5571
Prepared for
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Pollutant Assessment Branch and Chemicals and Petroleum Branch
Contract No. 68-01-6871
Assignment Nos. 7 and 13
EPA Project Officer EPA Task Officers
Jon R. Perry Michael Dusetzina
Kent C. Hustvedt
PRELIMINARY SOURCE ASSESSMENT
FOR HAZARDOUS WASTE AIR EMISSIONS
FROM TREATMENT, STORAGE AND
DISPOSAL FACILITIES (TSDFs)
Final Report
February 1985
Prepared by
William Battye
Charles Vaught
David Zimmerman
Michael Glowers
Elizabeth Ryan
GCA CORPORATION
GCA/TECHNOLOGY DIVISION
-------
450R85003
DISCLAIMER
This Draft Final Report was furnished to the Environmental Protection
Agency by the GCA Corporation, GCA/Technology Division, Bedford, Massachusetts
01730, in partial fulfillment of Contract No. 68-01-6871, Technical Service
Task Orders 7 and 13. The opinions, findings, and conclusions expressed are
those of the authors and not necessarily those of the Environmental Protection
Agency or the cooperating agencies. Mention of company or product names is
not to be considered as an endorsement by the Environmental Protection Agency.
-------
TABLE OF CONTENTS
Page
LIST OF TABLES iv
Chapter 1. Summary and Conclusions 1-1
Chapter 2. Introduction 2-1
Chapter 3. Calculation of Emissions 3-1
3. 1 Emission Models 3-1
3.2 Sources of Facility Specifications 3-1
3.3 Sources of Process Stream Composition 3-9
3.4 Sources of Pollutant Properties Data 3-11
Chapter 4. Calculation of Cancer Risk and Incidence 4-1
4.1 Calculation of Public Exposures 4-1
4.2 Risk Factors 4-4
Chapter 5. Nationwide Extrapolation 5-1
Chapter 6. Uncertainties and Limitations 6-1
Appendix A. Emissions and Risks for Model Plants A-l
Append :.x B. Emissions Model Equations B-l
Appendix C. Default Parameters for Emission Models C-l
Appendix D. Sample Calculations D-l
Appendix E. Development of Unit Risk Factors E-l
Appendix F. Nationwide Extrapolation Based on Mass Balance F-l
-------
LIST OF TABLES
Page
Table 1-1. Nationwide Emission and Risk Estimates Aggregated
by Facility Type 1-2
Table 1-2. Nationwide Emission and Risk Estimates Aggregated
by SIC Codes 1-3
Table 2-1. List of Case Study Facilities 2-2
Table 3-1. Summary of Analytical Results for Impoundments
and Open Tanks 3-12
Table 4-1. Risk Factors for the Top 10 Chemicals by Emissions
From GCA and RTI Rankings 4-6
Table 5-1. Nationwide Emission and Risk Estimates Aggregated by
Source Type for a Unit Risk Factor of 2 E-5 5-2
Table 5-2. Nationwide Emission and Risk Estimates Aggregated by
Source Type for a Unit Risk Factor of 2 E-7 5-3
Table 5-3. Nationwide Emission and Risk Estimates Aggregated
by SIC Code for a Unit Risk Factor of 2 E-5 5-4
Table 5-4. Nationwide Emission and Risk Estimates Aggregated
by SIC Code for a Unit Risk Factor of 2 E-7 5-5
Table B-l. Summary of Empirical Relationships to Determine the
Individual Liquid and Gas Phase Mass Transfer
Coefficients for a Nonaerated Impoundment B-2
Table B-2. Thibodeaux, Parker and Heck Model for Surface
Impoundments B-4
Table B-3. Thibodeaux-Hwang Model for Volatile Organic Emissions
from Landtreatment Operations B-7
Table B-4. Farmer Model for Volatile Organic Emissions from Covered
Landfills B-9
-------
ACKNOWLEDGEMENT
This preliminary source assessment benefitted from the contributions of a
number of individuals and organizations. Michael Dusetzina and Kent C.
Hustvedt served as Task Officers, and their guidance was especially helpful.
The facility visits which provided the basis for this analysis were conducted
by Versar, Inc., Engineering-Science, and Radian Corp. The assistance of the
following GCA staff members is also recognized. Rebecca Sommer conducted the
HEM modeling necessary for this assessment. Those involved in the large data
manipulation task included Tom Warn, David Mis-enheimer, Calvin Overcash, Scott
Osbourn and James Pierce.
-------
1. SUMMARY AND CONCLUSIONS
This report presents a preliminary assessment of organic emissions from
area sources at hazardous waste treatment, storage, and disposal facilities
(TSDFs), and of the risks to the general public associated with these
emissions. The types of potential emission sources which were studied are:
drum storage and handling; tanker unloading; storage and treatment tanks;
surface impoundments for treatment, storage, and disposal; land treatment
facilities; landfills; fugitive emission sources associated with incineration
and deep well injection; and distillation and other recovery operations. The
study includes only fugitive emissions from incinerators; emissions from
incinerator stacks were not included. The study also does not include
fugitive dust emissions from waste piles or from excavation.
This preliminary assessment is based on emissions estimates for 39 TSDFs
visited by EPA. Emissions for the selected facilities were quantified using
empirical TSDF emissions models; and human exposures associated with these
emissions were estimated using atmospheric dispersion modeling. Ranges of
cancer risks were then calculated for the visited plants using a range of
cancer risk factors for TSDF pollutants.
Nationwide emissions and cancer incidence for organic emissions from TSDF
area sources were extrapolated from emissions and risks for the selected model
plants. Two methods were used for this estimation. One was based on average
impacts for various facility types at the model plants, and nationwide
facility counts developed in a survey by Westat, Inc.^ The second method used
average impacts for model plants in various SIC groups, and plant counts from
the Westat survey. Table 1-1 presents nationwide emissions and cancer
incidence ranges aggregated by facility type, while Table 1-2 gives nationwide
emissions and incidence aggregated by the SIC code of the TSDF. Maximum
individual risk is also presented in the tables for various SIC codes and TSDF
operation types. The maximum individual risk is the cancer risk over a
70-year lifetime for an individual exposed at the point of maximum ambient
impact for the emission source.
As illustrated by the SIC code aggregation results given in Table 1-2,
emissions and incidences could not be determined for TSDFs handling fabricated
metal production waste (SIC 34) or electric equipment manufacturing waste
(SIC 36). Organic emissions from each of these categories are probably
significant because of due to use of solvents in cleaning parts made in both
categories. Also, because of the urban locations of such facilities, risks
and incidences associated with these emissions may be substantial. However,
it is known that the model plant sample chosen to represent the industry is
somewhat biased toward large, complex facilities, especially in the case of
those facilities not classified by SIC. Thus, the impact estimates for the
unspecified category in Table 1-2 may be overestimates.
-------
TABLE 1-1. NATIONWIDE EMISSION AND RISK ESTIMATES AGGREGATED BY SOURCE TYPE
Number of Maximum
visited individual
plants in risk for
category visited plants8'**
Drum handling 22 3 E-7 to 3 E-5
Storage tanks 19 1 E-6 to 1 E-4
Open treatment tanks and
surface impoundment 20 4 E-5 to 4 E-3
Land treatment 4 6 E-6 to 6 E-4
Landfills 5 1 E-6 to 1 E-4
Injection wells 1 4 E-9 to 4 E-7
Incineration
(area sources) 10 5 E-8 to 5 E-6
Distillation and
other recycling 13 9 E-7 to 9 E-5
Number in
category
nation-
wide
3577
1428
1687
70
199
87
240
392
TOTALSd
National
total
emissions
(Gg/yr)
10
20
800
100
20
0.05
0.2
5
960
estimate
total
annual
incidence8 »c
0.04
0.07
0.4
0.1
0.004
0.00002
0.0002
0.03
0.6
to 4
to 7
to 40
to 10
to 0.4
to 0.002
to 0.02
to 3
to 64
a Based on a unit risk factor range of 2 E-7 to 2 E-5 (probability of cancer incidence for
exposure to 1 ug/nr* over a 70 year lifetime).
b Maximum individual risk is the lifetime cancer risk to an individual exposed over a 70 year
lifetime to the highest ambient concentration outside the plant boundary.
c Annual incidence is the total cancer incidence for individuals living in the neighborhood
of the facilities under study.
-------
TABLE 1-2. NATIONWIDE EMISSION AND RISK ESTIMATES AGGREGATED BY SIC CODES
Number of Maximum Number in
visited individual category
plants in risk for nation-
category visited plants8*^ wide
National estimate
total total
emissions annual
(Gg/yr) incidence3*0
Chemicals and allied
products (SIC 28)
Fabricated metal
products (SIC 34)
Electrical equipment
(SIC 36)
16
0
Other metal-related products
(SIC 33,35,37) 3
All other manufacturing
(SIC 20-27,29-32,38-39) 6
Not otherwise specified
23
1 E-6 to 1 E-4
1 E-6 to 1 E-4
2 E-6 to 2 E-4
4 E-5 to 4 E-3
1249
547
540
804
878
800
TOTALS
50
40
200
800
1100
to 11
1 to 10
1 to 9
1 to 70
1 to 100
a Based on a unit risk factor range of 2 E-7 to 2 E-5 (probability of cancer incidence for
exposure to 1 ug/m^ over a 70 year lifetime).
k Maximum individual risk is the lifetime cancer risk to an individual exposed over a 70 year
lifetime to the highest ambient concentration outside the plant boundary.
c Annual incidence is the total cancer incidence for individuals living in^the neighborhood
of the facilities under study.
-------
The results of the nationwide analysis of emissions and risks from TSDF
area sources are subject to a number of limitations. These result from
uncertainties in the input data, and from simplifying assumptions made as part
of the general methodology. Sources of uncertainty in inputs and in the study
methodology include:
o the use of 39 visited plants to represent the industry as a whole;
o the use of models to estimate emissions and population exposures;
o uncertainties associated with development of pollutant risk factors;
o the unavailability of data or models to calculate emissions from some
source types;
o the use of default inputs where model inputs were not available; and
o the use of literature sources for impoundment and open tank waste
concentrations.
Appendix F provides a comparison of the national emission and incidence
estimates presented in Table 1-1 with estimates based on mass balance
calculations for the visited facilities. These mass balance calculations were
made only for source types where such a method was applicable (i.e.,
impoundments, land treatment and landfills). Results were extrapolated and
incidences were computed as in Tables 1-1 and 1-2, although the number of
facilities used in the analysis was smaller due to the lack of waste
throughput information for some facilities.
-------
References for Section 1
1. Dietz, S., M. Emmet, R. DiGaetano, D. Tuttle, and C. Vincent (Westat,
Inc.). National Survey of Hazardous Waste Generators and Treatment,
Storage and Disposal Facilities Regulated Under RCRA in 1981. Prepared
for U.S. Environmental Protection Agency: Office of Solid Waste.
Washington, DC. April 1984. 318p.
-------
2. INTRODUCTION
Air emissions from hazardous waste treatment, storage, and disposal
facilities (TSDFs) recently have received a great deal of attention. The
U. S. Environmental Protection Agency (EPA) Offices of Solid Waste, and
Research and Development have conducted studies of potential sources of air
emissions from hazardous waste TSDFs, and potential emission controls. The
EPA Office of Air Quality Planning and Standards recently received primary
responsibility for assessing emissions from TSDF area sources, and developing
standards for these sources as appropriate.
This report presents a preliminary assessment of TSDF area sources and of
the risks to the general public associated with these emissions. The types of
potential emission sources which were studied are as follows:
o drum storage and handling;
o tanker unloading;
o storage tanks;
o treatment tanks;
o surface impoundments for treatment, storage, and disposal;
o land treatment facilities;
o landfills;
o area sources associated with incinerators;
o deep well injection; and
o distillation and other recovery operations.
Nationwide-environmental impacts were estimated by extrapolating from
impacts calculated for 39 TSDFs visited by EPA and its contractors. The 39
visited facilities are listed in Table 2-1. Section 3 of this report
describes the techniques used to estimate emissions from the visited TSDFs.
Techniques used to estimate risks from the model facilities are described in
Section 4. Section 5 discusses the extrapolation techniques used to estimate
nationwide emissions and risks, and presents the results of the nationwide
assessment.
-------
TABLE 2-1. LIST OF VISITED FACILITIES
1. ABCO Industries, Inc.
2. ARCO-Cherry Point Refinery
3. Allied Corporation Fibers and Plastics Co.
4. American Cyanamid Co.
5. American Recovery Co., Inc.
6. Amoco Chemical Corporation
7. Celanese Chemical Corporation
8. Chem-Security Systems, Inc.
9. E.I. DuPont deNemours, Chamber Works
10. El Paso Products
11. Environmental Enterprises, Inc.
12. Environmental Protection Corporation
13. Fike Chemicals, Inc.
14. Fondessy/Aces
15. General Electric
16. Hamilton Standard Div., United Technologies
17. IT Corporation
18. Liberty Solvents and Chemical Company, Inc.
19. Marisol, Inc.
20. Metropolitan Sewer District of Cincinnati
21. Mobil Joliet Refining Corporation
22. Morflex Chemical (formerly Pfizer)
23. Occidental Chemical Corporation, Durez Resins
24. Rollins Environmental Services, Inc.
25. S and W Waste, Inc.
26. SCA Chemical Services, Inc.
(also listed as Earthline Co.)
27. SCA Services
28. Seaboard Chemical Corporation
29. Solvents Recovery Services
30. Southern Coating Co.
31. Tektronix, Inc.
32. The Atchison, Topeka and Santa Fe Railway Co.
33. U.S. Pollution Control, Inc.
Lone Mountain Facility
34. Union Carbide Agricultural Products Co.
35. Union Carbide Corp.
36. Union Carbide Agricultural
Products
37. CECOS International
38. Chem-Waste Management
39. Gulf Coast Waste Disposal Authority
Roebuck, SC
Ferndale, WA
Philadelphia, PA
Wallingford, CT
Baltimore, MD
Alvin, TX
Bay City, TX
Arlington, OR
Deepwater, TX
Odessa, TX
Cincinnati, OH
Bakersfield, CA
Nitro, WV
Oregon, OH
Lynn, MA
Windsor Locks, CT
Rio Vista, CA
Twinsburg, OH
Middlesex, NJ
Cincinnati, OH
Joliet, IL
Greensboro, NC
N. Tonawanda, NY
Baton Rouge, LA
South Kearny, NJ
Newark, NJ
Pinewood, SC
Jamestown, NC
Linden, NJ
Sumter, SC
Beaverton, OR
Somerville, TX
Waynoka, OK
Woodbine, GA
S. Charleston, WV
Institute, WV
Niagara Falls, NY
Kettleman City, CA
LaMarque, TX
-------
3. CALCULATION OF EMISSIONS
Emissions from the 39 selected TSDFs were calculated using empirical
emission models and emission factors. Appendix A lists estimated emissions
for individual emission sources at the 39 selected TSDFs. Because many of the
selected TSDFs-made blanket claims of confidentiality, plant names are not
given in the appendix. The emission source types, and the emission factors
and models used to estimate emissions, are discussed below. Specific
equations used in the emission models are given in Appendix B. The second and
third portions of this section discuss sources of input data for the models
and criteria used to select between different input data sources. Where
possible, input data for the models were based on site visit observations or
on published data; however, it was sometimes necessary to use default
parameters. Default parameters used in the models are listed in Appendix C,
while Appendix D gives sample emission calculations.
3.1 EMISSION MODELS
3.1.1 Drum Handling
Drum handling losses constitute emissions from drum filling (working
losses) and fugitive losses during drum storage. Filling losses were computed
after the method presented by Engineering-Science.1 Fugitive loss factors
were condensed into two categories in the E-S report: equipment leakage and
spillage. For equipment leakage, an emission factor of 0.0017 Ibs./drum was
used. For drum spillage it was estimated that 50 gallons per 100,000 gallons
stored are spilled, and that all of the spilled material is allowed to
evaporate. These factors are considered to represent a conservative measure
of loss.l
3.1.2 Tanker Unloading
The unloading of volatile organics from tank trucks was quantified using
two emission factors presented by Monsanto Research Corporation: one for
unloading (0.36 g/kg handled) and one for spillage (0.095 g/kg handled) .2
-------
3.1.3 Tank Storage
Emissions from storage tanks containing organic liquids are estimated
using equations presented in AP-42.3 For the tank designs of interest, fixed
roof and floating roof, there are four primary sources of emissions: fixed
roof breathing losses, fixed roof working losses, floating roof standing
storage losses, and floating roof withdrawal losses.
Emissions from fixed roof tanks are the result of breathing and working
losses. Breathing losses occur when the vapor in the tank expands due to
temperature and/or barometric pressure changes. Expansion of the vapor in the
tank results in the expulsion of organic vapors contained within the tank's
vapor space through the pressure/vacuum relief valve.
Fixed roof working losses are the result of the combined filling and
emptying losses. Losses due to filling result when an increase in liquid
level increases the vapor space pressure within the tank. As the tank
pressure approaches the pressure/vacuum valve relief pressure the valve
cracks. The valve allows the vapors to be released thus reducing the tank
pressure. Emptying losses occur as air is drawn into the tank during liquid
removal. The air becomes saturated with organic vapor and expands releasing
the vapor through the pressure/vacuum relief valves.
Standing storage and withdrawal losses are the primary sources of
emissions from external and internal floating roof tanks. Standing storage
losses result as air flows across the top of a floating roof and the tank
wall. Withdrawal losses occur when organic liquid clings to the tank wall and
is exposed to the atmosphere as the tank is emptied. Again, the emission rate
will vary depending on the seal type used.
