Você está na página 1de 14

Energy Vol. 17, No. 8, pp. 735-747, 1992 . 0360-5442/92 $5.00 + 0.

00
Printed in Great Britain. All rights reserved Copyright © 1992 Peigamon Press Ltd

ENVIRONMENTAL CONSEQUENCES OF
.HYDROELECTRIC DEVELOPMENT: THE ROLE OF
FACILITY SIZE AND TYPE
PETER H. GLEICK
Global Environment Program, Pacific Institute for Studies in Development, Environment, and
Security, 1681 Shattuck Avenue, Suite H, Berkeley, CA 94709, U.S.A.

(Received 28 October 1991)

Abstract—The development of hydroelectric power throughout the world is receiving


renewed attention as the economic, political, and environmental costs of conventional
energy production rise. There is currently a perception that hydro electricity has environ-
mental and economic advantages over electricity produced by conventional energy
technologies, but there is a paucity of information about the environmental impacts of
hydroelectric facilities as a function of size and type. We characterize the environmental
impacts of hydroelectric developments and quantify these impacts as a function of the size
and type of project. Several unexpected conclusions arise from our analysis. For most
hydroelectric facilities, size, as measured by installed capacity, is not necessarily a good
indicator of the severity of environmental costs. For impacts such as land flooded and
evaporative water lost, smaller facilities cause greater environmental disruptions per unit
of energy produced than do larger facilities. A more striking conclusion, however, is that
differences m the type of facility, as indicated by the relationship between dam height and
gross static head, are often far more important from an environmental perspective than are
differences in the installed electrical capacity of a facility. Another major conclusion is that
the development of hydroelectric facilities (independent of their size) such as dams at new
sites and dams operated to produce peak power are often, accompanied by environmental
and ecological disruptions comparable to or exceeding those of conventional non-
hydroelectric energy facilities. These results suggest that there is no justification to
expedite licensing procedures for hydroelectric facilities smaller than some arbitrarily
chosen installed capacity, as currently permitted by' some laws. Appropriate policies
should emphasize the development of dams on the basis of other favorable physical,
chemical, and biological characteristics.

INTRODUCTION

The growing recognition of the environmental impacts of conventional fossil-fuel facilities,


including the emission of greenhouse gases, has led to a search for methods of producing
energy that are more environmentally acceptable. Attention has focused on expanding the use
of hydropower by exploiting both undeveloped large sites, particularly outside of the U.S.,1
and numerous smaller sites, including dams originally built for purposes other than the
production, of electricity. Interest in hydroelectricity is also high because of the lack of technical
obstacles, the widespread ability to produce the necessary equipment, and its competitive
economic cost compared to other energy resources.
One of the major questions surrounding whether or not to expand the use of hydroelectric
facilities greatly throughout the world is the nature and severity of their environmental impacts.
There has been considerable interest in small hydroelectric facilities because of the assumption
that such units have fewer severe environmental impacts than large hydroelectric plants. No
comprehensive analysis of this assumption has been done, however.
We compare, the environmental and ecological impacts of small and large hydroelectric
facilities, with a focus on land requirements, evaporative water losses, seepage, and
sedimentation rates. The results suggest that many of the environmental impacts of small
facilities are comparable to or even worse than conventional large hydroelectric plants, when
735
736 PETER H. GLEICK

measured on a unit-energy basis. Perhaps even more important., the data presented here
suggest that the installed capacity of a facility may not be the most meaningful basis of
comparison. Another fundamental characteristic, the ratio of gross static head to dam height,
appears to separate environmental impacts in a more direct and clear manner. In particular,
hydroelectric plants for which the gross static head greatly exceeds the dam height may have
significant environmental advantages over other hydroelectric designs.
There are many other important environmental impacts, including downstream flow
alterations and impacts on terrestrial and aquatic biota, that are not quantified here. Indeed,
many of these impacts are poorly understood or extremely difficult to quantify. Nevertheless,
these environmental amenities and services must not be ignored when choosing among energy
alternatives. One of the real challenges facing energy planners may prove to be balancing
uncertain or largely unqualified ecological threats with the benefits of additional energy
production. ,

METHODS: LIMITATIONS AND UNCERTAINTIES

Energy production and use are accompanied by adverse environmental and ecological
consequences. Compiling, calculating, and comparing these diverse consequences has become a
major field of effort over the last several decades, spawning new methods of risk analysis,
economic assessment, and electric resource planning. The question of how environmental
externalities should be evaluated and compared is covered in detail in a variety of studies.2'4
Despite considerable progress in this area in the last few years, many important impacts remain
un quantified.
In this analysis, an attempt was made to quantify in a consistent manner some of the most
significant physical environmental impacts of hydroelectric facilities. No attempt was made to
translate impacts into common units such as dollars, since so little is known about the value of
the types of goods and services disrupted by hydroelectric facilities. These valuations are
inherently subjective, and so the evaluation of impacts is left in a form that permits the later
application of such values in the political process. Environmental impact's were, however, all
evaluated on the basis of the energy produced by the facility. Thus, some direct comparisons
between the different classes of hydroelectric plants are possible. In particular, impacts are
summarized as a function of size and type of plant.
The hydroelectric "fuel cycle" is considerably less complicated than fuel cycles for other
conventional electricity-generating systems. The resource, falling water, can be "mined"
without disturbing a quantity of land that continuously increases, and it requires no
concentrating, refining, or processing. The conversion of the resource into electricity emits no
gaseous, radioactive, or solid effluents, and little heat or noise. In normal operation,
hydroelectric systems entail little public or occupational, mortality or morbidity, no genetic
effects, and no undesirable political consequences, such as the proliferation of nuclear weapons
capability or links to war over strategic resources (although there is a risk of conflict over
shared international water resources, discussed later).
Despite these advantages, diverse impacts do arise from the construction and operation of
hydroelectric facilities. Significant physical, chemical, and biological transformations of local
ecosystems result from the creation of the reservoir and the displacement of wildlife, from the
storage of water in the reservoir, including consumptive water losses due to evaporation and
seepage, and from the way the plant is operated. In addition to these environmental impacts,
hydroelectric dams are subject to the risk of catastrophic failure with extensive loss of life and
property, an unusual risk associated with few other energy sources, most notably nuclear
power. And when hydroelectric facilities are developed on rivers that are shared by two or
more nations, political conflicts can arise.5 Conflicts have arisen over dams and water diversions
r

