United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S7-81-112 Sept. 1981
Project Summary
Executive Summary for
Power Plant Cooling System
Water Consumption and
Nonwater Impact Reports
M. C. Hu, G. F. Pavlenco, and G. A. Englesson
This executive summary contains
salient features of three studies
performed by United Engineers &
Constructors Inc. (UE&C) and Versar.
Inc. to aid EPA in assessing water
consumption and other nonwater
quality environmental impacts of
cooling systems used by the steam-
electric generating industry. After an
introduction followed by a summary
and major conclusions from the
reports, important details and data
serving as bases for major conclusions
are condensed separately in a section
for each report.
The two reports on water consump-
tion agree that the Leung-Moore
model satisfactorily predicts evapora-
tion rates of evaporative cooling
towers. Versar found that this model
predicted evaporation rates at base-
load plants within ±15 percent of
material balance values based on plant
data. UE&C addressed the Harbeck
and Brady models for predicting
evaporation rates from cooling ponds
and (using available data) could not
reach a definite conclusion on which
model predicted evaporation more
closely. Comparing predictions from
five models, including the Harbeck
and Brady models, Versar concluded
that the Harbeck-Koberg-Hughes and
Meyer models predicted evaporation
rates for five of six cooling ponds
studied within ±15 percent of material
balance values and appear suited to
preliminary studies. UE&C concluded
that limited data precluded general
conclusions on drift effects from
saltwater cooling towers, but that
cooling tower drift and deposition
apparently do not increase ambient
salt loading to an extent that a
significant impact has been observed
or is expected. Similarly, the effects of
wet cooling tower and stack plume
interactions on acid rain formation
and deposition could not be estab-
lished from available information.
This Project Summary was devel-
oped by EPA's Industrial Environmen-
tal Research Laboratory, Research
Triangle Park. NC, to announce key
findings of the research project that is
fully documented in a separate report
of the same title (see Project Report
ordering information at back).
Introduction
Three studies were performed by two
contractors to assist EPA's Effluent
Guidelines Division in addressing the
water consumption and nonwater
quality environmental impacts of cooling
system options available to the steam-
electric generating industry. Reports for
these studies, listed below, addressed
issues which the court (in remanding
effluent guidelines for the steam-
electric generating industry, point
-------
source category) instructed EPA to
reconsider.
1. Water Consumption and Costs for
Various Steam Electric Power
Plant Cooling Systems (EPA-
600/7-78-157, NTIS No. PB 285397)
by M.C. Hu, G.F. Pavlenco, and
G.A. Englesson, United Engineers
& Construction Inc. (designated as
the first UE&C study).
2. Nonwater Quality Impacts of
Closed-Cycle Cooling Systems
and the Interaction of Stack Gas
and Cooling Tower Plumes (EPA-
600/7-79-090, NTIS No. PB 80-
102387) by G.A. Englesson and
M.C. Hu, United Engineers &
Constructors Inc. (designated as
the second UE&C study).
3. Comparison of Model Predictions
and Consumptive Water Use of
Closed-Cycle Cooling Systems
(EPA-600/7-78-206, NTIS No. PB
80-148273) by Jerome B. Strauss,
Versar, Inc. (designated as the
Versar study).
The results of these reports are1
summarized in this Executive Summary
to facilitate easy access to the salient
features contained in these reports. The
major conclusions are presented in this
report, along with a summary which
provides an overview of the method-
ologies, data bases, and conclusions of
the three reports. The important details
and data which provide the basis of the
major conclusions drawn in the three
reports are condensed in separate
sections, one for each report.
Summary of UE&C Report:
Water Consumption and Costs for Various Steam-Electric
Power Plant Cooling Systems
General Description
The subject study reviewed and
evaluated available information and
assessed the state-of-the-art on the
following subjects: (a) water consump-
tion rates of various open- and closed-
cycle cooling systems used by moderate
and large capacity steam-electric gen-
erating stations in the U.S.; (b) costs of
cooling system alternatives; (c) the
estimated availability of water for all
uses, and, in particular, for power plant
water heat rejection in the various
regions of the U.S.; and (d) the impact of
regulatory guidelines on consumptive
water use in the U.S.
The primary objective of the study
was to better understand the water
consumption aspects of power plant
cooling systems as they may impact
water availability for power generation
and other uses. The literature search
included papers and reports published
since 1973.
Water Consumption* Rates
of Cooling Systems
The literature review on water con-
sumption revealed that two reports by
Espey, Huston and Associates, Inc.
(EH&A) and one by Hanford Engineering
Development Laboratory (HEDL) con-
tained comprehensive calculations of
water consumption requirements of all
the major cooling system alternatives
for all 18 water resource regions in the
conterminous U.S. Only one field study
*The water consumption of a power plant cooling
system is defined as that portion of the water
removed from and not returned to the surface water
resources of a given area
on a closed-cycle cooling pond was
found in the literature. Preliminary
results from the first phase of the
concurrent Versar study, using unpub-
lished field data supplied by utilities,
were also reviewed.
The cooling system alternatives
considered in the EH&A and HEDL
studies included the open-cycle once-
through cooling system and the closed-
cycle cooling towef and cooling pond/
lake systems. These studies emphasized
the latter systems. A cooling tower was
classified as: (a) a tower with or without
a supplemental water reservoir for
makeup supply, and (b) a tower with its
blowdown retained for on-site disposal
or a tower with its blowdown returned
to its water resource. A cooling pond
was classified as either a single-
purpose pond or a multi-purpose pond.
The former refers to a man-made pond
used primarily for power plant cooling
with other uses being incidental to its
construction. The latter refers generally
to a natural pond used for power plant
cooling and other purposes such as
recreation and flood control.
The water consumption rate of the
cooling systems is calculated by equa-
tions which may include some or all of
the following water consumption terms
as appropriate: (a) forced evaporation
loss; (b) natural evaporation loss; (c)
blowdown; (d) uncontrolled release
such as seepage, overflow, and drift
loss; (e) local runoff inflow; (f) makeup
water; and (g) precipitation impinge-
ment onto the cooling water surface.
Forced evaporation loss is specifically
attributed to the cooling process;
whereas, natural evaporation exists
whether the power plant is operating or
not. Of these terms, the evaporation
loss is the major component for calcu-
lating cooling system water consump-
tion rates. It can be calculated with
semi-empirical models simulating the
operating behavior of various cooling
systems.
The major differences between the
EH&A studies and HEDL study, which
provided the bulk of the calculated^
water consumption rates reported in|
this UE&C study, are: (a) the models
used to calculate the cooling tower and
cooling pond evaporation rates, (b) the
assumed cooling tower blowdown rate
as a percentage of the tower evaporation
rate, and (c) the assumed precipitation
which may be credited as local runoff to
a pond/lake or impoundment. The
details of these differences are sum-
marized in Table 1. The evaporation
prediction models used in these studies
are briefly described below.
Evaporation Prediction Models
1. Cooling Tower Models
An evaporative or wet cooling tower is
a device which cools hot water by heat
exchange at the air/water interface.
The process is primarily based on
evaporation (latent heat of vaporization
is absorbed by evaporating some water
from the cooling liquid) with a small
portion of sensible heat transfer.
Because both the air and water flows
are channeled through the tower, the
physical process involved in the tower
operation can be easily modeled to give
reasonably accurate predictions of the
forced oxidation rate, which dominates
the natural evaporation. A
-------
Table 1. Assumptions Used for Calculating Water Consumption of Cooling Towers and Cooling Ponds
Espey, Huston & Associates. Inc.
