often related to consumer costs as described in Chapter 4. A lump sum payment may be levied to cover the initial cost of providing the service connection, or the repayment period may be spread over several years to provide relief to customers. Recurrent fixed costs are charged to meet the costs of meter reading, billing, and other repetitive expenses. In some cases, the charge based on a consumer's connected kilovoltamps is also called a fixed charge, but this is really a proxy for the capacity or kilowatt cost, which is a variable charge. In general, the conversion of the strict LRMC into applicable kilowatt, kilovolt-amp, or equivalent kilowatt-hour charges during different pricing periods requires the knowledge of other customer characteristics, such as the load factor, diversity or coincidence factor, power factor, ratio of connected kilovolt-amps to maximum demand, and so on, as illustrated numerically in 2 the following case studies. Tariffs contain power factor (PF) penalty surcharges in excess of the regular price to encourage consumers whose PF drops below some acceptable
2. For a given group of consumers, Diversity factor = (I/Coincidence factor) = (Sum of individual maximum demands)/ (simultaneous maximum demand of group). For a given consumer, p.f.
=

kilowatts/kilovolt-amps

2 2 kilowatts/(kilowatts + kilovars ) '.

EXPERIENCE WITH PRICINGSTRUCTURES

87

limit to install capacitative correction. This is because the current flow (I) required to meet a given kilowatt demand increases as the PF falls, thus increasing resistive or PR losses in the power system as well as the generation requirement. Fuel surcharge or fuel adjustment clauses are also becoming increasingly common. This permits the utility to quickly pass on to the consumer any unforeseen increases in fuel costs, especially of liquid fuels. Ideally, any changes in relative input prices would require reestimating the strict LRMC followed by changes in the tariff structure, but the legislative procedure to achieve the latter may take a long time. A convenient short-run 3 fuel adjustment clause can meanwhile provide much needed financial relief.

Recent Empirical

Evidence

Peak-load pricing or 7OU tariffs have been applied in varying degrees for 4 many years in Europe and more recently in some developing countries. In general, HV and MV industrial and commercial customers have been faced with separate capacity and energy charges, varying by time of day or season. Greater deviations are allowed for LV consumers, because of lifeline tariff considerations (as discussed earlier), or simplicity of metering-for example, demand charge based on maximum load limited by breaker or fuse, or demand charge rolled-in to energy charge using a typical user's load factor. Recurrent charges at a flat rate are also used to recover fixed costs. The conditions imposed on interruptible loads are an extreme form of peakload pricing, in that the price charged to these customers during periods of capacity shortage is either exceptionally high (if the consumer has the option to continue receiving power at an increased price) or infinite (if the load is automatically shed). Some typical power tariffs recently applied in several countries are summarized in Table 6-1. The Hv and MV tariff structures in European countries, such as England and France where substantial thermal generation exists, reflect the importance of both seasonal and time-of-day demand peaks as well as the variations in energy costs of the least efficient or marginal plants operated during the various pricing periods. In the Scandinavian countries where hydroelectric generation is dominant, seasonal variations in the availability of waterthe additional costs of storage in the winter months-tend to dictate the
3. For more practical details regarding tariff structures, see Gerhard F. K. Herman, Electric-

ity Tariffs (South Africa: University of Stellenbosch, 1974).

4. For details of the European experience with peak-load pricing, see Bridger M. Mitchell, William J. Manning, and Jan Paul Acton, Peak Load Pricing (Cambridge, Mass.: Ballinger Publishing Company, 1978); and Jan Paul Acton, Ellen K. Gelbard, James R. Hosek, and Derek J. McKay, "British Industrial Response to the Peak Load Pricing of Electricity," Report no. R-2508-DOE/DWP (Santa Monica, Calif.: Rand Corporation, February 1980).

88

THEORY

Table 6-1. Examples of Peak Load Tariffs

Hv and MV consumers

Country a. Europe
England (Northwestern Area Electricity Board, 1980)

Demand charge (dollars per kilowatt per month)

Energy charge (cents per kilowatt-hour)

Fixed charge (dollars per month)

>650 volts 0.0 5.91 (April-October) (0700-2400 daily) 6.33 2.78 (November-March) (2400-0700 daily)

1.0 per kilovolt-amp.

France (Electricite de France 1980)'

5-30 kilovolts 3.21 10.36 (November-February;b 0700-0900 and 1700-1900) 6.08' (October-March; 0600-2200) 3.06 (October-March; 2200-0600 and Sunday) 3.87b (April-September; 0600-2200) 2.42 (April to September; 2200-0600 and Sunday) 0.86 (October-April contract) 0.86 (all year contract; October-April) 0.44 (all year contract; May-September) 38-75

Norway (Water Resources and Electricity Board, 1975)

1.65 (October-April contract) 1.89 (all year contract)

Sweden (State Power Board, 1980)

70 and 130 kilovolts 0.27 1.35 (contractual) (May-August) 2.71 1.56 (peak load) (September-April)

WITH PRICING sTRUCrURrtS EXPERIENCE

89

LV

(domestic) consumers

Energy charge (cents per kilowatt-hour)

Fixed charge (dollars per month)

Percentage hydro'

8.48 (two part) 9.37 (white meter, 0900-2200) 4.0 (white meter, 2200-0900) 6.45 (3-kilowatt breaker) 6.45 (6-kilowatt breaker) 6.45 (9-kilowatt breaker) 6.45 (12-kilowatt breaker) 6.45 (15-kilowatt breaker) 6.45 (18-kilowatt breaker)

2.82

1-2

3.75

3.90 6.14 38-40 10.27 14.37 18.49 22.61

0.85 (regular rate) 3.34 (in excess of subscribed level)

2.78 per subscribed kilowatt

99

4.37 (0600-2200) 2.21 (2200-0600)

3.88 plus 2.24 to 0.59 per ampere of fuse rating (from 16 to 200 amps)

58-60

90

THEORY

Table 6-1 (continued)

HV

and

MV

consumers Fixed charge (dollars per month)

Country b. Developingcountries
Indonesia (Perusahaan Umum Listrik Negara, 1980)

Demand charge (dollars per kilowatt per month)

Energy charge (cents per kilowatt-hour)

12 and 20 kilovolt; 200-5000 kilovolt-amps 2.56d 4.64 (1800-2200) 3.20 (2200-1800)

..

Kenya (Kenya Power Company, 1980)

11 and 33 kv, > 100 megawatt-hrs per month 6.74d 3.75 (Monday-Friday; 0800-2200) 2.25 (all other times) 66 and 132 kilovolts 7.11 2.64

53.93

Pakistan (Water and Power Development Authority, 1980)

Note: To facilitate comparison, all prices have been converted into constant mid-1980 U.S.
dollars, using exchange rates prevailing in the year that the tariffs were applicable, and, where necessary, the U.S. implicit deflator of GNP up to 1980. ... None. a. Energy generated from hydroelectric sources in 1978-79, as a percentage of total generation. b. Excluding Sunday. c. Excluding Sunday and November-February; 0700-0900 and 1700-1900. d. Per connected kilovolt-amp. e. Includes 17.6 percent value-added tax. Source:World Bank data.

EXPERIENCE WITHPRICING STRUCTURES

91

LV

(domestic) consumers Fixed charge (dollars per month) Percentage hydroa

Energy charge (cents per kilowatt-hour)

18.5d

24

(0-200 volt-amps) 4.64 (201-500volt-amps) 6.24 (501-2200volt-amps) 7.20 (2201-6600volt-amps) 9.20 (> 6600 volt-amps) 3.30

4.48d 4.48d 4.48d 4.48d 2.25 75

(< 30 kilowatt-hours)7.49 (> 30 kilowatt-hours)

3.15
(< 40 kilowatt-hours)
4.57

...

80-82

(41-200 kilowatt-hours) 7.61 (> 200 kilowatt-hours)

92

THEORY

structure of prices. The HV and MV tariffs adopted in the developihg countries are quite recent and somewhat simpler because of the need to introduce these new pricing structures gradually, the shortage of appropriate metering, and so on. The LV domestic tariffs in all countries are also relatively uncomplicated particularly because less complex metering is used. For example, the demand charges (if any) are levied on the basis of subscribed kilovoltamps or maximum size of fuse or breaker. Off-peak night discounts are offered in some systems dominated by thermal generation. The generally low level of domestic tariffs, especially of lifeline rates, reflect the importance of sociopolitical considerations. Where these modern pricing structures have been in existence long enough to show results, their effect has been favorable. Industrial and commercial customers in Europe have responded to peak load prices in many ways, including: (a) changing their pattern of production by increasing activity during off-peak periods and vice versa, (b) using waste heat or combustible residual materials to generate their own energy during peak periods, (c) storing heat and energy for use during peak periods, and so on. The dominant response has been to shift load from the peak to the off-peak period rather than to cut back total electricity consumption. For example, peak loads of a wide range of European industries have been reduced by amounts ranging from 30 to 90 percent, whereas the group load factor of 70 percent of industrial load in the United Kingdom increased from 45 to 70 percent from 1961 to 1975.5 In the case of LV residential consumers, the price signals are less clearly conveyed because of the simpler metering used. Therefore, these tariffs have been effectively supplemented in several European countries by a coordinated package of other domestic load management techniques, including storage (space and water) heating during off-peak hours and central control of specific domestic loads, such as ripple control. Some problems have resulted in Europe, especially in France, because of peak shifting. A combination of techniques including use of a wider peak period, redefining the peak period to include both actual and potential peaks, direct control of certain loads, and so on may be used in order to overcome this difficulty. U.S. regulatory bodies and utilities have generally hesitated for a long time to give up conventional accounting or embedded cost approaches to

5. See J. A. Burchnall, "A Review of Off-peak Tariff Incentives and Their Effect in England and Wales," Report no. EP/SEM.5/R17 (Brussels: Economic Commission for Europe, Committee on Electric Power, March 1977); Y. Pelletier, "Application des Methodes d'Analyse de Donnees a l'Etude d'un Echantillon de Courbes de Charge Moyenne Tension," (Paris, Electricite de France, April 1977; processed) and Y. Pioger, "Elements de Reflexion sur la Deformation des Courbes de Charge Globales et Sectorielles: Evaluation de l'Effet Induit par la Tarification" (Paris, Electricite de France, July 1977; processed).

EXPERIENCE

WrTH PRICING STRUCTURES

93

6 rate making, and to adopt more modem techniques based on LRMC. This conservatism could generally be attributed to the steady decline in historical electricity pnices resulting from technological improvements and economies of scale. The recent worldwide increases in inflation rates and energy costs have reversed the trend, however, and these developments have stimulated interest in marginal cost pricing. The recent Public Utilities Regulatory Policies Act of 1978 has been an important step forward in helping to rationalize electricity tariffs in the lUnited States.' It may be interpreted in the spirit of the LRMC methodology, to structure tariffs and use marginal costs based on long-run considerations. The European experience and preliminary results from ongoing rate studies in the United States indicate that the benefits of implementing TOU tariffs (based on LRMC) for large Hv and Mv consumers could be substantial.8 Some specific results of TOU pricing experiments in the United States are summarized below, but should be interpreted cautiously. First, changes in load shape caused by TOU rates are difficult to analyze because demand in a given rating period will be affected not only by the price in that period but also by the price changes in other periods. That is, the own- and crossprice elasticity effects are difficult to separate. Second, short- and long-run responses may be quite different and cannot be easily distinguished. Finally, the effects of many other factors such as climate and user tastes also complicate the analysis. Typical results indicate that 515 large industrial and commercial users in five service areas of California, Michigan, New York, and Wisconsin with minimum demands ranging from 300 kilowatts to 5,000 kilowatts responded to TOU prices as follows.9 About 20 percent of the customers surveyed reduced their peak period, mainly by shifting equipment use, whereas the rest reduced their peak demand without shifting to off-peak use by more efficient operation of air-conditioning, ventilation, and lighting. Fifty-five percent of users reported no change in peak demand; on average 82 percent of these did not respond to TOU tariffs because their production process was inflexible or the change was not cost effective, and only 18 percent made no effort to

6. For a recent and more detailedreview of the application of marginalcost pricingin Europe and its implications for the UnitedStates, see MohanMunasinghe,"Principlesof Modern ElectricityPricing," Proceedings of the IEEE (March 1981);also availableas Reprintno. 185 from the World Bank, Washington,D.C. 7. U.S. Congress, House of Representatives, Public Utilities Regulatory Policies Act (PURPA),Reportno. 95.1750, 95thCong. 2 sess., October10, 1978. 8. See Mitchell and others, Peak Load Pricing, chap. 8; and S. P. Reynoldsand T. E. Creighton, "Time of Use Rates for Very Large Customerson the PacificGas and Electric Company System," IEEE Transactions on Power Apparatus and Systems, vol. PAS-99 (January-February 1980), pp. 147-51.
9. See NARUc Electricity Rate Design Study, Report no. 79 (Palo Alto, Calif.: EpRI, April

1980).

94

THEORY

change. Finally, 25 percent of the large users increased peak demand because of stepped up production. About 90 percent of all users surveyed indicated that they understood the TOU rates well, and nearly two-thirds had made studies of the effect of the new price structures. Studies involving TOU metering for residential consumers are now being carried out in several U.S. states. Some broad but tentative conclusions have already been reported for Arizona, Connecticut, Ohio, North Carolina, Rhode Island, and Wisconsin. These include reduced kilowatt demand during the peak hour, reduced kilowatt-hour consumption for the entire peak period, more favorable diversity factor for TOU consumers, reduction in intermediate period (or shoulder) demand, and no evidence of needle peaking outside the peak period.'" Another set of results from Wisconsin indicates that an eight-to-one ratio in the peak to off-peak price would reduce the average peak to off-peak energy consumption during the summer by 24 percent. The degree of response would be positively related to the number of large household appliances owned." Although all of these studies report on short-run results, additional long-run consumer responses may be anticipated given the increasing ability of households to change their patterns of appliance ownership and time of use. Application of TOU prices to Lv residential and commercial customers in the developing countries may be premature, however, and may require more study to determine whether the significant cost of additional metering is outweighed by the net savings from improved patterns of consumption. For example, a recent British study indicates that on the basis of such a costbenefit analysis of metering, TOU prices for domestic consumers would not 2 be justified. 1 More generally, techniques for load management, such as control of domestic water and space storage (heating and cooling), have been 3 very effective in Europe and show promise in the United States.' Available
10. Allen K. Miedema, "Overview of DOE Analyses of Residential TOU Data" (North Carolina, Research Triangle Institute, April 1980; processed); and Allen K. Miedema, S. B. White, C. A. Clayton, and D. P. Lifson, "Analysis of Experimental Time-of-Use Electricity Prices" (North Carolina, Research Triangle Institute, December 1978; processed). 11. Douglas W. Caves and Laurits R. Christensen, "'Residential Substitution of Off-peak for Peak Electricity Usage under Time of Use Pricing," Energy Journal, vol. I (April 1980), pp. 85-142. 12. Domestic Tariffs Experiment, Load and Market Research Report no. 121 (London: Electricity Council). 13. J. G. Asbury and A. Kouvalis, "Electric Storage Heating: The Experience in England and Wales and in the Federal Republic of Germany," Report no. ANLIES-50 (Argonne, Ill.: Argonne National Laboratory, May 1976); R. V. deGrasse, "Electric Storage Heating after Two Years," Public Utilities Fortnightly, vol. 23 (January 6, 1977); and B. F. Hastings, "Ten Years of Operating Experience with a Remote Controlled Water Heater Load Management System at Detroit Edison," IEEE Transactions on Power Apparatus and Systems, vol. PAS-99 (July-August 1980), pp. 1437-41.

EXPERIENCE

WITH PRICING

STRUCTURES

95

technologies for load control include radio and ripple control, power line carrier, telephone line, time switches, interlocks, load limiters, and so on. The usefulness of these methods in the developing countries, where marginal cost pricing is just 'being introduced, however, has to be verified further in light of wide differences in patterns of energy use, tastes, and climate. Except for the middle-income countries where air-conditioning and space heating is significant, such techniques would not be relevant in the near future. More important, even simple initial steps such as eliminating the promotional decreasing block tariffs would constitute a significant improvement in tariff structuring. The implementation and effectiveness of future pricing structures will depend critically on the development of new hardware and techniques. For example, advances in solid state technology may permit cheaper metering and justify more complex tariff structures even at the LV level. Also, as load management techniques improve, centralized control of customer appliances by the utility could be used more effectively and could be coupled with more accurate and practically instantaneous pricing signals. In fact, the complexity of the tariff structure would be limited only by the ability of customers to comprehend it. These ideas are embodied in the concept of homeostatic utility control, in which sophisticated devices including microprocessors would be used to perform a wide range of switching, metering, and signaling functions. Such developments presage an era in which the new philosophy of adjusting the demand to meet the available supply would increasingly supplement the existing practice of merely adjusting supply to follow variations in the demand. 4
14. MIT Energy Laboratoryt, New Utility Management and Control Systems (Cambridge, Mass., 1979).

PartTwo

Case Studies

Chapter7

Overview ofthe Case Studies

were made during a period of unprecedented increases in the real costs of electric power. These increases emphasized the need for more conservation than ever before. The focus on marginal cost pricing, which stresses real economic resource costs, is therefore particularly appropriate now. Despite the differences among the marginal cost pricing studies discussed in the following chapters, some broad patterns and issues may be identified. The studies follow the pattern advocated in Chapter 1. The first step involves estimating the strict long-run marginal cost (LRMC) of supplying different consumption categories, classified by time, voltage, and geographical area. The costs are estimated both for the power authority and for the economy as a whole. Therefore, some shadow pricing based on economic opportunity costs has been used. The second step in each case is to use the strict LRMC estimate to determine a tariff structure in light of several objectives and constraints. These include treatment of second-best issues, basic energy needs, access to service of low-income groups, regional variations, sector financial viability, metering problems, and so on. It is generally recognized that large price changes cannot be introdluced immediately; gradualism is, therefore, invariably part of each country's strategy in reforming its tariffs.
THE TARIFF ANALYSES PRE.SENTED IN THIS BOOK

Institutional, Technical, and Physical Characteristics

Although the general rnethodology used in these tariff studies is the same, the detailed application of those principles varies considerably. This is not surprising given the institutional, technical, and physical differences, as well as the variation in the quiality and quantity of available data. Thus, in determining tariff policy, Thailand has to consider the tariff structure of three authorities, one of which, EGAT, generates power and sells in bulk to the others, PEA and MEA, as well as to individual industrial consumers. Similarly, Pakistan has two main electric power sector institutions and several smaller ones, but Indonesia, Korea, and Sri Lanka essentially have only one electric power authority. In the Philippines numerous authorities are in99

100

CASE STUDIES

volved. NPC is the primary generating authority, selling power to large distributors and to a host of municipal, private, and cooperative utilities-some of which also have generation facilities. All cost calculations in the case studies are carried out in the local currencies. The relevant official exchange rates used, relative to the U.S. dollar, are: Country
Indonesia Pakistan Philippines Sri Lanka Thailand

Currencyunits
rupiah (Rp) rupee (Rs) peso (P) rupee (Rs) baht (B)

Officialexchangerate (valueof US$1)

625 9.9 7.4 15.5 20.4

The strict LRMC calculations are generally straightforward, except for the estimation of generation capacity costs. The Indonesian, Pakistani, and Sri Lankan case studies attempt to identify these costs based on the type and pattern of plant operation, especially during the peak period. In Thailand an average capacity cost is calculated based on all types of generating plants. It is then corrected for the net fuel savings, which offsets some of the higher capacity costs of base-load machines (see Appendix C). In the Philippine case, no such correction is made, so that the kilowatt charges imposed on peak-period consumers may be somewhat overestimated. The type of generation that is critical to the methodology used to estimate marginal generation costs also varies considerably. Thus the percentage of installed capacities of various types at the time of the studies were:
Type of generatintgplant Country
Indonesia Pakistan Philippines Sri Lanka Thailand

Hydro
22 62 72 87 33

Steam (coal/ligniteloil)
15 24 22 13 60

Diesel
23 0 5 0 1

Gas turbine
40 14 0 0 6

Other
0 0 1 0 0

Further, the countries' physical characteristics also determine whether or not one grid or several would be the most cost-effective means of supplying electricity. Although small systems operate separately from the main grid in all these countries, the Thai, Sri Lankan, and Pakistani studies essentially use one grid. This is also true of the Indonesian study only because data on the large grids serving islands outside Java are unreliable. A comprehensive national tariff study should eventually consider tariffs for various grids throughout Indonesia. This raises the thorny question of regional variations in tariff. Although this issue is also a factor in Thailand, it is much more

OVERVIEW OF THE CASESTUDIES

101

prominent in the Philippines, where a comprehensive examination of supply costs has been undertaken, involving several island grids that NPc serves. If a policy of uniform national tariffs is imposed for political reasons, the analysis of geographic variations in the marginal costs of supply permits the policymaker to estimate roughly the efficiency costs of this decision.

System Losses With system losses the question is whether it is fair to pass all these costs on to consumers. The issue is especially pertinent in Indonesia and Pakistan, where losses in the starting year averaged 25 and 34 percent, respectively, of gross generation. These values are relatively high. Normal loss levels, however, are not easy to define because of the individual characteristics of each power system. Technical losses may be excessive because of inadequate planning, poor operation and maintenance, or even government constraints on the utility's budget and revenue raising capability. Theft and unpaid bills make up the nontechnical component of losses. Since consumers' willingness-to-pay is undefined in this case, it is not clear whether the economic benefits of electricity use are comparable with the corresponding benefits a paying customer receives. In all the case studies presented in this book, costs resulting from both technical and nontechnical losses are imposed on the paying consumer. This practice becomes questionable, especially when losses are exceptionally high and could be partially corrected by using straightforward measures. Such a policy often constitutes a disincentive to improving the efficiency of utility management. Where significant losses exist, as in Indonesia and Pakistan, it is clear that a program should be implemented at the earliest opportunity to reduce losses to acceptable limits.' The strict LRMC calculations help to pinpoint losses in various parts of the system, so that remedial measures may be taken. More generally, the presence of high marginal costs at various voltage levels (for example, EHV/HV transmission, MV distribution, and LV distribution) indicate that losses are high or that investments are too great relative to incremental demand. Thus, in addition to providing the basis for the tariff structure, the LRMC calculation enables the analyst to comprehensively examine a wide range of information about the system and therefore to identify weak areas and inconsistencies in both planning and operations.
1. Average generation losses from station use may vary from about I percent for hydro plant to more than 5 percent for coal-fired steam plant. Normal transmission and distribution losses usually average between 7 to 12 percent of gross generation. Theft in U.S. systems has been estimated at about 2 percent of gross generation, but norms in the developing countries may have to be set at somewhat higher levels. See Alan J. Donziger, "'The Underground Economy and the Theft of Utility Services," Public Utilities Fortnightly (November 22, 1979), pp. 23-27.

102

CASE STUDIES

System Planning
The need for sophisticatedsystemplanningto optimizeinvestment scheduling and plant operationis widelyrecognizedin the electricpower industry. The authoritiesresponsiblefor the systemsdescribedin this book are no exception. Uncertaintiesand data limitationsabound, however. On the supply power are inadequate, side, historicaldata on stream flows for hydroelectric and future oil prices are uncertain.On the demandside, problemsof pricing stemmingfrom the foregoingand other uncertainties are particularly difficult to handle in a developingcountry. Demandforecasting,which is hazardous even in a developedcountry, becomesespeciallyproblematicbecauseof inadequatedata. In particular,past trendsare frequentlya misleading guide to the future, since consumptionis typicallyconstrainedby supplyand demand is unsatisfied, makingit very difficult to determinethe full potentialor underlying demand. A common approach to demand forecastingseems to be used in all the countries, in whichforecastsbased on correlation with macroor sector-by-sector economicvariablesare reconciledwith a microeconomic buildup of consumerdemand. Greater relianceappears to be placedon the latter method, particularlywhere shortagesare endemic and wheredemand must be tailored to suit supply (on a sectoralconsumptionbasis).

Financial Viability
As noted in Chapter 6, financial viability is a necessary condition for achievingthe various objectives of utility operations, and one which is of continuingand pressing concernto the utilities.In countrieswherefinancial
viability is of some concern, such as Indonesia and Pakistan, the proposed tariff structure is a scaled down version of the strict LRMC, which meets traditional financial targets, usually defined as a rate of return on assets. The imposition of the strict LRMC, even if modified considerably to reflect social or other economic goals, would imply a significant increase in prices, usually well above those required to meet traditional financial criteria. Financial viability in a situation of increasing real unit costs is, therefore, quite compatible with LRMC pricing. Recent developments indicate, however, that increasing financial difficulties are being encountered by power utilities all over the world as the real costs of energy rise and as consumers become more articulate in voicing their disapproval of tariff increases. In this situation the imposition of strict LRMC pricing to increase tariffs would place an additional sociopolitical or public relations burden on the utility management. The problem of consumer education in this situation is generally recognized. Paradoxically, the- economic rationale is strongest for raising prices to control demand in the short run, when shortages are prevalent

OVERVIEW OF THE CASE STUDIES

103

and the quality of service is especially poor. In such a situation pure economic theory suggests price rationing, in which prices would be dictated not by the LRMC of supply, but rather by the highest amount bidders would be willing to pay for scarce electric power. For public relations and political reasons, however, raising prices-even temporarily-during periods of shortage when the quality of service is poor would probably cause the greatest possible public outcry. Little attention is paid in the studies to the possibility of pricing electricity to generate public revenues. It is perhaps not surprising that none of the case studies refer explicitly to this issue, since the studies have been prepared by power utility staff members who would traditionally not consider it appropriate for the power authority to play the role of tax collector.

Divergences from the Strict LRMC

Although the need to increase the average level of tariffs is apparent in all cases, the extent to which different consumer categories receive preferential or unfavorable treatment varies considerably. Each case study, therefore, considers the question of tariff structures closely. Given the difficulty of raising average tariff levels, revising tariff structures appears to offer the main scope for improvement in the near term. Thus in the Indonesia and Thailand studies attention is drawn specifically to the distortions that arise because of the government's policy of levying a national uniform tariff, which in effect involves large cross-subsidy between metropolitan and more remote provincial electricity consumers. In the Pakistan study, however, it is implied that a move toward uniformity might be desirable in the interest of subsidizing rural consumers in a way that is more equitable to urban customers than is the present system. As an interim step, a similar policy is advocated in the Indonesia study. Conversely, the Philippine marginal cost calculations show what is generally considered to be a surprising result: namely, that real economic costs of power do not in fact vary significantly from region to region and that even more uniform tariffs would result from applying marginal cost pricing in the country. This is an extremely important finding, especially in view of the government policy of encouraging industry to locate in the Mindanao region by subsidizing electricity prices because of the assumed relatively low costs of hydropower in that area. In all the tariff studies the LRMC of capacity greatly exceeds the existing kilowatt charges. In fact, the allocation of costs between capacity and energy receives much attention. The Philippine study is especially noteworthy in this regard, since it points out that the failure to adequately recognize marginal capacity costs in the tariff structure of the principal bulk supplier compounds the distortions inherent in the tariff structures of distribution au-

104

CASE STUDIES

thorities. This is because the prices passed on to the ultimate consumer are based on bulk purchases and, therefore, the contribution of the underlying capacity costs at the bulk level may be disregarded at the retail level. In Sri Lanka and Indonesia LV residential consumers are given favorable treatment. In the Sri Lankan case study a declining block rate is specifically cited as encouraging the waste of energy by consumers. In the Pakistan study the favorable treatment accorded to residential consumers, tubewell operators, and those consumers with relatively poor load factors is significant. The desirability of using the increasing block rate for domestic consumers is generally recognized in the studies. Clearly, however, care must be exercised in determining the size of the first block. Thus the recommendation of 30 or 40 kilowatt-hour subsidized blocks in the Thai and Pakistani studies is based on careful analysis of household electricity use, whereas the first block of 200 kilowatt-hours suggested for Manila households implies a substantial subsidy that is hard to justify on the grounds of basic need.

Special Metering
An important issue in determining an appropriate tariff structure concerns the introduction of special metering to record the consumption during peak and off-peak periods. The case studies reveal a general recognition of the need to shift loads from peak to off-peak periods; time-of-day metering receives some consideration. For example, the Pakistani study strongly recommends time-of-day metering for all Mv and HV consumers. The cost-benefit analysis for time-of-day metering, which may be a simple exercise for large consumers, becomes more difficult the smaller the individual load that is being considered. The need for analysis is stressed in the Pakistani study and implied in others. In particular the need to obtain more accurate information on load factors of individual consumers and large consumer groups is stressed.

Market Distortions
Market distortions are recognized and treated in all the studies. In calculating the strict LRMC, estimates are made with and without shadow pricing. Given the prevalence of policies in which the domestic prices of oil, natural gas, and coal tend to diverge from their opportunity costs, the result of the adjustment in all cases is the same. Although marginal capacity costs tend to be slightly greater when calculated in shadow-priced economic terms than in market-priced financial terms, the shadow-priced marginal energy costs are usually considerably greater than financial estimates, especially where the fuel used for thermal electricity is heavily subsidized, as in Indonesia. The

OVERVIEW OF THE CASE STUDIES

105

difficulty of estimating shadow prices is recognized; special problems arise in estimating the economic costs of domestic, depletable primary energy sources, such as the natural gas resources of Pakistan. The determination of the opportunity cost of natural gas, in light of options such as leaving it underground, liquefying and exporting it, using it for new industrial activities, or replacing imported or domestic oil or coal in other uses, is complex. The need for careful analysis is stressed in the Pakistani study. Similar problems are encountered with regard to Thailand's depletable natural gas and coal resources and Indonesia's oil, coal, and natural gas resources.

Second-best and Political Considerations

Second-best issues are also discussed in several of the studies and appear as constraints to a strict LRMC tariff for electric power. Thus in the Thai, Indonesian, and Pakistani studies the subsidization of diesel fuel is recognized as setting an upper bound on the pricing of electric power. The pricing limit occurs because a LRMC tariff might prompt users to switch to captive plant (already a particularly serious issue in Indonesia), to run small industry or tubewells by using subsidized diesel, or to take other actions that might often not be in the best interests of efficient allocation of resources. Other issues, including the possible inflationary effect of electricity price increases and the adverse effect on domestic industrial development or the attractiveness of the country to multinationals, are also discussed and essentially dismissed, as in the Thai study. The tariff studies contained in this book represent conscientious efforts to achieve precision in estimating marginal costs. But the political and other difficulties inherent in attempts to significantly change tariff levels or structures have never been far from the minds of the authors. This consideration has resulted in a healthy balance between the economic and noneconomic considerations in setting tariffs, as represented by the effort devoted to the calculations of the strict LRMC and to subsequent adjustments. Although the unwillingness of governments to pass on all the increase in the economic costs of power to consumers is understandable, these studies help to emphasize that failure to do so will make it increasingly difficult for their successors-and future generations-to handle the problem, especially when the standards of service are already low.

Governmental Attitudes toward Reforms

Social and governmental resistance to tariff reform is prevalent throughout the world. In the countries discussed in this book, however, several adjustments to tariff levels and structures that are consistent with the LRMC ap-

106

CASE STUDIES

proach have recently been observed. For example, in Indonesia and Pakistan, tariff structures based on LRMC, which were similar to those described in the respective case studies, were introduced in 1980. In the Philippines, Thailand, and Sri Lanka existing tariffs have been partially adjusted or restructured to avoid the obvious economic inefficiencies highlighted by the studies. Thus the studies have been associated with a general environment of change in this area. As illustrated in the preceding chapters, it appears that throughout the world similar types of reforms are taking place, often as an intuitive response to shortages and the rising costs of energy. Sometimes such changes, which may be generally consistent with the marginal cost pricing approach, are made without any detailed knowledge of the rational and theoretical underpinnings of the principle. It is, however, our contention that although such a broad-brush approach to tariff reform may be desirable, much work remains to be done in refining the necessary analysis, particularly with regard to systematically determining the adjustments to the LRMC that might be warranted. Although they are not necessarily model investigations in all respects, the efforts demonstrated in the following case studies are welcome evidence of progress in this direction.

Chapter 8

Indonesia
COST PRICING CALCULATIONS were used to determine an appropriate level and structure of electricity prices in Indonesia in 1979. A revised power tariff for Indonesia was particularly needed at the time for several reasons. First, the existing tariff was originally introduced in 1973 following an earlier power pricing study. Because of administrative regulations governing tariff changes, however, subsequent increases raised the average price by almost 100 percent between 1973 and 1978 by applying a surcharge to the energy price alone. Therefore, the structure of tariffs was badly distorted; in particular, the ratio of kilowatt-hours to kilowatt charges was far too high. Second, since price increases were introduced in an ad hoc manner and were based almost exclusively on financial requirements, the level and structure of tariffs did not reflect the economic costs of supply for the various types of consumers. Third, until quite recently, the government utility, Perusahaan Umum Listrik Negara (PLN), could not satisfy the rapid growth of demand of the Indonesian power market, and, therefore, its policies were not designed to encourage new consumers. For example, payment of the full connection charge was required in advance; for medium and large consumers this amounted to the substantial figure of Rpl20 per voltamp. Even some customers who had paid the necessary fees were not connected because of lack of system capacity ahd administrative bottlenecks. As a result, there was a rapid increase in captive generation, amounting to almost half the total Indonesian power market by 1.977. Since 1978 the situation has improved considerably, however. PLN has embarked on a program of balanced system expansion, including building more based-load generating plants, strengthening transmission interconnections, relieving distribution bottlenecks, and upgrading existing local networks. PLN has also tried to increase its market share by achieving high rates of load growth (values of more than 20 percent projected for the next five years) and, in particular, MARGINAL

Note: This chapter was prepared in collaboration with Soejitno Soejanto and Djaya Santoso
(PLN).

The views expressed in this chapter do not necessarily reflect those of

PLN

or the govern-

ment of Indonesia.

107

Rol

'0

'0

INDONESIA

109

they have moved toward a more liberal connection policy. Therefore, it was important that the revised tariffs, including the connection policies, reflect this new outlook.

Organization of the Sector

The electric power sector in Indonesia includes PLN-the governmentowned enterprise responsible for all generation, transmission, and distribution of electricity in the country-captive plant installations, and some municipal franchises.' It was estimated in 1979 that the capacity of all captive plants equalled that of PLN. The total generating capacity available to PLN at the end of fiscal year 1978 was approximately 2,000 megawatts, 40 percent of which was in the form of gas turbines, 23 percent diesel, 22 percent hydro, and 15 percent steam.2 In Java, gas turbines accounted for 50 percent of the total 1,255 megawatts in 1978. This development of generating capacity evolved from a crash program of gas turbjine installations in the early 1970s when PLN suffered large breakdowns in generating capacity. Then gas turbines were seen as the only way to add new generating capacity quickly. A better balance will be reached with the commissioning of new steam generating capacity at Perak and Gresik (East Java), Semarang (Central Java), and Muara-Karang (West Java). These new steam stations will add 1,200 megawatts of baseload capacity by 1982. At that time, the PLN system in Java will have 53 percent steam, 20 percent hydro, and 27 percent gas turbine plants; all small diesel plants will be retired. The proportion of steam, hydro, gas turbines, and diesel in the entire country (including the islands outside Java) will be 41 percent, 18 percent, 20 percent, and 21 percent, respectively. PLN operates about 3,800 kilometers of transmission lines, mostly in Java. About 2,000 kilometers of high voltage (70/150 kilovolt) transmission lines are currently being constructed and will be completed by 1982. Map 1 shows the existing 150-kilovolt system and extensions in progress in Java. The currently isolated systems of West, Central, and East Java will be interconnected in 1980 to form an all-Java grid. In addition to its main transmission lines, PLN now operates distribution systems totaling some 40,000 kilometers. PLN'S plans to expand capacity, particularly in Java, are based on tight reserve margins. The existing diesel captive installations with any remaining economic life will continue to be used by their owners as emergency standby
1. The captive plants were installed by private parties to produce electricity for their own use. 2. This includes the 125-megawatt hydroelectric station owned by the Jatiluhur Authority.

110

CASE STUDIES

or during peak periods when supply is restricted. Captive power installations in Java, with a few exceptions, are too small to be interconnected with the PLN system; they will continue to be poorly used. Where the equipment is in satisfactory condition, PLN can appropriate it for limited use in isolated areas where public supply must be introduced. Although these installations are old, and it is difficult to obtain spare parts for repair, any usable equipment should not be left idle when the sector needs it. Indonesia's population has low access to electricity compared with other countries. Approximately 6 percent of the households were connected by mid-1977 compared with 27 percent in Thailand and 32 percent in the Philippines. Regional imbalances exist as shown below by the percentages of households having connections in various regions.
Region
Ambon Bali and other islands Java (includes Jakarta at 20.5 percent) Kalimantan Sulawesi Sumatra West Irian All Indonesia

Domestic connections (percentage of all households)

3.2 2.1 9.6 5.0 4.2 4.3 4.6 5.9

Rural electrification has been undertaken in a small way from government funds provided to PLN. A rural electrification division was formed in 1976 in PLN to promote such projects, and the division has several projects in view. Demandforecast and system characteristics Several attempts have been made in recent years to prepare a usable demand forecast for Indonesia. This has been difficult because of low levels of consumption and the fact that the public power supply did not cover the total market. Furthermore, the restrictions on PLN'S ability to provide supply made its past record of sales an unreliable base for future projections. In 1973, an American consultant, Charles T. Main, applied an indirect method of measuring growth probabilities in selected sensitive economic sectors and their probable effect on other sectors. A more direct detailed assessment of the underlying demand in Java was made from 1974 to 1976 by Preece, Car3 This study quantified the effects of dew, and Rider of the United Kingdom. restricting supply during peak hours, at low voltage, at self-generation, and on PLN'S waiting lists. There have been no similar comprehensive surveys of demand for the other islands.
3. This is the demand PLN could have served if its installations had been adequate and if it served self-supplied customers.

INDONESIA

111

Assuming that PLN'S sales would not immediately rise to the levels of future underlying demand, Preece, Cardew, and Rider estimated in Java that a certain segment of this rnarket will continue to be met by captive installations for some time. The researchers also determined what it considered realistic sales targets on which PLN could base its future development program. The targets are substantially lower than the underlying high market forecast, but are of the same order of magnitude as a 1976 independent macro estimate by the International Atomic Energy Authority (IAEA) that used per capita GNP projections. These studies show that in the short and medium term, PLN'S future sales will depend entirely on the rate at which PLN can satisfy the existing demand. In the long term, however, sales are expected to reflect more closely the growth of the power market based on growth of sales. All forecasts are given in Table 8-1. PLN'S

Table 8-1. Forecasts of Future Generation Requirements

(gigawatt-hours)
Year 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990
PCR

(high),

PCR

(low)a

IAEA

(low)

PLN/World

Bank (1978)b
5,780 7,005 8,491 10,291 12,388 14,771 17,332 20,478 22,643 36,039 29,945 34,437

13,826 16,078 18,538 21,513 24,894 27,212 31,393 36,187 41,684 47,986 55,266 53,669

7,500 9,728 11,639 13,285 15,054 16,999 19,198 21,646 24,370 27,393 30,739 34,439

9,165 12,926 20,142 43,400

Not applicable. a. Estimates by Preece, Cardew, and Rider, U.K. b. The figures in this column are for PLN'S fiscal year, which ends in March of the succeeding year. To compare with the figures in the preceding columns, they should be decreased by about 5 percent up to 1984 and 4 percent thereafter.

In this study, the disaggregate Java demand forecast is determined at three (iiv), medium (Mv), and low (LV)-assuming that voltage levels-high there is enough interconnection to treat the three grids-West, Central, and East Java-as a single system for the 1979-87 period. The system peak demand at the generators is determined by taking PLN'S current energy forecast and applying the projected load factor of 0.68. As described in Chapter 3 (see Figure 3-2) and summarized in Tables 8-2 to 84, the peak demand at each lower voltage level is estimated by deducting the

112

CASE STUDIES

Table 8-2. Demand and Loss Forecast at Time of System Peak for Java
(megawatts)

Year
1977-78 1978-79 1979-80 1980-81 1981-82 1982-83 1983-84 1984-85 1985-86 1986-87

Generation (gigawatthours)
3,909 4,769 5,780 7,005 8,491 10,291 12,388 14,771 17,332 20,478

System peak at generation (1)

656.2 800.6 970.3 1175.9 1425.4 1727.6 2079.6 2479.7 2909.6 3437.8

EHV

Station use and HV System HV transmispeak consumer peak sion loss at HV (2) (3) = (1) - (2) (4)
63.98 82.06 106.73 135.23 174.61 220.27 280.75 347.16 407.34 481.29 592.2 718.54 863.57 1040.67 1250.79 1507.33 1798.85 2132.54 2502.26 2956.51 13.00 43.76 68.13 146.4 196.51 229.27 262.02 294.77 327.52 360.27

Note: The columns (1) to (9) correspond to power flows measured at the same points in Figure 3-2. Source: PLN data.

Table 8-3. Incremental Demand Forecast at Time of System Peak for Java
(megawatts)
HV

Year
1977-78 1978-79 1979-80 1980-81 1981-82 1982-83 1983-84 1984-85 1985-86 1986-87 Total
-

Discount factor (A)

1 0.8929 0.7972 0.7118 0.6355 0.5674 0.5066 0.4523 0.4039 -

Generation peak
656.2 800.6 970.3 1175.9 1425.4 1727.6 2079.6 2479.7 2909.6 3437.8 -

Peak
592.2 718.54 863.57 1040.67 1250.79 1507.33 1798.85 2132.54 2502.26 2956.51 -

Increment (B)
126.34 145.03 177.1 210.12 256.54 291.52 333.69 369.72 454.25 -

Discounted increment (A) x (B)

126.34 129.50 141.18 149.56 163.03 165.41 169.05 167.22 183.47 1394.76

Notapplicable. Source: Derived from Table 8-2.

INDONESIA

113

MV losses (substations + MV lines) (5)

System peak at MV (6) = (3)-(4)-(5) 527.07 614.05 723.85 813.79 959.39 1182.2 1421.57 1699.94 2011.63 2401.52

MV

consumer peak

At system peak time Own peak (7)a (7)b 60.55 69.86 88.68 97.24 123.18 165.82 220.45 288.93 374.80 447.68 121.10 139.71 177.35 194.47 246.35 331.64 440.90 577.85 749.60 895.36

System LV losses (substations peak at Lv + lines) (9) = (8) (6)-(7)a-(8) 155.23 179.58 200.08 225.71 250.86 304.91 342.32 402.14 441.94 527.54 315.15 364.61 435.09 490.84 585.35 711.47 858.80 1008.87 1194.89 1426.30

52.13 60.73 71.59 80.48 94.89 95.86 115.26 137.83 163.11 194.72

MV

LV

Peak 527.07 614.05 723.85 813.79 959.39 1182.2 1421.57 1699.94 2011.63 2401.52 -

Increment (C) 86.98 109.8 89.94 145.6 222.81 239.37 278.37 311.69 389.89 -

Discounted increment (A) x (C) 86.98 98.04 71.70 103.64 141.60 135.82 141.02 140.98 157.48 1077.26

Peak 315.15 364.61 435.09 490.84 585.35 711.47 858.80 1008.87 1194.89 1426.30 -

Increment (D) 49.46 70.48 55.75 94.51 126.12 147.33 150.07 186.02 231.41 -

Discounted increment (A) x (D) 49.46 62.93 44.44 67.27 80.15 83.60 76.03 84.14 93.47 641.49

Table 8-4. Average and Peak Loss Factors for Energy and Power at Time of System Peak for Java
(percentage of incomingenergyand power)

Year
1977-78 1978-79 1979-80
1980-81 1981-82 1982-83

Total losses as a perLV losses centageof Mv losses + EHV/HV generation; EHV/Hv average Average Average Average Average energyand Peak Peak energyand energyand Peak Station energyand energy and power loss power loss power loss power loss power loss power loss power loss power loss use (1) (2) (3)=(1)+(2) (4)=(1)+[ax (2)] (5) (6)=cox (5) (7) (8)=a x (7) (9) Station use
3 3.5 3.5
4 4 4.5

4.5 4.5

5
5 5.5 5.5

7.5 8 8.5
9 9.5
10

9.75 10.25 11.00

11.50 12.25

6 6 6
6 6 5

9 9 9
9 9 7.5

22 22 21
21 20 20

33 33 31.5
31.5 30 30

24.9 23.6 21.8

21.2 21.2 21.2

1983-84 1984-85 1985-86 1986-87

4.5 5 5 5

6 6 6 6

10.5 11 11 11

12.75 13.50 14.00 14.00 14.00

5 5 5 5

7.5 7.5 7.5 7.5

19 19 18 18

28.5 28.5 27 27

21.1 21.4 20.9 20.7

megawatts 1.5. Column(9) Note:a = peak megawatts/average Source: Peak losses from Table 8-2.

generated). losses)/(total kilowatt-hours kilowatt-hour (total system

INDONESIA

115

system losses and consumer demand that is drawn off directly at the higher voltage level. The HV and LV peaks coincide with the system's peak hoursapproximately from 1800 to 1900 hours in the evening. The contribution of MV consumers to the system's peak demand is estimated to be about 0.5 of their own peak demand, which occurs during the day. Typical daily load duration curves (LDC) for West, Central, and East Java, and the composite all-Java annual curves are given in Figure 8-1. The system peak occurs everywhere at or around 1900 hours. Seasons did not significantly affect the shape of the LDC as shown in Figure 8-2. Also the dry season peak of 611 megawatts is 14.2 percent greater Figure 8-1. Typical Daily Load Curves for Java
600 550
-

Dry season (September 1977) Wet season (February 1977)