Environmental consequences of hydroelectric development 737

on the Han River between North and South Korea, the Euphrates River shared by Turkey,
Syria, and Iraq, the Nile River shared by nine nations, and many other rivers throughout the
world.6
The operating regime of hydroelectric dams is an important factor in determining the type
and extent of the impacts of particular facilities. The mode of operation itself is determined by
a set of economic criteria, together with competing upstream and downstream demands for the
water.-On occasion, the;mode of operation is determined by minimum flow requirements set to
maintain certain ecological conditions. In addition, electric utilities value .firm peaking power
(power requirements during periods of maximum demand), which hydroelectric plants are
usually well suited to provide. At the other extreme, some hydroelectric plants are constrained
to operate in "run-of-river" mode, where existing water demands take precedence over the
desire to hold water back for more lucrative peak periods. These different characteristics must
be taken into account in any overall environmental assessment.
In evaluating and comparing the environmental and ecological impacts of small and large
hydropower facilities, a number of other criteria that cut across the issue of size must also be
evaluated. These include the temporal and spatial distribution of harm, the possibility of
irreversible effects, .the coincidence of risks and benefits, and the uncertainty surrounding the
nature of the evidence of environmental harm. These issues have been widely discussed over
the last decade,- but many unresolved issues still remain.4'7
In addition, since hydroelectric facilities must be built where the resource is available, as
opposed to where the demand for electricity exists, there is a problem of non-coincident costs
and benefits. Hydroelectric dams also pose the risk of a single catastrophic event, i.e. dam
failure, not associated with most other energy facilities. A quantitative analysis of the risk of
such a high-consequence, low-probability event is extremely difficult given the mix of stochastic
and predictable causal factors involved.

The role of size


The delineation and identification of the environmental effects of hydroelectric facilities as a
function of size is complicated by a definitional problem; how is "small" hydro defined? Several
definitions have been promulgated by various official bodies interested in the subject, but these
definitions vary over time and are "often ambiguously formulated. Furthermore, since different
organizations define small hydro in different ways, few of the data related to large and small
hydroelectric facilities are in consistent, comparable forms.
One of the earliest official definitions classified dams as low-head if their height was less than
20 m.8 The existence of dams with large gross static heads due to the use of overland conduits
and penstocks led to a redefinition of small dams as hydroelectric facilities with an installed
capacity less than 5MWe.9 As part of the Federal program to encourage the development of
small hydroelectric facilities, the Federal Energy Regulatory Commission (FERC) developed
accelerated licensing requirements for certain facilities. The FERC definition of small hydro
was originally limited to units smaller than 1.5 MWe.
In 1978, the Public Utility Regulatory Policies Act (PURPA) was signed into law as part of
the National Energy Act. Title IV of PURPA authorized loans for feasibility studies and loans
for the construction of hydroelectric facilities at existing dams under 15 MWe installed capacity.
This act changed the official Department of Energy definition to 15 MWe (but not the FERC
definition) and small hydroprogram operated with this limit for 2yr. :PURPA also included
hydroelectric power in its general definition of small power projects as any project under
80 MWe.
In late 1979, there was some discussion within the Department of Energy about increasing
the definition to include dams with an installed capacity of 25 MWe or less, and in 'January
1980, the General Accounting Office recommended that the National Energy Act be amended
by redefining a small hydro project as one that could have up to 100 MWe of installed
738 PETER H. GLEICK

capacity.10 The Energy Security Act, passed in July 1980, increased the official Department of
Energy definition to 30 MWe or less at existing dams.
Other definitions include micro-hydro (100 kWe or less), small scale high-head hydropower7
small .low-head hydro power, and small hydro (10 MWe or less).11"13 '
As even this incomplete history suggests, a consistent classification of data, or even an
accepted definition of large or small, has an arbitrary component to it. For the purposes of this
analysis, most impacts data on facilities were compared on the basis of energy production,
rather than installed capacity. The basis for comparison here is the annualized impact per
1018 joules (J) of electricity produced. The choice of alternative common denominators is, of
course, possible, but a comparison on the basis of energy produced appears to be the most
useful. Some additional comparisons are made by separating facilities above and below
25 MWe. This value was chosen because it is the median installed capacity for the data set of
approximately 100 California facilities studied here, thus permitting an analysis of two groups
of units with equal numbers of facilities.14"19 - • -
One final distinction serves a useful purpose in this analysis. Hydroelectric facilities can be
separated into two other categories: those facilities with a gross static head (GSH) less than the
height of the dam (DH); and those facilities with a gross static head greater than the height of
the dam. The gross static head is the vertical distance from the top of the water in the reservoir
to the top of the water in the tailrace below the dam or powerhouse. The first group includes
dams with .the powerhouse typically located at the foot of the dam, while the second group
consists, of dams with long penstocks leading downhill to a powerhouse often far removed from
the dam itself. . .
Data from a large number of California hydroelectric facilities were.compiled and compared:
Data from a smaller set of 40 more facilities in Pennsylvania were also evaluated as a
comparison. California is blessed with substantial hydroelectric potential. Hydroelectric plants
satisfy nearly 20% of total state electrical demand in a normal year. A wide range of facilities
are in operation, from enormous dams and multi-purpose reservoirs to tiny run-of-river
facilities built on aqueducts and irrigation canals. The data used in this analysis represents
those facilities for which good data. could be collected on a wide range of physical
characteristics, including dam height, installed capacity, average annual energy production,
gross static head, reservoir area, reservoir volume, average annual flow, and drainage area.
The nearly 100 facilities studied are operated by federal, state, local, and private owners and
range in size from less than 1 MWe to nearly 500 MWe; the total capacity of these facilities is
just under 6000 MWe.

