ORDES
Volume lll-H
Special Study Report
Pollutant Transport Models
for the ORBES Region
R. E. Bailey, R. G. Barile, D. D. Gray, R. B. Jacko,
P. O'Leary, R. A. Rao, and J. E. Reinhardt
Purdue University
May 15, 1977
PHASE I
OHIO RIVER OASIN ENERGY STUDY
-------
OHIO RIVER BASIN ENERGY STUDY
VOLUME III-H
SPECIAL STUDY REPORT
POLLUTANT TRANSPORT MODELS
FOR THE ORBES REGION
R. E. Bailey
R. G. Barile
D. D. Gray
R. B. Jacko
P. O'Leary
R. A. Rao
J. E. Reinhardt
Purdue University
May 15, 1977
Prepared for
Office of Energy, Minerals, and Industry
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D. C.
Grant Number R804849-01
-------
CONTENTS
1- INTRODUCTION III-H-1
2. S02, N02 AND PARTICULATES
R. B. Jacko III-H-3
2.1. INTRODUCTION III-H-3
2.2. DESCRIPTION OF THE MODEL III-H-3
2.3. COMPUTER MODEL RESULTS III-H-5
3. COOLING TOWER PLUME MODELS
R. E. Bailey and J. E. Reinhardt III-H-11
3.1. INTRODUCTION III-H-11
3.2. DESCRIPTION OF PLUME MODELS III-H-11
3.3. CONCLUSION III-H-20
4. DISPERSION OF RADIOACTIVITY
R. E. Bailey and P. O'Leary III-H-27
4.1. INTRODUCTION III-H-27
4.2. ATMOSPHERIC DISPERSION OF RADIONUCLIDES III-H-32
4.3. TERRESTRIAL TRANSPORT OF RADIONUCLIDES III-H-34
4.4. ESTIMATION OF EXTERNAL RADIATION DOSE TO MAN III-H-37
4.5. ESTIMATION OF INTERNAL RADIATION DOSE TO MAN III-H-41
4.6. COMPUTER MODELING AND SAMPLE PROBLEM III-H-43
4.7. DISCUSSION OF MODIFICATIONS III-H-49
5. REFINING AND CONVERSION
R. G. Barile III-H-51
5.1. INTRODUCTION - A TYPICAL PROCESS III-H-51
5.2. ALTERNATIVE DATA FOR COMPARISON III-H-57
5.3. MODELING III-H-62
5.4. RESEARCH NEEDS III-H-63
6. ACID MINE DRAINAGE
R. A. Rao III-H-67
7. AQUATIC THERMAL PLUMES
0. D. Gray III-H-71
7.1. INTRODUCTION III-H-71
7.2. FAR-FIELD MODELS III-H-72
7.3. ASSESSMENT OF POTENTIAL FOR MARBLE HILL-TRIMBLE
COUNTY THERMAL PLUME INTERACTION III-H-77
III-H-iii
-------
TABLES
£§ae.
2-1 STACK PARAMETERS FOR TRIMBLE COUNTY
GENERATING STATION III-H-7
4-1 LISTING OF SYMBOLS FOR TERMOD METHODOLOGY III-H-38
4-2 HALF-LIVES OF THREE NUCLIDES III-H-47
4-3 DOSE RATE (MREM/YR) FROM SUBMERSION IN
CONTAMINATED WATER III-H-48
5-1 STACK GASES FROM COAL GASIFICATION-ENGLISH UNITS. . III-H-56
5-2 ENVIRONMENTAL FACTORS FOR SOME EMERGING
ENERGY TECHNOLOGIES III-H-59
5-3 ESTIMATED TRACE ELEMENT DISPOSITION FOR WESCO
COAL-FIRED BOILER 111-H-60
5-4 TRACE ELEMENT DISPOSITION GASIFICATION SECTION. . . III-H-61
6-1 REPRESENTATIVE RATES OF EROSION FROM
VARIOUS LAND USES III-H-68
III-H-v
-------
FIGURES
em.
2-1 PROPOSED LOCATION OF TRIMBLE COUNTY STATION III-H-6
2-2 CONTOURS OF SULFUR DIOXIDE (MICROGRAMS PER
CUBIC METER) III-H-8
3-1 PLUME CONTROL VOLUME III-H-21
3-2 CHALK POINT - DECEMBER 19, 1975 (1200) III-H-22
3-3 CHALK POINT - DECEMBER 15, 1975 (0930) III-H-23
4-1 RADIATION PATHWAYS TO MAN III-H-28
4-2 ATMOSPHERIC DISPERSION FLOW DIAGRAM III-H-29
4-3 TERRESTRIAL TRANSPORT FLOW DIAGRAM III-H-30
4-4 INTERNAL DISTRIBUTION FLOW DIAGRAM III-H-31
5-1 PROCESS FLOW DIAGRAM: SNG FROM COAL III-H-52
5-2 CONTOURS OF SULFUR DIOXIDE (MICROGRAMS PER
CUBIC METER) III-H-64
7-1 COMPUTED AND MEASURED TEMPERATURES AT HUNTINGTON,
WEST VIRGINIA - COLHEAT MODEL III-H-75
7-2 MEASURED-COMPUTED TEMPERATURE SCATTER DIAGRAM
FOR HUNTINGTON, WEST VIRGINIA - COLHEAT
MODEL III-H-76
7-3 OHIO RIVER TEMPERATURE PROFILE, JULY 25 -
PART I III-H-78
7-4 OHIO RIVER TEMPERATURE PROFILE, JULY 25 -
PART II III-H-79
III-H-vi
-------
1. INTRODUCTION
This report describes the program to date in making operational
a set of pollutant transport programs which will be available for the
ORBES project. The general approach used in this "special study"
is to:
1. Collect review and state-of-art papers relating to
pollutant transport—both the modelling and the actual
data.
2. Write a synoptic paper integrating the above.
3. Decide upon the model or technique that will be made
operational here at Purdue to estimate pollutant transport.
4. Make the model or technique operational for use in the
technology assessment.
5. Delineate steps in an order of priority as to what
needs to be done to improve upon existing models and
techniques.
6. Where possible, make improvements.
The pollutants of interest and the respective approaches to
modelling are presented in separate chapters, each with its own
author.
This report is not a final report, for the special study will
continue until September 1, 1977.
III-H-1
-------
2. S(>2, NO2 and Particulates
2.1 Introduction
Burning of coal for the generation of electrical power results in the
release to the atmosphere of a number of pollutants, the most notable of
which are particulates, sulfur dioxide and nitrogen oxides.
The emission of particulates is a function of the type of boiler, ash
content of the coal,and configuration of particulate removal devices located
downstream from the boiler. Sulfur dioxide is a function of the original
sulfur content in the coal. Theoretically, for every gram of sulfur com-
busted along with the coal there are 2 grams of SOg generated:
S + Og SO2
MW=32 64
However, some of the sulfur is oxidized to SO3. Usually about 5% of the
sulfur in the coal is converted into SC^.with the remainder staying in
the SO2 form. The SO3 can be a problem in terms of visible emissions
assuming that most of the particulates have been removed from the stack
gas stream. Sulfates (SO4) formed from further oxidation of the SO2 and
SO3 forms have also been of concern since they have been shown to produce
marked deleterious health effects on test animals. The rate and tempera-
ture of the fuel combustion process in the boiler fixes atmospheric
nitrogen with oxygen to form nitric oxide. The nitric oxide (NO) usually
remains in this form until it is exhausted into the atmosphere at which
time further oxidation to NO2 is enhanced by the relatively oxygen rich
atmosphere. It is the NO2 gas molecule which absorbs uv radiation and
breaks down into nitrogen and an oxygen radical. The oxygen radical,
beinq extremely oxidative, is a precursor of photochemical smog. It is
generally considered that automotive emissions are the predominant con-
tributor to photochemical smog,as has been the classic situation in the
Los Angeles area. This report, however, concerns itself primarily with
the dispersion of SO2 gas in the atmosphere and computer simulation tech-
niques for estimating respective SO2 concentrations at selected receptors.
2.2 Description of the Model
A model which is widely accepted and used in the United States is
based on the premise that the dispersion of SO2 can be approximated with
a normal Gaussian mathematical distribution function. Copious information
in the literature documents its development,but the most well-known pub-
lication which made the model available to practicing enqineers was the
"Workbook of Atmospheric Dispersion Estimates" by Turner (1).
An extension and refinement of the Gaussian model into an extensive
computer simulation of point sources is the CIimatological Dispersion
Model (CDM) developed by the EPA in the early 1970's (2). This particular
model utilizes actual meteorology for the prediction of SO2 corcentrations
on an annual basis. It also uses an improved plume rise equation de-
veloped by Briggs (3) and the joint probability function of wind directions.
111 - H -3
-------
wind speed.and atmospheric stability.
2.2. 1 CDM Model Definition
The basic equation for estimating long-term, ground-level concen
trations, x> due to an elevated point source is
16 n 6
* = 2^x k=l i=l h \ Si]
where: ^-j. is the joint probability function value for stability
J class i in appropriate wind direction j
is average source emission rate for source k
S- is the diffusion function for stability class i and is
1 of the form
s = 27iht7 expT [oj]
or
S = — for a > 0.8 L
u L z
and: x is the distance between source and receptor
H is the effective plume height
U is the mean wind speed at stack height for stability class i
in wind direction j
a is the vertical diffusion coefficient
L is the mixing height
Basic assumptions included above are: (1) SO? decay half Ti'fe is large,
(2) x" independent of oy, (3) mean wind speed u" is sufficient for each
wind direction and stability class (a simplification on EPA's CDM which
considers 6 wind speed classes), (4) mixing height, L, is the upper
limit of dispersion and acts like an impervious barrier, and (5) back-
ground contributions are negligible.
The value for the vertical dispersion coefficient, which is a function
of downwind distance and stability class, was calculated using the model
proposed by McMullen® to approximate the graphs shown in Turner's Work-
book of Atmospheric Dispersion Estimates (1). The model is of the form
c = exp j! + J (In x) + K (In x)2J
where a = horizontal (cr^) and vertical (az) dispersion coefficients
x = downwind distance (km)
I, J, K = empirical constants for given stability class and a
Mean wind speeds at stack height were determined by interpolation of
data collected at five surrounding upper air radiosonde stations located in
Salem, 111., Dayton,0hio, Nashville, Tenn., Flint, Mich., and Peoria, 111.
III-H-4
-------
The interpolation scheme weighs each contribution by the inverse of the
distance between the radiosonde station and the power station under study.
Symbolically, the interpolation is given by
tr
5
E
i=l
S~~
z
i=l
ui
di
where u_ is the interpolated mean stack-top wind speed for the power
station under study, IT is the mean wind speed at the elevation corresponding
to stack height obtained from the radiosonde sounding and d is the distance
between the radiosonde station and the power station. Since radiosonde data
is available only twice daily, linear interpolation was used to estimate
stack height wind speed for those hours between soundings.
2.3 Computer Model Results
The CDM model was then utilized with the meteorology from a southern
Indiana county for 1975 along with the emission characteristics from the
proposed Trimble County Generating Station near Wises landing alonq the
Ohio River,as shown in Figure 2-1. The emission parameters used In this
modeling were taken directly from the EPA1s pre-construction and review
report dated October 14, 1976. Table 2-1 contains the pertinent stack
parameters.
The resulting annual average S0£ isopleths are shown in Figure ?-2.
Note the maximum annual ground level SO2 concentration is 1.1 ug/irr and
occurs 4.6 km northeast of the stacks. Secondary maximums are seen in
other quandrants surrounding the stacks.
It must be kept in mind that the CDM model assumes the surrounding
topography is flat and that the terrain surrounding Wises landing is
anything but flat. Referring back to Figure 2.-1, it is seen that terrain
surrounding the stack is, in plans, at the same elevation as the stack
outlet. Therefore, ground level SO? concentration could be much greater
than indicated by the computer model.
III-H-5
-------
rt7i
frVi
Av, \Wit/,-'sy 1 ' A \ vr£?rrt^=| c^r
-v-'t
~'^j
V/W
Mr
_r£s%i'- c\^. ' v^il $'t MiVife^ ¦" ^"v,'!)): •
rfW ®
%i \
m»^iAc« m<^J};fy 1 ,^>
£4r*:
0
>! • I
/>¦ :viu1. ^ "=• 'v^n
* \wX
r.l& " \\fl/
^W^%&&
.Cto^>y;,
/£TST^\/ *C."i ^-' 1
* .-i^" - ' ' i
- v.,
A '"<
v. "JV^ ¦•
w hN. v •
-vs.- 'j ^"IXiV / L'a s' {v -' "> -> °\r,i
=d l";f(«wrr:f '• V ¦ c->'
^ / ¦) >i\\ !ll\ 4 '• -.J it:;
A ¦f.\ Ar-i-^JJ ViVl % • J - - | v 7' l' I
tA^~V-_\\ \-?!«s-{r ' l^'j '/
= ?.U>
Figure 2.-1
Proposed Location of Trimble County
2340 MW Generating Station
^'¦Wytfr;. >
0 "n ' L ' 'P/J • . '
III-H-6
-------
TABLE 2.-1
Stack Parameters for Trimble County Generating Station
Unit Stack Exit Temp. Emissions*
Height Velocity °« g/sec
Meters m/sec. SO2 TSP
#1 243.8 30.5 334 520 51.2
§2 243.8 30.5 334 520 51.2
§3 243.8 30.5 334 709 83.4
#4 243.8 30.5 334 709 83.4
* emission rates listed are those which would result from
operation at 100% capacity
HI-H-7
-------
CONTOURS OF SULFUR DIOXIDE. MICROGRAMS PER CUBIC METER
V
CLlHflTCLGGICflL DISPERSION MODEL AT T0C. 1375 WNO ROSE
77QH2S 16110
Figure 2-2
III-H-8
-------
References
1. Turner, D. Bruce, Workbook of Atmospheric Dispersion Estimates,
U.S. Dept. of Health, Education and Welfare, PHS #999-AP-26, 1969.
2. Busse, A.D., and J.R. Zimmerman, User's Guide for the Climato-
logies! Dispersion Model, EPA-R4-73-024, December 1973.
3. Briqqs, G.A., Plume Rise, AEC Critical Reviews Series, TID 25075,
1969.
4. McMullen, R.W., "The Change of Concentration Standard Deviations with
Distance", JAPCA, 25, p. 1057-1058, October 1975.
III-H-9
-------
3. COOLING TOWER PLUME MODELS
3.1 Introduction
A number of plume models are currently available; some are in the
open literature whereas others are proprietary. A current study by Ar-
gonne National Laboratory provides a comparison between different
models. Data is available from a number of field observations from
which one can tune a given model. Current plume models are more de-
tailed than the plume rise formulas described by Briggs.
Plume properties may be expressed in either a Gaussian or top-hat
form. Neither of these distributions is entirely representative of the
true properties which have been observed. Many models are classified loose
by different criteria.and a listing is provided below. Basically, there
are three promising physical approaches:
(1) Cloud Physics Models
Motor-Columbus Incorporated (1 ,2 ) in Switzerland has
developed several models along this line by using the Wein-
stein model as a basis. These models are supplemented by
field data and wind tunnel modeling. They are proprietary.
