Draft



          Regulatory  Impact Analysis



for the Final Rulemaking  on Corrective Action



      for Solid Waste Management  Units
      Proposed Methodology for Analysis
                APPENDICES
            Office of Solid Waste




    U.S. Environmental  Protection  Agency
                 March 1993
                                        Printed on Recycled Paper

-------
                              TABLE OF CONTENTS
EXECUTIVE SUMMARY

1.     INTRODUCTION	 1-1
        1.1   The Need for Regulation	 1-1
        1.2   Description of Corrective  Action Regulatory Impact Analysis  	 1-5
        1.3   RIA Organization	 1-6

2.     REGULATORY OPTIONS  	2-1
        2.1   Description of the Baseline  	2-1
        2.2   Description of the Subpart S Proposed Rule 	2-2
        2.3   Other Options to be Examined  	2-3

3.     SAMPLE SELECTION, FACILITY CHARACTERIZATION, AND MODELING OF
      RELEASES	3-1
        3.1   Approach	 3-1
        3.2   Results	 3-27
        3.3   Limitations	 3-39

4.     REMEDY SELECTION AND MODELING OF REMEDY EFFECTIVENESS ... 4-1
        4.1   Background  	4-1
        4.2   Approach	4-2
        4.3   Results	,:	 4-14
        4.4   Limitations	 4-22

5.     COSTS 	5-1
        5.1   Approach	5-1
        5.2   Results	5-7
        5.3   Sensitivity Analysis	 5-23
        5.4   Largest Federal  Facility Cost Analysis  	 5-29
        5.5   Limitations	 5-32

6.     OVERVIEW OF BENEFITS  	 6-1
        6.1   Potential Benefits	 6-1
        6.2   Studies Conducted for this RIA	 6-6

7.     HUMAN HEALTH BENEFITS	7-1
        7.1   Background  	7-1
        7.2   Approach	7-4
        7.3   Results	 7-27
        7.4   Limitations	 7-56

-------
                       TABLE OF CONTENTS (CONTINUED)
8.     ECOLOGICAL BENEFITS	8-1
        8.1   Approach	8-1
        8.2   Results	8-6
        8.3   Discussion 	 8-14
        8.4   Limitations	 8-14

9.     AVERTED WATER USE COSTS  	9-1
        9.1   Economic Framework  	9-2
        9.2   Analytic Approach	9-8
        9.3   Results	 9-15
        9.4   Limitations	 9-29

10.    NONUSE BENEFITS OF GROUND WATER REMEDIATION 	 10-1
        10.1 Approach	 10-2
        10.2 Results	  10-29
        10.3 Sensitivity Analyses  	  10-31
        10.4 Limitations	  10-36

11.    RESIDENTIAL PROPERTY ANALYSIS  	 11-1
        11.1 Expected Linkages Between TSDFs and Residential
             Property Markets	,	 11-1
        11.2 Statistical Specifications and Data Employed in
             Current Analysis  	 11-5
        11.3 Overview of Findings	  11-10
        11.4 Conclusions 	  11-17

12.    CHANGES IN THE VALUE OF FACILITIES 	 12-1
        12.1 Economic Framework  	 12-2
        12.2 Analytic Approach	 12-7
        12.3 Example of Facility Value Benefit Calculations  	  12-15
        12.4 Results	  12-17
        12.5 Sensitivity Analysis	  12-20
        12.6 Limitations	  12-22

13.    COMPARISON OF BENEFITS AND COSTS 	 13-1
        13.1 Introduction	 13-1
        13.2 Review of Costs and Benefits	 13-2
        13.3 Comparison of Costs and Benefits	 13-6
        13.4 Limitations	 13-9

BIBLIOGRAPHY

-------
                      TABLE OF CONTENTS (CONTINUED)
APPENDIX A.  DEVELOPMENT OF FACILITY SAMPLE	A-l
       A.1  Creating the Federal Frame	A-l
       A.2  Creating the Non-Federal Sample Frame	A-3
       A.3  The Full Corrective Action Sample  	A-8

APPENDIX B.  PREDICTING RELEASES AND EXPOSURES WITH MMSOILS	B-l
       B.1  Model Selection	B-l
       B.2  Parameter Selection and Assumptions	B-3
       B.3  Model Application Assumptions and Limitations	B-13
       B.4  Post-MMSOILS Processing  	B-17

APPENDIX C.  SIMULATION OF REMEDY EFFECTIVENESS	C-l
       C.I  Source Control Technologies	C-l
       C.2  Waste Treatment Technologies  	C-3
       C.3  Ground-Water Remediation Technologies	C-7

'APPENDIX D.  COST ANALYSIS  	D-l
       D.I  Expert Panel Cost Estimation  	D-l
       D.2  Additional Analysis of Results	D-13

APPENDIX E.  HUMAN HEALTH BENEFITS ANALYSIS	E-l
       E.1  Hazard Identification and Dose-Response Assessment	E-l
       E.2  Exposure Analysis  	E-6
       E.3  Risk Characterization	E-25

APPENDIX F.  ECOLOGICAL THREATS: METHODOLOGIES AND
              CASE STUDIES	F-l
       F.I  Methodology for Proximity Analysis	F-l
       F.2  Methodology for Deriving Screening Ecological
              Benchmark Levels	F-l
       F.3  Methodology for Estimating Extent  of Contamination  	F-8
       F.4  Proximity Analysis Results	F-9
       F.S  Concentration-Based Screening Analysis Results  	F-9
       F.6  Time Sequence Results  	F-9
       F.7  Extent of Contamination Results  	F-9
       F.8  Qualitative Case Studies	F-17

APPENDIX G.  KEY PARAMETERS MATRIX	G-l

APPENDIX H.  FACILITY AND SWMU DATA FORMS  	H-l

APPENDIX I. CHARACTERISTICS OF FACILITY AND SWMU
              POPULATIONS  	  1-1
       I.I   Facility Characteristics 	  1-1
       1.2   SWMU Characteristics	1-17

-------
                  APPENDIX A. DEVELOPMENT OF FACILITY SAMPLE1


        This appendix provides additional detail on the process EPA used to identify the frame of
 facilities subject to the Subpart S Corrective Action requirements. It documents the federal and
 non-federal facility sampling frame creation, the sampling design, and the sample  allocation and
 selection processes.

 A.1 Creating the Federal Frame

        In the winter of 1991, EPA used the 1990 Inventory of Federal Agency Hazardous Waste
 Facilities (Inventory) to identify the universe of federal facilities  potentially subject to corrective
 action under RCRA. Based on this  universe (i.e., the sampling frame), EPA selected the federal
 facility sample for the corrective action RIA.

        According to facility responses to Part III of the Inventory2, 395 federal  facilities were
 identified as being potentially subject to corrective action. To verify this characterization of the
 sampling frame, EPA presented the preliminary list of facilities to officials in each of the relevant
 Federal departments to identify any facilities that should not be listed among this universe (i.e., are
 not subject to corrective action). As a result of this verification process, 36 of the 395 facilities were
 deleted for a corrected population of 359 federal facilities potentially subject to corrective action.
 Because the Inventory  represents data from 1990 and appeared  to be more complete than  the
 Hazardous Waste Data Management  System (HWDMS), EPA used.the Inventory-derived list as the
 sampling frame for the corrective action RIA.

        EPA grouped the 359 facilities in the universe into three size categories,  (very large, large,
 and other)  defined  by the  magnitude  of potential costs of corrective actions.  Facilities with
 estimated costs greater than $1 billion were classified as "very large"; those with estimated costs
 between $100 million and $1 billion were classified as "large". The remaining facilities were classified
 as "other." Exhibit A-l characterizes  the 359 federal facilities subject to corrective action by these
 three size categories and Federal  departments.

       A.1.1 Determining the Federal Sample Size and Allocation

       The selection of the federal facilities sample underwent several changes before arriving at
 the final sample size. From the 359 federal facilities in the universe, 30 were initially randomly
 selected, across strata, for the analysis.
      For more information on the sample plan see Selecting Samnle Facilities for the Corrective
Action Regulatory Impact Analysis; Draft Report prepared for U.S. EPA by Research Triangle
Institute. Research Triangle Park, NC:  June 1991.

   2    Part in contains information on RCRA treatment, storage, and disposal facilities that
managed hazardous waste on or after November 19,1980.
                             » »
                                 DRAFT - March 24,1993 • • •

-------
                                          A-2
                                      EXHIBIT A-l

                DISTRIBUTION OF FEDERAL FACILITIES POTENTIALLY
           SUBJECT TO CORRECTIVE ACTION BY SIZE AND DEPARTMENT

Department
Defense
Energy
Other
Total by Size Category
Size Category
Very Large
2
7
0
9
(3%)
Large
9
13
0
22
(6%)
Other
284
14
30
328
(91%)
Total By
Department
295 (82%)
34 (10%)
30 (8%)
359
(100%)
       Following the selection of the 30 federal facilities, a pilot phase of the study was conducted.
The experiences gained during the  pilot phase convinced the Agency of the impracticability of
providing expert panel review for a large number of facilities given the time constraints of the study.
Furthermore, it became clear that the number of facilities for which adequate data existed for the
analysis had been overestimated  and that several  facilities that may not be subject to RCRA
authority had been inaccurately included in the population.

       The Agency determined that  inadequate data existed for the analysis if, after contacting the
appropriate regional office or federal agency, either no facility documents were available or the
existing documents were of poor quality. For example, lists of the SWMUs and their locations were
often unavailable.

       The facilities which may have been inaccurately included in the population fell  into two
categories (1) protective filers, and (2) ninety-day converters. Protective filers represent those firms
that did not manage hazardous wastes as of May 19,1980, and were therefore not required to submit
a RCRA Part A permit application notifying EPA of hazardous waste activity, but that nevertheless
filed a Part A application to EPA as a legal  precaution.  Ninety-day converters represent those
facilities that stored hazardous wastes for longer than 90 days at the time of their permit application
(and were therefore subject to permitting as a storage facility), but who subsequently changed their
waste storage practices to less than 90 days and no longer require a hazardous waste storage permit.
While protective  filers and ninety-day converters are  not subject  to  the RCRA permitting
requirements, some of these facilities may be subject to RCRA corrective action authorities under
some circumstances.
                            * * * DRAFT •- March 24,1993
                                                         * * *

-------
                                            A-3
       To ensure that this RIA does not understate the population of facilities that may be required
to conduct corrective action, protective filers and ninety-day converters were included in the universe
but were  excluded from the sample.  Facilities that  did not have adequate data to conduct the
analysis were also left in the population but were excluded from the sample. Excluding the facilities
with inadequate data decreased the sample size and therefore allowed the Agency to better analyze
the facilities in the sample.  As a result, however, error was introduced into the sample in the form
of non-response error. The exclusion of these facilities from the sample resulted in a final federal
sample of 9 facilities.

       A.1.2 Description of the Federal Sample

       The 9 federal facilities in the final  sample included :

              0 "Very Large" DOD/DOE Facilities;
              3  "Large" DOD/DOE Facilities; and
       •      6 "Other" Facilities.
       No "Very Large" DOD/DOE facilities were included in the final sample for two reasons.
First, because of their large size and complexity, their inclusion in the analysis would have greatly
reduced the number of non-federal and other federal facilities that the Agency would have been able
to include  in the RIA.  Second, the agencies responsible for these nine  facilities  are currently
expending a high level of effort towards their characterization and remediation, and therefore, are
best equipped to estimate the potential corrective action costs and benefits at each of these facilities.
In order to maximize the sample size, these facilities were excluded from the sample of facilities
analyzed.

       EPA weighted the sample with a higher percentage of "large" facilities (33 percent) than is
present in the universe (6 percent). This weighting was done for two reasons.  First, the universe
of facilities that will require corrective action is likely to contain a larger proportion of facilities in
this size category than the universe of all  federal facilities.  Second, because these "large" facilities
are expected to drive the cost and benefit analyses, sampling them at a  higher rate provides more
precision in these groups, thus making the analysis more accurate.
A.2 Creating the Non-federal Sample Frame

       Creating the sampling frame involved obtaining lists of facilities from  EPA databases,
evaluating the suitability of the resulting set of facilities, and revising both the databases and the
selection criteria used. During this exploration, several databases were accessed to locate facilities
belonging to the sampling frame.  The components of this frame were three  EPA databases:
HWDMS; the Corrective Action Reporting System (CARS); and the Resource Conservation and
Recovery Information System (RCRIS) which now encompasses both HWDMS and CARS.
                            • • * DRAFT - March 24, 1993 * * *

-------
                                            A-4
       HWDMS, CARS, and RCRIS were accessed to identify the universe of facilities potentially
subject to corrective  action.  Facilities in one or more of the following seven  categories were
considered potentially subject to corrective action and assigned to the sampling frame:

       •      Treatment, storage, or disposal facilities;

       •      A facility that previously stored and/or treated hazardous waste but is no longer pan
              of the universe;

       •      Permit by rule facilities (e.g., underground injection facilities and publicly owned
              treatment works (POTWs));

       •      Clean-closed facilities; and

       •      Other facilities with specific permit status not listed above.

Applying these criteria yielded a universe of 8,284 facilities potentially subject to corrective action.
A flag variable was used to identify and delete Federally-owned facilities.  The resulting sampling
frame of non-federal facilities potentially subject to corrective action contained 5,432 facilities.

       A.2.1 Choosing the Non-federal Sampling Design

       Selecting a sample of facilities from the sampling frame Involved several steps.  First, the
Agency developed a sample design based on the key population parameters of interest.  Then, given
specified precision  requirements  for the selected parameter estimates, the sample allocation was
determined.  Finally, EPA used a randomization procedure based on the sampling design to obtain
the distribution of sample facilities specified in the  allocation.
             <7
       EPA considered several factors in  deciding to employ a stratified, one-stage cluster design
for the non-federal  facility sample.  First, greater precision was desired in the parameter estimates
for large facilities, which were  expected to incur higher corrective action costs ceterisparibus. At the
same time, EPA wanted to minimize  the cost of collecting the data needed to  make parameter
estimates for each type of facility. For these reasons, EPA decided to use strata defined by facility
size and data availability.

       In order to  stratify by data availability, EPA decided to make use of data collected during
RCRA facility assessments (RFAs). Typically, before a permit is issued, EPA or an authorized state
would conduct a RCRA Facility Assessment (RFA) to determine whether a potential problem exists
(the RFA is  analogous to the  Superfund Preliminary Assessment/Site Investigation [PA/SI]). If a
likely release is found,  the  permit  would contain a schedule of compliance requiring  the
owner/operator to conduct a Remedial Field Investigation (RFI) to characterize the nature and
extent of contamination. Because much of the information required for this RIA is collected during
an RFA, data collection costs were expected to be much lower for facilities with RFAs completed.
Furthermore, since EPA regions tend to complete RFAs for the highest priority  facilities first, the
                            • * * DRAFT - March 24, 1993
                                                           * * *

-------
                                            A-5
facilities with RFAs completed could be assumed to be among those most likely to trigger corrective
action.  Following this rationale, the Agency established three strata based on size and RFA status.
Within  each stratum, sample facilities were selected with equal conditional probabilities, given the
stratum, and without replacement.

        The facilities identified as "Large" formed one stratum. These are the facilities that were
identified by EPA Regions as being the largest (with estimated costs between $100 million and 1$
billion). The remaining 5,345 "Not Large" facilities were divided into two strata:

        •      Those with RFAs completed; and
        •      Those that do not have RFAs completed.

For these two strata 1,711 facilities were "Not Large" and had a completed RFA, and 3,634 facilities
were "Not Large" and had no RFA.

        To further characterize the sampling frame facilities, EPA also profiled them according to
•industry group, using the facilities' 2-digit Standard Industrial Classification (SIC) code. However,
SIC code  49, was divided into 4953, Refuse Systems, and 49 Other as SIC 4953 is frequently used
by commercial hazardous waste management facilities. Facilities for which there was no SIC code
information were grouped into "Missing/Other."

        Although stratifying by SIC code was not practical given the overall sample size planned for
the survey, EPA wanted to control the distribution of the sample with respect to the industries
represented in each stratum.  To provide this control, a sequential selection procedure was used to
select the sample.  That is, dividing the sampling frame into strata, each stratum was sorted by the
SIC code groupings (industry groups) shown in  Exhibit  A-2.  Then the Agency used a selection
procedure that selected facilities over the entire sequence of SIC codes within each stratum. This
assured a proportional representation of industries in the sample frame up to the limitations imposed
by the stratum.

        A.2.2 Determining the Non-federal Sample Size and Allocation

        The optimal sample allocation procedure minimizes the total variable costs of the study,
while meeting predetermined  precision requirements (variance  constraints).   EPA  produced
estimates of the average cost of collecting data from a facility for each of the  strata. In addition,
EPA determined the level of precision  required for estimates for facilities in the "Large" stratum
and overall. Using this information, a system of equations was developed that described the variable
costs and the variances in terms of the characteristics  of the sampling design and the sample sizes
of the strata. These equations were solved simultaneously to determine the sample size for each
stratum that would minimize the total variable cost of conducting the study while achieving the
desired  level of precision.  The solution of this system of equations resulted  in the stratum-level
sample  sizes shown in Exhibit A-3.
                               • DRAFT - March 24, 1993 * • *

-------
               EXHIBIT A-2
NON-FEDERAL SAMPLING FRAME BY SIC CODE
INDUSTRY GROUP
Milling/Other
28
Chemical ProducU Manufacturing
29
Petroleum and Coal ProducU
33
Primary MeUls Manufacturing
34
Fabricated Mclal* Manufacturing
35
Nonelectrical Machinery
36
Electrical and Electronic
Machinery and Equipment
37
Transportation Equipment
49, Other
Electric, Gas, and Sanitary Services
4953
Refuse Syitems
50
Wholesale Trade
Stratum Total
LARGE
12-
(0.6)'
(13-8)'
(0.2)'
35
(3.0)
(40.2)
(0.6)
16
(6.9)
(18.4)
(03)
5
(1.4)
(5.8)
(0.1)
1
(0-2)
(1.2)
(0.0)
0
(OX))
)
2
(0.4)
(23)
(0.0)
10
(«)
(11.5)
(02)
0
(Od)
(OJ))
(0.0)
6
(3.9)
(6.9)
(0.1)
0
(0.00)
(OM)
(OM)
87
(U)
•NOT
RFA
515
(M-8)
(30.1)
(9.5)
430
(373)
(25.1)
(7.9)
123
(52*)
C-2)
(23)
161
(43.9)
(9.4)
(2.9)
100
(22-8)
(5*)
(1.8)
56
(23.9)
(33)
(1-0)
82
(17.6)
(«)
(»-S)
57
(23-8)
(3J)
(10)
62
(28J)
(3-6)
(1.1)
88
(56-8)
(5.1)
0-6)
37
(26*)
(")
(0.7)
1711
(31.5)
LARGIT
NO RFA
1262
(70-S)
(34.7)
(23i)
689
(59.7)
(18.9)
(12.7)
94
(403)
(2A)
(1.7)
201
(54*)
(5.5)
(3.7)
338
(76.9)
(93)
(«)
178
(76.0)'
(4.9) --
(3J)
383
(82.0)
(10.5)
(7.1)
173
(72.1)
(4-8)
(3.2)
153
(71-2)
(«-2)
(18)
61
(39^)
(1.7)
(I.I)
102
(73.4)
(")
(1.9)
3634
(66.9)
INDUSTRY TOTAL
(INDUSTRY PERCENT)
1789
(32.93)
1154
(2124)
233
(1-2°)
367
(676)
439
(6M)
234
(431)
467
(8.60)
240
(4.42)
215
(3.96)
155
(185)
139
(2S6)
5432
(100)
      • • * DRAFT -• March 24,1993 • • *

-------
                                          A-7
* In each row, the first number is the number of facilities in the sampling frame.
b In each row, the second number represents that cell's percentage of the row, for example, 12/1789.
c In each row, the third number represents that cell's percentage of the column, for example, 12/87.
d In each row, the fourth number represents that cell's percentage of the table, for example, 12/5432.
                          • • • DRAFT - March 24,1993 * * •

-------
                                           A-8
                                      EXHIBIT A-3

                     INITIAL NON-FEDERAL SAMPLE ALLOCATION
Stratum
"Large" facilities
"Not Large" facilities with RFAs
"Not Large" facilities without
RFAs
Total
Number of Facilities
56.13
54.86
56.56
167.55
       Following the random selection of the 167 non-federal facilities, EPA conducted the pilot
phase of the study to test the methodology. As described in Section A.I.I (Determining the Federal
Sample Size and Allocation), the experiences gained during the pilot phase convinced the Agency
of the impracticability of providing expert panel review for a large number of facilities given the time
constraints of the study. Subsequently, the number of non-federal facilities was randomly dropped
to 70, distributed across the three strata.

       A.2J  Description of the Non-federal Sample

       The sample of 70 non-federal facilities was allocated among the three strata and selected as
described in Section A.2.1. The sample of 70 included 26 facilities from the "Large" stratum, 27
from the stratum of facilities that are "Not Large" and have a completed RFA, and  17 from the
stratum of facilities that are "Not Large" and do not have a completed RFA.

       As is the case among hazardous waste management facilities in general, the largest share of
the sample came from SIC code group 28, Chemicals Manufacturing. SIC code 29, Petroleum and
Coal Products, accounted for the next largest share. And, roughly one-fourth of the sample either
had no known SIC code or fell into an industry group other than the ones specified by EPA. Thus,
the distribution of the  sample facilities among  SIC codes reflected the distribution of hazardous
waste management facilities generally.

AJ The Full Corrective Action Sample

       The sample of facilities studied in analyzing the impact of the Subpart S Corrective Action
Rule comprises two separately selected frameworks: the federal facility frame and the non-federal
facility frame.  The sample frames were constructed separately  for these two samples, and the
samples were selected using different sampling designs and sample allocations.  Of the 359 federal
and 5,432 non-federal facilities in the overall population, 9 federal and 70 non-federal facilities were
                            • • • DRAFT « March 24,1993

-------
                                            A-9
selected for the final corrective action sample. Exhibit A-5 summarizes the number of facilities in
the corrective action sample and the associated number in the overall population. For population-
level data, the designation of facilities either requiring corrective action or needing no further action
was estimated based on the sample results. The exhibit presents the facility weights for each stratum
that were used to extrapolate the facility-level data to national results.
                           * • • DRAFT - March 24, 1993 * * •

-------
                                                    A-IO
                                               EXHIBIT A-4
                         FACILITIES SUBJECT TO CORRECTIVE ACTION BY STRATA

Sample
CAR1
NFA4
Total
Population Weight per
Sample Facility
Population
CAR
NFA
Total
Very
Large
DOD/DOE
0
0
0
n/a
9
0
9
Large
DOD/DOE
3
0
3
7.3
22
0
22
Other
5
1
6
54.7
273
55
328
Federal
Total
8
1
9
n/a
304
55
359
Large
26
0
26
3.3
87
0
87
Not
Large
RFA
19
8
27
63.4
1204
507
1711
Not Large
No RFA
6
11
17
213.8
1283
2351
3634
Non-
Federal
Total
51
19
70
n/a
2574
2858
5432
TOTAL
59
20
79
n/a
2878
2913
5791
3 Designates facilities where corrective action was required.