ANALYTICAL RESULTS

Physical effects
The amount of land flooded by a reservoir following completion of a hydroelectric dam is
useful as an index of ecosystem disruption. Flooding land displaces or eliminates terrestrial
biota by destroying habitat and altering the aquatic ecosystem of the formerly free-flowing
river. The actual land area flooded, however,.is not completely satisfactory as a measure of the
environmental disruption. The surface area of a reservoir is less than the surface area flooded
in a way that depends upon the geographic basin and the variation in terrain at the site.
Moreover, the ecological disturbance extends over an area that exceeds the area of land
directly flooded. Fauna displaced to habitat surrounding the reservoir disturb the stability of
that habitat by increasing competition for the remaining resources.
It can be argued that the land flooded for a reservoir is not consumed in the sense that the
surface area'of a reservoir can then be used for human activities, such as recreational boating.
The development of still-water fisheries has also been cited as a beneficial use of the reservoir.
738 PETER H. GLEICK

capacity.10 The Energy Security Act, passed in July 1980, increased the official Department of
Energy definition to 30 MWe or less at existing dams.
Other definitions include micro-hydro (100 kWe or less), small scale high-head hydropower,
small .low-head hydro power, and small hydro (10 MWe or less).11"13
. As even this incomplete history suggests, a consistent classification of data, or even- an
accepted definition of large or small, has an arbitrary component to it. For the purposes of this
analysis, most impacts data on facilities were compared on 'the basis of energy production,
rather than installed capacity. The basis for comparison here is the annualized impact per
1018 joules (J) of electricity produced. The choice of alternative common denominators is, of
course, possible, but a comparison on the basis of energy produced appears to be the most
useful. Some additional comparisons are made by separating facilities above and below
25 MWe. This value was-chosen because it is the median installed capacity for the data set of
approximately 100 California facilities studied here, thus permitting an analysis of two groups
of units with equal numbers of facilities.14"19 • • • • •
• One final distinction serves a useful purpose in this analysis. Hydroelectric facilities can be
separated into two other categories: those facilities with a gross static head (GSH) less than the
height of the dam (DH); and those facilities with a gross static head greater than the "height of
the dam. The gross static head is the vertical distance from the top of the water in the reservoir
to the top of the water in the tailrace below the dam or powerhouse. The first group includes
dams'with .the powerhouse typically located at the foot of the-dam, while the .second group
consists of dams with long penstocks .leading downhill to a powerhouse often far removed from
the dam itself-. . . .
Data from a large number of California hydroelectric facilities were compiled and compared.
Data from a smaller set, of 40 more facilities'in Pennsylvania were also evaluated as a
comparison. California is blessed with substantial hydroelectric potential. Hydroelectric plants
satisfy nearly 20% of total state electrical demand in a normal year. A wide range of facilities
are in operation, from .enormous dams and multi-purpose reservoirs to tiny run-6f-river
facilities built on aqueducts and irrigation canals. The data used in this analysis represents
those facilities for which good data. could be collected on j a wide range of physical
characteristics, including dam height, installed capacity, average annual energy production,
gross static head, reservoir area, reservoir volume, average annual flow, and drainage area.
The nearly 100 facilities studied are operated by federal, state, local, and private owners and,
range in size from less than 1 MWe to nearly 500 MWe; the total capacity of these facilities is
just under 6000 MWe. ' •

ANALYTICAL RESULTS

Physical effects
The amount of land flooded by a reservoir following completion of a hydroelectric dam is
useful as an index of ecosystem disruption. Flooding land displaces or eliminates terrestrial
biota by destroying habitat and altering the aquatic ecosystem of the formerly free-flowing
river. The actual land area flooded, however, is not completely satisfactory as a measure of the
environmental disruption. The surface area of a reservoir is less than the surface area flooded
in a way that depends upon the geographic basin and the variation in terrain at the site.
Moreover, the ecological disturbance extends over an area that exceeds the area of land
directly flooded. Fauna displaced to habitat surrounding the reservoir disturb the stability of
that habitat by increasing competition for the remaining resources.
It can be argued that the land flooded for a reservoir is not consumed in the sense that the
surface area of a reservoir can then be used for human activities, such as recreational boating.
The development of still-water fisheries has also been cited as a beneficial use of the reservoir..
Environmental consequences of hydroelectric development 739

