Escolar Documentos
Profissional Documentos
Cultura Documentos
Edition 2.1
July 2003
F.T. Sparrow
Brian H. Bowen
Shimon K. Modi
PURDUE UNIVERSITY
Power Pool Development Group
Institute for Interdisciplinary Engineering Studies
500 Central Drive, Suite 270
West Lafayette, IN 47907-2022 USA
Fax: 765-494-2351
ii
Table of Contents
Page
Introduction
Chapter 1 1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
Chapter 2 2.1
2.2
2.3
2.4
2.5
2.6
2.7
Chapter 3 3.1
3.2
3.3
Chapter 4 4.1
4.2
4.3
1
Definition of Economic Terms
2
2
2
3
5
5
5
5
6
7
8
10
11
13
15
16
16
17
18
21
21
26
27
27
29
30
31
32
Chapter 5 5.1
5.2
Chapter 6 6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
6.10
6.11
iii
40
40
43
45
45
45
45
46
46
47
47
47
48
48
49
50
69
Bibliography
76
Introduction
SUFG (Purdue Universitys State Utility Forecasting Group) electricity and natural gas systems
planning models have been employed for the past 20 years. Many countries around the world
have yet to develop the capacity for construction and use of these analytical tools and so
Purdues modeling team is being instrumental in promoting programs of collaboration with
governments, utilities, and universities internationally. These collaborative international
modeling activities encourage regional cooperation and provide a substantial quantitative basis
on which to build improved regional electricity trading policies with potential enormous cost
saving options from collective construction and closer regional integration. This introductory
training manual is intended for general planners who wish to better understand and use these
economic models for electricity market assessment and for support in training for data collection
for the models.
Purdues electricity trade models are used in Indiana, USA Mid-West and internationally since
1997 (Southern African Power Pool, SAPP). The SAPP work was funded by the USAID with
much interest and general support from the DOE and the World Bank. Following the successful
work with SAPP the Purdue energy modeling group in now playing a significant role if the
infrastructure building of the West Africa Power Pool (WAPP). The organization of these
international projects is administered through Purdue Universitys Power Pool Development
Group (PPDG). Both the SUFG and PPDG are housed at Purdue Universitys Institute for
Interdisciplinary Engineering Studies (IIES).
SUFG advises and is supported by the Indiana State Government since the early 1980s. SUFGs
forecasting and marketing models provide quantitative analysis of many electricity policy
scenarios. All of the interested stakeholders have full and equal access to the SUFG model
formulation and so there is a transparency to the analysis that promotes in-depth study of options
for construction of new capacity (generation and transmission), deregulation, and tariff structures
that face government and the utilities.
The SUFG forecasting and long-term planning models use mathematical programming and
operations research techniques (linear and mixed integer) to combine the many economic and
technical objectives and constraints into clearly defined algorithms for optimal (cost
minimization) solutions.
There is a fully detailed description of the SUFG long-term (LT) electricity-planning model in
the Long-Term Model User Manual, Edition 7, June 2000. This can be freely downloaded
from the web page: http://engineering.purdue.edu/IIES/PPDG/
The LT Model User Manual provides a full description of the objective function, load balance
equation, capacity and reliability constraints, and technical operating instructions. It is written
with the technical user and operations research specialist in mind. Over the past three years a
user-friendly windows interface has been developed that helps all energy planners to easily use
the complex LT model. This current document, the General Training Manual, is for the general
electricity policy decision maker. It is for the person who is not so involved or interested in the
precise line by line description of the program structure and who does not have the time to
investigate it and to look into all the technical details of the model.
Chapter 1
Definition of Economic Terms
As we start to look at energy modeling, lets first be sure of some of our essential economic
terms. Lets first look at some definitions.
1.1
An economist thinks of cost differently from an accountant, who is concerned with the firms
financial statements. Accountants tend to take a retrospective look at a firms finances because
they have to keep track of assets and liabilities and evaluate past performance.
Economists take a forward-looking view. They are concerned with what cost is expected to be
in the future, and how the firm might be able to rearrange its resources to lower its cost and
improve its profitability. They must therefore be concerned with opportunity cost, the cost
associated with opportunities that are foregone by not putting the firms resources to their
highest value use. [1]
1.2
Opportunity Cost
The benefit foregone by using a scarce resource for one purpose instead of for its next best
alternative use.
An opportunity cost is incurred because of the use of limited resources, such that the opportunity
to use those resources to monetary advantage in an alternative use is foregone. Thus, it is the
cost of the best rejected (i.e., foregone) opportunity and is often hidden or implied.
Example:
Suppose that a construction project involves the use of a storage space presently owned by a
company. The cost for that space to the project should be the income or savings that possible
alternative uses of the space may bring to the company. In other words, the opportunity cost for
the space should be the income derived from the best alternative use of it. This may be more
than or less than the average cost of that space obtained from the accounting records of the
company [3].
1.3
Fixed costs are those unaffected by changes in activity level over a feasible range of operations
for the capacity or capability available. Typical fixed costs include interest costs on borrowed
capital, insurance and taxes on facilities, general management and administrative salaries, and
license fees. Of course, any cost is subject to change, but fixed costs tend to remain constant
over a specific range of operating conditions. When large changes in usage of resources occur,
or when plant expansion or shutdown is involved, fixed costs will be affected.
Variable costs are those associated with an operation that vary in total with the quantity of
output or other measures of activity level. If you were making an engineering economic analysis
of a proposed change to an existing operation, the variable costs would be the primary part of the
prospective differences between the present and changed operations as long as the range of
activities is not significantly changed. For example, the costs of material and labor used in a
product or service are variable costs because they vary in total with the number of output units
even though the costs per unit stay the same.
1.4
Marginal Cost
An incremental or marginal cost is the additional cost, or revenue, that results from increasing
the output of a system by one (or more) units. Marginal cost is often associated with go/no go
decisions that involve a limited change in output or activity level. For instance, the incremental
cost per mile for driving an automobile may be $0.27, but this cost depends on considerations
such as total mileage driven during the year (normal operating range), mileage expected for the
next major trip, and the age of the automobile. Also, it is common to read of the incremental
cost of producing a barrel of oil. The incremental cost (or revenue) is often quite difficult to
determine in practice.
With electricity generation the marginal cost is a function of how much advance notice is
given for demand. One additional MW in a minutes time horizon is a very different cost to an
additional MW in one months time.
The data in Table 1.4 describe a company with a fixed cost of $50. Variable cost increases with
output, as does total cost. The total cost is the sum of the fixed cost in column (1) and the
variable cost in column (2). The marginal cost of increasing from output from 2 to 3 units is
$20, because the variable cost of the firm increases from $78 to $98. Total cost of production
also increases from $128 to $148. The average total cost of producing at a rate of five units is
$36, $180/5. Average cost tells us the per unit cost of production.
Table 1.4 Short-Run Costs
Rate of
Output
0
1
2
3
4
5
6
7
8
9
10
11
Fixed Cost
(FC)
(1)
50
50
50
50
50
50
50
50
50
50
50
50
Variable
Cost
(VC)
(2)
0
50
78
98
112
130
150
175
204
242
300
385
Total Cost
(TC)
(3)
50
200
128
148
162
180
200
225
254
292
350
435
Marginal
Cost
(MC)
(4)
50
28
20
14
18
20
25
29
38
58
85
Average
Fixed Cost
(AFC)
(5)
50
25
16.7
12.5
10
8.3
7.1
6.3
5.6
5
4.5
Average
Variable
Cost (AVC)
(6)
50
39
32.7
28
26
25
25
25.5
26.9
30
35
Average
Total Cost
(ATC)
(7)
100
64
49.3
40.5
36
33.3
32.1
31.8
32.4
35
39.5
Example:
A team of four colleagues live in the same geographical area and intend to travel together to a
conference (a distance of 400 miles each way). One of the team has a car and agrees to take the
other three if they will pay the cost of operating the car for the trip. When they return from the
trip, the owner presents each of them with a bill for $102.40, stating that he has kept careful
records of the cost of operating the car and that, based on an annual average of 15,000 miles,
their cost per mile is $0.384. The three others felt that the charge is too high and ask to see the
cost figures on which it is based. The owner shows them the following list:
Cost Element
Gasoline
Oil and lubrication
Tires
Depreciation
Insurance and taxes
Repairs
Garage
Total
The three riders, after reflecting on the situation, form the opinion that only the costs for
gasoline, oil and lubrication, tires, and repairs are a function of mileage driven (variable costs)
and thus could be caused by the trip. Because these four costs total only $0.198 per mile, and
thus $158.40 for the 800-mile trip, the share for each person would be $158.40/3 = $52.80.
Obviously, the opposing views are substantially different. Which, if either, is correct? What are
the consequences of the two different viewpoints in this matter, and what should the decisionmaking criterion be?
Solution:
In this instance assume that the owner of the automobile agreed to accept $52.80 per person for
the three riders, based on the variable costs that were purely incremental for the conference trip
versus the owners average annual mileage. That is, the $52.80 per person is the with a trip
cost relative to the without alternative.
Now, what would the situation be if the team, because of the low cost, returned and proposed
another 800-mile trip the following weekend? And what if there were several more such trips on
subsequent weekends? Quite clearly, what started out to be a small marginal (and temporary)
change in operating conditions from 15,000 miles per year to 15,800 miles soon would
become a normal operating condition of 18,000 or 20,000 miles per year. On this basis, it would
not be valid to compute the extra cost per mile as $0.198.
Because the normal operating range would change, the fixed costs would have to be considered.