The equations presented in AP-42 for estimating emissions from storage
tanks containing organic liquids were developed for tanks containing pure
organic compounds. For tanks containing dilute mixtures of organics in heavy
oil or water the equations are still valid; however the total pressure in the
tank and the average vapor molecular weight must be calculated for a mixture
-------
of organics. In calculating emissions from tanks containing mixtures, it was
assumed that the tank vapor space was at equilibrium with the stored liquid.
The stored liquid was also assumed to be an ideal solution.
3.1.4 Treatment Tanks
The emission factors for API separators were based on AP-42 emission
factors for oil/water wastes. The value given is 0.6 kg of VOC emitted per
1000 liters wastewater throughput.4 Techniques used to estimate emissions for
open aerated and non-aerated treatment tanks are the same as those used for
aerated and non-aerated surface impoundments (below).
3.1.5 , Surface Impoundments
The emission rate of a component from the liquid into the gas phase is
given by the following equation:6
Q = K A (X£ - X*) MWi
where:
Q = emission rate (mass/time)
K = overall mass transfer coefficient (mole/length^ time)
A = surface area of impoundment (length^)
X^ = mole fraction of i in liquid phase
X. = mole fraction of i in gas phase
W^ = molecular weight of component i ( mass/mole)
It is reasonable to assume that the concentration of i in the gas phase is
negligible compared to that in the liquid phase. Thus, the above equation
becomes:
Q = K A Xi MW£
-------
The overall mass transfer coefficient, K is a combination of the
individual mass transfer coefficients on the liquid and gas side. The
following equation describes the relationship between the overall and
individual mass transfer coefficients:
1 1
— = +
K kL Keq kg
k^ = individual liquid phase mass transfer coefficient
kg = individual gas phase mass transfer coefficient
Keq = vapor liquid equilibrium constant
The vapor liquid equilibrium constant is the ratio of the molar vapor
concentration to the molar liquid concentration at equilibrium for a
particular compound. Methods have been proposed by a number of authors
concerning the calculation of individual liquid and gas phase mass transfer
coefficients (Table B-l). These methods constitute emission models for
surface impoundments. There are differences in basic assumptions used to
develop each model.
Quiescent Impoundments. A correlation proposed by Mackay and Matsugu (1973)
was used to calculate the gas phase resistance (k_) over a smooth liquid
surface.5 This coefficient is a function of windspeed and effective diameter
of the liquid surface.
Liquid phase mass transfer coefficients were based on one of three mass
transfer theories presented in the literature:
- stagnant film theory,
penetration or surface renewal theories, and
boundary layer theory.
Stagnant film theory identifies kj^ as a linear function of diffusivity.
Penetration or surface renewal theories identify the dependency to be to the
0.5 power. Finally, the boundary layer theory states the dependency to be to
-------
the two-third power. The exact dependence of UL on diffusivity remains to be
established.
In 1978 Cohen, et al., presented a liquid phase mass transfer coefficient
developed from laboratory wind wave tank studies.5 Their equation was
developed for benzene and toluene and indicated that UL is dependent primarily
on windspeed for velocities between 3 and 10 meters per second. Below 3 m/s,
UL was influenced by subsurface agitation, while at windspeeds above 10 m/s,
UL increases due to the presence of spray, bubble entrainment and white
capping. Their equation, developed from stagnant film theory, suggested a
linear dependency of !CL to diffusivity.
In 1964 Owens,et al, presented an equation based on reaeration stream
studies.5 Unlike the wind velocity dependency analyzed by Cohen, Owens'
correlations show k^ to be dependent on water velocity and depth. However,
like the equation developed by Cohen, the Owens model also suggests a linear
dependency of k^ to diffusivity.
Turbulent Impoundments. Estimating emission i-ates from aerated impoundments
is more complex in the calculation of the overall mass transfer coefficient.
In reality, only a small fraction of any impoundment is effectively aerated.
As a result, both quiescent and turbulent mass transfer coefficients must be
used. The first step involves the calculation of an overall mass transfer
coefficient in the quiescent area of the impoundment from individual mass
transfer coefficients as follows:5
KQ -
Likewise, an overall mass transfer coefficient on the turbulent side is
also calculated:^
T
" 1? * ~T~?
KL e
-------
From these two overall mass transfer coefficients on the quiescent and
turbulent side an actual overall mass transfer coefficient is obtained on a
per area basis as follows:
AT RT + AQ KQ
JC = ______________
AT + AQ
where A^ + AQ = Total area of impoundment.
The previous section addressed the models used to calculate !CL and kg for
the quiescent portion of the impoundment. The literature provides very little
information on the calculation of individual turbulent mass transfer
coefficients. Nonetheless, there are two correlations found, one on the gas
and one on the liquid side, that correlate kg and kj, to aeration parameters
(Table B-2).
In 1977 Reinhardt developed an empirical expression to approximate the
gas-phase mass transfer coefficient under mechanically aerated conditions.
The correlation indicates that the dependency of kg to diffusivity is linear.5
In the literature, the only relationship found to calculate the liquid
phase mass transfer coefficient under agitated conditions was developed by
Thibodeaux in 1978.5 Considerable research has been conducted on scale models
with agitated water surfaces to determine the absorption rate of oxygen.
These oxygen absorption rates, being liquid phase controlled, have been
transformed to yield a liquid phase mass transfer coefficient under agitated
conditions.
Disposal Impoundments. The above models for impoundments do not apply to
drying impoundments for oily sludges. For these cases, the Hartley model was
used to determine emissions.6 The Hartley model was developed to determine
the evaporative loss of pure volatile compounds. The model assumes that the
rate of mass transfer is controlled by resistance in the gas phase and is
proportional to the saturated vapor concentration. The liquid phase
resistance, which plays an important role for multicomponent liquid mixtures
-------
containing volatile compounds, is completely ignored in the model development.
The Hartley model calculates the flux rate of a component with respect to the
flux rate of a known reference compound, usually water. Although the
Hartley model represents a simple method of calculating air emissions it has
several drawbacks.
For the case where a surface impoundment is allowed to dry, while
occasionally being tilled, the land treatment model developed by Thibodeaux
and Hwang was used to determine emissions.6 Because disposal operations are
generally considered batch operations, emission rate estimates based on mass
balance calculations have been favored where the necessary information has
been provided. Having reasonable data about concentration and throughput
provides upper limit emission estimates assuming no absorption on soil
surfaces and no biological activity.
3.1.6 Land Treatment
The Thibodeaux-Hwang model estimates emissions from land-treatment
facilities (Table B-3).6 The model assumes that the emission rate is
controlled by diffusion of vapor through the air-filled pores of the soil.
With the assumptions that the soil column is isothermal, that no vertical
movement of waste occurs by capillary action, no adsorption of material occurs
on soil particles and no biochemical oxidation occurs, the model describes the
evaporation rate of a chemical species from a soil matrix. The model, as it
appears in the literature, gives an instantaneous emission rate. Integration
of the model with respect to time gives an average emission rate over a
specified time range.
3 .1 .7 Covered Landfills
The Farmer model was used to estimate emissions from covered landfills
(Table B-4). This model estimates emissions for cases where diffusion through
air-filled pore spaces in the landfill cover is the rate limiting step. The
model assumes: no degradation occurs from biological activity, no adsorption
of the compound or transport in moving water occurs, and there is
-------
no landfill gas production.6 The vapor concentration of a component in the
waste is the driving force for diffusion. The addition of water to the soil
will reduce the air-filled porosity, thus reducing the vapor flux from the
soil surface because the diffusion rate through a liquid is generally several
orders of magnitude less than diffusion through air.
3.1.7 Recycling. Deep Well Injection, and Incineration
Emissions from vents, pump seals, valves, flanges, relief valves, and
other fugitive emission sources were calculated using published emission
factors as follows:
o for distillation vents, 1.65 g/kg throughput;7
o for pumps in light liquid service, 98.8 g/hr-pump (assuming
2 seals/pump);°
o for pumps in heavy liquid service, 42.8 g/hr-pump (assuming
2 seals/pump);^
o for valves in liquid service: light liquid, 7.1 g/hr-valve;^
heavy liquid, 0.23 g/hr-valv.;8
o for valves in gas service, 5.6 g/hr-valve;^
o for relief valves, 104 g/hr-valve;^
o for flanges, 0.83 g/hr-flange;^
o for open-ended lines, 1.7 g/hr-line;8 and
o for sampling connections, 15 g/hr-connection.8
3.2 SOURCES OF FACILITY SPECIFICATIONS
Emissions from each TSDF and each operation within that facility were
computed using the models and emission factors presented in the previous
section. This section describes the process whereby input values were derived
for the estimation process. These values were obtained from site visits, or
from various literature sources when site data was unavailable.
-------
3.2.1 Site Visits
Site visits to the 39 model TSDFs were made by EPA and its contractors.
These visits assessed the processes involved in hazardous waste treatment,
storage and disposal through review with plant personnel and inspection of the
processes and -operations. These visits provided the primary source of data
for subsequent emission estimation, including physical specification of
equipment and areas, operating practices and emission controls. Whenever
available, values given by or derived from these visits were used
preferentially to any other source.
3.2.2 Default Parameters
Where site data for certain parameters were unavailable or inadequate
from the site visit reports, default values were used as applicable. These
fell mainly into the area of physical specifications of equipment or
operations (e.g., aerator specifications, impoundment depth). Generally,
every attempt was made to accurately portray the plant's operation. Default
parameters, values and refei^nces are provided in Appendix C.
3.3 SOURCES OF PROCESS STREAM COMPOSITION
Process stream composition is among the most important parameters for
estimating emissions from all types of TSDFs. Process stream compositions
were obtained or estimated using five basic sources. In order of preference,
they are as follows:
o analyses given or estimated by plant personnel and
preliminary process testing results;9-12
o waste codes obtained during visits with generic waste code
analyses;
o waste codes obtained from Part A data base with generic
waste code analyses;
o W-E-T model waste characterization;
-------
o Effluent Guidelines Document waste characterization for
industry type; and
o default waste concentrations.
3.3.1 Plant Estimates
Wherever possible compound specific waste composition data were based on
site visit reports. Analyses estimated or given by plant personnel were
assumed to be the best available wastestream information.
3 .3.2 Preliminary Process Testing Results
For a limited number of sites with impoundments and open-top tanks,
results from EPA sampling programs were available. These programs identified
major organic constituents and estimated total organic carbon in an
impoundment or open tank.
3.3.3 Waste Codes Obtained During Visits
Frequently EPA Waste Codes constituted the most detailed listing of
wastestream specification available. For some EPA Waste Codes, typical
compositions are available in the literature. References 13 to 15 provided
proportional compositions necessary to estimate emissions.
3.3.4 Waste Code Obtained from Part A Data Base
The computerized RCRA Part A data base supplies the 15 largest volume
waste codes and volumes by treatment type for each RCRA facility.1^ These
data were used in the same manner as waste codes obtained during visits
(above).
3.3.5 W-E-T Model Waste Characterization
The W-E-T Model data base for hazardous wastes^ gives waste
characterization for industry types or waste codes. When this source was
-------
used, the EPA waste code associated with the process wastestream was located
in the W-E-T data base. Then the organic composition of the model wastestream
was substituted for the process wastestream.
3.3.6 Effluent Guidelines Documents
Effluent Guidelines Documents provide wastestream compositional data
solely for priority pollutants by industry type. This source was used for
acrylic polymer production wastestreams.18
3 .3 .7 Default Wastestream Compositions
Where wastestream data were unavailable for drums and storage tanks,
their contents were assumed to be pure trichloroethylene. Trichloroethylene
was selected as typical of waste code F001, which was selected based on the
distribution of waste handled in tanks and drums from the Westat data
base.1 »17
For surface impoundments and open top treatment tanks, where process
stream composition data were not available from any of the above sources,
emissions were based on model wastes. The total organic concentrations for
the model wastes were selected based on preliminary analytical results for a
number of impoundments and treatment tanks sampled by EPA. Preliminary
analytical results for impoundments and open tanks are summarized in
Table 3-1. The average concentration from Table 3-1, 47 ppm, was used in
emission calculations where no process stream composition data were available.
3.4 SOURCES OF POLLUTANT PROPERTIES DATA
Most of the emission models discussed in Section 3.1 require physical
parameters for pollutants. Parameters required include Henry's Law constants,
diffusivities in air and water, vapor pressures, densities, and molecular
weights. Generally such parameters were obtained from measurements or
estimates given in literature sources.19-20 However, gaps in physical
properties data are common for many chemicals of environmental concern.
-------
TABLE 3-1. SUMMARY OF ANALYTICAL RESULTS FOR IMPOUNDMENTS AND OPEN TANKS
Facility
TNMHG
(ppb)
Facility
TNMHC
(ppb)
Treatment and Evaporation
Final clarifier
Waste water discharge
Final clarifiers
Secondary clarifier
Primary aeration basin
Aeration basin
Biomass basin
Bio basin; aeration tank
Evaporation pond
Evaporation pond
Evaporation pond
Evaporation pond
Evaporation pond
Evaporation pond
Evaporation pond
Evaporation pond
Evaporation pond
Reducing lagoon
Spray evaporation pond
Evaporation pond
Evaporation pond
Evaporation pond
Evaporation pond
Evaporation pond
308
37100
620
2882
370
4120
690
2812
1810
2250
2610
4760
6560
8190
11000
15600
18300
3250
8399
4500
6500
16000
18800
30000
Receiving and Holding
Primary clarifiers 4020
Primary clarifier 35560
Primary clarifier overflow 195000
Oxidizing lagoon 654231
Energency holding pond 30470
Sludge holding pond 2480
Leachate collection pond 5276
Equalization Pond 41640
Holding pond 1050
Liquid waste holding pond 181000
Liquid waste holding pond 14700
Receiving pond 269000
Neutralization pond 38700
Equalization basin 6890
Lagoon recycle from filters 79300
Inlet canal 40920
Neutralization tanks 115280
AVERAGE 100913
AVERAGE
-------
Standard methods were used where necessary to estimate physical properties.
For Henry's Law constant, Almgren's equation was used.21 Diffusivities in
water were estimated using the Hayduk and Laudie method; and diffusivities in
air were estimated using the Wilke and Lee method.22
For model-wastes discussed in Section 3.3.7, the physical parameters used
were averages of parameters for chemicals which were estimated in previous
ranking studies to represent large fractions of TSDF emissions.19,23 Chemicals
ranking among the top ten in these studies are listed in Section 4.2. The
Henry's Law constant used for model wastes was 0.01 atm-m^/mol. Diffusivities
were 0.07 cm2/sec for air, and 10~5cm2/sec for water. The molecular weight
used was 100 grams per mole.
-------
References for Section 3
1. Engineering-Science. National Air Emissions from Tank and Container
Storage and Handling Operations at Hazardous Waste Treatment, Storage and
Disposal Facilities (Draft). Prepared for U.S. Environmental Protection
Agency: Office of Air Quality Planning and Standards. Research Triangle
Park, NC. September 1984. p. 5-9.
2. Tierney, D.R. and T.W. Hughes (Monsanto Research Corporation). Source
Assessment: Reclaiming of Waste Solvents, State-of-the-Art. Prepared for
U.S. Environmental Protection Agency. Washington, DC. August 1977.
3. U.S. Environmental Protection Agency: Office of Air Quality Planning and
Standards. Compilation of Air Pollution Emission Factors, Third Edition.
AP-42. Research Triangle Park, NC. August 1977. Section 4.3.
4. Reference 3. Section 9.1.
5. Hwang, Seong T. Toxic Emissions From Land Disposal Facilities.
Environmental Progress. ^:46-52. February 1982.
6. Farino, William et al. (GCA Corporation). Evaluation and Selection of
Models for Estimating Air Emissions From Hazardous Waste Treatment,
Storage and Disposal Facilities (Draft). Prepared for U.S. Environmental
Protection Agency. Office of Solid Waste, Land Disposal Branch. May
1983. p. 5-1 to 6-5.
7. U.S. Environmental Protection Agency: Office of Air Quality Planning and
Standards. Compilation of Air Pollutant Emission Factors, Third Edition.
AP-42. Research Triangle Park, NC. April 1977. Section 4.7-3.
8. U.S. Environmental Protection Agency: Office of Air Quality Planning and
Standards. Control of Volatile Organic Compound Leaks from Synthetic
Organic Chemical and Polymer Manufacturing Equipment. EPA-450/3-83-006.
Research Triangle Park, NC. March 1984. p. 2-21.
9. Engineering-Science. Preliminary Process Testing Results for Surface
Impoundments and Open-Top Tanks at Six Hazardous Waste TSDFs. (Draft)
Prepared for U.S. Environmental Protection Agency: Office of Air Quality
Planning and Standards. Fairfax, VA. September 1984. 16p.
-------
10. Radian Corporation. Summary of Process Data and Test Results - Upjohn
(Draft). Prepared for U.S. Environmental Protection Agency: Industrial
Environmental Research Laboratory. Austin, TX. October 1983. 13p.
11. Radian Corporation. Evaluation of Air Emissions from Hazardous Waste
Treatment, Storage and Disposal Facilities in Support of RCRA Air Emission
Regulatory Impact Analysis (RIA): Sites 5, 6, 8, and 9. Prepared for
U.S. Environmental Protection Agency: Industrial Environmental Research
Laboratory. Cincinnati, OH. 1984.
12. Radian Corporation. Hazardous Waste Treatment, Storage, and Disposal
Facility Area Sources—VOC Air Emissions: Emission Test Reports Chemical
Waste Management, Inc. Kettleman Hills Facility, and IT Corporation
Benicia Facility (Draft). Prepared for U.S. Environmental Protection
Agency: Office of Air Quality Planning and Standards. Research Triangle
Park, NC. 1985.
