&EPA
United States
Environmental Protection
Agency
Office of Research and
Development
Washington DC 20460
EPA600-6-90001
August 1990
Development of Risk
Assessment
Methodology for Surface
Disposal of Municipal
Sludge
-------
EPA/600/6-90/001
August 1990
DEVELOPMENT OF RISK ASSESSMENT METHODOLOGY
FOR SURFACE DISPOSAL OF MUNICIPAL SLUDGE
Environmental Criteria and Assessment Office
Office of Health and Environmental Assessment
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
U s Environmental Protection
*j» $
crt2Sf.it60604-3590
-------
DISCLAIMER
This document has been reviewed in accordance with U.S. Environmental Protection Agency
policy and approved for publication. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
11
-------
PREFACE
This is one of a series of reports that present methodologies for assessing the potential risks
to humans or other organisms from the disposal or reuse of municipal sludge. The sludge
management practices addressed by this series include land application practices, distribution and
marketing programs, landfilling, surface disposal, incineration and ocean disposal. In particular,
these reports provide methods for evaluating potential health and environmental risks from toxic
chemicals that may be present in sludge. This document addresses risks from chemicals associated
with surface disposal of municipal sludge.
These proposed risk assessment procedures are designed as tools to assist in the development
of regulations for sludge management practices. The procedures are structured to allow calculation
of technical criteria for sludge disposal/reuse options based on the potential for adverse health or
environmental impacts. The criteria may address management practices (such as site design or
process control specifications), limits on sludge disposal rates or limits on toxic chemical
concentrations in the sludge.
The methods for criteria derivation presented in this report have been submitted to the U.S.
EPA Office of Water Regulations and Standards (OWRS) to provide scientific background for the
development of technical criteria for toxic chemicals in sludge. However, the methods used by
OWRS to develop regulations could differ from this guidance in some respects; therefore, the
Technical Support Documents provided by OWRS should be consulted for explanation of the
regulations.
in
-------
DOCUMENT DEVELOPMENT
Document Managers
R.J.F. Bruins,
N.E. Kowal
Environmental Criteria and Assessment
Office
Office of Health and Environmental
Assessment
U.S. Environmental Protection Agency
Cincinnati, OH 45268
Authors
K. O'Neal,
J. Weisman,
V.A. Hutson,
S.E. Keane
Abt Associates, Inc.
Cambridge, MA 02138
Dr. S.G. Buchberger
Department of Civil Engineering
University of Cincinnati
Cincinnati, OH 45221
B.H. Lester
GeoTrans, Inc.
Herndon, VA 22070
Reviewers
Dr. Clark Allen
Research Triangle Institute
Research Triangle Park, NC 27709
Dr. Carl Anderson
Department of Agricultural Engineering
Iowa State University
Ames, IA 50011
Dr. Robert M. Sykes
Department of Civil Engineering
Ohio State University
Columbus, OH 43210
Document Preparation
Karen Sweetlow
Dynamac Corporation
Rockville, MD 20852
Judith Olsen
Environmental Criteria and Assessment
Office
Cincinnati, OH 45268
IV
-------
TABLE OF CONTENTS
1.0. INTRODUCTION 1-1
1.1. PURPOSE AND SCOPE 1-1
1.2. DEFINITIONS AND COMPONENTS OF RISK ASSESSMENT 1-2
1.3. RISK ASSESSMENT IN THE METHODOLOGY DEVELOPMENT
PROCESS 1-5
1.3.1. Hazard Identification and Dose-Response Assessment 1-5
1.3.2. Exposure Assessment 1-6
1.3.3. Risk Characterization 1-7
1.4. POTENTIAL USES OF THE METHODOLOGY IN RISK
MANAGEMENT 1-8
1.5. LIMITATIONS OF THE METHODOLOGY 1-9
2.0. DEFINITION OF DISPOSAL PRACTICES 2-1
2.1. INTRODUCTION 2-1
2.2. LAGOONS FOR LONG-TERM STORAGE OF SLUDGE 2-2
2.3. SURFACE IMPOUNDMENTS AS PERMANENT DISPOSAL 2-3
2.4. SLUDGE STORAGE AS A COMPONENT OF WASTEWATER
TREATMENT 2-3
2.5. ASSUMPTIONS USED IN RISK ASSESSMENT 2-3
2.6. CONDITIONS AFFECTING RISK 2-4
2.7. SUMMARY 2-5
3.0. EXPOSURE PATHWAYS AND MOST-EXPOSED INDIVIDUALS (MEIs) 3-1
3.1. INTRODUCTION 3-1
3.2. GROUNDWATER PATHWAY 3-4
3.3. SURFACE WATER PATHWAY 3-4
3.4. VOLATILIZATION PATHWAY 3-5
4.0. DERIVATION OF CRITERIA FOR THE GROUNDWATER PATHWAY 4-1
4.1. OVERVIEW OF METHOD 4-1
4.2. ASSUMPTIONS 4-3
4.3. CALCULATIONS 4-7
4.3.1. Source Term 4-7
4.3.2. Unsaturated Zone Flow and Transport 4-12
4.3.3. Contaminant Transport in the Saturated Zone 4-20
4.4. INPUT PARAMETER REQUIREMENTS 4-25
4.4.1. Source Term 4-25
4.4.2. Unsaturated Zone 4-26
4.4.3. Saturated Zone 4-26
4.5. HEALTH AND ENVIRONMENTAL EFFECTS 4-27
4.5.1. Threshold-Acting Toxicants 4-27
4.5.2. Carcinogens 4-34
-------
4.6. DERIVING CRITERIA 4-37
4.7. SAMPLE CALCULATIONS 4-38
4.7.1. Analysis of Exposure for the Most Exposed Individual 4-40
4.7.2. Analysis of Exposure for the Most Exposed Populations 4-72
5.0. DERIVATION OF CRITERIA FOR THE SURFACE WATER PATHWAY 5-1
5.1. OVERVIEW OF THE METHOD 5-1
5.2. ASSUMPTIONS 5-5
5.3. CALCULATIONS 5-8
5.4. INPUT PARAMETER REQUIREMENTS 5-14
5.4.1. Mean Annual River Flow 5-14
5.4.2. Cell Area 5-14
5.4.3. Net Contaminant Loss Rate 5-15
5.4.4. Hydraulic Characteristics 5-22
5.4.5. Wind Speed 5-23
5.5. HEALTH AND ENVIRONMENTAL EFFECTS 5-23
5.5.1. Aquatic Life Protection 5-25
5.5.2. Threshold-Acting Toxicants 5-25
5.5.3. Carcinogens 5-32
5.6. SAMPLE CALCULATIONS 5-32
5.6.1. Analysis of Exposure for the Most Exposed Individual 5-32
5.6.2. Analysis of Exposure for the Most Exposed Populations 5-62
6.0. DERIVATION OF CRITERIA FOR THE AIR 6-1
6.1. OVERVIEW OF THE METHOD 6-1
6.2. ASSUMPTIONS 6-2
6.3. CALCULATIONS 6-5
6.3.1. Influent and Effluent Flow 6-6
6.3.2. Contaminant Mass Lost to Biodegradation 6-7
6.3.3. Contaminant Mass Lost to Volatilization 6-8
6.3.4. Contaminant Mass Lost to Volatilization from Diffused Air . . . 6-9
6.3.5. Change in Total Contaminant Mass Contained Within the
Lagoon 6-9
6.3.6. Mass of Contaminant Transferred from Adsorbed to
Dissolved Phase 6-10
6.3.7. Estimation of Volatile Emissions 6-12
6.3.8. Estimation of Wind Transport 6-20
6.3.9. Deriving Criteria 6-21
6.4. INPUT PARAMETER REQUIREMENTS 6-26
6.4.1. Chemical Characteristics 6-26
6.4.2. Site Characteristics 6-26
6.4.3. Meteorological Conditions 6-26
VI
-------
6.5. HEALTH AND ENVIRONMENTAL EFFECTS 6-27
6.6.1. Threshold-Acting Toxicants 6-27
6.6.2. Carcinogens 6-30
6.6. SAMPLE CALCULATIONS 6-30
6.5.1. Analysis of Exposure for the Most Exposed Individual 6-30
6.5.2. Analysis of Exposure for the Most Exposed Populations 6-55
7.0. SIMULTANEOUS CONSIDERATION OF MULTIPLE PATHWAYS
OF EXPOSURE 7-1
8.0. REFERENCES 8-1
vn
-------
LIST OF TABLES
No. Title Page
4-1 Assumptions for Methodology to Analyze the Groundwater Pathway 4-4
4-2 Summary of Measured Seepage Rates From Municipal Lagoon Systems .... 4-9
4-3 Water Content of Sludges from Various Treatment Processes 4-11
4-4 Water Ingestion and Body Weight by Age-Sex Group in the
United States 4-29
4-5 Daily Intakes of Drinking Water by Adults 4-31
4-6 Illustrative Categorization of Evidence Based on Animal and
Human Data 4-35
4-7 Input Parameters for VADOFT Simulation of Flow and Contaminant
Transport Through the Unsaturated Zone: Antrim, New Hampshire 4-42
4-8 Results of VADOFT Execution: Antrim, New Hampshire 4-45
4-9 Input Parameters for AT123D Simulation of Contaminant Transport
Through the Saturated Zone: Antrim, New Hampshire 4-46
4-10 Results from AT123D Simulation of Contaminant Transport Through
the Saturated Zone: Antrim, New Hampshire 4-47
4-11 Input Parameters for AT123D Simulation of Contaminant Transport
Through the Saturated Zone to Surface Water:
Antrim, New Hampshire 4-50
4-12 Results from AT123D Simulation of Contaminant Transport Through
the Saturated Zone to Surface Water: Antrim, New Hampshire 4-51
4-13 Input Parameters for VADOFT Simulation of Flow and Contaminant
Transport Through the Unsaturated Zone: Tulsa, Oklahoma 4-54
4-14 Results from VADOFT Execution: Tulsa, Oklahoma 4-55
4-15 Input Parameters for AT123D Simulation of Contaminant Transport
Through the port Through the Saturated Zone: Tulsa, Oklahoma 4-57
4-16 Results from AT123D Simulation of Contaminant Transport Through
the Saturated Zone to Surface Water: Tulsa, Oklahoma 4-58
4-17 Input Parameters for AT123D Simulation of Contaminant Transport
Through the Saturated Zone to Surface Water: Tulsa, Oklahoma 4-61
4-18 Results from AT123D Simulation of Contaminant Transport
Through the Saturated Zone to Surface Water: Tulsa, Oklahoma 4-62
4-19 Input Parameters for VADOFT Simulation of Flow and Contaminant
Transport Through the Unsaturated Zone: Portland, Oregon 4-64
vin
-------
LIST OF TABLES (cont.)
No. Title Page
4-20 Results from VADOFT Execution: Portland, Oregon 4-65
4-21 Input Parameters for AT123D Simulation of Contaminant Transport
Through the Saturated Zone: Portland, Oregon 4-66
4-22 Results from AT123D Simulation of Contaminant Transport Through
the Saturated Zone: Portland, Oregon 4-68
4-23 Input Parameters for AT123D Simulation of Contaminant Transport
Through the Saturated Zone to Surface Water: Portland, Oregon 4-70
4-24 Results from AT123D Simulation of Contaminant Transport Through
the Saturated Zone to Surface Water: Portland, Oregon 4-71
5-1 Assumptions for Methodology to Analyze Surface Water Pathways 5-6
5-2 Properties of Selected Organic Chemicals 5-18
5-3 U.S. Annual Per-Capita Consumption of Fish and
Shellfish 1960-1984 5-30
5-4 Fish Consumption by Demographic Variables 5-33
5-5 Chemical Properties of Benzene and Lead 5-34
5-6 Site-Specific Parameters: Contoocook River, New Hampshire 5-39
5-7 Model Results for Benzene: Contoocook River, New Hampshire 5-45
5-8 Model Results for Lead: Contoocook River, New Hampshire 5-46
5-9 Site-Specific Input Parameters: Bird Creek, Oklahoma 5-53
5-10 Model Results for Benzene: Bird Creek, Oklahoma 5-55
5-11 Model Results for Lead: Bird Creek, Oklahoma 5-56
5-12 Site-Specific Input Parameters: Columbia Slough, Oregon 5-59
5-13 Model Results for Benzene: Columbia Slough, Oregon 5-60
5-14 Model Results for Lead: Columbia Slough, Oregon 5-61
6-1 Assumptions Required for Methodology to Analyze the Air Pathway 6-3
6-2 Parameter Values Used to Calculate CTZ 6-22
6-3 Daily Respiratory Volumes for "Reference" Individuals (Normal
Individuals of Typical Activity Levels) and for Adults with
Higher-than-Normal Respiratory Volume or Higher-than-Normal
Activity Levels 6-29
IX
-------
LIST OF TABLES (cont.)
No. Title Page
6-3 Total Air Inhalation Rates 6-29
6-4 Input Parameters for Estimating Emissions of Benzene 6-31
6-5 Derivation of Criteria for the Air Pathway: Antrim, New Hampshire 6-36
6-6 Input Parameters for Execution of ISCLT 6-38
6-7 Derivation of Criteria for the Air Pathway: Tulsa, Oklahoma 6-50
6-8 Derivation of Criteria for the Air Pathway: Portland, Oregon 6-54
-------
LIST OF FIGURES
No. Title Page
1-1 Relationship of Risk Assessment Methodology to Other Components of
Regulation Development for Sewage Sludge Reuse/Disposal Options 1-3
5-1 Definition Sketch for Cascading Cells Model 5-9
5-2 Mean Annual Wind Speed in the United States 5-24
5-3 Longitudinal Profile of Contoocook River Study Reach 5-36
5-4 Three Cell Cascade for Contoocook River between Antrim and
Hillsboro, NH 5-38
5-5 Typical Cross Section of Contoocook River 5-40
5-6 Five Cell Cascade for Bird Creek Near Tulsa, OK 5-54
5-7 Five Cell Cascade for Columbia Slough Near Portland, OR 5-58
5-8 Ratio of Exposure to Sludge Concentration By Size of MEP Near
Atrim, NH 5-64
5-9 Ratio of Exposure to Sludge Concentration By Size of MEP Near
Tulsa, OK 5-66
5-10 Ratio of Exposure to Sludge Concentration By Size of MEP Near
Portland, OR 5-68
6-1 Cumulative Distribution of Population by Distance from Facility:
Antrim, NH 6-57
6-2 Air Concentration Per Unit Emissions for Selected Directions
Antrim, NH 6-58
6-3 Transport Ratio by Population: Antrim, New Hampshire 6-59
6-4 Cumulative Distribution of Population by Distance from Facility
Tulsa, OK 6-61
6-5 Air Concentration Per Unit Emissions for Selected Directions
Tulsa, OK 6-62
6-6 Transport Ratio by Population: Tulsa, Oklahoma 6-63
6-7 Cumulative Distribution of Population by Distance from Facility:
Portland, OR 6-64
6-8 Air Concentration Per Unit Emissions for Selected Directions
Portland, OR 6-65
6-9 Transport Ratio by Population: Portland, Oregon 6-66
XI
-------
LIST OF ABBREVIATIONS
T Cumulative transport factor for cascade model (unitless)
If Transport factor for cell in cascade model (unitless)
6 Temperature correction factor (unitless)
CTZ Standard deviation of contaminant plume above the ground (m)
OL Longitudinal dispersivity (m)
T) Effective water storage capacity (m"1)
Effective porosity (unitless)
A First-order contaminant loss due to atmospheric decay and deposition
(unitless)
A First-order decay constant (sec"1)
Ma Viscosity of air (g/cm-sec)
A*w Viscosity of water (g/cm-sec)
V> Pressure head (m)
V>a Air entry pressure head (m)
pb Bulk density of the wet soil (g/m3)
p Air density (g/m3)
a
pw Water density (g/m3)
6 Volumetric water content (unitless)
7Q10 Minimum average stream flow expected to occur once every 10 years
(m3/sec)
A Area of air-exposed surface of water (m2)
AC Upper-bound estimate of contaminant concentration in air at property
boundary (g/m3)
AWQC Ambient Water Quality Criteria
B Width of cell in surface water model (m)
BCF Unadjusted bioconcentration factor in fish (//kg)
BCF Adjusted bioconcentration factor (//kg)
xn
-------
BI Background intake of pollutant from a given exposure route (mg/day)
BW Human body weight (kg)
C Contaminant concentration (g/m )
c Concentration of solute in unsaturated zone (g/m3)
Cb Background concentration of contaminant in the environmental medium
(mg// or mg/m3)
Cd Concentration (dissolved form) (g/m3)
CE Concentration of contaminant in effluent (g/m3)
Cf Fluid compressibility (m-sec2/g)
C Concentration in groundwater recharged to river (g/m3)
Cou Groundwater concentration of contaminant (mg//)
gw
Ct Contaminant concentration in liquid (g/m3)
Cmax Maximum allowable contaminant concentration in impoundment (mg//)
CQ Inlet contaminant concentration (g/m3)
C Concentration (particulate form) (g/m3)
Cs Contaminant concentration in solid (g/kg)
Cu Concentration in river upstream of sludge disposal site (g/m3)
d Aquifer depth (m)
D Dispersion coefficient (m2/sec)
D* Effective molecular diffusion coefficient (m2/sec)
{D} Dispersion tensor (m2/sec)
Dca Diffusivity of contaminant in air (cm2/sec)
Daf Anti-dilution factor (unitless)
d Effective distance (m)
e
>eth
Df Dilution factor (unitless)
D ,.. Diffusivity of ether in water (cm2/sec)
D Diffusivity of contaminant in water (cm /sec)
cw
2>
E Maximum allowable air emission of contaminant (g/sec)
Xlll
-------
E
Fa
FD
oc
GEMS
H
Hc
He
HETOT
HHAG
I.
ICRP
If
ISCLT
\
K
k
kagg
Emission rate of contaminant to air from site surface (g/sec)
Volume of fluid passing through a vertical cross-section of an aquifer
oriented perpendicular to the direction of flow (m3/sec)
Fetch-to-depth ratio (unitless)
Dissolved fraction of contaminant (unitless)
Frequency of specific stability array parameters for class indication i,j
(stability class, windspeed)
Weight fraction of organic carbon in total solids (unitless)
Volume of fluid leaving surface impoundment (m3/sec)
Gravitational acceleration (m/sec2)
Graphical Exposure Modeling System
Mean flow depth (m)
Henry's law constant for the contaminant (atm-m3/mol)
Dimensionless Henry's law constant (unitless)
Total human exposure to contaminant (mg-day/kg)
Human Health Assessment Group
Impoundment depth (m)
Effective height (m)
Total air inhalation rate (m3/day)
International Commissions on Radiological Protection
Total fish ingestion rate (kg/day)
Industrial Source Complex Long-Term
Total water ingestion rate (day"1)
Vertical hydraulic conductivity (m/sec)
Overall mass transfer coefficient for volatilization (m/sec)
"Aggregate" decay rate, or the arithmetic sum of k , kh and kb (sec )
First-order biodegradation constant (sec"1)
Equilibrium distribution coefficient for concentration (//kg)
Loss rate due to contaminant decay (m/sec)
xiv
-------
K
ow
rw
ks
ksl
L
LCd
LCe
LGW
M
m
MCLG
ME
MEI
MEP
MEU
MOD
MI
Mo
MSL
AMT
M,
VA
Gas phase mass transfer coefficient (m/sec)
Loss rate due to hydrolysis (sec"1)
Liquid phase mass transfer coefficient (m/sec)
Octanol-water partition coefficient (unitless)
Loss rate due to photolysis (sec"1)
Effective permeability (unitless)
Loss rate due to interaction with sediment (sec"1)
Equilibrium distribution coefficient for mass (unitless)
Net contaminant loss rate (m/sec)
Cell length (m)
Lipid content of dietary seafood (kg/kg)
Lipid content of experimental organism (kg/kg)
Contaminant loadings to groundwater (g/sec)
Longitudinal mixing distance (m)
Mass flux (g/m2-sec)
Mass of contaminant (g)
Mass of contaminant removed by biodegradation (g/sec)
Maximum Contaminant Level Goal
Mass of contaminant removed by effluent or seepage (g/sec)
Most exposed individual
Most exposed population
Most exposed unit
Million gallons per day
Mass of contaminant entering the facility (g/sec)
Mass of contaminant leaving facility (g/sec)
Mass transfer between solid and liquid phases (g)
Change in total contaminant mass within the facility (g/sec)
Mass of contaminant removed by volatilization due to wind (g/sec)
xv
-------
MVD Mass of contaminant removed by volatilization of resulting from diffused
air (g/sec)
N. Total concentration of contaminant i in sludge (mg/kg or g/Mg, dry
weight)
Nmax Maximum allowable concentration of contaminant in sludge (mg/kg or
g/Mg, dry weight)
n Manning's roughness coefficient (unitless)
NR»C National Research Council
OHEA Office of Health and Environmental Assessment
OWRS Office of Water Regulations and Standards
Q Liquid volumetric flow rate (m3/sec)
Qa Volumetric flow rate of diffused air (m3/sec)
qd Discharge per unit width (m2/sec)
Qf Mean annual flow (m3/sec)
Q Groundwater discharge to river (m3/sec)
q,- Infiltration rate from impoundment (m/sec)
Qs Quantity of sludge entering facility (Mg/year, dry weight)
Qu Upstream discharge (m3/sec)
q,,* Human cancer potency ((mg-day/kg)"1))
r Radius of circle with same area as impoundment (m)
R Ideal gas constant = 8.206xlO"5 (m3-atm/mol-K)
RF Retardation factor in groundwater (unitless)
RAC Reference air concentration (mg/m3)
RfD Reference Dose (mg/day-kg)
RE Relative effectiveness of exposure route (unitless)
RL Risk level (unitless)
rgl Ratio of total solids to liquid in flow (kg//)
RWC Reference water concentration (mg/p
S Longitudinal slope of the cell (unitless)
Sc Schmidt number on gas side (unitless)
xvi
-------
Sc, Schmidt number on liquid side (unitless)
Se Effective water saturation (unitless)
Ss Specific storage (unitless)
SH Water saturation (unitless)
Swr Residual water saturation (unitless)
SRRQU Source-receptor ratio for groundwater (kg//)
SRRSU Source-receptor ratio for surface water (kg//)
SRRVOL Source-receptor ratio for volatilization (kg/m3)
STAR Stability array
T Temperature (°K)
t Time (sec)
TBI Total background intake rate of contaminant (mg/day) from all other
sources of exposure
U10 Mean windspeed at 10 meters above lagoon surface (m/sec)
U* Friction velocity
Us Mean flow velocity of water (m/sec)
UH Mean windspeed (m/sec)
V Volume of liquid (m3)
VD Darcy velocity (m/sec)
V. Regional velocity of horizontal groundwater flow (m/sec)
Vj Increase in groundwater velocity due to the impoundment (m/sec)
VT Transport velocity (m/sec)
Vy Vertical velocity due to the impoundment (m/sec)
W Contaminant mass loading (g/sec)
w Width of impoundment perpendicular to direction of flow (m)
X Downstream distance (m)
Xa Distance in the x-coordinate direction (parallel to wind velocity, u .) from
source to point of interest (m)
Xj Maximum expected dissolved concentration of contaminant i (mg//)
xvii
-------
x . Virtual distance required for point source plume to spread to width of site
(m)
X Distance from center of impoundment to surface water (m)
sw
Ya Distance in the y-coordinate direction (perpendicular to wind velocity Ug,)
from source to point of interest (m)
z Vertical coordinate in the unsaturated zone (m)
Za Distance in the z-coordinate direction (perpendicular to wind velocity,
U ,) from source to point of interest (m)
W J
xvin
-------
1. INTRODUCTION
1.1. PURPOSEANDSCOPE
This report is one in a series of documents that describe methodologies used to evaluate the
potential health and environmental risks resulting from the management (i.e., disposal or reuse) of
municipal sewage sludge, including application to land for beneficial purposes, landfilling,
incineration and ocean dumping. This document addresses health and environmental risks associated
with human and non-human exposure to chemical contaminants in sludge from the surface disposal
of municipal sludge, and explains the development of a methodology by which criteria for sludge
surface disposal mav be derived.
The term "surface disposal" will refer in this document to the permanent disposal or long-
term storage of sludge in uncovered lagoons or impoundments. Short-term storage of sludge during
wastewater treatment (i.e., less than one year) is not included in this definition. The land application
of sludge, the distribution and marketing of sludge and the disposal of sludge in covered landfills,
are also excluded from this definition. Methodologies for deriving criteria for these sludge
management practices are described by other documents in this series (U.S. EPA 1986a, 1989d). In
addition, disposal or long term storage of sludge in waste piles, which might also be described
appropriately as "surface disposal," is not included in the present discussion.
The risk evaluation methods described in this document are intended to aid in the
development of regulations for the surface disposal of sludge. The procedures are structured to allow
calculation of technical criteria for regulating sludge disposal/reuse, based on potential adverse health
and environmental impacts. These criteria may include restrictions on the concentrations of chemical
contaminants in sludge and restrictions governing the design and management of sludge disposal sites.
Restrictions on concentrations of chemical contaminants consist of maximum allowable dry-weight
concentrations for each contaminant in the sludge.
The methods for deriving criteria presented in this report are intended to be used by the
Office of Water Regulations and Standards (OWRS) to develop national technical criteria for toxic
chemicals in municipal sludge. This document is not intended to be a manual to guide users in
1-1
-------
estimating potential risks from a particular sludge management site. A user's manual based on the
methods described in this document should be developed separately.
This report is part of a series of methodology background documents. This series does not
address potential health risks from the presence of pathogenic organisms in municipal sludge, which
are examined in separate U.S. EPA analyses. These documents also do not address health and
ecological risks from the treatment, transport, handling or accidental release of sludge.
1.2. DEFINITIONS AND COMPONENTS OF RISK ASSESSMENT
The National Research Council (NRC) defines risk assessment as the process of characterizing
the potential adverse human health or environmental effects of exposures to environmental hazards
(NRC, 1983). Risk management, on the other hand, is defined as the process of evaluating alternative
regulatory actions that may be taken to reduce risk and choosing among them, based on consideration
of costs, availability of technologies, and other factors.
NRC defines four components of risk assessment. The first step is hazard identification. In
this process, relevant data are gathered and assessed to evaluate whether exposure to a particular
agent poses a health or environmental hazard. The next step, the dose-response assessment, is used
to estimate the likely level of response observed given a particular level of exposure to a toxic agent.
The evaluation of dose-response data involves quantitatively characterizing the relationship between
the amount of exposure and extent of toxic injury or disease. The U.S. EPA has broadened the
definitions of hazard identification and dose-response assessment to include the nature and severity
of the toxic effect in addition to the incidence. Procedures for hazard identification and for
developing dose-response assessments have been established by the Agency and are followed in this
methodology document. Exposure assessment, the third step in risk assessment, is the process of
measuring the intensity, frequency and duration of exposure to an agent currently present, or of
estimating hypothetical exposures that might arise. Risk characterization, step four in risk
assessment, is performed by combining the exposure and dose-response assessments to estimate the
likelihood of an effect (NRC, 1983).
Figure 1-1 shows the relationship of the development of risk assessment methodologies to
sludge management practices. The figure further shows how each method may be used to develop
1-2
-------
PARTI: RISK ASSESSMENT
Existing EPA
Health and
Ecological Risk
Methodologies
of Existing
Sludge
Management
Practices
I
u>
Define Specific
Management
Practices for
Risk Assessment
Develop
Appropriate
EXPOSURE
ASSESSMENT
Methods: Define
Pathways, Exposed
Subjects and
Calculation
Methods
Sludge
Analytical
Data
SLUDGE
METHODOLOGY
DEVELOPMENT
Adapt Existing
HAZARD
IDENTIFICATION
and
DOSE-RESPONSE
ASSESSMENT
Methods
Select Health
Effects Data
and/or
Existing EPA
Assessment Values
(Potency, RID)
Select Fate
and
Transport
Input Data
Conduct Initial
Screen to
Select Candidate
Chemicals
RISK
CHARACTERIZATION
Criteria Derivation,
Sensitivity Analysis
CHEMICAL-SPECIFIC
DATA SELECTION
FIGURE 1-1
Relationship of Risk Assessment Methodology to Other Components of Regulation Development
for Sewage Sludge Reuse/Disposal Options
-------
PART II: RISK MANAGEMENT
Alter or
Control
Management
Practice
option
REGULATION
DEVELOPMENT
PERMITTING
Use Site-Specific
Inputs to
Rerun Criteria
Derivation
Issue
or
Deny
Permit
cirnnr i i / i
-------
national technical criteria, and how the risk manager may use or modify these criteria to develop
site-specific regulations or permits. The methodology to be discussed in this document falls within
the area of Figure 1-1 marked "SLUDGE METHODOLOGY DEVELOPMENT."
1.3. RISK ASSESSMENT IN THE METHODOLOGY DEVELOPMENT PROCESS
As can be seen from Figure 1-1, the methodology development process begins by defining
the management practice to be evaluated. Even within a single reuse or disposal option, the manner
in which the practice is carried out is highly variable. A definition that encompasses the variety of
methods used must be developed. The definition should not be limited to the ideal engineering
practice; it should include the types of practices most frequently used. It should not necessarily
include practices that are poor or substandard, unless such practices are widespread. The definition
of sludge surface disposal, included in Chapter 2 of this document, helps to determine the limits of
applicability of the methodology and to identify exposure pathways that may be of concern.
However, as shown in Figure 1-1 and as discussed in Section 1.4 below, the methodology described
in this document may help to redefine and regulate the practice; as a result, the definition of the
practice may be modified.
1.3.1. Hazard Identification and Dose-Response Assessment. Hazard identification requires the
evaluation of data that affect whether a chemical poses a specific hazard. It is a qualitative
determination, based on information regarding the type of effect produced by exposure, the
conditions of exposure, and the metabolic processes within the body that govern chemical effects.
Hazard identification includes the determination of whether effects observed under one set of
conditions (e.g., laboratory experiments) are likely to occur in other settings (e.g., environmental
exposures).
Information on the toxic properties of chemical substances is obtained principally from
animal studies and controlled epidemiologic investigations of exposed populations. The use of animal
toxicity studies are based on the assumption that effects observed in animals can be extrapolated to
effects in humans. Epidemiological studies are also useful in identifying hazards to humans. These
studies involve comparing the health status of a group of persons exposed to a causal agent with a
comparable unexposed group. In most cases, however, estimates of dose-response relationships are
1-5
-------
based on animal studies because even good epidemiological studies rarely have reliable information
on exposure.
Based on available evidence, the hazard identification requires a decision on whether a
chemical should be treated as a carcinogen. Procedures for evaluating the weight of evidence of
carcinogenicity have been described in U.S. EPA (1986c). If a chemical is a carcinogen, then a dose-
response assessment would consist of the use of Agency-accepted values (U.S. EPA, 1990). If values
are not available, the cancer risk estimation procedures published by the Agency may be used (U.S.
EPA, 1986c).
If a chemical is not carcinogenic, then the hazard identification and dose-response assessment
consist of identifying the critical systemic effect, which is the adverse effect occurring at the lowest
dose, and the Reference Dose (RfD), which is "an estimate (with uncertainty spanning perhaps an
order of magnitude) of the daily exposure to the human population (including sensitive subgroups)
that is likely to be without appreciable risk of deleterious effects during a lifetime" (U.S. EPA,
1990). Methods for deriving RfDs are given in U.S. EPA (1990).
For risks to nonhuman organisms, existing Agency methodologies for assessing possible
deleterious effects are used. For aquatic organisms, the methodology uses the existing U.S. EPA
water quality criteria for the protection of aquatic life (U.S. EPA, 1984c).
1.3.2. Exposure Assessment. The first step in an exposure assessment is to identify the pathways
through which exposure may occur. Exposure pathways are the routes by which chemicals migrate
from the reuse/disposal site to the target organ. Individual characteristics of human behavior that
affect the likelihood and extent of exposure are given particular scrutiny in the development of
models for human exposure pathways. Diversity in individual behavior patterns that affect potential
exposure will lead to variation in individual risk. The methodologies described in this document
focus on developing estimates of exposure to the "most exposed individual" (MEI)1, that is, the
individual in the exposed population who would experience the greatest health risk. However, a
1The MEI definition does not include workers exposed through the handling of sludge. It is
assumed that workers can take steps to minimize exposure through the use of protective gear.
However, agricultural workers may be included in the MEI definition.
1-6
-------
broader spectrum of human behaviors may also be modeled to define the distribution of possible
exposure levels and the number of persons exposed to each level. To this end, this report also
presents methods for estimating the distribution of exposures within a "most exposed population"
(MEP).
Chapter 3 discusses the pathways to be examined and defines the MEI separately for each
of these pathways. Specific data used to quantify exposure to the MEI will be given in Chapters 4,
5 and 6, which describe each pathway in detail. For pathways involving risks to nonhuman
organisms, the methodology uses the term "most exposed unit" or MEU rather than MEI, but still
uses conservative assumptions regarding exposure. This approach should ensure the development of
protective criteria.
1.3.3. Risk Characterization. Risk characterization consists of combining exposure assessment and
dose-response assessment results. Usually, exposures are estimated by tracing the movement of
chemicals from the source to the receptor. In this methodology, deriving the criteria requires the
reverse calculation. First, acceptable exposures to the receptor are determined. Next, the
concentration in the medium of concern (such as air or water) that would result in acceptable
exposures to the receptor are derived. The fate and transport calculations used to estimate chemical
concentrations in affected media are then back-calculated in order to estimate the corresponding
source concentrations and/or management practices that would result in those media concentrations
for the purpose of designing protective regulatory strategies. These calculations are carried out on
a chemical-by-chemical basis. Criteria are derived for each chemical assessed and for each pathway.
Example calculations for two chemicals, lead and benzene, are provided in this document.
Compiling the data to be used in the methodology is an exercise separate from developing
the model. Data on health effects of individual chemicals must be gathered from the scientific
literature. In many cases, the U.S. EPA has published values for cancer potency or the Reference
Dose. The chemical-specific input parameters for the fate and transport models, such as solubility,
Henry's Law constants and bioconcentration factors, must be collected separately from the scientific
literature. This information does not appear in the methodology development, except for those
chemicals used in the example calculations.
1-7
-------
Once data on chemicals of interest have been collected, it is useful to prioritize chemicals for
risk characterization. Prioritization may be based on the occurrence of a chemical in sludge, on the
likelihood of a chemical to migrate through pathways of concern and on the existence of data gaps
that would preclude application of the methodology.
Risk characterization or criteria derivation may be conducted after specific data have been
gathered. Criteria may include limits on application rates or concentrations, or restrictions on
management practices. When deriving criteria, it is advisable to vary input values over their plausible
range to determine the sensitivity of the results to the input values selected.
1.4. POTENTIAL USES OF THE METHODOLOGY IN RISK MANAGEMENT
Risk assessment can be used as input for the risk management process, as shown in part 2
of Figure 1-1. This document does not specify how the risk management should be conducted but
does suggest some possible uses of the methodology. A risk manager may evaluate the feasibility of
a set of criteria values based on consideration of costs, available technology or other factors unrelated
to risk. If certain restrictions on sludge concentrations, as specified by the calculations, would be
difficult to achieve, the disposal method could instead be regulated through design standards or
required management practices. Following the promulgation of the criteria, it may be possible to
evaluate sludge reuse or disposal practices on a site-specific basis, using site-specific data to derive
locally applicable standards that would reflect local conditions. Thus, the risk manager may use the
methodology as a tool to develop the most reasonable and effective regulatory control strategy for
particular sites.
The following chapters discuss one such approach for using the risk assessment methodology.
The approach consists of a first step, in which ("Tier 1") numerical criteria are developed for national
application, followed by a second step, in which ("Tier 2") site-specific calculations are conducted
where appropriate. These criteria are used to determine whether the sludge from each individual
facility is of acceptable quality for disposal or long-term storage in an impoundment.
1-8
-------
1.5. LIMITATIONS OF THE METHODOLOGY
The limitations of the methods used to derive the criteria are discussed in detail in the text
and in tables in the chapters where calculations are presented. Limitations common to all methods
are presented here.
Municipal sludges are highly variable mixtures of residuals and by-products of the wastewater
treatment process. Chemical interactions could affect the fate, transport and toxicity of individual
components; risk from the entire mixture may be greater than the sum of the risks of individual
chemicals considered separately. This methodology ignores possible synergistic effects. It should be
noted that the U.S. EPA's risk assessment guidelines for mixtures (U.S. EPA, 1986e) caution that a
great deal of dose-response information is required before a risk assessment that accounts for
chemical interactions may be performed. Future revisions to these documents are likely to include
only qualitative discussions of possible toxic interactions.
Transformation of chemicals during sludge disposal (i.e., during combustion or composting)
or following release into the environment (i.e., through biodegradation) may result in exposure to
chemicals other than those originally found in sludge. In some cases, the methodologies presented
in this document may not adequately characterize risks from chemical transformation products.
The methods described in this document do not consider differences attributable to the source
of the sludge. For example, the composition of the sludge matrix can be expected to differ between
sludge from primary treatment, and sludge from biological treatment processes (e.g., activated sludge
systems, trickling filters, and other attached growth systems); these differences are not considered
by the methodology. Similarly, potential impacts of lime stabilization on sludge are not discussed in
this document, but could be accommodated by the proposed models if sufficient data are available.
To the extent that these differences affect the mobility of toxic contaminants in the sludge, the
methods proposed here may under- or overpredict actual risks. Similarly, plants receiving a higher
contribution of wastewater from industrial sources are treated the same as those that receive no
industrial wastewater; the sludge criteria to be derived are independent of the source of the sludge,
and refer only to maximum allowable (dry weight) concentrations.
1-9
-------
The approach of the methodology described in this report can be used to examine each
exposure pathway separately, or to examine simultaneous exposure to more than one pathway. In
many cases, it will be unlikely that any single individual would be the most exposed individual
through more than one pathway, and adding up the risks across pathways is not recommended.
However, if such an individual could receive exposures through more than one pathway, estimates
restricted to a single pathway would underestimate risk. Chapters 4, 5, and 6 will outline methods
for separate calculation of criteria through individual pathways of exposure. Chapter 7 will provide
a simple method for considering these pathways simultaneously.
For nonhuman organisms, the methodology examines risks only to individual specimens and
only to aquatic organisms. Population-level or ecosystem-level effects are not examined, because it
is assumed that criteria sufficiently protective of individual organisms will also provide sufficient
protection at the population or ecosystem level. Although risks to aquatic species are considered
where appropriate, risks to non-human terrestrial species have not been examined; it is assumed that
criteria sufficiently protective of humans will also protect other terrestrial species for the appropriate
pathways of exposure.
1-10
-------
2. DEFINITION OF DISPOSAL PRACTICES
2.1. INTRODUCTION
Municipal wastewater treatment works may use one or more levels of treatment (i.e. primary,
secondary, or tertiary) to treat wastewater. Each level of treatment provides both greater wastewater
clean-up and greater amounts of sludge. Primary treatment processes remove the solids that settle
out of wastewater by gravity. These sludges contain 3-7%, 60-80% of which is organic matter. The
water content of primary sludge can easily be reduced by thickening or by removing water.
Secondary treatment produces a sludge generated by biological treatment processes. These processes
remove up to 90% of the organic matter in the wastewater and produce a sludge that typically
contains from 0.5-2% solids. These solids are generally more difficult to dewater than primary
sludges. The organic content of the solids ranges from 50-60%. Advanced wastewater treatment
processes, such as chemical precipitation and filtration, produce an advanced or tertiary sludge.
Chemical precipitation uses chemicals to remove organic contaminants and nutrients and to separate
the solids from the wastewater. Because these sludges typically contain considerable amounts of
added chemicals, the solids content will vary from 0.2-1.5%, while the organic content of the solids
will be 35-50%.
All three levels of treatment produce sludge for which reuse or disposal is required. In some
circumstances, long-term storage is also desirable. This document discusses a risk assessment
methodology for the long-term storage or permanent disposal of sludge in impoundments or lagoons.
Included in the methodology are facilities at which sludge accumulates and is stored for long periods
as a result of the use of lagoons for wastewater treatment. A lagoon or impoundment is an earth
basin used to deposit untreated or digested sludge. Anaerobic and aerobic digestion stabilize organic
solids in untreated-sludge lagoons. Stabilized solids settle and accumulate at the bottom of the lagoon
(Metcalf & Eddy, Inc., 1979).
No national survey has yet been calculated to identify the number of plants using surface
disposal or to estimate the volumes of sludge stored or disposed in this manner. However, 679 plants
participating in the 1986 Needs survey reported using forms of sludge treatment and disposal other
than landfilling, distribution and marketing, land application, incineration and ocean disposal. Some
2-1
-------
of these plants may have engaged in practices that fit the definition of surface disposal.
Furthermore, 727 plants reported having sludge lagoons, although some of these may have been used
for temporary storage or treatment and are thus not included in the surface disposal definition.
The impoundments to be considered in this document can be categorized into three groups:
lagoons for long-term storage, lagoons for permanent disposal, and wastewater treatment lagoons in
which sludge is stored for long periods. Each of these types of facilities is discussed briefly below.
2.2. LAGOONS FOR LONG-TERM STORAGE OF SLUDGE
Wastewater treatment plants may store sludge from primary, secondary, or tertiary treatment
processes in lagoons, with the intention of later exhuming the sludge for ultimate reuse or disposal
elsewhere. In most cases, sludge is periodically removed from the facility, although the frequency
of such removal will vary among facilities.
Long-term sludge storage may serve a number of purposes. Sludge storage may be an integral
part of a plant's overall sludge management plan. For example, sludge may be stored in
impoundments over the winter until weather conditions permit land application. Storage of sludge
will also tend to decrease water content, thereby decreasing hauling costs when the sludge is exhumed
for permanent disposal elsewhere. Smaller plants may store sludge to accumulate quantities that can
be practically disposed or reused. Storage impoundments may also be used intermittently, in
emergency situations, when normal sludge management operations are overwhelmed.
Although they are not technically "disposal" facilities, lagoons for long-term storage of sludge
have been included in methodologies for deriving criteria for "surface disposal" for two reasons.
First, this practice may pose health and environmental risks as severe as or worst than those associated
with permanent disposal. Second, future plans for removing sludge from "storage" impoundments
may be uncertain. As economic or regulatory conditions change, plans for the removal of sludge
from "temporary" impoundments have also been known to change. Since the ultimate fate of sludge
stored in such impoundments cannot be known with certainty, analysis of risks associated with their
continued or permanent existence is appropriate.
2-2
-------
2.3. SURFACE IMPOUNDMENTS AS PERMANENT DISPOSAL
In certain cases, sludge may be deposited in on-site impoundments without further planning
for removal. Permanent surface disposal may be the only practical disposal option for plants that
have accumulated very large quantities of sludge on-site over a number of years. It may also become
the de facto disposal method for plants that fail to identify satisfactory final disposal options for the
sludge currently stored in lagoons.
2.4. SLUDGE STORAGE AS A COMPONENT OF WASTE WATER TREATMENT
Wastewater treatment lagoons are often used by small wastewater treatment facilities because
they are less expensive than alternative treatment processes. Because such lagoons typically receive
relatively low volumes of wastewater, they can accumulate sludge in a bottom layer for years (or
even decades) before it begins to interfere with the treatment process. Although the main purpose
of these lagoons is to treat wastewater before discharge, the lagoons also serve as long-term storage
for the sludge, and may fall within the definition of surface disposal of sludge. In addition, the
lagoons are sometimes aerated, leading to increased volatilization of sludge contaminants.
2.5. ASSUMPTIONS USED IN RISK ASSESSMENT
The purpose of a sludge storage or disposal impoundment influences the manner in which
the sites are managed. For example, sludge in impoundments used for storage between land
application seasons will be emptied more frequently than sludge from wastewater treatment lagoons
at small plants. All of the surface disposal practices described here pose potential risks from human
and non-human exposure. Each practice also possesses unique features that affect its associated
risks. Long-term storage impoundments undergo periodic sludge removal, which may increase or
decrease movement of sludge contaminants into the environment. Permanent disposal impoundments
might present risks through future land uses of the disposal site (although these risks are not
considered by the present methodology). There follows a brief listing of assumptions used for
estimating risks from these sludge management practices:
1. Sludge storage or disposal occurs on plant property, and the public has no access
to the affected areas.
2. Sludge impoundments will be required to have berms, dikes or other surface
runoff controls that effectively eliminate significant risks of exposure from
flooding or accidental releases.
2-3
-------
3. Periodic removal of sludge from long-term storage lagoons allows their continuous
use over extended periods of time.
4. Sludge that accumulates in long-term storage or wastewater treatment lagoons is
removed prior to plant closure, and before any conversion of the land to other
uses can take place.
5. Sludge deposited in permanent disposal lagoons is never exhumed.
2.6. CONDITIONS AFFECTING RISK
A number of environmental conditions and management practices affect the risks posed by
long-term sludge storage impoundments. Management or design requirements can be developed,
based on an evaluation of the following factors, to minimize risks from surface disposal:
1. Concentrations of contaminants in the sludge will influence the magnitude of risk
posed by long-term sludge storage or permanent disposal.
2. Physical characteristics, such as distance to groundwater and surface water, soil
type and other geohydrologic features of the site, and proximity to human or
non-human populations will influence the rate of migration of contaminants and
the potential for exposure.
3. The length of time sludge is stored (i.e., the frequency of sludge removal from
the impoundments) may affect the mobility of sludge contaminants. Frequent
removal may decrease the length of time sludge contaminants are available for
leaching, but could also disturb the underlying stabilized sludge layer that may
inhibit contaminant movement from unlined impoundments.
4. The use of synthetic or clay liners will reduce the potential for contaminant
movement into groundwater.
5. The use of berms and proper siting can reduce risks associated with surface
runoff and flooding.
6. For permanent disposal facilities, restrictions on future access and land use
through a notice in the deed may be required once the facility is closed in order
to prevent inadvertent direct human contact with disposed sludge. Alternatively,
the sludge disposal area may be encapsulated at closure in order to prevent future
direct contact with sludge if land is converted to another use.