Hanna ( 3,4 ) and Briggs have developed an approach also based
on the Weinstein model which is supplemented by the analysis
of data by Briggs.
(2) Eularian Set of Conservation Equations
Slawson and Wigley have developed a model based on the
approach of Morton ( 5,6 ) et al., and Csanady ( 7,8 )»
Tennessee Valley Authority.
(3) Puff Models
Wlniarski and Frick ( 9 ) have developed a model along this
line which they describe as essentially Lagrangian.
The above is by no means an exhaustive list of plume models or a
complete list of plume model types.
SHnn has analyzed the Eularian conservation equation approach in
a conventional engineering context and arrives at order of magnitude
effects due to postulated drag and heat transfer coefficients.
3.2 Description of Plume Models
Slawson-Wiqley Plume Model (10,11)
This model consists of a set of Eularian conservation equations
of the integral form (12,13). Top-hat properties are commonly used
but it is possible to use the model with Gaussian properties HI).
Three equations of plume rise, as used here, are the conservation
equations of mass, momentum, and heat energy. They are derived by
integrating general conservation equations across a control volume
III-H-11
-------
perpendicular to the plume and of thickness AS. As shown in Figure 3-1,
the plume entering the control volume is characterized by p,W,V, and R,
the characteristics density, vertical velocity, longitudinal velocity,
and radius of the plume, respectively. After passing through the con-
trol volume and entraining ambient air of density p with an entrap-
ment velocity into the plume of Ve, the parameters p,W,V, and R have
changed incrementally to p + Ap, and so forth. It is assumed that pa
and Ve are constant around the perimeter of the plume.
Conservation of Mass
An equation of conservation of mass is
/pV-dA = - I pdv (3.1)
For steady state conditions, the term on the right equals zero, so
that
/pV-dA = 0. (3.2)
As indicated in Figure 3-1, the control volume over which this equation
1s evaluated has three surfaces. The surface into which the plume
flows, characterized+by p,V^. and R, has mass flux of -irR2pV. Its sign
is negative because V and dA are oppositely directed. The surface out
of whj.ch the^plume flows, characterized by p + Ap, V + AV, R + AR, with
both V and dA directed outward on this surface, has mass flux of n
(R + AR)2(p + Ap)(V +AV). Multiplying this expression, dropping all
terms involving the products of differences (such as ApAV), and adding
the flux term, -R2pirV, we have as the sum of the mass fluxes across
these two surfaces, irpV2RAR +TiR2pAV + nR2VAp, which equals A(irR2pV).
Because ambient air of density pa 1s flowing into the control volume
through its outer surface and because the entrainmen£ velocity of this
air (Ve) is parallel to and oppositely directed to dA, the mass flux
across this surface is -2irRpaVeAS. Thus, the total flux across the
surface enclosing the control volume is
A(ttR2pV) - 2irRpaVeAS = 0 (3.3)
Dividing by AS and in the limit as AS approaches 0, we have
^ (R2pV) = 2RPaVe. (3.4)
It is assumed that Ve is proportional to the vertical velocity of the
plume, so that Ve = ctW. According to the Boussinesq approximation,
variations in density are negligible except where such density varia-
tions involve the operation of gravity. Thus, p may be regarded as
a constant, and because p and pa are approximately equal, they may be
canceled from Equation 3.4. Under these assumptions the equation of
conservation of mass is
III-H-12
-------
(R2V) = 2RaW.
(3.5)
Conservation of Momentum
An equation for conservation of fluid momentum is
Fs + / Bdv = 2- J pVdv + / p(V)V-dA (3.6)
where the surface integral is taken over the surface of an arbitrary
control volume and the volume integrals are taken throughout this con-
trol volume, Fs is the surface force vector (pressure and shear), and
B is the body force per unit volume. For steady state conditions the
first term on the right of Equation 3.6 is zero. Assuming that all
surface forces can be neglected, = 0. The only body forge o| imgor-
tgnce is tha£ of gravity, so that B = g(pa - p)k. Because V = U + W =
Ui + wk and V'dft is a scalar, Equation 3.6 is
/ g(Pa " p)^dv = / pflfl + wtO^-dft. (3.7)
Equation 3.7 is a vector equation and reduces to the following two
scalar equations:
J pltf-dft = 0, and (3.8)
J g(pa " P)dv * W?-dX. (3.9)
Equation 3.8 reduces to an equation similar to Equation 3.2. Integra-
ting Equation 3.9 over the volume shown in Figure 3-1 we have
A(irR2WVp) - g(pa - p)irR2AS, or (3.10)
^ (R2WVp)= R2g(pa - p). (3.11)
Using the 8oussinesq approximation, we can regard p on the left side
of Equation 3.11 as a constant that can be approximated by pa, the
plume density at stack exit. Under these assumptions, Equation 3.11
becomes
^ (R2WV) = R2g (3.12)
III-H-13
-------
Letting b = g P3 " P. ) the bouyant acceleration per unit buoyant fluid,
Equation 3.12 Pa is written as
^
-------
BLir-a? (r2V9) = a? (£riiea-|'JVp) - R2vp3s (£Lira''- <3-19'
Equating Equation 3.18 and 3.19 and combining the first term on the
right side of Equation 3.19 with the left side of Equation 3.18 we
have
3S f R2V(pr - p) - R2V ({£- 1)] = - R2Vp^- (^- 1) (3-20)
This equation simplifies to
ar«,*8,5rfi»-«'»£* It*
If it is assumed that the atmosphere is neutral, so that = 0, then
Equation 3.21 may be written as
fa (R2Vb) = 0 (3.22)
Equation 3.22 integrates to
R2Vb = F, (3.23)
where F is the constant flux of buoyancy.
Winarski-Frick Plume Model
The model uses the average or top-hat properties in solving a set
of algebraic equations. It employs a method of entrainment which re-
quires essentially no adjustment. Plume bending is primarily due to
the momentum of the air mass that passes through the projected area of
the plume. Momentum of the air mass is imparted to the plume in two
ways: (1) the air mass entrained gives its momentum directly to the
plume; and (2) the deflection of some of the air produces a strong
horizontal pressure force near the source. Entrainment is due to the
air mass impinging on the plume and aspiration or shear (plume segment
moving with a velocity relative to the wind). The trajectory of a
plume puff is traced in time with the puff gaining mass by entrainment.
The plume 1s assumed to be essentially a cylindrical segment whose
radius grows due to entrainment.
Model Development
The initial plume mass, M0, is the mass issuing from the tower with
radius b0.
M0 = b0H0 (3.24)
III-H-15
-------
where H0 is the length of the plume mass (comparable to b0) and equals
Ho " V*
(3.25)
An Increment in the plume mass is evaluated from an assumed rate of
entrainment:
DM = Ei (rate of entrainment) iAt
(3.26)
Entrainment occurs due to impingement and shear. To find the rate of
entrainment, one solves the following equation:
Total horizontal momentum flux = Local horizontal pressure
force + Horizontal momentum due to entrainment.
The local entrainment is assumed to equal the horizontal momentum flux
due to entrainment divided by the wind velocity. The local horizontal
pressure force is approximated as the force required to decelerate the
available mass flux t& the velocity of the plume.
Horizontal
Pressure =
Force
Mass flow Aspirated
through + mass
projected flow
area
Wind Horizontal
velocity - plume
velocity
The total horizontal momentum flux is approximated as
Total
Horizontal =
Momentum
Flux
Mass flow
through
projected
area
Aspirated
mass
flow
Wind
velocity
Thus,
[pApW+(aspiration)] W - [pApW+(aspiration)] (W-U) = (3.27)
[pApW+(aspiration)] U = (Entrainment) W
Entrainment = (pApW+aspiration) jj-
(3.28)
This permits a solution for the term DM. The horizontal momentum of
the plume is solved for by the following equation:
(Mt+DM)Ut+At = MV + DM-W + FhpAt
where (M^+DM)U^+A^ = the new horizontal momentum
(3.29)
and
MtUt = the old horizontal momentum
DM-W = the horizontal momentum of the entrained air
F^At = the impulse of the horizontal pressure forces Dn the
plume.
III-H-16
-------
One can readily solve for the horizontal velocity. Assuming the total
pressure force vector acts normal to the plume, it is possible to
approximate the vertical pressure force (provided the plume is not
close to being horizontal) as
F = . Fph (3.30)
pv tan 0
0 = angle between the plume centerline and the (ir/2 to 0)
The vertical pressure force must vanish as the tan Q + 0; thus ^pv is
set to zero when the tan 8 = 0.1. The vertical momentum equation is
(M^+M14^) Vt+At = MtVt+F At. (3.31)
x pv
Since the model solves the case only for a horizontal wind, the en-
trained mass does not carry any vertical momentum with it. The new
plume mass is given by:
Mt+At = ftt+oM (3.32)
The new plume temperature is:
Tt+At = JLI_£9!LIi = (ambient lapse rate) Az (3.33)
l^+DM
The new plume density is:
P
pP = ST (3.34)
The plume vertical acceleration is calculated from a dynamic considera-
tion of the virtual mass concept where the motion in plume is mirrored
in the air.
a » Pa2~pPp g (3.35)
The vertical velocity change is AV = aAt and vt+At = Vt+At+AV (3.36)
The new plume location:
(3.37)
(3.38)
The speed of the puff is:
V = / U^+V2 (3.39)
III-H-17
-------
The average radius of the plume puff:
pirb2H = M b = / (M/pirH)
The plume puff length is given by:
(3.40)
Ht+At = Ht + (V* - Vt"At) HlAt
(3.41)
/ (dX)2 + (dY)2
Where dX
The Winiarski-Frick model has been compared with three different
types of data:
1. Laboratory tests with air jets.
2. Actual field data taken on a large single cell cooling tower.
3. Laboratory tests with plumes in water.
Comparison with these data types was quite good as is noted by Winiar-
ski. The current ANL study (14,15) rates the model superior to all
other models tested to date. Winiarski is continuing research in
this area to add more physical realism to the model.
Cloud Physics Model (16 ,17)
The Cloud Physics Model is adapted from numerical models of cumu-
lus convection. A one-dimensional model developed by Weinstein is
utilized in the plume models of Bogh and Hanna. It accounts for the
processes of horizontal mixing, evaporation, precipitation, generation,
and freezing, as well as the standard thermodynamic and dynamic pro-
cesses in isolated cumuli.
The results which Weinstein stresses are: (1) the liquid water
content and vertical velocity undergo coupled damped oscillations in
time; (2) the rates at which precipitation-sized water drops are
formed and grow are not sensitive parameters in the determination of
rainfall; and (3) the amount of liquid water in the form of cloud
droplets required before a few large drops can be developed is a
crucial parameter in determining rainfall.
Weinstein's model is broken down into four areas: thermodynamics,
dynamics, moisture balance, and overall mass balance.
The thermodynamics of the Weinstein model deals with the heat
balance in a rising parcel of air of mass M which entrains a mass dM
of environmental air. Thus, one utilizes the 1st Law of Thermodynamics,
III-H-18
-------
dh = (CpdT-adp)M. Where
(3.42)
h = heat
Cp = specific heat
T = temperature
p = pressure
a = (ff)T - V
H = enthalpy
V = volume.
The processes considered in the heat balance are condensation, mixing,
evaporation,and freezing.
The dynamics of the Weinstein model is described by the equation
of vertical momentum:
w = vertical velocity
p = density
p = pressure
z = vertical coordinate
g = gravitational constant
t = time.
Drag is due to two mechanisms: (1) the weight of the liquid water;
and (2) the transfer of vertical momentum from the undiluted parcel
to the entrained air.
Moisture 1s balanced by the following approach: one considers
the mass of water substance, Fo, in the original parcel at time t = 0,
the mass of the water substance in the entrained air, Fm, and the mass
of water substance at time t = t+dt, Ft. This yields
Fo = M(q+Q)
Fm = dMqe
Ft = (M+dM)(q+dq+Q+dQ)
Ft = Fm + Fo
q = water vapor content
Q = liquid water content
e - refers to entrained air.
Upon dividing Ft = Fo + Fm by Mdt and dropping second-order
terms, we obtain:
= "9'drag- Where
(3.43)
where
(3.44)
III-H-T9
-------
1 rlM
Where u = ry, the entrainment rate. Entrairiment which is of the
, M dz 0 2
order -5-as suggested by Simpson is introduced as y = . Where R =
the uporaft radius.
The above is the approach used by Weinstein's Cloud Physics
Model. Plume models which have adopted this basic approach have been
modified somewhat. Both the models of Bogh and Hanna rely heavily on
empirical parameters. Hanna employs an empirical entrainment relation-
ship derived from Briggs' analysis of observed field data. Bogh has
been fortunate enough to have field observations and wind tunnel test-
ing to enable him to tune the plume model SAUNA.
Hanna's model has been used to analyze American Electric Power's
John E. Amos Power Plant with promising results; however, a current
study conducted by Argonne National Laboratory under the direction of
Policastro (14,15) rates the Slawson-Wigley and Winiarski-Frick plume
models superior to Hanna's.
Bogh's model has yielded good results as is presented in the pub-
lished literature. Unfortunately, this model has not been employed in
the current ANL survey. It is proprietary.
3.3 Conclusion
A comparison of various plume models is currently in progress at
Argonne National Laboratory (14,15). Quarterly reports to date have
provided analysis of about fifteen models. Of these, the Winiarski-
Frick and Slawson-Wigley models have been shown to be the most repre-
sentative of observed plumes. (See Figures 3-2 and 3-3.) Argonne
National Laboratory has conmented on the strong and weak points of
each model. In addition, data has been subjected to critical analysis
to enable future research to gather data sets which are representative
of plume and atmospheric parameters which are in accord with visible
plume observations.
John Norman of Pennsylvania State University (18) has compiled
data for the Keystone Power Plant with surprising results. Plume ve-
locity measurements have yielded a peak velocity well in excess of the
tower outlet value. This is somewhat disturbing in light of the cur-
rent velocity distributions which are deployed in most plume models.
It is well known that turbulence is incompletely accounted for in most
current models; however, Simpson 091 has given an excellent summary of
cloud physics models and stresses the aspect of the inadequacy of cur-
rent computers to handle this phenomena. In addition, she mentions
the exorbitant costs which one may well run up against. ($10,000 or
more for a single case.)
III-H-20
-------
^V+aV W+*W
\ J?-
P+AP
»—*
I
X
I
ro
Figure 3-1. Plume Control Volume
-------
600
— 400 -
XJ
c
3
O
c_
0
0)
1
g> 200
a>
x
¦— Observed
— Winiarski - Frick
Slawson-Wigley
Cooling
Tower
300 400 500 600
Distance from Tower (meters)
700
800
900 1000
Figure 3-2. Chalk Point - Dec. 19, 15/b (1200)
-------
600
500
Observed
Winiarski-Frick
Slawson-Wigley
300
200
100
Cooling
Tower
200
400 500
Distance from Tower (meters)
700
800
300
600
100
900
1000
Figure 3-3. Chalk Point - Dec. 15, 1975 (0930)
-------
REFERENCES
1. Bogh, P., "Experience with Combined Wind-Tunnel-Plume Model
Analysis of Cooling Tower Environmental Impact," Cool inq Tower
Environment, 1974 CGNF-74032 Proceedings of a Symposium held at
College Park,"W March 4-6, 1974 (1975) Springfield, VA, NTIS
(265-290).