13. Industrial Economics, Incorporated. Interim Report on Hazardous Waste
Incineration Risk Analysis. Prepared for U.S. Environmental Protection
Agency: Office of Solid Waste. Washington, DC. August 1982.
14. U.S. Environmental Protection Agency: Office of Solid Waste. Background
Document, Resource Conservation and Recovery Act: Subtitle C -
Identification and Listing of Hazardous Waste. PB81-190035. November
1980. 853p.
15. Environ Corporation. Characterization of Selected Waste Streams Listed in
40 CFR Section 261 Draft Profiles. Prepared for U.S. Environmental
Protection Agency: Waste Identification Branch. Washington, DC. August
1984.
16. SCS Engineers. W-E-T Model Hazardous Waste Data Base: Final Draft.
Prepared fot U.S. Environmental Protection Agency: Office of Solid
Waste. Publication pending. Section 2.2.
17. Dietz, S., M. Emmet, R. DiGaetano, D. Tuttle, and C. Vincent (Westat,
Inc.). National Survey of Hazardous Waste Generators and Treatment,
Storage and Disposal Facilities Regulated Under RCRA in 1981. Prepared
for U.S. Environmental Protection Agency: Office of Solid Waste.
Washington, DC. April 1984. 318p.
18. U.S. Environmental Protection Agency: Office of Water Regulations and
Standards. Development Document for Proposed Effluent Limitations
Guidelines and New Source Performance Standards for the Organic Chemicals
and Plastics and Synthetic Fibers Industry: Volume II (BAT). Washington,
DC. February 1983. Table V-16 and VI-2.
19. Nunno, Thomas, et al. (GCA Corporation) Assessment of Air Emissions From
Hazardous Waste Ranking. Prepared for U.S. Environmental Protection
Agency Office of Solid Waste Land Disposal Branch. Washington, DC.
September 1983. p.97-102.
-------
20. Versar Inc. Physical-Chemical Properties and Categorization of RCRA
Wastes According to Volatility. Prepared for U.S. Environmental
Protection Agency Mission Standards and Engineering Division Office of Air
Quality Planning and Standards. Research Triangle Park, NC. September
28, 1984.
21. Mackay, D., W.Y. Shiu, A. Bobra, J. Billington, E. Chau, A. Yeu, C. Ng,
and F. Szeto. Volatilization of Organic Pollutants from Water.
Environmental Research Laboratory. Athens, GA. Publication No.
EPA-600/3-82-019. April 1982. p.42.
22. Lyman, W.J., W.F. Reehl, and D.H. Rosenblatt. Handbook of Chemical
Property Estimation Methods. New York, McGraw-Hill. 1982. Chapter 17.
23. Spivey, J.J., C.C. Allen, D.A. Green, J.P. Wood, and R.L. Stallings
(Research Triangle Institute). Preliminary Assessment of Hazardous Waste
Pretreatment as an Air Pollution Control Technique. Prepared for U.S.
Environmental Protection Agency: Industrial Environmental Research
Laboratory. Cincinnati, OH. March 15, 1984.
-------
4. CALCULATION OF CANCER RISK AND INCIDENCE
The maximum individual cancer risk and the total annual cancer incidence
were estimated for each of the visited facilities. The maximum individual
risk is the risk of cancer over a 70-year lifetime for an individual exposed
to the maximum, pollutant concentration outside the plant boundary. The total
annual incidence is the aggregate cancer risk for all individuals living in
the vicinity of the plant.
In order to calculate maximum individual risk and total annual incidence,
it is first necessary to calculate the maximum off-site pollutant
concentration, and the total population exposure for each plant. Maximum
risks and total annual incidences can then be calculated using the risk factor
for the pollutant of concern.
4.1 CALCULATIONS OF PUBLIC EXPOSURES
4.1.1 General
The EPA's Human Exposure Model (HEM) was used to calculate public exposures to
ambient air concentrations of pollutants emitted from stationary sources. The
HEM contains (1) an atmospheric dispersion model, with included meteorological
data, and (2) a population distribution estimate based on Bureau of Census
data. The only input data needed to operate this model are source data, e.g.,
plant location, height of the emission release point, and temperature of the
off-gases. Based on the source data, the model estimates the magnitude and
distribution of ambient air concentrations of the pollutant in the vicinity of
the source. The model is programmed to estimate these concentrations within a
radial distance of 50 km (30.8 miles) from the source. If other radial
distances are preferred, an over-ride feature allows the user to select the
distance desired. The selection of 50 km (30.8 miles) as the programmed
distance is based on modelling considerations, not on health effects criteria
or EPA policy. The dispersion model contained in HEM is felt to be reasonably
accurate within 50 kin (30.8 miles).
-------
4.1 .2 Pollutant Concentrations Near a Source
The dispersion model within the HEM is a gaussian diffusion model that
uses the same basic dispersion algorithm as the EPA's Climatological
Dispersion Model. The algorithm has been simplified to improve computational
efficiency. The algorithm is evaluated for a representative set of input
values as well as actual plant data, and the concentrations input into the
exposure algorithm are arrived at by interpolation. Stability array (STAR)
summaries are the principal meteorological input to the HEM dispersion model.
The STAR data are standard climatological frequency-of-occurrence summaries
formulated for use in U.S. EPA models and are available for major U.S.
meteorological monitoring sites from the National Climatic Center, Asheville,
NC. A STAR summary is a joint frequency-of-occurrence of wind speed,
atmospheric stability, and wind direction, classified according to Pasquill's
categories. The STAR summaries in HEM usually reflect 5 years of
meteorological data for each of 314 sites nationwide. The model produces
polar coordinate receptor grid points consisting of 10 downwind distances
located along each of 16 radials which represent wind directions.
Concentrations ara estimated by the dispersion model for each of the 160
receptors located on this grid. The radials are separated by 22.5-degree
intervals beginning with 0.0 degrees and proceeding clockwise to 337.5
degrees. The 10 downwind distances for each radial are 0.2, 0.5, 1.0, 2.0,
5.0, 10.0, 20.0, 30.0, 40.0 and 50.0 kilometers. The center of the receptor
grid for each plant is assumed to be the plant center.
4.1.3 Population Living Near an Emission Source
To estimate the number and distribution of people residing within 50 km
(30.8 miles) of each plant, the model contains for 1980, the Master Area
Reference File (MARF) from the U.S. Census Bureau. This data base is broken
down into enumeration district/block group (ED/BG) values. It contains the
population centroid coordinates (latitude and longitude) and the 1980
population of each ED/BG in the United States (50 States plus the District of
Columbia). The HEM identifies the population around each plant by using the
-------
geographical coordinates of the plant. The HEM identifies, selects, and
stores for later use those ED/BGs with coordinates falling within 50 km
(30.8 miles) of plant center.
4.1.4 Population Exposure Determinations
The HEM uses the estimated ground level concentrations of a pollutant
together with population data to calculate population exposure. For each of
160 receptors located around a plant, the concentration of the pollutant and
the number of people estimated by the HEM to be exposed to that particular
concentration are identified. The HEM multiplies these two numbers to produce
exposure estimates and sums these products for each plant.
A two-level scheme has been adopted in order to pair concentrations and
populations prior to the computation of exposure. The two level approach is
used because the concentrations are defined on a radius-azimuth (polar) grid
pattern with non-uniform spacing. At small radii, the grid cells are usually
smaller than ED/BGs; at large radii, the grid cells are usually larger than
ED/BGs. "he area surrounding the source is divided into two regions, and each
ED/BG is classified by the region in which its centroid lies. Population
exposure is calculated differently for the ED/BGs located within each region.
For ED/BG centroids located between 0.1 km (0.06 miles) and 3.5 km (2.2 miles)
from the emission source, populations are divided between neighboring
concentration grid points. There are 64 (4 x 16) polar grid points within
this range. Each grid point has a polar sector defined by two concentric arcs
and two wind direction radials. Each of these grid points and respective
concentrations are assigned to the nearest ED/BG centroid identified from
MED-X. Each ED/BG can be paired with one or many concentration points. The
population associated with the ED/BG centroid is then divided among all
concentration grid points assigned to it. The land area within each polar
sector is considered in the apportionment.
For population centroids between 3.5 km (2.2 miles) and 50 km
(30.8 miles) from the source, a concentration grid cell, the area
-------
approximating a rectangular shape bounded by four receptors, is much larger
than the area of a typical ED/BG. Since there is an approximate linear
relationship between the logarithm of concentration and the logarithm of
distance for receptors more than 2 km from the source, the entire population
of the ED/BG is assumed to be exposed to the concentration that is
logarithmically interpolated radially and arithmetically interpolated
azimuthally from the four receptors bounding the grid cell. Concentration
estimates for 96 (6 x 16) grid cell receptors at 10.0, 20.0, 30.0, 40.0, and
50.0 km from the source along each of 16 wind directions are used as reference
points for this interpolation.
In summary, two approaches are used to arrive at coincident
concentration/population data points. For the 64 concentration points within
3.5 km (2.2 miles) of the source, the pairing occurs at the polar grid points
using an apportionment of ED/BG population by land area. For the remaining
portions of the grid, pairing occurs at the ED/BG centroids themselves through
the use of log-log and linear interpolation. For a more detailed discussion
of the model used to estimate exposure, see reference 1.
4.2 RISK FACTORS
Maximum individual risk and annual incidence are calculated from
population exposures using the unit risk factor for the pollutant of concern.
The unit risk factor for a given pollutant is the probability of a cancer
incidence for an individual exposed to 1 yg/rn^ of the pollutant over a 70-year
lifetime. Thus, the maximum individual risk for a plant is given by the
product of (1) the unit risk factor for the pollutant emitted; and (2) the
maximum concentration of the pollutant to which an individual is exposed. The
annual incidence for a plant is given by the product of (1) the unit risk
factor divided by 70; and (2) the population exposure in units of
people-ug/m3 .
-------
Unit risk factors have been established only for a small percentage of
TSDF pollutants. For this reason, risks were not calculated for the 39
visited facilities on a pollutant-by-pollutant basis. Instead, for each
plant, total pollutant exposures were calculated, and the range of cancer
risks and incidences was calculated based on a range of unit factors. The
risk factor range for TSDF pollutants was based on two ranking studies of TSDF
pollutants.2>3 The top ten pollutants by emissions from the two ranking
studies are given in Table 4-1, along with their unit risk factors, where
avail able.^ These risk factors were developed based on animal studies and
occupational epidemiological studies. (Epidemiological studies have not been
conducted for TSDFs, but have been conducted for various manufacturing
industries emitting TSDF pollutants.) The techniques used to develop unit
risk factors are detailed in Appendix E.
The range of risk factors used for TSDFs was 2 X 10~7 to 2 X 10~5
(rounded to the nearest one significant figure). The upper end of the range
was based on the unit risk factor for carbon tetrachloride, the highest-risk
compound appearing on both rankings. The lower end of the range was based on
the unit risk factor for methylene chloride.
-------
TABLE 4-1. RISK FACTORS FOR THE TOP 10 CHEMICALS BY
EMISSIONS FROM GCA AND RTI RANKINGS
Ranking
Compound
Toluene
Methylene chloride
Trichloroethylene
Perchloroethylene
Cyanide
1,1, 1-trichloroethane
Benzene
Carbon tetrachloride
Acetonitrile
ftethanol
Xylene
1,2-dichloroethane
Formaldehyde
Ethyl acetate
Vinyl chloride
Acetone
Ethylene oxide
GCA
6
10
5
1
18
11
21
4
9
2
3
--
—
7
—
8
--
RTI
5
6
12
18
1
8
2
19
16
—
--
3
4
--
8
--
10
Unit
risk
/ug/a3
—
1.8E-7
4.1E-6
1.7E-6
--
--
6.9E-6
1.5E-5
--
--
--
7.0E-6
6.1E-6
--
--
--
3.6E-4
-------
References for Section 4
1. Systems -Application, Inc. Human Exposure to Atmospheric Concentration of
Selected Chemicals--Volumes 1 and 2. Prepared for the U.S. Environmental
Protection Agency. Research Triangle Park, North Carolina. EPA-2/250-1
and 2.
2. Breton, Marc, et al (GCA Corporation). Assessment of Air Emissions from
Hazardous Waste Treatment, Storage, and Disposal Facilities
(TSDFs)—Preliminary National Emissions Estimates. Prepared for U.S.
Environmental Protection Agency. Office of Solid Waste, Land Disposal
Branch. September 1983. pp. 103 to 104.
3. Spivey, J.J., et al (Radian). Preliminary Assessment of Hazardous Waste
Pretreatment as an Air Pollution Control Technique. Prepared for U.S.
Environmental Protection Agency Industrial Environmental Research
Laboratory. Cincinnati, OH. March 1984. pp. B-l to B-44.
4. Memorandum from Battye, W. , GCA Corporation, to Dusetzina, M., and
Hustvedt, K.C., EPA. October 11, 1984. 3p. Composite risk factor for
TSDF preliminary national risk assessment.
-------
5. NATIONWIDE EXTRAPOLATION
Nationwide emissions and cancer incidence for organic emissions from TSDF
area sources were estimated from emissions and risks for the visited plants.
Two methods were used for this estimation. Tables 5-1 and 5-2 present
nationwide emissions and incidences aggregated by source type for pollutant
unit risk factors of 10~5 an(j 10"?, respectively. Tables 5-3 and 5-4 give
nationwide emissions and incidences aggregated by the SIC code of the TSDF for
unit risk factors of 10~5 and 10"?. The unit risk factor for a given
pollutant is defined as the probability of cancer incidence for an individual
exposed to an average pollutant concentration of 1 yg/m^ over a 70-year
lifetime. In addition to the national estimates, these tables give average
emissions and incidences associated with the visited plants for each category
of source or SIC. Maximum individual risk is also given for visited
facilities and sources in the various SIC and source categories. The maximum
individual risk is the cancer risk over a 70-year lifetime for an individual
exposed at the point of maximum ambient impact for the emission source.
The results presented in Tables 5-1 through 5-4 are based on emissions,
risks, and annual incidences for the individual visited facilities, calculated
as described in Sections 3 and 4:
f c = **s [source and waste specifications, meteorological
data]
[source specifications, meteorological data,
population data, pollutant risk factor] • Es f c
Rs f c = HEM [source specifications, meteorological data,
pollutant risk factor] • Es f c
where Es f c = the emission estimate for all sources of type "s" at
facility "f" in SIC category "c";
^-s f c = t*ie cancer incidence estimated to result from the
emission, Es f c> in the neighboring population;
-------
Ui
I
TABLE 5-1. NATIONWIDE EMISSION AND RISK ESTIMATES AGGREGRATED BY SOURCE TYPE
FOR A UNIT RISK FACTOR OF 2 E-5 a
Number of For visited plants
visited Average Average Maximum
plants in emissions annual individual
category (Mg/yr) incidence0 risk"
Drum handling 22 31 E-3 3 E-5
Storage tanks 19 10 5 E-3 1 E-4
Open treatment tanks and
surface impoundment 20 300 3 E-2 4 E-3
Land treatment 4 1400 2 E-l 6 E-4
Landfills 5 100 2 E-3 1 E-4
Injection wells 1 12 E-5 4 E-7
Number in National estimate
category Total Total
nation- emissions annual
wide (Gg/yr) incidence0
3577 10 4
1428 20 7
1687 800 40
70 100 10
199 20 0.4
87 0.05 0.002
Incineration
(area sources)
Distillation and
other recycling
10
13
8 E-5
8 E-3
5 E-6
240
0.2
9 E-5 392
TOTALS
5
960
0.02
3
64
a The unit risk factor is the probability of cancer incidence for exposure to 1 ug/mj over a 70 year
lifetime.
k Maximum individual risk is the lifetime cancer risk to an individual exposed over a 70 year lifetime to
the highest ambient concentration outside the plant boundary.
c Annual incidence is the total cancer incidence for individuals living in the neighborhood of the facilities
under study.
-------
TABLE 5-2. NATIONWIDE EMISSION AND RISK ESTIMATES AGGREGRATED BY SOURCE TYPE
FOR A UNIT RISK FACTOR OF 2 E-7 a
Number of For visited plants
visited Average Average Maximum
plants in emissions annual individual
category (Mg/yr) incidence0 risk"
Drum handling 22
Storage tanks 19
Open treatment tanks and
surface impoundment 20
Land treatment 4
Landfills 5
Injection wells 1
3 1 E-5 3 E-7
10 5 E-5 1 E-6
300 3 E-4 4 E-5
1400 2 E-3 6 E-6
100 2 E-5 1 E-6
1 2 E-7 4 E-9
Number in
category
nation-
wide
3577
1428
1687
70
199
87
National estimate
Total Total
emissions annual
(Gg/yr) incidence0
10 0.04
20 0.07
800 0.4
100 0.1
20 0.004
0.05 0.00002
Incineration
(area sources)
Distillation and
other recycling
10
13
8 E-7
8 E-5
5 E-8
240
9 E-7 392
TOTALS
0.2 0.0002
5 0.03
960 0.64
a The unit risk factor is the probability of cancer incidence for exposure to 1 ug/mj over a 70 year
lifetime.
k Maximum individual risk is the lifetime cancer risk to an individual exposed over a 70 year lifetime to
the highest ambient concentration outside the plant boundary.
c Annual Incidence is the total cancer incidence for individuals living in the neighborhood of the facilities
under study.