These arguments, perhaps supportable using a conventional cost-benefit approach., fail to


account for the value and nature of the ecological goods and services foregone. No
widely-accepted economic method to assess this value exists that permits the incorporation.of
ecological losses into the decision-making -process. And the marginal benefit of additional
still-water recreation, particularly in areas with extensive reservoir development., shrinks while
the marginal cost of losing remaining riparian habitat and wild river ecosystems increases.
An analysis of reservoir area as a function of annual energy produced by hydroelectric
facilities reveals that certain small (under 25 MWe) hydroelectric facilities flood a land area
much greater than large dams. An analysis of the major hydroelectric facilities in California
shows an interesting trend (Table 1). First of all, the range of land flooded for a plant capable
of producing a given amount of energy per year is enormous.f For all facilities in the data set,
land requirements ranged from 10 square kilometers (km2) per 101S J per year to 58,000 kni2
per 1018J per year, with a median value of 1500km2 per 1018J/yr. When disaggregated by
installed capacity, i.e. greater and less than 25 MWe, the specific land area flooded (reservoir
area divided by annual energy production) for smaller facilities is considerably greater than that
for larger reservoirs. Dams larger than 25 MWe flooded a median value of 660 km2 per 1018
joules of electricity produced annually, while dams smaller than 25 MWe disturb six times as
much land, about 4000 km2 per 1018 J/yr.
If the data are broken down in a different way, a more interesting trend appears. .When the
gross static head of a facility exceed the dam height (i.e. when there is a penstock .carrying
water downslope to a power house), the median value of area of land flooded is considerably
smaller than when the dam height exceeds the gross static head, namely 340 vs 9500 km2 per
1018 J/yr, irrespective of installed capacity. . .
When both type and size of facility are evaluated, the results are mixed. Where the gross
static head exceeds the height of the dam, smaller dams flood nearly four times the land area of
larger dams (900 vs 200 km2/1018 J/yr). If the dam height exceeds or equals the gross static
head, large dams are more land consumptive by a factor of two (19,OOOkm2/10ls J vs
9400km2/1018J).. . . '
The 'primary reason for these distinctions is that many big dams .constructed for multiple

Table 1. Land requirements for California hydroelectric facilities in km2/1018 J/yr.


These values are computed from the surface area of the reservoir and the average
annual energy production of the facility; DH = dam height, GSH = gross static
head.
All Plants over Plants under
Category Facilities 25 MWe 25 MWe

Range: All Facilities 10 to 58,000 10 to 45,000 50 to 58,000

Median: All Facilities 1,500 660 4,000

Dam Height < GSH 10 to 34,000 10 to 34,000 50 to 23 ,000

Median: DH < GSH 340 200 900

Dam Height > GSH 540 to 58,000 1000 to 45,000 540 to 58,000

Median: DH > GSH 9,500 . 19,000 9,400

fThis measure, rather than one that uses total lifetime energy produced by a plant as the denominator, was chosen for
several reasons: (i) it does not require assumptions about plant lifetime; (ii) it permits normalization and
comparison with different types of plants; (in) it is easily converted to other indices without confusion. The annual
energy values are computed from long-term averages.
740 PETER H. GLEICK

Table 2. Land requirements for Pennsylvania non-


hydroelectric dams in km 2 /10 IH J/yr. These values are
computed from the surface area of the 40 reservoirs in the
sample from Ref. 30 (see text) and the average annual
energy production that would result from retrofitting each
facility with turbines.
Dams with potential
Category generating capacity
under 25 MWe '

Range: All Dams 3,130 to 4,400,000

Median: All Dams 34,000

Average: All Dams 213,000

purposes -have large reservoir storage areas and small annual energy production. Evaluating
non-hydroelectric dams that have the potential of producing energy supports the observation
that flood control reservoirs., irrigation reservoirs, and water supply reservoirs require greater
storage volumes and land areas in order to meet their more seasonal demands. An analysis of
40 non-hydroelectric dams in Pennsylvania (see Table 2), al! with potential electrical generating
capacity less than 25 MWe and with dam height greater than gross static head, indicates that
the specific land 'area for these dams is considerably greater than for the dams in California
under 25 MWe, a fact which also reflects differences in the nature of the terrain.
These results indicate that under certain circumstances, the type of dam may be a more
appropriate-indicator of land consumption than the size, and that the construction of large
numbers of small dams may lead to the disruption of considerably more land than developing
the same- capacity in larger facilities.
When evaluated on the basis of total lifetime energy produced, rather than average annual
energy, dams with gross static head greater than dam height have another major advantage.
Sedimentation rates (as measured by trap efficiency) for such facilities are much lower, leading
to longer lifetimes and larger life-cycle energy production. This is discussed further below.

Evaporative and seepage losses


Water is lost from reservoirs and other bodies of water as a result of evaporation from the
exposed surface area and from seepage through underlying porous rock formations. While
hydroelectric projects are often considered non-consumptive of resources, the evaporative loss
of water from a region due to the presence of large areas of artificial reservoirs represents the
consumption of a resource that would otherwise be available for downstream human and
ecological uses. Conflicts for water among agricultural users, industrial users, commercial
users, and ecological support functions are intensified by this water loss. As water has become
more precious in different regions of the United States and the world, evaporation and seepage
from artificial reservoirs have received more and more attention.
Evaporative loss. Evaporation occurs naturally from the surface of any body of water.
Indeed, an early aspect of hydrologic research included estimates and studies of evaporation
from regulated reservoirs and lakes.20 Gross evaporative losses from artificial reservoirs alone
appear to exceed consumptive water use by all other activities, though this loss is often
ignored. In part this is due to the perception that such reservoirs provide a net water benefit by
permitting water to be stored for use that would otherwise have been "lost" to runoff during
high flow periods.
The evaporation of water is directly related to the surface area of the body of water and
varies with the temperature, wind conditions, and humidity of a region. Average annual
Environmental consequences of hydroelectric development 741