A more valid incremental cost would be obtained by computing the total annual cost if the car
were driven, say, 18,000 miles, then subtracting the total cost for 15,000 miles of operation, and
thereby determining the cost of the 3,000 additional miles of operation. From this difference the
cost per mile for the additional mileage could be obtained. In this instance, the total cost for
15,000 miles of driving per year was 15,000 x $0.384 = $5,760. If the cost of service due to
increased depreciation, repairs, and so forth turned out to be $6,570 for 18,000 miles per year,
evidently the cost of the additional 3,000 miles would be $810. Then the corresponding
incremental cost per mile due to the increase in the operating range would be $0.27. Therefore,
if several weekend trips were expected to become normal operation, the owner would be on
more reasonable economic ground to quote an incremental cost of $0.27 per mile for even the
first trip. [3]
1.5
Sunk Cost
A cost incurred in the past that cannot be retrieved as a residual value from an earlier investment.
It is not an opportunity cost. In economics the sunk cost is equivalent to fixed cost in short-term
decision making.
A classic example of sunk cost involves the replacement of assets. Suppose that your firm is
considering the replacement of a piece of equipment. It originally cost $50,000, is presently
shown on the company records with a value of $20,000, and can be sold for an estimated $5,000.
For purposes of replacement analysis, the $50,000 is a sunk cost. However, one view is that the
sunk cost should be considered as the difference between the value shown in the company
records and the present realizable selling price. According to this viewpoint, the sunk cost is
$20,000 minus $5,000, or $15,000. Neither the $50,000 nor the $15,000, however, should be
considered in an engineering economic analysis except for the manner in which the $15,000
may affect income taxes.
1.6
Market Price
The market price is the price at which a good or service is actually exchanged for another good
or service (as an in kind payment) or for money (in which case it is a financial price) [2].
Example:
The market clearing price of electricity in a power pool is the price at which the most expensive
unit is dispatched to meet demand. The results from the Purdue power pool model gives a
pattern of expansions that occur if a tight power pool were to operate a power exchange, where
every hour, a market clearing price was set.
1.7
Shadow Price
Shadow price technically implies a price that has been derived from a complex mathematical
model (for example, from linear programming). See the discussion that follows in Chapter 2.4.
1.8
The annual payment that will repay a loan of 1 currency unit in n years with compound
interest on unpaid balance permits calculating equal installments necessary to repay (amortize)
a loan over a given period at a stated interest rate i. Such that: crf = i(1+i)n/[(1+i) n-1]
Chapter 2
Costing and Computing Concepts
Further to the general definitions, let us now consider more carefully some of these concepts.
2.1
TC(Q)
DRS
Q2
dTC(Q) dTVC(Q)
=
dQ
dQ
2.2
Marginal Costs are what are critical in decision making, not average
costs
, which maximizes
Example: Suppose that a company wishes to select the level of output, Q
profits.
such that P = dTC(Q) , or
Total profit = TR TC; max PQ TC(Q) Q
Q
dQ
such that Price = incremental cost.
Q
Note the irrelevance of average anything, except, after the fact, to indicate profit/unit, cost/unit,
price/unit.
Example:
Price = $35/unit, Cost = 50,000 + 20.2x + 0.0001x2
Total Cost
Total Revenue
TC
TR
50,000
x1
What x maximizes profit? max
x2
35 - 20.2 - 0.0002x = 0
14.8
= 0.0002 = 74,000
Questions:
Question:
A common average cost is depreciation. Equipment lasts 10 years, costs $10,000; want to cost it
on a yearly basis. Use straight-line depreciation of $1,000/year. While depreciation is important
for tax purposes, it is irrelevant for decision making, except as it affects after-tax profits.
Example:
Trying to decide which plant should produce a new product.
(a)
(b)
$1 x 106
$2 x 106
$3 x 106
$1.5 x 106
$1 x 106
$2.5 x 106
2.3
Sunk Costs: A cost irretrievably incurred in the past that cannot be altered by any action taken
from now on.
Examples:
Binding contractual agreement to purchase a specialized piece of equipment with
no salvage value.
I purchased IBM stock @ $130/share; it is now $80/share. Should the fact that I
purchased @ $130 enter into the decision to keep or sell the stock?
Question: What is the relation between fixed and sunk costs?
Fixed Costs:
Once incurred, remain invariant for all alternative courses of action under
consideration.
Examples:
Cost of a factory of fixed size prior to decision to construct
Suppose Im trying to figure out what type of assembly line to build: (A) an expensive, highly
automated one, or (B) a cheaper, less automated one.
A
Equipment: 1 x 106
Now: Before I choose: both equipment costs and operating costs are variable.
Which to choose depends on expected sales; < 1 x 106, choose B, > 1 x 106, choose A.
TC x 10 6
Plant B
Plant A
Q x 10 6
After I choose and construct, fixed costs become sunk costs to the extent that I cannot recover
equipment investment, and they become irrelevant for decision-making.
Question: Do all sunk costs arise from fixed costs?
Answer: No. Even fuel costs can be sunk costs, if a take or pay contract is signed.
2.4
10
LaGrange Multipliers
Demand theory is based on the premise that consumers maximize utility subject to a budget
constraint. Utility (U, the level of satisfaction that a person gets from consuming a good or
undertaking an activity) is assumed to be an increasing function of the quantities of goods
consumed, but marginal utility is assumed to decrease with consumption. The consumers
optimization problem when there are two goods, X and Y, may then be written as
Maximize U(X,Y)
(1)
subject to the constraint that all income is spent on the two goods:
PXX + PYY = I
(2)
Here, U( ) is the utility function, X and Y are the quantities of the two goods that the consumer
purchases, PX and PY are the prices of the goods, and I is income. (To simplify the mathematics,
we assume that the utility function is continuous (with continuous derivatives) and that goods are
infinitely divisible.) [3]
To determine the individual consumers demand for the two goods, we choose those values of X
and Y that maximize Equation (1) subject to Equation (2). When we know the particular form of
the utility function, we can solve to find the consumers demand for X and Y directly. however,
even if we write the utility function in its general form U(X,Y), the technique of constrained
optimization can be used to describe the conditions that must hold if the consumer is maximizing
utility.
To solve the constrained optimization problem given by Equations (1) and (2), we use the
method of Lagrange multipliers, which works as follows. We first write the LaGrangian for
the problem. To do this, rewrite the constraint in Equation (2) as: PXX + PYY I = 0. The
LaGrangian (L) is then:
L = U(X,Y) - (PXX + PYY I)
(3)
(4)
11
Here, MU is short for marginal utility (i.e., MUX(X,Y) = U(X,Y)/X, the change in utility from
a small increase in the consumption of good X).
The third condition is the original budget constraint. The first two conditions of Equation (4) tell
us that each good will be consumed up to the point at which the marginal utility from
consumption is a multiple of () of the price of the good. To see the implication of this, we
combine the first two conditions to obtain the equal marginal principle:
= [MUX(X,Y)/PX] = [MUY(X,Y)/PY]
(5)
2.5
(6)
Operations research is sometimes also called OR, Management Science, or decision analysis. In
the process of taking major project decisions there will be thousands and probably millions of
options available. Consider a few simple examples:
Consider the options available with the unit commitment problem for one coal fired generating
unit during one time period. List the options of when it is switched on and off.
Example:
Condition 1
Condition 2
On
0
1
Off
1
0
Option No:
Condition:
Unit 1
Unit 2
Unit 3
1
On/Off
0/1
0/1
0/1
2
On/Off
1/0
0/1
0/1
3
On/Off
0/1
1/0
0/1
4
On/Off
0/1
0/1
1/0
5
On/Off
1/0
1/0
0/1
6
On/Off
1/0
0/0
1/0
7
On/Off
0/1
1/0
1/0
12
8
On/Off
1/0
1/0
1/0
With this simple example, in one time period (say one hour), there are already 8 different options
available.
With 2 conditions and 3 units there are:
23 = 8 possible operating options available.
Example:
Consider the example above again but this time let there be two time periods called hour 1 and
hour 2.
In hour 1 there is option 1 and following in hour 2 there would be 8 options.
In hour 1 there is option 2 and following in hour 2 there would be 8 options.
In hour 1 there is option 3 and following in hour 2 there would be 8 options.
Etc. etc. . . . . .
In hour 1 there is option 8 and following in hour 2 there would be 8 options.
With a second time period being involved there are now 64 possible operating options to
consider. The complexity of the problem increases exponentially.
There are now 23 x 23 = 64 conditions.
26 = 64
In one day with 24 one hour time periods the number of operating options available will be equal
to:
23 x 24 = 272
272 = 4.722366483 x 1021
272 = 4,722,366,483,000,000,000,000
= 4,722 trillion trillion options
Thus a relatively simple problem can quickly involve an unmanageable number of options.
Imagine the size or complexity of the decision process for the unit commitment in an electricity
utility or power pool in which there are several or more power generating stations with scores of
units to be switched on and off over hourly periods for days and weeks.
It can be appreciated that a computer solver for real problems in decision analysis will be
indispensable as a manual procedure would be extremely long and prone to mistakes
effectively impossible to do. The GAMS and CPLEX solvers are used for this type of problem
(using branch and bound techniques).
2.6
13
Harare
Plants
Inga
HCB
Demands
(MWh)
1.6
0.3
700
SETS - Indices
i = plants,
Wheeling Distances
(Thousands of miles)
1.3
2.2
0.6
1.0
400
2500
Supplies
(MWh)
2100
1600
j = markets
ij
Hi
Objective Function
Minimize
CijXij
i
Dj
j Harare
j Lusaka
i Inga
14
j Pretoria
i HCB
I
J
Generation plants /
Demand Centers
/
Inga, HCB
/;
Harare, Lusaka, Pretoria
PARAMETER
/
H(I)
PARAMETER
/
D(J)
/;
TABLE L(I,J)
Distance in thousands of miles from I to J
Harare
Lusaka
Pretoria
Inga
1.6
1.3
2.2
HCB
0.3
0.6
1.1
;
SCALAR
/ 2 /;
PARAMETER C(I,J);
C(I,J) = W*L(I,J);
VARIABLE
VARIABLE
X(I,J)
Z
POSITIVE VARIABLE X ;
EQUATION
EQUATION
EQUATION
COST
SUPPLY(I)
DEMAND(J)
COST..