4 Designates facilities requiring no further action.



                                       •*• DRAFT - March 24,1993 •••

-------
           B.  PREDICTING RELEASES AND EXPOSURES WITH MM SOILS

       Appendix B provides supplementary information concerning the prediction of constituent
releases, transport, and exposures. Section 1 discusses the model selection process.  Section 2
discusses MMSOILS parameter selection and assumptions.  Section 3 discusses model application
assumptions and limitations. Finally, Section 4 discusses post processing of the model outputs,
particularly the aggregation of SWMU-specific data into facility-wide estimates.  This appendix
does not provide a detailed explanation of the MMSOILS model. For further information on
the MMSOILS model, refer to the model documentation.1

B.1 Model Selection

       This section discusses the criteria and process for selecting a fate and transport model for
the RIA analyses.2  EPA reviewed many models for potential use in the RIA:

             LLM;
             MMSOILS;
             SLUDGEMAN/SLAPMAN;
             EPACML/EPACMS;
             UST;
             TRAM;
             FECTUZ/SAFT3D;
             ISO
             MICHRIV;
             EXAMS; and
             RAMMS.  .

The Agency evaluated the models using  three broad, interrelated criteria to determine the
appropriate level of complexity for the corrective action RIA: objectives criteria, implementation
criteria, and technical criteria. The MMSOILS model was selected  for several reasons:

       •      The model comprehensively addresses multimedia chemical release, fate,
             transport, and exposure;

       •      The mode] can simulate failure and constituent release from a variety of waste
             management units found  at RCRA facilities;

       •      EPA could obtain the input data required by the model within the Agency's time
   1 U.S. Environmental Protection Agency. MMSOILS: Multimedia Contaminant Fate.
Transport, and Exposure Model — Documentation and Users Guide. Draft.  Washington, D.C.:
U.S. Environmental Protection Agency, 1992.

   2 See also: U.S. Environmental Protection Agency/ORD/ERL Athens. Summary of Review
and Evaluations of the Technical Approach to Corrective Action Regulatory Impact Analysis:
Fate and Transport. Washington, D.C.: U.S. Environmental Protection Agency, 1992.

                            *** DRAFT: March 22,1993 •**

-------
                                          B-2
              and budget constraints;

       •      The screening-level complexity of the model was appropriate for the analysis of a
              large sample of facilities; and

       •      EPA had already applied the model and was familiar with its strengths and
              weaknesses.

MMSOILS can simulate releases from five type of units commonly found at RCRA facilities:

       •      Surface impoundments;
       •      Landfills;
       •      Waste piles;
       •      Tanks; and
       •      Underground injection wells.

The model simulates the fate and transport of constituents released from these units within four
main pathways — atmospheric, surface water, ground water, and soil erosion. The model also
tracks the mass balance of constituents released through the pathways and accounts for cross-
media transfers, such as ground-water discharge to surface water and atmospheric deposition of
contaminants to the soil.  MMSOILS also simulates the foodchain/bioaccumulation pathway for
estimating concentrations of contaminants in fish,  crops, meat, and milk.

       MMSOILS was initially developed as a screening tool  to assist EPA in setting priorities
for hazardous waste management at Superfund sites.  It was later modified to model multiple
chemicals released from RCRA land-based  units and underground injection wells. Following
extensive review for the corrective action RIA,3 the Agency made numerous additional
modifications (e.g., improved mass balance) to improve the accuracy and flexibility of the
model.4
   3  U.S. Environmental Protection Agency/ORD/ERL Athens. Summary of Review and
Evaluations of the Technical Approach to Corrective Action Regulatory Impact Analysis: Fate
and Transport. Washington. D.C: U.S. Environmental Protection Agency, 1992.

   4 See also U.S. EPA, MMSOILS: Multimedia Contaminant Fate. Transport, and Exposure
Model — Documentation and Users Guide. Draft.  Washington, DC: September 1992.

                             ••• DRAFT: March 22,1993 ••*

-------
                                           B-3
B.2 Parameter Selection and Assumptions

       This section discusses the selection of MMSOILS model parameters for the RIA. The
first subsection provides a listing of all the model parameters. The second subsection discusses
the distinctions between central  tendency and high-end parameter values.  Finally, the third
subsection discusses the model used to estimate tank release rates.

       B.2.1  Input Parameters

       Exhibit B-l below presents the parameters contained in the input files to the MMSOILS
model.  The parameters are organized into sets corresponding to different input files. Under any
given set, each parameter may fit into one of three categories: (1) SWMU-specific parameters;
(2) facility-specific parameters; or (3) global parameters that are  constant over the entire
modeling effort, including those  in the chemical property database (see Appendix E) or soil
property table.5 The list does not include those parameters that are only used for control
purposes (e.g., beginning and ending times) or those that are not used for RIA modeling
calculations.6
   5 U.S. Environmental Protection Agency. Superfund Exposure Assessment Manual. Draft
Report. Washington, D.C.: U.S. Environmental Protection Agency, 1987.

   6 Additional information on input parameters for both transport and risk modeling may be
found in Appendix G, and facility- and SWMU-specific data collections forms may be found in
Appendix H.

                             •*• DRAFT: March 22,1993 ••*

-------
                                               B-4
                   EXHIBIT B-l MMSOILS INPUT PARAMETERS
Constituent-specific Chemical Properties
Henry's Law Coefficient
Molecular  Diffusivity  in Air
Molecular  Diffusivity  in Water
Vapor Pressure
Molecular  Weight
Adsorption  Coefficient on
Organic Carbon,  Koc
Landfill Leachate  Concentration
Concentration  in Soil or Fill
Concentration  in Surface
Pore Liquid
Reference  Dose Level,  Oral Route
Reference  Dose Level,
Inhalation  Route
Carcinogen  Potency  Factor,
Oral Route
Carcinogen  Potency  Factor,
Inhalation  Route
Skin Permeability Factor
Kd for waste materials
Kd for ground water flow
Decay  rate
Concentration  in Leachate
Concentration  in Influent
Concentration  in Injection Well
Bioconceniration  in  Fish
Transfer  Factor  for Cattle
Uptake  from Soil to Plant
Soil Moisture  to Root  Factor
Action Level for Air
Action Level for Soil
Action Level for Ground  Water
Action Level for Surface  Water
From chemical  database.
From chemical  database.
From chemical  database.
From chemical  database.
From chemical  «tat«ha«»
From chemical  database.

For the central  tendency, EPA used the Organic  Leachate  Model (OLM)
to obtain leachate concentrations  for organics as a function  of the
chemical concentration  in the unit (from the SWMU  data form) and the
solubility (from the  chemical database).  For the high-end,  EPA used
chemical database values for solubility   For  inorganics,  EPA took central
tendency and high end values from the database.  These values are
entered in the input file as the solubility limit.
From SWMU data form.  Used for landfills.
From SWMU data form.  Used for volatilization

From chemical  database
From chemical  database.

From chemical  database

From chemical  database.
From chemical  database.
From chemical  database.
From chemical  database.
From chemical  database.
From SWMU  data form.
From SWMU  data form.
From SWMU  data form.
From chemical  database.
From chemical  database.
From chemical  database.
From chemical  database.
From chemical  database.
From chemical  database.
From chemical  database.
From chemical  database.
Varies with pH
Varies with pH.
Used only for  tank releases.
Used only for  surface  impoundments
Used only for  injection wells
Atmospheric Pathway Parameters

Volatilization  Model  Flag

Depth  of Clean  Cover
From SWMU  data form.  Depends on whether waste is covered or
exposed.
From SWMU  data form.
                              *•• DRAFT: March 22,1993 •••

-------
                                               B-5
                                    EXHIBIT B-l  (ConL)
Depth of Contamination
Average  Temperature
Mizing Depth of Site Soil
Deposition  Velocity  of Particulates
Wind Velocity
Threshold Friction Velocity
Fraction  of Vegetative Cover
Roughness  Height
Silt Content of Road Surface
Mean Vehicle Speed
Mean Vehicle Weight
Number  of Wheels  per Vehicle
Number  of Days  per Year With
at Least  0.01 Inches  Rain
Vehicle  Kilometers  on
Unpaved Surface
Drop Height for  Loading
Operations
Capacity  of  Loading  Machinery
Daily Load of Material
Area  of Daily Spreading
Operations
Distance  to  Model Observation
Point
Number  of Stability  Classes
Stability  Class Wind  Velocity
Stability  Class Type
Stability  Class Frequency
 From SWMU  data  form.
 From facility data form
 Global
Global.
 From facility data form.
 Global.
 From SWMU  data  form.
 From facility data form.
 Global.
 Global.
 Global
 Global
 From facility data form.

 From SWMU  data  form  Function of SWMU  area

 Global.

 Global.
 From SWMU  data  form  Function of SWMU  area.
 From SWMU  data  form.  Function of SWMU  area.

 Varies with  facility size, SWMU location, etc

 Global
 Global
 Global
 Global.
Surface Water Pathway Parameters
Flow  Rate  in River
Downstream  Distance  to
Point  of Use
Average River Velocity
Distance from SWMU to River
USLE Rainfall  Factor
USLE Soil Erodibility  Factor
USLE Length-Slope  Term
USLE Cover Factor
USLE Erosion  Control Factor
Ground Water  Discharge  to
River Flag
Sediment Delivery Fraction
to Lake
 From facility data  form.
 Global.  Equal  to point of discharge.
 From facility data form
 From SWMU  data  form.
 From facility data form
 From facility data form.
 From SWMU  data  form.
 From facility data form.
 Global.
 From facility data form.
Function of soil properties.
 Function  of SWMU size.
 From facility data form.
                             ••• DRAFT: March 22,1993 ***

-------
                                                B-6
                                     EXHIBIT B-l (Cent)
Fraction  of Organic  Carbon
in Lake Sediments
Wind Velocity  Above  the Lake
Average  Water  Depth of Lake
Surface Area of Lake
Lake  Sediment   Porosity
Diffusion  Path  Length for
Contaminated  Sediments
Global.

From facility data form.
From facility data form.
From facility data form.
Global
Global.
Ground Water Pathway Parameters
Gradient
Hydraulic Conductivity
Aquifer Porosity
Aquifer Bulk Density
Fraction  of Organic Carbon
in Aquifer
Dispersivity  in  X-Direction
Dispersivity  in  Y-Direction
Dispersivity  in  Z-Direclion
Thickness of Aquifer
Times  of Ground  Water
Use  Points
Positions of Ground Water
Use  Points
Depths of Ground  Water
Use  Points
From facility data  form.
From facility data  form.
From facility data  form.
From facility data  form
From facility data  form.

         Global.
         Global.
         Global.
From facility data  form.
From SWMU data form.
Depends, .upon the start-up lime of the aquifer.
From SWMU data form.  Depends upon  the location of the SWMU
relative  to the facility boundary.
From facility data form for the central tendency.
Set to the lop of the aquifer  for the high-end.
Infiltration and Meteorological Parameters
Field Capacity  of Surface Soil
Wilting Point of Surface Soil
Depth of Root  Zone
Number  of Soil Layers in
Unsaturated  Zone
Saturated Conductivity  for
Each Soil Layer
Saturated Water Content for
Each Soil Layer
Bulk Density for Each  Soil Layer
Exponent *b* for Moisture  Curve
in Each  Soil Layer
From facility data form.
From facility data form.
Global.
From facility data form

A toil type is specified in the facility data form.  Each soil type has
global properties.
Global soil properties  according to soil type.

Global soil properties  according to soil type.
Global soil properties  according to soil type.
                              ••• DRAFT: March 22,1993 ***

-------
                                                B-7
                                     EXHIBIT B-l  (ConL)
Percentage of Organic  Mailer  in
Each  Soil Layer
Percentage of Clay in
Each  Soil Layer
Percentage of Sill in
Each  Soil Layer
Percentage of Sand in
Each  Soil Layer
Decay Ratio for  Each Soil Layer
Depth of Each Soil Layer
Pan Factor for Convening
Ep to PET
Latitude  of Site
Curve Number for Surface  Soil
Monthly  Precipitation
Number  of Days  per  Month
with Precipitation
Monthly  Average  Temperature
Monthly  Pan Evaporation  (Ep)
Starting  Month of Growing Season
Ending Month of Growing  Season
From facility data form.

Global soil properties  according  to soil type

Global toil properties  according  lo soil type.

Global coil properties  according  to soil type.

Global.
From facility data form.  May be decreased  if SWMU is depressed
Global

From facility data form
From facility data form
From facility data form
         From facility data form

From facility data form.
From facility data form
From facility data form
From facility data form.
Foodchain Pathway Parameters
Deposition  Interception  Fraction
Pasture Production
Vegetable  Production
Soil  Intake  Rate for Cattle
Feed Rate for Cattle
Fraction of Cattle Feed
from Pasture
Water Intake  by Cattle
Sediment  Delivery Fraction  (SDF)
to Agricultural Field
Area of Agricultural  Field
SDF to Human Exposure  Field
Area of Human Exposure  Field
Fraction of Organic Carbon in
Offsite  Fields
Bulk Density  of Soil  in
Offsite  Fields
Depth  of Applied Irrigation
Source  of Irrigation  Water  From
Source  of Cattle  Water
Global
From facility data form.
From facility data form.
Global
Global.
Global

Global.
From facility data form.

From facility data form.
From facility data form.
From facility data form.
From facility data form.
From facility data form.

                  form.

                  form.
   From facility data
facility data form.
   From fad lily data
                               ••• DRAFT:  March 22, 1993

-------
                                                B-8
                                     EXHIBIT B-l (Cent.)
SWMU Description Parameters
Source of Human  Water  Intake
IVpe  of SWMU

SWMU Width
Number  of Constituents  of Concern
Global.
From SWMU data form.
impoundment,  or lank.
From SWMU data form.
From SWMU data form.
Choices are injection well,  landfill, surface
Landfill Parameters
Month when Capture  System Fails
Cover  Type
Liner Type


Thickness of Layer
Area  of Layer
Hydraulic Conductivity  of Layer

Field  Capacity  of Layer

Saturation  Limn of Layer

Capture  System Effectiveness
Flag for  Clay or Synthetic Layer

Month of Failure

Degree of Failure

Waste Layer Initial
Moisture  Content
Method  of Calculating  Leachate
Concentration
End of Waste  Additions
Incremental Area of Waste
Added each Year
Bulk  Density of the Waste
Global.
From SWMU data form.   Each type of cover has a different number of
layers.   Each layer of the  liner  system  has a set of characteristics  described
below.
From SWMU data form.   Each type of liner has a different  number  of
layers.   Each layer of the  liner  system  has a set of characteristics  described
below.
From SWMU data form.
From SWMU data form.   Equal  to SWMU area.
Based on global soil properties for soil and drainage  layers and on remedy
effectiveness  assumptions  for clay or synthetic layers  (see Appendix  C)
Based on global soil properties for-soil and drainage  layers and on remedy
effectiveness  assumptions  for clay or synthetic layers  (see Appendix  C).
Based on global soil properties for soil and drainage  layers and on remedy
effectiveness  assumptions  for clay or synthetic layers  (see Appendix  C)
Global.
From SWMU data form.   If a  layer is clay or synthetic,  it is assigned a
series of five failure events  as described  below.
The liming and  magnitude  of a liner or cover failure are based on globally
applied  deterministic  failure probabilities  (see Appendix  C).
The timing and  magnitude  of a liner or cover failure are based on globally
applied  deterministic  failure probabilities  (see Appendix  C).
Global.

Global.

From SWMU data form.
From SWMU data form.

Global.
                               •*•  DRAFT:  March 22, 1993 ••*

-------
                                               B-9
                                    EXHIBIT B-l (ConL)
Surface Impoundment  Parameters
Month when Capture System Fails
Cover  Type
Liner Type


Thickness  of Layer
Area of Layer
Hydraulic  Conductivity of Layer

Field Capacity of Layer

Saturation  Limit of Layer

Capture System  Effectiveness
Flag for Clay or Synthetic Layer

Month  of Failure

Degree of Failure

Year of Conversion  to Landfill
Initial Impoundment   Volume
Maximum   Impoundment  Volume
Monthly  Influx  to Impoundment
Area of Impoundment  Surface
Yearly  Evaporation
Waste Layer Initial
Moisture   Content
Method  of Calculating Leachate
Concentration
Bulk Density of the Waste

Tank Release Parameters
Global
From SWMU  data  form.  The model assumes that when an impoundment
is taken out of use it is similar to a landfill,  and may be covered  to prevent
infiltration.  Each type of cover has a different  number  of layers.  Each
layer of the cover system has a set of characteristics  described below.
From SWMU  data  form.  Each type of cover has  a different number of
layers.  Each layer of the cover system has a set of characteristics
described  below.
From SWMU  data  form.
From SWMU  data  form.  Equal to SWMU area
Based  on  global soil properties  for soil  and drainage layers and on remedy
effectiveness assumptions  for clay  or synthetic layers (see Appendix  C).
Based  on  global soil properties  for soil  and drainage layers and on remedy
effectiveness assumptions  for clay  or synthetic layers (see Appendix  C)
Based  on  global soil properties  for soil  and drainage layers and on remedy
effectiveness assumptions  for day  or synthetic layers (see Appendix  C)
Global
From SWMU  data  form.  If a layer is clay or synthetic,  it  is assigned a
series of five failure events as described below.
The  timing and magnitude  of a liner or cover  failure are based  on globally
applied deterministic  failure probabilities  (see Appendix C).
The  timing and magnitude  of a liner or cover  failure are based  on globally
applied deterministic  failure probabilities  (see Appendix C)
From SWMU  data  form.
From SWMU  data  form.  Taken  as 80% of the maximum
From SWMU  data  form
From SWMU  data  form.
From SWMU  data  form.
From facility data form.
Global.

Global

Global.
Number  of Years with Releases
Release Volume for Each Year
From SWMU  data form.
From SWMU  data form.  This may be calculated  using release  profile
models  that use the tank volume  or other characteristics  of the unit. See
Section  B3.1.
                              ••*
                                  DRAFT: March 22,  1993
                                                                 •**

-------
                                            B-10
                                   EXHIBIT B-l (Cent.)
       Injection Well Parameters

       •type of Well Failure             From SWMU  data form.
       Difference Between Water Table   From SWMU  data form
       and Injected Water Column
       Injection Rate                  From SWMU  data form.
       Pump Head                    From SWMU  data form.
       Type of Well                   From SWMU  data form.
       Flag for Best Case or Wont Case   From SWMU  data form.
       Solution
       B.2.2 Central Tendency and High-end Assumptions

       To account for uncertainties in the risk assessment process, EPA estimated risks for both
"central tendency" and "high-end" scenarios when conducting long-term modeling. The following
section discusses the data elements and assumptions that changed between central tendency and
high-end modeling scenarios.  Each change falls into one of three categories: global, facility-
specific, and SWMU-specific.  For each change, the source of the data and a brief description of
the nature of the change is presented.7

       Global Changes

       These global changes affect all of the facilities in a uniform manner.

       Soil/Water Partition Coefficient for Inorganics fKdV The Agency decreased Kd values
       for inorganics from mid-range values supplied by EPA's Office of Research and
       Development (ORD) to lower ORD-supplied values for the high-end. The lower Kd
       values will simulate less sorption to subsurface soils and will increase the contaminant
       velocity in the unsaturated and saturated zones.

•      Landfill Leachate Quality:  The Agency used water solubilities supplied by ORD as
       central-tendency concentration values for inorganics and used estimates from the Organic
       Leachate Model for central-tendency values for organics.  For the high-end, the Agency
       increased the landfill leachate concentrations for all constituents and set them equal to
       water solubility values supplied by ORD. Higher leachate concentrations increase the
       exposure concentrations, and increase the release rate of constituent mass from the  unit.

•      Ground-Water Well Depth: The Agency set the vertical location of the ground-water
       exposure points to the top of the saturated zone for all cases in the high-end. The
    7  Additional information on central tendency and high-end parameter values may be found
in Appendix G.

                               **• DRAFT: March 22,1993 •••

-------
                                           B-ll
       Agency set central-tendency depths based on facility-specific well data or, if no actual
       supply wells were present, set the depths to 75 percent of the depth to the bottom of the
       aquifer.  The high-end well depth corresponded to the maximum ground-water
       concentration because modeled concentrations are highest at the surface of the aquifer.

       Saturated Hydraulic Conductivity:  The central-tendency value for hydraulic conductivity
       was based on the best available facility-specific data (e.g., from RFAs or RFIs).  For the
       high-end, the Agency increased the hydraulic conductivity at each facility by one order of
       magnitude from the central-tendency value.   Increasing conductivity increases
       contaminant velocities in the saturated zone, but may decrease well concentrations
       through  increased dilution.

       Karst and Fractured Bedrock Simulation:  Assumptions for simulating the effects of karst
       terrain or fractured bedrock on transport and exposures varied for the central tendency
       and high-end cases. A complete discussion of all specific assumptions for karst terrain
       and fractured bedrock follows in Section B.3.1 of this appendix.

       Facility-Specific Changes

       Percentage of Organic Matter in Soil Layers:  For the high-end, the Agency selected a
       lower range value for the percentage of organic matter in the unsaturated soil layers
       based on ranges obtained from the DBAPE database according to location and soil
       type.8 Decreasing the organic matter content decreases the retardation of organics  in the
       soils and increases their transport velocity.

       Fraction of Organic Carbon in Off-Site Fields: The Agency varied this value based on
       ranges of values supplied by DBAPE. Decreasing the fraction of organic carbon (f^)
       increases the mobility of organic constituents  and therefore the  lower f^ was used in the
       high-end analysis.

       Sediment Delivery Fraction (SDF) to Off-Site Fields:  The Agency varied this value based
       on professional judgement  using topographic maps.  For the high-end, a larger SDF was
       employed to increase the simulated amount of contaminated soil delivered to offsite
       fields.

       River Flow Rate: The Agency varied the flow rate based on ranges obtained from the
       REACH files from the STORET database. For the high-end, a lower range for the flow
       rate was employed to increase the constituent concentrations in the river by decreasing
       the dilution rate.
   8 U.S. Environmental Protection Agency. Data Base Analyzer and Parameter Estimator
(DBAPE1 Interactive Computer Program User's Manual. EPA/600/3-89/083. Washington, D.C.:
U.S. Environmental Protection Agency, 1990.

                             ••• DRAFT: March 22,1993 •••

-------
                                           B-12
       SWMU-Specific Changes

•      Waste Concentration: The Agency varied the waste concentration for each constituent
       based on the best available background information (e.g, facility documents, TSDR survey
       data, and professional judgement). Where the Agency knew the waste concentrations
       with greater certainty (e.g., from waste sampling), the central-tendency values typically
       reflected the average of the reported concentrations, and the high-end values reflected
       the maximum of the reported concentrations. The Agency used best professional
       judgment and available data where SWMU-specific data were not available.  Increasing
       the concentration increases the overall constituent mass available  for transport and
       exposure, and directly corresponds to the leachate concentration for tanks and surface
       impoundments.

•      Depth of Unit:  The Agency varied the unit depth based on the best available background
       data, with larger unit depths corresponding to the high-end analysis.  For units with
       known dimensions, this parameter remained constant in the high-end and central
       tendency analyses. Increasing this value increases the constituent mass added to landfills,
       decreases the depth  to the aquifer for all units, and can increase the volatilization release
       rate.

•      Area of Unit:  The Agency varied the unit area based on the best available background
       data, with larger unit areas corresponding to the  high-end'analysis.  For units with known
       dimensions, this parameter remained constant in the  high-end and central tendency
       analyses. Increasing this parameter increases the mass added to landfills, and allows for
       greater total leachate volume from all units.

•      Tank Release Rates: The Agency varied the tank release rate  based on central tendency
       and high-end estimates of tank volumes or other parameters (such as the drum capacity
       of a storage area or a leakage or spillage rate) for a particular  SWMU, with greater
       release rates corresponding to the high-end analysis.  Increasing the release  rate increases
       the total constituent mass available for transport and exposure.