(Document No 7775, September 1977)
Han ford Engineering Development Laboratory
(HEDL-TME 76-82, September 19761
1 . Forced Evaporation Model
2. Storage Volume
3. Seepage
4 Precipitation (P)
Cooling pond
Harbeck
{heat budget method)
No Change
Neglected
Included for single purpose
pond only but adjusted down-
ward by runoff rainfall ratio
Cooling tower
Leung S Moore
N/A
N/A
N/A
Cooling pond
Brady, Graves & Geyer
(mass transfer analysis)
No Change
Neglected
Included without any
correction for runoff
Cooling tower
Merkel Equation
N/A
N/A
N/A
5 Natural Evaporation (EN)
6. Forced Evaporation (EF)
7 Miscellaneous Plant
Use Water
8 Ambient Conditions
9 Wind Speed
10. Slowdown IB)
11. Pond Size
12. Water Surface
Temperature
13 Water Consumption (C)
Equation Used
14. Cycles of Concentration/
Percent Slowdown
15. Capacity Factor
Included for single purpose
pond only; data taken from
mean annual lake evapora-
tion
Included
Neglected
Ma/or city annual averages
for dry bulb and wet bulb
temperature and relative
humidity in the region
Annual average wind speed
adjusted to wind speed at 2
meters above water surface
Neglected
1 acre/M We and 2
acres/MWe
Same as average dry bulb
temperature
C=EN+EF-1 (l-r)P
(for single purpose pond)
C = EF
(for multipurpose pond)
N/A
80% for calculating heat re-
jected annually by a 1000
MWe fossil fuel plant
N/A
Included
Neglected
Same as for cooling pond
N/A
Case 1 = Neglected
Case 2 = 25% of evaporation
N/A
N/A
C= EF + B (blowdown retained)
C = EF (blowdown returned)
5/25% of evaporation loss
Same as for cooling pond
Included for manmade
(single purpose) pond only
Included
Neglected
Mean of monthly average
temperatures
Annual average
N/A
Included
Neglected
Same as for cooling pond
N/A
Same as for cooling pond
1 acre/MWt and 3 acres/MWt N/A
5% of evaporative
requirements
Determined by iteration
using the equilibrium
temperature concept
C=EN+EF-P
(for single purpose pond)
C = EF
(for multipurpose pond)
N/A
80% for calculating heat
rejected annually by a
1000 MWe fossil fuel plant
N/A
C = EF + B (blowdown retained)
C = EF (blowdown returned)
21/5% of evaporation loss
Same as for cooling pond
16 Effect of Elevation on
Water Consumption
1 7. Cooling Range
Neglected
Not specified
Considered
Not specified
Neglected
Considered for the slug
flow pond model only
Considered
Not specified
EH&A used the Leung-Moore model
for calculating tower evaporation rates,
given the tower heat rejection rate, wet
bulb temperature, relative humidity,
and elevation. HEDL used the Winiarski
model which involves solving Merkel's
equation to obtain tower performance
and then using the performance data to
calculate evaporation rate.
2. Cooling Pond Models
Heat from the pond surface is dis-
jipated through evaporation, convec-
tion, conduction, and radiation. It is
highly dependent on local meteorolog-
ical conditions (solar radiation, dry bulb
temperature, relative humidity or dew
point, wind speed, and cloud cover).
Determining the evaporative loss of a
cooling pond is considerably more
complex than for a wet cooling tower,
because quantitative estimates of
evaporation for cooling ponds involve
many parameters, making the modeling
difficult. There are two basic approaches
for estimating the forced evaporation
from cooling ponds. The energy budget
method, based on the First Law of
Thermodynamics, accounts for all
incoming, outgoing, and stored energy
at the pond surface layer and enables
the calculation of the energy available
for evaporation. The mass transfer
method is based on the Law of Conser-
vation of Matter. Numerous models
have been developed based on these
approaches. The Harbeck model, based
on the energy budget approach, was
used by EH&A, while the Brady model,
based on the mass transfer approach,
was used by HEDL
-------
The Harbeck model is represented by
a nomograph which gives the ratio of
the heat energy input used for forced
evaporation to the total heat energy
input to the pond/lake as a function of
the pond water surface temperature,
with the 2-m wind speed as the only
parameter. (Thermal loading of the
pond, that is, heat rejection rate per unit
pond surface area, need not be specified.)
According to Harbeck, the pond water
surface temperature required for the
application of the nomograph can be
considered approximately equal to the
air temperature above the pond. Harbeck
states that, in areas where ice cover
does not occur, the average annual
water surface temperature is usually
slightly lower than the average annual
air temperature because of the cooling
effect of natural evaporation. The
addition of heat by a power plant may
cause the water surface temperature to
more nearly equal the air temperature,
unless the thermal load is large relative
to the size of the lake. Harbeck also
states that if large air/water temper-
ature differences exist, the procedure
using his nomograph becomes of
questionable value* because of probable
errors in the conducted energy term of
the energy budget equation.
The Brady model and all other models
based on the mass transfer approach fit
the following general mass transfer
equation:
E = A f(u)
(D
where:
E = total evaporation rate,
A = pond surface area,
f(u) = wind speed function, and
ev = water vapor pressure potential
for mass transfer between the
saturated air at the pond water
surface and the ambient air
above.
*ln the two studies reviewed by UE&C, EH&A used
air temperatures as pond water surface temper-
atures when applying the Harbeck nomograph to
calculate forced evaporation rates of cooling ponds
No analysis was given concerning the validity of
this assumption when high thermal loading ponds
are considered in which large air/water temper-
ature differences potentially exist. In responding to
this question raised in the UE&C report after it was
published, EH&A performed an analysis which was
documented in an unpublished memorandum
EH&A states that its analysis shows that the
"Harbeck nomograph is a viable means of
estimating forced evaporation for a cooling pond
even at load levels in the EH&A report (i e , 1 and 2
ac/MWe)."
The Brady model requires an iterative
procedure based on an equilibrium
temperature concept for estimating the
pond water surface temperature in
order to determine ev, given the heat
rejection per unit pond surface area
(often called pond thermal loading) and
local meterological conditions (dew
point, wind speed, and gross solar
radiation).
HEDL used the Brady model to
determine both the total evaporation
rates and natural evaporation rates
using estimated pond water surface
temperatures and pond equilibrium
temperatures, respectively. The forced
evaporation rates were obtained by
subtracting the calculated natural
evaporation rates from the total evap-
oration rates. The natural evaporation
rates used in the EH&A estimates were
based on pan evaporation rates obtained
from weather stations.
Water Consumption Rates of-
Cooling Systems
Tables 2 and 3 present a portion of the
predicted water consumption rates of
cooling towers and cooling ponds
compiled in the UE&C study. These
results are the basis of the observations
and conclusions drawn in the study.
As indicated earlier, the predicted
cooling system water consumption
rates presented in the UE&Cstudy were
taken mainly from the EH&A and HEDL
studies. The EH&A results are given as
single values for each of the 18 water
resources regions; whereas, the HEDL
results are given as low and high values
resulting from the consideration of a
range of design parameters and oper-
ating conditions. All the results were
adjusted by UE&C to the same plant
heat rejection basis and to the same
data units. No other adjustments were
made. Therefore, the comparison of the
EH&A and HEDL results as presented in
the UE&C report reflect the differences
in the models for calculating evaporation
and the assumptions on blowdown for
cooling towers, water consumption
credit for cooling ponds due to local
runoff and precipitation, etc. as shown
in Table 1.
The single field study reported in the
literature on a closed-cycle cooling pond
did not provide measured water con-
sumption rates. This study concluded
that the Meyer model predicts pond
evaporation rates in excellent agree-
ment with field measurements obtained
in the same study.
Observations and Conclusions
Based on the separate comparisons of
the EH&A results and the HEDL results
on cooling towers versus cooling ponds
as given in Tables 2 and 3, UE&C made
the following observations with respect
to whether the cooling tower or the
cooling pond is more water consump-
tive:
1. The results of the EH&A study
(Figures 1 and 2) clearly show that
a cooling tower is more water
consumptive than a cooling pond.
(Figures 1 and 2 are the graphic
representations of the EH&A
results in Tables 2 and 3.)
2. The results of the HEDL study
(Figures 3 and 4) show that in
some regions a cooling tower is
more water consumptive than a
cooling pond, while in other
regions the opposite is true.
(Figures 3 and 4 are the graphic
representations of the HEDL re-
sults in Tables 2 and 3.)
The major conclusions drawn by the
first UE&C study with respect to water
consumption and evaporation rates of
cooling systems are:
1. The two cooling tower evaporation
prediction models (Leung-Moore|
used by EH&A and Wmiarski used"
by HEDL)gave comparable results.
2. The two evaporation rate predic-
tion models for cooling ponds/
lakes (Harbeck model* used by
EH&A and Brady model used by
HEDL) gave disparate results; the
rates predicted by the Brady model
were consistently higher than
those given by the Harbeck
model.**
3. Because of the disparity between
the predicted cooling ponds/lakes
evaporation rates and the unavail-
ability of consistent sets of field
data, a definitive conclusion can-
not be drawn as to which of the
models, Harbeck or Brady, gives
more accurate results.
4. Without the consideration of the
actual cooling pond water surface
temperature, the use of the
Harbeck model for calculating the
forced evaporation losses of cool-
ing ponds may result in an under-
estimation of these losses. The
assumption that the pond water
surface temperature is equal to
Referred to as the "Harbeck nomograph" in the
Versar study.