-------
TABLE 5-3. NATIONWIDE EMISSION AND RISK ESTIMATES AGGREGRATED BY SIC CODES
FOR A UNIT RISK FACTOR OF 2 E-5 a
Number of For visited plants Number in National estimate
visited Average Average Maximum category Total Total
plants in emissions annual individual nation- emissions annual
category (Mg/yr) incidence0 risk'3 wide (Gg/yr) incidence0
Chemicals and allied
products (SIC 28) 16 40
Fabricated metal
products (SIC 24 0
Electrical equipment
(SIC 36) 0
Other metal-related products
(SIC 33,35,37) 3 50
All other manufacturing
(SIC 20-27,29-32,38-39) 6 200
Not otherwise specified 23 1000
9 E-3 1 E-4 1249 50 10
547
540
1 E-2 1 E-4 804 40 10
1 E-2 2 E-4 878 200 9
9 E-2 4 E-3 800 800 70
TOTALS d 1100 99
a The unit risk factor is the probability of cancer incidence for exposure to 1 ug/mj over a 70 year
lifetime.
k Maximum individual risk is the lifetime cancer risk to an individual exposed over a 70 year lifetime to
the highest ambient concentration outside the plant boundary.
c Annual incidence is the total cancer incidence for individuals living in the neighborhood of the facilities
under study.
-------
TABLE 5-4. NATIONWIDE EMISSION AND RISK ESTIMATES AGGREGRATED BY SIC CODES
FOR A UNIT RISK FACTOR OF 2 E-7 a
Ul
I
Ln
Number of For visited plants Number in National estimate
visited Average Average Maximum category Total Total
plants in emissions annual individual nation- emissions annual
category (Mg/yr) incidence0 risk^ wide (Gg/yr) incidence0
Chemicals and allied
products (SIC 28) 16 40
Fabricated metal
products (SIC 24 0
Electrical equipment
(SIC 36) 0
Other metal-related products
(SIC 33,35,37) 3 50
All other manufacturing
(SIC 20-27,29-32,38-39) 6 200
Not otherwise specified 23 1000
9 E-5 1 E-6 1249 50 0.1
547
540
1 E-4 1 E-6 804 40 0.1
1 E-4 2 E-6 878 200 0.09
9 E-4 4 E-5 800 800 0.7
TOTALSd 1100 0.99
a The unit risk factor is the probability of cancer incidence for exposure to 1 ug/mj over a 70 year
lifetime.
k Maximum individual risk is the lifetime cancer risk to an individual exposed over a 70 year lifetime to
the highest ambient concentration outside the plant boundary.
c Annual incidence is the total cancer incidence for individuals living in the neighborhood of the facilities
under study.
-------
^s f c = t^ie risk of cancer for an individual exposed to the
maximum ambient concentration due to Es f c;
Ms = the emission model for source type "s" (Chapter 3);
HEM = the human exposure model (Chapter 4); any [ ] denote
parameters used in Ms or HEM.
Emissions, maximum individual risks, and annual cancer incidences for the
visited plants are given in Appendix A. Because a number of the selected
plants made confidentiality claims on some or all of the information used to
estimate emissions, the emission and risk information given in Appendix A is
not matched with plant names. The maximum individual risks presented for
various source or SIC categories are simply the highest risks calculated for
any visited source or facility in the category:
M
R = Maximum Rs,f,c ^or any f or c
T
R- = z Rs f -
f,c s s,r,c
R
3
M
= Maximum R, for any f
c t, c
M
where R = the highest individual risk for a visited source in
S category "s";
T
R- = the sum of the maximum individual risks for all
>C sources at visited facility "f" in SIC "c"; and
M
R = the highest individual risk for a visited facility in
C SIC category "c".
The same basic methodology was used to develop national impact estimates
from model plant impacts in the facility aggregation (Tables 5-1 and 5-2) and
the SIC code aggregation (Tables 5-3 and 5-4) . In the facility type
aggregation, average emissions and annual incidences were first calculated for
each basic type of operation used at the visited plants:
-------
]7 — y -p* /—
f.c S' 'C S>
I_ T T / —
S "~ *^ S f C' ^ S f
f,c ' ' '
where Es = the average emission rate for source type "s" (i.e.,
drum handling, tank storage, etc.);
Is = the average annual cancer incidence due to emissions
from a source of type "s";
ns f = t*16 number of visited facilities which included
operations of type "s".
The total number of operations of each type nationwide was then obtained from
the Westat survey of plants impacted by RCRA.l Total nationwide emissions
were obtained for each source type by multiplying the average emissions for
the source type from the visited plant calculations by the total number of
facilities using that type of source nationwide. Annual cancer incidences
were estimated similarly for each source type by taking the product of the
average incidence for the source type and the nationwide number of facilities
using the source type. Total nationwide emissions and incidences were then
obtained by summing the national emissions and incidences for the various
source categories:
Ns
SET = SET
s s
SIT = HT
s s
T
where E = the total nationwide emission rate for source type
-------
T
I = the total annual cancer incidence due to emissions
from source type "s";
Ng = the number of sources of type "s" nationwide;
s T
E = the total nationwide aggregate emission estimate, by
the source type method;
s T
I = the total nationwide annual cancer incidence due to
TSDF emissions, aggregated by the source type method.
For the aggregation by SIC code, emissions and incidences were first
averaged for the various plants in each SIC code. Total impacts for each SIC
code were calculated by multiplying the average emissions and incidences for
the SIC categories by the nationwide numbers of plants falling into the
categories. Again the total numbers of plants falling into the various SIC
categories were obtained from the Westat survey.1 Total nationwide impacts
were obtained by summing the national impacts for the various SIC categories:
Er = I EC f r/nf -
C fa ° > r » C r » C
I , £>
TC - £ZB IS,f,c/nf,c
ET = ¥c • Nc
T —
i: - ic - NC
where EC = the average emission rate for a visited facility in
SIC category "c";
Ic = the average annual cancer incidence due to emissions
from a facility in SIC category "c";
-------
nf,c = t*16 numt>er of visited facilities in SIC category "c";
T
E = the total nationwide emission rate for TSDFs in SIC
category "c";
T
I = the total annual cancer incidence due to emissions
from TSDFs in category "c";
NC = the number of facilities in SIC category "c"
nationwide;
c T
E = the total nationwide emission estimate, aggregated by
the SIC category method;
c T
I = the total nationwide annual cancer incidence due to
TSDF emissions, aggregated by the SIC category method.
As Tables 5-1 and 5-2 show, the national estimates made using the two
aggregation techniques are similar. As illustrated by the SIC code
aggregation results given in Table 5-2, emissions and incidences could not be
determined for fabricated metal production facilities (SIC 34) or electrical
equipment manufacturers (SIC 36). Organic emissions from each of these
categories may be significant because of the use of solvents in cleaning parts
made in both categories.
Appendix F provides a comparison of the national emission and incidence
estimates presented in Table 5-1 with estimates based on mass balance
calculations for the visited facilities. These mass balance calculations were
made only for source types where such a method was applicable (i.e.,
impoundments, land treatment and landfills). Results were extrapolated and
incidences were computed as in Tables 5-1 and 5-2, although the number of
facilities used in the analysis was smaller due to the lack of waste
throughput information for some facilities.
-------
References for Section 5
1. Dietz, S., M. Emmet, R. DiGaetano, D. Tuttle, and C. Vincent (Westat,
Inc.). National Survey of Hazardous Waste Generators and Treatment,
Storage and Disposal Facilities Regulated Under RCRA in 1981.
Prepared for U.S. Environmental Protection Agency: Office of Solid
Waste. Washington, DC. April 1984. 318p.
-------
6. UNCERTAINTIES AND LIMITATIONS
The results of the nationwide analysis of emissions and risks from
TSDF area sources are subject to a number of limitations. These result
from uncertainties in the input data, and from simplifying assumptions
made as part of the general methodology. In addition, the emission
models and dispersion models used in the study are subject to some
uncertainty.
A major assumption in the analysis was that the sample of visited
facilities is representative of the nationwide TSDF population. In fact,
a study of the visited plants showed the sample to be biased in some
respects.^ A major bias is that the sample does not cover some SIC
categories. Also, the model plant group is somewhat biased toward large,
complex facilities, especially for plants unspecified by SIC. Thus, the
average volume treated among the visited TSDFs is much larger than the
average estimated by the Westat survey for the total TSDF population.2
Although these biases may cancel one another, they indicate a high degree
of uncertainty in the SIC code impact aggregation.
It should be noted that the national emissions estimates include only
emission sources covered by the Resource Conservation and Recovery Act
(RCRA). Many hazardous waste treatment, storage, and disposal operations
are currently exempt from RCRA requirements, although they do emit air
pol?ntants. (For instance, small facilities and some open tank1* are
exempt from RCRA). In fact, some non-RCRA sources were studied at the
visited facilities, and emissions calculations for these were used in
making estimates of emissions for similar RCRA sources. The use of
non-RCRA source data in estimating RCRA source emissions is not expected
to bias the national estimates substantially. However, the exclusion of
non-RCRA sources from the national aggregations certainly results in an
underestimate of overall TSDF emissions.
In addition, the study methodology involved the tacit assumption that
the sources for which emission factors or models have been developed are
the major emission sources. Some sources, especially transfer and
handling points, were not included in this study because of the
unavailability of emission models or emission factors. This lack of
emissions data for some sources may also have resulted in an
underestimate of nationwide impacts for these sources.
The use of default parameters for model inputs is another potential
source of uncertainty, although these were used only where no data were
available. For surface impoundments or open tanks, another major source
of uncertainty is in the waste concentrations which were input to
emission models. Waste analyses were not available for most plants, thus
analyses were typically based on estimates in the literature for generic
waste types.
-------
The emission models used in this study are empirical models, and many
have undergone preliminary verification studies; however any model is
subject to some uncertainty. In addition, some of the models were
extended to cases for which they have not been verified. For instance,
the unverified Hartley land disposal model was extended to cover disposal
surface impoundments in some cases, while the Thibodeaux-Hwang land
treatment model was extended to cover disposal surface impoundments in
other cases.
-------
References for Section 6
1. Memo from Andrew Baldwin, GCA/Technology Division, to William Battye
GCA/Technology Division. Case Study Facility Group
Representativeness. September 26, 1984. 16p.
2. Dietz, S., M. Emmet, R. DiGaetano, D. Tuttle, and C. Vincent (Westat,
Inc.). National Survey of Hazardous Waste Generators and Treatment,
Storage and Disposal Facilities Regulated Under RCRA in 1981.
Prepared for U.S. Environmental Protection Agency: Office of Solid
Waste. Washington, DC. April 1984. 318p.
-------
APPENDIX A
EMISSIONS AND RISKS
-------
• L nil I all
t Ciilit
1 4;
i 4;
1 4 =
: 4=
i it'
2 29
3 35
3 35
; 35
4 4?
4 49
4 45
4 49
5 28
t 25
: 28
7 26
7 25
8 37
B 37
5 37
B 37
'• 4<
I >'; '•*.
! .'. "'c
i'! 25
Iv 2:
!! !„
11 !o
f^CiLi: i
Driia Filiina
DisulUiiDn
Tiriter Unlc-adina
StcrsoE TsiiiiE
Draft storaae
Distillation
Drim sicrase
yE = tc-)3.I?r Irrat^eFit
Ei-tilUtiop
Distillation
Druti linlosjin;:
Ie3uurij*erit=
Storsae Tar;t=
jtasieHiier Triitaent
Irooundwnt
Incineration
incineration
HaEteHater Trsata°nt
Druii Starane
Incineration
Stance iants
Distillitic.il
Lindtreatflsr.t
Dr-, nar-K
:'.C-r = ;r Toilr.5
Oi5tilliti-:n
Iacoi.no»6nt
LY:is atorsas
HasteHatsr IreaMent
EniEiiuiii
uf.J, vf .
B.wE-00
5,!5E*Oi
>.65E *•.•(•
2.2=E»vi
TSIn, 9.271,01
2.00S-01
3.sfrE»"w
f-t ^4t ^^
TOTAL 3.SOE»00
3. ["•'.'!•
i . 29E*02
!.i'7tvO;;
TOTAL l.ioltw'
2.74E»00
4.00E-01
1 . 4BE+00
3.40E-61
ifttiitftti
TOTAL 4.?BE»w
4.7vE*Oi
|-*+ttt-
TOTnL 4.-f'E*0!
1. 90S *60
4.00E-OI
********
TOTAL 2.50E*00
1.21E-00
4.80E*0!
*******
TOTAL 4.92E+01
1.90E-01
6.00E-01
3.SOE-03
4.2'5E*00
******
TCTnl 5.04E*00
1.I8E+03
******
TOTAL 1.15E-03
4,!SPM^
l.i2E-"i
!.4:EH'l
i.tOEni!
TOT,: :.i«tJI
5.41E*«
2.23E+02
TC^L 2.33E*02
il-ul'.'Iii'jiiL Flil.
tilih ijiii ! l-hL-'O'r
CF 2,v'E"r
4.3;.E--D
4.i2E-i!'"
7.32E-0^
1.82E-07
f t *t *!*+ +
1.PE--
2.01E-V5
i»{»^^f »
2.12E-08
',--:-0?
i . 4 :E-"~
1. !"£-:• 5
1.4I£-Oi
1.19E-03
1.74E-05
4.42E-09
1.5;E-i)9
********
2.14E-OS
2.S8E-J7
^f ^*^t*«
2.o5E-07
2.98E-OV
9.42E-1')
*********
3.53E-09
3. 93E-09
1.54E-07
*********
l.iOE-07
L.47E-OS
4.»4E-OB
2. ~ !£-!',!
3.29E-07
*********
3.90E-07
5.94E-04
********
5.94E-14
i .•-,/:-.-,-
,' i "TE-" ,'
S.:!E-'}:
9.71E-07
2.9JE-06
2.39E-05
l.'"i!E-'''6
l.OJE-Oi
r**«***
6.E4E-04
7.04E-04
»*«**»**
7.'JiE-04
l.iOE-05
3. J'!'E-OE
3. m-'-j
3.!2E-OE
'TIiE-^
3.25E-04
l.;7E-:-4
1.40E-04
nii!!1'.!^ lilCj!"'E!'-';
hi !jiii i ''•'i':
OF i. •'..;- 5
E.50E-.3
c - • .- • -
j. . :;".-t
!.>I'c-vl
^. ~'it" .'.
l.«2c-0l
i.OOE-v:
i.OBE-.^
******¥ t**
I.I4E-04
B.fcE-:.;
i . r : : " v _
2.9=E-'Ji
3.5JE-0?
O.rtOE-'.-.
6!oOE+00
O.C"JE*Ovi
O.OQE-K'i
***********
O.OOEtOO
1.41E-«
*********
!.41E-;0
1.14E-54
3.6i-E-»5
»t»»***t
1.50E-J4
1.45E-54
5.7oE-03
tit*******
5.91E-/3
2.47E-05
7.30E-OE
4.55E-J7
5.53E-M
»<»*<«**»**
4.5oE->»
7.04E-02
*********
7.0tE-v2
i.t •:•£-"
3.40E-';
:, si?-..,:
3.52E-vT
"la".!
3.25E-)i
1.37E-<
1.40E-;.;
2.40E+04 P:JF^..
4.52E-0; URi--:,
3.i3E«05 DSEiti
-------
i:
[;
i ~
i ~
i:
13
ij
M
14
!,:
14
U
13
15
15
It
16
In
It
17
17
17
17
18
IS
ie
21.1
^t
Li
21
22
n
1-r
I.1
4?
49
i!
49
25
23
26
28
25
23
28
25
23
29
29
29
28
28
28
28
28
28
28
28
49
49
49
lift
28
28
42
tin
Nn
HA
IitDoundveni;
Drue Storssj
Landfill
Stors:? TarA'i
t'lstillatiar-
E'rar, Stcra:s
Storaae Tan.:;
Tanker Ur.loadino
Btoraoe tanks
Distillation
Incineratic^
Drus Stora:e
Va;tei\
4. liEti'i
i,'jSt'0!
t *** ft)r*ftV
l.BIE'O:
;.?3E+"i
3.35EOO
5.4lEt01
4.20E»w
9.30EMJ1
1 . 20E+00
1.71E+0:
6.00E-M
I.'^E'O^
3."2E*v!
***-**»v*
5.S5E*OI
S.vOE-O!
8.54Et02
1.53E*01
**tf ttttt
B.70Et02
4.80E-91
1.65E+01
4.0C/E-02
3.20EtOO
»*** **»
2.04E-01
4.50E-02
1.52EVOO
2.20E»02
4.00E-02
itHUH
2.21E»02
4.30E-01
9.19E+.W
9.20E-91
f ^fr4*t *
l.'ME^l
Z.lOEtOl
If »>**»»
2.10Etvl
6.00E-01
6.00E+00
1 1 tff * Jrf
i.OOE-0!
3.4iE+02
*4^t ^*
3.4itt»2
i."t-v2
2. ;3E»'»
1.2!E"»
*»f »*i
i.JvE+02
2.22E-07
5.3BE-:?