7. Aeration of wastewater treatment lagoons will affect the volatilization and
biodegradation rates of the organic sludge contaminants.
Improper management practices at a particular facility can result in additional risks
to human health or the environment; the methodology presented here does not include
techniques for assessing risks associated with such practices. Differences in potential risk
associated with siting of surface disposal facilities are considered by the methodologies insofar
as they can be represented by the input parameters required for model calculations.
2-4
-------
2.7. SUMMARY
The methodology presented in the following chapters is designed to derive criteria for lagoons
used for both long-term storage and permanent disposal of sludge. The methods are intended to
apply to all types of municipal wastewater sludges, regardless of the extent of industrial contribution
to the treated wastewater, and regardless of the type of wastewater treatment processes that generate
the sludge. Risks from these surface disposal facilities can occur through a variety of pathways of
potential human and non-human exposure. Chapter 3 of this document will discuss the selection of
pathways for inclusion in this methodology, and the identification of "most exposed individuals" and
"most exposed populations" at greatest risk through each pathway. Chapters 4, 5, and 6 will provide
detailed discussion of methods for assessing potential risks and deriving criteria for the three
exposure pathways thought to be of most concern: groundwater, surface water, and air.
2-5
-------
-------
3. EXPOSURE PATHWAYS AND MOST EXPOSED INDIVIDUALS (MEIs)
3.1. INTRODUCTION
Human or environmental exposure to contaminants from wastewater sludge can occur through
a number of pathways. The pathways of greatest concern for surface sludge disposal will depend on
the purpose of surface disposal (e.g., storage vs. wastewater treatment) and the management practices
employed. This chapter describes pathways of potential concern for each of the three types of
surface disposal, using the definitions developed in Chapter 2, and identifies management or design
practices that may influence exposure through these pathways. When a particular exposure pathway
is identified as a pathway of concern for a given management practice, the assessment methods
described later in this report should be used to determine whether criteria are needed. Identification
of pathways of concern for a given management practice should not be interpreted to mean that
criteria will necessarily be required for that practice.
The methodology described in this document considers three types of pathways of potential
exposure: a pathway involving the transport of sludge constituents through groundwater, pathways
involving subsequent transport through surface water, and a pathway involving the volatilization of
organic constituents to ambient air. Risks associated with each of these pathways will be examined
later in this report. For each pathway identified, risk to the most exposed individual (MEI) will be
evaluated. The MEI is defined as the individual in the general population who would experience the
greatest health risk, either because of the expected magnitude of the individual's exposure to
contaminants or because of physiological or behavioral characteristics that make the individual
particularly susceptible or sensitive to contaminant exposure. For certain pathways, aquatic species
may be the organisms at greatest risk from exposure to sludge contaminants; for these pathways, the
most sensitive aquatic species will be considered the MEI or most exposed unit (MEU) of interest.
Although the MEI (or MEU) is a hypothetical individual, care should be taken to make the definition
realistic. For each disposal option and exposure pathway, formulas are presented to calculate
regulatory criteria that adequately protect the appropriate MEI or MEU. It is reasoned that as long
as the risk assessment procedures can reasonably estimate the risks to these individuals, then the
quantification of lesser risks experienced by other individuals is not required.
3-1
-------
Two problems may arise from reliance on such an approach. First, physical identification
of the actual MEI is rarely feasible; the MEI to be considered when setting criteria is therefore a
theoretical construct. Where large modeling uncertainties exist, the compounding of conservative
assumptions in the MEI definition or exposure models can produce extreme MEI scenarios that
result in unnecessarily strict criteria. Conversely, seemingly reasonable "worst case" scenarios may
be found, upon closer inspection, to predict lower levels of exposure than those encountered by
actual individuals in the exposed population, especially when the size of the exposed population is
large and potential variability in individual behavior is great. Choices for defining the "most exposed
individual" are directly linked with the ultimate stringency of regulatory criteria; inappropriate
choices are likely to lead to inappropriate criteria.
A second problem with using an MEI approach to set regulatory criteria is the difficulty of
maintaining comparability in MEI definitions across all sludge management practices and exposure
pathways. If compliance with criteria for one sludge management practice is more difficult than
compliance with criteria for another, sludge managers will be encouraged to use the practice with the
more lenient criteria. Changes in management practice will serve the public health and environmental
quality only if the more strictly regulated option for sludge reuse or disposal is indeed riskier than
the less strictly regulated one. Over-regulation of one practice coupled with under-regulation of
another practice may actually increase health or environmental risks if sludge managers shift from
alternatives with lower risks to those with higher risks. Such disparities, and their consequences, are
most likely if MEI definitions for different sludge reuse or disposal practices are not comparable.
Avoidance of such unwanted consequences requires a systematic approach to defining the
exposure scenarios and the most exposed individuals for which sufficiently protective criteria are
to be derived. The approach should have two goals: (1) to derive realistic MEI definitions that are
both realistic and sufficiently protective, and (2) to achieve comparability across MEI definitions
applicable to different sludge management practices. One such approach involves consideration of
"most exposed populations" (MEP) for each sludge management practice and for each potential
pathway of exposure. The approach involves estimating the size of the populations exposed to
various levels of sludge contaminants through each exposure pathway. Based on estimates of national
3-2
-------
distributions of various relevant parameters, criteria can be derived that are sufficiently protective
of populations of specified (non-zero) estimated sizes. The natural extension of these techniques is
to use these distributions to derive an MEI definition appropriate for a single (theoretical) individual
with the highest true potential for exposure. Application of these methods to the derivation of
criteria for surface disposal facilities will be discussed in more detail in Chapters 4, 5 and 6.
The precise definition of the MEI will depend on the assumptions and requirements for each
management practice described in Chapter 2. This chapter will qualitatively define the MEI for each
potential pathway of concern. The MEI will be described quantitatively, using reasonable worst-
case exposure assumptions, in the following chapters where criteria calculation methods are given.
As discussed in Chapter 1, occupational exposures will not be considered in this analysis. Neither
will the analysis consider hazards that are not related to routine migration of sludge contaminants
from the impoundment into the environment (e.g., exposure to pathogens, potential surface water
contamination by flooding, risks of releases caused by earthquakes, underground gas migration and
risks of explosion, nuisance concerns, and general environmental concerns like global warming, ozone
depletion, or vegetative distress).
Possible pathways through which humans or other organisms may be exposed to contaminants
from surface-disposed sludge include:
1. sludge-groundwater-wells-drinking water;
2. sludge-groundwater-surface water-drinking water;
3. sludge-groundwater-surface water-consumption of fish;
4. sludge-groundwater-surface water-exposure of aquatic life;
5. sludge-volatilization-downwind transport-inhalation.
6. sludge-surface runoff-surface water-drinking water;
7. sludge-surface runoff-surface water-consumption of fish;
8. sludge-surface runoff-surface water-exposure of aquatic life; and
9. sludge-wind transport of particulates or aerosols-inhalation
10. sludge-soil (site conversion) - humans
3-3
-------
The methodology presented in this document will be restricted to consideration of those pathways
associated with the transport of contaminants through groundwater (and subsequently through surface
water) or with the volatilization of organic contaminants. Potential exposure associated with surface
runoff from the facilities, from possible wind transport of particulates or aerosols, and from pathways
associated with future conversion of the site to residential uses will not be included.
3.2. GROUNDWATER PATHWAY
Contaminants in sludge currently stored or disposed in impoundments can leach into the
groundwater underlying the site. Sludge contaminants may subsequently migrate into wells from
which drinking water is obtained. Although all three types of surface disposal units under
consideration potentially pose risks through the groundwater pathway, the rate and extent of
contaminant migration from sludge impoundments may be greater than from other kinds of sludge
disposal sites because of the presence of liquid in the impoundment. The presence of synthetic
liners may retard or delay migration of contaminants. The MEI for this pathway would be an
individual who obtains all drinking water over the course of a lifetime from a well adjacent to the
wastewater treatment plant boundary.
If contamination of groundwater resources is sufficiently widespread, a number of wells may
be affected by sludge contaminants. If more than one well is potentially affected, and the level of
contamination differs significantly among the different wells, then a more detailed analysis of the
extent to which exposure varies among individuals in the exposed population (an analysis of the MEP)
can provide additional information of potential usefulness to the regulatory process.
3.3. SURFACE WATER PATHWAY
Sludge contaminants may enter surface water through direct surface runoff from sludge
impoundments or waste piles. However, it is assumed that since the purpose of the impoundments
is sludge containment, berms and other surface runoff controls will be required at all surface disposal
sites. However, sludge contaminants may enter groundwater and subsequently discharge into surface
water. Contaminants in surface water can accumulate in the body tissues of aquatic organisms.
Endpoints of concern include the direct exposure of aquatic life to contaminated water, and human
ingestion of contaminated surface water and fish. The ecological endpoint of greatest concern would
-------
be the aquatic species most sensitive to the presence of the contaminant in the water. The human
MEI would be defined as that individual who obtains all drinking water from an intake downstream
from the point of contaminant discharge into the surface water body, or who consumes fish caught
near the point of contaminant discharge into the water body. Criteria sufficiently protective of
aquatic species and of the human MEI are assumed to be sufficient for protection of nonhuman
terrestrial species.
If more than one human is exposed, then levels of potential exposure are likely to vary within
the exposed population. In this case, a more detailed analysis of the MEP can provide guidance in
selecting an appropriate definition for the human MEI, and can provide other information useful to
the regulatory process.
The groundwater-surface water-drinking water, groundwater-surface water-fish
consumption, and groundwater-surface water-exposure of aquatic life pathways may be important
pathways of exposure for any of the surface disposal practices under consideration. These pathways
may be particularly important whenever surface disposal is practiced in close proximity to surface
water resources. When using the methodology for the purpose of setting national criteria, it should
be assumed that surface waters near surface disposal sites may be used for both drinking and fishing.
3.4. AIR PATHWAY
Volatile sludge contaminants may be emitted into the air and may be dispersed downwind,
where they may be inhaled by receptors. The MEI is defined as the downwind receptor closest to
the sludge surface disposal area. The distribution of potential exposure over the entire exposed
population may also be of interest. To obtain an estimate of the distribution of exposure, air
modeling can be used to derive estimates of air concentrations surrounding a facility as a function
of distance; this information is coupled with population data to derive a distribution of air
concentrations as a function of the number of persons exposed.
While volatilization may occur from any of the sludge surface disposal sites described in
Chapter 2, this pathway may be especially important for aerated wastewater treatment impoundments.
Aeration will increase the rate of volatilization, resulting in higher downwind concentrations.
3-5
-------
-------
4. DERIVATION OF CRITERIA FOR THE GROUND WATER PATHWAY
This chapter describes a methodology for deriving criteria, based on human exposure through
the groundwater pathway, for the storage or disposal of sludge in surface impoundments. The
approach is based on methodologies for assessing risks to human health, and can be used for both
national and site-specific applications. A risk assessment methodology for groundwater
contamination resulting from the landfilling of municipal sludge (Development of Risk Assessment
Methodology for Municipal Sludge Landfilling. U.S. EPA, 1986a) has also been developed as a part
of this document series; the present methodology differs from that earlier effort in two important
respects. Surface disposal facilities differ from landfills in that the sludge deposited in an active
lagoon is likely to contain more liquid than does sludge deposited in a landfill. Consequently, the
"source term" for estimating groundwater contamination beneath a surface disposal facility will differ
from the source term for a landfill. In addition, the methodology proposed in this document departs
from that proposed for landfills because it selects a different mathematical model for estimating
flow and contaminant transport through the unsaturated soil zone. Nevertheless, some elements of
the two risk assessment methodologies are identical; to avoid lengthy repetition of text, this document
occasionally refers the reader to the landfill methodology document for additional discussion of
details shared by the two methodologies.
4.1. OVERVIEW OF THE METHOD
Water leaking from the bottom of surface disposal sites can contain dissolved contaminants
from the sludge. These contaminants can be carried downward by water percolating through the
unsaturated zone beneath the facility, until they reach an underlying aquifer. Within the aquifer,
they can be transported beyond the property boundaries of the facility where the contaminated water
may be withdrawn through private or public wells for human consumption. As with other pathways
of potential human and wildlife exposure to be discussed in this document, the methodology for
deriving criteria based on the groundwater pathway involves a two-tiered approach. Tier 1 derives
numerical sludge criteria for national application, whereas Tier 2 provides criteria tailored to
conditions at specific surface disposal sites.
4-1
-------
Tier 1 analysis begins with the calculation of reference water concentrations (RWC) for each
sludge contaminant of concern; these are calculated using methods to be described in Section 4.5.
Next, the analysis uses a mathematical model of a "generic" surface impoundment scenario to estimate
the maximum concentration of each sludge contaminant that can be allowed in the impoundment if
contaminant concentrations in groundwater beneath or near the facility are not to exceed the RWC.
These maximum allowable concentrations are used as criteria for regulating sludge quality for all
surface disposal facilities. The scenario is selected to represent reasonable worst case conditions, so
that the associated risks are as high as or higher than those to be expected at actual facilities. If
desired, these maximum allowable concentrations, or sludge criteria, can be estimated separately for
different aquifer classes.
If sludge produced by a particular publicly owned treatment works (POTW) fails the
numerical criteria derived by Tier 1 analysis, then a site-specific (Tier 2) analysis is required. Tier
2 analysis begins with the reference water concentrations derived in the first step of Tier 1. Using
the same mathematical model employed in Tier 1, but with site-specific input parameter values, the
analysis computes maximum sludge concentrations that can be allowed at the particular facility if
reference water concentrations are not to be exceeded. As with Tier 1, separate requirements can
be imposed according to the class of aquifer involved.
Both methodologies are based on a mathematical model that describes the migration of sludge
contaminants from an impoundment to an underlying aquifer, lateral movement of contaminants
through the aquifer to a nearby well, and withdrawal of drinking water from the well for human
consumption. The property boundary may be selected as the point of compliance, since drinking
water wells could be constructed at that location and outward, and contamination of the wells could
result in potential impacts to public health. This selection would be based on the premise that any
potable water supply beyond the property boundary must be maintained at healthful levels for
potential future uses even if there are no current wells immediately off site. Alternatively,
groundwater directly beneath the site might be selected as the point of compliance if the site is
located over a Class I aquifer.
4-2
-------
The mathematical model includes five analytical steps:
1. estimation of the quantity of water seeping from the bottom of the impoundment and
of contaminant concentrations in that water,
2. estimation of the extent to which contaminant concentrations are reduced during
transport through the unsaturated zone,
3. estimation of the extent to which concentrations change during horizontal contaminant
transport to a receptor well,
4. estimation of potential human exposure based on well-water concentrations, and
5. evaluation of potential human health risk, based on comparisons of expected exposure
to risk reference doses for non-carcinogenic contaminants, and on the use of potency
values for carcinogens.
Details for each step in the methodology are provided in Sections 4.3 and 4.4.
Numerical criteria are specified as limits to the dry-weight concentrations of contaminants
in sludge. These sludge criteria must be set so that any environmental contamination resulting from
surface disposal of sludge would not exceed health-based reference concentrations. To derive criteria
for the storage or disposal of sludge at a particular site, the five steps of the methodology must be
linked in reverse order: based on reference water concentrations in a drinking water well, the
methodology is used to estimate maximum allowable concentrations in sludge as it enters the
impoundment. Facilities managing sludge must compare the sludge concentration of each
contaminant with the Tier 1 criteria. If all contaminant concentrations pass the criteria, then the
facility's application is accepted. If any one contaminant exceeds the Tier 1 criteria, then a Tier 2
analysis can be performed using site-specific data to derive (presumably less conservative) sludge
criteria more applicable to conditions at the particular site.
4.2. ASSUMPTIONS
Application of the methodology presented in this document requires several simplifying
assumptions; these are summarized in this section and in Table 4-1, but will be discussed in greater
detail in Section 4.3. As can be seen from Table 4-1, calculations are based on a source term of
seepage volume and contaminant mass flux into the unsaturated and saturated soil zones beneath a
lagoon. Representation of this source term depends on the certainty with which the future operation
of the lagoon can be predicted; for storage facilities from which sludge is occasionally removed,
4-3
-------
TABLE 4-1
Assumptions for Methodology to Analyze the Groundwater Pathway
Functional Area
Assumption
Ramifications
Source Term
-c-
-t-
Unsaturated Zone Flow
Unsaturated Zone Transport
At sludge storage facilities for which the
active lifetime cannot be predicted with
certainty, contaminant flux through seepage
can be described as steady-state, with a
renewable supply of contaminant mass.
For permanent disposal facilities at which
the timing of the final sludge deposit is
known, contaminant flux through seepage can
be described as a square wave of duration
determined by the mass of contaminant
contained within the impoundment.
Ratio of solids to liquids at bottom of
impoundment is 10:1. Dry-weight
concentration of contaminant in sludge of
bottom layer equals that of sludge entering
the facility. Dissolved concentration can
be predicted from an equilibrium
distribution coefficient.
One-dimensional flow in the vertical
direction.
Water flow is steady state; upper boundary
has constant flux of contaminant mass.
Soil characteristics are constant with depth
for any layer.
Attenuation of organics is related to soil
organic fraction only.
All adsorption is reversible.
Degradation of organic contaminants is first
order.
May overpredict loading to aquifer after
facility ceases to accept sludge.
Overpredicts duration of high concentration
period, but may underpredict peak
concentration if mass is conserved.
Neglects presence of low concentration tail.
Unknown.
Overpredicts concentration since it ignores
horizontal dispersion.
May overpredict loading after impoundment
ceases to accept sludge.
Unknown.
Overpredicts contaminant velocity for soils
with low organic content where mineral
interaction may predominate.
Overpredicts concentration arriving at
aquifer.
Unknown.
-------
TABLE 4-1 (cont.)
Functional Area
Assumption
Ramifications
Saturated Zone Transport
Transport to Surface Water
I
vn
Potential effects of "mounding" can be
approximated with a superimposed velocity
term, estimated from quantity of seepage,
perimeter of site, and depth of aquifer.
Geochemical reactions and other "pseudo-
decay" processes are ignored for metals.
Degradation of organic contaminants is first
order.
Receptor well is located directly down-
gradient of the site, at a distance defined
by the property boundary.
Direction of groundwater flow is toward
surface water, perpendicular to the bank of
surface water body.
Average distance of travel is approximated
by shortest distance from site to surface
water.
Total mass of contaminant in seepage from
the impoundment eventually reaches the
surface water body, except that lost to
degradation.
Vertical and transverse dispersion of
contaminant (parallel to stream bank) can be
ignored when calculating average loading.
Likely to overestimate flow velocity in
aquifer. Likely to overestimate velocity
and concentrations at the receptor well.
Overpredicts concentrations of metals.
Unknown.
Likely to overpredict concentrations in
actual wells.
Will overpredict loadings where actual
direction is different.
Likely to overpredict loadings.
May overpredict loading to surface water
body.
Should not affect results if loading along
entire stream bank is to be considered, and
if the entire contaminant plume is
intercepted by the stream.
-------
operation can continue indefinitely, and the source of seepage and contaminant mass can be
renewable. For permanent disposal facilities, the moisture and contaminant mass supplied by the
sludge are considered finite. Maximum loading to ground water is assumed to occur once a sludge
layer has accumulated on the floor of the impoundment, and solids concentrations are highest.
Dissolved concentrations of contaminant within this sludge layer, and within seepage from the floor
of the lagoon, are assumed to be predictable based on equilibrium distribution coefficients and a
conservative 10:1 ratio of solids to liquids within the sludge layer. As proposed, the methodology
does not consider potential losses of contaminant mass to biodegradation or volatilization during
containment within the impoundment. Although methodologies proposed in Chapters 4 and 6 could
be linked to provide a mass balance between these competing loss processes, the methodologies
proposed in Chapter 6 for estimating volatilization and biodegradation are not intended to provide
an accurate depiction of conditions within the sludge layer on the floor of an impoundment.
As will be explained in further detail in Section 4.3, flow and contaminant transport through
the unsaturated zone are represented by a one-dimensional model that allows consideration of
multiple soil layers, each with soil characteristics that do not vary with depth. With the unsaturated
zone, the attenuation of organic contaminants is predicted based on longitudinal dispersion, an
estimated retardation coefficient derived from an equilibrium partition coefficient, and a first-order
rate of contaminant degradation. Potential effects from local elevation of the water table are
considered by the model when estimating contaminant attenuation during transport within the
unsaturated zone. Concentrations of metals are assumed to be conserved during transport through
the unsaturated zone.
Contaminant migration through the saturated soil zone is estimated with a model that includes
adjustments for likely effects of "mounding" associated with significant seepage beneath some
facilities. Consideration of these effects within a one-dimensional model of contaminant transport
through the saturated zone requires some simplifying assumptions about the velocity of contaminant
transport and the extent of contaminant dilution by the aquifer. As in calculations for the
unsaturated zone, degradation of organic contaminants are assumed to be first order during transport
through the aquifer. Geochemical reactions are ignored for metals, leading to the likely
k-6
-------
overestimation of expected concentrations of metals in groundwater at a receptor well location. The
receptor well is conservatively assumed to be located in the down-gradient direction at a minimum
distance defined by the property boundary.
Estimation of expected loadings of contaminant to nearby surface water bodies is based on
the assumptions summarized above, as well as a few additional assumptions particular to this pathway
of potential exposure. The direction of groundwater flow is assumed to be toward the surface water;
the time of travel and extent of contaminant dispersion during transport to the surface water are
determined by the aquifer medium and by the shortest distance from the site to the lake or stream.
Contaminant loading to the surface water body is predicted without regard to dispersion of
contaminant in directions parallel to the shoreline of the surface water body. Further details for the
assumptions required for the methodology will be provided in Sections 4.3 and in Chapter 5.
4.3. CALCULATIONS
This section provides a description of the calculations and model used for deriving criteria.
The methodology is described within the context of a site-specific application for a particular surface
disposal site. As mentioned above, however, the same methods can also be used for Tier 1
calculations to derive national, numerical criteria.
4.3.1. Source Term. Each contaminant transport pathway begins with a source term and ends
with a receptor or point of exposure. In evaluating sludge disposal alternatives, the sludge itself is
the source term through which the various contaminants are introduced into the environment. To
quantify risks attendant to the storage or disposal of sludge in surface disposal facilities, it is
necessary to determine the mass of contaminant released as a function of time.
To characterize the behavior of sludge in an impoundment, it is useful to conceptualize the
lifetime of a surface disposal facility in two phases. In the first (or "active") phase, sludge is
continuously or periodically added to the facility. For the special case of a wastewater treatment
lagoon, the sludge will accumulate as a result of the settling of solids from wastewater. For other
types of impoundments discussed in Section 2.1, sludge solids simply accumulate as a result of
repeated sludge deposits. During the active phase of the impoundment, the reservoir of contaminant
mass contained in the impoundment can increase as a result of sludge accumulation. An accumulating
4-7
-------
sludge layer on the bottom of the impoundment might result in increased concentrations of dissolved
sludge contaminants localized at the bottom of the impoundment. On the other hand, the presence
of an anaerobic sludge layer at the bottom of a lagoon may reduce contaminant concentrations in
leachate and may slow the rate of seepage from an impoundment (Kehew et al., 1983; U.S. Geological
Survey, 1988b). Average rates of seepage from a particular lagoon can be estimated from a mass
balance that includes inflow, outflow, precipitation and evaporation. Typical rates of seepage
beneath municipal lagoons are listed in Table 4-2, taken from U.S. EPA (1987b).
In the second, or "inactive," phase of the facility's lifetime, new additions of sludge are no
longer received by the impoundment. For those impoundments used only for storage of sludge (e.g.,
wastewater treatment lagoons that accumulate sludge, or impoundments used for temporary holding
and dewatering), accumulated sludge is periodically removed, and the facility may never reach the
inactive phase without being emptied. Impoundments used for permanent sludge disposal, however,
are eventually closed without removal of their contents. Contaminant mass within the impoundment
may decrease as a result of volatilization, biodegradation, seepage to groundwater or other loss
processes. If water seepage beneath the facility exceeds local precipitation less evaporation, the
contents of the impoundment will lose moisture until the rate of seepage reaches equilibrium with
the rate of recharge. As the moisture content of the impoundment decreases, the reduced rate of
seepage beneath the facility might result in a reduced flux of contaminants to groundwater, but that
reduction could be offset at least partially by increased concentrations of dissolved contaminants
that might leach from sludge with a higher solids content.
Because of uncertainties about the behavior of the bottom layer of sludge in impoundments,
and because future plans for sludge removal or addition cannot be known with certainty for many
surface impoundments, separate modeling of the expected active and inactive phases of a particular
impoundment is in many cases infeasible. For these reasons, the methodology presented here
combines estimates of steady-state seepage from an active impoundment with conservative estimates
of leachate concentrations that may be more typical of inactive lagoons.
Calculations for both Tier 1 and Tier 2 begin with three conservative assumptions for
estimating expected concentrations of contaminant in seepage from a lagoon. First, they assume that
-------
TABLE 4-2
Summary of Measured Seepage Rates From
Municipal Lagoon Systems3
Water
depth
(feet)
5
6
5
6
6
-
-
-
-
-
5
5
5
-
Lagoon Type
Facultative
Facultative
Facultative
Facultative
Facultative
--
—
—
--
Maturation
Facultative
Facultative0
Evaporationd
Facultative
Underlying Soil
Heavy silty clay
Light silty clay
Alkaline silt
Fine sand
Gravel and silt
Sandy soil
Sand and gravel
Sandy soil
Clay loam and shale
Mica and schist
Silt, sand, marl
Sand, silt, marl
Sand, silt, marl
Sandy soil
Seepage
in/day
0.3
0.29
0.65
1.2
1.3
0.35
0.61b
0.34
0.3
0.06-0.23
0.18
1.07
0.04-0.11
0.12
Rate
//m2-hour
0.32
0.31
0.69
1.3
1.4
0.37
0.65
0.36
0.32
0.06-0.24
0.19
1.13
0.04-0.12
0.13
a Source: U.S. EPA, 1987b
b Includes net precipitation/evaporation
c Used intermittently
d Sealed with bentonite and soda ash
-------
sludge at the bottom of an active or inactive impoundment contains a low water content. Values for
the water content of drained sludges are presented in Table 4-3; most sludges described by the table
retain water at a ratio of 12:1 to 24:1. The use of a 10:1 ratio might therefore represent a reasonable
"worst case" value appropriate for the bottom layer of an inactive impoundment. Second, the
calculations assume that the contaminant mass in the sludge and water mixture on the lagoon floor
is partitioned at equilibrium between dissolved and adsorbed phases. It is assumed that this
partitioning is described by an appropriate partition coefficient (krf) for each contaminant of concern.
Then the dissolved concentration of contaminant can be related to the total (dry weight) sludge
concentration by:
CI/N = l/(kd+rsl'1) (4-1)
where:
C, = concentration of the contaminant in liquid phase at the bottom of the
I -9
impoundment (mg// or g/m )
N = dry weight concentration of contaminant i in sludge at the bottom of the
impoundment (mg/kg)
kd = equilibrium partition coefficient for the contaminant (//kg)
r = ratio of solids to liquid in sludge at the bottom of the impoundment,
conservatively assumed to be 0.1 (kg//)
Values for the partition coefficient (kd) for each contaminant of concern can be derived from the
literature, obtained from Appendix C of U.S. EPA 1986a, calculated with the CHEMEST procedure
from the Graphical Exposure Modeling System (GEMS) maintained by the U.S. EPA Office of
Pesticides and Toxic Substances (U.S. EPA 1988c), or can be estimated empirically for sludge using
methods described in DeWald and Phillips (1989).
An alternative method for predicting leachate concentrations would be to use actual
measurements of contaminant concentrations in water drained from each individual facility's sludge.
This method avoids the necessity of relying on theoretical estimates of leachate concentrations, but
results are likely to depend on the initial solids content of the sludge received by the impoundment,
and might not reflect conditions at the bottom of a lagoon that has accumulated a sludge layer.
-------
TABLE 4-3
Water Content of Sludges from Various Treatment Processes*
Water
Sludge Type Content (%) Water: Solids
Primary sludge 95 19
Digested primary sludge 94 16
Trickling filter 92 12
Chemical precipitation 92 12
Primary and activated sludge 96 24
Digested primary and activated sludge 94 16
Activated sludge 98.5 66
Septic tank-digested activated sludge 90 9
Imhoff tank-digested activated sludge 85 6
*Source: U.S. EPA, 1978
-------
Characterization of the source term also requires estimating the time over which the
contaminant will be present in the leachate. Sludge contains a finite mass of contaminant that can
be mobilized in leachate. For some contaminants, that mass is less than the total mass in the sludge
because of irreversible adsorption or other binding mechanisms.
As long as an impoundment is active, the mass of sludge contaminant it contains can be
periodically restored, and will not be permanently depleted. For an inactive permanent disposal
impoundment, however, the mass of available sludge contaminant will be depleted over time. As
discussed above, this methodology conservatively assumes that the end of the active phase for most
impoundments cannot be predicted with certainty and, therefore, assumes that an impoundment
continues operation indefinitely. This assumption is represented by steady-state loadings of
contaminant to the unsaturated zone beneath an impoundment.
For cases where a reasonable determination of the duration of an impoundment's active
lifetime can be accomplished, however, the model will allow the applicant to model loadings to
groundwater as a pulse, or series of pulses, of finite duration. For such a facility, calculation of pulse
time requires determination of total contaminant levels in the contents of the inactive facility, rates
of decay for contaminant concentrations, and the time path of seepage volume as seepage rates
decrease with decreasing moisture content in the impoundment. Derivation of pulse times is based
on mass conservation considerations. Details of the necessary calculations are provided in U.S. EPA
(1986a).
4.3.2. Unsaturated Zone Flow and Transport. No one method or model for estimating water
flow and contaminant transport through the unsaturated zone is appropriate for application to all
cases. In general, criteria used to select models for this methodology are as follows:
1. the method should be simple and amenable to widely available equipment;
2. the data required by the method should be generally available; and
3. the method should be applicable to a wide range of problems or sites.
The first criterion is important if the model is to be used for numerous site-specific
applications in a variety of locations. As model complexity increases, demands on the skill and
judgement of the modeler also increase, as does the difficulty of applying the model consistently
4-12
-------
and equitably from location to location. In addition, large numerical model codes can require
expensive or specialized computing equipment that may be unavailable in some locations. The second
criterion, that the data required by the method should be generally available or that the data can be
estimated, is important if expensive, specialized site studies are not to be required in support of each
application. For each site, required data should be already available, should be easily obtainable or
measurable, or should be amenable to reasonable approximation based on literature sources. The
third criterion is that the method be generally applicable to a wide range of problems or sites.
Potential sites exist across the United States and, therefore, may be characterized by a range of
values. It is impossible to select any single method that is optimal for all sites, especially given the
constraints of the first two criteria. Methods should be selected that can be used for the wide range
of values potentially encountered, without requiring different approaches for each setting.
Simulation of contaminant transport through the unsaturated zone can be accomplished by
models with varying degrees of sophistication. Complex three-dimensional models provide the best
available simulation of contaminant transport beneath a surface impoundment, but these models
require more detailed geohydrological data than are typically available at existing sludge management
facilities. In addition, they require considerable modeling expertise for proper execution and
interpretation. A one-dimensional vertical unsaturated zone model is simpler to use, requires fewer
input data and is easily applied to a wide variety of sites. The methodology proposed herein
therefore relies on a one-dimensional model for simulating contaminant movement through the
unsaturated zone.
The present methodology links two well-accepted models for unsaturated and saturated zone
flow and transport. The Vadose Zone Flow and Transport finite element module from the RUSTIC
model (U.S. EPA, 1989a,b) is used to estimate flow and transport through the unsaturated zone, and
the AT123D analytical model (Yeh, 1981) is used to estimate contaminant transport through the
saturated zone. The combination of these two codes creates a simple-to-use model appropriate for
deriving sludge criteria.
Three candidate models for estimating flow and transport through the unsaturated zone were
considered for use in the present methodology: the CHAIN model (van Genuchten, 1985) as used in
4-13
-------
U.S. EPA (1986a), the LEACHM model (Hutson and Wagenet, 1985), and the VADOPT module from
RUSTIC. CHAIN was rejected because of previous problems encountered when using the model to
derive criteria for sludge monofills. Review of supporting documentation suggests that both
LEACHM and VADOFT are well suited for adaptation to model the unsaturated zone beneath a
surface impoundment; VADOFT was selected in part because it is supported by the U.S. EPA Office
of Research and Development in Atlanta. U.S. EPA (1989a) reports that tests on benchmark problems
show that results from VADOFT are in good agreement with results from the more sophisticated two
dimensional finite element codes UNSAT2 (Davis and Neuman, 1983) and SATURN (Huyakorn et
al., 1984). The Pesticide Root Zone Model (PRZM) is used by the RUSTIC model to simulate the
movement of pesticides through the soil surface, and runoff to surface water. It is not appropriate
for use in modeling contaminant migration beneath a surface impoundment. The remaining
component of RUSTIC, or SAFTMOD, simulates the movement of contaminant through the saturated
zone, and could have been appropriate for this methodology, but was rejected in favor of AT123D,
for consistency with other methodologies described in this series.
The ultimate goal of this methodology is to determine the maximum contaminant
concentration in sludge that will result in contaminant concentrations in groundwater that do not
exceed reference water criteria. As mentioned earlier, the proposed model executes in two steps;
results from the unsaturated zone flow and transport module are passed as input to the saturated zone
module. The input requirements for the unsaturated zone module include various site-specific and
geologic parameters and the leakage rate from the bottom of the surface disposal facility. It is
assumed that the concentrations of various contaminants entering the unsaturated zone beneath a
facility can be represented by the concentrations derived with methods discussed in Section 4.3.1.
Results from the unsaturated zone analysis give the flow velocity and concentration profiles for each
contaminant of interest. These velocities and concentrations are evaluated at the water table,
converted to a mass flux and used as input to the saturated zone module.
A complication in estimating contaminant movement beneath a surface impoundment arises
from the inherent multi-dimensionality of water flow and contaminant transport. Seepage from the
surface impoundment can cause local elevation of the water table if rates of seepage from the lagoon
-------
exceed natural rates of aquifer recharge in the surrounding area. Such elevation of the water table,
or mounding, has two implications for the expected concentrations of sludge contaminants at a
receptor well. First, the reduced vertical distance between the impoundment and the local water table
will result in decreased time of travel for water moving between the impoundment and the saturated
zone, and changes the extent of soil saturation between the impoundment and the floor of the aquifer.
The second effect of mounding in the water table is an increased hydraulic gradient in the aquifer
between the disposal site and the down-gradient receptor well. This change in the gradient will
increase the expected rate of horizontal transport of the contaminant through the saturated zone.
To accommodate these two effects in the model calculations, this methodology adapts the
approach used in the RUSTIC model. The first component of the model performs calculations for
a vertical column containing both unsaturated and saturated zones, and predicts the extent to which
the elevation of the water table will be increased by the flux of water from the impoundment. Once
the vertical column problem has been solved for mass and water fluxes at the water table elevation,
the second model component (AT123D) simulates the movement of contaminants through the
saturated zone, with adjustments to represent increased elevation of the water table. Unlike RUSTIC,
however, the present methodology does not allow for partial feedback between the unsaturated and
saturated zone components of the model; the saturated zone is represented separately by an analytical
transport model.
The flow system in the vertical column is solved with the VADOFT component of the
RUSTIC model, which is based on an overlapping representation of the unsaturated and saturated
zones. The user must specify a water flux at the soil/liquid interface at the bottom of the
impoundment, which defines the top of the unsaturated zone in the model. In addition, a constant
pressure-head boundary condition must be specified for the bottom of the unsaturated zone beneath
the lagoon. This pressure-head is chosen to be consistent with the expected pressure head at the
bottom of the saturated zone, without consideration of the added flux leaking from the impoundment.
Transport in the unsaturated zone is then determined using the Darcy velocity (Vd) and saturation
profiles from the flow simulation. From these, the transport velocity profile can be determined.
-------
Although limited to one-dimensional flow and transport, the use of a rigorous
finite-element model in the unsaturated zone allows consideration of changes in water-table level and
also allows consideration of depth-variant physical and chemical processes that would influence the
mass flux entering the saturated zone. Among the more important of these processes are advection
(which is a function of the Darcy velocity, saturation and porosity), mass dispersion, adsorption of
the leachate onto the solid phase, and both chemically and biologically induced degradation. In the
unsaturated zone, groundwater seepage and associated transport of dissolved leachate are estimated
with one-dimensional Galerkin and upstream-weighted finite-element methods, respectively.
One-dimensional advective-dispersive transport is estimated with VADOFT based on the
estimated mass flux of contaminant into the top of the soil column, and a zero concentration
boundary condition at the bottom of the saturated zone. The mass flux of contaminant into the
saturated zone is evaluated at the water table based on the derived concentration distribution and the
Darcy velocity. The resulting mass flux from the VADOFT simulation is used as input for the
AT123D model, which simulates contaminant transport through the saturated zone. It is represented
as a mass flux boundary condition applied over an area representative of the facility's air-exposed
surface. Because of input requirements for the AT123D code, the source area of contaminant loading
to the saturated zone must be represented as rectangular in shape.
The transient nature of the flux can be represented by one of two model specifications.
For most applications, the flux can be estimated by a steady-state transport analysis in the
unsaturated zone. For those facilities for which closure can be reliably predicted, contaminant mass
flux to the saturated zone can be represented by (1) a plug source applied at the peak level given by
a transient analysis with pulse time defined by mass conservation constraints; or (2) a transient source
with time-dependent flux levels interpolated from the unsaturated zone simulation results. The
theory and methodology for estimating contaminant transport through the unsaturated zone are
discussed in greater detail below.
-------
4.3.2.1. FLOW OF WATER THROUGH THE UNSATURATED ZONE -- The one-
dimensional downward flow of groundwater through the unsaturated zone beneath an impoundment
is described by Richard's equation:
where:
K = saturated hydraulic conductivity (m/sec)
kpw = effective permeability (unitless)
V> = pressure head (m)
z = vertical coordinate in the unsaturated zone (m)
t = time (s)
and r) is the effective water storage capacity (m"1), defined as
>? = SHSS +
w s at
where:
Sw = the water saturation (unitless)
Ss = specific storage (m"1)
= the effective porosity (unitless)
This governing equation is solved subject to user-specified initial conditions and boundary
conditions. The initial (t = 0) pressure-head profile must be specified for all depths, and the fluid
flux at the top of the system and the pressure head at the bottom of the system must be assigned. All
this information must be supplied to obtain a unique solution to Richards' equation.
The system as described by these equations is non-linear, because saturation depends on
pressure-head, and permeability depends on saturation. To solve the governing equation, the
nonlinear relationships must be defined for the soil types of interest. VADOFT allows the use of
either of two functions to define the relationship between relative permeability and saturation (U.S.
EPA, 1989a). These two functions, which are taken from Brooks and Corey (1966) and van
Genuchten and Weirenga (1976), can be written as:
If —
-------
krw = se°-5 [ i - (i - se(i/T>n2
where n and T are empirically derived exponents. The effective water saturation (Se) is defined as:
S =(S -S ) / (1 - S )
where Swr is the residual water saturation. The empirical function in VADOFT that describes the
relationship between pressure head and saturation can be written as:
(1 / [1 + (a | V>-t/>a | f]T for V < i>a, or
Se =
[\ for V > V>
a
where i>a is the air entry pressure head, a and B are empirically derived and T is defined by
r = 1 - I//?. Statistical distributions for a, 13, r, and SH(. as derived by Carsel and Parrish (1988) are
given in U.S. EPA (1989b).
To represent the variably saturated soil column beneath the floor of the lagoon, the model
discretizes the column into a finite-element grid consisting of a series of one-dimensional elements
connected at nodal points. Elements can be assigned different properties for the simulation of flow
in a heterogenous system. The model generates the grid from user-defined zones; the user defines
the homogeneous properties of each zone, the zone thickness and the number of elements per zone,
and the code automatically divides each zone into a series of elements of equal length. The governing
equation is approximated using the Galerkin finite element method and then solved iteratively for
the dependent variable (pressure-head) subject to the chosen initial and boundary conditions.
Solution of the series of nonlinear simultaneous equations generated by the Galerkin scheme is
accomplished by either Picard iteration, a Newton-Raphson algorithm or a modified Newton-
Raphson algorithm. A detailed description of the solution scheme is found in U.S. EPA (1989a).
Once the finite-element calculation converges, the model yields estimated values for all the
variables at each of the discrete nodal points. The Darcy velocity and saturation values are used as
-------
input to the unsaturated zone transport module, which is used to derive the source term for the
saturated zone. From this source term, the saturated zone module determines contaminant
concentrations in groundwater at the selected well location.
4.3.2.2. UNSATURATED ZONE TRANSPORT -- The governing equation (U.S. EPA,
1989a) for one-dimensional advective-dispersive transport of a dissolved species in a variably
saturated soil can be written as:
where:
D = the dispersion coefficient (m2/sec)
c = the solute concentration (g/m3)
= the porosity (unitless)
RF = the retardation factor (unitless)
VD = the Darcy velocity (m/sec)
A = the first-order decay coefficient (sec"1)
D is the dispersion coefficient, defined as:
D = aL VD + ^ D*
where:
aL = longitudinal dispersivity (m)
D* = the effective molecular diffusion coefficient (m2/sec)
and RF is the retardation factor (unitless) defined as
RF = 1 + (pbkd / Sw)
where:
pb = the bulk density of wet soil (g/m3)
kd = the linear adsorption coefficient (m3/g)
SH = the water saturation (unitless)
-------
The governing differential equation for unsaturated zone transport is solved by VADOFT
subject to specified boundary and initial conditions (concentrations). The initial concentration
profile must be specified for all depths below the bottom of the site, down to the bottom of the
unsaturated (and saturated) zone.
In VADOFT, the one-dimensional advective-dispersive transport equation is approximated
with an upstream-weighted finite-element method, which reduces to a Galerkin approximation when
the weighting parameter is set to zero. The upstream weighting term is used to circumvent numerical
instabilities that may occur if the problem is advective dominant, as the weighting effectively adds
dispersion to the system (Huyakorn and Finder, 1983). Linear elements are used in the spatial
discretization, and time integration can be performed using either backwards or central differencing.
A more detailed description of the mathematical theory and numerical approximations used in
VADOFT can be found in U.S. EPA (1989a). Spatial discretization for the transport analysis parallels
the discretization described above for the flow analysis. Successful model execution requires that the
definition of elements and nodes be the same for the transport and flow analyses.
4.3.3. Contaminant Transport in the Saturated Zone. Two basic approaches are commonly used
for estimating contaminant concentrations and travel time (or velocity) in the saturated groundwater
flow system: analytical solutions and numerical modeling. Analytical solutions are relatively quick
and simple to use. However, they are based on a variety of simplifying assumptions related to
contaminant characteristics and the subsurface environment, and so provide only order-of-magnitude
estimates of contaminant travel time and concentration. Numerical models, on the other hand, are
far less restrictive with regard to simplifying assumptions; however, they typically require more data,
are time consuming to set up and run, and require expensive and specialized equipment and expertise.
Based on the selection criteria discussed earlier, the use of numerical models is not appropriate for
the methodology proposed in this document.
Prior to selecting the AT123D model for simulating contaminant transport through the
saturated zone near a surface disposal facility, the performance of that model with respect to
conditions representative of surface impoundments was compared with that of a fully three-
dimensional numerical model, and with that of an alternative analytical code. The numerical model
k-20
-------
used for the tests was SEFTRAN, a finite element flow and transport code, which is available from
Holcomb Research Institute and has been benchmarked for HRI and the U.S. EPA Office of Solid
Waste (Huyakorn et al., 1985). Its numerical solution allows contaminant migration underneath the
source to be influenced by a flow field gradient and surficial infiltration effects. The alternative
analytical code is similar to the analytical solutions currently used by the U.S. EPA Office of Solid
Waste to evaluate solute transport in the saturated zone for generic Monte Carlo studies (Huyakorn
et al., 1984; Lester et al., 1986; U.S. EPA, 1988d). It uses a partially penetrating vertical plane, or
"patch source," to represent the source term for unsaturated zone transport (Lester et al., 1986).
Unlike AT123D, this model represents its source term as a constant concentration boundary
condition; AT123D uses a mass flux. Over a range of distances from the source, depths, and
contaminant half-lives, the AT123D model compared quite favorably with the numerical code, and
generally out-performed the patch source analytical solution. This methodology therefore follows
U.S. EPA (1986a) in selecting the AT123D model for its simulation of the transport of contaminants
through the saturated zone.
Detailed description of the AT123D model are provided by U.S. EPA (1986a) and by Yeh
(1981) and will not be repeated here. In general, the model provides an analytical solution to the
basic advective-dispersive transport equation. One advantage of AT123D is its flexibility: the model
allows the user up to 450 options and is capable of simulating a wide variety of configurations of
source release and boundary conditions. For further details concerning AT123D, the reader is
referred to the landfill methodology document or to Yeh (1981). Some modifications to AT123D that
are specific to the modeling of surface impoundments are briefly described below.
As mentioned above, seepage from a sludge lagoon can result in an increased local influx of
water to an underlying aquifer, and can result in local elevation or mounding of the water table.
The AT123D model accepts as input the flux of pure contaminant mass entering the top of the
saturated zone, and does not consider the extent of the contaminant's dilution by water from the
source area, or the impact of that water on groundwater flow within the saturated zone. When the
vertical movement of contaminant through the unsaturated zone is due only to infiltration throughout
the area, the gradient within the aquifer is a function of the water entering the saturated zone, and
k-2]
-------
neglect of the diluted state of the source term may be valid. For the case of a surface impoundment,
however, neglect of the extent of the contaminant's original dilution could result in non-trivial
overestimation of the source concentration, leading to an overestimation of contaminant
concentrations at the receptor well. Furthermore, neglect of mounding effects could lead to incorrect
assumptions about the velocity of groundwater flow near the site.