SUPPLY(I)..
DEMAND(J)..
Z =E= SUM((I,J),C(I,J)*X(I,J))
SUM(J,X(I,J)) =L= H(I)
;
SUM(I,X(I,J)) =G= D(J)
;
MODEL ELEC /
ALL
SOLVE
ELEC USING LP
DISPLAY X.L, X.M
/
MINIMIZING
;
Z
15
GAMS OUTPUT:
ITERATION COUNT, LIMIT
6
10000
Cplex 6.0, GAMS Link 12.0-7, 386/486 DOS
Optimal solution found.
Objective :
VAR X
Inga.Harare
Inga.Lusaka
Inga.Pretoria
HCB .Harare
HCB .Lusaka
HCB .Pretoria
10480.000000
Shipment quantities in MWh
LOWER
LEVEL
UPPER
.
.
+INF
.
400.000
+INF
.
1600.000
+INF
.
700.000
+INF
.
.
+INF
.
900.000
+INF
MARGINAL
0.400
.
.
.
0.800
.
The total shipment cost (minimized) for meeting the demand at the three markets amounts to
$10480. The optimal shipments are obtained by Inga sending out 400MWh to Lusaka and
1600MWh to Pretoria and by HCB sending out 700MWh to Harare and 900MWh to Pretoria.
2.7
Computing Requirements
The computing requirements to run the Purdue power pool models can be met with a new PC
(Pentium 3, etc.) or the latest laptop. The speed of the processor is important for an efficient use
of the model and it is recommended that at least 500MHz is used. A large memory is also an
important requirement. The Purdue Generic Seven Country Model is free. The regional models
that have already been tested by Purdue have confidential data populating them and for this
reason can not be distributed. Two commercial solvers are needed to run the model and these
are GAMS and CPLEX. More details can be obtained from the LT Model User Manual. A total
expenditure of about $16,000 will be adequate to fully install the system (hardware and
software) for the commercial user. Educational institutions are eligible for substantial cost
reductions with the solvers.
16
Chapter 3
Basic Electricity Modeling Formulation
3.1
In the short-run model I the objective is to minimize the total costs that arise from the cost of
operations (fuel and maintenance), distributed generation costs, and the cost of unserved MW.
i.e.: Minimizing over all hours, all stations, and all countries, the sum of fuel costs (cost/MW
times MW) plus demands met by distributed generation plus unsatisfied reserve requirements.
c(i, z)
PG(i,z,t)
DGcost
DG(z,t)
UMcost
UM(z)
All generation in country z plus all imports from other countries (adjusted for line loss) is equal
to the demand in country z plus exports to all countries.
PG ( i,z,t ) PGinit ( i,z )
PGinit(i, z)
The generation at station i, in country z, at any time t, is always less than or equal to the initial
generating capacity of that station i in country z.
17
The power flow from country z to country zp will always be less than or equal to the initial
power flow capability along the transmission line connecting country z to country zp.
PGinit ( i,z )
1 + res
i
( i,z )
res(i,z)
D(z,peak)
+ UM ( z ) D ( z,peak )
The sum of total capacity of all the plants in country z, derated for their reserve margins, plus the
unmet MW in country z will always exceed or be equal to the peak demand in country z.
A(z)
The sum total of initial generating capacities from stations i, in country z, will be greater than or
equal to the peak demand in country z times the autonomy of country z.
3.2
In the short-run models I and II the objective is to minimize the total costs that arise from the
cost of operations (fuel and maintenance), distributed generation costs, and the cost of unserved
MW.
min c(i, z)PG(i, z, t) + DG cos tDG(z, t) + UM cos tUM(z)
t
18
zp
zp
3.3
In the long-run model the objective is to minimize the total cost of operations (fuel,
maintenance), distributed energy (or unmet energy), unmet reserve, and the cost of capital (crf,
capital recovery factor) for capacity expansion. There is a time horizon of y years and a
discount rate in this model.
Y
min
1+disc
y=1
y =1 = y
(1+disc )
PGexp(i,z,y) = MW added in y at i in z
expcost(i,z)
disc
crf
= cost/MW of expansion at i in z
= discount rate for present value purposes
= capital recovery factor
19
The Model II load balance and PF equations with the y (yearly) variable added.
PG(i,z,t,y) < PGinit(i,z) + =1-y PGexp(i,z,)
The power generation at station i, in country z, at hour t, and in year y will be less than or equal
to the initial generating capacity at station i in country z plus the sum of all new expansions at
the station i for the years up to year y.
y
1+res ( i,z )
zp
Same as before, except total generating capacity now includes additions up to and including year
y.
iPGinit(i,z) +
t =1
The model is a cash flow model; cash outflows entered into the model in the year in which
they take place.
No need to collect data on sunk costs (costs of past investments, etc.), only incremental costs.
Model assumes equipment purchases financed by borrowed money hence equipment
purchase cost shows up as an annualized cost, equal to the capital recovery factor times the
Engineering, Procurement, and Construction (EPC) cost, in each year subsequent to the
purchase date.
Plant operating costs (fuel, variable O&M, water costs) should be average incremental costs
for each plant, not marginal costs which might be lower due to say, take or pay fuel
contracts. Ignore variable heat rates for existing thermal plants assume heat rate at 100%
load.
Plant equipment costs should be EPC costs, not including financing costs.
Fixed O&M ($/kW/yr) should be considered only for new plants; they are sunk costs for
existing plants (unless plants mothballed).
20
Reserve margins, autonomy factors, discount rate, crf, unserved energy and reserve costs are
policy decisions; get them, if you can but dont spend a lot of time.
Line losses should be average incremental, not marginal.
Line capacities should be maximum transfer capability, not maximum capacity.
Generation capacities should be net effective (dependable) sent out capacity, not nameplate
capacity.
Demands (D(z,t,y)) should be sent out demands, not received demands.
21
Chapter 4
The Generic Seven Country Regional Model
4.1
The generic seven-country model has been constructed for demonstration and training purposes.
Electricity demand patterns for each time period and for each country are primary data
requirements for populating the regional models. The electricity demand drives the model and
is further explained in Chapters 5 and 6. Taken into account are the load variations on an hourly,
daily, weekly, season, yearly, and national basis. The forecasting of annual growth rates in
demand needs special attention during the data collection process.
Table 4.1.1 Existing and Proposed Generation Stations
Country
Station
Details of Station
Name
Country1
PG(1A)
PG(1B)
NH(1C)
NH(1D)
GT(1E)
Country2
Country3
PG(2A)
PG(3A)
GT(3B)
Country4
PG(4A)
PG(4B)
CC(4C)
GT(4D)
Country5
PG(5A)
CC(5B)
Country6
H(6A)
NH(6B)
Country7
H(7A)
NH(7B)
22
The existing and proposed new generation and transmission infrastructure within this seven
country regional model is shown in Figures 4.1 and 4.2. Peak demand in Country1 is 3000MW
(Figure 4.1) and this country has an existing thermal generating capacity of 2800MW (station 1A
is 1200MW and station 1B is 1600MW). This country has a generation deficit of 200MW and
has proposed plans to construct a new hydropower station of 900MW (1C), a new hydropower
station of 600MW (1D), and a gas turbine station of 600MW (1E).
Country1 is interconnected with Country2 (Figure 4.2) and can therefore import electricity at
peak times. The proposed new generation and transmission project capacities are shown in
italics in Figures 4.1 and 4.2. Table 4.1.1 lists the names of the existing thermal generating
stations as PG(1A) and PG(1B) and shows the potentially new hydropower plant called
NH(1C). Similarly the existing and proposed international transmission lines from Country1 are
shown in Table 4.1.2.
Table 4.1.2 Existing and Proposed International Transmission Lines
From Country Interconnector
Details of International Interconnector
To Country
Name
1 to 2
OT(1-2)
2 to 3
OT(2-3)
3 to 4
OT(3-4)
4 to 5
NT(4-5)
5 to 6
NT(5-6)
6 to 2
NT(6-2)
6 to 7
NT(6-7)
7 to 1
NT(7-1)
Similar to Country 1 the information about the existing and proposed generation stations for each
country is summarized in Table 4.1.1. It is noted that there are no existing proposals for new
generating capacity in Country 2. Countries 4 and 5 both have excess generating capacity with
significant generation coming from combined cycle stations (using natural gas). Countries 6 and
7 are dominant hydropower countries and also have excess capacity but this time it is excess
23
hydropower. Both countries 6 and 7 have proposals for the construction of new further
hydropower generating capacity.
It is possible at present for Country 1 to import up to a maximum of 100MW from Country 2
(Figure 4.2). The existing line can have its existing load carrying capability increased up to a
maximum of 2000MW. There is also a proposed totally new international transmission line for
connecting Country 1 with Country 7. The initial capacity of this new line is 300MW. The
capacities of the existing international lines are shown for between countries 1, 2, 3, and 4, in
Figure 4.2. All of the existing and proposed new transmission lines can be expanded up to a
maximum load carrying capability of 2000MW.
Table 4.1.3
There are no options to construct natural gas pipe-lines to all parts of the region. Table 4.1.3
summarizes the status of existing and proposed maximum supplies of natural gas to the region.
The mix of fuels in the region is broad. Existing old thermal stations (1A, 1B, 2A, 3A and 4A)
can be described as either oil or coal fired. The fuel costs and heat rates will reflect the fuel
characteristics. Old thermal station 4B is the only existing combined cycle station using natural
gas. Station 4B can expand its capacity to use more gas and Country 5 also has the potential for
large combined cycle generation. The capital costs (fixed and variable) for generation expansion
will determine whether hydro, solid fuels or gas usage is to be further increased. The costs of the
interconnecting transmission lines will determine the ease or difficulty of trade, depending on the
capital fixed and variable costs for increased transmission capacity.