500

400 -

i300

E250
200 -

15(

12
Hours of the day

16

20

24

116

CASE STUDIES

Figure 8-2. Typical Daily Load Duration Curves for Java

sa600

611
~~~~~~~r

seaso (Seteber

1977)

Wet season (February 1977) E

U

_,200

12 Hoursof the day

16

20

than the wet season peak of 535 megawatts. The actual growth of demand in that year was 23 percent, however. This means that demand grew from 535 to 656 megawatts, and, therefore, during the seven-month period from February to September, the prorated increase in the system peak would have been 13.5 percent. Thus, the difference in magnitude between the wet and dry season peak demands may be accounted for essentially by normal load growth, implying no seasonality of demand. Therefore, the demand for electric power varies principally by time of day, and it is assumed that the evening peak, caused mainly by residential consumers, would remain dominant during the 1979-87 period. The rapid growth of HV demand at high load factor would increase the demand uniformly throughout the day, whereas the increase in MV demand would not be large enough to make the day peak the dominant one.

Computationof the Strict LRMC for Java

The long-nin marginal cost (LRMC) is calculated from the following components: peak capacity cost, peak and off-peak energy costs, and customer costs. All costs have to be adjusted for losses up to the point of delivery to consumers. Capacity costs per kilowatt are converted to annual costs per kilowatt by annuitizing over the lifetime of the equipment. The costs are computed in constant 1978-79 rupiah (Rp), at border prices (BP). Because of inadequate data, a relatively simple shadow pricing frame-

INDONESIA

117

work is applied. The new official exchange rate (OER) of Rp625 = US$1 is used to convert foreign costs to rupiahs (in BP). Local costs are converted to BP by the conversion factor (CF) 0.9.4 The total costs-that is, foreign plus local-in BP are transformed into domestic-priced costs (DP) by dividing by the standard conversion factor (SCF), which is assumed to be 1.0 following the recent devaluation of the rupiah (from Rp410 to Rp625 = US$1). In comparison, the costs without shadow pricing are also shown. The discount factor is 12 percent, the opportunity cost of capital (OCC). Peak capacity costs Capacity costs for each peak kilowatt delivered are first determined at the voltage level at which they are incurred. They are then converted to costs per kilowatt annually by annuitizing at 12 percent over the plant's life-span. The annual capacity costs for every kilowatt at lower voltage levels are determined by dividing by the appropriate power loss factor (see Table 8-4). As described in Chapter 4, the LRMC of capacity may be as the ratio of the change in system capacity costs AC associated determined with an incremental increase AD in the long-run peak demand. The capacity cost of generation ACIAD would be annuitized over the lifetime of the plant in terms of its present value. This figure must be adjusted for reserve margin and losses at the various voltage levels. In an optimally planned system, the expansion program would normally be changed to meet the new incremental load by advancing the commissioning date of new plant or inserting units such as gas turbines or peaking hydro. In Java the generating system is imbalanced. In 1977-78, there were almost 600 megawatts of gas turbines with a total installed capacity of some 1,300 megawatts; the total is 1,060 megawatts when only dependable capacity is considered. Consequently, gas turbines usually used for peaking duty are currently being used to serve base-load. Therefore, to restore the generating balance by 1983 or 1984, about 1,500 megawatts of base-load steam plant will be added to the system. There will be large reserve margins, sometimes exceeding 35 percent of the installed capacity, until 1983 or 1984. The fuel savings alone from displaced gas turbines will help to justify these additions to the base-load plant. Consequently, in the short run an inGENERATION.

4.

Percentage of local costs

CF

Unskilledlabor Other Localcosts CF

35 65

x 0.7 = 0.25 x I = 0.65 = 0.9

Table 8-5. Generation Capacity Cost for Java (Saguling Hydro Plant First Unit of 175 Megawatts)
(constant billions of rupiahs)

Year
1978-79 1979-80 1980-81 1981-82 1982-83 1983-84 1984-85 1985-86 1986-87 Total

Foreign Local (1) (2)

2.1 6.15 7.65 10.2 7.65 4.05 2.1 39.9 2.04 3.5.04 3.72 3.2.04 0.96
-

Not shadow priced, total (3) =(1) +(2)

4.14 9.15 12.69 13.92 10.65 6.09 3.06
-

Shadowpriced, Shadowpriced, Discount Not shadowpriced, Shadowpriced, local total factor discounttotal discountedtotal (BP) (4)=(2) x (09) (5) =(l) +(4) (6) (7) =(3) x (6) (8) =(5) x (6)
1.84 2.7 4.54 3.35 2.7 1.84 0.86 _ 17.82 57.72 3.94 8.85 12.19 13.55 10.35 5.89 2.96 _ -

1 0.8929 0.7972 0.7118 0.6355 0.5674 0.5066 _

4.14 8.17 10.12 9.91 6.77 3.46 1.55 _ 44.12

3.94 7.9 9.72 9.64 6.5 3.34 1.50 _ 42.61

19.8

59.7

Note:The total costs include the full darn costs (equal to 13.5 percent of the total costs). Only half of dam costs (or 6.75 percent of total costs) should be allocated to capacity costs. Therefore, adjusted total discounted capacity costs are: (a) shadow priced, 42.61 x (I - 0.0675) = 39.73; and (b) not shadow priced, 44.12 x (I - 0.0675) = 41.14. Not applicable. Source:PLN data.

iDONESIA

119

cremental block of peak period customers can be served from gas turbines that have been freed at essentially no capacity cost. New peaking capacity will be required only by about 1984, when the reserve margin will have become low and the Saguling hydro units will come on-stream for peaking duty. Therefore, the LRMC of generation capacity will be represented by the costs of the Saguling units discounted to the present time, appropriately annuitized, corrected for losses, and so on. The average costs of the first 175megawatt unit from Table 8-5 are used to calculate the annual generation capacity costs per kilowatt at Hiv, Mv, and LV in Table 8-6. All transmission and distribution (T&D) or kilowatt, costs. The present capacity, to allocated costs are investment of incremental consumpkilowatt every costs for T&D value of discounted incremental cost (AIC) the average using tion served may be computed method. This involves the discounted ratio of future investment costs to incremental kilowatt served, as defined in Chapter 4. Transmission investment costs at EHv/HV are computed from the transmission investment plan, as shown in Table 8-7. Distribution investment costs are allocated between MV and LV levels and are similarly tabulated. The incremental costs and demand at each voltage level are summarized in Table 8-8. The AIC is calculated as the ratio of the above quantities for T&D.
TRANSMISSION AND DISTRIBUTION. TOTAL PEAK CAPACITY COSTS. These costs are summarized in Table 8-9 at different voltage levels in domestic prices. The method of derivation involves use of the incremental costs in Table 8-8 combined with peak loss factors in Table 8-4, as described in Chapter 4.5

O&M and A&G costs Operation and maintenance (O&M) as well as administrative and general (A&G) costs are taken from PLN financial projections and are calculated as annual costs per kilowatt. Ideally, they should be apportioned between kilowatt and kilowatt-hour costs. Since these costs are small, however, amounting to only a few percent of investment costs, and since no accurate data on allocation are available, they are apportioned only among the capacity costs of generation, transmission, and distribution. A typical computation for O&M costs per kilowatt annually is shown in Tables 8-10 and 8-11. Since this is a recurrent item of annual expenditure, annuitization is not necessary.

5. For details, see Mohan Munasinghe, 'Indonesia:

PLN

Power Tariff Study," Energy De-

partmentReport, World Bank, Washington, D.C., May 1978.

120

CASE STUDIES

Table 8-6. Calculation of Generation Capacity Costs for Java

(Saguling Hydro Plant First Unit of 175 Megawatts)

a. Shadow priced HV consumers 39-73 x 107Rp x (1.2 / 0.927) 175 x 103 kilowatts Per kilowatt = Rp293,888 Per kilowatt per year = Rp293,888 x 0.120925 = Rp35,538
MV

consumers Per kilowatt = Rp293,8888 / 0.918 = Rp320,139 Per kilowatt per year = Rp320,139 x 0.120925 = Rp38,713 consumers Per kilowatt = Rp320,139 / 0.7 = Rp457,341 Per kilowatt per year = Rp457,341 x 0.120925 = Rp55,304

LV

b. Not shadow priced HV consumers 41.14 x 10 Rp x 1.2 / 0.927 175 x 103 Per kilowatt = Rp304,318 Per kilowatt per year = Rp304,318 x 0.120925 = Rp36,799
MV

consumers Per kilowatt = Rp304,318 /0.918 = Rp331,501 Per kilowatt per year = Rp331,501 x 0.120925 = Rp4O,086

LV

consumers
= Rp331,501 / 0.7 = Rp473,572

Per kilowatt

Per kilowattper year = Rp473,572x 0.120925= Rp57,266

Note: Lifetime = 43 years (from PLN data); discount rate = 12 percent = 0.12; annuity factor = 0.120925; 0.927, 0.918, 0.7 are peak power loss factors at EHV/HV, MV, and LV averaged over 1977-84 (from Table 8-4); and 1.2 is factor for 20 percent average reserve margin.

Table 8-7A. Transmission Capacity Costs Based on the EHV/HV Investment Plan
(constant billions of rupiahs) Not shadow priced, discounted total (7) = (3) x (6) 37.98 55.93 38.46 38.26 60.43 42.70 20.71 17.04 29.99 341.50 Shadow priced, discounted total (BP) (8) = (5) x (6) 36.97 54.80 37.72 37.21 59.08 41.85 20.34 16.64 29.39 334.00

Year 1978-79 1979-80 1980-81 1981-82 1982-83 1983-84 1984-85 1985-86 1986-87 Total
--

Foreign (1) 27.9 49.92 39.08 38.95 73.89 60.25 33.55 28.90 59.36 411.80

Local (2) 10.08 12.72 9.16 14.81 21.20 15.01 7.33 8.77 14.89 113.97

Not shadow priced, total (3) = (1) + (2) 37.98 62.64 48.24 53.76 95.09 75.26 40.88 37.67 74.25 525.77

Shadow priced, local (4) =(2) x (0.9) 9.07 11.45 8.24 13.33 19.08 13.51 6.60 7.89 13.40 102.57

Shadow priced, total (5) = (1) + (4) 36.97 61.37 47.32 52.28 92.97 73.76 40.15 36.79 72.76 514.37

Discount factor (6) 1 0.8929 0.7972 0.7118 0.6355 0.5674 0.5066 0.4523 0.4039 -

Source:

Not applicable. PLN data.

Table 8-7B. Distribution Capacity Costs Based on the MV and

(constant billions of rupiahs)

LV Investment

Plan
Not shadow priced, discounted total (7) = (3) x (6) Shadow priced, discounted total (BP) (8) = (5) x (6)

Year
MV

Foreign (1)

Local (2)

Not shadow priced, total + (2) (3) = (1)

Shadow priced, local (4) = (2) x (09)

Shadow priced, total + (4) = (1) (5)

Discount factor (6)

investment plan 8.04 14.76 18.43 19.18 21.33 18.53 21.73 29.82 16.70 168.52 10.04 15.92 18.82 21.30 24.60 21.74 28.14 35.14 17.09 192.27 18.08 30.68 37.25 40.48 45.41 40.27 49.87 64.96 33.79 360.79 9.04 14.33 16.94 19.17 21.67 19.56 25.33 31.63 15.38 173.05 17.08 29.09 35.37 38.35 43.00 38.09 47.06 61.45 32.08 341.57 1 0.8929 0.7972 0.7118 0.6355 0.5674 0.5066 0.4523 0.4039 18.08 27.39 29.69 28.81 28.86 22.85 25.26 29.38 13.65 223.97 17.08 25.97 28.20 27.30 27.33 21.61 23.84 27.79 12.96 212.08

1978-79 1979-80 1980-81 1981-82 1982-83 1983-84 1984-85 1985-86 1986-87 Total

Year
LV

Foreign (1)

Local (2)

Not shadow priced, total (3) = (1) + (2)

Shadow priced, local (4) = (2) x (0.9)

Shadow priced, total (5) = (1) + (4)

Discount factor (6)

Not shadow priced, discounted total (7) = (3) x (6)

Shadow priced, discounted total (BP) (8) = (5) x (6)

investment plan 38.25 52.74 63.87 76.48 91.90 101.73 117.76 142.02 137.19 821.94 19.42 29.76 35.05 40.45 47.33 46.54 58.10 70.46 47.97 395.08 57.67 82.50 98.92 116.93 139.23 148.27 175.86 212.48 185.16 1,217.02 17.48 26.78 31.54 36.40 42.59 41.88 52.29 63.41 43.17 355.54 55.73 79.52 95.41 11288 134.49 143.61 170.05 205.43 180.36 1,177.48 1 0.8929 0.7972 0.7118 0.6355 0.5674 0.5066 0.4523 0.4039 57.67 73.66 78.86 83.23 88.48 84.13 89.09 96.10 74.78 726.00 55.73 71.00 76.06 80.35 85.47 81.48 86.15 92.91 72.85 702.00

1978-79 1979-80 1980-81 1981-82 1982-83 1983-84 1984-85 1985-86 1986-87 Total
-

Not applicable.
PLN data.

Source:

124

CASE STUDIES

Table 8-8. Summary of Incremental Capacity Costs for Java Annuitized Ratio Incremental Incremental cost demand (AIC) cost 9 P) (103 kilowatts) (RplJY/kilowatt)' (RplO3 1kilowattlyear) (Rp1O
35.54 30.53 25.10 139.53 36.80 31.22 26.51 144.30

Item
Shadow priced Generation HV transmission MV distribution LV distribution Not shadow priced Generation HV transmission MV distribution LV distribution

334 212 702

1,394.76 1,077.26 641.49

294.89 239.5 196.9 1,094.0 304.32 244.8 207.9 1,131.4

341.5 224 726

1,394.76 1,077.26 641.49

a. Present discounted value from 1978-79 to 1986-87, calculated from Tables 8-7A, 8-7B, and 8-3. b. Except generation cost, which is calculated in Table 8-6. c. Annuitization factors: generation, 0.1209 (discount rate of 12 percent for 43 years; and HV transmission, MV and LV distribution, 0.1275 (discount rate of 12 percent for 25 years). Source:PLN data.

Table 8-9. Summary of Peak Capacity Costs for Java

(constant billions of rupiahs)

Transmission Generation Supply voltage

Shadow priced HV (70/150 kilovolts) MV (6/12/20 kilovolts) LV (120/380 volts) Not shadow priced HV (70/150 kilovolts) MV (6/12/20 kilovolts) LV (120/380 volts)
EHV/HV

Per kilowatt
293,888 320,139 457,341 304,318 331,501 473,572

Per kilowatt Per kilowatt per year Per kilowatt per year
35,538 38,713 55,304 36,799 40,086 57,266 239,467 260,857 372,652 244,845 266,715 381,021 30,532 33,259 47,513 31,217 34,006 48,580

Not applicable. a. From Table 8-8. b. For generation and EHV/HV transmission: MV cost = peak MV loss factor of 0.082).

HV

cost/0.918, where 0.918

(I -

INDONESIA

125

Peak and off-peak energy costs

The energy cost of mneetinga block of incremental peak consumers will essentially be the cost of running the gas turbines. Even after 1984, when the peaking hydro at Saguling has come on-stream, gas turbines still will be run as the marginal units. Owing to the variations in fuel prices, the cost per kilowatt-hour of generation has been computed using two values for highspeed diesel (HSD) costs, US$15.5 and US$13 a barrel, at international prices. The LRMC for energy during the off-peak period would be the cost of running the marginal base-load steam plants, for example, the MuaraKarang units. The cost for every kilowatt-hour has been computed using two prices for Bunker C fuel corresponding to the HSD prices, that is, US$12 and US$10 a barrel. Typical calculations for both peak and off-peak energy costs at HV, MV, and LV levels are summarized in Tables 8-12 and 8-13.
Customer costs

These costs are available from PLN costs. They include connection, labor, meter, service drop line costs, and so on. The costs of meter reading and billing are already included in the A&G costs.

Primary distribution
Mv

Secondary distribution
LV

Total
Per kilowatt per year Per kilowatt Per kilowatt per year

Per kilowatt

Per kilowatt per year

Per kilowatt

196,869a 281,241
-

25,100 35,858 26,508 37,868

1,094,327a -

139,526' 144,296'
MV

533,355 777,865 2,205,561 549,163 806,123 2,283,343

66,070 97,072 278,201 68,016 100,600 288,010 cost/

207,907' 297,010

1,131,740'

c. For generation, EHV/HV transmission, and 0.7, where 0.7 = 1 - peak loss factor of 0.3. Source: PLN data.

(primary) distribution:

LV coSt = MV

Table 8-10. Calculation of Operation and Maintenance Costsfor Java

(constant billions of rupiahs)

Total O&M costs Incremental cost of O&M (2)

9.88 8.21 4.72 6.74 2.30 6.57 4.13 7.55 8.50 -

Allocation of O&M costs Incremental costs (109 Rp)

11.96 4.78 5.18 17.90 39.87

Year
1977-78 1978-79 1979-80 1980-81 1981-82 1982-83 1983-84 1984-85 1985-86 1986-87 Total

O&M (1)
28.15 38.03 46.24 50.95 57.69 60 66.57 70.70 78.25 86.95 -

Discount O&M increfactor mental discount (3) (4) (2) x (3)

1 0.8929 0.7972 0.7118 0.6355 0.5674 0.5066 0.4523 0.4039 9.87 7.33 3.75 4.8 1.46 3.73 2.09 3.41 3.43 39.88

Item
Generation Transmission Mvdistribution LV distribution Total

Allocation (percent)
30 12 13 45 100

Incremental AIC demand (103 Rp per (megawatts) kilowatt)

1,639.67 1,394.76 1,077.26 641.49
-

7.29 3.43 4.81 27.9

-

Not applicable. Source: PLN data.

Table 8-11. Summary of Operation and Maintenance Costs for Java

(constant billions of rupiahs per kilowatt)

Shadowpriced supply voltage

HV

Transmission Generation
7.870' 8.574 12.248
EHV/HV

Primary distribution
MV

Secondary distribution
LV -

Total
11.302 17.125 52.448

(70/150 kilovolts) for peak/Saguling

sa

MV (6/12/20 kilovolts) for peak/Saguling LV (120/380 volts) for peak/Saguling

3.432 3.739 5.341

4.812 6.875

27.984

- Not applicable. a. Includes an EHV/HV transmission loss factor of 7.25 percent.

Source: PLN data.

128

CASE STUDIES

Table 8-12. Energy Generation Cost Calculation for Table 8-13

Gas turbine (peaking) Fuel use = 0.456 liters per kilowatt-hour Lube oil use = 3.6 x 10 liters per kilowatt-hour at Rpl30.92 per liter, which is a negligible cost 1. Shadow priced (international price for high-speed diesel-HSD)' Fuel price = US$15.5 per barrel 15.5 ($) x 625 (Rp;$) - Rp60.93 per liter 159 (litersibarrel) Fuel cost = Rp6O.93 x 0.456 = Rp27.78 per kilowatt-hour 2. Not shadow priced (local price) Fuel price = Rp43.03 per liter HSD for gas turbine (local price) Fuel cost = Rp43 x 0.456 = Rpl9.62 per kilowatt-hour Steam (boase-load) Fuel use = 0.303 liters per kilowatt-hour 1. Shadow priced (international price for medium fuel oil-bunker Fuel price = US$12 per barrel Fuel cost = 12 x 625 x 0.303 = Rpl4.29 per kilowatt-hour 159

C)'

2. Not shadow priced Fuel cost = 30.93 (Rp per liter) x 0.303 (liters per kilowatt-hour) = Rp9.37 per kilowatt-hour a. Ex-Singapore prices, February 1979.
Source: PLNdata.

Summary The calculations of the strict LRMC for Java are summarized in Table 8-14. The capacity costs have been converted to equivalent kilowatt-hour costs, assuming different consumer coincidence and load factors, in Table 8-15. Ideally, the LRMC for each of the islands outside Java should be computed separately, because the system parameters are widely different in each case. This was not done, however, because the available data on demand, investment, operating costs, and losses were unreliable.

Adjustment of the Strict LRMC

The strict LRMC of supply for the Java system, based on the most recent investment program and appropriately shadow priced, is as follows. The

Table 8-13. Energy and Fuel Cost Type of price

and consumer
Shadow priced
Hv

Based on gas turbine using HSD

Rp27.78 per kilowatt-hour (output from generator) Station use = 2 percent = 4.5 x 1.5 Transmission losses EHV/HV = 8.75-8.8 percent Total = Rp3O.46 per kilowatt-hour 27.78/0.912 = 6 percent x 1.5 = 9 percent MV losses 30.46/0.91 = Rp33.47 per kilowatt-hour
LV

Based on steam plant using bunker C

Rpl4.29 per kilowatt-hour (output from generator) Station use = 4.5 percent = 4.5 x 0.8 Transmission losses EHV/HV = 8.1 percent Total = Rpl5.55 per kilowatt-hour 14.29/0.919 = 6 percent x 0.8 = 4.8 percent losses 15.55/0.952 = Rpl6.33 per kilowatt-hour = 20 percent X 0.8 = 16 percent LV losses 16.33/0.84 = Rpl9.44 per kilowatt-hour
Mv

consumers

Mv consumers

LV

consumers

losses 33.47/0.7

= 20 percent x 1.5 = 30 percent = Rp47.82 per kilowatt-hour Rpl9.62 per kilowatt-hour (output from generator)

Not shadow priced

HV MV LV

Rp9.37 per kilowatt-hour (output from generator) 9.37/0.919 = RplO.20 per kilowatt-hour 10.20/0.952 = RplO.71 per kilowatt-hour 10.71/0.84 = Rpl2.75 per kilowatt-hour percent - 8.8 percent and so = 0.8 x average losses; 0.912 = 100

consumers consumers consumers

19.62/0.912 = Rp2l.51 per kilowatt-hour 21.51/0.91 = Rp23.64 per kilowatt-hour 23.64/0.7 = Rp33.77 per kilowatt-hour

losses; for base load, the losses Note:For peak load, the losses = 1.5 x average on for loss factors. Source: Table 8-12.

Table 8-14. Strict LRMC for Java

Fuel cost (rupiahs per kilowatt-hour)' Kilowatt cost (rupiahs per kilowatt-hour per year) Supply voltage
Shadow priced HV consumers peak MV consumers peak LV consumers peak Not shadow priced HV consumers peak MV consumers peak LV consumers peak

Oil price (A) Facilities

2,875 4,356 13,344 2,875 4,356 13,344

Oil price (B) Base

12.96 13.61 16.20 8.50 8.93 10.63

Capacity cost
66,070 97,072 278,201 68,016 100.600 288,010

O&M

A&G

Total kilowatt cost

82,659 122,205 355,178 85,605 125,733 364,987

Base
15.55 16.33 19.44 10.20 10.71 12.75

Peak
30.46 33.47 47.82 21.51 23.64 33.77

Peak
25.54 28.06 40.08 18.04 19.83 28.32

11,302 17,125 52,448 11,302 17,125 52,448

2,412 3,652 11,185 2,412 3,652 11,185

a. Fuel prices in U.S. dollars per barrel are: shadow priced, base (fuel oil) = 12, and peak (HSD) = 15.5; and not shadow priced, base (fuel oil) 10, and peak (HSD) = 13. Source: PLN data.

INDONESIA

131

peak period costs would be approximately Rp6,200 per kilowatt and Rp3O.5 per kilowatt-hour for HV consumers, and Rp7,640 per kilowatt and Rp33.5 per kilowatt-hour for MV consumers. The off-peak costs would be Rpl5.5 and Rpl6.3 per kilowatt-hour for HV and MV consumers, respectively. Therefore, the best estimates of the average costs per kilowatt-hour of serving Hv and MV consumers are approximately Rp29 and Rp36, respectively. The corresponding LRMC for LV consumers averages nearly RplOO per kilowatt-hour. The proposed new tariff structure contains seventeen customer categories. In calculating from the strict LRMC to the proposed tariff, the main considerations were: a. An average tariff of approximately Rp4O per kilowatt-hour to meet financial requirements b. A substantially subsidized tariff for low-income consumers, accompanied by progressively smaller subsidies for higher income households c. An upper bound of some Rp27 per kilowatt-hour on charges to large HV and MV industrial consumers to be competitive with autogeneration costs. The new average tariff level would be approximately 50 percent higher than the existing mean tariff of Rp27 per kilowatt-hour. Further relief also should be provided to alleviate the high connection cost. This can be done by subsidizing possibly from 50 to 90 percent of the cost and by lenient payment options, perhaps extending for several years. PLN is now reconsidering details of its connection policy, given the effect on its cash flow and general financial situation. Preliminary pricing structure based on the strict LRMC Given a four-hour peak period, from 1745 to 2145 hours (see Figure 8-1), a possible strict LRMC tariff structure is calculated below.
HV CONSUMERS

With time-of-day metering and assuming a CF of 0.9. Peak charges: Capacity cost = (82.659/12) x 0.9 = Rp6,199 per kilowatt per month Energy cost = Rp30.46 per kilowatt-hour Off-peak charges: Energy cost = Rpl5.55 per kilowatt-hour With maximum demand metering. The HV consumers' peak is assumed to fall within the system peak period. Therefore, the charges would be the same as the above figures.

132

CASE STUDIES

Table 8-15. Equivalent Strict LRMC for Java

(rupiahs per kilowatt-hour)

Supply voltage

Capacity cost, per year

Equivalent energy cost with own load factors 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Shadow priced HV consumers peak 82,659 MV consumers peak 122,205 LV consumers peak 355,178 Not shadow priced Hv consumers peak 85.605 Mvconsumers peak 125,733 LV consumers peak 364,987

28.31 34.87 94.6 29.32 35.88 97.22

21.23 26.15 70.95 21.98 26.91 72.91

MV =

16.98 20.92 56.76 17.59 21.53 58.33 0.75, and

14.15 17.43 47.3 14.66 17.94 48.61

LV =

12.13 14.95 40.54 12.56 15.38 41.66

10.61 -

9.43 -

Note: Coincidence factor (cf), HV = 0.9,

Not applicable.

0.7.

a. Equivalent rupiahs per kilowatt-hour = rupiahs per kilowatt per year x cf/8760 x own load factor. Own load factor = Load factor with respect to own individual peak demand. For MV consumers equivalent rupiahs per kilowatt-hour decreased by factor 0.5 because MV demand at time of system peak is one-half the MV own peak demand (for example, their peak falls outside system peak), if necessary: e.q., max demand metering.

MV CONSUMERS

With time-of-day metering and assuming a CF of 0.75. Peak charges: Capacity cost = (122.205/12) x 0.75 = Rp7,638 per kilowatt per month Energy cost = Rp33.47 per kilowatt-hour Off-peak charges: Energy cost = Rpl6.33 per kilowatt-hour Average energy charge = Rp2O.62 per kilowatt-hour (assuming a 25 percent peak and a 75 percent off-peak). With maximum demand metering. The MV consumers' peak is assumed to fall outside the system peak period, and their contribution to the system peak is 0.5 times their own peak. Therefore, peak capacity charges should be half of the above figure. Peak charges: capacity cost = Rp3,819 per kilowatt per month Peak and off-peak energy charges are the same as the above figures.
LV CONSUMERS

Only kilowatt-hour metering.

INDONESIA

133

Fuel cost' Base

15.55 16.33 19.44

Total cost with own load factorc 0.3 0.4

104.58

Peak
30.46 33.47 47.82

Average
19.28 20.62 33.63

0.5
41.54 90.39

0.6
33.43 38.05 80.93

0.7
31.41 35.57 -

0.8
29.89 -

0.9
28.71 -

128.23

Mv,

b. Oil price A from Table 8-14. Average fuel cost: HV, 75 percent base + 25 percent peak; 75 percent base + 25 percent peak; and LV, 50 percent base + 50 percent peak. c. These are only approximate total costs per average kilowatt-hour. Source: PLN data.

Capacity equivalent kilowatt-hour costs are calculated as follows. The LV consumer peak falls within the system peak period. A LF of 0.4 and a CF of 0.7 is assumed. Capacity charge = Rp355.178 per kilowatt per year
= Rp29.598 per kilowatt per month

Capacity equivalent cost per kilowatt-hour p, = (282.829 x 0.7)/(8760 x 0.4)

= Rp56.50 per kilowatt-hour

Energy-based average kilowatt-hour costs are calculated as follows. Consider the simplified load curve for LV consumers shown in Figure 8-3. Peak kilowatt-hours per day = 4(x + y/2) Off-peak kilowatt-hours per day = 20x Total kilowatt-hours per day = 24x + 2y LF = [24x + 2y]/[24(x + y)] = 0.4 (assumed)
Therefore, ylx = 1.895.

Define the ratio: R = peak kilowatt-hours per day/off-peak kilowatt-hours per day
= (4x + 2y)/20x = 0.2 + 0.1 ylx Therefore, R = 0.3895. Suppose energy costs are p, and P2 per kilowatt-hour during peak and off-

134

CASESTUDIES

Figure 8-3. Daily Load Profile for a

+ -

LV

Consumer

Four-l
hour
system I

peak
II I

i
20 24

12

16

Hours of the day

peak periods. Then the average price p per kilowatt-hour is given by p (total) = p, (peak) + P2 (off-peak). p = p, (0.3 8 9 5 /1.3 8 9 5 ) + P2 (1/1.3895) Take p, = 47.82 and P2 = 19.44; then p = Rp27.40 per kilowatt-hour. Total kilowatt-hour cost to be charged: PT = p, + p = Rp83.90 per kilowatt-hour Transition to a practical tariff Generally, the average level of tariffs is constrained. An average price of approximately Rp4O per kilowatt-hour is assumed to be necessary to ensure adequate values for projected financial performance indicators and, therefore, the tariff structure had to have an average electricity price near this value. Other tariff constraints that are more consumer specific will be discussed according to the appropriate tariff category. The proposed new tariff is given in Table 8-16. In general, tariff levels are below the strict LRMC to keep within the bounds for the overall average price per kilowatt-hour. The residential and LV categories receive the greatest reductions as a fraction of the strict LRMC.
LOW-INCOME RESIDENTIAL-LV (Al/R). Government policy requires that this category be heavily subsidized. There is no kilowatt-hour metering; only a load limiter. Suppose the average charge to the consumer is Rpx for each connection monthly, but average consumer cost per kilowatt-hour per month is Rp7O. Average price per kilowatt-hour = Rpx/70 per kilowatt-hour. Full LRMC tariff = RplO5.3 per kilowatt-hour.

Full equivalent charge = 105.3 = x/70. Therefore

x = Rp7,371 per connection per month.

INDONESIA

135

Assume that an Al/R consumer's income is 0.25 times the national average. Therefore, reduce x by a factor 0.25 to obtain the subsidized social tariff (see Chapter 4 for theory). The subsidized tariff is approximately Rpl,850 per connection per month.
MIDDLE- TO HIGH-INCOME RESIDENTIAL-LV (A2/R, A3/R, A4/R). In these categories, kilowatt metering exists, and a separate capacity charge is levied per month as a flat tariff on each connection. Also, there is a separate energy charge. The total price per kilowatt-hour (column 13) lies between the Al/R tariff and the full LRMC, and increases with increasing ability to pay, that is, the difference between price and the strict LRMC decreases as income level increases. To compare the tariff with the strict LRMC, assume that the utilization factor = (contribution to system peak kilowatt)/(installed kilovolt-amp) = 0.3 LRMC capacity cost = Rp29,598 per kilowatt per month = Rp8,880 per kilovolt-amp per month LRMC average energy cost (F) = Rp34.4 per kilowatt-hour VERY-HIGH-INCOME RESIDENTIAL-MV (A4/M). Assume that time-of-day metering exists for measuring peak demand. Both the capacity and energy charges are effectively the full LRMC. The actual tariff of Rp8,000 per peak kilowatt is assumed to be equivalent to Rp5.330 per kilovolt-amp in estimating the projected revenue, SPECIAL-LV AND MV (Bl/R, BUM, B2/R, B2/M). Tariffs are generally less than the strict LRMC, but the deviation is smaller for the MV than for the LV consumers, who are assumed to be less well-off. There is a large difference between capacity charges per kilovolt-amp and per peak kilowatt for MV consumers because their own peak occurs during the day, thus making the contribution to the systern peak rather small. COMMERCIAL AND INDUSTRIAL-LV (Cl/R, C2/R, Dl/R).

Average tariff levels

are comparable with high-income residential users.

INDUSTRIAL-MV AND LV (D2/M, D3/T). The upper bound on average tariffs and energy charges will be competitive with autogeneration. Tariffs can be modified given mnoredetailed information on private generation costs. TEMPORARY AND SPECIAL CONTRACTS (F). Temporary users pay extremely high kilowatt-hour rates. Special contracts will be user specific. The total average tariff level for all consumers = Rp39.7 for every kilowatt-hour sold, which is within the bounds discussed at the beginning of this section.

Table 8-16. Typical New Tariff

Category Kilowatt- Volt-amps Con- Connected hoursconper sumers volt-amps sumed consumer (4) (5) (6) (7)
595,700 83,465,840 41,186,430 140 715 6,034 18,535 400,803

1973 (1)
Al Bi Bl Bl E (RT)

New (2)
AI/R A2/P A31R A4/R A41M

Class (3)
up to 200 VA 250-3700 VA 3.8 kVA 10 135 kVA 630 > 100 kVA

569,068 406,907,918 69,833,926 21,470 129,551,163 15,076,317 4,426 82,038,951 13,713,044 257 103,006,500 13,874,883

D + A2 E (Social)

BI/R BIIM

250 VA 630 > 100 kVA

1,195 15,065 55

23,168,163 40,329,370 21,690,000

8,514,174 6,276,552 2.777,959

3,905 394,364

C2 E (Ktr)

B2/R B21M

250 VA 630 kVA kVA 100

19,340 226

93,658,317 89,844,500

18,907,764 10,907,566

4,841 397,542

B2 B2 E (Kom)

CIIR C21R C2/M

250 VA 1300 1301 VA 630 kVA kVA 100

82,206 54,648,386 8,984,952 34,720 198,600,212 23,087,212 260 106,447,240 15,242,531

665 5,720 409,412

Cl

DI/R

6.6 kVA 630

5,770

146,142,230

9,309,224

25,328

E (Ind)

D2/M

100 kVA 1000

532 232,709,667

40,342,660

437,424

E (Ind)

D3/T

> 5000 kVA -

3 66
-

36,673,333 7,089,900
-

7,812,728 12.224,444 1,314,140 2,215,236 307,163,484 -

F F Special Special contracts contracts Total -

Note: MV and HV consumers who have low power factors (PF) should be penalized by a percentage increase in the tariff for every percentage point their PF falls below an initial value, for example, 0.85. Column definitions: (1) Original 1973 consumer categories. (2) New consumer categories that do not always correspond exactly to the old ones. The main criteria for structuring the new tariff are type of electricity use, voltage level, and connected kilovolt-amps. Therefore the categories A, B, C, D, and F represent residential, special (for example, mosques), commercial, industrial, and temporary consumers, respectively. The letters R, M, and T indicate low, medium, and high voltage supply. (3) The range of connected volt-amps. (4) Total number of consumers, 1977-78.

136

Tariff Per kiloPer kiloAverage Kilowatt- volt-amp watt-hour rate (rupiahs hours per (or peak (p = peak; Revenue per kiloconsumer kilowatt) op = off peak) (Rp x 106) watt-hour) (8) (9) (10) (11) (12) 69 123 702 3,098 53,988 909 50,508 977 48,264 109 665 58,625 1,613 75,832 2,604,242
-

Comments (13) Flat rate Rpl,850 per month 25 percent(p) Rpl8.3 per kilowatt-hour 75 percent(op) (av.) 33 percent(p) Rp9.3 per kilowatt-hour 67 percent(op) (av.) 20 percent(p) RpIl.2 per kilowatt-hour 80 percent(op) (av.) 33 percent (p) RplO.6 per kilowatt-hour 67 percent(op) (av.) 25 percent(p) Rpl6.9 per kilowatt-hour 75 percent(op) (av.) 25 percent(p) Rpll.7 per kilowatt-hour
75 percent (op) (av.)

3,200 3,200 6,400 5,330 (8,000) 3,200 (8,000) 4,670 3,330 (8,130) 4,800 3,330 2,400 3,200
(8,000)

20 20 20 25.3 (p) 16 (op) 31.3 14.7 (p) 6.7 (op) 16 16 (p) 10 (op) 18.7 18.7 16 (p) 8 (op) 26.7 (p) 13.3 (op) 16.7 (p) 10 (op) 10.7 75

1,112 2,699 716 799 808 463 95 740 416 430 1,384 517 500 1,218 196 98

27.0 38.6 47.5 58.3 58.3 31.5 34.3 39.1 38.1 47.9 60.0 34.0 53.7 27.5 25.1 75 39.7

3,070 (5,070) -

12,191

(5) Totalconnectedvolt-amps, 1977-78. (6) Totalkilowatt-hour consumption(monthly). (7) Volt-amps per consumer. (8) Kilowatt-hours per consumer(monthly). (9) Capacity chargeper monthper kilovolt-amp.Thechargeper peakkilowattis givenin parenthesesfor MV and HV consumerswith adequate metering;the equivalent chargeper kilovoltamp is givenonly to estimaterevenue. (10) Energychargeper kilowatt-hour dividedinto peakand off-peakcomponents for MV and
Hv consumers.

(11) Expectedrevenueon the basis of 1977-78consumption and new tariff. (12) Averagerate per kilowatt-hlour. Notapplicable. 137

138

CASE STUDIES

Policy Issues and Conclusions

The tariff study raises several issues. The first and probably the most important is whether or not to continue a uniform national tariff policy. Strong practical considerations would appear to support implementation of a uniform tariff as a first step in tariff revision. In theory, the LRMC of supply should be estimated for each separate island grid, and a corresponding geographically differentiated tariff should be devised. LRMC should be based on a least-cost expansion program, however, and the Java system investment plan is the only one that may be considered acceptable. The planning for additions to capacity and other data for the other islands are far less reliable. Therefore only the LRMC for Java may be calculated with sufficient confidence at the present time. Furthermore, Java constitutes about 80 percent of the total Indonesian power market, and will continue to be the dominant component during the planning period (1979-87). Therefore, given that geographically distinct electricity prices would be unacceptable to the government and that tariff reform is urgent, a uniform national tariff based on the LRMC for Java alone would be a constructive first step. As system planning, least-cost investment policy, and other data on the outer islands are improved, it may be possible to calculate geographically disaggregate LRMCs and introduce these in subsequent tariff revisions. A uniform tariff would suit Java now; the bulk of the power market would tend to somewhat subsidize the outer island consumers because a rough estimate of the islands' LRMC appears to be generally higher than the Java LRMC. Thus, based on current rather unreliable data, to subsidize the outer island costs the average Java consumer would pay approximately 20 to 25 percent higher tariffs. The government justifies this on socioeconomic and equity grounds, because residents of Java are better off than those living on the other islands. The second noteworthy point raised by the tariff calculation is the high strict LRMC for the LV consumers. The incremental costs of distribution appear to be high compared with the incremental consumption. These high distribution costs may be caused by the network rehabilitation costs and voltage conversion that add little new consumption, or simply by overinvestment in distribution facilities. The high level of losses at the LV level also tends to penalize these consumers, although this situation is expected to improve. For comparison, the LRMC was calculated with and without shadow prices. The two figures are quite similar with regard to capacity costs: the nonshadow-priced values were less than 3 percent higher because of the shadow pricing of unskilled labor. In the case of energy costs, however, the shadow-priced LRMC is about 40 percent higher, because of the government

INDONESIA

139

subsidies on domestic fuiel prices relative to world market prices. The shadow-priced values have. been used as the basis for the strict LRMC calculated in this chapter. As discussed in Part One, no perfect tariff exists. Therefore, PLN should recalculate the LRMC regularly and revise tariffs periodically, as often as is practically possible. The tariff levels should at least be raised to keep pace with the general inflation. The LRMC for Java should be refined by improving the data relating to the load forecast, demand characteristics, losses, investments, and operating costs. Similarly, the LRMC' for the other island grids should be computed after improving the data base. Since the load outside of Java is only about 20 percent of the total load for Indonesia, an estimate of average transmission and distribution costs may be made for all the islands outside Java. Generation costs would have to be determined on the basis of capacity and fuel costs of the specific plant types used for each of these smaller grids. O&M, A&G, and other costs would also have to be estimated from improved PLN data. In subsequent tariff calculations, it may also be necessary to specifically examine rural electricity supply costs and pricing policy. This is not a problem now because the number of these customers is negligible.

Chapter9

Pakistan
BETWEEN 1970-71 AND 1976-77, the average annual growth rates of generation, sales, and maximum demand for electricity in Pakistan were 6.9 percent, 5.1 percent, and 10 percent, respectively. By the end of 1976-77, the combined generation and maximum demand of the Water and Power Development Authority of Pakistan (WAPDA) and the Karachi Electric Supply Corporation (KESC) totaled 10,742 gigawatt-hours and 2,003 megawatts, respectively, whereas sales to 2,325,000 consumers equaled 7,150 gigawatt-hours. Thus, the benefits of electricity were available to about 22 percent of Pakistan's total population.

Organization of the Sector

WAPDA is the principal partner in the electric power sector, accounting for about 80 percent of the public electric power supply in Pakistan. By the end of 1977-78, WAPDA had an installed capacity of 2,635 megawatts and served 2.26 million customers. KESC serves the Karachi metropolitan area and accounts for the remaining 20 percent of the public electric power supply in Pakistan. At the end of 1977-78, KESC supplied 310,000 consumers with an installed capacity of 645 megawatts. Finally, several other companies distribute electricity, which they purchase in bulk from WAPDA. Such bulk purchases account for about 13 percent of WAPDA'S sales. The largest distribution companies are the Multan Electric Supply Company and the Rawalpindi Electric Power Company. The Pakistan power system is shown in Map 2. The average annual load factor varied from 69.7 percent in 1970-71 to 61.2 percent in 1976-77. The relatively high load factors in the early 1970s are attributable to generation capacity shortages and resulting load shedding during periods of peak demand. System losses (as a percentage of generation) increased from 26.4 percent in 1970-71 to 33.4 percent in 1976-77.

Note: This chapter was prepared in collaboration with Muhammed Jahangir and Iqbal Khan (wAPDA). The views expressed in this chapter do not necessarily reflect those of WAPDA or the government of Pakistan.

140

PAKISTAN

141

Table 9-1. Maximum Demand, Generation, and Sales, 1970-71 to 1976-77

Maximum demand Growth Megarate watts (percent)
1,188 1,288 1,424 1,540 1,718 1,762 2,003 8.4 10.6 8.3 11.8 2.4 13.7 10 combined.

Generation Gigawatthours
7,203 7,561 8,471 9,055 10,003 10,198 10,742 -

Sales Losses as percentage of generation

26.4 28.2 29.2 29.9 31.7 31.9 33.4
-

Year
1970-71 1971-72 1972-73 1973-74 1974-75 1975-76 1976-77 Average Note:
-

Growth Giga- Growth rate wattrate (percent) hours (percent)

5.0 12.0 6.9 10.5 1.9 5.3 6.9 5,300 5,430 6,000 6,350 6,830 6,940 7,150 2.5 10.5 5.8 7.6 1.6 3.0 5.1

WAPDA and KESC

Not applicable.

a. Maximum demand was constrained by shortages of generation capacity during these years. Source:WAPDA data.

This increase was highly correlated with the continued extension of transmission and distribution lines into the rural areas of Pakistan during this period. Table 9-1 summarizes the annual development of the total power market for 1970-71 to 1976-77.
Forecast growth of demand and energy

Overall, generation and maximum demand in Pakistan are forecast to grow at about 9 percent ainually between 1977-78 and 1982-83. Although these forecasts may be adequate for short-term planning, the current economic situation and the uncertainty underlying the forecast make it essential that a longer term, more detailed forecast be prepared. Currently, load forecasting within WAPDA is the responsibility of the Power Market Survey Division. A survey of prospective consumers is made, and relevant data from the public and private sectors on proposed agricultural and industrial development are reviewed. A forecast on this basis, without correlation with historical and forecasted economic development may result in an imprecise load projection. The forecast in this report, which is based on WAPDA projections and other recent information, assumes that sales will grow about 11.9 percent a year for the next five years. Allowing for a reduction in system losses, the annual growth of generation would be about 9.1 percent. These growth rates are based on extrapolations of histori-

142

CASE STUDIES

io

70

~~~~~~~U.S.S.R. _
;CINA

MAP2.

Pakistan: Principal Power Stations

and TransmissionLines
in the WAPDA System

4
;C~~

A
JAMMU

I.
~7
g
.

?.
34~~ Warsak~~~~ -34-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4 War" akSLAM

/Malakand

-.

...

AND 4Appra~nr
Li fCa

or.r>

Daga Tarbel

..,.~~~r.
SAMABAD

KASHMIR

DEMOCRATICREPUBLIC
OF AFGHANISTAN

It

angla

,, ,

A _

NORTH
h

iShah
h

Kak*haaa
d

al~~~~~~~~Sahd

30

uetta

30

1'-.' i
CENTRA L . Guddi
Shikrpur

4<;

~~INDIA

\.
EXISTING FUTURE

.4 W r. f --.- 26*

500 KV Lines 220 KV Lines 66/132KVtLines t^_ -' ! _ 2Hydroelectric

SOUTH -O
SOUTH

\'
.+

----

0m
| * l l l >

Hydefa.had

~~~~~~~~Nuclear Plant Substations

Power Dtstrict Boundaries International

Thermal Stations

#**,t*

~~~~~~~~Boundaries

22'j

.74'

PAKISTAN

143

cal trends, excluding 1975-76 and 1976-77, when the rates were affected by economic stagnation and political unrest. A revised load forecast should include the results of broader energy studies undertaken to reduce the uncertainties underlying the present forecast. This will enable the investment program for generation plants scheduled to be commissioned during the sixth plan period to be firmed up, indicating necessary investments during the fifth plan. Table 9-2 summarizes the expected growth of WAPDA'S system. One might question whether an assumed 12 percent annual growth of WAPDA sales is realistic in, view of an average historical growth rate of 8 percent for 1970-71 to 1977-78. The years 1975-76 and 1976-77, however, were periods of industrial stagnation and political difficulties, and the supply situation was compounded by WAPDA'S financial and technical constraints. The technical constraints were to a large degree overcome by the recent commissioning of the first four Tarbela units, and the financial constraints will be reduced by funds from the World Bank's Third Power Project. This is designed not only to rehabilitate the existing secondary transmission system, but also to expand this system to new areas to meet potential demand. The growth rate of 19 percent in 1977-78 indicates this latent demand and suggests that the assumption of a 12 percent annual growth of sales during 1978-79 to 1982-83 is likely to prove conservative.
Future development

The sector expansion program for the fifth plan (1978-79 to 1982-83) has been adjusted to confomi to the above load forecast and assumes an effec-

Table 9-2. Forecast of Maximum Demand, Generation, and Sales,

1977-78 to 1982-83 Maximum demand Growth Megarate watts (percent)
1,972 -

Generation Gigawatthours
10,576

Sales Growth rate (percent)

-

Year
1979

Growth Gigarate watt(percent) hours

6,980

Losses as percentage of generation

34

1980 1981 1982 1983 Average

Note:
-

1,149 1,342 1,553 1,783 only.

9.0 9.0 9.0 9.0 10

11,576 12,617 13,754 14,993 -

9.5 9.0 9.0 9.0 6.9

7,756 8,706 9,765 10,945 -

11.1 12.2 12.7 12.1 5.1

33 31 29 27
-

WAPDA

Not applicable.
data.

Source:WAPDA

144

CASESTUDIES

Table 9-3.

WAPDA

Generation Expansion Program

Installed capacity (megawatts) 50 50 25 210 200 80 350 350
700'

Year 1978-79 1980-81 1980-81 1980-81 1980-81 1980-81 1981-82 1982-83

1983-84-85

Project Kotri, gas turbines Kotri, gas turbines Quetta, gas turbines Guddu thermal Manglahydro, units 7 and 8 Warsackhydro, units S and 6 Tarbelahydro, units 5 and 6 Tarbelahydro, units 7 and 8
Tarbela hydro, units 9 to 12

a. Size of units is likely to be increased. Source: WAPDA data.

tive loss reduction program. ' The generation expansion program includes hydro, 1,480 megawatts (WAPDA); thermal, 410 megawatts (WAPDA, 210 megawatts; KESC, 200 megawatts); and gas turbines, 350 megawatts (WAPDA, 125 megawatts; KESC, 225 megawatts) for a total of 2,240 megawatts, increasing installed capacity to 5,520 megawatts. Reserve capacity during low water months will vary between 81 megawatts in March 1979 to 353 megawatts in March 1980. The reserve capacity in subsequent years is adequate and reasonable. In addition private generation capacity is expected to increase from 230 megawatts to about 400 megawatts. The WAPDA generation expansion program up to 1985 is shown in Table 9-3. Expansion and rehabilitation of transmission and distribution (T&D) systems is vital to achieve the objectives set by the government. Studies by Meta Systems of Canada of EHV systems have shown the need for a 500-kilovolt transmission system. This system should be built from Tarbela to the Hyderabad/Karachi area to fully integrate the hydroelectric sources in the north with the thermal plants and market in the center and south. Such an interconnection would allow the hydro capability to be used fully during the high water months (July to September) and would give thermal support to the northern area from the center and the south during low water months (March to May). This interconnection, which constitutes a "bulk power expressway," would also reduce fuel expenses (for spinning reserves) and generating capacity requirements by pooling the reserves of WAPDA and
KESC.

1. Modifications include postponing beyond 1982-83 some of the proposed thermal developments and dropping some of the proposed gas turbine plants from the program.

PAKISTAN

145

During 1978-79 to 1982-83, two 500-kilovolt lines will be completed from Tarbela to Faisalabad. The first is in operation but is energized at 220 kilovolts, and the second will be operational in 1982-83. Financing is being provided by the Canadian International Development Agency. One of these lines will be extended from Faisalabad to Hyderabad/Karachi, with intermediate substations at MultaLnand Guddu and is to be completed in 1982-83. Again, initial operation is envisaged at 220 kilovolts. Financing of this line is being provided by the World Bank and by the Kuwait Fund. The entire 500-kilovolt interconnection will be converted to a 500-kilovolt operation when the last Tarbela unit, Unit 12, is commissioned in November 1983. In addition, a substantial program to extend, convert to higher voltage, and reinforce the 220-kilovolt, 132-kilovolt, and 66-kilovolt system is required. These secondary transmission systems are essential to improve voltage conditions, reduce losses, and supply new loads. Distribution systems also will have to be reinforced and expanded to connect about one million new consumers during the fifth plan. By 1982-83, this will bring the benefits of electricity to about 27 percent of the population, compared with 22 percent in 1976-77. Computation of the Strict LRMC

The strict long-run marginal cost (LRMC) may be defined broadly as the incremental cost of optimnaladjustments in the system expansion plan and system operations attributable to an incremental increase in demand that is sustained into the future. The three broad categories of marginal costs are: capacity costs, energy costs, and consumer costs. By examining the system load duration curves (LDC), it is possible to determine periods during which demand presses on capacity, for example, at a particular time of day or in a given season of the year. The WAPDA system does not exhibit marked seasonality of demand. Variations in hydro capability are also not significant to justify considering seasonal changes in price. Therefore, only two pricing periods by time of day are selected-peak and off-peak. Other aspects of structuring will be introduced later in the analysis.
Marginal capacity costs

The WAPDA grid system uses a combination of hydroelectric units, gasfired thermal plants, and gas turbines to generate electricity. The system is basically thermal dominated, with hydroelectric generation accounting for only about 20 to 25 percent of total energy output during the recent past. This situation is likely to continue at least for the next few years. The order of operation according to merit (that is, operating cost) is such that peak demand, which typically consists of a four-hour period between 1600 and 2000

146

CASESTUDIES

Figure 9-1. Daily Load Curve for December 20, 1978

2,000 -

t 1,500
E

o
Source:
WAPDA

4
data.

12
Hoursof the day

16

20

24

(see Figure 9-1), is met by operating both gas turbines and gas-fired thermal units. The annual LDC for the WAPDA system in 1978 is represented in Figure 9-2. It shows that the four-hour peak demand is met primarily by operating a mix of gas turbines and steam units. Specifically, calculation based on Figure 9-1 indicates that approximately 55 percent of peak period demand is met by gas-fired steam plants like Guddu; the balance is met by the gas turlines. This mix of peaking plants and pattern of operation are expected to continue for the next few years; and, therefore, generation capacity costs are calcu7ated on the basis of these two types of plants.2 The capacity costs for a representative gas-fired steam plant and a gas turbine are summarized in Table 9-4. The weighted capacity cost per kilowatt installed in 1979 border prices (BP) is Rs8,875. This measures the LRMC of meeting a kilowatt increment in peak demand at the busbars, including the reserve margin. The percentage of gas turbine (G1) capacity used for peaking and the reserve margin may be higher than 45 percent by 1983. For example, see the generation expansion program described earlier. The effect of these future changes in the peaking plant mix on the present value of costs will be reduced by the effect of discounting, however. Furthermore, even if the relative weights assigned to steam and gas turbine capacity costs are changed somewhat in Table 9-4, the relative effect on the LRMC of capacity will be much smaller. This is because generation capacity costs are only
2. For further details of system operation, see Mohan Munasinghe, "Pakistan: WAPDA Power Tariff Study," Energy Department Report, World Bank. Washington, D.C., February 1980.

PAKISTAN

147

Figure 9-2. Annual Load Duration Curve for the WAPDA Grid System (Northern Upper Sind and Quetta Grid Systems), 1979
100

Ga

80
Guddu

Lyallpur\ and Multan 60 Sukkur

jX
&o
o_

Mangla and Tarbela Warsak

_ _ _ ___ _ _ _ _

a. 40 -

fMangla and Tarbeta

Mangla and Tarbela 20 -

WarsakHvdels Guddu Multan 0 1,000 2,000 3,000 4,000 5,000 6,000 Lyallpur 7,000 8,000 8,760

Source: WAPDA data.

Hours

148

CASESTUDIES

Table 9-4. Costs of Generation Capacity at the Busbar

(rupees per kilowatt-hour installed, 1979 prices)

Unit
Gas-fired steam Gas turbine Weighted total
-

Capacity cost in market prices,

16,100 5,610 -

Capacity cost, BP'

12,558 4,376 -

Weighted capacity cost, BPC

6,905 1,970 8,875

Not applicable. a. Capacity costs are adjusted to account for a 30 percent reserve margin and 2.1 percent station use of energy. b. This assumes a 50 percent foreign exchange component and a 50 percent local cost component of capacity costs. Conversion factors for the foreign exchange component (which includes 40 percent import duty) and the local component (which consists of local materials and skilled labor) are 0.71 and 0.85, respectively. c. The relative weights are 0.55 for steam and 0.45 for gas turbine.

Source: WAPDA data.

about 45 percent of total capacity costs, including T&D (as shown in Table 9-12). For example, taking weights of 40 percent for steam plant and 60 percent for gas turbine, the generation capacity cost is Rs7,670 per kilowatt (BP), or Rsl,150 per kilowatt per year (annuitized DP). This would decrease the LRMC of capacity of the LV level only by about 6 percent. Next, the LRMC of T&D are calculated. Generally, all investment costs for T&D are allocated to incremental capacity, since the designs of these facilities are determined principally by the peak kilowatts that they carry rather than by the kilowatt-hour. Structuring the LRMC by voltage level may be introduced here by considering three supply voltage categories: HV, MV, and LV. Consumers at each voltage level are charged only upstream costs. Therefore, capacity costs at each supply voltage must be identified. The planned investment program for EHV/HV transmission capacity (66/ 132 kilovolt) for 1980-90 is summarized in columns (1) to (3) of Table 9-5, and projected peak HV demand is derived in Table 9-6 and in columns (1) to (3) of Table 9-7. Using the results presented in the last two tables and the average incremental cost approach, the incremental EHV/HV capacity cost can be calculated as Rs6,471 x 106 for 1,079 megawatts = Rs5 ,997 per kilowatt. The planned investment program for MV distribution capacity (11/33 kilovolt) is summarized in columns (4) to (6) of Table 9-5, and projected peak megawatt demand is derived in Table 9-6 and in columns (4) to (6) of Table 9-7. Based on these estimates, the incremental MV capacity cost is Rsl,451 x 106for 982 megawatts = Rsl,478 per kilowatt. Similarly, the planned investment program for LV distribution capacity (220/400 volts) is summarized in columns (7) and (8) of Table 9-5, and forecast peak LV demand is derived in Table 9-6 and in columns (7) and (8) of Table 9-7. Therefore, the incremental LV capacity cost is Rs(2,175 x 106/679) per megawatt = Rs3,203 per kilowatt.

Table 9-5.

WAPDA Investment (millions of 1979 rupees)

EHV/HV

Program, 1980 to 1990

Mv

transmission

distribution Discounted cost, BP (6)

163 184 172 173 158 144 130 119 108 100 89
1,451

Lv

distribution Discounted cost, BP (9)

243 277 258 261 236 215 195 179 162 149 134
2,175

Year
1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990
Total, 1980-88 Total, 1980-89
-

Cost, DP (1)
1,224 1,226 1,587 1,700 1,632 1,672 1,698 1,729 1,847 1,740 1,669
-

Cost, BP, (2)

967 970 1,279 1,344 1,289 1,322 1,343 1,367 1,461 1,375 1,319
-

Discounted cost, BP' (3)

863 774 854 254 731 670 607 552 527 443 379
6,471
-

Cost, DP (4)
215 274 286 323 330 335 341 349 354 366 368
_

Cost, BPC (5)

182 231 242 273 278 284 288 294 299 310 311
-

Cost, DP (7)
322 412 428 485 493 503 511 522 531 549 552
-

Cost, BPc (8)

272 348 362 410 416 425 431 442 449 464 467
-

Not applicable. a. This assumes a 40 percent foreign exchange component, a 55 percent share for local materials and skilled labor, and a 5 percent share for unskilled labor. Conversion factors for the foreign exchange component (which includes a 40 percent import duty), local materials, and unskilled labor are RsO.71, RsO.85, and RsO.75, respectively. b. The assumed discount rate is 12 percent. c. The same as a., except that there is assumed to be no foreign exchange component of distribution investment. Source: WAPDA data.

150

CASE STUDIES

Table 9-6. Projected System Peak Demand at 1979 to 1990

(megawatts)

EHV/HV,

MV,

and

LV,

HV EHV/HV

Year
1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990

Generation peak (1)

1,972 2,149 2,342 2,553 2,783 3,033 3,306 3,538 3,821 4,127 4,457 4,813

Station use (2a)

39 54 54 56 56 61 66 71 76 83 89 96

transmission Peak loss (3) = Consumption (2b) (1)-(2a)-(2b) (4)

316 301 304 306 278 243 264 283 306 330 357 385 1,617 1,794 1,984 2,191 2,449 2,729 2,976 3,184 3,439 3,714 4,011 4,332 221 246 276 310 347 392 413 442 478 516 557 602

Note: The power loss factors used here measure losses as a percentage of incoming power during periods of peak demand as summarized in Table 9-9. The columns (1) to (9) correspond to power flows measured at the same points in Figure 3-2. Source: WAPDA data.

Marginal energy costs Peak and off-peak energy costs equal the cost of running the units used during each of these periods to supply incremental amounts of energy. Typically, this will be the running cost of the most inefficient plant used to meet demand during each period. As discussed previously (see Figure 9-2), increments in peak demand are met by increased use of a mix of gas-fired thermal units (Lyallpur, Multan, and Guddu) and gas turbines. The relative importance of each of these units in meeting peak demand can be determined from the annual LDC in Figure 9-2. Using this information-together with the heat rates of the respective plants, the BTU'S per cubic meter of gas, and the economic cost of producing gas-peak period energy costs at the busbars can be determined. Off-peak energy costs can be determined in the same fashion. In this case, however, it is assumed that the Guddu unit, or a similar gas-fired thermal unit, supplies incremental amounts of off-peak energy to supplement hydroelectric energy. An upper bound on the marginal opportunity cost (MOC) or economic value of natural gas would be the value of imported liquid fuels it displaces.

PAKISTAN

151

MV

LV

Loss (5)
145 161 162 173 183 203 223 239 258 278 300 325

Peak (6) = (3)-(4)-(5)

1,251 1,387 1,546 1,708 1,919 2,134 2,340 2,503 2,703 2,920 3,154 3,405

Consumption (7)
177 197 221 248 278 313 331 354 382 413 446 481

Loss (8)
317 351 370 387 415 441 486 520 561 607 655 707

Consumption (9) = (6)-(7)-(8)

757 839 955 1,073 1,226 1,380 1,523 1,629 1,760 1,900 2,053 2,217

Again, for lack of informrration on alternative high-value uses of gas, the cost of high-speed diesel (HSD) fuel, valued at Rs2.483 per liter and having a beat value of 35,500 BTLU per liter, could be used. Peak and off-peak energy costs in rupees per kilowatt-hour, based on the above values, are summarized in Table 9-8. To determine capacity and energy costs at various voltage levels, line losses must be considered. Projected loss factors for generation, EHV/HV transmission, MV distribution, and LV distribution are summarized in Table 9-9.
Customer costs

Customer costs and charges are summarized in Table 9-10.

Summary

Now the long-run marginal capacity and energy costs can be determined. Specifically, the capacity costs calculated earlier are converted back to DP and are annuitized in Table 9-11.

Table 9-7. Incremental Peak Demand at

(megawatts)
EHV/HV

EHV/HV, MV,

and

LV,

1979 to 1990
MV

LV

Year
1979

Discounted Discounted Discounted Peak Changein changein Peak Changein changein Peak Changein changein demand' peak demand peak demandb demandc peak demand peak demand demandd peak demand peak demand (1) (2) (3) (4) (5) (6) (7) (8) (9)
1,617 1,251 757 -

1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 Total,
1981-90

1,794 1,984 2,191 2,449 2,729 2,976 3,184 3,439 3,714 4,011 4,332
-

177 190 207 258 280 247 208 255 275 297 321
-

158 151 147 164 159 125 94 103 99 96 92

-

1,387 1,546 1,708 1,919 2,134 2,340 2,503 2,703 2,920 3,154 3,405
-

136 159 162 211 215 206 163 200 217 234 251
-

121 127 115 134 122 104 74 81 78 75 72

982

839 955 1,073 1,226 1,380 1,523 1,629 1,760 1,900 2,053 2,217
-

82 116 118 153 154 143 106 131 140 153 164
-

73 92 84 97 87 72 48 53 50 49 47
679

Total,
1982-90
-

1,079

Notapplicable. a. Column (1) is taken from column (3) in Table 9-6. b. The assumed discount rate is 12 percent.

c. Column (4) is taken from column (6) in Table 9-6. d. Column (7) is taken from column (9) in Table 9-6. Source: Derived from Table 9-6.

Table 9-8. Marginal Energy Costs

(rupees per kilowatt-hour)

At market price of gas Voltage level

Generatorsd HV (66/132 kilovolts) MV (11/33 kilovolts) LV (400 kilovolts)

At

HSD

equivalent price Off-peakb

0.8908 0.9578 1.0300 1.3037

Best estimate' Peak,

0.266 0.292 0.321 0.428

Peak,
0.0862 0.0947 0.1052 0.1403

Off-peakb
0.0420 0.0452 0.0486 0.0615

Peak'
0.9286 1.0204 1.1214 1.4951

Off-peakb
0.255 0.275 0.296 0.375

a. Relative weights of Guddu, Lyallpur/Multan, and gas turbines are 0.354, 0.531, 0.115, respectively, and corresponding heat rates are 12,488, 11,815, and 20,084 BTu/kilowatt-hour. b. Relative weight of Guddu is 1.0. c. Gas valued at Rs2O per 106BTU, based on value in alternative uses (data from Pakistan Oil and Gas Sector Memo of October 5, 1978, EWT Dept., World Bank, Washington, D.C.). d. Including station use of 2.1 percent.
Source:
WAPDA data.

154

CASESTUDIES

Table 9-9.
(percent)

WAPDA

System Loss Factors, 1979 to 1990

Voltage level
Generation Generation Incoming (average energy) Incoming (peak power)
EHV/HV

1979
2.0 2.0 2.0

1980
2.5 2.5 2.5

1981
2.3 2.3 2.3

1982
2.2 2.2 2.2

1983
2.0 2.0 2.0

1984
2.0 2.0 2.0

Generation Incoming (average energy) Incoming (peak power)

MV

12.0 12.0 16.0

10.5 10.5 14.0

10.0 10.0 13.0

9.0 9.0 12.0

8.0 8.0 10.0

6.5 6.5 8.0

Generation Incoming (average energy) Incoming (peak power)

LV

6.0 8.0 10.4

6.0 8.0 10.4

5.5 7.3 9.5

5.3 7.1 9.2

5.0 6.7 8.7

5.0 6.7 8.7

Generation Incoming (average energy) Incoming (peak power)

Not applicable. Source: WAPDA data.

14.0 25.4 29.5

14.0 25.4 29.5

13.2 24.0 27.9

12.5 22.7 26.5

12.0 21.8 25.3

11.5 20.9 24.2

PAKISTAN

155

1985

1986

1987

1988

1989

1990

Average, 1979-90

2.0 2.0 2.0

2.0 2.0 2.0

2.0 2.0 2.0

2.0 2.0 2.0

2.0 2.0 2.0

2.0 2.0 2.0

2.1 2.1

6.5

6.5

6.5

6.5

6.5

6.5

6.5
8.0

6.5
8.0

6.5
8.0

6.5
8.0

6.5
8.0

6.5
8.0

7.0
9.0

5.0

5.0

5.0

5.0

5.0

5.0

6.7
8.7

6.7
8.7

6.7
8.7

6.7
8.7

6.7
8.7

6.7
8.7

7.0
9.0

11.5 20.9 24.2

11.5 20.9 24.2

11.5 20.9 24.2

11.5 20.9 24.2

11.5 20.9 24.2

11.5 20.9 24.2

21.0 25.0

156

CASE STUDIES

Table 9-10. Customer Costs

(1979 rupees)

Average cost per connection Category to WAPDA

Domestic and commercial 334
WAPDA

Customer contribution
bears full costs for up to 100 feet of secondary lines, and customer pays excess, if any, as a lump sum or monthly service rent. Customers who supply their own meters are given priority among waiting list of about 17,000. LV customers pay line costs only. MV and HV customers pay all costs: line, transformer, substation, and so forth. WAPDA bears full costs up to Rsl5.000. If transformer is required, the cost will exceed Rsl5,000, and customer must pay the excess.

Industrial Agricultural

9,200 8,500

Source: WAPDAdata.

Table 9-11. Capacity Costs

(1979 rupees)

Voltage level
Generation HV transmission MV distribution LV distribution

Capacity cost per kilowatt, BP

8,875 5,997 1,478 3,203

Annuitized Capacity cost capacity costs per kilowatt, DPa per kilowatt per year'
10,440 7,055 1,739 3,768 1,330 856 222 480

a. The CF is 0.85. b. The assumed lifetimes for generation, HV, MV, and LV capacities are 25, 40, 25, and 25 years, respectively. With an assumed opportunity cost of capital of 12 percent, the resulting annuitizing factors are 0.1275, 0.1213, 0.1275, and 0.1275 for the respective capacity investments. Source: WAPDA data.

Finally, the results from Table 9-11 are combined in Table 9-12 with estimates of line loss and expenditures for operation and maintenance (O&M) and administration and general (A&G) to determine strict long-run marginal capacity costs at various voltage levels. The underlined figures indicate the LRMC of peak period capacity in rupees per kilowatt per year at each voltage level. The corresponding values of peak and off-peak LRMC of energy in rupees per kilowatt-hour have already been indicated in Table 9-8.

Table 9-12. Summary of Strict Long-Run Marginal Capacity and Energy Cost (DP)
(rupees)

Capacity cost (per kilowatt per year) Generation

Voltage Annual O&M level, capacity + A&G Generation
'-

EHV/HV

MV

LV

Total' 1,486 1,633

-

Annual O&M Cumulative Annual O&M Cumulative Annual u&M Cumulative capacity + A&G total capacity + A&G total capacity + A&G total 856
-

Energycost (per kilowan hour) Peak Off-peak 0.266 0.255 0.275 0.296 0.374

1,330
-

156
-

EHV/HV

2.12
-

2,701
2,968
-

0.292
122
-

'

Mv
LV

222
-

3,312 4,416

480

264

5,160

0.321 0.428

Note: Underlined figures indicate the total LRMC of peak period capacity at each voltage level. Not applicable. a. The losses of incoming peak power, as presented in Table 9-9, equal 9 percent for EHV/Hv and Mv, and 25 percent for LV. Station use equals 2.1 percent. Average losses of incoming equal 7 percent for EHV/HV and MV, and 21 percent for LV. b. O&M plus A&G costs equal 1.5 percent, 3 percent, 5 percent, and 7 percent of capacity costs per kilowatt for generation, EHv/Hv, MV, and LV capacities, respectively. Source: Derived from previous tables.

158

CASESTUDIES

Adjustment of the Strict LRMC

The methodsused for adjustingstrictLRMC are describedin this section. Comparisonof existingtariffs with the LRMC The existing WAPDA tariffs and LRMC for large categories of consumers are summarizedin Table 9-13. The tariff structureis complex,with various energyand minimumcharges, which are generallynot directlyrelatedto the costsof supplyingelectricityto consumers.For example,the ratio of energy to capacity charges is highly distorted comparedwith the LRMC. Thus, in general,the kilowattchargesare smallerthan the LRMC by an order of magnitude. Also, existing tariffs are lowest for domesticconsumersand tubewells. Yet, accordingto the LRMC estimatesin the table, the costs of supplying both types of consumers are relatively high. These and other considerations suggest that the structure of electricity demand might be significantlyinfluencedif these tariffs were made more consistentwith the LRMC. Proposednew tariff structure The new tariff is summarizedin Table 9-14. The principalchangesin the tariff categoriesare: a. An additionalblock of consumption(more than 200 kilowatt-hours) has been added to the domesticcategory, representing higher income households b. In the bulk supply categories C.1 and C.2, the number of subcategories has been reduced from three to two by consolidating (b) and (c) (see Table 9-13) c. In the bulk supply category C.3, only a single subcategorywith the demand charge is retained. There are three broad considerations that underliethe adjustments madeto the strictLRMC. First, althoughmore ambitiousfinancialtargetsmay be required in later years, the target averageprice for 1980-81 basedon WAPDA'S financialrequirements (as agreedunder the WorldBank's Third PowerProject loan) is RsO.482per kilowatt-hoursold. This value is 10 percent higher than the 1979-80 target of RsO.438per kilowatt-hour,which was itself 25 percentabove the averageactual revenueof RsO.3503 per kilowatt-hour sold in 1978-79. Second, the tariff for low-incomeresidentialand private tubewells is most sensitive, and any increasesmust be kept to a minimum.And third, the principalchanges in tariff structureshould aim to increasethe low existingcapacity chargesnearer to the LRMC. Increasesin tariff level, however, should not raise the averageprice too much higher than the RsO.482

PAKISTAN 159

per kilowatt-hour indicated above. Detailed considerations relevant to particular customer categories follow.
DOMESTIC (A. 1). LOW-INCOME CONSUMPTIONBLOCK.

Appliance
Lights Fan Total

Watts
120 80

Summer Kilowatt-hours Hours per month

120 300 14.4 24.0 38.4

Winter Kilowatt-hours Hours per month

180 0 21.6 0 21.6

Therefore, the first block of consumption is set at 40 kilowatt-hours per month. The price of the first block is to be maintained at its present level of RsO.31 per kilowatt-hour. The mean income of low-income households is about Rs410 per month, and electricity payments would be Rs5 to RslO per month.
DOMESTIC

(A. 1).

MEDIlM-INCOME

CONSUMPTION BLOCK.

Appliance

Watts

Summer Kilowatt-hours per month Hours

150 600 0 30 180 45.0 96.0 0 18.0 27.0 186.0

Winter Kilowatt-hours Hours per month

180 0 90 30 180 54.0 0 90.0 18.0 27.0 189.0

Lights 120 Fan 80 Heaters 1000 Iron 600 Refrigerator 150 Total

Therefore, the second block of consumption is set at 200 kilowatt-hours per month, at a price of RsO.45 per kilowatt-hour, or about 30 percent higher than the present level.
DOMESTIC (A. 1). HIGH-INCOME CONSUMPTION BLOCK. The tariff for electricity use in excess of 200 kilowatt-hours per month is set at RsO.75 per kilowatt-hour or about 115 percent higher than the existing level to discourage excessive consumption. COMMERCIAL

(A.2).

A SMALL SHOP.

Appliance
Lights Fan (surmmer) Total

Watts
200 80

Hours
360 360

Kilowatt-hours per month

72.0 28.8 100.0

Table 9-13. Summary of Main Features of the Existing WAPDA Tariffs

Existing tariff Energy charge (paisa per kilowatt-hour)'
31 first 40 kilowatt-hours 35 balance (Rs5 per month minimum) 85 first 100 kilowatt-hours 95 balance (Rsl5 per month minimum) 52 36 35 32 20.00-25.00 (minimum charge) 50.00 45.00 42.00 {321(p)
29 2

LRMC tariff" Energy charge, (paisa per kilowatt-hour) Capacity charge (rupees per kilowatt per month)

c,+ Consumer category

A. 1. Domestic

Capacity charge (rupees per kilowatt per month)

A.2. Commercial

42.8(p) 37.5(op)

430

B. Industrial B. 1. 70 kilowatts connected load B.2. Up to 500 kilowatts connected load B.3. Supply at 11 kilovolts B.4. Supply at 33, 66, or 132 kilovolts

276 225

. (p) 127.5 (op) (66/132 kilovolt)

C. Bulk supply C. 1. Supply at 400 volts a. d b. d40 C. C.2. Supply at 11 kilovolts a.d2 b.3

32 430 35 45.00 37.5(op)

41.003 34410

32.1(p)

27 5(op)
2

276

C.3. Supply at 33, 66, and 132 kilovolts a. b. D. Tubewells D. I. For reclamation and drainage D.2. For agriculture and irrigation a. Northwest frontier b. Other areas 34 33 31.0 39.50

9.2(p)25 7.5(op) (661132 kilovolt)

2

225

42.8(p) 13.0 19.0 9 11

430

Note: As of July 1, 1979. a. A fuel adjustment surcharge is added to the energy charges listed in this table. b. No adjustment is made for diversity factor. c. (p) = peak; (op) = off-peak. d. Includes licensees and nonlicensees. e. Includes railways, cantonment boards, and other institutions with their own distribution facilities. Source: WAPDA data.

162

CASE STUDIES

Table 9-14. LRMC and Proposed New

WAPDA

Power Tarifffor

1980-81

LRMC'

Consumer category
A. l. a. b. c. Domestic 30 kilowatt-hours 30-200 kilowatt-hours 200 kilowatt-hours

Capacity (rupees per kilowatt per month)

Energyb (rupees per kilowatt-hour)

Average effective cost' (rupees per kilowatt-hour)

A.2. Commercial a. <100 kilowatt-hours b. >100 kilowatt-hours B. Industrial B.l. 70 kilowatt-hours B.2. 70-500 kilowatt-hours B.3. 11/33 kilovolts B.4. 66/132 kilovolts C. Bulk supply C.l. 400 volts a.
b.

43 430

88 0.43-0.38

to

276 225 430

0.32-0.30 0.29-0.28 0.43-0.38

1.07 0.77 2.76

I
276 225 0.32-0.30 0.29-0.28 1.26 0.88

C.2. 11/33 kilovolts a. b. C.3. 66/132 kilovolts D. Tubewells D.l. Government D.2. Private
a. NWFP

430

0.43-0.38

2.76

b. Other G. Public lighting H. Traction Average price (new tariff)

-Not =

430

0.43-0.38

1.82

48.5 paisa per kilowatt-hour.

applicable.

a. No adjustment is made for diversity factor. b. LRMC of energy: peak/off-peak. c. LRMC of capacity/(730 x LF x diversity factor) + LRMC of energy. d. Using time-of-day metering for capacity charge. Peak demand metering is less desirable but adequate. e. For (LF x diversity factor) = 20 to 50 percent. f. Ten hours of lighting per day.

PAKISTAN

163

New Capacity' (rupees per kilowatt per month)

WAPDA

tariff Average effective price LF x (rupees diversity per kilofactor watt-hour) (percent)

Energy (rupees per kilowatt-hour)

0.31 0.45 0.75 0.85 0.95 0.52 0.29 0.27 0.26

0.31 0.399 0.587 0.85 0.90 0.52 0.60 0.49 0.42

90 80 70

40 50 60

0.50

0.50

65 60 55 31 33 -

0.25 0.43 0.23 0.21 0.48 0.15 0.20 0.70 0.30

0.60 0.43 0.44 0.36 0.48 0.32 0.38 0.70 0.30

25 40 50 25 25 -

(Tablecontinueson thefollowingpage)

164

CASE STUDIES

Table 9-14 (continued)

Application of new tariff to 1978-79 consumption Average consumption per Thousands customer of (kilowatthours) customers
1,183.77 599.91 82.87 462.95 17.9 82.64 394.24 60.41

Consumer category
A. 1. Domestic a. 30 kilowatt-hours b. 30-200 kilowatt-hours c. 200 kilowatt-hours A.2. Commercial a. <100 kilowatt-hours b. >100 kilowatt-hours B. Industrial B.i. 70 kilowatt-hours B.2. 70-500 kilowatt-hours B.3. 11/33 kilovolts B.4. 66/132 kilovolts C. Bulk supply C.1. 400 volts a. b. C.2. 11/33 kilovolts a. b. C.3. 66/132 kilovolts D. Tubewells D. 1. Govemment D.2. Private a. NWFP b. Other G. Public lighting H. Traction

Total revenue (thousands of rupees)

6,568.7 19,781.0 19,177.7 24,233.9

Revenue per customer (rupees)

5.55 32.97 231.41 52.35

98.21 2.33 0.403 0.009

946.94 20,285.7 194,289 1,353,417

48,357.7 28,359.4 38,366.3 5,115.9

492.39 12,171.4 95,201.7 568,433

0.529

54,875

15,967

30,183.4

0.236 0.014

174,314 85,643

17,895 432 23,272.8

75,826.3 30,857.1

11.14

4,352.7 25,163.6 4,465.1

4,748.8

1.315 0.008

4,417 444,896

4,065.8 1,067.8

3.091.9 133,475

Therefore, the first consumption block is set at 100 kilowatt-hours. Existing tariff levels are retained at RsO.85 per kilowatt-hour for consumption up to 100 kilowatt-hours per month, and RsO.95 for greater use.
INDUSTRIAL (B). In general, the capacity charges are raised closer to the LRMC for the appropriate voltage level, whereas energy charges are somewhat reduced. The proposed capacity charges, however, are still consid-

PAKISTAN

165

erably below the strict LRMC to avoid a large price shock. Subsequent tariff revisions should focus on raising kilowatt charges still higher.
BULK SUPPLY (C). Tari:ffs are set to encourage a shift to higher bulk supply voltages, that is, the average price per kilowatt-hour falls as voltage rises. Sufficient margin must exist, however, between average bulk price and resale price to domestic and commercial customers (A. 1 and A.2) to allow a reasonable profit margin. TUBEWELLS (D). These tariffs are raised as high as social and political considerations will allow. There is much less freedom to raise prices to private farmers than to government tubewells.

PUBLIC LIGHTING AND TRACTION (G, H). The highest levels that are considered acceptable to these customers are set.
TOTAL AVERAGE TARIFF. Assuming that the 1978-79 consumption pattern remains essentially unchanged, the proposed new tariff structure yields an average price of RsO.485 per kilowatt-hour sold, which is within 1 percent of the target financial tariff level of RsO.482 per kilowatt-hour.

Policy Issues and C'onclusions

Calculation of the strict LRMC has shown that both the level and structure of existing tariffs are badlly distorted. In particular, the present capacity or kilowatt charges are well below the LRMC, but the kilowatt-hour charges are not too different from the best estimates of the LRMC of energy. In general, the low-voltage consumer charges were significantly below the LRMC of supply for this category. The main considerations used in the second stage of adjusting the strict LRMC to determine a practical tariff structure were: a. To achieve a financial target price averaging RsO.482 per kilowatthour during 19808 1 b. To provide some relief in the sensitive LV category, particularly in determining social-subsidized tariff blocks for poor households, small shops, and private tubewells c. To increase kilowatt charges wherever possible to bring them closer to the LRMC. In this regard, some formi of time-of-day metering for all
mers is strongly warranted.
MV and HV

custo-

As stressed earlier, setting tariffs is a continual process. The new tariff structure proposed here should be recognized as a transition tariff between

166

CASE STUDIES

the existing one and a tariff that more accurately reflects the LRMC. Thus, several important issues remain that have to be resolved in any future tariff revisions. First, the question of losses needs further analysis. Although losses are expected to fall from 34 percent of generation in 1979 to 25 percent by 1990, it is debatable whether such losses, which are considerably higher than normally accepted levels, should be incorporated routinely in the calculation of supply costs. Automatically passing on technical losses to the consumer removes the utility's incentive to reduce these losses. At the same time although stolen kilowatt-hours may constitute an economic benefit to the user, it is unclear how this benefit should be valued in the absence of willingness-to-pay and whether paying consumers should bear the cost of this supply, assuming that losses will be the same proportion of incremental demand as they are relative to average demand today. In future tariff revisions the LRMC based on some normal or accepted level of both losses and theft should also be calculated and compared with the actual LRMC. Second, the MOC or economic value of natural gas, a domestic energy resource used in the power sector, must be more accurately determined to pro3 vide a better estimate of the cost per kilowatt-hour. A study has been undertaken by the gas authorities in Pakistan to help resolve this problem. Third, the question of uniform national tariffs and appropriate tariffs in the KESC area should be explored. Since KESC has a well-established market with high load density and relatively small needs for new investment, their tariffs are lower than WAPDA'S, and the quality of supply is at least as good. This could cause discontent among WAPDA consumers, who have to subsidize rural and tubewell consumers while also helping to finance a heavy investment program. Fourth, WAPDA'S connection policy must be re-examined. Both a simultaneous change in tariff structure as well as connection policy would have greatly increased customer resistance. The tariff revision was considered as a first priority. Therefore, whether the full connection charges should be passed on even to poor consumers or whether some of these costs should be included in the kilowatt-hour charges must be reviewed in the future. Another issue is the extent to which tubewell customers should be encouraged through subsidies to convert from diesel to electric pumps. Finally, in subsequent tariff revisions, many items of data including the demand forecast and customer information involving load factors and so on must be more accurately estimated.

3. For a detailed economic analysis of the interactions between different energy subsectors. see Mohan Munasinghe, "An Integrated Framework for Energy Pricing in Developing Countries," Energy Journal, vol. I (July 1980), pp. 1-30.

Chapter 10

PhNippines
in the Philippines was not well organized. The fragmented operation of the sector was evidenced by stark regional imbalances in electric service coverage, wide disparities in production costs and selling rates, poor service in rural areas, and ill-matched levels of power supply and demand. The institutional framework is described below.
UNTIL RECENTLY THE ELECTRIC POWER SECTOR

Organization of the Sector

The electric power sector is composed of both government and private utility firms. The National Power Corporation (NPC) is wholly owned by the national government and is primarily responsible for generating and transmitting power to meet the bulk power requirements of other government and private electric utilities, cooperatives, and principal industries. Private and municipally owned utilities and electric cooperatives are basically responsible for distributing electricity to users in both urban and rural areas, although they occasionally generate power. In 1978, in addition to NPC, an estimated 282 electric utilitjieswere operating in the Philippines. Of this total, 194 were privately or municipally owned, and 88 were cooperatives. To support the government's objective of total electrification, in 1972 a rural electrification program was planned to complete the country's backbone electricity distribution system by 1980, the electrification of all barrios by 1984, and, finally, total electrification by 1990. The National Electrification Administration (NEA) was set up to organize electric cooperatives in rural areas. At the end of 1978, aggregate installed power capacity of utilities in the Philippines was 3,200 megawatts. Sixty-seven percent was operated by NPC, 16 percent by the Manila Electric Company (MEco), and 12 percent by other private industrial firms operating their own plants with more than 230 megaNote: This chapter was prepared in collaboration with the NPC Power Tariff Sector Study Team led by Ms. Malaine Manzo. The views expressed in this chapter do not necessarily reflect those of any Philippine power sector organizations or the Philippine government.

167

168

CASE STUDIES

Tuguegarao

an

,Lbuagan

auang -Ar'solano

Stiago

n t atuan v ex 0 0

_.kngat apang Pai tawak al e Binan IP ak-ban ..faga,.