2. Bogh, P.,and Bhargava, "Combined Dry/Wet-Cooling Towers: Their
Environmental Promise and Their Problems," Environmental Effects
of Coollnq Systems at Nuclear Power Plants"! IAEA-SM-187/44,
1975 Vienna (45-62).
3. Hanna, S. R., Rise and Condensation of Large Cooling Tower Plumes,
Journal of Applied Meteorology, Volume 11, 793-799; 1972.
4. Hanna, S. R., "Predicted and Observed Cooling Tower Plume Rise and
Visible Plume Length at the John E. Amos Power Plant." Atmospheric
Environment, Volume 10, 1976; 1043-1052.
5. Morton, B. R., Taylor, G. I., and Turner, 6. S., "Turbulent
Gravitational Convection from Maintained and Instantaneous Sources"
Proceedings of Royal Society of London; 1956; A234 1-23.
6. Morton, B. R., "Buoyant Plumes in a Moist Atmosphere." Journal
of Fluid Mechanics. Volume 2 (March 1957), pp. 127-144.
7. Csanday, G. T. "Bent-Over Vapor Plumes" Journal of Applied Mete-
orology, Volume 10, February 1971 , 36-42.
8. Slawson, P. R.,and Csanady, G. T. "The Effect of Atmospheric
Conditions on Plume Rise," Journal of Fluid Mechanics, Volume 47,
May 1971 (33-50).
9. Wlnlarski, L.,and Frlck, W., Cooling Tower Plume Model, Program
Element 1BA032, R0AP/TASK No. 21AJH 31 EPA, Corvallis, Oregon,
1976.
10. Slawson, P. R., Natural Draft Cooling Tower Plume Behavior at
Paradise Steam Plant (part II) Tennessee Valley Authority 1974.
11. Slawson, P. R., Vapour Plume Theory and Example Computer Programs,
Tennessee Valley Authority. Contract TV 3836A, January 19761
12. Slawson, P. R., "Buoyant Moist Bent-over Plumes." Cooling Towers,
Chemical Engineering Progress Technical Manual. American Insti-
tute of Chemical Engineers 1972 (87-90).
13. Slawson, P. R., Coleman, J. H.»and Frey, J. W., "Some Observations
on Cooling-Tower Plume Behavior at the Paradise Steam Plant-"
Cooling Tower Environment, 1974, C0NF-74032; Proceedings of a Sympo-
sium held at College Park, MD; March 4-5, 1974, 1975, Springfield,
VA, NTIS 147-160.
III-H-24
-------
14. Pollcastro, A. J., Generic Report on Atmospheric Effects of Evapora
tlve Cooling Systems, Progress Report, August 1 - October 31, 1976.
Argonne National Lab, Chicago, Illinois.
15. Pollcastro, A. J., Carhart, R. A., Shobrys, D., DeVantler, B.,
and Ounn, W., Validation of Cooling Tower Plume Rise and Salt
Drift Deposition Models, Progress Report, November 1, 1967 -
January 31, 1977, Argonne National Lab, Chicago, Illinois 77 pp.
16. We1nste1n, Alan I., "A Numerical Model of Cumulus Dynamics and
MlcrophyslcsJournal of the Atmospheric Sciences, Volume 27,
March 1970 246-255:
17. Bogh, et. al., "A New Method of Assessing the Environmental In-
fluence of Cooling Towers as First Applied to the Kalseraugst and
Lelbstadt Nuclear Power Plants," International Nuclear Industries
Fair, Technical Meeting No. 9/25, 16/21, October 1972, Basel,
Switzerland.
18. Thomson, D. W., Norman, J. M.,and Miller, R. L., "Airborne Mea-
surements of Turbulent Temperature and Velocity Fluctuations 1n
Cooling Tower Plumes," Proceedings of the American Meteorological
Society, Symposium on Atmospheric Turbulence, Diffusion, and
A1r Quality,Raleigh, N.C., October 19-22, 1976 576-580.
19. Simpson, Joanne, "Precipitation Augmentation from Cumulus Clouds
and Systems: Scientific and Technological Foundations, 1975,"
Advances 1n Geophysics, Volume 19, 1976, 1-72.
Ill-H—25
-------
4. DISPERSION OF RADIOACTIVITY
4.1 INTRODUCTION
The Intent of this report is to present a detailed methodology for
tracing radioactive effluents from a nuclear facility through the en-
vironment to man. This will subsequently lead to critical organ doses
resulting from radionuclidic uptake and deposition.
The first part of this report, Sections 4.10 through 4.50, will
deal with the methodology and theoretical modeling for tracing radio-
active effluents through the environment to man. The latter part of
this report, Sections 4.60 through 4.70, will deal with the computer
modeling being used for the ORBES Project and present some modifica-
tions and probable adaptations to the computer models.
There are five major pathways to man considered: (a) immersion in
air, (b) immersion in water, (c) contact with contaminated surfaces,
(d) inhalation of contaminated air, and (e) ingestion of contaminated
foodstuffs. See Figure 4-1.
Three computer codes from Oak Ridge National Laboratory (1),
will be used to trace the radionuclides through the environment to man
and estimate dose commitments for each radionuclide considered. These
codes are: INREM, EXREM III, and AIRDOS II (2,3,4).
INREM estimates Dose Conversion Factors (DCF's) from ingestion
and/or inhalation of a unit source of radioactivity (i.e. lyCi). DCF's
for eleven organs are estimated in INREM, i.e., whole body, lungs,
G.I. tract, spleen, kidney, thyroid, bone, testes, ovaries, muscle,and
the liver.
EXREM III estimates DCF's for persons Inmersed in air and water
and in contact with contaminated surfaces.
Figures 4-2,3,4 show how gasses and particulates travel through
the environment. This flow chart is valid for both stable and unstable
nuclides. It is the chemical and physical properties which dictate
the materials transportation through the environment.
AIRDOS II uses the DCF's from INREM and EXREM III to adjust the
total dose commitments that it estimates from site specific data.
AIRDOS II uses Pasquills Gaussian Dispersion Model (5) and Briggs'
Plume Rise formulae (6) to estimate dispersion and deposition of
radionuclides throughout the environment.
Terrestrial transport models and coefficients (7) are used to
follow the radionuclides through the environment to man. Subsequent
doses and DCF's are then estimated in AIRDOS II.
This methodology, INREM, EXREM III, and AIRDOS II, estimates
DCF's to within a factor of + 2 times observed DCF's resulting from
III-H-27
-------
Gaseous
Effluents
Liquid
Efft
|1 Air Submersion
I? 11 Air
yL Inhalation
Shoreline Irradiation ^
(fishing & picnics)
Immersio^^^,
Consumption
(boating & swimming)^ fj
^ T ^^Consumption (milk)
Deposition
Ingestion F°"^mP,ion,
3 — (fish & clams)
1
' n
//Deposition
^—\ Ingestion
Figure 4.1 Radiation Pathways to Man.
III-H-28
-------
Rainout
Snowout
Scavenging
Plume depletion
Particulate
Gravitational
Fallout
Bnggs plume rise equations
Pasquills dispersion formula
Radioactive gases
dissolved in air
Inhalation of
contaminated
air.
100%
Radioactive stack
effluents uniform
emmission rate
Gaussian atmospheric
dispersion model
Radionuclide surface deposition ground and water
Exposure to man
from immersion in
contaminated water.
1%
Exposure to man
from contact with
contaminated surfaces.
100%
External
exposure
to man due
to immersion in
contaminated air.
100%
to TERMOD
Figure 4.2 Atmospheric Dispersion Flow Diagram.
111-H-29
-------
e.m
g.c
p.nri
i.m
c.rn
Milk
Beef
Soil sink
Pasture soil
Water table
Above surface
food crop
Absorption into
digestive tract
Absorption into
blood stream
Soil surface
below food crop
Subsurface soil
pool for food crop
Pasture grass,
milk and beef
production
Terrestrial transport model TERMOD
Radionuclide uptake by man via ingestion
Subsequent deposition
of radionuclides in
critical organs
Passed out of body at
rate proportional to
biological half-life
Collection in stomach, subsequent
dispersion throughout body
Figure 4.3 Terrestrial Transport Flow Diagram.
III-H-30
-------
25% exhaled
25% exhaled
Soluble isotopes
50% swallowed
50% swallowed
Insoluble isotopes
Absorption into
blood stream
Absorption into
digestive tract
Slowly absorbed
by body fluids
12.5% stay in
lungs for 120
days
25% in lower
respitory
passage
12.5% stay 2U
hours then
swallowed
Inhalation of
contaminated air
100%
25% in lower
respiratory
absorbed into blood
Subsequent deposition of
nuclides in critical organs
Insoluble nuclides
pass through body
after 36 hours
Passed out of body at rate
proportional to biological half-life
Collection m stomach, subsequent
dispersion throughout body
Figure 4.4 Internal Distribution Flow Di
-------
field tests.
The AIRDOS II methodology 1s being investigated to see how it can
be adapted to estimate uptakes of stable nuclides, such as heavy metals.
4.2 ATMOSPHERIC DISPERSION OF RADIONUCLIDES
The basic equation used to estimate atmospheric dispersion in
AIROOS II is Pasquills Equation (5) as modified by Gifford (8), i.e.,
*¦" "P l(cy) exp ~l(if) * exp (4-1)
where:
X = concentration in air at the centerline of a plume x meters down-
wind from the point of discharge (curies/m3).
Q = uniform emission rate from the stack (Ci/sec).
U = mean wind speed (m/sec).
ay = horizontal dispersion coefficient (m).
°z = vertical dispersion coefficient (m).
H = effective stack height (physical stack height h plus the plume
rise Ah) (m).
Y = crosswind distance (m).
Z = vertical distance (m).
Downwind distance, x> comes into Eq. (4-1) through oy and az, which
are functions of x 35 we^ as the atmospheric stability category appli-
cable during emission from the stack. Pasquill (4) described six
atmospheric stability categories ranging from A (very unstable) to
F (very stable). A seventh category, G (extremely stable) has been
added to AIRDOS II. Values for cry and az as functions of x for each
of the six original Pasquill categories are the most recent values
reconniended by the AIR Resources Atmospheric Turbulence and Diffusion
Laboratory (9). The values for category G were extrapolated from
values 1n categories in A through F. Figure (4-2) shows the logical
flow of radionuclide dispersion and deposition throughout the
environment.
4.2.1 PLUME RISE ESTIMATIONS
Three options are available in AIRDOS II to estimate plune rise,
Ah, for Eq. (4-1). Only one option will be considered in this report.
The equation of RUPP e_t aj_ (10) will be used to estimate plune rise.
111—H- 32
-------
The plume rise is caused mainly by the momentum of emitted stack gases.
This seems ideal for nuclear facilities since the gaseous, effluents are
near ambient temperatures. The equation is
Ah = 1.5 Vd/y (4-2)
where:
Ah = Plume Rise (m),
V = effluent gas velocity (m/sec.),
d = stack diameter (m), and
u = wind velocity (m/sec.).
For part of each year a stable layer will exist above a less stable
layer. This effect, referred to as a lid, restricts vertical disper-
sion of the plume and results in a higher air concentration at ground
level than would exist in the absence of a lid.
4.2.2 PLUME DEPLETION MODES
There are three principal modes of plume depletion accounted for
in the AIRDOS II methodology: (A) Gravitational Fallout, (B) Gases
Dissolved in Air, (C) Washout Modes.
(A) Gravitational Fallout.
Radionuclides released as particulates may be substantially
affected by gravity during plume travel. A value for the gravi-
tational, fall velocity, Vg, for each radionuclide is necessary
in AIRDOS II. This value is zero for gases and usually zero for
most particulates.
Particulates will deposit on ground or water surfaces at a
rate that is the product of their concentration in air at ground
level and the deposition velocity (m/sec.).
(B) Gases Dissolved in Air.
Radionuclides released as gases will largely stay within the
plume and disperse throughout the atmosphere at a rate propor-
tional to the plume dispersion. Some gases will intimately mix
with air as a result of turbulence at the interface of the plume
with the atmosphere.
(C) Scavenging Modes.
Scavenging is a process by which rain or snow washes out
particulates or dissolves gases and deposits them on grcund or
water surfaces. The fraction of particulates or soluble gases
removed by scavenging from a vertical column of air per unit
time during rain or snow is with the units of sec-1, the rate of deposition on a
III-H-33
-------
surface is;
R = (fiCH pci/cm2-sec (4-3)
where:
C = average concentration of radionuclide in a vertical column
(pci/cm3),
H = height of the vertical column (cm)
For each radionuclide considered, AIRDOS II uses a scavenging coeffi-
cient which is the sum of the rainout and snowout coefficients. The
average concentration in the vertical column used in Eq. (4-3) is
computed through the use of Eq. (4-1).
4.3 TERRESTRIAL TRANSPORT OF RADIONUCLIDES
The Terrestrial Transport model, TERMOD (7), is a generalized
systems analysis coupled-compartment code intended for predictions of
radionuclidle intakes by man. The intakes considered are, consumption
of milk, beef, and plant parts contaminated directly by deposited
radionuclides as well as by uptake from the soil. See Figure (4-3) for
the logical flow of radionuclides through the environment to man.
Radioactivity is transferred to man via three main food sources: food
crops, beef, and milk. These foods could be produced at different
locations. The fractions of foodstuffs produced in the area of con-
sideration which are consumed by man are necessary input values for
AIRDOS II.
A fraction, Si, of the initial radionuclide deposition affects
the above surface food crop. A fraction S? is deposited on the soil
surface not protected by this food crop. A fraction S3 is deposited
on the pasture grass. Finally, a fraction S4 deposits itself in open
water areas. This aspect is not treated in the TERMOD methodology
because of such small dose contribution to man. It will be treated
later.however, 1n considering the external dose to man.
Each "d" factor, di, d2, d3, and d
-------
derived from contaminated land, and (4) the only loss mechanisms
included were radioactive decay, movement to the soil sink, and metabolic
turnover in the cattle.
4.3.1 MODEL EQUATIONS
Three transfer coefficients, xg^, tg c? and rp m (Figure 4-3),
were considered a function of the element Seing modeled (i.e., the numer-
ical value assigned each of these coefficients is different for each
element). Conversely, the transfer coefficient re m was not considered
a function of each element, but is held a constant! A single value,
or a set of values,was selected for the rest of the transfer coeffi-
cients (i.e.,they were not considered to be a function of the radio-
nuclide). These were selected with the objective that (1) the model
be realistically conservative, (2) the dependent variables simulate
a realistic dynamic behavior, and (3) the biomass and radioactivity be
conserved. The fact that all of the transfer coefficients were not
different for each element or radionuclide reflects a lack of quantita-
tive information about the movement of all radionuclides to man by the
pathways considered.