       B.23 Tank Release Assumptions

       The Agency simulated releases from tanks based on the results of EPA's Hazardous
Waste Tank Failure Model  (HWTFM).' The HWTFM is a stochastic, Monte Carlo simulation
model that generates release profiles for each of several different tank technologies based on the
probability of events such as overflows, corrosion of tanks and pipes, natural  catastrophes, and
accidental  spills.  For each technology, the HWTFM generated several representative release
profiles. A single profile (with the greatest  quantity released for an assumed 20-year operating
    9  U.S. Environmental Protection Agency. "Hazardous Waste Tank Risk Analysis, Draft
Report," prepared by ICF Incorporated and Pope-Reid Associates Incorporated, June 1986.

                               ••• DRAFT: March 22,1993 ••*

-------
                                           B-13
life) was then selected from among the representative profiles. The Agency then scaled the
profiles according to the tank volume to generate a release profile for specific tanks.  For a given
tank size and tank type, this process created a series of release volumes corresponding to each
year of the operating life (e.g., 0.5 m3 in the first year, 1.2 mj in the second year, etc.). The tank
types addressed by the HWTFM included the following:

       •      Closed, 3700 gallon, in-ground, concrete/steel treatment tank;

       •      Closed, 2100 gallon, in-ground, concrete storage or accumulation tank;

       •      Closed, 5500 gallon, above-ground, cradled, carbon steel storage or accumulation
              tank;

       •      Closed, 210,000 gallon, above-ground, ongrade, carbon steel storage or
              accumulation tank;

       •      Open, 2300 gallon, above-ground, cradled, carbon  steel treatment tank;

       •      Closed, 4000 gallon, underground, carbon steel storage or accumulation tank;

       •      Closed, 4000 gallon, underground, stainless steel storage or accumulation tank;

       •      Closed, 2300 gallon, above-ground, cradled, carbon steel treatment tank;

       •      Closed, 5500 gallon, above-ground, cradled, carbon steel storage or accumulation
              tank;

       •      Closed, 60,000 gallon, above-ground, ongrade, carbon steel treatment tank; and

       •      Closed, 4000 gallon, underground, stainless steel storage or accumulation tank.

       For the corrective action RIA, the Agency selected the tank type from the list above that
corresponded most closely to each tank being simulated at sample facilities.  The release rate
profile for the tank was then used as an MMSOILS input during the operational  period of each
tank being modeled. The Agency assumed no further releases after tank closure.

BJ Model Application Assumptions and Limitations

       This section  covers specific model application assumptions and limitations.  The first
subsection presents the assumptions made to simulate transport in karst  or fractured media, the
second covers  assumptions made for Non-Aqueous Phase Liquids (NAPLs), and the third covers
assumptions for non-standard waste management units (e.g., container storage areas and process
sewers. The last subsection discusses several other limitations of the MMSOILS model.
                             ••• DRAFT: March 22,1993 •**

-------
                                            B-14
       B .3.1 Assumptions for Karst Terrain and Fractured Bedrock

       Karst terrain and fractured bedrock systems pose particular ground-water modeling
problems.  Karst aquifers are a special form of carbonate aquifer, and differ from other aquifers
in that they contain an integrated system of pipe-like solution channels that act as underground
drains for the highly localized transport of water. Fractured bedrock systems can occur within
many different types of rock and are characterized by an abundance of cracks, joints, faults and
other fissures in the rock.  In both these cases, the permeability of the primary material (i.e., the
rock) is usually very low.  However, the secondary permeability introduced by fractures or
solution channels is usually several orders of magnitude greater, resulting in relatively high flow
through the system.

       Although contaminant transport in fractured/karst geologic materials is governed by the
same processes as in granular media (i.e., advection, mechanical dispersion, and chemical
reactions), the results in fractured media can be quite different. Because conventional ground-
water flow equations (such as those in MMSOILS) were developed primarily for homogeneous
and isotropic granular aquifers,  they are limited in describing flow in fractured/karst rock.  For
example, the average velocity of ground water through the aquifer is calculated using Darcy's law
based on the bulk hydraulic conductivity, the hydraulic gradient, and the bulk porosity. Although
this equation is accurate for granular porous media, it becomes more uncertain as the  size and
number of the fractures or channels increase and the flow becomes dominated by secondary
transport through  these fractures or channels.

       Though mathematical  equations developed for homogeneous and isotropic granular
aquifers do not accurately describe flow through karst/fractured systems, they are often used for
lack of suitable alternatives. For the RIA, the Agency developed special  assumptions for  the
more uncertain karst aquifer conditions, but  did not use any alternative approaches for fractured
media.

       Additionally, EPA differentiated those sites where karst or fractured bedrock represented
the primary aquifer of concern from those where it did not. For many facilities, while karst or
fractured zones were present, they were either confined or were situated beneath a porous
medium  aquifer that dominated regional flow.  Based on research of the hydrogeology at  each
facility, EPA identified several  facilities requiring corrective action with karst or fractured
conditions where the karst or fractured aquifer was not the primary aquifer of concern. In these
cases, the Agency did  not model the karst aquifer.  EPA also determined that a few facilities are
significantly affected by karst conditions, and that some are significantly affected by fractured
bedrock conditions.

       For those  facilities where transport was affected by karst conditions, the Agency modeled
them using the following approach:

               (1) For high-end estimates, EPA assumed downgradient concentrations were
               equivalent to the concentration at the bottom of the unsaturated zone.  Exposures

                               ••• DRAFT: March 22,1993 •••

-------
                                          B-1S
              at all downgradient receptor points at any given year were set equal to
              concentrations equal to those exiting the unsaturated zone in that year. Time
              delay and mixing with upgradient flow were not accounted for.

              (2)  For central-tendency estimates,  EPA set porosity equal to 0.01, assumed no
              contaminant retardation, and used best estimates for conductivity and dispersion.

       The Agency modeled the fractured bedrock facilities using the representative porous
medium methodology, (i.e., used the existing best estimates of porosity, conductivity, and
dispersivity at each facility) for central-tendency calculations, and differentiated high-end
calculations by increasing the conductivity by one order of magnitude. Both approaches are
identical to those used to simulate granular porous media.

       BJ.2 Non-Aqueous Phase Liquids (NAPLs)

       Dense and light non-aqueous phase liquids  (DNAPLs and LNAPLs) are believed to be
present at several  facilities within the RIA sample. Typically, a DNAPL release migrates through
the unsaturated and saturated zones and forms a lens at the bottom of the aquifer. Constituents
dissolve from the DNAPL-contaminated areas of the aquifer into the surrounding ground water
such that the DNAPL-contaminated area acts as a  source of contamination within the aquifer.
LNAPLs, on the other hand, usually migrate through the unsaturated zone and form a floating
lens on the water table.  An LNAPL lens also acts  as a source of contamination as constituents
dissolve into the underlying water.

       Direct simulation of NAPL behavior requires complex algorithms not usually found in
screening-level models such as  MMSOILS.  While models have been developed that simulate
transport of LNAPLs adequately, effective simulation of DNAPL fate and transport has been   .
less successful.  MMSOILS does not simulate the multiphase flow that characterizes NAPLs, but
instead simulates the transport of these constituents as if they behave like aqueous leachate
releases.  While this method does not duplicate the behavior of lens formation, it does ensure
mass conservation and provides a reasonable screening-level approximation of the extent of
contamination.10

       BJJ Non-Standard Waste Management Units

       MMSOILS is capable of simulating five different types of waste management units:
landfills, waste piles, surface impoundments, tanks, and injection wells.  Most waste management
practices use one of these types of units and may be simulated directly.  In other cases, however,
actual disposal practices can only be approximated  by these units.  For example, often the only
evidence of past waste disposal were areas of soil contaminated by an unknown source. In these
    10 Methods for simulating the effectiveness of remedies at NAPL sites are discussed in
Appendix C.

                             ••• DRAFT: March 22,1993 ••*

-------
                                           B-16
cases, EPA simulated the contaminated soil as a landfill provided that sufficient information was
available concerning the extent of contamination and the concentrations of hazardous
constituents in the soil.  Another common source of contamination was releases from container
storage areas. EPA generally modeled these units as "tanks" due to the similarity of the release
behavior (i.e., each releases liquid-state waste to the ground).  For these and similar scenarios,
the Agency used the flexibility of the MMSOILS model to approximate releases from these non-
standard units or releases due to non-standard management practices.  In other cases, however,
the waste management practice was too complex or uncertain to be accurately simulated.  These
cases included direct dumping of wastes into lakes, rivers, or wetlands, and unknown releases
from process sewers or other systems. Because of the high degree of uncertainty, the Agency did
not model cases such as these for the RIA.

       BJ.4 Other MMSOILS  Limitations

       Overall, the principal limitation is that MMSOILS is a screening-level model selected for
reasons including the ability to simulate several pathways, readily obtainable input requirements,
and limited  computational requirements. While these characteristics are appropriate in a model
used for a national-level  screening analysis, they do contribute to uncertainty in the modeling
process.

       Several other general limitations of the MMSOILS model and the modeling process
should also  be discussed.  The most basic involves uncertainty arid estimation in the model input
data. While the Agency used the best possible data sources available (e.g., RCRA Facility
Investigations (RFIs), and facility-specific hydrogeologic surveys), the best of these investigations
still involve uncertainty in their  findings.  In other cases, such  detailed  information did not exist
for some facilities, and the Agency used professional judgement to estimate certain key
parameters.  This uncertainty in the input data results in corresponding uncertainty in the model
results.

        The level of exposure estimate detail in space and time is limited by computational and
data management restrictions.  MMSOILS calculates exposure concentrations at a pre-specified
number of points  for each pathway.  The RIA risk methodology requires exposure estimates
from 1992 through the end of the  modeling period in year 2120.  To encompass this large period,
the Agency made  estimates at ten-year intervals at each exposure point.  Furthermore, ground
water concentrations were calculated out to two miles and atmospheric concentrations were
calculated out to ten kilometers beyond the facility boundary. Exposures were generally
calculated near the unit, at the facility boundary, and at regular intervals out to the desired
distance. If ground water contamination was indicated at two miles, an additional run was
performed to determine concentrations out to five miles. While this approach is useful for
capturing potential exposures at relatively distant points, it provides limited resolution for points
within or near to the facility boundary.

        Not all of the MMSOILS pathway algorithms are equally precise or accurate. Precision
in time and space varies from the atmospheric algorithm, which calculates annual concentrations

                               •*• DRAFT:  March 22,1993  ***

-------
                                           B-17
and any given point, to the soil transport algorithm, which must use an average (i.e., not time-
variant) source concentration.  Accuracy refers to the overall vigor and complexity of the model
equations.  The ground-water and atmospheric algorithms are most accurate.  Other pathway
algorithms, however, contain certain key limitations.  For example, the surface water pathway is
limited because downstream concentration decreases due to sedimentation and degradation are
not accounted for. Additionally, the soil pathway is not capable of simulating local transport  of
contaminants within the facility; it is limited to deposition and erosion to off-site fields.

B.4 MMSOILS Post-Processing

       The MMSOILS model simulates releases, transport, and exposure on a SWMU-specific
basis. However, the Agency estimated long-term risk and the extent of contamination  at off-site
points on a facility-wide basis because of uncertainty in SWMU and well locations, modeling
resource limitations, and because each facility exposure point was potentially influenced by many
SWMUs. Different methods were used to aggregate SWMU-specific exposure concentrations for
different pathways, as described below.

       B.4.1  Ground-Water Pathway

       To  aggregate the SWMU-specific ground-water results into facility-wide values, EPA
developed  an aggregation scheme encompassing three components: creation of a facility-wide
grid, aggregation of the SWMU-specific plumes, and interpolation of the data to give annual
results.

       Downgradient Grid

       As  depicted in Exhibit B-2, the Agency developed an exposure grid for each facility
extending out to a distance two-miles in the direction of ground water flow from the facility
boundary.  The Agency constructed the grid to span a 90 degree sector centered in the direction
of ground-water flow.  The Agency divided this grid into four radial sectors of 22.5 degrees and
six  rings at distances of 0.25,0.5,0.75,1.0,1.5 and 2.0 miles from the facility boundary. If initial
modeling indicated exposures beyond two miles, the Agency performed a supplementary
MMSOILS run to estimate exposures out to five miles using a similar grid (depending upon  local
hydrogeologic conditions and topography).  For  each element of the grid, the Agency gathered
residential well information (e.g., number of wells and population served). In addition, the
Agency determined the exact location of public/municipal wells, agricultural wells, and closest
potential receptors in relation to the grid origin.

       Plume Aggregation

       The Agency aggregated ground-water concentrations at offsite downgradient points using
one method for private/residential wells and another method for the remaining wells.  These
methods may be demonstrated using Exhibit B-2. For both cases, the aggregation method
assumed that each SWMU lay on the centerline of the grid.  This assumption overestimated

                              *•* DRAFT: March 22,1993 •••

-------
                                            B-18
exposures in some cases, while underestimating exposures in others.  In general, if the size of the
facility was small compared to the size of the grid, the magnitude of the error was small because
the actual distance from the SWMU to the centerline would be small compared to the scale of
the off-site grid.  If the facility was larger, however, this assumption would tend to underestimate
exposures to points that are  actually more directly in line downgradient (e.g., exposures to the
public well from SWMU A in Exhibit  B-2), and it would overestimate exposures if the SWMU
and exposure point were actually farther apart (e.g., exposures to the public well from SWMU
B).  The Agency assumed that these errors would generally cancel each other when determining
aggregate exposure concentrations from several SWMUs at a facility.  In addition, if the facility
was very large and the SWMUs were not evenly distributed, the Agency focused the grid on the
portion of the facility where  the SWMUs were located in order to minimize the error.

       For public wells, agricultural wells, and other explicitly defined  points, the  Agency
determined  the downgradient distance (specified in the input file) from the unit to the exposure
point by taking the downgradient distance from the unit to the facility boundary (e.g., distance L,
for SWMU A) and adding the downgradient distance from the facility boundary to the specific
exposure point (e.g., distance L,). The lateral (i.e., distance perpendicular to the gradient)
coordinate of the exposure point was taken as the distance from the grid centerline to the
exposure point (e.g., distance L,). Because the Agency assumed that the SWMUs lay on the
centerline, the lateral distance was not a function of the SWMU location.  To determine
aggregated exposures at the explicitly defined points, a post-processor summed  the SWMU-
specific concentrations at each point.

       For private wells, average exposures were calculated for each ring instead  of at each well.
For these calculations, the distances to the exposure points were not determined according to the
exact well locations, but were dictated by the geometry of the data collection grid  (as shown in
Exhibit B-2).  To represent exposure concentrations within any one ring of the grid (e.g. all
points between 1.0 and  1.5 miles downgradient), two points were defined: the first exposure point
(e.g., C,) lay on the grid centerline at the midpoint of the ring (e.g., at  1.25 miles), and the
second exposure point lay 33.75 degrees off the centerline at the midpoint of the ring. To
calculate average aggregate exposures, post-processing routines summed the SWMU-specific
concentrations for these points (generated by MMSOILS) to give aggregated concentrations at
the points. The routines then averaged these concentrations [(Cl +2xC2)/3] and multiplied the
average concentration by the total ring population to produce an aggregate exposure level for
each ring of the grid. Because of symmetry, the offset point represented concentrations on both
sides of the  centerline.  Note that the  exhibit only shows  points within one of the six rings, but
the post-processor made separate calculations for each ring.
                              ••• DRAFT: March 22,1993 •••

-------
                        EXHIBIT  B-2
                GRID FOR SWMU AGGREGATION
                  (HYPOTHETICAL EXAMPLE)
                                 ^^^" Facility Boundry
• — 1
SWMUA
LI
SWMHB
	 	 •
                                  Miles Downgradient
                                  from
                                  Facility Boundry
Shaded area represents one cell of the grid.

C = exposure points

-------
                                           B-20
       Interpolation

       MMSOILS calculated both private and public well exposure concentrations at ten-year
 intervals over the modeling period (i.e., 1992 through 2120).  A post-processing routine was used
 to interpolate these values to annual values across the modeling period, using a cubic spline
 routine based on Press, et al.." This non-linear routine allowed the interpolation process to
 capture concentration peaks that may have occurred between the ten-year time-steps, preventing
 the underestimation of peak concentrations due to the ten-year modeling constraints. Additional
 routines then used these annual values as inputs to estimate risks associated with each of the
 exposure points. The risk calculations are discussed further in  Appendix E.

       B.4.2 Atmospheric Pathway

       Aggregation of SWMU-specific atmospheric concentrations was more straight-forward
 because the modeling and risk methodology averaged exposures over entire (360 degree) grid
 rings instead of over partial (90 degree) sectors (as for ground  water).  The model calculated
 annual exposures at various radial  distances from the facility boundary, and the exposure point
 distance in the input file were defined as the sum of the distance from the facility boundary to
 the desired exposure point and the minimum distance from the SWMU to the facility boundary.
 Because only one point was defined for each risk calculation (i.e., for each ring), aggregate
 concentration were determined by  simply summing the SWMU-specific concentrations at each
 exposure point for each constituent.

       B.4.3 Surface Water Pathway

       The  MMSOILS model calculated surface water concentrations at only one exposure
 point.  For rivers and streams, annual average concentrations were aggregated simply by adding'
 the constituent concentrations by year across all the SWMUs.  For lakes, MMSOILS generated  a
 single concentration already averaged over the exposure period. In this case, aggregation was
 performed simply by adding the concentration associated with each  unit.

       B.4.4 Off-site Soil Pathway

       The MMSOILS model calculated a single average concentration  at each of the two off-
 site soil points (i.e., the agricultural field and the human exposure field). For each point,
 aggregation was performed simply  by adding the concentration  associated with each unit.
    11 Press, W.H.; Flannery, B.P; Teukolsky, S.A; and Vetterling, W.T. Numerical Recipes in C.
Cambridge: Cambridge University Press, 1988.

                              »** DRAFT: March 22,1993 •••

-------
                     C. SIMULATION OF REMEDY EFFECTIVENESS

       This appendix provides a detailed discussion of the methodology for simulating the
effectiveness of remedies more briefly addressed in Chapter 4. In order to evaluate the benefits
of the corrective measures implemented under Subpart S, it was necessary to estimate the extent
of contamination and associated human and environmental exposures remaining after
implementation of corrective measures. Because not all corrective measures are equally
effective, the Agency developed approaches for simulating the effectiveness of specific corrective
measures technologies, focusing on several key technologies selected by the expert panels.

       The key technologies may be divided into three categories:

       (1)     Source  Control Technologies

              •      Covers and Liners

              •      Excavation

       (2)     Waste Treatment Technologies

                     Stabilization/Solidification
                     Soil Vapor Extraction (SVE)
                     Soil Washing
                     Landfarming
                     On-Site Incineration

       (3)     Ground-Water Remediation Technologies

              •      Pump and Treat Systems
              •      French Drains
              •      Slurry Walls or HDPE Barriers

       The Agency developed quantitative approaches for simulating each of these  remedial
technologies.  When available, the Agency used performance levels specified by the  expert panels
as the basis for determining the remedy effectiveness. In other cases, the Agency used default
values based on technology-specific performance data.  EPA assumed some corrective measures
(e.g., off-site disposal and on-site disposal  in a Subpart F unit) to be completely effective  and
were not simulated.  The technology-specific assumptions are described below.

C.1 Source Control Technologies

       Source control  technologies physically restrict the release of contaminants from a  waste
management unit. The source controls simulated in the RIA include covers, liners,  and
excavation.
                             ••• DRAFT: March 24,1993 *•*

-------
                                             C-2
       C.I.I  Covers and Liners

       The MMSOILS model directly simulates the installation  of covers and liners on landfills
and surface impoundments  through  water balance routines. Exhibit  C-l  shows the thicknesses
and hydraulic conductivities of clay and synthetic cover or liner materials simulated  in the RIA.

                                       EXHIBIT C-l

                          PROPERTIES OF COVERS AND LINERS

Thickness
Hydraulic Conductivity
Clay
60 cm if not specified by the
expert panel '
1 x 10'7 cm/sec '
Synthetic
40 mils '
1 x 10'12 cm/sec 2
In addition, the Agency developed  a series of time-dependent  efficiencies for each type of cover
or liner.3  The  efficiencies  simulate the ability of the liner or coyer to effectively block flow after
accounting for  failure and  aging.  In developing  these efficiencies;- the Agency assumed that new
caps would be  added  on top of the old caps after  100 years of operation; therefore the efficiency
values for caps, shown in Exhibit C-2, increase at  year 100.
    1  U.S. Environmental Protection  Agency. Technical Guidance  Document:  Final Covers on
Hazardous Waste Landfills  and Surface Impoundments. EPA/530-SW-89-047.  Washington, D.C.
U.S. Environmental  Protection Agency, July 1989.

    2  GeoServices Inc. Liner/Leak  Detection Rule Background Document.  April  1987.

    3  U.S. Environmental Protection  Agency. Indexing of Long-term Effectiveness of Waste
Containment  Systems for a  Regulatory Impact  Analysis (draft). Washington, D.C.: U.S.
Environmental  Protection Agency, December  1992.
                               •** DRAFT: March 24,1993 ••*

-------
                                             C-3
                                        EXHIBIT C-2

                           EFFICIENCIES OF CAPS AND LINERS

EfTiciency
when
installed
EfTiciency
after 10
years
EfTiciency
after 30
years
Efficiency
after 100
years
Clay Cap
80%
75%

60%
85%

Synthetic
Cap
90%
85%

75%
90%

Clay/
Synthetic
Composite
Cap
95%
92%

80%
98%

Clay Liner
70%
60%

40%
5% :

Synthetic
Liner
85%
75%

35%
0%

RCRAC
Liner
System
98%
95%

85%
60%

       C.I.2 Excavation

       The Agency simulated excavation by revising the  MMSOILS  model to allow removal of a
specified percentage  of waste mass from a unit at a given time (i.e., the year of excavation).  If
the expert  panels specified that  the entire  waste source was excavated, the MMSOILS inputs
were modified to specify that no waste remained in the unit after excavation.  Contaminants that
had already entered  the unsaturated  zone  continued  to migrate, but no new contamination
entered  the subsurface.  If the expert panels only excavated  hot spots (10 percent  of the waste,
for example), the inputs were modified to reflect the percent removal  (i.e., the mass would be
reduced to 10% of the original value).
C.2  Waste Treatment Technologies

       Waste treatment  technologies,  by definition, reduce the mobility, toxicity, persistence,  or
concentration  of contaminants  in waste materials.  The waste treatment technologies simulated
for this RIA  include stabilization/solidification,  soil vapor extraction, soil washing, landfarming,
and on-site incineration.   The primary effect of stabilization  is to reduce the constituent  mobility
                               ••* DRAFT: March 24,1993 •*•

-------
                                               C-4
by reducing the leachate flow and by causing some constituents  to bind more tightly to the
stabilized material,  thus reducing the leachate concentration.  For the remaining technologies,
the primary effect is to reduce the amount of constituent  mass in the waste  material.  Because  of
these differences, the Agency adopted  different  methods for simulating these waste treatment
technologies.

        C.2.1 Stabilization/Solidification

        To simulate stabilization/solidification, the Agency modified  the  MMSOILS model to
allow changes  to  the conductivity of the waste (affecting the quantity of leachate generated)  and
the concentration  of hazardous  constituents  in the leachate at a specific point in time during the
simulation period.  At the time of remediation,  the model  switches the values of these
parameters  from  the baseline values to new input values reflecting  the effectiveness of the
technology.