"These models are also compared in the Versar
study
-------
rable 2. Consumptive Water Use for Cooling Towers
Without makeup pond (106 gal. /day)
Slowdown returned
Water
resource
region
1 . New England
2. Mid-Atlantic
3. South Atlantic -Gulf
4. Great Lakes
5 /"I i
. Ohio
6. Tennessee
7. Upper Mississippi
8. Lower Mississippi
9. Souris-Red-Rainy
10. Missouri Basin
1 1. Arkansas-White-Red
12. Texas -Gulf
13. Rio Grande
14. Upper Colorado
15. Lower Colorado
16. Great Basin
17. Pacific Northwest
18. California
Location (1974)
Boston, MA
Concord, NH
Bangor, ME
Richmond, VA
Philadelphia, PA
Tampa, FL
Atlanta, GA
Detroit, Ml
Cleveland, OH
Columbus, OH
Louisville, KY
Knoxville, TN
Chattanooga, TN
Twin Cities, MN
St. Louis, MO
Jackson, MS
New Orleans, LA
Bismarck, ND
Duluth, MN
N. Plane, NE
Great Falls, MT
Tulsa, OK
Garden City, KS
Dallas, TX
Houston, TX
Albuquerque, NM
El Paso, TX
Farmington, NM
Gr. Junction, CO
Phoenix, AZ
Yuma, AZ
S. Lake City. UT
Reno, NV
Seattle, WA
Portland. OR
Los Angeles, CA
Sacramento, CA
EH&A
(1977)
7.91
8.27
8.98
8.03
8.50
8.62
8.27
8.86
7.79
8.39
8.74
9.09
8.98
8.15
8.98
8.50
8.27
8.75
9.10
Slowdown retained
EH&A HEDL EH&A
Low High (1977)
8.51
8.87
9.34
8.75
8.39
8.87
9.22
8.51
8.87
8.75
9.34
9.34
8.99
8.39
8.75
9.10
6.73
6.77
6.24
7.31
7.06
8.08
7.85
6.79
6.90
7.01
7.27
7.40
7.49
6.44
7.28
7.81
7.92
6.46
6.26
7.14
6.74
7.57
7.57
7.85
7.91
7.48
8.02
7.53
7.11
8.21
8.33
6.88
7.30
6.64
6.77
7.69
7.45
7.77
7.81
7.58
8.06
7.94
8.48
8.35
7.83
7.88
7.93
8.06
8.10
8.13
7.66
8.06
8.33
8.38
7.67
7.59
8.01
7.78
8.19
8.19
8.33
8.40
8.10
8.41
8.16
7.95
8.58
8.77
7.84
8.01
7.81
7.84
8.21
8.08
EH&A HEDL
Low High
10.64
11.09
11.68
10.94
10.49
11.09
11.53
10.64
11.09
10.94
11.68
11.68
11.24
1O.49
10.94
11.38
7.07
7.10
6.54
7.68
7.41
8.49
8.24
7.13
7.24
7.36
7.63
7.77
7.86
6.76
7.65
8.20
8.31
6.79
6.57
7.50
7.07
7.95
7.95
8.24
8.31
7.86
8.42
7.91
7.47
8.62
8.75
7.22
7.67
6.97
7.11
8.07
7.82
8.16
8.20
7.96
8.46
8.34
8.90
8.77
8.22
8.28
8.32
8.46
8.50
8.54
8.04
8.46
8.75
8.80
8.05
7.97
8.41
8.17
8.60
8.60
8.75
8.82
8.51
8.83
8.57
8.35
9.00
9.21
8.24
8.42
8. 2O
8.23
8.62
8.49
the average air temperature is not
satisfactory for high thermal
loading ponds which correspond
to 1.0 acre/MWe or less.*
5. The Brady model appears to result
in more credible cooling pond
forced evaporation losses than
does the Harbeck nomograph
because it considers the actual
*For a plant with a thermal efficiency of 38 percent,
1.0 acre/MWe is approximately equivalent to 2.5
. MWt/acre.
thermal loading of the pond in
order to estimate the pond surface
temperature.
6. A general conclusion cannot be
drawn as to whether a cooling
tower or a cooling pond is more
water consumptive.
Costs of Cooling System
Alternatives
Cost information on various cooling
system alternatives wascompiledforall
18 water resource regions in the
conterminous United States for both
nuclear and fossil power plants. The
cooling system alternatives considered
included two wet/dry tower systems
designed to conserve water. All costs
were adjusted to 1978 dollars.
Two main categories of costs were
presented in the UE&C report: (a) the
capital cost for equipment (i.e., cooling
device, circulating water system in-
cluding condenser and electric equip-
-------
Table 3. Consumptive
Water
resource
region
1. New England
2 Mid-Atlantic
3 South Atlantic-Gulf
4 Great Lakes
5. Ohio
6. Tennessee
7. Upper Mississippi
8. Lower Mississippi
9. Souns-Red-Rainy
10. Missouri Basin
11 Arkansas-White-Red
12. Texas -Gulf
13 Rio Grande
14. Upper Colorado
15. Lower Colorado
16. Great Basin
17 Pacific Northwest
18. California
Water Use for Cooling Ponds and Once-Through Cooling
Multipurpose (natural} pond (10*gal /day) Single-purpose (manmade) pond (10B gal /day}
EH&A HEDL EH&A HEDL
Location (1977) Low High 1 ac/MWe 2 ac/MWe Low High
Boston, MA
Concord, NH
Bangor, ME
Richmond, VA
Philadelphia, PA
Tampa. FL
Atlanta, GA
Detroit, Ml
Cleveland, OH
Columbus, OH
Louisville, KY
Knoxville, TN
Chattanooga, TN
Twin Cities, MN
St. Louis, MO
Jackson, MS
New Orleans, LA
Bismarck, NO
Duluth, MN
N. Platte, NE
Great Falls, MT
Tulsa, OK
Garden City, KS
Dal/as, TX
Houston, TX
Albuquerque, NM
El Paso. TX
Farmington, NM
Gr. Junction, CO
Phoenix, AZ
Yuma, AZ
S. Lake City, UT
Reno, NV
Seattle. WA
Portland, OR
Los Angeles, CA
Sacramento, CA
3.66
4.49
5.53
4.01
4.01
4.61
508
4 13
4.37
425
5.55
484
4.72
4.61
401
4.61
573
584
535
658
624
776
7.07
583
592
6 13
6.46
661
6.70
542
6.48
7.18
7.31
5.25
5 13
605
5.43
682
6.48
7 14
742
6.26
694
6.25
600
741
7.32
5.55
6.07
585
6.07
696
6.69
721
7.34 5 10
6.88
845 573
800
9.42 757
783
7.46
747
793 5.17
8.24
8.61
8.62
6.90 5.41
8. 19 6.11
8.93 6.56
9.07
6.69 6.89
6.52
766 855
7.00 7. 19
844
7.89
8.65 9 89
904
8.07 12.90
8.58
8.06
7.82
9.37
9 10
735 7.52
8.12
779 481
799
8.90 891
8.45 10.29
1.81*
438 2 84*
2.39*
5 1 1 5.47*
3.82*
6 50 10 28
8.62
635
5.22*
459 5.72*
5.73*
4.38*
524*
4 71 7.38
5.36 8 04
5.82 7.16*
2.30*
5,51 8. 10
542
646 948
5.72 859
878
10.71
7.72 10.65
9.78
8.87 11.30
13.65
11.44
980
1477
14.90
612 880
11 11
4.71 1.20*
-755*
6.46 1 1 28
7.45 1005
5.91*
6.34*
589*
8.07*
7 19*
13.28
954
7.62
7.23*
7.78*
8.00*
7.85*
8.12*
9.02
8.88
890*
7.40*
11.59
662
13.90
12.31
10.50
17.85
16.16
1251
18.27 I
2492
1868
14.17
26.11
2743
11.92
17.11
623*
3.45*
16.47
13.98
* Average precipitation exceeds natural evaporation.
ment) and (b) the total evaluated cost,
including both the capital cost and the
capitalized penalty cost.
While the capital costs can be easily
identified, the penalty costs are less
definitive and can vary considerably,
depending on the economic factors,
analysis methods, and penalty items
included. Because most cost data
sources reviewed lacked much of the
design, performance, and cost informa-
tion needed to calculate these penalty
costs, only the capital cost estimates
were reported for most of the 18 water
resource regions. The total evaluated
costs were available for less than half of
the water resource regions and were
extracted from references where the
basic information needed for proper
adjustment were available.
Conclusions
1. There is no discernible trend of
capital costs either by water
resource regions or by the types of
cooling systems (see Table 4).
2. With respect to total evaluated
costs (sum of capital and capital-
ized operating costs), the observed
cost trend for a specific plant
generally remains as expected: a
dry cooling system hasthe highest
cost, a once-through cooling
system has the lowest cost, and
conventional closed-cycle cooling
systems, including wet/ dry tower
systems, lie between the two
extremes.
3. The economic impact of using dry
tower cooling systems to conserve
water may be significantly re-
duced by the use of wet/dry tower
cooling systems. When water is
available, however, wet tower
cooling systems will continue to
be the economic choice under
most circumstances.
-------
^
§
\
(ft
§)
10
o
cf
~
Q.
u>
Z
Q
0
Qv
^
16
15
14
13
12
11
10
9
8
7
6
5 .