Radionuclides can be transferred from pasture grass to man through
his intake of beef and milk. In Figure 4-3 a feedback loop to the grass
through the soil compartment is included to account for the movement
of radioactivity into the soil and its subsequent uptake by grass roots.
The subsurface soil pool and above surface food compartments simu-
late two pathways (possibly by the same crop) by which man can ingest
radioactivity through consumption of foods other than milk and beef.
The subsurface soil pool compartment simulates uptake of radioactivity
via the soil-root pathway, with equilibrium transfer to the edible
food-crop, and final consumption by man. The edible parts of the
crop are implied in the transfer coefficient connecting the pool to
man. Man's intake rate of radioactivity via this pathway is dominated
by the dynamics of the subsurface soil pool compartment.
The above-surface food compartment simulates truck crops, fresh
fruits, and other perishable foodstuffs growing at the time of
initial deposition. Radioactivity enters the above-surface food crop
by deposition and is retained either by external adhesion to, or foliar
absorption into.the crop's edible parts. This crop is assumed to be
continually harvested during and after radionuclide deposition, with
biomass removal balanced by crop growth.
The milk compartment variable C(t) (yCi/litre) is the concentra-
tion of radioactivity in milk produced by a cow grazing on the con-
taminated pasture. This compartment simulates the udder of a cow. In
this model the radioactivity is transferred from the grass directly
to the milk. This simulation is used because the milk of grazing cows
can be in transient equilibrium with the forage radioactivity after
two days (11).
The beef compartment variable B(t) (uCi/kg) is the radioactivity
in the meat or muscle of cattle grazing on the contaminated pasture.
III-H-35
-------
The transfer of radioactivity is taken to be from the grass directly
to the beef. Radioactivity leaves the beef compartment because of
radioactivity decay, biological elimination, and slaughter of the
cattle. The mass of the beef compartment is assumed to be constant,
because uncontaminated cattle are added to the herd, thereby balancing
the mass of the slaughtered cattle.
4.3.2 SYSTEM EQUATIONS
The equations (definitions of the terms, their units, and sample
numerical values) if constant are given in Table (4-1) describing the
transfer of deposited radionuclides to man as depicted in Figure 4-3
are:
dE Te,m
dT = (St) F(T) - (XR + A + xe>s) E, (4-4)
^ = Te,s E'(XR + Tsp)S + (S2} F(T)' (4"5)
dP
3T= (S)(A) tS)P - (AR + tp>m + Tpjd)P, (4-6)
dG Vc
3T = (S3) F(T) - G(Ar + xg>r + A^g) + R trjg> (4-7)
dR
3T" G Tg,p " *(XR + Tr,g + Tr,d'-
(4-8)
dD Vd
<3T = R t d + A P - ArD, (4-9)
dC
dT.GT9ic-C(XR + Tram), (4-10)
dB
"t ¦s Vb -Btx*+ w- (4-11)
1 * "• * Pd2 + Bd3 TM * Cd4 rc,m <4",2>
dm
dT = I - M(Xr - \e), (4-13)
F(T) = Fq 6(T). (4-14)
These equations express the change with time of the radioactivity in
each compartment as the difference between inputs and losses from the
III-H-36
-------
compartment. Each gain or loss term in the equation for a given compart-
ment is represented in Figure (4-3) as a connecting arrow between that
compartment and another compartment with its corresponding transfer
coefficient.
Many of the coefficients in this model were assigned numerical
values that were the same for all elements and radionuclides because
(1) data are not available for each element from which
unique transfer coefficients can be calculated; and (2) the transfer
for which the coefficient applied is a biomass transfer that is indepen-
dent of the element or radionuclide under study. The main objective in
the selection of transfer coefficients not treated as a function of the
element was that they be reasonable and conservative.
A 14-day environmental half-time is assumed for the transfer from
the grass to the soil and from the food crop to the soil surface (12).
This assumption enabled the determination of t_ „ and t_ .. Feedback
y > r c j s
of the radioactivity from the soil compartment was assumed to be 1% a
year (13), which determined t„ _. A loss rate of 4X/year from the soil
" >9
compartments to the soil below the roots was included in the model as
transfer to the soil sink (13), which enabled determination of x d and
Tr,d*
The parameter (2.0/day) signifies that the udder is emptied
of milk twice each day. Tfjeef signifies that a fraction of the beef
herd is continuously being slaughtered. The total mass in the beef
compartment, as stated earlier, is assumed to be constant. The transfer
coefficients t- _ and xh _ apply to a standard diet consisting of a
C jlll U jlTI
daily intake of 1 litre (approx. 1 quart) of milk and 0.3kg
(approx. 10 oz.) of beef. The numerical values for the fixed coeffi-
cients and their units are included in Table 4-1.
The environmental transport and dosimetry of Carbon-14 and
Tritium (H-3) require special treatments which are discussed in detail
in Ref. (1). Special note is made here to make the reader aware that
there are special problems in working with their transport and dosi-
metry calculations without detracting from the purpose of this report.
4.4 ESTIMATION OF EXTERNAL RADIATION OOSE TO MAN
External dose to man is considered to result from three basic
exposure models: (1) immersion of an individual in contaminated water;
(2) immersion in contaminated air; and (3) irradiation from contact
with contaminated surfaces.
The EXREM III computer code (3) will be used to estimate dose rate
and the total dose to man from beta, positron, electron, x-ray, and
gamma radiation through the three exposure modes listed above. The
dosimetric information was computed in EXREM III with the following
assumptions: (1) Dose-reducing factors such as shielding pro\ided by
dwellings are not considered, (2) Exposure via mode(l) occurs one
percent of the year, (3) Exposure from mode(2) occurs 100 percent of
111-H-37
-------
TABLE 4.-1 LISTING OF SYMBOLS FOR TERMOD METHODOLOGY
eq
B
S
C_
soil surface area requ1red3tOjfurn1sh F
food crops for one man (10 m )
% 2
pasture area per cow (10 m )
concentration of radioactivity 1n F
the beef (yC1/kg) 0
equilibrium concentration of a fu
stable Isotope of the nuclide
1n wet beef (g/kg)
concentration of radioactivity In
the milk (uC1/liter) f
concentration of an element in man
(ppm)
concentration of an element in
meat (ppm)
concentration of an element in
forage (ppm)
concentration of an element in
soil (ppm)
radioactivity present 1n th^ soil
below the root depth (yCi/m \
dry-weight areal density of pn's
above-surface food (0.1 kg/m )
dry-weight areal grass density
(0.15 kg/m )
depth of the plow layer (20 cm)
di,d2,dj,di,dietary factors that correct the
transfer coefficients to nan from
those for reference man to those
for the population under study
(dimensionless) (di and dz = 0.25;
d) and d* 3 1.0)
Kronecker delta function
radioactivity present on above-
surface food per m of surface
on which the food crop 1s grown
(yCi/m )
IT,
eq
eq
aR
deposition source, which for
this calculation was assumed
to be present only at time
t = 0 (uC1/m2-day)
ground deposition source
(uCi/m )
fraction of a steer's
elemental uptake in foraqe
that is converted into meat
(dimensionless)
fraction of an isotope that is
ingested daily by a cow and
secreted per liter of milk
(days/liter)
fraction of the total uptake
of a nuclide by a cow which
is transferred to the milk
(dimensionless)
radioactivity present in the
grass compartment (yCi/m2)
equilibrium concentration of
a stable isotope of the nuclide
in wet grass (g/m )
time between harvest and con-
sumption of food associated
with the kth pathway (days)
equilibrium stable elemental
uptake by standard man
input source to man (yCf/day)
turnover rate of the stable
isotop^ of the nuclide in man
(days* ) except for Sr where
was isotope dependent
radioactive decay rate of the
nuclide under study (day- )
rate of increase in the wet
weight muscle mass of a steer
with respect to time (0.4
kg/day)
III-H-38
-------
TABLE 4.-1 (continued)
radioactivity present 1n man
(vCi)
mass of muscle on a steer at the
time of slaughter (200 kg/steer)
equilibrium mass of a stable
element in the whole body of
standard man (g)
radioactivity present in the
subsurface pool associated with
one man's food supply (yd)
equilibrium mass of a stable
element in soil from the
ground surface to the depth
of the plow layer (g)
3
density of the soil (1.4 g/cm )
r.g
s,P
radioactivity present in the soil
from the ground surface to the2
root depth of the grass (yCi/m ) U
radioactivity present at the soil
surface (yCi/m ) V
correction factor to account
for different depositions to the
above-surface food, Si (0.1), the
soil below this food, S2 (0.9), Vc
and the pasture, Ss (0.25)
(factors are dimensionless).
fraction of the beef herd slaughtered V
per day (0.00381 day" )
excretion rate of a stable isotope
of the nuclide fr^m the muscle V
of a steer (days* )
transfer rate of milk from the
udder (2.0 days-1)
transfer coefficient from com-
partment i to compartment j
amount of meat eaten by a man
each day (0.3 kg/day)
amount of milk consumed by a
man each day (1.0 liter/day)
2
m /day
0.0495 day"
2
m /kg day
2
m /liter-day
0.0495 day"
1.096 x 10" day"
.:
day
1.096 x 10" day"
2.74 x 10"S day"'
0.0693, 0.00693, and
0.000693 day-1;
usually each problem
was run three times,
once for each value of
t „ listed
s.p
milk capacity of the udder
(5.5 liters)
dry-weight amount of
above-surface food con-
sumed by a man each day
(0.25 kg/day)
dry weight grass con-
sumption per day by a
cow (10 kg/day)
amount of milk produced
by a cow per day (11
liters/day)
amount of soil available
per man for food production
(4 x 108 g)
III-H-39
-------
the year, (4) Exposure from mode(3) occurs 100 percent of the year, with
the receptor being a point 1 metre above the surface.
In general, the external dose to man associated with a radionuclide
depends upon the distribution of the radionuclide in the environmental
medium to which the individual is exposed, upon its decay and build-up
of radioactive daughter products, and upon the dynamics of environmental
removal and replacement of the nuclide and each daughter product. The
dosimetric quantities computed assume a homogeneous distribution of
radioactivity throughout a large portion of the environment . The cal-
culation of the external dose to an individual requires that the con-
centration of each nuclide in the medium be determined as a function of
time. When radioactive daughter products occur, the concentrations of
the various species in the chain are interrelated through the usual
radioactive decay chains and their associated differential equations.
4.4-1 MODEL EQUATIONS
The dose equivalent rate at time t, including the contribution from
radionuclide decay chains, is given as:
i 3 Radionuclide index,
q = Radiation index (0 for beta radiation, p for positrons, e for
electrons, y for gamma and X for radiation),
p 3 Exposure index (w for water, a for air, and s for surface),
L = Location index,
Cj_i (t) = Concentration (yCi/cm3 for p=w and p=a, yCi/cm2 for p=s) of
p thei'th radionuclide for the p'th mode of exposure at the
L'th location at time t,
D. = Dose-rate factor [(mrem*cm3/uCi•hr)] for p=w and p=a;
qp [(mrem-cm2/uCi-hr)] for p=s, and
t - Time (hr).
The total dose equivalent from t^ to tg is simply:
where tj and t2 are given in hours.
These two equations, Eq.(4-15) and (4-16), form the basis of esti-
mating external dose to man via several separate or combined exposure
(4-15)
where
¦ °iqp /'2 CipL(t) "
1
(4-16)
III-H-40
-------
modes and decay radiations. For a detailed treatment of the model equations
and their respective derivations see (1).
4.5 ESTIMATION OF INTERNAL RADIATION DOSE TO MAN
Intake of a radionuclide into the body results in internal exposure
of man that may be quantified in terms of dose coimitment. The duration
of intake is assumed to be continuous in this study. The dose received by
a specified organ is proportional to the amount of radioactivity in that
organ at any time. The concentration of a given radionuclide in that
specified organ is a function of the uptake rate, the elimination rate, and
the half-Hfe of that radionuclide.
The term "dose coiranitment," as it is used in this report, is associated
with a specified intake of a radionuclide and is defined to mean the total
dose to a reference organ. The total dose is that dose which will accrue
during the remaining lifetime of the individual from that intake. This
definition is intended to include the contribution of any radioactive
daughters in the body as the parent decays. The exposed individual is
assumed to be an adult (20 years of age) at the time of intake who will
live to an age of 70 years. Thus "dose commitment" is to be interpreted
as "50 year dose commitment."
The total dose commitment to an organ resulting from the intake of a
radionuclide is resolved into two modes, inhalation and ingestion. INREM
computes DCF's for 259 radionuclides and eleven reference organs for each
intake mode. In this report the term "DCF" means the dose commitment to
an organ that will result from a single intake of 1 uCi of a radionuclide
via a specified exposure mode.
Retention of the radionuclides in the reference tissue is assumed to
be described by a single exponential function. Uniform distribution of
the radionuclides throughout the tissue is assumed for all reference organs
except the gastrointestinal tract. In the case of the G.I. tract, uniform
distribution in the contents of the tract segment is assumed. It is
further assumed that the dose to the segment wall is, on the average, half
that dose to the contents.
4.5.1 MODEL EQUATIONS
If q(t) denotes the activity burden (yCi) of a radionuclide in an organ
at time t (days), the resulting absorbed dose rate to the organ is:
P(t) = (Rem/day) (4-17)
where
51.2 = a conversion factor relating the energy imparted to tissue
from radiation to the activity of the radionuclide emitting
the radiation;
e(t) = the effective absorbed energy (Mev/dis) at time t;
m(t) = mass (g) of the organ at time t;
t = time (days).
The effective absorbed energy for a single radionuclide vi defined
as the sum:
111-H-41
-------
e =1 Ej(QF); Nj,
J
(4-18)
where j indexes the various decay processes (B,P,e, and r), and where,
Ej = Total Energy (MEV) from thej'th radiation type absorbed by
J the organ per disintegration,
QFj = Quality factor for the jth radiation type,
No 3 Relative damage factor for nuclides deposited in the bone;
J its values are treated in detail in Ref. 15, sec. IV.2,
basic rule (b). The factor Nj may be taken as unity for organs
other than bone.
All internal dose estimates in this study are based on the following
formulation.
The total dose commitment to an organ may be obtained by integrating
Eq. 4-17, yielding;
D(T) = SLLi / q(t) dt (Rem) (4-19)
m q
where T = 18250 days (50 years).
The mass m of the organ is assumed constant in view of the earlier
assumption that the reference individual is an adult during the dose
integration period.
Thus carrying Eq. (4-19) one step further, the total absorbed dose
to the organ from any number of radionuclides and their daughters is
given by:
D(T) = t D.(T). (4-20)
i=l 1
where, 1 is the radionuclide index and n is the n'th radionuclide under
consideration.
4.5.2 ORGAN DOSE MODELS
The models used to calculate the accumulated radiation dose to the
body organs were adapted from (14). There are three basic models to
consider; one for the G.I. tract, one for the lungs, and one which is
applied to all other organs. These models are as follows in the next
three sections.