        Decreasing the conductivity  of the waste layer restricts  the flow of water through the
waste, and thus decreases the volume of leachate generated  by the unit.  This serves to limit the
mass flux from the unit after remediation  (although  it does not affect the total mass within  the
unit that may leach out in the future).  Where the expert panels  specified performance  standards
for stabilization,  those values were used in the remedy  simulation.  If standards were  not
specified, the Agency used default values.  Because of differences in effectiveness  between in-situ
stabilization  (where the binding  material  is added to  the unit and 'mixed  in with augers  or
backhoes)  and ex-situ stabilization (where the waste is excavated, mixed  with the binding material
in a pug-mill or other mixing device, and then replaced), the Agency established two different
default  performance  levels based on literature values: lxlO"7cm/s for in-situ stabilized  material
and  IxlO'8 cm/s for ex-situ stabilized  material.4   Ex-situ stabilization  is generally considered  more
effective than  in-situ stabilization because of greater  mixing efficiency.

        For metals,  solidification/stabilization  has been  shown to decrease leachate  concentration
from wastes due to chemical bonding with the waste  matrix. To simulate this effect, the analysis
employed a one order  of magnitude reduction in leachate  concentration   for inorganic compounds
and  no reduction  for organic compounds.

        The parameter  changes  used to simulate stabilization are summarized  in Exhibit C-3.
    4  U.S. Environmental  Protection Agency/ORD/RREL, Handbook  on In Situ Treatment  of
Hazardous  Waste-Contaminated  Soils. EPA/540/2-90/002. Washington, D.C.: U.S. Environmental
Protection Agency, January 1990.

        U.S. Environmental  Protection  Agency/ORD/RREL,  The Superfund  Innovative
Technology Evaluation  (SITE) Program: Technology  Profiles. EPA/540/5-90/009. Washington,
D.C.: U.S. Environmental  Protection Agency, November  1990.

                                **• DRAFT: March 24,1993 **•

-------
                                          C-5
                                     EXHIBIT C-3

                       DEFAULT STABILIZATION PARAMETERS
Treatment
Technology
In-Situ Stabilization
Ex-Situ Stabilization
All Stabilization
Waste/Constituent
Type
All
All
Inorganics
Organics
Parameter
Conductivity
Conductivity
Leachate
Concentration
Leachate
Concentration
Change
Set to 10'7 cm/sec
Set to 10* cm/sec
Decrease 1 order of
magnitude
None
       C.2.2 Other Waste Treatment Technologies

       The remaining treatment technologies simulated in the RIA (SVE, soil washing, land
farming, and incineration) are all treatment technologies that reduce chemical mass in the waste.
To simulate these treatment technologies, the Agency revised the MMSOILS model to adjust the
constituent concentrations in the unit at a given point in time. At the time of remediation,
constituent concentrations within the unit are reset to target  cleanup concentrations specified by
the expert panels. If the expert panels did not specify target  concentrations,  the model used
default inputs based on other performance data.5  In general, these types of treatment
   5 Many overlapping sources were used to supply performance data for treatment
technologies.  These include the following:

       • Environmental Research and Technology (ERT), The Land Treatabilitv of Appendix
VIII Constituents Present in Petroleum Industry Wastes. February 1984.

       • U.S. Environmental Protection Agency/ORD/MERL. Hazardous Waste Land
Treatment. Washington, D.C.: U.S. Environmental Protection Agency, April 1983.

       • U.S. Environmental Protection Agency. Seminar Publication — Corrective Action:
Technologies  and Applications. EPA/625/4-89/020. Washington, D.C.: U.S. Environmental
Protection Agency, September 1989.
                             •** DRAFT: March 24,1993 •••

-------
                                           C-6
 technologies are only effective on organic constituents.  While these treatment technologies may
 have some effects on inorganic constituents, the Agency assumed that these effects would be
 negligible and did not change the inorganic concentration values.  The default effectiveness
 values for each technology are presented in Exhibit C-4.
       • U.S. Environmental Protection Agency/ORD/RREL. Handbook on In Situ Treatment
of Hazardous Waste-Contaminated Soils. EPA/540/2-90/002. Washington, D.C.: U.S.
Environmental Protection Agency, January 1990.

       • U.S. Environmental Protection Agency/OERR.ORD. Engineering Bulletin: Soil
Washing Treatment. EPA/540/2-90/017. Washington, D.C.: U.S. Environmental Protection
Agency, September 1990.

       • U.S. Environmental Protection Agency/ORD/RREL. The Superfund Innovative
Technology Evaluation rSITEi Program: Technology Profiles. EPA/540/5-90/006. Washington,
D.C.: U.S. Environmental Protection Agency, November 1990.

       • U.S. Environmental Protection Agency/OSWER/Technology Innovations Office.
Innovative Treatment Technoloaes: Overview and Status — Preliminary Draft. Washington,
D.C.: U.S. Environmental Protection Agency, March 1991.

       • U.S. Environmental Protection Agency/OERR.ORD. Engineering Bulletin: In Situ
Steam Extraction Treatment. EPA/540/2-91/005. Washington, D.C.: U.S. Environmental
Protection Agency, May 1991.

                             *** DRAFT: March 24, 1993 •••

-------
                                             C-7
                                       EXHIBIT C-4

                      DEFAULT WASTE TREATMENT EFFICIENCIES
Treatment
Technology
Soil Vapor
Extraction (SVE)
Soil Washing
Landfarming
On-site Incineration
Waste/Constituent
Type
VOCs
SVOCs
VOCs
SVOCs
Organics
Organics
Parameter
Waste
Concentration
Waste
Concentration
Waste
Concentration
Waste
Concentration
Waste
Concentration
Waste :
Concentration
(reflects residuals)
Change
Decrease by 90%
Decrease by 50%
Decrease by 90%
Decrease by 50%
Decrease by 95%
Decrease by four
orders of
magnitude.
C.3  Ground-Water Remediation Technologies

       Pump and treat  systems, french  drains, slurry walls, and  HDPE  barriers are all
technologies that restore or contain contaminated  ground water. The approach for simulating
these technologies  focused on determining  the effectiveness of a particular  system at restoring or
containing a dissolved plume,  while taking  into account hydrologic  conditions at the site and the
potential  presence of non-aqueous phase liquids (NAPLs).  The Agency simulated ground-water
remediation  using post-processing routines  that modify the extent of contamination  profiles
remaining after  simulating source controls  and waste treatment.

       C.3.1 Restoration of Dissolved Plumes

       For most contaminants  in ground-water systems, the majority of the total  constituent
mass is not initially dissolved in the water,  but is sorbed to soil  particles.  As contaminated  water
is pumped from the aquifer during ground-water  remediation, it is replaced  by clean water that
has been re-injected  into the aquifer or that has naturally recharged  the aquifer.  As contaminant
mass is removed through  pumping, additional  mass will desorb  from the solid  phase into the
aqueous phase.  Eventually, little mass will remain  in the sorbed phase and the ground-water
                               ••* DRAFT: March 24,1993 ***

-------
                                              C-8
concentrations  will have been reduced.

        The effectiveness of this ground-water  remediation  process can be expressed in terms of
first order coefficient (ft) based  on mass conservation, the sorption strength  of the individual
chemical, and  the  rate of water  flowing through the contaminated zone.  This methodology  is
consistent  with similar applications presented  in EPA guidance sources.6*7 By using this
available  Agency approach  for simulating restoration effectiveness, EPA avoided the additional
modeling and  data collection  complications  required when using more sophisticated  numerical
models, and remained  consistent with the screening-level complexity of the  MMSOILS model.

The coefficient (k) is calculated  as follows:

                                                                                          (1)
where R equals a retardation  factor for the specific chemical in the aquifer, and PV is the pore
volume exchange frequency.

        The retardation  factor (/?) is equivalent to the  inverse of the fraction of total contaminant
mass that  is in aqueous solution.  It is defined as follows:

                                                                                          (2)
where Kt equals the partition coefficient  between soil and water concentrations,  p equals  the
bulk density of the  aquifer material, and  ij equals the aquifer porosity.

        The pore  volume  exchange frequency (PV) represents the frequency at which one pore
volume is pumped from the  aquifer.  The pore volume  is defined as the volume of water within
.the contaminant  plume and  may be calculated as the product of the plume  area, the
contaminated  depth  (i.e., the aquifer thickness), and the aquifer porosity.  To determine the
exchange frequency, the total pumping rate (specified by the expert panels) was divided by the
pore  volume.

        The aqueous concentration of a contaminant  in ground  water (Q as a function of time
    6 U.S. Environmental  Protection Agency/Exposure Assessment Group/Office of Health and
 Environmental  Assessment. Guidance for Establishing Target Cleanup  Levels for Soils at
 Hazardous  Waste  Sites. Washington. D.C.: U.S. Environmental Protection  Agency, 1988.

    7 U.S. Environmental  Protection Agency/Office of Emergency and  Remedial Response.
 Guidance  on Remedial  Actions for Contaminated  Ground  Water  at Superfund  Sites.
 Washington,  D.C.: U.S. Environmental  Protection  Agency, 1988.

                                ••• DRAFT: March 24,1993 **•

-------
                                              C-9
 (r) may be expressed as:

                                           C=Cf-*                                       (3)

 where C, equals the base aqueous concentration  without pumping,  and t equals the time since
 pumping  began.

        These equations are based on several  key assumptions presented below.

 (1)     Any additional  contaminant releases from the source must be small compared to the total
        initial mass.  While source controls were not totally or instantaneously  effective, EPA
        believes the additional mass contributions  were relatively small.  The Agency included the
        effectiveness  of the source controls by adjusting the C0 term.

 (2)     All of the constituent  mass must be initially distributed  between the soil and  water only.
        This  implies that the  presence of NAPLs within the system  invalidate the equations.  As
        contaminant mass is removed  from the aquifer through  extraction,  more contaminant
        mass will dissolve from the non-aqueous phase into the aqueous  phase.  As long as
        NAPLs  are present, the extraction system  will likely not  have a significant affect on the
        aquifer concentrations.  The approach  for simulating  plume restoration when NAPLs are
        present  is discussed in Section C.3.3.

 (3)     Contaminant  mass is  freely able  to transfer from  the  sorbed  to the aqueous phase.   In
        reality, the concentration  often does not indefinitely  decrease, but approaches  a near-
        steady concentration  determined  by kinetic limitations of mass transfer to the aqueous
        phase.  These limitations  are likely to  be dependent  upon both the constituent of concern
        and  the  physical characteristics  of the  aquifer  media.

 (4)    The rate of mass removal at a simulated extraction well  is proportional to the
        concentration at that  well without pumping.   While this conserves mass within the plume,
        it  does  not allow the possibility that  the concentration distribution within the  plume may
        change  as a result of the pumping.

 (5)     All of the contaminant mass is drawn  toward the  extraction  wells and does not escape the
        boundaries  of the plume.  In the case  of karst or  fractured bedrock, the effectiveness of
       the extraction system  in capturing all of the contaminated  water may be questionable.  In
        these cases, the contaminant plume may spread beyond  its initial boundaries,  or fingers of
        contaminated  water may escape  through fractures  or  solution channels.  The  ability of
       ground-water remediation  systems to contain  the  plume under  different  hydrogeologic
       conditions is discussed in Section C.3.2.

       The Agency applied the exponential  remediation   equation  at each well downgradient
from each facility with  ground-water  contamination  using a synthesis of two methods  to estimate
the initial concentration  (CJ.  In the first approach,  C0 varies with  time and  is set  equal to the

                               ••• DRAFT: March  24,1993 •••

-------
                                              C-IO
time-varying concentration  at each point after simulating  source controls  and waste treatment.
As demonstrated  by a hypothetical case shown in Exhibit C-5, this  approach  (depicted  as
Approach  1) captures  the effects of source controls, but may underestimate  the impacts of source
controls and waste treatment  in the near-term.   The second approach  (depicted as Approach 2)
sets  C0 equal  to a constant concentration (the concentration  at the time of remedy
implementation).  While this approach  is more-consistent with the assumptions of the
exponential equation  (because the total mass in the system is fixed), long-term  concentrations
may actually be greater than  in the case with source controls alone  (e.g., at times after year 2054
in Exhibit C-S).

        To minimize the effects of the limitations in these two approaches, the  Agency estimated
concentrations   using a synthesis of both approaches.  The post-processing  routine calculates  a
ground-water  concentration   using both  equations,  and then sets the final  concentration at each
point for each time-step  equal to the  lower of the two estimates  (e.g., in  Exhibit C-5, Approach  2
was used from  years 2000 to 2036, and  Approach  1 was used at times  after year 2036). When
applying this formula,  the Agency assumed  that the extraction wells operated until the end of the
simulation period or until the cleanup objective was achieved at  each  well (whichever  came  first).
                                 ••• DRAFT: March 24,1993 •••

-------
                         Exhibit C-5
            Hypothetical Comparison of Different
                   Restoration Approaches
                                            No Corrective Action
                                  Source Controls Only
Remediation Begins
2010
                         2020
                                   2030
2040
          2050
                            Year

-------
                                             C-12
       CJ.2 Containment of Dissolved Plumes

       Ground-water plumes can be contained through  extraction  wells, French drains, HDPE
barriers,  or slurry walls, often in conjunction with pumping.  The expert panels typically specified
HDPE barriers and  slurry walls when the objective was to contain  the contamination  rather  than
to restore the aquifer to a beneficial  use.  Situations  requiring containment  were typically the
result of very high concentrations  of contaminants  in the aquifer.  The Agency's assumptions
concerning the effectiveness of these systems are summarized  in Exhibit C-6.

                                        EXHIBIT C-6

                    DISSOLVED PLUME CONTAINMENT EFFICIENCIES
Effectiveness of
containment in
various media
Granular
porous media
Fractured rock
Karst
Extraction wells
100%
75%
50%
French drains
100%
NA
NA
HDPE barrier
with ground-
water extraction
100%
--'NA
NA
Slurry wall with
ground-water
extraction
100%
NA
NA
NA:     Not applicable
        For situations  with  100 percent effective containment,  the Agency assumed that the
plume would not expand downgradient after  the  year of installation  of the containment system.
For karst and fractured  systems, after simulating  the  effects of source controls, the exponential
equation  was applied  to all potential exposure  points within the dissolved plume  (including any
points beyond the boundaries  of the extraction  system).  The  Agency then multiplied  the  ground-
water  concentrations  at exposure  points located beyond the initial plume boundary by the
appropriate  percent effectiveness  (i.e., 25 percent |or 1-.75] for fractured and 50  percent  for
karst).  The Agency assumed that French  drains, HDPE barriers, and slurry walls would not be
constructed  in fractured bedrock or karst  conditions.

        French drains decrease ground-water  concentrations at wells located downgradient of the
barrier.1  This effect  was simulated  by using the exponential equation  and setting C0 as a
constant  equal to the concentration  at the time of remediation.   The time term within the
    * This is based on the assumption that low-permeability barriers  were installed  down-gradient
 from the french  drains to prevent  ground water from being pulled up-gradient  into the drains.

                                •*• DRAFT: March 24,1993 •*•

-------
                                             C-13
exponential  equation  was then adjusted to account for delay associated  with the time of travel
from the drain to each well. In this manner, the effect of the French drain at each well would
not be apparent until the year when ground water traveling from the drain at  the time of its
installation reaches each successive well.

       C33  Non-Aqueous Phase Liquids

       The MMSOILS  model  is a semi-analytical  solution of the advection-dispersion  equation
for a single non-conservative constituent and cannot  directly simulate the fate and transport  of
NAPLs.  Because the MMSOILS model cannot directly  simulate  the behavior of NAPLs,
particularly when remedial technologies  were applied to the contaminated  area, the Agency
developed a simplified approach  for simulating the effectiveness  of NAPL remediation.  For  light
non-aqueous  phase liquids (LNAPLs), the Agency assumed that the floating product  could  be
recovered based on field experience with numerous  hydrocarbon  spills. Current field experience
with dense non-aqueous  phase liquids (DNAPLs), however,  indicates that remediating the
DNAPL  source may be technically  impracticable in many  cases.  Because  of this difference, a
separate  approach  was used for cases  where DNAPLs were  present.

       The approach  for simulating DNAPL remediation  assumed  that the DNAPL was the
principal  source of contamination.  With this approach,  source  controls, such  as caps or
excavation, specified for a unit with DNAPL releases  would not be effective at reducing overall
releases to the aquifer if it was assumed that DNAPL had already' entered the aquifer.  In these
cases, the Agency did not simulate  any source  controls or  waste treatment  technologies  with
MMSOILS.

       This approach for ground-water remediation   with DNAPLs  required the division of the
site into  two regions (see Exhibit C-7). The interior  region represents  the area surrounding  the
DNAPL-contaminated  portion of the aquifer.  The interior region was separated  from the
exterior region by the containment  system specified  by the expert panels for each facility; the
containment system (whether  extraction wells, a slurry wall, or  a HDPE wall) represented  a
hydraulic barrier between the two regions.

       Because MMSOILS cannot  directly  simulate  ground-water remedies, the Agency
simulated the  effect of the barrier through the MMSOILS cover  routine.  In order  to determine
the year  of installation of the  simulated  barrier, the travel time of the  DNAPL constituent was
estimated  from the time of release  from the unit to the  point at which the actual barrier would
be placed in the ground-water  system.  MMSOILS was then  run with the  cover installed in  order
to generate a  modified set of ground-water  concentrations  reflecting the barrier.  Using this
approach, the  effect  of the hydraulic barrier would be reached during the year of its installation.

       For example, if the expert panels specified installation  of a  slurry wall in year 2005
located 20 meters  from  a unit boundary, the Agency estimated  the  time it would take for the
DNAPL  constituent to flow from the SWMU,  through the unsaturated  zone, and to the slurry
wall located 20 meters from the unit boundary (say  20 years for this example).  In this case, the

                               ••• DRAFT: March  24,1993 •••

-------
          EXH  B T C-7
HYPOTHETICAL DNAPL SITE
                          Aqueous Plume
            Extent
            DNAPL
                  Containment
                     Wells
            Ground Water

-------
                                             C-15
Agency simulated  the slurry wall by installing a cover in year 1985; the resulting exposure
concentrations  would show a concentration  decline (attributable  to the slurry wall) at the 20
meter point at approximately year 2005. The efficiencies of these barriers (and, therefore  the
efficiencies of the covers used to simulate them) varied  with the site geology and were based  on
containment  efficiencies presented   previously in Exhibit C-6 (e.g., a slurry wall  would have a
100% efficiency, while extraction wells in fractured bedrock would have a 75%  efficiency).

       After running  MMSOILS to simulate the post-remediation  scenario,  the Agency  used  a
post-processor  to  calculate  concentrations inside and outside  the containment zone.  In the
interior zone, the  Agency assumed  that the  DNAPL  lens would  continue  to  act as a  source of
constituent releases  to ground water and that the ground water  would not be restored  to use
(and thus baseline concentrations  would be unaffected).   Outside the containment  zone, the
Agency used the exponential  equation  (described  in  Section C.3.1) to simulate  the effectiveness
of any ground-water extraction  wells specified  by the expert panels for plume restoration.
                                *** DRAFT: March 24,1993 ***

-------
                             APPENDIX D. COST ANALYSIS
       This appendix addresses the development of cost estimates for remedial activities,
 focusing on unit costs estimation.

 D.1 Expert Panel Cost Estimation

       The costs of corrective action presented in Chapter 5 were based on simulated remedy
 selections for the proposed corrective action rule. For each remedy selected at each unit, the
 expert panels broke out the detailed components of the remedy.  For example, a pump and treat
 system to address contaminated ground water might include:

              Extraction well installation;
              Pump installation;
              Pump operation;
              Piping installation (from wells to treatment system);
              Construction of treatment system (i.e., air stripper and vapor phase carbon unit);
              Construction of treatment enclosure;
              Operation of treatment system;
              NPDES permit; and
              Discharge through NPDES outfall.

 For each step, the experts developed a cost estimate using either professional judgment to
 estimate a lump sum cost or a unit cost calculation.  Wherever possible, the experts used unit
 cost calculations (e.g., $8 per yd3 to  excavate soil).  Examples of costs estimated by lump sum
 include permits, debris removal, steam cleaning of equipment and structures, and repair of
 existing structures.  Where tmit costs were  applicable  and available, the experts were able to base
 cost estimates on volumes and areas of waste and contaminated  media addressed at each unit.
 Unit costs were taken from EPA guidance  documents for hazardous waste remediation and
 engineering manuals (e.g., Means Site Work Cost Data).

       Exhibit D-l  provides the unit cost ranges used by the expert panels. The expert panels
 chose unit costs from the sources listed in Exhibit D-2 based on  facility-specific characteristics
 and professional judgment. For example, excavation of a very deep unit may require different
 removal techniques than excavation  of a shallow unit, with a corresponding difference in unit
 cost.  For some frequently used remedies, typical unit costs are not identified in this exhibit due
 to the complexity of the cost calculations (i.e., the cost estimate is dependent upon a number of
 facility-specific characteristics.) For example, the cost of a ground-water extraction well is
 partially determined by the depth of well required, necessary screening precautions, type of well
 casing used, and the characteristics of the soil and/or bedrock overlying the aquifer. In such
cases, the expert panels would have  first designed the well and then estimated its cost.  Finally,
the unit cost ranges in Exhibit  D-l may group dissimilar activities together. For example, the
category "Other Object Removal" may include unit costs for activities as different as excavating
and crushing a pipeline and removal of nerve gas canisters.
                            * • • DRAFT-March 23,1993 • • •

-------
               Exhibit D-1
Unit Cost Ranges Used by the Expert Panels
Media = AIR

Remedial Activity
Active landfill

Catalytic

Tarps

Vapor phase carbon



Cost Type (Unit
Capital blower
meters
Capital (unit
O&M (unit
Capital (square meters
O&M (hours
Capital (unit
O&M (cubic meters

Low

S50.000.00
S93.21
S200.000.00
S20.000.00
S10.76
S20.00
$500,000.00
S2.210.53
Unit Cost
High

$50,000.00
$93.21
S200.000.00
S200.000.00
$118.40
$20.00
$500,000.00
| $2,210.53

Median

$50,000.00
$93.21
$200.000.00
$20.000.00
$10.76
$20.00
$500.000.00
| $2.210.53
Media = GENERAL
Remedial Activity (Cost Type (Unit
Buildings (Capital
square meters
Clear and/or grub (Capital (square meters
Concrete
Electromagnetic
survey
Fencing
Capital
cubic meters
meters
square meters
O&M (square meters
Capital (square meters
Investigation (square meters
Capital
meters
unit
Medical moni toring|Capital (hours
Misc.
construction/repa-
ir
Capital
O&M
cubic meters
meters
unit
cubic meters
square meters
Sewer system study) Investigation (hours
Site preparation
Tank replacement
Wetlands
restoration
Capital
Capital
Capital
cubic meters
hours
square meters
cubic meters
hours
square meters
Unit Cost
Low
$118.40
High
$118.40
Median
$118.40
$0.27| $2,870.38) $3.46
$523.18) $653. 98 | $588.58
S32.81
$82.02) $57.41
S32.29) $53.82) $53.82
$32.29
$32.29) $32.29
S10.76) $10.76) $10.76
$1.24
S32.81
$10.76) $6.00
$49.21
$49.21
$1,000.00) $1.000.00) $1,000.00
$144.23) $144.23) $144.23
$741.61) $741.61
$741.61
$57.41 | $1.640.42) $656.17
$800.00) $800.00) $800.00
$588.58) $588.58) $588.58
$1.16) $86.11) $43.64
$60.00) $150.00) $105.00
$7.85) $10.46) S9.16
$125.00) $125.00) $125.00
$0.11) $0.11
$0.11
S37.04) $789.47) $37.04
$64.00) $64.00] $64.00
$0.00
$17.. 30
$0.00