4
3 -
2
7 .
0
Coo/ing Tower with Slowdown Returned
O Multipurpose Coo/ing Pond
* *
*
0 0
O
o ° ooo
00°^ 0
o
Figure 1.
' 23 45 6 7 8 9 10 11 12 13 14 15 16 17 18
Water Resource Region
Comparison of water consumption rates obtained by EH&A study for
multipurpose cooling ponds and cooling towers with blowdown returned.
Water Availability for Steam-
Electric Generation and
Other Uses
At present, the availability of en-
vironmentally acceptable sites for
electric power generating plants,
whether fossil or nuclear, is strongly
influenced by the availability of cooling
water. The water availability presented
in the UE&C report was based on the
then-unpublished results of a compre-
hensive 4-year study completed in 1978
by the U.S. Water Resources Council
(WRC). In the study, the conterminous
U.S. is divided into 18 water resource
regions (see Table 2 for listing) and
further subdivided into 99 subregions.
The results of the study included actual
water consumption and availability data
for 1975 and estimated values for 1985
and 2000.
Conclusions
The major conclusions on water
availability are:
1. Under dry-year conditions, there is
insufficient water in most of the
18 water resource regions to
satisfy all users at current and
projected rates of use. This water
shortage situation is particularly
critical in the Southwest.
2. Relative to total water consump-
tion, the percentage consumption
for steam-electric generation was
1.23 percent in 1975 and is
estimated to be 2.1 percent in
1985 and 7 22 percent in the year
2000.
3. The greatest potential for consump-
tive water savings lies in the
agricultural sector. Since this
sector is the largest consumptive
water user in most regions, sub-
stantial water savings can be
obtained with only small percent-
age reductions in this user cate-
gory.
Legal Constraints and Their
Impact on Consumptive
Water Use
To determine the availability of water
from a certain water body for consump-
tive use by power systems, it is first
necessary to examine the laws and
regulations that govern water allocation
and use. The legal right to use water
from a water body is the first determ-
inant of water availability for a partic-
ular use, even though water may be
physically available.
The review study compiled the major
features of the institutional framework
within which water for energy conver-
sion uses will be sought and developed.
The features include the constitutional
basis for water laws, important Federal
statutes, international treaties, and
interstate compacts. Also addressed are
the potential impacts of Federal water
rights, Indian water rights, and State
water laws and policies.
Conclusions
The major conclusions are:
1. There is no simple way to classify
all the differing laws and accom-
panying rules and regulations,
since they vary from state to state
and depend on court decisions.
Constraints on the legal availabil-
ity of water form a complex web
which involves Federal rights,
Indian rights, State rights, riparian
rights, appropriation rights, bene-
ficial uses, international treaties,
and others. Disregard of or any
attempt to abrogate these rights
(or arrangements) is certain to
raise serious objections and entail
lengthy litigation.
-------
19
18 '
17
16 -
15
14
13 '
^ 72
Consumption, 106 g
**k ~**
Qj Co O ~»
1 ''
6
5
4
3
2
1
0
Cooling Tower with Slowdown Returned
1 Single-Purpose Cooling Pond
.-'.*. *.f
1 I
III 1
Figure 2.
2345 6 7 8 9 10 11 12 13 14 15 16 17 18
Water Resource Region
Comparison of water consumption rates obtained by EH&A study for
single-purpose cooling ponds and cooling towers with blowdown
returned.
2. No comprehensive body of law
exists, either on the Federal or
State level, on the regulation and
consumptive use of water. Present
Federal and State laws and regu-
lations need codification and, in
some cases, rewriting to enhance
understanding and to meet societal
needs.
8
-------
I
I
c
o
I
3
v>
C
o
Cj
17
16 -\
15
14
13 "\
12
11
10
9 '
8
6
5
4 '
3
2 '
1 '
0
\ !
j Coo/ing Tower with Slowdown Returned
Multipurpose Coo/ing Pond
123 45 6 7 8 9 10 11 12 13 14 15 16 17 18
Water Resource Region
Figure 3. Comparison of water consumption rates obtained by HEDL study for multipurpose cooling ponds and cooling
towers with blowdown returned.
9
-------
22 '
27 '
20
19 '
18
17
16 '
15 '
t M "
\
1 13 '
o
* 72
c'
.0
i 77
S
3
£ 9 '
i
5
7
6 '
5
4
3 '
2
1
1
1
Cooling Tower with Slowdown Returned
Single-Purpose Cooling Pond
I
o
I
\
I
\
\
I
'!
f
\
1.
!!'
\
2345678
\
!
!
\ \
II
t
\
I
!
|
I
9 10 11 12 13 14 15 16 1
!''«
; 18
Water Resource Region
Figure 4. Comparison of water consumption rates obtained by HEDL study for
single-purpose cooling ponds and cooling towers with blowdown
returned.
10
-------
Table 4. Capital Costs of Cooling System Alternatives - Fossil Plants ($/kW, 1978 dollars)
Dry Tower
Water Resource Region
High Back Low Back
Once Through Wet Tower Cooling Pond 40% Wet/Dry* 10% Wet/Dry* Pressure Turbine Pressure Turbine
1 . New England
2. Middle Atlantic
3. Atlantic-Gulf
4. Great Lakes
5. Ohio
15 22-28
26-27
19 24-26, 28,
22-37
25, 21-23,
44-57
25, 19-25
39
63
22.44
43 56
53
27
34-38
45
62
39-45
85
6. Tennessee
7. Upper Mississippi
8. Lower Mississippi
9. Souris-Red-Rainy
JO. Missouri Basin
11. Arkansas-White-Red
12. Texas-Gulf
13. Rio Grande
14. Upper Colorado
15. Lower Colorado
16. Great Basin
17. Pacific Northwest
18. California
36
19
23
21-26.
22-24
25,25-27
62
38-43
39-44
52-67
47-52
49-57
69
30
47
46-49
73-87
59
101-103
88-108
*40% (10 %) wet-dry has 40% (10%) of the water consumption of a wet system designed to reject the same quantity of heat.
Summary of UE&C Report:
Nonwater Quality Impacts of Closed-Cycle
Cooling Systems
And the Interaction of Stack Gas and Cooling
Tower Plumes
General Description
The objective of this study was to
collect, analyze, and correlate the
information available since 1973 on the
following topics: (a) the environmental
impacts on biota of drift from evaporative
saltwater or brackish water cooling
towers, (b) the impacts of cooling tower
plumes on weather, and (c) the en-
vironmental impacts of the interaction
of cooling tower plumes with the
combustion gases emitted from power
plant stacks.
The air emitted from a wet cooling
tower is characterized by a water vapor
plume exiting the top of the tower. This
thermally buoyant plume typically
contains 1 to 3 percent of the water
circulating through the cooling tower. In
general, it is a visible plume which can
rise to several times the height of the
^cooling tower, depending on the pre-
vailing meteorological conditions. When
appropriate meteorological conditions
exist, the plume may cause an increase
in the frequency of ground level fogging.
The direct contact process of heat
transfer in a wet cooling tower allows
macroscopic droplets of condenser
cooling water to become entrained in
the air flowing out of the tower. These
droplets, commonly referred to as drift,
begin to fall to the ground once the
plume leaves the tower. During winter,
these droplets can freeze on the
surfaces of nearby structures and
transportation corridors. Since the drift
droplets contain the same chemical
constituents as the water circulating
through the tower, their presence in the
atmosphere may also have significant
environmental consequences.
The presence of wet cooling towers,
especially large hyperbolic natural draft
cooling towers, at fossil-fuel-fired
steam-electric generating stations
raises the possibility that the cooling
tower plumes will interact with the
combustion gases emitted from the
station's smoke stacks. The commin-
gling of a cooling tower plume with a
stack plume, which contains high
concentrations of sulfur oxides, nitrogen
oxides, and fly ash in addition to carbon
dioxide and water vapor, can potentially
enhance the formation of acid mist in
the atmosphere, causing adverse en-
vironmental impacts.
Cooling Tower Drift and Its
Environmental Impacts
A major concern of using saltwater
and brackish water cooling towers is
that the drift from these towers will
produce significant adverse environ-
mental impacts, especially to the biota
-------
in the vicinity of the towers. An
assessment of the potential environ-
mental impacts on biota should include
a determination of:
1. The naturally occurring salt depo-
sition rates and the atmospheric
salt concentrations in the vicinity
of the cooling tower.
2. The total salt deposition rates and
the atmospheric salt concentra-
tions that result when the cooling
tower is in operation.
3. The salt concentration levels in
vegetation before and after the
cooling tower is m operation.
4. The salt tolerance of each indig-
enous species of the biota.
5. The cumulative effects of salt load
on biota.