4.5.3 ALL ORGANS EXCEPT THE 6.1. TRACT AND LUNGS
The rate of change of the burden q of a radionuclide in an organ is
satisfied by;
^1= If - xq (pCi/day), (4-21)
where
III-H-42
-------
I = daily intake rate (pCi/day),
f = fraction of I deposited in organ,
x = effective elimination constant (day" ).
The effective elimination constant is given as the sum of radioactive
half-life of the nuclide and the biological half-life of the organ tissue.
The integration of this differential equation is the basis of estimating
the cumulative dose to all organs except the G.I, tract and lungs.
4.5.4 G.I. TRACT
For dosimetry purposes, the G.I. tract is divided into four connected
segments (stomach, small intestine, upper large intestine, and lower large
intestine); this model (14) requires that a different equation be written
to estimate the dose to each segment as a result of the passage of
radioactive material. The equations must consider the time required for
the segment to empty, the mass of the segment plus its contents, and the
fraction of radioactivity absorbed into the blood. See (14) for the
derivations and treatments of the equations used in the G.I. tract model.
4.5.5 THE LUNGS
INREM implements the model from Table 10 of (14) for particulate
matter 1n the respiratory tract. Inhaled particulates are classified as
either soluble or insoluble (.see Fig. 4-4). In either case 25% of inhaled
matter is immediately exhaled, 50% is deposited in the upper respiratory
passages and subsequently swallowed. The remaining 25% is initially
deposited in the lower respiratory passages. If the material is soluble,
the latter fraction is assumed to be absorbed into the blood at a rate
derived from the biological half-life of the lung. For insoluble matter
the residual 25% is equally divided into two compartments: the material in
one of these compartments has a biological stay time of 24 hours after which
the contents are swallowed; the remaining 12.5% stays in the lungs with a
biological half-life of 120 days. Exceptions to this are isotopes of
Plutonium and Thorium, in which the half-lives are one and four years, res-
pectively. Once the activity q(t) 1n the lungs is predicted in this fashion,
the dose commitment is estimated from integration of equation (4-17). The
next section, section 4.60, will briefly overview how the calculations intro-
duced in the preceding sections will be implemented in the computer analysis
for the ORBES project.
The modeling of the computer codes follows directly with the methodology
for the dispersion and deposition of radioactivity in man.
4.5 COMPUTER MODELING AND SAMPLE PROBLEM
4.5.1 THE AIRDOS II COMPUTER CODE
The AIRDOS II computer code estimates annual population doses (Man-Rems)
and the maximum annual individual doses in an area resulting from a continu-
ous release of radionuclides from a nuclear facility.
The AIRDOS II Code builds a 20 x 20 square grid with the ruclear
facility or source at the center. The input data for the gric include the
length of a grid on a side. All of the 400 grids are the same size. For
III-H-43
-------
each one of the 400 grids, the human, beef cattle, and dairy cattle popula-
tions are necessary input data. The type of terrain contained in each grid
is necessary input, i.e., water area, farmland, residential areas, and so
forth.
Meteorological data averaged over a typical year in the area of the
nuclear facility as well as physical and chemical properties of the
radionuclides being released are needed as input data. These are used
1n the atmospheric dispersion model and the terrestrial transport model to
compute the distribution of the radioactive effluents throughout the grid
area and their uptake into the environment and man's food chain.
The code calculates the radiation exposure to man using the DCF's cal-
culated by INREM and EXREM III. Exposure to man directly from the environ-
ment and through the food chain are considered. Ingestion of fish from
contaminated water areas and gaimia radiation from overhead plumes are not
considered in AIRDOS II. These sources of exposure can generally be neg-
lected in the exposure estimations due to their realistic insignificance
with respect to the other exposure modes considered.
The output summarizes population doses in 3 ways: (1) by nuclides,
(2) by pathways to man, and (3) by each of the 11 organs treated in
AIRDOS II. The highest individual doses in the area for each organ are
tabulated for each radionuclide, and the highest organ doses from all
radionuclides 1n the source term are listed. The location of the highest
dose to an individual is specified in the output.
4.6.2 THE INREM COMPUTER CODE
INREM estimates DCF's to body organs following a continuous intake,
through inhalation and/or ingestion, of one or more radionuclides.
The DCF's are calculated for different age groups. DCF's for each
age group are estimated for the 11 reference organs and 259 possible
radionuclides. The desired age groups, organs, and radionuclides are
necessary input data. The code is programmed to assume a constant intake
of lyC1/day.. The many various values pertaining to specific age groups,
organs, and radionuclides, are contained within the program's data base.
INREM measures time relative to a specified reference event. Time
data include the age of the individual, the period over which intake
occurs, and the time of dose evaluation.
The output table consist of dose commitments from each radionuclide
for each organ, and the cumulative dose from all of the radionuclides
under consideration.
4.6.3 THE EXREM III COMPUTER CODE
The EXREM III computer code estimates the DCF's from beta, positron,
electron, gamma,and X radiation. The code calculates the DCF's for a unit
source (lyCi) of radioactivity using a data base similar to tha~ in INREM.
The codes output tables tabulate the dose rates, and total doses for each
type of radiation, for each radionuclide, and the totals of all of the
radionuclides collectively.
Ill-H- 44
-------
This preceding discussion will now be illustrated through the use of
a sample problem.
III-H-45
-------
4.6.4 SAMPLE PROBLEM
PROBLEM
Liquid Mo-99 released into a creek which flows into a lake. Mo-99
and daughter product Tc-99m reach the lake at 1 mCi/day for each day. The
lake has a volume of 2Q million liters, An individual swims in the lake
15S of the time for 30 years. What is the dose received from electron
radiation?
SOLUTION
Tables 4-2 and 4-3 are excerpts of the output of the three programs.
From Table 4-2, the half-lives of the three nuclides can be calculated
using a, the decay constant.
Mo-99 - 66.2 Hours
Tc-99m - 6.02 Hours
Tc-99 - 1.26 x 109 Hours
Note that transient equilibrium is established quickly, and that
Tc-99m will make little contribution to the total dose. This is con-
firmed by the rates given in Table 4-3.
The total rate of radiation increase per unit volume is
2.08 x 1G"9 uCi/ml-hr, assuming no losses from the lake.
Using Table 4-3 and the inflow rate the surface dose estimates can be
calculated for the 30 year time span. For water immersion, the EXREM III
code assumes inmersion 1% of the time, so this factor is already considered.
Mo-99 .241 mrem
Tc-99m .00087 mrem
Total 3
-------
TABLE 4-2
Half-Lives of Three Nuclides
INDEX ATOMIC IWE DECAY CONSTANT NO. BETA- POSITRON NWER GONVERSION/AUGER
NUMBER O/HRS) PARTICLES PARTICLES PHOTONS ELECTRONS
V 42 rt)9tf 1.0470 E-02 6 0 10 13
»78 43 TC99H 1.1514 E-01 0 0 6 12
&9 43 TC99 3.7300 E-10 10 0 0
-------
TABLE 4-3
Dose rate (m?em/yr) from submersion in contaminated water, Continuous inflow of 1 microcurie/ml/hr
Radiation type: Electron
NUCLIDE T = 0 In?. 1 day 1 week 1 month Iyr 5yr 10 yr 20 yr 30 yr 50 yr
IU-99 0.U0E+00 3.89E-K34 3.48E-K5 3.19E+06 3.85E06 3.36E+06 3.36E-HD6 3.86E+06 3.86E-HB 3.86E+06 3.86E-KE
TC-99M 0.00E-+00 1.51E-+03 1.30E-KW 1.39E-KX* 1.39E+04 1.39E+04 1.39EKW 1.39E-K34 1.39E-K34 1.39E-KW 1.39E+M
TC-99 O.OOE+OO 3.97E-HJ3 2.15E-HE 1.51E+06 6.46E-KJ6 7.86E+07 3.93E+03 7.86E-KB 1.57E+09 2.36E+09 3.93E-KD9
-------
4.7 DISCUSSION OF MODIFICATIONS
4.7.1 MODIFIED ATMOSPHERIC DISPERSION PARAMETERS
Recently in 1976 an updated set of values of horizontal and vertical
dispersion coefficients, ay and <*z« respectively, were made available. Two
reportr (15, 16) recapitulate the Gaussian Plume Model and PasQuills (5)
technique of assessing the model sensitivity. The result of this effort
has been to determine more accurate values for
-------
REFERENCES
1. G. G. Killough et al. "A Methodology for Calculating Radiation Doses
from Radioactivity Released to the Environment." ORNL-4992,
Environmental Sciences Division.
2. G. G. iCHlough et al. "INREM-AFOXTRAN Code Which Implements ICRP2
Models of Internal Radiation Dose to Man." ORNL 5Q03 [Feb. 1975).
3. Trubey et al, "The EXREM III Computer Code for Estimating External
Radiation Doses to Populations from Environmental Releases," ORNL-
ORNL-TM-4322 (Dec. 73).
4. Documentation not yet available.
5. F. Pasquill, Meteorol. Mag. 90 1063 (1961).
6. G. A. Brlgqs, Plume Rise, AEC Critical Review Series. TID-25075
(Nov. 1968).
7. R. S. Booth et al. "A Systems Analysis Methodology for Predicting
Dose to Man from a Radioactively Contaminated Terrestrial Environment,"
Proc. Third Symposium on Radioecology.
8. F. A. Gifford J., "Use of Routing Meteorological Observations for
Estimating Atmospheric Dispersion," Nucl. Safety 2 (4), 47 (1961).
9. AIR Resources Atmospheric Turbulence and Diffusion Laboratory,
National Oceanic and Atmospheric Administration, Oak Ridge, Tenn.
10. A. F. Rupp et al. "Dilution of Stack Gasses 1n Crosswinds," U.S.A.E.C.
Report AECD-1811 (CE-1620), Clinton Laboratories (1948).
11. R. S. Booth et al. "Dynamics of the Forage-Cow-Milh Pathway for
Radioactive Iodine, Strontium, and Cesium to Man," Proc. of A.N.S.
Special Symposium, U. of Missouri, Columbia, (Aug. 22-25, 1971)
pp. 127-143.
12. J. D. Z1mbr1ck, 1967 CERT Progress Report, Controlled Radioiodine
Tests at the National Reactor Testing Station, Proqress Report No. 4,
ID0-12065 (1568).
13. R. G. Menzel, "Factors Influencing the Availability of Radionuclides
for Plants," Fed. Am. Soc. Exptl. Biol. Proc. 22, 1398-1401 (1963).
14. International Commission on Radiological Protection, Recommendations
of the International Cotnnission on Radiological Protection, ICRP#2,
Pergamon Press, London, 1959.
15. A. H. Weber, "Atmospheric Dispersion Parameters in Gaussian Plume
Modeling," Part I, U.S.E.P.A.-600/4-76-030A (July 76).
16. A. H. Weber, "Atmospheric Dispersion Parameters in Gaussian Plume
Modeling," Part II, U.S.E.P.E.-600/4-76-030B (June 76).
III-H-50
-------
5.0 REFINING AND CONVERSION
5.1 INTRODUCTION - A TYPICAL PROCESS
The purpose of this section is to describe coal gasification
processes and their environmental impacts. Emphasis is placed on the
high - Btu Lurgi process which is most likely to be widely employed in
the conversion of coal to clean-burning synthetic natural gas (SNG)
for residential and industrial heating. This gas will be distributed in
existing pipelines. Conversion to low-Btu gas is also briefly
discussed. It may be less widely used because its product is most suit-
able for fueling electric power plants. Since no commercial gasification
plant is complete at present, the data in this section are all based on
published estimates.
Coal gasification is expected to have considerably less impact
on air quality than a comparable coal-fired power plant. The nominal unit-
size coal gasification plant being planned commercially is 250 million
standard cubic feet per day (250 mm SCFD) of high-Btu synthetic natural
gas, using proven Lurgi technology. This amounts to 310 x 10^ Btu/day
energy output resulting from about 25,600 tons/day coal input. The
overall thermal efficiency is about 70%. Approximate emissions would be:
tons/day
AIR: S02* 14
COS 0.5
Particulates 2.4
N0X 17.4
WATER: Zero
SOLID WASTE: Ash 5,900
Total water and oxygen consumption would be about 30,500 t/d (5100 gpm)
and 5,700 t/d, respectively. (Emissions estimates from other
references are discussed in detail below.)
Figure 5-1 is a schematic flow diagram of a coal gasification plant
utilizing Lurgi technology to produce high-Btu SNG from strip-mined coal
in New Mexico. It and the data which follow are illustrative only, since
designs and numbers are site-and coal-specific. The following process
description is excerpted from reference 5-1.
*Based on 0.91 wt.% sulfur in coal, the sulfur input is 200 t/i and
this output amounts to 0.7% of the input.
III-H-51
-------
Tail Gas
to treater
Byproduct
Sulfur
Steam
•- C0„ Vent
Coal
SNG
to
compression
-p»Tars & oils
Phenolic Waters
¦>- Oils
Ash
Nitrogen
i Vent
Naphtha
Ammonia
S. Water
Byproduct
Phenols
Air
Reclaimed Water
(to cooling towers)
Byproduct Water
(to boilers)
[~ AUXILIARY "J
SERVICES 1
Steam
Coal
Figure 5-1
Process Flow Diagram
Cooling Water
. System
Reclaimed
Water
SNG FROM COAL
COOLING
TOWER
BOILER
PLANT
GAS
COOLING
SCRUBBING
CLAUS
SULFUR
RECOVERY
OXYGEN
PLANT
SHIFT
CONVERSION
PHENOSOLVAN
EXTRACTION
GASIFIERS
RECTISOL
ACID GAS
REMOVAL,
SNG DRYING
MET11ANATI0N
-------
5.1.1. THE PROCESS
Coal gasification involves a series of chemical reactions in
which carbon from coal is combined with hydrogen from steam to
form methane, which constitutes 97 vol % of the product SNG. The
heat required by the process is supplied by partial oxidation of the
gasification coal with pure oxygen. Much of the heat is subsequently
recovered by in-process generation of steam which is then reused as
reaction steam and to supply equipment-driving energy (augmented by
additional steam generation in auxiliary coal-fired boilers).
The reaction and conversion steps in Figure 5-1 include
Pressure Gasification ~ coal, steam,and oxygen are reacted under
controlled conditions of temperature and pressure to produce a crude
gas containing methane, hydrogen, carbon monoxide, carbon dioxide, excess
steam,and various by-products and impurities. Only some 40% of the
plant's end-product methane (SNG) is produced in the gasifiers. The
remainder of the methane is produced in subsequent reaction steps.
Gas Scrubbing — the crude gas is scrubbed and cooled with water, which
removes tar and oil by-products and phenolic waters. The phenolic
waters are subsequently processed for recovery of by-product phenols.
Shift Conversion — excess carbon monoxide in the crude gas is 'shifted'
(converted) to carbon dioxide to provide the 3-to-l ratio of
hydrogen-to-carbon monoxide needed for the subsequent synthesis of
additional methane.
Gas Cooling — the shifted gas is cooled again to remove additional
hydrocarbon oil by-products and residual phenolic water.
Recti sol — low temperature methanol is used to selectively absorb and
remove HgS and CO2 from the cooled gas. Pre-cooling at the Recti sol
unit entry also recovers by-produce naphtha.