24
Figure 4.1 Training Model with Peak Demand (D) & Existing Generation (PG, CC, H) for Each
Country
Boundary of region
for power pool
Country 2
Country 3
Country 4
D = 1000
PG(4A) = 500
PG(4B) = 1200-2600
(CC(4C) = 300-2100
GT(4D) = 300)
D = 300
PG(3A) = 260
(GT(3B) = 600)
Country 5
D = 2000
PG(5A) = 2400
(CC(5B) = 350-2800)
D = 500
PG(2A) = 550
Country 6
D = 300
H(6A) = 600
(NH(6B) = 150-900)
Country 1
D = 3000
PG(1A) = 1200
PG(1B) = 1600-2500
(NH(1C) = 300-900
NH(1D) = 600
GT(1E) = 800)
Country7D = 400
H(7A) = 450
(NH(7B) = 200-600)
Figure 4.2 Training Model with Existing International Transmission Lines and Proposed New Lines
Boundary of region
for power pool
100
150
100
4
150
300
350
300
300
25
4.2
26
0
0
0
0
0
0
0
0
0
0
0
0
Scenario #4
No Trade
10 years
4% growth
18.14
11.13
2.42
1.59
2.86
955
3,553
2,342
0
900
3,111
1,479
427
833
2,031
0
0
There is a 23% saving in total regional costs when the countries adopt a free trade
policy (4% electricity annual demand growth rates in all countries) .
Free trade reduces the massive unmet reserve margin penalty costs.
Generation capacity increases by 16% for free trade.
Construction of 3,611MW of load carrying transmission capacity causes enormous
cost savings.
Net exporters: Countries 4, 5, 6, 7, 8
Net importers: Countries 1, 2, 3
Some of the benefits and projects (generation and transmission) optimally selected under
these three scenarios are summarized in section 4.3. The user-friendly PLTETM windows
interface facility provides some of the output results.
4.3.1 Consequences of trade with County 7 being a major exporter and County 1 being a
major importer. Note the revenues and costs to each country over the 10 year planning
horizon:
Figure 4.3.a
Figure 4.3.b
27
28
4.3.2 Construction of key international transmission lines takes place in periods 4 and ?
Figure 4.3.c Scenario 3 Old Transmission Expansion for Period 1 (2004-2005)
29
30
31
4.3.5 Example Output file for County 1 from the 10 year long-term model.
5 Periods.
Each Period =
2 years
32
33
==================================================================
NEW TRANSMISSION EXPANSION
___________________________________________________________________
Period |
Between
| Capacity Added
___________________________________________________________________
per2 | country1 and country7 |
499 MW
per5 | country1 and country7 |
47 MW
===================================================================
OLD TRANSMISSION EXPANSION
___________________________________________________________________
Period |
Between
| Capacity Added
___________________________________________________________________
per1 | country1 and country2 |
102 MW
per3 | country1 and country2 |
13 MW
per4 | country1 and country2 |
128 MW
per5 | country1 and country2 |
108 MW
===================================================================
B) MW RESERVES
========================================================================
(Generation Capacity + Firm Import Reserve - Firm Export Reserve = Peak
Demand)
----------------------------------------------------------------------Year
|
2005 |
2007 |
2009 |
2011 |
2013
(a)
PEAK DEMAND (MW)
(From input files)|
3245
3510
3796
(b)
PEAK LOAD CARRYING CAPABILITY (MW)
(Adjusted by Decay Rate & Reserve Margin)
Old Thermal
|
3103 |
3097 |
New Hydro
|
0 |
1133 |
Total
|
3103 |
4229 |
3091 |
1133 |
4223 |
(f)
TOTAL MW CAPACITY
[ (b) + (c) - (d) = (e) ]
|
3243 |
4106
4441
3084 |
1133 |
4217 |
3078
1155
4233
0 |
0 |
0 |
0 |
0 |
0 |
0
206
206
4223 |
4217 |
4439
(c)
FIRM IMPORT RESERVE (MW)
(Adjusted by line loss, forced outage rate)
Imports From:
country2
|
140 |
0 |
country7
|
0 |
0 |
Total
|
140 |
0 |
(d)
NONE
4229 |
(g)
TOTAL RESERVE MARGIN (%)
[(b)(Adjusted only by decay)+(c)(Adjusted only by line loss)-(d)(a)]/[(a)-(c)+(d)]=(f)
Total
|
18.9%|
40.5%|
29.7%|
19.7%|
16.5%
(i)
AUTONOMY FACTOR
[Generation Reserve(adjusted for FOR, Decay) /
Peak Demand]
Actual
| 1.036 | 1.307 | 1.207 | 1.114 | 1.034
Required
| 0.000 | 0.000 | 0.000 | 0.000 | 0.000
(j)
MAXIMUM IMPORTS (MW)
[Maximum hourly flow of energy during the year unadjusted]
Imports from:
country2
|
0 |
183 |
191 |
306 |
country7
|
0 |
558 |
557 |
556 |
(j)
MAXIMUM IMPORTS (MW)
[Time period from which the maximum import flow come from]
Maximum Import time:
FRM | YEAR | SEASON |
DAY
| HOUR |
MW
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
per2
per2
per2
per2
per2
per2
per2
per2
per2
per2
per2
per2
per2
per3
per3
per3
per3
per3
per3
per3
per3
per3
per3
per3
per4
per4
per4
per4
per4
per4
per4
per4
per4
per4
per5
per5
per5
per2
per2
per2
per2
per2
per2
per2
per2
per2
per3
per3
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
summer
summer
summer
summer
summer
summer
winter
winter
winter
winter
winter
winter
winter
summer
summer
summer
summer
summer
winter
winter
winter
winter
winter
winter
summer
summer
summer
summer
summer
winter
winter
winter
winter
winter
summer
summer
summer
summer
summer
summer
summer
summer
winter
winter
winter
winter
summer
summer
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
peak
peak
peak
peak
peak
average
peak
peak
average
average
average
average
average
peak
peak
average
average
average
peak
average
average
average
average
average
average
average
average
average
average
peak
average
average
average
average
peak
average
average
peak
average
average
average
average
peak
average
average
average
peak
average
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
hr9
hr19
hr20
hr21
avdy
hr20
hr19
avdy
hr9
hr19
hr20
hr21
avdy
hr19
avdy
hr9
hr20
hr21
avnt
hr9
hr19
hr20
hr21
avdy
hr9
hr19
hr20
hr21
avdy
avnt
hr9
hr19
hr21
avdy
avnt
hr19
avdy
avnt
hr9
hr19
hr20
hr21
avnt
hr9
hr19
avdy
avnt
hr9
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
182.8
182.8
182.8
182.8
182.8
182.8
182.8
182.8
182.8
182.8
182.8
182.8
182.8
191.0
191.0
191.0
191.0
191.0
191.0
191.0
191.0
191.0
191.0
191.0
306.5
306.5
306.5
306.5
306.5
306.5
306.5
306.5
306.5
306.5
401.1
401.1
401.1
557.9
557.9
557.9
557.9
557.9
557.9
557.9
557.9
557.9
556.8
556.8
34
401
608
cou
cou
cou
cou
cou
cou
cou
cou
cou
|
|
|
|
|
|
|
|
|
per3
per3
per3
per3
per3
per4
per4
per5
per5
|
|
|
|
|
|
|
|
|
summer
summer
summer
summer
winter
summer
summer
summer
summer
|
|
|
|
|
|
|
|
|
average
average
average
average
peak
average
average
average
average
|
|
|
|
|
|
|
|
|
hr19
hr20
hr21
avdy
avnt
hr9
avdy
hr19
avdy
|
|
|
|
|
|
|
|
|
35
556.8
556.8
556.8
556.8
556.8
555.7
555.7
607.6
607.6
(k)
MAXIMUM EXPORTS (MW)
[Maximum hourly flow of energy during the year unadjusted]
Exports to:
country2
|
196 |
192 |
201 |
287 |
165
(j)
MAXIMUM EXPORTS (MW)
[Time period from which the maximum export flow come from]
Maximum Export time:
TO | YEAR | SEASON |
DAY
| HOUR |
MW
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
cou
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
per1
per1
per1
per1
per1
per1
per1
per1
per1
per1
per1
per1
per1
per1
per1
per1
per1
per2
per2
per2
per2
per2
per2
per3
per3
per3
per4
per5
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
summer
summer
summer
summer
summer
summer
summer
winter
winter
winter
winter
winter
winter
winter
winter
winter
winter
summer
summer
summer
winter
winter
winter
summer
summer
summer
summer
winter
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
offpeak
peak
peak
peak
peak
peak
average
peak
peak
peak
peak
peak
average
average
average
average
average
offpeak
offpeak
average
offpeak
offpeak
offpeak
offpeak
offpeak
offpeak
offpeak
offpeak
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
avnt
hr9
hr19
hr20
hr21
avdy
hr20
hr9
hr19
hr20
hr21
avdy
hr9
hr19
hr20
hr21
avdy
avnt
hr19
avnt
hr9
avnt
avdy
hr9
avnt
avdy
avdy
avnt
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
196.3
196.3
196.3
196.3
196.3
196.3
196.3
196.3
196.3
196.3
196.3
196.3
196.3
196.3
196.3
196.3
196.3
192.4
192.4
192.4
192.4
192.4
192.4
201.1
201.1
201.1
286.7
165.0