evaporation from standing water in the United States varies from 0.5 m in northeastern regions
to over 2 m in the desert regions of southeastern California. For 700 reservoirs and regulated
lakes in the western 17 states with a total effective surface area of approximately 14,000 km2,
annual evaporative losses based on variations in location and climate were approximately
15.2cubic kilometers (km3), or an average annual evaporation -of l.lm. 21 Evaporative losses
can be much higher, depending on the climate of a region. At the Aswan High Dam on the
Nile, about 14km3 of water (11% of reservoir capacity) are lost annually to evaporation from a
surface area of 5200 km2.22'23 This is equivalent to nearly 3 m of evaporative loss per year.
The total surface area for the California hydro facilities in our data set is nearly 600km2.
Assumiag an average annual evaporation of 1 m from the surface of these reservoirs, 0.6 km3 of
water are lost annually. These facilities produce about 24 billion kilowatt hours of electricity
per year. Apportioning all the evaporative losses to energy production, over 7 km3 of water are
lost per 1018 joules of electricity produced. This is nearly 10 times the consumptive water use of
nuclear, or fossil-fuel facilities per unit energy produced. Table 3 summarizes, the evaporative
losses from reservoirs. , -
Table. 3 breaks down, evaporative losses by size and type of facility. The median evaporative
loss for all the plants in the data set is 1.5 km3 per 1018 J, with an enormous difference between
types of plants. Where the GSH exceeds the dam height, the median loss is 0.3 km3 per 101S J.
Where GSH is smaller than the dam height, median, evaporative loss, is 30 .times larger:
9.5km3/1018J. When the size of the'plant is considered, small units lose more water to
evaporation than large ones per unit energy generated: 4.0 vs 0.7 km3 per 1018 J.-
Combining both characteristics (size and type), however, shows that evaporative losses from
both large and small facilities are relatively small when the gross static head exceeds the dam
height, while much greater evaporative losses appear when dam height exceeds gross static
height. Evaporative losses for hydroelectric plants exceeding 25 MWe, with gross static head
less than the dam height, lose an average of 19km3 per-1018 joules of electricity produced,
while, the same type of facilities with an installed capacity less than 25 MWe lose less than half
as much-water per unit energy: 9.4km3/1018 J. Where the gross static head is larger than the
height of the dam, however, evaporative losses are reduced considerably to 0.2km3/lQ18 J for
large dams and 0.9 km3/l018 J for smaller dams.
Seepage loss. Seepage losses from porous foundations underlying hydroelectric reservoirs
can be a serious problem. It has been estimated that an average of 5% of the volume of
reservoirs is lost annually to seepage,24 and seepage losses at some facilities have become big

Table 3.. Annual evaporative losses from California hydroelectric facilities in


km3/101S J(e)/yr. These values are computed from the surface area of the
reservoirs, annual evaporative loss, and the average annual energy production of
the facility; DH = dam height, GSH = gross static head.

All Plants over Plants under


Category Facilities 25 MWe 25 MWe

Range: All Facilities 0.01 to 58 CIO 1 to 45 0.05 to 58

Median: All Facilities 1.5 0.7- • 4.0

Dam Height < GSH 0.01 to 34 0.01 to 34 0.05 to 23

Median: DH < GSH 0.3 •0.2 0.9

Dam Height > GSH 0.54 to 58 1 to 45 - 0.54 to 58

Median: DH > GSH 9.5 • 19.0 9.4


742 PETER H. GLEICK

political and environmental problems. The Anchor Dam in Wyoming, for example, is built in a
location that is so porous that the reservoir has never totally filled in 30 yr of operation.25
There is an important qualitative difference between seepage and evaporative losses. Water
lost to evaporation usually leaves the hydrologic basin and thus is a true loss. Water lost to
seepage may still enter downstream portions of the river or become available for ground water
pumping.
Evaluating data from Ref. 9 on reservoir storage in the 18 hydrographic regions of the
United States reveals that there are approximately 210km3 of storage in hydroelectric
reservoirs and over 2100km3 of storage in 49,000 reservoirs. Hydroelectric production from
these facilities averages about 290 billion kWh annually. Assuming seepage of about 5%
.annually from these reservoirs, losses of 10km3 per 1018 J can be expected. For the California
reservoirs studied here, we estimate that about 11 km3 of water are lost to seepage for every
1018J of hydroelectricity produced, in good agreement with the national estimate.
. The storage requirements, and hence seepage losses, of hydroelectric reservoirs vary as a
function of the size of the generating facility.-Tables 4 and 5 summarize the range of storage
requirements and seepage losses for dams greater and less than 25 MWe and with gross static
heads greater and less than the height of the dam. In general, dams with dam height greater
than gross static head lose 50 times more water to seepage per unit energy produced than
dams with large GSH/DH ratios, regardless of installed capacity (a median of 11 vs 0.2 km3 per
1018J). - - . • .
. Facilities most likely to experience large seepage losses are dams with large reservoirs (as
measured by physical size, not energy production). An evaluation of the California data shows
that dams with a gross static head less than the dam height have larger losses per unit energy
produced than dams with large gross static heads. Where GSH<DH, small facilities appear to
require less storage volume and have fewer losses to seepage per unit energy produced than do
larger facilities. This is consistent with the earlier observation that many larger reservoirs have
been built with flood control storage, which cannot be used for power production during parts
of the year. • . .

Sedimentation
. Every free-flowing body of water carries a quantity of suspended matter—the sediment load
and .bed load. Organic and inorganic materials in the river provide many of the nutrients
required for aquatic biota. Sediments also play a role in protecting downstream stretches from

Table 4. Storage requirements for California hydroelectric facilities in


km3/1018 J/yr. These values are computed from the storage volume of the
reservoirs and the'average annual energy production of the facility; DH = dam
height, GSH = gross static head.