The present methodology addresses these concerns with three simple adjustments to the
execution of the AT123D model. First, to correct for AT123D's potential overestimation of the
original concentration of contaminant at the aquifer's boundary, the mass flux estimated from
VADOFT results is adjusted by a dilution factor (Df) as follows:
where Fa is the volume of fluid passing through a vertical cross section of the aquifer (m3/sec),
oriented perpendicular to the direction of flow, and having a width equal to the source width and a
depth equal to the saturated thickness of the aquifer. FS is the volume of fluid leaving the
impoundment (m3/sec). In cases where seepage from the lagoon is not significant compared with the
natural, regional rate of aquifer recharge, this dilution adjustment is inappropriate, and can be
inactivated for program execution.
The excess water released by leakage from a surface impoundment can also result in a
mounding of the water table beneath the impoundment, so that the lagoon superimposes a radial
velocity field on the background or regional velocity field of groundwater flow. In other words, the
horizontal velocity of water within the aquifer can be slowed up-gradient of the lagoon, and
accelerated down- gradient of the site. This change in the velocity field might result in reduced
time of travel for contaminants moving to receptor wells down-gradient of the impoundment site.
A reduction in time of travel could in turn lead to reductions in contaminant degradation prior to
human exposure. Accurate accounting of the influence of mixing and degradation would require a
fully three-dimensional flow and transport model; this methodology uses a simpler approach to
estimate a conservative limit to contaminant decay within the system. The limit is estimated by
4-22
-------
increasing the estimated velocity of groundwater flow to account for the maximum downgradient
increase in velocity due to the source. The velocity increase can be approximated by idealizing the
lagoon as a circular source, so that the rate at which seepage passes outward through a cylinder
beneath the perimeter of the lagoon's floor will be V{ = q,-r/d, where q. is the infiltration rate from
the impoundment, r the radius of a circle of area equal to that of the impoundment, and d the
aquifer depth.
In addition to increasing the expected velocity of contaminant transport through the aquifer,
this superimposed velocity would also have the effect of increasing AT123D's estimate of
contaminant dilution within the aquifer. This additional dilution effect must be subtracted back out
of the model calculations, since the true dilution is explicitly included in the factor introduced by
Equation 4-2. The model performs this calculation automatically, based on the following equation
for the anti-dilution factor:
Daf - / Vh
where Vv is the vertical velocity due to the source, and Vh is the regional velocity of horizontal
groundwater flow.
It should be noted that the above methodology is conservative, since it overestimates the
velocity beneath the source and does not allow for decreases in the superimposed velocity beyond the
source. As a result, the methodology is more conservative than a three-dimensional model. In
comparison with a two-dimensional cross-sectional flow and transport model, the model is more
conservative beneath the source, but less conservative (but still conservative) beyond the source.
By combining the VADOFT model with AT123D, and by adjusting calculations in AT123D
to accommodate the dilution and superimposed velocity described above, concentrations of a
contaminant in groundwater at a receptor well (Cgw) can be predicted as a function of the liquid
concentration of contaminant near the floor of the lagoon, the rate of seepage from the lagoon, and
geohydrological characteristics of the area. It should be noted that all of the calculations described
above are linear with respect to contaminant concentrations in liquid seeping from the lagoon. As
-------
will be illustrated in Section 4.7, it is convenient to perform the calculations based on the assumption
of a "unit" concentration of dissolved contaminant within the facility.
The above discussion described a method for estimating concentrations of contaminant in
groundwater at a specified well location down-gradient of the surface disposal site. As mentioned
in Chapter 3, an additional pathway of potential exposure associated with groundwater contamination
involves the discharge of contaminants from the aquifer to the surface water body, with potential
subsequent exposure to humans and wildlife. Estimation of potential exposure through these
pathways requires a method for estimating the expected loading of contaminants to the surface water
body. This loading can occur throughout the area at which the plume of contaminated aquifer flow
intersects the stream bed. Concentrations of contaminant in groundwater entering the stream can
be expected to diminish toward the edge of the plume. As will be explained in Chapter 5, however,
only the total mass loading of contaminant to the stream is of concern for methodologies to estimate
potential exposure to humans and wildlife. With some simplifying assumptions based on a mass-
balance approach, results from AT123D can be used to predict steady-state mass loadings to the
surface water body.
Unless available geohydrological data suggest otherwise, it is assumed that the direction of
groundwater flow is perpendicular to the stream, and that the stream intercepts the entire plume of
contaminated groundwater. For contaminant species without decay (i.e. metals), it is assumed that
the steady-state loading to the stream is the same as that seeping from the lagoon and loaded into the
aquifer. For non-conservative species, contaminant mass will be lost to degradation during transport
through the aquifer, so that the loading to the stream will be less than the predicted loading to the
aquifer.
One approach to estimating the extent of this degradation would be to estimate the average
time of travel for contaminant mass through the aquifer to the surface water body. This time
estimate could be derived from the retarded Darcy velocity calculated for contaminant in the aquifer,
together with an assumption regarding the average distance traveled (represented, perhaps, by the
shortest distance from the lagoon to the surface water body). Contaminant loss to degradation could
-------
then be estimated based on assumed rates of first-order decay for the subject contaminant in the
saturated zone.
A flaw in that approach is that it ignores the influence of longitudinal dispersion, which can
affect the average time of travel for contaminant mass in the aquifer. An alternative approach is to
use the AT123D model, constrained such that the effects of transverse and vertical dispersion are
ignored. The loading of contaminant to the aquifer is represented by a rectangular prismatic source
with width and length corresponding to the width and length of the lagoon, and with depth
corresponding to ihe depth >f the aquifer. If the aquifer is further specified as having a width
corresponding to the width rf the lagoon, then the estimated contaminant concentrations at the
selected distance will approximate the expected average concentration of contaminated groundwater
entering the stream. This concentration can then be multiplied by the corresponding volume of
vater flow in the cross-section of the aquifer under consideration, which can be estimated based on
the Darcy velocity of the relevant aquifer medium. Results from AT123D (based on unit
concentrations of the contaminant in the lagoon) can be used to derive a ratio between expected
loading of contaminant to :iie stream iWy ana the concentration of the contaminant in the lagoon
sC,,. Use of this ratio for deriving criteria for surface water pathways will be further discussed ia
Chapter 5.
4.4. INPUT PARAMETER REQUIREMENTS
Various input Darameters are needed to define the geohydrologic system through which the
dissolved species :s transported, and to define the behavior of individual chemical contaminants in
'he sludge. Reasonaole worst case parameter values should be selected when deriving national (Tier
.) criteria, but .nore accurate, site-specific values should be used for Tier 1 calculations. Four
categories of input parameters are required: (1) parameters defining the source term, (2) parameters
describing characteristics of the unsaturated zone, (3) parameters describing characteristics of the
saturated zone, and (4) chemical-specific parameters.
4.4.1. Source Term. The following parameters are needed to estimate the loading of
contaminant to the unsaturated zone beneath a site:
i. The rate of seepage from the bottom of the surface disposal facility should be estimated
for each site; alternatively, conservative default values can be used.
-------
2. Linear adsorption coefficients for each contaminant in sludge can be taken from
Appendix C of U.S. EPA (1986a) or other available literature, can be determined
empirically for the sludge matrix received by a particular facility, using methods
described in DeWald and Phillips (1989), or can be estimated with the CHEMEST
procedures available within GEMS (U.S. EPA, 1988c).
4.4.2. Unsaturated Zone. The following parameters are needed to estimate water flow and
contaminant transport through the unsaturated zone:
1. Stratigraphic variation in soil and rock types within the defined property zones should
be determined from the site plan and borings.
2. Depth to groundwater (without consideration of local mounding caused by seepage)
should be determined from the site plan and borings.
3. Saturated hydraulic conductivity and effective porosity of various soil and rock types
should be calculated from field measurements, or can be taken from Freeze and Cherry,
1979) or other sources.
4. Empirical parameters needed to define the effective permeability-saturation relationship
and saturation-pressure head relationship or sets of data points describing the effective
permeability-saturation curves and the saturation-pressure head curves for each soil
and rock type being modeled are available from U.S. EPA (1989b), or Carsel and Parrish,
(1988).
5. Distributions of residual saturations and air entry pressure heads for each soil and rock
type are available from U.S. EPA (1989b).
6. In the absence of site-specific values, longitudinal dispersivity can be estimated as one-
tenth the thickness of the vadose zone (Carsel and Parrish, 1988).
7. Linear adsorption coefficients and decay coefficients for each material zone can be
taken from Appendix C of U.S. EPA (1986a); alternatively, absorption coefficients can
be derived empirically, using methods described in Appendix B of that reference, or can
be estimated with the CHEMEST procedures in GEMS. If desired, and if sufficient data
are available, decay coefficients can also be used to represent pseudo-decay processes
for heavy metals, including the formation of stable precipitates with carbonates, sulfates,
hydroxides, etc.
4.4.3. Saturated Zone. The following input parameters are needed to simulate contaminant
transport through the saturated zone:
1. The geometry of the region of interest including thickness of the saturated zone and
lateral extent of the aquifer should be determined from the facility site plan, or from
geologic survey reports.
2. The areal extent of the source can be determined from the facility site plan.
3. Effective porosity and bulk density of the aquifer material can be determined from soil
measurements, or taken from Appendix C of U.S. EPA (1986a).
k-26
-------
4. Linear adsorption coefficients and decay coefficients for each material zone can be
taken from Appendix C of U.S. EPA (1986a); alternatively, adsorption coefficients can
be derived empirically (U.S. EPA, 1986a) or other available literature, or can be
estimated with the CHEMEST procedures in GEMS.
5. Hydraulic conductivity can be determined from field tests or taken from Appendix C
of U.S. EPA (1986a).
6. Longitudinal, transverse and vertical dispersivities should be calibrated against field data,
extrapolated from values calculated for similar aquifer systems, or drawn from ranges
supplied in Yeh (1981), calculated with methods presented in U.S. EPA (1986d).
7. Mass flux as determined by the unsaturated zone analysis.
8. Distance from the impoundment to the property boundary or surface water can be
determined from the site plan.
4.5. HEALTH AND ENVIRONMENTAL EFFECTS
Toxic pollutants that leach from surface disposal sites into groundwater can cause adverse
health effects to nearby residents who use wells as a source of drinking water. Wildlife could be
exposed directly to the surface of the lagoon, but are unlikely to be exposed to leached contaminants
until the contaminated groundwater discharges into streams or other surface water bodies. Potential
contamination of surface water through this pathway, and its resulting risks to human health and
wildlife, will be discussed in Chapter 5.
The first step in deriving sludge criteria to prevent risks to human health from contaminated
groundwater is to determine a reference water concentration (RWC) for each sludge contaminant of
concern. The procedure for determining the RWC differs according to whether the pollutant acts by
a threshold or non-threshold mechanism of toxicity. Each of these two types of toxicant will now
be discussed.
4.5.1. Threshold-Acting Toxicants. Threshold-acting toxicants are those for which a dose can
be identified below which no adverse effects are expected to occur. The Agency assumes that all
non-carcinogenic chemicals act according to threshold mechanisms. The RWC is derived as follows
for threshold-acting toxicants:
RWC = [(RfD BW RE'1) - TBI] / Iy
-------
where:
RWC = reference water concentration (mg//)
RfD = reference dose (mg/kg/day)
BW = human body weight (kg)
RE = relative effectiveness of ingestion exposure (unitless)
TBI = total background intake rate of contaminant from all other sources of exposure
mg/day)
Iw = total water ingestion rate (//day)
The definition and derivation of each of the parameters used to estimate RWC for threshold-acting
toxicants are further discussed in the following sections.
4.5.1.1. REFERENCE DOSE (RfD) -- For threshold-acting toxicants, the U.S. EPA
establishes a "Reference Dose" (RfD), defined as the lifetime dose of a chemical that is likely to be
without appreciable risk of deleterious effects. The RfD is expressed in milligrams of dose per
kilogram of body weight. Procedures for estimating reference doses from various types of
toxicological data were outlined in U.S. EPA (1980c), and more recently in U.S. EPA (1990). Values
of RfD for noncarcinogenic or systemic toxicity have been derived by several groups within the
Agency. An intra-Agency Reference Dose Work Group now verifies these values by consensus for
use by the various Agency programs. Most of the non-carcinogenic chemicals that are currently
candidates for sludge criteria for surface disposal are included in the Agency's verified RfDs, and
thus no new effort will be required to establish RfDs for deriving sludge criteria. For any chemicals
not so listed, RfD values should be derived according to established Agency procedures (U.S. EPA,
1990).
4.5.1.2. HUMAN BODY WEIGHT (BW) -- Values for human body weight vary widely
among individuals according to age and sex. The variations in mean body weight with age and sex
for the U.S. population are illustrated in Table 4-4. The choice of values for use in the risk
assessment depends on the individual defined as the most exposed individual, or MEI, which in turn
depends on exposure and susceptibility to adverse effects. In cases where effects on which the RfD
is based may occur after cumulative lifetime exposure, it would be reasonable to base derivation of
criteria on adult values of body weight. In circumstances where effects have a shorter latency, or
where children are known to be at special risk, it may be more appropriate to base the derivation of
4-28
-------
TABLE 4-4
Water Ingestion and Body Weight by Age-Sex Group in the United States
Age-Sex Group
6- 1 1 months
2 years
14-16 years, females
14 — 16 years, males
25-30 years, females
25-30 years, males
60-65 years, females
60-65 years, males
Mean Water
Ingestion
(m//day)
308
436
587
732
896
1050
1157
1232
Median
Body Weight
(kg)
8.8
13.5
51.3
54.2
58.5
67.6
67.6
73.9
Water Ingestion
per Unit Body Weight
(m//kg-day)
35.1
32.2
11.4
13.5
15.3
15.5
17.1
16.7
*Source: U.S. EPA, 1986a
-------
criteria on the body weight of toddlers or infants. The approach used in the derivation of Maximum
Contaminant Level Goals (MCLGs) by the U.S. EPA Office of Drinking Water is to assume an adult
body weight of 70 kilograms.
4.5.1.3. TOTAL WATER INTAKE RATE (IJ -- Total daily ingestion of water varies
widely among individuals according to age and sex. Table 4-5 shows the variation of adult drinking
water intake within and among several studies. Mean intakes in New Zealand, Great Britain, the
Netherlands and Canada varied from 0.96 to 1.30 liters per day. More recently, U.S. EPA (1989e)
has presented conservative and average water consumption values of 2 //day and 1.4 //day,
respectively. The choice of values for use in the risk assessment depends on the individual defined
as the MEI, which in turn depends on exposure and susceptibility to adverse effects. As shown in
Table 4-4, the total water intake on a body-weight basis is substantially higher for infants and
toddlers than for adults. Therefore, infants and toddlers would be at greater risk of exceeding the
RfD when exposure is by drinking water. However, in cases where effects on which the RfD is
based may occur after cumulative lifetime exposure, it would be reasonable to base derivation of
criteria on adult values of water intake. In circumstances where effects have a shorter latency, or
where children are known to be at special risk, it may be more appropriate to base the derivation of
criteria on the daily consumption rate of toddlers or infants. The approach used in the derivation
of Maximum Contaminant Level Goals (MCLGs) by the U.S. EPA Office of Drinking Water is to
assume a total water ingestion rate (IJ of 2.0 liters per day for adults. This value exceeds the 90th
percentile estimates from the studies presented in Table 4-5 and thus represents a relatively high
estimate of exposure, although the total range can extend well above this value, at least on a given
sampling day. On a body-weight basis, however, this value represents an intake of 28.6 ml/kg/day,
a value lower than the mean intake for infants and toddlers. Therefore, water intake values for
children should be used in cases where children are at greater risk than adults. For example, the
MCLG for lead, a toxicant known to threaten children, was based on exposure for infants rather than
adults (U.S. EPA, 1985e).
4-30
-------
.t-
I
Maan± sd
1.25± 0.39
0.96± 0.57
1.08
1.26
1.30
TABLE 4-5*
Daily Intakes of Drinking Water by Adults
(I/day)
Bercentiles
Range 10th 50th 90th n
0.26-2.80 0.42 1.26 1.90 109
0.00-6.47 0.40 0.87 1.64 1960
0.50 — 1.68 2596
1472
342
Method
Duplicate
sampling
24-hour
recall
Diary
Country
New Zealand
New Zealand
Great Britain
Netherlands
Canada
Source: Gillies and Paulin, 1983
-------
4.5.1 .4. TOTAL BACKGROUND INGESTION RATE (TBI) - - Sources of exposure other
than sludge reuse or disposal practice may exist, and total exposure from all sources should be
maintained below the RfD. Other sources of exposure include background levels (whether natural
or anthropogenic) in drinking water, food or air. Other types of exposure that are due to occupation
or habits such as smoking might also be included, depending upon data availability and regulatory
policy. These exposures are summed to estimate the total background ingestion rate.
Data for estimating background exposure usually are derived from analytical surveys of
surface, ground or tap water, from Food and Drug Administration market basket surveys and from
air-monitoring data. These surveys may report estimates of central tendency, percentiles or ranges.
Estimates of the TBI may be based on either estimates of central tendency or upper-bound estimates.
Data chosen to represent the TBI of the most exposed individual should be consistent with other
characteristics of the MEL For example, if the body weight of a child is used in the derivation of
criteria, then the TBI used should be based on the TBI estimated for children. Background data
reported as concentrations in air, food or water can be converted to values for background intake by
using appropriate estimates of ingestion or inhalation rates. Data reported for adults can be used to
estimate intakes for children, using adjustments based on relative food, water or air intakes.
The TBI is the sum of all possible background exposures except those exposures resulting
from the sludge disposal practice. The "effective" TBI is the sum of all background exposures
weighted by the relative effectiveness of the route through which each exposure occurs. The
effective TBI can be derived as follows:
.a*,. 5W.2-. ..... ^
where:
TBI = total background intake rate of contaminant from all other sources of exposure
(mg/day)
BI = background intake of pollutant from a given exposure route, indicated by the
subscript (mg/day)
RE = relative effectiveness, with respect to ingestion exposure, of the exposure route
indicated by the subscript
4-32
-------
When TBI is subtracted from the weight-adjusted RfD, the remainder (after adjusting for
RE) defines the increment that can result from sludge disposal without exceeding the RfD. The
magnitude of the impact on the exposed population resulting from sludge-related exposures that
exceed this increment will depend on the data used to determine the TBI. If the TBI is based on
upper-bound data, such as the 95th percentile data, then sludge-related exposures exceeding this
increment would imply that about 5% of the exposed population may approach or exceed the RfD.
If the TBI is based on an estimate of central tendency, such as the median, then about 50% of the
exposed population would be expected to approach or exceed the RfD.
4.5.1.5. RELATIVE EFFECTIVENESS OF EXPOSURE (RE) — The relative
effectiveness (RE) of exposure shows the relative toxicological effectiveness of an exposure by a
given route when compared with another route. The value of RE may reflect observed or estimated
differences in absorption between the inhalation and ingestion routes that can significantly affect the
quantity of the pollutant that reaches the target tissue, the time required for the contaminant to reach
the target tissue and the degree and duration of the effect. The RE factor may also reflect
differences in the occurrence of critical toxicological effects at the point of entry. For example,
carbon tetrachloride and chloroform were estimated to be 40% and 65% as effective, respectively, by
the inhalation route as by the ingestion route (U.S. EPA, 1984a,b). In addition to route differences,
RE can also reflect differences in bioavailability resulting from the exposure matrix. For example,
absorption of nickel via the inhalation route has been estimated to be 5 times the absorption when
ingested in the diet (U.S. EPA, 1985a). The presence of food in the gastrointestinal tract may delay
absorption and reduce the availability of orally administered compounds, as demonstrated for the
halocarbons (NRC, 1986).
Since exposure from drinking contaminated water is through oral ingestion, the RE factors
applied in the above equations represent relative effectiveness of exposure routes and matrices from
which the RfD was derived when compared with exposure via food or drinking water. In these
equations, the RE factors show the relative effectiveness, with respect to the oral route, of each
background exposure route and matrix.
14-33
-------
An RE factor should be applied only where well-documented, well-referenced information
is available on the contaminant's observed relative effectiveness. When such information is not
available, the RE is assumed to be equal to 1.
4.5.2. Carcinogens. For carcinogenic chemicals, the Agency considers the excess risk of cancer
to be linearly related to dose, except at high dose levels (U.S. EPA, 1986c). The threshold
assumption, therefore, is not applicable to carcinogens, since incremental carcinogenic risk from the
exposure diminishes with dose but does not become zero until dose becomes zero.
The decision whether to treat a chemical as a threshold-acting or carcinogenic agent depends
on the weight-of-evidence that the chemical is carcinogenic in humans. Methods for classifying
chemicals as to their weight-of-evidence have been described in U.S. EPA (1986c), and most of the
chemicals that are candidates for sludge criteria development have been classified in Health
Assessment Documents or other reports prepared by EPA's Office of Health and Environmental
Assessment (OHEA), or in connection with the development of Maximum Contaminant Level Goals
(MCLGs). To derive values of the adjusted reference intake, a decision must be made as to which
classifications constitute sufficient evidence for using a quantitative risk assessment on the
presumption of carcinogenicity. Chemicals are classified as follows: an "A" designation is given to
a known human carcinogen; a "B" classification is given to a probable human carcinogen; a possible
human carcinogen is designated as "C"; a "D" chemical is "not classifiable because of inadequate
animal and human data"; and an "E" designation is given to chemicals for which there is evidence of
non-carcinogenicity in humans. In general, "A" and "B" chemicals are viewed as important hazard
levels for public health concerns, while "D" and "E" chemicals are not. "C" chemicals have received
varying treatment. For example, the regulatory action on lindane, classified by the Human Health
Assessment Group (HHAG) of the U.S. EPA as a "B2-C" (that is, between the lower range of the B
category and the C category), has been based on both carcinogenic risk, (U.S. EPA, 1980b) and
threshold effects (U.S. EPA, 1985e). Table 4-6 gives an illustration of the EPA classification system
based on available weight-of-evidence.
The use of weight-of-evidence classification must also be coupled with information regarding
the route by which the chemical demonstrates carcinogenicity. For example, some forms of nickel
-------
VjJ
vn
TABLE 4-6
Illustrative Categorization of Evidence Based on Animal and Human Dataa
-------
have been shown to be carcinogenic by inhalation but not by ingestion (U.S. EPA, 1985a). Similarly,
arsenic has been shown to cause carcinogenic effects when certain inorganic forms are ingested in
water, but no carcinogenic potential has been demonstrated for the organic forms of arsenic
commonly present in many foods. The issue of whether to treat a chemical as carcinogenic by
ingestion is controversial for many chemicals.
For pollutants assessed as carcinogenic (non-threshold acting) agents, the reference water
concentration (RWC, in mg//) in groundwater used for drinking is derived as follows:
RWC = (RL BW)/(q1* RE Iy)
where:
RWC = reference water concentration (mg//)
RL = risk level (unitless)
BW = human body weight (kg)
q,,* = human cancer potency (mg/kg-day)"1 or (kg-day/mg)
RE = relative effectiveness of ingestion exposure (unitless)
IH = total water ingestion rate (//day)
The RWC, in the case of carcinogens, is thought to be protective, because the estimate of
carcinogenicity is an upper limit value. The parameters BW, RE and Iw are the same for carcinogens
and for threshold-acting toxicants; these parameters are defined and described in Section 4.5.1. The
definition and derivation of q^ and RL are discussed in the following sections.
4.5.2.1. HUMAN CANCER POTENCY (q.,*) -- For most carcinogenic chemicals, the
linearized multistage model is recommended for estimating human cancer potency from animal data
(U.S. EPA, 1986c). When epidemiologic data are available, potency is estimated based on the
observed relative risk in exposed vs. unexposed individuals, and on the magnitude of exposure.
Guidelines for the use of these procedures have been presented in U.S. EPA (1980c) and U.S. EPA
(1985e) and in each of a series of Health Assessment Documents prepared by OHEA (such as U.S.
EPA, 1985b). The true potency value is considered unlikely to be above the upper-bound estimate
of the slope of the dose-response curve in the low-dose range, and is expressed in terms of risk per
dose, where dose is in units of mg-day/kg. Thus, the q,* has units of (mg-day/kg)"1 or kg/mg-day.
14-36
-------
OHEA has derived potency estimates for each of the potentially carcinogenic chemicals that are
currently candidates for sludge criteria. Therefore, no new effort will be required to develop potency
estimates to derive the sludge criteria.
4.5.2.2. RISK LEVEL (RL) -- Since by definition no "safe" level exists for exposure to non-
threshold agents, values of RWC are calculated to reflect various levels of cancer risk. If the RL is
set at zero, then the RWC will be zero. If RL is set at 10~6, the RWC will be the concentration that,
for lifetime exposure, is calculated to have an upper-bound cancer risk of 1 case in 1 million exposed
individuals. This risk level refers to excess cancer risk, that is, risk over and above the background
cancer risk in unexposed individuals. By varying RL, the RWC may be calculated for any level of
risk in the low-dose region (i.e. for RL less than 10"2). Specification of a given risk level on which
to base regulations is a matter of policy, as is the usefulness of the upper bound concept in risk
derivation. Therefore, it is common practice to derive criteria representing several levels of upper
bound risk without specifying any risk level as "acceptable."
4.6. DERIVING CRITERIA
Section 4.3.1 described a method for predicting contaminant concentrations in seepage as a
function of the dry-weight concentration of contaminant in sludge received by an impoundment.
Sections 4.3.2, 4.3.3, and 4.3.4 described calculations that used the expected contaminant
concentration in seepage, together with the quantity of seepage, to predict expected contaminant
concentrations in groundwater at the property boundary. Section 4.5 presented methods for deriving
reference water concentrations (RWC) for groundwater. The final step in criteria derivation is to
combine these results to link dry-weight contaminant concentrations in sludge to expected well-
water concentrations and health effects, or conversely, to link reference water criteria to maximum
allowable dry-weight concentrations. In this context, it is useful to define a "source-receptor ratio"
or SRRGU, that represents the ratio of contaminant concentration in well-water (mg//) to dry-weight
concentration of contaminant in sludge received or accumulated in a facility (mg/kg). Since all of
the calculations described in Sections 4.3 are linear with respect to contaminant concentrations,
SRRGU (kg//) can be easily derived by combining results from the calculations and model results
described above:
-------
SRRGU = (Cgw/Ct)(Ct/N) = Cgw/N (4-3)
where:
SRRQW= the source-receptor ratio (kg//)
CHU = predicted concentration of contaminant in groundwater at a receptor well location
y w
(mg//)
Ct = predicted concentration of contaminant in liquid phase near the floor of the lagoon
(mg//)
N = dry-weight concentration of contaminant in sludge received by or accumulated in
the facility (mg/kg)
This ratio can then be used to estimate the maximum allowable concentrations of contaminant
in sludge received by a facility:
Nmax = RWC/SRRGU (4-4)
where:
Nmax = the maximum allowable dry-weight concentration of contaminant to be received
by or accumulated in the impoundment (mg/kg)
RWC = reference water concentration (mg//)
If background groundwater concentration levels for the contaminants of interest are
measurable, these should be incorporated into the calculation:
- C)/SRR
GU
where Cb is the background concentration for the contaminant in groundwater. Average or
reasonable worst-case background concentrations for groundwater in the United States can be used
to derive national criteria.
4.7. SAMPLE CALCULATIONS
Sample calculations are provided for Tier 2 analysis of three actual surface disposal facilities:
a wastewater treatment and sludge storage facility in New Hampshire, a sludge storage facility in
Oklahoma and a sludge disposal facility in Oregon. These three sites were chosen for sample
4-38
-------
calculations because they were thought to represent the considerable variability in facility types that
must be considered for actual application of the methodologies presented in this document. Two
contaminants, lead and benzene, will be considered for the sample calculations. Lead was chosen to
represent the use of the methodology for deriving criteria for a heavy metal, and because lead
concentrations may be limiting in many actual applications. Benzene has been chosen to illustrate
the application of the methodology for volatile organic contaminants; its volatility will be of special
interest in sample calculations presented in Chapters 5 and 6. It will be assumed that all three
facilities fail national (Tier 1) criteria, so that Tier 2 calculations can be described.
Tier 2 calculations rely on computer models that estimate expected contaminant concentrations
in groundwater as a function of liquid contaminant concentrations in sludge entering the
impoundments. Although the computer model is designed to estimate concentrations in groundwater,
results can also be used for "reverse" calculations of sludge concentrations as a function of RWC, since
ultimate estimates of groundwater concentrations are in all cases a linear function of concentrations
in the impoundments, which are in turn assumed to be a linear function of dry weight sludge
concentrations. For simplicity, site-specific simulations of contaminant transport in groundwater
derived for these sample calculations will be based on unit concentrations of contaminant in seepage
beneath the facility to be considered. Results from a simulation based on unit concentrations will
be used to derive a source-receptor ratio (SRRGU) that describes the ratio of contaminant
concentrations in groundwater at the receptor well to contaminant concentrations in sludge received
by the lagoons. From this ratio, criteria for maximum allowable sludge concentrations at these sites
will be derived.
Execution of the VADOFT and AT123D models for groundwater simulations requires values
for numerous input parameters descriptive of the site under consideration. For the sample
calculations presented here, some of the required data were taken from site visits, interviews with
plant operators, inspection of site plans and core sample results. Other data were derived from
literature values judged to be representative of site conditions; many of the ground material
parameters have been estimated from data obtained from general data for soils of similar
-------
characteristics (Carsei and Parrish, 1988;. The following secnoni cnef/y discus^ input parameters.
model results and criteria derivation for each of the three surface disposa. sites considered.
4.7.1. Analysis of Exposure for the Most Exposed Individual
4.7.1.1. ANTRIM, NEW HAMPSHIRE -- The Antrim wastewater treatment plant is a small
facility with a design capacity of 0.23 million gallons per day (MOD), that uses aerated lagoons as
its method of wastewater treatment. Unlike the other impoundments tr be considered in sample
calculations, those in the Antrim facility receive wastewater, not sludge. Since sludge remains in the
bottom of the lagoon for years or decades between emptying, however, these wastewater treatment
lagoons have been included in the sample calculations presented by this document.
The Antrim facility contains three lagoons lined with polyvinyi chloride (PVC): one with top
dimensions of 125 by 232 feet (2700 m2 area) and two smaller lagoons, each with dimensions of 101
by 125 feet each (1173 m2 area). The facility was built into the side of a hill, on a slope leading to
the Contoocook River (about 100 meters from the site). Each lagoon is about 10 feet deep (3 meters)
and surrounded by about 1 meter of vertical freeboard. Observation wells near the perimeter of the
lagoons allow monitoring of groundwater level. The water table occasionally rises to a level above
the bottom of the lagoons; a sump pump is available to lower the water table beneath the facility if
groundwater threatens to damage the liner of a partially filled lagoon.
After 8 years of wastewater treatment, the floors of the impoundments have collected about
14-18 inches (36-46 cm) of sludge. The facility operator does not plan to remove the sludge until
it begins to interfere with wastewater treatment; that condition is not expected for another 20 years
or more.
Both Tier 1 and Tier 2 calculations require the derivation of a reference water concentration,
or RWC. Since benzene is a carcinogen with an established potency value, the RWC (in mg/day) for
drinking is derived as follows:
RWC = (RL BW) / (q,* RE IH)
-------
For this sample calculation, a risk level of 10"6 is selected. With an assumed body weight of
70 Kg. an average rate of total water ingestion of 2 liters per day, a relative effectiveness of 1 for
benzene through ingestion, and a human cancer potency (q.,*) of 2.9 x 10"2 (kg-day/rag), the RWC
for benzene is calculated to be:
RWC = [(1(T6)(70)] / [(2.9xi(T2Xl)(2)] - i.2xl
-------
TABLE 4-7
Input Parameters for VADOFT Simulation of Flow and Contaminant Transport
Through the Unsaturated Zone
Antrim, New Hampshire
Parameter
Source area
Distance to bottom of saturated zone
Input Parameters for Flow Calculations
Flux at Top Node
Head at bottom node
Hydraulic conductivity
Effective porosity
Specific Storage
Residual water saturation
Power index (N)
Leading coefficient
Power index (j9)
Power index (7)
Input Parameters for Transport Calculations
Concentration of Benzene, Lead
Mass flux of contaminant to top node
Head at bottom node
Longitudinal dispersivity
Effective porosity
Retardation coefficient
Molecular diffusion coefficient
Default Darcy velocity
Default water saturation
Solute decay constant (benzene)
Solute decay constant (lead)
Value
5040
17
1.3
15
0.30
0.43
0
0.105
-1.0
14.5
2.68
0.62
1
1.3xlO'6
0
1.0
0.43
1.28
0
0
1.0
2.64xlO"4
0
(Units)
(m2)
(m)
(//m2-hr)
(m)
(m/hr)
(unitless)
(unitless)
(unitless)
(unitless)
(unitless)
(unitless)
(unitless)
(mg/f)
(kg/m2-hr)
(m)
(m)
(unitless)
{unitless)
(unitless)
(m/hr)
(unitless)
(hr'1)
(hr'1)
-------
benzene, the retardation factor is calculated based on an assumed distribution coefficient (kd) of 0.08
//kg (Hounslow, 1983), a bulk density of 1.51 kg//, and a porosity of 0.43:
RF = 1 + (ktf) = 1 + [(0.08)(1. 51)/(0.43)] = 1.28
Baes et al. (1984) reports that kd values for lead range from 4.5-7640 //kg. For these sample
calculations, a value of 234 //kg (appropriate for sand) is taken from Appendix C of U.S. EPA,
1986a, so
RF = 1 + (kb/0) = 1 + [(234)(1.51)/(0.43)] = 823
Benzene is assumed to have a decay coefficient of 2.64xlO~4 (hr~1) in the unsaturated zone; lead is
assumed to have a decay coefficient of zero.
The flux of contaminant mass entering the unsaturated zone from the impoundment can be
described by the rate of seepage (in kg/m2-hour) multiplied by the concentration of contaminant in
the seepage (in kg/kg). These calculations assume that the seepage contains contaminant at a unit
concentration of 1 mg//, so that each liter of seepage contains IxlO"6 kg of contaminant.
Contaminant flux into the unsaturated zone is then equal to the rate of seepage (1.3 kg/m2-hour)
multiplied by the contaminant concentration (in kg/kg) to yield 1.3 x 10"6 kg/m2-hour. For
V ADOPT and AT123D calculations, the three lagoons in Antrim are modeled in aggregate and are
idealized as a single square lagoon with the appropriate total surface area (5040 m2).
VADOFT simulates vertical flow and contaminant transport from the floor of the
impoundment to the bottom of the saturated zone. The total thickness of the soil layers to be
simulated by VADOFT is therefore equal to the sum of the thicknesses of the unsaturated and
saturated zones. For Antrim, it is assumed that the depth to groundwater is about 2 meters and that
the aquifer thickness is 15 meters, resulting in a total distance of 17 meters from the floor of the
lagoon to the bottom of the saturated zone.
-------
Based on these and other parameter values listed in Table 4-7, the VADOFT model produces
the results contained in Table 4-8. As can be seen from the table, VADOFT predicts that the water
table beneath the site will be elevated an additional 0.3 meters as a result of seepage from the
impoundment. The flux of contaminant mass leaving the unsaturated zone is estimated to be
1.2xlO"6 kg/m2-hour for benzene. This value, multiplied by the area of the site (5040 m2) yields
the total rate of benzene release to the saturated zone, or 5.9x10"3 kg/hour. For lead, the mass flux
is l.SxlO"6 kg/m2-hour (the same as the flux of lead in seepage from the lagoon), which yields a total
mass release rate of 6.4xlO"3 kg/hour. These values are used as input for the simulation of
contaminant transport through the saturated zone, as performed by AT123D.
Table 4-9 lists values for input parameters required for the AT123D simulation. As can be
seen from the table, the waste release rate for both benzene and lead is taken from results generated
by the VADOFT model. Values for the beginning and ending of the x-source and y-source locations
represent the dimensions of an idealized square impoundment with area defined by the total area of
the three lagoons at the Antrim facility. Other values are selected to be representative of an aquifer
consisting of sand. The superimposed velocity term (1.7xlO'3 m/hr) represents a conservative
estimate of the increment to groundwater velocity caused by seepage from the impoundment.
Groundwater concentrations at a well location 152 meters downgradient of the site (or 221 meters
from the far edge of the lagoon) were assumed to reflect maximum concentrations likely to be
encountered by a most exposed individual. For estimating potential exposure through the drinking
water pathway, it is assumed that groundwater does not flow in the direction of the Contoocook
River, which is located on the property boundary at a lesser distance.
Results from the AT123D simulation are presented in Table 4-10. The table shows that
steady-state concentrations of 4.2xlO"3 mg// for benzene and 0.38 mg// for lead are predicted at the
receptor well, based on 1 mg// unit concentrations of each contaminant in the impoundment. These
concentrations can be used to generate ratios of well concentrations to impoundment concentrations
for each of the two contaminants. For benzene:
Cnu / C, = 0.0042 / 1 = 0.0042
y w i
-------
TABLE 4-8
Results from VADOFT Execution
Antrim, New Hampshire
Parameter
Value
(Units)
Results from Flow Simulation
Elevation of water table
0.3
(m)
Results from Simulation of Benzene Transport
Net dispersive flux -2.1x10
Net advective flux 1.2xlO"6
Cumulative mass decay l.OxlO"6
Cumulative mass inflow 1.3x10"
Cumulative mass outflow -2.7x10"
Mass flux 1.2xlO"6
Waste release rate 5.9x10
-7
-3
(kg/m2-hr)
(kg/m2-hr)
(kg/m2-hr)
(kg/m2-hr)
(kg/m2-hr)
(kg/m2-hr)
(kg/hr)
Results from Simulation of Lead Transport
Net dispersive flux
Net advective flux
Cumulative mass decay
Cumulative mass inflow
Cumulative mass outflow
Mass flux
Waste release rate
1.3xlO"6
1.3xlO"6
0
l.SxlO"6
1.3xlO"6
1.3xlO"6
6.4xlO"3
(kg/m2-hr)
(kg/m2-hr)
(kg/m2-hr)
(kg/m2-hr)
(kg/m2-hr)
(kg/m2-hr)
(kg/hr)
-------
TABLE 4-9
Input Parameters for AT123D Simulation of Contaminant Transport
Through the Saturated Zone
Antrim, New Hampshire
Parameter
Distance to receptor well
Aquifer width
Aquifer depth
Begin point of x-source location
End point of x-source location
Begin point of y-source location
End point of y-source location
Porosity
Hydraulic conductivity
Hydraulic gradient
Longitudinal dispersivity
Lateral dispersivity
Vertical dispersivity
Distribution coefficient (benzene)
Distribution coefficient (lead)
Decay constant (benzene)
Decay constant (lead)
Density of soil
Density of water
Waste release rate (benzene)*
Waste release rate (lead)
Superimposed velocity term
Value
221
infinite
15
0
71
-35.5
35.5
0.43
7.13
0.01
15.3
5.1
1.0
0.08
234
2.6xlO'4
0
1.51
1.0
5.9xlO'3
6.4xlO"3
1.7xlO"3
(Units)
(m)
(m)
(m)
(m)
(m)
(m)
(m)
(unitless)
(m/hr)
(unitless)
(m)
(m)
(m)
(I/kg)
(I/kg)
(hr"1)
(hr'1)
(kg/1)
(kg/1)
(kg/hr)
(kg/hr)
(m/hr)
Generated by VADOFT component of model code.
-------
TABLE 4-10
Results from AT123D Simulation of Contaminant Transport
Through the Saturated Zone
Antrim, New Hampshire
Parameter
Results from Simulation of Benzene Transport
Retardation factor
Retarded Darcy velocity
Retarded longitudinal dispersion coefficient
Retarded lateral dispersion coefficient
Retarded vertical dispersion coefficient
Steady state concentration at receptor well
Results from Simulation of Lead Transport
Retardation factor
Retarded Darcy velocity
Retarded longitudinal dispersion coefficient
Retarded lateral dispersion coefficient
Retarded vertical dispersion coefficient
Steady state concentration at receptor well
Value
1.28
8.5xlO"3
0.13
0.04
8.5xlO'3
4.2xlO'3
823
1.3xlO*5
2.0xlO"4
6.7xlO'5
1.3xlO'5
0.38
(Units)
(unitless)
(m/hr)
(m2/hr)
(m2/hr)
(m2/hr)
(mg//)
(unitless)
(m/hr)
(m2/hr)
(m2/hr)
(m2/hr)
(mg//)
-------
In other words, for each 1 mg// of contaminant in seepage beneath the impoundment, 0.0042 mg//
of benzene are expected in the aquifer 500 feet (152 m) downgradient from the site. From Equation
4-1, it follows for benzene in a sludge layer with 10% solids:
Ct/N = l/(kd+rsl-1) = l/(37+10) = 0.021 kg//
Then from Equation 4-3:
t/N) = (0.0042)(0.021 kg//) = 8.9 x 10'5 kg//
and from Equation 4-4:
N«« - RWC/SRR-.. « (1.2xlO*3 mg//)/(8.9xlO"5 kg//) = 13 mg/kg
Similar calculations can be performed for lead:
Cai/Cl - °-38 / l = °-38
Ct/N = l/(kd+rsl'1) = l/(234+10) = 4.1xlO'3 kg//
(0.38X4.1xlO'3) = 1.6xlO"3 kg//
Nmax = RWC / SRRGU " <°-005 mg//)/(1.6xlO'3 kg//) = 3.1 mg/kg
If actual concentrations in the sludge exceed these criteria, then long-term storage of sludge in the
lagoons would not be allowed.
-------
An additional pathway of potential human and nonhuman exposure to contaminants from
surface disposal facilities involves the discharge of contaminants from groundwater to surface water,
with subsequent exposure to sludge contaminants for aquatic life, or to humans, through ingestion
of the water or of fish caught in the contaminated area. As explained in Section 4.3.3, the VADOFT
and ATI23D computer codes can be used to predict estimated loadings of contaminant to surface
water as a function of the dry weight concentration of contaminants in sludge. For Antrim, the
distance from near edge of the (idealized square) lagoon to the Contoocook river is approximately
100 meters, or 170 meters, from the far edge of the lagoon. To estimate loadings of contaminant to
the river, the linked VADOFT-ATI23D model is executed with input parameters listed in Table 4-
7, combined with slightly modified input parameters for AT123D, as listed in Table 4-11. Table 4-
11 differs from Table 4-9 in that the distance to the receptor well in Table 4-9 has been replaced by
a lesser distance to surface water in Table 4-11, and that the x-, y- and z-source locations have been
changed to reflect an even loading of contaminant throughout the entire section of aquifer beneath
the impoundment. In addition, the width of the section of aquifer to be modeled has artificially
constrained to that of the site. Results from execution of the linked model are presented in Table
4-12. The estimated concentration of benzene is 0.018 ing/kg (or g/m3) throughout a cross-section
of the aquifer as it intersects the Contoocook. As shown in the table, this value is multiplied by the
volume of water expected to discharge to the stream from this section of aquifer (3.2 m3/hr) to yield
an estimated 5.8xlO~5 kg/hr of benzene loading to the river. Since the linked model adjusts
calculations within AT123D by Equation 4-2 to account for extra dilution from seepage, and since
this dilution is irrelevant for the total loading of contaminant expected to reach the stream, the
calculated loading is divided by the dilution factor to yield (5.8x10~5)/(0.33) = 1.8x10** kg/hr of
loading, or about 0.003% of the expected loading to the aquifer beneath the site. For convenience,
this value is converted to 4.9xlO"2 mg/sec, and divided by the unit concentration of benzene used
as input to VADOFT to derive a ratio of loading to liquid concentration:
W0/Cl = (4.9xlO'2 mg/sec)/(1.0 mg//) - 4.9xlO'2 //sec.
-------
TABLE 4-11
Input Parameters for AT123D Simulation of Contaminant Transport
Through the Saturated Zone to Surface Water
Antrim, New Hampshire
Parameter
Distance to surface water
Aquifer width
Aquifer depth
Begin point of x-source location
End point of x-source location
Begin point of y-source location
End point of y-source location
Begin point of z-source location
End point of z-source location
Porosity
Hydraulic conductivity
Hydraulic gradient
Longitudinal dispersivity
Lateral dispersivity
Vertical dispersivity
Distribution coefficient (benzene)
Distribution coefficient (lead)
Decay constant (benzene)
Decay constant (lead)
Density of soil
Density of water
Waste release rate (benzene)*
Waste release rate (lead)*
Superimposed velocity term
Value
170
71
15
0
71
0
71
0
15
0.43
7.13
0.01
15.3
5.1
1.0
0.08
234
2.64xlO'A
0
1.51
1.0
5.9xlO'3
6.4xlO'3
1.7xlO"3
(Units)
(m)
(m)
(m)
(m)
(m)
(m)
(m)
(m)
(m)
(unitless)
(m/hr)
(unitless)
(m)
(m)
(m)
d/kg)
(I/kg)
(hr'1)
(hr'1)
(kg/1)
(kg/1)
(kg/hr)
(kg/hr)
(m/hr)
Generated by VADOFT component of model code.