Methanation — carbon monoxide and hydrogen are catalytically combined
to produce methane and by-produce water. About 60% of the end-product
methane is produced in the methanation step.
Compression -- the dry, purified SNG is compressed and delivered to the
pipeline with a heating value of 980-1000 Btu/SCF.
Phenosolvan — a selective solvent (isopropyl ether) extracts by-
product phenols from phenolic waters. The reclaimed water is stripped
of H?S and NH_, and is further processed for complete reuse within the
plant.
Claus Unit — H2S is catalytically converted to by-product sulfur and any
residual gaseous sulfur compounds are incinerated to SO2 and removed in
a subsequent 'tail gas' treating unit.
Oxygen Plant — pure oxygen is cryogenically extracted from atmospheric
air.
Auxiliary Services — these include coal-fired steam boilers with electro-
static precipitators to remove fly ash and stack gas scrubbers to
remove SO^. A closed loop, evaporative cooling water system as well
as extensive air cooling is provided. Extensive wastewater treating
and reuse is also provided.
III-H-53
-------
5.1.2 OVERALL PROCESS MATERIAL BALANCE
The overall material balance (not including the auxiliary steam
boilers) for the gasification process producing 250 MM SCFD of SNG
can be summarized as follows:
Tons/day wt%
INPUTS:
Gasification coal 21,860 41.48
Steam and water 25,160 47.74
Oxygen 5,680 10.78
TOTAL 52,700 100.00
OUTPUTS:
Product SNG 5,440 10.32
By-product phenols 105 0.20
By-product sulfur 175 0.33
Claus unit tail gas^) 617 1.17
Ammonia plus water 800 1.52
By-product and reclaimed water 21,581 40.95
By-product hydrocarbon fuels 1,475 2.80
C02 vent gas 16,631 31.56
Gasification ash 5,876 11.15
TOTAL 52,700 100.00
(1) Does not include air used to incinerate the tail gas, and the
material balance excludes the subsequent tail gas treating
inputs and outputs.
The amount of coal in the above material balance for producing 250 MM SCFD
of SNG is very specific to the particular coal being used. For other
coals, the amount may range from 18,000 to 36,000 tons per day. The steam
and oxygen requirements may also vary over a wide range.
5.1.3 WATER BALANCE
Coal gasification requires large amounts of water as a source of
hydrogen for producing methane SNG, for process cooling, and for
generating steam energy. The gasification design under discussion will
require about 5,100 gpm of raw water intake (8,200 acre-ft/year) to produce
250 MM SCFD of SNG. This amounts to 1.2 pounds of water intake* per pound
of coal, including the boiler plant coal and including water used by the
strip-mining operation. Very little of the raw water intake is destroyed.
In fact, about 80% of the water will be returned to the atmosphere, when
we include the combustion water that will be returned wherever the product
SNG is burned.
*This water requirement is very dependent on the particular coal
involved and upon the specific plant design. It might range as high as
2 to 2.5 pounds per pound of coal.
III-H-54
-------
Various water flows are broken down as follows:
TOTAL INPUT
5100 gpm
%
Process Consumption
ToT2
Return to Atmosphere
(Evaporation, blowdown, scrubbing)
69.6
Disposal to Mine
(in sludge, boiler ash, gasifier ash)
8.4
Others
(slurry pond and mine uses)
11.8
TOTAL
100.00
There will be no discharge of effluent wastewater. Water not evaporated
or converted to SNG is ultimately buried as wet ash and sludge in the
strip-mine pits as they are filled and graded for land reclamation.
5.1.4 STACK GASES
A detailed summary of stack gases from the various plant subunits
including stack height and heat release is shown in Table 5-1.
5.1.5 OTHER ENVIRONMENTAL FACTORS
Aside from the major environmental impacts which are air
emissions, water consumption, and mining operations, other environmental
factors are minor by comparison. The following is a brief survey of these
factors.
All wetted ashes, water treatment sludges, and blowdowns are
ultimately disposed of in the mining operation for land reclamation. The
specific process proposed in this design for sulfur plant tail gas treating
and boiler stack gas scrubbing is the Chiyoda Thorobred Process developed
in Japan. It converts SO2 to dry calcium sulfate (CaS04).
The CaS04 is of high enough quality to perhaps find a market in competition
with natural CaS04 (gypsum). Other S0£ removal processes may or may not
produce a marketable end-product. In any event, if the selected process
produces a 'throw-away' end-produce sludge, it could be disposed of in the
mining land reclamation along with the wetted ashes and water treatment
sludges.
The plant will require about 30 MW of electric power for lighting,
instruments, air-cooler fans, pumps, and other uses. With 250,000 horsepower
of steam turbines and a multitude of air-cooler fans, in-plant noise will
be a distinct problem but not an insurmountable one. A limit of 60-70 dBA
at the plant property line should be realistically attainable. A very large
emergency flare system will be required. When flaring at maximum
emergency conditions, the flame will be quite high and very nrsy. This
condition, however, should occur only rarely.
III-H-55
-------
TABLE 5-1 -- STACK GASES FROM COAL GASIFICATION-ENGLISH UNITS
Total Stack
Gases
Tons/day of Emissions
Height,
feet
MM SCFD
tons/day
OF
ft/sec
S02
N0„ Partic.
Claus tail gas
35
1,486
250
80-100
2.8
150-300
8o1ler stacks
845
32,800
250
80-100
9.4
15.9 1.8
150-300
Superheater stack
138
5,370
600
80-100
1.7
1.5 0.026
150-300
Nitrogen vent
502
18,500
100
80-100
(almost pure nitrogen)
150-300
CO2 vent
211
12,200
65
80-100
(0.5 tons/day of COS)
150-300
CO2 vent
78
4,400
50
80-100
(99% CO2 and 1% CH4)
150-300
Coal lock vent
77
2,900
65
80-100
(99+%
air, lOppmv H2S)
150-300
Ash lock vent
245
9,400
175
60-80
-
0.5
100-200
Coal conveyor vent
39
1,470
65
60-80
-
0.1
100-200
Emergency flare
(design
data unavailable)
-
-
250-350
Heat Release lbs/MM Btu Heat Release
MM Btu/day SO? NOy Partic.
Claus tail gas 363,000O) 0.015
Boiler stacks 70,8000<2) 0.27 0.45 0.05
Superheater stacks 10,200(3) 0.33 0.29 0.995
(1) Based on heating value of coal fed to Lurgi gasifiers
(2) Based on heating value of coal fines burned in boilers
(3) Based on heating value of by-product oils burned in superheater
-------
A good many storage tanks for chemicals, catalysts, and liquid
by-products will be required. Coal and sulfur by-product storage piles
will also be needed. There should be no problem with the periodic
disposal of spent catalysts via burial in the strip-mine pits.
As can be noted from Table 5-1, electrostatic precipitators on the
boiler stacks and other dust control measures will provide particulates
emission levels (Ibs/MM Btu of heat release) that will satisfy any
anticipated regulations. Most gasification plants will be located at the
'mine mouth,1 i.e.,adjacent to the coal fields, and should therefore be in
fairly remote sites. Such sites will probably require the concurrent
construction of:
--access roads and perhaps railroads
--water supply pipelines
--gas product pipelines
5.1.6 OTHER PLANT CONFIGURATIONS
Other design configurations may vary significantly from that
described above. In particular, the choice of whether to burn coal fines
to produce steam (as described herein), or whether to gasify the fines to
produce low-Btu gas for steam generation or for compressor drive energy
is one which is very difficult to assess. Whether to utilize very extensive
air-cooling as in the specific design discussed herein, or whether to use
less costly water-cooling, is another difficult choice to face. Finally,
coals of other sulfur and ash contents as well as other heating values
would result in different emission spectrums. Although most of the coal
gasification projects now underway in the U.S. are planning to use Lurgi
process technology, this may not always be the case. Two other processes,
dating back to the older low-Btu 'towns gas' era, have about the same
amount of coirenercial experience as the Lurgi process. These are the
Koppers-Totzek gasifiers with about 16 commercial installations, and the
Winkler gasifiers also with about 16 commercial installations. Both of
these processes were developed to operate at essentially atmospheric
pressure, as contrasted with the Lurgi gasifier's operating pressure of
350-450 psi.
In addition to the Lurgi, Koppers-Totzek, and Winkler processes,
there are a whole host of 'second-generation' process research programs
underway in the U.S. to develop more advanced gasifiers. These may be
found in reference 5-1, pages 51, 52.
5.2 ALTERNATIVE DATA FOR COMPARISON
The material in Section 5.1 is based primarily on an EPA report
oriented toward providing approximate environmental data and very
general process information. Emissions from coal-fired power plants,
found in Section 2, may be compared with gasification data. The
following is a summary of other environmental impact estimates associated
with gasification.
III-H-57
-------
A comparison of environmental factors between high- and low-
Btu gasification process (as well as other energy technologies) is
shown in Table 5-2. This table is based on 10'^ Btu/day energy output
from each process, whereas the previous example was based on about one-
third of that output. Low-Btu plants have significantly smaller
environmental influences except in land use. The main difference
between western coal and Illinois coal is the four-fold greater SO2
emission and about 40% greater solid wastes issuing from the latter.
Another proposed plant using the Lurgi process is the El Paso
National Gas Company's gasification project, to be located at Burnham,
N.M.5-3 The first complex, scheduled to be operational by 1978, would
have an initial output of 288 MM SCFD of SNG. The combined output of this
facility and a second complex, due to be operational by 1981, would be
785 MM SCFD. At that time the total project emissions would be about
20 t/d each of SO2 and N0X.
5.2.1 WESCO PROJECT
A detailed environmental impact statement was presented recently
by Western Gasification Company (WESCO).5-4 if built, the project will
also use a Lurgi gasifier and initially produce 250 MM SCFD (annual
average) of SNG by 1980. Later, 3 more identical plants would be
added, totaling 1000 MM SCFD by 1985. The projected emissions from one
plant unit are:
t/d
S02 9.8
N0X 18.6
Hydrocarbons 11.9
Particulates 0.88
COS 0.072
H2S Trace
These emissions added to those due to strip mining (not shown) would
adversely affect ambient air quality and visibility since monitoring at
the WESCO site showed extremely low levels of pollutants much of the time.
Since the operation would have no liquid effluents, there should be no
pollutants discharged to directly affect surface water quality. Ground
water quality could be affected slightly over the long term by the
disposal of ash and other solid wastes (containing waste water) in the
mined-out area.
II-H-58
-------
Table 5-2: Source: Council on Environmental
Quality Sixth Annual Reports'1975 (Ref. 5-2)
Environmental Factors for Some Emerging Energy Technologies
Low Dlu gasification
high Utu gasification
Coal liquefaction
Oil shalo
Western coat
Illinois coal
Western coal
Illinois cost
Western coal
Illinois coal
Colorado
Air (Ib/h)
Particulars
0 8C
0 66
727
944
633
612
453 G
SO,
580
2,250
l.UOO
10,400
1,493
1.957.7
5.324.1
NO.
1.130
1.130
7.110
7,770
8.507.5
8.507.5
1.966.7
CO
32 3
32 3
377
414
340
340
174 5
»tc
32 6
32.5
lis
126
2,607.6
2,607.6
2.652 3
NHi
SC 0
45 4
34 7
54 7
—
Water (Ib/h)
Suspended sotlds
0
0
0
0
0
0
0
Dissolved solids
0
0
0
0
0
0
0
Organic material
0
0
0
0
0
0
0
Thermal (Qtu/h)
Negligible
Ncgllglbia
Negligible
Negligible
Negligible
Negligible
Negllglblo
Solid wastes (Ion/day)
5,350
7,320
5.560
7,930
5,519
8.423
1G4.3X10"
Lond use (acres)
3,190
3,190
1.400
1.400
3.254
3.254
2,000
Water requirements {gal/day)
7.46X10*
7.4GX 10'
64 2X10'
64.2X101
33.3X10*
33.3X10*
21 1x10*
Occupational health (par year)
Deaths
0.71
0 71
1 a
1.8
0 511
0.511
0 755
Injuries
14 2
14 2
61
61
9 9
9 9
79 2
Man days tost
7,1.00
7,500
16,600
16.600
2.372
2,372
77 0
Efficiency (percent)
Primary product efficiency
75.B
75 a
66 2
67 9
62 5
62 5
66 7
Total products efficiency
86 S
83.9
68 2
67.9
62 5
62 5
79 7
Overall elhclency
B6 5
83 9
68 2
67.9
62 5
02.5
76 9
Ancillary energy (CJtu/day)
0
0
0
0
0
0
5.S9X 10«
On the basis of o lO" Btu/doy output from each process.
Sourco lladian Corporation* A Western Region*! Enotgy Dovotopmonl Study: Pilmory Envltonmcntal impacts. Vol. 2« prepared for the Council on Environ*
mental Quality and Iho federal Energy Administration under contract no EQ4AC037 (Springlield, Va : National Technical Information Servlco, 1975).
-------
5.2.2 TRACE ELEMENTS
The following is taken from the WESCO report. A study of trace
element disposition in the gasification plant and coal fired boilers was
performed by Battelle-Columbus. Correlation of the theoretical estimates
generated by Battelle were conducted through sampling and extrapolation
by SASOL and Commercial Testing and Engineering. Table 5-3 presents the
estimated trace element disposition for the boilers. Trace elements are
expressed as the percent by weight of the total input to the boilers.
Table 5-3
Estimated Trace Element Disposition
for WESCO Coal
-Fired Boiler
Component in
Inlet
Coal
Parts/million
Bottom
Electrostatic
Stack Gas
Emissions to./?/
by weight
Ash^/
Precipitator!/
Scrubberl/
Atmosphere
Ash
24.7
74.5
0.65
0.15
Arsenic
1.2
4.4
94.6
0.8
0.20
Cadmium
0.66
16.0
82.7
1.0
0.30
F1uorine
210
1.2
26.8
57.6
14.4
Mercury
0.01
4.4
13.0
0.1
82.51/
Lead
5.5
9.7
89.3
0.8
0.2
Antimony
0.42
10.0
89.0
0.8
0.2
Selenium
0.74
6.0
58.0
34.7
1.3
Beryl Hum
3.4
16.9
82.2
0.7
0.2
Tellurium!/
0.20
~ —
1/ Numbers expressed as percent of total input weight of individual elements
2/ Mainly enriched on fly ash (except mercury)
3/ 82.48% as vapor, 0.02 enriched on fly ash
4/ No data exists for coal similar to that which would be used by WESCO.
As can be seen from these calculations, the primary point of deposition
for the boilers is expected to be in the electrostatic precipitator for
all of the trace elements except mercury, fluorine, and possible selenium.
Primary deposition of selenium and fluorine would probably occur in the
stack gas scrubbers. The remaining fly ash could be somewhat enriched in
selenium and fluorine. Most of the mercury would exit the stack in the
gaseous form. The coal which would be used for the proposed WESCO plants
has a relatively low content of mercury and the net emission of mercury
should be correspondingly low for the boilers.