All • .Plants over Plants under


Category Facilities , 25 MWe 25 MWe

Range: All Facilities , 0.03 to 2000 . 0.03 to 2000 0.2 to 1100

Median: All Facilities 11 6 23

Dam Height < GSH 0.03 to 1700 • 0.03 to 1700 0.2 to 690

Median: DH < GSH • 3 2 4

Dam Height > GSH • 1.6 to 2000 . 7 to 2000 1.6 to 1100

Median: DH > GSH . 220 .. 500 110


Environmental consequences of hydroelectric development 743

Table 5. Annual seepage losses from California hydroelectric facilities in


km3/1018 J/yr. These values are computed from the storage volume of. the
reservoirs, an annual loss of 5% of total storage volume, and the average annual
production of the facility; DH ~ dam height, GSH — gross static head.
All Plants over Plants under
Category Facilities 25MWe 25MWe

Range: All Facilities 0.002 to 100 0.002 to 100 0.01 to 55

Median: All Facilities 0.6 0.3 1.1

Dam Height < GSH • 0.02 to 85 0.02 to 85 0.01 to 35

Median: DH < GSH 0.2 0.1 0.2

Dam Height > GSH 0.08 to 100 0.35 to 100 0.08 to 55

Median: DH > GSH 11.0 25.0 5.5

erosion by increasing the viscosity of the water, decreasing the velocity and souring effect of
clear, rapidly-flowing water, and by depositing sediments along river banks. The'transfer of
sediments from upstream regions to estuaries and bays is a valuable source of nutrients and
materials for estuarine ecosystems. The sediment load of a river varies with numerous factors,
including the climate of the region, the nature and extent of vegetative cover in the watershed,
geology and soil characteristics, the slope of the land and river, and the basin land use.,
Sedimentation is an important environmental issue in hydroelectric development for two
reasons: hydroelectric facilities affect the flow of sediments (and hence nutrients and materials)
downstream of dams; and sediments affect the storage volume and operation of reservoirs by
displacing useful capacity. Loss of storage capacity affects the quantity of energy that can be
produced, the mode of operation of the facility, the ability of. the development to provide
flood-control benefits, and the quality of the water in the reservoir. The loss of storage also
shifts a darn from peak-power production to run-of-river operation by reducing the amount of
water that can be stored for release at a later time. " . -- ,
Every reservoir experiences silting and sediment accumulation, although the extent and
severity of the sedimentation varies widely. The U.S. Bureau of Reclamation has estimated
that the proposed Auburn Reservoir in California, if completed, would have a lifetime of
9000 yr, while other reservoirs lose useful capacity at the rate of 2 or 3% annually, losing their
entire capacity within 30-50 yr.26-27 Crowder (Ref. 28) estimates that 0.22% of U.S. reservoir
storage capacity is lost annually to sedimentation.28 At the High Aswan Dam in Egypt, for
example, the mean annual storage of silt has been estimated to be 110 million tons.23 Although
at present siltation rates it would take 500 yr to occupy the dead storage of the reservoir, silt is
already encroaching upon live storage in the upper reaches of the reservoir.
The extent of sediment settling that occurs depends upon the sediment load of the river, the
size and settling characteristics of the sediments, the method of operation, and the "trap
efficiency" of the reservoir. The trap efficiency is measured as the ratio between reservoir
capacity and the annual volume of inflow; it is also a measure of the average residence time. In
his classic paper, Brune describes the correlation between the settling of particles and the
period of residence in a reservoir.29
Hydro facilities with gross static head less than the height of the dam tend to have high trap
efficiency; those with small reservoir surface areas and where the gross static head exceeds the
height of the dam have low trap efficiencies that permit sediments to remain in suspension and
pass through the reservoir.
744 PETER H. GLEICK

+
0.01 =
+
4-
+4- *4 4 4
0.001
200 400 600 800
Gross Static Head (m)
Fig. 1. The trap efficiency of reservoirs is shown as a function of gross static head (GSH is the vertical
distance from the top of the water in the reservoir to the top of the water in the tailrace below the
dam or powerhouse) for reservoirs with the GSH exceeding the dam height.

Trap efficiency exhibits 'another- dependency. The efficiency with which reservoirs trap
sediments increases with increasing gross static head for those dams where the gross static head
is less than the height of the dam. Figures 1 and 2 plots the trap efficiency versus the gross static
head for dams where the gross static head exceeds the dam height and where dam height
exceeds the gross static head. . .
No obvious trends separating units by installed capacity were found..Nevertheless, some
characteristics of hydroelectric dams suggest that numerous small developments on a river may
have more serious effects on sediment retention and erosion than single large units. The bed
load carried by a river is. that portion of the sediment suspended by turbulent action of the
water. This load usually makes up about 65-70% of the total sediment load by volume, and
settles'out quite rapidly upon entering still water, since it is composed of coarse material with
short settling times. A series of small dams on a river will result in the settling of the bed load
behind the first dam, resuspension of a new bed load by the clear water flowing more rapidly

10

~~ ~?f,- 'S'-V'"**'^
—/ '-f ^" -.-.•»
s y.-+"..-'-a*' • •"' '-*.'
L •' j-• -' '

0.1

0.01

0.001
50 100 150 200
Gross Static Head (m)
Fig. 2. The trap efficiency of reservoirs is shown as a function.of gross static head (GSH is the vertical
distance from the top of the 'water in the reservoir to the top of the water in the tailrace below the
dam or powerhouse) for reservoirs with the dam height exceeding the GSH.
Environmental consequences of hydroelectric development 745

below the first dam, a settling of the new bed load behind the second dam, and so forth. Thus
heavier erosion and greater sedimentation may occur along the length of the river than might
have occurred with a single larger dam.