MIN

PHILIPPINES

169

watts of capacity. The utilities generated more than 15.5 billion kilowatthours of electricity, 80 percent of which was produced by plants owned by NPC and MECO. The generation mix was 80 percent oil-fired, approximately 18 percent hydro-based, and 2 percent coal-fired and geothermal.
Electricity supply

The conditions of electricity supply vary markedly among the Philippines' three principal island groups, shown in Map 3. Luzon has by far the most developed power facilities and boasts an extensive integrated power network that traverses the entire Luzon mainland. In Mindanao, power development is expanding as NPC's hydroelectric grid reaches more urban and rural communities. In contrast, except for its large cities, the Visayas region has lagged behind in power infrastructure but is gradually building up its capacity to meet the region's growing electricity requirements. The Luzon grid accounts for more than 80 percent of total power generating capacity in the entire country. In 1978 the system had an aggregate installed capacity of 2,427 megawatts and produced 11.2 billion kilowatthours of electric energy with a generation mix of 84 percent thermal and 16 percent hydro. Leading suppliers of power in the island have been MECO and NPC. In 1978 MECO supplied electricity to more than 800 thousand customers in metropolitan Manila, accounting for more than 75 percent of total electric energy deliveries in Luzon. Whereas MECO caters to the power needs of urban Manila, NPC brings electricity to the provincial towns and municipalities of Luzon. NPC services were expanded primarily by exploiting the island's hydroelectric potential as well as by developing some thermal plants and, more recently, by using geothermal steam. With the integration of MECO's facilities in 1979, NPC'S grid connected all but two provinces in Luzon. The latter are scheduled to be linked to the system by 1981. Mindanao island is the site of Npc's second largest integrated grid operation. NPC supplies power to seven provinces along the northern coast of the island. The Mindanao grid is predominantly hydro (97 percent) and had a total installed capacity of 213 megawatts and a generation level of more than 1 billion kilowatt-hours in 1978. In addition to NPC, several private utilities and cooperatives operate isolated power systems consisting of minihydroelectric plants and diesel stations. The private utilities supply power to the landlocked urban centers in central and southern Mindanao. The Visayas region is made up of five principal islands-Panay, Cebu, Negros, Leyte, and Samar-plus several smaller islands such as Bohol. Before 1972, development of electricity supply in the Visayas was left to the private sector, and the government played a minor role. With the launching of the government's rural electrification program, however, NPC, in concert with electric cooperatives sponsored by NEA, embarked on a power develop-

170

CASE STUDIES

ment program. Diesel-based centralized electric power systems were installed in each island to support the growth of existing loads and to encourage the development of new markets. In 1978 total installed capacity in the Visayas was estimated at 240 megawatts; NPC, electric cooperatives, and private utilities each accounted for one-third of the total. In contrast to Luzon and Mindanao, where large hydro resources have been harnessed to produce power, hydroelectric potential in the Visayas is limited, and generating facilities are predominantly diesel. During the 1970s electricity service in the Philippines continued to expand. By 1978 all 83 cities and more than 64 percent of the country's 1,536 towns were electrified; this represented barely one-third of the country's total households, however. Electricity demand In Luzon electric energy consumption in 1978 was 9.4 billion kilowatthours. This translates into the highest per capita electricity consumption in a country in which the estimated annual consumption is 380 kilowatt-hours. Of its 3.5 million households, 47 percent had electric connections and consumed 2.6 billion kilowatt-hours. Industrial and commercial enterprises absorbed 3.9 billion and 2.5 billion kilowatt-hours, respectively. In Mindanao less than 14 percent of the region's 1.5 million households enjoyed electric service in 1978. In areas where service was available, however, per capita energy consumption levels were relatively high. The region's power market was characterized by a well-developed industrial load base that probably would expand rapidly as the hydroelectric grid of NPc extended to more communities in the island. The power market in the Visayas is at a nascent stage of development. This is partly attributable to the relatively higher cost of electricity production in the region. In 1978 only 16 percent of the less than 1.5 million households in the region shared the benefits of electric service. The Visayas has the lowest per capita electricity consumption in the region. With the scheduled installation of more diesel units plus the development of geothermal and coal-fired power plants, electricity consumption in the Visayas is expected to increase more than ten-fold during the next decade. At the start of 1978 there were almost 200 private and municipally owned utilities in the Philippines. More than half of these were self-generating; 20 percent relied on their own generation as well as on NPC or other utility firms. Altogether, these utilities accounted for more than 2,000 megawatts of installed capacity, mostly based on diesel power. Twenty private utilities relied on NPC as sole source of power, and another seven depended exclusively on MECO power deliveries. MECO is the largest private utility in the Philippines. Before it turned over its plant generation to NPc in late 1978, MECO had a total installed capacity

PHILIPPINES

171

of 1,672 megawatts, excluding a soon-to-be commissioned 350-megawatt thermal capacity. The second largest private utility is the Davao Light and Power Companies in Mindanao, with aggregate capacity of 80 megawatts, followed by the Panay Electric and Visayas Electric Companies in Panay and Cebu islands with a 70-megawatt capacity each. Other large private utilities with capacities ranging from 4 megawatts to 18 megawatts operate in principal cities, although several are connected to NPc's grids. Of the total energy delivered by the private utilities, more than 40 percent is absorbed by the industrial sector, 31 percent by commercial establishments, and 24 percent by residential consumers. The remaining 5 percent is used for street lighting and other miscellaneous consumption, including resale to other utilities. In 1978, of the eighty-eight existing electric cooperatives administered by NEA, forty-five operated in Luzon, twenty-five in Mindanao, and eighteen in the Visayas. Most of the cooperatives in Luzon purchased their power requirements from NPC; in contrast, those in Visayas and Mindanao mainly relied on self-generation. Still, other cooperatives partly generate and partly purchase power from NPC, but these will ultimately fully link up to NPC'S various power grids. Since cooperatives were established primarily to bring electricity to the countryside, they cater to the rural areas where commercial, industrial, and manufacturing activities are at a minimum. Eighty-seven percent of the 790,000 customers are residential, but they accounted for only 45 percent of energy sales. Commercial, industrial, and other sectors absorbed 23 percent, 19 percent, and 13 percent, respectively.
NPC

operations

NPC operates and maintains twenty power plants with an aggregate installed capacity of 1,006.5 megawatts.' These supply bulk power to various areas in Luzon, Visayas, and Mindanao through a distribution network of 5,645.2 kilometers of transmission lines connected to some 125 substations having a total capacity rating of 2,272.2 megawatt-amps. NPC operations involve two principal interconnected systems in Luzon and Mindanao and four subgrids in the Visayas, plus a few isolated plants. In 1977 plants operated by NPC generated 3,288.37 gigawatt-hours with a generation mix of 64.56 percent hydro, 33.79 percent oil-fired steam, 1.61 percent diesel, and 0.04 percent geothermal. An additional 97.40 gigawatthours of purchased energy from MECO in Luzon permitted a total of 3,101.06 gigawatt hours to be delivered to some 216 customers. The net revenue derived by the company for the year was P446.88 million (Table 10-1).

1. As of December 1977 and before integration of MECO'S therrnal plant into the NPC systeM.

172

CASE STUDIES

Table 10-1.

NPc

System Operations for 1977

Luzon
445.00 2,332.80 81.41 2,251.34

Item
Peak demand (megawatts) Gross generation (gigawatt-hours) Less: station use (gigawatt-hours) Net generation (gigawatt-hours)

Visayas
29.77 55.02 1.375 53.649 2.30 51.35 15.31

Mindanao
153.61 900.55 2.283 898.27 30.67 867.60 36.35

Total
628.38 3,288.37 85.06 3,203.31 97.4 199.65 3,101.06 446.89

Plus: purchased energy (gigawatt-hours) 97.4 Less: transmission energy losses (gigawatthours) 166.68 Energy sold (gigawatt-hours) 2,182.11 Net revenue (millions of pesos) 395.23 - Not applicable. Source: NPC data.
NPC'S

expansion program

The ten-year (1978-87) power expansion program for NPC can be used as the basis for developing estimates of the projected financial and economic cost structures of NPC power supply. The expansion program proposes to develop and commission three types of facilities in the three principal regions between 1978 and 1987 (Table 10-2). As the expansion program allows power generated by NPC to reach farther and wider markets, inefficient plants currently owned by self-generating utilities and industries will be phased out in favor of more economical NPC power supply. It will also be necessary to integrate other generating facilities into the system. Since power from NPC will still not be available in certain areas, the NEA and some firms may continue to install their own generating facilities. A significant reduction in the relative share of privately owned plant generation is expected within the coming decade. NPc's proposed ten-year development program will require investments of P73.5 billion, representing US$6.6 billion in foreign exchange costs and
Table 10-2. Physical Targets in 1978 to 1987 Addition
Additional plant capacities (megawatts) Additional transmission lines (kilometers) Additional substation capacities (megawatts) Source:
NPC

NPC'S

Power Expansion Program,

Luzon
4,382.0 3,956.5 5,705.0

Visayas Mindanao
916.6 1,677.8 1,333.2 1,801.2 2,464.0 750.0

Total
7,099.8 8,098.3 7,788.2

data.

PHIUPPINES

173

P2.4 billion in local costs. Of the total investments involved, 84 percent will finance the construction and installation of generating plant facilities, and the other 16 percent will cover the expansion of the transmission networks. The bulk of the capital investments (66 percent) will be channeled to Luzon; the remainder will be shared between the Visayas (10 percent) and Mindanao (24 percent).
LUZON. The growing electricity requirements of existing and potential customers in the Luzon grid is projected to increase the system's energy generation level from 10,298 gigawatt-hours in 1977 to 28,520 gigawatthours in 1987.2 The system peak demand is estimated to increase from about 700 megawatts in 1977 to 4,690 megawatts by 1987. To meet this increasing load requirement, the program proposes to install and operate six hydroelectric plants, twelve geothermal units, one nuclear, and two dual-fired thermal power plants in Luzon. The first geothermal plant came onstream in December 1978 and will be followed by other geothermal units in the succeeding two years. From 1981 to 1982, three hydro and one dual-fired thermal plant will be commissioned, followed by a nuclear plant in 1984. Additional hydro, dual-fired thermal, and geothermal plant facilities are programmed from 1985 to 1987. The Luzon expansion program also provides for the government to purchase and rehabilitate five MECO oil-fired thermal plants and to integrate them into the NPC system. VISAYAS. Because this region is made up of islands, the Visayas power expansion program involves the simultaneous development of several subgrids to meet the region's power requirements on an island-to-island basis. In the long run, however, when the market load in the region expands to a scale that would economically justify connection of the island subgrids (perhaps by submarine cable), an integrated Visayas Power Grid would eventually be developed. NPC began operating in Panay in four subgrids-Cebu, Negros, Bohol, and Leyte-Samar-in 1979. The hydroelectric potential in this area is limited, but there are several possible sites for geothermal development. Consequently, the Visayas program to expand generation is dominated by diesel, dual-fired thermal, and geothermal projects, supplemented by only a few small hydro plant additions. Projects for the five grids over the ten-year period are 228.8 megawatts diesel, 194.0 megawatts hydro, 420.0 megawatts (diesel-fired) thermal, and 108.0 megawatts geothermal. Total system generation is expected to increase from 750 gigawatt-hours in 1978 to about 4,200 gigawatt-hours by 1987.

2. The Luzon grid described in this section is an integration of both NPC and cilities and markets.

MECO power

fa-

174

CASESTUDIES

MINDANAO. During 1978 to 1982, the power required in Mindanao is expected to more than triple. Before 1983 the load forecast for Mindanao reflects certain restrictions in demand because of supply constraints. Under the proposed program, the existing system can provide 207 megawatts of peak power and 1,137 gigawatt-hours of energy in 1978, but at relatively low reserve levels. The potential power required will be adequately supplied with no suppression of demand and with the reserve capacity maintained at 24 percent of total system load only during the second half of the planning period. Altogether, 514 megawatts of installed capacity will be available in 1979, raising the reserve capacity level to 29 percent of the system load requirement. To sustain the increasing load growth of the grid, 115.25 megawatts of diesel capacity are scheduled to be commissioned within 1980 to 1982. An existing 22-megawatt diesel plant is also to be integrated into the system. The Agus River Basin hydro scheme will continue to be developed. As a result of the hydro plant addition, the hydro/thermal generation mix probably will increase from 56/44 in 1982 to 68/32 in 1985. Dual-fired thermal plants involving the flexible combination of coal and bunker-C as primary fuel will be introduced into the system to provide 150 megawatts of additional capacity in 1984 and another 150 megawatts in 1986. Finally, a further 331 megawatts of hydro capacity are programmed to be operational in 1987, resulting in a hydro/thermal generation mix of 67/33 for the year with a capacity reserve margin of 24 percent. EXPANSION OF TRANSMISSION. In the coming ten years, NPC plans to construct 6,820.85 kilometers of transmission lines, 45 percent in Luzon, 19 percent in the Visayas, and 36 percent in Mindanao. Regarding voltage level, 230-kilowatt lines will predominate in Luzon, and 69- and 138-kilovolt lines will be set up in the Visayas and Mindanao grids. The program also calls for the expansion or installation of substation facilities involving 4,145 megawatt-amps of additional capacity in Luzon, 1,383.2 megawattamps in the Visayas, and 750 megawatt-amps in Mindanao.

Computation of the Strict LRMC

The strict long-run marginal cost (LRMC) of supply associated with electricity generation, transmission, and distribution by NPC in its grids and subgrids, three sample private utility firms, and six sample electric cooperatives were calculated. In developing the estimates of capacity, energy, and customer costs, the long-run average incremental cost (AIC) method was addpted. Cost computations were based on the expansion programs of each sample utility supplemented by historical data on the utility's operating performance and system characteristics.

PHILIPPINES

175

Electric power usage was classified according to the voltage levels at which power was delivered: high voltage (HV), 69.0 kilovolts and above; medium voltage (Mv), 23.0 kilovolts to 34.5 kilovolts; and low voltage (LV), 13.8 kilovolts and below. NPC is responsible for HV deliveries, whereas private utilities and cooperatives account for LV deliveries. Estimates of LRMC are presented in 1978 prices, and shadow prices are used. A 15 percent opportunity cost of capital (OCC) is assumed because that is the current rate used to assess the economic viability of public infrastructure projects. The shadow price of foreign exchange is estimated to be 20 percent higher than the current exchange rate of P7.50 = US$1. This is 3 applied to the foreign cost component of the project. Expenses for unskilled labor are minimal relative to other cost components such as materials, equipment, and technical expertise. Thus shadow prices for unskilled labor were not applied. LRMC of bulk supplyfrom
NPC

Marginal cost schedules were separately determined for each NPC grid and subgrid-the Luzon grid; the Mindanao grid; and the Visayas subgrids of Cebu, Negros, Bohol, Panay, and Leyte-Samar-on the basis of the tenyear power expansion program (1978-87). Investment and operating costs (estimated in 1978 prices) associated with each grid's expansion program and the corresponding additions in capacity and energy demand supplied were basic to the LRMC' calculations. Operational characteristics of each system as indicated by present and past performance as well as anticipated improvement in operations were included in the estimation to reflect more realistic cost patterns.
CAPACITY COSTS. The iincrementalcapacity cost of NPC bulk power supply consists of investments in additional generation and HV transmission plant facilities plus the increases in operation and maintenance (O&M) costs involved in plant expansicon. Annual investment streams for both generating and transmission plant projects were developed for the plan period (1978-87) and then were discounted at 15 percent to arrive at the 1978 4 worth of the programmed capital outlays. The resulting present worth of investments was then annuitized over the estimated composite economic life of the plants in each respective grid. Similarly, the current worth of the annual increase in generation and transmission O&M expenses was derived.

3. The border pricing apprcach discussed in Chapter 3 is not used here. Instead, the strict LRMC has been calculated dilrectly in domestic prices, applying the traditional shadow exchange rate. 4. Since the investment costs of both peak and base-load generating plants are included in this calculation, the capacity cost of peak-period generation may be somewhat overestimated (see Chapter 4 and Appendix C).

176

CASE STUDIES

Costs at the generator and HV transmission levels were then compared with the present value of the added maximum demand to be met during the planning period, yielding annual incremental capital and O&M cost estimates. Finally the derived incremental generation capital and O&M costs were adjusted for power losses from transformation and transmission and were added to the total annual incremental capacity cost at HV levels. The results are summarized in Table 10-3. Table 10-3. Annual Incremental Capacity Cost at the
(pesos per kilowatt per year of coincident
HV

Hv

Levelfor
Total
1,606 1,805 1,828 1,557 2,302 2,062 2,051 1,866 1,673

NPc

maximum demand)

Grid
Luzon grid Mindanao grid Visayas subgrids Cebu Negros Bohol Panay Leyte-Samar Country average

Capital cost
1,555 1,735 1,710 1,437 2,182 1,896 1,887 1,777 1,610

O&M cost
51 170 118 120 120 166 164 89 63

Source:NPC data.

Results indicate that the highest incremental capacity costs per kilowatt per year are in the smaller island grids of the Visayas (Negros, Bohol, and Panay). Although investments in generating plant equipment in these subgrids are relatively small because of the predominance of less expensive diesel engines, additional demand for capacity in the areas is expected to expand slowly, thus raising unit cost. Moreover, the extensive spread of the HV transmission line network to reach small rural loads contributes to the high capacity cost. In contrast, the Cebu grid registers the lowest capacity cost reflecting combined effects of low-cost diesel investments during initial years as well as the substantial increments in capacity demand of the predominantly urban and industrialized market. Luzon, the largest grid, has a lower incremental capacity cost compared with that of Mindanao and the composite of Visayas. The value of future investments for both generating and transmission plants is highest in Luzon, but these costs are easily spread over the region's dense market. Although development of the Mindanao grid will involve twice as much investment as in the Visayas, its capacity is lower because of a wider and stronger market base. The marginal cost for peak and off-peak power supply was not calculated because of the difficulty in differentiating between peaking and base-load

PHuippiNEs

177

plants. Currently, most of the plants in NPC'S Luzon grid operate both as base-load as well as peaking plants, depending on the season. Moreover, studies indicate that application of peak and off-peak pricing may not yet be warranted in view of the existing high system daily load factor. In view of the possibility that Luzon's system load factor may deteriorate in the coming decade, however, further studies on peak pricing are being scheduled.
ENERGY COSTS. The energy-related costs of NPc's power supply were derived from the costs of fuel associated with the various types of plants scheduled for operation over the ten-year planning period, plus the generation and HV transmission plant O&M expenses required for energy production and deliveries. The fuel costs assumed for diesel, oil-fired, dual-fired, nuclear, and steam-geothermal plants are 0.2423, 0.1772, 0.1326, 0.0430, and 0.10 pesos per kilowatt-hour, respectively. These fuel costs were applied to the projected annual stream of plant generation to arrive at the annual fuel cost for each grid. Increases in fuel expense over the period were then discounted and compared with the increases in energy generation requirement to obtain the incremental energy fuel cost at the generation level. Increases in O&M expenses related to energy were similarly discounted, and the sum of incremental energy costs at the generation level was appropriately adjusted for transmission losses to yield customer cost estimates. The results are summarized in Table 10-4. The lowest incremental energy cost is indicated for Mindanao because hydro-based generation is expected to predominate throughout the plan period. The fuel cost register-edfor Mindanao reflects the initial reliance on diesel generation between 1980 and 1982 to sustain the growing energy demand in the region. The cost also allows for services to be extended to landlocked

Table 10-4. Incremental Energy-Related Costs at the Hv Levelfor

(pesos per kilowatt-hour)

NPC

Grid
Luzon grid Mindanao grid Visayas subgrids Cebu Negros Bohol Panay Leyte-Samar Country average Source: Npc data.

Fuel cost
0.0995 0.0725 0.1667 0.2175 0.1244 0.1663 0.1886 0.0937 0.0967

O&M cost
0.0019 0.0025 0.0033 0.0034 0.0034 0.0118 0.0060 0.0025 0.0023

Total
0.0924 0.0750 0.1700 0.2209 0.1278 0.1781 0.1946 0.0750 0.0990

178

CASE STUDIES

urban markets pending the full development of the hydroelectric potentials of the Agus River. Highest incremental energy costs are expected for the Visayas subgrids, except Leyte, because diesel-based generation is predominant there. Unit costs are slightly lower in Bohol and Negros where part of the generation is hydro-based.
CUSTOMER COSTS. NPC'S customer-related costs include administration and general (A&G) expenses and meter installation. Although, strictly speaking, some A&G expenses are not directly related to customers, for this study entire A&G expenses were considered, since a considerable amount is associated with accounting, customer billing, collection, and other related services. The incremental A&G costs were calculated by taking the projected yearly increases in A&G costs, computing their 1978 present worth, and obtaining their corresponding annuity value over the ten-year projections of the number of customers. Since NPC is primarily a bulk supply authority, such costs are insignificant compared with generation and energy costs, forming less than 0.05 percent of the total.

SUMMARY. Table 10-5 presents a summary of the marginal cost structure, in terms of unit value per kilowatt-hour delivered, estimated for the three principal regions of NPC operations. The figures indicate that the economic costs of bulk power supply in the three regions are not significantly different. In particular, incremental costs per kilowatt-hour are virtually equal for the Luzon and Mindanao grids. The relatively higher costs estimated for the Visayas, and the significant difference in the structure of costs compared with the two principal grids, is at-

Table 10-5. Marginal Cost Structure for

(constant 1978 pesos per kilowatt-hour)

NPC

Cost category'
Capacity Energy Total

Luzon
0.24 (72) 0.09 (28) 0.33
(100)

Mindanao
0.26 (78) 0.08 (22) 0.34
(100)

Visayas
0.20 (55) 0.17 (45) 0.37
(100)

Country total
0.24 (71) 0.10 (29) 0.34
(100)

Note: Percentages are given in parentheses. a. Customer costs are negligible. Source: NPC data.

PHILIPPINES

179

tributable to the higher energy-related cost, which is double that of the other two islands. Continuing increases in the price of fuel oil may widen the disparity in costs among the regions, with Mindanao possibly trailing behind because of its favorable hydro mix. LRMC of retail supply from private utilities Incremental cost schedules were developed for three sample private utilities operating in the principal urban centers of Luzon, Visayas, and Mindanao. The utilities, referrecdto as private utility 1, 2, and 3, belong to the top ten private electric firms in the Philippines. Private utility I (MECO) serves more than 800,000 customers in Manila with an annual sales volume in 1978 of 7,335 million kilowatt-hours. Expansion of the utility's services has been made possible by continued investments in generation and transmission plants as well as through the acquisition of smaller distribution systems of neighboring franchise-holders. The firm's customer base is projected to expand to nearly 1 million between 1978 and 1982 and will thus require heavy investments in distribution facilities. For its power requirements the utility will increase its reliance on NPC'S grid operations in Luzon as it tums over the ownership and operation of its large generating plants to the Philippine government through Npc. Private utility 2 (Cagayan Electric Light and Power Company) on the northern coast of Mindanao presently serves a highly diversified market involving 19,000 customers with an annual electric energy consumption of 80 million kilowatt-hours. The utility purchases all of its power from NPC'S Mindanao grid. Although it currently has a balanced distribution of sales among sectors, a heavier industrial load volume is anticipated in the coming years. Growth of private utility 2 will be significantly influenced by the regional industrialization program as well as by increased social and political stability. Private utility 3 (Visayas Electric Company) is located in the main trading port of the Visayas. The firm owns and operates 60 megawatts of thermal capacity supplemented by power purchases from the recently installed NPC diesel plant. Commercial and industrial loads account for more than 50 percent of its market. To meet the expected increase in power demand, private utility 3 plans to set up more distribution facilities as well as to augment its generating capacity through more thermal plant investments.
CAPACITY COSTS. Incremental capacity-related costs for retail power supply involve capital expenditures on transmission and distribution (T&D) facilities, including associated O&M. Of the T&D costs, however, only a segment is recognized as related to capacity; the rest is considered a customer-related expense. The allocation is determined by the number and spread of customers seived. Thus for HV and MV supplies involving rela-

180

CASESTUDIES

tively few customers, 100 percent of capital and O&M costs are categorized as capacity related. In contrast, for LV deliveries 70 to 90 percent of investment and O&M costs are allocated to customer-related costs when distribution facilities are more directly associated with customer connection. Capacity-related capital investments were then compared with additional power demand to be supplied and were discounted at 15 percent. Incremental O&M costs were added to the incremental capital costs to obtain the total incremented capacity costs at varying delivery voltages (Table 10-6). Data used to develop the marginal cost structures of the private utilities were obtained from the firms' respective forecasts of capital investment, financial performance, and sales. A planning horizon of ten years was used for private utilities 2 and 3; only a five-year plan period was considered for private utility 1. Procedures similar to that applied for bulk power supply from NPC were used to derive marginal costs of private utility operations. The supply voltage levels considered were HV for 138 kilovolts, 115 kilovolts, and 69 kilovolts; MV for 34.5 kilovolts; and LV for 13.8 kilovolts and below. Two sets of calculations are presented. The first distinguishes between capacity and energy costs as they are perceived by the utility. Thus, purchases of energy from NPC are treated solely as energy costs. In the second calculation the distinction is made on the basis of the costs to the economy, so that the cost of energy purchased from NPC is allocated to the underlying capacity and energy costs incurred by NPC in production. Since the bulk of energy requirements are purchased directly from NPC, this adjustment makes a considerable difference. Table 10-6 indicates that the bulk of the total incremental LV capacityrelated costs for all utilities results from the HV or MV capacity-related investment costs. At the LV delivery level, the highest value of capacityrelated cost is registered for private utility 3. This is due to the low incremental demand averaging 2.56 megawatts, over which a relatively large capital outlay (P 18 million) and O&M costs (average of P0.9 million a year) have to be spread. Private utility 2 has a lower incremental capacity cost because of a larger added demand level. Finally, although it will make the largest investment of PIll million, the capacity-related cost of private utility 1 is lowest because of a favorable additional demand averaging 106 megawatts annually.
ENERGY COSTS. Estimates of incremental energy-related costs of retail deliveries of private utilities are summarized in Table 10-7. Energy costs for private utilities 1 and 3 reflect the costs of both purchases and self-generated energy; for private utility 2 the cost represents energy purchased entirely from NPC. The incremental cost of self-generation is significantly lower than that for power supplied by NPC, as indicated below:

PHILIPPINES

181

Table 10-6. Annual Incremental Capacity Cost for Private Electric Utilities
(1978 constant pesos per kilowatt per year of maximum noncoincident demand)

Private utility
I-Luzon 2-Mindanao 3-Visayas

Total Voltage Incremental Incremental Added HV or MV incremental level capital costs O&M costs capacity cost capacity cost
LV MV LV HV LV

33.63 47.10 67.88 146.69 101.54

4.24 49.11 35.92 8.09 59.00

84.64 109.33 178.00

122.51 96.21 213.13 154.78 338.54

-Not

Source:

applicable. NPC data.

Table 10-7. Annual Incremental Energy-Related Costs for Private Electric Utilities
(constant 1978 pesos per kilowatt-hour)

Private utility
l-Luzon 2-Mindanao 3-Visayas Source:
NPC

HV

MV

LV

0.3250 0.3351 0.3221

0.3350 0.3402 0.3426

0.3593 0.3736 0.3928

data.

Private utility
I-Luzon 3-Visayas

Self-generation (pesos per kilowatt-hour)

0.2132 0.2278

Purchased energy (pesos per kilowatt-hour)

0.3989 0.3349

The weighted average value of energy cost to the utility is high because of the relative mix of its own energy and purchased energy, which is 1:10 for private utility 1 and 2:3 for private utility 3.
CUSTOMER COSTS. For the retail utilities, customer costs are the second largest component of total incremental cost, that is, 26 percent for private utility 1, 23 percent for 2, and 26 percent for 3. The incremental customer cost can be broken down as shown in Table 10-8. Private utility 2 has the fewest additional customers, accounting for its highest incremental customer-related cost. Although private utility 1 will be incurring the largest expenses in customer-related distribution facilities, these costs are spread widely over a fairly large market. Finally, private util-

182

CASE STUDIES

Table 10-8. Annual Incremental Customer-Related for Private Electric Utilities

(pesos per customer)

Costs at the

LV

Level

Private utility
l-Luzon 2-Mindanao 3-Visayas Source:

Incremental Incremental Incremental capacity costs O&M and A&G costs meter costs
573.22 337.29 141.67 608.33 1,698.20 185.95 44.15 594.20 65.55

Total
1,285.70 2,629.69 393.17

NPC data.

ity 3's low customer-related cost is attributable to investments projected for its distribution network that will be shared by many additional customers.
SUMMARY. The marginal cost structure for the sampled private utilities is summarized in Table 10-9. Shadow pricing aside, these costs are presented as the utilities perceive them in regard to the distinction between energy and capacity. The adjustments to the cost structure when NPC energy sales are allocated between capacity and energy costs are considerable. This is illustrated by comparing Tables 10-9 and 10-10. For example, for public utility 1, incremental energy costs for LV consumers, as perceived by the utility, account for 71 percent of total costs (Table 10-10).

Table 10-9. Marginal Cost Structure at the LV Level for Private Electric Utilities, from the Utilities' Viewpoint
(constant 1978 pesos per kilowatt-hour of sales)

Cost category
Capacity Energy Customer Total

Private utility 1, Luzon

0.02 (4) 0.36 (71) 0.13 (25) 0.51
(100)

Private utility 2, Mindanao

0.03 (6) 0.37 (71) 0.12 (23) 0.52
(100)

Private utility 3, Visayas

0.06 (10) 0.39 (64) 0.16 (26) 0.61
(100)

Note: Percentages are given in parentheses. Source: NPC data.

PHILIPPINES

183

LRMC of retail supply from sample cooperatives

Operations of six sample electric cooperatives were analyzed to determine the marginal costs of their respective power supply. Of the six, three cooperatives (A, B, and C) are located in the Luzon mainland, two (D and E) in Mindanao, and one (F) in the Visayas. Aside from their geographical spread, the sample was selected to get a fair representation of the varying conditions under which cooperatives operated in the Philippines. For instance, whereas coop B is typical of most cooperatives with a mix of urban and rural customers, coop A has a heavy industrial load, and coop C has a rural market base. Coop D represents a utility that purchases power from another retail utility. Coop E operates in a region where peace and order is a crucial concern. Finally, coop F is a selfgenerating utility strategically located in a first-class province and serves 2 cities, 49 municipalities, and more than 1,400 registered barangays or local villages. The five-year development plan of each cooperative as submitted to NEA served as the primary data source for determining the cooperatives' marginal cost structure. Resulting costs are given at the Lv level (that is, below 13.8 kilovolts) at which the cooperatives deliver power. As in the case of the private utilities, the average marginal cost of a kilowatt-hour and the structure of costs are based on the cooperatives' operating viewpoint and how NPC sales are allocated between capacity and energy.

Table 10-10. Marginal Cost Structure Cost Allocation for Private Electric Utilities, from the Economy's Viewpoint
(constant 1978 pesos per kilowatt-hour)

Cost category
Capacity Energy Customer Total

Private utility 1, Luzon

0.26 (52) 0.11 (22) 0.13 (26) 0.51 (100) are for
LV

Private utility 2, Mindanao

0.32 (61) 0.08 (16) 0.12 (23) 0.53 (100)

Private utility 3, Visayas

0.18 (30) 0.26 (43) 0.16 (26) 0.61 (100)

Note:For example, estimates

Source:
NPC data.

consumption. Percentages are given in parentheses.

184

CASE STUDIES

CAPACITY COSTS. The incremental capacity costs of power supply by cooperatives were based on the projected investment requirements for developing primary and secondary distribution lines and for attendant facilities. The annual stream of these investments was matched with the forecast of additional annual load demand. Applying a 15 percent discount rate and assuming an average plant life of twenty-five years, the resulting incremnental capacity cost per kilowatt was converted into its annual equivalent. Further, 30 percent of total incremental O&M as well as A&G expenses were assumed to be capacity related. Taken together, the annual incremental capacity costs per kilowatt per year of maximum capacity demand for each sampled cooperative is summarized in Table 10-11. Coop F, which has self-generating capability registers the highest level of capacity-related costs, amounting to more than P1,000 per additional kilowatt per year of satisfied demand. This is attributable to its sizeable investments in power plant improvements and associated O&M expenses totaling more than P2.5 million annually. Moreover, the relatively low volume of annual incremental demand of 395 kilowatts implies a limited spreading of these costs. The lowest capacity cost per kilowatt per year is exhibited by coop A. Although it has the largest projected investments among the utilities (averaging P6 million a year), costs are adequately spread over a fast-expanding market that anticipates an average annual load increase of 79,000 kilowatts. Its relatively good load factor of 50 percent further explained coop A's low unit capacity cost of P0.03 per kilowatt-hour. Coops B, C, and E record similar values of incremental capacity costs, although each reflects varying levels of capacity investments and demand structures. Coop D has relatively high average capacity costs because of its large investments and low incremental demand.

Table 10-11. Annual Incremental Capacity Cost at the Lv Level for Electric Cooperatives (pesosper kilowattper year of maximumnoncoincident demand)
Incremental Incremental Cooperative capital cost O&M and A&G costs
A B C D E F Source:
NPC

Total
126.17 263.27 287.14 455.04 326.58 1,046.91

120.67 205.32 236.69 368.18 280.43 906.08

5.50 57.88 48.45 86.86 46.15 140.83

data.

PHILIPPINES

185

Of the six sample cooperatives, four purchase power from Coops A, B, and C are served by NPC'S Luzon grid and coop E by the Mindanao (Agus) grid. Accordingly, the energy-related costs for these cooperatives are based on the total marginal costs derived for NPC bulk power supply adjusted for T&D losses down through the LV delivery level. Coop D is expected to purchase its power requirements from a private utility firm in the initial two years and from NPc's Mindanao grid thereafter. Coop F, however, will be generating part of its power and will supplement that by purchasing from a private utility. Table 10-12 presents the incremental energy-related costs of power delivered by cooperatives.
ENERGY COSTS. NPC.

Table 10-12. Incremental Energy-Related Costs at the LV Level for Electric Cooperatives
(pesosper kilowatt-hour) Cooperative Source of power A B
C
NPC NPC NPC

Total 0.3983 0.3958 0.3676

0.4004

Luzongrid Luzongrid Luzongrid

D E
F

Privateutility and
Mindanao grid
NPC

Mindanaogrd

0.3781 0.6874

Self-generation and

privateutility Source: NPC data.

A purchase price of P0.3349 per kilowatt-hour equivalent to the marginal cost of NPC's bulk power supply was assumed for the cooperatives operating in Luzon. This energy cost is, however, passed on to the customers at a higher cost to reflect the losses incurred in distribution. For instance, coop C suffers a 9 percent energy loss, whereas coops A and B incur losses of up to 16 percent and 15.5 percent, respectively, of their purchased energy. Coop E purchases its power from NPC-Mindanaoat P0.3351 but passes it on to its customers at P0.3781 because of an 11.4 percent distribution loss. Similarly, coop D has an average energy purchase cost of P0.3555 per kilowatt-hour, which translates to P0.4004 per kilowatt-hour at the delivery end because of a 1.2 percent loss. Finally, coop F's combined generation and purchase cost is P0.5525, but its losses of nearly 20 percent mean a further upward adjustment of costs.
CUSTOMER COSTS. Components of customer-related costs of power supplied by cooperatives include total costs of additional secondary lines, trans-

186

CASE STUDIES

formation facilities, service drops, and meters. Also included are substantial percentages of the O&M expenses; general administrative and consumer accounts expenses; and some provision for the cost of architectural and engineering services, street lighting, and general plant and other contingencies. As in the computation of capacity-related costs, the stream of projected investments in distribution facilities was discounted at 15 percent to arrive at its 1978 value annuitized over an average plant life of twenty-five years. Other customer-related incremental expenses were then matched with the forecast of additional customers to be connected. Results of these estimations are summarized in Table 10-13.
Table 10-13. Annual Incremental for Electric Cooperatives
(pesos per customer)

Customer-Related

Costs

Incremental Incremental Cooperative capital cost O&M and A&G costs

A B C D E F 82.61 45.17 72.71 62.01 161.97 79.50 13.15 48.36 84.71 48.64 95.81 49.63

Total
95.76 93.53 157.42 110.65 257.78 129.13

Source: NPC data.

Coop B is expected to incur the highest volume of investments in customer-related facilities, amounting to an annual average of P3.9 million. Nevertheless, it has the lowest incremental cost per customer of P3.53. This is made possible by its wide market base, which is projected to increase by an average of almost 7,000 customers annually between 1978 and 1982. Although coop A ranks only second on the basis of cost per customer, its equivalent unit cost (pesos per kilowatt-hour) is lowest, at less than P0.01 per kilowatt-hour, because of a relatively higher load factor coupled with a favorable market expansion at a rate of more than 5,000 new customers a year. Coop E requires over P2 million of annual investment in customer-related facilities, but customer additions are estimated at less than 2,000 a year, thus the high cost per customer. On the basis of per kilowatt-hour sales, coop D will have the highest unit cost (P0.04 per kilowatt-hour) because of poor energy sales.
SUMMARY. The marginal private financial cost structure of the sampled coops, as perceived from the coops' viewpoint (excluding shadow pricing

PHILIPPINES

187

adjustments), is presented in Table 10-14. Energy-related costs appear to constitute the bulk of the coops' incremental cost. Apart from coop F, this reflects the cost of purchased energy from NPC. Except for coop F, the picture changes dramatically when energy costs are broken down into the cost categories that Npc incurs. This is demonstrated in -Table 10-15, which presents a more accurate indication of the real contribution of energy and capacity to the coops' cost structures. On this basis, except for coop F where self-generated energy costs predominate, the cost of power delivered by cooperatives is chiefly determined (60 to 80 percent) by capacity-related investments, rather than the range of 7 to 24 percent indicated in Table 10-14. Table 10-14. Private Marginal Cost Structure at the LV Level for Electric Cooperatives, from the Cooperatives' Viewpoint
(constant 1978 pesos per kilowatt-hour)

Cost category
Capacity Energy Customer Total

Coop A Coop B Coop C Coop D Coop E Coop F

0.03 (7) 0.39 (91) 0.01 (2) 0.43 (100) 0.08 (16) 0.39 (80) 0.02 (4) 0.49 (100) 0.08 (17) 0.37 (80) 0.01 (2) 0.46 (100) 0.14 (24) 0.40 (69) 0.04 (7) 0.58 (100) 0.09 (18) 0.38 (78) 0.02 (4) 0.49 (100) 0.20 (22) 0.69 (76) 0.02 (2) 0.91 (100)

Note: Percentages are given in parentheses. Source:NPC data.

Table 10-15. Private Marginal Cost Structure at the LV Level for Electric Cooperatives, from the Economy's Viewpoint
(constant 1978 pesos per kilowatt-hour)

Cost category
Capacity Energy Customer Total

Coop A Coop B Coop C Coop D Coop E Coop F

0.32 (73) 0.11 (25) 0.01 (2) 0.44 (100) 0.36 (74) 0.11 (22) 0.02 (4) 0.49 (100) 0.35 (76) 0.10 (22) 0.01 (2) 0.46 (100) 0.36 (62) 0.18 (31) 0.04 (7) 0.58 (100) 0.39 (80) 0.08 (16) 0.02 (4) 0.49 (100) 0.20 (22) 0.69 (76) 0.02 (2) 0.91 (100)

Note: Percentages are given in parentheses. Source:Npc data.

188

CASE STUDIES

Adjustment of the Strict LRMC

In the Philippinesessentiallythree independently determinedsets of electricity rates exist: the bulk power rates of NPC, and the retail rates of private utilitiesand those of electriccooperatives.In general,rate designsrecognize several objectives including attainment of financial viability, provision of electric serviceat affordablerates, promotionof efficientuse of capacity and energy, and promotionof rational industrialexpansion. Existingtariffs NPC bulk power tariffs vary considerably from grid to grid. The prices essentiallyreflect the cost of power supply, which in turn is highly influenced by the generationmix. Consequently, the Luzon grid with an energy mix of 84 percent thermal, 16 percent hydro, and a growing share of geothermal has a standardprice of 20 centavosper kilowatt-hour.This price is 67 percent higher than the standard tariff for the predominantly hydro-basedMindanao grid but is 33 percent lower than that for the diesel subgrid in the Cebu (Visayas)area. Except for two subgridsin the Visayasregion, NPC presentlyappliestwo sets of tariff schedulesin each grid, one for utilitiesand anotherfor nonutilities. The latter category includes industrial customers and miscellaneous bulk power users such as militarybases and governmentcenters. As shown in Table 10-16,utility prices are set lower than nonutilityratesto discourage direct connection of industriesand to promote socializedpricing. Utilities charging high retail tariffs are usually those which incur heavy distribution losses and unaccountable energylosses, have predominantly ruralloads with extremely low load factors, cover extensive but thinly populated service areas, or operate inefficientand expensiveoil-firedgeneratingsets. Retail tariffs are classifiedaccording to the type of power use, namely, residential, commercial, industrial (general power), public buildings and street lighting, irrigation, and resale (wholesale).Among private utilities, average residential prices are higher than for other types of customers. Chargesfor street lightingare invariablylowest, followedby industrialand resale prices. Larger private utilities charge lower tariffs; MECO offers the lowestretail rates. In general, utilitiesin Luzonusuallycharge lower tariffs than those in Visayasand Mindanao. Prevailingprices of cooperativesare set such that tariffs for residences, public buildingsand streetlightingare on a par with each other. Commercial customerspay from 10 to 15 percentmore than residentialusers. Industries, irrigation,and water systemsare charged10 to 15 percent less than residential prices.

PHILIPPINES

189

The NPC price structure consists of a demand charge based on the measured maximum demand and an energy charge based on the quantity of energy consumed during the billing month. In general, both demand and energy charges are bracketed into blocks. Before 1977 these charges were promotional, reflecting lower unit capacity costs for higher loadings as well as savings from the high load factor or prolonged use of capacity. Worldwide inflation and limited energy supplies caused an abrupt rise in the cost of capacity and fuel, and to account for these changes in cost components, utility and nonutility tariffs in Mindanao and prices in Cebu and Luzon took on an inverted structure. The tariff structure was reformed to minimize the cost of investments in additional capacity and to promote energy conservation among large users of oil-based electric energy. Table 10-16. Average Electricity Tariffs in 1978
(constant 1978 centavos per kilowatt-hour) Luzon Type of rates Bulk power rates
NPC

Visayas Utilities Nonutilities

Mindanao Utilities Nonutilities

Utilities

Nonutilities

19-20 20 (grid average = 20)

MECO SG

22-27 22-32 (grid average = 30)

SG NSG

9 13 (grid average = 12)

SG NSG

Retail power rates Private utilities Residential Commercial Industrial

Resale

26 35 31
20

32-53 33-49 23-48

33

60-82 64 50-51
-

37-71 47-98 -

46-74 50-60 40-52

45

16 17 22
-

Street lighting

35
SG

13-32
NSG

47-73
SG

22-51
NSG

24-69
SG

14
NSG

Cooperatives
Residential Commercial Industrial Public buildings Street lighting rrigation

70-90 63-90 68-90 71-82 16-34 68-87

NiSG,

30-60 35-65 27-60 39-60 6-21 30-37

68-120 70-120 60-120 70-120 14-38 58-67

70-82 70-80 64-75 45-80 25-32 57

56-100 60-100 100 100 14 -

23-94 25-96 14-93 23-96 9-26 14-68

Note: SG, self-generating; Not applicable. Source: Npc data.

nonself-generating or connected to the NPC.

Charges of private utilities come in various forms. In general, flat rates are applied for street lighting and sometimes unmetered residential and commercial users. Residential prices are charged on the basis of energy consumed (kilowatt-hours), and these energy charges are progressively lower

190

CASE STUDIES

with increased consumption, with the notable exception of MECO in which residential consumption of less than 200 kilowatt-hours and commercial consumption of less than 90 kilowatt-hours a month are heavily subsidized. Metered tariffs to commercial users are classified according to loads and also take the form of energy charges occasionally coupled with demand charges for larger consumers. Again, most private utilities apply promotional energy prices to commercial users as their load and load factors improve. Industrial tariffs consist of both demand and energy charges and are invariably promotional. Finally, a minimum charge is usually applied to all types of metered customers. Tariffs of cooperatives follow a standard NEA design. Cooperative prices to residential and commercial customers start with a minimum charge for the first 10 kilowatt-hours and 20 kilowatt-hours, respectively. Energy consumption beyond these levels is subject to flat rate energy charges. Occasionally, energy blockings are used featuring promotional rates. Industrial prices of cooperatives are usually in the form of a simple demand charge plus a flat rate energy charge. Estimatedfinancial cost structures On the basis of current and projected financial statements of NPC and of the sampled private utilities and cooperatives, estimates of financial cost structures of wholesale and retail electric power supply were developed. The findings summarized in this section are compared with the marginal cost structures derived earlier for the same utilities. NPc. In addition to the capital requirements of NPc's expansion program, operating cost data for various NPC plants as well as present value estimates of existing assets in service were considered in projecting costs of future operations. An exchange rate of P7.50 = US$1.00 and an annual price escalation rate of 8 percent were also applied to estimate future values of assets and costs. Disregarding the return on capital, during the initial year, 1978, NPC expected to incur an average cost of 11.12 (centavos) per kilowatt-hour generated from its total operations. Costs are lowest in Mindanao at 4.27 (or 62 percent less than the aggregate average) per kilowatt-hour, and highest in the Visayas at 26.7 per kilowatt-hour. Within the next five years, the cost of power supplied by NPc is expected to more than double from the 1978 level, reaching an aggregate average of 23.4 per kilowatt-hour. Significant cost increases are expected in Luzon and Mindanao, where an equivalent rise of about 11.6 per kilowatt-hour is anticipated from additional oil-fired generation and higher plant operational expenses. In contrast, a relatively lower cost increase of 7.8 per kilowatt-

PHILIPPINES

191

hour is expected in the Visayas, which is attributable to increases in capital charges and fuel expenses. The second half of the plan period, 1983 to 1987, is forecasted to experience a slower rise in costs. By 1987 the aggregate average cost is anticipated to be 27.5 per kilowatt-hour, reflecting an additional 4. 10per kilowatthour over the 1982 projected level. The corresponding cost increases in each grid are 5.6 in Luzon, 3. 1 in Visayas, and 2.1 in Mindanao. The return on capital of 8 percent on the revalued rate base contributes substantially to the cost of service, and it accounts for an average of 35 percent of total cost nationwide. The cost components can be further classified corresponding to the cost categories used previously to present marginal cost estimates: capacity-, energy- and customer-related costs. Included in capacity costs are 80 percent of O&M expenses incurred in generation and transmission of power as well as depreciation. Energy-related costs consist of fuel oil and steam costs in addition to 20 percent of C)&M expenses. Finally, customer-related expenses refer to metering costs, general and administrative expenses, and other cost items associated with customer service. Table 10-17 summarizes the cost breakdown per kilowatt-hour of energy sales, including provision for a return on the rate base of 8 percent. Table 10-17. Financial Cost Structure of Energy Sales for
(pesos per kilowatt-hour)

NPC,

1978

Cost category
Capacity Energy Customer Sub-total Plus 8 percent return on rate base Total

Luzon
0.0399 (31.5) 0.0863 (68.0) 0.0006 (0.5) 0.1268
(100.0)

Mindanao
0.0288 (64.4) 0.0153 (34.2) 0.0006 (1.4) 0.0447
(100.0)

Visayas
0.0358 (12.8) 0.2442 (87.1) 0.0004 (0.1) 0.2804
(100.0)

0.0789
(-)

0.0492
(-)

0.0549
(-)

0.2057
(-)

0.0939
(-)

0.3353
(-)

Note: Percentages are given in parentheses. Not applicable. Source: NPC data.

192

CASE STUDIES

The average cost of power delivery in the Luzon grid will increase from 20 in 1978 to more than 48 by 1987. The relative contribution of capacity-related costs will shift from 32 percent in 1978 down to 20 percent in 1982, rising again to nearly 33 percent by 1987. The movement reflects the capital investment during the plan period. Likewise the expected rise in fuel costs will result in a relative weight increase of energy-related expenses during the first five years of the planning period, as heavy reliance on the government-acquired MECO thermal plants continues. Customer-related costs, however, will remain minimal at less than 1 percent. In Mindanao NPC'S cost of delivered power is projected to increase fourfold or by 290 during the planning decade. Initially, expenses will be largely attributable to capacity investments. But rapid buildup of energy-related (diesel) costs in excess of capacity expenses is anticipated until the hydroelectric potential of the region is amply developed by 1987. In the Visayas these costs will exceed capacity until costs rise by 23 per kilowatt-hour between 1978 and 1987. Energy-related expenses will predominate except toward the end of the planning period, when several hydroelectric projects are to be commissioned.
PRIVATE ELECTRIC UTILITIES. Not all the private utilities sampled were able to provide financial projections on which financial cost structures of their power deliveries at LV levels could be based. Nevertheless, attempts to

Table 10-18. Financial Cost Structure of Energy Sales for Private Electric Utilities, 1978
(pesos per kilowatt-hour)

Cost category
Capacity Energy Customer Sub-total
Plus 12 percent return

Luzon
0.0390 (15.6) 0.1532 (54.9) 0.0741 (29.5) 0.2663 (100.0) 0.0029
(-)

Mindanao
0.0364 (26.4) 0.0700 (50.8) 0.0315 (22.6) 0.1379 (100.0) 0.0390
(-)

Visayas
0.1038 (18.8) 0.3842 (69.7) 0.0632 (11.5) 0.5512 (100.0) 0.0766
(-)

on rate base Total

0.2692
(-)

0.1769
(-)

0.6278
(-)

Note: Percentages are given in parentheses. Not applicable. Source: NPC data.

PHILIPPINES

193

Table 10-19. Financial Cost Structure of Energy Sales for Electric Cooperatives, 1978
(pesos per kilowatt-hour)

Luzon Cost category

Capacity Energy Customer Sub-total

Mindanao Coop C
0.0339 (10.1) 0.2260 (67.3) 0.0757 (22.6) 0.3356 (100.0) 0.1191 0.4547

Visayas Coop F
0.1496 (16.6) 0.6546 (72.7) 0.0958 (10.7) 0.9000 (100.0) 0.1056 1.0056

Coop A
0.0123 (4.1) 0.2683 (89.8) 0.0183 (6.1) 0.2989 (100.0) 0.0153 0.3142

CoopB
0.0914 (21.4) 0.2499 (58.5) 0.0856 (20.1) 0.4269 (100.0) 0.1125 0.5394

CoopD
0.1358 (17.3) 0.4944 (63.2) 0.1526 (19.5) 0.7828 (100.0) 0.4511 1.2340

CoopE
0.0985 (34.1) 0.1107 (38.3) 0.0799 (27.6) 0.2891 (100.0) 0.1990 0.4792

Plus 12 percent return

onratebase Total

Note: Percentages are given in parentheses. - Not applicable. Source: NPC data.

construct the required cost profile from actual 1978 performance were made and matched with other costings provided by the utilities themselves. Table 10-18 summarizes the derived financial cost structures with separate provision for return on operating assets set at the maximum allowable rate of 12 percent. The relative distribution of costs is similar among the three utilities, although actual levels vary significantly. Energy-related expenses invariably account for over 50 percent of total costs; the level for Visayas is five times more than that for Mindanao and twice more than that for Luzon. Capacityrelated costs are also highest for Visayas, where the firm's power plants continue to operate.
ELECTRICCOOPERATIVES. The cost structure of power supply from the cooperatives' private financial viewpoint is summarized in Table 10-19. Capacity costs consist essentially of fixed depreciation charges. Energy costs reflect power purchase costs as well as the fuel expenses of self-generating cooperatives. Customer-related costs refer to A&G, O&M, and miscellaneous service expenses. ][naddition, capital costs corresponding to a 12 percent return on net utility plants in service were imputed to levels of costs comparable to that of private utilities.

194

CASE STUDIES

Results indicate that total costs incurred by cooperatives purchasing power directly from NPC are significantly lower than those that generate their own power (coop F) or purchase from other utilities (coop D). Among NPC customers, regional costs are not substantially different despite the marked difference in NPC selling rates. For instance, coop A's cost is highly competitive with coop E's, although the latter purchases power at a much lower rate from NPC. Energy-related expenses account for the bulk of capacity costs incurred by the cooperatives. The disparities in energy costs among coops in Luzon connected to NPC are attributable to varying levels of distribution losses that raise the costs of purchased power from P0.20 per kilowatt-hour at the HV delivery level to as high as P0.27 per kilowatt-hour on the LV supply side.
Rates under alternative costing approaches

NPC. Table 10-20 summarizes the resulting price levels indicated by the marginal and financial cost structures developed for NPC. Existing tariffs are generally in line with financial costs. The bulk of the shortfall in the price for the Visayas is easily obtainable through the fuel cost adjustment clause. The disparity between the 1978 year-end rate for Mindanao of P0.12 per kilowatt-hour and the resulting P0.09 per kilowatt-hour rate is attributable to efforts by NPC management to stabilize price levels in the region. The rate of P0. 12 per kilowatt-hour was implemented in the latter half of 1978; previously an average tariff of P0.09 per kilowatt-hour was charged. Anticipated increases in diesel-based generation in the Agus grid were expected to increase the cost of production substantially to over P0. 17 per kilowatt-hour in the coming years. Thus an intermediate price level of P0. 12 per kilowatthour was introduced and continued through 1979.
Table 10-20. Average Tariffs for NPc, 1978
(pesos per kilowatt-hour)

Grid or region
Luzon Mindanao Visayas Philippines

Existing year-end price level Financial Ratio 1978 cost (1) (2) (2)1(1)
0.20 0.12 0.30 0.19 0.20 0.09 0.32 0.19 1.0 0.75 1.07 1.0

LRMC (3)
0.33 0.34 0.37 0.34

Ratio (3)1(1)
1.65 2.83 1.23 1.79

LRMC versus financial costs (3)1(2)

1.65 3.77 1.16 1.79

Source: NPC data.

PHMIPPMS

195

In contrast, strict LRMC is an average of 79 percent higher than existing price levels. The largest disparity of 183 percent is indicated for Mindanao tariffs when marginal capacity cost charges are over three times that reflected in financial capacity costs. Differences in Luzon's tariff levels are also significant, requiring a 13 increase on existing 1978 rates. A relatively small deviation occurs in the Visayas; the high fuel cost component of production costs is automatically recovered by NPC through the existing fuel cost adjustment clause. Regarding tariff design, the resulting LRMC structures require a significant shift in the relative distribution of rate charges from the energy charges to the demand charges. Allocations under the two costing approaches are compared with the actual distribution of revenue charges as embodied in prevailing rate schedules in Table 10-21.

Table 10-21. Allocation of Charges for

(percent)

NPC

Luzon Charge
Demand (kilowatts) Energy (kilowatt-hours) Customer

Existing
5 95 0

Financial
57.8 42.0 0.2

LRMC
74.2 25.5 0.2

Mindanao Charge
Demand (kilowatts) Energy (kilowatt-hours) Customer

Existing
15 85 0

Financial
83.1 16.3 0.6

LRMC
77.5 22.4 0.1

Visayas Charge
Demand (kilowatts) Energy (kilowatt-hours' Customer
-Not

Existing
10 90 0

Financial
27.1 72.8 0.1

LRMC
54.5 45.4 0.1

applicable.

Source: Npc data.

196

CASE STUDIES

Table 10-22. Average Tariffs for Private Electric Utilities, 1978

(pesos per kilowatt-hour)

Grid or region
Luzon Mindanao Visayas Source:
NPC

Existing Financial average average price, 1978 cost (1) (2)

0.31 0.21 0.64 0.27 0.18 0.63

Ratio (2)1(1)
1.13 1.14 1.02

LRMC (3)
0.51 0.53 0.61

Ratio (3)1(1)
1.64 2.52 1.05

LRMC versus financial costs (3)1(2)

1.89 2.94 1.03

data.

PRIVATE ELECTRIC UTILITIES. Disparities between existing and derived rates for private utilities are shown in Table 10-22. Prevailing rates are slightly higher than those derived using financial cost projections, but marginal cost-based rates are significantly higher in Luzon and Mindanao, reflecting the substantial marginal costs associated with the power purchases from NPC. In contrast, the marginal cost of power supply by the selfgenerating utility in the Visayas is adequately covered by its prevailing electricity rates. In regard to rate structures, Table 10-23 indicates that to suitably reflect the economic costs of power delivered by the utilities, demand charges should account for the bulk of revenues of firms connected to NPC, customer charges for more than 23 percent, and energy charges for as low as 16 percent in Mindanao to as high as 43 percent for the self-generating utility in Visayas. In practice, the private utilities apply a minimum charge representing customer-related costs as well as energy rate schedules that vary according to customer group. Only bulk industrial power users are subject to demand charges. In addition, rate schedules of certain private utilities such as MECO embody subsidized rates for relatively small power users.

Table 10-23. Allocation of Charges for Private Utilities

(percent)

Luzon Charge
Demand Energy Customer Source:
NPC

Mindanao

Visayas

Financial LRMC Financial LRMC Financial LRMC

15 57 28 52 22 26 42 40 18 61 16 23 29 61 10 30 43 26

data.

PnMIwPPES

197

Table 10-24. Average Tariffs for Electric Cooperatives, 1978

(pesosper kilowatt-hour) Financial Existing average average rate rate Ratio (1) (2) (2)1(1) 0.29 0.40 0.42 0.76 0.25 0.77 0.31 0.54 0.45 1.23 0.48 1.00 1.07 1.35 1.07 1.62 1.92 1.30 LRMC LRMC average versus finanrate Ratio cial costs (3) (3)1(1) (3)1(2) 0.43 0.49 0.48 0.58 0.49 0.91 1.48 1.22 1.10 1.24 1.96 1.18 1.39 1.09 1.02 1.53 1.02 109

Cooperative and region A-Luzon B-Luzon C-Luzon D-Mindanao E-Mindanao F-Visayas Source: NPC data.

In general, the sampled cooperatives are selling ELECTRIC COOPERATIVES. power at rates that allow them to break even. This is made possible because of the concessionary credit that the cooperatives have been able to obtain. As Table 10-24 shows, however, if a 12 percent margin were applied to the rate base, the average rate would have to increase significantly, except for coop A. A shift to rates based on marginal cost would also require price increases for all coops except for coop D, in which prices would decline. Clearly, considerable variation in the relation between financial and marginal costs exists among cooperatives. Table 10-25. Allocation of Charges for Electric Cooperatives
(percent) Coop A Charge Demand Energy Customer Financial 8 85 6 Coop D Charge Demand
Energy

Coop B Financial 37 46 16 Coop E Financial 60

23

Coop C Financial 34 50 16 Coop F Financial 25

65

LRMC 73 25 2

LRMC 54 22 24

LRMC 76 22 2

Financial 48
40

LRMC 62
31

LRMC 80
16

LRMC 22
76

Customer Source:Npc data.

12

17

10

198

CASE STUDIES

Table 10-25 shows that the cooperatives are similar to the private utilities, in that a distribution based on marginal cost shows a markedly higher value given to demand-related costs. Also, both maintain a distinct priority for recovering energy costs over customer-related expenses.

Policy Issues and Conclusions

Underlying the application of marginal cost pricing is the assumption that it will lead to a more efficient allocation of resources than under the prevailing electricity pricing structure. Whether this indeed will happen has been the subject of controversy. One obstacle is created by the fact that the industry is currently troubled by several price distortions because of policies adopted in response to peculiar needs of different groups of power consumers. Thus, capital is available at heavily subsidized rates for rural electrification, reduced fuel cost is offered to selected utilities, and some utilities are exempt from taxes and duties. In addition, prices at which electricity is sold vary considerably by levels as well as by structure depending on independently set criteria of each utility firm including established subsidy considerations. Even if marginal cost pricing can be strictly applied in the power sector, distortions in the broader energy sector will still have to be addressed. For instance, by policy, selected petroleum products have been made to bear most of the additional cost burden imposed by recent escalation in the price of imported crude oil. Consequently, fuel oil, a significant component of power generation, has enjoyed relative protection from price increases in relation to alternative oil-based forms of energy supply. Further, pricing policies in other sectors of the economy should be based on marginal costs to ensure that power consumers view electricity rates based on marginal cost in the right perspective, that is, comparable with prices of other commodities. Unless marginal costing is applied in other sectors of the economy affecting power development, the value of applying a marginal cost tariff in the power sector is reduced. At best, users of subsidized electricity will shift to less expensive forms of energy. As an example, the current price of power for residential customers in Manila using less than 200 kilowatt-hours of electricity a month is less than the cost of the fuel used to generate these kilowatthours. Consequently, homeowners prefer cooking with electricity than the higher-priced but economically cheaper liquid petroleum gas. Under a marginal cost rate consumers are made aware of the higher cost of electricity and will thus shift to other forms of energy for cooking, such as liquid petroleum gas, coal, and so on. A danger of strict LRIMC pricing is that it may price electricity out of the reach of the general populace. The high cost of pursuing rural electrification

PHMLIPPMNES

199

that is indicated by marginal costing would negate government efforts to use electrification as a way to achieve rural employment, income distribution, and industrial dispersion. Nevertheless, marginal costing may give valuable insights about the degree of subsidies and burdens placed on the various producers and users of power. To the extent that such measures are provided, government planners and policymakers may better appreciate the economic ramifications of their fiscal and pricing policies on the sector. The introduction of marginal cost pricing as an alternative way to set tariffs in the electric power sector is, therefore, considered timely in view of the recent emphasis given to development of the sector. The current policy of the Philippine government to accelerate power development and to intensify electrification activities has led to a closer investigation of existing pricing policies and to a search for an improved and more efficient mechanism to generate revenue for the sector. Adjustment of prices The concept of marginal cost pricing may not sell well with the public, particularly since costs are expected to increase over time. Moreover, consumers will not be amenable to being charged costs of future services now. On the contrary, it has always been clear that before any rate increases, better and wider service first should be provided. It is, therefore, particularly difficult to raise rates in areas where outages are frequent. Stable power supply to existing consumers may be a prerequisite to instituting marginal cost tariffs, which at the bulk level would have to be raised by an average of 80 percent. The derived marginal costs indicate that rates are expected to dramatically increase in the urban areas served by NPC. Levels will still be below those charged by self-generating private utilities, however. Although the urban populace is likely to be the most vocal against any tariff increase, they are generally better off than their rural counterparts and would have greater ability to adapt to price changes. Moreover, the urban users of power are more aware of factors responsible for cost increases in power production. They can be reminded that they place the heaviest pressure on the power industry. The low-income urban users of power would nevertheless still require some form of subsidy that may be financed by higher income groups, provided the subsidy would still price rower appropriately relative to alternative available fuels. In contrast, a large increase in power tariffs in the rural areas may eventually kill the market. Prevailing prices in coops serving these areas are already much higher than in urban areas. Fortunately, the adjustments indicated by marginal cost pricing are moderate for cooperatives and would not significantly increase rate levels, except for those already connected to NPC'S

200

CASE STUDIES

Mindanao power grid. For coops that will be connected to NPC in the future, lower price levels are anticipated even with the use of marginal cost pricing. Effect on finances of utility firms The application of marginal cost pricing at the various levels of power delivery will shift the flow of revenue funds away from the retail utility firms and toward NPC. The shift will reflect the relative concentration of investment activities in NPc during the coming decade. This will improve NPC'S capability to finance its numerous capital projects as well as its capacity to decrease its growing debts. At the same time, this will increase NPC'S income beyond the statutory limit of 10 percent return on the base rate. Revising the charter provisions on rates of return may not be too difficult, however, since NPC is basically a nonprofit institution, and additional funds are reinvested. Accordingly, using the marginal cost rate will mean reduced allocations from the national budget, releasing more funds for the development of other sectors of the economy. Although more funds will flow toward NPC, private utilities are expected to maintain their 10 to 12 percent profit margin. Sales revenues net of payment to NPC for power purchases will still yield sufficient reinvestment funds. The financial cash position of rural cooperatives may be eroded, however. The moderate increases in rate levels will be channeled to NPC; greater portions of existing revenues will also have to be diverted to NPC to pay the increased cost of power purchases. The cooperatives may improve their financial position if they minimize system losses that are currently double those of private utility finns. Investments required to improve system networks can be reflected in marginal cost tariffs. Adjustment for different voltage levels Under marginal cost pricing, tariff levels should reflect the losses incurred because of lower voltage deliveries. In general, rates of private utilities are such that lower voltage deliveries are charged higher rates and vice versa. For cooperatives, disparities caused by voltage differences are not apparent since most coops sell to LV users only. For NPC, present price differences between utilities and industries run counter to indications of marginal costing. Again, policies to support rural electrification and to protect infant electric cooperatives have caused preferential prices to be set for utility firms. This difference in tariffs will gradually be reduced as recent developments in the government's plans and policies tend to advocate the return to a single bulk price schedule. To reflect the savings realized because of various delivery voltages, NPC applies voltage discount adjustment clauses. In this connection the power

PHLIPPNES

201

and energy loss studies required by the marginal cost calculations have served as the basis for recalculating these discount values to more meaningful levels. Adjustmentfor different regions Marginal cost pricing witll reduce the differences in power tariffs among the principal regional power grids of NPC to obtain a national single wholesale rate. Likewise, a narrower range of retail prices (P0.43 to P0.53 per kilowatt-hour) is indicated for utilities connected to NPC'S grids. In contrast, on the basis of present tariff designs, the price differences among the large islands are expected to increase particularly as the cost of oil-based generation increases with each round of OPEC price announcements. Accordingly, the price in Visayas will exceed what consumers can afford, and prices in Luzon will also soar, but Mindanao grid tariffs will be maintained until the costs of new hydroelectric plants are included in the base rate. Although recent policies have minimized the differences in tariffs by allowing cross-subsidies in the allocation of increases in fuel cost burden, a single national rate is not anticipated until the island's grid facilities are physically integrated. Effect on income distribution policies The implementation of strict LRMC price levels may negate efforts of the government to make available to all Filipinos the economic opportunities afforded by electricity. Currently, less than 35 percent of all households in the Philippines enjoy electricity service, and most of these are in the more economically active areas of the country. To make electricity inaccessible to the rest of the population because of prohibitive costs induced by marginal costing will undermine the egalitarian principles of power development policies. Thus it may be necessary to modify relative rates to produce the subsidies required to make power accessible to low-income consumers. Policies for load development
NPC has subsidized the tariffs charged to power users in subgrids where the load has yet to be developed. This is the case for Bohol, Negros, and the Panay islands in the Visayas. Because the market is still at its nascent stage of development and capacity and energy use is low, pricing power at the strict cost of service would be prohibitively high. Under marginal cost pricing, tariffs in these islands would be much lower than that indicated by traditional costing, because underuse of existing power facilities would yield low capacity cost values. These prices are still higher than those established through the previously mentioned policies.

202

CASE STUDIES

Marginal-cost tariffs may be acceptable as a basis for such preferential prices, however. Dispersion of industrial development The government has been encouraging industries to move away from the urban areas toward new regional growth centers. On a larger scale it has promoted industrial development in Mindanao and Visayas as opposed to Luzon. Uniform pricing between urban and rural power supply will help bring industries to rural areas. Further, with rates in Visayas competitive with those in Luzon and Mindanao, and given the limited power supply in Luzon, more industries will likely consider Visayas as an alternative location. In contrast, Mindanao will lose its comparative advantage as a site for power-intensive export finns. With the new tariffs indicated by marginal cost for Mindanao, the exportation of expensive energy at low prices may warrant a review by policymakers. Capacity and energy costs In general, existing tariffs understate demand or capacity charges relative to energy- and customer-related tariffs. This tendency occurs because NPC'S bulk rates do not adequately reflect the high cost of peak demand. This is compounded by the tendency of coops and private utilities to classify all NPC energy sales as energy costs. For NPC,the change to a marginal cost structure would mean that demand price would represent 75 percent of total revenues in Mindanao, 72 percent in Luzon, and 55 percent in the Visayas. Considering that absolute average price levels will increase as well, the effective demand charges will represent dramatic increases over present designs. In Luzon, for example, this would involve raising the average charge per kilowatt more than twentythree times, from the current P5 per kilowatt to P 116 per kilowatt. Yet, a marginal cost structure indicates lowered energy-related charges, which would reflect the anticipated improvement in generation mix because of development of indigenous energy fuels. Consequently, energy charges would drop from P0.27 per kilowatt-hour to P0. 17 per kilowatt-hour in Visayas, from P0.10 per kilowatt-hour to P0.092 per kilowatt-hour in Luzon, and from P0. 10 per kilowatt-hour to P0.075 per kilowatt-hour in Mindanao. The marginal cost approach differentiates more meaningfully among regional energy price levels than do the current tariff structures. But lowered charges may not be practical since short-run marginal costs tend to rise rapidly in periods of oil price uncertainties. Maintaining energy charges at existing levels may be possible, gradually increasing demand charges over time. Yet, as price adjustments in the future

P-uupppiEs

203

are expected to coincide with fuel price increases, which are necessarily related to energy, the suggested adjustment strategy may not be acceptable to the public. Instead, increases in both demand and energy charges may have to be applied, and the demand charges would register larger hikes. This has been the trend in tariff design proposals for NPC. In addition to demand and energy charges, marginal costing points to a third charge: a customer service charge. This charge will range from an average of P2,230 per customer annually in Luzon, to P2,520 in Visayas, and to P2,600 in Mindanao. Alternatively, these charges can continue to be integrated with energy and demand tariff schedules inasmuch as their relative unit is minimal. Pricing by time of use At the moment time-of-day pricing by NPC is not yet warranted. Specifically, analysis of the Luzon system's daily load characteristics shows that high system load levels occur for as long as fifteen consecutive hours during the day with off-peak hours occurring between 2300 and 0500 hours. Moreover, the daily system load factor is as high as 87 percent. Consequently, any substantial reduction in peak load during the day can be achieved only by curtailing power from twelve to fifteen hours. Yet peak pricing duringyselected months of peak demand may be more suitable. During the summer months when hydroelectric power capacity is reduced because of low reservoir levels, outage rates in Luzon are high. Imposing a premium on the use of power supplied by Npc during these months may significantly reduce air-conditioning loads and may reduce or eliminate power outages in Luzon. Residential and commercial loads will have to be managed in addition to selected industries. A related survey indicated that the industrial group's potential to shift their activities to off-peak hours was limited. Specifically, most industries are alreadly operating at sustained levels of power demand lasting as long as from twenty to twenty-four hours with no discernible seasonal peak patterns. For The few industries with deferrable loads, the additional wage payments and supervision required are considered to be significant deterrents to shifting to off-peak hours. Policy recommendations The principal policy recommendations resulting from this analysis are: 1. Marginal cost pricing should be pursued, but phased-in over time. 2. Tariffs among the regions should be made more uniform within the decade, even before physically integrating the regional grids. 3. Retail price levels should be rationalized, particularly within the same grids and eventually throughout the country.