========================================================================
C)
INCREMENTAL AND CUMULATIVE GENERATION CAPACITY
========================================================================
(a)
GENERATION CAPACITY (MW)
Incremental
Old Thermal
|
900 |
New Hydro
|
0 |
[Unadjusted]
(b)
GENERATION CAPACITY (MW)
Cumulative
[Unadjusted]
0 |
1246 |
0 |
2 |
0 |
2 |
0
27
Old Thermal
New Hydro
|
|
3700 |
0 |
3700 |
1246 |
3700 |
1248 |
36
3700 |
1251 |
3700
1278
(c)
NEW TRANSMISSION CAPACITY (MW)
[Unadjusted]
Incremental
country7
|
0 |
587 |
0 |
0 |
56
(d)
NEW TRANSMISSION CAPACITY (MW)
[Unadjusted]
Cumulative
country7
|
0 |
587 |
587 |
587 |
643
(e)
OLD TRANSMISSION CAPACITY (MW)
[Unadjusted]
Incremental
country2
|
102 |
0 |
13 |
128 |
108
(f)
OLD TRANSMISSION CAPACITY (MW)
[Unadjusted]
Cumulative
country2
|
202 |
202 |
215 |
343 |
451
========================================================================
D)
DEMAND/SUPPLY
========================================================================
(Energy Demand + Energy Exported + Energy Dumped = Energy Generation +
Energy Imported + Energy Unserved)
----------------------------------------------------------------------(a)
ENERGY DEMAND (MWh/Year)
[Yearly Total]
Local Demand
| 1.66E+7 | 1.80E+7 | 1.95E+7 | 2.11E+7 | 2.28E+7
Exports
| 1072285 | 578376 | 215287 | 207405 |
19594
Total
| 1.77E+7 | 1.86E+7 | 1.97E+7 | 2.13E+7 | 2.28E+7
(b)
ENERGY SUPPLY (MWh/Year)
Old Thermal
| 1.73E+7 |
New Hydro
|
0 |
Imports
|
0 |
Unserved Energy
| 418408 |
Total
| 1.77E+7 |
(c)
[Yearly
3224126 |
1.08E+7 |
4559889 |
22000 |
1.86E+7 |
Total]
4070822
1.08E+7
4772071
69032
1.97E+7
|
|
|
|
|
4719077
1.08E+7
5600568
173267
2.13E+7
|
|
|
|
|
5153559
1.08E+7
6534672
335391
2.28E+7
8.61 |
17.65 |
9.44
19.30
Station]
7095600 | 7095600 | 7095600
3679200 | 3679200 | 3679200
1.08E+7 | 1.08E+7 | 1.08E+7
[Yearly, by Station]
0.00 |
99.00 |
99.00 |
99.00 |
96.10
newh2
0.00 |
99.00 |
99.00 |
99.00 |
37
99.00
(D)
ENERGY IMPORTED (MWh/Year)
Imported From:
country2
|
0 | 449500 | 872305 | 1850751 | 2543801
country7
|
0 | 4110389 | 3899766 | 3749817 | 3990871
Total
|
0 | 4559889 | 4772071 | 5600568 | 6534672
(e)
ENERGY EXPORTED (MWh/Year)
Exported To:
country2
| 1072285 | 578376 |
Total
| 1072285 | 578376 |
215287 |
215287 |
207405 |
207405 |
19594
19594
========================================================================
E)
COST & REVENUES
========================================================================
(a)
FUEL COSTS ($/Year)
Undiscounted
Old Thermal
| 8.12E+8 | 1.53E+8 | 1.98E+8 | 2.34E+8 | 2.55E+8
Total Discounted Fuel Cost for Horizon = $2403593687
FUEL COSTS ($/MWh)
[Yearly fuel costs/yearly MWh generation]
Undiscounted
Old Thermal
|
46.91 |
47.37 |
48.62 |
49.53 |
49.54
(b)
O & M COSTS ($/Year)
Undiscounted
Old Thermal
| 1.12E+7 | 2201690 | 3153679 | 3970978 | 4340079
New Hydro
|
0 | 1.04E+7 | 1.04E+7 | 1.04E+7 | 1.05E+7
Total Discounted O & M Cost for Horizon = $83319607
(c)
WATER COSTS ($/Year)
Undiscounted
New Hydro
|
0 | 5387400 | 5387400 | 5387400 | 5387400
Total Discounted Water Cost for Horizon = $24846678
(d)
CAPITAL COSTS
Station
newh1 |
newh2 |
Station
newh1 |
newh2 |
2005 |
0
0
|
|
0
0
|
|
2007 |
2009 |
2011 |
2013 |
Total
| 2.7E+8
| 2.9E+8
| 1.5E+8
| 1.7E+8
0
0
| 2.4E+8
| 1.4E+8
38
| 1.1E+7
| 6.2E+6
2007
29876495|
2009
15647596|
2011
14181512|
2013
1078256
0|
0|
0|
0|
Year
|
2005 |
2007 |
Average Buying Price Present Value
|
0.00 |
22.89 |
Average Selling Price Present Value
|
47.14 |
27.19 |
Average Buying Price Real Dollars
|
0.00 |
30.51 |
Average Selling Price Real Dollars
|
51.93 |
36.24 |
2009
Year
|
2005 |
2007 |
Average Buying Price Present Value
|
65183|
0|
Average Selling Price Present Value
|
0|
0|
Average Buying Price Real Dollars
|
71810|
0|
Average Selling Price Real Dollars
|
0|
0|
0|
0|
2011
0
8
2013
20.22 |
18.19 |
15.72
38.25 |
35.99 |
28.96
32.61 |
35.50 |
37.12
61.70 |
70.24 |
68.40
2009
2011
2013
0|
0|
0|
0|
0|
0|
0|
0|
========================================================================
G)
OBJECTIVE FUNCTION BREAKDOWN
(Present Value)
========================================================================
39
Year
|
2005 |
2007 |
2009 |
2011 |
2013
FIXED COSTS
(a) Capital Costs
OT
| 98033328| 81019279| 66958082| 55337258| 45733271
LC
|
0|
0|
0|
0|
0
CC
|
0|
0|
0|
0|
0
CT
|
0|
0|
0|
0|
0
SC
|
0|
0|
0|
0|
0
Old H
|
0|
0|
0|
0|
0
New H
|
0| 295607305| 244791921| 202710978| 171048648
Pumped H |
0|
0|
0|
0|
0
New Line |
0| 11491288|
9496932|
7848704|
7103356
Old Line |
2228255|
1841533|
1712032|
2990355|
3570158
(b) Unserved MegaWatts
UM
|
0|
0|
0|
0|
0
(c) Fixed O&M
LC
|
0|
0|
0|
0|
0
CC
|
0|
0|
0|
0|
0
CT
|
0|
0|
0|
0|
0
SC
|
0|
0|
0|
0|
0
New H
|
0|
7476804|
6191529|
5127175|
4329596
Pumped H |
0|
0|
0|
0|
0
VARIABLE COSTS
(d) Fuel Costs
Old T
| 1.4733E+9| 229153561| 245421630| 239518781| 216214731
LC
|
0|
0|
0|
0|
0
CC
|
0|
0|
0|
0|
0
CT
|
0|
0|
0|
0|
0
SC
|
0|
0|
0|
0|
0
(e) Unserved Energy
Un En
| 106342797|
4621011| 11983610| 24858143| 39766516
(f) Water Costs
New H
|
0|
8083023|
6680185|
5520814|
4562656
Old H
|
0|
0|
0|
0|
0
(g) Variable O&M
Old T
| 20389071|
3303321|
3910450|
4069316|
3675666
LC
|
0|
0|
0|
0|
0
CC
|
0|
0|
0|
0|
0
CT
|
0|
0|
0|
0|
0
SC
|
0|
0|
0|
0|
0
New H
|
0|
8083023|
6680185|
5520814|
4562656
Old H
|
0|
0|
0|
0|
0
Old PH
|
0|
0|
0|
0|
0
New PH
|
0|
0|
0|
0|
0
Total
| 1.7003E+9| 650680150| 603826555| 553502337| 500567254
TOTAL COST FOR HORIZON =
4008854730.34
MWh Exp
MWh Imp
Res Exp
Res Imp
|
|
Total
0|
18285242|
0|
0|
0|
0|
0|
0|
0
8
4681236746.80
40
41
Chapter 5
Inputs and Outputs to the Model
From the perspective of the general user of the long-term planning model the actual model
with its formulation and coding can be treated with a black box. The general policy user
is to be concerned only with the data inputs and the resulting outputs (Figure 5.1). Detailed
descriptions of results from the Southern Africa are available [4-6].
Figure 5.1 General Users Approach to the Long-Term Model
Inputs
LT Model
Outputs
The long-term model has been tested extensively with the input data supplied by the
Southern African Power Pool (SAPP) and Figure 5.2 shows the input and output file names
as they were organized in June 2000. With each world region or power pool that is
modeled then the names of the output files will of course also change in order to report the
results for each new country that is in the model. The structure and number of the input files
will not change so significantly. One significant addition since June 2000 is the inclusion of
a natural gas sub-model to the LT electricity trading and expansion model.
A brief description of the functions of the input and output files follows. It is recommended
that the windows interface is used by the general user. This interface is described at the end
of the User Manual together with illustrations of the windows options that are available.
The use of the interface makes the model very user friendly and prevents the introduction of
errors from inexperienced editors of the basic model coding.
5.1
42
(1) June21.gms - Main program, contains all optimization constraints, optimizes model, no
changes will be made to this file.
Figure 5.2 The Files that Comprise the Southern African Long-Term - Model June 21 2000
Inputs
June21.gms
Angola.out
Outputs
June21.1st Projects.out
Data.inc
Botswana.out
Sixhr.inc
DRC.out
Projects.out
Lesotho.out
Hydro.inc
Trade.out
Malawi.out
Thermo.inc
Uncertain.inc
NMoz.out
Hyd_exp.out
SMoz.out
Namibia.out
Trans_exp.out
Lines_sapp.inc
NSA.out
Reserve.inc
Output.inc
Therm_exp.out
SSA.out
Swaziland.out
SAPP.out
Tanzania.out
Prices.out
Zambia.out
Zimbabwe.out
Flows.out
43
Data Files:
(2) Thermop.inc Contains data on the cost to expand new thermal stations, data on existing
capacities, maximum expansion of existing capacities, and the capital recovery factor on the
thermal stations.