Operating mode and instreamflow requirements


Plants operated in the peaking mode cause alterations in flow patterns downstream that are
highly stochastic over short periods of time. While river flows are naturally stochastic, the
periodicity of natural, changes is typically monthly or seasonally. The periodicity of flow
fluctuations resulting from operating hydroelectric dams for energy production during periods
of peak demand is typically on the order of hours to days. Peaking operations can also result in
significant fluctuations in reservoir levels over short periods of time. Such fluctuations are
disruptive of aquatic biota.
Hydroelectric plants operated in the run-of-river mode do not restrict the time of energy
production. An exception to this is, of course, dams that restrict or otherwise alter streamflow
in order to meet non-hydroelectric water requirements. These darns cannot be classified as pure
run-of-river. When there is little change in the natural stochasticity of flow, run-of-river plants
have fewer ecological impacts. For facilities in California, most very small units are operated in
run-of-river mode, since they do not have sufficient storage volume to retain water for peak
periods. This advantage also accrues to larger plants where the GSH>DH, though not as
consistently.
Certain problems arise with this generalization. Some rivers are so developed that the
operation of one hydro dam affects the water available for downstream dams. Perhaps the most
extreme example is the Columbia River system in the Pacific Northwest, which has some of the
largest dams in the world spaced so that the reservoirs of some downstream dams come up to
the tailrace of the dam immediately upstream. If the dam farthest upstream were to retain
water in its reservoir, and if it had sufficient storage, the downstream dams would be unable to
produce significant power. Conversely, if the upstream dam released more water in a short
period of time than could be run through the turbines of downstream dams, water would have
to be spilled without generating electricity. Generation of electricity in such systems is thus
highly regulated among the different dams, with most of them operating as run-of-river plants.
Such a system greatly complicates any comprehensive effort to protect fisheries and natural
ecosystems. v
One last vital consideration must be mentioned. Facilities where the gross static head greatly
exceeds the height of the dam require long penstocks to remove water from a river and feed it
downslope and downstream to a powerhouse. This very act of removing water from a river
alters its fundamental biological characteristics. In the United States, strict regulations are
often, though not always, imposed to ensure that sufficient flows remain in the river to maintain
biological communities., but in many regions of the world, rivers and streams are entirely
dewatered by such facilities. The environmental advantages accruing to facilities where
GSH>DH ; described above, may thus be completely outweighed by habitat and species
destruction unless these minimum flow requirements are identified and enforced.

DISCUSSIONS AND CONCLUSIONS

The construction and operation of hydroelectric plants cause significant environmental


changes in the characteristics of running-water ecosystems. The magnitude and extent of these
changes vary with the climatic, geophysical, and ecological character of the site and with the
type and mode of operation of hydroelectric facilities. This study focused on a set of
quantifiable physical effects of hydroelectric facilities in an effort to evaluate the role of size,'
and some significant distinctions among facilities were found. In addition to these effects, there
746 PETER H. GLEICK

is a wide range of other biological and chemical impacts not studied here that must be taken
into account in any final environmental assessments.
Historically, it has been assumed that clear environmental advantages among hydroelectric
facilities could be discerned based on the installed capacity. This assumption has been used to
justify accelerated or simplified licensing policies worldwide. "While modest differences on this
basis were identified here, other facility characteristics appear to offer more definitive
environmental advantages or disadvantages. In particular, the relationship between gross static
head and darn height appears to be a far better indicator of land consumption, evaporative
water loss, and seepage and sedimentation rates than power production.
Four categories were distinguished in this paper: those with an installed capacity greater and
less than 25 MWe and those with a gross static head greater and less than the height of the
dam/Both large and small facilities, where the gross static head is greater than the dam height,
have considerable environmental advantages over dams whose gross static head is smaller than
the dam height. Large, dams whose gross static head is smaller than the height of the dam
consume more land, more water, and require more storage volume, per unit energy produced,
than .smaller dams whose gross static head is smaller than the height of the dam. The situation
is reversed completely, however, where the gross static head exceeds the height of the dam, i.e.
in facilities with long penstocks running down to a powerhouse. In these cases, smaller dams
consume more land, more water, and require more storage volume per unit energy produced
than dams with an installed capacity greater than 25 MWe,
Sedimentation, which affects both downstream ecosystems and the storage capacity of a
reservoir, is a problem in larger dams with high trap efficiencies and hi reservoir systems with
dams whose height exceeds the gross static head, regardless of installed capacity. The mode of
operation of a hydroelectric facility also has environmental impacts independent of size, though
more small facilities tend to operate in run-of-river mode. Peaking facilities cause more severe
chemical and biological impacts • than facilities operated as run-of-river plants, as do new
facilities, when compared to retro-fitting existing dams with energy-generating facilities. These
results, though limited to specific physical environment impacts, have important implications
for future hydroelectric development.
Hydroelectric potential exists at thousands, of dispersed sites throughout the United. States
and the world, and in some regions, such as Africa, there is increasing need and pressure to
develop these sites. Identifying classes of facilities with .environmental advantages or
disadvantages will permit 'development decisions to be made with far more environmental
sensitivity than has been shown to date.