204

CASE STUDIES

4. Price differences by NPC between utilities and industries should be discontinued if they do not reflect differences in marginal cost. 5. NPC'S social pricing policy should be gradually eliminated, since retail utility firms are in a better position to implement such plans. 6. The use of subsidized power as incentive to industries, particularly export firms, should be limited; rather, the incentives should remain either fiscal or financial.

Chapter 11

SriLanka

THE CEYLONELECTRICITY BOARD (CEB) is a government corporation entrusted with developing and coordinating the generation, supply, and distribution of electric power in Sri Lanka. The CEB operates as a semiautonomous, commercially viable organization.

Organization of the Sector

Most of the better deve]loped sections of the country are connected to the transmission and distribution (T&D) system (Map 4). A few remote towns that are not connected to the CEB grid have their own diesel generating plant and distribution system, operated by the CEB or local authorities. Some local authorities purchase electricity in bulk from the CEBand operate their own distribution system. Large consumers are, however, supplied directly by the CEB.In 1969, the CEBembarked on a rural electrification program financed by grants from the central government. The foreign cost component of generation and principal transmission projects have been financed by the World Bank, the Asian Development Bank, and similar lending institutions. Local costs are financed from earnings of the CEBas well as from central government loans.
CEB

Past trends of supply and demand

The CEB system is predominantly hydroelectric as the following summary indicates.
Hydro capacity 329 megawatts, of which 16 megawatts are not operable in certain seasons. 50 megawatts

Thermal capacity (steam + diesel)

Note: This chapter was prepared in collaboration with K. K. Y. W. Perera, Nelson Wijemanne, and S. M. Chlisty (CEB).The views expressed in this chapter do not necessarily reflect those of the CEB or the government of Sri Lanka. 205

206

CASE STUDIES

11

--- I

,61j; T,inco

Anuradhapua I ii Puttalam a

--urune

'1415

In BaAlla

iyagala 5

-------

da Walawe

SRI LANKA

207

System peak demancd Daily load factor Total generation Total sales Losses as a percentage of generation

261 megawatts 53 percent 1,216.6 gigawatt-hours 1,040.5 gigawatt-hours 14.5 percent

Energy and power demand increased by approximately 6 percent from 1973 to 1977. Demand rose 14 percent during 1977 and is expected to continue to increase an average of approximately 11 percent per year. Currently, hydro power stations meet most of the energy and peak power demand; thermal stations are used during droughts and hydro plant outages. The typical daily and annual load duration curves (LDC) for the system are shown in Figures 11-I and 11-2. The peak period is quite pronounced from 1800 to 2200 hours, whereas the daily load factor is approximately 65 percent, which is equivalent to an annual load factor of about 54 percent. Forecasts of supply and demand The present and future supply-demand balance for the cEB system is summarized in Table 11-1 and Figure 11-3. The demand forecast is based on a review of six load forecasts made between September 1977 and January 1979. They include those of the Pereira Committee, Shankar, CEB, SOGREAH, PCR, and NEDECO. A detailed analysis of projections by consumer categorythat is, residential, industrial, commercial, rural, and so on-is also involved. Fairly high growth rates, averaging about 11 percent, are expected during the next decade. This should be accompanied by a slight improvement in the system load factor, since the main load growth will occur in the industrial category. The likely configuration of the CEB system and its consequences are best analyzed in the short-, medium-, and long-run context, as discussed below. The unforeseen and sustained rapid load growth after 1977 is attributable to government decontrol of the economy and to the fact that the principal hydro schemes are not scheduled to come onstream until 1985. Thus, serious shortages of electricity supply will occur before then. As shown in Table 11-1, the actual energy requirement will be 1,720 gigawatt-hours in 1980, with approximately 1,500 gigawatt-hours as an average load, assuming no forced outages. Adding gas turbine (GT) capacity is the only feasible way to augment the cEB's supply in the short time available. The government of Sri Lanka is considering the addition of three 30-megawatt GT units in 1980-81, followed by a steam unit by 1982-83. A 50-megawatt unit is advisable to meet energy requirements. After a rapid load increase during the early 1980s, the growth rate is expected to stabilize at 10 to 11 percent annually after 1984-85. Throughout the 1980s, the system will be energy constrained. Substantial thermal

208

CASE STUDIES

Figure 11-1.
300

Daily Load Curves for

CEB,

September 1978

24
260-

20, September ~~~Wednesday,

1978

240-

280
220 -

100Sunay,Weptnmesd17y,1Sptmbr208 ,200

17

12
Hours of the day

16

20

SRI LANKA

209

Figure 11-2. Synthesized Annual Load Duration Curve, 1978

100

- -------------

-----

50

1(0

Percentage of time

generation-up to 26 percent of total generation-will be required until 1985, when the Victoria hydroelectric project comes onstream. From 1985 to approximately 1990, several newly commissioned large hydro schemes should greatly reduce the need for therrnal generation. With the possible exception of dry years (see Table II- 1), the hydro additions will eliminate the need to use the gas turbines that incur high fuel costs. Currently, estimates for the hydro power potential of Sri Lanka are based primarily on a 1958 report by Pfeiffer that was updated by a United Nations Development Program (UNDP/PAO) study (1967-69). These two surveys concentrated on large hydro schemes, often ignoring smaller sites of from 15 to 20 megawatts. Also, some sites considered unveconomicat pre-1973 oil prices and not mentioned in these surveys may be more attractive today, given increased costs of oilfired generation. I Therefoare,a thorough hydro survey of the island and subsequent site investigatiornsshould occur as soon as possible. By the early 1990s, all planned and existing large hydro schemes will be fully utilized. In the early 1990s, some formn of base-load thermal generation also will be required, followed by more capacity addition to some of the existing hydro sites, such as Victoria, for peaking use. Coal-fired steam units
1. These remarks apply to rivers other than the Mahaweli, because the latter was extensively studied during the planning of the Mahaweli Project.

210

CASE STUDIES

Table 11-1.

CEB

Energy and Power Balance

1977 1978 1979 1980 1981 1982

Energy or power
Energy (gigawatt-hours) 1. Required generation 2. Growth rate (percent) 3. Hydro capabilitya 4. Hydro deficit 5. Thermal capabilityb 6. Total capability 7. Surplus (6. - 1.) Power (megawatts) 1. Peak demand 2. Hydro capacity 3. Thermal capacity 4. Total capacity 5. Margin (4. - 1.)' 6. Margin (100 x 5./1.) (percent)

1,217 7.4 1,500 -283 220 1,720 503 261 313 50 363 102 39

1,384 13.7 1,500 - 116 220 1,720 336 291 313 50 363 72 25

1,508 9.0 1,500 8 220 1,720 212 319 313 50 363 44 14

1,718 13.9 1,500 218 220 1,720 2 360 313 50 363 3 1

1,924 12.0 1,608 (B) 316 457 (GT) 2,065 141 403 353 140 493 90 22

2,145 11.5 1,752 (C) 393 457 2,209 64 445 383 140 523 78 18

a. Based on an average year. Firm gigawatt-hour hydro capability (that is, for a dry year) is unavailable, but would generally be about 15-20 percent lower; plants coming on stream are: B - Bowatenne (108 gigawatt-hours, 40 megawatts); C = Canyon (144 gigawatt-hours, 30 megawatts); K = Kotmale (376 gigawatt-hours, 200 megawatts); V = Victoria (680 gigawatthours, 195 megawatts); S = Samanalaweva (400 gigawatt-hours, 120 megawatts); R = Randenigala (300 gigawatt-hours, 100 megawatts); M = Moragahakande (150 gigawatt-hours, 40 megawatts). K/2 = half the capability of Kotmale. and so on.

appear to be an attractive option.2 Therefore, the possibilities of establishing long-term coal import contracts with nearby countries, such as india, and of training personnel to operate large steam stations should be pursued in the 3 1980s. Some smaller hydro sites of 15 megawatts still will be undeveloped by 1990. Since the annual growth of peak demand will be approximately 90 to 100 megawatts by then, however, many of these sites will also have to be developed. The skills acquired during the building of the large hydro
2. Another possibility is the direct burning of heavy crude oils, which are likely to be much more readily available than light crude during the next decade. For details of interactions between electricity and oil prices, see Mohan Munasinghe and Gunter Schramm, "Power and Energy Pricing Case Studies," Energy Department Report, World Bank, Washington, D.C., February 1980, chapter 3. 3. Since system peak demand will be less than 1,000 megawatts, nuclear stations with minimum economic unit sizes of 600 megawatts are likely to be too large until the twenty-first century. CEB will monitor this option and have a few trained personnel in this area, however.

SRI LANKA

211

1983

1984
2,643 11.0 1,940 (K/2) 703 754 (-D) 2,694 51 549 583 180 764 215 39

.1985
2,931 11.0 2,808 (V + K12) 123 754 3,562 631 597 778 180 958 361 60

1986

1987
3,582 10.5 3,352 (S + R)12 230 754 4,106 524 717 998 180 1,178 461 64

1988

1989

2,381 11.0 1,752 629 764 (ST) 2,516 135 494 383 190 573 79 16

3,242 10.5 3,008 (S/2) 234 754 3,762 520 655 898 180 1,078 423 65

3,958 10.5 3,658 (R/2 + M) 300 754 4,412 454 793 1,038 180 1,218 425 54

4,374 10.5 3,658 716 754 4,412 38 876 1,038 180 1,218 342 39

b. Thermal stations coming on stream are: GT = gas turbines (237 gigawatt-hours, 90 megawatts); ST = steam turbines (307 gigawatt-hours, 50 megawatts). - D = old diesel derated (10 gigawatt-hours, 10 megawatts). c. Largest units are hydro (50 megawatts up to 1984 and 67 megawatts 1985 onwards) and thermal (20 megawatts up to 1981, 30 megawatts up to 1982, and 50 megawatts after 1983). Source: CEB data.

schemes during the 1980s will be useful for these projects. The CEB plans to develop the capability for the parallel processing of many smaller hydro schemes by the late 1980s. Computation of the Strict LRMC

The principal marginal cost categories for long-run marginal cost (LRMC) 4 calculation are capacity, energy, and consumer costs. The first step in structuring LRMC is to select appropriate rating periods, such as peak and offpeak, by examining the typical system LDC shown in Figure 11-1. As discussed before, the peak: period during which demand presses on capacity appears to be from about 1800 to 2200 hours. If the duration of the peak period is defined more narrowly, peak-load pricing is likely to cause a shift from the original peak to the off-peak period (see Chapter 4). There is no marked seasonality of demand. On the supply side, however, spilling of
4. Unless otherwise stated, all costs are in constant 1979 prices.

212

CASESTUDIES

Figure 11-3.

CEB

Energy and Power Balance of Supply and Demand

1,200

I 1,000
POWER (RIGHT SCALE) _

Peakdemand \0X
4.4 w a

00

0~~~~~~~~~~~~~~~~~~0

3.6

S~~~~~~~~ydro

capabiity I/o
( S

2.4 0 ctD.2 _a

Total capabiliiy

2.0 -

Required energy

1977 1978

1980

1982

1984

1986

1988

1990

excess water is small and limited to a few months of the year. Therefore,

the complication of structuring LRMC by season would not appear to be justified, although this aspect should be studied further in subsequent tariff
revisions.

SRILANKA

213

Table 11-2. Peaking Hydro Capacity Costs

Total peak capacity costs = = = = = Rs3334 x 106in domestic prices (DP) Rs3334 x 106 x 0.836a in border prices (BP) Rs2787 x 106 390 x 103kilowatts Rs7146 per kilowatt

Installed capacity Cost per kilowatt Annuitization factor at 10 percent over 40 years Annual cost per kw annually

= 0.1023 = Rs7146 x 0.1023 = Rs731 per kilowatt annually (in 1985) or Rs731/(I.1) = Rs4l3 per kilowatt annually (discounted to 1979)

Note: High Victoria is used as a representative scheme. a. Cost item Domestic material Skilled labor Unskilled labor Fraction 0.13 0.10 0.10 100 x Conversion factor = 1.0 1.0 0.7 Total conversion factor = Weighted value 0.13 0.10 0.07 0.836

Source: CEB data.

Marginal capacity costs Since the CEB will be installing GT units in the early 1980s because of the short-run need for capacity, the costs of these units may be used as a proxy for marginal generation capacity costs. Assume GT costs of approximately US$275 for each kilowatt installed multiplied by a factor 0.1175, representing annuitization at a discount rate of 10 percent during a twenty-year lifetime.5 Then, multiplying again by a factor of 1.2, representing a 20 percent reserve margin, yields the value of Rs600 a year at the point of generation. The considerable hydro capacity with large reserve margins coming onstream after 1984 will displace the GT units that now serve as the reserve margin with consequent savings in fuel. Thus, the incremental cost of peaking hydro units in 1984 as calculated in Table 11-2, but without the reserve margin, must also be included in the capacity costs. Therefore, total generation capacity costs are Rs(600 + 413) = Rsl,013 per kilowatt per year. Although the later development of hydro is effectively added to the capacity costs of GT, a corresponding saving in fuel will occur that will be reflected in lower incremental energy costs, as shown later. Next, the LRMC of T&D is calculated. All T&D costs are allocated to incremental capacity because the facilities' designs are determined principally by the peak kilowatts that they carry rather than by the kilowatt-hours.
5. These costs are based on alconsultant's study of the GT project and on crB estimates.

214

CASE STUDIES

Table 11-3. Load, Demand, and Loss Forecast

(megawatts)

Year
1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988

Generation (gigawatthours) (1)

1,384 1,508 1,718 1,924 2,145 2,381 2,643 2,931 3,242 3,582 3,958

Load factor
(LF)

(2)
0.54 0.56 0.545 0.545 0.55 0.55 0.55 0.56 0.565 0.57 0.57

HV After HV conDemand demand 6 percent summers' after HV peak loss' peak demand consumption (3) (4) = 96 x (3) (5) (6) = (4) - (5)

291 319 360 403 445 494 549 597 655 717 793

274 300 338 379 419 464 517 561 616 674 745

13 13 21 28 35 35 35 35 35 35 35

261 287 317 351 384 429 482 526 581 639 710

a. Station use and HV transmission (66/132 kilovolts) b. Primary feeders and substation (11/33 kilovolts)

Table 11-4. Incremental Load or Demand Forecast

(megawatts)

Demand peak
HV MV LV

Demand peak increment

HV MV

Year
1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988
Discounted

(1)
291

(2)
235

(3)
93

(4)
-

(5)
-

319 360 403 445 494 549 597 655 717 793
-

258 285 316 347 386 434 473 523 575 639
-

96 102 111 120 131 148 155 168 181 200

-

28 41 43 42 49 55 48 58 62 76
-

33 27 31 31 39 48 39 50 52 64
-

totals Note: Increment with respect to previous year (derived from Table 11-3.). Not applicable.

SRI LANKA

215

After 10 percent
MV

lossb

(7)=9x(6)
235 258 285 316 347 386 434 473 523 575 639

LV demand Demand peak after consumers" after MV 13 percent peak demand consumption lossc (8) (9)=(7)-(8) (10)=87x(9) Mv

Lv

demand LV demand peak, peak, nonColombo Colombo (11) (12)

55.8 58.6 61.5 64.6 67.8 71.2 74.8 78.5 82.4 86.6 90.9 37.2 37.4 40.5 46.4 52.2 59.8 73.2 76.5 85.6 94.4 109.1

128 148 168 188 209 235 264 295 330 367 409

107 110 117 128 138 151 170 178 193 208 230

93 96 102 111 120 131 148 155 168 181 200

c. Secondary lines and distribution transforrners (230/380 volts)

Source: CEB data.

Demand peak increment

LV

(6)
3 6

Discount factor at 10 percent (7)

L.OOC 0.9091 0.8265 0.7513 0.6830 0.6209 0.5645 0.5132 0.4665 0.4241 -

Discounted increment
HV MV LV

(8) =(4) x (7)

28.0 37.3 35.5 31.6 33.5 34.1 27.1 29.8 28.9 32.2 220.5 (1981-87)

(9) =(5) x (7)

33.0 24.5 25.6 23.3 26.6 29.8 22.0 25.7 24.3 27.1 177.5 (1980-86)

(1O)=(6) x (7)
3.0 5.5 7.4 6.8 7.5 10.6 4.0 6.7 6.1 8.1 48.5' (1980-86)

9
9 11 17 7 13 13 19 -

a. Colombo = 15.8; Non-Colombo = 32.7. Source: cEBdata.

216

CASE STUDIES

T&D capacity costs may be estimated using the average incremental cost (AWC) method (see Chapter 4). The estimation of the incremental demand at the HV transmission (66/132 kilovolts), MV primary distribution (11/33 kilovolts), and LV secondary distribution (230/400 volts) voltage loads is shown in Tables 11-3 and 11-4. The corresponding incremental costs are summarized in Tables 11-5 to 11-7, and the AIC ratios are given in Table 11-8.
Marginal energy costs

Until 1985, the thermal plants will be used throughout the year to meet energy requirements in excess of the total hydro supply. Clearly, any incremental energy, irrespective of when it is required, will ultimately involve increased thermal generation. The cost of producing an additional kilowatthour with GT units is shown in Table 11-9. After 1985, there will be little thermal generation, and the incremental energy costs will be hydro costs estimated on the basis of hydro storage costs. Therefore, the LRMC of generating an additional kilowatt-hour is a mix of GT and hydro costs; the appropriately weighted average is RsO.32 per kilowatt-hour at the point of generation. The total LRMC of capacity and energy is summarized in Table 11-10, where adjustments for losses, O&M costs, and so on have been made. At

Table 11-5.

HV

Transmission Investment Costs Total costs (DP) (1)

62.06 91.22 69.34 152.56 246.30 334.84 343.74 -

((Rs x 106) - 66/132 kilovolts)

Year
1979 1980 1981 1982 1983 1984 1985 Discounted total for 1979-85

Total Discount costs' factor (BP) (10 percent) (2) = (1) X O.836 (3)
51.88 76.26 57.97 127.54 205.91 279.93 287.37 1.000 0.9091 0.8265 0.7513 0.6830 0.6209 0.5645 -

Total discounted costs (4) = (2) x (3)

51.88 69.33 47.91 95.82 140.64 173.81 162.22 741.71

- Not applicable. a. Same conversion factor (0.836) as derived in Table 11-2.

Source: CEB data.

SRILAINKA

217

Table 11-6. MV Distribution Investment Costs

(Rs X 106)

Year 1979 1980 1981 1982 1983 1984

1985

Urban costs (1) 119.92 244.6 22.40 21.95 72.20 72.62

66.38

Rural costs (2) 57.0 57.0 57.0 57.0 57.0 57.0

57.0

Total costs (DP) (3) 176.92 301.00 79.40 78.95 129.20 129.62
123.38

Total Discount Total costs factor discount (BP)a (10 percent) costs (4)=0.873x(3) (5) (6)=(4)X (5) 154.45 263.30 69.32 68.92 112.79 113.16
107.71

1.000 0.9091 0.8265 0.7513 0.6830 0.6209

0.5645

154.45 239.37 57.29 51.78 77.04 70.26

60.80

Discounted total for 1979-85

-

710.99

Notapplicable. a. Includescosts of consumersubstations and spur lines, that is, averagecost to meetall MV peak demand.
Cost item Foreign(15 percent duty) Domesticmaterial Skilledlabor Unskilledlabor
Source: cEBdata.

Fraction x 0.25 0.225 0.21 0.315 100 percent

Conversion factor 0.87 1.0 1.0 0.7 Totalconversionfactor =

Weighted value 0.218 0.225 0.21 0.22 0.873

any given voltage level (HV, MV, or Lv), consumers should only bear the burden of upstream incremental costs.
Consumer-related costs

Incremental consumer-related costs are considered the costs of providing service connections, reading, and maintaining meters, and billing. These costs are given in Table 11-11.

Adjustment of the Strict LRMC

The existing tariff (Table 11-12) is significantly different from the strict LRMC in both level and structure. First, current charges are well below the strict LRMC, and the greatest discrepancy is in the LV category. Second, the structure of tariffs by voltage level is distorted because the ratio of LV to HV

Table 11-7.

LV Distribution ((Rs x 101) - 230/400 volts)

Investment Costs Total costs non-Colombo (DP) (2)

-

Year
1979

Total costs Colombo (DP) (1)

2.50

Total costs Colomboa (BP) (3) =0.873 x (1)

2.22

Total costs non-Colombo' (4) 0.873 x (2)

-

Discount factor (10 percent) (5)

1.000 0.9091 0.8265 0.7513 0.6830
0.6209 0.5645

Total discounted costs (6) = (3) x (5)

2.22

Total discounted costs (7) = (4) x (5)

-

t980 1981 1982 1983

1984 1985

2.63 2.79 2.89 3.04

3.19 3.35

140 140 140 140

-

2.30 2.44 2.52 2.65

2.78 2.92

122.22 122.22 122.22 122.22

-

2.09 2.02 1.89 1.81

1.73 1.65

111.11 101.01 91.82 83.48

-

Discounted total for

1979-85
-13.4

Not applicable. a. Same conversion factor (0.873) as derived in Table 11-6. Source: CEB data.

SRILANKA

219

Table 11-8. Incremental Costs Summary (at Source)

(rupees) Item Generation
HV

Cost per kilowatt GT, Hydro, 7,147 742 x 106- 3,360 221 x 1O1 711 x 106
1,775 x 103 1

Cost per kilowatt per year 1,013 357

transmissiona

MV distributionb
LV

425 90

distribution-Colomlbob

13.4 x 106 850

15.8 x 103

(220/380 volts)

a. Allow two-yearlag betweeninvestmentsand incrementalmegawatt benefits.Annuitization factor = 0.1061(10 percent over thirty years). b. Allow one-year lag betweeninvestments and incrementalmegawatt benefits.Annuitization factor = 0.1061 (10 percent over thirty years). Non-Colombo costs, especiallyfor rural electrification, are too arbitraryto provideaccurateresults. Source: CEB data.

Table 11-9. Marginal Energy Costs

GT costs (usingfurnaceoil) Heat rate = 0.312 kilograms of fuel per kilowatt-hour Fuel cost = Rsl.40 per kilogram, Cost per kilowatt-hour = RsO.43per kilowatt-hour Hydrocosts (using storageor dam costs of Victoriaand Samanalawewa schemes) Total costs (DP) = Rs (1677 + 773) x 106= Rs2450 x 106 Total costs (BP)' = Rs2450 x 106 X 0.836 = Rs2048 x 106 Annuitizedcostsannuallyc= 2048 x 106 X 0.1023 = Rs209.5 x 106per year Energycapability = (400 + 680) = 1080gigawatt-hours per year Energycosts = (209.5 x 106) (1080 x 106)= RsO.19 per kilowatt-hour LRMC Year 1979-85 1986 + Costsper kilowatt-hour Kilowatt-hours 0.43 1 0.19 1

Discounted ratio of costs to kilowatt-hours is RsO.29 per kilowatt-hour. a. International price = $91 per ton. b. Same conversion factor (0.836)as derivedin Table 11-2. c. Annuitization factor = 0.1023 (10 percentfor forty years).
Source: CEB data.

220

CASE STUDIES

Table 11-10. Summary of Strict LRMC of Capacity and Energy (BP)

(rupees)

Capacity costs Generation Transmission

MV

distribution

Voltage level Generation

Annual Total Annual Total Annual Total (per (per kilowatt (per (per kilowatt (per (per kilowatt kilowatt) per year) kilowatt) per year) kilowatt) per year) 12,262
-

1,111 a
1,182 1,313 1,509

transmission (66/132 kilovolts)' MV distribution (11/33 kilovolts) LV distribution (230/400 volts)b

HV -

3,360 -

411' 457 525

4,010 -

5858 672

Not applicable. a. Includes 0.8 percent of total investment costs as annual O&M cost. b. Costs increased at each voltage level with respect to preceding voltage level by including losses. Losses as a percent of incoming are: Peak Off-peak Station use and HV transmission = 6 4 MV distribution 10 7
LV distribution =

13

c. Includes 1.6 percent of total investment costs as annual O&Mcost. d. Includes 4 percent of total investment costs as annual O&M cost.

Table 11-11. Standard Costs for Service Connections

(rupees)

Rating
Single phase, 15 amperes Single phase, 30 amperes Three phase, 30 amperes Three phase, 60 amperes Loop service, single phase, 15 amperes

Fixed cost
315 375 710 1,000 225

Variable cost per meter of the cost

9 12 22 33 11 570 each 600 each 400 each 315 each year.

Extra charges (all inclusive): Poles for a single-phase line Poles for a three-phase line Pole used as a strut Stay The average cost per customer for meter reading and billing is Rs37.50 per

SRI LANKA

221

Capacity costs
LV

distribution Energy costs (per kilowatt-hour) Peak

0.29

Aggregate Total Annual capacity costs (per (per kilowatt (per kilowatt kilowatt) per year) per year)
-

Off-peak
0.29

Total equivalent costs' (per kilowatt-hour)

-

1,111

1,593 2,355 2,847

0.31 0.34 0.39

0.30 0.32 0.35

0.79 0.71 1.49

850

141'

e. Includes 6 percent of total investment costs as annual O&M cost. f. Total equivalent cost per kilowatt-hour = [(Aggregate capacity costs per kilowatt per year)/(mf x 8760)] + [(peak energy costs) x (fp)] + [(off-peak energy costs) x (fo)]. Assumption
Hv Mv
LV

Mf
0.37 0.7 0.29

fp 0.2 0.2 0.5

fo

0.8 0.8 0.5

Source:

CEB

data.

charges is much less than the same ratio with respect to the strict LRMC. Third, the ratio of capacity to energy charges is much smaller than for the strict LRMC. Finally, there are undesirable features in the existing tariff, such as decreasing block rates, that discourage energy conservation. Although some of the discrepancies between the level of the existing tariff and LRMC will be reduced by adjusting the strict LRMC, the structure of the 6 present tariff clearly would have to be altered substantially.

6. A new tariff structure was being considered for implementation by the govemment in 1981. In light of the sensitive issues involved, details of the adjustment procedure and final proposed tariff were not available at the time of publication.

222

CASE STUDIES

Table 11- 12. Summary of Existing Tariff Consumer category

Domestic Religious premises Street lighting Bulk supply Less than 400 volts More than 400 volts Industrial or commercial Less than 400 volts and less than 50 kilovolt-amps

Monthly charges (rupees)

31 per kilowatt-hour (up to 100 kilowatt-hours) 0.21 per kilowatt-hour (more than 100 kilowatt-hours) 0.18 per kilowatt-hour 0.25 per kilowatt-hour 17 per kilovolt-amp and 0.20 per kilowatt-hour 15 per kilovolt-amp and 0.19 per kilowatt-hour

Less than 400 volts and less than 50 kilovolt-amps Less than 400 volts

10 (up to 10 kilovolt-amps) 5 per kilovolt-amp (less than 10 kilovolt-amps) and 0.36 per kilowatt-hour 19 per kilovolt-amp and 0.25 per kilowatt-hour 17 per kilovolt-amp and 0.24 per kilowatt-hour

Very large consumers and nongrid consumers can negotiate special rates. A fuel surcharge is applicable when thermal generation is required.
Source: CFBdata.

Chapter 12

Thailand

includes three state enterprises. The Electricity Generating Authority of Thailand (EGAT) is responsible for both generation and transmission. The Metropolitan Electricity Authority (MEA)distributes electricity to the Bangkok metropolitan area, serving about 5.9 million people. The Provincial Electricity Authority (PEA)distributes electricity to the rest of the country, serving approximately 38 million people.
THE POWER SECTOR OF THAILAND

Organization of the Sector

Several government agencies directly or indirectly influence the operations of the three authorities, including the Ministry of Finance, Office of the National Economic Social Development Board, the Budget Bureau, National Energy Administration, and the Power Policy Committee. Members of the Power Policy Committee, which is chaired by the Deputy Prime Minister, represent the three electric utilities and some of the government agencies. The committee has an active and direct role in policies relating to electric power development and tariffs. Proposals for revising power tariffs must be reviewed and recommended by this committee and passed on to the cabinet for approval. EGAT'S electric power system can be divided into four regions. Regions 1, 2, and 4 cover the central, northeastern, and northern regions, respectively. These are interconnected by 230-kilovolt, 115-kilovolt, and 69-kilovolt transmission lines. Region 3 in the southern region is still isolated from the others. The 115-kilovolt central-southern tie line linking Region 3 with Region 1 was constructed in 1980 (Map 5). MEA and PEA buy most of their electricity supply from EGAT. At the time of this study, PEA still owned and operated 308 small diesel plants that served areas not yet within the reach of EGAT's transmission grids, including many remote village areas.
Note: This chapter was prepared in collaboration with the Electricity Tariff Committee of Thailand. The views expressed in ibis chapter do not necessarily reflect those of any Thai power sector organizations or the government of Thailand.

223

BURMA 20'

~Ch rj~.J%) ~ i a
DamL_ j Chiangmai REGION

J-'

<. -r

MAPS.

Thailand: Principal PowerStations

and Transmission Lines

in the EGATSystem

Mh M

O
j

/
NamNg
>

Dag am Mong

LAO PEOPLE'S DEMOCRATIC Pg B

RGIO4 Dnam
Bhumibol

b
,QYai

Udonthani
UboIlantan

E~~~~~~~~~~N
Na P
Da

\&WREGION

anD\

Thajan

224

THAILAND

225

The objectives of this tariff study are to analyze the incremental costs of generating, transmitting, and distributing electricity at different places, times, and voltage levels to different consumers of the three energy authorities during fiscal years 1978 to 1986; to examine to what extent the existing tariff structure reflects costs; to derive a tariff structure that reflects as closely as possible the marginal costs to the economy of meeting the demand for electricity-subject to certain social, economic, and political constraints; and to analyze the problems, advantages, disadvantages as well as the applicability of the marginal cost pricing, using a calculated cost structure to simplify the existing tariff structure in a way that would be fair to both utility and consumer groups. This study analyzes the economic costs of electricity and determines the level and structure of tariffs and metering on the basis of long-run marginal costs (LRMC). Financial, social, and other factors are introduced in such a way as to minimize the distortion of the ideal tariff. The economic analysis of costs and tariffs involves the following five procedures: a. The incremental costs of generating, transmitting, and distributing electricity are estimated according to different places, times, and voltage levels to different consumers during fiscal years 1978 to 1986. Most relevant are the economic costs to the various Thai power authorities of expanding and operating their systems to meet the projected demand. Shadow prices for capital, labor, and foreign exchange, rather than their actual market prices, are considered appropriate for measuring costs. Any taxes on sector inputs are deducted. b. The existing tariff structures are examined and are compared with the structure of marginal costs of supply. c. A set of proposals for changing the existing tariff system is formulated to correspond to the marginal cost of electricity. d. The advantages and problems of implementing the system of marginal cost pricing are assessed. e. The existing tariff is revised for simplicity and equity by using the calculated marginal cost structures as a guideline.
Electricity supply

In 1977 the total installed generating capacity of the EGAT'S system was 2,742 megawatts. This consisted of seven hydro plants supplying 909 megawatts, five thermal plants (oil- and lignite-fired) producing 1,633 megawatts, five gas turbine plants yielding 165 megawatts, and five diesel plants giving 35 megawatts. EG2AT'S transmission system serves customers throughout the country and consists of 230-, 115-, and 69-kilovolt lines. In the MEA'S distribution system, the primary and secondary networks are essen-

226

CASE STUDIES

tially the overhead radial type. These are standardized as both 12 and 24 kilovolts in the primary distribution system, and 400/230 volts in the secondary distribution system. PEA'S subtransmission and distribution systems are predominantly overhead radials using 11, 22, and 33 kilovolts in primary distribution and 400/230 volts in secondary distribution. In fiscal year 1977 the total energy supply was 11,048.2 gigawatt-hours. EGAT'S share of generation amounted to 10,777.3 gigawatt-hours or approximately 97.55 percent, the rest being supplied by isolated PEA diesel plants and some energy purchased from Laos. The percentage of energy generated by the various types of EGAT's generating plants were:
Hydro plant Oil-fired plant Lignite-fired plant Diesel plant Gas turbine plant Total 30.58 65.51 2.71 0.36 0.84 100

In the future natural gas will be available for power generation, and thus some of the oil-fired plants can be replaced largely by natural gas. In the long term coal and perhaps nuclear generation will become more significant. Electricity demand During the past ten years, electricity consumption in Thailand increased five-fold, growing at an average rate of 17 percent per year. The rate reached a high of 25 percent in 1968/69 and a low of 5.3 percent in 1973/ 74, the year of the energy crisis and conservation measures. In the MEA area sales grew 15 percent during the period. The high was 19 percent in 1972/73 and the low 1.7 percent in 1973/74. In the PEA area, the average growth was 24 percent during 1967 to 1976. By the end of fiscal year 1977, approximately 9,640.36 gigawatt-hours of electric energy were consumed in Thailand, with a peak load of 1,909 megawatt-hours. The numbers and distribution of the customers are shown in Table 12-1. Energy consumption for each consumer type is shown in Table 12-2 as a percentage of total national consumption. System characteristics The variations in monthly maximum demand and energy from October 1976 to September 1977 are given in Figure 12-1. The highest maximum demand in the year occurs in July. Daily load curves for the EGAT interconnected system plus the isolated region 3 are also given in the same diagram. Seasonal variations in consumption are not significant. For EGAT'S current system, the base load is generally provided by the oil-

THAILAND

227

Table 12-1. Number of Customers of the Electric Power Sector, by Type of Customer
Customer category Residential Smallbusiness Mediumand largebusiness Others Streetlighting Total
MEA PEA EGAT

Total 1,575,584 117,729 5,500 2,648 2,694

Percent 92.46 6.90 0.32 0.16 0.16

451,657 1,122,927 103,150 14,579 4,567 916 2a 2,646b 619 2,079 559,995 1,144,143

17 17

1,704,155 100

Not applicable. a. Off-peak rate consumers. b. Water works, agricultural pumping, and temporary consumers. Source: EGAr data.

Table 12-2. Energy Consumption, by Type of Customer

(gigawatt-hours)

Percentageof Customercategory
Residential Smallbusiness Mediumand largebusiness Others Streetlighting Total
MEA PEA EGAT

national consumption 18.77 13.92 64.81

11.16 8.80 42.09 0.42a 0.34 62.81

7.61 5.12 17.97 1.3 3b 0.41 32.44

4.75 4.75

1.75
0.75 100

Not applicable. a. Off-peak rate consumers. b. Water works, agricultural pumping, and temporary consumers. Source: EGAT data.

fired thermal plants. The small hydro and lignite plants are operated more or less continually. The Sirikit hydro plant has a 375-megawatt capacity and operates at a constant intermediate level during the daytime; the Bhumibol hydro plant has a 420-megawatt capacity and supplies the balance of the load. The spinning reserve during the peak is provided by the unused capacity of these two main hydro plants. Existing tariffs
EGAT's bill supply tariff applies to MEA, PEA, and direct industrial customers with different tariff levels. It costs EGAT more to supply electricity to PEA areas than it does to supply MEA. Given longer distances and less load density, however, the bulk tariff charged to PEA iS much lower than that

Figure 12-1. Characteristics of the

DAILY LOAD CURVE 2,0001 _

EGAT System

tl,600

-<

01 7)

E
cl)