-------
                                           Exhibit D-1
                            Unit Cost Ranges Used by the Expert Panels
Media = GROUND WATER
Remedial Activity (Cost Type (Unit
Activated alumina |0&M (cubic meters
Aerobic biological O&M
treatment |
Air stripping
liostimulation
>ioremediation
Carbon
adsorption/GAC
:atalytic
ncineration
(oxidation)
Chemical
precipitation
Discharge to POTU
(ground water)
Discharge to
surface water
(ground water)
Capital
cubic meters
blower
cubic meters
O&M (cubic meters
O&M (cubic meters
Capital (cubic meters
O&M (cubic meters
O&M
O&M
O&M
Capital
cubic meters
cubic meters
cubic meters
meters
O&M (cubic meters
Equalization (Capital (cubic meters
Extraction wells
Ground water
treatment
(unspecified)
Capital
O&M
O&M
meters
well
kilowatt hour
well
cubic meters
hours
ki lowatt hour
HOPE wall (Capital (square meters
Hydrocarbon
col lection/ recove-
ry
Interception
trench/ French
drain
O&M
Capital
cubic meters
meters
unit
Ion exchange |0&M (cubic meters
Leachate Capital
collection |
Monitoring labor
Monitoring wells
Offsite RCRA
landfill
O&M
Capital
meters
hours
sample
meters
well
O&M (well
O&M
(cubic meters
Unit Cost
Low
SO. 61
SO. 26
High
$0.61
$0.26
Median
SO. 61
$0.26
S22.000.00| S22.000.00j S22.000.00
SO. 08
S0.08| SO. 08
S0.04) $13.16]
SO. 42
SO. 42
S13.16| $13.16
$13.16
$0.42
$13.16
$13.16| S16.578.95| S16.578.95
S1.435.41
SO. 03
SO. 26
SI, 435. 41
S7.89
SO. 26
$1.435.41
$0.03
$0.26
$95. 14 | S95.14J S95.14
S0.26J SO. 26) SO. 26
$394. 74 | S394.74| $394.74
$164. 04 | $738.19) $328.08
$100. 00| $50.000.00) $7,500.00
$0.10| $0.10| $0.10
$1,000.00) $11,500.00) $1,000.00
$0.53) $7.89) $1.97
$28.85
$28. 85 | S28.85
$0.10) $0.10) $0.10
$53.82) $53.82) $53.82
$118.90
$118.90
$118.90
S145.46) $5,314.96) $984.25
$25,000.00) $105.000.00) $25.000.00
$0.79) $0.79) $0.79
$57.41
S984.25
$520.83
$62.50) $625.00) $125.00
$250.00) $2,500.00) $2,500.00
$328.08) $656.17) $328.08
$500.00) $28,250.00) $2,000.00
$2.500.00) $2.500.00) $2,500.00
$130. 8C
) $261. 5^
> $261.59
(CONTINUED)

-------
                                          Exhibit 0-1
                           Unit Cost Ranges Used by the Expert Panels
Media = GROUND WATER
Remedial Activity (Cost Type
ffsite disposal
ground water)
Offsite
ncineration
O&M 1
O&M
Pier for wells (Capital
Piping for ground-
water extraction
Product recovery
Njnps
Product recovery
wells
Pumps for ground-
water extraction
Pumps for
monitoring wel Is
Sampl i ng/ana I ys i s
Sealing/capping
abandoned wells
Sheet pile
Sludge
stabi lization/vit-
rif i cat ion
Capital
O&M
Capital
O&M
Unit
cubic meters
cubic meters
Unit Cost
Low
$356.71
$1.435.41
meters SI .640.42 |
meters
meters
pump
kilowatt hour
pump
Capital (well
O&M (well
Capital
O&M
Capital
O&M
pump
hours
kilowatt hour
pump
pump
hours
meters
report
sample
well
Capital [meters
Capital
cubic meters
square meters
Capital (cubic meters
O&M (cubic meters
Slurry wall (Capital (cubic meters
Stabilization/sol-
idification (in-
situ)
O&M
Subsurface drains (Capital
cubic meters
unit
Truck transport (O&M (kilometers
Truck transport
(offsite
treatment)
Ultraviolet
oxidation
O&M
O&M
Vapor phase carbon|Capital
kilometers
trip
cubic meters
unit
$15.00|
$124. 67|
$2, 000.00|
$0.10|
$2,000.00
$4.000.00
High
$356.71
$1,435.41
$1,640.42)
$3, 280. 84 |
$124. 67 |
$10,000.00]
$0.10
Median
$356.71
$1.435.41
$1,640.42
$96.78
SI 24. 67
$2,000.00
$0.10
$2,000.00) $2,000.00
$25,000.00) $25.000.00
$2, 000.00| $2.000.00) $2.000.00
$500. 00 | $50.000.00) $5,000.00
$2.88| $11.54| $11.54
$0.10| $0.10| $0.10
$500.00) $10,000.00) $5,000.00
$1,200.00
$2,500.00
$1,500.00
$2.31 | $250.00) $250.00
$500. 00 1 $500. 00 1 $500.00
$1.000.00) $15,000.00) $5,000.00
$40.00) $6.200.00) $1,200.00
$500.00) $500.00) $500.00
$2,624.67
$1,171.21
$2.624.67
$2,624.67
$1.171.21| $1,171.21
$22.97) $22.97) $22.97
$27.47| $27.47) $27.47
$19.62) $832.33) $156.95
$19.22) $38.44) $23.07
$1,105.26
$1,105.26
$1,105.26
$3,425.00) $3,425.00) $3,425.00
$2.49) $2.49) $2.49
$2.49) $2.49) $2.49
$600.00) $600.00) $600.00
SO. 26
$1,105.2*
> $0.79
$6,000.00] $6,000.00) $6,000.0
(CONTINUED)

-------
                                            Exhibit D-1
                             Unit Cost Ranges Used by the Expert Panels
Media = GROUND WATER

Remedial Activity (Cost Type (Unit
Vapor phase carbon |0&M (cubic meters
Water infi It rat ion Capital meters
basin


Low

S2.210.53
$492.13

Unit Cost
High

$198,947.37
$492.13


Median

$2.210.53
$492.13

Media = SOIL
Remedial Activity
Cost Type
Unit
Asphalt kilns (Capital (cubic meters
BEST solvent
extraction
Backfilling (soil)
Backhoe excavation
(soil)
Biostimulation
bioremediation (in
situ)
Confirmatory
sampling (soil)
Capital (cubic meters
Capital
Capital
Capital
Capital
O&M
cubic meters
hours
truck miles
cubic meters
hours
cubic meters
hours
sample
square miles
hours
sample
Consolidation (Capital (cubic meters
Disposal in
offsite Subtitle C
landfill
D i sposa I i n
offsite Subtitle D
landfill
Disposal in onsite
Subtitle D
landfill
Excavation and
hauling (soil)
Grade or compact
Capital
Capital
Capital
Capital
Capital
cubic meters
cubic meters
cubic meters
cubic meters
cubic meters
hours
Landf arming (Capital (cubic meters
Landf arming (in-
situ)
Low temperature
thermal recovery
Capital
Capital
hours
cubic meters
Unit Cost
Low
$118.90
$1,307.95
High
$118.90
$1,307.95
Median
$118.90
$1,307.95
$3.27| $52.32| $16.00
$100.00) $100. 00 | $100.00
$4.00| $4.00| $4.00
$3.27| $2.179.05) $7.85
$70.00) $100.00) $83.00
$117.72
$117.72
$117.72
$30.37) $125.00) $125.00
is. oo| $1,500.00) $300.00
$200.00) $1,200.00) $675.00
$125.00) $125.00) $125.00
$100.00) $2,000.00) $250.00
$2.94) $10.46) $7.85
$130.80
$98.10
$3.27
$3.01
$1,586.54
$163.49
$3.27
$1.046.36
S261.59
$118.90
$3.27
$10.46
$0.88) $5.23) $5.23
$90.00) $90.00) $90.00
$11.12) $11.12) $11.12
$25.00
$118.90
$25.00
$118.90
$25.00
$118.90
(CONTINUED)

-------
                                            Exhibit  D-1
                             Unit  Cost  Ranges Used by the Expert  Panels
Media = SOIL
Remedial Activity (Cost Type (Unit
Metals recovery
(soil)
Capital
cubic meters
Offsite disposal (Capital (cubic meters
Off site
incineration
Offsite
treatment/disposal
of haz solids
On-site treatment
plant
Capital
cubic meters
Capital (cubic meters
O&M (cubic meters
Capital
cubic meters
meters
Onsite disposal (Capital (cubic meters
Ons ite
treatment/disposal
of haz solids
Replace in unit
(soil)
Replace in unit
(waste)
Revegetation
Rotary kiln
incinerator
SVE blowers (in-
situ)
SVE monitoring (in
situ)
SVE vents (in-
situ)
Capital
Capital
Capital
Capital
Capital
O&M
Capital
O&M .
Capital
cubic meters
cubic meters
cubic meters
cubic meters
square meters
cubic meters
horsepower/hour
unit
sample
unit
cubic meters
meters
Unit Cost
Low
$5.23
High
$5.23
Median
$5.23
$392.39) $392.39) $392.39
$3.27
$1,046.36
$524.82
$261.59) $1, 046.36| $518.60
$523.18) $523.18) $523.18
$1.31
$1.31) S1.31
$98.43) $98.43) $98.43
$3.27) $5.23) $4.25
$5.23
$7.19
$1.31
$5.23
$10.46
$1.31
$5.23
$7.19
$1.31
$0.50) $7.85) $4.17
$0.17) $21,527.82) $0.49
$980.96
$0,10
$2,153.11
$0.10
$1.567.04
$0.10
$500.00) $1,000.00) $1.000.00
$100.00) $100.00) $100.00
$400.00) $400.00) $400.00
$26.16| $65.40| $65.40
$328.08) $328.08) $328.08
Samp I ing/ ana lysis (O&M (sample $375.00] $375.00) $375.00
Soil flushing (m-
situ)
Capital
cubic meters
Soil gas survey (investigation (hours
Soil Capital
neutralization |
Soi I vapor
extraction (SVE)
piping
Soil vapor
extraction (in-
situ)
Capital
Capital
O&M
cubic meters
meters
cubic meters
well
cubic meters
well
Soil washing (Capital (cubic meters
Stabilization/sot- Capital
idif ication
cubic meters
$130.80
$130.80
$130.80
$60. 00 | $250.00) $187.50
$26.16
$57.41
$26.16
$95.14
$26.16
$57.41
$23.78) $4,784. 69| $4,421.05
$6, 000. 00 | $6,000.00) $6,000.00
$7.93 | $6,078.95) $1,657.89
$1,000.00) $1,000.00) $1,000.00
$65.40) $392.39) $261.59
$23.54
$359.69
$113.79
(CONTINUED)

-------
                                            Exhibit D-1
                             Unit Cost Ranges Used by the Expert  Panels
Media = SOIL

Remedial Activity
Stabi I izat ion/sol-
idification (in-
situ)
Thermal desorbers
Truck transport
Truck transport
CSOl I)

Unspecified
incinerator
Vitrification


Cost Type
Capital
Capital
Capital
Capital


Capital
Capital


Unit
cubic meters
cubic meters
kilometers
cubic meters
kilometers
truck miles
cubic meters
cubic meters

Low

S2.55
$118.90
$2.49
$3.01
$2.49
$4.00
$356.71
$588.58
Unit Cost
High

$130.80
$261.59
$2.49
$4,203.75
$3.11
$5.00
$951.24
$588.58

Median

$24.20
$118.90
$2.49
$6.54
$2.49
$4.00
$951.24
$588.58
Media = SOURCE CONTROL
Remedial Activity (Cost Type (Unit
Asphalt cap
Capital (square meters
Unit Cost
Low
$0.84
High
$53.82
Median
$32.29
O&M (square meters $0.54| $0.54| $0.54
Asphalt liner (Capital (square meters $21.53| $27.02| $27.02
Backfilling (Capital (cubic meters $3.27) $130. 80 1 $17.66
Backhoe excavation
Capital
cubic meters ,$1.70| $130.80| $10.46
hours $75. 00 | $75.00| $75.00
O&M (cubic meters $7.85| $7.85| $7.85
Bags (Capital (unit
Berm/containment
wall
Clay Liner
Clay cap
Clay cap with vent
Compos ite/RCR A
cover
Compost i ng/Landfa-
rming (in-situ)
Concrete pad/ liner
Capital
cubic meters
hours
$50.00| $50.00| $50.00
$5.11 1 $1. 307.95| $19.08
$125.00) $125.00) $125.00
Capital (square meters $14.35| $14.35| $14.35
O&M (square meters $14.35| $14.35| $14.35
Capital (cubic meters $18.31) $31.39| $19.62
O&M (cubic meters $0.59| $0.59| $0.59
Capital (square meters $10.00| $50.00| $49.50
O&M (square meters $0.54| $0.54| $0.54
Capital
square meters $10.00| $101.66) $64.58
O&M (square meters $0.27| $21.53| $1.18
Capital
Capital
cubic meters $10.46| $11.12| $11.12
meters $9.84 | $9.84 | $9.84
cubic meters
hours
meters
square meters
$653.98) $653.98) $653.98
$64.00) $64.00) $64.00
$4.10) $4.10) $4.10
$10.76) $64.58) $64.58
(CONTINUED)

-------
                                           Exhibit D-1
                            Unit Cost Ranges Used by the Expert Panels
Media = SOURCE CONTROL
Remedial Activity (Cost Type (Unit
Concrete pad/liner|0&M (square meters
Consolidation (Capital (cubic meters
Construction of
onstte Sub C
landfill
Capital
square meters
Oewatering (Capital (cubic meters
Drum removal
Capital
cubic meters
truck miles
Equalization (Capital (cubic meters
Excavation and
Hauling
Capital
cubic meters
Floating cover (Capital (square meters
Grade or compact (Capital (square meters
Gravel cover
Landf arming
Leak detection
Capital
cubic meters
square meters
Capital (cubic meters
Unit Cost
Lou
$1.70
High
$1.70
Median
$1.70
$6. 54 | $20.00) $10.46
$22.00
S83.53
$22.00
$118.90) $118.90) $118.90
$9.81
$1.593.30) $9.81
$4.00) $4.00) $4.00
$400.00) $400.00) $400.00
$10.46
$20.00
$20.00
$7.21) $35. 31 | $7.21
$0.50) $32.81
$1.31
$1.31
$2.48
$1.31
$1.09) $16.15| $10.76
$11.12| $11.12) $11.12
O&M (cubic meters $1.03) $11.12) $6.07
Capital
hours
$62.50) $125.00) $93.75
unit $3.000.00) $3.065.00) $3.000.00
Mercury retorting (Capital (cubic meters
Neutralization (Capital (cubic meters
Off site Subtitle C
landfill
Offsite Subtitle D
landfill
Capital
Capital
cubic meters
cubic meters
Offsite disposal (Capital (cubic meters
Offsite
treatment/disposal
of haz liquid
Offsite
treatment/di sposal
of haz solids
Onsite Subtitle C
landfill
Onsite Subtitle D
landfill
Capital
Capital
Capital
cubic meters
cubic meters
cubic meters
hours
Capital (cubic meters
O&M (cubic meters
Onsite disposal (Capital (cubic meters
Onsite
treatment/disposal
of haz liquids
Other object
removal
Capital
Capital
cubic meters
cubic meters
meters
square meters
$17. 003.. 36) $17,003.36) $17,003.36
$95.69) $95.69| $95.69
$130.80
$130.80
$237.81
$13.16
$1,151.00
$237.81
$356.71
$261.59
$237.81
$478.47
$1.151.00
$237.81
$261.59
$237.81
$13.16
$1.151.00
$956.96) $478.47
$150.00) S150.00) $150.00
$3.27) $3.27) $3.27
$2.46) $2.46) $2.46
$1.31 | $1.31
$0.26
$0.26
$1.31
$0.26
$7.42 | $570.74) $227.99
$4.76| $16.73) $4.92
$40.36) $4.036.47) $45.75
(CONTINUED)

-------
                                            Exhibit D-1
                             Unit  Cost Ranges Used by the Expert Panels
Media = SOURCE CONTROL
Remedial Activity (Cost Type |Unit
Overpack drums (Capital (cubic meters
Unit Cost
Low
$2.392.34
High
$2,392.34
Median
$2,392.34
Pack in drums (Capital (cubic meters $2. 392. 34 | S2.392.36) S2.392.34
Pug mi 1 1 ing
Capital
cubic meters
meters
RCRA vault (Capital (cubic meters
Removal of free
liquids
Replace in unit
(waste)
Revegetation
Capital
Capital
Capital
$19.62) $50.00) $50.00
$4.76) $4.92| $4.84
$47.90) $47.90) S47.90
cubic meters S0.26| $197.37) SO. 26
hours SI. SO) S88.00) $33.44
meters
$47.41) $328.08) $187.75
square meters S9.1S) S9.15) S9.15
tank
cubic meters
square meters
O&M (square meters
Rip rap (Capital (square meters
Rotary ki In
incinerator
Capital
Run on/run off Capital
controls |
Seal coating
(asphalt)
Soi I cap
Soi I vapor
extraction (in-
situ)
Stabilization/sol-
idification
Stabilization/sol-
idification (in-
situ)
Steam
cleaning/washing
Synthetic cap
cubic meters
cubic meters
Capital (square meters
O&M (square meters
Capital
O&M
Capital
cubic meters
square meters
cubic meters
square meters
cubic meters
Capital (cubic meters
O&M (cubic meters
Capital
Capital
Capital
O&M
cubic meters
hours
meters
square meters
square meters
cubic meters
square meters
Synthetic liner (Capital (cubic meters
$6,667.00) $6.667.00) $6,667.00
$1.31
SI. 31
$1.31
$0.16) S0.50J $0.49
$0.27) $0.27
$0.27
$47.84) $86.11 $66.98
$951.24
$1.37
S2.049.56
S352.39
$1.500.40
$35.13
f 0.89 | $0.89) $0.89
$0.11
$7.85
$0.11) $0.11
$19.62) $11.44
$16.44) $16.44) $16.44
$0.14) $0.59) $0.59
$0.54) $1.18) $0.54
$4,421.05
$4.421.05
$4,421.05
$23.54) $2.392.34) $24.00
$52.32) $52.32) $52.32
$23.54
$130.80
$31 .39
$30.00) $203.00) $115.00
$6.56) $6.56) $6.56
$10.76) S10.76) $10.76
$10.76) $53.82) $37.67
$0.14) $0.14) $0.14
$0.01
$21.53) $0.54
$3.42) $34.17) $7.16
(CONTINUED)

-------
               Exhibit D-1
Unit Cost Ranges Used by the Expert Panels
Media = SOURCE CONTROL

Lo
Remedial Activity (Cost Type |Unit
Tank integrity
test
Capital (tank $5,
O&M (tank $1,
Tank location | Investigation
Tank removal
Capital
square meters
cubic meters
hours
Unit Cost
w High Median
000.00 $5,000.00 $5,000.00
000.00) $1,000.00) $1,000.00
$1.24 $1.24) $1.24
$20.27| $717.70) $32.26
$75.00) $150.00) $150.00
tank $5,000.00) $20. 000. 00 | $5.000.00
Thermal desorbers (Capital (cubic meters $555.88 $555.88 $555.88
Truck transport
Capital
O&M
cubic meters
kilometers
truck miles
kilometers
truck miles
Unspecified cover (O&M (square meters
Unspecified
incinerator
Vacuum extraction
(in-situ)
Vegetative cover
Vitrification (in-
situ)
Waste sampling and
analysis
Capital
Capital
Capital
cubic meters
hours
cubic meters
square meters
O&M (square meters
O&M
Capital
O&M
cubic meters
cubic meters
$5.52) $65.40) $15.70
$2.49) $31.07) $2.49
$4.00) $5.00) $4.00
$2.«9| $2.49) $2.49
$3.00) $3.00) $3.00
$17.66| $17.66) $17.66
$57.62 $2,400.00 $1,046.36
$88.00 $88.00 $88.00
$19.62) $19.62 $19.62
$a.45| $0.54) $0.49
$0.27) $0.54 $0.54
(594.52 $594.52 $594.52
$24.71 $239.23) $239.23
hours $125.00] $125.00) $125.00
report $3
.000.00) $3.000.00) $3,000.00
sample $100. 00| $2,000.00) $500.00
hours $125.00) $125.00) $125.00
sample $100.00) $1,500.00| $900.00
well $7
.000.00) $7.000.00) $7,000.00

-------
               Exhibit D-1
Unit Cost Ranges Used by the Expert Panels
Media = SURFACE WATER
Remedial Activity (Cost Type |Unit
Diversion and
col lection
(surface water)
Sediment sampling
Stabi I i zat i on/ero-
sion prevention
Surface water
monitoring
Capital
O&M
O&M
Capital
O&M
meters
pump
meters
hours
sample
square meters
hours
sample
Unit Cost
Lou
S9.02
High
$656.17
Median
$328.08
$1,025.00) S1.025.00j SI. 025. 00
$82.02) $82.02) 182.02
S125.00| $125. 00| S125.00
$1. 000.00) $1,000.00) $1,000.00
$107.64
$107.64
$107.64
$125.00) $125.00) $125.00
$1,000.00) $1, 000.00) $1,000.00

-------
                                   D-12
                               Exhibit D-2
                          Sources for Unit Costs
R.S. Means Company. MEANS Site Work Cost Data. Kingston, Massachusetts: R.S.
Means Company. Published annually.

Sobotka & Company. Experience Curves. Innovation and Diffusion: Application to
RCRA. Washington, D.C.: Sobotka & Company, April 1992.

U.S. Environmental Protection Agency. A Compendium of Technologies Used in the
Treatment of Hazardous Wastes. Washington, D.C.: U.S. Environmental Protection
Agency, September 1987.

U.S. Environmental Protection Agency. Guide to Treatment Technologies for Hazardous
Wastes at Superfund Sites.  Washington, D.C.: U.S. Environmental Protection Agency,
March 1989.

U.S. Environmental Protection Agency. Handbook - Remedial Action at Waste Disposal
Facilities.  Washington, D.C.: U.S. Environmental Protection Agency, 1985.

U.S. Environmental Protection Agency. In Situ Treatment of Hazardous Waste-
Contaminated Soils. Washington, D.C.: U.S. Environmental Protection Agency, January
1990.

U.S. Environmental Protection Agency. Seminar Publication - Corrective Action:
Technologies  and Applications. Washington, D.C.: U.S. Environmental Protection
Agency, September 1989.

U.S. Environmental Protection Agency. Subsurface Contamination Reference Guide.
Washington, D.C.: U.S. Environmental Protection Agency, October 1990.