4-50
-------
TABLE 4-12
Results from AT123D Simulation of Contaminant Transport
Through the Saturated Zone to Surface Water
Antrim, New Hampshire
Parameter
Darcy velocity in aquifer
Depth of cross section
Width of cross section
Flow through cross section
Results from Simulation of Benzene Transport
Steady state concentration at discharge
Time required to reach steady state
Unadjusted loading to surface water
Dilution adjustment
Adjusted loading to surface water
w0/cl
CI/N
W /N
Results from Simulation of Lead Transport
Steady state concentration at discharge
Time required to reach steady state
Unadjusted loading to surface water
Dilution adjustment
Adjusted loading to surface water
W0/Cl
Ct/N
WO/N
Value
3.0xlO'3
15
71
3.2
0.018
IxlO5
5.8xlO"5
0.33
l.SxlO'4
4.9xlO"2
2.1x10'*
l.OxlO"3
0.67
9xl07
2.1xlO'3
0.33
6.4xlO"3
1.8
4.1xlO"3
7.3xlO'3
(Units)
(m/hr)
(m)
(m)
(m3/hr)
(8/m3)
(hr)
(kg/hr)
(unitless)
(kg/hr)
(//sec)
(kg//)
(kg/sec)
(g/m3)
(hr)
(kg/hr)
(unitless)
(kg/hr)
(//sec)
(kg//)
(kg/sec)
4-51
-------
This result is combined with the value estimated above for Ct/N:
wo/N = (Wo/qXq/N) = (4.9xlO-2 //sec)(0.021 kg//) = l.OxlQ-3 kg/sec
The calculations are similar for lead. Not surprisingly, the estimated loading of lead to the
Contoocook River is the same as the expected loading to the aquifer, since lead is assumed to be
conserved during transport. The expected concentration of lead within the artificially confined
aquifer is 0.67 mg//, about 2/3 the unit concentration assumed within the impoundment. This
concentration can be multiplied by the expected volume of water passing through the cross-section
of aquifer beneath the impoundment to derive an estimated loading of 2.1xlO'3 kg/hr of lead to the
river, which can be adjusted for the dilution factor to derive an expected loading of 6.4xlO'3 kg/hr
or 1.8 mg/sec (the same loading of contaminant calculated as the waste release rate to the aquifer).
This estimated loading must be converted into a ratio of loading to dry-weight sludge concentration:
W0/Ct = (1.8 mg/sec)/(1.0 mg//) = 1.8 //sec
and:
W0/N = (1.8 //sec)(4.1x!0-3 kg//) = 7.3xlO'3 kg/sec
These results show that for every mg/kg of benzene or lead in sludge accumulating in the lagoons in
Antrim, an expected l.OxlO"3 and 7.3xlO'3 kg/sec of benzene or lead, respectively, are expected to
be loaded into the Contoocook River. As will be discussed in Chapter 5, these ratios can be used to
derive criteria for exposure pathways related to contamination of surface water.
4.7.1.2. TULSA, OKLAHOMA -- The Northside municipal wastewater treatment facility
is designed to process 36 million gallons of wastewater per day. Sludge generated by the facility is
continuously piped to three clay-lined lagoons. When a lagoon fills with sludge solids, its contents
are removed for land application. Oklahoma law requires that such cleaning take place every 6
4-52
-------
months or less; this requirement is intended to stop plants from stockpiling sludge in their lagoons.
A two foot layer of sludge is usually left in the bottom of a lagoon when it is emptied, however, to
prevent damage to the floor of the facility. Upon removal from the impoundment, sludge typically
contains about 7-9 percent solids. Solids content in a filled lagoon range from about 6% at one foot
depth to about 8% percent at the bottom of the impoundment. Sludge in the lagoons is routinely
sampled for metals content, and recently was tested for the first time for organic contaminants.
Groundwater monitoring wells surround the site.
The surface area of each of the three sludge lagoons at the Tulsa facility is about 2 hectares;
depths of the lagoons range from about 9.5-11 feet (2.9-3.3m). For the purposes of these sample
calculations, the three facilities are modeled as a single impoundment with an area of about 6
hectares. Edges of the impoundments slope inward at about a three to one slope, but for simplicity,
the sample calculations described below model the impoundments as having identical bottom and
surface dimensions.
Tier 2 calculations for this facility are quite similar to those for Antrim, but several of the
parameter values used for the groundwater simulations are changed. Test borings beneath the site
have revealed mostly silty clay to a depth of 30-40 feet (9-12 meters), but measurements of the depth
to groundwater were not available. Based on data from the DRASTIC data base U.S. EPA (1985d)
in GEMS, however, average depths to groundwater in the Tulsa area range from 9-15 meters. A
distance of 12 meters was assumed for these calculations. Estimated aquifer thickness is based on
data from the GRNDWAT data base in GEMS that report aquifer thickness ranging from 1-61
meters in the region of the Tulsa facility. A midpoint value of 30 meters is used for these sample
calculations. Depths of these two layers total 42 meters. This value is used as input for the VADOFT
model, as shown by Table 4-13. Characteristics of both the unsaturated and saturated zones are
selected to be typical of silt. Seepage from the facility is assumed to be 0.12 //m2-hour, a value from
the top of the range in Table 4-2 for an impoundment sealed with bentonite.
Results of the VADOFT calculations are summarized in Table 4-14. As shown in the table,
the water table beneath the site is predicted to be about 1 meter higher than it would have been
4-53
-------
TABLE 4-13
Input Parameters for VADOFT Simulation of Flow and Contaminant Transport
Through the Unsaturated Zone
Tulsa, Oklahoma
Parameter
Source area
Distance to bottom of saturated zone
Thickness of Layer I
Thickness of Layer II
Input Parameters for Flow Calculations
Flux at Top Node
Head at bottom node
Hydraulic conductivity of Layer I
Hydraulic conductivity of Layer II
Effective porosity of Layer I
Effective porosity of Layer II
Specific Storage
Residual water saturation
Power index (N)
Leading coefficient
Power index (j9)
Power index (7)
Input Parameters for Transport Calculations
Contaminant concentration
Flux at top node
Head at bottom node
Longitudinal dispersivity
Effective porosity of Layer I
Effective porosity of Layer II
Retardation coefficient of Layer I
Retardation coefficient of Layer II
Molecular diffusion coefficient
Default Darcy velocity
Default water saturation
Solute decay constant (benzene)
Solute decay constant (lead)
Value
6.1xl05
42
12
30
0.12
30
2.0xlO'4
3.6xlO"3
0.36
0.10
0
0.19
-1.0
0.5
1.09
0.083
1
1.2xlO'7
0
1.0
0.36
0.10
1.38
1.91
0
0
1.0
2.6xlO"4
0
(Units)
(m2)
(m)
(m)
(m)
(//m2-hr)
(m)
(m/hr)
(m/hr)
(unitless)
(unitless)
(unitless)
(unitless)
(unitless)
(unitless)
(unitless)
(unitless)
(mg//)
(kg/m2-hr)
(m)
(unitless)
(unitless)
(unitless)
(unitless)
(unitless)
(unitless)
(m/hr)
(unitless)
(hr'1)
(hr'1)
-------
TABLE 4-14
Results from VADOFT Execution
Tulsa, Oklahoma
Parameter Value (Units)
Results from Flow Simulation
Water table elevation 2.6 (m)
Results from Simulation of Benzene Transport
Net dispersive flux 4.8xlO"8 (kg/m2-hr)
Net advective flux 7.3xlO"8 (kg/m2-hr)
Cumulative mass decay 1.2xlO~7 (kg/m2-hr)
Cumulative mass inflow 1.2xlO"7 (kg/m2-hr)
Cumulative mass outflow -4.0xlO~15 (kg/m2-hr)
Mass flux 2.7xlO'10 (kg/m2-hr)
Waste release rate 1.6xlO"5 (kg/hr)
Results from Simulation of Lead Transport
Net dispersive flux -1.2xlO"7 (kg/m2-hr)
Net advective flux 1.2xlO"7 (kg/m2-hr)
Cumulative mass decay 0 (kg/m2-hr)
Cumulative mass inflow 1.2xlO"7 (kg/m2-hr)
Cumulative mass outflow -1.2xlO"7 (kg/m2-hr)
Mass flux 1.2xlO~7 (kg/m2-hr)
Waste release rate 7.4x10"3 (kg/hr)
4-55
-------
without the impoundment. Almost all of the benzene is removed by decay processes during its
transport through the unsaturated zone, but lead concentrations are unchanged.
Table 4-15 shows inputs for the AT123D calculations. Parameter values have been selected
as appropriate for an aquifer that is 30 meters thick and consists primarily of sand. The
impoundment site is represented as a square of width 247 meters aligned with the direction of
groundwater flow. A superimposed velocity term of 2.8 x 10"4 m/hour is estimated from the rate of
seepage beneath the facility. Mass release rates of 1.6 x 10'5 kg/hour and 7.4 x 10~3 kg/hour, for
benzene and lead, respectively, are taken from calculations by the VADOFT module.
Table 4-16 lists results from the ATI23D model run. Because the Darcy velocity at the Tulsa
site is much lower than for Antrim, a longer time interval is required to reach steady state, and there
is more opportunity for contaminant decay between the site and the receptor well. As a result,
expected benzene concentrations are insignificant. For every milligram per liter of concentration
at the bottom of the impoundment, 2.0 x 10~9 nig// are expected at the receptor well. Predicted lead
concentrations, on the other hand, actually exceed the unit concentration used as input for the
simulation. This physical impossibility reflects the fact that lead is not expected to decay in transport
through the unsaturated zones, that the impact of seepage from the impoundment on the flow of
groundwater is assumed to be significant near the site, and that the model here described relies on
several conservative assumptions made necessary by reliance on a one-dimensional flow model for
the unsaturated zone, and an analytical code for transport in the saturated zone.
The results described above can be used to estimate maximum dry-weight concentrations of
benzene and lead to be allowed in sludge stored in the impoundments. For benzene:
/ C, = 2.0xlO'11 (unitless)
I
From Equation 4-1 it follows for benzene in a sludge layer with 10% solids:
= l/(37+10) = 0.021 (kg//)
4-56
-------
TABLE 4-15
Input Parameters for AT123D Simulation of Contaminant Transport
Through the Saturated Zone
Tulsa, Oklahoma
Parameter
Distance to receptor well
Aquifer width
Aquifer depth
Begin point of x-source location
End point of x-source location
Begin point of y-source location
End point of y-source location
Porosity
Hydraulic conductivity
Hydraulic gradient
Longitudinal dispersivity
Lateral dispersivity
Vertical dispersivity
Distribution coefficient (benzene)
Distribution coefficient (lead)
Decay constant (benzene)
Decay constant (lead)
Density of soil
Density of water
Waste release rate (benzene)
Waste release rate (lead)*
Superimposed velocity term
Value
400
infinite
30
0
247
-123.5
123.5
0.10
0.86
0.01
15.3
5.1
1.0
0.08
234
2.64xlO'4
0
2.38
1.0
1.9X10"4
9.2xlO'3
2.8xlO"4
(Units)
(m)
(m)
(m)
(m)
(m)
(m)
(m)
(unitless)
(m/hr)
(unitless)
(m)
(m)
(m)
(I/kg)
(I/kg)
(hr'1)
(hr'1)
(kg/1)
(kg/1)
(kg/hr)
(kg/hr)
(m/hr)
Generated by VADOFT component of model code.
4-57
-------
TABLE 4-16
Results from AT123D Simulation of Contaminant Transport
Through the Saturated Zone
Tulsa, Oklahoma
Parameter
Value
(Units)
Results from Simulation of Benzene Transport
Retardation factor 2.90
Retarded Darcy velocity 1.3xlO"3
Retarded longitudinal dispersion coefficient 0.02
Retarded lateral dispersion coefficient 6.75xlO"3
Retarded vertical dispersion coefficient 1.3xlO"3
Steady state concentration at receptor well 2.0x10
-9
(unitless)
(m/hr)
(m2/hr)
(m2/hr)
(m2/hr)
(mg//)
Results from Simulation of Lead Transport
Retardation factor
Retarded Darcy velocity
Retarded longitudinal dispersion coefficient
Retarded lateral dispersion coefficient
Retarded vertical dispersion coefficient
Steady state concentration at receptor well
5570
6.9xlO'7
l.OSxlO"5
3.5xlO'6
6.9xlO'7
1.1
(unitless)
(m/hr)
(m2/hr)
(m2/hr)
(m2/hr)
(mg//)
If-58
-------
Then from Equation 4-3:
SRRGW = (C^qXCyN) = (2.0xl(r11)(0.02l) = 4.3X10'13 (kg//)
and from Equation 4-4:
Nmax " RWC / SRRGW = (l-2xlO'3)/(4.3xlO'13) = 2.8xl09 (mg/kg)
From these calculations it could be concluded that no possible concentration of benzene in the
impoundment would be expected to result in groundwater concentrations that exceed the RWC.
Similar calculations can be performed for lead, for which the concentration in well water is predicted
to be the same as the dissolved concentration at the floor of the lagoon:
C w/Cl = 1.0 (unitless)
4.1xlO~3kg//
SRRGU = (Cqxq/N) = (l)(4.1xl
-------
the aquifer to the creek will be the same as the estimated loading of lead to the aquifer (7.4 kg/hr
or 2.0 mg/sec) so that W0/Ct = (2.0 mg/sec)/(l mg//) = 2.0 //sec. From this result it follows that:
W0/N = (Wo/qXq/N) = (2.0 //sec)(4.1xlO"3 kg//) = 8.3xlO'3 kg/sec.
For benzene, the linked V ADOPT- AT 123D model is used to predict the concentration of
contaminant within an artificially confined aquifer with width corresponding to that of the site.
Inputs for the VADOFT model as listed in Table 4-16, combined with inputs for the AT123D model,
as listed in Table 4-17, are supplied to the model to yield results listed in Table 4-18. As before,
AT123D is used to estimate expected loadings to the creek, based on a scenario in which the
contaminant loading is uniformly distributed throughout the aquifer beneath the site, and in which
transverse dispersion is constrained by confining the aquifer to a width corresponding to that of the
site. Results from AT123D indicate an expected steady-state concentration of l.SxlO"7 g/m3 (or
mg//) at a distance of 50 meters from the edge of the site. Based on a hydraulic conductivity of 0.86
m/hr and a hydraulic gradient of 0.1, the Darcy velocity of water within the aquifer is 3.6xlO"5
m/hr, suggesting 0.27 m3 of flow per hour through a cross-section of 30 x 247 meters. The estimated
concentration is multiplied by the expected volume of flow to yield 3.9xlO"11 kg/hour of loading to
the creek. This estimate is adjusted for a dilution factor of 0.035, converted to convenient units and
divided by the unit concentration used for the calculations to yield
W0/Ct = (3.8xlO"7 mg/sec)/(l mg//) = 3.1xlO'7 //sec
and
WQ/N = (WO/C^CL/N) = (3.1xlO'7 //sec)(2.1x!0'2 kg/1) = 6.6xlO'9 kg/sec
This ratio will be used in Chapter 5 to derive criteria for surface water pathways.
i»-60
-------
TABLE 4-17
Input Parameters for AT123D Simulation of Contaminant Transport
Through the Saturated Zone to Surface Water
Tulsa, Oklahoma
Parameter
Distance to surface water
Aquifer width
Aquifer depth
Begin point of x-source location
End point of x-source location
Begin point of y-source location
End point of y-source location
Begin point of z-source location
End point of z-source location
Porosity
Hydraulic conductivity
Hydraulic gradient
Longitudinal dispersivity
Lateral dispersivity
Vertical dispersivity
Distribution coefficient (benzene)
Distribution coefficient (lead)
Decay constant (benzene)
Decay constant (lead)
Density of soil
Density of water
Waste release rate (benzene)*
Waste release rate (lead)
Superimposed velocity term
Value
300
247
30
0
247
0
247
0
30
0.10
0.86
0.01
15.3
5.1
1.0
0.08
234
2.64xlO'4
0
2.38
1.0
1.9xlO"4
9.2xlO"3
3.5xlO'4
(Units)
(m)
(m)
(m)
(m)
(m)
(m)
(m)
(m)
(m)
(unitless)
(m/hr)
(unitless)
(m)
(m)
(m)
0/kg)
d/kg)
(hr'1)
(hr'1)
(kg/1)
(kg/1)
(kg/hr)
(kg/hr)
(m/hr)
Generated by VADOFT component of model code.
4-61
-------
TABLE 4-18
Results from AT123D Simulation of Contaminant Transport
Through the Saturated Zone to Surface Water
Tulsa, Oklahoma
Parameter
Darcy velocity in aquifer
Depth of cross section
Width of cross section
Flow through cross section
Results from Simulation of Benzene Transport
Steady state concentration at discharge
Time required to reach steady state
Unadjusted loading to surface water
Dilution adjustment
Adjusted loading to surface water
WO/CL
q/N
W0/N
Results from Simulation of Lead Transport
Steady state concentration at discharge
Time required to reach steady state
Unadjusted loading to surface water
Dilution adjustment
Adjusted loading to surface water
WO/CL
Ct/N
W0/N
Value
3.6xlO'5
30
247
0.27
1.5xlO"7
4xl06
3.9xlO'11
0.035
l.lxlO'9
3.1xlO"7
0.021
6.6xlO'9
0.97
8xl09
2.6xlO'4
0.035
7.3xlO'3
2.0
4.1xlO'3
8.3xlO'3
(Units)
(m/hr)
(m)
(m)
(m3/hr)
(g/m3)
(hr)
(kg/hr)
(unitless)
(kg/hr)
(//sec)
(kg//)
(kg/sec)
(g/m3)
(hr)
(kg/hr)
(unitless)
(kg/hr)
(//sec)
(kg//)
(kg/sec)
4-62
-------
4.7.1.3. PORTLAND, OREGON -- The Columbia Boulevard waste water treatment plant
in Portland has been in operation since the 1940's and was updated in the 1970's to include secondary
treatment. In the past, sludge was discharged directly to an on-site sludge lagoon; some of the sludge
from the lagoon was then pumped out and processed in a composting facility. Now, almost all of the
plant's sludge goes directly to the in-vessel composting facility. The impoundment is used only in
emergency situations, upon failure of the composting process. Recently, for example, problems with
the composting process resulted in the pumping of 260,000 gallons of sludge to the lagoon.
The lagoon in Portland is estimated to cover about 32 acres (13 hectares) to a depth of about
16 feet (5 meters), and is currently estimated to contain about 80,000 dry tons of sludge. Plans for
future operation of the facility are not certain; the city hopes to exhume all of the sludge eventually
for composting, but the in-vessel composting system can handle only the current volume of sludge
generated. Sludge quality in the impoundment varies by location, with older deposits typically
containing higher concentrations of metals and PCBs. The impoundment is located about 50 meters
from the Columbia Slough, which empties into the Columbia River.
Soil borings for the Portland site were not available. These sample calculations therefore rely
on data from the DRASTIC data base accessed through GEMS (U.S. EPA, 1988c). Based on
hydrogeological data for the county containing the lagoon, these sample calculations assume that the
vadose zone is composed of sand to a depth of 15 meters, followed by a saturated zone of sand with
a thickness of 15 meters. These two layers sum to a total distance of 30 meters from the floor of the
lagoon to the bottom of the saturated zone.
Table 4-19 contains values for input parameters used to simulate flow and contaminant
transport beneath the Portland facility. Table 4-20 lists results from the VADOFT simulation, and
shows that the water table beneath the site is expected to show an increase of 0.6 meters in elevation
as a result of seepage from the impoundment. Benzene concentrations decrease by about 13% as a
result of travel through the unsaturated zone.
Table 4-21 lists inputs for estimating contaminant transport through the saturated zone. As
for Antrim, the saturated zone is assumed to consist of a layer of sand 15 meters thick and infinitely
wide. Waste release rates are taken from VADOFT, and the distance to the well is set to equal the
-------
TABLE 4-19
Input Parameters for VADOFT Simulation of Flow and Contaminant Transport
Through the Unsaturated Zone
Portland, Oregon
Parameter
Source area
Distance to bottom of saturated zone
Input Parameters for Flow Calculations
Flux at Top Node
Head at bottom node
Hydraulic conductivity
Effective porosity
Specific Storage
Residual water saturation
Power index (N)
Leading coefficient
Power index (ft)
Power index (7)
Input Parameters for Transport Calculations
Contaminant concentration
Flux at top node
Head at bottom node
Longitudinal dispersivity
Effective porosity
Retardation coefficient
Molecular diffusion coefficient
Default Darcy velocity
Default water saturation
Solute decay constant (benzene)
Solute decay constant (lead)
Value
l.SOxlO5
30
1.3
15
0.30
0.43
0
0.105
-1.0
14.5
2.68
0.62
1
1.3xlO"6
0
1.0
0.43
1.28
0
0
1.0
2.6xlO'4
0
(Units)
(m2)
(m)
(//m2-hr)
(m)
(m/hr)
(unitless)
(unitless)
(unitless)
(unitless)
(unitless)
(unitless)
(unitless)
(mg//)
(kg/m2-hr)
(m)
(unitless)
(uni&ess)
(unitless)
(unitless)
(m/hr)
(unitless)
(hr'1)
(hr'1)
-------
TABLE 4-20
Results from VADOFT Execution
Portland, Oregon
Parameter
Results from Flow Simulation
Water table elevation
Results from Simulation of Benzene Transport
Net dispersive flux
Net advective flux
Cumulative mass decay
Cumulative mass inflow
Cumulative mass outflow
Mass flux
Waste release rate
Results from Simulation of Lead Transport
Net dispersive flux
Net advective flux
Cumulative mass decay
Cumulative mass inflow
Cumulative mass outflow
Mass flux
Waste release rate
Value
0.6
-l.lxlO'7
1.2xlO"6
l.lxlO'6
l.SxlO'6
-1.6xlO'7
6.9xlO'7
9.0xlO'2
-l.SxlO'6
1.3xlO"6
0
1.3xlO'6
-1.3xlO'6
l.SxlO'6
l.7xlO"1
(Units)
(m)
(kg/m2-hr)
(kg/m2-hr)
(kg/m2-hr)
(kg/m2-hr)
(kg/m2-hr)
(kg/m2-hr)
(kg/hr)
(kg/m2-hr)
(kg/m2-hr)
(kg/m2-hr)
(kg/m2-hr)
(kg/m2-hr)
(kg/m2-hr)
(kg/hr)
-------
TABLE 4-21
Input Parameters for AT123D Simulation of Contaminant Transport
Through the Saturated Zone
Portland, Oregon
Parameter
Distance to receptor well
Aquifer width
Aquifer depth
Begin point of x-source location
End point of x-source location
Begin point of y-source location
End point of y-source location
Porosity
Hydraulic conductivity
Hydraulic gradient
Longitudinal dispersivity
Lateral dispersivity
Vertical dispersivity
Distribution coefficient (benzene)
Distribution coefficient (lead)
Decay constant (benzene)
Decay constant (lead)
Density of soil
Density of water
Waste release rate (benzene)*
Waste release rate (lead)
Superimposed velocity term
Value
510
infinite
15
0
361
-180
180
0.43
0.30
0.01
15.3
5.1
1.0
0.08
234
2.64xlO~4
0
1.5.1
1.0
8.9xlO'5
1.65xlO"4
8.6xlO"3
(Units)
(m)
(m)
(m)
(m)
(m)
(m)
(m)
(unitless)
(m/hr)
(unitless)
(m)
(m)
(m)
(I/kg)
(I/kg)
(hr'1)
(hr'1)
(kg/1)
(kg/1)
(kg/hr)
(kg/hr)
(m/hr)
*Generated by the VADOFT component of the model code.
4-66
-------
distance to the property boundary. With these input values, AT123D produces the results listed in
Table 4-22. Concentrations of benzene are expected to decrease by about 98% between the
impoundment and the receptor well, but concentrations of lead decrease only 9%. These results can
be used to derive maximum allowable concentrations of lead and benzene in sludge deposited in the
Portland facility. For benzene,
Cgu/Cl = 0.019 (unitless)
CI/N = 0.021 (kg//)
Then from Equation 4-3:
SRR
GU
(0.019)(0.021) = 4.0 x 10
'4
and from Equation 4-4:
Nmax = RWC / SRRGU = (1.2xlO'3)/(4.0xlO-4) = 3.0 mg/kg
Similar calculations can be performed for lead:
SRRGW = (Cg,
C/Cl = 0.91
Cl/N = 4.1xlO
"3
(0.91)(4.1xlO'3) = 3.7xlO'3
N = RWC / SRRGU = (0.005)/(3.7xlO'3) = 1.3 mg/kg
4-67
-------
TABLE 4-22
Results from AT123D Simulation of Contaminant Transport
Through the Saturated Zone
Portland, Oregon
Parameter
Value
(Units)
Results from Simulation of Benzene Transport
Retardation factor
Retarded Darcy velocity
Retarded longitudinal dispersion coefficient
Retarded lateral dispersion coefficient
Retarded vertical dispersion coefficient
Steady state concentration at receptor well
1.28
0.02
0.32
0.11
0.02
19
(unitless)
(m/hr)
(m2/nr)
(m2/hr)
(m2/hr)
Results from Simulation of Lead Transport
Retardation factor
Retarded Darcy velocity
Retarded longitudinal dispersion coefficient
Retarded lateral dispersion coefficient
Retarded vertical dispersion coefficient
Steady state concentration at receptor well
823
3.3xlO'5
5.0xlO'4
1.7xlO'4
3.3xlO'5
907
(unitless)
(m/hr)
(m2/hr)
(m2/hr)
(m2/hr)
(fog/0
4-68
-------
If actual concentrations in the sludge exceed these criteria, then long-term (or permanent) storage
of sludge in the lagoons would not be allowed.
Results from VADOFT calculations are next used to estimate expected loadings of benzene
and lead to the Columbia Slough, located about 50 meters from the lagoon. It is assumed that the
loading of lead to surface water is the same as the estimated loading to groundwater, or 0.16 kg/hr.
This value is converted to mg/sec and divided by the unit concentration in the impoundment (1 mg//)
to derive W0/CL = 46 //sec, which is then combined with Ct/N to yield:
W0/N = (Wo/qxq/N) = (46)(4.IxlO'3) = 0.19 kg/sec
For benzene, methods identical to those described in sections 4.6.1 and 4.6.2 are used to predict
expected loadings to the Columbia Slough, based on input parameters listed in Table 4-23. As shown
in Table 4-24, the model predicts a steady-state concentration of 0.055 mg// (or g/m3). This result
is multiplied by the estimated flow of water through a cross section of the aquifer, and adjusted to
compensate for dilution calculations in the linked model:
W0 = (0.055 mg//)(3.0xlO'3)(15)(361)/(0.089) = 2.8 mg/sec
so:
WQ/Ct = (2.8 mg/sec)/(l mg//) = 2.8 //sec
This ratio is combined with an estimate of Ct/N to compare loading to the dry- weight concentration
of benzene in sludge received by the lagoon:
= (2.8 //sec)(0.021 kg//) = 0.059 kg/sec
These values will be used to derive criteria in Chapter 5.
-------
TABLE 4-23
Input Parameters for AT123D Simulation of Contaminant Transport
Through the Saturated Zone to Surface Water
Portland, Oregon
Parameter
Distance to surface water
Aquifer width
Aquifer depth
Begin point of x-source location
End point of x-source location
Begin point of y-source location
End point of y-source location
Begin point of z-source location
End point of z-source location
Porosity
Hydraulic conductivity
Hydraulic gradient
Longitudinal dispersivity
Lateral dispersivity
Vertical dispersivity
Distribution coefficient (benzene)
Distribution coefficient (lead)
Decay constant (benzene)
Decay constant (lead)
Density of soil
Density of water
Waste release rate (benzene)*
Waste release rate (lead)*
Superimposed velocity term
Value
410
361
15
0
361
0
361
0
15
0.43
0.30
0.01
15.3
5.1
1.0
0.08
234
2.64xlO"4
0
1.51
1.0
8.9xlO'5
1.6xlO"4
8.6xlO'3
(Units)
(m)
(m)
(m)
(m)
(m)
(m)
(m)
(m)
(m)
(unitless)
(m/hr)
(unitless)
(m)
(m)
(m)
(I/kg)
(I/kg)
(hr'1)
(hr'1)
(kg/1)
(kg/1)
(kg/hr)
(kg/hr)
(m/hr)
"Generated by the VADOFT component of the model code.
4-70
-------
TABLE 4-24
Results from AT123D Simulation of Contaminant Transport
Thorough the Saturated Zone to Surface Water
Portland, Oregon
Parameter
Darcy velocity in aquifer
Depth of cross section
Width of cross section
Flow through cross section
Results from Simulation of Benzene Transport
Steady state concentration at discharge
Time required to reach steady state
Unadjusted loading to surface water
Dilution adjustment
Adjusted loading to surface water
W0/Cl
Ct/N
w0/N
Results from Simulation of Lead Transport
Steady state concentration at discharge
Time required to reach steady state
Unadjusted loading to surface water
Dilution adjustment
Adjusted loading to surface water
Wo/Cl
Ct/N
WO/N
Value
0.003
15
361
16
0.055
IxlO5
8.9xlO"4
0.089
IxlO'2
2.8
0.021
0.059
0.91
8xl07
0.015
0.089
0.16
46
4.1xlO'3
0.19
(Units)
(m/hr)
(m)
(m)
(m3/hr)
(g/m3)
(hr)
(kg/hr)
(unitless)
(kg/hr)
(//sec)
(kg//)
(kg/sec)
(g/m3)
(hr)
(kg/hr)
(unitless)
(kg/hr)
(//sec)
(kg//)
(kg/sec)
it-71
-------
4.7.2. Analysis of Exposure for the Most Exposed Populations. As discussed in Chapter 3, an
analysis of exposure to most exposed populations (MEP) can be useful for understanding the human
health risks associated with a particular sludge disposal option or exposure pathway. For the
groundwater pathway, potential exposure is unlikely to extend beyond a small geographical area
surrounding the facility, and is likely to be confined to a small population. In cases where public
water supply systems withdraw water from the contaminated aquifer, the size of the potentially
exposed population will be larger, but variation in individual exposure will be determined primarily
by variation in individual water consumption behavior. Methodological difficulties in quantifying
individual variation in water ingestion rates (Iu) complicate efforts to determine the distribution of
potential exposure among those ingesting water from a contaminated aquifer. As a result of these
constraints, no analysis of most exposed populations will be presented for the groundwater pathway.
4-72
-------
S. DERIVATION OF CRITERIA FOR THE SURFACE WATER PATHWAY
This chapter describes and demonstrates a method for modeling the long-term fate
and transport in surface water of selected contaminants from a surface disposal site for
wastewater sludge.
5.1 OVERVIEW OF THE METHOD
Municipal sewage treatment plants are usually found near natural surface water
courses like rivers, lakes or estuaries. For simplicity of presentation, it is assumed that the
site is located next to a river, although the methodology is equally applicable to other types
of water bodies. The sludge disposal facility is assumed to be an unlined impoundment
enclosed by an engineered embankment, probably constructed from on-site material. It is
assumed that the site is properly graded so that rainfall runoff from upstream areas is
diverted around and not into the sludge disposal impoundment. Therefore, since runoff
from the site is unlikely to carry contaminants to nearby surface water, the methodology
considers only one pathway of surface water contamination: contaminants from the
impoundment are transported by seepage into an aquifer beneath the site. They are then
transported within the aquifer and released to the nearby surface water body.
To set criteria for sludge based on protection of surface water, three pathways must
be considered: exposure of aquatic wildlife, human exposure through drinking water and
human exposure through fish consumption. At sites where consumption of groundwater is
also a concern, criteria are developed for the protection of drinking water uses of
groundwater, based on methods outlined in Chapter 4. If the point of compliance for the
groundwater pathway is of no greater distance than the nearest distance to surface water,
then sludge criteria derived to maintain groundwater at healthful levels will also serve as
criteria for protection of drinking water uses of surface water, since contaminant
concentrations can only decrease upon further dilution from mixing with the stream. At
sites where groundwater is not considered usable for current or future needs (i.e., at sites
where no criteria need be developed based on the groundwater pathway) and at sites where
surface water is closer than all possible well locations, surface water modeling can be used
to derive sludge criteria for the protection of drinking water. For some contaminants,
5-1
-------
potential risks to aquatic life or to humans through fish consumption may occur at lower
levels of surface water concentrations than those considered acceptable on the basis of
drinking water criteria alone. For these, estimated exposure through the surface water
pathway must be considered even if estimated concentrations in the groundwater are
considered acceptable for drinking by humans.
Since the river and the sludge impoundment are assumed to be in close proximity,
the river is taken as the primary surface water pathway for transporting sewage sludge
contaminants away from the disposal site. Sludge contaminant loading to the river can be
represented as either an instantaneous pulse or a continuous step input. For example, an
instantaneous pulse could include sludge released during an embankment failure,
impoundment overflow, pipeline rupture or other catastrophic accident at the disposal site.
Instantaneous pulses are useful when investigating acute impacts. In contrast, continuous
inputs describe loadings of extended duration, more likely to be appropriate for the loading
of contaminants to a stream from groundwater inflow. Continuous inputs are useful in
studies of chronic effects.
The surface water methodology assumes that contaminants migrate in seepage from
an unlined impoundment to the groundwater, which in turn discharges to the river as a
continuous distributed source. Consequently, the methodology considers continuous rather
than instantaneous sludge loadings and is concerned with long-term chronic, rather than
short-term acute effects. This mass loading of contaminant from groundwater is estimated
from the steady-state response of a groundwater contaminant transport model, as described
in Chapter 4.
If potential impacts from shorter-term exposure to higher, transient concentrations
of contaminant are also to be considered, the methodology described in this chapter can be
modified slightly to estimate potential acute exposure based on conditions of low stream
flow. For example, values of the lowest 7-day flow expected to occur every 10 years
(7Q10) are available for many streams from district offices of the U.S. Geological Survey.
It is assumed that short-term peak loadings of contaminant from the aquifer to the stream,
5-2
-------
if they occur, are unlikely to coincide with periods of low stream flow; estimation of
potential acute exposure is thus based on average, steady state loadings to the stream.
The proposed methodology for deriving criteria based on exposure through surface
water pathways relies on a two-tiered approach. As in Chapter 4, Tier 1 involves the
derivation of national, numerical criteria based on a "generic" scenario assumed to represent
a reasonable worst case. For example, such a scenario might involve the discharge of
groundwater into a lake or stream with relatively low flow; and the withdrawal of surface
water for human consumption, the consumption of freshwater fish, and the exposure of
aquatic wildlife in an area close to the zone of recharge. Tier 2 involves site-specific
modeling of potential exposure. Those facilities that fail to satisfy the Tier 1 criteria can
elect to perform a site-specific Tier 2 analysis.
Both tiers rely on a relatively simple yet flexible analytical approach in which the
river is modeled as a cascade of well-mixed cells (Stefan and Demetracopoulos, 1981).
Output from each upstream cell is used as input to the next downstream cell.
Computations proceed downstream along the cascading cells from the point of contaminant
entry to the location of the most exposed individual (MEI) or most exposed unit (MEU).
This "cascading cells" model accounts for degradation processes and tributary inflows. Input
parameters can be obtained from site-specific reports, maps and other existing sources, or
from the literature. In addition, the cascading cells model can easily accommodate "reverse
application": if the maximum allowable contaminant concentration at the MEI location is
specified, the cascading cells model can be used to determine the corresponding allowable
mass loading from the contaminated aquifer to the stream. This in turn can be used to
calculate the maximum dry weight concentration for the contaminant in the surface
impoundment.
Computer models are available for more sophisticated analyses of the fate and
transport of chemical contaminants in surface water bodies (U.S. EPA, 1988b). For
example, the Water Quality Analysis Simulation Program, or WASP4 model (U.S. EPA,
1988a), employs a network of segments connected by channel links to simulate the transport
of contaminants in dissolved and solid phases in rivers, ponds, lakes, reservoirs and
5-3
-------
estuaries. Model segments include up to four zones (epilimnion, hypolimnion, upper
benthic and lower benthic), which can be used to divide the water column and the sediment
bed laterally, vertically and longitudinally. WASP4 can be employed at various levels of
complexity involving increasing sophistication in solids behavior, equilibrium reactions and
kinetic reactions. Transfer processes include sorption, ionization and volatilization;
transformation processes include biodegradation, hydrolysis, photolysis and chemical
oxidation. Sorption and ionization are treated as equilibrium reactions described by a single
partition coefficient. All other processes are described by kinetic rate equations, each
characterized by constant half-life or rate constant.
A similar model, the Exposure Analysis Modeling System (EXAMSII), is available
within GEMS (U.S. EPA, 1988c). Chemical fate processes considered by the EXAMSII
model include volatilization, hydrolysis, oxidation, reduction, photo-oxygenation, biolysis
and photolysis. Like WASP4, this model simulates the transport of contaminants between
multiple compartments in the stream, and considers interactions between sediment and the
water column. For many locations, values for input parameters required by the model can
be drawn from data bases accessible within GEMS.
Effective use of the capabilities of WASP4, EXAMSII, or similar models requires
numerous site-specific input parameter values that may be difficult or expensive to obtain
for certain sites, however. In addition, discharge of some species of contaminant from the
aquifer to the stream may occur years or even decades after the contaminants first migrate
into the aquifer. Since exact locations of future withdrawal of drinking water, or of
contact between surface water and aquatic species cannot be known with certainty, the
extra precision gained by using WASP4, EXAMSII or equivalent models is not necessarily
advantageous for deriving criteria for sludge disposal facilities. Without consideration of
additional contaminant loss processes associated with the stream floor, adsorption and
accumulation of metals and high molecular weight materials at the bottom of the stream,
WASP4 reduces to the equations described below. Those equations provide simpler, but
more conservative, estimates of contaminant concentrations in surface water. If sufficient
-------
data are available, and more accurate (less conservative) calculations are desired, the
additional capabilities of a more complete mathematical model should be utilized.
5.2. ASSUMPTIONS
A number of hydraulic and water quality assumptions are required for the cascading
cells model; these are summarized in Table 5-1. The primary motivation for these
assumptions is to formulate a surface water transport model that provides a reasonable
representation of the receiving water body and yet can be solved analytically.
The hydraulic geometry of each cell is assumed to be fixed; that is, the length,
width, depth, slope and volumetric flow of each cell in the cascade are constant, although
these parameters will usually vary among cells. For computing yearly average human
exposure, or for an evaluation of chronic exposure for aquatic species, the discharge
through any given cell is assumed to be steady and uniform at a rate equal to the mean
annual discharge of the river at the point in question. Comparisons with criteria for short-
term or acute exposure should be based on seven-day, ten-year flow rates (7Q10). To
simplify the calculations, it is conservatively assumed that the suspended sediment
distribution is constant and that the sediment bed is stationary, although neither of these
two conditions is likely to be observed in actual streams. Consideration of storm conditions
is not included in the analysis of chronic exposure since transient events are assumed to
have a negligible effect over the long term. Dilution effects of tributary inflows are
accommodated by defining cell boundaries at each confluence along the study reach.
Contaminant mass loading from the sludge disposal site to the river is assumed to
occur at a constant rate. Neither upstream nor instream sources of contaminant are
considered. To reduce input requirements and simplify the calculations, it is assumed that
there is no net loss or gain of contaminant through interactions with the sediment bed.
The input mass is assumed to mix instantaneously and completely (i.e., vertically, laterally
and longitudinally) throughout each cell. First-order kinetics with a constant decay
parameter are assumed to provide an adequate description of organic chemical
transformation and degradation mechanisms. It is assumed that heavy metals neither decay
5-5
-------
TABLE 5-1
Assumptions for Methodology to Analyze Surface Water Pathways
Functional Area
Assumption
Ramifications
Contaminant Transport in Surface Water
Loading of contaminant to the stream is
steady-state.
Dpstream/instream sources of contaminant are
not considered.
Hydraulic geometry of each cell is fixed.
Steady-state loading may not be reached for
many decades for some contaminants at some
sites. Will over-predict loading in period
before steady-state is reached.
Only incremental concentration is
considered.
Unknown.
For estimating chronic exposure, it is
assumed that mean annual steady state
conditions prevail for all hydraulic,
sediment and water quality input parameters.
For estimating acute exposure, it is assumed
that flow is defined by the lowest expected
7-day flow per 10 years (7010), but loading
to stream is steady-state.
There is instantaneous and complete mixing
of the contaminant in cross section.
The sediment bed is stationary and there is
no contaminant loss or gain due to sediment
and/or benthic interaction.
Partitioning of contaminants between
dissolved and adsorbed phases is
instantaneous, and can be described by a
linear isotherm characterized by a single
partition coefficient for each cell.
Does not account for short-term fluctuations
or seasonal variations in contaminant
concentrations that occur during the year.
Acute exposure could be higher if low stream
flow coincides with period of high discharge
from aquifer.
Underpredicts concentrations, especially
near the loading source.
If the sediment bed is a sink, this
assumption overpredicts contaminant
concentrations; if the sediment bed is a
source, it underpredicts concentrations.
May underpredict dissolved concentrations if
equilibrium is not attained.
-------
TABLE 5-1 (cont.)
Functional Area
Assumption
Ramifications
Drinking Water Pathway of Exposure
ui
i
Fish Consumption Pathway of Exposure
Degradation and transformation of organic
contaminants can be described by first order
kinetics.
Heavy metals neither decay nor volatilize.
MEI ingests water at concentrations
predicted for specified location downstream
of the zone of aquifer discharge.
Exposure can be predicted based on dissolved
concentration of contaminant.
MEI ingests fish with contaminant
concentrations predicted on the basis of
total contaminant concentration (dissolved
and adsorbed) in the first modeled cell
below the zone of aquifer discharge.
Unknown.
This assumption overpredicts heavy metal
concentrations to the extent that
transformations are important.
May over- or under-predict actual exposure
if future location of MEI differs from
specified location.
Underpredicts exposure to the extent that
suspended solids are not removed during
water treatment prior to ingestion.
Likely to overpredict concentration of
contaminant in fish tissue if dissolved
concentration is better predictor of
bioconcentration under field conditions.
Use of estimated concentrations from first
cell may overpredict concentrations in fish
tissue if fish inhabit a broader range
within the stream.
-------
nor volatilize. Adsorption/desorption processes are assumed to attain instantaneous
equilibrium according to a linear isotherm characterized with a single partition coefficient.
5.3. CALCULATIONS
Derivation of criteria through either Tier 1 or Tier 2 involves four steps for each
contaminant:
1. derive reference water concentrations (RWC) for the surface water body;
2. determine the loading of contaminant from the aquifer to the stream, as a
function of the dry-weight concentrations of contaminant in sludge received
by the lagoon;
3. determine downstream concentrations of contaminant in the stream as a
function of loadings from the aquifer and
4. link the results of steps (1) through (3) to determine the maximum allowable
concentration of the contaminant in sludge received by the surface disposal
facility.
The first of these steps is the derivation of reference water concentrations. These
describe the levels of contaminant concentrations in a surface water body that are not
expected to cause adverse effects on human health or aquatic wildlife. Steps for deriving
the RWC for surface water will be discussed in Section 5.5.
Next, the model described in Chapter 4 can be used to determine expected loadings
of each contaminant to the stream, based on dry-weight concentrations of contaminant in
sludge received by or accumulating in the surface impoundment. As explained in Chapter
4, results from VADOFT and AT123D simulations are linear with respect to contaminant
concentrations in the sludge, so that a ratio (W0/N) can be derived for each contaminant
that describes the relationship between loading to the stream and the concentration of the
contaminant in sludge.
Once a downstream point of compliance has been selected for the calculations, the
cascading cells model can estimate the downstream concentrations of a contaminant as a
function of its loading from the aquifer to the stream. Within the model, the study reach
is depicted as a series of well-mixed cascading cells, as shown in Figure 5-1. The
contaminant outflow from an upstream cell represents the contaminant inflow to the next
downstream cell. From the principle of mass conservation, the steady-state concentration
5-8
-------
AQ
FIGURE 5-1
Definition Sketch for Cascading Cells Model
5-9
-------
in the nth cell of the cascade is:
C = "J (5.!)
Qfn + AnKtotn
where:
Cn = concentration of contaminant in nth cell of cascade (g/m3)
Wn.1 = contaminant mass inflow from upstream cell (g/sec)
Qfn = river discharge in nth cell of cascade (m3/sec)
Ap = top surface area of nth cell of cascade (m2)
Ktotn = net contaminant loss rate of nth cell of cascade (m/sec)
Equation 5-1 shows that the concentration of contaminant Cn, in the downstream cell is
directly proportional to Wn.1, the contaminant mass loading received from the cell
immediately upstream. Further, cell concentrations are inversely proportional to the sum
of two processes in the cell, dilution and degradation, as represented by Qfn and AnKtotn,
respectively. For convenience, it is assumed that the loading of contaminant into the first
cell, W0, originates entirely from the aquifer; background contaminant levels from
upstream sources are ignored. Within this context, Cn is best interpreted as the
concentration of contaminant relative to background or upstream values.
For purposes of application, it is more convenient to express the cell concentration,
Cn, in terms of W0, the contaminant loading from the sludge disposal facility to the first
cell of the cascade. This can be accomplished as follows. From Equation 5-1, the steady-
state concentration in cell (n-1) of the cascade can be written as:
Qfn-1 + An-1Ktotn-1
(5-2)
The loading of contaminant mass from cell (n-1) to cell n, or Wn.1, is found as the
product of concentration and discharge,
(5-3)
5-10
-------
Combining Equations 5-2 and 5-3 with 5-1 gives:
C = Wn-2 Qfn-1
Qfn + AnKtotn Qfn-1 + An-1Ktotn-1
With (n-1) repetitions, this procedure leads to the following expression for contaminant
concentration in the nth cell:
c . _ __ fn, _ ... __fi _ (5_4)
" Qfn + AnKtotn Qfn-1 + An-1Ktotn-1 Qf1 + A1Ktot1
With the introduction of a dimensionless cell "transport factor,"
-yf
(5-5)
Qfi + AiKtoti
and the definition of the cumulative cascade transport factor,
rn . nf (5-6)
Equation 5-4 can be written more compactly:
r wn
r =_o (5.7)
Qfn + An
The parameters defined in Equation 5-7 have clear physical interpretations. The term 7.
represents the proportion of contaminant mass that enters cell i and is transported
downstream to cell i+1. By definition, T?0=l. Conversely, the quantity I-TJ is the
5-11
-------
percentage of contaminant mass that enters cell i and is degraded there. The weighting
factor Tn gives the percentage of the initial contaminant load, WQ that is transported
downstream through the cascade to the nth cell in the sequence. The values fn for
n = 1,2,3... form a decreasing geometric type of series signifying the cumulative transport
effect of the cascading cells model.