REFERENCES

1. C. O. Okidi, "Environmental Stress and Conflict in Africa: Case Studies of African International
Drainage Basins," Workshop on Environmental Stress and Acute Conflicts, American Academy of
Arts and Sciences/University of Toronto, (15-16 June 1991).
2. EPRI (Electric Power Research Institute), "Benefit of Environmental Controls: Measures, Methods,
and Application," EPRI EA-6030, Palo Alto, CA (1988).
3. O. Hohmeyer, Social Costs of Energy Consumption: External Effects of Electricity Generation in the
Federal Republic of Germany, Springer, Berlin (1989).
4. Pace University, Environmental Costs of Electricity, Center for Environmental Legal Studies, Pace
• .University, New York; Oceana Publications, New York, NY (1990).
5. P. H. Gleick, "Environment, Resources and International Security and Politics", in Science and
International Security: Responding to a Changing World, pp. 501-523, E. Arnett ed., American
Association for the Advancement of Science, Washington, DC (1990).
6. P. H. Gleick, "Water and Conflict", Workshop on Environmental Stress and Acute Conflicts,
American Academy of Arts and Sciences/University of Toronto, (15-16 June 1991).
Environmental consequences of hydroelectric development 747

7. J. P. Holdren, K. Anderson, P. M. Deibler, P. H. Gleick, I. Mintzer, and G. Morris, "Health and


Safety Aspects of Renewable, Geothermal, and Fusion Energy Systems," pp. 141-208, in Health
Risks of Energy Technologies, C. C. Travis and E. L. Etnier eds., American Association for the
Advancement of Science, Westview Press, Boulder, CO (1983).
8. U.S. Department of Energy, "Low-Head Hydroelectric Power Developments: Management Review
. and Control Document, Fiscal Year 1978", Division of Geothermal Energy, Washington, DC (1977).
9. U.S. Army Corps of Engineers, "Estimate of National Hydroelectric Power Potential at Existing
Dams," The Institute for Water Resources, Ft. Belvoir (1977).
10. U.S. General Accounting Office, "Hydropower—An Energy Source Whose Time has Come Again",
EMD-80-30, Washington, DC (1980).
11. T. W. Mermel, "Dams and Hydro Plants", Water Power and Dam Construction, pp. 67-77 (June
• 1991).
12. U.S. Department of Energy, "Small Low-Head Hydroelectric Power," Proc. from the Midwest
Regional Conf., IDO-10076 (23-25 May 1978).
13. Y. Takahasi, "Small Scale High-Head Hydropower Development in Japan, unpublished (1 March
1977).
14. Federal Power Commission, "Hydroelectric Power Resources of the United States: Developed and
Undeveloped," U.S. Government Printing Office, Washington, DC (January 1976).
15. Pacific Gas and Electric, "Hydroelectric Powerhouses as of 31 December 1988: Data Base," Pacific
Gas and Electric Company, San Francisco, CA (1989).
16. State of California, "Hydroelectric Energy Potential in California," Department of Water Resources,
Bulletin 194, Sacramento, CA (1974). .
17. State .of California, "A Survey of Small Hydroelectric Potential at Existing Sites in California,"
Department of Water ResourceSj'Bulletin 205, Sacramento, CA (1979).
18. State of California, "Small Hydroelectric Potential at Existing Hydraulic Structures in California,"
Department of Water Resources, Bulletin 211,. Sacramento, CA (1981).
19. State of California, "Dams Within the Jurisdiction of the State of California," Department of Water
Resources, Bulletin 17-88, Sacramento, CA (1988). ' .
20. U:S. Geological Survey, "Evaporation from the 17 Western States," Geological Survey Professional
Paper, 272-D, Washington, DC (1962). ' ' "
21. A. R. Golze ed., Handbook of Dam Engineering, Van Nostrand-Reinhold, New York, NY (1977)'.
22." M.-Shahin, Hydrology of the Nile Basin, Elsevier, Amsterdam, (1985). • •
23. G. F. White, "The Environmental Effects of the High. Dam at Aswan," Environment 30 (1988). '
24. J. Ingersoll, "Hydroelectric Power in California", Lawrence Berkeley .Laboratory,-.unpublished,
University of California, Berkeley, CA (1978). •
25. II.S. Water News, "After reservoir was nearly filled, Wyoming dam just doesn't hold water", Vol. 8,
No. 3, p. 4(1991). •" ' • ' ' '
26. L. V. Yakovleva, "Characteristics of. Silting of Small Reservoirs of the Central Chernozem Provinces
and Computation of Density of Bottom Deposits", in G. V. Lopatin ed., Water Balance and Silting of
Small Reservoirs in the Central Chernozem Provinces of the Russian Soviet Federal Socialist Republic,
Academy of Sciences, Moscow, Russia (1965).
27. U.S. Department of Interior, "Auburn Dam Environmental Impacts Study, Volume 2,4," Prepared
for the Bureau of Reclamation by Kennedy Engineers: Royston, Hanamoto, Beck, and Abey; Jara
• Applied Science (August 1971).
28. B. M. Crowder, "Economic Costs of Reservoir Sedimentation: A Regional Approach to Estimating
Cropland Erosion Damages," /. Soil Water Conserv. 42, pp. 194-197 (1987).
29. G. M. Brune, "Trap Efficiency in Reservoirs," Trans. Am. Geophys. Union 34, pp. 407—418 (1954).
30. John Hopkins University, "Problems of Hydroelectric Development at Existing Dams: An Analysis
of Institutional, Economic, and Environmental Restrains in Pennsylvania, New Jersey, and
Maryland," Applied Physics Laboratory, APL/JHU He-0043 Vol. 1 and 11, QM-79-237/133 (1979).
31. United States Department of Energy, "Technology Characterizations: Environment Information
Handbook," DOE/EV-0072, Washington, DC (1980).

Você também pode gostar