D. '9E-'!'S
6.S"r-)V
*• Itt t ^** f
3.04E-0-
2.'3E-07
:.i?E-OS
4.30E-0'
4.92E-OE
$##*$£ f fc
7.3EE-07
5.44E-il*»•(»
5.2AE-03
l.OOE-05
1.71E-02
3.04E-04
**«(«(>«•«»»
1.74E-02
8. 14E-04
1.98E-02
4.89E-05
3.B4E-03
»*>****»« *
'..45E-02
9.00E-06
3.04E-04
4.39E-02
1.20E-05
HHHHtt
4.42E-02
3.B3E-04
8. 18E-03
8. 19E-04
9i3BE-03
2. 10E-04
in* i**n
2. 10E-04
4.00E-04
O.OOE«)0
it#^+^i n f
6.00E-Oc
O.OoE»')0
*,;,., -,/,Etl,y
t.ltE-01
3.8:E-:»i
2. !BE-v4
1.17E-J1
4.25E-05 RU-ii.
t.07EKi5 RURSL
4.90EKI6 URBAN
2.I9E+04 URBAN
2,=3E+')6
fL'?.rt
-------
24
24
?j
24
25
25
7i
26
27
27
27
28
29
29
31
32
32
32
33
33
34
34
34
35
35
7C
J^
36
:'&
3o
26
2B
2B
26
29
29
49
49
28
28
28
17
35
35
49
28
28
28
49
49
2?
29
29
02
02
>)2
49
49
49
4'
Iniictior: iiell
Dm* Fiihno
Storage TanKs
Incid'rstion
Storaoe Tarn
Druft Storao*
Landtrestasnt
Landfill
Drill Stcrioe
Nastewitrr Treatssnt
luooundaent
Lanatreataent
Storaos Tanks
Drui Storaoe
lnoQundienU
Drut Storaae
Distillation
Storaae Tank;
Tanker Unloadino
Incineration
Dru* Slgraoe
H«steiiter freataent/API
Treatment Tank:
Distillation
Tam Storioe
Incineration
Landiill
*laste»iter Treatment
Incineration
Storaae Tank
TOTAL
TOTAL
TOTAL
TOTAL
TOTAL
TOTAL
TOTAL
TOTAL
TOTAL
TOTAL
(DIAL
TOTAL
*. DOE -.'I
I.80E-03
1.73E*vO
l.rittOO
^t»* r ff »
3.54E'i»'
9.90E-02
I.80E+W
1.90E»00
3. 10E+02
8.30E-02
#t fUrt
3.10E+02
2.90E-01
4.10E+00
7.30E-08
t^» t #*
4.39E»00
4.01E*03
4
9.00E-03
2.30E*00
1.19E+00
*K*f
3.50E+00
3.04EKH)
l.!lEtv«
1.24Ff)0
ft* **»*
5.6!E*W
1.31£*v2
5.53E*'.':
4.«E-01
5.45E":"i
*** + •*
5.?iE"::
3.75E-09
1.13E-I1
[.<>eE-:>8
7.eJ3£-0?
2.21E-08
o.64£-19
1.21E-OB
1.27E-08
1.74E-07
4.46E-11
1.74E-07
2.20E-09
3.12E-OS
5.55E-14
3.34E-OB
4.50E-04
4.50E-04
5.17E-09
9.05E-09
tttlrttH
1.42E-08
1.49E-09
1.49E-09
1.43E-OB
1.94E-07
1.03E-04
»»*«*t*t
1.24E-04
5.95E-08
2.0BE-06 '
nmut
3.03E-OB
3.32E-11
9.75E-09
5.05E-09
(»i»*(H
1.48E-08
2.75E-OB
1.33£-')8
1.12E-:i«
*#**•** ft
5.20E-oa
!.:3E-Ot
3.7t£-05
4.08E-09
3.71E-08
3.J9E-05
3.75E-07
1.13E-OV
i.-m-ie
7.56E-07
2.2lE-Ob
4.64E-OB
1.21E-04
1.27E-06
1.74E-65
4.46E-0'?
f Ht t#Hf *^>
1.74E-05
2.20E-07
3.12E-04
5.55E-I4
< t-^*+ ti-+*
3.34E-04
4.50E-04
ff Hf #4ttf
4.50E-V4
5.17E-07
9.05E-07
tit*4f^(4l
1.42E-04
1.49E-07
f/tf •ftffHtf
1.49E-07
1.43E-04
1.94E-05
1.03E-04
HHHtH
1.24E-04
5.95E-06
2.08E-04
»**»««*»
S.03E->6
3.82E-09
9.75E-07
5.05E-07
»Hf»M"
1.43E-06
2.75E-04
1.33E-Ot
1.12E-04
t**** vt»j
5.2vc-vb
1.23E-04
3.76E-03
4.06E-07
3.71E-04
3.89E-03
1.30E-07
5.40E-10
5.i?£-'7
3.43E-07
l.C'aE-'io
2.i8c-*"
4.32E-06
nm tt1**
4.56E-04
4.20E-05
l.ioE-OB
f f *< f tt*
4.20E-05
8.70E-08
i.2!E-Ot
2.19E-14
1.32E-06
4.42E-C3
fHftmi
4.42E-03
4.40E-04
8.04E-06
l**»*»*«
1.27E-05
2.00E-04
tHrKfrft
2.00E-06
9.10E-07
1.23E-'J5
4.53E-05
»»*>**»»
7.85E-05
3.27E-04
1.14E-04
f »**t«4*
4.4IE-06
1.26E-03
3.22E-06
1.47E-04
>»•»•**«
4.VOE-06
2.75E-08
l.:!E-JB
l.ilt-vfi
^ J *1**'**t
5.20E-OS
-, ,-ITF -.',«.
2.Jsi-C-3
3. WE -07
2.73E-06
*********
2.3t£-«
1.80E-0:
5.40E-05
5. 1SE-05
'.43E-05
l.OtE-i'4
2.3SE-05
4.32E-0-
4.54E-04
4.20E-03
l.tcE-06
4.20E-03
8.7C'E-(-=
1.23E-04
2.IiE-i:
i.3;E-(>4
4.42E-0!
6.42E-01
4.40E-04
8.04E-04
• JJHtl'tH
1.27E-03
2.00E-»4
2.00E-04'
9.10E-05
1.23E-03
6.53E-03
»4**»»«»<1
7.85E-03
3.27E-04
I.14E-04
kti>t«»*t
4.41E-04
1.24E-04
3.22E-04
1.67E-:'4
tf4^4*t»t<
4.90E-04
2.75E-Oi
1.3'.;E-Oo
S.I2E-V*
5.20E-vi
••.o:-£->:
2.76E-i,'l
3.JOE-0!
2.'3E-0-
2.66E-OI
1.77E+06 URBiii
3.1S£*05 RURfi^
e.35E»05 RURAL
3.i3EK>6 URSfiN
l.ZOEtOo RURAL
2.50E»06 RURAL
1.51E*04
-------
" 2s Distillation ;.30E'(") 2.71E->'S 2.71E-06 1.96E-06
3; 2S Storaoe Tanks 2.70E-OI 2.22E-0? 2.22E-0? 1.62E-0T
fr««*-ft ******** ******** **(»-**** *i
W'~i 3.57E-OC' 2.93E-05 2.93t-0i 2.14E-Oo
3S 42 Storsos Tanks I.02E*OI B.IOE-OS 8.1C-E-06 1.I2E-04
3i 4; Dm? BtsriOc 5.3<-t*00 4.21E-OB 4.21E-04 5.83E-0!
«*f^«- tir*««f>» ifi^titfi f*ft^4»^ (<
TCTAt 1.55E*01 1.23E-07 1.23E-05 1.71E-04
3° 2B Drua Starioe 1.54E+00 1.54E-08 1.54E-04 2.13E-05
tt«»* m*t**t »K*H*I jtfttui »>»K*(«m
TOTAL l.»4E*00 1.54E-08 1.54E-06 2.13E-05 2.1;E->! 1.30E«*r
40 HA Landfill 7.50E-01 3.92E-10 3.92E-06 2.25E-07
4C Nn ISDCU'idaent'ldMr,; fit 1.58E+02 B.26E-OB 8.24E-06 4.'4":-(5
M>tf l.t'tvt' H*K*lf» ttrl.tti m
8.30E-OE B.30E-06 4.74E-05
-------
APPENDIX B
-------
TABLE B-l. SUMMARY OF EMPIRICAL RELATIONSHIPS TO DETERMINE THE
INDIVIDUAL LIQUID AND GAS PHASE MASS TRANSFER
COEFFICIENTS FOR A NONAERATED IMPOUNDMENT l
Liouid Phase' Coefficient (k.^ .
MW \ 0.5
[1] k, = 1 11. 4 Re*°'195- 5) be°Z,ene I (Cohen, Cocchio and Mackay, 1978)
* MU.
for 0.11 < Re* _<_ 102, where k_ is in cm/hr and
7.07 x 10"3 (Z ) (U )1>25
Va ex? (56.6/U10°-25)
if Re* < 0.11, then k_'= 2.4 cm/hr
whe re;
Re* = roughness Reynolds number
K"rtv,erzBn° = "1o-£Cu-2-r weight of benzene (78.1 g/g-mole)
MWc = molecular weight of compound i (g/g-mole)
U]_Q = wind velocity (cm/s) measured at height ZJ_Q (10 m) above the water
surface .(ca)
Va = air kinematic viscosity (cm /sec)
r
D.
[2] k. = (1.3 Re*0^^ - 0.57)1- — ) (Cohen, et al. as presented
" • ** M
TOL,H2Oy by Hwang, 1982)
where;
k^ is in Ib-mol/ft2-hr. (Note: This equation is a modified form of
equation 1 to obtain the k^ value in units of Ib-mol/f t^-hr).
Re* = roughness Reynolds number (determined as above)
D, ^ Q = diffusion coefficient of compound i in water (cm-/sec)
^TOL F?0 = diffusion coefficient of toluene in water (cai-/sec)
-------
TABLE B-l. (CONTINTJED)
[3J k, = 3.12 (1.024)6"20 U °-67 H ~°'85\U / (Owens, Edwards and Gibbs..
° 2' 2 / 1964, as presented by
Ewang, 1982)
where;
k^ is in Ib-mol/ft^-hr
3 = temperature (°C)
Uo = surface velocity, ft/sec, normally 0.035 x wind speed (ft/sec) for
natural surface, and 0.1 ft/sec for outside region of effect
of aerators in the biological treatment.
Ho - effective depth of surface impoundment (ft)
DO~ H90 = diffusion coefficient of oxygen in water
Gas Phase Coefficient
[i ]
kr = 0.095S U . °'7S N "°'67 d "°-U -£i£- (MacKay and Matsugu, 1972)
v» air sc e Mw .
air
where ;
k^ is in Ib-aol/f t^-hr
Ua,-r = wind speed (ta/hr)
%c = 2as Schmidt number = Ug/PgD^ a^r
pg = absolute gas viscosity (g/ca-sec)
Pg = density of gas (g/cra^)
Dj_ a£r = diffusion coefficient of compound i in air (cm^/sec)
d^ = effective diameter of the quiescent area of the impoundment
(m) =
•\
Ac = area of convective (natural) zone of impoundment surface (m )
Pa£r = density of ci-r (lb/ft3)
MWair = molecular weight of air (28.8 Ib/lb-mole)
-------
TABLE B-2. THIBODEAUX, PARKER AND HZCK MODEL FOR SURFACE IMPOUNDMENTS
Type of Model:
Basis:
.Predictive
Mass transport calculations based on the two-film
resistance theory
Fora:
Ei ' Mi
and, for each volatile component i
T A + KT A )/(A + A )
iL t il n t n
Svrabol
A.,
Ei
kiG>
Symbol/Parameter Typical
Definition Precision3
Surface areas of the +.252
turbulent and natural
zones, respectively, (cm )
Flus o'f component i Unk
from the impoundment
surface, • (g/cai -s)
Henry's law constant Unk
in mole fraction form,
H - y/x
Individual gas phase Unk
mass transfer coeffi-
cients for the turbu-
lent and natural zones,
respectively, (moI/cm. -s)
' Overall liquid-phase Unk
mass transfer coeffi-
cient for component i,
(mol/cm2-s)
Source of
In-put Parameter
Measured
Literature data
or estimation
Calculated from
empirical corre-
lations*
Calculation
-------
Individual liquid
phase mass transfer
coefficients for the
turbulent and natural
zones, respectively,
(mol/cm -s)
Unk
Calculated from
empirical corre-
lations*
M,-
Overall liquid phase
mass transfer coeffi-
cients for the turbu-
lent (aerated) and
natural (un-aerated)
zones of the impound-
ment, respectively,
(mol/cm -s)
Molecular veight of
component i,(g/g-mol)
Mole fraction of com-
ponent i in the aq-
ueous phase
Mole fraction of com-
ponent i in equilib-
brium vith the mole
fraction of component
i in the air, y^. If
y- is assumed to be
negligibly small,
X,-* - 0.
Unk
Calculated
0%
+21
+22
Literature data
Calculated from
concentration
measurements
Calculated from
measured concen-
trations in the
atmosphere
Empirical Correlation for Individual Mass Transfer Coefficients are:
,1/2
t 0.823 J (POWR)a (1.024)
8"20 D
iL
c -5 PgSi,air 1<42 Q.4Q 0.5 -0.21
k. - 1.35x10 5 N N N N
iG a Re P Sc Fr
NT, » Reynolds Number, d^v p / y
^ O O
(b)
N = Power Number, Pr
^w
-------
NSc = Schmidt Number, Vg/I>i>air Pg
M
Npr = Froude Number, dw /g
iL
(c)
„ ,0 0.78 -0.67 -0.11
0.482 U . N. d
air Sc e
air
(d)
Svmbol
POWR
Div
02*
Symbol/Parameter Typical
Definition Precision
orygen-transfer rating of -
surface aerator,' normally
in the range of about 2-4
Ib 02/hp-hr
total power input to aera- +_102
tors in aerated surface
impoundment, Hp
correction factor for
wastewater/clean water
ozygen transfer (0.80 to
0.85)
water temperature, (°C) +11
diffusion coefficient ±152
for component i in water,
cm2/sec
diffusion coefficient • • ±1-52
for ozygen in water,
(cm2/sec)
surface area per unit .+102
of volume of surface
impoundment, (ft /ft )
volume of surface im- +252
poundment in region
affected by aeration,
(ft3)
density of air, (g/cm ) Dnk
Source of
Input Parameter
Estimated
Measured
Estimated
Measured
Laboratory data
or estimated
Laboratory data
or estimated
Calculated
Estimated
Literature
-------
i.air
U •
air
air
diffusion coefficient
for component "i" in
air, (cm /sec)
diameter of aerator
turbine or impeller,
(ft) (cm in Eq. for
kiG and NRe)
rotational speed of
turbin-e imp el lor,
(rad/sec)
viscosity of air,
(g/cm-sec)
power to impeller,
acceleration of gravity
32.2 ft/sec2
density of liquid,
(g/cm3)
surface velocity,
(m/sec)
effective depth
of surface impound-
ment , (m)
wind speed, (m/sec)
effective diameter
of quiescent area
of surface impound-
ment , (m)
molecular weight of
air, (g/g-mole)
non-aerated impoundment
area (m )
±102
Unk
±101
±0%
±22
±102
±102
±102
+252
+02
+102
Laboratory aaca
or estimated from
correlations
Measured
Measured
Literature
Calculated
Literature
Laboratory data
or literature
Measurements and
calculated
Measured or esti-
mated
Measured
Calculated
Literature
Measurements and
calculations
aln many cases these are rough estimates. Values can be refined with future
data. Unk = unknown.
-------
TABLE B-3. THIBODEAUX-HWANG MODEL FOR VOLATILE ORGANIC EMISSIONS FROM
LANDTREATMENT OPERATIONS 2
Basis: Emission rate is controlled by diffusion rate of vapor
through the air-filled pores of the landtreated soil.
Form:
E
D C
ei ig
i /2 D t A (h -h )C \ 1/2
M
x io
H C
c. c io
ig / H D . Z
D . A f(y)
wi s
and
f(y) = (h2, + hphs - 2h2,)/6
Typical Source of
Symbol SvmboI/Parameter Definition Precision Input Parameter
A surface area over which waste is applied, +_ 21 measured
car
A interfacial area per unit volume of soil - calculated
for the oily waste, cm /cm
C- effective wet zone pore space concentration - calculated
of component i, g/cm
C- concentration of component i in oil, g/cm +. 25Z calculated
D • effective diffusivity of component i in the i 252 published
air-filled soil pore spaces, cm /s data; esti-
mation
d soil clump diameter, cm - average cal-
culated from
measurements
or estimated
(Continued)
-------
TABLE B-3. (CONTINUED)
D • effective diffusivity of compound i in the +. 5%
wi
oil, cm /s
E^ flux of component i from the soil surface,
g /cm -sec
f(y) (h* + hphs-2h^)/6 accounts for the
lengthing dry zone
H Henry's Law constant in concentration
cm3 oil/cm'' air
depth of soil contaminated or wetted with
landtreated waste, cm
Mio
wf
Z.
initial mass of component i incorporated
into the zone (h -hg), g
time after application, sec
height of wetted soil remaining after
partial drying, cm
weight fraction oil in film 'form in soil
oil layer diffusion length, cm
f fraction of oil in film form
p soil clump density, g/cm
*j
waste-oil density, g/cm"
15%
102
52
+ 102
102
published
data; esti-
mation
calculated
calculated
published
data or
measurement
measured
measured
measured
measured
calculated
calculated
or esti-
mated
estimated
measured or
estimated
measured or
estimated
-------
TABLE B-4. FARMER MODEL FOR VOLATILE ORGANIC EMISSIONS FROM
COVERED LANDFILLS2
Basis: Emission rate is assumed to be mass transfer controlled by
diffusion of gases through the air-filled soil pores.
Form:
10/3
(P ) , v
E. = K D. C A —-—
i D is ,„ N2
P M
C =
w
s RT _
v-'-(-y
• -m
TS = P _ Q
-a rc c
Symbol Svsbol/Parameter Definition Precision Inout Parameter
A Surface area of the landfill (ca^) +0.12 File data or
direct measure-
ment
B Soil bulk density (g/c:^) ±3" Varies from 1 to
2 g/cc. Need
direct measure-
ment for accuracy,
C Saturation vapor concentration (g/mO .051 Calculated from
gas lav and
species vapor
pressure.