(3) Lines_sapp.inc Contains cost of expanding new lines and cost of new lines. Loss of
energy due to resistance in old lines, loss of energy due to resistance in new lines, initial
capacity of new lines, capital recovery of new lines, and cost of additional capacity on new
lines.
(4) Hydro.inc Contains data on the cost to expand new hydro stations, data on existing
capacities, maximum expansion of existing capacities, and the capital recovery factor on the
hydro stations.
(5) Sixhr.inc Peak demand for each region: highest demand for one hour for current year.
(6) Uncertain.inc Contains: data on uncertainties (i.e. expected rainfall).
(7) Reserve.inc Contains: Autonomy factor self reliance of each country, reserve margin
for each country, forced outage rate for both transmission lines and for all plant types in
country, unforced outage rate for all plant types in country, and largest generator station for
each country.
(8) Data.inc Contains data on the demand growth, and domestic growth, which can be
changed by user.
(9) Output.inc Generates the output files which contain the necessary data used for
analysis.
Output Files:
(10) June21.lst Generic output file created by gams.
(11) Therm_exp.out Thermal expansion plans from running the model.
(12) Hyd_exp.out Hydropower expansion plans from running the model.
(13) Trade.out Trade quantities from running the model.
(14) Trans_exp.out - Transmission expansion plans from running the model.
(15) Projects.out All of the chosen projects are defined in this file.
(16) Country.out The expansion results as they pertain for each country, and SAPP as a
whole. (Angola.out, Botswana.out, etc.)
44
5.2
Consider how do we change the number of day types in each year? We do not actually
change the number of day types but we can change the weightings. The model has three
day types peak day, off-peak day, and average day. The total number of days must always
add up to 365. Weightings for the days are shown in Table 5.1.
The SAPP model uses a 25:75 weighting for the winter and summer seasons. There are
three types of days; peak day, average day, and off-peak day. There are12 average night
hours and 12 day hours; 8 average hours, and 4 peak hours.
Season
Summer
Summer
Winter
Winter
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Winter
Winter
Winter
Winter
Winter
Winter
Winter
Winter
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Winter
Winter
Winter
Winter
Winter
Winter
Winter
Winter
Season
SAPP Winter
SAPP Summer
Day
Peak
Average
Offpeak
Hour
Avnt
Hr9
Avdy
Hr19
Hr20
Hr21
Day
Average
Average
Average
Average
Peak
OffPeak
Peak
OffPeak
Average
Average
Average
Average
Peak
OffPeak
Peak
OffPeak
Average
Average
Average
Average
Peak
Peak
Peak
Peak
OffPeak
OffPeak
OffPeak
OffPeak
Peak
Peak
Peak
Peak
OffPeak
OffPeak
OffPeak
OffPeak
Weights
Hour Season Day
Avdy 0.75
260
Avnt
0.75
260
Avdy 0.25
260
Avnt
0.25
260
Avdy 0.75
52
Avdy 0.75
52
Avnt
0.75
52
Avnt
0.75
52
Hr9
0.75
260
Hr19
0.75
260
Hr20
0.75
260
Hr21
0.75
260
Avdy 0.25
52
Avdy 0.25
52
Avnt
0.25
52
Avnt
0.25
52
Hr9
0.25
260
Hr19
0.25
260
Hr20
0.25
260
Hr21
0.25
260
Hr9
0.75
52
Hr19
0.75
52
Hr20
0.75
52
Hr21
0.75
52
Hr9
0.75
52
Hr19
0.75
52
Hr20
0.75
52
Hr21
0.75
52
Hr9
0.25
52
Hr19
0.25
52
Hr20
0.25
52
Hr21
0.25
52
Hr9
0.25
52
Hr19
0.25
52
Hr20
0.25
52
Hr21
0.25
52
Hour
12
8
12
8
12
12
8
8
1
1
1
1
12
12
8
8
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Total Hours
2340
1560
780
520
468
468
312
312
195
195
195
195
156
156
104
104
65
65
65
65
39
39
39
39
39
39
39
39
13
13
13
13
13
13
13
13
8736
Percent Cumm.
26.79% 26.79%
17.86% 44.64%
8.93%
53.57%
5.95%
59.52%
5.36%
64.88%
5.36%
70.24%
3.57%
73.81%
3.57%
77.38%
2.23%
79.61%
2.23%
81.85%
2.23%
84.08%
2.23%
86.31%
1.79%
88.10%
1.79%
89.88%
1.19%
91.07%
1.19%
92.26%
0.74%
93.01%
0.74%
93.75%
0.74%
94.49%
0.74%
95.24%
0.45%
95.68%
0.45%
96.13%
0.45%
96.58%
0.45%
97.02%
0.45%
97.47%
0.45%
97.92%
0.45%
98.36%
0.45%
98.81%
0.15%
98.96%
0.15%
99.11%
0.15%
99.26%
0.15%
99.40%
0.15%
99.55%
0.15%
99.70%
0.15%
99.85%
0.15% 100.00%
100.00%
0.25
0.75
52
260
52
8
1
12
1
1
1
45
46
Chapter 6
Electricity Policy Analysis Scenarios
6.1
Short-term (ST) modeling can be for almost any length of time less than 12 months. It can
be a period of hours, days, weeks, or months. Long-term (LT) modeling is normally
referring to several years. LT models are typically anywhere between 5 years and 20 years.
6.2
The state Governments creation of Purdues State Utility Forecasting Group (SUFG) was to
ensure that the State of Indiana had a dependable, accurate, and impartial assessment of the
demand for electricity in Indiana and the Mid-West USA (1985, Senate Bill 29). These
forecast figures for growth in electricity demand are vitally important for any utility or state.
Across the United States and in the industrialized national generally a growth rate of about
2% is typical. In the developing economies the growth rates are often quoted as being
double or triple this 2% growth rate and even more. A policy of needing to accurately
forecast the growth rate is especially important for the electricity market modeling work.
Very different long-term consequences occur when a 4% increase to growth is made instead
of a 3%. It can be appreciated therefore what the consequences might be when an 8% or
12% value is given.
Over a 20 year planning horizon, with a constant growth rate of 2% per year, there is a
compounded total increase in growth of 48%. Note the enormous differences in total
expected growth for a 20 year planning horizon when rates of 4%, 8%, and 12% are used:
1.0220 = 1.48, 1.0420 = 2.19, 1.0820 = 4.66, 1.1220 = 9.64
6.3
The Purdue long-term electricity trade model (PLTETM) trades in two commodities:
(a) Megawatt reserve power (MW)
(b) Megawatt hour energy (MWh)
During each hour modeled the amounts of energy or reserve power traded will vary. This
takes place for each hour over the total planning horizon (typically this can be 10 years).
Most utilities will be interested in trading MW but in situations where a utility cannot meet
its production demand, or where much cheaper supplies are available (hydropower etc) then
MWh are traded.
6.4
47
In many situations, especially in the developing economies, the power supply, in a certain
hour might not be able to meet the demand. Should there be some policy regarding unmet
reserve MW (UM) such as a penalty charge for not being able to meet demand? In the
industrialized counties there is normally a legal requirement that reserve margins must
always be securely maintained. Power shortage can be very costly to an economy and so the
PLTETM places a penalty charge of $2 million per MW if a country fails to meet its daily
peak MW demand.
The policy towards distributed generation (DG) is also represented in the PLTETM. If the
daily production level of electricity (for each hour of each day) cannot be met then it is
assumed that small stand-by generators (DG) will be available. This is particularly relevant
for rural areas with intermittent supplies and in developing economies when there is a short
fall. The estimated cost for purchasing and operating these small generators has been
estimated at 14 cents per kWh ($140 per MWh). This is the cost given in the model for
unserved energy (UE).
6.5
The PLTETM can be used as a short-term model by limiting the length of the planning
horizon. Typically it is used as a 10 year model (5 time periods with each period being 2
years long or 10 periods with each period being 1 year long). If however only a 1 year
planning horizon is specified then there will be no capacity expansions taking place and the
model will effectively become a short-term or dispatch model only. The amount of trading
taking place (using a cost minimization objective) will be subject to a demand constraint:
Such that over time: Generation + Imports + DG = Demand Exports
{ie:
The cost of trade is initially met by the supply generator in the model but in the final
analysis the model will also allocate the gains from trade, between the exporting and
importing country, on a equal basis (50% - 50%). The exporting country therefore obtains
revenues (post optimality) for the exports and the importing country makes costs savings
from not generating with its own more expensive generators. The present default
arrangement with the PLTETM is such that a trade tariff of 6 cents/kWh will take place
when the marginal cost of the exporting country is 2 cents/kWh and the marginal cost of the
importing country is 10 cents/kWh.
Trade Tariff = Marginal cost {(Exporter cost + Importer cost)/2}
6.6
48
The reliability policy for supplies in electricity is incorporated into the model with two
parameters which specify:
A reserve margin for thermal power generation
A reserve margin for hydropower generation
Each country reserve parameter may be given a specific value depending upon the level of
reliability required in that country or utility. In coordinated centrally controlled power pools
the reliability of supplies can be much more cost effective than with several isolated utilities.
With economies of scale and centralized coordination a single reserve unit can provide
service to several utilities. Reliability policy can therefore be significantly proved to be
much cheaper within a coordinated power pool. Typically a reserve margin value of 10% is
used in hydropower generation and a value of 19% is used in thermal generation.
The reliability policy is implemented such that:
Generation /(1+ Reserve Margin)
6.7
>
Peak Demand
With international electricity trading, or just trading between utilities, a decision must be
taken at some stage regarding the level of dependency that there is to be on the amount of
purchases that are to take place. Considerable attention will be given to this policy as it will
affect the planning for new capacity and the type of trading contract that is agreed upon
between the buyer and seller. The PLTETM represents this important trade policy issue
with two parameters that are called autonomy factors. There is an autonomy factor for
trading reserves and another one for trading energy:
Reserves trading of MW, autonomy factor AF
Energy trading of MWh, autonomy factor ENAF
The autonomy factors are implemented such that:
Generation
>
If an autonomy factor is set at 100% independence (AF=1.0) then this will mean that a
policy requirement will exist that insists that the country must, at all times, be capable of
meeting all of its own electricity demand.