Table 5-4 presents the estimated trace element disposition for the
gasification section in percent of total input weights. As can be seen
from this table, it is estimated that virtually all of the trace element
constituents of the coal would be deposited with the system. The only
emission would be 9 percent of the input mercury due to the low temperatures
in the methanol scrubbers.
III-H-60
-------
Table 5-4
Trace Element Disposition
Gasification Section
18009F
650aF
Vapor* Condensed Vapor Condensed
45"F
Vapor Condensed
-50 aF
Vapor Condensed
Hg Major Ash Com- Hg
ponents plusBe&cv
5b As (estimate 5b
Se 4"9 PPn) Se
Te
Cd
Fb
P
Te
Cd
Pb (Pb5) Hg Cd
F** Te(6.7%) 5c
Sb
Te(93.37.)
Hg-(8.6%) Hg(91.47.)
Te(6.7%)
-4 -2
* Initial gas-phase concentrations- Hg =* 4.6 x 10 ppm, 5b = 3.1 x 10 ppm,
-2 -2 -1
Se = 8.4 x 10 ppm, Te = 1.4 x 10 ppm, Pb = 2.4 x 10 ppm,
Cd = 5.4 x 10
** See text
-2
ppm, P = 99.8 ppm.
Note: Numbers are 1007. unless otherwise noted.
Arsenic would be retained in the condensed products of the rasifier.
Lead would form solid lead sulfide and be contained primarily in the bulk
ash. Fluorine would probably react with basic ash components. Essentially
all cadmium and antimony would condense in the 45°F temoeratura ranqe of the
process. Tellurium would be condensed finally in cold methanol scrubber.
The remaining trace elements in the coal, except mercury, are expected
to be retained in the gasifier ash as tfie insoluble oxide or sulfate form.
nr-H-61
-------
The importance of trace element release from coal gasification
can be gauged by related experience with coal combustion. It appears
that there will be no new or unfamiliar releases with regard to our
wide experience with coal-fired plants. However, the long-term effects
of trace elements from direct combustion of coal are only now being
studied and are not fully documented.
5.3 MODELING
Section 2, Central Station Conversion, includes description
of a Climatological Dispersion Model for SO2 movement in air. The model
yields SO2 isopleths averaged for one year. Specifically, calculations
simulated the emissions from a coal-fueled power plant projected for
Trimble County, Kentucky. The model is constructed such that the average
SO2 emission rate is a coefficient of the model, i.e., the resulting
concentrations are directly proportional to the average emission rate.
Thus, the result shown for coal combustion may be scaled by the emissions
ratio to apply correctly to gasification or to other sources situated
in the same area, e.g.:
SO2 concentration SO? rate for gasification x SO2 concentration
for gasification = SO2 rate for combustion for combustion
For approximate calculations, the numbers below indicate relative SO2
emission rates based on Western coal.
Relative yearly average SO2
emission rate (unitless)
Low-Btu Gasification 1
High-Btu Gasification 3
Low-Btu Gas Power Plant 7
Coal-Fired Power Plant 30 to 75
Of course, these figures are approximate and can be used only for very
rough estimates. As a precaution, if the wind data, stack heights,
exit gas temperature, and exit gas velocities are substantially different
for two cases, computed isopleths for one will not be accurately scalable
to the other case.
5.3.1 CASE STUDY: WARRICK COUNTY, INDIANA
The climatological Dispersion Model was used to simulate SO2
isopleths (concentrations) in Warrick County assuming a 250 MM SCFD,
high-Btu, coal gasification plant were located there. Total emissions put
into the model were taken from the WESCO environmental statement,
summarized in reference 5-3. Specifically, three different gas stacks
III-H-62
-------
were simulated in one calculation assuming close proximity of the
stacks. The data are:
S02
Source Emission,g/sec Stack Height, ft Exit Velocity, ft/sec Exit Temp,QF
Flare 0.89 300 100 +800
Coal Lock 8.6 200 100 300
Gasifier 93.6 300 100 240
Warrick County weather data also entered into the calculations (explained
elsewhere). The background SO2 concentration was assumed to be zero even
thougtj there is a source there at present. In case it is desired to know
the effects of other sources of SO2 in the same area, a simple addition of
isopleths will give an approximate answer for local concentrations.
Figure 5-2 shows the results of a calculation for the projected
Warrick County high-Btu coal gasification plant located at a single site.
The circle represents the source (3 stacks) location of SO2. Note that
maximum concentration occurs some distance above the source. The plus
signs with associated numbers represent local maxima of concentration
strengths indicated (by the numbers). All locations within the graph
border but outside the 0.05 isopleth have a concentration of 0.025
micrograms per cubic meter. The accuracy of these predictions is
discussed in Central Station Conversion.
5.4 RESEARCH NEEDS
At this time, the greatest uncertainty in estimating environmental
impact of gasification is in the data on emissions. No full size plant is
on stream presently. European experiences are not readily available,
perhaps in part because of less environmental activities there. American
companies will not release pilot plant data. Only estimates are made avail-
able in the form of impact statements. Thus, the first priority of
future research is to visit various pilot plants and gather as much new
data as possible. Better communication with European plants would also
ease this problem.
With regard to modeling air and water impacts, it appears that all
gasification and low-Btu combustion may be treated in the same way as
coal-fired plants. Location of gasification plants in areas without
coal-fired plants will usually not violate air quality standards, thus
detailed modeling would not be warranted. If new plants are added in
already burdened areas, the air effects (at least) will be additive or
scalable by simple ratio multiplication as cited above. However, there is
no doubt that further improvements in models are needed, as well as
more field tests of SO2 dispersion and more detailed weather data.
See Section 2 for a detailed discussion.
The question of long term impacts is much harder to answer. The
effects of leaching from ash-filled pits, water impoundments, and trace
element releases are not all well understood and are sensitive to
specific locations.
III-H-63
-------
Figure 5-2. CIimatological Dispersion Model at EVY. 1975 Wind Rose 770510 23:06
CONTOURS OF SULFUR DIOXIDE, MICROGRAMS PER CUBIC METER
o.i
0.07
0.05
0.025
0.0
1 KM
v..
/
/
—• ¦_ »•.
/
\
/
/'
\
/
\
\
\
/
/
r' +
t
s r'"
\V—-
.1
\
\
N .•
/
/
/\
/ \
/
\
)
/'
/
/
\
/
/
III-H-64
-------
REFERENCES
5-1. Milton R. Beychok, Process and Environmental Technology for Producing
SNG and Liquid Fuels, EPA-660/2-75-011, May, 1975.
5-2. Council on Environmental Quality, Sixth Annual Report, 1975.
5-3. L. K. Stone, "Emissions From Coal Conversion Processes", Chem. Eng.
Proqr. 71 (12), Dec. 1976.
5-4. Western Gasification Company (WESCO), Coal Gasification Project and
Expansion of Navajo Mine by Utah International, Inc., Final Environmental
Statement, Vol. I, Bureau of Reclamation, USDI, Jan., 1976.
I1I-H-65
-------
6. ACID MINE DRAINAGE
The drainage from coal mines has considerably dissolved as well as
suspended contaminants. The quantity and nature of these contaminants
depends on the type of mine, the mechanism producing the drainage from
the mine and the disposal of coal after it is mined.
In the so called acid mine drainage, the water quality parameters
which produce considerable environmental impact are the acids, the total
dissolved solid contents as revealed by conductivity of the water from
the mines, the hardness which reflects the presence and degree of occur-
rence of calcium, magnesium,or iron and sulfates as the decomposition of
pyrites produces large amounts of sulfate.
The principal water pollution problems associated with strip mining
of coal are increases in suspended solids, alkalinity and salinity-and
heavy metals in water bodies within the mine watershed. The acid mine
drainage has been observed and extensively studied since late 19th cen-
tury. The formation of sulfuric acid from pyrite (Fes2) has been a major
problem in the Ohio River basin, particularly in Ohio, West Virginia,and
Pennsylvania. The effects of acid mine drainage range from ruining
streams and lakes to corrosion of bridge piers, locks, boat hulls,and
pumps.
Extensive literature exists on various aspects of acid mine drain-
age. However, development of mathematical models to estimate the acid
load discharge into streams seems to have received some attention only
recently. The engineering aspects of acid mine drainage were discussed
by Smith (1) as a starting point in the development of mathematical
models of acid mine drainage. Shumate, et al. (2) appear to have suggested
the first mathematical model of acid drainage. A model of drainage in a
surface mined area has been developed by Sternberg and Agnew (3). Stern-
berg and Agnew's model deals with changes in ground water flow due to
uniform rate of deep percolation over the spoil bank. Morth et al. (4)
expanded on the ideas of Shumate et al. (2) and developed a mathematical
model of acid mine drainage. This model may be used to estimate the run-
off from the mined area as well as the acid load. Validation results of
the model by Morth et al. (4) indicate that the model performs satisfac-
torily for the data from one mine which was used in the study.
In conclusion, the mathematical modelling of acid mine drainage is
in its initial stages. Considerable work remains to be done in vali-
dating the models already proposed and in improving them.
Another important coal-mine-related environmental problem is
coal stockpiling. Although there are a few isolated instances where
significant quantities of coal are stockpiled, this condition is expected
to change. The coal pile leachates have been observed to change the water
quality characteristics, although these changes depend upon the hydrologic
characteristics of the site, most importantly that of the precipitation,
and the site conditions. The water quality changes due to ccal pile
leachates have not received much attention.
III-H-67
-------
In the surface mining operations,large volumes of sediment are gen-
erated. Some of the amounts and orders of magnitudes of the sediments
generated by surface mining are given below (5).
Table 6.-1 Representative Rates of Erosion from Various Land Uses
Metric Tons/ Tons/ Relative to
sq. km./year sq. mi./year Forest - 1
Forest 8.5 24 1
Grassland 85 240 10
Abandoned Surface Mines 850 2,400 100
Cropland 1,700 4,800 200
Harvested Forest 4,250 12,000 500
Active Surface Mines 17,000 48,000 2,000
Construction 17,000 48,000 2,000
Source: "Methods for Identifying and Evaluating the Nature and Extent of
Nonpoint Sources of Pollutants," U.S. EPA, Office of Air and
Water Programs, Washington, D.C., EPA-430/g-73-014, October 1973.
Table 6.-2 Comparison of Sediment Yields in Beaver Creek Watershed
Area Yield (tons/sq. mi.) Factor
Unmined Watershed 28 1
Mined Watershed 1,930 69
Spoil Bank 27,000 968
Haul Road 57,600 2,025
Although these impacts of coal mining on water quality have been
well known for quite some time and numerous mines have been monitored,
the extent and severity of the impacts are difficult to estimate pre-
cisely. Only rough estimates of the impacts can be given. The reasons
for this state of affairs are that the geology, hydrology, the soil
characteristics, and the chemical composition of the mined coal affect
the quality of water from the mines as well as from coal stockpiles.
III-H-68
-------
Another important reason for the difficulty in making precise esti-
mates of the changes brought about by mining operations is that
of the water quality parameters are sampled for short periods of time,
and often without the associated measurements of flow rates. This type
of sampling will not enable reliable estimation of the seasonal variation
in water quality parameters, establish definitive relationships between
water quality and quantity, and in fact makes it almost impossible to
construct and validate mathematical models. Consequently, long term sys-
tematic measurements of hydrologlc and water quality parameters at well
distributed mine sites are needed.
REFERENCES
1. Smith, E.E. "Engineering Aspects of Acid Mine Drainage," proc. Second
Annual Symposium, Water Resources Research, The Ohio State University,
1966.
2. Shumate, K.S., Smith, E.E., and Brant, R.A. "A Model for Pyritic
Systems" 157th National Meeting, Am. Chem. Soc., Div. of Fuel Chem-
istry, Minneapolis, MN. 1969.
3. Sternberg, Y.M., and Agnew, A.F. "Hydrology of Surface Mining - A
Case Study," Water Resources Res., Vol. 4, No. 2, 1969.
4. Morth, A.H., Smith, E.E., Shumate, K.S., "Pyritic Systems, A Mathe-
matical Model" Rept. No. EPA-R2-72-002, U.S.E.P.A., Washington, D.C.
Nov. 1972.
5. Mills, T.R., and Clar, M.L. "Erosion and Sediment Control during
Surface Mining in the United States." Sixth Symposium on coal mine
drainage research, sponsored by National Coal Association, Bituminous
Coal Research, Inc., Louisville, Kentucky, 1976.
III-H-69
-------
7. AQUATIC THERMAL PLUMES
7.1 INTRODUCTION
Conventional steam-electric technology results in the production of
about 2 units of thermal energy for every unit of electrical energy gen-
erated. Because efforts to make beneficial use of this by-product heat
have generally been uneconomic, this heat is usually considered a waste
product and is dissipated to the environment.
Systems for the dissipation of waste heat may be classed as once-
through or open cycle. In once-through cooling, the heat is rejected
directly to the aquatic environment. Water is withdrawn from a natural
waterbody, heated in the condensers, and discharged back to the water-
body. Although each phase of this operation involves aquatic environ-
mental impacts, the concern of this section is with the means necessary
to predict the configuration of the resulting plume of heated water.
In closed-cycle systems, the water heated in the condensers is
passed through some type of artificial cooling device which transfers
the heat to the atmosphere. Among the cooling devices in common use are
cooling ponds, spray canals, and various types of cooling towers. In
most of these devices, there is direct contact between the cooling water
and the atmosphere so that some of the water is lost through evaporation.
This does not occur in dry cooling towers which operate on the same prin-
ciple as the automobile radiator. Whenever evaporation does take place,
the content of dissolved solids in the remaining water tends to increase,
a phenomenon which could lead to precipitation and corrosion problems.
These difficulties are avoided by continuously discharging some of the
cooling water to the natural water source. This discharge is called
blowdown. In order to replace the cooling water lost to evaporation and
blowdown, additional makeup water is continuously withdrawn from the
natural water source and added to the circulating cooling water. Con-
sequently, in any conventional evaporative closed cycle cooling system
there will be a thermal plume discharged to the natural waterbody. Of
course, the size of this plume will be much less than for a comparable
once-through system. To cool a single 1200 MWe nuclear reactor with a
typical condenser temperature rise of 20°F requires about 1800 cfs of
water if once-through cooling is used. The same power plant using evap-
orative cooling towers would require a makeup of about 36.5 cfs, evapo-
rate about 30 cfs, and blowdown about 6.5 cfs. (These figures would
vary somewhat depending on weather conditions.) Although the volume of
heated effluent from a cooling tower plant is only about 0.4? as large
as in the comparable once-through situation, it may still have signifi-
cant impacts if discharged into a small stream.
In order to assess potential ecological impacts of a plume or
plumes, it is necessary to know the configuration of the plume under
various environmental and plant operating conditions. For an operating
plant, field studies may provide suitable information. To study a pro-
posed plant or a proposed modification to an existing plant, 'Some pre-
dictive technique is needed. Among the predictive techniques in common
use are physical and analytical models. For the purpose of the ORBES
III-H-71
-------
project, the use of physical models is considered impractical. Conse-
quently the main effort has been directed toward recommending appropri-
ate analytical models. The present usage of the term analytical model
is broader than is the case in the fluid mechanics literature. Here the
term is meant to include empirical correlations derived from field data,
as well as closed form and numerical solutions to the relevant equations.