(Continued)
-------
TABLE B-4. (CONTINUED)
a£r Diffusion coefficient of the species of
• interest in air (cm /sec)
+5%
Ei
M-
Pi
R
Mass emission rate (g/sec)
Codisposal factor. Use 1.0 for isolated
toxic waste disposal and 6.0 for toxic
waste codisposed with biologically de-
gradable wastes.
Depth of soil cover (cm)
±10%
-2%
+ 172
Molecular weight of the species (g/mole) -
Air-filled porosity (dimensionless) ±30%
Vapor pressure of the species in interest' 5%
(mm Hg)
Soil particle density (g/cnr) ±82
Soil porosity (dimensionless)
Density of water (g/cm )
+13;
+23
Gas constant 62,300
mm flg - cm"
°K - mole
Calculated value
from literature
or ratio to a
compound with a
known D by
molecular weight,
Literature
File data or
measurement
Literature
Calculated
Literature or
direct measure-
ment.
Recommends 2.65
g/cm3 for most
mineral material.
Can be estimated
based on soil bulk
density and soil
particle density. .
Literature
Given
(Continued)
-------
TABLE B-4. (CONTINUED)
W^/W Weight fraction of the species of
.interest in the disposed waste (g/g)
WW/W Weight fraction of water in the soil cover +5%
T Temperature (°K)
Volume fraction of water in the soil
cover (g/g)
+.202 Direct measure-
ment.
Direct measure-
ment.
+_1°K Direct measure-
ment.
+172 Direct measure-
ment.
-------
REFERENCES FOR APPENDIX B
1. DeWolf, G.B. and R.G. Wetherold (Radian Corporation). Protocols for
Calculating VOC Emissions from Surface Impoundments Using Emission
Models: Technical Note. Prepared for U.S. Environmental Protection
Agency: Office of Air Quality Planning and Standards. Research Triangle
Park, NC. September 1984. 32p.
2. Farino, W., P. Spawn, M. Jasinskis and B. Murphy (GCA/Technology
Division). Evaluation and Selection of Models for Estimating Air
Emissions from Hazardous Waste Treatment, Storage and Disposal
Facilities. Prepared for U.S. Environmental Protection Agency: Office of
Solid Waste. Washington, DC. May 1983. Chapters 4, 6 and 7.
-------
APPENDIX C
-------
APPENDIX C. DEFAULT PARAMETERS FOR EMISSION MODELS
Model
Parameters
Default Value
Reference (default)
Comment
Drum handling - working loss
Saturation factor
Vapor pressure
Molecular Weight
Temperature
0.2
NA
NA
25°C
1 (p 5-6)
Assumed bottom filling
Compound specific
Compound specific
Assumed
Fixed-roof tank: AP-42
Tank diameter
Average vapor space height
Average diurnal temperature change
Paint factor
Adjustment for small diameter
Product factor
Turnover factor
Annual throughput
Surface Impoundments: Cohen, Cocchio, Mackay, Owens
Aerated
Air temperature
Water temperature
Liquid density
Wind speed
Height of wind speed measure
Effective depth
Surface area
Turbulent area fraction
Aerator parameters
diameter of impeller
rotational speed
volume affected
power to Impeller
total power to aerators
oxygen transfer rating
correction factor for
oxygen transfer
based on capacity
0.5 (height) in ft
20°F
1.0
NA
NA
NA
NA
25°C
25°C
1 g/cm3
variable
1000 cm
410 cm
NA
0.2%
60 cm
126 radiens/second
50-1124 mj
5-100 Up
NA
3 Ib 02/hr-hp
0.825
1 (p. 5-12)
2 (p. 7-3)
3 (p. 38)
2 (p. 4-25)
2 (p. 4-25)
2 (p. 4-25)
2 (p. 4-25)
3 (p. 37)
3 (p. 37)
2x [v/(23.5'height)]°/5 where height is in 8'
Intervals, or manufacturer's specifications
Average of national estimates for large (0.75)
and small (0.25) facilities.
White, good condition tank. Best case.
Compound specific
Assumed
Assumed
Density of water
Closest available STAR aerometric data
Assumed
Site specific
Based on aerators required
Based on turbulent area required
Based on turbulent area required
Typical range 2-4
Typical range 0.8 - 0.85
(continued)
-------
APPENDIX C. DEFAULT PARAMETERS FOR EMISSION MODELS (continued)
Model . Parameters Default Value Reference (default)
Non-aerated See Aerated Impoundments, except
Effective depth
treatment
storage
disposal
Landtreatment
Thibodeaux Initial concentration of component
Soil contamination depth
Depth of subsurface injection
Duration of waste on ground
Surface area of application
Fxn of oil in film form
Initial mass of component
Soil clump or particle diameter
Soil clump density
Oil density
Bulk density of soil
Porosity of soil
Hartley Evaporative rate
Relative humidity
Vapor pressure
Molecular weight
Vapor pressure (water)
Molecular weight (water)
226 cm
194 cm
30 cm
NA
23 cm
NA
NA
NA
1.0
NA
0.005 cm
2.65 g/cm3
0.78 g/cm3
1.5 g/m3
0.43
34.78 g/cm2-mo
0.25
0.6
NA
NA
Constant
Constant
4
4
4
3
3
3
3
3
2
2
5
5
3
--
(P-
--
--
--
(P-
--
(P-
(P-
(P-
(P-
(P-
(p.
--
--
--
17)
17)
17)
17)
17)
6-4)
6-3)
12)
Comment
Wastestream specific
Site specific
Site specific
Site specific
Site specific
Kings County, CA
Kings Co. , CA for West Coast or Texas
Other sites
Compound specific
Compound specific
23.756 mm llg @ 25°C
18
(continued)
-------
APPENDIX C. DEFAULT PARAMETERS FOR EMISSION MODELS (continued)
Model
Parameters
Default Value
Reference (default)
Comment
Landfill
(covered): Farmer
Incinerator fugitives
Deep well
injection
Air temperature
Surface area
Depth of soil cover
Bulk density
Soil particle density
Height fxn of water
Weight fxn of component in waste
Source inventory
Source inventory
fugitives
25°C
NA
30
1.
2.
0.
NA
2
1
2
1
Assumed
Site specific
cm
5 g/cm3
65 g/cm3
2
pumps
relief valve
pumps
relief valve
3
2
3
3
(P-
(p.
(p.
(P-
--
--
—
--
22)
5-7)
22)
22)
Wastestream
Engineering
Engineering
Engineering
specific
judgment
judgment
judgment
Recycling fugitives
Source inventory
2 light liquid pumps
1 heavy liquid pump
27 gas valves 6
35 light liquid valves 6
17 heavy liquid valves 6
2 safety relief valves
21 open-ended lines 6
5 sampling connections 6
120 flanges 6
Engineering judgment
Engineering judgment
Engineering judgment
-------
References for Appendix C
1. Engineering-Science. National Air Emissions from Tank and Container
Storage and Handling Operations at Hazardous Waste Treatment, Storage
and Disposal Facilities. Prepared for U.S. Environmental Protection
Agency: Office of Air Quality Planning and Standards. Washington,
DC. September 1984. Chapter 5.
2. Farino, W., P. Spawn, M. Jasinski, and B. Murphy (GCA/Technology
Division). Evaluation and Selection of Models for ' stimating Air
Emissions from Hazardous Waste Treatment, Storage and Disposal
Facilities. Prepared for U.S. Environmental Protection Agency:
Office of Solid Waste. Washington, DC. May 1983. Chapters 4, 6 and
7.
3. Wetherold, R.G. and D.A. Dubose (Radian Corporation). A Review of
Selected Theoretical Models for Estimating and Describing Atmospheric
Emissions from Waste Disposal Operations. Prepared for U.S.
Environmental Protection Agency: Industrial Environmental Research
Laboratory. Cincinnati, OH. June 1982. Chapter 3.
4. Breton, M., T. Nunno, P. Spawn, W. Farino, and R. Mclnnes
(GCA/Technology Division). Assessment of Air Emissions from
Hazardous Waste Treatment, Storage and Disposal Facilities (TSDFs)
Preliminary National Emission Estimates. Prepared for U.S.
Environmental Protection Agency: Office of Solid Waste. Washington,
DC. August 1983. p. 34-38.
5. Scheible, M., G. Shiroma, and G. O'Brien. An Assessment of the
Volatile and Toxic Organic Emissions from Hazardous Waste Disposal in
California. State of California Air Resources Board. February
1982. p. D-4.
6. U.S. Environmental Protection Agency: Office of Air Quality Planning
and Standards. Control of Volatile Organic Compound Leaks from
Synthetic Organic Chemical and Polymer Manufacturing Equipment.
EPA-450/3-83-006. Research Triangle Park, NC. 262p.
-------
APPENDIX D
-------
DRUM STORAGE AND HANDING
Drum storage losses are considered as working losses and fugitive losses.
In the case of working losses
Lw(lbs/1000 gallons) = 12.46 SPM
T
where,
S = saturation factor,
P = vapor pressure of contents in PSIA,
M = molecular weight of contents,
T = temperature in degrees Rankin.
For the drums handled at facility #18, the following values were used:
S = 0.2 (assumes bottom filling, see Appendix C),
P = 2.5 PSIA (vapor pressure of trichloroethylene; the drum contents had
to be assumed, see Appendix C)
M = 131 (molecular weight of trichloroethylene)
T = 536.4°R (25°C)
Throughput = 100 drums weekly, or 260,000 gallons yearly @50 gallons per
drum. Substituting into the equation yields 1.532 ,lbs/1000 gallons or
0.18 Mg/year.
Fugitive losses stem from equipment leakage and spillage. These factors,
respectively, are 0.0017 Ibs/drum and 50 gallons per 100,000 gallons. Again
substituting the throughput figures, this yields 8.84 Ibs/yr and
130 gallons/yr, respectively. At a density of 1.5 g/ml (trichloroethylene),
the latter figure translates to 0.738 Mg/yr. Total fugitive losses are
0.74 Mg/yr.
Total drum storage losses total 0.92 Mg/yr.
-------
TANK UNLOADING
1
Two factors are used in estimating emissions from tanker unloading:
0.095 g/kg handled for spillage, and
0.36 g/kg handled for unloading.
At facility #13, approximately 1.4 x 10? kg/yr are unloaded from tankers.
Using the factors above,
spillage = 1.3 Mg/yr emitted, and
unloading = 5.0 Mg/yr emitted.
-------
FIXED-ROOF STORAGE TANKS2
Fixed roof tank breathing losses can be estimated from:
LB = (2.26 x 10"2)M [P/U4.7-P)])0-68 D1-73!!0^! TO-50FP CKc
where: Lg '= fixed roof breathing loss (Ib/yr)
M = molecular weight of vapor in storage tank (Ib/lb mole)
P = true vapor pressure at bulk liquid conditions (PSIA)
D = tank diameter (ft)
H = average vapor space height, including roof volume
correction (ft)
T = average ambient diurnal temperature change (°F)
Fp = paint factor (dimensionless)
C = (adjustment factor for small diameter tanks (dimensionless)
KQ = product factor (dimensionless).
For a tank at facility #18 storing toluene, the relevant values are as
follows:
M = 92 (toluene)
P = 0.56 PSIA (toluene at 25°C)
D = 11.4 (given)
H = 16.3 (assumed half full, on average)
T = 20°F (see Appendix C)
Fp = 1.33 (light gray tank, given)
C = 0.6 adjustment for small diameter)
KC = 1.0 (all organic liquids except crude oil)
-------
Substituting these values yields a breathing loss estimate at 105 Kg.
Fixed roof tank working losses can be estimated from:
Lw = 2.40 x 10'2 MPKNKC (2)
where: Ly = fixed roof working loss (lb/10^ gal throughput)
M = molecular weight of vapor in storage tank (Ib/lb mole)
P = true vapor pressure at bulk liquid conditions (PSIA)
KJJ = turnover factor dimensionless)
K£ = product factor (dimensionless).
The fixed roof working loss (Lw) is the sum of the loading and unloading
losses. Special tank operating conditions may result in losses which are
significantly greater or lower than the estimates provided by Equation 2.
For the toluene tank at facility #18, the values are as follows:
M = 92 (toluene)
P = 0.56 PSIA (toluene at 25°C)
% = 1 (less than 36 turnovers per year)
KC » 1
Throughput = 34,605 ga/yr.
Substituting, losses are 0.56 kg/1000 gallons or 19.4 kg/year.
Total loss of toluene from this tank is therefore estimated at 124.4 kg/yr.
-------
TREATMENT TANKS - OIL/WATER SEPARATORS
Oil/water separators were treated as uncontrolled emitters. An emission
factor (Reference 3) of 0.6 kg per 1000 liters throughput was used. At
facility #15, emissions were calculated using a throughput of 806,250 liters
annually for the API separator. This actually is the sludge volume generated
by the separator and is therefore an underestimate. No total throughput was
available. Using this data, emissions are calculated as 0.5 Mg per year for
the API separator.
-------
DISPOSAL IMPOUNDMENTS AND LAND SPREADING
The Hartley model for landtreatment emissions is:^
Ea = Ew/(l-RH)[Pa Ma °-5/Pw Mw °-5](A) (Wi/W)
where:
Ea = emission of chemical a in g/month
A = surface area (cm^)
Wi/W = weight fraction of a in waste
Ew = evaporate rate of water (g/cm2 - month)
Mw = molecular weight of water
Pw = vapor pressure of water (mm Hg)
RH = relative humidity
Pa = vapor pressure of a (mm/g)
Ma = molecular weight of a
For facility #9, the appropriate values for these variables are given
below. The site is in Texas and landtreats K wastes. Typical waste code
compositional breakdowns were used to determine the organics and their
concentrations.
A = 3.2 x 108 cm2 (given)
Ew = 34.78 g/cm2 - month (California data, see Appendix C)
Mw = 18 g/mole (given)
Pw = 23.756 mm Hg @25°C (given)
RH = 25% (California data, see Appendix B)
-------
Volatile organic Pa(mmHg) Ma Wi/W Emission (mg/yr)
Benz(a)pyrene 7.5 x 10"9 252 7.4 x 10'8 0.0
Cyanide (as HCN) 726 27 1.7 x 10'4 1140
Phenol 0.34 92 7.4 x 10~5 7.2
Oil (as crude) 6.0 x 10"3 189 9.0 x 10"2 13.2
Results of substituting the values are tabulated above. Each chemical
constituent is treated separately. Total emissions from landtreatment at the
site are 1160 Mg/yr.
-------
SOLVENT RECOVERY STILL
Facility Number 3
Throughput: About 1000 gal/yr of solvent is processed, principally
freon-113 and trichloroethylene. It is assumed that
50 percent of the solvent is freon-113, and 50 percent is
trichloroethylene. This corresponds to about 2760 kg/yr of
trichloroethylene and 2950 kg/yr 1,1,2-trichloro-
1,2,2-trifluoroethane (freon-113).
Process Fugitive Emission Sources:
2 light liquid pumps - 98.8 g/hr-pump (assuming 2 seals/pump)
1 heavy liquid pump - 42.8 g/hr-pump (assuming 2 seal/pump)
27 gas valves - 5.6 g/hr valve
35 light liquid valves 7.1 g/hr-valve
17 heavy liquid valves - 0.23 g/hr-valve
2 safety relief valves - 104 g/hr-line
21 open ended lines 1.7 g/hr-line
5 sampling connections - 15 g/hr-connection
120 flanges - 0.83 g/hr-flange
Note: The numbers of pumps and relief valves were estimated based on a
recovery operation configuration comprising a batch distillation
column with a condenser and with no tank storage (solvent is
pumped directly to and from drums). The numbers, relief valve
inventories, and standard ratios of process valves to pumps,
etc., are taken from Reference 5. Emission factors for process
fugitive sources are also from this reference. The solvent
still is assumed to be operated about 1000 hrs/yr.
-------
Other Emission factors:"
Still condenser vent: 1.65 g/kg
Drum loading: 0.36 g/kg
Spillage: 0.10 g/kg
Calculated Emission
Process Fugitive:
Still condenser vent;
drum loading; and
spillage
Totals:
0.96 Mg/yr total VOC
0.48 Mg/yr trichloroethylene
0.48 Mg/yr 1,1,2-trichloro-
1,2,2-trifluoroethane
0.012 Mg/yr total VOC
0.0058 Mg/yr trichloroethylene
0.0062 Mg/yr 1,1,2-trichloro-
1,2,2-trifluoroethane
0.97 Mg/yr total VOC
0.49 Mg/yr trichloroethylene
0.49 Mg/yr 1,1,2-trichloro-
1,2,2-trifluoroethane
-------
INJECTION WELLS
Emissions from the associated pumps and valves in a deep well injection
unit were estimated as follows. Two pumps (two seals per pump) and one relief
valve, all in light liquid service, were assumed per unit. Total operation
time was assumed to be 2000 hours per year. Emission factors for these
sources were taken from Reference 5.
98.8 g/hour - pump, and
104 g/hour - valve.
For example, the injection well at facility #21 was treated in this
manner. Multiplying the 2000 hours by the emission factors yields 0.395 Mg
per two pumps and 0.208 Mg per valve. Total emissions for the unit are
therefore 0.60 Mg per year.
-------
INCINERATORS
For the purposes of estimating fugitive emissions associated with
incinerators, there were several assumptions made in all cases. Two pumps
(two seals per pump) and one relief valve, all in light liquid service, were
assumed per unit. Emission factors for these sources were taken from
Reference 5i
98.8 g/hour - pump, and
104 g/hour - valve.