6.8
Each utility and power pool will develop its own strategy and policy for attracting
investment. This becomes even more important when a power system becomes less
controlled by government constraints and becomes more subject to operating under market
forces. A significant attraction to the power pool structure is that it provides increased
49
stability and a reduction in risk for interested investors. The construction therefore of new
capacity (generation or transmission) becomes less costly when borrowing capital under the
power pool infrastructure with central coordination. New projects are more likely to be
implemented at an earlier date.
The PLTETM uses a capital recovery factor (crf) to reflect this policy concern. When $1
million is borrowed over a 20 year time horizon with a 12% interest rate then the crf is 0.133
but when borrowed over the same time period with a 10% interest rate the crf is 0.117.
This means that the annual amortized payment with the 12% rate is $0.133 million and with
the 10% rate is $0.117 million.
6.9
In many parts of the world when there is a drought then hydropower reservoirs become
severely depleted and can dramatically reduce the amount of electrical generation that is
available. The PLTETM allows for drought scenarios to take place in any chosen area and
time period. This shows the consequences of low water inflows and the extent to which
spare reserve capacity will be needed or the extra electricity imports that will be called
upon. The drought factor in the model reduces the amount of energy that is stored behind
the reservoir dam wall such that:
Energy in Reservoir = Drought Factor x Normal Energy level in Reservoir
6.10
Where a third party country is involved for wheeling electricity from Country A to Country
C then a wheeling policy will be needed.
Country A
Country B
Country C
In the existing PLTETM there is an equal sharing in the gains from trade between any two
countries. This means that Country B presently will receive too high a revenue from being a
wheeling node. (The supply and demand all take place at the same point or node that
represents each country or utility, in the present model. The model is being improved to
allow model users to specify more than one node to represent each country).
50
51
Appendix 1
Template Data Collection Sheets
Supply, Demand, and Shipment (Existing & Proposed)
The Purdue electricity and gas trade model optimizes the minimum cost to meet the demands
for electricity and natural gas within one region over a long-term horizon (e.g., 20 years).
The region consists of several or more countries (indexed as z or zp). Normally each country
is modeled as one node. Free trade is permitted to take place between all of the countries in
the specified region. The total demand and supply of energy (electricity and natural gas) has
to be known for each node/country. The shipping capacity (of electricity and natural gas)
between any two nodes/countries has to be known. Data for the existing and potentially new
supply points (new generation stations, new gas wells, new transmission and new gas
pipelines) are all needed. The existing demand at each node and the forecast for electricity
growth in demand is required. The percentage of the natural gas supplied to each node for
electricity generation and the percentage for other needs are required for each year in the
model. (More than one node for each country can be created if shown to be necessary).
52
A2
MW
1998
1999
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
B
GWh
1998
1999
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
31
32
33
34
35
336
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
Hourly Data (MW) for a Representative Week, in the most recent year
(24 x 7 = 168 values)
Year: ,
Hour
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Sun
Week Number:
Mon
Thurs
Fri
Sat
8760 Hour Load (MW) Data File for the most recent year
As an alternative to B and C this one year 8760 (52 x 7 x 24) hour data file can
be supplied. Please attach the appropriate sheets with all 8760 hours of data.
Hour
1
2
3
etc
Load
(MW)
Hour
8758
8759
8760
Load
(MW)
54
Station Name:
Plant Type: ..
(Coal, oil fired combustion turbine, gas fired combustion turbine, gas fired combined cycle)
Operating Status: .
(in operation, shut-down, mothballed, other)
Number of Units: .
Comment
Value
Parameter
PGOinit
Oexpcost
PGOexpstep
PGOmax
FORPGO
UFORPGO
crfi
VarOMoh
HRO
fpO
fpescO
decayPGO
PGmin
Fdecom
55
Station Name:
Plant Type:
..
Operating Status:
Number of Units: .
.
Load type:
Comment
Value
Parameter
Hoinit
HOVcost
Hoexpstep
HOVmax
HOLF
FORoh
Crfih
VarOMoh
DecayHO
Reshyd
MinH
FdecomH
56
Station Name:
Plant Type:
Operating Status: ..
Number of Units: .
Comment
Load type: .
Value
Parameter
FGSC
NSCexpcost
NSCexpstep
PGNSCmax
FORNSC
UFORNSC
Crfni
OMSC
FixOMSC
HRNSC
FpNSC
13. Escalation rate of fuel cost of new small coal plants (fraction/year)
FpescNSC
DecayNSC
AtSC
BefSC
AftSC
MinSC
57
Value
Parameter
FGLC
PGNLCinit
NLCexpcost
NLCexpstep
PGNLCmax
FORNLC
UFORNLC
Crfni
OMLC
FixOMLC
HRNLC
FpNLC
FpescNLC
DecayNLC
AtLC
BefLC
18. Large coal MUST NOT be built BEFORE or AT period ty. 0 if unconstrained
AftLC
MinLC
58
Value
Parameter
FGGT
Ntexpcost
Ntexpstep
PGNTmax
FORNT
UFORNT
Crfni
OMT
FixOMT
HRNT
FpNT
FpescNT
DecayNT
AtT
BefT
AftT
MinT
59
FGCC
NCCexpcost
NCCexpstep
PGNCCinit
PGNCCmax
FORNCC
UFORNCC
Crfni
OMCC
FixOMCC
HRNCC
FpNCC
14. Escalation rate of fuel cost of new combined cycle plants (fraction/year)
FpescNCC
DecayNCC
AtCC
BefCC
AftCC
MinCC
60
Station Name:
Plant Type:
..
Operating Status:
Number of Units: .
.
Load type:
Comment
Value
Parameter
Hninit
HNFcost
HNVcost
HNVmax
Hnexpstep
HNLF
FORnh
Crfnh
FixOMnh
VarOMnh
DecayHN
AtHn
BefHn
AftHn
MinHN
61
Value
Parameter
PFOinit
2. Type, AC or DC
3. Cost per MW of expanding existing line in (mill $)
PFOVc
crf
PFOloss
PFOVmax
FORICO
DecayPF
O
minPFO
62
Value
Parameter
PFNinit
2. Type, AC or DC
3. Line Voltage (kV)
4. Route length (km)
5. Capital recovery factor for transmission lines (fraction/year)
Crf
PFNFc
PFNVc
PFNVmax
PFNloss
FORICN
13. Annual unforced outage rate for new transmission line (%)
14. Or Annual maintenance duration (weeks per year)
15. Decay rate of new lines (fraction/year)
DecayPFN
MinPFN
Atlines
Aftlines
Beflines
63
Station Name:
Plant Type:
..
Operating Status:
Number of Units: .
.
Load type:
Comment
Value
Parameter
Decay PHO
PSOloss
PGPSOinit
64
Station Name:
Plant Type:
..
Operating Status:
Number of Units: .
.
Load type:
Comment
Value
Parameter
DecayPHN
PSNloss
PHNinit
HDPSNmw
h
Crfphn
FixOMph
VarOMph
Phncost
65
Station Name:
Load type: .
Comment
Value
Parameter
PGOinit
Oexpcost
PGOexpstep
PGOmax
FORPGO
UFORPGO
Crfi
VarOMoh
HRO
FpO
FpescO
DecayPGO
Pgmin
Fdecom
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
66
67
Country (Node): ..
Value
Parameter
GasInit(z,gw)
GWCost(z,gw)
Countries (z,zp):
Value
Parameter
PipeCap(z,zp)
GSpCst(z,zp)
InvCst(z)
PFGloss(z,zp)
68
69
Parameter
PipeCap(z,zp)
GSpCst(z,zp)
InvCst(z)
PFGloss(z,zp)
PipeCost(z,zp)
Crfg(z,zp)
70
Appendix 2
Modeling Notation
LT-Model Notation (April 14, 2000)
Definition-Comment
DecayHN
DecayHO
DecayNCC
DecayNLC
DecayNSC
DecayNT
DecayPFN
DecayPFO
DecayPGO
DecayPHN
DecayPHO
dgr(z,ty)
dgrowth1(z)
dgrowth2(z)
dgrowth3(z)
dgrowth4(z)
dgrowth5(z)
dgrowth6(z)
dgrowth7(z)
dgrowth8(z)
dgrowth9(z)
dgrowth10(z)
disc
DLC(z)
DW
Dyr(ty,ts,td,th,z)
E
Enaf(z,ty)
F
fdrought(ty,z)
Fdecom(z,i)
FdecomH(z,ih)
FGCC(z,ni)
FGLC(z,ni)
FixOMCC(z,ni)
FixOMLC(z,ni)
fixOMnh(z,nh)
fixOMph(z,phn)
FixOMSC(z,ni)
FixOMT(z,ni)
Fmax(ty,zp,z)
Fmax(ty,z,zp)
FORICN(z,zp)
FORICO(z,zp)
FORNCC(z,ni)
FORnh(z,nh)
FORNLC(z,ni)
FORNSC(z,ni)
FORNT(z,ni)
FORoh(z,ih)
FORPGO(z,i)
fpescNCC(z)
71
fpescNLC(z)
fpescNSC(z)
fpescNT(z)
fpescO(z, i)
fpNCC(z, ni)
fpNLC(z,ni)
fpNSC(z,ni)
fpNT(z,ni)
fpO(z,i)
H
H(ty,ts,td,th,z,ih)
HA(ty)
HDPSNmwh(z,phn)
HDPSOmwh(z)
HNcapcost(ty)
Hnew(ty,ts,td,th,z,nh)
HNexpstep(z,nh)
HNFcost(z,nh)
HNinit(z,nh)
HNLF(z,nh)
HNVcost(z,nh)
HNVexp(ty,z,nh)
HNVexp(tye,z,nh)
HNVmax(z,nh)
HOcapcost(ty)
HOexpstep(z,ih)
HOinit(z,ih)
HOinitty(z,ih,ty)
HOLF(z,ih)
HOVcost(z,ih)
HOVexp(ty,z,ih)
HOVexp(tye,z,ih)
HOVmax(z,ih)
HOVmaxTY(z,ih,ty)
HRNCC(z,ni)
HRNLC(z,ni)
HRNSC(z,ni)
HRNT(z,ni)
HRO(z,i)
72
Escalation rate of fuel cost of new large coal plants (fraction per year).