While the UE&C study reviewed all
available information, it emphasized
available field studies. These field
studies are summarized below.
Cooling Tower Drift and
Deposition Rate Measurements
Available field studies include: (a) the
B.L. England Station of the Atlantic City
Electric Company, (b) the Chalk Point
Station of the Potomac Electric Power
Company, and (c) the Turkey Point
Station of the Florida Power and Light
Company
B.L. England Station, Unit 3
The B.L. England Station is located
near Marmora, New Jersey, on Great
Egg Harbor Bay and about 4.5 miles
from the Atlantic Ocean. The station has
two coal-fired units and one oil-fired
unit with a generating capacity of 476
MWe. The oil-fired Unit 3 is cooled by a
closed-cycle cooling tower system
which employs a hyperbolic counterflow
natural-draft cooling tower designed
and built by Research-Cottrell. The
tower has a base diameter of 180 feet
and a height of 208 feet. The design
circulating water flow rate is 200,000
gpm, and the concentration of total
dissolved solids in the circulating water
ranges from 24,000 to 32,000 ppm.
Field measurements of drift rate were
conducted by the tower manufacturer.
The mean drift rate from 15 measure-
ments was 0.000424 percent of the
circulating water flow rate with a
standard deviation of 0.000123 percent.
Field measurements of atmospheric
salt concentration and drift deposition
rate were conducted by Environmental
Systems Corporation. Two studies were
undertaken during two consecutive
72
years. The first study performed mea-
surements during the year when the
tower was not in operation, and the
second study performed measurements
during the year when the tower was in
operation.
Although the field data collected at
the B.L. England Station indicate that
the saltwater cooling tower increases
the atmospheric salt concentration and
deposition rate, anomalies exist, indi-
cating uncertainties with regard to the
effects of the cooling tower. For in-
stance, during the periods when the
cooling tower was in operation, the
annual average salt deposition rates
measured at every sampling location
increased relative to those measured
during the pre- and operational periods.
However, the annual average salt
concentrations at these sampling loca-
tions did not show a similar increase
when the cooling tower was in opera-
tion. Possible explanations of this
apparent anomaly include: (a) the
differences in meteorological condition
between the periods (1973-1974 and
1974-1975) when the pre-operational
and operational measurements were
conducted, (b) the difference in sampling
times for the deposition rate measure-
ments (approximately 24 hours) and the
salt concentration measurements (ap-
proximately several hours), and (c)
"blow-through" occurred when high
winds swept through the cooling tower
basin carrying with them droplets of the
circulating water.
Chalk Point Station, Unit 3
The Potomac Electric Power Com-
pany's Chalk Point Station is located at
the confluence of the Patuxent River
and Swanson Creek, about 40 miles
southeast of Washington, D.C. The
station has three operational units with
a total generating capacity of 1329
MWe. Units 1 and 2 are coal-fired base-
load units with once-through cooling.
Unit 3, on which the Chalk Point Cooling
Tower Project (CPCTP) has been con-
ducted, is an oil-fired cycling unit. The
unit is cooled by a closed-cycle system
using a crossflow natural-draft cooling
tower designed by the Marley Company.
The cooling water for this system is
taken from the discharge of the once-
through cooling for Units 1 and 2 and
the flow rate is 260,000 gpm. The
salinity of the circulating water ranges
from 14,000 ppm to 21,000 ppm during
the summer. The tower has a base
diameter of 374feet and a heightof 400
feet. The CPCTP is an on-going and the
most comprehensive study to date of
cooling tower drift and its environmental
effects. Summarized are the tests and
results obtained so far.
Two independent groups (Johns
Hopkins University and Environmental
Systems Corporation) measured the
cooling tower drift at ground level. The
two sets of deposition rate measure-
ments differed by factors of 5 to 10 for
the same distances from the cooling
tower. Droplets less than 100/urn were
not necessarily recorded, although they
were observed. The explanation of this
difference in deposition rates may be
that one of the studies was performed
under slightly stable conditions. Argonne
National Laboratory has attempted to
use these data to validate mathematical
salt deposition models.
Turkey Point Station
Florida Power and Light Company's
Turkey Point Station is located about 30
miles south of Miami, on Biscayne Bay.
It consists of two 430 MWe fossil-fueled
units and two 730 MWe nuclear units.
Condenser cooling of all four units is
provided by a closed-cycle salt water
cooling canal system. The salinity of the
circulating water ranges from 3,0001
ppm to 30,000 ppm. "
A single-cell mechanical-draft cross-
flow tower with improved drift eliminator
design was also installed at the station
to evaluate the effects of drift from a
mechanical-draft cooling tower. The
tests indicated that the increase in
ambient salt concentrations resulting
from the tower operation was less than
the instrument accuracy of 3 to 5
/ug/m3.
Field Studies of Biological
Impacts of Cooling Tower Drift
At Chalk Point and Turkey Point
Stations, studies have also been con-
ducted to assess the vegetative impacts
of salt drift from the cooling towers. The
UE&C study found no field study
investigating the impacts of salt drift on
animal species. The available field'
studies on vegetative impacts are
summarized below.
Chalk Point Station-
Studies have been conducted at the
station to investigate the cumulative
short-term effects of salt loading on
native vegetation. Four tree species
native to the Chalk Point area wer
-------
studied: Virginia pine, black locust,
sassafras, and dogwood. The test data
indicate that the naturally occurring salt
loading (Na+ and CT) on the four tree
species does not approach the threshold
concentrations of these ions when mar-
ginal necrosis occurs. The threshold
concentrations of Na+ and Cl~for many
woody species are 2000 ppm and 5000
ppm, respectively. Operational test data
obtained in 1975 for the four tree
species did not show significant in-
creases of Na* and Cl~ concentrations.
Similar results were obtained for
soybea ns, tobacco, a nd corn with rega rd
to the naturally occurring salt loading
and the increase in salt concentration
resulting from tower operation.
Turkey Point Station
At Turkey Point, cultivated plants
(bush beans) were introduced to the
area near the station's saltwater cooling
tower. Using glass slide collectors,
different salt deposition rates were
observed on the windward and leeward
sides of the plants. The excessive
accumulation of Na+ and Cl~ in the
tissue on one side of a plant's foilage
can result in "molding;" i.e., stunted
growth on the windward side of the
plant.
UE&C found an abundance of in-
formation on the effect of salt aerosol
deposition on vegetation. But, the
results generally cannot be directly
applied to power plant cooling towers
because the experimental conditions do
not necessarily correspond to power
plant environments.
Conclusions
Based on the information compiled,
the following major conclusions were
drawn concerning the measurements
and the biological impacts of drift from
saltwater and brackish water cooling
towers.
1. Cooling tower drift and salt depo-
sition apparently do not increase
ambient salt loading to the extent
that a significant impact has been
observed or is expected. Field data
from the operational Chalk Point
and B.L. England Stations indi-
cated that, although drift and its
deposition from cooling towers
may be significant within the
immediate region of the station,
contributions from cooling towers
beyond 1 km may be inseparable
from the natural salt loading of the
local region.
2. The degree of salt injury to biota
from cooling tower drift and salt
deposition cannot be firmly de-
termined using presently available
data.
Weather Modifications by
Evaporative Cooling Towers
The atmospher c effects of evapor-
ative cooling towe rs operating at steam-
electric generatin j stations include the
potential modific
locally with respe
and intensity o
conditions, (b) tie enhancement of
cloud formation,
ment of precipita
ind (c) the enhance-
Fogging and I sing
Fogging and
resulting from visi jle, wet cooling tower
plumes travelling
have been obse
tower operation.
derived various m ithematical models to
predict the frequ
fogging and icing from wet cooling
tower operation, a nd these models have
been summarized
reported. Howeve
the operation of a
Tennessee Valley
served during a 5-
no evidence wa
cooling towers a
Power Station of
Power Service Co
The occurrence
fogging is due
of the water vapo
tion of the weather
it to: (a) the frequency
fogging and icing
on.
ing, as phenomena
near ground level,
ved during cooling
Meteorologists have
ncy of occurrence of
in a model validifica-
tion effort by Arg >nne National Labor-
atory. Although t le model predictions
differ, they indie
icing are not likely
km radius of the i
Field observat
icing near wet coo ing towers have been
te that fogging and
to occur beyond a 10-
ooling tower.
ons of fogging and
, no ground fog from
natural draft cooling
tower at the Paradise Station of the
Authority was ob-
year study. Similarly,
> found of induced
ground fogging from the natural draft
the John E. Amos
he American Electric
poration.
of icing rather than
:o the presence of
temperatures bek w the freezing points
and drift droplets in
the cooling tower plume. Meteorological
conditions conducive to fogging include
wind speeds greater than 2 m/sec,
stable temperature lapse rates, relative
humidity greater than 95 percent, and
low dry bulb temperatures. These
conditions create downwash, low
dispersion, low evaporation, and con-
densation.