800 -Sunday 400 -

12

16

20

24

Hours of the day

MONTHLY PEAK DEMAND AND OPERATION

-v,

Oct Nov Dec

Jan

Feb

~~~~~~~Months f Mar Apr |MayfJune July

Aug

Sep

1,800 _

1,700_

00

81

ca

600

...

50CZ
Note: IDPurchased from
EDL

(Laos),

*diesel
228

and gas turbine,

hydro, and Qthermal.

THAILAND

229

charged to MEA. This is because of the government policy designed to equalize the residential tariffs between PEA and MEA. To cover the reduced PEA revenue, EGAT has been required to subsidize PEA from the proportionately greater revenues obtained from MEA and its direct industrial customers. Apart from its residential tariff, PEA's retail price levels for other consumer categories are higher than those charged in the metropolitan areas served by MEA. Both wholesale and retail tariffs are the block rate type. There is an increasing block pattern for residential consumers, and a declining block type for other categories. The rate for the demand and energy charge of some business and industrial consumers is also of the declining block form. Energy and demand forecasts Two different methods of forecasting are used in the load projection. The first is based on the historical growth of consumption and on the known project proposals of industry that will be established in each utility area. The second is based on a correlation analysis between historical consumption and economic indicators that double checks the accuracy of the forecast. Although there are some differences between the two methods, the investment necessary to meet the forecast load would not be greatly affected by these differences. The load projections from the first method are used here. For 1978 to 1986 energy consumption for all categories in the MEA area is expected to grow 7.88 percent per year. The comparable PEA figure, including rural electrification load, is 14.28 percent. EGAT'S direct customers, large industrial establishments, are expected to increase their energy consumption by 4.69 percent annually. Total EGAT sales are the sum of the estimated requirements of MEA and PEA as well as of direct customers. The electric energy and demand requirements of the whole power sector in Thailand for 1978 to 1986 are estimated to grow 10.33 percent and 10.5 percent annually, respectively. The forecast up to 1986 of EGAT'S total generation requirement and EGAT'S total sale to MEA, PEA, and direct customers is summarized in Table 12-3. Power development plan
EGAT'S power development plan was formulated to supply adequate power and energy to meet the load forecast. For 1978 to 1986, the principal plant additions will consist of 1,013 megawatts of hydro plants and 1,667 megawatts of thermal plants with lignite, oil, natural gas, or nuclear plants as well as 240 megawatts of gas turbine. MEA's power distribution development program for 1978 to 1986 includes expansion of the subtransmnission and distribution systems to the outskirts of the existing service area. It also includes system reinforcement and expansion of both the primary voltage and the low-voltage systems.

Table 12-3. Sales of Power and Energy by EGAT to Consumers

EGAT

Year
Power (megawatts) 1978 1979 1980 1981 1982 1983 1984 1985 1986

MEA

PEA

Direct

Total

requirements

Percentage increase

Load factor (percent)

1,158.02 1,250.57 1,347.53 1,450.08 1,556.70 1,668.40 1,785.78 1,908.84 2,038.16

790.20 928.40 1,087.90 1,301.50 1,515.50 1,763.90 1,978.30 2,204.30 2,431.60

88.50 95.65 102.20 110.25 112.85 115.48 118.12 120.80 123.50

2,036.72 2,274.62 2,537.63 2,861.83 3,185.05 3,547.78 3,882.20 4,233.94 4.593.26

2,098.0 2,343.0 2,614.0 2,948.0 3,274.0 3,636.0 3,966.0 4,314.0 4,667.0

11.99 11.68 11.57 12.78 11.06 11.06 9.08 8.77 8.18

t-i

Average increase of total peak demand (1977-86) is 10.90 percent. Energy (gigawatt-hours) 1978 1979 1980 1981 1982 1983 1984 1985 1986 6,890.05 7,484.03 8,111.60 8,774.64 9,475.63 10,216.59 11,000.43 11,824.77 12,707.58 3,863.20 4,538.70 5,275.00 6,169.00 7,146.90 8,176.60 9,088.10 10,028.90 10,991.80 482.80 516.00 554.50 603.62 625.75 642.19 658.96 675.67 692.56 11,236.05 12,538.73 13,941.10 15,547.26 17,248.28 19,035.38 20,747.49 22,529.34 24,391.94 12,135.0 13,542.0 15,056.0 16,748.0 18,631.0 20,533.0 22,361.0 24,265.0 26,243.0 11.60 11.59 11.18 11.24 11.24 10.21 8.90 8.51 8.15 66.03 65.98 65.75 64.85 64.96 64.47 64.36 64.21 64.19

Average increase of total peak demand (1977-86) is 10.33 percent.

-

Not applicable.

Source: EGAT data.

THAILAND

231

PEA'S power distribution development plan is made for the short and long range. The latter includes the rural electrification program for the period. The investment program consists of reinforcing and extending the existing system, revamping diesel plants, and electrifying 12,000 villages.

Computation of the Strict LRMC

The disaggregated calculation of strict LRMC is described in this section. Marginal costs for generation and transmission by
EGAT'S EGAT

system of power generation is very complicated, and marginal costs for MEA, PEA, and direct industrial customers will be different because of varying locations and voltage levels. Therefore, the model shown in Figure 12-2 simplifies the calculations. All generation is assumed to be from a power plant and represents the aggregation of various plants. System pools are used to represent the common transmission network shared by various EGAT consumers. In the model, separate transmission systems-to MEA and to PEA and the direct industrial customers-are assumed to go out from the pool to the corresponding consumer area. The peak demand used to calculate the marginal cost is obtained from the Figure 12-2. Diagram of the
EGAT

System

Aggregated generation of various

plants
Point of generation requirement Part of the HV interconnected system Point of demand at pool

EE1~~~~~~~~~4 11
Delvery point for I E
I

Delivery point for and direct customers

PEA

232

CASE STUDIES

total peak generation requirements minus the associated power loss in the transmission grid considered as a pool. From the past records and current experience it is clear that overall power and energy losses within the system, including station services, are 3 and 8 percent for MEA and PEA, respectively. Losses are allocated to various parts of the system. Pool power losses that involve transmission line losses and station services have been estimated at 5 percent of generation requirements. This loss is common to MEA, PEA, and direct customers. According to past load flow analysis, power loss from the pool to MEA'S and PEA's delivery points is estimated at 2 percent and 6 percent, respectively. The percentage of peak energy loss is considered equal to that of the peak power loss. Estimates of off-peak energy losses are 1.46 percent from pool to MEA's delivery point, and 3.4 percent from pool to PEA'S and direct customer delivery point. EGAT'S marginal costs of providing for incremental electricity demand and energy are calculated separately for MEA and PEA areas as well as for the direct customers. The two marginal costs estimated are capacity and energy costs.
MARGINALCAPACITYCOSTS. EGAT'S marginal capacity costs are considered to be those incidental to increasing generation and transmission capacity, plus incremental costs of operation (excluding the fuel cost) and 70 percent of total maintenance and administration for 1978 to 1986. The investment cost was estimated at 1978 prices, excluding inflation, government contribution, import duty, and interest during construction.' In estimating the transmission investment, all network costs have been lumped together. The length and cost of transmission line extensions required to serve the increasing demand, however, are quite different for MEA and PEA as well as for some industry consumers. So EGAT'S transmission costs have not been directly allocated to MEA, PEA, and industrial consumers. Given that demand density in each area is quite different, it is not possible to relate the transmission investments to the increasing demands of MEA and PEA separately. For the maintenance cost, which partially depends on the energy generated, 70 percent of total incremental maintenance and the administrative costs are considered the capacity component. The operation and maintenance (O&M) and administrative costs are assumed to be correlated with the amount of additional energy generated. Then the estimated additional maintenance can be obtained by using past experience regarding maintenance cost per kilowatthour as a multiplier against the additional kilowatt-hour of energy generation for each year.

1. The government contribution refers to the irrigation component of multipurpose water projects estimated to be equal to the economic benefits of irrigation.

THAILAND

233

First, the average incremental costs (WC) for every added kilowatt are calculated (see Chapter 4). The full marginal capacity costs for MEA and PEA as well as for direct industrial consumers are obtained by adjusting AIC with the power loss multiplier during system peak to cover the transmission losses. The results of calculation are shown in Table 12-4.
MARGINAL ENERGY COSTS. The basic characteristics of EGAT'S operating schedule to meet the load changes is similar throughout the year. The thermal plants are operated for base-load generation fairly steadily for the three daily periods, although hydro plants are operated more or less continuously. The marginal energy cost calculated consists of the fuel cost, to which is added 30 percent of maintenance cost for future plants. Marginal energy costs of the EGAT connected system varies slightly from year to year depending on the hydrological conditions. Thus, the marginal energy costs are estimated as average fuel costs from all kinds of fuel expenses that are used in the various plants and the additional expenses of total system energy sales. Most of the fuel oil used by EGAT'S plants comes principally from a local refinery, but some has to be imported, for which EGAT pays a higher price. The weighted average fuel consumption for all thermal plants is 0.26 liter per kilowatt-hour. Using this result, the expected value of long-term marginal energy cost for MEA and PEA as well as for direct customers including the peak and off-peak loss are as follows, in baht per kilowatt-hour:

Peak period
MEA
PEA

Off-peak period
0.3691 0.3762

and direct customers

0.3711 0.3856
MEA

Marginal costs for distribution by

and

PEA

The marginal costs for

MEA

and

PEA

distribution are incremental costs for

EGAT

Table 12-4. Marginal Capacity Costs for

(baht per kilowatt)

Cost for
Generation at pool
MEA' PEA

Generation Transmission cost cost (per year) (per year)

1,951.68 1,990.71 2,068.78 92.73 215.83

O&M and Total A&G costs (per year) Per year Per month
24.91 29.27 1,976.59 2,112.71 2,319.98 164.72 176.06 193.33

and direct customersb

Not applicable.

35.37

a. From pool to MEA delivery point, 2 percent of power is lost. b. From pool to PEAand direct customers at delivery point, 6 percent of power is lost.
Source:
EGAT data.

234

CASE STUDIES

the expansion of their distribution systems and the reinforcement of their subtransmission and distribution systems to meet the expected load growth. Marginal costs are calculated at two and three different voltage levels at which consumers are supplied for MEA and PEA. The estimated investment at different voltage levels is also the economic cost excluding inflation, government contribution, import duty, and interest during construction. The marginal capacity costs of the distribution system vary greatly from one area to another because of the considerable variation in load density. Some areas require substantially more circuit-kilometers of line for every consumer and for every kilowatt of demand than others. Reinforcement costs usually are higher than expansion costs. Therefore, an average incremental cost for the study period must be determined in the same way as EGAT'S marginal cost calculation. By relating the estimated increase in demand to the corresponding investments required at different voltage levels for 1978 to 1986, the average incremental capacity cost for every kilowatt of demand at each voltage level can be obtained. This is annuitized over a thirty-year period at 11 percent per year, and annual maintenance and administration costs are added to obtain the annual marginal capacity cost per kilowatt at each voltage level. The full costs at the various voltage levels are obtained by using loss multipliers during system peak in each voltage stage. MEA and PEA marginal capacity costs are estimated below.
MARGINAL CAPACITY COSTS FOR MEA. MEA'S marginal capacity costs consist of expenditures for reinforcing and expanding the distribution system, including the incremental O&M and administration costs for 1978 to 1986. The costs are calculated at three different voltage levels: 69 kilovolts, 12-24 kilovolts, and 220-380 volts. The increasing demands for each consumer voltage level at system peak are the integrated demands of each consumer category as estimated by load research. The estimated operation and administration costs are based on past experience and are estimated at 10.3 percent of the additional investment. The power-loss factors at each voltage level used in this study are: 1.01 14 at 69 kilovolts; 1.0114 at 12 and 24 kilovolts; and 1.0871 at 220 and 380 volts. The results of the calculation are summarized in Table 12-5. MARGINAL CAPACITY COSTS FOR PEA. The marginal capacity costs for PEA are also calculated on the basis of their investment program for 1978 to 1986. Given considerable variations in load density, the long distribution lines required per customer, as well as the high equipment and operation costs, it is clear that, within the study period, the marginal cost for each project will be different and subject to much more variation than that of MEA. Although the method of calculation is the same, the annual operation and administrative costs are estimated on the basis of each project. Given the length of the distribution line and the load density in PEA areas, there is a

THAILAND

235

Table 12-5. Marginal Capacity Costs for

(baht per kilowatt)

MEA

Voltage level
69 kilovolts 12-24 kilovolts 220-380 volts Source: EGATdata.

Capital cost (per year)

37.31 282.84 1,015.38

O&M and A&G costs (per year)

3.84 29.14 104.59

Total Per year

41.15 311.98 1,119.97

Per month
3.43 26.00 93.33

Table 12-6. Marginal Capacity Costs for

(baht per kilowatt) per year)

PEA

Voltage level
High voltage (11-33 kilovolts) Low voltage (220-380 volts) Source:
EGAT

Capital cost
384.38 1,578.37

O&M and A&G costs

42.39 89.67

Total
426.77 1,668.04

data.

higher percentage of subt-ansmission and distribution losses than in the MEA system. The losses in PEAk'S low voltage system are assumed to be 12 percent. The costs are calculated at two different voltage levels: the low voltage (220-380 volts) and the high voltage (11-33 kilovolts). The results of the calculation are summarized in Table 12-6.
CONSUMER-RELATED COSTS. Incremental consumer-related expenses include the costs of reading and maintaining meters, as well as billing. The consumer and the utility agree on the amount that the consumer should contribute to cover all or some percentage of the cost of a new connection. This includes extensive service connections, distribution transformers, meters and meter installation. Some administrative costs for metering and billing are lumped together in the total administrative cost. Estimates of consumerrelated costs are based on the records of the average cost of reading and billing for both the MEA and PEA systems in 1976 and the cost of types of different meters. The consumer-related costs thus calculated are shown in Table 12-7.

Shadow pricing

To compensate for the effect of distortions on the economic resource costs

236

CASE STUDIES

Table 12-7. Consumer-Related

(baht per kilowatt)

Costs for

MEA

and

PEA

Total cost Consumer category

Residential Small business 220-380 volts 11-33 kilovolts Large business and industrial 220-380 volts 11-35 kilovolts 69 kilovolts
Not applicable.
MEA PEA

17.31 32.15 316.01 109.67 334.38 354.38

15.97 29.87 313.73 103.90 348.61 -

Source: EGAT data.

of inputs used to produce electricity, the appropriate shadow pricing is applied to the relevant inputs. The shadow prices of inputs are converted to border prices, then the border-priced marginal costs must be divided by a standard conversion factor to arrive at the domestic-priced marginal costs. In the calculation, the directly imported inputs converted at the official exchange rate for the c.i.f. value are already in border prices. The domestically purchased inputs have been converted to the border prices by a conversion factor of 0.905. A conversion factor for unskilled labor of 0.685 is used at EGAT and MEA. The factor for skilled laborers at MEA is 0.957. The standard conversion factor used is 0.926.
Summary

The long-run capacity and energy marginal costs of each utility computed with and without shadow prices are summarized in Tables 12-8 and 12-9. For the total marginal costs of MEA and PEA, their distribution capacity costs were combined with EGAT'S capacity costs and also included transmission and distribution power losses at each voltage level. Adjustment of the Strict LRMC

Preliminary pricing structures for the various categories of consumers are described below.
EGAT The ideal tariffs based on marginal cost concept can be charged during two pricing periods: peak and off-peak. During the off-peak period, a consumer is charged the basic marginal energy cost per kilowatt-hour used.

THAILAND

237

Table 12-8. Total Marginall Cost of

(with Shadow Prices)
MARGINAL CAPACITY COST (bahlt per kilowatt per month)

EGAT, MEA,

and

PEA

Utility
EGAT

Consumer category
MEA PEA

Marginal Marginal generation transmission cost, or or distribupurchase tion cost

168.95, 175.58, 175.58, 180.80b 182 .8 6b 19 8 .7 8b
199.49b

O&M and A&G costs

2.50 3.01 3.01 0.43 2.69 8.86 3.53 7.49

Total marginal capacity cost

178.75 195.58 195.58 185.44 211.65 293.60 235.83 365.62

7.30 16.99 16.99 4.21' 26.10, 85.96, 32.81, 134.70,

Direct consumers
MEA

69 kilovolts 12-24 kilovolts 220-380 volts

11-33 kilovolts

PEA

220-380 volts MARGINAL ENERGY COST (baht per kilowatt)-hour)

223.43b

Utility
EGAT

Consumer category
MEA

Peak period
0.3834
0.3985

Off-peak period
0.3814
0.3887

PEA

Direct consumers
MEA

0.3985 0.3878 0.3922 0.4264 0.4066 0.4620

HV loss,

0.3887 0.3846 0.3878 0.4120 0.3931 0.4213

69 kilovolts 12-24 kilovolts 220-380 volts

11-33 kilovolts 220-380 volts

PEA

a. Including loss due to station service, b. Including subtransmission. c. Including distribution loss. Source: EGAT data.

and loss from pool to delivery point.

In the peak period, a consumer is responsible for.both kilowatt-hour usage and kilowatt demand as reflected in the basic marginal energy and capacity charges, respectively. Each pricing period is determined from the typical system's daily load curve. The seasonal variation is insignificant, and the evening load during the weekend is only slightly different from that of the working day. There-

238

CASE STUDIES

Table 12-9. Total Marginal Cost of (without Shadow Prices)

MARGINAL CAPACITY COST (bahtper kilowattper month)

EGAT, MEA,

and

PEA

Utility
EGAT

Consumer category
MEA

Marginal Marginal generation transmission cost, or or distribupurchase tion cost

165.89a

O&M and A&G costs

2.44

Total marginal capacity cost

176.06

7.73

PEA

172.40a 172.40a
178.05b

17.98 17.98
3.11c

2.95 2.95
0.32

193.33 193.33
181.48

Direct consumers
MEA

69 kilovolts

12-24 kilovolts 220-380 volts

PEA

180.08b 195.76b 197.20b

220.86b

23.57' 84.61c 32.03c

131.53'

2.43 8.72 3.53

7.49

206.08 289.09 232.76

359.88

11-33 kilovolts
220-380 volts

MARGINAL ENERGY COST (baht per kilowatt-hour)

Utility
EGAT

Consumer category
MEA PEA

Peak period
0.3711 0.3856

Off-peak period
0.3691 0.3762

Direct consumers
MEA

0.3856 0.3753 0.3796 0.4127 0.3925 0.4472

HV

0.3762 0.3722 0.3853 0.3987 0.3804 0.4077

69 kilovolts 12-24 kilovolts 220-380 volts 11-33 kilovolts 220-380 volts

PEA

a. Including loss due to station service, b. Including subtransmission. c. Including distribution loss.
Source:
EGAT data.

loss, and loss from pool to delivery point.

fore, the pricing structure does not vary by season. For each daily pricing period, the peak period lasts from 1830 to 2200 hours, and the off-peak period lasts from 2200 to 1830 hours (of the following day). The tariff for industrial direct customers is not the same level and structure as that of bulk supply rates.
MEA AND PEA In setting the ideal tariffs to reflect the marginal cost, the charges would cover the three components of the costs: a basic energy

THAILAND

239

charge covering the eneirgy purchased from EGAT; a basic demand charge covering the kilovolt demand rates of EGAT; and a basic demand charge covering distribution capacity cost of MEA or PEA. For the residential and small business consumers whose average consumption per customer is small, it would be more complicated and too expensive to introduce time-of-day metering instead of a single kilowatt-hour meter. Medium- and large-size businesses and industrial consumers, however, would be charged according to the time of day. The pricing periods for both MEA and PEA tariffs would be the same as that defined in the EGAT tariffs. A new tariff structure that would reflect the marginal costs would have the following general form.
RESIDENTIAL CONSUMERS There would be a monthly fixed charge to reflect consumer-related costs anid three blocks for energy charges. The first block, covering consumption units from 0 to 30 kilowatt-hours per month, would be charged at a relatively low lifeline rate. The second block, covering the amount of consumption from 31 to 500 kilowatt-hours, would be charged at a rate that fully reflects its marginal cost. The amount of energy beyond 500 kilowatt-hours per month would be charged at a relatively high rate to cover its marginal cost and the subsidy of the lifeline rate. SMALL BUSINESS CONSUMERS (demand of less than 30 kilowatts). There would be a monthly fixed charge to reflect consumer-related costs and an energy charge at a flat rate. The energy charges are calculated on the basis of marginal cost at the low voltage level and 12, 22, or 24 kilovolts. Some discounts would be given to consumers supplied at these voltage levels. MEDIUM AND LARGE BUSINESSES AND INDUSTRIAL CONSUMERS (demand of 30 kilowatts or higher) Time-of-day metering would be appropriate for these consumers. The rnonthly fixed charge reflecting consumer-related costs should be supplied at 220 volts, 12-22 kilovolts, and 69 kilovolts. During the peak period, itherewould be a demand charge based on registered maximum demand within the period and a flat energy charge rate. During the off-peak period, only flat-rate energy charge would apply. SPECIAL TARIFFS Time-of-day metering would be applied to electrolytic processing chemical industries and metal smelting industries, which have a high level of power consumption. IRRIGATION PURPOSES There would be a monthly fixed charge and a flat rate energy charge reflected in marginal energy cost during the off-peak period. STREET LIGHTING The monthly fixed charge would be eliminated. A flat rate energy charge woulcl reflect the marginal capacity cost and marginal energy cost at the 12-kilovolt level.

240

CASE STUDIES

Policy Issues and Conclusions

There are several important policy issues that need to be considered. Time-of-use metering A reduction in peak load and a corresponding increase in the use of energy during the off-peak period will result in savings by deferring investment as well as reducing annual operating and maintenance costs. Since the reduction in consumption likely to result from additional metering is usually uncertain, one way to determine whether a more complex tariff is justified would be to estimate what percentage reduction in demand consumption would make the investment in metering worthwhile (see Chapter 5). Residential and small business customers Considerable data are needed to estimate the net savings resulting from a shift in residential demand from the peak period to the off-peak period. The necessary information would include use profiles, cost effects of changing use profiles, saturation in the use of electricity, types of appliances, and the load pattern of residential and small businesses. Since these data are not usually available, the likely changes in consumption by electrical appliances for households and small businesses cannot be analyzed. It is possible, however, to identify the types of electrical appliances that most residential consumers use at various times of the day and to determine the possibilities for shifting loads during different periods of the day. The 1977 billing records show that the average monthly energy consumption for each household was approximately 200 kilowatt-hours and 60 kilowatt-hours in the MEA and PEA areas, respectively. Approximately 67 percent of all MEA residential customers and 70 percent of PEA residential customers used energy at this average level. Most households have no washers, dryers, dishwashers, or water heaters. Energy use involves lighting, refrigerators, electric fans, irons, and some low-power appliances. This implies that normally only irons can be shifted from the peak to the off-peak period. Kilowatt demand involving air conditioners of some residential and small business consumers, who work primarily during the off-peak period, may also be shifted. Given the appliances used by both residential and small business customers, peak load pricing would only slightly change the patterns of residential and small business electricity consumption. Medium and large business customers The medium and large business customers include industrial manufacturers, office buildings, hotels, theaters, and so on. The possibility of shifting demand from the peak to the off-peak period would focus on industrial

THAILAND

241

manufacturers. Given the nature of their services, the other businesses consume primarily during the working day, and shifting their loads would be
difficult.

Modification of the load pattern incurs some costs for the consumers as well as the inconvenience of shifting activities to different hours. Industrial consumers may incur higher costs of production; they often have different pay scales to compensate employees for the inconvenience of working during off-peak periods. The response of the industrial users depends on several important factors including the relative costs of capital, labor, and other forms of energy, as well as the technological process involved. The pattern of electricity consumption by industrial consumers is needed to estimate the potential amount of industrial load shifting. To justify time-of-use metering, how much peak demand could be reduced to make metering installation worthwhile must be estirnated. One general problem may arise. Time-of-day pricing may induce consumers to shift to private plants to meet peak use. This would be desirable only if the incremental costs of private power generation to the economy were less than the incremental costs of supplying peak energy through the public system. Such a conclusion is unwarranted if shadow prices are used. Since diesel fuel is subsidized, however, private generation may be financially attractive (see Chapter 5). This implies that there is an artificial ceiling above which one should not raise the peak price. The unit cost of electricity of typical high-speed diesel plants based on the actual private diesel fuel cost of B 1.64 per liter at a load factor of 15 percent (1,300 hours annually), is B 1.60 per kilowatt-hour. This figure should be the ceiling for peak period rates. The peak period LRMC of generation capacity was calculated using the AIC formula. To avoid difficulties associated with the large increase in demand charges and to maintain the energy charges at a level not less than the fuel cost of oil-fired thermal plants, EGAT'S AIC of capacity and energy were adjusted with the formula shown in Figure 12-3. The adjusted marginal costs are shown in Table 12-10. Social equity The 1977 records on the distribution of electricity bills for MEA and PEA show that average monthly consumption for each residential unit is 200 and 60 kilowatt-hours, respectively, Since the poor tend to spend a larger percentage of their income on electricity than others, across-the-board increases in power tariff affect the low-income users more than higher-income users. Introducing a new power tariff based on marginal cost pricing may prevent the low-income groups ifrom obtaining a minimum supply of electricity. A proposal that recognizes the increasing cost of services provides for equity by subsidizing electricity consumption of the low-income group with a life-

Figure 12-3. Flowchartfor Calculation of the Strict LRMC

Lfaad fotecast

Development plan
(least-cost solution)

Added demand M

Added energy

cE a

stmept

a Operati expense

,~

valus~ ~~~~~ ~~ ~Total~ ~ ~

~Cpct-eae

Eneresentte

,t

Total present value of added demand

ocacity

cost

r_ Ir _

r
_ _(AICE)

au Ttlpresent of added eneg

AIC of capacity (AICC)

AIC of eeg

| |~~~Aduse

~~

-C--

IF

;C

adjustment

duse C

energy MC, = cuffent incremental adjusted Note:Adjusted MC, = AIC, - MC adjustment;

cost; and *MC adjustment = [(AICE-adjustedMCE) X total present value of energy]/totat present value of added demand.

THAILAND

243

Table 12-10. Adjusted AMarginalCost of EGAT,

Marginal capacity cost (baht per kilowatt per month)
120.73 135.37 135.37 126.75 156.99 265.48 174.47 340.45

MEA, and PEA

Marginal energy cost (baht per kilowatt-hour) Peak period

0.5229 0.5434 0.5434 0.5288 0.5349 0.5815 0.5545 0.6301

Utility
EGAT MEA

Consumer category

Off-peak period
0.5115 0.5212 0.5212 0.5158 0.5201 0.5528 0.5271 0.5649

PEA

Direct consumers
MEA

69 kilovolts 12-24 kilovolts 220-380 volts 11-33 kilovolts 220-380 volts

EGAT data.

PEA

Source:

line rate entirely supported by the high-volume residential users. The tariff schedule would consist of three blocks or steps. The first would be a relatively low lifeline rate, an intermediate block would be at marginal cost, and the last step would be higher than marginal cost to recover the subsidy given to low users. The intermediate rate step is necessary to avoid unduly penalizing low-income residential users whose minimum electricity needs are somewhat above those covered by the lifeline rate because of such factors as family size. Before setting up the subsidized block of energy uses for the low-income group, a poverty threshold should be defined. A study by the National Statistical Office, "A Poverty Determination for Thailand," suggests that one threshold line occurs where the household's total consumption expenditure for essentials equals its income. Those groups whose incomes are below this level are defined as poor. Data from the Thai National Statistics Office's 1968-69 socioeconomic survey is used as reference. In the study, income is broken down into wages, salaries, compensation for working overtime, income earned as a broker, bonus, or net profit from the household's earnings. Expenditures refer to expenditures for food, housing, clothing, education, medical care, transport, electricity, and so on. The report presents average upper and lower limits of income for the poverty band for representative households in 1974, which were adjusted to the 1977 consumer price index to obtain the 1977 figures. The details of average representative households by regions are shown in Table 12-11. According to the calculation, 46 percent of the people in Thailand can be considered poor. This group, however, is not homogeneous. Most people in the low-income group are able to use electric power effectively only if they purchase other necessary complementary factors of production, thereby al-

244

CASE STUDIES

Table 12-11. Average Upper and Lower Limit of Poverty Band of Representative Households, Adjusted by 1977 Consumer Price Index
(1977 baht)

Region
Whole country Upper limit Lower limit Northern region Upper limit Lower limit Central region Upper limit Lower limit Northeastern region Upper limit Lower limit Southern region Upper limit Lower limit Bankok-Thonburi Upper limit Lower limit
Source:

Municipal area

Nonmunicipal area

2,536.34 1,320.43 2,367.77 1,265.82 2,038.97 1,106.35 2.184.96 1,187.10 2,863.71 1,486.55 4,440.24 1,367.48

1,494.18 877.23 1,578.14 918.49 1,378.00 769.07 1,451.68 883.35 1,704.37 991.94 1,548.09 860.89

EGATdata.

lowing them to increase their output of goods and services. At least in the long run, those who have access to productive factors such as land probably will benefit by the augmented supply of electric power. Their ability to pay will be increased. In this case it might be argued that the reduced rates at which power is supplied to them may be limited to a specific period of time. This discussion acquires more significance as the rural electrification program accelerates and as more consumers are affected by price increases. It may not be feasible to provide electricity in the immediate future to individual households of the absolutely poor people, whose number is as yet unknown. The amenity may be supplied to this group initially on a community basis. For both the relatively poor and the absolutely poor, it may also be necessary initially for the government to pay for connection and installation charges. To determine the minimum basic needs of households for electricity, the kilowatt-hour consumption by electric appliances required for living-to maintain health and a reasonable level of comfort-was considered. These uses include lighting, cooking, electric fans, radios, and irons. From the kilowatt-hours consumed for these residential uses, the appropriate size of the subsidized block is estimated to be 30 kilowatt-hours per month. The tariff level to be charged for the subsidized block should be considered in re-

THAILAND

245

lation to thle average income of the poor. If the lifeline rates were implemented, 11 percent of ME^Aresidential users and 51 percent of PEA residential users would be subsidized. To consider the effect of the lifeline rate structure on the kilowatt-hour usage and revenues, it is assumed that residential consumers will not respond to the rate changes applied by the lifeline rate. But the price elasticity of demand may not be zero. The low-usage consumer group whose electric bills will be reduced or a group that pays less under the new lifeline rate than under the current structure most probably will increase its electricity consumption. The high-usage consumer group will most probably decrease its consumption, since it will have to pay more if the lifeline rate structure is applied. The additional use by the low-usage group, whose kilowatt-hour use exceeds the initial kilowatt-hour block, will be subject to the intermediate rate. Since the intermediate rate will cover marginal cost of service, there will be little or no increase in the revenue deficiency to be recovered from the high-usage group. Yet, the adjustment in kilowatt-hour use by the high-usage group in response to the new lifeline rate will help revenue recovery. A properly designed lifeline proposal will serve two objectives: it will enable the basic electricity needs of the low-income users to be fulfilled at low cost while encouraging conservation in households using more electricity.
Inflationary effects of power price increases

A tariff based on marginal cost will be at a higher level than the existing tariff. For most households and for normal uses, however, only a small and often insignificant amount of total income is spent on electric power. Such demand for power may be inelastic with respect to a change in its price, but there will doubtless be some instances in which consumers will economnize on electricity. This effect would help to counter inflation. For most industrial and commercial users, electricity costs would amount to no more than from 5 to 10 percent of total costs. For power-intensive industries, however, electricity may be a significant input and may constitute approximately from 5 to 30 percent of costs, as in chemical industries using electrolytic processing, cement industries, iron smelting industries, and so on. In these examples, increases in the price of power might tend to have an inflationary effect, but this will be offset by the following factors. First, desirable technology change may be induced by price increases; and second, the price increases may hasten the exit of old firms from the industries. This is desirable since it will improve the efficiency of the entire industry. The increase in the puice of power may have a psychological effect, however, and may contribute to a general increase in the prices of commodities. This will be true only in a generally inflationary situation, which now exists.

246

CASE STUDIES

Yet, for reasons of resource allocation, it is necessary in such a situation that the price of power should increase so that it remains at least constant in real terms and increasingly reflects changes in relative prices. Care should be exercised, however, in adjusting tariffs; an increase in the price of power may bring about unexpected and erratic changes in demand. If so, unexpected demand changes may adversely affect the load factor during the off-peak period, for example, regarding the use of air conditioners. The inflationary effect of increasing the price of power was calculated by using 1975 input-output coefficient data. If the electricity tariff were increased by 30 percent, the general price level is estimated to increase by only 0.1 percent. Industrial and export promotion It is often argued that cheap power is an instrument to promote industrial development. The argument is valid but narrowly limited. As mentioned earlier, electricity charges are a small fraction of the cost of industrial production. Industrial development should be viewed more as a function of entrepreneurship, available investment resources, and technology. The net contribution of cheap power to industrial development is not necessarily positive. Cheap power is not a very effective way to promote industrial development; without basic resources and skills, cheap power will not alone induce industrial development. The same argument applies to the relation between cheap power and export promotion. Nevertheless, it is a popular belief that cheap power might induce the export industry to reduce its costs. Although it is doubtful that the reduction in the cost of the export industry can readily be brought about by cheap power, the electricity industry will probably lose by pricing its output below cost to the export industry. The incidental loss to the electricity industry will have to be replaced by charging higher rates to the other consumers. The practice of cross substitution among consumers will violate the marginal cost principle. A better policy would be for the government to create necessary conditions, such as full availability of raw materials and other inputs and favorable administrative action, to encourage the export industry. Multinational corporations should be treated the same way as domestic industry. In practice, multinational industry is attracted to a country by the market potential and inexpensive labor. Also, special agreements with local companies in the country and governments' policies affect their decisions. Rural and regional development In the context of the government's antipoverty program, rural and regional development have taken an important role. Power sector representatives have set the tariff rate in the PEA area, currently the same as mEA area

THAILAND

247

for residential households, but higher than that in the MEA area for industrial and commercial uses. The difference in the industrial and commercial tariff between MEA and PEA can be established on the basis of the difference in the marginal cost of power. This, however, results in a higher tariff in the PEA area, where, in general, people are poorer than in the MEA jurisdiction. Under these circumstances, if the government were to decide to have a tariff structure that does not exactly correspond to the marginal cost in the interest of regional balance and rural development, the cost of such a departure from marginal cost pricing should be calculated and actions taken to minimize it. Tariffs and public revenue Since the marginal costs of incremental power are much higher than the average accounting cost of the existing capacity, marginal cost prices would result in revenues well in excess of traditionally accepted fair rates of return. Some methods to solve this problem should be considered. One approach would be for the government to impose a tax on the power sector. Alternatively, the profits tax may be redistributed to the utilities' customers. The amount received by each customer can be based on a variety of formulas, but the redistribution should not reflect current use as a redistribution base, as this would tend to alter the price signals of marginal cost pricing. The third method is to charge the marginal cost only for marginal use. This would require determination of a baseline of electricity use for each class of customer that would be charged at low rates. All uses beyond the base would be charged at marginal costs. The form of the tariff structure would be a two-block rate in each rating period. The first block rates are set to meet the revenue requirement and to implement a fair allocation of the existing capital costs between rating periods. For the terminal block, rates are not relative to the time-of-use pricing. It will be based on the marginal cost associated with increased use in the two rating periods, and it will be significantly higher in the peak: period than in the off-peak period. The revenues in each rating period must equal its fuel and O&M costs plus its allocated annual capacity costs. Thus a block rate schedule in each rating period enables the utility to meet the revenue requirement. Summary Although the rationale for marginal cost pricing is accepted as a general principle to guide the electricity supply industry in setting tariffs, several adjustments must be made. Some of them imply that marginal cost pricing may not be implemented immediately, but rather should be introduced gradually over several years, with the following adjustments. a. Ensure that poor c-onsumershave access to at least a minimum supply at a subsidized rate

248

CASE STUDIES

b. Allow for the fact that the cost of special metering (particularly time of day) may exceed the benefits c. Allow time-of-day metering to be phased in (where appropriate) so that peak shifting is avoided d. Allow all substantial changes in tariff levels or structure to be phased in so that those who have recently made investment decisions based on prevailing rates would get fair consideration e. Ensure financial viability of the utilities (this is not likely to be a problem since marginal cost usually exceeds financial costs) f. Achieve fiscal goals g. Consider divergences between prices and marginal costs of competing products, for example, kerosene, diesel, firewood h. Ensure that any inflationary effect is minimized. The general conclusion is that Thailand should proceed gradually toward a tariff structure based on marginal cost. Where other objectives prevail, they should be analyzed with full knowledge of the economic costs of diverging from the strict marginal cost approach. Marginal cost, in other words, should continue to be a benchmark by which other pricing policies should be evaluated. Furthermore, analysis of marginal cost must be an ongoing exercise because the data are continually changing.

Appendixes

AppendixA

Seminar Participants
THE CASE STUDIESIN PART Two

were developed from material presented at two World Bank Seminars on electric power pricing that were organized and conducted by the authors and attended by about seventy representatives from twelve Asian countries as well as several regional institutions. Seminar participants are listed belowa.

First World Bank Asian Regional Power Tariff Seminar, Bangkok, Thailand (June 27-July 3, 1978)
Country delegates
BURMA

Mr. U. Tin Tun Director (Finance) Burma Electric Power Corporation Rangoon, Burma
Mr. Bruce Harold Williams

FIJI

Company Secretary Fiji Electricity Authority Private Mail Bag, Lautoka, Fiji
INDONESIA Mr. Soejanto Soejitno

Deputy Director of Budgeting

Perusahaan Umuni Listrik Negara (PLN)

Jakarta, Indonesia Mr. Djaya Santoso Chief, Division for Electricity Tariffs and Revenue, PLN Jakarta, Indonesia Mr. Sastrowidjoyo Soepangkat Chief of Planning Division, PLN Jakarta, Indonesia
KOREA Mr. Yoon Hi Woo

Deputy Manager of Planning and Controlling Department

251

252

APPENDIXES

Korea Electric Company Seoul, Korea

(KECO)

Mr. Ha Soo Gil Assistant Chief of Rate Study Section, Planning & Controlling Department
KECO

Seoul, Korea
MALAYSIA

Mr. Wan Ali Bin Wan Jaafar Assistant Chief Accountant (Finance) National Electricity Board Kuala Lumpur, Malaysia Mr. Razali Ismail Economics & Statistics Office Bembaga Letrik Negara Petaling Jaya, Malaysia

NEPAL

Mr. Krishna Man Shreshtha Manager, Finance & Economic Analysis Department Nepal Electricity Corporation Kathmandu, Nepal Mr. Muhammed Jahangir Director (Tariffs and Statistics) Water and Power Development Authority Lahore, Pakistan Mr. Ignatius Namake Assistant General Manager (Commercial) Electricity Commission of PNG Boroko, Papua New Guinea Mr. Ildefonso S. Bayani Manager, Rates & Power Economics Division National Power Corporation Quezon City, Philippines Mr. Wee Kian Kok Acting General Manager, Public Utilities Board Singapore Mr. Goh Tiak Theng Chief Planning Officer Public Utilities Board Singapore

PAKISTAN

(WAPDA)

PAPUA

NEW GUINEA

PHILIPPINES

SINGAPORE

SEMINAR

PARTICIPANTS

253

SRI

LANKA

Mr. Edward Nelson Wijemanne Commercial Manager, Ceylon Electricity Board Colombo, Sri Lanka Mr. Chrisantha Ratnayake Commercial Engineer, Commercial Division Ceylon EJlectricityBoard Colombo, Sri Lanka

THAILAND

Mr. Pravit Ruyabhorn Director--General National Energy Authority Bangkok, Thailand

(NEA)

Mr. Nibhat Bhukkanasut Director, Fiscal Planning Division, Fiscal Policy Office, Ministry of Finance Bangkok, Thailand Mrs. Amphorn Lekuthai Senior Economist Fiscal Planning Division, Fiscal Policy Office Ministry of Finance Bangkok, Thailand Mr. Yongyuth Sricharoen Deputy Manager Metropolitan Electricity Authority Bangkok, Thailand Mr. Vanrn Chaiyasuta Chief of Section, MEA Bangkok, Thailand Mr. Tophong Vachanavasti Senior Economist and Chief of Energy Planning National Economic and Social Development
(NESDII)

(MEA)

Board

Bangkok, Thailand Miss Chintana Sakulprahm Senior Energy Economist Infrastructure Projects Division, Bangkok, Thailand

NESDB

Mr. Pratin Pathanaporn Assistant Director, Planning Department Electricity Generating Authority of Thailand Bangkok, Thailand

(EGAT)

254

APPENDIXES

Mr. Viroj Nopkhun Chief, Power Economics & System Operation Planning Division, EGAT Nonthaburi, Thailand Mr. Khien Vongsuriya Chief, Economic & Investment Policy Division, Nonthaburi, Thailand Mr. Sunthorn Tanthavorn Chief, Project & Planning Division, Provincial Electricity Authority (PEA) Bangkok, Thailand Mr. Sudchai Lophongse Assistant Chief, Power Economics Division, Nonthaburi, Thailand Mr. Tammachart Sirivadhanakul Director, Regulatory Division National Energy Authority (NEA) Bangkok, Thailand Mr. Viset Choopiban Chief of Energy Administration Section Regulatory Division, NEA Bangkok, Thailand Miss Attaya Chiewattakee Chief of Energy Tariff Section Regulatory Division, NEA Pibultham Villa Kasatsuk Bridge Bangkok, Thailand Special invitees and observers Mr. V. Desai Asian Development Bank Manila, Philippines Dr. D. V. Ramana Asian Development Institute Bangkok, Thailand Mr. K. Yamada Economic and Social Commission for Asia and the Pacific Bangkok, Thailand Dr. Psisit Pakkasem National Economic and Social Development Board
PEA EGAT

SEMINAR

PARTICIPANTS

255

Bangkok, rhailand Mr. Alan Walters Urban Adviser World Bank Washington, D.C.

Second World Bank Asian Regional Power Tariff Seminar, Bali, Indonesia (January 7-13, 1979)
Country delegates
BURMA

Mr. U. Tin Tun Director (Finance) Burma Electric Power Authority Rangoon, ]3urma
Mr. Bruce Harold Williams

FIJI

Company Secretary Fiji Electricity Authority Lautoka, Fiji

INDONESIA

Dr. Samaoen Samadikoen Director General of Power Ministry oi Mines and Energy Jakarta, Indonesia Mr. Abdul Kadir Director for the Development of Power Industries Ministry oi Mines and Energy Jakarta, Indonesia Prof. Suryono
President-Director

Perusahann Umum Listrik Negara Jakarta, Indonesia Mr. Sardjono Director of Planning, Jakarta, Indonesia
PLN

(PLN)

Mr. Bambang Sara Director of Operations, Jakarta, Indonesia

PLN

Mr. Imam Dharmono Director of Finance, PLN Jakarta, Indonesia

256

APPENDIXES

Mr. Soejanto Soejitno Deputy Director of Budgeting, Jakarta, Indonesia

PLN

Mr. Djaya Santoso Chief, Division of Electricity Tariffs and Revenue, Jakarta, Indonesia Mrs. Agostina Economist, PLN Jakarta, Indonesia
KOREA

PLN

Mr. Yoon Hi Woo Deputy Manager, Planning & Controlling Department Korea Electric Company Seoul, Korea Mr. Ha Soo Gil Assistant Chief of Rate Study Section, Planning & Controlling Department Korea Electric Company Seoul, Korea

MALAYSIA

Mr. Ahmad Abid Abidin Deputy Commercial Manager National Electricity Board Tanah Melayu Kuala Lumpur, Malaysia Mr. Razali Ismail Economics and Statistics Officer Bemhaga Letrik Negara Petaling Jaya, Malaysia

NEPAL

Mr. Krishna Man Shreshtha Manager, Finance and Economic Analysis Department Nepal Electricity Corporation Kathmandu, Nepal Mr. Muhammed Jahangir Director (Tariffs & Statistics) Water and Power Development Authority Lahore, Pakistan Mr. Ignatius Namake Assistant General Manager (Commercial) Electricity Commission of PNG Boroko, Papua New Guinea

PAKISTAN

(WAPDA)

PAPUA

NEW GUINEA

SEMINAR

PARTICIPANTS

257

PHILIPPINES

Mr. Demetrio C. Paz Vice Presidlent, Corporate Planning National Power Corporation Metro Manila, Philippines Mr. Asisclo T. Gonzaga Manager, Pricing and Tariff Division National Power Corporation Metro Manila, Philippines

SINGAPORE

Mr. Wee Kian Kok Deputy General Manager, Public Utilities Public Utilities Board Singapore Mr. Goh T iak Theng Director, Corporate Planning Public Utilities Board Singapore

SRI LANKA

Prof. K. K. Y. W. Perera Chairman, Ceylon Electricity Board Sri Lanka Mr. Tophong Vachanavasti Senior Economist and Chief of Energy Planning National Economic and Social Development Board Bangkok, Thailand Mr. Pratin Pathanapom Assistant Director, Planning Department Electricity Generating Authority of Thailand (EGAT) Bangkok, Thailand Mr. Viroj Nopkhun Chief, Power Economic and System Operation Planning
EGAT

THAILAND

Nonthabuii, Thailand Mr. Khien Vongsuriya Chief, Economic and Investment Policy Division Planning Department, EGAT Bangkok, Thailand Mr. Viset Choopiban Chief of E,nergy Administration Division National Energy Administration Bangkok, Thailand

258

APPENDIXES

Special invitees and observers Mr. V. Desai Economist Asian Development Bank Metro Manila, Philippines Mr. J. R. Frisch Assistant to the Director-General Electricit6 de France Paris, France

B Appendix

Power SectorStatistics forDeveloping Countries

WITHIN THE CATEGORIES "developing countries," "developed market economies," and "centrally planned economies," in Table 1-1, are listed below. THE COUNTRIES AGGREGATED

Developing Countries
Algeria, Libya, Morocco, Western Sahara, Sudan, Tunisia, Egypt, United Republic of Cameroon, Central African Republic, Chad, Congo, Gabon, Angola, Burundi, Cape Verde, Comoros, Zaire, Benin, Equatorial Guinea, Ethiopia, Djibouti, Gambia, Ghana, Guinea, Ivory Coast, Kenya, Liberia, Madagascar, Malawi, Mali, Mauritania, Mauritius, Mozambique, Niger, Nigeria, Guinea Bissau, Reunion, Rwanda, Sao Tome and Principe, Senegal, Seychelles, Sierra Leone, Somalia, Zimbabwe, Togo, Uganda, United Republic of Tanzania, Upper Volta, Zambia, Bermuda, Greenland, St. Pierre and Miquelon, Argentina, Bolivia, Brazil, Chile, Ecuador, Falkland Islands, French Guiana, CGuyana, Paraguay, Peru, Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, Antigua, Bahamas, Barbados, Belize, British Virgin Islands, Cayman Islands, Colombia, Cuba, Dominica, Dominican Republic, Guadaloupe, Haiti, Jamaica, Martinique, Mexico, Montserrat, Netherlands Antilles, Panama, Puerto Rico, St. Kitts, St. Lucia, St. Vincent, Virgin Islands, Venezuela, Bahrain, Cyprus, Iran, Iraq, Jordan, Kuwait, Lebanon, Oman, Qatar, Saudi Arabia, Democratic Yemen, Syrian Arab Republic, United Arab Emirates, Turkey, Yemen, Afghanistan, Bangladesh, Brunei, Burma, Democratic Kampuchea, Sri Lanka, Hong Kong, India, Indonesia, Republic of Korea, Laos, Macau, Malaysia, Maldives, Nepal, Pakistan, Philippines, East Timor, Singapore, Thailand, American Samoa, Solomon Islands, Christmas Islands, Cook Islands, Fiji, French Polynesia, Kiribatin, Guam, Nauru, New Caledonia, New Hebrides, Niue, Papua New Guinea, Samoa.
259

260

APPENDIXES

Developed Market Economies

Australia,Canada,Israel, Japan, New Zealand, SouthAfrica, U.S.A., Belgium, Denmark,France, Federal Republicof Germany,Ireland, Italy, Luxembourg,Netherlands,United Kingdom, Austria, Faeroe Islands, Finland, Iceland,Norway, Portugal, Sweden, Switzerland,Gibraltar, Greece,Malta, Spain, Yugoslavia.

Centrally Planned Economies

Albania, Bulgaria, China, Czechoslovakia, Democratic People'sRepublicof Korea, German DemocraticRepublic, Hungary, Mongolia, Poland, Romania, Viet Nam, USSR. The followingtables contain useful socioeconomic and power sector statistics for some of the developingcountries. The data used in these tables were gatheredfrom the World Bank Atlas, World Bank AppraisalReports, U.N. World EnergySupplies"Series J," and other country-specific reports. Electricalenergy units were convertedto oil equivalentunits in item 5, using a factor of 4,000 gigawatt-hours per milliontons of oil equivalent,representingthe averageefficiencyof conversionin a thermal power station and thus reflectingmore-trulythe electricityshare in total energy.

AFGHANISTAN

WHICH DA AFGHANISTAN 8RESHNA MOASSESSA OPERATES 28 REGIONAL SYSTEMS. LARGEST CAPACITY) OF INSTALLED (60% IS KABUL 1973-78 AVG ANNUAL GTH RATE (%) 2.20 4.70 2,70 1.50 7.00 S. COMMERCIALENERGY CONSUMPTION X MTOE 58.1700 0.388 : 3.7500 0.025 17.3900 0.116 : 0.138 20.6900

1. SOCIO-ECONOMIC DATA POPULATION (000) : GNP (USSM) GNP/CAPITA (USS) : ENERGYCONSUMED(MTOE) : ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) (X) POPULATION BREAKDOWN ACCESS TO ELECTRICITY (X) 1461t 3530 240 0.667 806 0.201 30.1 46 55 URBAN 14 URBAN 21

OIL GAS COAL LIGNITE HYORO NUCLEAR GEOTHERMAL OTHER TOTAL

RURAL 86 0 RURAL

0.667

100.0000

TOTAL

6- UTILITY DATA 50 FREQUENCY(HZ): 110/44/35 TRANSMISSION (XV): 11 DISTRIBUTION (KV): LOW VOLTAGE (VOLTS): 220/380 44.0 MAIN SYSTEM LOAD FACTOR (%): 28.0 TOTAL SYSTEM CAPACITY FACTOR (%): 33.0 TOTAL SYSTEM T8D LOSSES (%): 98300 NUMBEROF CONSUMERS: 5000 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 20 2.4 AVERAGE REVENUE (US CENTS/KWH) N ASSETS REVALUED (POLICY)(Y.YES,N=NO): 0.0 RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 10.0 7. SALES DWMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .3SLF) INDUSTRIAL (200KW.70LF) OTHER GWH 213 46 86 345 % 81.7400 13.3300 24.9300 100.0000 KWH/MONTH US CENTS/KWH . 2 1.7 25 1.7 100 1.9 400 1.9 400 2.6 2000 2.6 100000

2. INSTALLED CAPACITY & GROSS GENERATION MW THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERMAL SUPPLY SECTOR TOTAL GWH

(a,

42 42 258

32 32 514 21 567 239 806

300 54 354

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE AVE 78 USsi*4U.5 AFOHANIS AFO AVE 1980 U551-44.8 X

ALGERIA

SOCIETE NATIONALE DE L'ELECTRICITE ET DU GAZ WHICH OPERATES ONE MAIN INTERCONNECTED SYSTEM (89% OF GENERATION) AND FEW ISOLATED SYSTEMS 1973-78 AVG ANNUAL GTH RATE (%) 3.20 5.30 2.10 11.30 9.40 S. COMMERCIALENERGYCONSUMPTION MTOE % OIL 4.300 49.1900 GAS : 4.163 47.6200 COAL 0.153 1.7500 LIGNITE HYDRO 0.128 1.4400 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 80 S. UTILITY DATA FREQUENCY(HZ): 50 TRANSMISSION (KV): 220/90/60 DISTRIBUTION (KV): 15/11/5 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (): 57.5 TOTAL SYSTEM CAPACITY FACTOR (%): 38.3 TOTAL SYSTEM TBD LOSSES (%): 12.6 NUMBEROF CONSUMERS: 1129000 NUMBEROF EMPLOYEES: 12000 CONSUMERS/EMPLOYEE: 94 AVERAGEREVENUE (US CENTS/KWH) 4.0 ASSETS REVALUED (POLICY)(Y=YES,N=NO): N RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 0.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS 1.991 RATES OF EXCHANGE 100.0000 DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW,.70LF) OTHER GWH 1176 1783 1425 4364 % 26.9500 40.4000 32.6500 100.0000 KWH/MONTH US CENTS/KWH 7.0 3.3 i.7 8.742 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) GNP/CAPITA (USS) : ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) : ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) : POPULATION BREAKDOWN (%) : ACCESS TO ELECTRICITY (%) : 2. 17701 22290 1260 8.742 5441 1.360 15.6 494 307 URBAN 40 URBAN 100

RURAL 60 RURAL 33

INSTALLED CAPACITY B GROSS GENERATION MW 787 216 55 1038 287 GWH 4457 90 4547 244

t',

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY

1325

4791 850 5441

AUTO PRODUCERS HYDRO THERMAL 295 TOTAL SECTORSUPPLY 1820

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL MTOE 0.857 1.134 % 43.0400 56.9600

AVE 78 USS1-3.97 ALO DINARS AVE 1980 USS1-3.97 ALO D

ARGENTINA

FRAGMENTEDPOWER SECTOR REGULATEDBY GOV'T FIVE OF THE NINE ELECTRICAL REGIONS ARE INTERCONNECTED 1973-78 AVG ANNUAL GTH RATE (%) 1.30 3.20 1.80 1.40 5.10 5. COMMERCIAL ENERGYCONSUMPTION MTOE X OIL 23.360 65.6000 GAS : 8.560 24.0400 COAL : 1.080 2.9800 LIGNITE : 1.9tO 5.3600 HYDRO 0.720 2.0200 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 80 S. UTILITY DATA 35.610 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) : GNP (US$M) : GNP/CAPITA (USS) : ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) : POPULATION BREAKDOWN (X) : ACCESS TO ELECTRICITY ) : 26371 50250 1910 35.610 34025 8.510 23.9 1350 129O URBAN 80 URBAN 99

RURAL 20 RURAL 5

2. INSTALLED CAPACITY & GROSS GENERATION MW THERMAL-STEAM 3813 -GAS TURBINES 1486 -COMBINED CYCLE -DIESEL 765 -SUBTOTAL 6064 HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY 2911 370 9345 GWH 15283 1892 1226 18401 7618 2896 28915 78 5212 34205 FREQUENCY (HZ): 5o TRANSMISSION (KV): 500/220/132 DISTRIBUTION (KV): 13.2 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR Ct): 59.0 TOTAL SYSTEM CAPACITY FACTOR (%): 35.3 TOTAL SYSTEM T&D LOSSES (%): 14.0 NUMBEROF CONSUMERS: 6800000 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: AVERAGEREVENUE (US CENTS/KWH) 5.0 ASSETS REVALUED (POLICY)(Y'YES,N=NO): Y RATE OF RETURN(%): 5.2 CONTRIBUTION TO INVESTMENT: 14.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW..35LF) INDUSTRIAL (200KW,.70LF) OTHER GWH 8188 2799 11583 3344 25914 X 31.6000 10.8000 44.7000 12.9000 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 20.7 12.1 11.5 14.9 7.8 6.4

AUTO PRODUCERS HYDRO 24 THFRMAL 1906 TOTAL SECTORSUPPLY 11275

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL MTOE 9.815 %

RATES OF EXCHANGE

AVE 78 USS1=796 PESOS AVE 1980 US$1=1837.2 PESOS

BANGLADESH

BANGLADESHPOWERDEVELOPMENT BOARD GENERATES, TRANSMITS ANO D1STRIZUTES. RURAL ELECTRIFICATION BOARD STARTIsBO 1973-78 AVG ANNUAL GTH RATE (%) 2.80 2.30 -0.20 8.90 14.00 5. COMMERCIALENERGY CONSUMPTION MTDE % : 1.238 48.7000 : 0.909 35.7E00 : 0.257 10.1100 : 0.138 5.4300

1. SOCIO-ECONOMIC DATA POPULATION (000) : GNP (USSM) GNP/CAPITA (USS) : ENERGYCONSUMED(MTOE) : ELECTRICTY PRODUCE(GWH) : ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (x): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (x) ACCESS TO ELECTRICITY (X) :

83641 7630 80 2.542 2482 0.616 25.1 29 29 URBAN 9 RURAL 91 URBAN 18 RURAL 0

OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL

2.542

1oo.OOOO

TOTAL

2 8. UTILITY DATA FREQUENCY(HZ): 50 TRANSMISSION (KV): 132/66/33 DISTRIBUTION (KV): 33/11 LOW VOLTAGE (VOLTS): 240/415 MAIN SYSTEM LOAD FACTOR (1): 55.5 TOTAL SYSTEM CAPACITY FACTOR (x): 33.7 TOTAL SYSTEM T&D LOSSES (%): 29.0 NUMBEROF CONSUMERS: 451718 NUMBEROF EMPLOYEES: 20200 CONSUMERS/EMPLOYEE: 22 AVERAGEREVENUE (US CENTS/KWH) 2.9 ASSETS REVALUED (POLICY)(Y=YES,N=NO): N RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 0.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS GWH 228 220 824 S5 1338 X 17.0400 16.4400 81.5800 4.9300 99.9900 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 10000 2.6 2.8 2.8 4.9 3.9 4.2 2.6

2. INSTALLED CAPACITY & GROSS GENERATION THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTOR SUPPLY MW 394 GWH 1103 363 89 1535 587

18

k"J

88 838 8O

718 30 B5 833

2122 190 1SO 2462

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL MTOE 0.173 0.350 X 33.0800 86.9200

0.523

100.0000

4. OFFICIAL RATES OF EXCHANGE AVE 78 USS1=15.5 TAKA AVE 1980 US1=18.25 TAKA

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUJSTRIAL (8KW, .35LF) INDUSTRIAL (200KW, .70LF) OTHER

BARfADOS

BARBADOS LIGHT AND POWER CO. LTD SINGLE LARGELY PRIVATE OWNED UTILITY SUPPLIES ALL PUBLIC SUPPLY 1973-78 AVG ANNUAL GTH RATE (X) 0.50 3.10 2.60 3.30 4.60 5. COMMERCIALENERGY CONSUMPTION MTOE X OIL : 0.180 96.7700 GAS 0.006 3.2300 COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 90 6. UTILITY DATA 0.186 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) : GNP (US$M) : GNP/CAPITA (USS) ENERGY CONSUMED(MTOE) : ELECTRICTY PRODUCE(GWH) : ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (X) : ACCESS TO ELECTRICITY (X) 250 490 1940 0.188 314 0.079 42.5 744 1256 URBAN 80 URBAN 100

RURAL 20 RURAL 50

2. INSTALLED CAPACITY & GROSS GENERATION THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTOR SUPPLY MW 40 17 39 98 GWH 210 3 74 287 FREQUENCY(HZ): 50 TRANSMISSION (KV): 24 DISTRIBUTION (KV): It LOW VOLTAGE (VOLTS): 110/220 MAIN SYSTEM LOAO FACTOR (X): 68.5 TOTAL SYSTEM CAPACITY FACTOR (X): 33.8 TOTAL SYSTEM T&D LOSSES (X): 11.0 NUMBEROF CONSUMERS: 63761 NUMBEROF EMPLOYEES: 422 CONSUMERS/EMPLOYEE: 151 AVERAGEREVENUE (US CENTS/KWH) 7.9 ASSETS REVALUED (POLICY)(Y-YES,N=NO): Y RATE OF RETURN(%): 1.6 CONTRIBUTION TO INVESTMENT: 0.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL S. TARIFFS 0.083 100.0000 DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW,.35LF) INDUSTRIAL (200KW, .70LF) OTHER GWH 74 167 4 245 % 30.2000 68.1600 1.6300 09.9900 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 17.0 12.5 12.1 14.1 13.4 12.3

ON
>JI

98 0 B 104

287 0 28 315

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL MTOE 0.083 X 100.0000

4. OFFICIAL RATES OF EXCHANGE AVE 78 USS1=2.01 BARBADOS S AVE 1980 USS1=2.01 BAR S

BOLIVIA

MAdOR PRODUCERS OF ELECTRICITY (BPC, ENDE, AND COMIBOL) ENDE GENRATESAND TRANSMITS TO ALL EXCEPT LA PAZ. 4 ISOLATED SYSTMS 1973-78 AVG ANNUAL GTH RATE (X) 2.70 5.50 2.90 1.20 5.90 S. COMMERCIALENERGYCONSUMPTION MTOE X : 1.336 79.3300 0.117 6.9500 0.231 13.7200

1. SOCIO-ECONOMIC DATA POPULATION (000) : GNP (USSM) GNP/CAPITA (USS) ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (%) : ACCESS TO ELECTRICITY (%) 2. 5290 22690 510 1.584 1354 0.339 20.1 318 250 URBAN 34 URBAN 78

OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL

RURAL 88 RURAL 10

1.884 DATA

100.0000

TOTAL

33 8. UTILITY FREQUENCY( HZ): 50 TRANSMISSION (KV): 220/155/89 DISTRIBUTION (KV): 25/13.2 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (%): 52.0 TOTAL SYSTEM CAPACITY FACTOR (%): 36.7 TOTAL SYSTEM T&D LOSSES (X): 12.0 NUMBEROF CONSUMERS: 231838 NUMBEROF EMPLOYEES: CoNSUMERS/EMPLOYEE: AVERAGEREVENUE (US CENTS/KWH) 2.8 ASSETS REVALUED (POLICY)(Y-YES.N:NO): Y RATE OF RETURN(X): 9.2 CONTRIBUTION TO INVESTMENT: 30.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW,.70LF) OTHER GWH 298 120 308 454 1180 % 25.2500 10.1700 28.1000 38.4700 99.9900 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 3.2 5.2 3.2 2.8

INSTALLED CAPACITY & GROSS GENERATION MW GWH 143 98 241 810

1si :> ON

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTOR SUPPLY

e8 45 113 214

327 27 73 427

1051 154 149 1354

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE X

AVE 78 USS1=20 PESOS AVE 1980 USSl-24.5 PESOS

BOTSWANA

CORPORATION- PRINCIPAL POWER BOTSWANA ENTITY RESPONSIBLE FOR POWERGENERATION AND DISTRIBUTION 1973-78 AVG ANNUAL GTH RATE (%) 1.90 18.30 16.10 5. COMMERCIAL ENERGYCONSUMPTION % MTOE OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL S. UtILITY DATA

1. SOCIO-ECONOMIC DATA : POPULATION (000) : GNP (USSM) GNP/CAPITA (US$) : ENERGYCONSUMED(MTOE) : ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): (X): ENERGY ELECTRICITY/TOTAL ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) : (X) POPULATION BREAKDOWN (X) TO ELECTRICITY ACCESS 747 466 620 299 0.075 399 URBAN 11 URBAN

RURAL 89 RURAL

TOTAL

2. INSTALLED CAPACITY & GROSSGENERATION THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO N,UCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERmAL TOTAL SECTORSUPPLY MW es 15 86 GWH

K)

86

FREQUENCY (HZ): 50 TRANSMISSION (KV): 66 DISTRIBUTION (KV): 33/Il LOW VOLTAGE (VOLTS): 220/380 70.1 MAIN SYSTEM LOAD FACTOR (%): 39.4 TOTAL SYSTEM CAPACITY FACTOR (X): 1.5 TOTAL SYSTEM TAD LOSSES (X): 4500 NUMBEROF CONSUMERS: 517 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 9 4.9 AVERAGE REVENUE (US CENTS/KWH) N ASSETS REVALUED (POLICY)(Y=YES,N=NO): 7.1 RATE OF RETURN(%): 0.0 CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW,.7OLF) OTHER GWH 'A 5.8S00 88.6200 5.5200 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000

86 87

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE AVE 78 US$1=.83 PULA AVE 1980 USS1'.77 PULA X

257 18 290

8.2 3.6

BRAZIL

FOUR MAJOR SUPPLIERS ARE SUBSIDIARIES OF ELECTROBRAS. DISTRIBUTION IS MADE BY CESP,CPEL,CEMIG,COPEL,ESCELSA 1973-78 AVG ANNUAL GTH RATE (X) 2.90 9.60 8.70 7.90 11.40 5. COMMERCIAL ENERGYCONSUMPTION MTOE % OIL 46.180 52.5800 GAS 0.530 0.6000 COAL LIGNITE 9.080 10.3400 HYDRO 25.770 29.3400 NUCLEAR GEOTHERMAL OTHER 8.260 7.1300 TOTAL TOTAL 62 6. UTILITY DATA FREQUENCY(HZ): 60 TRANSMISSION (KV): 500/440/345/230/138 DISTRIBUTION (KV): 138/69/34.5 LOW VOLTAGE (VOLTS): 110/220 MAIN SYSTEM LOAD FACTOR (%): 56.0 TOTAL SYSTEM CAPACITY FACTOR t%): 47.9 TOTAL SYSTEM TBD LOSSES (%): 17.0 NUMBEROF CONSUMERS: 14268164 NUMBEROF EMPLOYEES: 105700 CONSUMERS/EMPLOYEE:135 AVERAGE REVENUE(US CENTS/KWH) 3.2 ASSETS REVALUED (POLICY)(Y=YES,N=NO): Y RATE OF RETURN(%): 8.1 CONTRIBUTION TO INVESTMENT: 26.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 1.000 2.870 34.8400 100.0000 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW..70LF) OTHER GWH 18221 11041 48017 10594 87873 % 20.7400 12.5600 54.6400 12.0600 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 3.5 6.9 6.8 8.2 6.7 3.0 87.820 99.9900

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (US$M) GNP/CAPITA (USS) : ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EOUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (%) ACCESS TO ELECTRICITY (%) 2. 119430 187190 1570 87.820 113890 28.470 32.4 735 953 URBAN 60 URBAN 95

RURAL 40 RURAL S

INSTALLED CAPACITY & GROSSGENERATION MW GWH

Xo

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL ofi -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY 2158 9999

2158

9999 3104 4856 17959

AUTO PRODUCERS HYDRO 425 THERMAL 1025 TOTAL SECTOR SUPPLY 3608

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL MTOE 1.870 % 65.1600

4. OFFICIAL RATES OF EXCHANGE AVE 78 USS1-20.9 CRUZ AVE 1980 USS1:B5.5 CRUZ

BURMA

ELECTRIC POWERCORPORATIONOPERATES SYSTEM A MAIN INTERCONNECTEO (93% OF SALES) 1973-78 AVG ANNUAL GTH RATE (X) 2.20 3.50 1.30 5.90 8.40 5. COMMERCIALENERGYCONSUMPTION MTOE % 1.030 70.3100 0.153 10.4400 : 0.144 0.138 9.8300 9.4200

1. SOCIO-ECONOMIC DATA : POPULATION (000) : GNP (USSM) GNP/CAPITA (USS) ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE (GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) (X) POPULATION BREAKDOWN ACCESS TO ELECTRICITY (X) 2. 32205 4900 150 1.465 1102 0.276 18.5 46 34 URBAN 22 URBAN

OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL

RURAL 78 RURAL

1.485

100.0000

TOTAL

7 S. UTI LITY DATA 50 FREQUENCY(HZ): TRANSMISSION (KV): 220/132 DISTRIBUTION (KV): 66/33/11/6.6/3.3 LOW VOLTAGE (VOLTS): 220/380 83.0 MAIN SYSTEM LOAD FACTOR (X): 29.6 TOTAL SYSTEM CAPACITY FACTOR (%): TOTAL SYSTEM T&D LOSSES (%): 27.0 NUMBEROF CONSUMERS: 469000 NUMBEROF EMPLOYEES: 14000 CONSUMERS/EMPLOYEE: 34 3.8 AVERAGE REVENUE(US CENTS/KWH) ASSETS REVALUED (POLICY)(Y=YES.N=NO): N RATE OF RETURN(%): 5.1 CONTRIBUTION TO INVESTMENT: 16.2 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS GWH 196 91 378 28 693 % 28.2800 13.1300 54.5500 4.0400 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 6.9 8.9 6.9 2.2 1.4 9.6

INSTALLED CAPACITY & GROSS GENERATION MW 83 104 55 222 168 GWH 38 177 el 274 703

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTOR SUPPLY 3.

390 20 32 442

977 60 a5 1102

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE 0.080 0.060 X 50.0000 50.0000

OIL GAS COAL LIGNITE OTHER TOTAL

100.0000

4. OFFICIAL RATES OF EXCHANGE AVE 78 US$16.7 KYATS AVE 1980 US51=B.76 KYATS

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW, .70LF) OTHER

CAMEROON

ELECTRIC SUPPLY IS MANAGEDBY A SINGLE UTILITY SOCIETE NATIONALE D'ELECTRICITE DW CAMEROUN, A PARASTATAL AGENCY 1973-78 AVG ANNUAL GTH RATE ( 2.20 3.20 1.00 5. COMMERCIALENERGY CONSUMPTION MTOE % 0.603 57.2800 OIL : 0.122 11.5900 GAS COAL LIGNITE HYDRO : 0.328 31.1500 NUCLEAR GEOTHERMAL OTHER TOTAL 0 TOTAL B 6. UTILITY DATA 50 FREQUENCY(HZ): TRANSMISSION (KV): 90/30 DISTRIBUTION (KV): 5.5/10/15/30 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (%): 75.0 TOTAL SYSTEM CAPACITY FACTOR (%): 58.0 TOTAL SYSTEM T8D LOSSES (%): NUMBEROF CONSUMERS: 104000 2129 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 49 8.3 AVERAGEREVENUE (US CENTS/KWH) ASSETS REVALUED (POLICY)(Y=YES,N=NO): Y RATE OF RETURN(%): 11.2 CONTRIBUTION TO INVESTMENT: 50.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS 0.022 RATES OF EXCHANGE 100.0000 DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .SSLF) INDUSTRIAL (200KW, .70LF) OTHER GWH 163 11 1019 16 1209 X 13.4800 0.9100 84.2800 1.3200 99.9900 KWH/MONTH US CENTS/KWH 17.7 9.4 0.7 3.0 1.053 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (US$M) : GNP/CAPITA (USS) : ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (%) : ACCESS TO ELECTRICITY (%) : 806b 3700 460 1.053 1434 0.359 34.0 130 178 URBAN URBAN 28

RURAL RURAL

2. INSTALLED CAPACITY & GROSS GENERATION MW THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDP.O THERMAL TOTAL SECTOR SUPPLY GWH

55 55 155

74 74 1310

210 23 233

1384 SO 1434

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL MTOE 0.022 X 100.0000

AVE 78 US$1=226 CFAF AVE 1980 USSi=225.8 CFAF

CHAD

SOCIETE TCHADIENNE D'ENERGIE ELECTRIQUE RESPONSIBLE FOR PUBLIC SERVICE IN N'OJAMENA a ABECHE MAIN SYSTEMANDIN SAHR. MOUNDOU 1973-78 AVG ANNUAL GTH RATE MX 1.90 2.50 0.60 4.00 2.00 5. COMMERCIAL ENERGY CONSUMPTION MTOE 100.0000 0.065 OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 10 6. UTILITY DATA 50 FREQUENCY(HZ): TRANSMISSION (KV): DISTRIBUTION (KV): LOW VOLTAGE (VOLTS): 220/380 50.0 MAIN SYSTEM LOAD FACTOR (X): 34.0 TOTAL SYSTEM CAPACITY FACTOR (%): 10.0 TOTAL SYSTEM T&D LOSSES (%): 4500 NUMBEROF CONSUMERS: 320 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 20 20.0 AVERAGE REVENUE (US CENTS/KWH) Y ASSETS REVALUED (POLICY)(Y-YES,N=NO): 9.9 RATE OF RETURN(X): CONTRIBUTION TO INVESTMENT: 100.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW,.35LF) INDUSTRIAL (200KW, .70LF) OTHER GWH 18 17 2 37 % 48.6500 45.9500 5.4100 100.0100 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 22.0 15.0 18.0 0.065 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) GNP/CAPITA (USS) : ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EOUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%1: ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) (%) POPULATION BREAKDOWN : ACCESS TO ELECTRICITY (% 2. 4320 620 140 0.065 68 0.017 26.1 15 iS URBAN 14 URBAN 70

RURAL 86 RURAL 0

INSTALLED CAPACITY & GROSS GENERATION MW GWH

,x, '4

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERMAL TOTAL SECTOR SUPPLY 3.

21 21

63 63

21 2 23

63 5 68

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE X

OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE

AVE 78 USS1225 CFA FRANCS AVE 1980 USS1=225.8 CFA FR

CHILE

EMPRESANACIONAL DE ELECTRICIDAD SA GENERATES77% OF PUBLIC ELECTRICITY CHILECTRA-PRINCIPAL RETAIL SUPPLIER 1973-78 AVG ANNUAL GTH RATE M% 1.70 -1.80 -1.40 3.10 S. COMMERCIAL ENERGYCONSUMPTION MTOE % 5.060 59.6000 0.630 7.4200 1.120 13.1900 1.680 19.7900

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) GNP/CAPITA (USS) ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) : POPULATION BREAKDOWN (%) ACCESS TO ELECTRICITY (X) 10734 15180 1410 8.490 10360 2.590 30.5 793 965 URBAN 80 URBAN 99

OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL

RURAL 20 RURAL 30

8.490

100.0000

TOTAL

85 8. UTILITY DATA FREQUENCY(HZ): 50 TRANSMISSION (KV): 220/154/110 DISTRIBUTION (KV): LOW VOLTAGE (VOLTS): 380/220 MAIN SYSTEM LOAD FACTOR (%): 61.0 TOTAL SYSTEM CAPACITY FACTOR (%): 38.1 TOTAL SYSTEM T&D LOSSES (%): 13.1 NUMBEROF CONSUMERS: 283075 NUMBEROF EMPLOYEES: 4763 CONSUMERS/EMPLOYEE: 59 AVERAGEREVENUE (US CENTS/KWH) 2.4 ASSETS REVALUED (POLICY)(Y=YES,N=NO): Y RATE OF RETURN(S: 2.3 CONTRIBUTION TO INVESTMENT: 6.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS GWH 1584 478 3376 2047 7485 % 21.1600 6.3900 45.1000 27.3500 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 5.5 5.8 7.3 7.3 8.0 4.5

2. INSTALLED CAPACITY & GROSS GENERATION THERMAL-STEAM -GAS TURBINES L-COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY MW 626 103 109 838 1363 GWH 1103 44 203 1350 5993

t,

2201

7343 829 2188 10360

AUTO PRODUCERS HYDRO 111 THERMAL 627 TOTAL SECTOR SUPPLY 2939

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL MTOE 0.810 0.030 0.290 0.930 RATES OF EXCHANGE PESOS % 85.5900 3.2300 31.1800 100.0000

AVE 78 USS1:33.9 AVE 1980 US$1=39.0

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW, .70LF) OTHER

COLOMBIA

INTECONEXION ELECTRICA SA PROVIDES PUBLIC GENERATION &TRANSMISSION TO NATIONAL ENTERPRISES(ICEL,CVC,CORELCA) 1973-78 AVG AN*IJAL GTH RATE (%) 2.10 5.90 3.80 3.90 4.20 5. COM4MERCIAL ENERGY CONSUMPTION % MTOE 47.SSOO 6.700 OIL 14.4100 2.030 GAS 19.3000 2.720 COAL LIGNITE 18.7400 . 2.640 HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 52 6. UTILITY DATA FREQUENCY (HZ): 60 TRANSMISSION (KV): 230/115 DISTRIBUTION (KV): 34.5/13.8/13.2/11.4 LOW VOLTAGE (VOLTS): 115/230 50.0 MAIN SYSTEM LOAD FACTOR (%): 20.3 TOTAL SYSTEM CAPACITY FACTOR (%): 20.3 TOTAL SYSTEM T8D LOSSES (%): 2300000 NUMBER OF CONSUMERS: 13445 NUMBEROF EMPLOYEES: 171 CONSUMERS/EMPLOYEE: AVERAGEREVENUE (US CENTS/KWH) Y ASSETS REVALUED (POLICY)(Y=YES.N=NO): RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW, .70LF) OTHER GWH 5725 1741 4131 1274 12871 X 44.4800 13.5300 32.1000 9.9000 100.0100 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 0.6 0.6 0.8 5.0 4.4 4.4 3.1 14.090 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (US$M) GNP/CAPITA (USS) : ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE (GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPlTA LKGOE) ELECTRICITY/CAPITA (KWH) (%) POPULATION BREAKDOWN ACCESS TO ELECTRICITY (X) 2. 25138 21790 870 14.090 17377 4.340 30.8 581 691 URBAN 64 URBAN 8B

RURAL 38 RURAL 19

INSTALLED CAPACITY & GROSS GENERATION MW 717 285 43 1045 2642 GWH 2928 1191 115 4234 11893 25 16152 88 1137 17377

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY

3687

20 AUTO PRODLUCERS HYORO THERMAL 223 3930 TOTAL SECTORSUPPLY 3.

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE %

OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE AVE 75 US$1=39.3 PESOS AVE 1980 US$1=50.9 PESOS

COSTA RICA

INSTITUTO COSTARRICENSEDE ELECTRICIDAD GENERATES85% AND DISTRIBUTES 24% OF PUBLIC ELECTRICITY. INTERCONNECTED SYSTEMS 1973-78 AVG ANN%AL GTH RATE (%) 2.50 5.70 3.20 5. COMMERCIALENERGYCONSUMPTION MTOE OIL: 0.680 63.5500 GAS COAL LIGNITE HYDRO 0.390 36.4500 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 70 5. UTILITY-DATA FREQUENCY (HZ): 80 TRANSMISSION (KV): 138/34.5 OISTRIBUTION (KV): 34.5/24.9113.2/4.2 LOW VOLTAGE (VOLTS): 110/220 MAIN SYSTEM LOAD FACTOR (%): 58.7 TOTAL SYSTEM CAPACITY FACTOR (%): 45.2 TOTAL SYSTEM T8D LOSSES (): 13.7 NUMBER OF CONSUMERS: 298414 NUMBEROF EMPLOYEES: 1841 CONSUMERS/EMPLOYEE: 162 AVERAGEREVENUE (US CENTS/KWH) 17.4 ASSETS REVALUED (POLICY)(Y-YES,N=NO): Y RATE OF RETURN(%): 12.7 CONTRIBUTION TO INVESTMENT: 21.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW..35LF) INDUSTRIAL (208KW, .7LF) OTHER GWH 718 299 528 44 1585 % 45. 1700 18.8600 33.1800 2.7800 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 5.7 1.070 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000): GNP (US$M) GNP/CAPITA (USS) ENERGY CONSUMED (MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) : POPULATION BREAKDOWN (%) ACCESS TO ELECTRICITY (%) : 2. 2110 3250 1540 1.070 2079 0.520 48.6 507 747 URBAN 45 URBAN 95

RURAL 55 RURAL 52

INSTALLED CAPACITY & GROSS GENERATION MW 10 80 170 294 1453 GWH

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE N -SUBTOTAL

~~~~~-DIESEL 80

HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTOR SUPPLY

484 55 519

1453 240 1693

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE %

AVE 79 US $1=B.57 COLONES AVE 1980 USS1=8.57 COL

S.4

CYPRUS

ELECTRICITY AUTHORITY OF CYPRUS WHICH DOES NOT HAVE ACCESS TO THE 36% OF ITS TOTAL AREA UNDER TURKISH CYPRIOT CONTROL 1973-78 AVG ANNUAL GTH RATE (%) 0.70 1.00 0.30 1.30 2.10 S. COMMERCIAL ENERGY CONSUMPTION X MTOE 100.0000 : 0.824

1. SOCIO-ECONOMIC DATA 646 POPULATION (000) 1370 GNP (USSM) 2110 : GNP/CAPITA (USS) 0.824 : ENERGYCONSUMED(MTOE) 926 ELECTRICTY PRODUCE(GWH) 0.230 ELECTRICITY OIL EQUIV (MTOE): 27.9 ELECTRICITY/TOTAL ENERGY 1%): 1275 : ENERGY/CAPITA (KGOE) 1433 : ELECTRICITY/CAPITA (KWH) IR8AN 5e 1XI POpULATION BREAKDOWN URBAN 100 ACCESS TO ELECTRICITY (%) 2. INSTALLED CAPACITY & GROSS GENERATION THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERMAL SUPPLY SECTOR TOTAL MW 264 GWH 914 :

OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL

RURAL 42 RURAL 98

0.824 DATA

100.0000

TOTAL 99 S. UTILITY 50 FREQUENCY(HZ): TRANSMISSION (KV): 132/60 DISTRIBUTION (KV): LOW VOLTAGE (VOLTS): 240/415 58.0 MAIN SYSTEM LOAD FACTOR (%): 40,0 TOTAL SYSTEM CAPACITY FACTOR (%): 10.8 TOTAL SYSTEM T&D LOSSES (X): 158000 NUMBEROF CONSUMERS: 1425 NUMBEROF EMPLOYEES: 111 CONSUMERS/EMPLOYEE: 5.8 AVERAGE REVENUE(US CENTS/KWH) N ASSETS REVALUED (POLICY)(Y=YES,N=NO): 5.5 RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 100.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL B. TARIFFS GWH 163 183 248 38 532 X 25.7900 28.9600 39.2400 5.0100 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 7.2 8.1 7.3 13.4

-P1

264

914

264 5 269

914 12 926

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL MTOE 0.259 X 100.0000

O.259 RATES OF EXCHANGE

100.0000

AVE 78 USS1=.373 CYPRUSL AVE 1980 US$1=.353 C L

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW,.7OLF) OTHER

DOMINICAN REPUBLIC

CORPORACIONDOMINICA DE ELECTRICIDAD ALMOST TOTALLY INTERCONNECTED SYSTEM

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) GNP/CAPITA (USS) ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRrCrTY oIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) : POPULATION BREAKDOWN (x) ACCESS TO ELECTRICITY (X) 4980 4680 91o 1.630 3624 0.906 55.5 327 728 URBAN 46 URBAN

1973-78 AVG ANNUAL GTH RATE (x) 2.90 6.20 3.30 3.60 4.20

5.

OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER: TOTAL

COMMERCIAL ENERGYCONSUMPTION MTOE X 1.610 98.7700

0.000 0.020 1.2300

RURAL 54 RURAL

1.630 DATA

100,0000

TOTAL

35 S. UTILITY FREQUENCY(HZ): 60 TRANSMISSION (KV): 69 DISTRIBUTION (KV): LOW VOLTAGE (VOLTS): 110/220 MAIN SYSTEM LOAD FACTOR (x): 64.0 TOTAL SYSTEM CAPACITY FACTOR (x): 41.0 TOTAL SYSTEM T8D LOSSES (%): 8.4 NUMBEROF CONSUMERS: 343518 NUMBER OF EMPLOYEES: 4687 CONSUMERS/EMPLOYEE: 74 AVERAGEREVENUE (US CENTS/KWH) 5.2 N ASSETS REVALUED (POLICY)(Y=YES,N=NO): RATE OF RETURN(X): 0.0 CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS GWH 639 214 618 203 1674 % 38.1700 12.7800 36.9200 12.1300 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 200 100000 4.6 6.1 5.1 12.7

2. INSTALLED CAPACITY & GROSS GENERATION THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERMAL TOTAL SECTOR SUPPLY MW 363 121 24 508 159 GWH 1726 313 7 2046 55 293 2394 1230 3624

667 320 987

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL MTOE 0.580 X 100.0000

0.580

100.0000

4. OFFICIAL RATES OF EXCHANGE AVE 78 USS1=1 PESO AVE 1980 USS1I1 PESO

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (BKW,.35LF) INDUSTRIAL (200KW, .70LF) OTHER

ECUADOR

INSTITUTO ECUATORIANO0D ELECTRIFICATION IS RESPONSIBLE FOR POWERSECTOR. 84 DIFF SUPPLIERS OPERATE SMALL ISOLATED SYSTEMS 1973-78 AVG ANNUAL GTH RATE (% 3.00 9.10 B.10 13.00 14.20 S. COMMERCIAL ENERGYCONSUMPTION MTOE 91.6400 2.520 OIL 1.4500 0.040 GAS COAL LIGNITE 6.9100 : 0.190 HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 38 6. UTILITY DATA SO FREQUENCY(HZ): TRANSMISSION (KV): 230/138/69 13.8/6.6/4.4/2.2 DISTRIBUTION (KV): LOW VOLTAGE (VOLTS): 110/220 52.0 MAIN SYSTEM LOAD FACTOR (X): 50.0 TOTAL SYSTEM CAPACITY FACTOR (%): 15.2 TOTAL SYSTEM T&D LOSSES (%): NUMBEROF CONSUMERS: 581000 5800 NUMBEROF EMPLOYEES: 100 CONSUMERS/EMPLOYEE: 4.4 AVERAGE REVENUE (US CENTS/KWH) ASSETS REVALUED (POLICY)(Y=YES,N'NO): 4.0 RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS 0.514 100.0000 DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW,.70LF) OTHER GWH 792 300 863 226 2181 % 36.3100 13.7600 39.5700 10.3600 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 4.7 5.4 5.0 8.9 0.2 4.2 2.750 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) GNP/CAPITA (USS) ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE (GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): : ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) (%) POPULATION BREAKDOWN ACCESS TO ELECTRICITY (%) 7550 6890 910 2.750 2570 0.642 23.4 364 333 URBAN 42 URBAN 72

RURAL RURAL

58 30

2. INSTALLED CAPACITY A GROSS GENERATION MW 138 174 243 553 204 759 GWH

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY HYORO AUTO PRODUCERS THERMAL TOTAL SECTOR SUPPLY

757 10 90 857

759 225 984

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL MTOE 0.514 % 100.0000

4. OFFICIAL RATES OF EXCHANGE AVE 78 USS125 SUCRES AVE 1980 USS1=25 SUCRES

EGYPT

EGYPTIAN ELECTRICITY AUTHORITY-BULK SUPPLY TO THE WHOLECOUNTRYAND DISTRIBUTION TO CAIRO & ALEXANDRIA 1973-78 AVG ANNUAL GTH RATE (X) 2.10 7.30 5.20 11.90 10.80 S. COMMERCIALENERGY CONSUMPTION MTOE % OIL 9.894 70.5900 GAS 0.951 6.7800 COAL 0.862 6.1500 LIGNITE HYDRO 2.310 16.4800 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 38 6. UTILITY DATA FREQUENCY(HZ): 50 TRANSMISSION (KV): 500/220/132/66 DISTRIBUTION (KV): 33/11.6/6.6/3 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (%): 67.1 TOTAL SYSTEM CAPACITY FACTOR (%): 49.2 TOTAL SYSTEM T&D LOSSES (%): 9.9 NUMBEROF CONSUMERS: 4390000 NUMBEROF EMPLOYEES: 48000 CONSUMERS/EMPLOYEE: 91 AVERAGEREVENUE (US CENTS/KWH) 2.1 ASSETS REVALUED (POLICY)(Y=YES,N=NO): RATE OF RETURN(%): 5.8 CONTRIBUTION TO INVESTMENT: 6.8 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS GWH 3051 8068 1488 12607 X 24.2000 64.0000 11.8000 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 2.3 2.4 2.6 3.6 2.9 2.7 1.4 14.017 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) : GNP/CAPITA (USS) ENERGY CONSUMED(MTOE) : ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (%) ACCESS TO ELECTRICITY (X) 38688 15520 400 14.020 16899 4.220 30.2 362 437 URBAN 45 URBAN 62

RURAL 55 RURAL 18

2. INSTALLED CAPACITY & GROSS GENERATION MW THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY 2445 GWH 6393 94B 7339 9000

kJ Oc

2445

16339 0 560 16899

AUTO PRODUCERS HYDRO 0 THERMAL 153 TOTAL SECTORSUPPLY 2598 3.

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE X

OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE

AVE 78 USS1=.39 E L AVE 1980 USS1=.70 E L

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW,.70LF) OTHER

EL SALVADOR

COMISION EdECITIVA HIDROELECTRICA DEL RIO LEMPA-GENERATES 95% OF COUNTRY'S SUPPLY SELLS IN BULK TO S PRIVATE COMPANIES 1973-7a AVG ANJWAL GTH RATE M% 3.10 5.20 2.10 5.00 9.70 5. COMMERCZAL ENERGYCONSUMPTrON M4TOE OIL 0.734 70.6400 GAS COAL LIGNITE HYDRO 0.207 19.9200 NUCLEAR GEOTHERMAL : 0.098 9.4300 OTHER TOTAL TOTAL 36 6. UTILITY DATA 1 039 9900

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (US$M) GNP/CAPITA (US$) : ENERGYCONSUMED(MTOE) : ELECTRICTY PRODUCE (GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (%) : ACCESS TO ELECTRICITY (%) : 2. 4382 2810 600 1.039 1450 0.362 34.9 237 331 URBAN 39 URBAN 75

RURAL 61 RURAL 12

INSTALLED CAPACITY a GROSS GENERATION MW 63 65 128 232 60 420 I 54 475 GWH 152 4 156 829 391 1376 74 1450

Nz

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTOR SUPPLY

FREQuENcy (HZ): 60 TRANSMISSION (KV): 115/69/44 DISTRIBUTION (KV): 46/23/13.2 LOW VOLTAGE (VOLTS): 120/240 MAIN SYSTEM LOAD FACTOR (%): 61.0 TOTAL SYSTEM CAPACITY FACTOR (%): 31.4 TOTAL SYSTEM TBO LOSSES (%): 8.8 NUMBEROF CONSUMERS: 274000 NUMBEROF EMPLOYEES: 1873 CONSUMERS/EMPLOYEE: 138 AVERAGEREVENUE (US CENTS/KWH) 3.6 ASSETS REVALUED (PDLICY)(Y=YES,N=NO): N RATE OF RETURN(%): 16.0 CONTRIBUTION tO INVESTMENT: 50.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 GWH %

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL MTOE 0.043 X 100.0000

0.043

100.0000

4. OFFICIAL RATES OF EXCHANGE AVE 78 USS1=2.5 COLONES AVE 1980 USSI=2.5 COL

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW,.7OLF) OTHER

3.6

ETHIOPIA

ETHIOPIAN ELECTRIC LIGHT AND POWER AUTHORITY OPERATES ONE INTERCONNECTED SYSTEM (88% OF INSTALLED CAPACITY) 1973-78 AVG ANNUAL GTH RATE (%) 2.50 2.70 0.20 -6.00 1.00 S. COMMERCIAL ENERGYCONSUMPTION MTOE % : 0.364 79.8500

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) GNP/CAPITA (USS) : ENERGY CONSUMED(MTOE) : ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (%) : ACCESS TO ELECTRICITY (X) : 2. 31011 3640 120 0.457 690 0.173 37.9 Is 22 URBAN 11 URBAN 33

OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL

0.093

20.3500

RURAL 89 RURAL 0

0.457

t00.0000

TOTAL

4 5. UTILITY DATA FREQUENCY (HZ): 50 TRANSMISSION (KV): 230/132/45 DISTRIBUTION (KV): LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (%): 60.0 TOTAL SYSTEM CAPACITY FACTOR (%): 19.0 TOTAL SYSTEM T&D LOSSES (X): NUMBER OF CONSUMERS: 199700 NUMBEROF EMPLOYEES: 3330 CONSUMERS/EMPLOYEE: 60 AVERAGEREVENUE (US CENTS/KWH) 5.3 ASSETS REVALUED (POLICY)(Y=YES,N=NO): N RATE OF RETURN(%); 2.8 CONTRIBUTION TO INVESTMENT: 73.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW..35LF) INDUSTRIAL (200KW,.70LF) OTHER GWH 163 a8 172 8 429 % 38.0000 20.0500 40.0900 1.8600 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 7.3 7.3 5.4 6.6 8.4 7.2 5.3

INSTALLED CAPACITY & GROSSGENERATION MW 8 40 45 249 374 GWH

00

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTOR SUPPLY 3.

295 47 342

374 200 574

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE X

OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE

AVE 78 USS1=2.07 BIRRS AVE 1980 USS1=2.07 BIRRS

FIJI

UNTIL MID-78 SUVA CITY COUNCIL ELEC DEPT AND FIJI ELEC AUTHORITY PRODUCED & DISTRIBUTED ALL ELECTRICITY 1973-78 AVG ANNUAL GTH RATE (X) 1.80 5.60 3.70 3.70 7.70 5. COMMERCIALENERGYCONSUMPTION MTOE X *0.18 95.8800 OIL GAS COAL 0.008 4.1200 LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 50 S. UTILITY DATA FREQUENCY(HZ): 50 TRANSMISSION (KV): DISTRIBUTION (KV): 33/11 LOW VOLTAGE (VOLTS): 230/400 MAIN SYSTEM LOAD FACTOR (%): 65.0 TOTAL SYSTEM CAPACITY FACTOR (%): 28.5 TOTAL SYSTEM TBD LOSSES (X): 8.0 NUMBEROF CONSUMERS: 33000 1092 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 30 AVERAGE REVENUE (US CENTS/KWH) 10.9 ASSETS REVALUED (POLICY)(Y=YES,N=NO): Y RATE OF RETURN(%): 6.2 CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW,.35LF) INDUSTRIAL (200KW,.70LF) OTHER GWH 44 139 183 X 24.0400 75.9500 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 13.3 13.3 13.3 17.0 14.8 10.5 0.194 100.0000

1. SOCIO-ECONOMIC DATA 599 POPULATION (000) : 880 GNP (USSM) GNP/CAPITA (US$) 1440 ENERGY CONSUMED(MTOE) 0.194 ELECTRICTY PRODUCE(GWH) 287 ELECTRICITY OIL EQUIV (MTOE): 0.072 ELECTRICITY/TOTAL ENERGY (X): 37.0 324 ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) : 479 39 POPULATION BREAKDOWN (X) :URBAN URBAN ACCESS TO ELECTRICITY (X) 2. INSTALLED CAPACITY 8 GROSS GENERATION MW THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERMAL TOTAL SECTORSUPPLY GWH

RURAL 61 RURAL

-.

8S 85

197 197

85 27 112

197 90 287

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE %

AVE 78 USS1=.84 FS AVE 1980 USSI=.82 FS

GHANA

VOLTA RIVER AUTHORITY SUPPLIES BULK POWERTO ELECTRICITY CORP OF GHANA TO THE VOLTA ALUMINUM COMPANY MINES 1973-78 AVG ANNUAL GTH RATE ()MTOE 3.00 1.00 -2.00 3.00 2.00 5. COMMERCIAL ENERGYCONSUMPTION OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 13 6. UTILITY DATA FREQUENCY(HZ): 50 TRANSMISSION (KV): 161/35 DISTRIBUTION (KV): 34.5/33/11 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAO FACTOR (X): 65.0 TOTAL SYSTEM CAPACITY FACTOR (%): 46.1 TOTAL SYSTEM T&D LOSSES (%): 12.2 NUMBER OF CONSUMERS: 150000 NUMBEROF EMPLOYEES: 8520 CONSUMERS/EMPLOYEE: 23 AVERAGEREVENUE (US CENTS/KWH) 0.8 ASSETS REVALUED (POLICY)(Y-YES,N=NO): Y RATE OF RETURN(%): 1.9 CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS 0.014 RATES OF EXCHANGE 100.0000 DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW,.35LF) INDUSTRIAL (200KW, .70LF) OTHER GWH 318 215 2780 13 3326 % 9.5600 6.4600 83.5800 0.3900 99.9900 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 3.8 2.7 2.1 6.7 B.O 2.0 0.5 0.880 : 0.005 1.029 45.4100 0.2600 54.3300

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) GNP/CAPITA (USS) ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (%) ACCESS TO ELECTRICITY (X) 2. 10972 4250 390 1.890 3541 0.885 46.7 173 325 URBAN 33 URBAN 30

RURAL 67 RURAL 4

1.894

100.0000

INSTALLED CAPACITY & GROSS GENERATION MW GWH

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTORSUPPLY

83 83 790

SO SO 3718 -233 3535 a 3541

873 25 898

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL MTOE 0.014 X 100.0000

AVE 78 USS1=1.51 CEDIS AVE 1980 USS1=2.75 CEDZS

GUATEMALA

INSTITUTO NATIONAL DE ELECTRIFICATION 6sV OF TOTAL GENERATION EMPRESADE GUATEMALA-30%OF TOTAL GENERATION 1973-78 AVG ANNUAL GTH RATE (%) 2.90 5.20 3.30 4.40 5.10 S. COMMERCIAL ENERGY CONSUMPTION % MTOE 95.6700 1.990

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) GNP/CAPITA (USS) ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY i%}! : ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) t%( POPULATION BREAKOOWN ACCESS TO ELECTRICITY (%) 2. 6627 6207 937 2.0P0 1576 0.394 31.0 314 239 URBAN 37 URBAN 26

OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL

0.090

4.3300

RURAL 63 RURAL 5

2.080

100.0000

TOTAL

13 6. UTILITY DATA e0 FREQUENCY (HZ): TRANSMISSION (KV): 138 69/50/34.5/22 (KV): DISTRIBUTION LOW VOLTAGE (VOLTS): 110 61.4 MAIN SYSTEM LOAD FACTOR (V): 44.9 TOTAL SYSTEM CAPACITY FACTOR (%): 13.7 TOTAL SYSTEM T&8 LOSSES (X): 256219 NUM8ER OF CONSUMERS: 3128 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 82 6.4 AVERAGE REVENUE (US CENTS/KWH) ASSETS REVALUEO (POLICY)(Y=YES,N=NO): 5.9 RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 15.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .3SLF) INDUSTRIAL (200KW, .70LF) OTHER KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 14.7 13.0 10.8 10.6 13.3 11.3 GWH V

INSTALLED CAPACITY S GROSSGENERATION MW 82 80 60 11 233 121 GWH 854 151 1005 341

,s,

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERMAL TOTAL SECTOR SUPPLY 3.