The cascading cells model given in Equation 5-7 is straightforward to apply yet
sufficiently flexible to accommodate a wide variety of situations. The user need only
define the cells constituting the cascade and then, for each cell, estimate three parameters:
the river discharge (Qf), the cell top surface area (A), and the net contaminant loss rate
(Ktot). Further discussion of the determination of input parameters is given in Section 5.4.
The number of cells used depends on the physical conditions of the study reach and
on the judgment of the modeler. As a general rule, individual cells should be established
for reaches of the river over which conditions are reasonably uniform. Whenever there is
an appreciable change in cross-sectional geometry, bed slope, river discharge or net
contaminant loss rate, a new cell should be defined. This is particularly important with
tributary confluences, where cell boundaries must be established to properly account for
dilution effects. It is worthwhile to point out that the cascading cells model is the discrete
analog of the classic continuous plug flow model. If any cell in the cascade were
subdivided many times with transverse slices, the response for this "sliced" cell would
approach the exponential response of the plug flow model. In fact, either approach
(cascading cells or plug flow) could have been adopted for the surface water model.
One advantage of the model is the ease with which it can be used to calculate sludge
criteria. That is, given a maximum allowable contaminant concentration at some
downstream location (assumed to equal the RWC) the cascading cells model can determine
the corresponding mass loading from the sludge disposal site to the river. As can be seen
from Equation 5-7, the estimated concentration of contaminant in the last cell (Cn) is
directly proportional to the estimated loading of contaminant to the stream. Rearranging
Equation 5-7 yields a fixed ratio between estimated concentrations in the last cell of the
stream, and estimated loadings from the aquifer:
5-12
-------
C/W =
- ~
(Qfn + AnKtotn)
Since all terms on the right-hand side are known or can be estimated, Equation 5-8
provides a simple, direct method for estimating the maximum allowable loading of
contaminant to the river (W0) based on a maximum allowable value of Cn, which equals the
RWC.
The final step in the derivation of criteria is to combine results from earlier
calculations to yield maximum values for allowable contaminant concentrations in sludge
accumulated or deposited in the lagoon. In this context it is useful to define a "source-
receptor ratio" or SRRSW that represents the ratio of expected contaminant concentrations
at the selected downstream location to the dry-weight concentration of contaminant in
sludge received by the facility. Since all of the calculations described thus far are linear
with respect to contaminant concentrations:
SRRSW = Cn/N = (Cn/W0)(W0/Cl)(Cl/N) (5-9)
where:
SRRSH = source-receptor ratio, or the ratio of downstream concentrations of
contaminant to dry-weight concentrations of contaminant in sludge
(kg//)
C = downstream contaminant concentration in the surface water (mg//)
N = dry-weight concentration of the contaminant in sludge received by the
impoundment (mg/kg)
W0 = contaminant loading to the surface water (mg/sec)
Ct = contaminant concentration in liquid within the impoundment (mg//)
The first term on the right side of Equation 5-9 can be calculated from Equation 5-8; the
second term can be calculated with methods described in Section 4.3.1, and the third term
can be calculated from Equation 4-1.
Equation 5-9 can be used to derive maximum allowable concentrations of a
contaminant in sludge received by the surface disposal facility, as a linear function of the
reference water concentrations for that contaminant:
5-13
-------
Nmax = RWC/SRRSW (5-10)
where:
Nmax = the maximum allowable dry-weight concentration of this contaminant
for sludge received by this facility (mg/kg)
RWC = reference water concentration (mg//)
Derivation of reference water concentrations for surface water will be discussed in Section
5.5. As an alternative approach, SRRSW can be combined with other results to derive
criteria on the basis of simultaneous exposure through more than one pathway.
Consideration of multiple pathways of exposure will be discussed in Chapter 7.
5.4. INPUT PARAMETER REQUIREMENTS
In order to use the cascading cells model, the user must first define a series of cells;
guidance for selecting the number and size of cascading cells has been provided in Section
5.3. Once the configuration of cells has been established, it remains for the user to
specify three parameters for each cell: (1) the mean annual river flow, Qf; (2) the top
surface area, A; and (3) the net contaminant loss rate, Ktot.
5.4.1. Mean Annual River Flow. If the study reach is far from a gage, so that
records of mean annual flow rates are not available, then a simple regression analysis using
stream flow records from nearby rivers could be performed to estimate a regional
relationship between discharge and watershed area. Based on the watershed area along the
study reach, this regional relationship can provide an estimate of the mean annual flow.
For assessing potential risks to human health or wildlife from short-term, acute
exposures to contaminants, the 7Q10, or minimum average 7-day flow expected to occur
every 10 years, can be used in the equations in place of the mean annual flow rate. Values
for these low-flow rates can be obtained from district offices of the U.S. Geological
Survey, or from the GAGE data set accessible through the Graphical Exposure Modeling
System (U.S. EPA, 1988c).
5.4.2. Cell Area. The top surface area of the cell, A, is computed as the product
of the length, L, and the width, B, of each cell or,
-------
A = L B (5-11)
The cell length is the longitudinal distance between the upstream and downstream
boundaries of the cell. The cell width is the hydraulic top width of the river when
conveying the mean annual discharge. If possible, the top width measurement for each cell
should be based on actual river cross sections obtained from field surveys. It is quite
possible that surveyed cross sections along the study reach already exist as part of flood
insurance studies performed for the U.S. Army Corps of Engineers, or the Federal
Emergency Management Agency. This possibility should be explored before embarking on
an expensive field survey program. If, however, there are no existing site-specific field
data, then (as a minimum) cross sections should be obtained at the upstream and
downstream boundaries of all anticipated cells. In principle, the top width of each cell can
be found using a discharge rating curve and a typical cross section for each cell. For
purposes of this analysis, however, it may be sufficient to approximate the cell top width
as the transverse distance from left bank to right bank. If feasible, hydraulic field work
should also include verification of Manning's roughness coefficient and a measurement of
the suspended sediment concentration. On-site measurements can also help determine or
verify several of the other key parameters (e.g. kd, foc, and Ktot) to be discussed below.
5.4.3. Net Contaminant Loss Rate. It is convenient to express the net contaminant
loss rate, Ktot, as the sum of two distinct loss mechanisms:
Ktot = Kd + ks-
Here Kd represents the loss of the dissolved fraction due to all decay and exchange
processes in the water column and ks represents the loss due to interaction with the
sediment.
As mentioned previously, it is assumed that the sediment bed is stationary and that
storm events have a negligible effect oh long-term average concentrations of contaminant
5-15
-------
in the stream for cases where interaction with sediment does occur. This assumption
implies that k_. = 0 , meaning that loss of contaminants to the stream bottom is not
o
considered. This restriction has important ramifications for modeling heavy metals. If the
sediment interaction pathway is eliminated, then the overall loss rate for heavy metals will
be zero, since heavy metals neither decay nor volatilize. Stated another way, over the long
term, there is no accumulation of heavy metals in the sediment bed. Instead, the total mass
of the heavy metal transported in the water column is conserved. Hence, with the proposed
modeling approach, dilution is the only mechanism affecting the total concentration of a
heavy metal. It should be noted that under actual conditions, the sediment bed is likely
to be removed, at least in part, by routine losses and flood events. By assuming that the
sediment bed is stationary and at equilibrium, the methodology ignores contaminant losses
through this route, leading to a conservative estimate of contaminant concentrations in the
stream. Use of a more sophisticated model of contaminant transport in surface water (for
example, the WASP4 model) allows the modeler to consider these processes in more detail,
but does not seem warranted for the present application.
Given the suspended sediment concentration and the partition coefficient, the total
concentration of a heavy metal in the water column can be resolved into dissolved and
particulate fractions. Possible speciation associated with ionization, complexation or other
reactions is ignored.
For toxic organic chemicals, it is helpful to express the dissolved contaminant loss
rate as the weighted sum of two loss rates,
Kd =
5-16
-------
where:
Kd = dissolved contaminant loss rate (sec"1)
kn = "aggregate" decay rate (sec"1)
H = average depth of flow in the cell (m)
k = overall mass transfer coefficient for volatilization (m/sec)
f. = dissolved fraction of the total mass of contaminant (unitless)
The fraction dissolved is given by:
fd = (5-14)
1 + kd rsl
where:
f = the fraction of contaminant mass in dissolved phase (unitless)
kd = the partition coefficient of the linear isotherm describing the
adsorption-desorption process at equilibrium (unitless)
r = the ratio of the mass of solids to the volume of liquid in the flow
(kg//)
In surface water systems and other systems with a relatively low ratio of solids to liquid
on a volume basis, rsl can be approximated by the concentration of suspended sediment.
The partition coefficient for organic chemicals can be derived from organic carbon
partition coefficients presented in U.S. EPA (1986a), can be estimated with the CHEMEST
procedures in GEMS (U.S. EPA, 1988c), or can be estimated from octanol-water partition
coefficients with the following empirical relation (Thomann and Mueller, 1987):
2-8(focKow)
kd= - 2£_21! - (5-15)
where:
f = weight fraction of organic carbon of the total solids concentration
(unitless)
K = octanol-water partition coefficient of the contaminant (unitless)
Values of Kow and the corresponding kd for 10 critical toxic organic chemicals are
summarized in Table 5-2.
5-17
-------
TABLE 5-2
Properties of Selected Organic Chemicals
Organic Chemical
Benzene
Benzo(a)pyrene
BEHP
Chlordane
DDT
Dimethyl nitrosamine
Lindane
PCBs (Aroclor 1248)
Toxaphene
Trichloroethylene
L°glOKow
(unitless)
2.11-2.48
4.05-6.04
8.73
2.78
4.89-6.91
0.06
3.11-3.72
5.75-6.11
3.30
2.29
Partition
Coefficient
(//kg)
30
11,700
28,000
120
19,400
0.23
620
24,500
390
40
Henry's
Law Constant
(m3-atm/mol)
4.4xlO'3-5.5xlO'3
4.9xlO"7
3.0xlO'7
2.8xlO'6
1.6xlO"5-4.8xlO'5
3.3xlO'5
7.5xlO"5-4.8xlO"7
1.2xlO"2-2.4xlO'3
2.lxlO'1
8.8xlO"3
-4
From Equation (5-15) with rsl=10"* kg// and foc=0.10
5-18
-------
The overall mass transfer coefficient (k) in Equation 5-13 is calculated based on a
two-layer resistance model. Because the contaminant must pass through both liquid and
air to be released into the atmosphere, the overall resistance equals the sum of the liquid
and gas phase resistances. The overall mass transfer coefficient can thus be defined in
terms of individual liquid and gas coefficients:
1 _ 1 . RT , ,,
k ' *i H?g (5-16)
where
kt = the liquid phase mass transfer coefficient (m/sec),
k = the gas phase mass transfer coefficient (m/sec),
R = the ideal gas constant = 8.21xlO~5 (m3-atm/mol-K),
T = temperature (°K) and
H = Henry's law constant for contaminant (m3-atm/mol).
Numerous methods for calculating kt and k for water surfaces have been proposed
in the literature (Hwang, 1985; Hwang and Thibodeaux, 1985; MacKay and Leinonen, 1975;
MacKay and Yeun, 1983; Shen, 1982; Springer et al., 1984, U.S. EPA, 1987a; U.S. EPA,
1989c). This methodology follows an approach selected by U.S. EPA (1987a, 1989c) for
estimating volatilization from surface impoundments. That approach, which should also be
applicable to other surface water bodies, calculates mass transfer coefficients according to
two types of site characteristics: (1) the ratio of the surface's effective diameter (or "fetch")
to its depth and (2) the local average wind speed. Effective diameter is defined as
2(A/7r)°'5, where A is the area of the water surface.
For sites at which the ratio of fetch:depth is less than 14 and for which local average
wind speeds 10 meters above the surface (U10) are greater than 3.25 m/sec, U.S. EPA
(1987a) uses equations taken from MacKay and Yeun (1983) to calculate k.. These
equations are based on laboratory tests with a wind-wave tank, and estimate kt as a
function of the "friction velocity" of wind, defined as
U* = 0.01 U10 (6.1 + 0.63 U10)°'5 (5-17)
5-19
-------
where U10 is the average wind speed 10 meters above the liquid surface. If U* >_ 0.3
m/sec, then
kt = 1.0 x 10'6 + 34.1 x 10'4 U* ScL'°'5 (5-18)
otherwise, if U* < 0.3,
kt = 1.0 x 10"6 + 144 x 10'* U* 2.2 ScL'°'5 (5-19)
where ScL equals the Schmidt number on the liquid side, defined as
ScL = ^ I (pHDcw) (5-20)
and where:
Hu = viscosity of water (g/cm-sec),
pw = density of water (g/cm3) and
DCH = diffusivity of constituent in water (cm2/sec).
For most cells likely to be specified to represent surface water bodies, the ratio of
fetch:depth will be greater than 14. For water surfaces in areas where the fetchrdepth
ratio is greater than 14 or U10 is less than 3.25 m/sec, the methodology uses three different
expressions for kt taken from Springer et al. (1984) as reported in U.S. EPA (1987a). For
all sites where U10 < 3.25 m/sec,
kt = 2.78 x 10'6 [Dcw / Dether}2/3 (5-21)
where Dether is the diffusivity of ether in water (8.5 x 10"6 cm2/sec). For those with U10
> 3.25 m/sec and fetch-to-depth ratios (FD) between 14 and 51.2,
= [2.605 x ID"9 FD + 1.277 x 10'7] U1Q2 [DCH/Dether]2/3 (5-22)
5-20
-------
where FD is the fetchrdepth ratio. Finally, for facilities where U10 >_ 3.25 m/sec and FD
>_ 51.2,
kt = 2.611 x 1(T7 U102 [Dcw/Dether]2/3 (5-23)
Calculation of the mass transfer coefficient for the gas phase is based on Hwang
(1985). For all values of FD and U10, k (in m/sec) is calculated from
kg = 1.8 x 10"3 U10°'78 ScG"°-67de"°-11 (5-24)
where ScG equals the Schmidt number on the gas side, defined as
ScG = Ma / (PaDca) (5-25)
u a o \fO x
and where:
/ia = viscosity of air (g/cm-sec),
p = density of air (g/cm3) and
a
D = diffusivity of constituent in air (cm /sec)
2,
Default values for /ia and pa (1.8xlO"4 g/cm-sec and 1.2xlO"3 g/cm-sec at STP,
respectively) can be taken from Incropera and DeWitt (1985). Equations 5-16 through 5-
25 are sufficient to estimate k, the overall mass transfer coefficient for the dissolved
fraction of the contaminant.
The transformation processes of photolysis, hydrolysis and biodegradation are assumed
to follow linear kinetic reactions. The combined effect of these transformations is
represented with an "aggregate" loss rate, defined as:
kagg = kp + kh + kb (5-26)
where:
kagg = aggregate loss rate (sec"1)
k = overall loss rate due to photolysis (sec )
kh = overall loss rate due to hydrolysis (sec"1)
kb = overall loss rate due to biodegradation (sec"1)
5-21
-------
Estimates of organic chemical loss rates due to photolysis, hydrolysis and biodegradation
are tabulated by Schnoor et al. (1987); biodegradation rates are also tabulated by Fitter,
1976. The overall loss rate due to biodegradation is sensitive to water temperature. When
the temperature is different from 20°C, the value for kb should be adjusted using:
where the temperature correction factor, ©, is 1.06 (Thomann and Mueller, 1987) and the
temperature is expressed in degrees Celsius.
5.4.4. Hydraulic Characteristics. Several of the transfer rate expressions require
estimates of hydraulic characteristics such as mean depth and/or mean velocity. If steady
uniform flow is assumed for the length of the cell, the depth and velocity of the flow can
be estimated using Manning's formula. If Qf is the total flow through the cell expressed
in metric units (i.e., cubic meters per second), then the discharge per unit width, qd is
computed as:
qd - — (5-28)
B
where:
qd = discharge per unit width (m /sec)
Qf = mean annual flow (m3/sec)
B = width of cell (m)
and the depth of flow in the cell can be estimated with
(nqd)0'6
H = (5-29)
S0.3
where:
H = mean depth of flow (m)
n = Manning's roughness coefficient (unitless)
S = longitudinal slope of the cell (unitless)
5-22
-------
The longitudinal cell slope can be obtained from the bed slope of the river. Values for
Manning's roughness coefficient vary considerably among streams. For natural streams,
values range from 0.025-0.033 for smoothest beds, 0.045-0.060 for roughest beds, and
0.075-0.150 for very weedy streams (Thomann and Mueller, 1987). Guides for determining
Manning's n have been compiled by Chow (1959) and Barnes (1967).
Combining Equations 5-28 and 5-29 gives the average flow velocity:
U = -- (5-30)
8 H
The cascading cells model assumes well-mixed conditions in each cell. The model will
better represent well-mixed conditions if the length of the cell, L, is greater than Lm , the
length required for complete mixing. To check this assumption, a rough approximation of
the distance Lm (measured from a side-bank discharge to the zone of complete mixing)
can be obtained from:
B2
— (5-31)
H
Where possible, cell lengths should be selected so that L >_ Lm.
5.4.5. Wind Speed. The wind speed (U10) needed in Equations 5-17, and 5-22 through
5-24, can be obtained from Figure 5-2, which gives the prevailing direction and the mean
annual speed of wind for site3 around the nation (Environmental Sciences Service
Administration, 1968).
5.5. HEALTH AND ENVIRONMENTAL EFFECTS
Contamination of surface water can cause adverse effects for both wildlife and
humans:
1. Contaminants can cause adverse effects on fish and other biota inhabiting
streams, lakes and estuaries. If the concentration of a particular pollutant
5-23
-------
FIGURE 5-2
Mean Annual Wind Speed in the United States
5-24
-------
exceeds threshold levels, fish and other biota may die or suffer adverse health
or reproductive effects.
2. Adverse effects on human health can be direct (by water consumption) or
indirect (by fish or shellfish consumption). If the human diet includes fish
or shellfish that have bioconcentrated a toxic pollutant from surface waters,
the indirect mode of toxicant consumption may pose more risk than the direct
mode.
The first step in deriving criteria for surface disposal of sludge is the derivation of a
reference water concentration (RWC) for each contaminant, below which adverse effects
are not expected. Such reference concentrations should be protective of both the
environment (e.g., aquatic life) and human health.
5.5.1. Aquatic Life Protection. Protection of aquatic life from long-term effects
should be based on the Ambient Water Quality Criteria (U.S. EPA, 1980c). The AWQC
consist of two concentrations: criteria for chronic exposures that should not be exceeded
as a yearly average; and criteria for acute exposures that should not be exceeded, on
average, in a 24-hour period. Estimated yearly average concentrations should be compared
with criteria for chronic exposures, and estimates based on 7Q10 flows should be compared
with acute criteria. For chemicals for which criteria are not available, the literature should
be evaluated to determine whether useful data have become available since the AWQC were
developed.
5.5.2. Threshold-Acting Toxicants. If the only source of potential exposure is
drinking water, then the RWC for a threshold-acting toxicant is derived with methods
described in Section 4.5.1. If the only source of potential exposure is fish living in
contaminated surface waters, then the RWC is calculated as:
RWC = [(RfD BW RE'1) - TBI] / (BCF If) (5-32)
where:
RWC = reference water concentration (mg//)
RfD = reference dose (mg-day/kg)
BW = human body weight (kg)
RE = relative effectiveness of ingestion exposure (unitless)
TBI = total background intake rate of contaminant from all other
sources of exposure (mg/day)
BCF = unadjusted bioconcentration factor in fish (//kg)
L = total rate of fish ingestion (kg/day)
5-25
-------
If humans are exposed to sludge contaminants through both drinking water and consumption
of fish, the reference concentration is calculated as:
RWC = [(RfD BW RE'1) - TBI] / [IH + (BCF If)]
where:
Iw = total water ingestion rate (//day)
The definition and derivation of bioconcentration factors and fish ingestion rates are
provided below. Other parameters in Equation 5-32 are described in Section 4.5.
5.5.2.1. BIOCONCENTRATION FACTOR (BCF) -- Bioconcentration is the
ability of living organisms to accumulate substances to higher than ambient level
concentrations. The degree to which a chemical accumulates in an aquatic organism above
ambient concentrations is indicated by the bioconcentration factor (BCF), which is defined
as the quotient of the concentration of a substance in all or part of an aquatic organism
(mg/kg fresh weight) divided by the concentration in the water to which the organism has
been exposed (mg//). The BCF is usually determined at equilibrium conditions, or for 28-
day exposures, is based upon the fresh weight of the organism, and is specified in units
of mg/kg per mg//, or //kg (or unitless, since I/ of water has a mass of 1 kg). Other
terms are used to describe increases of environmental pollutant in organisms, such as
biomagnification, bioaccumulation or ecological magnification. Bioconcentration is
distinguished from these other processes in that BCF considers only the uptake of
contaminant from physical surroundings (i.e., the contaminated water in which the organism
lives) and does not consider uptake from food sources. Although BCF has been
documented as the primary pathway for accumulation in numerous studies (Marcelle and
Thome, 1984; Bahner et al., 1977; Clayton et al., 1977), there is also evidence that
biomagnification by aquatic food chains can be important under certain environmental
circumstances (Lee et al., 1976).
5-26
-------
Bioconcentration factors are specific for the pollutant and for the species absorbing
the pollutant. Pollutants that are lipophilic and resistant to biodegradation are most likely
to bioconcentrate in living organisms. Initial diffusion into the organism occurs by rapid
surface adsorption or partitioning to the lipoprotein layer of cell membranes. Once in the
bloodstream, subsequent accumulation of the pollutant into particular compartments of the
organism is dependent on the metabolic capabilities of the species and the lipid content of
the individual organism. With continued exposure, an equilibrium condition is eventually
reached where the rate of contaminant excretion is equal to the rate of uptake.
Bioconcentration factors can be estimated through laboratory experiments, field
studies, correlations with physicochemical factors such as octanol-water partition
coefficients, and models based on pollutant biokinetics coupled with fish energetics. In
developing the ambient water quality criteria, the U.S. EPA used laboratory data in the
calculation of BCFs. Field data are often less useful than laboratory data because of the
difficulty of establishing the field pollutant concentrations over time and of determining
the range of territory inhabited by the organism. BCFs calculated from field data tend to
be greater than those calculated from laboratory data because they may also include
ingestion of the pollutant through the consumption of prey, sediment and water, as well
as direct absorption from water.
Where laboratory and field data are not available, bioconcentration factors can be
estimated by several methods. Correlations between BCFs and octanol-water partition
coefficients, water solubility and soil adsorption coefficients have been documented. Veith
et al. (1979) developed the following equation using the correlation between BCF and the
(unitless) octanol water partition coefficient (KQW):
Iog10 BCF = 0.85 Iog10 Kow - 0.70
Using this equation, Veith et al. were able to estimate BCFs to within 60% before
laboratory testing. The equation was developed using data from the whole-body analyses
of fish with approximately 7.6% lipids (U.S. EPA, 1980c). The U.S. EPA adopted the
5-27
-------
equation for use in determining BCFs in the exposure sections of the health effects chapters
of the Ambient Water Quality Criteria Documents in cases where the BCF was not available
from other data. In a later study, Veith et al. (1980) used the results of their own
laboratory experiments and data from other laboratories for a variety of fish species and
84 different organic chemicals to obtain the following modification of their original
equation:
Iog10 BCF = 0.76 Iog10 Kou - 0.23
Equations similar to the ones developed by Veith et al. (1979, 1980) have been developed
for more specific chemical classes and particular aquatic species (Veith et al., 1979; Neeley
et al., 1974). Other investigators (Norstrom et al., 1976) have developed more elaborate
models using pollutant biokinetics and fish energetics in addition to using octanol-water
partition coefficients to predict bioconcentration factors.
Since bioconcentration for lipophilic compounds is based on the lipid content of the
organism, the estimated or measured BCF for these lipophilic compounds must be adjusted
by the average lipid content of seafood consumed in the U.S. diet. In 1980, the U.S. EPA
determined that the average lipid content of the freshwater and estuarine species, weighted
by average daily consumption, was about 3% (U.S. EPA, 1980c). Since fresh and estuarine
waters would be affected by surface water contamination from surface disposal sites, a lipid
content of 3% should be assumed when deriving the reference water concentration. The
bioconcentration factor should be adjusted as follows:
BCFa = BCF (LCd/LCe)
where:
BCFa = adjusted bioconcentration factor (//kg)
BCF = unadjusted bioconcentration factor (//kg)
LCd = lipid content of dietary seafood (kg/kg)
LCe = lipid content of experimental organism (kg/kg)
5-28
-------
5.5.2.2. FISH CONSUMPTION RATE (If) — The most recent fish consumption
document from the U.S. Department of Commerce (1985) reports that total per capita fish
and shellfish consumption ranged from 12.8 pounds per year in 1980 to 13.6 pounds per
year in 1984, as shown in Table 5-3. The latter value corresponds to a daily intake of 16.9
grams of fish (all kinds). These figures do not include recreational catch, of which average
consumption is estimated to be an additional 3-4 pounds per year or 3.7-5 grams per day
(U.S. EPA, 1980a). If 3.5 pounds per year, or 4.4 grams per day, is consumed from
recreational fishing, the total average per capita intake of all types of seafood is about 21.3
grams per day.
Surface water contamination from surface disposal sites could affect freshwater and
estuarine species, but is unlikely to affect the marine species that constitute a substantial
portion of seafood in the U.S. diet. To estimate the average daily consumption of
freshwater and estuarine species, the U.S. EPA analyzed data from a survey of fish
consumption in 1973-74 (as reanalyzed by U.S. EPA, 1980a) and eliminated all species not
taken from fresh or estuarine waters (Stephan, 1980). Per capita consumption of these
species was estimated to be 6.5 grams per day, or about half the estimate of total fish
consumption derived from that study, 13.4 grams per day. Based on this information, it
is reasonable to conclude that fresh and estuarine species constitute about 50% of the total
fish consumption in the U.S. diet. Combining this assumption with an estimated average
total consumption of fish of 21.3 grams yields an estimate of daily consumption of
freshwater and estuarine species of about 10.6 grams.
It should be noted that average daily rates of fish consumption vary substantially by
region, age, race and religion. An analysis of fish consumption data by SRI International
found that per-capita fish intake by black and Jewish populations to be double the average
value for the total U.S. population (U.S. EPA, 1980a). The New England and East South
Central regions of the United States had the highest regional rates of fish consumption.
Consumption levels at the upper 95th percentile of the national distribution were 300-400%
of the national average. The Asian population in the United States showed the highest 95th
percentile consumption rate; this value was 502% of the reported national average.
5-29
-------
TABLE 5-3
U.S. Annual Per Capita Consumption of
Commercial Fish and Shellfish, 1960-1984a-b
Per Capita Consumption
(pounds of edible meat)
Year
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977e
Civilian Resident
Population
(million persons)
178.1
181.1
183.7
186.5
189.1
191.6
193.4
195.3
197.1
199.1
201.9
204.9
207.5
209.6
211.6
213.8
215.9
218.1
Fresh
& Frozenc
5.7
5.9
5.8
5.8
5.9
6.0
6.1
5.8
6.2
6.6
6.9
6.7
7.1
7.4
6.9
7.5
8.2
7.7
Cannedd
4.0
4.3
4.3
4.4
4.1
4.3
4.3
4.3
4.3
4.2
4.5
4.3
4.9
5.0
4.7
4.3
4.2
4.6
Cured
0.6
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.4
0.4
0.5
0.5
0.4
0.5
0.4
0.5
0.4
Total
10.3
10.7
10.6
10.7
10.5
10.8
10.9
10.6
11.0
11.2
11.8
11.5
12.5
12.8
12.1
12.2
12.9
12.7
5-30
-------
TABLE 5-3 (cont.)
Per Capita Consumption
(pounds of edible meat)
Year Civilian Resident Fresh
Population
(million persons)
Population & Frozen0 Cannedd Cured Total
1978e
1979e
1980e
1981e
1982e
1983e
1984e
220.5
223.0
225.6
227.7
229.9
232.0
234.0
8.1
7.8
8.0
7.8
7.7
8.0
8.3
5.0
4.8
4.5
4.8
4.3
4.8
5.0
0.3
0.4
0.3
0.3
0.3
0.3
0.3
13.4
13.0
12.8
12.9
12.3
13.1
13.6
aSource: U.S. EPA, 1989d. Adapted from U.S. Department of Commerce, 1985
bThese consumption estimates refer only to fish and shellfish entering commercial channels, and
they do not include data on consumption of recreationally caught fish and shellfish, which since
1970 is estimated to be between 3 and 4 pounds (edible meat)/person annually. The figures are
calculated on the basis of raw edible meat (e.g., excluding bones, viscera, shells). The U.S.
Department of Agriculture (USDA) consumption figures for red meats and poultry are based on
the retail weight of the products, as purchased in retail stores. The USDA estimates are the net
edible weight to be -70-95% of the retail weight, depending on the cut and type of meat. From
1970 through 1980, data were revised to reflect the results of the 1980 census.
cBeginning in 1973, data include consumption of artificially cultivated catfish.
Based on production reports, packer stocks and foreign trade statistics for individual years.
eDomestic landings data used in calculating these data are preliminary.
5-31
-------
Applying the same percentage increase to an estimated average consumption rate of 10.6
grams per day yields an estimated consumption rate of about 53 grams of freshwater fish
per day, or 43 pounds per year. Rates of fish consumption by various demographic
variables are listed in Table 5-4.
5.5.3. Carcinogens. Section 4.5.2 describes an approach for deriving reference water
concentrations for carcinogens in drinking water. If the only source of pollutant exposure
is ingestion of contaminated fish, then the RWC is derived as:
RWC - (RL BW)/(q1* RE BCF If)
where:
q. = upper bound human cancer potency (kg/mg-day)
RL = upper bound risk level (unitless)
If sources of human exposure include both ingestion of water from the affected surface
water body and consumption of fish that live in the affected surface water body, then the
RWC is calculated as:
(RL BW)
RWC =
(q/ RE)[IW + (BCF If)J
This method for deriving the RWC for carcinogens is thought to be conservative, because
the human cancer potency value (q,,*) is based on a 95th percentile confidence interval.
All parameters have been defined in previous sections.
5.6. SAMPLE CALCULATIONS
5.6.1. Analysis of Exposure for the Most Exposed Individual. Use of the proposed
surface water quality model is illustrated with sample calculations for one organic chemical
(benzene) and one heavy metal (lead) at three locations in the United States: the Contoocook
River near Antrim, New Hampshire, Bird Creek near Tulsa, Oklahoma, and the Columbia
Slough near Portland, Oregon.
Table 5-5 provides input parameters used to describe the chemical characteristics of
benzene and lead. As can be seen from the table, zero rates of photolysis and hydrolysis
5-32
-------
TABLE 5-4
Fish Consumption by Demographic Variables*
Demographic Category
Mean Consumption
(g/day)
Upper 95th Percentile
(g/day)
Race
Sex:
Age
!•
£.•
Caucasian
Black
Oriental
Other
Female
Male
(years):
0-9
10-19
20-29
30-39
40-49
50-59
60-69
70+
14.2
16.0
21.0
13.2
13.2
15.6
6.2
10.1
14.5
15.8
17.4
20.9
21.7
13.3
41.2
45.2
67.3
29.4
38.4
44.8
16.5
26.8
38.3
42.9
48.1
53.4
55.4
39.8
Census Region:
New England
Middle Atlantic
East North Central
West North Central
South Atlantic
East South Central
West South Central
Mountain
Pacific
16.3
16.2
12.9
12.0
15.2
13.0
14.4
12.1
14.2
46.5
47.8
36.9
35.2
44.1
38.4
43.6
32.1
39.2
Source: U.S. EPA, 1989d. Adapted from U.S.EPA, 1980a
5-33
-------
TABLE 5-5
Chemical Properties of Benzene and Lead
Chemical Property
Kou
kd
Hc
kb
kP
kh
Dca
Dcw
^w
RWC (drinking)
RWC (fish)
Benzene
135
*
5.5xlO"3
l.VxlO'6
0.0
0.0
8.8xlO'2
9.8xlO'6
l.SxlO'4
1.2xlO'3
8.8xlO'3
Lead (Units)
(unitless)
234 (//kg)
(atm-mol/m3)
0.0 (sec"1)
0.0 (sec'1)
0.0 (sec'1)
(cm2/sec)
(cm2/sec)
(kg//)
5.0xlO'3 (mg//)
(mg//)
*Varies according to suspended solids concentration and fraction of organic carbon
in each stream
-------
have been assumed for both contaminants. Of the remaining parameters, only the partition
coefficient is relevant to model calculations for lead. As in Chapter 4, Lead is assumed
to have a partition coefficient of 234 //kg, appropriate for sand (U.S. EPA, 1986a).
Partition coefficients for benzene are calculated separately for each site with Equation 5-
15, based on the octanol-water partition coefficient and assumed values for fQC and rgl.
Most of the hydraulic input parameters for the sample calculations have been derived from
site-specific data and the U.S. Geological Survey (1982, 1983, 1988a); water quality
parameters have been obtained from available literature. The following sections present a
Tier 2 analysis for these sites.
5.6.1.1. SAMPLE CALCULATION: ANTRIM, NEW HAMPSHIRE -- This study
area encompasses a 25-kilometer reach of the Contoocook River between the towns of
Antrim and Hillsboro in southern New Hampshire. The Antrim wastewater treatment plant
accumulates and stores sludge in three impoundments located next to the Contoocook River.
Hillsboro was selected as the location of the MEI since it is the first community
downstream from the sludge impoundment that would use the Contoocook River as a source
of water.
5.6.1.1.1. Defining Cells for the Cascading Cells Model -- To represent
contaminant transport in the Contoocook River, the study reach was divided into three cells
as follows:
• Cell 1 extends from the Antrim sludge impoundment downstream for
about 17 kilometers to the confluence of Beards Brook;
• Cell 2 extends from Beards Brook downstream for about 2.4 kilometers
to the Hillsboro Paper Company dam;
• Cell 3 extends from the Hillsboro dam for about 5.1 kilometers to the
downstream limits of Hillsboro.
The process of selecting these cells was guided by the dual objectives of minimizing
the total number of cells in the cascade, and yet maintaining uniform geometry and
discharge through each cell. Selection of the final cascade configuration was based on a
review of the longitudinal profile of the Contoocook River (Figure 5-3), river cross-
5-35
-------
CONTOOCOOK RIUER, NEW HAMPSHIRE
BETWEEN ANTRIM AND HILLSBORO
O
id
8
m
-------
sections, and topographic quad maps showing tributary inflows along the entire study reach.
A proportional sketch of the three-cell cascade is shown in Figure 5-4.
To apply the cascading cells model given in Equations 5-7 and 5-8 to the
configuration shown in Figure 5-4, it is necessary to estimate the three parameters Qf, A
and Ktot for each cell. Estimates of the mean annual discharge, Qf for the main stem
of the Contoocook River and for its tributaries were obtained from the U.S. Geological
Survey district office in Concord, New Hampshire. Flow values for all three cells are
summarized in Table 5-6. As expected, the mean annual flow increases in the downstream
direction.
The top surface area of each cell, A, was computed with Equation 5-11 after first
measuring the width, B, of a rectangular channels fitted onto representative cross-sections
of the Contoocook River, as shown in Figure 5-5 for Cell #2 of the cascade. At Cell 2,
for example, with length L2 = 2410 meters and an estimated top width of B2 = 36.6
meters (from Figure 5-5), Equation 5-11 gives the top surface area for Cell 2 as:
A2 = (L2)(B2) = (2,410)(36.6) = 88,200 m2
With a mean annual discharge Qf2 = 17.1 m3/sec and a cell width of B2 = 36.6 m, the
discharge per unit width for Cell 2 is given by Equation 5-28:
B2 36.60
•>
0.47 m2/sec
Substituting this result into Equation 5-29, with an assumed Manning's n2 = 0.06 and a
measured bed slope S2 = 0.00027, gives the depth of uniform flow in Cell 2 as:
6 [(0.06K0.47)]0-6
= - —
S20'3 (0.00027)0'3
= 1.37 m
5-37
-------
1
FIGURE 5-4
Three Cell Cascade for Contoocook River
Between Antrim and Hillsboro, NH
5-38
-------
TABLE 5-6
Site-Specific Input Parameters
Contoocook River, New Hampshire
Parameter Cell 1
L 16,890
B 30.5
H 1.27
A 515,000
V 654,000
Qf 12.6
qd 0.41
Us 0.32
n 0.06
S 0.00027
Lm 2000
U10 3.2
T 25
rsl 5xlO'5
foe °'18
Cell 2
2410
36.6
1.37
88,200
121,000
17.1
0.47
0.34
0.06
0.00027
2800
3.2
25
5xlO'5
0.18
Cell 3
5077
45.7
1.24
232,000
287,000
18.0
0.39
0.31
0.06
0.00027
4500
3.2
25
5xlO'5
0.18
(Units)
(m)
(m)
(m)
(m2)
(m3)
(m3/sec)
(m2/sec)
(m/sec)
(unitless)
(unitless)
(m)
(m/sec)
(°C)
(kg//)
(unitless)
5-39
-------
183 -
181
\
U
179
CONTOOCOOK RIVER, NEW HAMPSHIRE
BETWEEN ANTRIM AND HILLSBORO
177
176
CroM S«ction
Riuvr Kilomi
•t«r 18.7
j_
I
40 88 120
TRANSVERSE DISTANCE
160
FIGURE 5-5
Typical Cross Section of Contoocook River
5-40
-------
Substituting the estimated values for qd2 and H2 into Equation 5-30 gives the flow
velocity through Cell 2:
0.47
U , = - = = 0.34 m/sec
82 H2 1.37
These calculations are performed for all three cells of the cascade; the resulting hydraulic
input parameters for the Contoocook River are summarized in Table
5-6.
From Equation 5-31 with Us1 = 0.32 m/sec, B, = 30.5 m, and H1 = 1.27 m, the
approximate distance downstream from the "point" of contaminant loading to the zone of
complete mixing is:
(30.5)2
Lm1 = (8.53)(0.32) 2000 m
1.27
Since L = 2000 < L1 = 16,890 m, the contaminant should be reasonably well-mixed
in the cross section before it moves from Cell 1 to Cell 2.
The mean annual wind velocity at the site is estimated from Figure 5-2 to be U10
= 3.2 m/sec. The surface water temperature is taken to be T = 25 °C. The Department
of Environmental Services in Concord, New Hampshire, has measured the suspended solids
concentration in the Contoocook River at S.OxlO"5 kg// along the study reach; this value
is used for rgl. Estimates for U10, T and rgl are assumed to be the same for all three
cells in the cascade and are listed in Table 5-6.
5.6.1.1.2. Estimating Net Contaminant Loss Rates -- Determination of the net
contaminant loss rate, Ktot for benzene requires application of Equations 5-12 through 5-
15 using a four-step procedure:
1. Estimate fd, the dissolved fraction of benzene.
2. Estimate k, the overall mass transfer rate for benzene.
5-41
-------
3. Estimate k , the "aggregate" decay rate for benzene.
4. Estimate Ktot, the net contaminant loss rate for benzene.
From Table 5-2, the log1Q (KQH) is taken as 2.13 and, hence, the octanol-water partition
coefficient for benzene is KQW = 135. The suspended solids concentration in the study
reach is estimated to be rsl = 5.0xlO~5 kg//. It is also reported by the New Hampshire
Department of Environmental Services that the concentration of suspended organic carbon
is 9xlO~6 kg//, so the weight fraction of organic carbon is foc = (9xlO"6/5xlO~5) = 0.18.
Substituting these values for KQW, foc and rsl into Equation 5-15 gives,
(2.8)(0.18)(135)
1.4 + (5xlO'5)(0.18)(135)
= 48.6 //kg
Substituting this result into Equation 5-14 gives the dissolved fraction,
kd rsl
f = —
d 1 + (48.6)(5xl(T5)
= 0.9976
which indicates that virtually all (i.e., 99.76%) of the benzene will be in solution in the
water column. This result is not surprising, since the values for both kd and rsl are
relatively low.
Derivation of the overall mass transfer coefficient k for Equation 5-13 requires
Henry's constant and the two individual mass transfer coefficients kg and kt. From Table
5-42
-------
5-2, Henry's constant for benzene is assumed to be 0.0055 atm-mole/m3. The liquid film
coefficient is estimated with Equation 5-21 for the wind velocity of 3.20 m/sec, DCW =
9.8xlO'6 cm2/sec, and Deth = 8.5xlO'6 cm2/sec:
kt = 2.78xlO'6(Dcw/Deth)2/3
= 2.78xlO"6 (9.8xlO'6/8.5xlO'6)2/3
= S.lxlO"6 m/sec
The Schmidt number for benzene is calculated from Equation 5-25 as:
Scg = Ma/(Pa Dca)
(1.8xlO'4)/[(1.2xlO'3)(8.8xlO~2)]
1.70
Next, the mass transfer coefficient for gas is estimated from Equation 5-24 with an
effective distance estimated as (2)(88,200/7r)°-5 = 355 m:
kg = 1.8X10'3 U10°-78 ScG-°'67 de-°'11
l.8xlO"3(3.2)°-78(l.70)-°-67(355)'°-11
1.6xlO"3 m/sec
Finally with Hc=5.5xlO~3 atm-m3/mol, R=8.21xlO"5 atm-m3/mol-K, and
T=298° K, the overall mass transfer rate for Cell 2 is found using Equation
5-16:
1 RT
kl Hckg
1 (8.21xlO'5)(298)
3.1xlO'6 (5.5xlO'3)(1.6xlO~3)
-------
which gives k2 = 3.0xlO"6 m/sec, as listed in Table 5-7.
The aggregate decay rate, k , represents the summed effects of contaminant loss
due to photolysis, hydrolysis and biodegradation. Of these three decay mechanisms, only
biodegradation appears to significantly impact benzene (Schnoor et al. 1987). The benzene
decay rate at 20°C due to biodegradation is kb = 1.27xlO"6 sec"1. With Equation 5-27 and
© = 1.06, this loss rate is corrected for the assumed temperature of the river (25°C):
[kb]25 = 1.27xlO"6(1.06)(25"20)
= 1.70xlO~6 sec"1
With k = 0.0, kh = 0.0 and kb given above, Equation 5-26 gives the aggregate decay rate
for benzene as k = 1.7xlO"6 sec"1.
The overall net loss rate for benzene is now computed with Equation 5-12. In Cell
2, with H2 = 1.37 m, fd = 0.9976, k = S.OxlO"6 m/sec and kagg = 1.7xlO'6 sec"1, Equation
5-12 gives,
Ktot~ Kd + ks
= [(1.7xlO'6)(1.4)+(3.0xlO"6)](0.9976)
= 5.4xlO"6 m/sec
The four steps outlined above are applied to all three cells in the cascade. Results for
benzene are summarized in Table 5-7.
These steps are not needed for heavy metals. As mentioned previously, it is assumed
that heavy metals neither decay nor volatilize. Further, it is assumed that the sediment
bed does not act as a source or a sink of heavy metals. Hence the net loss rate for lead
in the Contoocook River is zero as summarized in Table 5-8. The only parameter to be
estimated for lead, then, is the partition coefficient, which in this case is taken to be
-------
TABLE 5-7
Model Results for Benzene
Contoocook River, New Hampshire
Parameter Cell 1
kd 48.6
fd 0.9976
kt 3.1xlO'6
kg 1.5xlO'3
k 3.0xlO'6
kagg 1-7x10"6
Ktot 5.2x10-6
-y 0.82
r i.o
C/W0 6.6xlO'2
Cd/W0 6.5xlO'2
W0/N l.OxlO'3
SRRSW 6.6xlO'5
Nmax 18
Cell 2
48.6
0.9976
S.lxlO"6
1.6xlO"3
3.0xlO'6
1.7xlO'6
5.4xlQ-6
0.97
0.82
4.7xlO'2
4.7xlO"2
l.OxlO'3
4.7xlQ-5
25
Cell 3
48.6
0.9976
3.1xlO'6
1.6xlO'3
3.0xlO'6
1.7xlO'6
5.1xlO'6
0.94
0.80
4.2xlO'2
4.2xlO'2
l.OxlO'3
4.2X10'5
28
(Units)
(//kg)
(unitless)
(m/sec)
(m/sec)
(m/sec)
(sec'1)
(m/sec)
(unitless)
(unitless)
(sec//)
(sec//)
(kg/sec)
(kg//)
(mg/kg)
-------
TABLE 5-8
Model Results for Lead
Contoocook River, New Hampshire
Parameter Cell 1
kd 234
fd 0.99
kt 0.0
kg 0.0
k 0.0
kb 0.0
kagg 0.0
*tot 0.0
7 1.0
r i.o
C/W0 S.OxlO'2
Cd/W0 7.9xlO'3
W0/N 7.3X10'3
SRRSW 5.7xlO"4
Nmflx 9
Cell 2
234
0.99
0.0
0.0
0.0
0.0
0.0
0.0
1.0
1.0
5.9xlO'2
5.8xlO'2
7.3xlO'3
4.2xlO"4
12
Cell 3
234
0.99
0.0
0.0
0.0
0.0
0.0
0.0
1.0
1.0
5.6xlO"2
5.5xlO'2
7.3xlO'3
4.0xlO"4
12
(Units)
(//kg)
(unitless)
(m/sec)
(m/sec)
(m/sec)
(sec'1)
(sec'1)
(sec'1)
(unitless)
(unitless)
(sec//)
(sec//)
(kg/sec)
(kg//)
(mg/kg)
5-46
-------
kd = 234 //kg, an appropriate value for sand (U.S. EPA, 1986a) within the range reported
by Baes, et al. (1984). From Equation 5-14 with rgl = 5xlO~5 kg//, the fraction of lead
dissolved is:
fd =
kd rsi
1
1 + (234)(5xlO'5)
= 0.99
This result indicates that lead will be distributed throughout the water column in dissolved
(99%) and paniculate (1%) form.