Field studies determined that icing
from the natural draft cooling towers at
the Paradise Station was insignificant.
At the Chalk Point Station, the natural
draft cooling tower caused no more than
a few millimeters of ice on structures
located on the plant site. Only in rare
instances could icing be observed at
distances more than 5 km from the
natural draft cooling towers at Amos
Station.
Cloud Enhancement
The meteorological conditions which
are conducive to the formation of
additional clouds include a stable
stratification of the vertical temperature
profile, high relative humidity, cool
temperatures, and low insolation.
Under these conditions the clouds will
be typically of the stratus variety.
Theoretical considerations indicate
that wet cooling towers can modify
these natural cloud formations. How-
ever, there have been no field studies to
establish the extent of any cloud
enhancement.
Precipitation Enhancement
A maximum of 2.5 cm of snowfall
from natural draft cooling towers at the
John E. Amos Station was measured
when no other precipitation existed in
the area. Snow accumulations were
found as much as 43 km from the
cooling tower, and visibility was re-
stricted to less than 1600 m in the area
where the cooling tower plumes ap-
proached ground level. However, no
statistically meaningful change in
rainfall was found near a 2000 MW
electric generating station with eight
natural draft cooling towers in England.
Convective precipitation enhancement
by wet cooling towers has been ob-
served only once.
Conclusions
The following conclusions were
drawn concerning weather modifica-
tion by evaporative cooling towers.
1. Field observations on the fogging
and icing from natural draft cool-
ing towers suggest that these
effects are not significant envi-
ronmental problems. At most, only
a few millimeters of ice have been
observed to accumulate on sur-
rounding structures, while fog-
ging from natural draft towers has
not been observed under a variety
of meteorological conditions. Fog-
ging and icing from mechanical
draft towers have been observed,
but the tower design significantly
13
-------
affects the probability that the
plume will intersect the ground.
2. Precipitation and cloud enhance-
ment by cooling towers have been
insignificant although they have
been observed in a few cases.
Previous studies indicate that
precipitation and cloud enhance-
ment for any sector downwind
from a tower are increased by only
a few percent.
Cooling Tower Plume and
Stack Gas Interaction
The presence of wet cooling towers,
especially large hyperbolic natural draft
cooling towers, at fossil-fuel-fired
steam electric generating stations
raises the possibility that the cooling
tower plumes will interact with the
combustion gases emitted from the
stations' stacks. The commingling of a
cooling tower plume with a stack plume
containing high concentrations of sulfur
oxides, nitrogen oxides, and fly ash in
addition to carbon dioxide and water
vapor, can enhance the formation of
acid mist in the atmosphere. The
presence of acid mist may be evident as
increased concentrations of aerosols
and fine particulates containing sul-
fates and nitrates in the commingled
plume, acid rainfall, and a reduction in
atmospheric visibility.
A complete description of the inter-
action of a cooling tower plume with a
stack plume requires an understanding
of : (a) the physical, chemical, and
geometrical characteristics of each
plume as it leaves the source; (b) the
meteorological conditions which affect
the rise of each plume in the atmosphere
and the spatial extent of the interaction
of the plume and the ambient at-
mosphere; and (c) the complex physico-
chemical reactions that can take place
when the plumes commingle. At
present, there are insufficient data to
give a comprehensive description and
quantitative assessment of cooling-
tower-plume/stack-plume interactions.
Meteorological conditions conducive
to the merging of a visible cooling tower
plume with a stack plume have been
established reasonably well in the
literature. The longer than usual visible
cooling tower plumes occur during light
winds, high relative humidity (greater
than 80 to 85 percent), low ambient
temperatures, and stable lapse rates.
The light winds and stable lapse rates
ensure low atmospheric dispersion of
the plume. The high relative humidity
reduces the rate of evaporation of the
water droplets in the cooling tower
plume, thereby enhancing the visible
plume length. Finally, the low temper-i
ature may also help to keep the water in
the cooling tower plume in a condensed
state.
Quantitative data concerning acid
precipitation enhancement resulting
from plume/stack gas interaction are
scarce. Existing literature generally
provides unsophisticated explanations
and non-quantitative descriptions.
Three field studies, one at the Chalk
Point Station and two at the Keystone
Station, obtained limited quantitative
data, but these data are site-specific and
cannot reliably be extrapolated to
general circumstances.
Conclusions
The major conclusions concerning
cooling tower plume and stack gas
interaction are:
1. There is no mathematical model
for predicting acid precipitation or
deposition due to cooling-tower-
plume/stack gas interaction.
2. Neither the enhancement of acid
precipitation due to cooling-
tower-plume/stack-gas interac-
tion nor the quantification of
resulting environmental impacts^
has been demonstrated in the
field.
Summary of Versar Report:
Comparison of Model Predictions and Consumptive
Water Use of Closed-Cycle Cooling Systems
The primary objectives of this study
conducted by Versar, Inc. were: (a) to
survey and identify simple generic
models for predicting evaporation rates
from power plant closed-cycle cooling
systems using wet towers and ponds/
lakes, (b) to verify and calibrate these
generic evaporation models with field
data from operating power plants, and
(c) to determine which of the two types
of cooling systems is more water
consumptive in terms of evaporation.
To achieve these objectives the
available technical literature was
reviewed, and available field data were
solicited and acquired for a number of
operating power plants. Evaporation
prediction models satisfying certain
criteria were selected and evaluated
using available field data provided by
utilities and data estimated by Versar
14
when the field data were insufficientfor
the intended analyses. Tower and pond
evaporation rates were then compared
on a regional basis, to determine which
(the tower or the pond) is more water
consumptive.
Selection of Evaporation
Models for Evaluations
The evaporation prediction models
selected by Versar for validation, one for
cooling towers and five for cooling
ponds/lakes, were selected because
they are simple, non-iterative, and
generic (i.e., not intended for site- or
plant-specific applications). For cooling
towers, the model selected was the
mechanical-draft cooling tower model
developed from studies performed for
the Navajo Station in northern Arizona.
The model, called the Leung-Moore
model, is represented by a set of four
nomographs.
Five evaporation prediction models
were selected for cooling ponds/lakes.
The models are identified in the Versar
report as: (a) Marciano-Harbeck (Lake
Hefner); (b) Harbeck-Koberg-Hughes
(Lake Colorado City); (c) Meyer; (d) Brady
et al.; and (e) Harbeck nomograph.
Models (a) through (d) are based on
mass transfer concepts and fit a general
mass transfer equation of the form
noted earlier; see Equation (1). Each has
a different empirically developed wind
speed function. They were selected as
they had been evaluated in an earlier
study for EPA.
The Brady model for cooling ponds
was also used in the HEDL study, while
the Leung-Moore model for cooling
tower and the Harbeck nomograph for
-------
cooling ponds were used in the EH&A
studies. The HEDL and EH&A studies
were summarized earlier in this sum-
mary.
Acquisition of Field Data and
Verification of Evaporation
Prediction Models
To verify the evaporation prediction
models, using available field data of
cooling towers and cooling ponds,
Versar obtained data from 12 utilities in
time to be included in the study. The
data supplied by the utilities include 15
wet cooling tower systems, 7 cooling
ponds, and 1 cooling canal system.
These data are summarized in Tables 5
and 6 for cooling towers and cooling
ponds/lakes, respectively.
The data in Tables 5 and 6 are not
entirely field data. They include some
design data provided by utilities and also
some estimates by Versar. Since in
most cases the data provided were
incomplete for the intended evaluations,
Versar made approximations or estima-
tions to make up for the missing data.
However, the estimations were done on
a consistent basis with the best available
information. The data were then used as
inputs for both the models and the water
budget equations to determine evapora-
tion rates.
When using the material balance
method, the evaporation rate was
determined from the difference between
the inflow to, and the outflow from, the
system. The inflow for towers is the
makeup water; inflow for ponds/lakes
may include stream flow into the lake
(makeup), local runoff, and direct
precipitation. The outflow for towers
may include blowdown and drift, and
that for ponds/lakes may include
blowdown (outflow) and seepage. The
tower drift and pond seepage, being
usually small compared to the other
terms, were neglected in Versar's
evaluation. Local runoff was determined
from direct precipitation by multiplying
it by an estimated runoff coefficient
based on local U.S. Geological Survey
information.
The comparisons of the evaporation
results obtained in the Versar study for
cooling towers and cooling ponds are
shown in Tables 7 and 8, respectively.