354 138 492

1346 230 1576

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE %V

OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE

AVE 78 USS1=1 QUETZAL AVE 1980 USSI=1 QUET

GUINEA

SOCIETE NATIONALE D'ELECTRICITE FULLY OWNEDSTATE ENTERPRISE BY GOV'T OPERATES GUINEANPOWER UTILITIES 1973-78 AVG ANNUAL GTH RATE ()MTOE 3.00 5.50 2.50 2.40 1.20 S. COMMERCIAL ENERGYCONSUMPTION OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 5 6. UTILITY DATA 50 FREQUENCY (HZ): TRANSMISSION (KV): 110/60 DISTRIBUTION (KV): 30/15/6.3 LOW VOLTAGE (VOLTS): 220/440 MAIN SYSTEM LOAD FACTOR (%1: 63.0 26.3 TOTAL SYSTEM CAPACITY FACTOR (%): TOTAL SYSTEM T8D LOSSES (%): NUMBEROF CONSUMERS: 23000 1260 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 18 AVERAGEREVENUE (US CENTS/KWH) 11.1 N ASSETS REVALUED (POLICY)(Y=YES,N=NO): 57.0 RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COSMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS 0.003 100.0000 DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW,.7OLF) OTHER GWH 40 8 42 32 122 % 32.7900 6.53 00 34.4300 26.2300 100.0100 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 0.288 89.7200

1. SOCIO-ECONOMIC OATA POPULATION (000) GNP (USSM) GNP/CAPITA (USS) ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): : ENERGY/CAPITA (KGODE) ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (%) : ACCESS TO ELECTRICITY (%) : 2. 5133 1070 210 0.321 549 0.137 42.8 63 107 URBAN URBAN

0.033

10.2800

RURAL RURAL

0.321

100.0000

INSTALLED CAPACITY & GROSS GENERATION MW GWH

t'i

00

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERMAL TOTAL SECTOR SUPPLY 3.

25 25 18 131 174 e6 240

25 25 40

65 400 465

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE 0.003 % 100.0000

OIL GAS COAL LIGNITE OTHER TOTAL

4. OFFICIAL RATES OF EXCHANGE AVE 78 USS1=19 SYLIS AVE 1980 US51=19 SYLIS

GUYANA

GUYANA ELECTRICITY CORP-MAJORGENERATOR AND DISTRIBUTOR OF ELECTRICITY OPERATES 5 SYSTEMS 1973-78 AVG ANNUAL GTH RATE (I)MTOE 2.00 2.40 0.40 0.00 3.30 S. COMMERCIALENERGYCONSUMPTION IC OIL 0.597 100.0000 GAS COAL LIGNITE HYORO 0.000 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL S0 6. UTILITY DATA FREQUENCY(HZ): 50 TRANSMISSION (KV): 69 DISTRIBUTION (KV): 13.8/11 LOW VOLTAGE (VOLTS): 110/220 MAIN SYSTEM LOAD FACTOR (X): 85.0 TOTAL SYSTEM CAPACITY FACTOR (X): 23.8 TOTAL SYSTEM T&D LOSSES (IC: 18.0 NUMBEROF CONSUMERS: 84100 NUMBEROF EMPLOYEES: 1500 CONSUMERS/EMPLOYEE: 56 AVERAGEREVENUE (US CENTS/KWH) 9.5 ASSETS REVALUED (POLICY)(Y=YES,N=NO): N RATE OF RETURN(%): 0.0 CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS 0.010 0.120 8.3300 100.0000 DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW,.70LF) OTHER GWH ES 98 18 179 X 35.3100 54.7500 8.9400 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 14.3 14.3 14.3 18.1 15.1 0.5.7 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (US$M) GNP/CAPITA (USS) ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) : ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) : POPULATION BREAKDOWri(g) : ACCESS TO ELECTRICITY (X) : 2. 830 480 550 0.597 448 0.112 18.8 714 538 URBAN 80 URBAN 60

RURAL 40 RURAL 5

INSTALLED CAPACITY & GROSSGENERATION MW 30 22 GWH 85 40 127 232 B

X0 t-11

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY

s0 ~~~-SUBTOTAL 112 2

114 35 149

238 210 44E

AUTO PRODUCERS HYDRO THERMAL TOTAL SECTOR SUPPLY

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL MTOE 0.110 X 91.6700

4. OFFICIAL RATES OF EXCHANGE AVE 78 USSi=2.5S GUYANAS AVE 1980 USS1=2.55 G$

HAITI

ELECTRICITE D'HAITI-GOV'T AGENCY RESPONSIBLE FOR PUBLIC POWERSUPPLY OPERATES PORT AU PRINCE-PELIGRE 1973-78 AVG ANNUAL GTH RATE (M 1.70 3.70 2.10 7.40 14.40 S. COMMERCIALENERGY CONSUMPTION TOE K OIL : 0.168 75.6800 GAS COAL LIGNITE HYDRO : 0.054 24.3200 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 5 6. UTILITY DATA FREQUENCY (HZ): 60 TRANSMISSION (KV): 115 DISTRIBUTION (KV): 12.5/7.2/4.2/2.4 LOW VOLTAGE (VOLTS): 120/240 MAIN SYSTEM LOAD FACTOR (X): 53.0 TOTAL SYSTEM CAPACITY FACTOR (X): 32.0 TOTAL SYSTEM T&D LOSSES (X): 30.0 NUMBEROF CONSUMERS: 598si NUMBER OF EMPLOYEES: 901 CONSUMERS/EMPLOYEE: 66 AVERAGE REVENUE (US CENTS/KWH) 7.1 ASSETS REVALUED (POLICY)(Y=YES,N=NO): Y RATE OF RETURN(K: 4.9 CONTRIBUTION TO INVESTMENT: 38.0 7. SALES DOMESTIC COMMERCIAL INDUJSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW,.35LF) INDUSTRIAL (200KW,.70LF) OTHER KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 9.5 10.8 11.9 10.9 7.6 GWH K 0.222 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) : GNP/CAPITA (US$) ENERGY CONSUMED(MTOE) : ELECTRICTY PRODUCE(GWH) : ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) : POPULATION BREAKDOWN (X) : ACCESS TO ELECTRICITY () : 4831 1240 260 0.222 276 0.069 31.1 46 57 URBAN 23 URBAN 23

RURAL 77 RURAL 0

2. INSTALLED CAPACITY S GROSS GENERATION MW THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTOR SUPPLY GWH

X0

45 45 45

30 30 221

90 22 112

251 28 276

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE X

AVE 78 USS1=B GOURDES AVE 1980 USS1=S GOURDES

HONDURAS

EMPRESANACIONAL DE ENERGIA ELCTRICA PROVIDES 100% OF PUBLIC ELCTRICITY SERVICES-MAIN INTERCONNECTED SYSTEMS 1973-78 AVG ANNUAL GTH RATE (%) 3.30 3.30 6.40 11.60 S. COMMERCIAL ENERGYCONSUMPTION MTOE X OIL 0.608 83.1700 GAS COAL LIGNITE HYDRO 0.123 16.8300 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 21 6. UTILITY DATA FREQUENCY (HZ): 60 TRANSMISSION (KV): 138/69 DISTRIBUTION (KV): 34.5/13.8/4.2/2.4 LOW VOLTAGE (VOLTS): MAIN SYSTEM LOAD FACTOR (X): 64.6 TOTAL SYSTEM CAPACITY FACTOR (X): 43.5 TOTAL SYSTEM T&D LOSSES (X): 16.2 NUMBEROF CONSUMERS: 118000 NUMBEROF EMPLOYEES: 1747 CONSUMERS/EMPLOYEE: 68 AVERAGE REVENUE (US CENTS/KWH) 4.8 ASSETS REVALUEO (POLICY)(Y=YES,N=NO): Y RATE OF RETURN(X): 7.0 CONTRIBUTION TO INVESTMENT: 27.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW, .70LF) OTHER GWH 150 93 87 274 604 % 24.8300 15.4000 14.4000 45.3600 99.9900 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 6.7 7.9 6.5 6.4 6.1 3.8 0.73t 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) : GNP/CAPITA (USS) ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) : POPULATION BREAKDOWN (X) : ACCESS TO ELECTRICITY (%) 3441 1650 480 0.731 775 0.194 26.5 213 225 URBAN 32 URBAN

RURAL 68 RURAL

2. INSTALLED CAPACITY & GROSS GENERATION MW ,'.) X0 THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY AUTO PROWUCERS HYDRO THERMAL TOTAL SECTORSUPPLY 3. 29 52 81 108 647 8 655 5S 710 GWH

189 13 202

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE X

OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE

AVE 78 USS12 LEMPIRAS AVE 1980 USS1=2 LEMPIRAS

INDIA

21 STATE ELECTRICITY BOARDS GENERATE AND DISTRIBUTE 90%+ OF ELECTRICITY CENTRAL ELECTRICITY AUTH RESP NATIONWIDE 1973-78 AVG ANNUAL GTH RATE (%) 2.10 3.20 1.10 4.50 8.10 S. COMMERCIALENERGY CONSUMPTION MTOE % OIL 1.024 1.6700 GAS COAL 50.681 82.4700 LIGNITE HYDRO 9.747 15.8B00 0.875 1.4200 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 14 6. UTILITY DATA 50 FREQUENCY(HZ): TRANSMISSION (KV): 220/132 DISTRIBUTION (KV): 33/11 LOW VOLTAGE (VOLTS): 240/415 MAIN SYSTEM LOAD FACTOR (X): 72.0 TOTAL SYSTEM CAPACITY FACTOR (X): 44.0 TOTAL SYSTEM T&D LOSSES (%): NUMBER OF CONSUMERS: 28000000 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 3.5 AVERAGEREVENUE (US CENTS/KWH) N ASSETS REVALUED (POLICY)(Y=YES,N=NO): RATE OF RETURN(): 9.5 CONTRIBUTION TO INVESTMENT: 10.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUJSTRIAL (8KW, .3SLF) INDUSTRIAL (200KW, .7LF) OTHER GWH 7588 4801 47857 17192 77439 X 9.8000 8.2000 61.8000 22.2000 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 4.9 4.4 4.4 7.3 3.8 3.5 2.3 61.452 101.4200

1. SOCIO-ECONOMIC DATA 643896 POPULATION (000) 112660 GNP (USSM) GNP/CAPITA (USS) : 180 84.150 ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) :10342 27.300 ELECTRICITY OIL EQUIV (MTOE): 32.5 ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) : 131 170 (KWH) ELECTRICITY/CAPITA URBAN 21 POPULATION BREAKDOWN (%) : URBAN ACCESS TO ELECTRICITY (X) 2. INSTALLED CAPACITY & GROSS GENERATION MW
t')

RURAL 79 RURAL

GWH

00 00

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY 840 640 20 6890 6910

HYDRO AUTO PRODUCERS THERMAL 2225 2865 TOTAL SECTOR SUPPLY

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL MTOE 1.790 0.180 19.360 0.540 21.870 X 8.1800 0.8200 88.5200 2.4700 99.9900

4. OFFICIAL RATES OF EXCHANGE AVE 78 USSt=8.2 RUPEES AVE 1980 USS157.9 RUPEES

INDONESIA

PERUSAHAANUMUMLISTRIK NEGARA SOLE GENERATOR AND DISTRIBUTOR OF PUBLIC ELEC-NO INTERCONNECTED SYSTEM 1973-78 AVG ANNUAL GTH RATE (%) 1.80 7.60 5.70 18.40 8.20 CONSUMPTION S. COMMERCIAL ENERGY MTOE % 83.8600 OIL 23.430 3.984 14.2600 GAS 0.179 0.6400 COAL LIGNITE 1.2400 0.346 HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL B 8. UTILITY DATA FREQUENCY (HZ): 50 TRANSMISSION (KV): 150/70 DISTRIBUTION (KV): NUMEROUS LOW VOLTAGE (VOLTS): NUMEROUS MAIN SYSTEM LOAD FACTOR (%): 67.0 TOTAL SYSTEM CAPACITY FACTOR (%): 24.5 12.7 TOTAL SYSTEM TSD LOSSES (X): NUMBEROF CONSUMERS: 1784000 NUMBEROF EMPLOYEES: 28800 CONSUMERS/EMPLOYEE: 62 AVERAGEREVENUE (US CENTS/KWH) 0.2 Y ASSETS REVALUED (POLICY)(Y=YES,N=NO): 3.1 RATE OF RETURN(X): CONTRIBUTION TO INVESTMENT: 29.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS 1.103 100.0000 DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (BKW, .35LF) INDUSTRIAL (200KW,.70LF) OTHER GWH 2473 1664 146 4283 S 57.7400 38.8500 3.4100 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 2000 100000 100000 4.8 4.8 3.8 3.8 27.939 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) GNP/CAPITA (USS) : ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (%) ACCESS TO ELECTRICITY (%) 2. 135993 48820 360 27.939 5210 1.303 4.7 205 38 URBAN 18 URBAN

RURAL 82 RURAL

INSTALLED CAPACITY & GROSS GENERATION MW 556 882 500 1938 350 GWH 1523 1157 846 3526 1384

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY

2288

4910 300 5210

AUTO PRODUCERS HYDRO THERMAL 1855 TOTAL SECTOR SUPPLY 4143 3.

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE 1.103 X 100.0000

OIL GAS COAL LIGNITE OTHER TOTAL

4. OFFICIAL RATES OF EXCHANGE AVE 78 US$1=442 RUPIAHS AVE 1980 USS1=B27 RUPIAH

IRELAND

ELECTRICITY SUPPLY BOARD GENERATES AND DISTRIBUTES ALL PUBLIC UTILITIES 1973-78 AVG ANNUAL GTH RATE (X) 1.20 3.30 2.10 1.30 6.00 S. COMMERCIAL ENERGYCONSUMPTION MTOE % 5.003 67.5300 2.142 0.264 28.9100 3.5600

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) GNP/CAPITA (USS) ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) : ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (X) ACCESS TO ELECTRICITY (X) 3234 11210 3470 7.409 10000 2.500 33.7 2290 3092 URBAN 60 URBAN 100

OIL GAS COAL LIGNITE HYDRO NUCLEAR: GEOTHERMAL OTHER TOTAL

RURAL 40 RURAL 100

7.409 DATA

100.0000

TOTAL 100 6. UTILITY FREQUENCY (HZ); 50 TRANSMISSION (KV): 275/220/110 DISTRIBUTION (KV): 38/10/5/3.3 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (X): 56.9 TOTAL SYSTEM CAPACITY FACTOR (%): 38.8 TOTAL SYSTEM T&D LOSSES (5): 11.8 NUMBEROF CONSUMERS: 1015636 NUMBEROF EMPLOYEES: 11297 CONSUMERS/EMPLOYEE: 89 AVERAGEREVENUE (US CENTS/KWH) 5.8 ASSETS REVALUED (POLICY)(Y=YES.N'NO): N RATE OF RETURN(%): 8.5 CONTRIBUTION TO INVESTMENT: 20.2 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW. .35LF) INDUSTRIAL (200KW,.70LF) OTHER GWH 3403 1490 2747 7640 X 44.5400 19.5000 35.9600 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 4.8 4.7 4.7 8.9 8.9 7.4 2.7

2. INSTALLED CAPACITY & GROSS GENERATION MW THERMAL-STEAM 2270 -GAS TURBINES 85 -COMBINED CYCLE -DIESEL 28 -SUBTOTAL 2383 HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY 511 GWH

OZ)

1046

2894

1046 4 166 1216

AUTO PRODUCERS HYDRO 1 THERMAL 64 TOTAL SECTORSUPPLY 2959

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE X

AVE 78 USSI=.521 L AVE 1980 USS1 .485 L

IVORY COAST

ENERGIE ELECTTRIQUE DE LA COTE D'IVOIRE SYSTEM OPERATES THE MAIN INTERCONNECTED

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) GNP/CAPITA (USS) ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): : ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) : (%) POPULATION BREAKDOWN : ACCESS TO ELECTRICITY (%) 2. 7836 6580 840 1.334 1209 0.342 25.6 170 154 URBAN 35 URBAN 90

1973-78 AVG ANNUAL GTH RATE (%) 6.00 7.10 1.00 6.80 9.40

5.

OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL

COMMERCIALENERGY CONSUMPTION X MTOE 94.3800 1.259

0.075

5.8200

RURAL 65 RURAL 20

1.334 DATA

100.0000

TOTAL

45 6. UTILITY 50 FREQUENCY (HZ): TRANSMISSION (KV): 220/90 DISTRIBUTION (KV): 33/15 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (X): 72.0 33.0 TOTAL SYSTEM CAPACITY FACTOR (%): TOTAL SYSTEM T&D LOSSES (X): 205000 NUMBEROF CONSUMERS: 3043 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 87 12.0 AVERAGEREVENUE (US CENTS/KWH) N ASSETS REVALUED (POLICY)(YOYES,N=NO): O.S RETURN(%): OF RATE CONTRIBUTION TO INVESTMENT: 19.0 7. SALES GWH 302 142 855 38 1137 X 28.5600 12.4S00 S7.6100 3.3400 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 17.6

INSTALLED CAPACITY & GROSS GENERATION MW 240 45 285 130 GWH 1O90 75 11E5 204 -16O 1209

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERMAL TOTAL SECTOR SUPPLY

415

415

1209

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL MTOE 0.315 % 100.0000

DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS

0.315 RATES OF EXCHANGE

100.0000

AVE 78 USSI=228 CFA FRANCS CFA F AVE 1980 USS1211.3

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW, ,70LF) OTHER

11.1
5.7

JAMAICA

JAMAICA PUBLIC SERVICE CO LTD IS RESPONSIBLE FOR PUBLIC ELECTRIC SUPPLY PROVIDES 60% OF ELECTRIC CONSUMPTION 1973-78 AVG ANNUAL GTH RATE (%) 1.70 -0.20 -2.00 -0.10 -0.50 5. COMMERCIAL ENERGYCONSUMPTION MTOE % OIL 1.611 SB 1100 GAS COAL : 0.001 0.0600 LIGNITE HYDRO ; 0.030 1.8300 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 35 6. UTILITY DATA FREQUENCY(HZ): 50 TRANSMISSION (KV): 138/69/33 DISTRIBUTION (KV): LOW VOLTAGE (VOLTS): 380/220 MAIN SYSTEM LOAD FACTOR (X): 65.9 TOTAL SYSTEM CAPACITY FACTOR (%): 32.8 TOTAL SYSTEM T&D LOSSES (%): 15.9 NUMBEROF CONSUMERS: 204700 NUMBEROF EMPLOYEES: 1780 CONSUMERS/EMPLOYEE: 115 AVERAGEREVENUE (US CENTS/KWH) 9.4 ASSETS REVALUED (POLICY)(Y-YES,N=NO): Y RATE OF RETURN(%): S.8 CONTRIBUTION TO INVESTMENT: 0.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL B. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW, .7OLF) OTHER GWH 350 479 159 151 1139 X 30.7300 42.0500 13.9600 13.2600 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 9.5 8.0 5.6 8.5 3.6 2.8 1.642 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) : GNP/CAPITA (USS) ENERGY CONSUMED(MTOE) : ELECTRICTY PRODUCE(GWH) : ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (X) : ACCESS TO ELECTRICITY (X) : 2. 2131 2350 1110 2.642 2208 0.552 20.9 1239 1036 URBAN 40 URBAN 60

RURAL 60 RURAL IB

INSTALLED CAPACITY & GROSS GENERATION MW GWH 1120 108 1228 15 115 11 1354 854 2208

'C
tJ

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTOR SUPPLY

is 234 249

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE AVE 78 USS1=1.38 JAMAICAS AVE 1980 USSI=1.78 JAS X

JORDAN

JORDAN ELECTRICITY AUTHORITY BULK SUPPLY 30% SALES TO JEPCO JEPCO-MAJORDISTRIBUTOR OF ELEC 1973-78 AVG ANMJAL GYH RATE Ct 3.30 9.80 6.SO 12.90 15.60 S. COSERCIAL ENERGYCONSUMTION TOE 1.082 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) : GNP (USSM) : GNP/CAPITA (USS) : ENERGYCONSUMED(MTOE) : ELECTRICTY PRODUCE(GWH) : ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY Ct): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (X) : ACCESS TO ELECTRICITY (X) : 2. 2985 2270 1050 1.082 703 0.176 16.2 383 235 URBAN 42 URBAN 71

OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL

RURAL 58 RURAL 23

1.082

TOTAL

100.0000

43 8. UTILITY DATA FREQUENCY (HZ): 50 TRANSMISSION (KV): 132/66 DISTRIBUTION (KV): 33/6.6 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (%): 53.0 TOTAL SYSTEM CAPACITY FACTOR (X): 26.1 TOTAL SYSTEM T&D LOSSES (%): 13.0 NUMBEROF CONSUMERS: 179933 NUMBER OF EM4PLOYEES: 1127 CONSUMERS/EMPLOYEE: AVERAGEREVENUE (US CENTS/KWH) 4.2 ASSETS REVALUED (POLICY)(Y-YES,N-NO): RATE OF RETURN(%): 4.8 CONTRIBUTION TO INVESTMENT: 17.0 7. SALES DOMESTIC COMMERCIAL INDUJSTRIAL OTHER TOTAL S. TARIFFS GWH 224 74 171 146 615 X 36.4200 12.0300 27.8000 23.7400 99.9900 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 12.2 12.2 12.2 14.8 12.8 8.9

INSTALLED CAPACITY & GROSS GENERATION MW 99 72 8B 257 GWH

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTOR SUPPLY

257 25 282 115 115

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE AVE 78 USSI=.307 JOR DINARS AVE 1980 USSI=.298 U 0 X

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL C8KW, .35LF) INDUSTRIAL (200KW, .70LF) OTHER

KENYA

EAST AFRICAN POWERAND LIGHTING CO LTD THE SOLE DISTRIBUTOR OF POWER INTERCONNECTEDSYSTEM 1973-78 AVG ANNUAL GTH RATE (%) 3.80 4.70 0.90 3.70 7.10 S. COMMERCIALENERGY CONSUMPTION MTOE % OIL : 1.278 80.7600 GAS COAL : 0.046 2.9100 LIGNITE HYDRO 0.258 16.3300 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 6 B. UTILITY DATA FREQUENCY(HZ): 50 TRANSMISSION (KV): 132/33 DISTRIBUTION (KV): LOW VOLTAGE (VOLTS): 240/415 MAIN SYSTEM LOAD FACTOR (X): 68.0 TOTAL SYSTEM CAPACITY FACTOR (X): 37.0 TOTAL SYSTEM TAD LOSSES (X): 10.0 NUMBEROF CONSUMERS: 141727 NUMBEROF EMPLOYEES: 3525 CONSUMERS/EMPLOYEE: 40 AVERAGEREVENUE (US CENTS/KWH) B.0 ASSETS REVALUED (POLICY)(Y=YES,N=NO): Y RATE OF RETURN(X): 8.8 CONTRIBUTION TO INVESTMENT: 38.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS 0.086 RATES OF EXCHANGE 100.0000 DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW,.35LF) INDUSTRIAL (200KW,.70LF) OTHER GWH 194 321 758 28 1301 X 14.9100 24.6700 58.2600 2.1500 99.9900 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 17.6 14.8 4.5 11.8 7.9 4.4 1.580 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) : GNP (US$M) GNP/CAPITA (US$) : ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) : ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (%) : ACCESS TO ELECTRICITY (X) : 2. 15187 4830 320 1.579 1599 0.400 25.3 104 105 URBAN 13 URBAN

RURAL 87 RURAL

INSTALLED CAPACITY A GROSS GENERATION MW 98 30 32 160 306 GWH 252 2 5 259 1053 217 1529 20 50 1599

THERMAL-STEAM -GAS TURBINES L-COMBNED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (MET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTORSUPPLY

466 40 508

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL MTOE 0.088 0.000 X 100.0000

AVE 78 USSI=7.7 KENYA SHIL AVE 1980 USS1s7.4 K SH

KOREA

KOREA ELECTRIC COMPANYIS THE SOLE DISTRIBUTOR OF ELECTRICITY-THE COUNTRYIS ELECTRICALLY INTERCONNECTED 1973-78 AVG ANNUAL CTH RATE (X) 2.00 9.70 7.60 4.40 15.40 S. COMMERCIAL ENERGY CONSUMPTION MTOE X OIL 20.89t 61.1500 GAS COAL , 12.952 37.9OO LIGNITE HYDRO : 0.320 0.9400 NUCLEAR : 0.581 1.7000 GEOTHERMAL OTHER TOTAL TOTAL 95 6. UTILITY DATA FREQUENCY (HZ): 60 TRANSMISSION (KV): 345/154 DISTRIBUTION (KV): 66/23 LOW VOLTAGE (VOLTS): 100/200 MAIN SYSTEM LOAD FACTOR (X): 70.3 TOTAL SYSTEM CAPACITY FACTOR (X): 52.0 TOTAL SYSTEM TED LOSSES (V.): 9.3 NUMBEROF CONSUMERS: 4935000 NUMBEROF EMPLOYEES: 16576 CONSUMERS/EMPLOYEE: 295 AVERAGEREVENUE (US CENTS/KWH) 4.6 ASSETS REVALUED (POLICY)(Y-YES,N=NO): Y RATE OF RETURN(%): 6.7 CONTRIBUTION TO INVESTMENT: 13.0 7. SALES DOMESTIC COM1ERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW, .70LF) OTHER GWH 3959 5061 18116 190 27326 % 14.4900 18.5200 66.3000 0.7000 100.0100 KWH/MONTH US CENTS/KWH S.8 7.0 3.1 1.8 34.163 101.7000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) : GNP/CAPITA (USS) ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KOQE) : ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (X) : ACCESS TO ELECTRICITY (X) 2. 36616 42460 1160 34.744 33160 8.290 23.9 949 906 URBAN 49 URBAN 99

RURAL 91 RURAL 90

INSTALLED CAPACITY & GROSS GENERATION GWH 23393 1981 25374 1281 2324 2531 31510 1650 33160

MW THERMAL-STEAM 4692 -GAS TURBINES -COMBINED CYCLE -DIESEL 925 -SUBTOTAL 5617 HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY 712 587 6918

AUTO PRODUCERS HYDRO THERMAL 548 TOTAL SECTOR SUPPLY 7464 3.

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE X

OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE

AVE 78 USS 1485 WON AVE 1980 USS1.607.B W

LIBERIA

LIBERIA ELECTRICITY CORP OPERATES ONE MAIN SYSTEM (95% OF INSTALLED CAPACITY) AND FEW SMALL SYSTEMS 1973-78 AVG ANNUAL GTH RATE (%) 3.40 4.50 1.10 2.00 1.30 CONSUMPTION S. COMMERCIAL ENERGY MTOE X 0.470 86.2400 OIL GAS COAL LIGNITE 13.7600 HYDRO 0.075 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 25 6. UTILITY DATA 60 FREQUENCY(HZ): TRANSMISSION (KV): 69 DISTRIBUTION (KV): LOW VOLTAGE (VOLTS): 120/240 MAIN SYSTEM LOAD FACTOR (%): 70.0 31.0 TOTAL SYSTEM CAPACITY FACTOR (X): TOTAL SYSTEM TAD LOSSES (%): 28.0 15200 NUMBEROF CONSUMERS: NUMBEROF EMPLOYEES: 800 CONSUMERS/EMPLOYEE: 19 5.8 AVERAGEREVENUE (US CENTS/KWH) N ASSETS REVALUED (POLICY)(Y=YES,N=NO): 1.9 RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 0.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW,.35LF) INDUSTRIAL (200KW,.70LF) OTHER GWH 81 103 92 20 276 % 22.1000 37.3200 33.3300 7.2500 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 14.0 9.0 9.0 9.0 8.0 6.5 0.545 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) : GNP (USSM) GNP/CAPITA (USS) : ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) : (%) : POPULATION BREAKDOWN ACCESS TO ELECTRICITY (%) 1742 820 460 0.545 1021 0.255 46.8 313 Sa URBAN 28 URBAN 85

RURAL 72 RURAL 2

2. INSTALLED CAPACITY & GROSS GENERATION MW t'i ON THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERMAL TOTAL SECTOR SUPPLY 65 17 82 64 GWH 89 22 t11 287

146 143 289

398 623 1021

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE AVE 78 USS1=t LIBERIAN $ AVE 1980 USStI1 LIDS %

MADAGASCAR

MALAGASY ELECTRICITY AND WATER CORP OPERATES ALL ELECTRICITY AND WATER SUPPLIES SYSTEM ONE INTERCONNECTED 1973-78 AVG ANNUAL GTH RATE (X) 2.50 -0.20 -2.70 5.50 3.50 S. COMMERCIAL ENERGYCONSUMPTION % MTOE 87.4700 0.433 :
t

1. SOCID-ECONOMIC DATA POPULATION (000) GNP (US$M) : GNP/CAPITA (USS) : ENERGY CONSUMED(MTOE) : ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) (X) POPULATION BREAKDOWN : ACCESS TO ELECTRICITY (%) 8298 2050 250 0.495 365 0.092 18.4 50 92 URBAN 15 URBAN

OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL

0.018 0.044

3.6400 8.8900

RURAL 85 RURAL

0.495

100.0000

TOTAL

ig 50 FREQUENCY(HZ): TRANSMISSION (KV): 63 DISTRIBUTION (KV): LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (%): 47.6 TOTAL SYSTEM CAPACITY FACTOR (K): TOTAL SYSTEM T&D LOSSES (K): 91500 NUMBEROF CONSUMERS: 2500 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 35 7,6 AVERAGE REVENUE (US CENTS/KWH) ASSETS REVALUED (POLICY)(Y=YES,N=NO): RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUJSTRIAL OTHER TOTAL S. TARIFFS GWH 57 21 151 % 24.8900 9.1700 65.9400

2. INSTALLED CAPACITY & GROSS GENERATION MW THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERMAL TOTAL SECTORSUPPLY 3. GWH

NJ

25 25 40

90 90 191

e5 30 95

271 95 366

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE 0.027 K 100.0000

OIL GAS COAL LIGNITE OTHER TOTAL

229

100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 11.4 11.5 5.0 5.0

0.027

100.0000

4. OFFICIAL RATES OF EXCHANGE AVE 78 USS1-225 MALAGASYFRANC AVE 1980 US$1=211.3 M F

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW,.35LF) INDUSTRIAL (200KW, .7OLF) OTHER

MALAWI

ELECTRICITY COMMISSION OF MALAWI -GOVERNMENTOWNEDUTILITY RESPONSIBLE FOR PUBLIC ELECTRICITY SUPPLIES 1973-78 AVG ANNUAL GTH RATE (X) 3.10 6.20 3.10 2.60 9.00 5. COMMERCIALENERGYCONSUMPTION MTOE X OIL 0.120 49.5900 GAS COAL : 0.048 19.8300 LIGNITE HYDRO : 0.074 30.5800 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 6. UTILITY DATA FREQUENCY(HZ): 50 TRANSMISSION (KV): 132/66/33 OISTRIBUTION (KV): LOW VOLTAGE (VOLTS): 230/400 MAIN SYSTEM LOAD FACTOR (X): 84.5 34.0 TOTAL SYSTEM CAPACITY FACTOR (X): 7.9 TOTAL SYSTEM T&D LOSSES (10: NUMBEROF CONSUMERS: 19670 NUMBEROF EMPLOYEES: 1616 CONSUMERS/EMPLOYEE: 12 AVERAGEREVENUE (US CENTS/KWH) 4.2 ASSETS REVALUED (POLICY)(Y-YES,N=NO): Y RATE OF RETURN(): 8.0 CONTRIBUTION TO INVESTMENT: 0.0 7. SALES WMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL B. TARIFFS 0.002 RATES OF EXCHANGE 100.0000 DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .3SLF) INDUSTRIAL (200KW,.70LF) OTHER GWH 47 46 192 2 287 X 1E.3800 16.0300 o6.9000 0.7000 100.0100 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 14.8 9.0 5.8 7.0 4.9 3.5 0.242 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (US$M) : GNP/CAPITA (USS) : ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE (GWH) : ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) : (X) : POPULATION BREAKDOWN ACCESS TO ELECTRICITY (%) : 2. 5780 1010 180 0.242 344 0.086 35.5 42 59 URBAN URBAN

RURAL 95 RURAL

INSTALLED CAPACITY & GROSS GENERATION MW GWH 2 5 7 304 1 312 14 19 345

00

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYORO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYORO THERMAL TOTAL SECTORSUPPLY

15 8 23 65

88 2 14 104

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL MTGE 0.002 7. 100.0000

AVE 78 USS1=.84 KWACHA AVE 1980 US$1=.81 KWACHA

MALAYSZA

NATIONAL ELECTRICITY BOARDOF THE STATES OF MALAYA SUPPLIES 95% OF TOTAL ELECTRICITY 1973-78 AVG ANNUAL GTH RATE (%) 2.70 7.70 4.90 4.70 10.60 S. COMMERCIAL ENERGY CONSUMPTION % MTOE 88.5300 5.688 OIL 9.8900 0.648 GAS 0.4400 0.029 COAL LIGNITE 3.1300 0.205 HYDRaO NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 52 6. UTILITY oATA FREOUENCY(HZ): 50 6.550 99.9900

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) GNP/CAPITA (USS) ENERGY CONSUMED (MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL E.UIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) ELEcrRICITY/CAPITA (KWH) ' (X) POPULATION BREAKDOWN ACCESS TO ELECTRICITY (%) 2. 13300 14540 1090 6.550 7073 1.770 27.0 487 932 URBAN 30 URBAN

RURAL 70 RURAL

INSTALLED CAPACITY & GROSS GENERATION MW 970 20 98 1088 352 GWH 5525 225 5750 820 178 6748 11 314 7073

'C

THERMAL-STEAM -GAS TURBINES -COMINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY

1440

TRANSMISSION (KV): 275/132/86 DISTRIBUTION (KV): 33/22/11 LOW VOLTAGE (VOLTS): 380/220 70.8 MAIN SYSTEM LOAD FACTOR (%): 53.0 TOTAL SYSTEM CAPACITY FACTOR (%): 10.0 TOTAL SYSTEM T8D LOSSES (X): 977000 NUMBEROF CONSUMERS: 15590 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 83 5.5 AVERAGEREVENUE (US CENTS/KWH) Y ASSETS REVALUED (POLICY)(Y=YES,N=NO): 8.2 RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 35.0 7. SALES DOMESTIC COMMERCIAL INDUJSTRIAL OTHER TOTAL 8. TARIFFS GWH 987 2038 2448 1068 6541 % 15.0900 31.1600 37.4300 16.3300 100.0100 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 7.3 9.7 7.9 8.0

116 AUTO PRODUCERS HYORO THERMAL 155s TOTAL SECTOR SUPPLY

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL MTOE 1.700 % 100.0000

1.700

100.0000

4. OFFICIAL RATES OF EXCHANGE AVE 78 USS1:2.31 RINGGIT AVE 1980 USS1=2.2 RINGGIT

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW, .70LF) OTHER

MALI

ENERGIE OU MALI-A PUBLIC SECTOR UNDERTAKING RESPONSIBLE FOR ELEC & WATER SUPPLY-NOT INTERCONNECTED 1973-78 AVG ANNUAL GTH RATE (X) 2.50 4.40 1.90 7.10 4.60 S. COMMERCIAL ENERGYCONSUMPTION MTOE % OIL 0.124 91.1800 GAS COAL LIGNITE HYDRO : 0.012 9.8200 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 5 6. UTILITY DATA FREQUENCY(HZ): 50 TRANSMISSION (KV): 30 DISTRIBUTION (KV): 15/30 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (X): 82.0 TOTAL SYSTEM CAPACITY FACTOR (X): 45.0 TOTAL SYSTEM T8D LOSSES (X): NUMBEROF CONSUMERS: 27000 NUMBEROF EMPLOYEES: 970 CONSUMERS/EMPLOYEE: 31 AVERAGEREVENUE (US CENTS/KWH) 13.0 ASSETS REVALUED (POLICY)(Y=YES,N=NO): N RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL B. TARIFFS 0.022 100.0000 DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW, .70LF) OTHER GWH 34 30 19 92 X 41.4800 38.5900 21.9500 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 18.3 18.3 18.3 18.3 18.3 18.3 0.138 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (US$M) : GNP/CAPITA (USS) ENERGY CONSUMED(MTOE) : ELECTRICTY PRODUCE(GWH) : ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) : POPULATION BREAKDOWN (X) : ACCESS TO ELECTRICITY (%) : 8297 790 120 0.138 135 0.034 24.8 22 21 URBAN 17 URBAN 30

RURAL B3 RURAL 0

2. INSTALLED CAPACITY & GROSS GENERATION MW THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTORSUPPLY GWH

20 20 10 30 10 40

66 88 38 102 38 138

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL MTOE 0.022 X 100.0000

4. OFFICIAL RATES OF EXCHANGE AVE 79 USS1=450 MALIAN FR AVE 1980 USS1=422.B M FR

MAURITIUS

CENTRAL ELECTRICITY BOARDSOLE GENERATORAND DISTRIBUTOR OF PUBLIC ELECTRICITY 1973-78 AVG ANN)AL GTH RATE ()MTOE 1.30 8.30 6.90 -1.30 9.10 S. COMMERCIALENERGYCONSUMPTION OIL GAS COAL LIGNITE HYORO NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 85 S. UTILITY DATA 50 FREQUENCY(HZ): TRANSMISSION (KV): 66/22 DISTRIBUTION (KV): 22/6.6 LOW VOLTAGE (VOLTS): 220/380 56.0 MAIN SYSTEM LOAD FACTOR (X): 33.0 TOTAL SYSTEM CAPACITY FACTOR (%): TOTAL SYSTEM T&D LOSSES (%): 158900 NUMBEROF CONSUMERS: 1710 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 93 5.7 AVERAGEREVENUE (US CENTS/KWH) Y ASSETS REVALUED (POLICY)(Y-YES,N=NO): 1.7 RETURN(%): OF RATE 2.0 CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS 0.061 RATES OF EXCHANGE 100.0000 DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .3SLF) INDUSTRIAL (200KW,.70LF) OTHER GWH 127 62 77 15 281 % 45.2000 22.0600 27.4000 S.3400 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 8.4 8.4 8.4 8.4 8.4 8.4 0.242 0.001 0.015 93.8000 0.3900 5.8Ino

t.

SOCIO-ECONOMIC DATA 918 760 830 0.258 28 0.132 S1.0 281 573 URBAN 48 URBAN 90

POPULATION (000) GNP (US$M) : GNP/CAPITA (USS) : ENERGY CONSUMED(MTOE) : ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): : ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) (%) POPULATION BREAKDOWN ACCESS TO ELECTRICITY (%) 2.

RURAL 82 RURAL 80

0.258

100.0000

INSTALLED CAPACITY & GROSS GENERATION MW GWH

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUSTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERMAL TOTAL SECTOR SUPPLY

III 11 25

288 288 65 38 391 135 526

136 15 151

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL MTOE 0.061 100.0000

AVE 78 USS1s6.1 RUPPES AVE 1980 US$1=7.7 RUPEES

MEXICO

COMISION FEDERAL DE ELECTRICIDAD-LARGELY INTERCONNECTEDSYSTEM OF 8 REGIONS PLUS 3 ISOLATED REGIONS(TIOUANA-YUCATAN-LAPAZ) 1973-78 AVG ANNUAL GTH RATE (% 3.30 4.40 1.20 6.90 7.60 5. COMMERCIAL ENERGYCONSUMPTION MTOE 42.996 64.5500 OIL 12.873 19.3300 GAS 8.0200 COAL : 5.339 LIGNITE HYDRO NUCLEAR 8.1000 GEOTHERMAL 5.398 OTHER TOTAL TOTAL 80 S. UTILITY DATA FREQUENCY (HZ): 60 TRANSMISSION (KV): 400/230/115 DISTRIBUTION (KV): 115/13.2 LOW VOLTAGE (VOLTS): 220/127 MAIN SYSTEM LOAD FACTOR (X): 42.3 TOTAL SYSTEM CAPACITY FACTOR (%): TOTAL SYSTEM T&D LOSSES (%): NUMBEROF CONSUMERS: 8579000 NUMBEROF EMPLOYEES: 100300 CONSUMERS/EMPLOYEE: 86 2.5 AVERAGEREVENUE (US CENTS/KWH) Y ASSETS REVALUED (POLICY)(Y=YES,N=NO): RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUJSTRIAL (8KW. .35LF) INDUSTRIAL (200KW, .70LF) OTHER GWH 8269 33846 2935 45050 % 18.3800 75.1300 6.5100 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 3.3 9.0 8.8 6.8 5. 2.9 66.606 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) GNP/CAPITA (USS) ENERGYCONSUMED(MTOE) : ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) (%) POPULATION BREAKDOWN ACCESS TO ELECTRICITY (%) 2. 65470 84150 1290 66.605 56800 14.200 21.3 962 867 URBAN 83 URBAN 99

RURAL 37 RURAL 50

INSTALLED CAPACITY & GROSS GENERATION GWH 1259 233 1492 18100

o
k.)

MW THERMAL-STEAM 6708 1289 -GAS TURBINES -COMBINED CYCLE 720 233 -DIESEL -SUBTOTAL 8920 HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY 5219 150 14289

17592 155 3645 21392

AUTO PRODUCERS HYDRO 71 THERMAL 1679 16039 TOTAL SECTOR SUPPLY

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE AVE 78 USS1=22.7 PESOS AVE 1980 USS1=22.9 PESOS X

MOROCCO

OFFICE NATIONAL DE ELECTRICITE GENERATES 91% OF COUNTRY'S ELECTRIC POWER SYSTEM INTERCONNECTED 1973-78 AVG ANNUAL GTH RATE (%) 2.70 8.90 4.20 6.60 9.50 5. COMMERCIAL ENERGYCONSUMPTION % MTOE 77.7000 3.070 OIL 1.7200 0.068 GAS 11.9200 0.471 COAL LIGNITE 8.6600 0.342 HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 28 6. UTILITY GWH 2289 745 3031 1504 DATA 50 FREQUENCY(HZ): TRANSMISSION (KV): 225/150/60 DISTRIBUTION (KV): 30/22/8 LOW VOLTAGE (VOLTS): 220/380 59.8 MAIN SYSTEM LOAD FACTOR (%): 50.8 TOTAL SYSTEM CAPACITY FACTOR (X): 12.3 TOTAL SYSTEM T&D LOSSES (X): NUMBEROF CONSUMERS: 5173 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 4.7 AVERAGEREVENUE (US CENTS/KWH) N ASSETS REVALUED (POLICY)(Y=YES,N=NO): RETURN(%): RATE OF CONTRIBUTION TO INVESTMENT: 18.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW,.35LF) INDUSTRIAL (200KW, .70LF) OTHER GWH 1969 847 580 3396 X 57.9800 24.9400 17.0800 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 18.7 18.7 18.7 18.7 14.0 14.0 3.951 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (US$M) GNP/CAPITA (USS) ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): : ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) (%) POPULATION BREAKDOWN : ACCESS TO ELECTRICITY (%) 2. 18914 12610 E70 3.951 4785 1.19B 30.3 209 253 URBAN 40 URBAN eo

RURAL 60 RURAL 6

INSTALLED CAPACITY & GROSS GENERATION MW 409 160 569 449

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTMERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY

1018

4535 15 235 4785

16 HYDRO AUTO PRODUCERS 132 THERMAL 1166 TOTAL SECTOR'SUPPLY 3.

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE 0.039 0.470 0.509 % 7.6600 92.3400 100.0000

OIL GAS COAL LIGNITE OTHER TOTAL 4.

OFFICIAL RATES OF EXCHANGE

AVE 78 USS1=4.17 DIRHAMS AVE 1980 USS1-3.94 OIR

NEPAL

NEPAL ELECTRICITY CORP RESPONSIBLE FOR GENERATING AND DISTRIBUTING ELECTRICITY IN CENTRAL AREA 1973-78 AVG ANNUAL GTH RATE (%) 2.20 4.70 2.40 -0.50 15.70 5. COMMERCIAL ENERGYCONSUMPTION MTOE X 0.079 63.7100 OIL GAS 0.006 4.8400 COAL LIGNITE 31.4500 HYDRO : 0.039 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 5 B. UTILITY DATA GWH 50 FREQUENCY (HZ): TRANSMISSION (KV): 132/6 DISTRIBUTION (KV): 33/11/3.3 LOW VOLTAGE (VOLTS): 400/230 45.0 MAIN SYSTEM LOAD FACTOR (%): 38.0 TOTAL SYSTEM CAPACITY FACTOR (%): TOTAL SYSTEM T&D LOSSES (%): 31.0 NUMBEROF CONSUMERS: 77464 2103 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 37 AVERAGE REVENUE (US CENTS/KWH) 3.2 ASSETS REVALUED (POLICY)(Y=YES.N=NO): N RATE OF RETURN(%): 0.5 CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW,. 35LF) INDUSTRIAL (200KW, .70LF) OTHER GWH 60 13 25 7 lOB % 57.1400 12.3800 23.8100 6.6700 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 3.0 3.3 4.6 3.0 3.0 0.124 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) : GNP/CAPITA (USS) ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (%) ACCESS TO ELECTRICITY (%) 2. 13627 1580 120 0.124 201 0.051 40.8 9 Is URBAN URBAN

RURAL 95 RURAL

INSTALLED CAPACITY & GROSS GENERATZON MW

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTOR SUPPLY

15 IS 38

6 a 153 27 ls 1s 201

53 11 54

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES Of EXCHANGE AVE 78 USS1=12 RUPEES AVE 1980 US51=12 RUPEES %

NICARAGUA

EMPRESANACIONAL DE LUZ Y FUERZA IS RESPONSIBLE FOR ALMOST ALLPUBLIC GENERATION AND TRANSMISSION 85% OF TOTAL 1973-78 AVG ANNUAL GTH RATE (VX) 3.30 5.ao 2.50 -6.80 12.30 5. COMMERCIAL ENERGYCONSUMPTION X MTOE 89.7000 0.810 OIL GAS COAL LIGNITE 10.3000 0.093 HYORO NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 41 8. UTILITY DATA 0.903 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (US$M) GNP/CAPITA (USS) (MTOE) ENERGY CONSUMEO ELECTRICTY PRODUCE (GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): : ENERGY/CAPITA (KGOE) ELECTRXCITY/CAPITA (KWH) : (X) POPULATION BREAKDOWN ACCESS TO ELECTRICITY (X) 2490 2100 840 0.903 1210 0.303 32.5 362 486 URBAN 48 URBAN 64

RURAL 52 RURAL 20

2. INSTALLED CAPACITY 8 GROSS GENERATION THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY HYDRO AUTO PROWUCERS THERMAL TOTAL SECTORSUPPLY 3. MW 175 Is 7 197 94 350 -10 340 25 100 465 GWH FREQUENCY(HZ): 60 138/69 TRANSMISSION (KV): DISTRIBUTION (KV): LOW VOLTAGE (VOLTS): 110/220 57.8 MAIN SYSTEM LOAD FACTOR (X): 42.6 TOTAL SYSTEM CAPACITY FACTOR (X): 10.2 TOTAL SYSTEM T&O LOSSES (X): 170500 NUM8ER OF CONSUMERS: 1753 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 97 5.9 AVERAGEREVENUE (US CENTS/KWH) N ASSETS REVALUED (POLICY)(Y=YES,N=NO): 8.8 RATE OF RETURN(X): CONTRIBUTION TO INVESTMENT: 29.3 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS 0.239 RATES OF EXCHANGE 100.0000 DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW,.35LF) INDUSTRIAL (200KW,.70LF) OTHER GWH 222 79 443 228 972 X 22.8400 8.1300 45.5800 23.4600 100.0100 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 9.0 10.7 10.8 B.1 5.9

291 8 45 344

FOSSIL FUEL FOR ELECTRICITY PiROUCTION MTOE 0.239 % 100.0000

OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL

AVE 78 US$1=7.02 CORDOBAS AVE 1980 US51=10.O CORD

NIGERIA

NATIONAL ELECTRIC POWERAUTHORITY PROVIDES 95% OF PUBLIC SUPPLY-INTERCONNECTED SYSTEM

1. SOCIO-ECONOMIC DATA POPULATION (000) : GNP (USSM) GNP/CAPITA (USs) ENERGY CONSUMED(MTOE) : ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (x): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) : POPULATION BREAKDOWN (x) ACCESS TO ELECTRICITY (x) 81039 45720 560 8.197 6799 1.700 21.7 101 84 URBAN 28 URBAN

1973-78 AVG ANNUAL GTH RATE ()MTOE 2.80 4.40 1.80 8.50 5.90

s.

COMMERCIAL ENERGYCONSUMPTION

OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL

6.215 1.247 0.100 0.e05

76.1000 15.2700 1.2200 7.4100

RURAL 74 RURAL

8.167

100.0000

TOTAL 20 6. UTILITY DATA FREQUENCY (HZ): 50 TRANSMISSION (KV): 330/132 DISTRIBUTION (KV): 33/11 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (%): 81.0 TOTAL SYSTEM CAPACITY FACTOR (): 43.0 TOTAL SYSTEM TLD LOSSES (x): 20.0 NUMBER OF CONSUMERS: 829000 NUMBEROF EMPLOYEES: 16380 CONSUMERS/EMPLOYEE: 49 AVERAGEREVENUE (US CENTS/KWH) 8.1 ASSETS REVALUED (POLICY)(Y=YES,N=NO): Y RATE OF RETURN(): -0.5 CONTRIBUTION TO INVESTMENT: 10.3 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS GWH 2290 781 1825 4596 % 49.8300 18.9900 33.1800

2. INSTALLEo CAPACITY A GROSS GENERATION THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY AUTO PRODWCERS HYDRO THERMAL TOTAL SECTOR SUPPLY 3. MW 347 618 18 983 760 GWH 1401 1764 97 3282 3237

u,,

1743 100 1843

6499 300 6766

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE 0.198 0.960 0.028 1.186 RATES OF EXCHANGE % 16.6900 80.9400 2.3600 99.9900

OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL

100.0000 KWH/MONTH US CENTS/KWH 25 too 400 400 2000 100000 18.2 14.5 12.3 14.4 13.1

AVE 7aUsst=.e37 NAIRAS AVE 1980 US$ =.560 NAI

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW.. 35LF) INDUSTRIAL (200KW, .70LF) OTHER

OMAN

DIRECTORATEGENERAL OF ELECTRICITY OPERATES TWOMAIN SYSTEMS-CAPITAL AREA(76% SALES) SALALA AREA. DGE ALSO DISTRIBUTES WATER 1973-78 AVG ANNUAL GTHf RATE %)MTOE 3.20 7.20 4.00 15.90 26.00 5. COMMERCIAL ENERGYCONSUMPTION OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 20 6. UTILITY DATA GWH 400 98 31 144 498 FREQUENCY(HZ): 50 TRANSMISSION (KV): 132/33 DISTRIBUTION (KV): LOW VOLTAGE (VOLTS): 240/415 56.0 MAIN SYSTEM LOAD FACTOR (X): 40.3 TOTAL SYSTEM CAPACITY FACTOR (%): TOTAL SYSTEM T&D LOSSES (%): 18.0 NUMBER OF CONSUMERS: NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE; AVERAGEREVENUE (US CENTS/KWH) 5.8 ASSETS REVALUED (POLICY)(Y=YES,N=NO): N RATE OF RETURN(X): CONTRIBUrION TO INVESTMENT: 7. SALES DOMESTIC CDMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUJSTRIAL (8KW,..35LF) 37LF) INDUSTRIAL (208KW, OTHER KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 5.8 5.8 5.8 5.8 5.8 5.8 GWH % 1.150 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) : GNP/CAPITA (USS) ENERGYCONSUMED(MTOE) : ELECTRICTY PRODUCE (GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): : ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) (%) POPULATION BREAKDOWN : ACCESS TO ELECTRICITY (%) 2. 839 2160 2570 1.150 644 0.161 14.0 1370 767 URBAN 40 URBAN 50

RURAL 80 RURAL 0

1.150

100.0000

INSTALLED CAPACITY & GROSS GENERATION MW 63 50

'4

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRD THERMAL TOTAL SECTOR SUPPLY

144 38 182

498 146 644

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE X

AVE 78USS1=.345 RIALS OMAN AVE 1980 USSI=.345 R 0

PAKISTAN

WATER & POWERDEV AUTH GENERATES80% OF PUBLIC ELECTRICITY. KARACHI ELECTRIC CORP ACCOUNTSFOR 20% OF GENERATION 1973-78 AVG ANNUAL GTH RATE ()MTOE 3.10 3.90 0.80 3.30 4.30 5. COMMERCIAL ENERGYCONSUMPTION OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 22 6. UTILITY DATA FREQUENCY(HZ): 50 TRANSMISSION (KV): 500/220/132/66 DISTRIBUTION (KV): 33/11 LOW VOLTAGE (VOLTS): 240/415 61.0 MAIN SYSTEM LOAD FACTOR (%): 48.3 TOTAL SYSTEM CAPACITY FACTOR (%): 31.0 TOTAL SYSTEM T&D LOSSES (%): NUMBEROF CONSUMERS: 2530000 67000 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 38 3.5 AVERAGEREVENUE (US CENTS/KWH) N ASSETS REVALUED (POLICY)(Y=YES,N=NO): RATE OF RETURN(%): 11.8 CONTRIBUTION TO INVESTMENT: 35.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL B. TARIFFS 1.322 100.0000 DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW. .3SLF) INDUSTRIAL (200KW, .70LF) OTHER GWH 2091 676 3572 2637 8976 % 23.3000 7.5300 39.8000 29.3800 100.0100 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 3.1 3.4 3.5 9.3 8.2 4.8 : : 5.271 8.430 0.828 2.058 0.027 36.1300 44.0800 5.6800 14.1100 0.1900

1. SOCIO-ECONOMIC DATA POPULATION (000) : GNP (USSM) GNP/CAPITA (USS) ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) (%) POPULATION BREAKDOWN ACCESS TO ELECTRICITY (%) 77337 17530 230 14.614 14874 3.719 25.4 189 192 URBAN 26 URBAN

RURAL 74 RURAL

14.587

100.1900

2. INSTALLED CAPACITY & GROSS GENERATION MW 1177 THERMAL-STEAM 457 -GAS TURBINES -COMBINED CYCLE 15 -DIESEL 1649 -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY 1567 137 3353 GWH

Goi

8232 106 8338 320 380 9038

AUTO PRODUCERS HYDRO 95 THERMAL 230 3678 TOTAL SECTORSUPPLY 3.

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE 0.018 1.293 0.013 % 1.2100 97.8100 0.9800

OIL GAS COAL LIGNITE OTHER TOTAL

4. OFFICIAL RATES OF EXCHANGE AVE 78USS1=9.9 RUPEES AVE 1980 USS1=9.9 RUPEES

PANAMA

INSTITUTO DE RECURSOSHIDRAULICOS Y ELECTRIFICACION RESPONSIBLE FOR ALL PUBLIC POWERSUPPLY IN PANAMA-ALMOSTCOMPLETELY INTERCONNECTED 1973-78 AVG ANNUAL GTH RATE (%) 3.10 3.00 -0.10 4.90 4.00 5. COMMERCIALENERGY CONSUMPTION MTOE X OIL 1.205 94.1400 GAS COAL LIGNITE HYDRO 0.075 5.8600 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL SO S. UTILITY DATA FREQUENCY(HZ): 60 TRANSMISSION (Ky): 230 DISTRIBUTION (KV): 34.5/4.2 LOW VOLTAGE (VOLTS): 110/220 MAIN SYSTEM LOAD FACTOR (X): 65.0 TOTAL SYSTEM CAPACITY FACTOR (X): 38.4 TOTAL SYSTEM T&D LOSSES (%): 11.9 NUMBEROF CONSUMERS: 109700 NUMBEROF EMPLOYEES: 3101 CONSUMERS/EMPLOYEE: 67 AVERAGEREVENUE (US CENTS/KWH) 7.5 ASSETS REVALUED (POLICY)(Y=YES,N=NO): Y RATE OF RETURN(%): 7.9 CONTRIBUTION TO INVESTMENT: 8.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS 0.279 RATES OF EXCHANGE 100.0000 DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW. .35LF) INDUSTRIAL (200KW,.70LF) OTHER GWH 413 418 134 280 1245 % 33.1700 33.5700 10.7600 22.4900 99.9900 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 13.1 10.8 9.0 10.6 9.0 7.1 1.280 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) : GNP (USSM) GNP/CAPITA (USS) ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE (GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (X) ACCESS TO ELECTRICITY (X) 2. 1826 2350 1290 1.280 1657 0.414 32.4 689 906 URBAN 70 URBAN 77

RURAL 30 RURAL 36

INSTALLED CAPACITY & GROSS GENERATION MW 183 107 280 168 350 GWH

'C

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTOR SUPPLY

456 SO 51B

350 125 475

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL MTOE 0.279 % 100.0000

AVE 78 USSI=1 BALBOA AVE 1980 USS1=1 BALBOA

PERU

ELECTRICIDAD DEL PERU IS THE MAIN SUPPLIER OF ELECTRICITY-ELECTRICIDAD DE LIMA IS THE MAIN DISTRIBUTION CO(70% OF TOTAL) 1973-78 AVG ANNUAL GTH RATE (%) 2.80 4.50 1.80 3.90 5.00 S. COMMERCIAL ENERGYCONSUMPTION MTOE S OIL 6.296 74.1500 GAS 0.461 5.4300 COAL : 0.141 1.6600 LIGNITE HYDRO : 1.593 18.7600 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 35 B. UTILITY DATA FREQUENCY (HZ): 60 TRANSMISSION (KV): 220/60 DISTRIBUTION (KV): 60/30/10/2.3 LOW VOLTAGE (VOLTS): 110/220 MAIN SYSTEM LOAD FACTOR (): 596.0 TOTAL SYSTEM CAPACITY FACTOR (%): 45.6 TOTAL SYSTEM T&D LOSSES (5: 10.5 NUMBEROF CONSUMERS: 654000 NUMBEROF EMPLOYEES: 2867 CONSUMERS/EMPLOYEE: 228 AVERAGE REVENUE (US CENTS/KWH) 2.8 ASSETS REVALUED (POLICY)(Y=YES.N=NO): Y RATE OF RETURN(%): 8.6 CONTRIBUTION TO INVESTMENT: 18.9 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL B. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW,.35LF) INDUSTRIAL (200KW, .70LF) OTHER GWH 1042 538 1246 175 3001
%

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) GNP/CAPITA (USS) : ENERGYCONSUMED(MTOE) : ELECTRICTY PRODUCE(GWH) : ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) : POPULATION BREAKDOWN (%) : ACCESS TO ELECTRICITY (%) : 16820 12440 740 8.491 8B75 2.210 26.1 501 527 URBAN 55 URBAN

RURAL 45 RURAL

8.491

100.0000

2. INSTALLED CAPACITY a GROSS GENERATION MW THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY 1010 4925 GWH

1010

4926 1425 1900 8250

AUTO PRODUCERS HYDRO 405 THERMAL 785 TOTAL SECTORSUPPLY 2200

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE %

34.7600 17.9300 41.4900 5.8300 100.0100 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 1.2 2.1 2.2 7.2 2.8 3.5

AVE 78USS1=15 SOLES AVE 1980 US$1=288.6 SOLES

PHILIPPINES

NATIONAL POWERCORP IS THE MAIN PUBLIC 63% UTILITY. MANILA ELECTRIC CO ACCOUNTED 82 UTILITIES AND 73 COOPERATIVESSHARE REST 1973-78 AVG ANNUAL GTH RATE (%) 2.70 6.50 3.70 4.50 5.20 S. COMMERCIAL ENERGY CONSUMPTION % MTOE 86.9900 10.062 OIL 0.0000 GAS: 1.7000 0.197 COAL 0.0000 LIGNITE 11.3100 1.308 : HYDRO 0.0000 NUCLEAR 0.0000 GEOTHERMAL 0.0000 : OTHER TOTAL TOTAL iL 6. UTILITY DATA 60 FREQUENCY(HZ): TRANSMISSION (KV): 230/138/115/69 DISTRIBUTION (KV): 13.8 LOW VOLTAGE (VOLTS): 110/220 70.0 MAIN SYSTEM LOAD FACTOR (%): 48.0 TOTAL SYSTEM CAPACITY FACTOR (%): 7.0 TOTAL SYSTEM TBD LOSSES (%): NUMBEROF CONSUMERS: 2208000 30000 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 74 2.3 AVERAGE REVENUE (US CENTS/KWH) ASSETS REVALUED (POLICY)(Y-YES,N=NO): Y S.9 RATE OF RETURN(%): 5.0 CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 0.000 2.110 0.0000 100.0000 8. TARIFFS GWH 2122 2736 5756 621 11235 S 18.8900 24.3500 51.2300 S.5300 100.0000 KWH/MONTH US CENTS/KWH 25 1oo 400 400 2000 100000 4.5 4.5 4.5 7.3 7.0 7.0 2.7 11.567 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) GNP/CAPITA (USS) : ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): : ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) (%) POPULATION BREAKDOWN ACCFSS TO ELECTRICITY (X) 2. 45639 23250 510 11.567 14961 3.750 32.4 253 329 URBAN 30 URBAN

RURAL 70 RURAL

INSTALLED CAPACITY & GROSS GENERATION MW 1680 GWH 7842 3048 10887 2417 2 13306 1655 14961

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE 538 -DIESEL 2416 -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERMAL TOTAL SECTOR SUPPLY

746 3 3165 600 3765

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL MTOE 2.110 % 100.0000

4. OFFICIAL RATES OF EXCHANGE AVE 78US11-7.37 PESOS AVE 1980 USS157.51 PESOS

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW,. 35LF) INDUSTRIAL (200KW,.70LF) OTHER

PORTUGAL

ELECTRICIDADE DE PORTUGAL- PUBLIC UTILITY OPERATES AN INTERCONNECTEDSYSTEM

1. SOCIO-ECONOMIC DATA POPULATION (000) : GNP (US$M) GNP/CAPITA (US$) ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE (GWH) : ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) : POPULATION BREAKDOWN (X) : ACCESS TO ELECTRICITY (X) : 9653 19540 2020 8.431 14173 3.543 42.0 874 1468 URBAN 93 URBAN 100

1973-78 AVG ANNUAL GTH RATE (%) 0.80 3.90 3.1O 2.30 5.00

OIL GAS COAL LIGNITE HYORO NUCLEAR GEOTHERMAL OTHER TOTAL

S. COMMERCIALENERGY CONSUMPTION MTOE : 5.685 57.4300 0.397 : 2.349 4.7100 27.8600

RURAL 7 RURAL 93

8.431

100.0000

TOTAL

9S 8. UTILITY DATA FREQUENCY (HZ): 50 TRANSMISSION (KV): 400/220/150/60 DISTRIBUTION (KV): 40/30/15/10/6/5/4/3 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (): TOTAL SYSTEM CAPACITY FACTOR (X): 44.4 TOTAL SYSTEM T&D LOSSES (X): 12.6 NUMBEROF CONSUMERS: NUMBEROF EMPLOYEES: 13644 CONSUMERS/EMPLOYEE: AVERAGEREVENUE (US CENTS/KWH) 2.8 ASSETS REVALUED (POLICY)(Y=YES,N=NO): N RATE OF RETURN(%): 16,5 CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL B. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW,.70LF) OTHER KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 3.6 32.6 4.2 4.9 2.4 1.7 GWH %

2. INSTALLEO CAPACITY & GROSS GENERATION THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY MW 971 187 1138 2276 GWH 2966 57 3023 10750 200 13973 30 570 14573

3413

AUTO PRODUCERS HYDRO 215 THERMAL Is TOTAL SECTOR SUPPLY 3243

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE AVE 78USS1=43.9 ESCUDOS AVE 1980 USS1=S0.06 ESC X

ROMANIA

OF POWER SECTORUNDER THE MANAGEMENT MINISTANTI ENERGIEI ELECTRIC 97% OF TOTAL INSTALLED CAPACITY IS CONNECTED AVG ANNUAL 1973-78 GTH RATE (%) 0.90 1.80 0.90 3.60 5.70 S. COMMERCIAL ENERGYCONSUMPTION X MTOE 25.3400 15.503 OIL 49.7500 30.442 GAS 21.1800 : 12.960 COAL LIGNITE 3.7400 2.286 HYDRO NUCLEAR GEOTHERMAL OTHER TnTAL TOTAL 100 8. UTILITY DATA R1 191 100.0100

1, SOCIO-ECONOMIC DATA POPULATION (000) : GNP (US$M) GNP/CAPITA (USS) : ENERGY CONSUMED(NTOE) ELECTRICTY PRODUCE(GWH) (MTOE): EQUIV OIL ELECTRICITY ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) : (X) POPULATION BREAKDOWN ACCESS TO ELECTRICITY (X) 2. 21853 38170 1750 81.191 62822 15.410 25.8 2800 2875 URBAN 49 URBAN 100

RURAL 51 RURAL 89

INSTALLED CAPACITY & GROSS GENERATION GWH 100

MW 99 THERMAL-STEAM 9563 -GAS TURBINES 109 -COMBINED CYCLE -DIESEL 9771 -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY 3071

100 10521

12842

10621 3368 13989

50 FREQUENcY ( HZ): TRANSMISSION (KV): 400/220/110 DISTRIBUTION (KV): 35/20/15/10/5 LOW VOLTAGE (VOLTS): 220/380 75.8 MAIN SYSTEM LOAD FACTOR (X): 52.6 TOTAL SYSTEM CAPACITY FACTOR (X): 7.3 TOTAL SYSTEM T&D LOSSES (X): NUMBEROF CONSUMERS: 5810000 56500 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE:103 1.3 AVERAGEREVENUE (US CENTS/KWH) N ASSETS REVALUED (POLICY)(Y=YES,N=NO): 2.6 RATE OF RETURN(X): 47.0 CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL GWH 5349 466 37247 9459 52521 X 10.1800 0.8900 70.9200 18.0100 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000

HYDRO AUTO PRODUCERS THERMAL 787 13629 TOTAL SECTOR SUPPLY 3.