5.6.1.1.3. Estimating Downstream Concentrations of Contaminant --The
estimates of net contaminant loss rate (Ktot) derived in Section 5.6.1.1.2 can be combined
with values for the top area of each cell (A) and the flow rate for each cell (Qf) to derive
estimates of 7, T, and ultimately Cn/WQ. Values of the dimensionless transport factors, 7,
and the cumulative transport factor, T, are computed using Equations 5-5 and 5-6,
respectively: results are summarized in Table 5-7 for benzene and Table 5-8 for lead.
For benzene, Equation 5-5 gives:
12.6
7l = 0.82
12.6 + (515,000)(5.2xlO'6)
and
17.1
17.1 + (88,200)(5.4xlO'6)
= 0.97
Since I = 1 by definition, it follows from Equation 5-6 that
-------
n 75 = 7n Ti T, = (1.00)(0.82)(0.97) = 0.80
The above result shows that approximately 80% of the benzene entering the surface water
is transported downstream. Substituting F3 into Equation 5-8 yields:
C,/W
A3Ktot3)
(0.80)
18.0 + (232,000)(5.13xlO'6)
4.2xlO'2 (sec//)
This concentration must be resolved into dissolved and suspended fractions, using the value
of fd calculated above from Equation 5-14:
C/w — f ir* /\u \
en/ "n - 'H v^v "n^
(4.2xlO'2 sec//)(0.9976)
4.2xlO"2 sec//
This result can be interpreted to mean that each mg/sec of contaminant discharged from
the aquifer to the Contoocook river is expected to increase the dissolved concentration of
benzene at the location of the MEI by 4.2xlO"2 mg//.
Calculations are simpler for lead, since Ktot is assumed to equal zero. From
Equations 5-5 and 5-6 it follows that 7=! and T=l for all cells in the cascade. Then from
Equation 5-8,
AnKtotn)
5-48
-------
(1.0)
[18.0 + (232,000)(0)]
= 5.6xlO"2 (sec//)
The dissolved component of this concentration is then estimated from Equation 5-14:
Cd3/wo " fd (C3/W0) = (0.99)(5.6xlO-2)
= 5.5xlO"2 (sec//)
5.6.1.1.4. Deriving Criteria -- With Equations 5-9 and 5-10, the above results
can be used to derive the maximum allowable dry-weight concentrations of contaminant in
sludge allowed to accumulate in the Antrim facility. WQ/CN for benzene was estimated for
the Antrim facility in Section 4.7.1.1. It was estimated that for every mg/kg dry-weight
of benzene in the sludge accumulating in lagoons at that site, approximately l.OxlO"3
mg/sec is expected to be discharged to the Contoocook river from the aquifer beneath the
site, once the aquifer reaches steady state concentrations at the point of discharge. From
Equation 5-9,
SRRSU = C3/N = (C3/W0)(W0/N)
= (4.2xlO'2 sec//)(1.0xlO'3 kg/sec)
= 4.2xlO'5 kg//
From Equation 5-10, SRRSW can be applied to a RWC of 1.2xlO~3 for benzene to derive
the maximum allowable concentration of sludge to be allowed in the impoundment:
= RWC/SRRSW
= (1.2xl(T3 mg//)/(4.2xlO'5 kg//)
= 28 mg/kg
-------
Similarly, WQ/N was calculated to be 7.3xlO"3 kg// for lead in the Antrim lagoons. It
follows that for lead,
SRRSW = C3/N = (C3/W0)(WQ/N)
= (5.5xlO"2)(7.3xlO~3 kg/sec)
= 4.0xlO"4 kg//
and
Nn,ax = RWC/SRRSW
= (S.OxlO'3 mg//)/(4.0x!0'4 kg//)
= 12 mg/kg
If dry-weight concentrations in the impoundment are greater than the above values then
long-term storage should not be allowed in the facility.
For the exposure pathway involving human consumption of fish, the RWC for
benzene can be calculated with methods described in Section 5.5.3:
RWC = (RL BW)/(q1* RE BCF If)
For an arbitrary choice of RL of 10'6 per lifetime, the RWC will be:
RWC = (10'6)(70) / [(0.029)(1.0)(5.2)(0.053)] = 8.8xlO'3 mg//
Since fish could be caught anywhere in the river near the Antrim site, estimated dissolved
concentrations in Cell 1 are used for the calculations. From Equation 5-8 and Table 5-
7,
5-50
-------
(D
[12.6 + (515,000)(5.2xlO'6)]
6.6xlO'2 (sec//)
For comparison with RWC based on fish consumption, this concentration is not resolved
into dissolved and suspended fractions. From Equation 5-9:
SRRSW = C,/N = (cy
= (6.6xl(T3 sec//)(1.0xlO"3 kg/sec)
= 6.6xlO'5 kg//
From Equation 5-10:
Nmav = RWC/SRR-U
R13X dw
= (8.8xlO'3 mg//)/(6.6x!0'5 kg//)
f = 130 mg/kg
For this particular site, the fish pathway of potential human exposure results in less
stringent criteria for benzene than does the drinking water pathway. No bioconcentration
factor is available for lead, so potential exposure through the fish consumption pathway is
not considered.
5.6.1.2. SAMPLE CALCULATION: TULSA, OKLAHOMA -- The Northside
Wastewater Treatment Plant is located at the confluence of Bird Creek and Mingo Creek
in northeast Tulsa. Of five lagoons at the site, three are used for storage of sludge. The
selected study reach extends for 20 km from the treatment plant to the City of Catoosa,
Oklahoma. Catoosa was selected as the location of the MEI because it evidently is the
closest community downstream from Tulsa's Northside Treatment Plant. After review of
5-51
-------
topographical and cross-sectional data for Bird Creek, the 20-km study corridor was
modeled as the five-cell cascade described below:
• Cell 1 extends from the treatment plant boundary north along Mingo Creek
for about 0.5 km to its confluence with Bird Creek;
• Celt 2 runs for 1.6 km along Bird Creek to the Elk Creek confluence;
• Cells 3, 4 and 5 cover about 18 kilometers on Bird Creek beginning at the
Elm Creek confluence and ending at the Verdigras River confluence. The
need for three cells along this reach was determined on the basis of significant
changes on the hydraulic geometry of Bird Creek.
Mean annual flow values for these five cells were estimated from a regression of
discharge on drainage area developed from USGS stream gages located in the Verdigras
River watershed. Suspended sediment concentrations were obtained from field data reported
by the U.S. Geological Survey, the U.S. Army Corps of Engineers and the U.S. Soil
Conservation Service. Hydraulic geometry parameters were taken from HEC-2 input files
prepared by the U.S. Army Corps of Engineers for use in Bird Creek flood studies. Input
parameters for this site are summarized in Table 5-9, and represented schematically in
Figure 5-6. The resulting estimates of contaminant concentrations and allowable mass
loadings for benzene and lead are presented in Table 5-10 and 5-11. Methods used to
compute the entries in those tables are identical to those used in the study of the
Contoocook River, New Hampshire, as explained in detail in Section 5.6.1.1.
Results of those calculations show that benzene in sludge stored in the Tulsa facilities
is not expected to cause adverse effects through the surface water pathways, but that
criteria for lead (11 mg/kg) could be restrictive.
5.6.1.3. SAMPLE CALCULATION: PORTLAND, OREGON -- The City of
Portland's municipal wastewater treatment plant and sludge disposal lagoon are located next
to the Columbia Slough, which runs into the Willamette River approximately 7 kilometers
downstream from the plant. The Columbia Slough, which bounds the treatment plant on
the north, is considered a slackwater channel since significant flows are produced only as
a result of storms. The confluence of the Columbia Slough with the Willamette River is
approximately 1200 m upstream of the Columbia River. St. Helens, Oregon was chosen as
the site for the MEI since it is the closest community downstream of the site. St. Helens
5-52
-------
TABLE 5-9
Site-Specific Input Parameters
Bird Creek, Oklahoma
Parameter
L
B
H
A
V
Qf
id
us
n
S
Lm
U10
T
rsl
foc
Cell 1
550
6
0.35
3,300
1,155
0.28
0.05
0.14
0.075
0.0004
123
4.9
25
4xlO'4
0.1
Cell 2
1,585
18
2.1
28,530
59,910
16.7
0.93
0.44
0.075
0.0004
--
4.9
25
4xlO'4
0.1
Cell 3
6,430
22
1.9
141,460
268,770
17.0
0.77
0.41
0.075
0.0004
--
4.9
25
4xlO'4
0.1
Cell 4
8,580
32
2.4
274,560
658,940
17.5
0.55
0.23
0.080
0.0001
—
4.9
25
4xlO"4
0.1
Cell 5
3,220
22
1.3
70,840
92,092
17.5
0.80
0.62
0.065
0.0012
--
4.9
25
4xlO'4
0.1
(Units)
(m)
(m)
(m)
(m2)
(m3)
(m3/sec)
(m2/sec)
(m/sec)
(unitless)
(unitless)
(m)
(m/sec)
(°C)
(kg//)
(mg//)
5-53
-------
+*
3
4
FIGURE 5-6
Five Cell Cascade for Bird Creek Near Tulsa, OK
-------
TABLE 5-10
Model Results for Benzene
Bird Creek, Oklahoma
Parameter
kd
fd
kl
kg
k
kagg
Ktot
Qf
1
r
C/W0
cd/w0
WQ/N
SRRSW
Nn,ax
Cell 1
27
0.9894
6.9xlO'5
2.7xlO'3
6.2xlO'5
1.7xlO~6
6.2xlO'5
0.28
0.58
1.0
2.1
2.0
6.5xlO"9
1.3xl(T8
9000
Cell 2
27
0.9894
6.9xlO'5
2.4xlO'3
6.1xlO'5
1.7xlO"6
6.4xlO"5
16.7
0.90
0.58
3.1xlO'2
S.lxlO"2
6.5xlO"9
2.0xlO'10
> IxlO6
Cell 3
27
0.9894
6.9xlO"5
2.2xlO'3
6.1xlO'5
1.7xlO'6
6.3xlO"5
17.0
0.66
0.52
2.0xlO'2
2.0xlO"2
6.5xlO'9
1.3xlO'10
> IxlO6
Cell 4
27
0.9894
6.9xlO"5
2.2xlO"3
6.0xlO"5
1.7xlO"6
6.4xlO'5
17.5
0.50
0.34
9.8xlO'3
9.7xlO'3
6.5xlO'9
6.3xlO'11
> IxlO6
Cell 5
26.9
0.9894
6.89xlO'5
2.3xlO'3
6.1xlO'5
1.70xlO'6
6.2xlO'5
17.5
0.80
0.17
7.8xlO'3
7.7xlO"3
6.5xlO'9
5.0xlO"11
> IxlO6
(Units)
(//kg)
(unitless)
(m/sec)
(m/sec)
(m/sec)
(sec'1)
(m/sec)
(m3/sec)
(unitless)
(unitless)
(sec//)
(sec//)
(kg/sec)
(kg//)
(mg/kg)
5-55
-------
TABLE 5-11
Model Results for Lead
Bird Creek, Oklahoma
Parameter
kd
fd
kl
kg
k
kagg
Ktot
1
r
C/W0
cd/w0
w0/N
SRRSW
N
Cell 1
234
0.91
0.0
0.0
0.0
0.0
0.0
1.00
1.00
3.6
3.3
8.3xlO'3
2.7xlO'2
0.18
Cell 2
234
0.91
0.0
0.0
0.0
0.0
0.0
1.00
1.00
0.060
0.055
8.3xlO"3
4.6xlO'4
11
Cell 3
234
0.91
0.0
0.0
0.0
0.0
0.0
1.00
1.00
0.059
0.054
8.3xlO'3
4.5xlO'4
11
Cell 4
234
0.91
0.0
0.0
0.0
0.0
0.0
1.00
1.00
0.057
0.052
8.3xlO'3
4.4xlO"4
11
Cell 5
234
0.91
0.0
0.0
0.0
0.0
0.0
1.00
1.00
0.057
0.052
8.3xlO'3
4.4xlO"5
11
(Units)
(//kg)
(unitless)
(m/sec)
(m/sec)
(m/sec)
(sec'1)
(m/sec)
(unitless)
(unitless)
(sec//)
(sec//)
(kg/sec)
(kg//)
(mg/kg)
5-56
-------
is located approximately 24.5 km downstream of from the confluence of the Willamette and
the Columbia rivers.
Two unique factors influenced the formation of the conceptual model: (1) the fact
that the Columbia Slough is a slackwater channel which receives the majority of inflow
from agricultural drainage and (2) the fact that the Columbia River is influenced by the
tidal action of the Pacific Ocean. These factors were considered in specifying the
following 5-cell model, which is depicted in
Figure 5-7:
• Cell 1 represents the Columbia Slough and is assumed to have an extremely
low flow rate with a high detention time.
• Cell 2 covers the 1200 m section of the Willamette River.
• Cells 3, 4 and 5 represent different characteristics of the Columbia River.
Three segments are required for adequate representation due to the significant
changes in channel geometry and tributary inflow of the Lewis River at
approximately 1.6 km above the location of the MEI.
Most hydraulic data were taken from HEC-2 input files prepared by the U.S. Army
Corp of Engineers. The mean annual flow values were estimates established by the U.S.
Geological Survey. Site-specific input parameters are summarized in Table 5-12; the
resulting estimates of contaminant concentrations and allowable mass loadings for benzene
and lead are presented in Tables 5-13 and 5-14, respectively. As can be seen from these
tables, if Cell 5 is taken as the point of compliance, then dry-weight concentrations of
benzene in sludge will be restricted to 1000 mg/kg, and concentrations of lead will be
restricted to 160 mg/kg.
Human exposure through the drinking water pathway from the Portland site is
limited by significant dilution of benzene within the Williamette and Columbia rivers prior
to ingestion by humans. For the fish ingestion pathway, however, the first modeled cell
of the Columbia Slough is taken as the point of compliance. As can be seen from Table
5-13, the source-receptor ratio for benzene in Cell #1 is 3.7xlO"3. This value, derived for
the drinking water pathway, does not include contaminant adsorbed to suspended sediment;
for the fish ingestion pathway, SRRSW is approximately 3.8xlO"3. Then Nmax for this
5-57
-------
p
FIGURE 5-7
Five Cell Cascade for Columbia Slough Near Portland, OR
5-58
-------
TABLE 5-12
Site-Specific Input Parameters
Columbia Slough, Oregon
Parameter
L
B
H
A
V
Qf
Od
Us
n
S
Lm
U,o
T
rsl
foc
Cell 1
6,980
55
1.1
383,900
575,850
2.3
0.04
0.36
0.030
IxlO"6
8,445
3.6
25
IxlO"4
0.2
Cell 2
1,220
445
11.3
542,900
7,546,000
838
1.88
0.17
0.030
IxlO'6
--
3.6
25
IxlO'4
0.2
Cell 3
9,980
825
10.1
8,232,500
88,100,000
5,660
6.86
0.68
0.024
1.2xlO'5
--
3.6
25
IxlO'4
0.2
Cell 4
12,875
830
10.2
10,686,250
111,642,500
5,660
6.82
0.67
0.029
1.7xlO'5
--
3.6
25
IxlO'4
0.2
Cell 5
1,610
955
10.3
1,537,550
16,020,000
5,776
6.05
0.59
0.030
1.4xlO'5
--
3.6
25
IxlO'4
0.2
(Units)
(m)
(m)
(m)
(m2)
(m3)
(m3/sec)
(m2/sec)
(m/sec)
(unitless)
(unitless)
(m)
(m/sec)
(°C)
(kg//)
(unitless)
5-59
-------
TABLE 5-13
Model Results for Benzene
Columbia Slough, Oregon
Parameter
kd
fd
kt
kg
k
k
Ktot
7
r
C/W0
Cd/W0
W0/N
SRRSU
Nmax
Cell 1
54
0.995
3.7xlO"5
1.7xlO"3
3.4xlO"5
1.70xlO'6
3.6xlO'5
0.14
1.00
6.3xlO'2
6.2xlO'2
0.059
3.7xlO"3
0.33
Cell 2
54
0.995
3.7xlO'5
1.6xlO'3
3.4xlO'5
1.70xlO'6
5.3xlO'5
0.97
0.14
1.7xlO"4
1.7xlO'4
0.059
9.8xlO'6
120
Cell 3
54
0.995
3.7xlO'5
1.4xlO'3
3.3xlO"5
1.70xlO"6
5.0xlO"5
0.93
0.14
2.3xlO'5
2.3xlO"5
0.059
1.4xlO"6
890
Cell 4
54
0.995
3.7xlO'5
1.4xlO"3
3.3xlO'5
1.70xlO'6
5.0xlO'5
0.91
0.13
2.1xlO"5
2.1xlO'5
0.059
1.2xlO'6
970
Cell 5
54
0.995
3.7xlO'5
1.5xlO"3
3.4xlO"5
1.70xlO'6
5.1xlO"5
0.99
0.12
2.0xlO'5
2.0xlO"5
0.059
1.2xlO'6
1000
(Units)
(//kg)
(unitless)
(m/sec)
(m/sec)
(m/sec)
(sec'1)
(m/sec)
(unitless)
(unitless)
(sec//)
(sec//)
(kg/sec)
(kg//)
(mg/kg)
5-60
-------
TABLE 5-14
Model Results for Lead
Columbia Slough, Oregon
Parameter
kd
fd
kl
kg
k
kagg
Ktot
1
r
C/W0
Cd/W0
w0/N
SRRSU
N
Cell 1
234
0.98
0.00
0.00
0.00
0.00
0.00
1.00
1.00
4.3xlO'1
4.2xlO'1
0.19
0.08
0.06
Cell 2
234
0.98
0.00
0.00
0.00
0.00
0.00
1.00
1.00
1.2xlO"3
1.2xlO'3
0.19
2.2xlO'4
23
Cell 3
234
0.98
0.00
0.00
0.00
0.00
0.00
1.00
1.00
l.SxlO'4
1.7xlO'4
0.19
3.3xlO'5
154
Cell 4
234
0.98
0.00
0.00
0.00
0.00
0.00
1.00
1.00
l.SxlO'4
1.7xlO'4
0.19
3.3xlO'5
154
Cell 5
234
0.98
0.00
0.00
0.00
0.00
0.00
1.00
1.00
1.7xlO'4
1.7xlO'4
0.19
3.2xlO'5
157
(Units)
(//kg)
(unitless)
(m/sec)
(m/sec)
(m/sec)
(sec'1)
(m/sec)
(unitless)
(unitless)
(sec//)
(sec//)
(kg/sec)
(kg//)
(mg/kg)
5-61
-------
pathway can be calculated from Equation 5-10 for RWC = 8.8xlO"3 (as derived in Section
5.6.1.1):
Nmax = RWC/SRRSU = (8.8xl(T3 mg//) / (3.8xlO'3 kg//) = 2.3 mg/kg
Because this value is based on expected concentrations in Cell #1 of the system, it is
significantly more restrictive than criteria based on the drinking water pathway.
5.6.2. Analysis of Exposure for the Most Exposed Populations. Tier 2 calculations
provide for the MEI an upper-bound estimate of maximum concentrations likely to be
encountered at a specified receptor location downstream from the point of groundwater
discharge into the stream. They also provide a derivation of maximum allowable sludge
concentrations, based on reference water concentrations, site characteristics and a specified
receptor distance.
The distribution of potential exposure over the entire exposed population may also
be of interest. For the sample calculations presented here, the most exposed population
(MEP) to be considered will be those individuals who regularly fish in the surface water
body at the point of contaminated groundwater discharge.
5.6.2.1. MEP ANALYSIS: ANTRIM, NEW HAMPSHIRE -- To perform an
analysis of the most exposed population (MEP) for the fish ingestion pathway, two pieces
of information are needed: a distribution of fish consumption in the exposed population,
and the number of persons exposed to fish caught in the contaminated water. As discussed
in Section 5.5.2.2., U.S. EPA (1980a) reports data for fish consumption in the United States
by a number of demographic variables. These data were summarized in Table 5-4, and
included consumption of fish from all sources (freshwater and marine). This table provided
the mean and upper 95th percentile of fish consumption for New England (for the New
Hampshire site analysis), but were based on data analyzed in 1980. From the table, it can
be seen that average fish consumption for the United States was estimated to be about 14.4
g/day. As discussed in Section 5.5.2.2, more recent data (adjusted for recreational catch)
5-62
-------
suggest an average consumption rate of 21.3 grams, about 48% higher than values reported
in Table 5-4.
If it is assumed that the distribution of fish consumption within each population is
approximately log-normal, the values in Table 5-4 can be used to derive a geometric mean
and geometric standard deviation of fish consumption for each region in the U.S. For
New Hampshire, the appropriate distribution is described for the New England population,
which is reported to have a mean consumption rate of 16.3 g/day, and a 95th percentile
value of 46.5 g/day. These values suggest a geometric mean of 13.3 g/day and geometric
standard deviation of 1.89. If that mean is adjusted by 46% to correspond to the 1985
average value from Table 5-3, and if the geometric standard deviation is held constant,
then the revised distribution has a geometric mean of 19.7 and arithmetic mean of 24.1.
It will be assumed that 50% of the fish ingested is from freshwater sources, which yields
an estimated geometric mean of 9.8 g/day or arithmetic mean of 12.1 g/day of freshwater
fish consumed by the average resident of New England.
Next, the size of the exposed population must be determined. It will be assumed
that fishing behavior is uniformly distributed (by flow) over the freshwater resources in
New Hampshire. Estimated flow in the Contoocook river near the Antrim site is about 12.6
m3/sec, or about 2% of the total freshwater budget of New Hampshire (about 630 m3/sec).
If freshwater fish consumed within these waters are consumed by 2% of the New
Hampshire population, then about (978,000)(0.02)=20,000 persons are potentially exposed.
The estimated bioconcentration factor for benzene in fish is about 5.2 //kg. If this
estimate is combined with the source receptor ratio SRRSU = 6.8xlO"5 kg// estimated for
Cell #1 of the Antrim site, then a distribution can be derived for the ratio of exposure (in
mg/kg-day) to dry-weight concentration of benzene in sludge stored by the facility (in
mg/kg). Combining this distribution with the estimated size of the exposed population
yields the results presented in Figure 5-8. As can be seen from the figure, the estimated
ratio of potential exposure (in mg/kg-day) to sludge concentration (in mg/kg dry-weight)
is greater than 0.15 (day"1) for about 1000 persons. For about 100 persons, the ratio is
5-63
-------
LoglO of Population with Higher Ratio
O
CD
it)
13
0)
n
T3
O
c:
CL
O
O
3
O
CL
O
O
O
to
O
O
O
CO
0)
Q
>
Q
O
m
CD Q
^ CO
<: C
m -^
T) ^
-, CD
O
O
D
O
en
a
I
co
Q_
D
CD
-------
greater than 2.5. According to the figure, the ratio of exposure to sludge concentration
for the MEI would be about 0.6 day"1.
5.6.2.2. SAMPLE CALCULATION: TULSA, OKLAHOMA -- Calculations for the
Tulsa site are similar to those just described for Antrim. Table 5-7 showed that residents
of the West South Central region consumed 14.4 g/day, with an upper 95th percentile of
43.6 g/day. If the distribution is log-normal, then these estimates suggest a geometric mean
of 11.0 and geometric standard deviation of 2.1. This mean is adjusted by 1.48 to
correspond to estimates based on 1985 data, and is multiplied by 0.5 to include only
freshwater fish. The resulting geometric mean is (11.0)(1.48)(0.5) = 8.1 g/day,
corresponding to an arithmetic mean of 12.1.
The estimated rate of flow for the river system considered for Tulsa was about 0.28
m3/sec. This flow represents about 0.03% of the total freshwater budget of Oklahoma
(about 900 m3/sec), so it is assumed that about 0.03% of the total population of Oklahoma
(or about 1000 persons) consumes fish caught from the study area. This estimate is
combined with the estimated distribution of fish consumption described above, with a
bioconcentration factor of 5.2, and with the estimated SRRSH for Cell #1 of this site of
6.5xlO"9 kg// to yield the distribution graphed in Figure 5-9. As can be seen from the
figure, the ratios of predicted exposure to sludge concentration of benzene are much lower
for Tulsa than were estimated for Antrim. Of the 1,000 persons exposed, about 100 can
expect exposure to more than 2.5xlO"5 mg/kg-day of benzene for each mg/kg of benzene
in sludge stored at the site. The ratio could reach as high as 8xlO"5 for the MEI. As
discussed in Section 5.6.1.2, this low exposure is explained by the slow water flow (and
greater opportunity for contaminant decay) within the aquifer beneath the Tulsa site.
5.6.2.3. SAMPLE CALCULATION: PORTLAND, OREGON -- The same calculations
can be repeated for the Portland, Oregon site to derive an estimated distribution of
exposure to Oregon residents as a function of the concentration of benzene in sludge
deposited in the impoundment near the Columbia Boulevard plant. For the Pacific region,
Table 5-4 reported that the average rate of consumption of fish was 14.2 g/day, with a
95th percentile value of 39.2 g/day. If the distribution is log-normal, then it can be
5-65
-------
99-5
LoglO of Population with Higher Ratio
o
CO
n>
n
N
m
X
~o
o
w
c
— CD
ciw
03 _»
CD O
n
Q.
O
v;
00
p
In
O
O
cn
Cn
NO
CJi
O
cn
CD
m
T3
CD
Q
Q
o'
O
x
C
~i
(D
S"
00
o C
^ a.
o
o
D
O
73
m
vn
-------
described by a geometric mean of 11.8 and geometric standard deviation of 1.9 g/day.
These values are adjusted as described above to yield a revised geometric mean of
(11.8)(1.48)(0.5) = 8.7 g/day, corresponding to an arithmetic mean of 10.5 g/day.
The flow of the Columbia Slough is estimated to be about 2.3 m3/sec, or about 0.06%
of the estimated 4000 m3/sec total freshwater budget for Oregon. If 0.06% of the total
population of Oregon is potentially exposed, then the size of the exposed population is
about 1600. This estimate is combined with the distribution derived above, and with a
source-receptor ratio of SRRSW = 3.7xlO"3 for Cell #1 in the Columbia Slough to yield the
distribution described by Figure 5-10. As can be seen from the figure, about 100 persons
are exposed to a ratio higher than 6, and the MET is exposed to about 17 mg/kg-day for
every mg/kg of dry-weight concentration of benzene in sludge received by the
impoundment. Because for the relatively high source-receptor ratio for Cell #\ in this
system, estimated exposures through the fish consumption pathway are considerably higher
for the Portland site than for Antrim or Tulsa.
5-67
-------
89-S
LoglO of Population with Higher Ratio
CD
(Is
3
N
CD
rs
m
X
~D
O
(ft
C
n
n>
if}
c
o_
UD
fD
O
O
3
O
CL
Q
o
NO
4^
cn
00
O
CO
NO
O
NO
NO
O
In
O
O
In
Cn
NO
CJi
C*J
OJ
In
Q
O
CD
m
to X
S'TJ
•0
0)
T)
0
g
o
O
C
0)
c/2
c
O
O
D
O
o
-------
6. DERIVATION OF CRITERIA FOR THE AIR PATHWAY
6.1. OVERVIEW OF THE METHOD
This chapter describes a methodology for deriving criteria for the air pathway of exposure
from surface disposal facilities. The approach is based on the assessment of risks to human health
and is suitable for both site-specific and national application. As with the exposure pathways
discussed in Chapters 4 and 5, the methodology for deriving criteria based on volatilization involves
a two-tiered approach; both tiers seek to derive criteria sufficiently protective for the most exposed
individual (MEI). The first tier involves the use of a mathematical model to describe the migration
of sludge contaminants from liquid within a lagoon to air above the lagoon surface, and the transport
of contaminated air to downwind receptor locations. Based on this model, Tier 1 calculations prepare
national, numerical sludge criteria based on a reasonable worst case scenario. The scenario is chosen
such that conditions at actual facilities are unlikely to result in higher potential for human or non-
human exposure than conditions assumed for the derivation of Tier 1 criteria. Tier 2 calculations
derive limits to contaminant concentrations in sludge based on the same mathematical model but with
site-specific input values. The property boundary may be selected as the point of compliance, since
residential dwellings constructed near the property boundary could be affected by contaminated air,
with subsequent public health implications. Alternatively, the nearest actual downwind residential
location could be analyzed.
Both tiers seek to limit concentrations of volatile contaminants in sludge so that they will not
lead to air concentrations that exceed health-based limits. The first derives maximum allowable
contaminant concentrations in ambient air, based on health effects data for each contaminant under
consideration. The second determines expected rates of volatile emissions from the lagoon, as a
function of contaminant concentrations in sludge received by or accumulating in the impoundment.
This step uses a two-layer resistance model to calculate volatilization as a function of expected
contaminant concentrations within the impoundment; these concentrations are determined with a mass
balance approach. The third step predicts expected average contaminant concentrations in ambient
air at the specified location of the MEI, as a function of expected volatile emissions. Calculations
are based on a mathematical model of contaminant dispersion during wind transport. The fourth step
6-]
-------
combines results from the other three, to derive maximum allowable concentrations in sludge. Each
of these steps will be discussed in greater detail in Sections 6.3 and 6.5.
Chapters 4 and 5 described methods for deriving criteria related to potential exposure through
groundwater and surface water pathways; for the calculations described in those chapters, no effort
was made to discriminate between different kinds of active impoundments. For estimating potential
exposure to contaminants volatilizing from surface disposal facilities, however, it is useful to classify
lagoons into three categories: (1) well-mixed lagoons with continuous inflow, (2) poorly mixed
lagoons with continuous inflow, and (3) impoundments receiving occasional or periodic deposits of
sludge. As will be discussed below, differences in conditions at these types of lagoons suggest
different choices of mathematical models for estimating volatile emissions.
6.2. ASSUMPTIONS
Table 6-1 summarizes some key assumptions used in the methodology described below. These
assumptions can be grouped into two categories: those involved in the estimation of emissions from
surface disposal facilities, and those involved in the estimation of the transport of those emitted
contaminants by wind. Volatilization is predicted with a mass balance approach in which
biodegradation, volatilization, settling to the sludge layer, seepage, and removal through effluent are
the only significant loss processes for contaminant. It is assumed that adsorbed contaminant within
the sludge layer does not contribute to volatilization, but that adsorbed contaminant on suspended
solids maintains an equilibrium with the dissolved concentration in the liquid layer, and acts as a
source of contaminant as contaminant volatilizes from dissolved phase. That liquid layer is assumed
to cover the entire facility, and to maintain a constant volume throughout the lifetime of the lagoon.
It is assumed that biological degradation within the liquid layer is first order. Estimation of the
extent to which concentrations of contaminant are attenuated during wind transport to a receptor
location is based on a Gaussian plume dispersion model that assumes meteorologic conditions are
constant between the emissions source and the receptor. It is further assumed that atmospheric losses
of the contaminant can be described as a first order process. These and other assumptions will be
discussed in more detail throughout the remainder of this chapter.
6-2
-------
TABLE 6-1
Assumptions for Methodology to Analyze the Air Pathway
Functional Area
Assumption
Ramifications
Emissions from Impoundment
Biodegradation, volatilization, seepage, and
removal in effluent are the only significant
loss processes for contaminant.
Plug flow scenario assumes one-dimensional
flow of water and contaminant.
Equilibrium between dissolved and adsorbed
concentrations of contaminant in suspended
solids within the impoundment.
Adsorbed contaminant in sludge layer does
not contribute to volatilization.
Underpredicts emissions if photolysis,
hydrolysis, or other loss processes are
significant.
Unknown.
Underpredicts emissions to the extent that
bottom layer acts as a contaminant reservoir
for release through volatilization.
Underpredicts emissions.
i
OJ
Rate of volatile emissions of emissions from
impoundment can be described by two-phase
resistance model.
Average temperature of 25" is assumed for
all impoundments.
Entire surface of impoundment is assumed to
be in liquid phase.
Concentration of suspended solids is uniform
within the liquid layer and throughout the
active lifetime of the facility.
Use of diffused air in a wastewater
treatment impoundment does not affect solids
content in surface layer of liquid.
For facilities with continuous flow, inflow
of liquid is assumed to equal outflow
(including seepage).
Unknown.
Unknown.
Unknown in circumstances where accumulating
bottom layer of sludge solids reaches the
impoundment surface. Overpredicts emissions
from frozen impoundments.
Unknown.
Underpredicts emissions if diffused air
increases concentrations of suspended solids
near the impoundment surface.
Unknown.
-------
TABLE 6-1 (cont.)
Functional Area
Assumpti on
Ramifications
Atmospheric Transport
i
.e-
Biological degradation is first order.
The rate of contaminant loss through
effluent and seepage can be approximated by
QC^ for well mixed facilities, and by QC^(L)
for poorly mixed facilities.
Atmospheric conditions are constant between
the site and the receptor location.
Receptor is at ground level.
Emission rates for a particular site are
constant for each contaminant.
Contaminant plume follows a Gaussian
distribution.
Atmospheric losses of contaminant are first-
order.
Will overpredict degradation (and
underpredict emissions) at high
concentrations of contaminant.
Unknown.
Unknown.
Unknown.
May underpredict acute exposure.
Unknown.
Unknown.
-------
6.3. CALCULATIONS
Calculations for Tier 1 and Tier 2 are based on the assumption of mass balance for
contaminants entering the lagoon. The methodology assumes that the quantity of each contaminant
entering the lagoon in a specified time interval is equal to the sum of the quantity removed or
destroyed, plus the quantity retained:
M, = M0 +AMT (6-1)
where:
Mj = mass of contaminant entering the lagoon (g/sec)
MQ = mass of contaminant degrading or leaving the lagoon (g/sec)
AMT = change in total contaminant mass stored within the lagoon (g/sec)
Contaminants enter the impoundment through periodic or occasional deposits of sludge, or
through continuous inflow of sludge or wastewater. They leave through discharged effluent, seepage
beneath the lagoon in dissolved phase, biodegradation and volatilization. Other potential routes by
which contaminants might be eliminated include photodecomposition, hydrolysis, oxidation/reduction
and hydroxyl radical reactions. For surface impoundments and wastewater treatment lagoons,
however, these other pathways are assumed to be of either negligible or secondary importance (U.S.
EPA, 1987a). Equation 6-1 can thus be expanded to include those removal processes to be
considered:
M, = ME + MB + Mw + MVA + AMT (6-2)
where:
ME = the mass of contaminant removed with effluent, or with seepage from the
bottom of the impoundment (g/sec)
MB = the mass of contaminant removed by biodegradation (g/sec)
Myp = the mass of contaminant removed by volatilization that results from diffused
air (g/sec) and
MVA = the mass of contaminant removed by volatilization that results from wind
blowing across the impoundment (g/sec)
Each of the components of Equation 6-2 will now be discussed in more detail.
6-5
-------
6.3.1. Influent and Effluent Flow. The rate at which contaminants enter a sludge disposal
facility depends on the type of facility involved. The present methodology considers facilities in
which sludge is deposited continuously, and those in which sludge is deposited periodically or
occasionally. For impoundments for which inflow of sludge or wastewater is continuous, the mass
of contaminant entering the lagoon per unit of time is described by:
M, = Q C0
where:
Mj = mass of contaminant entering the lagoon (g/sec)
Q = the liquid volumetric flow rate (m3/sec)
CQ = the concentration of the contaminant (both dissolved and adsorbed) in liquid
at the inlet (mg// or g/m3)
Similarly, the rate of contaminant loss associated with liquid flow out of the system
is given by the relation:
ME = Q CE (6-3)
where:
ME = mass of contaminant removed from the lagoon through effluent or seepage
(g/sec)
Q = the volumetric flow of liquid leaving the impoundment as effluent or seepage
(m3/sec)
C_ = the concentration of the contaminant in effluent or seepage (g/m3)
For surface impoundment sites with continuous influent flow, liquid flow rates into and out
of the lagoon will be assumed to be equal (i.e., net precipitation will be ignored). This assumption
results in a constant-volume model for these systems. It should also be noted that contaminant
concentrations in effluent and seepage are here assumed to be identical. This assumption will be
further discussed below.
6-6
-------
For surface impoundments receiving periodic deposits of sludge, the mass of contaminant in
the facility decreases throughout the time interval between sludge deposits. Between deposits,
contaminant is lost to seepage, effluent, biodegradation, volatilization and other loss processes. For
some impoundments, sludge may be removed between deposits, or deposits may be sufficiently
infrequent that yearly average rates of emissions can be calculated based on time-averaged estimates
of emissions from a single deposit of sludge. For these facilities, the mass of contaminant received
in a single deposit can be described as:
MI
where Vd is the average volume of sludge received in a single deposit.
6.3.2. Contaminant Mass Lost to Biodegradation. The rate at which contaminant mass is lost
through biological degradation is approximated as a first-order process. For very high in-liquid
concentrations, the assumption of first-order biodegradation will tend to overestimate biodegradation
rates, and hence underestimate rates of contaminant volatilization. A first-order model, however,
can provide a reasonable approximation of biodegradation rates at low contaminant concentrations.
A more accurate relation valid for both high and low concentrations is described by Monod-type
kinetics (U.S. EPA, 1987a). The assumption of Monod-type kinetics, however, would result in a
nonlinear mass balance equation that would require numerical solution, and would require a more
complex model for volatile emissions than the one proposed here. For liquid concentrations of
contaminant lower than the Monod half-saturation constant, biodegradation can be approximated
with reasonable accuracy as a first-order process described by:
MB = kb v q
where:
MB = the loss rate of contaminant mass through biodegradation (g/sec)
kb = the first-order biodegradation rate constant (sec"1)
V = the volume of the liquid considered (m3)
Ct = the concentration of the contaminant in the liquid (g/m3)
6-7
-------
Biodegradation rates for organic contaminants are available from the literature, and can be adjusted
for different temperatures with Equation 5-27 from Chapter 5.
Biological activity might also generate mass of some toxic chemicals within the lagoon; for
lack of sufficient data, however, this analysis assumes that all potential exposure is to contaminants
contained in the sludge or wastewater as received by the impoundment.
6.3.3. Contaminant Mass Lost to Volatilization. The rate at which the contaminant volatilizes
to air is described by:
MVA = k A CL (6-4)
where:
MVA = the loss rate of contaminant mass through volatilization to ambient air (g/sec)
k = the overall mass transfer coefficient of the system (m/sec)
A = the area of the air-exposed surface of the lagoon (m2)
Ct = the concentration of the contaminant in the liquid (g/m )
As in Chapter 5, the overall mass transfer coefficient (k) in Equation 6-4 is calculated based on a
two-layer resistance model:
1 = 1 + El (6-5)
k kt Hckg
where:
kt = the liquid phase mass transfer coefficient (m/sec)
R = the ideal gas constant = 8.21x10 (m-atm/mol-°K)
T = temperature (°K)
H = Henry's law constant for contaminant (m -atm/mol)
k = the gas phase mass transfer coefficient (m/sec)
Numerous methods for calculating kt and k have been proposed in the literature. As in
Chapter 5, this methodology follows U.S. EPA (1987a) and U.S. EPA (1989c) in selecting methods
for calculating these coefficients; the methods are described in Chapter 5 by Equations 5-17 through
5-25.
6-8
-------
6.3.4. Contaminant Mass Lost to Volatilization from Diffused Air. Volatilization via this
pathway is applicable only to wastewater treatment facilities with diffused air systems. The
contaminant mass loss rate from the system through diffused air is described by:
MVD • Qa (Hc/RT) Ci
where:
Myjj = the emission rate of contaminant in diffused air (g/sec)
Qa = the volumetric flow rate of diffused air through the system (m3/sec)
Hc = Henry's law constant for the contaminant (atm-m3/mol)
R = the ideal gas constant = 8.21 xlO"5 (atm-m3/mol-°K)
T = temperature of the system (°K)
C = the concentration of the contaminant in the liquid (g/m3)
Volatilization through diffused air contributes to total volatilization from the lagoon and must
be included in estimates of potential exposure through the air pathway. Because this type of
volatilization is of concern for wastewater treatment lagoons only, discussion of this pathway will
be deferred until Section 6.3.7.4, where wastewater treatment lagoons will be considered. Estimation
of volatilized emissions from diffused air will not be included in mass balance calculations for other
types of facilities.
6.3.5. Change in Total Contaminant Mass Contained Within the Lagoon. The final term (AMT)
in Equations 6-1 and 6-2 represents the increase in contaminant mass stored within the lagoon that
will result if the quantity of contaminant flowing into the impoundment exceeds the quantity
removed or degraded. In most active lagoons, a bottom layer of sludge solids accumulates with time.
Contaminant adsorbed to solids in this sludge layer represents one component of AMT; the other
component includes any increase over time in the mass of dissolved contaminant within the
impoundment.
As will be discussed in Section 6.3.7.1, lagoons with continuous inflow and outflow of sludge
or wastewater are modeled as if AMT=0. The model for those lagoons ignores the possible
accumulation of contaminant mass in the lagoon over time, and instead assumes that the steady-state
concentrations of dissolved contaminant, adsorbed contaminant and total suspended solids all remain
6-9
-------
constant throughout the active lifetime of the lagoon. This assumption is accomplished by redefining
Equation 6-2 to refer to a "control volume" within the lagoon, defined as the layer of liquid between
the surface of the lagoon and the top of an idealized sludge layer on the impoundment's floor. Terms
on the right side of Equation 6-2 then refer to contaminant removal from this control volume, and
the definition of ME is expanded to include the loss of adsorbed contaminant through settling of
solids to the sludge layer at the bottom of the lagoon. Implicit in this re-definition is the assumption
that adsorbed contaminant within the sludge layer does not contribute to concentrations in the upper
layer of liquids, and so does not contribute to volatilization.
As redefined, M£ will not necessarily equal QCt, as assumed in Equation 6-3. However, the
methodology presented below will assume that QCL provides a reasonable approximation of the total
quantity of contaminant leaving the impoundment through effluent, seepage, or settling. To the
extent that the increasing reserve of sludge mass and adsorbed contaminant increases dissolved
concentrations of contaminant above the impoundment floor (and therefore increases rates of
volatilization), the rates of volatile emissions predicted by this methodology may underestimate actual
emissions.
Similarly, AMT is assumed to equal zero for poorly mixed impoundments with continuous
inflow. These impoundments, which are discussed in Section 6.3.7.2, are represented by a "plug flow"
model, in which it is assumed that a differential flow element can be modeled without regard to
accumulation of contaminant or sludge solids in the impoundment. As with well-mixed systems, this
assumption is accomplished by assuming that the flow of liquid within the lagoon occurs within a
control volume above the accumulating layer of sludge solids, and that the mass of contaminant that
the model predicts will leave the far end of the lagoon per unit of time is comparable with the actual
total of contaminant mass lost through effluent and seepage, plus the mass of contaminant lost from
the control volume through settling. To the extent that this assumption is incorrect, the model may
under- or overpredict actual emissions.
6.3.6. Mass of Contaminant Transferred from Adsorbed to Dissolved Phase. Within the control
volume, contaminant will exist both in dissolved phase and adsorbed to suspended solids; both phases
must be considered for a meaningful estimate of volatile emissions. As volatilization removes
6-10
-------
dissolved contaminant from the impoundment, additional contaminant mass will be transferred from
the suspended solids to dissolved phase. This transfer will not affect the overall mass balance
described by Equation 6-2, but needs to be considered when estimating rates of volatilization.
For each type of facility discussed below, two separate mass balance equations are derived:
one for the mass of a contaminant entering and leaving the liquid phase of the control volume and
the other for the mass of a contaminant entering and leaving the (suspended) solid phase. In addition
to various gain and loss terms, the equations contain a term describing the transfer of contaminant
mass (positive or negative) from the solid to the liquid compartment of the system. This mass
transfer will be denoted Msl and will be conserved in the system: the gain by the liquid phase of the
mass balance will equal the loss by the solid phase. As mentioned earlier, possible desorption of
contaminant from the sludge layer on the floor of an impoundment will be ignored.
In this analysis, it is assumed that the concentration of adsorbed contaminant is always in
equilibrium with the concentration of dissolved contaminant. For each contaminant, a single
solid/liquid equilibrium distribution constant describes the equilibrium relationship between these
concentrations. From Equation 4-1,
Ct/N =
where:
C, = concentration of the contaminant in liquid within the control volume of the
impoundment (mg// or g/m3)
N = dry weight concentration of contaminant in sludge contained within the control
volume (mg/kg)
kd = equilibrium partition coefficient for the contaminant (//kg)
rgl = ratio of the mass of solid to the volume of liquid (kg/1)
The partition coefficient kd describes the ratio between dissolved and adsorbed contaminant
concentrations, and is valid at low contaminant concentrations only, where concentrations do not
exceed solubility thresholds. For the equations that follow, it is useful to introduce an additional
variable ksl, that describes the ratio between adsorbed and dissolved contaminant mass:
ksl = Vsl
6-11
-------
where ksl describes the (unitless) ratio of adsorbed contaminant mass to dissolved contaminant mass.
Combining Equations 4-1 and 6-7 for CL=C0 yields:
C0/N = rst/(l + ksl) (6-8)
that describes the concentration of contaminant in the liquid fraction of influent, as a function of
the dry-weight concentration of contaminant in inflow. As will be discussed in Sections 6.3.7.1 and
6.3.7.2, it will be assumed that both rgl and kgt are constant within the control volume throughout
the active lifetime of a sludge lagoon.
6.3.7. Estimation of Volatile Emissions. Estimation of rates of contaminant volatilization from
a surface disposal facility requires estimation of the overall mass transfer coefficient (k) in Equation
6-5, and simultaneous quantification of each of the other contaminant loss processes included in the
mass balance described by Equation 6-2. The exact form of the mass balance equation will depend
on the type of facility involved. Four types of facilities will be considered: (1) well-mixed
impoundments with continuous inflow of sludge or wastewater, (2) poorly mixed impoundments with
continuous inflow, (3) sludge impoundments receiving occasional or periodic deposits of sludge and
(4) well-mixed impoundments in which sludge accumulates for long-term storage as a product of
wastewater treatment. For each type of facility, a mass balance approach yields expected
contaminant concentrations in liquid contained within the impoundment. Once these concentrations
are determined, total volatilized emissions are easily estimated from Equation 6-4.