U.S. Environmental Protection Agency. The Superfund Innovative Technology
Evaluation Program: Technology Profiles. Washington, D.C.: U.S. Environmental
Protection Agency, November 1990.
                    • • * DRAFT-March 23, 1993 • • •

-------
                                           D-13
D.2 Additional Analysis of Results

       Exhibit D-3 presents the total projected cost of corrective action broken out by media
and general activity type.
EXHIBIT D-3
CORRECTIVE ACTION COSTS BY ACTIVITY AND MEDIA
Total Net
Present Value Percent of Total
Activity Type Media (millions of $) Cost
RCRA Facility Investigation
RCRA Facility Investigation
Corrective Measures Study
Containment
Containment
Containment
Containment
Removal/Treatment of
Media
Removal/Treatment of
Media
Removal/Treatment of
Media
Removal/Treatment of
Media
Disposal
Disposal
Disposal
Institutional Controls
Other Activity
Other Activity
Other Activity
Other Activity
Other
Soil
Other
Waste
Ground Water
Surface Water
Air
Ground Water
Surface Water
Air
Soil
Waste
Ground Water
Soil
Other
Waste
Ground Water
Air
Other
1,800
1.3
210
220
1,400
26
4.1
5,800
0.4
15
4,000
160
360
160
11
0.3
31
4.5
91
9.7%
0.01%
1.12%
1.16%
7.6%
0.14%
0.02%
30.6%
< 0.01%
0.08%
21.62%
0.83%
1.94%
0.88%
0.06%
< 0.01%
0.16%
0.02%
0.49%
                            • • • DRAFT-March 23,1993 • • •

-------
              D-14
EXHIBIT D-3
CORRECTIVE ACTION COSTS BY ACTIVITY AND MEDIA
Total Net
Present Value Percent of Total
Activity Type Media (millions of $) Cost
Removal/Treatment of
Waste
Monitoring
Monitoring
Monitoring
Monitoring
Monitoring
Capping
Waste
Waste
Ground Water
Surface Water
Air
Soil
Waste
TOTAL
1,400
16
1,400
130
2.0
26
1,500
18,700
7.3%
0.09%
7.36%
0.7%
0.01%
0.14%
7.97%
100%
• * •
    DRAFT-March 23,1993 • • •

-------
                                      APPENDIX E

                         HUMAN HEALTH BENEFITS ANALYSIS
       This appendix discusses in detail the methodology EPA used for the human health
benefits analysis.  It lays out the analytical framework that the Agency used to arrive at
quantitative estimates of human health risk at the corrective action facilities, and lists all major
assumptions and values for variables used to quantify risk.  This appendix supplements the
summary of approach presented in Chapter 7.1

       This appendix is organized by major steps of EPA's risk assessment methodology.
Section E.1 presents topics related to the hazard identification and dose-response assessment.
Section E.2 presents topics of the exposure analysis. Topics related to the risk characterization
are presented in Section E.3.

E.1 Hazard IdentiGcation and Dose-Response Assessment

       In this section, EPA first presents the basis for determining which chemicals posed a
hazard at the corrective action facilities, and, therefore,  are included in the human health risk
assessment.  Next, the Agency discusses measures of hazard associated with each of these
chemicals. The discussion includes the types of hazard measures or health criteria used in the
RIA, and the sources that EPA used to obtain values for these toxicity measures.

       E.1.1 Selecting Chemicals of Concern

       EPA selected the chemicals to be evaluated for risk at each facility from a set of
approximately 220 chemicals. This set represents chemicals from Appendices VII (56 FR 7568,
Feb. 25, 1991) and IX (52 FR 25946, July 9,1987) that also have toxicity data (e.g., action levels,
maximum contaminant  levels (MCLs), slope  factors, and reference doses).  Of these, about 100
chemicals were  identified as major contaminants in SWMUs of concern at the 52 facilities
modeled for this benefits analysis.  These constituents, along with their toxicity data, are listed in
Exhibit E-l. The Agency used action levels from the corrective action proposed rule  (55 FR
30798, July 27,1990) when available, or derived them using the assumptions given in the
proposed  rule.  The MCLs listed in Exhibit E-l are as promulgated under the Safe Drinking
Water Act, up to July 1992.

       Exhibit E-l also lists the health criteria values representing the dose-response
information, i.e., slope factors and reference  doses,  that EPA used in the risk assessment.  EPA
did not develop new dose-response assessments for this  RIA, but instead, relied on the Agency's
existing standardized, peer-reviewed dose-response data. The slope factors and reference doses
are taken from EPA's Integrated Risk Information System  (IRIS), or, if not listed in IRIS, from
EPA's Health Effects Assessment Summary Tables  (HEAST, March 1992). The  following
section explains these measures of dose-response in greater detail.
   1 Related discussions of the MMSOILS model and its input parameters are presented in Appendix B.

                              ••* Draft -• March 23,1993 •••

-------
                 EXHIBIT E-l:
Chemicals Modeled and Health Criteria Values Used
CAS No
630206
79005
75343
75354
120821
95501
107062
156605
78875
106467
1746016
88062
120832
105679
121142
91941
108101
83329
67641
75058
107028
107131
107186
62533
120127
7440360
11141165
12672296
11097691
7440382
7440393
71432
92875
7440417
117817
75274
75252
85687
7440439
75150
56235
57749
108907
67663
Chancel
, 1 , 1 ,2-TETRACHLOROETHANE
,1.2-TRICHLOROETHANE
,1-DICHLOROETHANE
,1-OICHLOROETHYLENE
,2,4-TRICHLOROBENZENE
,2-DICHLOROBENZENE
,2-DICHLOROETHANE
,2-DICHLOROETHYLENE (trans-DCE)
,2-DICHLOROPROPANE
,4-OICHLOROBENZENE
2,3,7,8-TETRACHLOROOIBENZO-p-DIOXIN
2,4,6-TRICHLOROPNENOL
2,4-DICHLOROPHENOL
2.4-DIHETHYLPHENOL
2,4-DINITROTOLUENE
3,3'-DICHLOROBENZIDINE
4-METHYL-2-PENTANONE
ACENAPHTHENE
ACETONE
ACETONITRILE; METHYL CYANIDE
ACROLEIN
ACRYLONITRILE
ALLYL ALCOHOL
ANILINE
ANTHRACENE
ANTIMONY
AROCLOR 1232
AROCLOR 1248
AROCLOR 1254
ARSENIC
BARIUM AND COMPOUNDS
BENZENE
BENZIOINE
BERYLLIUM
BIS(2-ETHYLHEXYL)PHTHALATE
BROMODICHLOROMETHANE
BROMOFORM
BUTYL BENZYL PHTHALATE
CADMIUM
CARBON DISULFIDE
CARBON TETRACHLORIDE
CHLORDANE
CHLOROBENZENE
CHLOROFORM
NCL
m m
5.0E-03
--
7.0E-03
7.0E-02
6.0E-01
5.0E-03
1.0E-01
5.0E-03
7.5E-02
3.0E-08
--
--
--
--
--
--
--
..
--
..
..
--
.'
--
6.0E-03
5.0E-04
5.0E-04
5.0E-04
5.0E-02
2.0E+00
5.0E-03

4.0E-03
--
1.0E-01
1.0E-01
..
5.0E-03
5.0E-03
2.0E-03
1.0E-01
1.0E-01
AL AIR
1.0E+00
6.0E-01
3.5E*02
3.0E-02
1.0E+01
1.4E+02
4.0E-02
--
--
2.5E+03
2.3E-08
2.0E-01
--
--
--
--
7.0E+01
--
--
3.5E+01
3.5E-01
1.0E-02
--
1 .OE+00
--
--
-•
-,i .
-'
7.0E-05
4.0E-01
1.2E-01
2.0E-05
4.0E-04
--
9.0E-01
--
6.0E-04
1.0E+01
3.0E-02
3.0E-03
2.0E+01
4.0E-02
AL HgO
1.0E-02
6.0E-03
3.5E+00
5.8E-04
7.0E-01
3.2E+00
3.8E-04
7.0E-01
5.1E-04
1.5E-02
2.2E-10
2.0E-03
1.0E-01
7.0E-01
5. IE-OS
8.0E-05
2.0E+00
2.1E+00
4.0E+00
2.0E-01
1.0E+02
6.0E-05
2.0E-01
6.0E-03
1.1E+01
1.0E-02
5.0E-06
5.0E-06
5.0E-06
2.0E-05
1.7E+00
1.2E-03
2.0E-07
8.0E-06
3.0E-03
3.0E-05
7.0E-01
7.0E+00
1.8E-02
4.0E+00
3.0E-04
3.0E-05
7.0E-01
6.0E-03
AL SOIL
3.0E+02
1 .OE+02
8.0E+03
1.0E+01
2.0E+03
7.2E+03
8.0E+00
1.6E+03
1.0E+01
2.9E+02
4.5E-06
4.0E+01
2.0E+02
1.6E+03
1.0E+00
2.0E+00
4.0E+03
4.8E+03
8.0E+03
5.0E+02
--
1.0E+00
4.0E+02
1.0E+02
2.4E+04
3.0E+01
9.0E-02
9.0E-02
9.0E-02
8.0E+01
4.0E+03
2.4E+01
3.0E-03
2.0E-01
5.0E+01
5.0E-01
2.0E+03
2.0E*04
4.0E+01
8.0E+03
5.0E+00
5.0E-01
2.0E+03
1.0E+02
RFD 0
3.0E-02
4.0E-03
1.0E-01
9.0E-03
1.0E-02
9.0E-02
--
2.0E-02
--
--
--
•-
3.0E-03
2.0E-02
2.0E-03
•-
5.0E-02
6.0E-02
1.0E-01
6.0E-03
2.0E-02
--
5.0E-03
-•
3.0E-01
4.0E-04
•-
•-
-•
3.0E-04
7.0E-02
--
3.0E-03
5.0E-03
2.0E-02
2.0E-02
2.0E-02
2.0E-01
5.0E-04
1.0E-01
7.0E-04
6.0E-05
2.0E-02
1.0E-02
RFD 1

••
1.0E-01
* -
2.6E-03
5.7E-02
•-
•-
1.1E-03
2.0E-01
--
--
•-
--
--
--
2.3E-02
•-
--
1.0E-02
6.0E-06
5.7E-04
--
2.9E-04
••
--
••
•-
•-
--
1.0E-04

--
•-

--
--
•-
2.9E-03
6.0E-03
SF 0
2.6E-02
5.7E-02
--
6.0E-01
••
--
9.1E-02
--
6.8E-02
2.4E-02
1 .6E+05
1.1E-02
--
•-
6.8E-01
4.5E-01
--
-•
--
--
--
5.4E-01
••
5.7E-03
••
••
7.7E+00
7.7E*00
7.7E+00
1.7E*00
-•
2.9E-02
2.3E+02
4.3E+00
1.4E-02
1.3E-01
7.9E-03


1.3E-01
1.3E+00
6.1E-03
SF I
2.6E-02
5.7E-02
--
1.8E-01
••
•-
9.1E-02

•-
•-
1.5E+05
1.1E-02

•-
••
--


-•
--

2.4E-01
• "



•-
~-

1.5E+01

2.9E-02
2.3E+02
8.4E+00
• •
3.9E-03

6.1E+00
5.3E-02
1.3E+00
8.1E-02

-------
                    EXHIBIT E-l:
Chemicals Modeled and Health Criteria Values Used (cont.)
CAS No
7440473
1319773
57125
72548
72559
50293
84742
60571
122394
145733
72208
100414
106934
75218
86737
7782414
50000
118741
302012
74908
78831
78591
7439921
58899
108316
7439976
16752775
72435
74873
71556
78933
75092
86306
91203
7440020
98953
56382
82688
87865
108952
85449
1336363
129000
110861
7782492
7440224
93721
Chorical
CHROMIUM (VI)
CRESOLS
CYANIDES (SOLUBLE SALTS AND COMPLEXES)
ODD
DDE
DDT
D I BUTYL PHTHALATE
DIELDRIN
DIPHENYLAMINE
ENDOTHALL
ENDRIN
ETHVLBENZENE
ETHYLENE Dl BROMIDE
ETHYLENE OXIDE
FLUORENE
FLUORINE
FORMALDEHYDE
HEXACHLOROBENZENE
HYDRAZINE
HYDROGEN CYANIDE
ISOBUTYL ALCOHOL
ISOPHORONE
LEAD
LINDANE
MALE 1C ANHYDRIDE
MERCURY
METHOMYL
METHOXYCHLOR
METHYL CHLORIDE
METHYL CHLOROFORM
METHYL ETHYL KETONE
METHYLENE CHLORIDE
N-N1TROSODIPHENYLAMINE
NAPHTHALENE
NICKEL
NITROBENZENE
PAR AT HI ON
PENTACHLORONITROBENZENE (PCNB)
PENTACHLOROPHENOL
PHENOL
PHTHALIC ANHYDRIDE
POLYCHLORINATED 8IPHENYLS (PCBS)
PYRENE
PYRIDINE
SELENIUM
SILVER
S1LVEX (2,4.5-TP)
MCL
1.0E-01
--
2.0E-01
--
..
..
--
--
• -
1.0E-01
2.0E-03
7.0E-01
5.0E-05
--
--
..
..
1.0E-03
--
--
--
--
1.5E-02
2.0E-04
--
2.0E-03
..
4.0E-02
--
2.0E-01
5.0E-03
--
..
1.0E-01
--
..
--
1.0E-03
--
..
5.0E-04
..
..
5.0E-02
5.0E-02
5.0E-02
AL AIR
9.0E-05
--
--
--
-.
1.0E-02
--
2.0E-04
--
--
--
1.0E+03
5.0E-03
1.0E-02
--
--
8.0E-02
2.2E-03
2.0E-04
--
--
--
1 .5E+00
--
--
• •
--
--
^5.6E+00
1.0E+03
3.0E+02
3.0E-01
--
--
4.2E-03
2. OE+00
..
1.0E-01
--
--
--
--
--
L _
3.5E+00
--
--
AL HjO
1.8E-01
2. OE+00
7.0E-01
1.0E-04
1.0E-04
1.0E-04
4. OE+00
2.0E-06
9.0E-01
7.0E-01
1.1E-02
4. OE+00
4.0E-07
1.0E+02
1 .4E+00
2.1E+00
7. OE+00
2.2E-05
1.0E-05
7.0E-01
1.0E+01
9.0E-02
1.0E+02
2.7E-05
4. OE+00
1.1E-02
9.0E-01
1.8E-01
2.7E-02
3. OE+00
2. OE+00
5.0E-03
7.0E-03
1.4E-01
7.0E-01
2.0E-02
2.0E-01
1.0E-01
1. OE+00
2.0E+01
7.0E+01
4.5E-06
1.1E+00
4.0E-02
1.1E-01
1.1E-01
2.8E-01
AL SOIL
4.0E+02
4.0E+03
2.0E+03
3. OE+00
2. OE+00
2. OE+00
8.0E+03
4.0E-02
2.0E+03
2.0E+03
2.0E+01
8.0E+03
8.0E-03
--
3.2E+03
4.8E+03
1.6E+04
4.4E-01
2.0E-01
2.0E+03
2.0E+04
2.0E+03
5.0E+02
5.0E-01
8.0E+03
2.0E+01
2.0E+03
4.0E+02
5.4E+02
7.0E+03
4.0E+03
9.0E+01
1.0E+02
3.2E+02
2.0E+03
4.0E+01
5.0E+02
2.0E+02
2.0E+03
5.0E+04
2.0E+05
9.1E-02
2.4E+03
8.0E+01
2.4E+02
2.0E+02
6.4E+02
RFD 0
5.0E-03
--
2.0E-02
--
--
5.0E-04
1.0E-01
5.0E-05
2.5E-02
2.0E-02
3.0E-04
1.0E-01
--
--
4.0E-02
6.0E-02
2.0E-01
8.0E-04
••
2.0E-02
3.0E-01
2.0E-01
--
3.0E-04
1.0E-01
3.0E-04
2.5E-02
5.0E-03
--
--
5.0E-02
6.0E-02
-•
4.0E-02
2.0E-02
5.0E-04
6.0E-03
3.0E-03
3.0E-02
6.0E-01
2. OE+00
••
3.0E-02
1.0E-03
5.0E-03
5.0E-03
1.0E-02
RFD 1 SF 0
..
--
..
2.4E-01
3.4E-01
3.4E-01
..
1.6E+01
..
--
--
2.9E-01
8.5E+01
1. OE+00
--
--
--
1.6E+00
3. OE+00
--
--
9.5E-04
--
1.3E+00
- -
--
•-
--
1.3E-02
2.9E-01
2.9E-01
8.6E-01 7.5E-03
4.9E-03
..
-•
6.0E-04
- -
2.6E-01
1.2E-01

--
7.7E+00
..
--
1.0E-03
• -
~ ~
SF 1
4.2E+01
•-


--
3.4E-01
•-
1 .6E+01
--
•-
•-
••
7.6E-01
3.5E-01
•-
--
4.5E-02
1.6E+00
1.7E+01
••

--

•-




6.3E-03

1.6E-03

• ~
8.4E-01

• •
• •
• ~
* *
• •
" *

• •

* *
"

-------
                     EXHIBIT E-l:
Chemicals Modeled and Health Criteria Values Used (cnnt.)
CAS No
100425
127184
78002
108883
79016
7440622
1314621
108054
75014
1330207
7440666
319846
319857
108394
MCL
ALAIR
ALH20
AL SOIL
RFDO
RFD 1
SFO
SF1
Clinical NCL
STYRENE 1.0E-01
TETRACHLOROETHYLENE 5.0E-03
TETRAETHYL LEAD
TOLUENE 1 .OE+00
TR1CHLOROETHYLENE 5.0E-03
VANADIUM AND COMPOUNDS
VANADIUM PENTOXIDE
VINYL ACETATE
VINYL CHLORIDE 2.0E-03
XYLENES (NIXED) 1.0E+01
ZINC
alpha-BHC
beta-BHC
m-CRESOL
= Maximum Contaminant Level (mg/L)
= Action level Tor air (ug/m3)
= Action level for water (mg/L)
= Action level for soil (mg/kg)
= Oral reference dose (mg/kg-day)
= Inhalation reference dose (mg/kg-day)
= Oral slope factor (mg/kg-day)'1
= Inhalation slope factor (mg/kg-day)'1
AL AIR
1.7E+00
1.0E+00
--
7.0E*03
2.1E-01
--
--
2.0E*02
1.2E-02
1.0E*03
--
5.6E-04
1.9E-02
~ ~








AL HgO
7.0E+00
7.0E-04
4.0E-06
1.0E+01
3.2E-03
2.5E-01
3.0E-01
3.5E+01
1.8E-05
7.0E+01
7.0E+00
5.6E-06
1.9E-04
2.0E+00








AL SOIL
2.0E+04
1.0E+01
8.0E-03
2.0E+04
6.0E+01
5.6E*02
7.0E+02
8.0E*04
3.7E-01
2.0E+05
1.6E+04
1.1E-01
3.9E+00
4.0E+03








RFD 0
2.0E-01
1.0E-02
1.0E-07
2.0E-01
--
7.0E-03
9.0E-03
1.0E+00
--
2.0E+00
3.0E-01

5.0E-02








RFD 1 SF 0 SF 1
2.9E-01
. • • • - -
mm - • ™ -
1.1E-01
..
•-
--
5.7E-02
1.9E*00 3.0E-01
..
--
6.3E*00 6.3E+00
1.8E+00 1.8E+00
• • • • - •









-------
       E.1.2  Selecting Health Criteria Values

       For risk assessment purposes, individual chemicals are separated into two categories
based on whether they elicit non-carcinogenic or carcinogenic adverse health effects upon
exposure. This classification of chemicals into carcinogens and non-carcinogens (systemic
toxicants) is rooted in the concept of dose threshold.  For non-carcinogenic or systemic effects,
protective physiological mechanisms exist that must be overcome before the adverse effect is
manifested.  Such a threshold, however, is thought to be absent in the case of carcinogens, i.e.,
any level of exposure, however small, could result in cancer.

       Non-carcinogenic Chemicals

       To quantitatively assess risk for non-carcinogenic effects, EPA uses a hazard ratio
approach based on the reference dose (RfD). The RfD for a particular chemical is defined as an
estimate of the maximum daily exposure level,  for humans, that is likely to be without
appreciable risk of adverse health effects during a lifetime.  EPA currently develops RfDs for
oral exposure routes, and a closely related parameter, the reference concentration (RfC), for
inhalation exposure effects. For many chemicals, the RfD approach has produced useful
quantitative estimates of the toxicity threshold, and thus, has been used as a "benchmark" on
which to consider regulatory decisions in relation to potential impacts on human health.  In
essence, the purpose of the RfD is to provide a "safe" level to which chemical intakes into the
body (e.g., those projected from human exposure to contaminated environmental media) might
be compared. Intakes that are Jess than the RfD are not likely to be of concern. Intakes that
are greater than the RfD indicate an increased probability of adverse effects; however, that
probability is not a certainty.  (Note that the ratio of the intake to the RfD is not a measure of
the probability of the adverse effect.)

       Carcinogenic Chemicals

       Because the hypothetical mechanism for carcinogenesis is referred to as "nonthreshold,"
and no dose is thought to be risk-free, RfDs cannot be calculated for carcinogens. Instead,
lifetime cancer risks associated with various levels of exposure to carcinogens are estimated by
using a value called a cancer slope factor (SF)  expressed in units of [mg chemical/kg body
weight/day]'1. The slope factor is a plausible upper-bound estimate of the probability of an
individual developing cancer over a lifetime as a result of exposure to a potential carcinogen.
This probability is referred to as the excess lifetime cancer risk because it is in addition to  cancer
risks the individual incurs that are not related to the chemical. The actual risk can vary from a
lower bound of zero to the upper bound represented by the slope factor.  Upper bound in this
case denotes that the risk estimate is unlikely to be underestimated although it may very well be
overestimated.

       Lead

       In current risk assessments, EPA considers lead to be in a unique category of chemical
toxicity.  Although lead is a non-carcinogen, EPA  has determined that an RfD should not be
developed for lead for two primary reasons. First, an appreciable threshold  is absent for many of

                              •** Draft ~ March 23,1993 »**

-------
 the noncancer effects of lead. Second, humans are exposed to lead in a variety of multimedia
 exposure scenarios.

        (a)    Absence of an appreciable threshold.  Although there is no widely accepted
               theoretical basis for the absence of an appreciable threshold for many of the
               health effects associated with lead exposure, the data currently available are not
               sufficient to adequately define the dose-response relationship for many of the
               toxic effects of lead. That is, even though there are considerable human exposure
               and health effects data for lead, it is unclear what exposure level is without
               appreciable risk of adverse health effects.

        (b)    Multimedia exposure to lead.  Humans are exposed to lead from a variety  of
               media.  The'relative contribution of each medium to total lead uptake changes
               with age and can  vary greatly in magnitude on a site-specific basis.  Because they
               are route-specific, RfDs cannot incorporate either site-specific information on
               lead exposure sources (e.g., from multiple sources) or age-related data (e.g.,
               varying susceptibility with  age).  For example, an inhalation RfC is an estimate of
               the air concentration to which human populations can be exposed for a lifetime,
               in the absence of exposure from other sources, without appreciable risks of
               adverse health effects.  Inhaled lead, however, generally contributes only a
               fraction of the total lead intake; most adults are exposed primarily from dietary
               sources and through occupational exposures.  The RfC method would not be able
               to consider these  additional intakes.

       Although EPA has elected not to calculate an RfD for the noncancer effects of lead
exposure, analyses showing correlations between human blood levels and health effects indicate
that associations may persist through the lowest blood lead levels tested (^  10-15 ug/dL).
Therefore, it is possible that if a threshold for the noncancer effects of lead were to exist, it
would lie within or below the 10-15 ug/dL range.  Note, however, that the range 10-15 ug/dL is
regarded as a "level of concern," warranting attention from a medical viewpoint, and not a dose
level or threshold  below which no adverse health effects would be expected to occur (i.e.,  it is
not strictly parallel to the definition of an RfD).  For this RIA, the Agency used 10 ug/dL as the
effects threshold for lead.

E.2 Exposure Analysis

       In the context of estimating risk at the corrective action facilities, the key steps for
exposure assessment  include:

       •      identifying principal exposure pathways and routes;
       •      characterizing exposed human  populations;
       •      selecting appropriate exposure assumptions and contact rates; and
       •      calculating intakes at exposure points.
                               *** Draft - March 23,1993 •••

-------
 Discussed below are methods and sources that the Agency relied on to complete each of the
 above steps.

        E.2.1  Identifying Exposure Pathways, Exposure Routes, and Exposure Points

        At the offset of this RIA, EPA identified a variety of human exposure pathways and
 exposure routes that could possibly contribute to risk at the corrective action facilities.  The
 Agency selected five pathways for analysis: ground water, surface water, air, soil, and foodchain.
 Within each of these pathways, EPA evaluated multiple exposure subpathways (e.g., beef
 ingestion within the foodchain pathway) and exposure routes (e.g., ingestion, inhalation, and
 dermal contact in the ground water pathway) as relevant. Note that not all pathways are
 applicable at all facilities.  Exposure  pathways, exposure routes, and exposure points are shown in
 Exhibit E-2.