Incinerators were assumed to operate 2000 hours per year in each ease.
For example, the incinerator at facility #14 was treated in this manner.
Emissions for the two pumps are 0.395 Mg per year, and for the valve losses
are 0.298 Mg. Total emissions for the unit are therefore 0.60 Mg per year.
-------
SURFACE IMPOUNDMENTS
Non-Aerated (Facility #11, model wastestream - see Section 3.3.7)
KG = 2.920 Uair °-78 NSc ~0-67 de - °-n (cm/hr) (MacKay)7
Uair = wind speed (meters/hr) = 15,000 m/hr
Ngc = Mair = Schmidt Number (dimensionless)
Pair Di.air
Mair = 4.5686 x 10"7 T + 1.7209 x 10"4
= 4.5686 x 10~7 (25. °O + 1.7209 x 10'4 = 1.835xlO'4 §/cm.sec
Pair = MWair/.08206 (T+273)xlOOO
= 28.8/.08206(298)xlOOO = 1.178 x 10'3 g/cm3
Di.air = .07 cm2/sec
NSc = 1.835xlO"4/1.178xlO"3(.07) = 2.225
de=(4A/3.14)^'^ = effective diameter (meters)
= 4((1.17 x 103 m2)/3.14)°-5 = 38.6 meters
kG = 2.920 (15,000)°-78 (2.225)'-67 (38.6)'-11
= 2.920 (1808.64) (.585X.669)
= 2066.88 cm/hr
kL = (11.4 RRe'195 -5) (78.1/MWL)°-5 (Cohen, Cocchio, Mackay)7
RRe a 7.07 x 10"3 Z U ^-25 = Roughness Reynolds #
Va exp (56.6/U'z5)
Z = height of measurement = 10 meters = 1000 cm
U = windspeed = 15,000 m/hr = 416.67 cm/sec
Va = Mair/Pa = 1.835 x 10"4 g/cm • sec = .1558 cm2/sec
1.178 x 10'3 g/cm3
-------
RRe = 7.07 x 10-3 (1000) (416.6
.1558 EXP (56.67(416.67) )
= 8.54 x 104 = .31
2.759 x 105
kL = (11.4/.3D'195 -5) (78/100)0-5
= 4.07 (.884) = 3.60 cm/hr
Overall Mass Transfer Coefficient
_! = _! + 1 H = Henry's law const
k kL HkG
= 1 + 1
3.60 .01 (2066.88)
= .278 + .048
= 3.07 cm/hr
Emission Rate
Q = K (cm/hr) ppm x 10~6 (g/cm3) Area (cm2)
= 3.07 (40 x.10"6) (1.171 x 107)
= 1.438 x 103 g/hr
1.438 x 103 g/hr x Mg x 24 hr x 365 day
10°g day yr
= 12.60 Mg_
yr
Aerated (Facility #11, model wastestream - see Section 3.3.7)
Natural side--
kG - 2.920 U air °-78 Nsc-°-67 de'0-11 cm/hr (MacKay)7
Nair = 15,000 m/hr
NSc = 2.225
de = (4A/3.14)0'5 = 4 ((2.770 x 102/3.14))°-5 = 18.78 meters
kG = 2.920 (15.000)0-78 (2.225)"0-67 (18.78)'0-11
= 2.920 (1808.64) (.585) (.724)
= 2236.8 cm/hr
-------
kL = (11.4 RRg'195 -5) (78.1/MWi)0-5 (Cohen, Cocchio, Mackay)7
= 3.60 cm/hr
Convective Side—
kT = 1.90 x 10"4 Re1-42 NP-4 Sc-5 FR'-21 Pair Pi, air7
G d
Re = d2 w Pair = Reynolds #
Mair
NP = Pr gc/Pi (d/30.48)5 W3 = Powers #
NSc = Mair = Schmidt #
Di, air Pair
NFr = dw2 = Froude #
gc
d = impelled diameter = 60 cm
w = rotational speed of impeller = 126 rad/sec
Pair = 1.178 x 10"3 g/cm3 = density air (at 25°C)
Mair = 1.835 x 10~4 g/cm sec = viscosity air (25°C)
Pr = power to impeller = 4.5 Hp
gc = grav. const.
?l = density of water = Ig/cc or 62.37 lb/ft3
Di,air = .07 cm^/sec
Re = 2.91 x 106
NP = 6.88 x 10'4
SL = 2.225
FR = 9.53 x 105
kTG = 3.28 cm/hr
KLT = 1.509 x 103 J (POWR) a (1.024) e~20 (Diw/2.41 x 10"5)-5/(AV/30.48)V (Ref. 7)
J = 02 Transfer Rating = 3 Ib 02/Hp-hr
POWR = Total Power to Aerator = 5Hp
-------
a = correction factor for D£ Transfer = .825
6 = Temperature = 25°C
Ay = surface to volume ratio (meters'^) = 100
depth (cm) = .244
V = volume affected by aeration = 50 m3
kLT = 3643 cm/hr
K = 1 + 1
kL Keg kG
Keg = HP1MWair
RT Pair ((ppm (MWi) - (1-ppm) MWj_) x 10-6)
H = Henry's Law Const = .01 atm m3/mole
R = 8.2 x 10~5 m3-atm/mol K
PI = I g/cm3
MWair = 28.8 g/mole
Pair = 1.178 x 10'3 g/cm3
= 40
= 18.0 g/mole
Keq = 555.9
KT = 1 + 1
Keg KGJ
3643 (555.9X3.28)
= 1215.16 cm/hr
KN = 1 + 1
Keg KG"
1 + 1
3.60 (555.9X2236.8)
=3.60 cm/hr
KOVERALL = fKT +- d-f)KN
f = fraction aerated = .058
-------
= .058 (1215.16) + (1-058) (3.60)
= 70.48 + 3.39
K = 73.87 cm/hr
Emission Rate
Q = K (ppm x 10'6) Area (cm2)
= 73.87 (40 x 10'6) (2.770 x 106)
= 8185 g/hr
8185 g/hr x 8.76 x 10~3 = 71.70 Mg
yr
LAND TREATMENT (Thibodeaux-Hwang Model)7 (Facility #40, methylene chloride)
Q = Dei(f)/tA(hp_hs) (hs2 + 2 Dei A (hp-hs) Cig t/fMio)-5 x 315.36
De± = DiP1-33
Cig = He Cio
f = fraction of waste in film form = 1.0
t = duration of waste on ground (sec) = 86,400
A = surface area of landfarm (cm2) = 6.1 x 10°
hp = soil contamination depth (cm) = 60.96
hs = depth of subsurvace injection (cm) = 0
Mio = initial mass of i put of ground (g) = 22,497
Dei = effective diffusivity
Di = diffusivity of i in air (cm2/sec) = 8.08 x 10~2
Cig = effective wet zone pore space concentration (9/cm3)
C^o = initial concentration of i in the waste (g/cm3) = .01
HL = Henry's Law Cont in concentration form (cm3 oil/cm2 air) = 2.4 x 10"-*
P = soil porosity = .41
Q = Emission Rate (Mg/yr) = 5.54
-------
COVERED LANDFILL (Farmer Model)7 (Facility #28, methyl ethyl ketone)
E = KD D Cs A Pa 10/3 JL Wi
L W
PT = 1 - B
P
Pa = Pi-e
e = ww B
w PW
E = mass emission rate (g/sec)
Kj) = codisposal factor =1.0
M = molecular wt. of species = 85.0
R = Gas const. = 62,300 mm Hg - cm3/k mole
T = Temp (K) = 298
PT = soil porosity = .55
L = depth of soil cover = 320 cm
A = surface area of landfill = 1.00 x 108 cm2
Wi = weight fraction of species i in waste = 7.243 E-3
W
B = soil bulk density = 1.56 g/cm3
P = soil particle density = 2.65 g/cm3
Pvp = vapor pressure of i (mm Hg) = 438 mm Hg
Ww = weight fraction of water in soil cover = .25
w
Pw = density of water = 1 g/cm3
D = Diffusion coeff. of in i air cm2/sec = .1
6 = .25 (1.56) = .39
1.0
Cs = 438 (85) = 2.01 x 10'3 g/cm3
62,300 (298)
-------
PT = 1 - 1.56 = .41
2.65
Pa = .41 - .39 = .02
E = (!)(.! cm2/sec)(2.01 x 1(T3 g/cm3)(l x 108 cm2)(.02)10/3
(741)2
1 (7.243 x 10"3)
320 cm
E = 5.87 x 10~6 g/sec
= 1.85 x 10'4 Mg/yr
-------
REFERENCES FOR APPENDIX D
1. Tierney, D.R. and T.W. Hughes (Monsanto Research Corporation). Source
Assessment: Reclaiming of Waste Solvents, State-of-the-Art. Prepared for
U.S. Environmental Protection Agency. Washington, DC. Publication
No. EPA-600/12-78-004. August 1977. Section 4.3.
2. U.S. Environmental Protection Agency: Office of Air Quality Planning and
Standards. Compilation of Air Pollutant Emission Factors, Third Edition.
AP-42. Research Triangle Park, NC. August 1977. Section 4.3.
3. U.S. Environmental Protection Agency: Office of Air Quality Planning and
Standards. Compilation of Air Pollution Emission Factors, Third Edition.
AP-42. Research Triangle Park, NC. August 1977. Section 9.1-10.
4. Scheible, M. , G. Shiroma, and G. O'Brien. An Assessment of the Volatile
and Toxic Organic Emissions from Hazardous Waste Disposal in California.
State of California Air Resources Board. February 1982. p. D-4.
5. U.S. Environmental Protection Agency: Office of Air Quality Planning and
Standards. Control of Volatile Organic Compound Leaks from Synthetic
Organic Chemical and Polymer Manufacturing Equipment. Publication
No. EPA-450/3-83-006. March 1984. Research Triangle, NV. 262p.
6. Engineering-Science. National Air Emissions from Tank and Container
Storage and Handling Operations at Hazardous Waste Treatment, Storage and
Disposal Facilities. Prepared for U.S. Environmental Protection Agency:
Office of Air Quality Planning and Standards. Washington, DC.
September 1984. Chapter 5.
7. Hwang, Seong T. Toxic Emissions From Land Disposal Facilities.
Environmental Progress. 1^:^6-52. February 1982.
-------
APPENDIX E
-------
APPENDIX E
DEVELOPMENT OF UNIT RISK FACTORS
INTRODUCTION
Overview
The quantitative expressions of public cancer risks presented in this
report are based on (1) a dose-response model that numerically relates the
degree of exposure to certain airborne volatile organic compounds (VOC) to the
risk of getting cancer, and (2) numerical expressions of public exposure to
ambient air concentrations of VOC estimated to be caused by emissions from
TSDFs. Each of these factors is discussed briefly below and details are
provided in the following sections of this Appendix.
The Relationship of Exposure to Cancer Risk
The relationship of exposure to the risk of getting lung cancer is derived
from epidemiological studies in occupational settings rather than from studies
of excess cancer incidence among the public. The epidemiological methods that
have successfully revealed associations between occupational exposure and
cancer for substances such as asbestos, benzene, vinyl chloride, inorganic
arsenic and ionizing radiation, as well as for certain VOC, are not readily
applied to the public sector, with its increased number of confounding
variables, much more diverse and mobile exposed population, lack of
consolidated medical records, and almost total absence of historical exposure
data. Given such uncertainties, EPA considers it improbable that any
association, short of very large increases in cancer, can be verified in the
general population with any reasonable certainty by an epidemiological study.
Furthermore, as noted by the National Academy of Sciences (NAS), ". . . when
there is exposure to a material, we are not starting at an origin of zero
cancers. Nor are we starting at an origin of zero carcinogenic agents in our
environment. Thus, it is likely that any carcinogenic agent added to the
environment will act by a particular mechanism on a particular cell population
-------
that is already being acted on by the same mechanism to induce cancers." In
discussing experimental dose-response curves, the NAS observed that most
information on carcinogenesis is derived from studies of ionizing radiation
with experimental animals and with humans which indicate a linear no-threshold
dose-response relationship at low doses. They added that although some
evidence exists for thresholds in some animal tissues, by and large,
thresholds have not been established for most tissues. NAS concluded that
establishing such low-dose thresholds ". . . would require massive, expensive,
and impractical experiments ..." and recognized that the U.S. population
". . . is a large, diverse, and genetically heterogeneous group exposed to a
large variety of toxic agents." This fact, coupled with the known genetic
variability to carcinogenesis and the predisposition of some individuals to
some form of cancer, makes it extremely difficult, if not impossible, to
identify a threshold.
For these reasons, EPA has taken the position, shared by other Federal
regulatory agencies, that in the absence of sound scientific evidence to the
contrary, carcinogens should be considered to pose some cancer risk at any
exposure level. This no-threshold presumption is based on the view that as
little as one molecule of a carcinogenic substance may be sufficient to
transform a normal cell into a cancer cell. Evidence is available from both
the human and animal health literature that cancers may arise from a single
transformed cell. Mutation research with ionizing radiation in cell cultures
indicates that such a transformation can occur as the result of interaction
with as little as a single cluster of ion pairs. In reviewing the available
data regarding carcinogenicity, EPA found no compelling scientific reason to
abandon the no-threshold presumption for VOC emitted from TSDFs.
In developing the exposure-risk relationship for VOC emitted from TSDFs,
EPA has assumed that a linear no-threshold relationship exists at and below
the levels of exposure reported in the epidemiological studies of occupational
exposure. (It should be noted that no occupational studies have been
conducted for hazardous waste treatment, storage, and disposal facilities.
The studies referred to here are for other industries emitting TSDF
-------
pollutants.) This means that any exposure to VOC is assumed to pose some risk
of lung cancer and that the linear relationship between cancer risks and
levels of public exposure is the same as that between cancer risks and levels
of occupational exposure.
The numerical constant that defines the exposure-risk relationship used
by EPA in its analysis of carcinogens is called the unit risk estimate. The
unit risk estimate for an air pollutant is defined as the lifetime cancer risk
occurring in a hypothetical population in which all individuals are exposed
continuously from birth throughout their lifetimes (about 70 years) to a
3
concentration of 1 ug/m of the agent in the air which they breathe. Unit
risk estimates are used for two purposes: (1) to compare the carcinogenic
potency of several agents with each other, and (2) to give a crude indication
of the public health risk which might be associated with estimated air
exposure to these agents. The comparative potency of different agents is more
reliable when the comparison is based on studies of like populations and on
the same route of exposure, preferably inhalation.
Publie EXPOsure
The unit risk estimate is only one of the factors needed to produce
quantitative expressions of public health risks. Another factor needed is a
numerical expression of public exposure, i.e., of the numbers of people
exposed to the various concentrations of pollutants. The difficulty of
defining public exposure was noted by the National Task Force on Environmental
Cancer and Health and Lung Disease in their 5th Annual Report to Congress, in
2
1982. They reported that "... a large proportion of the American
population works some distance away from their homes and experience types of
pollution in their homes, on the way to and from work, and in the workplace.
Also, the American population is quite mobile, and many people move every few
years." They also noted the necessity and difficulty of dealing with
long-term exposures because of ". . . the long latent period required for the
development and expression of neoplasia [cancer]. . . ."
-------
EPA's numerical expression of public exposure is based on two estimates.
The first is an estimate of the magnitude and location of long-term average
ambient air concentrations of VOC in the vicinity of emitting TSDFs based on
dispersion modeling using long-term estimates of TSDF emissions and
meteorological conditions. The second is an estimate of the number and
distribution of people living in the vicinity of TSDFs based on Bureau of
Census data which "locates" people by population centroids in census tract
areas. The people and concentrations are combined to produce numerical
expressions of public exposure by an approximating technique contained in a
computerized model.
Public Cancer Risks
By combining numerical expressions of public exposure with the unit risk
estimate, two types of numerical expressions of public cancer risks are
produced. The first, called individual risk, relates to the person or persons
estimated to live in the area of highest concentration as estimated by the
dispersion model. Individual risk is expressed as "maximum lifetime risk."
As used here, the word "maximum" does not mean the greatest possible risk of
cancer to the public. It is based only on the maximum exposure estimated by
the procedure used. The second, called aggregate risk, is a summation of all
the risks to people estimated to be living within the vicinity (usually within
50 kilometers) of a source and is customarily summed for all the sources in a
particular category. The aggregate risk is expressed as incidences of cancer
among all of the exposed population after 70 years of exposure; for
statistical convenience, it is often divided by 70 and expressed as cancer
incidences per year.
There are also risks of nonfatal cancer and of serious genetic effects,
depending on which organs receive the exposure. No numerical expressions of
such risks have been developed; however, EPA considers all of these risks when
making regulatory decisions on limiting emissions of VOC emitted from TSDFs.
-------
UNIT RISK
Procedures for Determination of Unit Risk for Specific Chemicals
The data used for quantitative estimation are one or both of two types:
(1) lifetime animal studies, and (2) human studies where excess cancer risk
has been associated with exposure to the agent. In animal studies it is
assumed, unless evidence exists to the contrary, that if a carcinogenic
response occurs at the dose levels used in the study, then responses will also
occur at all lower doses with an incidence determined by the extrapolation
model.
There is no solid scientific basis for any mathematical extrapolation
model that relates carcinogen exposure to cancer risks at the extremely low
concentrations that must be dealt with in evaluating environmental hazards.
