&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

-------