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE 10.600 1.400 2.800 0.300 15.100 RATES OF EXCHANGE X 70.2000 9.2700 18.5400 1.9900 100.0000

OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL

8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW, .70LF) OTHER

AVE 78 US$1=18 LEI AVE 1880 US$1=18 LEI

2.0 1.3 1.1

SRI LANKA

CEYLON ELECTRICITY BOARD RESPONSIBLE FOR ALL GENERATION AND DISTRIBUTION OF ELECTRICITY IN SRI LANKA -ALMOST TOTALLY INTERCONNECTEDSYS 1973-78 AVG ANNUAL GTH RATE (%) 1.70 3.00 1.30 -1.50 5.50 5. COMMERCIAL ENERGYCONSUMPTION MTOE % OIL 0.998 75.8900 GAS COAL : 0.002 0.1500 LIGNITE HYDRO 0.315 23.9500 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 8 B. UTILITY DATA FREQUENCY (HZ): 50 TRANSMISSION (KV): 132/68 DISTRIBUTION (KV): 33/11 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (%): 54.0 TOTAL SYSTEM CAPACITY FACTOR (%): 40.3 TOTAL SYSTEM T&D LOSSES (%): 16.1 NUMBEROF CONSUMERS: 144569 NUMBEROF EMPLOYEES: 10000 CONSUMERS/EMPLOYEE: 15 AVERAGEREVENUE (US CENTS/KWH) 1.1 ASSETS REVALUED (POLICY)(Y=YES,N=NO): Y RATE OF RETURN(%): 2.1 CONTRIBUTION TO INVESTMENT: 16.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL S. 0.050 RATES OF EXCHANGE 100.0000 TARIFFS GWH 119 163 as 290 1181 % 10.2500 14.0400 5O.7300 24.9800 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 2.0 1.4 1.4 1.4 2.3 1.7 1.6 1.315 99.9900

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (US$M) : GNP/CAPITA (USS) : ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (%) ACCESS TO ELECTRICITY (%) : 14350 2720 190 1.315 1352 0.337 30.5 92 94 URBAN 24 URBAN 23

RURAL 76 RURAL I

2. INSTALLED CAPACITY & GROSS GENERATION MW ,,A, 41. THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTORSUPPLY 332 1230 GWH

332 3 Ia 351

1230 25 45 1300

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL MTOE 0.050 % 100.0000

AVE 75 USS1=15.6 SL RS AVE 1980 US$1=1.53 SL RS

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW. .35LF) INDUSTRIAL (200KW,.70LF) OTHER

SUDAN

PUBLIC ELECTRICITY & WATER CORP IS THE MAIN PRODUCER (90% OF TOTAL ELECTRICITY PRODUCED)& SOLE DISTRIBUTOR-INTERCONNECTED 1973-78 AVG ANNJAL GTH RATE (%) 2.60 5.10 2.50 0.40 12.70 5. COMMERCIALENERGY CONSUMPTION MTOE % 1.952 94.0700

1. SOCIO-ECONOMIC DATA POPULATION (000) : GNP (USSM) GNP/CAPITA (USS) : ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) : ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) : POPULATION BREAKDOWN (%) ACCESS TO ELFCTRCTITY (%) : 17390 5540 320 2.075 833 0.208 10.0 119 48 URBAN 13 URBAN

OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL

0.123

5.9300

RURAL 87 RURAL

2.075

100.0000

TOTAL

6. UTILITY DATA FREQUENCY(HZ): S0 TRANSMISSION (KV): 220/110 DISTRIBUTION (KV): 33/11 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (%): 81.0 TOTALSYSYSTEM CAPACITY FACTOR (%): 37.0 TOTAL SYSTEM T&D LOSSES (%): 18.0 NUMBEROF CONSUMERS: 165000 NUMBEROF EMPLOYEES: 7200 CONSUMERS/EMPLOYEE: 23 AVERAGEREVENUE (US CENTS/KWH) 7.2 ASSETS REVALUED (POLICY)(Y=YES,N=NO): Y RATE OF RETURN(%): 5.8 CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COSMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS GWH 206 48 215 105 574 % 38.8900 8.3600 37.4600 18.2900 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 '100000 9.3 9.3 9.3 7.6 S.5 S.5 4.9

2. INSTALLED CAPACITY & GROSS GENERATION THERMAL-STEAM -GAS TURBINES -cOMBINED CYCLE -DIMESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTOR SUPPLY MW 30 15 71 118 118 GWH

234 26 260 83 83

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL MTOE 0.225 % 100.0000

0.225

100.0000

4. OFFICIAL RATES OF EXCHANGE AVE 78 USS1=.375 SUDANESELBS AVE 1980 USS1=.500 SUD POUNDS

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW,. 70LF) OTHER

SURINAME

EBS(ENERGY CO OF SURINAME) RESPONSIBLE FOR GENERATION OF ELECTRICITY 0IST D OF GAS & ELEC TO PUBLIC 1973-78 AVG ANNUAL GTH RATE (%) -1.10 5.20 8.30 -2.90 S. COMMERCIALENERGY CONSUMPTION X MTOE 60.9400 0.546 OIL GAS 2.2300 COAL 0.020 LIGNITE HYDRO 0.330 36.8300 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 61 S. UTILITY DATA 60 FREQUENCY(HZ): TRANSMISSION (KV): 33 DISTRIBUTION (KV): LOW VOLTAGE (VOLTS): 220 MAIN SYSTEM LOAD FACTOR (X): 66.0 41.0 TOTAL SYSTEM CAPACITY FACTOR (X): 11.3 TOTAL SYSTEM T&D LOSSES (X): NUMBEROF CONSUMERS: 63870 550 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 116 9.8 AVERAGEREVENUE (US CENTS/KWH) ASSETS REVALUED (POLICY)(Y=YES,N=NO): Y RATE OF RETURN(X): 18.1 CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL S. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW. .70LF) OTHER GWH 78 40 77 17 212 % 36.7900 18.8700 36.3200 8.0200 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 21.3 23.2 19.5 23.8 21.5 18.9 0.898 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) : GNP (US$M) : GNP/CAPITA (USS) : ENERGY CONSUMED(MTOE) : ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): : ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) : (X) POPULATION BREAKDOWN : ACCESS TO ELECTRICITY (X) 2. 389 820 2110 0.898 1631 0.407 48.0 2303 4193 URBAN 40 URBAN 88

RURAL 6 RURAL 44

INSTALLED CAPACITY & GROSS GENERATION MW GWH

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTOR SUPPLY 3.

45 45

160 160

1B 81 189 181 411

SI 241 904 511 1656

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE X

OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE

AVE 78USS1=1.785 GUILDERS AVE 1980 US$1s1.785 0

SWAZILAND

SWAZILAND ELECTRICITY BOARD IS THE SOLE AUTHORIZEDPUBLIC SUPPLIER OF POWER FACTORIES HAVE OWN POWERSTAS SUGAR & WOOOPULP 1973-78 AVG AN()AL GTH RATE (% 2.50 8.10 5.80 5. COMMERCIALENERGYCONSUMPTION MTOE 40.0900 0.091 OIL GAS 35.6900 0.081 COAL LIGNITE 15.4200 : 0.035 HYDRO NUCLEAR GEOTHERMAL 8.8100 : 0.020 OTHER TOTAL 0 TOTAL 15 B. UTILITY DATA 0.227 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) : GNP/CAPITA (USS) ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): : ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) (X) POPULATION BREAKDOWN ACCESS TO ELECTRICITY () 2. 528 310 590 0.227 398 0.099 43.9 432 757 URBAN URB.AN 30

RURAL RURAL

INSTALLED CAPACITY & GROSS GENERATION MW GWH

N1

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE M-IESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERMAL TOTAL SECTOR SUPPLY

10 10 21

5 5 140 104 249 150 399

31 35 68

FREQUENCY (HZ): 50 TRANSMISSION (KV): 132/68 DISTRIBUTION (KV): LOW VOLTAGE (VOLTS): 400 54.0 MAIN SYSTEM LOAD FACTOR (X): 55.0 TOTAL SYSTEM CAPACITY FACTOR (%): 10.3 TOTAL SYSTEM T&D LOSSES (%): 8373 NUMBEROF CONSUMERS: 355 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 24 3.1 AVERAGEREVENUE (US CENTS/KWH) ASSETS REVALUED (POLICY)(Y=YES,N=NO): 8.0 RATE OF RETURN(X): CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS GWH 41 22 98 92 223 % 18.3900 9.8700 43.9500 27.8000 100.0100 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 3.4 3.4 3.4 4.8 2.1 2.1 13.8

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL MTOE 0.001 X 100.0000

0.001

100.0000

4. OFFICIAL RATES OF EXCHANGE AVE 78USS1-=.89 LILANGENI AVE 1980 USS1=.778 LIL

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW,.35LF) INDUSTRIAL (200KW, .70LF) OTHER

SYRIA

ETABLISSEMENT PUBLIC D'ELECTRICITE ALL PUBLIC ELECTRICITY SUPPLYSYSTEM INTERCONNECTED 1973-78 AVG ANNUAL GTH RATE (X) 3.30 9.40 6.10 21.20 11.70

PROVIDES

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) : GNP/CAPITA (USS) ENERGYCONSUMED(MTOE) : ELECTRICTY PRODUCE (GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): : ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) (%) POPULATION BREAKDOWN : ACCESS TO ELECTRICITY (%) 8088 7490 930 5.619 2090 0.523 9.3 695 258 URBAN 46 URBAN 80

OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL

S. COMMERCIAL ENERGYCONSUMPTION X MTOE 83.5000 : 4.692 0.486 8.6500 : 0.003 O.OSOO : 0.438 7.7900

RURAL 54 RURAL 20

5.619

99.9900

TOTAL

48 6. UTILITY DATA 50 FREQUENCY(HZ): TRANSMISSION (KV): 230/66 DISTRIBUTION (KV): 20 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (X): 58.7 31.5 TOTAL SYSTEM CAPACITY FACTOR (X): TOTAL SYSTEM TAO LOSSES (7)): 17.0 NUMBEROF CONSUMERS: 15773 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 3.4 AVERAGE REVENUE (US CENTS/KWH) ASSETS REVALUED (POLICY)(Y=YES,N=NO): RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS GWH 738 870 386 1994 V 37.0100 43.9300 19.3600 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 3000 100000 4.8 4.8 4.8 4.8 5.7 2.6 3.8

2. INSTALLED CAPACITY & GROSS GENERATION MW THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL 00 HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERMAL TOTAL SECTOR SUPPLY GWH 270 48 41 359 2134 -S9 2434 160 2594

~~~~-SUBTOTAL
523

523 95 BiB

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE AVE 78USS1=3.925 SYRIAN POUNDS AVE 1980 US$Si3.925 S POUNDS X

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .3SLF) INDUSTRIAL (200KW,.7OLF) OTHER

TANZANIA

TANESCO-A PUBLIC SECTORPOWERCO IS RESPONSIBLE FOR GENERATION, TRANSMISSION & DISTRIBUTION OF ELECTRIC POWER 1973-78 AVG ANNUAL GTH RATE (%) 3.00 8.10 2.10 -4.50 2.90 5. COMMERCIAL ENERGY CONSUMPTION MTOE X 86.0400 0.900 OIL GAS COAL LIGNITE 13.9600 0.146 HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 7 S. UTILITY DATA 1.046 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (US$M) : GNP/CAPITA (USS) : ENERGY CONSUMED(MTOE) : ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): : ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) : (X) POPULATION BREAKDOWN ACCESS TO ELECTRICITY (%) 2. 16871 3880 230 1.048 758 0.190 18.0 62 44 URBAN URBAN

RURAL 93 RURAL

INSTALLED CAPACITY & GROSS GENERATION MW GWH 1 95 98 584 3 683 75 758

'C

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERMAL TOTAL SECTOR SUPPLY 3.

12 72 84 151

235 4 239

50 FREQUENCY(HZ): TRANSMISSION (KV): 220/132/66 DISTRIBUTION (KV): 11/6.6/3 LOW VOLTAGE (VOLTS): 220/400 MAIN SYSTEM LOAD FACTOR (X): 33.2 TOTAL SYSTEM CAPACITY FACTOR (%): 13.3 TOTAL SYSTEM T8D LOSSES (%): NUMBEROF CONSUMERS: 105117 4000 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 26 5.8 AVERAGEREVENUE (US CENTS/KWH) Y ASSETS REVALUED (POLICY)(Y=YES,N=NO): 2.5 RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 20.7 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL B. TARIFFS GWH 112 64 406 6 588 % 19.0500 10.8800 69.0500 1.0200 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 10.4 8.1 5.1 24.4 9.9 5.4

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE 0.027 % 100.0000

OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL

0.027 RATES OF EXCHANGE

100.0000

AVE 78 USS1=8.1 SHILLINGS AVE 1980 USS1=8.2 SH

OOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW, .70LF) OTHER

THAILAND

ELECTRICITY GENERATING AUTHORITY OF THAILAND GENERATES94% OF TOTAL ELECTRICITY. MEA AND PEA STATE ENTERPRISES DISTRIBUTE ELEC IN BANGKOKMETRO 1973-78 AVG ANNUAL GTH RATE (%) 2.80 7.00 4.10 5.10 11.30 5. COMMERCIAL ENERGYCONSUMPTZON MTOE OIL 9.497 87.B800 GAS COAL LIGNITE 0.139 1.2900 HYDRO 1.171 10.8400 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 27 6. UTILITY DATA FREQUENCY(HZ): 50 TRANSMISSION (KV): 230/115/69 DISTRIBUTION (KV): 33/22 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (%): 70.7 TOTAL SYSTEM CAPACITY FACTOR (%): 40.0 TOTAL SYSTEM T&O LOSSES (%): 10.0 NUMBER OF CONSUMERS: 1957000 NUMBEROF EMPLOYEES: 33231 CONSUMERS/EMPLOYEE: 59 AVERAGE REVENUJE (US CENTS/KWH) 2.7 ASSETS REVALUED (POLICY)(Y=YES,N-NO): RATE OF RETURN(%): 0.5 CONTRIBUTION TO INVESTMENT: 87.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 0.230 2.550 9.0200 100.0000 8. TARIFFS GWH 2718 1856 7822 98 12492 % 21.7400 14.8600 62.6200 0.7500 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 6.1 6.1 8.1 5.9 5.4 4.8 10.807 100.0100

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (US$M) GNP/CAPITA (USS) ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) : ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (X): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (X) ACCESS TO ELECTRICITY (X) : 2. 44345 21790 490 10.807 12685 3.170 29.3 244 286 URBAN 17 URBAN 70

RURAL 83 RURAL 18

INSTALLED CAPACITY & GROSS GENERATION GWH

CZ

MW THERMAL-STEAM 2087 -GAS TURBINES 165 -COMBINED CYCLE -DIESEL 302 -SUBTOTAL 2584 HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY 910

4500 185 4665 535 5200

3484

AUTO PRODUCERS HYDRO THERMAL 507 TOTAL SECTOR SUPPLY 3971

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL MTOE 2.320 % 90.9800

4. OFFICIAL RATES OF EXCHANGE AVE 78USS1-20.4 BAHT AVE 1980 USSI=20.5 BART

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL I8KW,.3LF) INDUSTRIAL (280KW, .7OLF) OTHER

TUNISIA

SOCIETE TUNISIENNE DE L'ELECTRICITE SYTEM INTERCONNECTED

ET DU GAS

1. SOCIO-ECONOMIC DATA : POPULATION (000) GNP (US$M) GNP/CAPITA (uss) (MTOE) ENERGY CONSUMED ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) ELECTRICIT.Y/CAPITA (KWH) (x) POPULATION BREAKDOWN ACCESS TO ELECTRICITY (V) 6050 5760 950 2.251 1987 0.497 22.1 372 328 URBAN 47 URBAN 78

1973-78 AVG ANNUAL GTH RATE (%) 2o00 8.50 6.50 9.90 8.90

OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL

s. COMMERCIAL ENERGYCONSUMPTION % MTOE 84.6300 1.905 11.5900 0.261 3.3800 0.076 : 0.009 0.4000

RURAL 53 RURAL 10

2.251 DATA

100.0000

TOTAL 44 6. UTILITY

2. INSTALLED CAPACITY & GROSS GENERATION THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERMAL TOTAL SECTOR SUPPLY MW 230 333 17 580 30 30 GWH

610 so 66O

30 203 233

50 FREQUENCY(HZ): TRANSMISSION (KV): 225/150/90 DISTRIBUTION (KV): 30/15/10 LOW VOLTAGE (VOLTS): 110/220 MAIN SYSTEM LOAD FACTOR (%): 54.5 34.4 TOTAL SYSTEM CAPACITY FACTOR (): 13.3 TOTAL SYSTEM TBD LOSSES (%): 552800 NUMBEROF CONSUMERS: 4783 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE:116 6.4 CENTS/KWH) (US REVENUE AVERAGE ASSETS REVALUED (POLICY)(Y=YES,N=N0): 7.4 RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 37.2 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL S. TARIFFS OWH 515 755 267 1537 X 33.5100 49.1200 17.3700 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 7.0 7.0 7.0 7.0 7.0 7.0

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL MTOE 0.342 0.259 X 56.9100 43.0900

0.601

100.0000

4. OFFICIAL RATES OF EXCHANGE AVE 78Us$l= .418 DINARS AVE 1980 USS1=.40S DINARS

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW 3sLF) INDUSTRIAL (200KW, . 70LF) OTHER

TURKEY

TURKIYE ELECTRIK KURUMUIS RESPONSIBLE FOR BULK SUPPLY PLANNING AND OPERATIONS. IT ACOOUNTSFOR 91% OF TOTAL SALES 1973-78 AVG ANNUAL GTH RATE (%) 2.50 7.00 4.50 7.70 11.30 5. COMMERCIAL ENERGYCONSUMPTION MTDE % OIL 17.227 69.1200 GAS COAL 5.261 21.1100 LIGNITE HYDRO 2.436 9.7700 NUCLEAR GEOTHERMAL OTHER TOTAL TOTAL 72 B. UTILITY DATA FREQUENCY(HZ): 50 TRANSMISSION (KV): 380/154/66 DISTRIBUTION (KV): 34.5/15/6.3 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (%): 61.9 TOTAL SYSTEM CAPACITY FACTOR (%): 52.1 TOTAL SYSTEM T&D LOSSES (%: 10.3 NUMBEROF CONSUMERS: NUMBEROF EMPLOYEES: 5323 CONSUMERS/EMPLOYEE: AVERAGEREVENUE (US CENTS/KWH) 3.4 ASSETS REVALUED (POLICY)(Y=YES,N=NO): RATE OF RETURN(%): 3.9 CONTRIBUTION TO INVESTMENT: 11.0 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW..35LF) INDUSTRIAL (200KW,.70LF) OTHER GWH 2796 1198 13665 1141 19020 % 14.7000 8.3000 73.0000 6.0000 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 11.0 11.0 11.0 11.0 11.0 11.0 24.924 100.0000

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) GNP/CAPITA (USS) ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL ECUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (%) ACCESS TO ELECTRICITY (%) 2. 42925 51750 1120 24.923 22221 5.555 22.3 581 51g URBAN 43 URBAN 95

RURAL 57 RURAL 45

INSTALLED CAPACITY & GROSS GENERATION MW GWH

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY 1867 9330 600 9830 30 1669 11549

1867

AUTO PRODUCERS HYDRO 12 THERMAL 588 TOTAL SECTOR SUPPLY 2467 3.

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE %

OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL RATES OF EXCHANGE

AVE 78USS1=24.3 LIRAS AVE 1980 USS1=76.0 LIRAS

URUGUAY

ADMINISTRACION GENERAL DE USINAS Y TRANSMISIONES ELECTRICA IS RESPONSIBLE FOR GENERATION. TRANSMISSION SYSTEM. AND DISTRIBUTION. MAIN INTERCONNECTED AVG ANNUAL i973-78 GTH RATE ()MTOE 0.20 1.SO 1.30 0.04 5.80 S. COMMERCIAL ENERGYCONSUMPTION OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL O B. UTILITY DATA FREQUENCY (HZ): SO TRANSMISSION (KV): 150/110 DISTRIBUTION (KV): 80/30/8 LOW VOLTAGE (VOLTS): 220/380 S8.5 MAIN SYSTEM LOAD FACTOR (%): 51.2 TOTAL SYSTEM CAPACITY FACTOR (%): TOTAL SYSTEM T&D LOSSES (V.): 18.7 790335 NUMBEROF CONSUMERS: 9718 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: S1 4.8 (US CENTS/KWH) AVERAGEREVENUE Y ASSETS REVALUED (POLICY)(Y-YES,N=NO): 10.5 RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 7. SALES GWH 1089 357 897 82 2405 % 44.4500 14.8400 37.3000 3.4100 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 4.8 0.7 7.1 10.3 4.7 1.891 0.018 0.433 80.7400 0.7700 18.4900

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM4) GNP/CAPITA (USS) ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE (GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGODE) ELECTRICITY/CAPITA (KWH) (X1 POPULATION BREAKDOWN ACCESS TO ELECTRICITY )%. 2. 2892 4860 1610 2.342 3284 0.821 35.1 810 1135 URBAN 81 URBAN 95

RURAL 19 RURAL SS TOTAL

2.342

100.0000

INSTALLED CAPACITY & GROSS GENERATION MW 333 30 60 423 238 GWH 84 84 1236 1364 1830 29 3023 325 3348

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY HYDRO AUTO PRODUCERS THERMAL TOTAL SECTORSUPPLY

859 140 799

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL MTOE 0.341 % 100.0000

DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL 8. TARIFFS

0.341 RATES OF EXCHANGE

100.0000

AVE 78USS1=8. 13 NEW PESOS AVE 1980 US$1=9.18 N PESOS

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW, .35LF) INDUSTRIAL (200KW. .70LF) OTHER

YEMEN ARAB REPUBLIC

YEMEN GENERAL ELECTRICITY CORPORATION OPERATES THREE MAIN ISOLATED SYSTEMS

1. SOCIO-ECONOMIC DATA : POPULATION (000) GNP (USSM) GNP/CAPITA (USS) : ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE(GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) : ELECTRICITY/CAPITA (KWH) POPULATION BREAKDOWN (%) : : ACCESS TO ELECTRICITY (% 5098 2960 580 0.262 85 0.018 6.2 51 13 URBAN 7 URBAN 62

1973-78 AVG ANNUAL GTH RATE (%) 1.90 19.20 15.20

S. COMMERCIAL ENERGYCONSUMPTION MTOE % 100.0000 OIL 0.262 GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL 0.262 DATA 100.0000

RURAL 93 RURAL 5

TOTAL

9 6. UTILITY 50 FREQUENCY(HZ): TRANSMISSION (KV): DISTRIBUTION (KV): 15/11/10 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (%): 26.2 TOTAL SYSTEM CAPACITY FACTOR (%): TOTAL SYSTEM T&D LOSSES (%): 22.5 NUMBEROF CONSUMERS: 40600 888 NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: 46 13.3 AVERAGEREVENUE (US CENTS/KWH) ASSETS REVALUED (POLICY)(Y=YES,N=NO): RATE OF RETURN(%): 2.4 CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL B. TARIFFS KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 21.9 21.9 21.9 21.9 21.9 21.9 GWH %

2. INSTALLED CAPACITY & GROSS GENERATION MW THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY AUTO PRODUCERS HYDRO THERMAL TOTAL SECTOR SUPPLY 4 IB 23 GWH

23 5 28 15 Is

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL MTOE 0.059 % 100.0000

0.059

100.0000

4. OFFICIAL RATES OF EXCHANGE AVE 78US51=4.58 YEMEN RIALS AVE 1980 USS1=4.5B Y RIALS

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW,.35LF) INDUSTRIAL (200KW,.70LF) OTHER

YEMEN PDR

PUBLIC CORPORATIONOF ELECTRIC POWER OPERATES A MAIN SYSTEM IN ADEN (89% OF SALES) AND A FEW ISOLATED SYSTEMS 1973-78 AVG ANNUAL GTH RATE (%) 1.90 13.10 11.20 5. COMMERCIALENERGY CONSUMPTION MTOE X OIL 0,644 100.0000 GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL S. UTILITY 0.644 DATA 100.0000

1. SOCIO-ECONOMIC DATA : POPULATION (000) : GNP (USSM) GNP/CAPITA (USS) ENERGYCONSUMED(MTOE) ELECTRICTY PRODUCE (GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) : POPULATION BREAKDOWN (X) : ACCESS TO ELECTRICITY (X) 1749 740 420 0.644 262 0.087 10.2 368 150 URBAN 40 URBAN 45

RURAL 60 RURAL 6

TOTAL

22

2. INSTALLED CAPACITY & GROSS GENERATION THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL ,, -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED(NET) TOTAL UTILITY SUPPLY AUTO PROaDUCERS HYDRO THERMAL TOTAL SECTOR SUPPLY MW 13 13 40 68 GWH

68 26 94

36 38 70 106

50 FREQUENCY(HZ): TRANSMISSION (KV): 33 DISTRIBUTION (KV): 33/11 LOW VOLTAGE (VOLTS): 240/400 MAIN SYSTEM LOAD FACTOR (%): 52.0 TOTAL SYSTEM CAPACITY FACTOR (%): 31.8 12.0 TOTAL SYSTEM T&D LOSSES (X): 63000 NUMBEROF CONSUMERS: 1540 NUMBER OF EMPLOYEES: CONSUMERS/EMPLOYEE: 41 6.3 AVERAGE REVENUE (US CENTS/KWH) ASSETS REVALUED (POLICY)(Y5YES,N=NO): RATE OF RETURN(%): 0.0 CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL S. TARIFFS GWH 57 34 54 12 157 % 36.3100 21.6600 34.3900 7.6400 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000 5.8 4.9 7.1 7.9 7.8 7.8 7.8

3. FOSSIL FUEL FOR ELECTRICITY PRODUCTION OIL GAS COAL LIGNITE OTHER TOTAL 4. OFFICIAL MTOE 0.054 % 100.0000

0.054 RATES OF EXCHANGE

100.0000

AVE 78USS1=.345 DINAR AVE 1S80 USS15.345 DINAR

DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW,.35LF) INDUSTRIAL (200KW,.70LF) OTHER

YUGOSLAVIA

UNION OF YUGOSLAV ELECTRIC POWER INDUSTRY UNION OF ALL SELF-MANAGEDBOAL (BASIC ORG. OF ASSOCIATED LABOR) IN THE 8 REPUBLICS 1973-78 AVG ANNUAL GTH RATE (%) 0.90 6.00 5.10 5.20 7.80

1. SOCIO-ECONOMIC DATA POPULATION (000) GNP (USSM) : GNP/CAPITA (USS) ENERGY CONSUMED(MTOE) ELECTRICTY PRODUCE (GWH) ELECTRICITY OIL EQUIV (MTOE): ELECTRICITY/TOTAL ENERGY (%): ENERGY/CAPITA (KGOE) ELECTRICITY/CAPITA (KWH) : POPULATION BREAKDOWN (x) ACCESS TO ELECTRICITY (x) 2. 21933 52340 2390 34.437 55200 13.800 40.1 1572 2518 URBAN 40 URBAN 100

5.

OIL GAS COAL LIGNITE HYDRO NUCLEAR GEOTHERMAL OTHER TOTAL

COMMERCIALENERGY CONSUMPTION MTOE 14.022 40.7200 1.788 5.1900 12.477 38.2300 :.153 17.8700

RURAL 50 RURAL 84

34.438 DATA

100.0100

TOTAL

90 6. UTILITY FREQUENCY (HZ): 50 TRANSMISSION (KV): 400/220/110 DISTRIBUTION (KV): 35/10 LOW VOLTAGE (VOLTS): 220/380 MAIN SYSTEM LOAD FACTOR (%): 69.0 TOTAL SYSTEM CAPACITY FACTOR (%): 50.5 TOTAL SYSTEM T&D LOSSES (x): NUMBEROF CONSUMERS: NUMBEROF EMPLOYEES: CONSUMERS/EMPLOYEE: AVERAGEREVENUE (US CENTS/KWH) ASSETS REVALUED (POLICY)(Y=YES,N-NO): Y RATE OF RETURN(%): CONTRIBUTION TO INVESTMENT: 7. SALES DOMESTIC COMMERCIAL INDUSTRIAL OTHER TOTAL GWH 13075 1594 27141 1907 43717 % 29.9100 3,esoo 92.080o 4.3800 100.0000 KWH/MONTH US CENTS/KWH 25 100 400 400 2000 100000

INSTALLED CAPACITY & GROSS GENERATION MW GWH

THERMAL-STEAM -GAS TURBINES -COMBINED CYCLE -DIESEL -SUBTOTAL HYDRO NUCLEAR GEOTHERMAL PURCHASED (NET) TOTAL UTILITY SUPPLY 5396 25379 652 26031 100 2700 25831

5398

AUTO PRODUCERS HYDRO 120 THERMAL 545 TOTAL SECTORSUPPLY 8o01 3.

FOSSIL FUEL FOR ELECTRICITY PRODUCTION MTOE 4.080 6.250 10.330 % 39.SOOO 80.5000 100.0000

OIL GAS COAL LIGNITE OTHER TOTAL 4

B. TARIFFS DOMESTIC DOMESTIC DOMESTIC COMMERCIAL INDUSTRIAL (8KW,. 38LF) INDUSTRIAL (200KW,.70LF) OTHER

OFFICIAL RATES OF EXCHANGE

AVE 7SUSSl=18.B DINARS AVE 1980 USS1=24.9 DINARS

Appendix C

Allocation ofCapacity andEnergy CostsbetweenPeak andOff-Peak GCnsumers

typical electric power generation system presented in this appendix shows, conceptually, how a long-run marginal cost (LRMC) analysis based on the optimal system expansion plan yields the following idealized conclusions.'
THE SIMPLIFIED MODEL OF A

a. Peak users should. pay the full LRMC of capacity, as well as energy costs. b. Off-peak users should pay only the off-peak LRMC of energy. c. LRMC of peak capacity = LRMC of base-load capacity minus the net fuel savings from this base-load plant. Consider an all-thermal generation system having the annual load duration curve (LDC) shown in F'igure C-1. The linearized cost characteristics of two types of plant are given in the table below and also in the figure. For the moment, all losses, reserve margin, and so on are ignored. Capacitycost per kilowattinstalled (annuitized)
a b

Plant type
Peaking (gas turbines, GT) Base-load (steam)

Hourly operating costs

e f

1. A more realistic system model would have to consider several complicating factors, such as a larger number of rating periods and plant types, noncoincidence of the rating periods with the economic crossover points between different plant types, economies of scale and variable heat rates for a given plant type, hydroelectric plant (especially pumped storage, reserve margins and stochasticity of supply and demand, and so on). Generally, the most important difference is that consumers outside the peak period would have to bear some capacity costs. The bulk of the capacity costs would still be allocated to peak period users, however. A simplified exposition of this result is provided in J. T. Wenders, "Peak Load Pricing in the Electricity Utility Industry," The Bell Journal of Economics, vol. 7 (Spring 1976), pp. 232-41.

327

328

APPENDIXES

Figure C-1. Plant Costs and Annual Load Duration Curve

GT (slopee) Steam (slopef)

01

n'

tH
Hours of operation

+ '~G y I
yA

'N

~~B

~~~~~~~~~E

Hours per year

8,760

IE" Peak

"i

Off-peak

Total cost of 1 kilowatt, which is used h hours annually, is a + e h for GT and b + f h for base-load. Let H be the hours of operationthat correspondsto the crossoverpoint for whichpeakingand base-loadplant total costs are equal. Therefore,a + e H = b + f H, and (C.1) H = (b - a)I(e -f). The most economic use of plant can be determined by examining the LDC, OABCEF:for plannedbase-loadoperation(more than H hours annually), use base-loadplant (X kilowatts);and for plannedpeak operation(less

ALLOCATION

OF COSTS

329

than H hours annually), use GT (Y - X kilowatts). Total annual costs (C 0) of meeting the demand clepicted by the LDC is
CO = X(b + fT) + (Y
-

X) (a + eH)

where T = 8,760 hours denotes that the base-load plant is used all year. Case 1. Only peak-period demand increases by 1 kilowatt (as shown by shaded area AGNB in Figure C-1). The optimal system planning response is to increase GT by 1 kilowatt. Total annual cost is
C, = X (b + fT) + (Y + 1 - X) (a + e H).

Therefore, the increase in cost is

C, - CO= a + e H.

This is the increase in system costs incurred because of the 1 kilowatt increase in marginal (or incremental) demand during the peak period, and thus the peak period user must pay this cost. The peak costs consist of Capacity charge = a per kilowatt annually, and Energy charge = e per kilowatt-hour. The peak users' paymenit = a + e H = the increase in system costs. Case 2. Only off-peak: period demand increases by 1 kilowatt (as shown by shaded area CIJE in Figure C-1). The optimal system planning response is to add 1 kilowatt more of base-load plant. But now there is 1 kilowatt less of GT required than before. Total annual cost is C 2
=

(X + I)' (b + fT) + [Y-

(X + 1)] (a + e H).

Therefore, increase in cost is

C2 -

Co = (b + f7)

(a + e H).

Substituting for H from equation (C. 1), the increase in cost is

C2 -

C.

(b

a) + (f

e) [(b
-

a)l(e

f)] + f (T

H)

= f(T

H).

Therefore the increase in system cost incurred because of the 1 kilowatt increase in marginal off-peak demand is equal to the energy cost of operating the base-load plant during this period. Thus, off-peak users must pay only this energy charge f per kilowatt. There are no capacity costs incurred
by off-peak users, since off-peak users' payments = f(T
-

H) = the in-

crease in system cost. In particular, the base-load capacity cost b is exactly offset by the total cost saving because GT is not required any more (that is, capacity cost a

330

APPENDIXES

plus net fuel cost saving (e - f) H inside the shaded area LKIC). In other words, peak capacity cost = base load capacity cost - net fuel savings
a
=

b - (e

f) H.

Case 3. Both peak and off-peak demand increases by 1 kilowatt. This case is a linear combination of cases 1 and 2 and, therefore, consumer charges include Peak capacity charge: a peak kilowatts annually; Peak energy charge: e per kilowatt-hour; Off-peak energy charge: f per kilowatt-hour. Clearly, total peak and off-peak users' payments = b + f T = the increase in system cost. These results may be generalized to include more types of generating units and pricing time periods, for example, n plant types and n pricing periods, where these pricing periods are chosen to coincide with the economic crossover points between competing types of plant. In this case LRMC prices would be peak period: capacity charge a, per kilowatt annually, and energy charge e, per kilowatt-hour H, to H2 = 2nd period: only energy charge e2 per kilowatt-hour H__ to T = nth period: only energy charge e, per kilowatt-hour.
=

0 to H,

AppendixD

STLRMC: A Simple Computer Program forEstimating Strict Long-Run Marminal Costs

A SIMPLE COMPUTER PROGRAM, STLRMC, can be used to calculate the strict long-run marginal cost of electricity supply as described in Chapter 4. The program uses the following steps: a. Determination of loss multipliers b. Determination of incremental demand forecast at different voltage levels c. Calculation of marginal transmission and distribution capacity cost d. Calculation of marginal energy cost e. Calculation of marginal generation capacity cost. Further details regarding use of the program are given in the program listing and sample outputs at the end of the appendix.

Calculation of Loss Multipliers

There are two options, A or B, depending on the data inputs to the system. A. Both the average loss coefficients and the multiplying factor (that is, the proportion by which peak-period loss coefficient is greater than the average loss coefficient) are known at each intermediate voltage level. B. The average and the peak-period loss coefficient is known at each intermediate voltage level. If the set of inputs specified in A above are used, then the set of variables as specified in B are computed as follows: PPLF(I,J) = [SLF(I,J)] MF

Note: This appendix was prepared in collaboration with Rajesh Pradhan and Karl Jechoutek.

331

332

APPENDIXES

where PPLF(I,J) SLF(I,J)

MF

= peak period loss coefficient as a fraction of incoming

power or energy at voltage level I in year J,

= average loss coefficient as a fraction of incoming

power or energy at voltage level I in year J, and

= Multiplying factor.

The average peak and off-peak loss coefficient of the whole planning period are given by:
Peak: PLF(I) = ALF(I)
=

PPLF(I,J)

I (NYTD + 1)

Off-peak:

,jv,D+) SLF(I,J) / (NYTD + 1)

where NYTD = number of years of transmission and distribution (T&D) data considered in the planning period. The corresponding loss multipliers are then calculated for voltage level as follows:
AMLF(I) PMLF(I) = 1/ [1 - ALF(I)] = 1/ [1 - PLF(I)]

where AMLF(l) = average loss multiplier at voltage level I, and PMLF(I) = peak loss multiplier at voltage level I. The cumulative loss multipliers, which reflect full losses occurring at a particular voltage level of service, including those losses occurring upstream from the point of consumption, are calculated as follows: CAMLF(I) CPMLF(I)
= = HL,

IAMLF(I)
PMLF(I)

where CAMLF(I) = cumulative average loss multiplier at voltage level I, CPMLF(I) = cumulative peak loss multiplier at voltage level I, and L = 1, 2, 3 for EHV/HV, Mv, and LV, respectively.

Determination of Incremental Demand Forecast at Different Voltage Levels

The forecasts of peak demand at generation and of consumers at internediate supply voltage levels are used as inputs. The projected peak demand at different voltage levels is derived as follows: LOSSVL(1,J) = CPAVL(1,J) x PPLF(1,J) LOSSVL(2, J) = CPAVL(1, J) x PPLF(2, I)

PROGRAM

FOR ESTIMATING

STRICT LRMC

333

SPAVL(2, ) PAVL(2, J)

CPAVL(l, J) - LOSSVL(l, J) - LOSSVL(2, J) SPAVL(2, J) - CPAVL(2, J)

where LOSSVL(1, J) = station use in year J (megawatts) CPAVL(1, J) = generation peak in year J (megawatts) PPLF(1, J) = peak loss factor at the auxiliary in year J (in decimal) LOSS(2, J) = I_HV/HV transmission losses in year J SPAVL(2, J) = HV peak demand available in year J before consumption PAVL(2, J) = peak available after consumption CPAVL(2, J) = peak demand forecast at HV in year J. Similarly, the projected peak demand at successive voltage levels can be derived. LOSSVL(I, J) = PAVL(I - 1, J)
SPAVL(I, J)
=

PPLF(I, J)

PAVL(I -

1, J) - LOSSVL(I, J)

PAVL(I, J)
forI = 3, 4.

SPAVL(I, J) - CPAVL(I, J)

The projected incremental peak demand at different voltage levels is then derived as follows:
CUVL(I for I = 2, 1, J 1) = SPAVL(I, J) - SPAVL(I, J 1)
=

* 4 and for J = 1, * - * (NYTD + 1), where CUVL(I, J)

projected incremental peak demand at voltage level I in year J.

Calculation of M[arginalT&D Capacity Cost

Marginal T&D capacity cost is calculated using the average incremental cost (AIC) method. It uses the concept of structuring by voltage levels. Consumers at a voltage level are charged only upstream costs. The AIC of T&D capacity at a given voltage level is calculated as shown below: ADTINV(I, J) ADCUVL(I, J) ACT(I)
=

= =

AATINV(I, 1)1(1 + DR)J+K ACUVL(I, J)/(l + DR)`+ ADTINV(I, J)

37
LJI+L

NYA+LAGT(J)
G

AVL(I) =
where K = DYEAR

ADCUVL(I,J)

ACK(I) = ACT(I)IAVL(I)
- BYEAR - 1 DYEAR = beginning year for the demand forecast BYEAR = year in which study is being carried out

334

APPENDIXES

AATINV(I, J) ACUVL(I, J)

= =

incremental investment directly associated with T&D facilities for voltage level I in year J incremental peak demand transmitted at voltage level I in year J

DR = discount rate ACT(I) = net present value of annual incremental investment for T&D facilities at voltage level I AVL(I) = net present value of annual incremental peak demand transmitted at voltage level I ACK(I) = average incremental cost for T&D facilities at voltage level L. The marginal T&D capacity cost directly associated with voltage level I is then given by: AMCTV(I) = 1/12 [AAF(I) + AOMT(I)] ACK(I)ISSCF
CCTV(I) = AMCTV(I)

where AAF(I) = annuitizing factor for T&D facilities at voltage level I SSCF = standard commission factor OMT(I) = operation and maintenance (O&M) cost per year as a percentage of AIC of T&D. The total marginal T&D capacity cost at a voltage level is the sum of the marginal cost debited from higher voltage levels plus the marginal T&D cost directly associated at that level. The marginal T&D capacity cost debited to lower voltage levels is given by: M(I, J) = CCTV(I) [CPMLF (J + l)lCPMLF(I + 1)]
SM(J, I = M(V, J)

SM(I, n = CCTV(I)

for J = (I + 1), - - 3, and I = 1, 2. The total marginal T&D capacity cost at a given voltage level I is given by: MCTV(I) =

M,=, SM(I, J).

Calculation of Marginal Energy Cost

There are two options, A or B, depending on the input data to the system. A. NEC = 1, that is, marginal energy costs are prespecified (for example, hydroelectric system). B. NEC i 1. Inputs to the system are the fuel cost, calorific value of fuel, and thermal conversion factor for the marginal plant.

PROGRAM

FOR ESTIMATING

STRICT LRMC

335

If the set of inputs to the system is as specified in B, then the marginal energy costs are calculated aos shown below: MEC(I) = FCQI)/[PKCAL(I) x FKCAL(I)] where MEC(I) = marginal energy cost per kilowatt-hour for period I FC(1) = fuel cost for marginal plant used in period I (cost per liter) PKCAL(I) ==thermal conversion factor for marginal plant used in period I (kilowatt-hours per kilocalorie) FKCAL(I) = calorific value of fuel used for marginal plant in period I (kilocalories per liter). The marginal energy cost at peak and off-peak periods at different voltage levels is then obtained by adjusting the marginal energy cost by appropriate losses: EGV(I, 1) = MEC(l) x CPMLF(1) EGV(I, 2) = MEC(2)
X

CAMLF(I)

where EGV(I, 1) = marginal energy cost at peak period at voltage level I EGV(I, 2) = marginal energy cost at off-peak period at voltage level

Calculation of Marginal Generation Capacity Cost

There are two options, A and B, depending on the data inputs to the system. A. MCVG = 1, that is, marginal generation capacity costs of each of several types of peaking plant are known. B. MCVG i 1, that is, the installed capacity and the investment stream for several types of peaking plant are known. If the set of inputs specified in B are used, then the marginal generation capacity cost is calculated as follows: DICOST (1, J) SICOST (I) =
where K = BYG(I)

IIC'OST (1, J)/(1 + DR)VJ+' -G DICOST (I, J)

- BYEAR - 1 I = 1, NPP BYG(I) = beginning year of investment stream for plant I, BYEAR = the year of study DICOST (1, J) = present value referred to the year of study, of investment in year J for plant I IICOST (I, J) = incremental investment in border price in year J for plant I

336

APPENDIXES

HCAP(I, J)
where K = BYC(I)

ICAP(I, J)/(1 + DR)Y+-

-BYEAR

SICAP(I)

,,2'Y') IICAP(I, J)

CK(I) = NYG) (DICOST(I, J)ISICAP(I))

PPCK(I)
=

CK(I) x OMX(I)
jVP

CKK = ,
where NYG(I) NYC(I) ICAP(I, NPP =

PPCK(I)

= number of years of incremental investment data for plant I = number of years of installed capacity data for plant I J) = installed capacity by plant I in year J number of peaking plants.

The marginal generation capacity cost at supply voltage levels is then given by: CCK(I) = E,V,P PPCK(I) x [I + CM x (PMLF(I))]ICF
ACK(I) = CCK(I) x AF

ATOMG(I) = OMG(I) x CCK(I) CGBBY

CGBB CVG(I)
= =

= EjV

[ACK(I) + ATOMG(J)]

CGBBYI12 CGBB x CPMLF (I 1)

where CGBB

marginal generation capacity cost at the system busbar per month CVG(I) = marginal generation capacity cost per month at voltage level I CCK(J) = capital cost for plant I
=

CM

capacity margin

AF = annuitizing factor for given lifetime of plant and discount rate CPMLF = cumulative peak loss multiplier OMG(I) = yearly O&M plus other costs as a percentage of capital cost for plant I. Shadow Pricing The marginal costs are calculated in constant border prices to correct for readily quantifiable market distortions. The cost components are converted to border prices by using appropriate conversion factors. In cases where cost

PROGRAM

FOR ESTIMATING

STRICT

LRMC

337

components are not known accurately, composite weighted conversion factors are used. The standard conversion factor is used to convert the marginal cost in border prices back to a domestically priced tariff. In other words, the power tariffs are expressed in tenns of domestic prices.

Listing of the
c c c
c
-

STLRMC

Program
.

C C C C C
C C

-_

STLRMC IS A FORTRAN PROGRAM FOR ESTIMATING THE ELECTRICITY TARIFF BASED ON STRICT LONG RUN MARGINAL COST AT DIFFERENT SUPPLY VOLTAGE LEVELS

- -

--

---

- ----- - -

--

- -

--

----

---

--

--

----

- --- - - -

--

-- - -

- - - - - -

- - - - -

Co

DIMENSION VL(10),ALF(10),PLF(10),SLF(10,20),PPLF(10,20),OMT(10), 1CUVL(1O.20).TINV(10.20),FU(5).FP(5),TL(i0),CFT(10),CFC(10), 2SICOST(20),OICOST(10,20),M(9,9),CTV(10),CCTV(10),MCTV(10),CVG(10), 3MEC(iO),OMX(10),ICAP(9,20),ICOST(10,20),CK(1O).PCK(10).AFF(10), 4AMLF(10),PMLF(1O),CAMLF(10),CPMLF(10),EGV(10,5),SM(9,9),MCC(10), 5IICOST(9,20),NYG(9),CPAVL(9,20).A(15,20),LOSSVL(9,20),SPAVL(9,20) COMMON CURNR,LAGT(10),BYEAR,DYEAR COMMON PFCT(10).PLMT(10),PULT(l0),PFCG(10),PLMG(10),PULG(1O), 1PAVL(9,20),PPCK(10),BYG(10),MCVG(10).TOTALG(10),PL(10),OMG(i0), 2CCK(10),ACK(10),ATOMG(10),BTINV(10,20) 3,IICAP(9.20),SICAP(9),NYC(9),BYC(9) COMMON FCTINV(9,20),LMTINV(9,20),LOTINV(9,20),FCCOST(9,20), lLMCOST(9,20).LOCOST(9.20),PKCAL(9),FKCAL(9),FC(9) !NTEGERt4 TITLE(12) REAL*S CURNR

200 202 203 204 206 207 208

209 210 211 212 213 214 215 216

217 218 219 220 221

INTEGER NEC,NVL,NP,NYG,NYTD,PL.TL,LAGT,BYEAR,BYG,TOTALT,TOTALG i,BYC,DYEAR REAL ICAP.ICOST,MF,DR,OMG,OMT.CM.CFC,CFT,MEC,LMCF 1,MCTV,SCF,M,SM,MCC,LF,IICAP,SICAP 2,LOSSVL,LMCOST,LOCOST,LMTINV,LOTINV FORMAT(43X,' MARGINAL ENERGY COST ') VOLTAGE LEVEL ',10X,' PEAK PERIOD ',iOX FORMAT('0',15X,' 1,' OFF-PEAK PERIOD ') FORMAT(20X,' KV ',18X,A6,'/KWH',13X,A6.'/KWH') FORMAT(20X,A4,20X,F6.3,17X,F6.3) COST ' FORMAT(30X,' MARGINAL GENERATION TRANSMISSION&DISTRIBUTION FORMAT(45X.' BY VOLTAGE LEVELS ') GENERATION COST',T55, LEVEL',5X,'MARGINAL FORMAT(//,5X,'VOLTAGE I'MARGINAL TRANSMISSION & OISTRIBUTION COST'.T105, 2'MARGINAL CAPACITY COST') FORMAT(11X,'KV',16X,A6,'/KW/MONTH'.T70,A6,'/KW/MONTH',T108,A6. 1'/KW/MONTH') FORMAT(/,IOX,A4,15X,FiO.3,5X,4(2X,F1O.3),TiO8,FIO.3) FROM',T88,'TOTAL',1112,'TOTAL') FORMAT(/.T60,'CONTRIBUTION FORMAT(T54,3(A4,8X).TB9,'T&D',T1l,'GEN+T&D') FORMAT(A6) AND LOSS FORECAST'/T46, FORMAT('I',T45,'DEMAND I'AT TIME OF SYSTEM PEAK') FORMAT(/////,T12,'GFNERATION',T25,'STATION',T36,8(A6,4X)) 6X, FORMAT(3X,'YEAR',T15,'PEAK',T27,'USE',T35,2('LOSS',6X,'PEAK' I'CONS'.6X),'LOSS',6X,'PEAK',/T16, 2'MW',T27,'MW',T35,4(X,'MW',8X,'MW',7X),///) FORMAT(/,3X,14.T10,F1O.2,T22,F8.2,T31,9(F9.2,X),F9.2) COST',T65, COST',T45,'CAPACITY FORMAT(//,TIO,'UNIT',T25,'CAPACITY 1'WEIGHTED CAPACITY') FORMAT(T25,'IN MARKET PRICE',T45,'IN BORDER PRICE',T65, 1'COST IN BORDER PRICE') FORMAT(/,TIO,'PLANT'I2,T25,FIO.2,T45,FiO.2.T66,FIO.2) AND PEAK LOSS FACTORS FOR ENERGY/POWER' FORMAT('1',//,T30,'AVERAGE OF INCOMING)' I/T42,'(FRACTION USE',T48,4(A4,16X)) 1/////T1O,'YEAR',T22,'STATION

222 223 224

FORMAT(T20,4('AVERAGE',3X.'PEAK',6X)) FORMAT(/,T1O,I4,T20,10(F5.3,5X)) FORMAT('1',///,T40,'TRAMSMISSION CAPACITY COSTS',X,'(',A4, 1' COST)') 225 FORMAT('1',///,T40,'DISTRIBUTION CAPACITY COSTS',X,'(',A4, 1' COST)') 226 FORMAT(///,Ti5,'YEAR',T25,'FOREIGN',T45,'LOCAL MATERIAL',T65,
1'LABOR&OTHERS',T80,'TOTAL INVESTMENT'//T22,A6,'(MILL)',T46,A6,

227 228 229 230

231 232 233 300 301 302 303 304 305 306 350 400 401

2'(MILL)',T65,A6,'(MILL)',T82,A6,'(MILL)'//) FORMAT(/,TI5,I4,T21,FiO.2,T45.FlO.2,T64,F1O.2,T81,F1O.2) FORMAT(///,TIO,'AVERAGE',T20,i0(F5.3,5X)) FORMAT('1',///.T40,'GENERATION CAPACITY COST'/T48,'PLANT',X,I2/, 1T40,'(CAPACITY=',F8.2,' MW)',//) FORMAT(T10.'YEAR',T20,'FOREIGN',T35,'LOCAL MATERIAL',T55, 1'LABOR&OTHERS',T75,'TOTAL INVESTMENT',T95,'TOTAL INVESTMENT', 2T115,'DISCOUNTED'/T75,'DOMESTIC PRICES',T95,'BORDER PRICES',T115, 3'BORDER PRICES',//T1B,A6,'(MILL)',T36,A6,'(MILL)',T55,A6,'(MILL)', 4T77,A6,'(MILL)',T95,A6,'(MILL)',T115,A6,'(MILL)') FORMAT(/,TIO,I4,T17,FIO.2,T34,FIO.2,T52,FlO.2,T75,FiO.2,T93, 1F1O.2,T115,F1O.2) FORMAT(//,T1O.'TOTAL DISCOUNTED INVESTMENT',T115,FIO.2) FORMAT(//,T10.'WEIGHTED AVERAGE'.T66,F10.2) FORMAT('1',' ') FORMAT('0'.30X,' CAPACITY COST OF GENERATION AT BUSBAR ') FORMAT(/////,' GENERATION CAPITAL COST ',25X,F15.3,2X,A6,'/KW ') FORMAT('O',' CAPITAL COST PER YEAR ',27X,F15.3,2X,A6.'/KW/YEAR ') FORMAT('O'.' O&M '.45X,F15.3.2X,A6,'/KW/YEAR ') FORMAT('O',' TOTAL CAPITAL COST PER YEAR ',21X,F15.3,2X, 1A6.'/KW/YEAR ') FORMAT('O',' CAPACITY COST PER MONTH ',25X.F15.3,2X,A6,'/KW/MMN ') FORMAT(20X,' STRICT LONG RUN MARGINAL COST BASED TARIFF FOR i12A4) FORMAT('O',' FORMAT(12A4)

402

403 404 405 414 415 416

FORMAT(///.T10, 1'NUMBER OF YEARS OF T&D DATA=',14,Tg90.YEAR OF STUDY=',I5,//,T1O 2,'DISCOUNT RATE='.F5.3,T50,'CAPACITY MARGIN=',F5.3,T90, 3'MULTIPLYING FACTOR=',F5.3,//,TIO,'LOCAL MATERIAL CF=',F5.3,T50, 4'LABOR & OTHERS CF=',F5.3,T90,'STANDARD CF=',F5.3) FORMAT(///.T1O,'LIFE OF T&D FACILITIES=',615) FORMAT(//,TIO,'O&M FOR T&D FACILITIES=',6F7.3) FORMAT(//,T10,'LAG FOR T&D=',6I5) FORMAT(T25,'YEAR',T40.'INVESTMENT',T60,'INVESTMENT',T80, I'DISCOUNTED',/T40,A6,'(MILL)',T60,A6,'(MILL)',T80,A6,'(MILL)') FORMAT(/,T25.14,T40.FlO.2,T60,FIO.2,T80,FIO.2) FORMAT(//,TIO,'TOTAL DISCOUNTED INVESTMENT',T80.FIO.2) READ(5,401) (TITLE(I),I=1, 12) AUTHORITY

C TITLE=NAME OF THE ELECTRICITY READ(5,213) CURNR C CURNR=CURRENCY

READ/,NYTD,TOTALT,BYEAR,DYEAR

C C C C C C C C C C C C

NYTD=NUMBER OF YEARS OF TRANSMISSION AND DISTRIBUTION DATA TOTALT=1 T&D INVESTMENT STREAM IS NOT BROKEN DOWN BY LOCAL,FOREIGN ET BYEAR=YEAR OF STUDY DYEAR=BEGINNING YEAR OF DEMAND FORECAST READ/,DR,CM,MF DR=DISCOUNT RATE(IN DECIMALS) CM=CAPACITY MARGIN(IN DECIMALS) MF=MULTIPLYING FACTOR. IT IS THE PERCENTAGE BY WHICH PEAK LOSS EXCEEDS THE AVERAGE LOSS.(IN DECIMALS)
READ/.FECF,LMCF,ULCF,CF

FECF=FOREIGN EXCHANGE CONVERSION FACTOR LMCF=LOCAL MATERIAL CONVERSION FACTOR ULCF=UNSKILLED LABOR CONVERSION FACTOR CF=STANDARD CONVERSION FACTOR TO CONVERT

BORDER

PRICES TO DOMESTIC

C TL=T&D FACILITIES LIFE READ/,(OMT(I),I=1,3)

C OMT=YEARLY O&M AS A PERCENTAGE READ/,(LAGT(I),1=1,3) C LAGT=LAG FOR T&D FACILITIES

C C -- -------- - - -- - - ---------

OF T&D CAPITAL

COST(IN

DECIMALS)

-- -

---

- - -

- -

C
C C

- -

READ SUPPLY VOLTAGE

- - - -------- - --

LEVEL
- - - - - - - - -

- - - -

DO

100

1=1,3

C 100
C C

READ(5,213) VL(I) VL=SUPPLY VOLTAGE LEVEL CONTINUE

--------- -----------------________ ----

C
C

READ LOSS COEFFICIENTS

---------------------

DO 101 I=1.4 READ/,(SLF(I,J),d=1 ,NYTD) C SLF=SYSTEM LOSS COEFFICIENT.(IN DECIMAL) IF (MF.NE.O) GO TO 101 READ/,(PPLF(I,J),J=1,NYTD) C PPLF=PEAK LOSS COEFFICIENT.(IN DECIMAL) 101 CONTINUE
C C ---------_________ _____ _____-

C
C C

-------

READ T&D CAPACITY

--------

DATA

__- _________ __- _________

IF (TOTALT.EO.1) DO 1001 1=1,3

GO TO

1000

READ!,(FCTINV(I,J),J=1,NYTD) READ/.(LMTINV(I,J),J=1,NYTD)

READ/,(LOTINV(I,J),J=1,NYTD)

C FCTINV=FOREIGN COMPONENT OF INVESTMENT C LMTINV=LOCAL MATERIAL C LOTINV=LABOR AND OTHER 1001 CONTINUE DO 1002 I=1,3 DO 1002 J=1,NYTD TINV(I,U)=FCTINV(I,J)+LMTINV(I,J)+LOTINV(I,J) BTINV(I,O)=FCTINV(I,J)*FECF+(LMTINV(I,U)*LMCF)+(LOTINV(I,J)*ULCF) 1002 CONTINUE GO TO 1003 1000 DO 102 1=1,3 READ/,PFCT(I).PLMT(I),PULT(I) C PFCT=PERCENTAGE OF FOREIGN COMPONENT IN T&D COST(IN DECIMAL) C PLMT=PERCENTAGE OF LOCAL MATERIAL IN T&D COST.(IN DECIMALS) C PULT=PERCENTAGE OF UNSKILLED LABOR IN THE T&D COST.(IN DECIMALS) READ/,(TINV(I,J),J=i.NYTD) C TINV=INCREMENTAL YEARLY INVESTMENT FOR T&D FACILITIES FOR A GIVEN C SUPPLY VOLTAGE.(COST IN MILLIONS) 102 CONTINUE 1003 DO 103 I=1,4 READ/,(CPAVL(I,J),J=1,NYTD) C CPAVL(1.J)= GENERATION PEAK IN YEAR U C CPAVL(I,J)= PEAK DEMAND AT VOLTAGE (I-1) IN YEAR J 103 CONTINUE
C C -----------__ ___ ____ ___ ____ ___ _ -- ______________

C
C C

-----

READ GENERATION
---

CAPACITY

COST

_____ _______ ______ __ __-- ______________

READ/,NPP C NPP=NUMBER OF PEAK PERIOD MARGINAL PLANTS IF (NPP.EO.l) GO TO 104 READ/,(OMX(I),I=i,NPP) C OMX=OPERATING MIX OF THE PEAKING PLANTS 104 READ/,(MCVG(I),I=1,NPP)

C IF MCVG=1 MARGINAL GENERATION CAPITAL COST FOR THE MARGINAL PLANTS ARE C AVAILABLE DO 105 I=1,NPP IF (MCVG(I).EQ.1) GO TO 106 READ/.NYG(I),NYC(I),BYG(I),BYC(I),TOTALG(I),PL(I),OMG(I) C NYG=NUMBER OF YEARS OF INCREMENTAL ANNUAL INVESTMENT IN THE MARGINAL C NYC=NUMBER OF YEAR OF INCREMENTAL CAPACITY INSTALLED(MW) C BYG=FIRST YEAR INVESTMENT BEGINSFOR THE PLANT C BYC=FIRST YEAR OF INSTALLED CAPACITY C TOTALG=1 INVESTMENT STREAM FOR GENERATION IS NOT BROKEN DOWN C PL=PLANT LIFE C OMG=YEARLY O&M COST AS PERCENTAGE OF GENERATION CAPITAL COST.(IN DECIM READ/,(ICAP(I.,),U1.NYC(I)) C ICAP=CAPACITY OF THE MARGINAL PEAKING PLANT.(MW) IF (TOTALG(I).NE.1) GO TO 1005 READ/.PFCG(I),PLMG(I),PULG(I) C PFCG=PERCENTAGE OF FOREIGN COMPONENT IN GENERATION COST(IN DECIMALS) C PLMG=PERCENTAGE OF LOCAL MATERIAL FOR GENERATION COST.(IN DECIMALS) C PULG=PERCENTAGE OF UNSKILLED LABOR FOR GENERATION COST.(IN DECIMALS) READ/,(ICOST(I, ),J=i,NYG(I)) C ICOST=YEARLY INCREMENTAL INVESTMENT IN THE MARGINAL PLANT.(IN MILLIONS GO TO 105 1005 READ/,(FCCOST(I.J),J=l.NYG(I)) C FCOST=FOREIGN COMPONENT FOR GENERATION READ/,(LMCOST(I,J),J=l,NYG(I)) C LMCOST=LOCAL MATERIAL READ/,(LOCOST(I.J),J=1,NYG(I)) C LOCOST=LABOR AND OTHERS GO TO 105 106 READ/.PFCG(I),PLMG(I),PULG(I) C PFCG=PERCENTAGE OF FOREIGN COMPONENT IN GENERATION COST(IN DECIMALS) C PLMG=PERCENTAGE OF LOCAL MATERIAL FOR GENERATION COST.(IN DECIMALS) C PULG=PERCENTAGE OF UNSKILLED LABOR FOR GENERATION COST.(IN DECIMALS) READ/,PCK(I).PL(I),OMG(I)

C PCK=MARGINAL GENERATION CAPITAL COST FOR THE MARGINAL PLANT.(COST/KW) 105 CONTINUE DO 1006 I=1,NPP DO 1006 J=1,NYG(I) IF (MCVG(I).EO.l) GO TO 1006 IF (TOTALG(I).EQ.1) GO TO 1006 ICOST(I,J)=FCCOST(I, d)+LMCOST(I,J)+LOCOST(I,J) IICOST(I,U)=FCCOST(I,J)*FECF+LMCOST(I,J)*LMCF+LOCOST(I,J)*ULCF 1006 CONTINUE
C C ------ ___________ _-- _______________

C
C C

READ MARGINAL
----------

ENERGY

COST DATA

_________________ _________________

151 READ/,NEC C IF NEC=1 MARGINAL ENERGY COST ARE AVAILABLE.(COST/KWH) DO 108 I=1,2 IF (NEC.EO.l) GO TO 107 READ/,FKCAL(I),PKCAL(I),FC(I) C FKCAL= CALORIFIC VALUE OF FUEL(KCAL/LITER) C PKCAL= THERMAL CONVERSION FACTOR OF PLANT(KWH/KCAL) C FC= FUEL PRICE(COST/LITER) GO TO 108 107 READ/,MEC(I) C MEC=MARGINAL ENERGY COST OF THE MARGINAL PLANT IN THE DISCRETE 108 CONTINUE
C C ------------ ---------------- --

PERIODS

C
C C

WRITE INPUT DATA

--------- -- --- - --------------- - --

WRITE(6,402) WRITE(6,403) WRITE(6,404) WRITE(6.405)

NYTD,BYEAR,DR,CM,MF,LMCF,ULCF.CF (TL(I) ,=1 .3) (OMT(I),I=1,3) (LAGT(I),I=1,3)

c
c ----------------

C
C C

--

CALCULATION
----- -----------

OF CONVERSION
---------- ---

FACTORS
--------

DO

109

I=1,3

109

110
C C

IF (TOTALT.NE.1) GO TO 109 CFT(I)=(PLMT(I)*LMCF+PULT(I)*ULCF+PFCT(I)*FECF) CONTINUE DO 110 I=1,NPP GO TO 110 IF (TOTALG(I).NE.1.ANO.MCVG(I).NE.1) CFC(I)=(PULG(I)*ULCF+PLMG(I)*LMCF+PFCG(I)*FECF) CONTINUE
----------- - - - -- -- ---------- - - -- - -- -- - -