        Once the exposure  pathways and exposure routes were identified, EPA determined the
 exposure point locations at each facility using best available facility-specific data. Exposure
 points occur wherever humans ingest, inhale, or make dermal contact with contaminated media.
 The exposure routes and exposure point locations for each pathway are discussed below.

        Ground Water

        EPA estimated risks from ground water as the sum of risks from three exposure routes:

        •     ingestion of contaminated drinking water,
        •     inhalation of volatile compounds during household use of ground water,  and
        •     dermal exposure due to direct contact while showering.

 Ground water may also contribute to contamination of surface water (via ground water  discharge
 to surface water bodies), and foodchain contamination  (via use of contaminated ground water for
 irrigating crops and/or watering cattle).  EPA estimated these risks under the secondary
 pathways.

       Exposure points for this pathway were confined to a 90° sector for  up to 2 miles in the
 downgradient direction from the facility.  (Plumes were assumed to attenuate and dilute beyond
 2 miles, except at a few facilities where information to the contrary was available2.) Essentially,
 exposures occurred wherever there was a ground-water well in this sector.  EPA determined the
 number and general location of private wells within 2 miles downgradient of the facility and the
 number of residents that they serve, as well as the general location of and population served by
 public water supply wells. EPA assumed that any wells outside of this 90°  sector would not be
 affected by contaminated ground water.
   2 At facilities where the plume concentrations exceeded action levels at the two-mile boundary, EPA
determined the number and location of pubb'c and private wells up to five miles downgradient of the
facility. The Agency then considered these as additional exposure point locations, and estimated
concentrations of contaminants at these points as well.

                              ••• Draft ~ March 23,1993 •••

-------
             EXHIBIT E- 2
EXPOSURE MEDIA, PATHWAYS, AND ROUTES
Relec










| sw









ises










MU









On-Slte or
Adjacent
1 Air 1
	 H n \




r 	 H Surface Water
I


| 	 *] Surface Soil







	 I UaHnoa 7nna












' 	 1
|







































































At Point of
Exposure
-• Air 	
»
1 Biota
|
I .
1 . ^
-• Surface Water ^ 	
-H 	 LI^_J--^
T

•
[-•T Surface Soil [±**^

1
1

Btola |



Exposure
Route
— »| Inhalation 1 	 1
1 	 1


^ Ingestlon | —

— • Contact 1 —

^* Inhalation | —

^~*\ Ingestlon 1 —

— — J Contact 1


	 « Ingestlon [—

^^J IngesDon 1
	 » Contact 1 	
* Inhalation | —










_•.! Receptor 1
1 r J









                                          F2D02B-3

-------
       EPA defined the exposure point for calculating individual risk as the well (private or
public) located closest to the facility within the 90° sector. Exposure concentrations at all wells
within any distance range were estimated as being equal to the average concentration over that
distance  range.1 Exposure point locations for calculating population risk were defined as the
center points of the six distance ranges, and, where applicable, as the exact locations of public
wells.  For public wells, all persons served were assumed to be exposed at the well location.

       EPA estimated individual and population risks only where facility-specific data indicated
that wells are actually present and the ground water is being used.

       Air

       The only exposure route EPA considered for the air pathway was inhalation of airborne
contaminants.  To evaluate individual and population risk via air exposures, EPA identified
several exposure point locations for this pathway.

       First, EPA defined  the exposure point for individual risk as the location of the residence
determined to be closest to the facility. Next, data were collected for populations within a 10
kilometer radius of each facility, sectored into 8 directions and 5 distance ranges.  EPA assumed
every person within each range to be exposed to the concentration at the center point of that
range. (Note that MMSOILS predicts an average air concentration  at a specific distance, and is
not direction-specific, i.e., the same concentration would be predicted at any point  on the circle
defined by that distance.) Distances from sources (i.e., SWMUS with releases to air) to the
center points of distance ranges were calculated by adding the minimum distance between the
SWMU and facility boundary to the distance from the facility boundary to the center point of
each range.

       Surface Water

       EPA evaluated surface water risks for two different exposure scenarios:  (1) recreational
water use, and, where applicable, (2) domestic water use.  Surface water is also an intermediate
transport route for contaminants in several other exposure pathways, such as ingestion of
contaminated fish  and ingestion of agricultural products that have been irrigated with
contaminated surface water.  These risks  are accounted for in the foodchain pathway.

       Exposure locations  for the surface water pathway (recreational use)  were determined
based on known uses (swimming and/or fishing) of surface waters downstream of a facility. In
most cases, EPA defined the closer use point as the exposure location for both  swimming and
fishing. In cases where facility-specific information on the distance to these use points was not
   3 Note that the highest individual ground water risk could occur at residences other than the one
nearest the facility boundary. For example, plumes resulting from past releases could have moved
downgradient beyond the nearest residence.  Alternatively, households more distant than the nearest
household could be located closer to the downgradient direction of the contaminant plume.

                               ••• Draft ~ March 23, 1993 •••

-------
                                             10

 available, EPA assumed swimming and fishing to occur at a default distance of 10 meters
 downstream from the point where the surface water received contamination from the facility.

        For recreational use, the Agency estimated individual risk from surface water exposure
 caused by swimming only.  This risk was calculated as the sum of two exposure routes:

        •     dermal absorption via direct contact with surface water, and
        •     incidental ingestion of surface water while swimming.

 As mentioned above, individual risk due to ingestion of fish from the contaminated surface water
 is calculated under the foodchain pathway. Population risks for recreational uses (swimming and
 fishing) were not calculated because the populations exposed through those pathways are usually
 small, and it is difficult to devise reasonable assumptions for the number of people  that utilize
 surface water bodies for recreation.

        For domestic water use, the exposure routes are the same as those considered for ground
 water.  In this case, the exposure point for both individual and population risk (for  all affected
 populations) is defined as the point of intake for a municipal water supply identified on the
 surface water (i.e., exposure concentrations are estimated at the intake point).

        Soil

        EPA calculated risks through the soil pathway due to both incidental ingestion and direct
 contact (dermal absorption).  The contaminated  soil transport pathway also played  an
 intermediate role in  foodchain, surface water, and air pathways. These secondary exposures were
 accounted for  in the respective pathways.

        EPA defined two separate exposure point locations for calculating individual risk via the
 soil pathway. The nearest off-site field is  identified as the exposure point located at the distance
 from the facility to the nearest field likely for human exposure to soil (e.g., parks, playgrounds,
 and residential yards). The second exposure point, nearest off-site agricultural field, is the
 nearest field located at the distance from  the facility to the nearest field where human exposure
. to soil  through agricultural or gardening activities could take place.  This field could be actual
 agricultural land or a vegetable garden at  the closest  residence.

        EPA did not calculate population risk from exposure to contaminated soil due to the
 complexity of estimating spatial distribution of populations around these fields and  likelihood of
 population members using these fields for various activities.

        Foodchain

        The foodchain pathway accounts for exposure due to ingestion of contaminated
 vegetables, milk, beef, and fish.

        Exposure concentrations in contaminated vegetables, milk, and beef are estimated at the
 agricultural field identified as being closest to the facility for the soil pathway.  Individual risk is

                               *** Draft - March 23,1993 *»*

-------
                                            11

calculated for a person ingesting the contaminated vegetables, beef, and milk "originating" from
this field. EPA estimated individual risk for a person ingesting contaminated fish caught from
surface water contaminated by releases from a facility where recreational water use is identified.

       The Agency did not calculate population  risks for the foodchain pathway because of the
difficulties in estimating the number of persons potentially exposed.

       E.2.2  Characterizing Exposed Populations

       There are two categories of exposures that EPA characterized for risk assessment in this
RIA:  exposures to individuals and exposures to  populations.  As noted in Section E.2.1 above,
population exposures are evaluated only for the ground-water, air, and surface water pathways.
For this analysis, EPA assumed that both current and future populations would be potentially
exposed to contamination from the facilities.

       Current Populations

       Based on topographic maps and interviews with municipal officials, EPA determined the
number people served  by public and private wells in  1992 near the  corrective action facilities.
'EPA used these sources to identify numbers of public and private wells and the number of
people served by public wells. To estimate numbers  of people served by private wells, the
Agency multiplied the  number of private wells by an  assumed average of 2.6 persons per well
that is based on the 1990 Census  (2.6 persons per residence). The  total population estimated for
public and private wells was assumed  to be the current population  potentially exposed via ground
water.

       Likewise, for the air pathway,  the population  residing within a 10-kilometer radius of
each  facility was considered to be potentially exposed via air pollution.  This population was
identified using the Graphical Exposure Modeling System (GEMS) to count all individuals
residing within 10  kilometers of the facility boundaries (based on 1980 U.S. census; counts were
updated to  1992 numbers - see next section) and using topographic maps for distances closer to
the facility (i.e., individual  residences  were identified within a two-mile boundary of the facility).
Finally, EPA determined the population served by municipal water supplies with intakes on
contaminated surface waters using Federal Reporting Data System (FRDS). These populations
were then assumed to  be potentially exposed via household use of  surface water.

       Future Populations

       EPA predicted the size of future populations that would potentially be exposed via the
ground water, air, and surface water pathways based  on the size of current exposed populations.
This was accomplished by multiplying the current population counts in local areas surrounding
facilities (population within 10 kilometers of the facility or population served by private and/or
public wells, henceforth referred to as "local populations") by population growth rates.

       The Agency used commercially available population forecasts to calculate growth rates
for local populations.  These forecasts contained county-level population estimates for the years

                              ••• Draft - March 23,1993 *••

-------
                                            12

 1970 through 2015.  The county-level population estimates were used to calculate population
 growth rates between years 1980 and 2120. Because the population forecasts available extend
 only to 2015, the Agency calculated population growth rates for 1980 through 2015 differently
 from the growth rates for years beyond 2015 (i.e., 2016 through 2120).  For the years prior to
 and including 2015, the growth rate for any year i was calculated as the ratio of population count
 in year i to the population count in year i-L For example, the growth rate for the year 2001
 equaled:

              Population county, 200, -*- Population countyT20oo

 EPA estimated  the annual growth rate for years 2016 through 2120, using the average of all
 annual growth rates between 1970 and 2015.

        This method assumes that local populations grow at the same rates as the entire
 populations of the counties in which the facilities are located. Many of the facilities are  located
 near (i.e., within 5 miles of) county boundaries.  EPA purchased county-level population data for
 all counties within 5 miles of any facility, as determined using U.S.G.S  quadrangle maps  and the
 Rand McNally 1990 Road Atlas of the United States, Canada, and Mexico (66th edition).  For
 facilities with multiple counties, EPA first calculated growth rates for each county using the
 method described above.  The growth rates were then averaged using total county populations as
 weighting factors, and this average was considered  the growth rate for the facility.

        County Population  Projection Method

        This section discusses the underlying data that EPA used for calculating growth rates.
 The discussion includes models and assumption employed in the original forecasting.

        Population growth rates used in the human health risk assessment are based on
 demographic data that the Agency purchased from Woods & Poole Economics, Inc.  These data
 contain population estimates by county for each year from 1969 to 2015.  The basis of annual
 population counts varies as follows:

        •      1969 to 1980:  Population estimates from the U.S. Census Bureau;

        •      1981 to 1990:  U.S. Census data adjusted by Woods & Poole
                            Economics, Inc. to reflect 1990 Census counts; and

        •      1991 to 2015:  Forecasts by Woods & Poole Economics, Inc.

        Population estimates for the years 1981 to 1990 are adjusted estimates based on 1980 and
 1990 census counts.  The Census Bureau used 1980 census results to estimate populations for the
years 1981 to 1989. Woods & Poole Economics, Inc. adjusted the Census Bureau estimates
using 1990 census results. The purpose of this adjustment was to provide a smooth trend
between 1980 and 1990  and to increase the accuracy of population estimates for these years.
                              **• Draft •- March 23,1993 *•*

-------
                                            13

       The Census Bureau population statistics, on which all population estimates are based, are
residential populations (i.e., counts of persons residing in census areas). The census bureau
definition of residential population includes civilians, military personnel, college residents, institu-
tional residents (i.e., prison inmates, mental institution residents, hospital patients, and nursing
home residents).  These counts include estimated numbers of undocumented aliens and exclude
U.S. citizens living in foreign countries.

       The county-level population data purchased included projections for the years 1991 to
2015. The Woods & Poole projections integrate historical data, macroeconomic modeling, and
demographic  modeling to simulate population and economic trends for the U.S., economic
regions, and counties. Three sets of models are used for these estimates.  County models
estimate population based on county-level demographic and employment statistics.  Regional
models estimate populations using county-level demographic and employment statistics. A
national macroeconomic model produces national totals for demographic and economic statistics
(e.g., population, GNP, inflation, employment).  National totals are used as upper bounds for
aggregated regional model results, and regional estimates are used as upper bounds for
aggregated county results.  All counties are modeled simultaneously to allow trends in one
county to influence trends in other counties.  Comparison of forecast results to historical  data
show that national and regional  statistics produced by Woods &  Poole have  an average root-
mean-squared error under 3%.

       The county level demographic model forecasts population trends on the basis of rates of
birth, death, and migration. Migration patterns are related to projected employment and
earnings. The Woods & Poole demographic model is a traditional cohort-component model. In
such models,  the population at a given time is disaggregated by demographic sub-groups (e.g.
age, sex, and  race) into a matrix of population counts.  The population in the next time step is
calculated by  multiplying each sub-population by natality, fertility, and migration rates specific to
that sub-population.  Differential rates for sub-groups result in population trends over a number
of subsequent time steps.  Woods & Poole disaggregates population counts by five-year age
groups, sex, and race (black, white, and other). Population changes are modeled with a one-year
time step.

       Highly-Exposed Subpopulations

       When applicable, EPA considered three special cases for individual risk: a subsistence
farmer, a subsistence fisherman, and  a pica child. These individuals were assumed to represent
subpopulations whose exposure was expected to be significantly different from the larger
population due mainly to lifestyle and economic factors.  EPA defined subsistence farmers as a
subpopulation having higher intake of contaminated vegetables, beef, and milk (i.e, a higher
percentage of their daily consumption of these  foods is from the contaminated source). Subsis-
tence fishermen were  defined as a subpopulation having a higher consumption of fish from the
contaminated source.  The pica child represents a subpopulation of children having an
abnormally high soil ingestion rate.

       Individual risks to both subsistence lifestyles were calculated using the same exposure
concentrations in food as for the larger population, but intake assumptions such as exposure

                              ••• Draft ~ March 23,1993 •••

-------
                                             14

frequency and amount ingested were increased to account for higher exposures. Similarly, pica
children were assumed to be exposed to the same soil concentrations as other children, but their
ingestion rates were higher.

       E.2J Selecting Exposure, Uptake, and Transfer Factors

       In order to quantify human exposure to chemicals from the corrective action facilities,
EPA selected and used (1) rates  at which the exposed populations were expected to be in contact
with each contaminated medium, (2) human uptake factors (e.g., soil-to-skin absorption rate),
and (3) food transfer factors that would determine the amount of a chemical that is transferred
from an environmental media to a vehicle of exposure (e.g., from surface water to fish). These
rates and factors were used either in risk equations to calculate intake at media exposure points
or in the MMSOILS model to predict concentrations in the foodchain pathway.

       Exposure Factors

       The Agency identified exposure rates (i.e., data describing the extent, frequency, and
duration of exposure) for the ground water, surface water, air, soil, and foodchain pathways.
These  exposure  rates and other exposure assumptions such as  the body weight  and uncovered
skin area of the exposed individual, are referred to in this document as exposure factors.  Values
that EPA used for the standard exposure factors  (such as exposure frequency, average body
weight, and exposure duration), and their sources are listed by pathway in Exhibit E-3.  Values
and sources for  less  standard factors (e.g., skin surface area available for exposure and soil-to-
skin adherence factor) are stated in the notes to Exhibit E-3. For the central tendency risk
analysis, exposure factor values used were either  the average or the 50th percentile, depending
on availability.  Exposure factor values used in computing risks to highly-exposed subpopulations
were the 90th or 95th percentile  estimates.
                               ••• Draft - March 23,1993 •••

-------
                                        IIT E-3
EXPOSURE PARAMETER VALUES USED F\     • RRECTIVE ACTION RIA RISK ASSESSMENT  (cont.)
Exposure Pathwny1
Ground Water
Ingest ion of Contaminated
Drinking Water
Inhalation of Volatile
Contaminants from Indoor Air
Dermal Absorption of
Contaminants from Water while
Showering11
Air
Inhalation of Contaminants
from Air
Surface Water1
Dermal Absorption of
Contaminated Surface Water
while Swimming'1
Ingestion of Contaminated
Surface Water while Swimming
Soil '
mgcsiinn of Contaminated Soil
(Adult)
Exposure Rnte

1 4 liters/day
(avg.r
0.63 mVhour
(15 m'/day)
(avg)"<


0.83 mVhour
(20 m'/day)
(avg)»


0.05
liters/hour"

100 trig/day"
Exposure Time2

-
~
O.I 16 hours/day
(7 minutes)
(50%)"

24 hours/day

2.6 hours/day
(hours/event)
(avg.)"
2.6 hours/day
(hours/event)
(avg )"


Frequency or
Exposure

350 days^cjir*1
350 days/year11
350 daysfycarb

350 days/year8

26 days/year
(avg.)"'6
26 days/year
(avg.)"

350 daysftcar11
Exposure
Duration:
Carcinogen

9 years
(5Qlk%)a
9 years
(50th%)'
9 years
(50lh%)"

9 years
(50lh%)'

9 years
(50lh%)a
9 years
(50th%)*

9 years
(50ih%)"
Exposure
Duration:
Non-
carcinogen

9 years'
9 years"
9 years"

9 years*

9 years'
9 years*

9 years"
Body
Weight

70 kg"
(avg.)
70kg"
(avg.)
70kg"
(avg.)

70kg*
(avg.)

70kg*
(avg-)
70kg*
(avg.)

70kg8
(avg )
Averaging
Time:
Cnrcinogen

15,560 days
(70 years)
25,560 days
25,560 days

25,560 days

15,560 days
15,560 days

15,560 days
Averaging
Time:
Non-
cnrcinogen

3,285
days3
(9 years)
3,285 days
3,285 days

3,285 days

3,285 days
3,285 days

3,285 days
                      Draft - March 22, 1993

-------
                                   EXHIBIT E-3
EXPOSURE PARAMETER VALUES USED FOR CORRECTIVE ACTION RIA RISK ASSESSMENT (cont.)
Exposure Pathway1
Ingestion of Contaminated
Soil (Child)
Ingestion of Contaminated
Soil (Pica Child)
Dermal Absorption from Soil
(Adult)7
Dermal Absorption from Soil
(Child)7
Foodchain
Ingestion of Contaminated
Homegrown Root Vegetables
Ingestion of Contaminated
Homegrown Leaf Vegetables
Ingestion of Contaminated
Homegrown Beef
Ingestion of Contaminated Dairy
Products
Ingestion of Contaminated
Recreationally Caught Fish
Subsistence Fanner
Ingestion of Contaminated
Homegrown Root Vegetables:
Subsistence Farmer
Ingeslion of Contaminated Home-
grown Leaf Vegetables:
Subsistence Farmer
Exposure Rate Exposure Time2
200 mg/day
(avg.)*'
800 mg/day
(high end)c
-
-

46 g/day
(avg.f8
65 g/day
(avg.rR
44 g/day
(avg.)0-9
160 g/day
(avg.r10
7.6 g/day
(50th%fe-"

74 g/day0-8
103 g/day
(avg.)c'R
Frequency of
Exposure
350 daysfycarb
365 daystyearb
350 daysyyearb
350 daystyearb

350 days/year11
350 days/year1"
350 days/year11
350 daystyearb
350 days/year11

365 days/year"
365 days/year"
Exposure
Duration:
Carcinogen
5 years0
5 years0
9 years
(50th%)a
5 years0

9 years
(50th%)a
9 years
(50th%)'
9 years
(50th%)a
9 years
(50th%)a
9 years
(50th%)a

40 years'
40 years'
Exposure
Duration:
Non-
carcinogen
5 years0
5 years0
9 years'
5 years0

9 years"
9 years8
9 years"
9 years"
9 years"

40 years'
40 years'
Body
Weight
16 kg0
(50th%)
16 kg0
(50th%)
70kg"
(avg.)
16kgc
(50th%)

70kg"
(avg.)
70kg"
(avg.)
70kg"
(avg.)
70kg*
(avg.)
70 kg"
(avg.)

70kg'
(avg.)
70kg"
(avg.)
Averaging
Time:
Carcinogen
25,560 days
25,560 days
25,560 days
25,560 days

25,560 days
25,560 days
25,560 days
25,560 days
25,560 days

25,560 days
25,560 days
Averaging
Time:
Non-
carcinogen
1,825 days
(5 years)
1,825 days
3,285 days
1,825 days

3,285 days
3,285 days
3,285 days
3,285 days
3,285 days

14,600
days
14,600
days
                      Draft - March 22,

-------
                        EXPOSURE PARAMETER VALUES USED Fv
                                                              IIT E-3
                                                             *RRECTIVE ACTION RIA RISK ASSESSMENT  (cant.)
Exposure Pathway1
Ingcstion or Coniaminaied
Homegrown Beef-
Subsistence Farmer
Ingestion of Contaminated Dairy
Products:
Subsistence Farmer
Subsistence Fisherman
Ingestion of Contaminated
Rccrcalionally Caught Fish:
Subsistence Fisherman
Exposure Rate Exposure Time1
75 g/daycQ
300 g/day
(95th%)c'in

99 g/day
(95l\\%)"-"
Frequency at
Exposure
365 days/year*
365 daysfycar'

365 daysfycar"
Exposure
Duration:
Carcinogen
40 years'
40 years'

30 years
(90th%)a
Exposure
Duration:
Non-
carcinogen
40 years'
40 years'

30 years'
Body
Weight
70 kg'
(avg.)
70kg'
(avg.)

70kg'
(avg.)
Averaging
Time:
Carcinogen
25,560 days
25,560 days

15,560 days
Averaging
Time:
Non-
carcinogen
14,600
days
14,600
days

10,950
days
SOURCES:
(Note (hat the sources listed below may in turn refer to secondary documents as the original source Tor some of the parameter values.)

(a)      USEPA 1989. Risk Assessment Guidance for Supcrfund Volume I:  Human Health Evaluation Manual (Part A). Office of Emergency and Remedial
        Response.  EPA/540/1-89/002.

(h)      USEPA 1991. Human Health Evaluation Manual. Supplemental Guidance:  Standard Default Exposure Factors. Office of Emergency and Remedial
        Response.  OSWER Directive: 9285.6-03.

(c)      USEPA 1989. Exposure Factors Handbook.  Office of Health and Environmental Assessment  EPA/600/8-89/043.

(d)      USEPA 1991. Interim Guidance for Dermal Exposure Assessment.  Office of Research and Development. EPA/600/8-91/Oil A.
(g)
USEPA 1992. Dermal Exposure Assessment.  Principles and Applications  Office of IIcalih and Environmental Assessment. EPA-600/8-9I/OI in
                                                  Draft - March 22, 1993

-------
                                                                   EXHIBIT E-3
                        EXPOSURE PARAMETER VALUES USED FOR CORRECTIVE ACTION  RIA RISK ASSESSMENT (conL)


NOTES:

1. Intake via each exposure pathway is calculated using the general equation described in Section E.2.4.

2. The general equation is modified for the inhalation and dermal exposure pathways to include the "Exposure Time." This parameter (in hours/day) is included as
   a multiplicand in the numerator of the equation.