6.3.7.1. ESTIMATION OF EMISSIONS FROM WELL-MIXED SLUDGE
IMPOUNDMENTS WITH CONTINUOUS INFLOW -- In a well-mixed impoundment, contaminant
concentrations are assumed to be constant throughout the entire control volume, and an overall mass
balance for the entire facility is used to determine the rate of contaminant emissions. For the liquid
phase of contaminant mass,
Q CQ + MSL = k A C( + kfa V C{ + QCt (6-9)
6-12
-------
The first term on the left side of this equation represents the mass of dissolved contaminant entering
the control volume per unit of time. The second term represents the mass transferring from adsorbed
to liquid phase. On the right side of the equation, the first term represents volatilization from the
liquid, the second term represents losses to biodegradation and the final term represents losses
through effluent and seepage. For the solids mass balance,
ksi Q co = MSL + ksi kb v CL + ksi Q ci (6-l°)
where the term on the left side of the equation gives the mass of adsorbed contaminant flowing into
the system. This mass is equal to the contaminant mass in the liquid (Q C0) multiplied by a mass
partitioning factor (kgl), based on the assumption of local equilibrium between dissolved and
adsorbed contaminants within the lagoon. On the right side of the equation, the three terms indicate
the mass of contaminant transferring from adsorbed to dissolved phase, the solid contaminant
biodegraded and the solid contaminant lost to the system through effluent and settling.
Eliminating Mgl from these two equations and solving for the impoundment's concentration
yields:
Q c0 (i +k8l)
cl= (6_n)
+ksl) + kb V(l + ksl) + k A
Substituting into Equation 6-5 gives:
k AQCn(l+k.)
Esite ^-2^ ^ (6-12)
which can also be expressed as:
6-13
-------
k A Q(l+k .)
Esite/Co= ~
Equation 6-13 gives the estimated rate of volatile emissions (g/sec) as a function of inflow
characteristics, and of site- and chemical-specific parameters.
Implicit in Equation 6-10 is the assumption that ksl has the same value for influent to the
impoundment, for the impoundment's contents and for effluent and seepage from the impoundment.
Since ksl depends on the ratio of suspended solids to liquid (rgl), this assumption further implies that
effluent and seepage from the impoundment contain the same fraction of suspended solids as does
influent. For most impoundments, effluent will contain lower quantities of suspended solids than
influent, since sludge solids will settle and accumulate within sludge lagoons. QC^l+k,^) is therefore
likely to overestimate contaminant losses through effluent and seepage. As mentioned in Section
6.3.5, Equations 6-9 and 6-10 describe activity within a control volume that extends downward from
the surface of the impoundment to the top of the sludge layer on the lagoon floor. Conditions within
this control volume are assumed to be steady-state, so that liquid concentrations remain constant with
time. To the extent that actual contaminant losses from this control volume through effluent, seepage
and settling to the lagoon floor can be approximated by (l+k^QC^, Equation 6-13 may provide a
reasonable approximation of volatile emissions from the surface of the lagoon.
6.3.7.2. ESTIMATION OF EMISSIONS FROM POORLY MIXED IMPOUNDMENTS
WITH CONTINUOUS INFLOW -- Poorly mixed impoundments are better represented by a "plug
flow" model. For these, it is assumed that sludge enters the impoundment at known contaminant
concentrations, and flows with time toward the opposite end. Dissipative processes occur along the
facility's length and generate location-dependent concentrations. As the mixture flows along the
length of the facility, contaminant partitioning and losses via the various pathways described above
take place. For this type of impoundment, the mass balance relation is best written for a differential
volume element within the lagoon, and the contaminant concentration must be solved as a function
of distance from the inlet. The in-liquid contaminant mass balance for a differential volume element
in the facility can be described by:
-------
Q C, + dMSL = Q(C( + dC<) + k dA Ct + kb dV Ct (6-14)
where:
dA = w dx
and
dV = h w dx
and where:
dA = the surface area of a differential volume element (m2)
w = the width of the impoundment perpendicular to the direction of flow (m)
dx = the length of a differential volume element
dV = the volume of a differential volume element (m3)
h = the depth of a volume element (m)
The corresponding mass balance for solid-phase contaminant is described by:
ksl Q Ct = dMSL + ksl kb dV Ct + [ksl Q(CL + dCt)] (6-15)
where dMSL represents the movement of contaminant mass between solid and liquid phase within the
differential volume element (g/sec). Combining these equations and eliminating dMSL yields:
(l+ksl)Q dCt = -[k dA + kb dV (l+ksl)] Ct
Substituting for the differential elements and rearranging yields:
-[k w + khwh(l+k ,)]
dC./dx = fe ^- C,
This equation can be solved subject to the boundary condition that Ct(x) = C0, to yield concentration
as a function of downstream distance (x):
6-15
-------
Ct(x) = C0 exp
[-[kw + kbwh(l+k
Q(1+ksi>
,
si>
Substituting into Equation 6-4 yields the rate of volatile emissions as a function of distance:
dES)-te(x) = k w Ct(x) dx
To derive total emissions of contaminant for the entire facility, this function is integrated from the
inlet (x = 0) to the outlet (x = L), to obtain:
Esite " k A C0 [l-exp(-or)] / a (6-16)
or
E/C = k A C
where:
+ kbV(l+ksl)
a =
As in Section 6.3.7.1, the above equations implicitly assume that effluent and seepage have the same
suspended solids content as inflow (i.e., that rgl and kgl are constant throughout the impoundment).
To the extent that Q Ct(L) provides a reasonable approximation of total contaminant losses from the
control volume through effluent, seepage and settling to the lagoon floor, Equation 6-17 may provide
a reasonable estimate of volatilization.
6.3.7.3. ESTIMATION OF EMISSIONS FROM SLUDGE IMPOUNDMENTS RECEIVING
OCCASIONAL OR PERIODIC DEPOSITS OF SLUDGE --At some facilities, sludge does not enter
the impoundment continuously, but rather in discrete deposits. Between deposits of sludge,
6-16
-------
contaminant mass is degraded or released from the impoundment and concentrations decrease.
Contaminant concentrations in the impoundment are therefore a function of elapsed time between
sludge deposits, and the mass balance of contaminant within the impoundment is most conveniently
described for a discrete time interval between deposits of sludge. If the deposits are sufficiently
frequent that emissions from previous deposits are still significant at the time of subsequent deposits,
the system is best modeled with methods described in Sections 6.3.7.1 and 6.3.7.2. For those lagoons
receiving less frequent deposits of sludge, or those in which sludge is periodically removed, the
change of contaminant mass in liquid phase within the volume of sludge deposited can be described
as a function of time:
- d(ctvdep)
where:
Vd = the volume of sludge deposited (m3)
t = time elapsed since the sludge was deposited (sec)
For contaminant mass in adsorbed phase,
MSL + Kb Vdep Ksl
If it is assumed that the volume of the liquids in the system remains approximately constant with
time, these two relations can be combined to yield
(1+k .) V. ndC.
sl dep l = - [k A Ct + kb Vdep(l+ksl)Ct]
dt
that can be rearranged to yield an equation for the change in concentration per increment of time:
6-17
-------
dC i = -Kk A)/Vdep + kb(l+ksl)]
dt l+k
gl
This equation can be solved subject to Ct(t=0) = CQ to yield
c,(t),C0exp
L
or
Ct(t) = C0
where
/9 = [(k A/Vdep) + kb(l+ksl)]
It follows by substitution into Equation 6-4 that from time 0 to time T, average emissions can be
described by:
1
T C0kA [l-exp(-/3T)]
kAC,(t)dt= -S - ^JL^
t=o err
or
kA[l-exp(-^T)]
Several limitations in the above methodology should be noted. First, it was assumed that the
volume of liquid contained in the impoundment is constant with time. In many cases, this volume
will decrease as the lagoon loses moisture to evaporation, seepage, and effluent. Second, only the
most recent deposit of sludge is considered when estimating emissions. In other words, Equation 6-
18 estimates the rate at which contaminants are emitted from the most recent deposit of sludge, up
until the time when the sludge is removed or the impoundment receives a subsequent deposit. Where
sludge deposits are infrequent and rates of volatilization high, this model may provide a reasonable
6-18
-------
representation of expected emissions. In other applications the model may underestimate actual
emissions because of its neglect of potential emissions of contaminants from previous sludge deposits.
For impoundments with frequent deposits of sludge, the continuous inflow models described in
Section 6.3.7.1 and 6.3.7.2 should be used.
6.3.7.4. ESTIMATION OF EMISSIONS FROM WASTEWATER TREATMENT LAGOONS
USED FOR LONG-TERM SLUDGE STORAGE -- Emissions from wastewater treatment lagoons
can be modeled with methods similar to those described in Section 6.3.7.1. For both types of
facilities, sludge contaminants are assumed to enter the impoundment continuously. Wastewater
treatment lagoons differ from sludge impoundments in two respects, however. First, the solids
content of liquids entering a wastewater treatment lagoon is typically lower than that of water
entering a sludge lagoon. Second, wastewater treatment lagoons may use diffused air as a component
of wastewater treatment. The methodology described in Section 6.3.7.1 is equally valid for facilities
with lower concentrations of solids in inflow and can be applied to wastewater treatment lagoons that
do not use diffused air. This section, therefore, presents a methodology for estimating emissions for
wastewater treatment systems that include diffused air.
Diffused air acts as a medium for transporting contaminant directly from the impoundment
into the atmosphere. This methodology assumes that contaminant concentrations in diffused air reach
equilibrium with the impoundment's liquid before the air reaches the surface. Contaminant emissions
are then assumed to equal the volumetric flow rate of diffused air multiplied by the estimated
equilibrium concentration in diffused air. As discussed in Section 6.3.4, losses to volatilization in
diffused air are assumed to be described by M^ = Qa (Hc/RT) Ct. If this term is inserted into
Equation 6-8, and Equations 6-9 and 6-10 are combined as in Section 6.3.7.1, the following estimate
is obtained for the concentration of contaminant in the impoundment:
QCn(l +k.)
Cl= -^-^ ^ (6-19)
Q(l+ksl) + kb V(l+ksl) + kA + Qa(Hc/RT)
Total volatile emissions can then be calculated as:
6-19
-------
Esite
or
Esjte/C0 = - (6-20)
Q(l+ksl) + kbV(l+ksl) + kA + Qa(Hc/RT)
6.3.8. Estimation of Wind Transport
For both Tier 1 and Tier 2 calculations, the methodology includes consideration of the
dispersion of contaminants likely to occur during wind transport. This dispersion is estimated with
the Industrial Source Complex Long Term (ISCLT) model, as described in U.S. EPA (1979) and U.S.
EPA (1986b) and implemented in the Graphical Exposure Modeling System (U.S. EPA, 1988c) and
elsewhere. The model predicts long-term average contaminant concentrations in ambient air at
specified receptor locations near a site. It assumes that the plume of contaminated air follows a
Gaussian distribution, that the emission rate is uniform and continuous over the source, and that
meteorological conditions are constant between the source and the receptor location. The model
represents the average concentration within 16 directional sectors (each of 22.5°) based on the
following:
1. the mass flux from the source area,
2. the wind speed,
3. the distance from the virtual source,
4. the vertical standard deviation of the Gaussian plume based on the stability class and
5. deposition flux onto the ground surface and losses to other decay processes.
For estimating contaminant transport from area sources of emissions (such as surface
impoundments), the model represents emissions as a point source located at a specified virtual
distance upwind of the actual site. From ESE (1985), the atmospheric transport and dispersion model
for the sector-averaged concentration of a specific sector near an area source of emissions is:
E - A -Z 2 \ (-.
-.
Ca
-------
where:
C = the concentration of contaminant in air at the location defined by X. and Z.
a S3
Xg = distance in x-coordinate direction (parallel to velocity Uwj) from source to
point of interest (m)
Za = distance in z-coordinate direction (perpendicular to wind velocity, UHJ) from
source to point of interest (m)
Es-te = mass flux of contaminant into the atmosphere (gm/sec)
A = first-order source depletion due to atmospheric decay of contaminant, and wet
and dry deposition (dimensionless)
a = mixing coefficient in z direction, standard deviation of Gaussian plume (m)
xy. = virtual distance required for point source plume to spread to width of site (m)
{.. = frequency of the specific stability array parameters for classification i,j
(stability class, wind speed)
UM- = average wind speed in j direction (m/sec)
First-order contaminant losses are calculated as A = exp[-A (Xa/Uw)], where A equals a lumped
first-order loss constant (sec"1) and Xa/Uw equals the time required for transport of the contaminant
from the site to the receptor location. According to ESE (1985), contaminant losses to deposition and
atmospheric decay are usually negligible within the time required to reach the location of the most
exposed individual. The mixing coefficient in the vertical direction (ffz) is calculated from the
distance Xa, based on parameters that vary according to Pasquill-Guifford stability category.
Parameter values appropriate for stable atmospheric conditions, for example, are presented in Table
6-2.
The input requirements for ISCLT can be satisfied with local stability array (STAR) weather
data available for most locations in the United States. For each radial distance specified by the user,
the model provides estimated average ground level concentrations of contaminant in each of 16
directions from the source. For an analysis of exposure to an MEI, the modeler is interested in the
highest predicted air concentration beyond the property boundary. Because Equation 6-21 is linear
with respect to Egite, the ratio between air concentration and emissions will be constant for a single
contaminant, a specified lagoon location and area, and a single receptor location, so that a "wind
transport ratio" can be defined as Ca/Egite for the MEI.
6.3.9. Deriving Criteria. The final step in deriving Tier 2 criteria for the air pathway is to
relate the selected RAC at the location of the MEI to the maximum allowable dry-weight
6-21
-------
TABLE 6-2
Parameters Used to Calculate aza
Pasquil
Stability Category
x (km)
Very Unstable6
Unstable6
Slightly Unstable6
Neutral
Slightly Stable
0.10 - 0.15
0.16 - 0.20
0.21 - 0.25
0.26 - 0.30
0.31 - 0.40
0.41 - 0.50
0.51 - 3.11
3.11
0.10 - 0.20
0.21 - 0.40
0.40
0.10
0.10 - 0.30
0.31 - 1.00
1.01 - 3.00
3.01 - 10.0
10.01 - 30.0
30.00
0.10 - 0.30
0.31 - 1.00
1.01 - 2.00
2.01 - 4.00
4.01 - 10.0
10.01 - 20.0
20.01 - 40.0
40.00
a
158.080
170.222
179.520
217.410
258.890
346.750
453.850
(c)
90.673
98.483
109.300
62.141
34.458
32.093
32.093
33.504
36.650
44.053
23.331
21.628
21.628
22.534
24.703
26.970
35.420
47.618
P
1.04520
1.09320
1.12620
1.26440
1.40940
1.72830
2.11660
(c)
0.93198
0.98332
1.09710
0.91465
0.86974
0.81066
0.64403
0.60586
0.56589
0.51179
0.81956
0.75660
0.63077
0.57154
0.50527
0.46713
0.37615
0.29592
6-22
-------
TABLE 6-2 (cont.)
Pasquil
Stability Category x (km) az = a
a ft
Stable 0.10 - 0.20
0.21 - 0.70
0.71 - 1.00
1.01 - 2.00
2.01 - 3.00
3.01 - 7.00
7.01 - 15.0
15.01 - 30.0
30.01 - 60.0
> 60.00
15.209
14.457
13.953
13.953
14.823
16.187
17.836
22.651
27.084
34.219
0.81558
0.78407
0.68465
0.63227
0.54503
0.46490
0.41507
0.32681
0.27436
0.21716
aSource: Environmental Science and Engineering, 1985
ktf the calculated value of <7Z exceeds 5000 m, cr_ is set to 5000 m
ca=5000 m
z
6-23
-------
concentrations of contaminant in sludge received by the impoundment. As for other pathways
discussed in this document, this relation may be described by a "source receptor ratio" (in kg/m3) for
each impoundment site and MEI location evaluated:
SRRVOL = Ca/N (6-25)
where Ca represents the expected concentration of contaminant in ambient air (mg/m3) at the location
of the MEI, and N represents the concentration of a contaminant in sludge (mg/kg). For each sludge
contaminant to be considered, this ratio is easily calculated with Equations 6-22 and 6-8, together
with an estimate of Esite/C0 derived by Equation 6-13, 6-17, 6-18 or 6-20:
SRRVOL = (Ca/Esite>(Esite/Co)(C0/N) (6-26)
Once SRRyoL has been estimated from Equation 6-26, maximum allowable levels of each contaminant
in sludge received or accumulated in a surface disposal facility (in mg/kg dry-weight) can be
determined by:
= (RAC mg/m3)/(SRRVOL kg/m3) (6-27)
A worst-case scenario for wind transport of emitted contaminants, appropriate for Tier 1
criteria, could be created with results from an ISCLT run based on STAR data for Santa Barbara,
California, which produces the highest concentrations at the receptor site. With the conservative
assumption that contaminant loss through degradation is negligible during wind transport to the
location of the MEI, the transport ratio is the same for all volatile contaminants for a given site and
a given receptor location. For this reason, a single estimated value for the transport ratio (or perhaps
different values for different ranges of wind speed) could be used to derive all Tier 1 criteria.
A slightly different approach for deriving a transport ratio for Tier 1 would be based on a
set of simplifying assumptions described in ESE (1985) and incorporated into the methodology
-------
described in U.S. EPA (1986a). This approach assumes (1) that there are no contaminant losses to
deposition or other atmospheric decay processes (A = 1), (2) that the wind direction is constant in the
direction of the receptor, (3) that atmospheric conditions are always stable with no restriction on
vertical mixing, (4) that the receptor is located on the centerline of wind direction (Yfl = 0), (5) that
the receptor breathes air at concentrations predicted for ground level (Za = 0) and (6) that wind speed
is constant, described by the local average (U; = Uu for all j). With these assumptions, Equation 6-
J M
21 reduces to:
Ca = 2.032 Esjte /[az (Xa+xyj) UJ (6-23)
All of the above assumptions lead to a conservative estimate of average downwind contaminant
concentrations. The assumption of unlimited vertical mixing could lead to a lower estimate of
contaminant concentrations at the receptor location, but should not affect results under stable
conditions and over the short transport distances to be considered when deriving criteria. For a more
precise calculation, the full ISCLT procedure should be executed and site-specific stability array
(STAR) weather data should be used.
ESE (1985) recommends approximating xy- by assuming that all contaminant is contained in
a 22.5° angular sector that originates at the virtual source and for which the width matches the
effective diameter of the site. Based on this approach, virtual distance can be approximated by:
xvl. = -S- cot (22.5/2)
where de=2(A/7r)°'5 equals the effective diameter of the lagoon. This equation can be approximated
by xyi = 2.84 A0'5, and produces a conservative derivation of virtual distance when compared with
a more precise method based on atmospheric conditions and wind velocity. A nearly identical
approach has been found to give reasonable predictions of ambient air concentrations under
experimental conditions (Baker and MacKay, 1985).
6-25
-------
Equation 6-23 can be rearranged as:
Ca/Esite ' 2-032
6.4. INPUT PARAMETER REQUIREMENTS
Several input parameters are needed to describe conditions within the impoundment to be
modeled, and to describe meteorological conditions in the surrounding area. Reasonable worst case
parameter values should be used for deriving Tier 1 numerical criteria for national application, but
values appropriate for each individual site must be specified for Tier 2. Three types of parameters
are required: (1) those describing the chemical or physical characteristics of the volatile contaminant
under consideration, (2) those describing the surface disposal facility and the quantity of wastewater
or sludge it receives, and (3) meteorological characteristics of the area.
6.4.1. Chemical Characteristics. For each sludge contaminant to be modeled, the user must
specify four parameters. The diffusivity of the contaminant in water and its diff usivity in air may
be derived with methods described in Lyman (1982). Henry's Law constant can be obtained from
Table 5-2. Biodegradation rates for many organic chemicals are tabulated in Schnoor et al. (1987).
Partition coefficients for each contaminant can be derived with Equation 5-15, using values of KOU(
from Table 5-2, or can be derived from KOC values reported in Appendix C of U.S. EPA (1986a), or
can be estimated with the CHEMEST procedures in GEMS (U.S. EPA, 1988c).
6.4.2. Site Characteristics. Dimensions of each lagoon, rates of influent and effluent flows,
and rates of diffused air flow should be readily available at each site. The distance to the property
boundary can be determined from the site plan. Rates of seepage from the facility can be estimated
from a mass balance that includes inflow, effluent, evaporation and precipitation. The concentration
of solids in sludge or wastewater received by each facility can be obtained from site measurements.
6.4.3. Meteorological Conditions. Local average temperature, wind speeds, and atmospheric
stability data can be obtained from the STAR station nearest to each site.
6-26
-------
6.5. HEALTH AND ENVIRONMENTAL EFFECTS
Toxic pollutants that volatilize to air from surface disposal sites can cause adverse human
health effects to those living downwind from the surface disposal site. The reference air
concentration (RAC) is defined as the air concentration of pollutant used to evaluate the potential
for adverse effects on human health. If a particular concentration of pollutant in sludge results in
an air concentration that is greater than the RAC, adverse health effects may occur in a population
exposed to this concentration at a compliance point downwind from the disposal site.
The units used for the RAC in this methodology are mg/m3. To obtain the RAC in these
units, a reference intake in mg-day/kg is converted using a particular human body weight assumed
to represent the average in the exposed population, and using an assumed average daily air inhalation
rate. Furthermore, the reference intake is adjusted to account for the intake of the contaminant from
sources unrelated to contamination from sludge.
The procedure for determining the RAC varies according to whether the pollutant acts by a
threshold or nonthreshold mechanism of toxicity.
6.5.1. Threshold-Acting Toxicants. Threshold-acting toxicants are those for which a dose can
be identified below which no adverse effects are assumed to occur. The Agency assumes that all
noncarcinogenic chemicals act according to threshold mechanisms. The RAC is derived as follows
for threshold-acting toxicants:
RAC = [(RfD BW RE'1) - TBI] / Ia (6-28)
where:
RAC = reference air concentration (mg/m3)
RfD = reference dose (mg-day/kg)
BW = human body weight (kg)
RE = relative effectiveness of inhalation exposure (unitless)
TBI = total background intake rate of contaminant (mg/day) from all other
sources of exposure
Ia = total air inhalation rate (m3/day)
The parameters RfD, BW, and TBI have been discussed in previous sections. The definition
and derivation of Ia and RE are given below.
6-27
-------
6.5.1.1. TOTAL AIR INHALATION RATE (Ia) -- Table 6-3 shows values of Ig for a
typical man, woman, child and infant with a typical activity schedule, as defined by the International
Commission on Radiological Protection (ICRP, 1975). The Agency has used an assumption of 20 m3
per day for a 70-kg adult to represent a typical inhalation rate in many Agency risk assessments.
Values have also been derived for two additional scenarios: an adult with the same activity schedule,
but with upper limit rather than average respiration rates for each activity, and an adult with normal
respiration rates but whose work is moderately active and who practices 1 hour of heavy activity per
day (Fruhman, 1964; Astrand and Rodahl, 1977). Representative body weights have been assigned
to each of these individuals to derive a respiratory volume-to-body weight ratio. The resulting ratios
range from 0.33 to 0.47 m3-day/kg. Values in this range exceed the ratio value of 0.29 m3-day/kg,
which corresponds to a 70-kg adult inhaling 20 m3 per day. Therefore, use of the latter value may
underestimate actual exposure for certain adults. In circumstances where children are known to be
at special risk, it may be more appropriate to base the derivation of criteria on the inhalation rate of
toddlers or infants.
6.5.1.2. FRACTION OF INHALED AIR FROM CONTAMINATED AREA -- All
individuals exposed to emissions from a surface disposal site may not necessarily remain in proximity
to the surface disposal site for 24 hours a day. However, if residential areas near the site are
affected, then less mobile individuals (e.g., infants or the elderly) may be exposed for approximately
24 hours a day. These less mobile residents may be included among those determined to be at greatest
risk, whom the criteria are designed to protect. Therefore, it is reasonable to assume for this
methodology that 100% of the air inhaled by the MEI is from the area of the surface disposal site.
6.5.1.3. RELATIVE EFFECTIVENESS OF EXPOSURE (RE) -- RE, as discussed
previously, is a unitless factor that shows that relative toxicological effectiveness of an exposure by
a given route when compared with another route. Since exposure from the vapor pathway is through
inhalation, the RE factors applied in Equation 6-25 represent the relative effectiveness of exposure
though the media for which the RfD was derived when compared with exposure through inhalation.
Where no relevant data are available, RE should be assumed to equal 1.
6-28
-------
TABLE 6-3
Dai ly Respiratory Volumes for "Reference" Individuals (Normal Individuals at Typical Activity Levels)
and for Adults with Higher-than-Normal Respiratory Volume or Higher-than-Normal Activity Levels3
ro
vo
8-hour working, light activity
8-hour working, moderate activity
8-hour nonoccupational (light) activity
7-hour nonoccupational (light) activity
1-hour heavy activity
8-hour rest
Total dai ly respiratory volume
Body weight (kg)
Ratio of volume to body weight
"Reference" "Reference" Upper-Limit Active
Manb Womanb Adult Adult
(m3/day) (m3/day) (m3/day) (m3/day)
9.6 9.1 12. Oc
14. 4d
9.6 9.1 12. Oc
8.4
2.5d
3.6 2.9 7.0e 3.6
22.8 21.1 31.0 28.9
70 58 65 65
0.33 0.36 0.47 0.44
"Reference" "Reference"
Child: Infant:
10 years 1 year
(m3/day) (m3/day)
6.24 2.5
(10 hours)
--
6.24
..
..
2.3 1.3
(14 hours)
14.78 3.8
33 10
0.45 0.38
aSource: U.S. EPA, 1989d.
Upper-limit values for "ordinary man or woman," for "light" activity category (As t rand and Rodahl, 1977).
Averaged values for "ordinary man or woman," for "moderate" or "heavy" activity categories (As t rand and Rodahl, 1977)
Upper 95th percent! le for adults, age 20-39 years, at rest (Fruhman, 1964)
The inhalation volume to body weight ratios have been derived from referenced values for i I lustrati ve purposes only.
-------
6.5.2. Carcinogens. The general approach used by the Agency to assess carcinogens is discussed
in Section 4.5.2. For pollutants assessed as carcinogenic (nonthreshold-acting) agents, the reference
air concentration is derived as follows:
RAC = (RL BW) / (q,* RE Ig) (6-29)
where:
RAC = reference air concentration (mg/m3)
RL = risk level (unitless)
BW = human body weight (kg)
q. = human cancer potency [(mg-day/kg)'1]
RE = relative effectiveness of inhalation exposure (unitless)
Ia = total air inhalation rate (m3/day)
For carcinogens, the RAC as computed with Equation 6-29 is thought to be protective, since the
estimate of carcinogenicity is an upper limit value. All parameters have been defined and discussed
in previous sections.
6.6. SAMPLE CALCULATIONS
Sample calculations for benzene volatilization are provided for three actual surface disposal
facilities: a wastewater treatment and sludge storage facility in Antrim, New Hampshire, a sludge
storage facility in Tulsa, Oklahoma, and a sludge disposal facility in Portland, Oregon.
6.6.1. Analysis of Exposure for the Most Exposed Individual. Sample calculations will be
presented in this section for the derivation of criteria for the same three sites evaluated in Sections
4.7, and 5.6. These sample calculations determine the maximum allowable concentration of benzene
for sludge that accumulates or is deposited in these surface disposal facilities. Actual or estimated
benzene concentrations in the sludge are then compared with this value to determine whether sludge
characteristics are acceptable. Benzene has been chosen to illustrate the methodology because it is
highly volatile and because exposure via the air pathway may be of concern. Pertinent chemical-
specific data for benzene are provided in Table 6-4. It is assumed that the air pathway is not of
concern for lead, the other contaminant evaluated in Sections 4.7 and 5.5.
6-30
-------
TABLE 6-4
Input Parameters for Estimating Emissions of Benzene
Parameter
Value
(Units)
PB
Dca
Dcw
Deth
Hc
R
ScG
ScL
foe
koe
kd
[kb]2Q
[k 1
l.SxlO'4
1.2xlO"3
8.8xlO'2
9.8xlO'6
8.5xlO'6
5.5xlO'3
8.21xlO'5
1.7
918
0.50
74
37
1.27xlO"6
1.7xlO'6
(g/cm-sec)
(g/cm3)
(cm2/sec)
(cm2/sec)
(cm2/sec)
(atm-m3/mol)
(m3-atm/mol-°K)
(unitless)
(unitless)
(unitless)
(//kg)
(//kg)
(sec'1)
(sec"1)
6-31
-------
6.6.1.1. SAMPLE CALCULATION: ANTRIM, NEW HAMPSHIRE -- The three lagoons
at the Antrim wastewater treatment facility were described in Section 4.7.1.1. For the calculations
carried out in this section, it is assumed that the contents of the lagoons are well-mixed: that the
characteristics of the sludge and liquid are similar throughout the lagoons. The three lagoons are
modeled as a single aggregated facility with total area equal to the total of the three impoundments.
The calculations proceed in six steps:
1. Estimation of the reference air concentration (RAC, in mg/m3) for benzene,
2. Estimation of the ratio between the concentration of dissolved contaminant in the
lagoon and the dry-weight concentration in sludge received by or accumulated in the
lagoon (C0/N, in kg/m3)
3. Estimation of the ratio of emissions to contaminant concentrations in liquid contained
in the lagoon (ESjte/Ct, in m3/sec)
4. Estimation of the ratio between concentrations of contaminant in ambient air at the
selected receptor location for the MEI and the rate of total emissions from the site
'in
5. Estimation of the ratio between contaminant concentrations in ambient air and dry
weight concentrations in sludge (SRRVOL, in kg/m3), and
6. Estimation of the maximum allowable dry-weight concentration of contaminant in
sludge received by or accumulated in the surface disposal facility (N(nax, in mg/kg).
Each of these steps will be described below.
6.6.1.1.1. Derivation of the Reference Air Concentration — Tier 1 and Tier 2 calculations
begin with the derivation of a reference air concentration, or RAC. Since benzene is a carcinogen
with an established potency value (qt*), the reference air concentration is calculated with Equation
6-29:
RAC = (RL BW) / (q,* RE Ifl)
For illustration, this sample calculation arbitrarily selects a risk level of 10~6. Based on a body weight
of 70 kg, an inhalation volume of 20 m3/day and a relative effectiveness of 1.0 for inhalation,
6-32
-------
RAC = [(10*6)(70)]/[(2.9xlO'2)(l)(20)] = 1.2 x 10"4 mg/m3
6.6.1.1.2. Estimation of CQ/N -- The relationship between the concentration of the
benzene in the inflowing liquid and the dry-weight concentration of benzene in sludge accumulating
in the impoundment is estimated with Equation 4-1 based on a reported solids concentration of
5.4xlO"4 kg/1 in influent to the impoundment:
C0/N = l/(kd H-r^'1) = l/[37+(5.3xlO'4)'1] = 5.3xlO'4 kg// = 0.53 kg/m3
6.6.1.1.3. Estimation of Esite/C0 -- Expected emissions of benzene from the
impoundment are next related to the concentration of benzene in the lagoons' inflow, based on
Equation 6-19 for well-mixed systems with diffused air. Evaluation of that equation requires
estimation of the overall mass transfer coefficient. This coefficient is estimated from the liquid and
gas phase mass transfer coefficients, from Equation 6-5:
I _ 1 El
k kt Hckg
Selection of an equation for estimating kt requires calculation of the effective diameter (de) or
"fetch" of the idealized lagoon:
de = 2(A/7r)°-5 = 2(5040/3.142)0-5 = 80.1 m
Since the depth of the lagoons is approximately 3 meters, the fetch:depth ratio is about (80/3)=26.
Since average wind speed in Antrim is 3.2 m/sec (less than 3.25 m/sec), and the fetch:depth ratio
for the Antrim facility is 26 (which is between 14 and 52), kt can be calculated from Equation
6-21:
6-33
-------
kt = 2.78xlO-6(DH/Deth)2/3
2.78xlO'6 (9.8xlO'6/8.5xlO'6)2/3 - 3.1xlO'6 m/sec
The mass transfer coefficient for the gas side requires estimation of the Schmidt number for the
contaminant in air:
SCG = Ma/(pa Dca)
1.8xlO'V[(l-2xlO'3)(8.8xlO'2)] = 1.7
Substituting into Equation 5-24 yields:
kQ = l.SxlO'3 U°-78 Sc -0.67 d -0.11
y w y c
(l.8xlO-3)(3.2)°-78(l.7)-°-67(80)-°-11
1.9xlO"3 m/sec
Substituting values for kt and k into Equation 6-6 yields
1/k = l/(3.1xlO"6) + (8.21xlO'5)(298)/[(5.5xlO'3)(1.9xlO'3)]
or
k = 3.0xlO"6 m/sec
The biodegradation rate for Equation 6-20 is adjusted to a temperature of 25" C with
Equation 5-27:
kb = 1.27xlO'6 (1.06)(25'20) = 1.7xlO'6 sec'1
From Equation 6-20 and input values from Tables 6-4 and 6-5, emissions can be related to
contaminant concentrations in inflow to the lagoon:
6-34
-------
r ,„ [Qa(Hc/RT) + kA]Q(l+ksl)
Esite/C0 "
{Q(l+kst) + kbV(l+ksl) + kA + Qa(Hc/RT)}
[(6.2X0.22) + (3.0xlO'6)(5.040)](0.01)(l.Q2)
(0.01)(1.02) + (1.7xIO'6)(1.5xl04)(1.02)+(3.0xlO'6)(5040)+(6.14)(0.22)
9.9xlO"3 m3/sec
6.6.1.1.4. Estimation of the Transport Ratio (Ca/ESfte) -- To estimate the expected
dispersion of benzene during wind transport, the ISCLT model was executed with STAR data
retrieved through GEMS from the Concord, NH airport (latitude 43.2, longitude 71.5). These data
are converted into a form usable by ISCLT, as presented in Table 6-6. Execution of ISCLT with
these input values results a maximum concentration of benzene in ambient air beyond 500 meters
of the site of 4.6x10 g/m for each g/m -sec emitted from the site. This unit rate of emissions is
equivalent to 5040 g/sec from a site of 5040 m2, so:
Ca/Esite = (4.6xlO~2 g/m2)/(5040 g/sec) = 9.2xlO'6 sec/m3
6.6.1.1.5. Estimation of the Source-Receptor Ratio (SRRTOL) --
Results derived in Sections 6.6.1.1.2 through 6.6.1.1.4 (and summarized in Table 6-5) are combined
to estimate the source-receptor ratio, or Ca/N:
a
SRRVOL = (Ca/Esite)(Esite/C0XC0/N)
= (9.2x10'6 sec/m3)(9.9xlO'3 m3/sec)(0.53 kg/m3)
= 4.8xlO"8 kg/m3
6.6.1.1.6. Estimation of the Maximum Allowable Concentration in Sludge -- From
Equation 6-27, the maximum allowable dry-weight concentration of benzene in sludge accumulating
in the Antrim facility is then:
6-35
-------
TABLE 6-5
Derivation of Criteria for the Air Pathway
Antrim, New Hampshire
Parameter
A
h
V
Q
Qa
rsl
kst
u10
T
de
U*
kt
kg
k
C0/N
Esjte/CQ
Ca/Esite
SRRvol
mdV
Value
5040
3.0
15000
0.01
6.14
0.54
0.02
3.2
298
80
0.091
3.1xlO'6
1.9xlO"3
3.0xlO'6
0.53
9.9xlO'3
9.2xlO'6
4.8xlO'8
2500
(Units)
(m2)
(m)
(m3)
(m3/sec)
(m3/sec)
(kg/m3)
(unitless)
(m/sec)
(o -\r \
E^f
(m)
(m/sec)
(m/sec)
(m/sec)
(m/sec)
(kg/m3)
(m3/sec)
(sec/m3)
(kg/m3)
(mg/kg)
6-36
-------
TABLE 6-6
Input Parameters for Execution of ISCLT
Number of Source 1
Number of X Axis Grid System Points 30
Number of Y Axis Grid System Points 16
Number of Special Points 0
Number of Seasons 1
Number of Wind Speed Classes 6
Number of Stability Classes 6
Number of Wind Direction Classes 16
File Number of Data File Used for Reports 1
Program Mode Rural
Concentration (Deposition) Units Conversion Factor 1.0x10
Acceleration of Gravity (meters/sec ) 9.800
Height of Measurement of Wind Speed (meters) 10.000
Correction Angle for Grid System Versus Direction Data North (degrees) 0.000
Decay Coefficient 0.1155x10"2
Program option switches: 1, 2, 2, 0, 0, 3, 2, 3, 3, 0, 0, 0, 7,-8,-9, 0, 0, 1, 0, 0, 1, 1, 1, 1
Width of Area (M) 71.0
Source Strength (g/sec-m) 1.0
6-37
-------
TABLE 6-6 (cont.)
Ambient Air Tenperature (Degrees Kelvin)
Stability Stability Stability Stability Stability Stability
Category 1 Category 2 Category 3 Category 4 Category 5 Category 6
Season 1 283.9000 283.9000 283.9000 280.6000 277.6000 277.6000
6-38
-------
TABLE 6-6 (cont.)
Mixing Layer Height (meters)
Wind Speed
Category 1
Wind Speed
Category 2
Wind Speed
Category 3
Wind Speed
Category U
Wind Speed
Category 5
Wind Speed
Category 6
Stab. Category 1
Stab. Category 2
Stab. Category 3
Stab. Category U
Stab. Category 5
Stab. Category 6
0.18225x10* 0.18225x10* 0.18225x10* 0.18225x10* 0.18225x10* 0.18225x10*
0.12150x10* 0.12150x10* 0.12150x10* 0.12150x10* 0.12150x104 0.12150x10^
0.12150x10* 0.12150x10* 0.12150x10* 0.12150x10* 0.12150x104 0.12150x104
0.12150x10* 0.12150x104 0.12150x10* 0.12150x104 0.12150x104 0.12150X104
0.10000x105 0.10000x105 0.10000x105 0.10000x105 0.10000x105 0.10000x105
0.10000x105 0.10000x105 0.10000x105 0.10000x105 0.10000x105 0.10000x10
5
6-39
-------
TABLE 6-6 (cont.)
Frequency of Occurrence of Wind Speed, Direction and Stability
Stability Category 1
Wind Speed Category
Direction 123456
(degrees) (1.5 mps) (2.5 tips) (4.3 mps) (6.8 mps) (9.5 mps) (12.5 mps)
0.000 0.00022001 0.00016000 0.00000000 0.00000000 0.00000000 0.00000000
22.500 0.00002000 0.00002000 0.00000000 0.00000000 0.00000000 0.00000000
45.000 0.00012000 0.00009000 0.00000000 0.00000000 0.00000000 0.00000000
67.500 0.00008000 0.00005000 0.00000000 0.00000000 0.00000000 0.00000000
90.000 0.00013000 0.00016000 0.00000000 0.00000000 0.00000000 0.00000000
112.500 0.00033001 0.00030001 0.00000000 0.00000000 0.00000000 0.00000000
135.000 0.00030001 0.00021001 0.00000000 0.00000000 0.00000000 0.00000000
157.500 0.00018001 0.00011000 0.00000000 0.00000000 0.00000000 0.00000000
180.000 0.00058002 0.00030001 0.00000000 0.00000000 0.00000000 0.00000000
202.500 0.00023001 0.00023001 0.00000000 0.00000000 0.00000000 0.00000000
225.000 0.00052002 0.00027001 0.00000000 0.00000000 0.00000000 0.00000000
247.500 0.00032001 0.00018001 0.00000000 0.00000000 0.00000000 0.00000000
270.000 0.00051002 0.00037001 0.00000000 0.00000000 0.00000000 0.00000000
292.500 0.00033001 0.00025001 0.00000000 0.00000000 0.00000000 0.00000000
315.000 0.00042001 0.00041001 0.00000000 0.00000000 0.00000000 0.00000000
337.500 0.00023001 0.00018001 0.00000000 0.00000000 0.00000000 0.00000000
-------
TABLE 6-6 (cont.)
Frequency of Occurrence of Wind Speed, Direction and Stability
Stability Category 2
Wind Speed Category
Direction 1 23456
(degrees) (1.5 mps) (2.5 mps) (4.3 mps) (6.8 mps) (9.5 mps) (12.5 mps)
0.000 0.00209006 0.00128004 0.00043001 0.00000000 0.00000000 0.00000000
22.500 0.00105003 0.00050002 0.00009000 0.00000000 0.00000000 0.00000000
45.000 0.00108003 0.00048001 0.00027001 0.00000000 0.00000000 0.00000000
67.500 0.00097003 0.00039001 0.00027001 0.00000000 0.00000000 0.00000000
90.000 0.00123004 0.00055002 0.00027001 0.00000000 0.00000000 0.00000000
112.500 0.00127004 0.00048001 0.00021001 0.00000000 0.00000000 0.00000000
135.000 0.00173005 0.00112003 0.00059002 0.00000000 0.00000000 0.00000000
157.500 0.00149004 0.00123004 0.00080002 0.00000000 0.00000000 0.00000000
180.000 0.00214006 0.00201006 0.00126004 0.00000000 0.00000000 0.00000000
202.500 0.00197006 0.00146004 0.00075002 0.00000000 0.00000000 0.00000000
225.000 0.00313009 0.00237007 0.00121004 0.00000000 0.00000000 0.00000000
247.500 0.00223007 0.00176005 0.00103003 0.00000000 0.00000000 0.00000000
270.000 0.00284009 0.00237007 0.00128004 0.00000000 0.00000000 0.00000000
292.500 0.00279008 0.00210006 0.00185006 0.00000000 0.00000000 0.00000000
315.000 0.00379011 0.00263008 0.00183005 0.00000000 0.00000000 0.00000000
337.500 0.00309009 0.00196006 0.00100003 0.00000000 0.00000000 0.00000000
6-41
-------
TABLE 6-6 (cont.)
Frequency of Occurrence of Wind Speed, Direction and Stability
Stability Category 3
Wind Speed Category
Direction 1 23456
(degrees) (1.5 mps) (2.5 mps) (4.3 mps) (6.8 mps) (9.5 mps) (12.5 mps)
0.000 0.00126004 0.00146004 0.00151005 0.00027001 0.00000000 0.00000000
22.500 0.00062002 0.00064002 0.00064002 0.00005000 0.00000000 0.00000000
45.000 0.00090003 0.00075002 0.00094003 0.00000000 0.00000000 0.00000000
67.500 0.00061002 0.00043001 0.00059002 0.00009000 0.00000000 0.00000000
90.000 0.00076002 0.00080002 0.00071002 0.00000000 0.00002000 0.00000000
112.500 0.00059002 0.00048001 0.00089003 0.00025001 0.00000000 0.00000000
135.000 0.00115003 0.00148004 0.00247007 0.00016000 0.00002000 0.00000000
157.500 0.00144004 0.00153005 0.00304009 0.00027001 0.00000000 0.00000000
180.000 0.00157005 0.00201006 0.00288009 0.00043001 0.00000000 0.00000000
202.500 0.00104003 0.00162005 0.00222007 0.00023001 0.00000000 0.00000000
225.000 0.00156005 0.00267008 0.00354011 0.00057002 0.00000000 0.00000000
247.500 0.00101003 0.00153005 0.00354011 0.00050002 0.00002000 0.00000000
270.000 0.00133004 0.00201006 0.00452014 0.00059002 0.00002000 0.00000000
292.500 0.00195006 0.00301009 0.00621019 0.00119004 0.00009000 0.00005000
315.000 0.00305009 0.00562017 0.00916027 0.00212006 0.00023001 0.00002000
337.500 0.00211006 0.00313009 0.00438013 0.00114003 0.00009000 0.00002000
6-1*2
-------
TABLE 6-6 (cont.)
Frequency of Occurrence of Wind Speed, Direction and Stability
Stability Category 4
Wind Speed Category
Direction 123456
(degrees) (1.5 mps) (2.5 mps) (4.3 mps) (6.8 mps) (9.5 mps) (12.5 nips)
0.000 0.00701021 0.00797024 0.00573017 0.00375011 0.00030001 0.00000000
22.500 0.00470014 0.00591018 0.00537016 0.00206006 0.00014000 0.00005000
45.000 0.00778023 0.00920028 0.00820025 0.00283008 0.00034001 0.00007000
67.500 0.00418013 0.00523016 0.00525016 0.00244007 0.00037001 0.00032001
90.000 0.00560017 0.00649020 0.00537016 0.00267008 0.00030001 0.00005000
112.500 0.00462014 0.00553017 0.00502015 0.00233007 0.00018001 0.00002000
135.000 0.00857026 0.00993030 0.00767023 0.00272008 0.00016000 0.00000000
157.500 0.00625019 0.00895027 0.01062032 0.00308009 0.00009000 0.00000000
180.000 0.00654020 0.00929028 0.01023031 0.00343010 0.00014000 0.00002000
202.500 0.00266008 0.00361011 0.00514015 0.00233007 0.00011000 0.00002000
225.000 0.00337010 0.00413012 0.00468014 0.00276008 0.00009000 0.00002000
247.500 0.00222007 0.00331010 0.00498015 0.00411012 0.00014000 0.00009000
270.000 0.00269008 0.00377011 0.00728022 0.00808024 0.00112003 0.00005000
292.500 0.00440013 0.00596018 0.01238037 0.02206066 0.00393012 0.00057002
315.000 0.00992030 0.01304039 0.01984060 0.03494105 0.00820025 0.00210006
337.500 0.00629019 0.00824025 0.01028031 0.01886057 0.00377011 0.00057002
6-43
-------
TABLE 6-6 (cent.)