In principle, 1t is possible to obtain a complete numerical solu-
tion for a thermal plume under any known circumstances. In practice
such an undertaking is far beyond the capabilities of any present or
projected computer system. This has led to the development of models
which are valid only in restricted regions of the plume. These regions
are conventionally called the far-, intermediate-, and near-field
regions. The far-field region is loosely defined as the region far from
the outfall where the dynamics of the plume is controlled by the ambient
flow field and by buoyancy. In this region the details of the discharge
design are unimportant. The near-field is the region close to the dis-
charge structure where the momentum of the discharged flow and hence the
details of outfall geometry dominate the flow. The intermediate-field
is the transition zone where the flow shares some characteristics of the
other zones. Because the dynamics of flow in these zones are quite dif-
ferent, different models are available for each zone and the title be-
comes a convenient means of classification. Because the ORBES is in-
tended to examine basin wide effects rather than local effects, only
far-field models will be considered further.
7.2 FAR-FIELD MODELS
7.2.1 Review of Available Models
In the ORBES region, rivers are the only potentially significant
source of condenser cooling water. A review of the literature disclosed
five far-field thermal plume models suitable for use in rivers. These
models are: 1. Edinger-Geyer
2. STREAM
3. COLHEAT
4. Iowa Thermal Regime Model
5. ESTONE
The Edinger-Geyer model [1] is a simplified one-dimensional, steady-
state model based on a balance between advected heat and heat transfer to
the atmosphere. The model follows a fully mixed slug of water as it pro-
gresses downstream, allowing it to receive heat of artificial origin and
to transfer heat to the atmosphere. The model essentially consists of
solving the equation
T = E + (Tm-E) exp
kA
pCnQ
R
where T = fully mixed water temperature
E = equilibrium water temperature (the water temperature for
which heat exchange with the atmosphere is zero.)
T = fully mixed water temperature at a heat dise+rarge
III-H-72
-------
k = heat exchange coefficient
A = surface area of river reach
p = density of water
C = specific heat at constant pressure of water
Qjj = volumetric flow rate of river.
In order to allow the calculations to be made on a desk calculator, if
desired linearized equations and charts for the determination of k and E
are part of the Edinger-Geyer formulation. The model can be applied for
arbitrary time steps and the simulation of a series of heat sources is
readily accomplished.
STREAM is a water quality simulation model developed by the Ohio
River Valley Water Sanitation Commission (ORSANCO) [2]. The thermal
subroutine of STREAM is a simplified steady-state, one-dimensional, far-
field model. No formal energy balance equation is written. Rather,
heat of artificial origin 1s assumed to be dissipated in one day's
travel from the heat source. Natural water temperatures are input to
the model and excess temperatures due to heat of artificial origin are
assumed to decay exponentially. The decay constant, although formally
similar to a heat transfer coefficient, is a constant whose value is
independent of meteorological conditions.
COLHEAT [3] is a one-dimensional, far-field model in which the
river is divided into a number of fixed reaches. Water is allowed to
flow from one reach to the next where it exchanges heat with the atmos-
phere according to an explicit heat budget, gains heat of artificial
origin, and is mixed to uniform temperature. This process is repeated
in each time step. Although the model usually computes one-dimensional
temperatures averaged over a cross section, the velocity of flow is not
assumed constant. Rather each cross section is assumed to be made up of
an inner and an outer trough, the latter having a greater velocity of
flow. Subroutines are available to account for transient behavior,
tributary inflows, and density driven currents such as are encountered
in estuarlne situations.
The Iowa Thermal Regime Model [4] is based upon the numerical solu-
tion of the one-dimensional convection-diffusion equation. It predicts
the longitudinal variation of fully mixed temperature in the river. The
total river length is subdivided into smaller reaches which are solved
separately and then linked by boundary conditions at the cornnon nodes.
Each reach can have multiple thermal inputs and tributary inflows.
ESTONE [5] is a complete one-dimensional hydrodynamic and thermal
model which can predict cross section average values of velocity, tem-
perature, and salinity in rivers and estuaries. The model is based on a
discrete element formulation of the one-dimensional conservation equa-
tions for momentum, energy, and salinity. This model is particularly
suited for conditions 1n which tidal reversals and density currents are
present.
Ill-H-73
-------
Because each of these models is highly idealized, none can be
accepted for predictive use without adequate validation involving the
comparison of model predictions with field measurements. Validation of
Edinger-Geyer, STREAM, and COLHEAT was performed on a uniform basis as
part of the Ohio River Cooling Water Study [6]. Each of these models
was used to predict the daily average, one-dimensional, longitudinal
temperature profile for the 300 mi reach of the Ohio River from Pitts-
burgh, Pennsylvania, to Huntington, West Virginia. Predictions were
obtained for each day in the year 1964. Measured meteorological con-
ditions, river flow rates, and artificial heat additions from 11 power
plants and 7 Industrial plants were varied on a dally basis in accordance
with historical records. Predictions of daily average temperatures were
compared with measurements obtained at 4 widely spaced stations. Figure 7-1
shows the dally COLHEAT predictions at Huntington West Virginia, com-
pared with measurements obtained by an 0RSANC0 robot monitor. It is
seen that COLHEAT reproduces the trend of the observed data well through
the year. The maximum deviation occurs in May, June, and July when
COLHEAT overpredicts by about 2°C. Figure 7-2 is a scatter diagram of the
same data illustrating the tendency of COLHEAT to overpredict the highest
temperatures. Similar trends were noted in the COLHEAT-data comparisons
at 3 other stations. In comparisons of the Edinger-Geyer predictions
with the same data sets,a stronger tendency to underpredict low tem-
peratures and to overpredict high temperatures was observed. The Edinger-
Geyer predictions tended to diverge from the measurements as the location
of comparison was taken farther downstream. The STREAM predictions were
seen to consistently underpredict measured temperatures. Based on these
comparisons and consideration of the theoretical completeness of the
models, COLHEAT was recommended as the most appropriate model for simula-
tion of thermal loading of the Ohio River.
"TCie Iowa Thermal Regime Model has accurately predicted the tempera-
tures on June 2, 1976, at two locations separated by 110 mi on the upper
Mississippi River. ESTONE is said to have yielded good agreement in two
non-successive 6 month simulations of a 152 mi stretch of the Hudson
River estuary. Details of this comparison have not yet been published.
Consequently, at the present time it appears the COLHEAT remains
the most fully validated model available for prediction of far-field
thermal loading patterns in rivers.
7.2.2 Ohio River Temperature Prediction Survey
As part of the Ohio River Cooling Water Study [6], longitudinal pro-
files of one-dimensional, daily average temperature were obtained using
COLHEAT for the 705 mi reach of the Ohio River between Pittsburgh,
Pennsylvania, and river mile 705 (about 100 mi downstream of Louisville,
Kentucky). The river temperatures were obtained from May 24 thru
November 10 under loading conditions expected in the late 1970's. The
thirty-day, once-in-terr-year low flows were used throughout the months of
July, August,and September. Measured 1964 flows were used for the other
dates. Actual daily meteorological data measured during 1964 rfas also
used. The estimated heat additions from 21 power plants were included
in the simulation.
III-H-74
-------
Figure 7-1
Computed and Measured Temperatures at Huntington,
West Virginia COUIEAT Mode
COMPUTED
MEASURED
151 00
91.00
121 CO
IBl.GC
31.CO
61.CO
COTE
-------
Figure 7-2
LO
O)
Measured-Computed Temperature Scatter Diagram y
for Huntington, West Virginia — COLHEAT Model
;.-V/
* '• J /
/
I
re
o\
a
r—»
U)
0_
¦n
Ui
li
i
t
i
o •
Hi
u
I
•/
/
•
j •
Ui
o
o
s:
'/
, 7-
o
CJ
c.
r*
23.
/. *
A' ¦
/
'b.oo
3. SO
7.00
I0.S0
~i
17.SO
14.00
MEASURED
—i
21.00
2U.SQ
28.00
31.S3
3S.00
-------
Figures 7-3 and 7-4 show the predicted natural and thermally loaded
temperatures on July 25, a typical day during the extended low flow
period. The effect of heat additions due to the power plants 1s quite
evident. The effect of different scenarios of power development could
be easily compared by performing simulations of this type. In fact,
reference [6] does present calculations for 3 regulatory schemes intended
to mitigate some of the high temperatures revealed in the previous
figures.
7.2.3 Recoimiendatlons
Although the COLHEAT model has been demonstrated to yield reason-
able predictions 1n the Ohio River, its tendency to overpredict by 2°C
in the summer months 1s only marginally acceptable 1n view of current
thermal standards. It would be very useful to attempt a comparative
validation study of COLHEAT, E5T0NE, and The Iowa Thermal Regime Model.
Such a study would reveal whether the more rigorous formulation of the
later models has resulted in any gain in accuracy. Once the best model
has been selected, simulations of the river temperatures for the various
ORBES scenarios should be performed. The results of these simulations
should allow an evaluation of the relative thermally induced ecological
impacts which could be expected 1n each case.
7.3 ASSESSMENT OF POTENTIAL FOR MARBLE HILL - TRIMBLE COUNTY THERMAL
PLUME INTERACTION
The Marble Hill Nuclear Generating Station Is proposed for a site
1n Jefferson County, Indiana, at River Mile 570 of the Ohio River. The
Trimble County Power Plant is a coal-fired station proposed for a site
1n Trimble County, Kentucky, about 2 mi downstream of Marble H111. The
purpose of this section is to investigate the possibility that these
2 plants, acting in concert,might create a thermal problem in the Ohio
River.
At the Marble Hill site the Ohio River has a width of about 2200 ft
and a normal depth of 24 ft. Flow 1n this reach of the Ohio, known as
the McAlpine Pool, is regulated by McAlpine Dam (downstream) and Markland
Dam (upstream). Since the completion of the dam system, the maximum and
minimum flows at the Markland gaging station have been 465,000 cfs and
10,500 cfs. The average river flow at the site is estimated to be
112,000 cfs.
The Marble Hill Nuclear Generating Station [7,8] consists of 2
reactors with a total net electrical output of 2260 MWe at rated condi-
tions. The 4590 MWt waste heat will be dissipated primarily by mechani-
cal draft wet cooling towers. This cooling system requires the with-
drawal of between 53 and 57 cfs makeup and the discharge of 8 to 10 cfs
blowdown. This blowdown will be warmer than ambient river water by an
amount which depends on river water temperature and weather conditions.
The annual average excess blowdown temperature will be about 24°F and
the greatest value 1s expected to be 56.1°F during January. The highest
actual blowdown temperature expected is 93.9°F.
III-H-77
-------
i
oo
cc
uJ
>
5
c
£
u.
O
in
r\j
s
r-
Figure 7-3
Ohio River Teiqperature Profile,
July 25 - Part 1
LOADS
NO LOADS
55.
"b.co
25.00
SO. 00
75.00
100 00
-+-
125.00
150.00 175.00
MILE PT.
200.00
22S.0Q
250.00
275.00
300.00
325.OZ
-------
or
o
o
u.
o
Figure 7-4
Ohio River Temperature Profile, July 25 - Part 2
LOADS
NO LOADS
CT
r
t-t c
- P
i -*
y,
vO
l/»
r\.
c.
X
Q_
CD
G
s;
o
CJ
ca *
fl
Ijr
-1
3U
391.00 115.03 mo.00 465.(H) 430.00 SIS 00 SU0.00
MILE Pf.
565 00
S90 00
61S 00
MO. 00
ECS.00
69: C.'
-------
The Trimble County Power Plant is a proposed 2 unit coal-fired
plant with a total net electrical output of 990 MWe. Although exact
data is not available, Trimble County can be expected to dissipate about
1980 MWt waste heat primarily by evaporative cooling towers. Conse-
quently the Trimble County blowdown plume should contain about 40% as
much heat as that due to Marble Hill.
Assuming that both plants reject heat to the river at the maximum
expected rate, this would add 4.9x10^ Btu/s to a minimum flow of 10500 cfs.
Under these most severe conditions, the maximum fully mixed temperature
rise of the river would be less than 0.1°F, assuming that no heat is lost
to the atmosphere.
Actually, the 2 plumes are never expected to meet. The original
discharge design for the Marble Hill plant called for a surface discharge
normal to the shoreline. Plume configurations resulting from this design
were extensively studied by the utility and by the Nuclear Regulatory
Commission staff using entirely different models. The utility used a
finite element potential flow code and the NRC used a near-field integral
model developed by Shirazi and Davis. Monthly average and extreme cases
were analyzed using both models. The largest area within the 3°F excess
temperature isotherm predicted by the utility was 15.6 acres. The NRC's
maximum 3°F prediction was 0.6 acres. In all cases the plume was pre-
dicted to hug the shoreline, extending away from shore a maximum dis-
tance of less than 200 ft. Consequently, if the Trimble County outfall
was as large as Marble Hill and was located directly across the river.no
measurable interaction would take place. In fact, interaction is even
less likely than indicated by these model studies because the Marble Hill
outfall design has been changed to a submerged pipe. This design will
result in much more rapid dilution and much smaller areas within
isotherms.
It is concluded that there will be no detectable interaction be-
tween the Marble H111 and Trimble County blowdown plumes. No further
analysis of this question is needed.
III-H-80
-------
REFERENCES
1. J. E. Edlnger and J. C. Geyer, Heat Exchange In the Environment,
Edison Electric Institute, New York, June 1965.
2. Ohio River Valley Water Sanitation Coimrission, Automated Forecast
Procedures for River Quality Management - Volume I: Project Report,
June 1972.
3. Hanford Engineering Development Laboratory, The Col heat River Simu-
lation Model, HEDLTME 72-103, August 1972.
4. P. P. Pally and J. F. Kennedy, A Computational Model for Predicting
the Thermal Regimes of Rivers, IIHR Report No. 169, Iowa Institute
of Hydraulic Research, The University of Iowa, Iowa City, November
1974.
5. A. H. Eraslan, Systematic Application of Transient, Multi-Dimensional
Models for Complete Analysis of Thermal Impact in Regions with Severe
Reversing Flow Conditions, Proceedings of the Conference on Waste
Heat Management and Utilization, eds. S. S. Lee and S. Sengupta,
May 1977 , p. IV-A-43.
6. B. P. Butz, D. R. Schregandus, B. A. Lewis, A. J. Policastro, and
J. L. Relsa, Jr., Ohio River Cooling Water Study, Argonne National
Laboratory, Report No. EPA-905/9-74-004, 1974.
7. Office of Nuclear Reactor Regulation, U. S. Nuclear Regulatory Com-
mission, Final Environmental Statement, Marble Hill Nuclear Generat-
ing Station Units 1 and 2, Docket Nos. 50-546 and 50-547, NUREG-
0097, September 1976.
8. Public Service Indiana, Marble Hill Nuclear Generating Station,
Units land 2, Environmental Report, Construction Permit Stage,
Docket Nos. 50-5TN-546 and 50-STN-547, September 1975.
III-H-81
------- |