Escalation rate of fuel cost for new small coal plants (fraction per year).
Escalation rate of fuel cost for new gas turbines plants (fraction per year).
Escalation rate of fuel cost of existing thermal plants (fraction per year).
Fuel cost of new combined cycle plants ($/million BTU).
Fuel cost of new large coal plants ($/million BTU).
Fuel cost of small coal plants ($/million BTU).
Fuel cost of new gas turbine plants ($/million BTU).
Fuel cost of existing thermal plants ($/MWh).
Generating level of existing hydro plants (MW)[variable].
n times period ty (HA = n).
New pumped storage hydro reservoir volume capacity (MWh per day).
Existing pumped storage hydro reservoir volume capacity (MWh per day).
Construction cost of a new hydro plant ($).
Output for new hydro plants (MW) [variable].
Expansion step for new hydro stations (MW).
Fixed capital cost of new hydro stations ($).
Initial capacity of new hydro stations (MW).
Annual generation limit for new reservoir (GWh/year).
Capital cost of additional capacity to new hydro stations ($/MW).
Number of units of the given expansion step size installed in ty for new hydro
plants [integer or continuous variable].
Number of units of the given expansion step size installed in tye for new hydro
plants [integer or continuous variable].
Maximum MW expansion added to a new hydro station (MW).
Expansion cost for existing hydro plants ($) [variable].
Expansion step for existing hydro (MW).
Initial capacity of an existing hydro station (MW).
Initial capacity of an existing hydro station in ty (MW).
Annual generation limit for existing reservoir (MWh/year).
Capital cost of additional capacity for existing hydro stations ($/MW).
Number of units of the given expansion step size installed in ty for existing
thermal plants [integer or continuous variable].
Number of units of the given expansion step size installed in tye for existing
thermal plants [integer or continuous variable].
Maximum MW expansions that can be added to an existing hydro station
(MW).
Maximum MW expansions that can be added to an existing hydro station in ty
(MW).
Heat rate of a new combined cycle plant (million BTU/MWh).
Heat rate of a new large coal plant (million BTU/MWh).
Heat rate of new small coal plants (million BTU/MWh).
Heat rate of a new gas turbine plant (million BTU/MWh).
Heat rate of existing thermal plants (million BTU/MWh); set equal to 1, since
fuel cost for old plants is expressed in ($/KWh).
I
i
ih
J
j
L
LM(z,th)
M
maxfor(zp,z)
maxloss(zp,z)
Mday(td)
minCC(z,ni)
minH(z,ih)
minHN(z,nh)
minLC(z,ni)
minSC(z,ni)
_int(z,ni)
Mperiod(ty)
Mseason(ts)
Mtod(th)
N
n
NCCexpcost(z,ni)
NCCexpstep(z,ni)
nh
ni
NLCexpcost(z,ni)
NLCexpstep(z,ni)
NSCexpcost(z,ni)
NSCexpstep(z,ni)
NTexpcost(z,ni)
NTexpstep(z,ni)
O
Oexpcost(z,i)
OMCC(z,ni)
OMLC(z,ni)
OMO(z,i)
OMSC(z,ni)
OMT(z,ni)
ord(ni)
ord(ty)
ord(tya)
73
ord(tyb)
ord(tye)
ord(z)
P
PeakD(z)
PF(ty,ts,td,th,z,zp)
PF(ty,ts,td,th,zp,z)
PFNcapcost(ty)
PFnew(ty,ts,td,th,z,zp)
PFnew(ty,ts,td,th,zp,z)
PFNFcost(z,zp)
PFNinit(z,zp)
PFNloss(zp,z)
PFNVcost(z,zp)
PFNVexp(ty,z,zp)
PFNVexp(tye,z,zp)
PFNVmax(z,zp)
PFOcapcost(ty)
PFOinit(z,zp)
PFOloss(zp,z)
PFOVcost(z,zp)
PFOVexp(ty,z,zp)
PFOVexp(tye,z,zp)
PFOVmax(z,zp)
PG(ty,ts,td,th,z,i)
PGmin(z,i)
PGNcapcost(ty)
PGNCC(ty,ts,td,th,z,ni)
PGNCCexp(tyb,z,ni)
PGNCCinit(z,ni)
PGNCCmax(z,ni)
PGNLC(ty,ts,td,th,z,ni)
PGNLCexp(tyb,z,ni)
PGNLCinit(z,ni)
PGNLCmax(z,ni)
PGNSC(ty,ts,td,th,z,ni)
PGNSCexp(ty,z,ni)
PGNSCexp(tyb,z,ni)
PGNSCexp(tye,z,ni)
PGNSCmax(z,ni)
PGNT(ty,ts,td,th,z,ni)
PGNTexp(ty,z,ni)
74
PGNTexp(tyb,z,ni)
75
Number of units of the given expansion step size installed in tyb for new gas
turbine plants [integer or continuous variable].
PGNTexp(tye,z,ni)
Number of units of the given expansion step size installed in tye for new gas
turbine plants [integer or continuous variable].
PGNTmax(z,ni)
Maximum MW that can be added to a new turbine plant (MW).
PGOcapcost(ty)
Expansion cost of all existing thermal plant ($)[variable].
PGOexp(tyb,z,i)
Expansion of existing thermal plants in tyb [variable].
PGOexpstep(z,i)
Expansion step size for existing thermal plant units (MW).
PGOinitTY(z,i,ty)
Current capacity for existing thermal plants in ty (MW).
PGOmax(z,i)
Maximum MW that can be added to an existing thermal plant (MW).
PGPSN(ty,ts,td,th,z,phn) Electricity production level of a new pumped storage plant (MW) [variable].
PGPSO(ty,ts,td,th,z)
Electricity production level of an existing pumped storage plant (MW)
[variable].
PGPSOinit(z)
Existing pumped hydro capacity (MW).
phn
Indice for proposed new pumped hydro.
PHNcapcost(ty)
Cost of new pumped storage installed in ty ($)[variable].
PHNFcost(z,phn)
Pumped hydro fixed capital cost ($).
PHNinit(z,phn)
Initial capacity of proposed new pumped hydros (MW).
PSNloss(phn)
New pumped storage loss coefficient (fraction).
PSOloss
Existing pumped storage loss coefficient (fraction).
PUPSN(ty,ts,td,th,z,phn) Electricity consumption level of a new pumped storage plant (MW)
[variable].
PUPSO(ty,ts,td,th,z)
Electricity consumption level of an existing pumped storage plant (MW)
[variable].
R
reshyd(z)
Reserve margin of hydro plants for each country (fraction).
resthm(z)
Reserve margin of thermal plants for each country (fraction).
T
td
Indice for time in days (off-peak, average, peak).
th
Indice for the time in hours (hr9, avnt, hr19, hr20, hr21, avdy).
ts
Indice for the time in seasons (summer, winter).
ty
Indice for the period.
tya
Alias of ty.
tyb
Alias of ty.
tye
Alias of ty.
U
UE(ty,ts,td,th,z)
Unserved energy (MWh) [variable].
UEcost
Cost of unserved energy ($/MWh).
UFORNCC(z,ni)
Unforced outage rate for new combined cycle plants (fraction).
UFORNLC(z,ni)
Unforced outage rate for new large coal plants (fraction).
UFORNSC(z,ni)
Unforced outage rate for new small coal plants (fraction).
UFORNT(z,ni)
Unforced outage rate for new gas turbine plants (fraction).
UFORPGO(z,i)
Unforced outage rate for existing thermal plants (fraction).
UM(z,ty)
Unmet reserve requirement for country z in ty (MW) [variable].
UM(z,tye)
Unmet reserve requirement for country z in tye (MW) [variable].
UMcost
Cost of unmet reserve requirements ($/MW).
V
VarOMoh(z,ih)
O&M variable cost for old hydro ($/MWh).
VarOMnh(z,nh)
VarOMph(z,phn)
W
wcost(z,ty)
Y
YCC(ty,z,ni)
YCC(tya,z,ni)
YCC(tye,z,ni)
Yh(ty,z,nh)
Yh(tye,z,nh)
YLC(ty,z,ni)
YLC(tya,z,ni)
YLC(tye,z,ni)
Yper(ty)
Ypf(ty,z,zp)
Ypf(ty,zp,z)
Ypf(tye,z,zp)
Yph(ty,z,phn)
Yph(tye,z,phn)
Z
z
zp
76
77
Bibliography
[1]
[2]
[3]
[4]
F.T. Sparrow, Brian H. Bowen, Zuwei Yu, Modeling Long-Term Capacity Expansion
Options for the Southern African Power Pool (SAPP), Proceedings of the IASTED
International Conference, Power and Energy Systems, Las Vegas, Nevada, November 810, 1999.
[5]
Zuwei Yu, F.T. Sparrow, Brian H. Bowen, A New Long Term Hydro Production
Scheduling Method for Maximizing the Profit of Hydroelectric Systems, IEEE
TRANSACTIONS on Power Systems, Volume 13, Number 1, February 1998.
[6]
All the Southern African Power Pool (SAPP) proposals and reports can be downloaded
from:
http://engineering.purdue.edu/IIES/PPDG/SAPP/