Frequency of Occurrence of Wind Speed, Direction and Stability
Stability Category 5
Wind Speed Category •
Direction 123456
(degrees) (1.5 mps) (2.5 mps) (4.3 mps) (6.8 mps) (9.5 mps) (12.5 mps)
0.000 0.00000000 0.00185006 0.00137004 0.00000000 0.00000000 0.00000000
22.500 0.00000000 0.00116003 0.00034001 0.00000000 0.00000000 0.00000000
45.000 0.00000000 0.00151005 0.00007000 0.00000000 0.00000000 0.00000000
67.500 0.00000000 0.00089003 0.00011000 0.00000000 0.00000000 0.00000000
90.000 0.00000000 0.00121004 0.00014000 0.00000000 0.00000000 0.00000000
112.500 0.00000000 0.00187006 0.00050002 0.00000000 0.00000000 0.00000000
135.000 0.00000000 0.00512015 0.00132004 0.00000000 0.00000000 0.00000000
157.500 0.00000000 0.00438013 0.00137004 0.00000000 0.00000000 0.00000000
180.000 0.00000000 0.00523016 0.00215006 0.00000000 0.00000000 0.00000000
202.500 0.00000000 0.00219007 0.00091003 0.00000000 0.00000000 0.00000000
225.000 0.00000000 0.00288009 0.00162005 0.00000000 0.00000000 0.00000000
247.500 0.00000000 0.00139004 0.00247007 0.00000000 0.00000000 0.00000000
270.000 0.00000000 0.00206006 0.00443013 0.00000000 0.00000000 0.00000000
292.500 0.00000000 0.00256008 0.00854026 0.00000000 0.00000000 0.00000000
315.000 0.00000000 0.00457014 0.01142034 0.00000000 0.00000000 0.00000000
337.500 0.00000000 0.00258008 0.00530016 0.00000000 0.00000000 0.00000000
6-kk
-------
TABLE 6-6 (cont.)
Frequency of Occurrence of Wind Speed, Direction and Stability
Stability Category 6
Wind Speed Category •
Direction 1 23456
(degrees) (1.5 mps) (2.5 mps) (4.3 mps) (6.8 mps) (9.5 mps) (12.5 mps)
0.000 0.01015030 0.00331010 0.00000000 0.00000000 0.00000000 0.00000000
22.500 0.00803024 0.00212006 0.00000000 0.00000000 0.00000000 0.00000000
45.000 0.01573047 0.00327010 0.00000000 0.00000000 0.00000000 0.00000000
67.500 0.00701021 0.00126004 0.00000000 0.00000000 0.00000000 0.00000000
90.000 0.01304039 0.00256008 0.00000000 0.00000000 0.00000000 0.00000000
112.500 0.00662020 0.00176005 0.00000000 0.00000000 0.00000000 0.00000000
135.000 0.01879056 0.00459014 0.00000000 0.00000000 0.00000000 0.00000000
157.500 0.01343040 0.00432013 0.00000000 0.00000000 0.00000000 0.00000000
180.000 0.01859056 0.00667020 0.00000000 0.00000000 0.00000000 0.00000000
202.500 0.01022031 0.00290009 0.00000000 0.00000000 0.00000000 0.00000000
225.000 0.01405042 0.00539016 0.00000000 0.00000000 0.00000000 0.00000000
247.500 0.00687021 0.00354011 0.00000000 0.00000000 0.00000000 0.00000000
270.000 0.01136034 0.00443013 0.00000000 0.00000000 0.00000000 0.00000000
292.500 0.01126034 0.00612018 0.00000000 0.00000000 0.00000000 0.00000000
315.000 0.02234067 0.00998030 0.00000000 0.00000000 0.00000000 0.00000000
337.500 0.01206036 0.00518016 0.00000000 0.00000000 0.00000000 0.00000000
6-45
-------
TABLE 6-6 (cont.)
Vertical Potential Temperature Gradient (Degrees Kelvin/meter)
Wind speed
category 1
Wind speed
category 2
Wind speed
category 3
Wind speed
category 4
Wind speed
category 5
Wind speed
category 6
Stab. Category 1
Stab. Category 2
Stab. Category 3
Stab. Category 4
Stab. Category 5
Stab. Category 6
0.000000x10° 0.000000x10° 0.000000x10° 0.000000x10° 0.000000x10° 0.000000x10°
0.000000x10° 0.000000x10° 0.000000x10° 0.000000x10° 0.000000x10° 0.000000x10°
0.000000x10° 0.000000x10° 0.000000x10° 0.000000x10° 0.000000x10° 0.000000x10°
0.000000x10
,-1
0.000000x10
,-1
0.000000x10
-1
0.20000x10"' 0.20000x10 ' 0.20000x10 ' 0.20000x10 ' 0.20000x10 ' 0.20000x10
0.000000x10'
-1
0.000000x10°
,-1
0.000000x10"
-1
,-1
0.35000x10"' 0.35000x10"' 0.35000x10 ' 0.35000x10 ' 0.35000x10 ' 0.35000x10
-1
,-1
,-1
6-46
-------
TABLE 6-6 (cont.)
Wind Profile Power Law Exponents
Wind speed
category 1
Wind speed
category 2
Wind speed
category 3
Wind speed
category 4
Wind speed
category 5
Wind speed
category 6
Stab. Category 1
Stab. Category 2
Stab. Category 3
Stab. Category 4
Stab. Category 5
Stab, category 6
0.70000x10"
0.70000x10"
,-1
-1
0.70000x10 ' 0.70000x10 ' 0.70000x10 ' 0.70000x10
,-1
,-1
-1
0.70000x10 ' 0.70000x10 ' 0.70000x10 ' 0.70000x10
°
,-1
0.70000x10"
0.70000x10"
0.100000x10 0.100000x10 0.100000x10 0.100000x10 0.100000x10 0.100000x10
0.150000x10
0.350000x10'
0.150000x10 0.150000x10° 0.150000x10 0.150000x10u 0.150000x10u
0.350000x10° 0.350000x10° 0.350000x10° 0.350000x10° 0.350000x10°
°
0.550000x10 0.550000x10 0.550000x10 0.550000x10 0.550000x10 0.550000x10
0
-------
N™ =RAC/SRRVOL
= (1.2xlO"4 mg/m3 )/(4.8xlO"8 kg/m3)
= 2500 mg/kg
If dry-weight concentrations of benzene exceed this value, then sludge should not be allowed to
accumulate in the lagoons at this site.
6.6.1.2. SAMPLE CALCULATION: TULSA, OKLAHOMA -- The Tulsa, Oklahoma,
facility consists of three impoundments connected in series. The impoundments receive sludge in
continuous inflow and release decanted effluent for return to the wastewater treatment facility.
Sludge is removed from the impoundments about every 9-12 months. The lagoons are judged to be
poorly mixed, with the concentrations of sludge varying across the impoundment length. For these
sample calculations, the individual impoundments contained in the Tulsa facility will therefore be
modeled as a single aggregate impoundment. Table 6-7 contains input parameters appropriate for
modeling the Tulsa site, as well as results from the calculations described below. As before, a
reference air concentration (RAC) of 1.2xlO"4 is selected, corresponding to a risk level of 10"6.
The expected ratio between the concentration of dissolved benzene in sludge received by the
first lagoon and the dry-weight concentration of benzene in the sludge is calculated from Equation
4-1, where the ratio of solids to liquids in that sludge is reported to be about 0.016 kg//:
C0/N = l/(kd +rsl-1) = l/[37+(0.0!6)'1] = 0.01 kg// = 10 kg/m3
Expected emissions of benzene are next related to the concentration of dissolved contaminant
in sludge received by the first lagoon, based on equations developed in Section 6.3.7.2 for poorly
mixed systems. As before, the overall mass transfer coefficient (k) is estimated from the liquid and
gas phase mass transfer coefficients (k and kt). With a total area of 61,000 m2, the lagoons in Tulsa
have an effective diameter, or fetch, of:
6-48
-------
TABLE 6-7
Derivation of Criteria for the Air Pathway
Tulsa, Oklahoma
Parameter
A
h
V
Q
rsl
ksl
U10
T
de
U*
kl
k
k
C0/N
Esite/C0
Ca/Esite
SRRvol
N
Value
61,000
3.1
1.9xl05
0.01
16
0.59
4.9
298
280
0.15
6.9xlO"6
2.3xlO"3
6.8xlO'6
10
7.1xlO"2
6.5xlO"6
4.6xlO'7
260
(Units)
(m2)
(m)
(m3
(m3/sec)
(kg/m3)
(unitless)
(m/sec)
(°K)
(m)
(m/sec)
(m/sec)
(m/sec)
(m/sec)
(kg/m3)
(m3/sec)
(sec/m3)
(kg/m3)
(mg/kg)
6-49
-------
de = 2(A/7r)°-5 = 2(61,000/3.142)°'5 = 280 m
Since the depth of the lagoons is reported to be about 3.1 meters, the fetchrdepth ratio is
(280/3.1) = 90 m. Average wind speed in Tulsa (4.9 m/sec) is greater than 3.2 m/sec, and the
fetchrdepth ratio (93) is greater than 52, so kt (in m/sec) can be calculated from Equation 5-23.
kt = 2.61 IxlO'7 U1Q2 [DCH/Deth]<2/3> = 2.61 IxlO"7 (4.9)2 [9.8xlO'6)/(8.5xlO"6)](2/3) = 6.9xlO'6
The mass transfer coefficient for the gas side is estimated from Equation 5-24:
ka = l.SxlO'3 Uu°'78 Sc •°'67 d •°'11 = (l.8xlO'3)(4.9)°-78(l.70r°-67(279)'0-11 = 2.3xlO'3 m/sec
y wye
These estimates of kt and k are substituted into Equation 6-5:
l/(6.9xlO"6) + (8.21xlO~5)(298)/[(5.5xlO'3)(2.3xlO"3)]
or
k = 6.8x10'6 m/sec
Equation 6-17 for plug flow systems gives the ratio of expected benzene emissions to the dissolved
concentration of benzene in liquid received by the lagoons. First, a must be calculated:
kA + kbV(l+ksl)
(6.8xlO')(61,OOOHl.7xlO-K190,000)(1.59)
58
Substituting this value into Equation 6-17 yields:
6-50
-------
Esite/C0 = k A [l-exp(-a)] / a = (6.8xlO'6)(61,000)[l-exp(-32.5)]/(32.5) = 7.1xlO'3
To estimate the expected concentration of benzene in ambient air as function of emissions from the
Tulsa site, the ISCLT model is executed within GEMS, based on STAR data from the Tulsa airport
(latitude 36.2, longitude 95.9). Results from ISCLT execution reveal that for a unit emission rate of
1 g/m2-sec from the lagoons, the maximum expected concentration in ambient air beyond the
property boundary is 0.40 g/m3, so C /E •_ can be calculated as:
a SI tS
C/Esite = (0.40 g/m3)/(61,000 g/sec) = 6.5xlO'6 sec/m3
The source-receptor ratio is next calculated by combining the results derived above:
L = (Ca/Es,-teXEsite/CoXC0/N)
= (6.5xlO"6 sec/m3)(7.1xlO"2 m3/sec)(10 kg/m3)
- 4.6xlO'7 kg/m3
From the source-receptor ratio can be derived the maximum allowable dry-weight concentration of
benzene in sludge received by the lagoons in Tulsa:
Nn*x = RAC/SRRVOL = (1.2x10'* mg/m3)/(4.6xlO-7 kg/m3) = 260 mg/kg
6.6.1.3. SAMPLE CALCULATION: PORTLAND, OREGON -- The third sample
calculation involves an impoundment for permanent disposal of sludge. The Portland facility uses
its large (13 ha) impoundment only occasionally, when the primary means of sludge reuse
(composting) is inoperative. Such failures occur infrequently. For the sample calculations presented
here, it will be assumed that the composting operation fails once each year, and that upon failure of
that system, the impoundment receives 260,000 gallons of sludge, with 1.8% solids content. The
6-51
-------
calculations will be based on the conservative assumption that each sludge deposit spreads to a thin
layer that covers the entire surface of the impoundment. Because deposits to the impoundment are
sporadic and relatively infrequent, the facility will be modeled with methods presented in Section
6.3.7.3. Input parameters for the calculations are provided by Table 6-8. As before the calculations
will be based on a reference air concentration of 1.2xlO"4 mg/m3.
Based on the reported ratio of solids to liquids (1.8%) in each deposit of sludge, the ratio of
the concentration of dissolved benzene to the dry-weight concentration of benzene received by the
lagoon is estimated from Equation 4-1:
C0/N = l/(kd +rsl-1) = l/[37+(0.0!8)'1] = LlxlO"2 kg// = 11 kg/m3
Expected emissions of benzene are next related to the concentration of dissolved contaminant
in sludge deposited in the first impoundment, based on Equation 6-18 for impoundments with
infrequent deposits of sludge. First, the overall mass transfer coefficient (k) is estimated from the
liquid and gas phase mass transfer coefficients (k and kt). Based on the assumption that the sludge
spreads to cover the entire surface of the impoundment (with a total area of 130,000 m3), the
effective diameter, or "fetch" of the surface is:
d = 2(A/7T)0'5 = 2(130,000/3.142)°'5 = 410 m
e
The depth of such a layer of sludge would be only 7.7xlO"3 m, so that the fetch:depth ratio is much
greater than 51.2. Average wind speed in Portland (3.7 m/sec) is greater than 3.2 m/sec, so kt (in
m/sec) can be calculated from Equation 5-23.
kt = 2.61 IxlO'7 U102 [Dcw/Deth](2/3) = 2.61 IxlQ-7 (3.7)2 [9.8xlO'6)/(8.5xlO'6)](2/3) = 3.9xlO'6
The mass transfer coefficient for the gas side is estimated from Equation 5-24:
6-52
-------
TABLE 6-8
Derivation of Criteria for the Air Pathway
Portland, Oregon
Parameter
A
dep
rsl
ksl
U,o
T
de
u*
kl
kg
k
CO/N
Esite/c0
Ca/Esite
SRRvot
N
Value
130,000
1000
18
0.67
3.7
298
410
l.lxlO"1
3.9xlO"6
1.8xlO"3
3.9xlO"6
11
5.3xlO'5
3.4xlO'6
1.9xlO"9
63,000
(Units)
(m2)
(m3)
(kg/m3)
(//kg)
(m/sec)
(°^" \
•"•/
(m)
(m/sec)
(m/sec)
(m/sec)
(m/sec)
(kg/m3)
(m3/sec)
(sec/m3)
(kg/m3)
(mg/kg)
6-53
-------
kg= l.8xlO-3Uw°-78Scg'0-67de-°-11 = (L8xlO-3)(3.7)°-78(l.70)-°-67(4lO)-0-11 = l.SxlO'3 m/sec
These estimates of kt and k are then substituted into Equation 6-5:
= l/(3.9xlO"6) + (8.21xlO"5)(298)/[(5.5xlO'3)(1.8xl(r3)]
k = 3.9xlO"6 m/sec
or
^6
Equation 6-18 gives the ratio of expected benzene emissions to the dissolved concentration of benzene
in liquid received by the lagoons, but requires the calculation of a value for 0:
0< A Vdep-V kb(l+ksl)
1 + 1.7xlO"6(1.67)
1.67
Substituting this value into Equation 6-18 gives:
kA [l-exp(-/n")]
.
= 3.1xlO"4
Esite/C0
/TT
(3.9xlO'6)(1.3xl05){l-exp[(-3.1xlO"4)(3.2xl07)]}
(3.1xlO~4)(3.2xl07)
= 5.3xlO"5 m3/sec
To estimate the expected concentration of benzene in ambient air as function of emissions from the
Portland site, the ISCLT model is executed within GEMS, based on STAR data retrieved from the
Portland airport, for latitude 45.6 and longitude 122.6. Results from ISCLT execution reveal that for
a unit emission rate of 1 g/m2-sec from the lagoons, the maximum expected concentration in ambient
air beyond the property boundary is 0.44 g/m3. Ca/Es-te can then be calculated as:
-------
C/E .,„ = (0.44 g/m3)/( 130,000 g/sec) = 3.4xlO'6 sec/m3
a SI L™
The source-receptor ratio is calculated by combining previous results (as listed in Table 6-8):
SRRm = (Ca/Esjte)(Esjte/C0)(C0/N)
= (3.4xlO"6 sec/m3)(5.3x!0'5 m3/sec)(ll kg/m3)
= 1.9x10'9 kg/m3
This ratio is used to calculate the maximum allowable dry-weight concentration for benzene in sludge
deposited in the lagoon:
Nmax = RAC/SRRVOL = (1.2xlO'4 mg/m3)/(2.0xlO'9 kg/m3) = 60,000 mg/kg
6.6.2. Analysis of Exposure for the Most Exposed Populations.
6.6.2.1. SAMPLE CALCULATION: ANTRIM, NEW HAMPSHIRE -- Tier 2 calculations
provide an upper bound estimate of maximum concentrations likely to be encountered at a specified
receptor location. Conversely, as illustrated above, they provide a derivation of maximum allowable
sludge concentrations, based on specified reference air concentrations, site characteristics and a
specified receptor distance.
The distribution of potential exposure over the entire exposed population may also be of
interest. Execution of the complete ISCLT model can provide a more complete picture of the
distribution of potential exposure over the total exposed population (and a determination of how
criteria would vary with alternative definitions of the MEI). To derive such a distribution for the
Antrim facility, the ISCLT model was executed with meteorological data from the Concord, New
Hampshire, airport STAR station, and with census data for each area within 50 kilometers of Antrim,
New Hampshire. The model reported both expected concentrations and sizes of exposed populations
for each of 30 ring distances in each of 16 radial directions, for a total of 480 locations.
6-55
-------
Figure 6- 1 shows the estimated distribution of population surrounding the Antrim facility.
As can be seen from the figure, approximately 450,000 persons reside within 50 kilometers of the
facility; of these, over 95% reside at a distance greater than 10 kilometers. Approximately 1,000
people live within 1 kilometer of the Antrim facility. Figure 6-2 shows estimated ground-level
concentrations of contaminant (per unit of emissions from the site) for each of 4 radial directions
selected from the 16 examined by the model. Most of the contamination occurs within 500 meters
of the facility with similar concentration reduction patterns for the four radial directions. As can
be seen from a simple comparison of these two graphs, most of the exposed population will be
exposed to ambient air concentrations considerably lower than those within the innermost ring
distances. Figure 6-3 combines estimated air concentration with population data to derive a
distribution of air concentrations as a function of the number of persons exposed. Based on a unit
rate of emissions from the Antrim site (1 g/sec), the graph presents the number of persons expected
to breathe air at each level of contamination. From the graph, for example, it can be seen that
approximately 100 persons are expected to be exposed to air with at least 0.04 mg/m3 of contaminant
for each g/sec of emissions from the Antrim facility. Approximately 30 people are estimated to be
exposed to the relatively high contaminant concentration of 0.08 mg/m3 per g/m2-sec of emission
from the Antrim facility.
If:
Esite/N " (Esite/CoXCo/N) = ^xlO'3 m3/sec)(0.53 kg/m3) = 5.3xlO'3 kg/sec
then 5.3xlO"3 mg/sec (or 5.3xlO'6 g/sec) of benzene is expected to be emitted from the lagoons for
each mg/kg dry-weight of benzene in influent. Combining this result with data presented in Figure
6-3 shows that for each mg/kg of contaminant concentration in Antrim sludge, about 30 persons
are expected to be exposed to more than 4xlO"7 mg/m3 of contaminant concentration in ambient air.
6.6.2.2. SAMPLE CALCULATION: TULSA, OKLAHOMA -- Similar distributions of
populations by distance and exposure level have been derived for the area surrounding the Tulsa
facility. Calculations are based on weather and census data for the area within 50 kilometers of the
6-56
-------
Cumulative Population
(Thousands)
O
en
o
o
o
en
O
ro
o
o
ro
Ui
o
o
o
en
o
o
o
-fs.
en
o
o
o
c/)'
r-t-
Q
n
CD
—h
O
3
a
o
ro
o
o
en
o
o
-------
FIGURE 6-2
i
ui
oo
w
\
CM
\
-------
Log(lO) of Exposed Population
Q
D
in
~O
o
Q
5'
(ft
CD
O
o
o
o
o -
p
b
ro
o
b
p
o
CD
O
O
00
n
OJ
CD
CJi
Ln
Q
15
CO
"O
O
I
Q
O
~D
O
"D
C_
Q
O
D
33
m
ON
-------
center of Tulsa. Figure 6-4 shows the distribution of populations by distance from the Tulsa facility.
As shown by the graph, approximately 600,000 persons live within 50 kilometers of the facility. In
contrast to Antrim, about half of this population lives within 10 kilometers of the facility; about
10,000 people live within 1 kilometer of the facility.
Ambient air concentrations per unit emissions are shown in Figure 6-5 for four radial
directions. The air concentrations in Tulsa are generally higher than the air concentrations in
Antrim, and fall sharply within the first 500 meters. As before, the estimate of population
distribution has been combined with the estimates of expected ambient air concentrations per unit
emission to derive a distribution of the number of persons exposed as a function of air
concentrations. Results are provided by Figure 6.6. As can be seen from the graph, a larger
population (about 100,000 persons) is potentially exposed in the Tulsa area than in the Antrim area,
but the highest level of contamination per unit emissions is lower in the area surrounding Tulsa than
the area around Antrim.
Results from Section 6.6.1.2 can be combined to show that 7.2xlO"5 g/sec of benzene are
expected to be emitted from Tulsa impoundments for each mg/kg dry-weight of benzene in the
sludge. It can therefore be shown from Figure 6-6 that about 3 people would be exposed to a
concentration greater than 2x1 Q~6 mg/m3 per mg/kg concentration in sludge.
6.6.2.3. SAMPLE CALCULATION: PORTLAND, OREGON -- Calculations described in
Sections 6.6.2.1 and 6.6.2.2 were repeated for the Portland facility, based on meteorological data
from the STAR station at Portland and census data for persons living within 50 kilometers of the
center of Portland. The estimated distribution of population by distance from the facility is
presented in Figure 6-7, indicating a greater population density (about 1.3 million people) in the
50-kilometer radius surrounding the Portland area than in the area surrounding Antrim or Tulsa.
About 9,000 people live within 1 kilometer of the location modeled.
Attenuation of air concentrations in four selected directions is graphed in Figure 6-8. The
air concentrations fall substantially by a distance of 1 kilometer from the facility. Figure 6-9
provides a graph of the size of exposed populations by the ratio of ambient air concentrations per
unit of emissions from the Portland facility. Although the Portland population potentially exposed
6-60
-------
19-9
Cumulative Population
(Thousands)
O
o
o
ro
o
o
C/J
o
o
o
o
en
O
O
O
Cj
en'
r-f-
Q
D
O
CD
—h
O
3
~n
Q
g
t-i-
X
ro
o
o
o
o
o
o
O
c_
o"
<'
0)
o
(7)'
03
en
O
0)
I c
i-
I. o
TJ
O
"O
c_
o"
o*
13
I
Jr-
-------
ro
£
\
O>
i_
0)
a
CP c
is o
LJ
0)
Q_
6
C
0
o
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
FIGURE 6-5
Air Concentration per Unit Emissions
For Four Selected Directions, Tulsa OK
0
2
4
Distance from Facility (km)
-------
£9-9
Log(10) of Exposed Population
o
ro
CT>
U)
T3
O
Q
r-t-
o'
LO
ft)
O
OJ
o
o
o
o
p
b
o
p
o
o
oo
p
o
ro
p
b
en
p
b
p
b
p
b
ro
oo
p
b
oo
ro
Q
13
00
O
Q
g o
Q
rj CT
3
Q
"D
O
c_
Q~
O
13
CT>
m
ON
I
-------
FIGURE 6-7
o
D
0 w
°- o
0) '—
<
O
Cumulative Distribution of Popu ation
By Distance from Portland Facility
0
40
Distance from Facility (km)
-------
vn
E
en
^
-------
99-9
Log(10) of Exposed Population
o
ro
CD
o
Q
3
C/l
T>
O
Q
r-f-
5'
U)
0)
o
^
OJ
o
o
o
p
b
o
ro
p
b
o
p
b
o
CD
p
b
o
00
p
b
-H
~i
Q
D
(f)
O
§- Q
Q I l~
P- 5'
o
S CT
T)
O
C_
Q
15
CO
-------
is about 1 million people, the concentration to which the MEI is exposed per unit of emissions is
lower than was described above for Tulsa and Antrim. If 5.7xlO"7 g/sec are emitted per mg/kg of
contaminant in sludge disposed, then about 30 people are exposed to over 6x10~9 mg/m3 per mg/kg
of contaminant in sludge disposed in the Portland area.
6-67
-------
7.0 SIMULTANEOUS CONSIDERATION OF MULTIPLE PATHWAYS OF EXPOSURE
Chapters 4, 5, and 6 described methods for estimating potential exposure to a most exposed
individual (MEI) or to the most exposed populations (MEP). These methods consider each pathway
of potential exposure independently; they do not consider the possibility that a single individual could
be the most exposed individual for more than one pathway at a time, or even that exposure through
one pathway might add to the exposure from another. While it is probably unlikely that a single
individual could be the MEI through drinking water pathways associated with both groundwater and
surface water, a single individual living near a surface disposal facility might conceivably be the MEI
for both groundwater and volatilization pathways, and perhaps also for the fish ingestion pathway
associated with surface water contamination. Similarly, significant simultaneous exposure is possible
through the air pathway and the drinking water component of the surface water pathway. This
chapter briefly outlines an approach for joint consideration of these pathways.
Common to Chapters 4, 5, and 6 was the derivation of source-receptor ratios (SRRQW, SRRSU,
and SRRVOL, respectively) that relate the concentration of each contaminant in the respective
environmental medium to the dry-weight concentration of contaminant in sludge received by the
surface disposal facility. With equations adapted from Sections 4.5, 5.5, and 6.5, these three ratios can
be used to predict total exposure to a single MEI exposed through more than one pathway
simultaneously. For example, simultaneous exposure through volatilization and groundwater
pathways might be described as:
T SRRVOLIaREA + SRRGWIWREW + TBI
1
where:
HETOT = total exposure of the MEI to contaminant released from the surface disposal
facility (mg/kg-day)
N - = dry weight concentration of contaminant in sludge (mg/kg)
SRRVOL= source receptor ratio for volatilization pathway: the ratio of the air
concentration at the receptor location to the dry weight concentration of the
same contaminant in sludge received by the facility (kg/m3)
L = quantity of air inhaled per day (m3/day)
REA = relative effectiveness of exposure to inhaled contaminant (unitless)
7-1
-------
SRRGW = source receptor ratio for groundwater pathway: the ratio of the concentration
of the contaminant in well water ingested by the MEI to the dry weight
concentration of sludge received by the facility (kg//)
I, = quantity of water ingested per day (//day)
REy = relative effectiveness of exposure to ingested contaminant (unitless)
TBI = total exposure to this contaminant from all other sources (mg/kg-day)
BW = body weight of MEI (kg)
If a maximum allowable level of human exposure can be established from the human cancer potency
(q,,*) or risk reference dose (RfD) for the particular contaminant of concern, then the above equation
can be used to calculate the maximum dry weight concentration of the contaminant in sludge received
or accumulated by the facility.:
BW
HE
max SRRVOLIaREA + SRRGWIUREU + TBI J
where:
Nmax = maximum allowable dry weight concentration of this contaminant in sludge
received or accumulated by this surface disposal facility (mg/kg)
HEmax = maximum allowable exposure to this contaminant, derived with methods
adapted from Sections 4.5, 5.5, and 6.5.
For non-human exposure (i.e., comparison of surface water quality to water quality criteria),
the methods described in Chapter 5 could be applied exactly as described in that Chapter. An
integrated model code could then identify the limiting pathway or group of pathways for each
contaminant, and report the maximum allowable concentration of each contaminant to be accepted
by the facility.
7-2
-------
8. REFERENCES
Astrand, P.O. and K. Rodahl. 1977. Textbook of Work Physiology, 2nd ed. McGraw-Hill, New
York, NY. Cited in Fiserova-Bergerova, 1983. Modeling of Inhalation Exposure to Vapors:
Uptake, Distribution, and Elimination, Vol II. CRC Press, Inc., Boca Raton, FL.)
Bahner, L.H., A.J. Wilson, J.M. Sheppard, J.M. Parick, L.R. Goodman and G.E. Walsh. 1977.
Kepone, bioconcentration, accumulation, loss, and transfer through estuarine food chains.
Chesapeake Sci. 18: 297-308.
Baes III, C.F., R.D. Sharp, A.L. Sjoreen, and R.W. Shor. 1984. A review and analysis of
parameters for assessing transport of environmentally released radionuclides through agriculture.
Oak Ridge National Laboratory. Oak Ridge, TN. ORNL-5786.
Baker, L.W. and K.P. MacKay. 1985. Screening models for estimating toxic air pollution near a
hazardous waste landfill. J. Air Pollut. Contr. Assoc., 35(11).
Barnes, H.H. 1967. Roughness Characteristics of Natural Channels. U.S. Geological Survey
Water-Supply Paper 1849. U.S. Government Printing Office. Washington, D.C.
Brooks, R.H., and A.T. Corey. 1966. Properties of porous media affecting fluid flow. ASCE J.
Irrig. Drain Div. 92 (IR2): 61-68.
Carsel, R.F., and R.S. Parrish. 1988. Developing joint probability distributions of soil-water
retention characteristics. Water Resour. Res. 24(5): 755-769.
Chow, V.T. 1959. Open Channel Hydraulics. McGraw-Hill Book Company, New York, NY.
Clayton, J.R., S.P. Pavlou and N.F. Breitner. 1977. Polychlorinated biphenyls in coastal marine
zooplankton: Bioaccumulation by equilibrium partitioning. Environ. Sci. Tech. 11: 676-682.
Davis, L.A., and S.P. Neuman. 1983. Documentation and User's Guide: UNSAT2 - Variably
Saturated Flow Model. U.S. Nuclear Regulatory Commission Report, NUREG/CR-3390,
Washington, D.C.
DeWald, J. and R. Phillips. 1989. Determination of the fate of toxic organic materials in surface
sludge impoundments. Draft final report. Prepared by S-CUBED, La Jolla, CA for the Risk
Reduction Engineering Laboratory, Office of Research and Development, U.S. EPA, Cincinnati,
OH.
ESE (Environmental Science and Engineering). 1985. Exposure to Airborne Contaminants
Released From Land Disposal Facilities - A Proposed Methodology. ESE No. 85-527-0100-2140.
Prepared by Environmental Science and Engineering, Gainesville, FL, for Office of Solid Waste,
U.S. EPA, Washington, DC.
Environmental Sciences Service Administration. 1968. Climatic Atlas of the United States,
Washington, DC.
Freeze, R.A., and J.A. Cherry. 1979. Groundwater. Prentice-Hall, Inc. Englewood Cliffs, NJ.
Fruhman, G. 1964. Title not given. Z. Exp. Med. 138: 1. Cited in Diem, K. and C. Lentner,
1970. Scientific Tables. Ciba-Geigy, Ltd., Basle, Switzerland.
Gillies, M.E. and H.V. Paulin. 1983. Variability of mineral intakes from drinking water: A
possible explanation for the controversy over the relationship of water quality to cardiovascular
disease. Int. J. Epidemiol. 12: 45-50.
8-1
-------
Hounslow, A.W. 1983. Adsorption and Movement of Organic Pollutants. Proc. 3rd Nat.
Symposium on Aquifer Restoration and Groundwater Monitoring. NWWA, Worthington, OH.
p. 334-346.
Hutson, J.L., and Wagenet, R.J. 1985. LEACHM: A Model For Simulating The Leaching And
Chemistry of Solutes in the Plant Root Zone. New York State Agric. Exp. Stn.; Ithaca, NY.
Huyakorn, P.S., and G.F. Finder. 1983. Computational Methods in Subsurface Flow. Academic
Press. 473 pp.
Huyakorn, P.S., A.G. Kretschek, R.W. Broome, J.W. Mercer, and B.H. Lester. 1984. Testing and
Validation of Models for Simulating Solute Transport in Ground-Water. International Ground
Water Modeling Center, Holcomb Research Inst., HRI No. 35.
Huyakorn, P.S., M.J. Ungs, E.A. Sudicky, L.A. Mulkey, and T.D. Wadsworth. 1985. RCRA
Hazardous Waste Identification and Land Disposal Restrictions: Groundwater Screening
Procedures, Technical Report prepared for the Office of Solid Waste, U.S. EPA, under contract
No. 68-01-7075.
Hwang, S.T. 1985. Model prediction of volatile emissions. Environ. Prog. 4(2): 141-144.
Hwang, S.T., and L.J. Thibodeaux. 1982. Toxic emissions from land disposal facilities. Envir.
Prog. 1(1): 46-52.
Incropera, F.P. and D.P. DeWitt. 1985. Fundamentals of Heat and Mass Transfer. John Wiley
and Sons, Inc. New York, NY.
ICRP (International Commission on Radiological Protection). 1975. Report of the Task Group on
Reference Man (No. 23). Pergammon Press, Ltd. London.
Kehew, A.E., D.J. Brown and F.J. Schwindt. 1983. Effect of seepage from unlined municipal
waste stabilization lagoons on chemical quality of groundwater in shallow aquifers. Prepared for
the U.S. Department of the Interior, Washington DC. Project Number A-072-NDAK.
Agreement Number 14-34-0001-1136.
Lee, R.F., C. Ryan and M.L. Neuhausen. 1976. Fate of petroleum hydrocarbons taken up from
food and water by the blue crab, Callinectes saoidus. Mar. Biol. 37: 363-370.
Lester, B.H., P.S. Huyakorn, H.O. White Jr., T.D. Wadsworth, and J.E. Buckley. 1986. Analytical
Models for Evaluating Leachate Migration in Groundwater Systems. Prepared for the Office of
Solid Waste, U.S. EPA, Washington DC.
Lyman, WJ. 1982. Handbook of Chemical Property Estimation Methods. McGraw-Hill, New
York, NY.
MacKay, D. and P.J. Leinonen. 1975. Rate of evaporation of low solubility contaminants from
water bodies to atmosphere. Environ. Sci. and Technol. 20(5): 1178-1181.
MacKay, D. and A. Yuen. 1983. Mass transfer coefficient correlations for volatilization of
organic s'olutes from water. Environ. Sci. and Technol. 17(4): 211-217.
Marcelle, C. and J.P. Thome. 1984. Relative importance of dietary and environmental sources of
lindane in fish. Bull. Environ. Contam. Toxicol. 33: 423-429.
Metcalf & Eddy, Inc. 1979. Wastewater Engineering. McGraw-Hill Book Company, New York,
NY.
8-2
-------
NRC (National Research Council). 1983. Risk Assessment in the Federal Government:
Managing the Process. National Academy Press, Washington, DC.
NRC (National Research Council). 1986. Dose-route extrapolations: Using inhalation toxicity
data to set drinking water limits. In: Drinking Water and Health, Vol. 6. National Academy
Press, Washington, D.C.
Neeley, W.B., D.R. Bronson and G.E. Blau. 1974. Partition coefficient to measure
bioconcentration potential of organic chemicals in fish. Environ. Sci. Technol. 8: 1113-1115.
Norstrom, R.J., A.E. McKinnon and A.S.W. DeFrietas. 1976. A bioenergetics-based model for
pollutant accumulation by fish. Simulation of PCB and methylmercury residue levels in Ottawa
River yellow perch (Perca flavescens). J. Fish Res. Board Can. 33: 248-267.
Pitter, P. 1976. Determination of biological degradability of organic substances. Water Res.
10: 231-235.
Schnoor, J.L., C. Sato., D. McKechnie, and D. Sahoo. 1987. Processes, Coefficients, and Models
for Simulating Toxic Organics and Heavy Metals in Surface Waters. EPA/600/3-87/015.
Shen, T.T. 1982. Estimation of organic compound emissions from waste lagoons. J. Air Pollut.
Contr. Assoc. 32(4): 79-82.
Springer, C., P.D. Lunney and K.T. Valsaraj. 1984. Emission of Hazardous Chemicals from
Surface and Near Surface Impoundments to Air. U.S. EPA, Solid and Hazardous Waste Research
Division, Cincinnati, OH. Project Number 808161-02.
Stefan, H.G. and A.C. Demetracopoulos. 1981. Cells-in-series simulation of riverine transport.
ASCE J. Hydraulic Engineer. 107(6): 675-697.
Stephan, C.E. 1980. Memorandum to J.F. Stara, Environmental Criteria and Assessment Office,
U.S. EPA. July 30.
Thomann, R.V. and J.A. Mueller. 1987. Principles of Surface Water Quality Modeling and
Control. Harper & Row Publishers, New York, NY.
U.S. Department of Commerce. 1985. Fisheries of the United States. April 1985. Current
Fisheries Statistics No. 8360.
U.S. EPA. 1978. Process Design Manual, Municipal Sludge Landfills. EPA 625/1-78-010,
SW-705.
U.S. EPA. 1979. Industrial Source Complex (ISC) Model User's Guide, Volumes I and II. EPA
450/4-79-030 and EPA 450/4-79-031.
U.S. EPA. 1980a. Seafood Consumption Data Analysis: Final Report. Prepared by SRI
International, Menlo Park, CA, under Contract No. 68-01-3887. U.S. EPA, Washington, D.C.
U.S. EPA. 1980b. Ambient Water Quality Criteria Document for Hexachlorocyclohexane.
Prepared by the Office of Health and Environmental Assessment, Environmental Criteria and
Assessment Office, Cincinnati, OH, for the Office of Water Regulations and Standards,
Washington, DC. EPA 440/5-80-054. NTIS PB 81-117657.
U.S. EPA. 1980c. Water Quality Criteria Documents: Availability. Federal Register 45(231):
79318-79379.
8-3
-------
U.S. EPA. 1984a. Health Assessment Document for Carbon Tetrachloride. Office of Health and
Environmental Assessment, Environmental Criteria and Assessment Office, Cincinnati, OH. EPA
600/8-82-00IF. NTIS PB 84-124196.
U.S. EPA. 1984b. Health Assessment Document for Chloroform. Office of Health and
Environmental Assessment, Environmental Criteria and Assessment Office, Research Triangle
Park, N.C. EPA 600/8-84-004A. NTIS PB 84-195163.
U.S. EPA. 1984c. Methods for Estimating Fish Catch Sizes. Prepared by the Office of Health
and Environmental Assessment for the Office of Water Regulations and Standards, Washington,
D.C. EPA/600/6-84/007. NTIS PB84-210277.
U.S. EPA. 1985a. Drinking Water Criteria Document for Nickel. Prepared by the Office of
Health and Environmental Assessment, Environmental Criteria and Assessment Office,
Cincinnati, OH, for the Office of Drinking Water, Washington, D.C. EPA/600/X-84/193. NTIS
PB 86-117801.
U.S. EPA. 1985b. Health Assessment Document for Polychlorinated Dibenzo-p-Dioxins. Office
of Health and Environmental Assessment, Environmental Criteria and Assessment Office,
Cincinnati, OH. EPA 600/8-84-014F. NTIS PB 86-122546.
U.S. EPA. 1985c. Water Quality Assessment: A Screening Procedure for Toxic and Conventional
Pollutants in Surface and Ground Water—Parts I and II. EPA/600/6-85/002b. Athens, GA.
U.S. EPA. 1985d. DRASTIC: A Standardized System for Evaluating Ground Water Pollution
Potential Using Hydrogeologic Settings. Prepared by Robert S. Kerr Environmental Research
Laboratory for the Office of Research and Development, Washington, DC. EPA/600/2-85/018.
U.S. EPA. 1985e. National Primary Drinking Water Regulations; Synthetic Organic Chemicals,
Inorganic Chemicals and Microorganisms; Proposed Rule. (40 CFR Part 141) Federal Register
50(219): 46936-47022.
U.S. EPA. 1986a. Development of Risk Assessment Methodology for Municipal Sludge
Landfilling. Prepared by the Office of Health and Environmental Assessment, Environmental
Criteria and Assessment Office, Cincinnati, OH for the Office of Water Regulations and
Standards, Washington, DC.
U.S. EPA. 1986b. Industrial Source Dispersion Model User's Guide - Second Edition. Office of
Air Quality Planning and Standards, Research Triangle Park, NC. EPA/450/4-86-005a.
U.S. EPA. 1986c. Guidelines for Carcinogen Risk Assessment. Federal Register 51(185): 33992-
34003.
U.S. EPA. 1986d. Hazardous Waste Management System: Land Disposal Restrictions, Proposed
Rules. 51(9): 1652.
U.S. EPA. 1986e. Guidelines for the Health Risk Assessment of Chemical Mixtures. Federal
Register. 51(185):34014-34025.
U.S. EPA. 1987a. Hazardous Waste Treatment, Storage, and Disposal Facilities (TSDF) - Air
Emissions Models. Office of Air Quality Planning and Standards, Research Triangle Park, NC.
EPA-450/3-87-026.
U.S. EPA. 1987b. Report to Congress: Municipal Wastewater Lagoon Study, Volumes I and II.
Office of Municipal Pollution Control.
8-4
-------
U.S. EPA. 1988a. WASP4, A Hydrodynamic and Water Quality Model—Model Theory, User's
Manual, and Programmer's Guide. Office of Research and Development, Athens, GA.
EPA/600/3-87/039.
U.S. EPA. 1988b. Selection Criteria for Mathematical Models Used in Exposure Assessments:
Surface Water Models. Office of Health and Environmental Assessment, Washington, DC.
EPA/600/8-87/042.
U.S. EPA. 1988c. Graphical Exposure Modeling System: GEMS User's Guide. Prepared by
General Sciences Corporation for the Office of Pesticides and Toxic Substances, Exposure
Evaluation Division. Contract No. 68-02-4281.
U.S. EPA. 1988d. Background Document for EPA's Composite Landfill Model (EPACML).
Prepared by Woodward-Clyde Consultants, Inc. for the Office of Solid Waste, Washington, DC.
U.S. EPA. 1988e. Drinking Water Regulations; Maximum Contaminant Level Goals and National
Primary Drinking Water Regulations for Lead and Copper. Federal Register 53(160): 31516.
U.S. EPA. 1989a. Risk of Unsaturated/Saturated Transport and Transformation Interactions for
Chemical Concentrations (RUSTIC), Volume 1: Theory and Code Verification. Prepared by
Woodward Clyde Consultants, HydroGeologic, and AQUA TERRA Consultants for the Office of
Research and Development, Environmental Research Laboratory, Athens, GA. Contract No.
68-03-6304.
U.S. EPA. 1989b. RUSTIC Documentation, Volume II: User's Guide. Prepared by Woodward- .
Clyde Consultants, HydroGeologic, and AQUA TERRA Consultants for the Office of Research
and Development, Environmental Research Laboratory, Athens, GA. Contract No. 68-03-6304.
U.S. EPA 1989c. Background Document for the Surface Impoundment Modeling System (SIMS).
Control Technology Center. Research Triangle Park, NC. EPA/450/4-89/013b.
U.S. EPA. 1989d. Development of Risk Assessment Methodology for Land Application and
Distribution and Marketing of Municipal Sludge. Prepared by the Office of Health and
Environmental Assessment, Environmental Criteria and Assessment Office, Cincinnati, OH for
the Office of Water Regulations and Standards, Washington, DC. EPA/600/6-89/001. NTIS
PB90-135740/A5.
U.S. EPA. 1989e. Exposure Factors Handbook. Office of Health and Environmental Assessment.
EPA/600/8-89/043.
U.S. EPA. 1990. Integrated Risk Information System (IRIS). Online. Office of Health and
Environmental Assessment, Environmental Criteria and Assessment Office, Cincinnati, OH.
U.S. Geological Survey. 1967. Roughness Characteristics of Natural Channels. U.S. Geological
Survey Water Supply Paper 1849. US Government Printing Office, Washington, DC.
U.S. Geological Survey. 1982. Streamflow Characteristics Related to Channel Geometry of
Streams in Western United States. U.S. Geological Survey Water Supply Paper 2193. U.S.
Government Printing Office, Washington, DC.
U.S. Geological Survey. 1983. Sediment Data for Mid-Arkansas and Upper Red River Basins
through 1980. U.S. Geological Survey Open File Report 83-692. U.S. Government Printing
Office, Washington, DC.
U.S. Geological Survey. 1984. National Water Summary 1983: Hydrologic Events and Issues.
U.S. Geological Survey. Water Supply Paper Number 2250. U.S. Government Printing Office,
Washington, DC.
8-5
-------
U.S. Geological Survey. 1988a. Statistical Summaries of Streamflow Records in Oklahoma and
Parts of Arkansas, Missouri, and Texas Through 1984. U.S. Geological Survey Water Resources
Investigations Report 87-4205. Oklahoma City, OK.
U.S. Geological Survey. 1988b. Ground-Water Flow and Quality Beneath Sewage-Sludge
Lagoons, and a Comparison with the Ground-Water Quality Beneath a Sludge-Amended Landfill,
Marion County, Indiana. Prepared in cooperation with the Indianapolis Department of Public
Works. Indianapolis, IN. Water-Resources Investigations Report 88-4175.
van Genuchten, M.T. 1985. Convective-dispersive transport of solutes involved in sequential
first-order decay reactions. J. Computers Geosci. 11:129-147.
van Genuchten, M.T. and P.J. Weirenga. 1976. Mass transfer studies in sorbing porous media. I.
Analytical solutions. Soil Sci. Am. J. 40: 473-480.
Veith, G.D., D.L. Foe and B.V. Bergstedt. 1979. Measuring and'estimating the bioconcentration
factor of chemicals in fish. J. Fish Res. Board Can. 36: 1040-1048.
Veith, G.D., K.J. Macek, S.R. Petrocelli and J. Carroll. 1980. An evaluation of using partition
coefficients and water solubility to estimate BCFs for organic chemicals in fish. In: Aquatic
Toxicology. J.G. Eaton, P.R. Parrish and A.C. Hendricks, Ed. ASTM STP 707. p. 116-129.
Yeh, G.T. 1981. AT123D: Analytical Transient One-, Two-, and Three-Dimensional Simulation
of Waste Transport in the Aquifer System. ORNL-5602. Environmental Sciences Division, Pub.
No. 1439. Oak Ridge National Laboratory, Oak Ridge, TN.
*US.GO\TERNMENTPRINTINGOFFICE:! 990 -718-1 5^20 1*6
8-6
------- |