Table 7 illustrates the accuracy of the
Leung-Moore model relative to material
balance values. On the average, the
Leung-Moore model predicted evapora-
tion rates to within ±15 percent of the
material balance results when the
power plants with capacity factors over
50 percent were considered. Cooling
tower evaporation rates for peaking and
cycling units (capacity factor below 50
percent) are not well simulated by the
model. Table 8 compares the accuracy
of the five cooling pond/lake models
with material balance values. The
power plants associated with the
cooling ponds/lakes were mostly large
base-load units. The comparison shows
that the Harbeck-Koberg-Hughes and
Meyer models give better predictions
than the other three models. Further,
the Brady model gave closer predictions
than the Marciano-Harbeck model in all
cases and than the Harbeck nomograph
in most cases. The latter three models
underpredicted material balance values
of evaporation rate in all cases con-
sidered.
Regional Comparison of
Cooling System Evaporation
Rates
A primary objective of the study was
to compare water consumption rates of
cooling towers and cooling ponds/lakes
on a regional basis. In actual analysis,
Versar compared the evaporation rates
of the two types of cooling systems and
considered that this also represented
the comparison of water consumption
rates of these cooling systems.*
A total of 20 cooling systems for 16
plants investigated in the Versar study
provided comparisons for 7 water
resource regions as shown in Table 9.
The predicted evaporation rates for
towers are the values calculated by the
Leung-Moore model, and those for the
cooling ponds are the values calculated
by the model which gives the closest
prediction to the material balance value.
The results in Table 9 show that cooling
ponds in these cases evaporate more
water than cooling towers. The relation-
ship is most dominant in the southern
regions (Lower Colorado, Rio Grande,
Texas Gulf) where natural evaporation
rates are high.
Conclusions
The major conclusions drawn from
the Versar study concerning water
evaporation and consumption of cooling
towers and cooling ponds/lakes are:
1. For cooling towers at base-load
plants (blowdown returned and no
makeup impoundments), the pre-
dicted evaporation rates by the
Leung-Moore model are in good
agreement with those determined
from material balances (within ±
15 percent).
2. For cooling ponds/lakes, the
Harbeck-Koberg-Hughes model
(Lake Colorado City) and Meyer
model generally give predictions
within ± 15 percent of the mate-
rial balance results; the Brady
model, Harbeck nomograph, and
Marciano-Harbeck model (Lake
Hefner) give consistently lower
evaporation rates than the mate-
rial balance values. Generally, the
Brady model predictions approxi-
mated the material balance values
more closely than the Harbeck
nomograph and Marciano-Harbeck
model.
3. Cooling ponds/lakes generally
evaporate more water than cooling
towers.
4. The satisfactory comparison of the
cooling tower evaporation rates
(predicted by the Leung-Moore
model with material balance values
developed using field data sup-
plemented with estimations where
needed to complete the evaluation)
suggest that this model can be
used adequately for estimating
cooling tower evaporation rates.
5. Comparisons of the evaporation
rates (predicted with the five
cooling pond/lake models and the
results determined from material
balances developed using field
data supplemented with estima-
tions where needed to complete
the evaluation) indicate that pre-
dictions from the Harbeck-Koberg-
Hughes model and the Meyer
model are preferable over the
other models for estimating cool-
ing pond/lake evaporation rates.
"In the UE&C study, the water consumption of a
power plant cooling system is defined as that
portion of the water removed from and not returned
to the surface water resources of a given area. In
the Versar study, the water consumption is
implicitly defined as evaporation only Both
definitions have been used in previous water
consumption studies
15
-------
Table 5. Cooling Tower Data
Station Name Huntmgton NA VJO North Main - Unit 1
Unit No Unit ! Run I Run 2 Run 3 Run 4 Run 5
Utility Name Utah Power Arizona Public Service Texas Electric Service Company
& Light
Test Period
Plant Capacity, MWe
Capacity factor. %
Unit Heat Re/ection, Btu/kWh
Circulating Water Flow. GPM
Makeup Water Flow. GPM
Slowdown Rate, GPM
Range. °F
Approach. °F
Air Flow Rate. SCFM
Outlet Air Temperature. °F
Approximate Drift Loss, GPM
Station Name
Unit No
Utility Name
Location
IWater Resource Region!
Test Period
Plant Capacity, MWe
Capacity Factor, %
Unit Heat Rejection, Btu/k Wh
Circulating Water Flow, GPM
Makeup Water Flow, GPM
Slowdown Rate. GPM
Range. °F
Approach, °F
Air Flow Rate, SCFM
Outlet Air Temperature, °f
Approximate Drift Loss, GPM
1976
Annual
A verage
400
80
5100
185,800
4,000
320
239
175
18 x 1C*
82-97
372
Unit 1
86
593
5680
43,000
1580
375
25
24
23 x 10s
104
87
Test 1-A 11 hrl Test 2-A /2 hrsl 1 Week Performance Test
August 6, 1 977 August 20, 1977 January 21 -26, 1960
750
107
4480
145,326
3.482
0
281
20.3
28x10'
948
293
Unit 2
August,
90
855
5715
42,000
1484
350
25
20
3x10*
97
84
750 ss as as ss as as as as
100 48 63 35 76
4480 6018 5979 6315 5948
147,306 66.244 63.116 64.092 63,429
3,432 508 535 553 509
0 201 250 250 141
277 78 100 60 120
227 159 219 137 222
29x10' 35x10° 35 x 10s 35 x 10? 3 5 x 10*
92 0 66 71 5 64 78
295
Newman/Rio Grande Stations
Unit 3 Unit 6 Unit 7 Unit 8
El Paso Electric Company
1977 July. 1977
110 50 50 165
982 305 585 48
5310 6545 5615 5150
42,500 36,800 28,350 56,300
1672 500 60S 1627
397 145 175 407
28 10 15 22
18 12 16 17
41 x 10s 35 x 10s 30 x 10t 87x10°
97 86 91 94
85 74 57 113
8585
825
6063
64.189
644
141
140
221
35 x 70*
84
Moses
Units 1 and 2
Run 6
Permian
Texas Electric
6 hours
November 5, 1 958
8585
100
6119
63, 765
710
148
164
233
35 x 10s
91 5
Couch
Units 1 and 2
100
Not Known
4788
69.550
704
0
138
172
33 x 10s
92
12
Lynch
Units 1, 2 and 3
Arkansas Power Arkansas Power Arkansas Power
and Light and Light and Light
1976
Annual Data
126
11
76OO/II
7575(21
79.650
1030
221
84
14
89 x 10s
66
160
1976
Annual Data
161
29
7370(11
6650(21
1 16.500
622
160
12
14
11 3 x 10°
83
466
1976
Annual Data
239
12
9500(11
8090/21
7950(3)
164.500
1186
0 1
98
14
125 x 10°
68
66
I
\
16
-------
Table 5. (continued)
Station Name
Unit No
Utility Name
Location
Water Resource Region)
Homer City Plant
Pennsylvania Electric Company
(Natural Draft Tower)
Clay Boswell Plant
Unit 3
Minnesota Power & Light Company
Koshkonong Plant*
Unit 1
Wisconsin Electric
Power Company
Test Period
Plant Capacity, MWe
Capacity Factor, %
Unit Heat Rejection, Btu/kWh
Circulating Water Flow, GPM
Makeup Water Flow, GPM
Slowdown Rate. GPM
Range, °F
Approach, °F
Air Flow Rate, SCFM
Outlet Air Temperature. °F
Approximate Drift Loss, GPM
January
1328
4941
5238
205,500
9186
2595
349
48
1238x 10s
93
206
1977 Average
April
1328
3494
5576
205,500
8889
2660
281
24
825x10*
92
206
July
1328
5735
5685
205.500
14,150
2838
283
18
1444 x 10*
105
206
1977 Average
January August Annual
350
86
5130
130.800
-
-
20
329
12 x 10s
643