C
C >

CALCULATION
---------------

OF LOSS COEFFICIENTS
---- - - - - - - --- - - - --- - --- _---- ---

~~~~~C

III 112

DO 113 1=1,4 DO 112 J=1,NYTD IF (MF.EQ.O) GO TO 111 PPLF(I,J)=SLF(I,J)*MF ALF(I)=ALF(I)+SLF(I,d) PLF(I)=PLF(I)+PPLF(I,J) CONTINUE ALF(I)=(ALF(I)/NYTD) PLF(I)=(PLF(I)/NYTD)
AMLF(I)=t/(1-ALF(I))

113

PMLF(I)=1/(1-PLF(I)) CONTINUE CAMLF(1)=AMLF(1) CPMLF(I)=PMLF(I) DO 114 1=2,4 CAMLF(I)=CAMLF(I-1)*AMLF(I) CPMLF(I)=CPMLF(1-1)*PMLF(I)

114
C C

CONTINUE
- - -- --- ---------- - - - -- - - - - - - - - - __---------------------------

C
C C

-------------

WRITE AVERAGE
- ----------- --

AND PEAK LOSS FACTORS

- ------------------------------------

WRITE(6,221)
WRITE(6,222)

(VL(I).I=1,3)

K=DYEAR DO 1030 0=1,NYTD WRITE(6,223) K,((SLF(I,J),PPLF(I,J)),I=1,4) K=K+1

1.030-CONTINUE

WRITE(6,228)
C C -------- - - --- - - -

((ALF(I),PLF(I)),I=1,4)
- - ------------- - - - - - - ----------- - - - - -

C
C C C

CALCULATION
-------_ -------

OF T&D CAPACITY

COST

__________________ ____-_______ ________ -__ _ _ _ _ _ __-_________________ _ __ __-

CALL AFTL(DR,TL,AFF) C
C
-

CALCULATION
----

OF DEMAND
__---_--

FORECAST
______________

122

DO 122 J=1,NYTD ,d) LOSSVL(1,J)=CPAVL(1.J)*PPLF( LOSSVL(2,J)=CPAVL(1,d)*PPLF(2,J) SPAVL(2,J)=CPAVL(l,J)-LOSSVL(1,J)-LOSSVL(2,J) PAVL2,J)=SPAVL(2,J)-CPAVL(2,J) 0.J) A(1,J)=CPAVL( A(2,d)=LOSSVL( I,J) A(3,J)=LOSSVL(2Jd) A(4,D)=SPAVL(2,J) A(5,J)=CPAVL(2,J) CONTINUE

DO 123 J=I,NYTD L=3 DO 123 I=3,4 LOSSVL(I,J)=PAVL(I-1,J)*PPLF(I.J) SPAVL(I,U)=PAVL(I-1,U)-LOSSVL(I.U) PAVL(I,J)=SPAVL(I.J)-CPAVL(I,J) A(I+L,U)=LOSSVL(I,,J) A(I+L+1,U)=SPAVL(I,U) A(I+L+2,J)=CPAVL(I,J) L=(I+L+2)-3 123 CONTINUE WRITE(6,214) ((VL(I),VL(I),VL(I)),I=1.2),VL(3),VL(3) WRITE(6,215) WRITE(6,216) K=DYEAR DO 160 J=1,NYTD WRITE(6,217) K,(A(I,J),I=1,10) K=K+1 160 CONTINUE DO 124 I=2,4 DO 124 U=2,NYTD CUVL(I-t,J)=SPAVL(I,J)-SPAVL(I,U-1) 124 CONTINUE IF (TOTALT.NE.1) GO TO 1011 DO 1010 I=1,3 DO 1010 U=1,NYTD BTINV(I,J)=TINV(I,U)*CFT(I) 1010 CONTINUE GO TO 1015 1011 DO 1012 1=1,3 IF (I.NE.1) GO TO 1013 WRITE(6,224) VL(I) GO TO 1014 1013 WRITE(6,225) VL(I)

1014 WRITE(6,226) CURNR.CURNR.CURNR,CURNR K=DYEAR DO 1012 J=1,NYTD WRITE(6.227) K,FCTINV(I,J),LMTINV(I,U),LOTINV(I,0),TINV(I,J) K=K+1 1012 CONTINUE 1015 CALL AICTD(TINV,BTINV,CUVL,VL.AFF,DR,OMT,NYTD,PMLF,CF,CCTV) 177 DO 180 I=1,2 00 181 J=(I+I),3 M(I,J)=CCTV(I)*(CPMLF(J+1)/CPMLF(I+1)) SM(J,I)=M(I,j) 181 CONTINUE i80 CONTINUE 00 179 1=1,3 SM(I,I)=CCTV(I) MCTV(I)=O.O 179 CONTINUE DO 182 J=1,3 DO 183 I=1,3 MCTV(I)=MCTV(I)+SM(I,J) 183 CONTINUE 182 CONTINUE
C
C ------_-----__--------------------- -----------------------

C
C C

CALCULATION
--- ----

OF MARGINAL
-------_----_-------_----_---------

ENERGY COST
- -- - -- ----- -- --

185 310

315

IF (NEC.EQ.1) GO TO 310 DO 185 1=1,2 MEC(I)=FC(I)/(FKCAL(I)*PKCAL(l)) CONTINUE DO 315 I=1,3 EGV(I 1)=MEC(i)*CPMLF(I+1) EGV(I,2)=MEC(2)*CAMLF(I+1) CONTINUE

c c

-----

----------------------

C
C C

CALCULATION

OF MARGINAL
-----

GENERATION

CAPACITY

COST

____ ______ ______ ____ _ _______________-_- ---

DO 115 I=1,NPP IF (MCVG(I).EQ.1) GO TO 152 K=(BYG(I)-BYEAR+1) 00 116 J=1,NYG(I) IF (TOTALG(I).NE.1) GO TO 1020 IICOST(I,J)=ICOST(I,J)*CFC(I) 1020 OICOST(I.J)=IICOST(I,J)/((1+DR)**(J+K)) SICOST(I)=SICOST(I)+DICOST(I.J) 116 CONTINUE GO TO 115 152 CK(I)=PCK(I)*CFC(I) 115 CONTINUE DO 130 I=1,NPP IF (MCVG(I).EO.1) GO TO 130 K=(BYC(I)-BYEAR+1) DO 130 J=1,NYC(I) IICAP(I,J)=ICAP(I,J)/((1+DR)**(J+K)) SICAP(I)=SICAP(I)+IICAP(I,J) 130 CONTINUE 150 CONTINUE DO 120 I=1,NPP IF (MCVG(I).EQ.1) GO TO 154 WRITE(6,229) I,SICAP(I) K=BYG(I) IF (TOTALG(I).EQ.1) GO TO 118 WRITE(6,230) CURNR,CURNR,CURNR,CURNR,CURNR,CURNR DO 117 J=1,NYG(I) WRITE(6,231) K.FCCOST(I,J),LMCOST(I,J).LOCOST(I,J),ICOST(I.J) 1.IICOST(I,J),DICOST(I,J) K=(BYG(I)+J)

117 118

250

251 253

254 154

155 156 252

120

121

CONTINUE GO TO 251 WRITE(6,414) CURNR,CURNP,CURNR DO 250 J=1,NYG(l) WRITE(6,415) K,ICOST(I,J).IICOST(I,J),DICOST(I,U) K=(BYG(I)+J) CONTINUE WRITE(6,416) SICOST(I) GO TO 253 WRITE(6,232) SICOST(I) CK(I)=(SICOST(I)/SICAP(I))*1000 IF (TOTALG(I).EQIl) GO TO 254 PCK(I)=CK(I)/CF GO TO 154 PCK(I)=(SICOST(I)/(SICAP(I)*CFC(I)))*1000 IF (NPP.NE.1) GO TO 1,55 PPCK(I)=CK(I) GO TO 156 PPCK(I)=CK(I)*OMX(I) CKK=CKK+PPCK(I) CCK(I)=PPCK(I)*(1+CM)*CPMLF(1)/CF NPL=PL(I) CALL AFAC(DR,NPL,AF) ACK(I)=CCK(I)*AF ATOMG(I)=OMG(I)*CCK(I) CGBBY=CGBBY+(ACK(I)+ATOMG(I)) CCCK=CCK(I)+CCCK ACCK=ACCK+ACK(I) TOMG=TOMG+ATOMG(I) CONTINUE CGBB=(CGBBY/12) DO 121 I=1,3 CVG(I)=CGBB*CPMLF(I+1) CONTINUE DO 178 I=1,3 MCC(I)=CVG(I)+MCTV(I)

178

CONTINUE
WRITE(6,300)

161 165

WRITE(6,350) (TITLE(I),I=1,12) WRITE(6,301) WRITE(6,400) IF (NPP.EQ.1) GO TO 165 WRITE(6,400) WRITE(6,218) WRITE(6,219) WRITE(6,400) DO 161 I=1,NPP WRITE(6,220) I,PCK(I),CK(I),PPCK(I) CONTINUE WRITE(6,233) CKK WRITE(6,400) WRITE(6,400) WRITE(6,302) CCCK,CURNR WRITE(6,303) ACCK,CURNR WRITE(6,304) TOMG,CURNR WRITE(6.305) CGBBY,CURNR WRITE(6,306) CGBB,CURNR WRITE(6,300) WRITE(6,350) (TITLE(I).I=1,12) WRITE(6.207) WRITE(6,400)
WRITE(6,400)

WRITE(6,200) WRITE(6,400) WRITE(6,400) WRITE (6,202) WRITE(6,203) CURNR,CURNR WRITE(6,400) DO 138 I=1,3 WRITE(6,204) VL(I),(EGV(I,K),K=i.2)

138

WRITE(6,400) CONTINUE WRITE(6,400) WRITE(6,400) WRITE(6.206) WRITE(6,207)

WRITE(6,400)

WRITE(6,208) WRITE(6,209) WRITE(6,211) WRITE(6,212)

WRITE(6,400)

CURNR,CURNR,CURNR (VL(I),1=1,3)

DO 139 1=1,3 WRITE(6,210) VL(I),CVG(I),(SM(I,J),J=1,3),MCTV(I),MCC(I)

WRITE(6f,400)

139

CONTINUE WRITE(6.300) STOP END

C_ C -

---

------ -

- -

-- _ - -

_-- - - -

_ _ - - -

_ _ - - -

_-

- - --- -

-- - - - -

C C
C C_

SUBROUTINE GENERATION
----

FOR CALCULATING PLANTS

--------

THE ANNUTIZING
--------------

FACTOR
- ---- ---

FOR
- - ---- --

--------------

---

SUBROUTINE AFAC(RD,LP,FA) FA-O DO 190 I=1.LP AR=F/((1+RD)**I)

FA-FA+AR

190

CONTINUE
FA=1/FA RETURN

END

C C

---

- - - - - _______ __________ ______

C
C C C

SUBROUTINE
T&D

FOR CALCULATING
-------------

THE ANNUTIZING

FACTOR FOR
- ____________________ ___

FACILITIES

SUBROUTINE AFTL(RDD,LPP,FAA) DIMENSION FAA(5),LPP(5) 0O 195 1=1,3 D0 196 J=1,LPP(I) ARR=1/((1+RDD)**J)

FAA(I)=FAA(I)+ARR

196
195

CONTINUE
FAA(I)=1/FAA(I) CONTINUE RETURN END

c c

------------------------------

C C
C C

SUBROUTINE TO CALCULATE THE MARGINAL T&D CAPACITY USING THE AVERAGE INCREMENTAL COST METHOD
------

COST

-- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - _ _ _ _ _ _ _ _ ___ _ _ _ _ _ _ _-- _ _ _ _ _- ___

515 516
U'

517 518 519 520 530 600 601 602

SUBROUTINE AICTD(ATINV,AATINV,ACUVL,VLA,AAF,DDR,AOMT,NYA, ISPMLF,SSCF,AMCTV) COMMON CURNR,LAGT(10),BYEAR,DYEAR REAL*8 CURNR DIMENSION ATINV(10,20),ACUVL(10,20),AMCTV(10),ADTINV(10,20), IADCUVL(10,20),AVL(10),ACT(10),VLA(10),AATINV(10,20), 2AOMT(10).AAF(10).SCFT(10),SPMLF(10) FORMAT(/,T20,14,T26,FIO.2,T38,FIO.2,T53,FIO.2,T68,F1O.2,T83,F1O.2) TOTAL',T38,F10.2,2X,'(',14,'-'.I4,')',T83, FORMAT(///,T20,'LAGGED 1F10.2,2X,'(',I4,'-',I4,')') INCREMENTAL T&D CAPACITY COST') FORMAT('1',35X,'AVERAGE FORMAT(///,T20,'YEAR',T30,'PEAK',T40,'DISCOUNTED',T55, 1'INVESTMENT',T70,'INVESTMENT',T85,'DISCOUNTED') PRICE', PRICE',T69,'BORDER FORMAT(T43,'PEAK',T53,'DOMESTIC 1T85,'INVESTMENT') LEVEL',X,A4) FORMAT(//,T45,'VOLTAGE FORMAT(T31,'MW',T44,'MW',T54,A6,'(MILL)',T69,A6,'(MILL)'. 1T84,A6.'(MILL)') FORMAT('O') INCREMENTAL TRANSMISSION CAPITAL COST=', FORMAT(/,20X,'AVERAGE 1F12.3) INCREMENTAL DISTRIBUTION CAPITAL COST=', FORMAT(/,20X,'AVERAGE IF12.3) K=DYEAR-BYEAR-1 DO 501 I=lj3 DO 502 J=1,NYA ADTINV(I,J)=(AATINV(I,J)/((l+DDR)**(d+K))) ADCUVL(I,J)=((ACUJVL(I.U)/((1+DDR)**(d+K)))

502 501

503

504

506

507 505

CONTINUE CONTINUE DO 503 I=1,3 DO 503 J=(2+LAGT(I)).NYA AVL(I)=AVL(I)+ADCUVL(I,J) CONTINUE DO 504 I=1,3 DO 504 J=1,(NYA-LAGT(I)) ACT(I)=ACT(I)+ADTINV(I,J) CONTINUE DO 505 1=1,3 WRITE(6,517) WRITE(6,520) VLA(I) WRITE(6,518) WRITE(6,519) WRITE(6,530) CURNR,CURNR,CURNR WRITE(6,600) K=DYEAR D0 506 J=1,NYA WRITE(6,515) K,ACUVL(I,J),ADCUVL(I,d),ATINV(I,J),AATINV(I,J), iADTINV(I.J) K=K+i CONTlNUE ACK=ACT(I)/AVL(I)*1000 TOMT=(ACT(I)/AVL(I))*AOMT(I)*1000 AMCTV(I)=((AAF(I)+AOMT(I))*ACK)/(12*SSCF) LL=DYEAR+(NYA-LAGT(I)-1) KK=LAGT(I)+DYEAR+1 NN=DYEAR+NYA-1 MM=DYEAR+1 WRITE(6,516) AVL(I),MM,LL,ACT(I).KK,NN WRITE(6,600) IF(I.NE.1) GO TO 507 WRITE(6,601) ACK GO TO 505 WRITE(6,602) ACK CONTINUE RETURN END

Sample Outputs from the

STLRMC

Program for the WAPDA System in Pakistan

Some of the computer-generated numbers may differ by a few percentage points from the figures given in chapter 9 because of the effects of rounding off.
NUMBER OF YEARS OF T&D DATA= DISCOUNT LOCAL RATE=0.120 CF=0.850 12 CAPACITY MARGIN=0.300 YEAR OF STUDY= MULTIPLYING 1979

FACTOR=0.O0O

MATERIAL

LABOR & OTHERS CF=0.750

STANDARD CF=0.850

LIFE

OF T&D FACILITIESt.

40

25

25

O&M

FOR T&D FACILITIES=

0.030

0.070

0.070

LAG FOR T&D=

AVERAGE

AND PEAK LOSS (FRACTION

FACTORS FOR ENERGY/POWER OF INCOMING)

YEAR

STATION AVERAGE 0.020 0.025 0.022 0.022 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020

USE PEAK 0.020 0.025 0.023 0.022 0.020 0.020 0.020 0.020 0.020 0.020 0.020 0.020

HV AVERAGE 0.120 0.105 0.100 0.090 0.080 0.065 0.065 0.065 0.065 0.065 0.065 0.065 PEAK 0.160 0.140 0.130 0.120 0.100 0.080 0.080 0.080 0.080 0.080 0.080 0.080 AVERAGE 0.080 0.080 0.073 0.071 0.067 0.067 0.067 0.067 0.067 0.067 0.067 0.067

MV PEAK 0.104 0.104 0.095 0 092 0.087 0.087 0.087 0.087 0.087 0.087 0.087 0.087 AVERAGE 0.254 0.254 0.240 0.227 0.218 0.209 0.209 0.209 0.209 0.209 0.209 0.209

LV PEAK 0.295 0.295 0.279 0.265 0.253 0.242 0.242 0.242 0.242 0.242 0.242 0.242

1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990

AVERAGE

0.021

0.021

0.079

0.101

0.070

0.091

0.221

0.257

DEMAND AND LOSS FORECAST AT TIME OF SYSTEM PEAK

YEAR

GENERATION PEAK MW

STATION USE MW

HV LOSS MW

HV PEAK MW

HV CONS MW

MV LOSS MW

Mv PEAK MW

MV CONS MW

LV LOSS MW

LV PEAK MW

1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990

1972.00 2149.00 2342.00 2553.00 2783.00 3033.00 3306.00 3538.00 3821.00 4127.00 4457.00 4813.00

39.44 53.73 53.87 56.17 55.66 60.66 66.12 70.76 76.42 82.54 89.14 96.26

315.52 300.86 304.46 306.36 278.30 242.64 264.48 283.04 305.68 330.16 356.56 385.04

1617.04 1794.42 1983.67 2190.47 2449.04 2729.70 2975.40 3184.20 3438.90 3714.30 4011.30 4331.70

221.00 246.00 276.00 310.00 347.00 392.00 413.00 442.00 478.00 516.00 557.00 602.00

145.19 161.04 162.23 173.00 182.88 203.38 222.93 238.57 257.60 278.25 300.52 324.48

1250.85 1387.38 1545.44 1707.47 1919.16 2134.32 2339.47 2503.63 2703.30 2920.05 3153.78 3405.22

177.00 197.00 221.00 248.00 278.00 313.00 331.00 354.00 382.00 413.00 446.00 481.00

316.79 351.16 369.52 386.76 415.21 440.76 486.05 520.21 561.76 606.71 655.28 707.66

757.07 839.22 954.92 1072.71 1225.95 1380.56 1522.42 1629.42 1759.55 1900.34 2052.49 2216.56

TRAMSMISSION

CAPACITY

COSTS

(HV

COST)

YEAR

FOREIGN RUPEES(MILL)

LOCAL

MATERIAL

LABOR&OTHERS RUPEES(MILL)

TOTAL

INVESTMENT

RUPEES(MILL)

RUPEES(MILL)

1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990

0.00 489.60 490.40 G34.80 680.00 652.80 668.80 679.20 691.60 738.80 696.00 667.60

0.00 673.20 674.30 872.85 935.00 897.60 919.60 933.90 950.95 1015.85 957.00 917.95

0.00 61.20 61.30 79.35 79.35 81.60 83.60 84.90 86.45 92.35 87.00 83.45

0.00 1224.00 1226.00 1587.00 1694.3S 1632.CO 1672.00 1698.00 1729.00 1847.00 1740.00 1669.00

DISTRIBUTION

CAPACITY

COSTS

(MV

COST)

YEAR

FOREIGN RUPEES(MILL)

LOCAL

MATERIAL

LABORSOTHERS RUPEES(MILL)

TOTAL

INVESTMENT

RUPEES(MILL)

RUPEES(MILL)

1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990

0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

0.00 197.80 252.10 263.10 297.20 303.60 308.20 318.70 320.20 325.70 336.70 338.60

0.00 17.20 21.90 22.90 25.80 26.40 26.80 27.30 27.80 28.30 29.30 29.40

0.00 215.00 274.00 286.00 323.00 330.00 335.00 346.00 348.00 354.00 366.00 368.00

DISTRIBUTION

CAPACITY

COSTS

(LV

COST)

YEAR

FOREIGN RUPEES(MILL)

LOCAL

MATERIAL

LABOR&OTHERS RUPEES(MILL)

TOTAL

INVESTMENT

RUPEES(MILL)

RUPEES(MILL)

1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990

0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

0.00 296.24 379.00 393.76 446.20 453.56 462.76 470.12 480.24 488.52 505.08 507.84

0.00 25.76 33.00 34.24 38.80 39.44 40.24 40.88 41.76 42.48 43.92 44.16

0.00 322.00 412.00 428.00 485.00 493.00 503.00 511.00 522.00 531.00 549.00 552.00

AVERAGE

INCREMENTAL VOLTAGE

T&D CAPACITY HV

COST

LEVEL

YEAR

PEAK MW

DISCOUNTED PEAK MW

INVESTMENT PRICE DOMESTIC RUPFES(MTILL

INVESTMENT BOROER PRICE PUPEES(MILL

DISCOUNTED INVESTMENT RPUPEES(MILL)

1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1q90

0.00 177.37 189.26 206.80 258.57 280.66 245.70 208.80 254.70 275.40 297.00 320.40

0.00 158.37 150.88 147.20 164.32 159.25 124.48 94.45 102.87 99.31 95.63 92.11 1079.62

0.00 1224.00 1226.00 1587.00 1694.35 1632.00 1672.00 1698.00 1729.00 1847.00 1740.00 1669.00 (1980-1988)

0.00 965.74 967.31 1252.14 1337.06 1287.65 1319.21 1339.72 1364.18 1457.28 1372.86 1316.E-4

0.00 862.26 771.14 891.25 849.73 730.65 668.35 606.02 550.97 525.51 442.02 378.56 6455.88 (1982-1990)

IA(G(GrD TOTAL

AVERAGE

INCREMENTAL

TRANSMISSION

CAPITAL

COST=

5979.783

AVERAGE

INCREMENTAL VOLTAGE

T&D MV

CAPACITY

COST

LEVEL

YEAR

PEAK MW

DISCOUNTED PEAK MW

INVESTMENT PRICE DOMESTIC RUPEES(MILL)

INVESTMENT BORDER PRICE RUPEES(MILL)

DISCOUNTED INVESTMENT RUPEES(MILL)

1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 LAGGED

0.00 136.53 158.07 162.03 211.69 215.16 205.15 164.16 199.67 216.75 233.73 251.44 TOTAL

0.00 121.90 126.01 115.33 134.53 122.09 103.94 74.26 80.64 78.16 75.25 72.28 982.49

0.00 215.00 274.00 286.00 323.00 330.00 335.00 346.00 348.00 354.00 366.00 368.00 (1980-1989)

0.00 181.03 230.71 240.81 271.97 277.86 282.07 291.37 293.02 298.07 308.17 309.86

0.00 161.63 183.92 171.40 172.84 157.67 142.91 131.80 118.35 107.49 99.22 89.08 1447.23 (1981-1990)

AVERAGE

INCREMENTAL

DISTRIBUTION

CAPITAL

COST=

1473.019

AVERAGE

INCREMENTAL VOLTAGE

T&D

CAPACITY

COST

LEVEL LV INVESTMENT DOMESTIC PRICE RUPEES(MILL) INVESTMENT BORDER PRICE RUPEES(Mltl_) DISCOUNTED INVESTMENT RUPEES(MILL)

YEAR

PEAK MW

OISCOUNTED PEAK MW

1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 LAGGED

0.00 82.15 115.71 117.79 153.24 154.61 141.86 107.00 130.13 140.80 152.15 164.06 TOTAL

0.00 73.35 92.24 83.84 97.39 87.73 71.87 48.40 52.56 50.77 48.99 47.16 680.95

0.00 322.00 412.00 428.00 485.00 493.00 503.00 511.00 522.00 531.00 549.00 552.00 (1980-1989)

0.00 271.12 346.90 360.38 408.37 415.11 423.53 430.26 439.52 447.10 462.26 464.78

0.00 242.08 276.55 256.51 259.53 235.54 214.57 194.63 177.52 161.23 148.83 133.61 2166.98 (1981-1990)

AVERAGE

INCREMENTAL

DISTRIBUTION

CAPITAL

COST=

3182.299

STRICT

LONG

RUN MARGINAL CAPACITY COST

COST BASED

TARIFF

FOR

WAPDA

OF GENERATION

AT BUSBAR

UNIT

COST CAPACITY IN MARKET PRICE

CAPACITY C05T IN BORDER PRICE

CAPACITY WEIGHTED COST IN BORDER PRICE

0N

PLANT PLANT

1 2

12135.00 4225.00

9465.30 3295.50

5205.92 1482.98

WEIGHTED

AVERAGE

6688.89

GENERATION CAPITAL D6M TOTAL

CAPITAL PER

COST YEAR

10447.728 1332.085 156.716

RUPEES/KW RUPEES/KW/YEAR RUPEES/KW/YEAR RUPEES/KW/YEAR RUPEES/KW/MMN

COST

CAPITAL COST

COST PER

PER YEAR MONTH

1488.801 124.067

CAPACITY

STRICT

LONG

RUN MARGINAL

COST BASED BY VOLTAGE MARGINAL

TARIFF LEVELS

FOR WAPDA

ENERGY

COST OFF-PEAK PERIOD RUPEES/KWH

VOLTAGE KV

LEVEL

PEAK PERIOD RUPEES/KWH

HV

0.295

0.277

MV

0.325

0.298

LV

0.437

0.383

MARGINAL

GENERATION TRANSMISSION&DISTRIBUTION BY VOLTAGE LEVELS MARGINAL

COST

VOLTAGE LEVEL KV

MARGINAL GENERATION COST RUPEES/KW/MONTH

TRANSMISSION & DISTRIBUTION RUPEES/KW/MONTH FROM LV

COST

MARGINAL CAPACITY RUPEES/KW/MONTH TOTAL GEN+T&D

COST

HV

CONTRIBUTION MV

TOTAL T&D

HV

140.915

88.702

0.000

0.000

88.702

229.618

MV

155.008

97.573

28.522

0.000

126.095

281.103

LV

208.555

131.279

38.374

61.618

231.272

439.826

References

The word "processed" describes works that are reproduced from typescript by mimeograph, xerography, or similar means; such works may not be cataloged or commonly available through libraries; or may be subject to restricted circulation. Acton, Jan Paul, Ellen K. Gelbard, James R. Hosek, and Derek J. McKay. "British Industrial Response to the Peak Load Pricing of Electricity." Report No. R-2508DOE/DWP. Santa Monica, Calif.: The Rand Corporation, February 1980. Anderson, D. "Models for Determining Least Cost Investments in Electricity Supply." The Bell Journal of Economics and Management Science, vol. 3 (Spring 1972), pp. 267-301. Aoki, H. "New Method of Long Range or Very Long Range Demand Forecast of Energy, Including Electricity, Viewed from a Worldwide Standpoint." Tokyo, Electric Power Development Company, Ltd., 1974. Processed. Arvanitidis, N. V., and J. Rosing. "Composite Representation of a Multireservoir Hydroelectric Power System." IEEE Transactions on Power Apparatus and Systems, vol. PAS-90 (February 1970). . "Optimal Operation of Multireservoir System Using Composite Representation." IEEE Transactions on Power Apparatus and Systems, vol. PAS-90 (February 1970). Asbury, J. G., and A. Kouvalis. "Electric Storage Heating: The Experience in England and Wales and in the Federal Republic of Germany." Report No. ANL/ES50. Argonne, Ill.: Argonne National Laboratory, May 1976. Balassa, Bela. "Estimating the Shadow Price of Foreign Exchange in Project Appraisal.'" Oxford Economic Papers, vol. 26 (July 1974), pp. 147-68. Bator, Francis J. "General Equilibrium, Welfare, and Allocation." American Economic Review (March 1957), pp. 22-59. Baumol, William J., and David F. Bradford. "Optimal Departures from Marginal Cost Pricing." American Economic Review (June 1970), pp. 265-83. Bohn, Roger E., Michael C. Caramanis, and Fred C. Schweppe. "Optimal Spot Pricing of Electricity." Working paper no. MIT-EL-81-008WP. Cambridge, Mass., MIT Energy Laboratory, March 1981. Processed. Boiteux, M. ''La Tarification des Demandes en Pointe.'' Revue Generale de I'Electricite, vol. 58 (1949). Boiteux, M., and P. Stasi. "The Determination of Costs of Expansion of an Interconnected System of Production and Distribution of Electricity." Marginal Cost Pricing in Practice. Edited by James Nelson. Englewood Cliffs, N.J.: PrenticeHall, 1964. 368

REFERENCES

369

Burchnall, J. A. "A Review of Off-peak Tariff Incentives and Their Effect in England and Wales." Report no. EP/SEM.5/R17. Brussels: Economic Commission for Europe, Committee on Electric Power, March 1977. Burington, Richard S. Handbook of Mathematical Tables and Formulas, 5th ed. New York: McGraw-Hill, 1973. Canadian Electrical Association. Marginal Costing and Pricing of Electrical Energy. Proceedings of the State of the Art Conference. Montreal: Canadian Electrical Association, May 1978. Canadian International Development Agency. "Proceedings of the Regional Power Systems Engineering Seminar, Kathmandu, Nepal, February-March 1977." Processed. Caves, Douglas W., and Laurits R. Christensen. "Residential Substitution of Offpeak for Peak Electricity Usage under Time of Use Pricing." Energy Journal, vol. I (April 1980), pp. 85-142. Christensen, Laurits. "The FPoduction Function, Method II." Marginal Costing and Pricing of Electric Energy. Montreal: Canadian Electrical Association, 1978. Cicchetti, Charles J., William J. Gillen, and Paul Smolensky. The Marginal Cost and Pricing of Electricity. Cambridge, Mass.: Ballinger, 1977. CIRBD. International Conference on Electricity Distribution. Conference Publication no. 151. London: Institution of Electrical Engineers, 1977. Crew, Michael A., and Paul R. Kleindorfer. Public Utility Economics. New York: St. Martin's Press, 1979. Dasgupta, Partha, Amartya Sen, and Stephen Marglin. Guidelinesfor Project Evaluation (UNIDO). New York: United Nations, 1972. Dasgupta, Partha, and Joseph Stiglitz. "Benefit Cost Analysis and Trade Policies." Journal of Political Economy, vol. 82 (January-February 1974), pp. 1-33. deGrasse, R. V. "Electric St:orageHeating after Two Years." Public Utilities Fortnightly, vol. 23 (January 6, 1977). Dhrymes, Phoebus, J., and M. Kurtz. "Technology and Scale in Electricity Generation." Econometrica, vol. 32 (July 1965), pp. 287-315. Donziger, Alan. "The Underground Economy and the Theft of Utility Services." Public Utilities Fortnightly (November 22, 1979), pp. 23-27. Dupuit, P. "De l'Utilite et de Sa Mesure." La Reforma Soziale, Turin (1932). Electricity Council. Domestic Tariffs Experiment. Load and Market Research Report no. 121. London. Faust, G. Robert, and H. Gary Larson. "International Valuation Procedures for Electric Utilities." Public Utilities Fortnightly (August 14, 1980), pp. 19-25. Federal Power Commission. The Methodology of Load Forecasting, 1970 National Power Survey. Washington, D.C.- Department of Energy, 1970. Ford, L. R., and D. R. Fulkerson. Flows in Networks. Princeton, N.J.: Princeton University Press, 1962. Galatin, M. Economies of Scale and Technical Change in Thermal Power Generation. Amsterdam: North-Holland, 1968.

370

REFERENCES

Garfield, Peter J., and William F. Lovejoy. Public Utility Economics. Englewood Cliffs, N.J.: Prentice-Hall, 1974. Harberger, Arnold C. Project Evaluation. London: Macmillan and Company, 1972. Harris, John R., and Michael P. Todaro. "Migration, Unemployment, and Development." American Economic Review, vol. 60 (March 1970), pp. 126-42. Hastings, B. F. "Ten Years of Operating Experience with a Remote Controlled Water Heater Load Management System at Detroit Edison." IEEE Transactions on Power Apparatus and Systems, vol. PAS-99 (July-August 1980), pp. 1437-41. Havrylyshyn, Oli, and Mohan Munasinghe. "Interactions among Alternative Modes of Energy Production: Empirical Case Study for Sri Lanka." Economics Discussion Paper. Washington, D.C.: George Washington University, April 1980. Herman, F. K. Electricity Tariffs. South Africa: University of Stellenbosch, 1974. Holt, S. B. Jr, and D. M. Crawford. "Distribution Substation Planning Using Optimization Methods." Application of Optimization Methods in Power System Engineering. IEEE tutorial course, text no. 76CH1 1072-PWR. New York: IEEE, 1976. Hotelling, H. "The General Welfare in Relation to Problems of Railway and Utility Rates." Econometrica, vol. 6 (July 1938), pp. 242-69. Huck, G. E. "Load Forecast Bibliography: Phase I." Proceedings of the IEEE PES Summer Meeting, Vancouver, July 1979, paper no. F7950-3. IAEA. Nuclear Power Planning, Study for Indonesia. Vienna, 1976. IE.EE. Application of Optimization Methods in Power System Engineering. IEEE tutorial course, text no. 76CH1107-2-PWR. New York, 1976. Knight, Upton G. Power System Engineering and Mathematics. New York: Pergamon Press, 1972. Krohmn, Gregory C. "Growth and the Cost of Electric Power." Public Utilities Fortnightly (December 18, 1980), pp. 32-35. Lal, Deepak. Methods of Project Analysis. World Bank Staff Occasional Papers no. 16. Baltimore, Md.: The Johns Hopkins University Press, 1974. Little, Ian M. D., and James A. Mirrless. Project Appraisal and Planning for Developing Countries. New York: Basic Books, 1974. Luenberger, David G. Introduction to Linear and Non-Linear Programming. Reading, Mass.: Addison-Wesley, 1973. Marglin, Stephen. "The Social Rate of Discount and the Optimal Rate of Investment." Quarterly Journal of Economics (1963). McCoy, Fred. "The Production Function, Method I." Marginal Costing and Pricing of Electric Energy. Montreal: Canadian Electrical Association, 1978. Miedema, Allen K. "Overview of DOE Analysis of Residential TOU Data." North Carolina, Research Triangle Institute, April 1980. Processed. Miedema, Allen K., S. B. White, C. A. Clayton, and D. P. Lifson. "Analysis of Experimental Time-of-Use Electricity Prices." North Carolina, Research Triangle Institute, December 1978. Processed. Mishan, Ezra J. Cost-Benefit Analysis. New York: Praeger, 1976. Mrr Energy Laboratory. New Utility Management and Control Systems. Cambridge, Mass.: MIT, 1979.

REFERENCES

371

Mitchell, Bridger M., William J. Manning, and Jan Paul Acton. Peak Load Pricing. Cambridge, Mass.: Ballinger Publishing Company, 1978. Munasinghe, Mohan. "The Costs of Electric Power Shortages to Residential Consumers." Journal of Consumer Research, vol. 6 (March 1980), pp. 361-69. Also available as Reprint no. 128 from the World Bank, Washington, D.C. . The Economics of Power System Reliability and Planning. Baltimore, Md.: The Johns Hopkins University Press, 1979. . "Indonesia: PLN Power Tariff Study." Energy Department Report. Washington, D.C.: World Bank, May 1978. "An Integrated Framework for Energy Pricing in Developing Countries." Energy Journal, vol. I (July 1980), pp. 1-30. Also available as Reprint no. 148 from the World Blank, Washington, D.C. "Integrated ' National Energy Planning (rNEP) for the Developing Countries." Natural Resources Forum, vol. 4 (October 1980), pp. 359-73. Also available as Reprint no. 165 from the World Bank, Washington, D.C. "A ' New Approach to Power System Planning." IEEE Transactions on Power Apparatus and Systems, vol. PAS-99 (May-June 1980), pp. 1198-1209. Also available as Reprint no. 147 from the World Bank, Washington, D.C. - "Pakistan: WAPDA Power Tariff Study." Energy Department Report. Washington, D.C.: World Bank, February 1980. . "Planning for Electrical Power: Costs and Technologies." National Development (April 1980), pp. 75-84. Also available as Reprint no. 149 from the World Bank, Washington, D.C. .- ''Principles of Modem Electricity Pricing." Proceedings of the IEEE (March 1981), pp. 332-48. Also available as Reprint no. 185 from the World Bank, Washington D.C. "System ' Planning to Achieve Optimal Reliability Levels." Proceedings of the Third International Conference on Analysis, Forecasting, and Planning for Utilities, Paris, June 1980. Processed. Munasinghe, Mohan, and 'Mark Gellerson. "Economic Criteria for Optimizing Power System Reliability Levels." The Bell Journal of Economics, vol. 10 (Spring 1979), pp. 353-65. Also available as Reprint no. 112 from the World Bank, Washington, D.C. Munasinghe, Mohan, and Gunter Schramm. "Power and Energy Pricing Case Studies." Energy Department Report. Washington, D.C.: World Bank, February 1980. Munasinghe, Mohan, and Jeremy Warford. "Shadow Pricing and Power Tariff Policy." Marginal Costing and Pricing of Electrical Energy. Montreal: Canadian Electrical Association, 1978. Munasinghe, Mohan, and Colin J. Warren, ed. Power Tariffs Case Studies in Latin America and the Caribbean. Washington, D.C.: World Bank, 1981. . "Rural Electrification, Energy Economics, and National Policy in the Developing Countries." Future Energy Concepts, publication no. 171. London: Institution of Electrical Engineers, 1979, pp. 44-47.

372

REFERENCES

Electricity Rate Design Study, reports 66, 67, and 79. Palo Alto, Calif.: 1977, 1980. Pelletier, Y. "Application des Methodes d'Analyse de Donnees a l'Etude d'un Enchantillon de Courbes de Charge Moyenne Tension." Paris, Electricite de France, April 1977. Processed. Pindyck, Robert S. The Structure of World Energy Demand. Cambridge, Mass.: MIT Press, 1979. Pioger, Y. "Elements de Reflexion sur la Deformation des Courbes de Charge Globales et Sectorielles: Evaluation de l'Effet Induit par la Tarification." Paris, Electricite de France Study, July 1977. Processed. Prasad, N. R., J. M. Perkins, and G. Nesgos. "A Markov Process Applied to Forecasting. Part I-Economic Development." Proceedings of the IEEE Power Engineering Society Summer Meeting, Vancouver, July 1973. "A Markov Process Applied to Forecasting. Part II-The Demand for Electricity." Proceedings of the IEEE Power Engineering Society Winter Meeting, New York, January 1974. Renshaw, Edward F. "Expected Welfare Gains from Peak-Load Electricity Charges." Energy Economics (January 1980), pp. 3745. Reynolds, S. P., and T. E. Creighton. "Time of Use Rates for Very Large Customers on the Pacific Gas and Electric Company System." IEEE Transactions on Power Apparatus and Systems, vol. PAS-99 (January-February 1980), pp. 14751. Rischl, R. "Optimal System Expansion Planning: A Critical Review." Systems Engineering for Power: Status and Prospects. Proceedings of the USERDA Conference, Henniker, N. H., August 1975. NTIS no. CONF-750867, pp. 233-60. Ruggles, N. "Recent Developments in the Theory of Marginal Cost Pricing." Review of Economic Studies, vol. 17 (1949-50), pp. 107-26. . "The Welfare Basis of the Marginal Cost Pricing Principle." Review of Economic Studies, vol. 17 (1949-50), pp. 2946. Sassone, Peter G. "Shadow Pricing in CBA: Mathematical Programming and Economic Theory." The Engineering Economist, vol. 22 (Spring 1977), pp. 219-33. Saunders, Robert J., Jeremy J. Warford, and Patrick C. Mann. "Altemative Concepts of Marginal Cost Pricing for Public Utility Pricing: Problems of Applications in the Water Supply Sector." World Bank Staff Working Paper no. 259. Washington, D.C.: World Bank, May 1977. Schramm, Gunter, and Mohan Munasinghe. "Interrelationships in Energy Planning: The Case of the Tobacco Industry in Thailand." Energy Systems and Policy (December 1980), pp. 117-39. Schuler, Richard E., and Benjamin F. Hobbs. "Spatial Competition: Applications in the Generation of Electricity." Paper presented at the 1981 Annual Meeting of the Southern Regional Science Association, Washington, D.C., April 1981. Processed. Schwartz, Harry, and Richard Berney, eds. Social and Economic Dimensions of Project Evaluation. Washington, D.C.: Inter-American Development Bank, 1977.
NARUC. EPRI,

REFERENCES

373

Soares, S., C. Lyra, and H. Tavares. "Optimal Generation Scheduling of Hydrothermal Power Systems.' IEEE Transactions on Power Apparatus and Systems, vol. PAS-99 (May-June 1980), pp. 1107-18. Squire, Lyn, and Herman G. van der Tak. Economic Analysis of Projects. Baltimore, Md.: The Johns Hopkins University Press, 1975. Stanton, K. N. "Medium-Range, Weekly, and Seasonal Peak Demand Forecasting by Probability Methods." IEEE Transactions on Power Apparatus and Systems, vol. PAS-71 (May 1971), pp. 1183-89. Steiner, P. "Peak Loads and Efficient Pricing." Quarterly Journal of Economics (November 1957). Stott, Brian. "Review of Load-Flow Calculation Methods." Proceedings of the IEEE, vol. 62 (July 1974,), pp. 916-29. Suelflow, James. Public Utility Accounting: Theory and Application. East Lansing: Michigan State University, 1973. Sullivan, Robert L. Power System Planning. New York: McGraw-Hill International Book Company, 1977. "Symposium on Peak Load Pricing." The Bell Journal of Economics, vol. 17 (Spring 1976), pp. 197-250. Taylor, Lester D. "The Demand for Electricity: A Survey." The Bell Journal of Economics, vol. 6 (Spring 1975), pp. 74-110. "Thirtieth Annual Electrical Energy Forecast." Electrical World (September 15, 1979), pp. 75-84. Turvey, Ralph. Optimal Pricing and Investment in Electricity Supply. Cambridge, Mass.: mrr Press, 1968. Turvey, Ralph, and Dennis Anderson. Electricity Economics. Baltimore, Md.: The Johns Hopkins University Press, 1977. United States Congress, House of Representatives. Public Utilities Regulatory Policies Act (PURPA), Report no. 95.1750. 95 Cong., 2 sess., October 10, 1978. United States Department of Energy. "Report no. DOEIRG-0036." Washington, D.C., June 1980. Processed. Vardi, Joseph, Jacob Zahavi, and Benjamin Avi-Itzhak. "Variable Load Pricing in the Face of Loss of Load Probability." The Bell Journal of Economics, vol. 8 (Spring 1977), pp. 270-88. Vickrey, William. "Responsive Pricing of Public Utility Services." The Bell Journal of Economics, vol. 2 (Spring 1971), pp. 337-46. Warr, Peter. "On the Shadow Pricing of Traded Commodities." Journal of Political Economy, vol. 85 (1977), pp. 865-72. Wenders, J. T. "Peak Load Pricing in the Electric Utility Industry." The Bell Journal of Economics, vol. 7 (Spring 1976), pp. 232-41. Williamson, 0. E. "Peak Load Pricing and Optimal Capacity under Indivisibility Constraints." The American Economic Review, vol. 56, no. 4 (September 1966), pp. 810-27. Wismer, D. A. Optimization Methods for Large Scale Systems. New York: McGraw-Hill Book Company, 1971.

Index

Administrative and general costs (A&G): in Indonesia, 119; in Pakistan, 156; in the Philippines, 178, 184, 191, 193; in Thailand, 232 Afghanistan, 261 Agus River Basin (Philippines), 174, 178, 185, 194 Algeria, 262 Anderson, Dennis, 7 Appliances, tariffs in Thailand and, 240-41, 244 Argentina, 263 Bangladesh, 264 Barbados, 265 Base-load steam plants: generation and, 38-39; hydrothermal systemis and, 62; in Indonesia, 107, 125; in Thailand, 233 Billing, 79, 81, 83, 240 Boiteux, M., 15 Bolivia, 266 Botswana, 267 Brazil, 268 Burma, 269 Cagayan Electric Light and Power Company (Philippines), 179 Cameroon, 270 Capacity, 11; additions to power systems, 17-19; hydro and thermal (Sri Lanka), 205; hydrothermal systems and, 65; installed (Philippines), 170, 174; installed and reserved (Pakistan), 1.44; installed (Thailand), 225, 227; theoretical pricing model and, 23-24 Capacity costs: all-hydro systems and, 61;

case studies and, 103-04; generation, 100; in Indonesia, 117-19, 133, 139; LRMC computation and, 53-57; in Pakistan, 145-50, 151; peak and off-peak consumers and, 327-29; in the Philippines, 175-77, 179-80, 184, 189; shadow pricing and, 104; in Sri Lanka, 213-14; in Thailand, 232-33, 234-35, 236 Capital indivisibilities, 17-19 Capital, shadow price of, 50 Case studies: divergences from STLRMC in, 103-04; financial viability in, 102-03; market distortions in, 104-05; metering in, 104; methodology used in, 99-101; political considerations in, 105; second-best issues in, 105; system losses in, 101; system planning in, 102; tariff reforms in, 105-06. See also Specific country names Ceylon Electricity Board (CEB, Sri Lanka), 205, 207, 211 Chad, 271 Chile, 272 Columbia, 273 Commercial consumers. See Consumers, commercial Connection charges: in Indonesia, 107, 109, 131; in Pakistan, 166; in Thailand, 235 Consumer categories: of tariff structure in Indonesia, 131; of tariff structure in Pakistan, 158, 164-65 Consumer-related costs: in Indonesia, 125; LRMC computation and, 58-59; in Pakistan, 151; in the Philippines, 178-79, 181-82, 185-86, 191, 193; in Sri Lanka, 215; in Thailand, 235 Consumers: adjusting tariff structures and,

375

376

INDEX

78; commercial, 87-93, 135, 159, 170, 188, 190, 239, 245; industrial, 87-93, 135, 164-65, 170, 188, 196, 203, 227, 229, 231, 239, 240-41, 245; interruptible, 86; losses and, 101; marginal cost pricing and, 199; pricing and education of, 102; residential, 104, 134-35, 159, 188, 190, 227, 239; tariff application and, 87-95; tariffs and, 86, 95. See also Low-income consumer Consumption: electricity (Philippines), 170; electricity (Thailand), 226, 229, 240-41; pricing and, 71, 72; shadow pricing and, 43, 49-51 Cooperatives. See Electric cooperatives (Philippines) Correlation techniques (forecasting), 32-33 Costa Rica, 274 Cost categories (defined), 52-53 Customers. See Consumers Cyprus, 275 Demand. See Electricity demand Demand forecasting: approaches to, 102; in Indonesia, 110-16; load factor relation and, 27-29, 32; LRMC computation (theoretical) and, 19-21, 25; in Pakistan, 141-43; in Sri Lanka, 208-11; in Thailand, 229. See also Load forecasting Distribution: capacity costs and, 56-57; consumer costs and, 58-59; costs in Indonesia, 119, 139; costs in the Philippines, 179, 193, 198; defined, 29-31; expansion program in Pakistan and, 144-45; LRMC computation (Pakistan) and, 148; in Sri Lanka, 205, 213-14; system planning and, 42; in Thailand, 225-26, 233-34 Dominican Republic, 276 Ecuador, 277 Efficiency: in electricity sector, 5; LRMCbased tariffs and, 11-12 Egypt, 278 Electric cooperatives (Philippines), 167, 202; financial costs and, 193-94, 200; LRMC of retail supply and, 183-87; tariffs and, 188, 197-98 Electric Generating Authority of Thailand (EGAT): capacity costs and, 232; demand forecasts and, 229; electricity supply and, 226; energy costs and, 233, 234; genera-

tion and, 231; power development and, 229; power system of, 223; system characteristics of, 226-27; tariffs and, 227-29, 236-38 Electricity demand: generation planning and, 37-38; in Indonesia, 110-16; load forecasting and, 27-34; in the Philippines, 170-71; pricing and, 102; social goals and, 26; Sri Lanka and, 208-11, 229; structuring LRMC and, 53; in Thailand, 226, 229; theoretical model and, 19-21, 23-24. See also Demand forecasting Electricity supply: LRMC bulk power calculation (NPC, Philippines) and. 175-79; LRMC retail power calculation (cooperatives, Philippines) and, 183-87; LRMC retail power calculation (private sector, Philippines) and, 179-82; in the Philippines. 169-70; in Sri Lanka, 208-11; in Thailand, 225-26 Electric power sector: in Indonesia, 109-16; in Pakistan, 140-45; in the Philippines, 167-74; in Sri Lanka, 205-11; in Thailand, 223-31 Electric power systems: basic, 29-32; planning of, 34-42 El Salvador, 279 Energy: alternative, 72, 199; load forecasting and, 27-29; policy decisions, 26; power and, 27 Energy costs, 77; for all-hydro systems, 61; in Indonesia, 125, 134, 151; LRMC computation and, 57-58; in Pakistan, 150-51; for peak and off-peak consumers, 327-29; in the Philippines, 177-78, 180-81, 185, 187, 191-93, 202-03; in Sri Lanka, 215; in Thailand, 233-34. 236 Ethiopia, 280 Europe: response to prices in, 92-94; tariff structures in, 87 Expansion of systems: capacity costs and, 53; generation and, 37-40; in Indonesia, 107, 138; in Pakistan, 143-45, 148; peak-load pricing and, 17-18; in the Philippines, 172-74, 190-91, 192; planning of, 21-22, 36; in Thailand, 229-31 Export promotion (Thailand), 246 Exports, shadow prices and, 46, 47-48 Fiji. 281 Finances: case studies and, 102-03; in the Philippines, 190-94, 200; tariff policy and,

INDEX

377

73-76 Fuels: diesel fuel subsidizationin Pakistan, 105; diesel fuel subsidizationin Thailand, 241; domestic prices and, 104;energycosts in Pakistan and, 150-51; energy costs in the Philippinesand, 177,192;environmental damage and, 67; fossil, 5, 25; hydro power in Sri Lanka and, 209--10,213;kerosene (subsidy),67; naturalgas, 165; petroleum, 198, 233; prices of (Indonesia), 125 Generation,53; in all-hydrosystems,61; in basic power systems, 31; capacity costs, 100; in Indonesia, 107, 109, 117-19; in Pakistan, 141-43, 144;in the Philippines, 169, 173; in Sri Lanka, 209; systemplanning and, 36, 37-40, 41; in Thailand, 231-32, 233, 241. See also Geothermal generation; Hydroelectric systems; Thermal generation Geothermal generation,5, 173, 174, 188.See also Generation Ghana,282 Grid systems,100-01 Guatemala, 283 Guinea,284
Guyana, 285

in, 110-16; energycosts in, 125, 134;generatingcapacityin, 107, 109, 117-19; meteringin, 131, 132;off-peakperiodcharges in, 131-34; O&Mand A&Gcosts in, 119; peak period charges in, 131; policy and, 138-39; power sector organization in, 109-16; power sector statistics for, 289; price increases in, 107; pricing structure (STLRMC)and, 131-34, 138-39; regional imbalance in, 110; STLRMC adjustment and, 128-35; STLRMC for Java and, 116-28, 138, 139;subsidizationand, 131, 134, 138; tariffs in, 107, 128, 131-32, 134-35, 138; tariff uniformity and, 103, 138; transmissioncosts in, 119, 139. See also Java Industrialconsumers.See Consumers,industrial Industrial development, 77; in the Philippines, 202, in Thailand,246 Interruptible loads, 87 Investment,74, 77; energy pricing decisions and, 26; least-cost system of, 25; LRMC tariffs and, 11; in Pakistan, 143, 148; in the Philippines,175, 176, 179, 180, 186, 187; shadow price model and, 44, 50; in Thailand,232, 234
Ireland, 290

Haiti, 286 Honduras,287 Hotelling,H., 15 Hydroelectric systems:all-hydro(LRMCcalculation),59-61; generation planning and, 36; hydrothermal (LRMC calculation), 62-65; in Indonesia,109;in Pakistan,144, 145, 150; in the Philippines, 169, 173, 174, 188, 192, 203; in Sri Lanka, 205, 207-11, 213;systemplanningand, 102; in Thailand,227, 233. See also Generation
Imports: of generation plants (private), 67; shadow price application and, 46 Income distribution: pricing and, 72; shadow prices and, 42-43; subsidy and, 67, 201 Income: tariff system in Indonesia and, 134-35; tariff system in l'akistan and, 158-59. See also Low-incomreconsumer India, 288 Indonesia: capacity costs in, 117-19, 133; consumer costs in, 125; distribution costs in, 119, 139; electricity demand forecasting

Irrigation tariff charges(Thailand),239 Ivory Coast,291 Jamaica, 292 Java: electricity demand forecast and, 110, 111-15; electric power generation in, 109-10; policy recommendations for, 138, 139, STLRMCfor, 116-28, 138. See also Indonesia Jordan, 293 Karachi Electric Supply Corporation (KESC,
Pakistan), 140, 144 Kenya, 294 Korea, 295 Labor, 48-49, 175 Land, 48, 50 Liberia, 296 Load forecasting, 27, 34, 42, 141. See also Demand forecasting Long-run marginal costs (LRMC): defined, 23; shadow priced, 50-5 1; tariffs based on,

378

INDEX

11-12; theoretical model and, 19-24. See also Strict long-run marginal costs (STLRMC) Losses, 31, 101nl; analysis of system, 101; LRMC computation of, 58; in Pakistan, 166; T&D (Philippines), 185; in Thailand, 232 Loss-of-load probability (LOLP), 21, 24 Low-income consumers: kerosene subsidy for, 67; pricing policy and, 69-73; subsidized social prices for, 17n4; subsidized tariffs and, 67-69, 77, 82; tariffs in Indonesia for, 134-35; tariffs in Pakistan for, 159; tariffs in the Philippines for, 199, 201; tariffs in Thailand for, 241-45. See also Income LRMC. See Long-run marginal costs Luzon: bulk power costs in, 178; capacity costs in, 176; consumer service charge in, 203; electricity demand in, 170, 171; electricity supply in, 169; energy costs in, 193, 202; expansion program for, 173; financial costs in, 191; outages in, 203; tariffs in, 188, 195, 196, 201, 202. See also Philippines Madagascar. 297 Malawi, 298 Malaysia, 299 Mali, 300 Manila, 169, 179 Manila Electric Company (MECO, Philippines), 167, 171, 173, 179; electricity demand and, 170-71; electricity supply and, 169; tariffs and, 188 Marginal cost pricing: basic model of, 15-17; LRMC computations for, 19-24; peak load pricing and, 17-18; seminars on (World Bank), 9-10, 251-58 Mauritius, 301 Metering, 87; analysis of, 79-82; homeostatic utility control and, 82; in Indonesia, 132; in Pakistan, 165; in the Philippines, 190, 191; special, 104; in Thailand, 235, 239, 240, 241, 248; time-of-day, 79, 82, 104, 165, 239, 240, 241, 248; types of tariffs and, 84-86 Metropolitan Electricity Authority (MEA, Thailand), 223; capacity costs for, 234; demand forecasts and, 229; distribution sys-

tem of, 225-26; electricity supply by, 226; generation by, 231, 232; power development by, 229-31; tariffs of, 227-29, 238-39, 246-47; transmission by, 232 Mexico, 302 Mindanao (Philippines), 200; bulk power costs in, 178; capacity costs in, 175-76; consumer service charge in, 203; electricity demand in, 170, 171; electricity supply in, 169; energy costs in, 177, 193, 202; expansion program in, 174; financial costs in, 191; tariffs in, 188, 189, 196, 201, 202 Morocco, 303 Multan Electric Supply Company (Pakistan), 140 National Electrification Administration (NEA, Philippines), 167, 183; cooperatives and, 171; generating facilities of, 170 National Power Corporation (NPC, Philippines), 167, 183, 199; bulk power supply LRMC for, 175-79; cooperatives and, 194; electricity demand of, 170, 171; electricity supply by, 169; energy costs of, 202-03; expansion program of, 172-74, 190-91; financial costs of, 190, 200; operation of, 171; tariffs of, 188, 194-95; voltage level tariff adjustments and, 175 Nepal, 304 Nicaragua, 305 Nigeria, 306 Nontradable inputs: defined, 46; shadow prices and, 47, 48 Nuclear plants, 5, 173, 210n3 Numeraire, shadow pricing and, 44-46 Off-peak load periods, 92; capacity costs of, 327-29; charges in Indonesia of, 131-34; energy costs and, 327-29; energy costs in Indonesia of, 125; energy costs in Pakistan of, 150; policy in Thailand for, 240, 241; prices in Indonesia for, 131-34; STLRMC in Sri Lanka for, 211; structuring LRMC for, 53; tariff structure in Thailand for, 236-38 Oman, 307 Operation and maintenance costs (O&M): in Indonesia, 119; in Pakistan, 156; in the Philippines, 175-76, 177, 179-80, 184, 186, 191, 193; in Thailand, 232, 234

INDEX

379

Outages:costs of, 24; in Luzon, 203 Pakistan:capacitycostsin, 145--50,151;consumercosts in, 151;dieselsubsidyin, 105; demand and energy forecastsfor, 141-43; electric power sector in, 140-45; energy costs in, 150-51; expansion program in, 143-45, 148;generationin, 141-43, 144; policy in, 165-66; power sector statistics for, 308; STLRMCadjustmentin, 158-65; STLRMC computation in, 145-56; tariff structure (new) for, 158-65, 166; tariff uniformityin, 103;tubewelltariff in, 165 Panama,309 Peak load periods:for all-hydrcsystems,61; capacity cost charges for, 55, 327-29; charges in Indonesia for, 131; electricity demandforecast in Indonesiafor, 111-16; energy costs for, 57-58, 327-29; energy costs in Indonesiafor, 125;energycosts in
Pakistan for, 150; load factor for, 29; LRMC tariffs for, 11-12; policy in Thailand for, 240-41; STLRMC in Sri Lanka for, 211; structuring LRMC for, 53; tariff structure in Thailand for, 236-38 Peak load pricing, 78, 86; analysis of, 87-92; in Indonesia, 131; LRMC computation (theoretical) of, 21; model of, 18-19; in Pakistan, 145-46. See also Prices Peru, 310 Perusahaan Umum Listrik Negara (PLN, Indonesia): connection charges of, 131; electricity demand growth of, 107; power sector organization of, 109-16 Philippines: capacity cost in, 175-77, 179-80, 184, 189; consumer costs in, 178, 181-82, 185-86; electricity demand in, 170-71; electricity supply in, 169-70, 175-87; electric power sector in, 167-74; energy costs in, 177-78, 180-81, 185, 187, 191, 192, 193, 202-03; expansion program in, 172-74, 190-91, 192; financial cost structure in, 190-94, 200; policy in, 198-204; power sector statistics for, 311; regional electricity prices in, 201; regional power costs in, 103; STLRMC adjustment in, 188-98; STLRMC' computation in, 174-87; subsidies in, 198, 199, 201, 204; tariffs in, 188-90, 194-98, 199, 200, 201, 202; tariff uniformity for, 201; volt-

age level tariff adjustmentsand, 200-01. See also Luzon;Mindanao; NationalPower Corporation (NPC, Philippines); Visayas Planning,21-22, 25; in Indonesia,138; load forecasting and, 27; system,34-42, 102 Policies (government): national uniform tariff, 78, 103;tariffreformresistanceand, 105-06. See also specificcountry names; Pricingpolicy Portugal,312 Power:basicpowersystemand, 31; LRMCof all-hydro systems and, 61; regional costs (Philippines) and, 103; relation between energyand, 27-29 Powerfactor(PF) penaltysurcharges,86 Powersector tax, 77, 247 Prices: border, 46-47, 236; of bulk power supply (Philippines), 177, 188; energy charges in the Philippines and, 202-03; fuel (Indonesia),125;increases in cooperative (Philippines), 197; increases in Indonesian, 107; increases in Thai, 245-46; outages and, 24; pricing structure in the Philippines and, 198-99; social equity in Thailand and, 241-45; structure of Indonesian (STLRMC), 131-34, 138-39; tariff adjustment and, 75-76. See also Shadow prices; Peak load pricing Pricing periods: for LRMC computations, 53; metering decisions and, 81-82; in Thai tariff structure, 236-38 Pricing policy: capacity cost charges and, 55; financial viability (STLRMC) and, 102-03; lifeline, 68-69; optimal electricity, 69-73; second-best issues and, 105; for Thailand, 241, 248 Private utilities, 35; in the Philippines, 170-71, 179-82, 188, 189-90, 192-93, 196; in Thailand, 241 Production function, 17; of electric power system, 35 Provincial Electricity Authority (PEA, Thailand), 223; capacity costs for, 234-35; demand forecasts by, 229; electricity supply by, 226; generation by, 231, 232; power development by, 231; tariffs of, 227-29, 238-39, 246-47; transmission by, 232 Public utilities: financial viability and, 73-76; shadow pricing and, 35 Rate periods (LRMC computations), 53

380

[NDEX

Rates. See Tariffs Rawalpindi Electric Power Company (Pakistan). 140 Regional development, 76; in Thailand. 246-47 Reliability measures, 21, 22 Reservoir (all-hydro systems), 59-61 Residential consumers. See Consumers, residential Resource allocation, 73, 105, 246; LRMC based tariffs and, 11; pricing and, 71, 72; shadow pricing and, 42, 43 Romania, 313 Ruggles, N., 15 Rural electrification: in Pakistan, 103, 141, 166; in the Philippines, 167, 169, 198-99; in Thailand, 244, 246-47

Second-best considerations, 17n4, 66-67; in case studies, 105 Shadow prices, 5, 17n4; analysis of, 42-44; applying, 46-51: case study estimation of, 104-05; numeraire and, 44-46; in STLRMC computation (Philippines), 175; in system planning, 34, 35, in Thai study, 225, 235-36. See also Prices Short-run marginal costs (SRMC): in peak load pricing model, 17-18; in theoretical model, 23 Tanzania, 319 Social benefits, 26; electricity pricing and, Tariffs: billing and, 79, 81; block, 84-85, 69-73 104; changes in (reform), 105-06; financial Spillage (all-hydro systems), 61 viability and, 73-76, 102-03; in IndoneSri Lanka: capacity costs in, 213-14; consia, 107, 128, 131-32, 134-35, 138; inflasumer costs in, 215; electric power sector tionary effects of increasing, 78, 105, in, 205-11; energy costs in, 214; power 245-46; kilowatt-hour consumption and, sector statistics for, 314; STLRMC adjust83, 84, 85, 86; lifeline, 68-69, 76, 87, ment for, 215; STLRMC computation for, 243, 245; load forecasting problems and, 33; LRMC based, 11-12; LRMC computa211-15; tariffs in, 215 tion and, 53, 66; metering and, 79-82, SRMC. See Short-run marginal costs Steiner, P., 15 104; objectives of, 10-1 1; in Pakistan, 158-65; peak period, 84, 85-86; in the STLRMC. See Strict long-run marginal costs Philippines, 188-90, 194-98, 199, 200, Strict long-run marginal costs (STLRMC): 201; pricing considerations and, 102-03; basis of, 35; billing and, 79, 81; case study divergences from 103-04; computer prosecond-best considerations and, 66-67, gram estimation of, 330-65; consumer 105; in Sri Lanka. 215; STLRMC adjustcosts and, 58-59; cost categories and, ment and, 75-76; subsidies and, 67-69, 76-77, 105; in Thailand, 225, 227-29, 52-53; defined, 52; financial viability and. 236-40, 241-45, 247, 248; time-of-use 73-76; generation capacity costs and, 100; (TOU), 84, 85-86, 93-94, 203, 247; twohydroelectric generation systems and, period, 79-82; uniform national, 78, 103, 59-65; industrial development and, 77; in138, 201. See also Prices flationary effects and, 78-79; lifeline

prices and, 68-69; losses and, 101; marginal capacity costs and, 53-57; marginal energy costs and, 57-58: marginal energy losses and, 58; market distortions and, 104-05; metering and, 79-82; optimal electricity pricing and, 69-73; pricing and, 102-03; pricing periods and, 53; regional development and, 76-77; revenue generation and, 77, 247; second-best considerations and, 66-67, 105; structuring, 52-53; subsidized prices and, 67-69, 76-77, 81; tariff structure and, 78; tariff structure reform and, 105-06. See also Long-run marginal costs (LRMC); specific country names Subsidies, 17n4, 76; in Indonesia, 131, 134, 138; kerosene, 67; metering decisions and, 82; in Pakistan, 105, 165, 166; in the Philippines, 198, 199, 201, 204; tariff setting and, 67-69, 76-77; in Thailand, 241, 243-45, 247 Sudan, 315 Supply. See Electricity supply Suriname, 316 Swaziland, 317 Syria, 318 System expansion. See Expansion of systems

INDEX

381

Thailand:capacitycosts in, 232-33, 234-35, 236; consumer costs in, 235; electricity supplyin, 225-26; electricpowersector in, 223-31; energy costs in, 233--34,236; energy and demand forecastsfcr, 29; expansionprogramin, 229-31; exportpromotion in, 246; industrial development in, 246; load period policy for, 240-45; metering in, 239, 240, 241, 248; policy issues for, 240-48; power sector statistics for, 320; prices increasesin, 245-46; pricing policy for, 248; private generation in, 241; shadow pricing in, 235-36; STLRMC adjustment for, 236-39; STLRMC computation for, 231-36; subsidies in, 241, 243-45, 247; tariffs in, 225, 227-29, 236-40, 247, 248; tariff structure and social equity in, 241-45; tariff tniformityin, 103.See alsoElectricGenerating Authority of Thailand(EGAT); MetropolitanElectricity Authority (MEA, Thailand); Provincial ElectricityAuthority(PEA, Thailand) Thermalgeneration,87; mixed hydrothermal systems and, 62-65; in Pakistan, 144; in Sri Lanka, 207-09; in Thailand, 233. See also Generation Time-of-day metering.See Metering,time-ofday rime-of-use(TOU) tariffs.See Tariffs,timeof-use (TOU) Tradableinputs, 48 Transmission: capacity costs and, 56-57; costs in Indonesia, 119, 139; costs in the Philippines, 179; defined, 29-31; expansion programin Pakistanand, 144-45; expansion program in the Philippines and, 174; losses, 99nl; LRMC computation (Pakistan) and, 148; in Sri Lanka, 205, 213-14; system planning and, 40-41; in Thailand,225, 226, 231, 232

Tubewells(Pakistan),158, 165, 166 Tunisia, 321 Turbines: diesel, 64, 109; gas, 38-39, 40, 57, 62, 109, 125, 144, 146, 150; steam, 62, 64, 109 Turkey,322 Turvey,Ralph, 7 United States: rate making in, 92-93; TOU tariffs in 93-94 Uruguay,323 Utilities.See Privateutilities;Publicutilities Visayas (Philippines):bulk power costs in, 178;capacitycosts in, 176;consumerservice charge in, 203; electricity demandin, 170, 171;electricity supplyin, 169-70;energy costs in, 193, 202;expansionprogram in, 173; financialcosts in, 191;tariffs in, 188, 189, 194, 195, 196,201, 202 VisayasElectricCompany,179 Voltage levels, 29, 165; capacity costs in Pakistanand, 148-50; marginalcost pricing adjustment(Philippines)for; 200-01; pricing structure in Indonesia and, 131, 132-33; STLRMCcomputation in the Philippinesand, 175; tariffs,92 Waterflow (all-hydrosystems),59-61 Water and Power DevelopmentAuthorityof Pakistan(WAPDA), 144; capacitycosts for, 145-46; load forecasting for, 141-43; policy of, 166;power sectororganization and, 140-43; tariffsof, 158 Water, value of (hydrothermal systems), 62-63 YemenArabRepublic,324 Yemen, People's Democratic Republic of, 325 Yugoslavia,326

The full range of World Bank publications, both free and for sale, is described in the Catalog of World Bank Publications; the continuing research program is outlined in World Bank Research Program: Abstracts of Current Studies. Both booklets are updated annually; the most recent edition of each is available without charge from the Publications Unit, World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A. Mohan Munasinghe is a senior economistengineer in the Energy Department of the World Bank. Jeremy J. Warford is a consultant to the Bank's Energy Department and a division chief in the Population, Health, and Nutrition Department.

0 8(11 27035