For practical reasons such low levels of risk cannot be measured directly
either by animal experiments or by epidemiologic studies. We must, therefore,
depend on our current understanding of the mechanisms of carcinogenesis for
guidance as to which risk model to use. At the present time the dominant view
of the carcinogenic process involves the concept that most agents that cause
cancer also cause irreversible damage to DNA. This position is reflected by
the fact that a very large proportion of agents that cause cancer are also
mutagenic. There is reason to expect that the quantal type of biological
response, which is characteristic of mutagenesis, is associated with a linear
non-threshold dose-response relationship. Indeed, there is substantial
evidence from mutagenicity studies with both ionizing radiation and wide
variety of chemicals that this type of dose-response model is the appropriate
one to use. This is particularly true at the lower end of the dose-response
curve; at higher doses, there can be an upward curvature, probably reflecting
the effects of multistage processes on the mutagenic response. The linear
non-threshold dose-response relationship is also consistent with the
relatively few epidemiologic studies of cancer responses to specific agents
that contain enough information to make the evaluation possible (e.g.,
radiation-induced leukemia, breast and thyroid cancer, skin cancer induced by
-------
arsenic in drinking water, liver cancer induced by aflatoxins in the diet) .
There is also some evidence from animal experiments that is consistent with
the linear non-threshold model (e.g., liver tumors induced in mice by
2-acetylaminof luorene in the large scale EDQ.. study at the National Center for
Toxicological Research and the initiation stage of the two-stage
carcinogenesis model in rat liver and mouse skin) .
Because its scientific basis, although limited, is the best of any of the
current mathematical extrapolation models, the linear non-threshold model has
been adopted as the primary basis for risk extrapolation to low levels of the
dose-response relationship.
The mathematical formulation chosen to describe the linear non-threshold
dose-response relationship at low doses is the linearized multistage model.
This model employs enough arbitrary constants to be able to fit almost any
monotonically increasing dose-response data, and it incorporates a procedure
for estimating the largest possible linear slope (in the 95% confidence limit)
at low extrapolated doses that is consistent with the data at all dose levels
of the experiment.
Animals —
Description of the low-dose animal extrapolation model. Let P(d) represent
the lifetime risk (probability) of cancer at dose d. The multistage model has
the form
P(d) = 1 - exp [-(qQ
where
0, 1, 2, .... k
-------
Equivalently,
Pt(d) = 1 - exp [-(qjd + q2d2 + ... +
where
P(d) - P(0)
P (d) =
1 - P(0)
is the extra risk over background rate at dose d or the effect of treatment.
The point estimate of the coefficients q., i = 0, 1,2, ...,k, and
consequently the extra risk function P (d) at any given dose d, is calculated
by maximizing the likelihood function of the data.
The point estimate and the 95% upper confidence limit of the extra risk,
P (d), are calculated by using the computer program GLOBAL79, developed by
Crump and Watson (1979). At low doses, upper 95% confidence limits on the
extra risk and lower 95% confidence limits on the dose producing a given risk
are determined from a 95% upper confidence limit, q., on parameter q1 .
Whenever q. > 0, at low doses the extra risk P..(d) has approximately the form
*
P (d) = q, x d. Therefore, q. x d is a 95% upper confidence limit on the
*
extra risk and R/q. is a 95% lower confidence limit on the dose producing an
extra risk of R. Let L be the maximum value of the log-likelihood function.
* *
The upper limit, q., is calculated by increasing q1 to a value q, such that
*
when the log-likelihood is remaximized subject to this fixed value, q. , for
the linear coefficient, the resulting maximum value of the log-likelihood L.
satisfies the equation
2 (LQ - Lj) = 2.70554
where 2.70554 is the cumulative 90% point of the chi-square distribution with
one degree of freedom, which corresponds to a 95% upper-limit (one-sided).
This approach of computing the upper confidence limit for the extra risk P (d)
-------
is an improvement on the Crump et al. (1977) model. The upper confidence
limit for the extra risk calculated at low doses is always linear. This is
conceptually consistent with the linear non-threshold concept discussed
earlier. The slope, q.. , is taken as an upper-bound of the potency of the
chemical in inducing cancer at low doses. (In the section calculating the
risk estimates, P (d) will be abbreviated as P.)
In fitting the dose-response model, the number of terms in the polynomial
is chosen equal to (h-1), where h is the number of dose groups in the
experiment including the control group.
Whenever the multistage model does not fit the data sufficiently well,
data at the highest dose is deleted, and the model is refit to the rest of the
data. This is continued until an acceptable fit to the data is obtained. To
determine whether or not a fit is acceptable, the chi-square statistic
N.P.
is calculated where N. is the number of animals in the i dose group, X. is
^ th
the number of animals in the i dose group with a tumor response, P. is the
probability of a response in the i dose group estimated by fitting the
multistage model to the data, and h is the number of remaining groups. The
2
fit is determined to be unacceptable whenever X is larger than the cumulative
99% point of the chi-square distribution with f degrees of freedom, where f
equals the number of dose groups minus the number of non-zero multistage
coefficients.
Selection of data. For some chemicals, several studies in different animals
species, strains, and sexes, each run at several doses and different routes of
exposure, are available. A choice must be made as to which of the data sets
from several studies to use in the model. It may also be appropriate to
correct for metabolism .differences between species and absorption factors via
-------
different routes of administration. The procedures used in evaluating these
data are consistent with the approach of making a maximum-likely risk
estimate. They are listed as follows.
1. The tumor incidence data are separated according to organ sites or
tumor types. The set of data (i.e., dose and tumor incidence) used in the
model is the set where the incidence is statistically significantly higher
than the control for at least one test dose level and/or where the tumor
incidence rate shows a statistically significant trend with respect to dose
level. The data set that gives the highest estimate of the lifetime
carcinogenic risk, q, 0 , is selected inmost cases. However, efforts are made
to exclude data sets that produce spuriously high risk estimates because of a
small number of animals. That is, if two sets of data show a similar
dose-response relationship, and one has a very small sample size, the set of
data that has the larger sample size is selected for calculating the
carcinogenic potency.
2. If there are two or more data sets of comparable size that are
identical with respect to species, strain, sex, and tumor sites, the geometric
*
mean of q., estimated from each of these data sets, is used for risk
assessment. The geometiic mean of numbers A. , A~, ... , A is defined as
(A x A_ x ... x A ) m
1 / m
3. If two or more significant tumor sites are observed in the same
study, and if the data are available, the number of animals with at least one
of the specific tumor sites under consideration is used as incidence data in
the model.
Calculation of human equivalent dosages from animal data. Following the
suggestion of Mantel and Schneiderman (1975), we assume that mg/surface
area/day is an equivalent dose between species. Since, to a close
approximation, the surface area is proportional to the 2/3rd power of the
weight as would be the .case for a perfect sphere, the exposure in mg/day per
-------
2/3rds power of the weight is also considered to be equivalent exposure. In
an animal experiment this equivalent dose is computed in the following manner.
Let
L = duration of experiment
1 = duration of exposure
m = average dose per day in mg during administration of the agent
(i.e., during 1 ), and
W = average weight of the experimental animal
Then, the lifetime average exposure is
Oral. Often, exposures are not given in units of mg/day and it becomes
necessary to convert the given exposures into mg/day. lor example, in most
feeding studies exposure is in terms of ppm in the diet. Similarly in
drinking water studies exposure is in ppm in the water. In these cases the
exposure in mg/day is
m = ppm x F x r
where ppm is parts per million of the carcinogenic agent in the diet or water,
F is the weight of the food or water consumed per day in kg, and r is the
absorption fraction. In the absence of any data to the contrary, r is assumed
to be equal to one. For a uniform diet, the weight of the food consumed is
proportional to the calories required, which in turn is proportional to the
surface area or 2/3rds power of the weight. Water demands are also assumed
proportional to the surface area, so that
2/3
m = ppm x W x r
-------
or
m
As a result, ppm in the diet or in water is often assumed to be an
equivalent exposure between species. However, we feel that this is not
justified since the calories/kg of food are very different in the diet of man
compared to laboratory animals primarily due to moisture content differences.
Consequently, the amount of drinking water required by each species differs
also because of the amount of moisture in the food. Therefore, we use an
empirically-derived factor, f = F/W, which is the fraction of a species body
weight that is consumed per day as food or water. We use the following rates:
Fraction of Body
Weight Consumed as
Species
Man
Rats
Mice
W
70
0.035
0.03
food
C.028
0.05
0.13
water
0.029
0.078
0.17
Thus, when the exposure is given as a certain dietary or water concentration
in ppm, the exposure in mg/w is
m ppm x F ppm x f x W
= -273 ------- -— = X f
When exposure is given in terms of mg/kg/day = m/Wr = s, the conversion is
simply
T
x W
-------
Calculation of the unit risk from animal studies. The 95% upper-limit risk
2/3
associated with d mg/kg /day is obtained from GLOBAL?9 and, for most cases
of interest to risk assessment, can be adequately approximated by
P(d) = 1 - exp-(q.d). A "unit risk" in units X is simply the risk
corresponding to an exposure of X = 1. To estimate this value we simply find
2/3
the number of mg/kg /day corresponding to one unit of X and substitute this
value into the above relationship. Thus, for example, if X is in units of
3 1/3 —3
pg/m in the air, we have that for case 1, d = 0.29 x 70 ' x 10
2/3 3
mg/kg /day, and for case 2, d =1, when ug/m is the unit used to compute
parameters in animal experiments.
If exposures are given in terms of ppm in air, we may simply use the fact
that
molecular weight (gas) mg/m
1 ppm = 1 .2 x
molecular weight (air)
Note, an equivalent method of calculating unit risk would be to use mg/kg/day
for the animal exposures and then increase the j polynomial coefficient by
an amount
(Wh/Wa)j/3 j = 1, 2, .... k
and use mg/kg equivalents for the unit risk values.
Adjustment for less than lifespan duration of experiment. If the duration of
experiment, L , is less than the natural lifespan of the test animal, L, the
if **
slope, q. , or more generally the exponent, g(d), is increased by multiplying a
^ o
factor (L/L ) . We assume that if the average dose, d, is continued, the
age-specific rate of cancer will continue to increase as a constant function
of the background rate. The age-specific rates for humans increase at least
by the 2nd power of the age and often by a considerably higher power as
demonstrated by Doll (1971). Thus, we would expect the cumulative tumor rate
to increase by at least the 3rd power of age. Using this fact, we assume that
-------
the slope, q7 , or more generally the exponent, g(d), would also increase by at
*
least the 3rd power of age. As a result, if the slope q, [or g(d)] is
calculated at age L , we would expect that if the experiment had been
continued for the full lifespan, L, at the given average exposure, the slope
^u <5
qj [or g(d)] would have been increased at least (L/Lg) .
This adjustment is conceptually consistent with the proportional hazard
model proposed by Cox (1972) and the time-to-tumor model considered by Crump
(1979) where the probability of cancer by age t and at dose d is given by
P(d,t) = 1 - exp [-f(t) x g(d)]
Interpretation of quantitative estimates. For several reasons, the unit risk
estimate based on animal bioassays is only an approximate indication of the
absolute risk in populations exposed to known carcinogen concentrations.
First, there are important species differences in uptake, metabolism, and
organ distribution of carcinogens, as well as species differences in target
site susceptibility, immunological responses, hormone function, dietary
factors, and disease. Second, the concept of equivalent doses for humans
compared to animals on a mg/surface area basis is virtually without
experimental verification regarding carcinogenic response. Finally, human
populations are variable with respect to genetic constitution and diet, living
environment, activity patterns, and other cultural factors.
The unit risk estimate can give a rough indication of the relative
potency of a given agent compared with other carcinogens. The comparative
potency of different agents is more reliable when the comparison is based on
studies in the same test species, strain, and sex, and by the same route of
exposure, preferably by inhalation.
The quantitative aspect of the carcinogen risk assessment is included
here because it may be of use in the regulatory decision-making process, e.g.,
setting regulatory priorities, evaluating the adequacy of technology-based
controls, etc. However, it should be recognized that the estimation of cancer
-------
risks to humans at low levels of exposure is uncertain. At best, the linear
extrapolation model used here provides a rough but plausible estimate of the
upper-limit risk; i.e., it is not likely that the true risk would be much more
than the estimated risk, but it could very well be considerably lower. The
risk estimates presented in subsequent sections should not be regarded as an
accurate representation of the true cancer risks even when the exposures are
accurately defined; however, the estimates presented may be factored into
regulatory decisions to the extent that the concept of upper risk limits is
found to be useful.
Alternative methodoloeical approaches. The methods used by the Carcinogen
Assessment Group (CAG) for quantitative assessment are consistently
conservative, i.e., tending toward high estimates of risk. The most important
part of the methodology contributing to this conservatism is the linear
non-threshold extrapolation model. There are a variety of other extrapolation
models that could be used, all of which would give lower risk estimates.
These alternative models have not been used by the CAG. The CAG feels that
with the limited data available from these animal bioassays, especially at the
high-dose levels required for testing, almost nothing is known about the ti'ae
shape of the dose-response curve at low environmental levels. The position is
taken by the CAG that the risk estimates obtained by use of the linear
non-threshold model are plausible upper limits, and that the true risk could
be lower.
In terms of the choice of animal bioassay as the basis for extrapolation,
the general approach is to use the most sensitive responder on the assumption
that humans are as sensitive as the most sensitive animal species tested. The
average response of all of the adequately tested bioassay animals was used;
this is because three well-conducted valid drinking water studies using
different strains of rats showed similar target organs and about the same
level of response.
Extrapolations from animals to humans could also be done on the basis of
relative weights rather than relative surface areas. The latter approach,
-------
used here, has more basis in human pharmacological responses; it is not clear
which of the two approaches is more appropriate for carcinogens. In the
absence of information on this point, it seems appropriate to use the most
generally employed method, which also is more conservative. In the case of
the TCI study, the use of extrapolation based on surface area rather than
weight increases the unit risk estimates by a factor of 12 to 13.
Humans—Model for Estimation of Unit Risk Based on Human Data—
If human epidemiologic studies and sufficiently valid exposure
information are available for the compound, they are always used in some way.
If they show a carcinogenic effect, the data are analyzed to give an estimate
of the linear dependence of cancer rates on lifetime average dose, which is
equivalent to the factor B . If they show no carcinogenic effect when
positive animal evidence is available, then it is assumed that a risk does
exist, but it is smaller than could have been observed in the epidemiologic
study, and an upper limit to the cancer incidence is calculated assuming
hypothetically that the true incidence is below the level of detection in the
cohort studied, which is determined largely by the c.:nort size. Whenever
possible, human data are used in preference to animal bioassay data.
Very little information exists that can be utilized to extrapolate from
high exposure occupational studies to low environmental levels. " However, if a
number of simplifying assumptions are made, it is possible to construct a
crude dose-response model whose parameters can be estimated using vital
statistics, epidemiologic studies, and estimates of worker exposures.
In human studies, the response is measured in terms of the relative risk
of the exposed cohort of individuals compared to the control group. The
mathematical model employed assumes that for low exposures the lifetime
probability of death from lung cancer (or any cancer), P., may be represented
by the linear equation
P0 = A + BRx
-------
where A is the lifetime probability in the absence of the agent, and x is the
average lifetime exposure to environmental levels in some units, say ppm. The
factor, B , is the increased probability of cancer associated with each unit
increase of the agent in air.
If we make the assumption that R, the relative risk of lung cancer for
exposed workers compared to the general population, is independent of the
length or age of exposure but depends only upon the average lifetime exposure,
it follows that
P = A + BH (X;L
P0 A + BH xl
or
RP0 = A + BH (x: + x2)
where x. = lifetime average daily exposure to the agent for the general
population, x. = lifetime average daily exposure to the agent in the
occupational setting, and P. = lifetime probability of dying of cancer with no
or negligible TCI exposure.
Substituting PQ = A + BH x, , and rearranging gives
Pn (R - 1)
BH = -a --------
To use this model, estimates of R and x? must be obtained from the
epidemiologic studies. The value P_ is derived from the age-cause-specific
death rates for combined males found in the 1976 U.S. Vital Statistics tables
using the life table methodology. For lung cancer the estimate of P. is
0.036. This methodology is used in the section on unit risk based on human
studies.
-------
References for Appendix E
1. National Academy of Sciences, "Arsenic", Committee on Medical and
Biological Effects of Environmental Pollutants, Washington, B.C., 1977.
Docket Number (OAQPS 79-8) II-A-3.
2. U.S. EPA, et. al., "Environmental Cancer and Heart and Lung Disease,"
Fifth Annual Report to Congress by the Task Force on Environmental Cancer
and Health and Lung Disease, August, 1982.
-------
APPENDIX F
-------
It National estimates of the total air emissions! and annual incidence an calculated from data
pri?stvntud in tho draft final report, masis balance of throughput volume*.*, anil thw ri'.'.ft
preliminary National Air Emissions report.
Htnraqe Tank
Open Treatment Tanks
.ind Surface Impoundments
1. and treatment
Landfill
Totals
Total Air Emissions <6g/yr>
I II III IV
14 14* 14* 17
BIO 2292 1220 2580
97 82 5 70
a
24 343 29 906
31
945 2731 126B 3573
2690
Tohal Annual Incidence
I II Id
< 1-7 < 1 - /« -:. 1 - / »
< I - 44 2 - 23O 2 - 20O
< i - 1 3 < i -- 6 ; i
< 1 < 1 - 1 < 1 - 1
4 <1 - 240 <1 - 21O
I - Draft Final Report (risk assessment)
II - Mass balance of waste throughput based on BCA estimation of organics in site visit reports
III - Mass balance of waste throughput based on 0.367. volatile organic;;
IV - BCA National Air Emissions report
* - not analyzed by mass balance
a - It is not anticipated these projections are valid) see tent of BCA National Air Emissions report
------- |