13
350
93
4965
130,781
2616
- 500
156
21 8
12 x 10*
895
13
900
100
7383
524.100
12,500
1850
26
18
402 x 1O*
82
26
*AII of the data given are design values
Table 6. Cooling Pond/Lake Data
Station Name
Unit No
Utility Name
Test Period
Plant Capacity, MWe
Capacity Factor, %
Unit Heat Re/ection, Btu/kWh
Circulating Water Flow, GPM
Flow Rate into Pond, GPM
Flow Rate Out of Pond, GPM
Cooling Range, °F
Condenser Makeup Water
Temperature, °F
Effective Pond Surface
Area for Cooling Acres
Water Volume of Pond,
Acre-feet
Drainage Area, sq mi
Cholla Morgan Creek
Arizona Public Texas Electric
Service Company Service
1974-1976 1959-1960
Annual Average
120 102
70
4820
27.800 493.714
1696 6,075
313 860
Not Given
569 68
340 1100
Not Given 31,000
Not Given 326
Kincaid
Commonwealth
Edison Company
1977
Annual Average
1319
34
5200
479,981
28.800
19,300
138
62 O
2400
33,500
76.6
Powerton
No 5 No 6
Commonwealth Edison
Company
1973
Annual
840
51 7
454O
690,562
19.666
14,772
188
61 5
1426
15,600
1977
Annual
945
47 1
4540
690,562
19,666
14,772
193
606
1426
15.600
Belews
HB Robinson Creek
Carolina Power Duke Power
and Light Company Company
1975-1976
Average
885
67
4900
500.923
131.202
125.232
14 1
71 6
2250
41,000
173
1977
Annual Average
2286
66
4225
1.050,332
26,222
11,381
1840
679
3553
176,000
709
Mt Storm
Virginia Electric
and Power Company
January
1977
1662
69 (No. 11,
4280
889,020
182,743
7,676
333
422
1130
49.000
30
July
1977
1662
61 (No 21, 35 4 (No 31
4280
889.020
142.378
3.124
33.3
831
1130
49,000
30
17
-------
Table 7. Comparison of All Cooling Tower Evaporation Rates, as Calculated and Normalized
Plant/Unit Size (MW)
Huntmgton/400
Nava/o/750
N Mam/85
Permian/ 100
Newman-1 /B6
-2/90
-3/1 10
Rio Grande-6/50*
-7/50
-8/165'
Moses/ 126"
Lynch/ 239'
Couch/ 161'
Homer City/1328
Clay BosweH/350
Koshkonong Nuclear/900
Time
Period
annual
hourly-summer
hourly-summer
hourly -summer
August
August
August
July
July
July
annual
annual
annual
January
July
January
August
annual
annual
Material
Balance
(cu m/mmj
125"
126
/!22f
1 46
30
42
40
45
1 1
1 4
42
245
4 16
1 62
169
405
(18 6f
795
4O
Model Pre-
diction
(cu m/mml
128
138
1 96
3 1
37
39
45
24
2 1
65
72
(0 851"
144
(1 71'
76
12 2 f
147
260
561
841
42
Ratio
Model/
Material
Balance
1 02
1 10
1 34
1 03
088
098
1 00
2 18
1 50
1 55
30
10 34f
35
(04 If
4 7
11 35)°
087
067
(1 44)"
OSS
1 05
Normalized Evaporation
cu m/mm-MW
Material
Balance
0039
0016
0025
OO30
0082
OO52
OO42
0072
0050
0053
0177
0 145
0035
0030
0054
10 026 f
0031
OO44
Model
0040
0018
0034
0031
0072
0051
0042
0 157
0075
0082
0531
(0 061)'
0508
(0047f
0 163
(0 059 f
0026
0036
0019
0026
0046
Normalized Evaporation
cu m/106 kcal
Material
Balance
1 84"
085
OSS'
099
1 49
344
2 17
1 86
262
2 11
245
554
493
1 02
1 19
262
(1 20f
1 50
1 43
Model
1 88
096
1 32
1 54
302
2 15
1 86
573
3 17
3 79
165
(i 92r
173
11 60 f
477
n 73?
1 04
1 68
086
1 24
1 50
range 1 50-067 (baseload plants)
d Units with capacity factors less than 50 percent
b Based on constant outlet air temperature
c Marley test results
d Gilbert A ssoc curves
6 Results X capacity factor
Table 8.
Summary of Coo/ing Pond/Lake Mater/a/ Balance and Computer Model Evaporation Values on as-is
and Normalized Bases
Normalized Evaporation Rate
for Model Giving Closest Prediction
Plant/Unit or
Station Size (MW)
Cholla/120
Morgan Creek/ 102
(equivalent}
HB Robinson/885
Belews Creek/
2,286
Mt Storm/ 1,662
Kmcaid/1.319
Powerton/840
Material
Time Balance
Period feu m/min)
July
annual
August
annual
August
annual
August
annual
January
July
August
annual
August
annual
11973}
69
298
21 0
765
446
99 1
909
362
18.5
OH
68
43
226
122
381
260
685
378
77
109
44 7
264
188
126
Model
Predicted
feu m/mm)
QC
99
63
333
179
560
383
101
555
112
162
658
390
276
ISO
QM
87
60
30 O
150
629
402
109
587
106
243
631
344
261
157
OB
76
53
256
139
340
835
465
100
192
51 5
301
21 5
140
HN OH
50 062
O76
147 058
050
268 OSS
069
488 042
198 O 73
140 068
Evaporation
Rate Ratio Area/Power
Mode// Material Balance lacres/MWI
QC
091
1 12
085
073
086
1 02
061
1 08
097
QM
087
1 01
0 71
082
090
1 09
065
095
085
QB HN
077 O72
086
066 070
076 065
086
051 054
083 055
076 076
ha/MW
(2831
1 15
W
1 62
12541
1 03
11 641
066
(068)
0275
(1 641
0664
11701
0689
to the Material Balance Value
(cu m/min-MW)
0 1O3
0075
0294
0201
0089
0068
0062
0 039(0 060 f
(0012-0026)
0077
0046
(cu m/mm-ha)
0073
0046
0067
0043
0069
0046
O070
(0.041 /
(0063'
(0 019-0 0241
(0 024-0 0531
0035
cu m/10' kcal
581
370
11 1
760
437
331
345
219
0 45-0 68
066-1 5
351
240
'Based on material balance evaporation
OH - Mamano and Harbeck model {Lake Hefner)
QC - Harbeck et al model {Lake Colorado City)
QM - Meyer model
QB - Brady et al model
HN - Harbeck nomograph plus pan evaporation
18
-------
Table 9. Regional Comparison of Cooling System Evaporation Rate
Water Resource
Region
Lower Colorado
Texas Gulf &
Rio Grande
South Atlantic
Gulf
Upper
Mississippi
Onto
Mid- Atlantic
Plant
Huntington
Navajo
Cholla
Newman-Unit J
Newman-Unit 2
Newman-Unit 3
Rio Grande- Unit 6
Rio Grande- Unit 7
Rio Grande-Unit 8
North Main
Permian
Morgan Creek
H.B. Robinson
Lake Be Hews
Clay-Boswell
Koshkonong
Kincaid
Powerton-Unit 5
Homer City
Mt. Storm
Cooling
System'
T
T
P
T
T
T
T
T
T
T
T
L
L
L
T
T
L
P
T
L
Plant Size
(MW)
400
750
120
86
90
110
50
50
165
85
100
102
885
2.286
350
900
1,319
840
664
1.662
Capacity
Factor
80
100
07
59
86
98
30
58
48
100
12
67
66
93
100
34
47
57
55
Model Predicted/
Material Balance
Evaporation"''
(rrf/mm)
12.8/125
13.8/126
63/6.9'
37/42
39/40
4.5/4 5
2.4/1 1
2 1/1 4
65/42
29/2.5
3 1/30
20 5/20 9"
40.2/44 6
58 7/90 2'
8.4/795"
42/40
34 4/36 2'
8.0/185"
25.9/395"
Summer
Normalized Evap. Rate
(m3/mm-MWI (rrP/W6 Kcalj
0018
0 103
0.072
0.024
0025
0.292
0.075
0 '01
0.034
0031
0.29
0089
0.062
0026
0.036
0.072-0.026
1 68
5.81
2.88
200
1 86
573
3 17
3.96
1 89
1 54
11 1
4.4
35
1 35
1 68
0.66-1 5
Annual
Normalized Evap Rate
fm3/mm-MW> (m3/106 Kcal)
004
0.075
0207
0068
0.039
004
0077
0046
003
1 88
3.70
7.60
337
2 19
1 50
351
240
1 40
"Cooling Tower IT); Cooling Pond IP); and Cooling Lake (L)
''For cooling towers the Leung and Moore model was used. For cooling ponds, the Harbeck-Koberg-Hughes, or Meyer model, or the Harbeck nomograph was used
dependent upon which model more closely approximates material-balance values
"Annual values are shown, except for performance test results on cooling towers which are based on full capacity test.
"Summer value
"Harbeck-Koberg-Hughes model
'Meyer model
M. C. Hu, G. F Pavlenco, and G A. Englesson are with United Engineers and
Constructors, Inc. Philadelphia, PA 19101,
Theodore G. Brna is the EPA Project Officer (see below).
The complete report, entitled "Executive Summary for Power Plant Cooling
System Water Consumption and Nonwater Impact Reports," (Order No
PB 81-231 474; Cost: $8.00, subject to change) will be available only from.
National Technical Information Service
5285 Port Royal Road
Springfield. VA 22161
Telephone- 703-487-4650
The EPA Project Officer can be contacted at:
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
19
&U. S. GOVERNMENT PRINTING OFFICE: I98I/559-092/3326
-------
.
5) w
TJ
I
2
5
0= S?
^ -t< (D
3 3 s
'
iu T> m -n
-ocS 3 D
------- |