3. The averaging time to calculate noncancer hazard is equal to the exposure duration expressed in days.

4. The general equation is modified to calculate intake via dermal absorption in the water pathway:

   Intake    =  CW x CF x SA x PC x ET x EF x ED
                    BWxAT

   where factors unique to this pathway include:

      CF    =  Conversion factor (1 liter/lOOOcm1)
      SA    =  Skin surface area available for contact (cm'/day)
                (19,400 cm1 for adults (central tendency, source (g)))
      PC    =  Chemical-specific dermal permeability constant
      ET    =  Exposure time (hours/day)

5. Where applicable, exposure  due  to use of surface water as municipal water supply is evaluated using the same exposure pathways, routes, and parameter values
   as for ground water.

6. The exposure frequency value for dermal absorption and ingestion of contaminated surface water while swimming was provided by EPA's Office of Research
   and  Development.  This assumes that an individual swims 2 times a week during the summer (3 months) only.

7. The general equation is modified to calculate intake via dermal absorption in the soil pathway:

   Intake    =  CS x CF x SA x AF x ABS x EF x  ED
                     BWxAT

   where factors unique to this pathway include:

      CF    =  Conversion factor (10* mg/kg)
      SA    =  Skin surface area available for contact (cmVday)
                (5.000 cm2 for adults  (central tendency, source (g)), and 2,500 cm2 for children (default value, source (d)))
      AF    =  Soil-to-skin adherence factor (mg/cm2)
                (0.2 mg/cm2 (central tendency, source (g)))
      ABS   =  Absorption factor (chemical specific constant)


                                                   Draft - March 24,

-------
                                                                    E         E-3
                        EXPOSURE PARAMETER VALUES USED FOR v    .fiCTIVE ACTION RIA RISK ASSESSMENT (cont.)


8.    The general equation is modified for the foodchain pathways to include "fraction from contaminated source."  This parameter is included as a multiplicand in
      the numerator of the equation, and modifies the exposure rate. Thus, the exposure rate accounts for the fraction of total vegetables consumed that comes
      from the contaminated source.  The proportion of contaminated vegetables is assumed to be 25 percent for the general population (average, source (c)) and
      40 percent for subsistence farmers (reasonable worst case, source (c)). For example, the 65 g/day exposure rate for contaminated leafy vegetables represents
      25 percent of the total daily consumption of leafy vegetables of 260 g.

9.    Exposure rate accounts for the fraction of total beef consumed that comes from the contaminated source.  The proportion of contaminated beef is assumed
      to be 44 percent for the general population (average, source (c)) and 75 percent for subsistence farmers (reasonable worst case, source (c)).

10.   Exposure rate accounts for the fraction of total dairy products consumed that comes from the contaminated source. The proportion of contaminated dairy
      products is assumed  to be 40 percent for the general population (average, source (c)) and 75 percent for subsistence farmers (reasonable worst case, source
      (c)).

11.   Exposure rate accounts for the fraction of total fish consumed that comes from the contaminated source.  The proportion of contaminated fish is assumed to
      be 20 percent for the general population (average, source (c)) and  75 percent for subsistence fishermen (reasonable worst case, source (c)).
                                                    Draft - March 24.1993

-------
                                           20

       Dermal Uptake Factors

       Availability of a chemical to cause harm to humans after exposure or contact is
dependent upon uptake into the body. To further quantify human exposure through dermal
contact, the Agency calculated the portion of each chemical in the environment that would be
absorbed through a person's skin. To do this EPA considered two factors: the skin permeability
coefficient (SPC) for uptake from contact with water and the dermal absorption factor (DAF)
for uptake from soil.  Recent EPA guidance4 was used in developing chemical-specific values for
these factors for use in this RIA.

       EPA guidance documents provide specific SPCs for several inorganics and many organics.
In cases where an inorganic chemical is not specifically  listed, EPA used a default value of 0.001
cm/hr. For non-specified organics, the Agency calculated SPCs using the following equation

       Log SPC      =     - 2.72 + 0.71 log K^ - 0.0061 x (molecular weight)
where K^ is a unitless partition coefficient that represents the ratio of the concentration of the
penetrant in octanol to the concentration of the penetrant in water under equilibrium conditions.

       A DAF is a fraction expressing how much of a chemical will be absorbed when applied to
the skin in a soil matrix. The main factors controlling the DAF are (1) the chemical's physical
properties; (2) soil loading or the soil adherence factor; (3) concentration of the chemical in the
soil; (4) exposure duration; and (5) organic content of soil.

       There are only three chemicals for which EPA has made recommendations for the range
of DAFs to be used in risk assessments. The upper end of the  ranges for each of these three
chemicals have been used for analysis in this RIA:

                    Chemical                                              DAF

                    2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)              0.03
                    3,3',4,4'-Tetrachlorobiphenyl  (TCB)                      0.06
                    Cadmium                                              0.01

       Most chemicals are not specifically identified in the EPA guidance document, although
there are two chemicals for which experimentation supports a very wide  range of DAF values.
DAFs range up to 1.0 for benzo[a]pyrene and up to approximately 0.3 for DDT, among the
chemicals evaluated in the EPA guidance. In both cases upper end values were utilized in this
RIA. In most cases even ranges are not available and DAF values are inferred from data
collected for analogue chemicals. EPA used TCDD data  for all chlorinated dioxins and furans,
TCB data for all  PCBs, and Benzo[a]pyrene data for all PAHs.  EPA used the McKone model
   4 U. S. EPA (1992) Dermal Exposure Assessment: Principles and Applications. Exposure
Assessment Group, Office of Health and Environmental Assessment, Washington, D.C.
EPA/600/8-91/01 IB.

                             •** Draft - March 23,1993  •••

-------
                                            21

for all organics which do not come under the prior categories.5 The model uses the K^, and the
dimensionless form of the Henry's Law Constant (K«) to determine which DAF is most
appropriate for each chemical category.  Using this model for chemicals with K,, between 0.01
and 0.1, and K^, greater than 10 the DAF was estimated at 0.2.  For those organics not covered
by the model (i.e., those organics with K,, between 0.001 and 0.01, or with both K,, less than 0.001
and KO., greater than 10'), and for inorganic chemicals, EPA used a DAF default value of 1.0.

       Foodchain Transfer Factors

       Transfer factors describe, in mathematical terms, the availability of chemicals for uptake
by way of the foodchain.  For this RIA, transfer factors collectively refer to bioaccumulation
factors, partition coefficients, and food transfer factors. For environment-to-aquatic biota
transfers, bioaccumulation factors are  used to estimate the rate at which contaminants will collect
in the lipids of animals (in this RIA, finfish were used to represent all aquatic fauna). Chemical
transfers from soil-to-piant can be estimated  using separate partition coefficients for leaf
vegetables and root vegetables.  Further up the foodchain, food transfer factors are used to
estimate the rates at which chemicals are passed on through consumption.  For this RIA, this is
limited to the assessment of chemical  transfers to beef and milk.

       Transfer rates vary by chemical and according to placement along the foodchain.
Detailed below are the methods and sources that the Agency used to determine chemical-specific
transfer factors.

       For inorganic chemicals in fin fish, the bioaccumulation factors used in this analysis were
those listed  in the Multi-media Environmental Pollutant Assessment System (MEPAS) data base.
(MEPAS in turn references original literature for these values, e.g., Napier, 1980, Strenge, 1986,
or Spehar, 1980.)  EPA calculated bioaccumulation factors for organic chemicals in fin fish based
on Veith's (1980) formula presented in MEPAS:

                           =     10 <076 »*•"• *»>

                           where K^ is the octanol-water partition coefficient
       Chemicals taken up by crops accumulate at different rates within the anatomy of the
plant; therefore, the Agency calculated separately partition coefficients for vegetative (vine and
leaf) and root parts of the plant. For organic chemicals, partition coefficients specific to
vegetative plants were calculated as the arithmetic average of the values generated by the Topp
(1988), and Travis and Arms (1988) equations:
       Topp (1988):               PC,   =     (8.77 x 10s) x MW <•

                                         where MW  =  molecular weight
   5 Note that the guidance document states that EPA is in the process of evaluating the
reasonableness of the generalizations made in the Mckone model.

                              •** Draft ~ March 23,1993 •**

-------
                                            22

       Travis & Arms (1988):       PC,    =     10 o-5"**78 »'••*'•'>

The final partition coefficient (PQ,) was calculated as (PC, + PCJ/2.  EPA derived partition
coefficients for organics specific to root parts using Briggs's (1982) equation, modified to be
applicable to concentrations of the chemical in soil rather than  in soil solution:


       PC,           =     [0.03 x (K-..V "1 + 0.82
       where K^     =     organic carbon partition coefficient
       and   f^.      =     fraction organic carbon in soil

The Agency employed the constituent-specific partition coefficients for uptake of inorganic
chemicals compiled by Baes (1984), for both vegetative and root parts of plants.

       The Agency calculated food transfer factors between feed and beef using the Travis and
Arms (1988) equation for organic chemicals, and directly used the Baes (1984) values for most
inorganics. The Travis and Arms equation calculates transfer factors to beef for organics based
on the chemical-specific octonol-water partition coefficient K^ such that:

       TRF^,       =     10 <•"*•••«•">

Mercury is the only inorganic chemical for which the Baes value., was not incorporated; its food
transfer factor was based on Ng et. al., (1982).

       Similarly, feed-to-milk transfer factors for inorganic chemicals were taken from Baes
(1984), while values for organics were calculated according to the Travis and Arms (1988)
equation:
       TRFmak       =      JQC-ilO + logKo.)

Note that the transfer factors discussed above were inputs required for MMSOILS, and were not
used external to the model.

       E.2.4 Calculating Intakes at Exposure Points

       Once the exposure points were selected and exposure assumptions defined, the Agency
proceeded to calculate chemical-specific human intakes at the exposure points. An intake, often
referred to as a chronic daily intake (CDI), is the measure of exposure expressed as the mass of
a chemical in contact with the exchange boundary per unit body weight per unit time (mg of
chemical/kg  of body weight-day).  The generic equation for calculating chemical intakes is as
follows:

       Intake        =      C x CR x EFD x OF
       (mg/kg-day)                BW x AT

                              ••• Draft - March 23, 1993 •••

-------
                                           23


Where:
              C      =     chemical concentration contacted over the exposure period (e.g.,
                           mg/liter water)

              CR    =     contact rate; the amount of contaminated media contacted
                           per unit time or event (e.g., liters/day)

              EFD   =     exposure frequency and exposure duration; describe how long and
                           how often exposure occurs (calculated as EF x ED, days/year and
                           years)

              OF    =     other pathway-specific exposure factors; adjust the intake as neces-
                           sary, e.g., skin surface area available for dermal contact via soil or
                           surface water pathways; fraction ingested from contaminated
                           source for foodchain pathway

              BW    =     body weight; the body weight of the individual exposed over the
                           exposure duration  (kg)

              AT    =     averaging time; period  over which the exposure is averaged (days)

Essentially, the intake equation transforms the concentration of the chemical in an exposure
medium to human daily intake.  Note that the intake was calculated from an average  concentra-
tion over the duration specific to that intake.  For example, the  exposure duration used to
evaluate risk for this RIA is equal to 9 years, which is the national  median time of residence  at
one particular  location based on Census Bureau  data. An average  concentration applicable to all
9 years was used in calculating the intake for this exposure duration.

       For this analysis, EPA evaluated exposures from 1992 through 2120. The Agency used
MMSOILS to predict concentrations in the exposure media and pathways throughout this time
period.  Concentrations of chemicals are predicted as steady-state in the soil and foodchain
pathways; therefore, the exposure concentration  is constant over the entire exposure duration.
Concentrations in ground water, air, and surface waters, however, were time-varying.  For these
three pathways, the Agency used a methodology  to calculate 9-year moving averages from the
concentration profile from 1992 through 2120. An example of this  is shown in  Exhibit E-4. Such
a 9-year average concentration was then used to  calculate  the intake for that particular year.

       E.2.5 Assessing Lead Exposure

       To assess lead exposure, EPA used a model known as the Uptake/Biokinetic (UBK)
model developed by the Office of Air Quality  Planning and Standards.  The LEAD (version 5)
program, a PC software application of the UBK  model, was used to estimate blood lead levels in
                              ••» Draft « March 23,1993 •••

-------
                          EXHIBIT E-4
CALCULATION OF 9-YEAR AVERAGE CONCENTRATION
        O  Actual concentration for any given year
           (e.g., •for 2018)

        A  Corresponding 9-year average concentration for
           that year (e.g., A for 2018)

       11=1  9-year period over which concentration for that
           year la averaged (e.g., •• •• •• tor 2018)
                I CD CD CD CD CD CD CD CD^C3-O*CD CD CD CD CD CD i

          | CD CD CD CD CD CD CD O-CDCD CD CD CD CD CD CD CD
	1	1	1	h-
 2008   2009  2010   2011
2012  2013

   Time
                                         2014   2015  2016   2017   2018
                                                                       F2D02S-2

-------
                                            25

the target population (included in the model as children of ages 0-6) based on the concentration
of lead in each exposure media at the facilities. This modeling approach provides estimates of
blood levels of lead based on facility-specific lead concentrations via a variety of pathway
categories (e.g., air, diet, water, indoor dust, soil, paint, and mother's blood lead concentration).

        With this information, the model can be used to estimate blood lead levels based upon
total lead uptake via the various pathways. For this analysis, EPA first used the default  values
built in to the program to estimate a baseline concentration of lead  in the blood. Then the
Agency added facility- and media-specific  lead concentrations to each default, and ran the model
again. The Agency kept the default values in as baselines, because background lead concentra-
tions (due to both natural and anthropogenic sources) were not available for the facilities. If the
Agency had used only the lead concentrations directly attributable to the facility, the estimated
blood concentrations would likely have been understated .

        The Agency estimated concentrations of lead in air, groundwater, soil, homegrown
vegetables, home-raised beef, and locally caught fish using MMSOILS, based on central  tendency
assumptions of waste release to and fate and transport in environmental media.  Values
calculated by MMSOILS were added to the UBK model defaults for the air, soil, and  drinking
iwater exposure pathways.  At facilities where ground water was not  used as a drinking water
source, only the program default was used. At most facilities, the MMSOILS-predicted  values
for air and drinking water  concentrations  of lead were too low to raise the program defaults,  i.e.,
no additional risk would result  due to facility-related lead exposure  via these pathways.  For the
diet section, the Agency directly input the concentrations of lead predicted by MMSOILS for
specific food types, i.e., homegrown vegetables, home-raised beef,'and locally-caught fish. In the
program, the default concentration and percent diet for these individual food types  are both
zero, and, therefore, the Agency did not need to add to any default  lead intake via these
pathways.

        In addition to the above lead concentrations input into the program, the model accepts
user input of variables pertaining to facility-specific exposure to lead through paint and  through
placenta! transfer to the infant (maternal  route).  The Agency did not change the default values
for these parameters. The default uptake from paint was 0.0 ug/day, and the maternal blood
concentration,  which determines the mother's contribution to the child's lead levels at birth, was
7.50 ug/dL.

EJ  Risk Characterization

        The  human health benefits analysis for this RIA includes a quantitative  characterization
of carcinogenic and non-carcinogenic human health risks for  closest exposed individuals and for
populations living near corrective action facilities.  Individual and population risks are estimated
 for four media pathways, the foodchain pathway, and three highly-exposed subpopulations.  For
 each pathway,  the Agency aggregated risks resulting from each chemical to which the individual
;was exposed and aggregated risks  from each exposure route within  the exposure pathway.
 However, the Agency did not aggregate risks across exposure pathways.
                               *•* Draft - March 23,1993 ••*

-------
                                             26

        EPA based its individual and population risk characterizations on exposures modeled at
 the facility level.  In particular, the Agency estimated baseline risks for facilities among a
 stratified random sample of all RCRA facilities in the U.S. The Agency then extrapolated risk
 characterizations  for sample facilities to determine the aggregate national human health risks in
 the absence of corrective action. This process was then repeated for the corrective action
 scenario to simulate the effect of the rule on human health risks.

        EJ.l  Characterizing Off-site Individual Risk

        For each sample facility, the Agency estimated cancer and noncancer health effect6 for
 an individual  located at the closest off-site point where actual exposure is likely given current
 land uses. This "closest individual" may be located at different  points for each exposure
 pathway.7  The Agency used different methods to calculate cancer risks and noncancer health
 effects to individuals because carcinogens and non-carcinogens  have unique relationships  between
 contaminant exposure level and health effect. Many chemicals  have carcinogenic and non-
 carcinogenic effects. The Agency included both effects of such  chemicals in the risk characteriza-
 tion (i.e., the  contaminants' contribution to each type of health  effect is considered).  Also, for
 exposures via ground water,  surface water, and air, EPA calculated  the individual risk based on a
 time-varying concentration profile.  Exposure concentrations at a given exposure point in  any of
 these media vary from  1992  through  2120 (the modeling period), leading to time-varying risks.
 Therefore, EPA calculated both 130-year average and  9-year peak individual cancer risks  and
 noncancer effects in the ground-water, air, and surface water pathways.8 An average risk  is the
 130-year arithmetic  average of all the 9-year risks over the modeling period. Peak risk is the
 highest 9-year risk.  Methods used to characterize the cancer risks and noncancer effects are
 detailed below.

        Individual Cancer Risk

        For this RIA, the Agency estimated "excess cancer risk," the increase (i.e., above
 background) in the probability  of developing cancer over an individual's lifetime in response to
 contaminant exposures. To estimate excess cancer risks, the Agency multiplied the daily intake
 (i.e., exposure level) of each  carcinogen by the sK'-c factor. Daily intakes differ for each
 exposure pathway depending on the exposure routes included and the exposure assumptions.
    6 For this RIA, the Agency refers to noncancer human health effects as "noncancer effects"
for convenience.  In past RIAs and other documents, the Agency has often referred to these
effects as " noncancer hazards."

    7 See section E.2.1 for descriptions of these points.

    8 Risks from all other pathways are not subdivided into average and peak risks because
MMSOILS models steady state (i.e., constant) contaminant concentrations in these pathways
over the modeling period.

                               *•• Draft « March 23,1993 •••

-------
                                            27

       Individual risk for each exposure pathway is the sum of cancer risks calculated for each
carcinogen to which the "closest individual" is exposed, i.e., risk aggregated across chemicals and
across exposure routes within a pathway.

       Individual Noncancer Health Effects

       The Agency evaluated noncancer effects by determining whether contaminant exposures
exceed the RfD. Ratios of contaminant exposure level to the RfD for that contaminant, called
"hazard quotients," greater than one indicate increased likelihood of a non-carcinogenic health
effect. The Agency calculated total individual noncancer effects by adding the chemical-specific
hazard quotients. The sum of all hazard quotients is called the "hazard index," or HI, and is
aggregated across chemicals  and across all exposure routes within pathways.

       High-End and Central Tendency Risks

       For this RIA, the Agency characterized individual cancer risks and noncancer effects for
two contaminant release scenarios. These scenarios, the "high-end" and "central tendency"
scenarios, differ in assumptions regarding the mass of contaminants in SWMUs, rate of contami-
nant releases from SWMUs, and their fate and transport in the environment.  For example, the
"high-end" scenario assumes  a faster rate of ground water transport than the "central tendency"
scenario. As a result of these different assumptions, the two scenarios present different  exposure
concentrations reaching the exposed individual through each pathway. All other features of the
risk characterization including exposure routes and assumptions, individual exposure point loca-
tions, and risk calculations, are identical under the high-end and central tendency scenarios.9

       E.3.2  Characterizing Future On-site  Individual Risk

       Using central tendency exposure assumptions, EPA also calculated potential risk under
alternative  assumptions about future use of on-site land at the facilities. EPA defined this risk as
the risk to a hypothetical individual who is assumed to reside or grow food on-site at some point
in the future.  The Agency assessed on-site future risk to individuals due to exposures via the
ground-water, air, and soil, and foodchain pathways, and to pica children and subsistence
farmers.10

       E.33  Characterizing Population'Risk

       The population risk characterization estimates the number of people living near correc-
tive action facilities who will have a cancer or noncancer effect due to exposure.  The Agency
expressed population cancer risks as the number  of excess cancer cases expected in the exposed
population over the 129-year modeling period.  Similarly, the Agency expressed population
   9 See also Exhibit 7-4 and Appendix G.

   10 See also Section 7.23 in Chapter 7.

                              *•* Draft ~ March 23,1993 *•*

-------
                                             28

 noncancer effects as the number of people exposed to contamination exceeding noncancer effects
 thresholds (i.e., hazard index or HI greater than 1) during the 129-year modeling period.

       The Agency estimated total population risks attributable to three exposure pathways:
 ground water, air, and surface water. To do so, the Agency used the same exposure routes and
 exposure assumptions that it used to characterize central tendency individual risks from these
 pathways. Because the population risk characterization counts numbers of people affected over a
 129-year modeling  period, the Agency estimated risks for populations currently living near
 corrective action facilities and also for future populations.  Section E.2.2 describes the approach
 for forecasting future exposed populations.

       Risk Cohorts

       The Agency calculated population risks based on 9 years of exposure using a cohort-
 averaging approach.  Population risk in any given year is related not to the total exposed
 population in that year but to that portion of the total population that has received the full nine-
 year exposure.  This portion of the population is defined as the "risk-cohort" for that year.  The
 size of a risk-cohort for a given  year, which  remains constant  through time, is equal to the size of
 the cohort which entered (or was born into) the area nine years ago and is now leaving.  See
 Exhibit E-5 for a graphic representation  of a risk-cohort "filling up" with exposure years as it
 moves through time, until it has completed nine years of exposure and is ready to have its risk
 calculated.

       For the population risk characterization, the Agency assumed that exposures begin in
 1992 and continue only to the end of the modeling period (i.e., until 2120).  The Agency did not
 include risks due to exposures occurring prior to 1992 ("sunk  risks") in the risk characterization.

       Population  Cancer Risk

       The Agency measured population cancer risk as the number of people in exposed
 populations that develop cancer in response to exposure to facility contamination. Total
 population cancer risk for a facility over the 129 years is equal to  the sum of the cohort popula-
 tion risk from 1992 through 2120. EPA estimated cohort population risk or "excess cancer
 incidence" in a risk cohort for any given year by multiplying the individual cancer risk (i.e.,
 average lifetime excess cancer risk) for that year by the cohort size. The result of this multipli-
cation is the number of cancer cases expected in the risk cohort.

       Population Noncancer Effects

       The Agency measured population noncancer effects as the number of people living  near
facilities  with an individual HI above 1.  To estimate population noncancer effects, the Agency
calculated the individual HI and assigned that individual HI to all individuals in the risk cohort.
Thus, if the individual HI exceeded 1, indicating that the individual's exposure exceeded the
threshold, EPA counted the entire risk cohort as having a HI greater than 1. Again, the total
population for a facility having an exceedence of the HI over  the 129 years is equal to the sum of
the risk cohorts with HI greater than 1 from 1992 through 2120.

                               ••* Draft ~ March 23,1993 •••

-------
                                    EXHIBIT E-5
   USE OF RISK COHORTS TO CALCULATE POPULATION RISK
                                                          _Total population in any given
                                                            year consists ol 9 cohorts
                         Cohort moves Into
                        area at the beginning
                             of 2010
   2019
                                                          /   /   /   /.    s
                                                         	/	x	•;••'	-X	/
                                                            cohort starts In 2010, and at
                                                         the end of 2018, receives the 9-year
                                                         average concentration (AC Mit)
                                                         I.e., this la the "risk cohort" for 2018.

                                                         Cohort leaves area after
                                                         9 years of residence
 or
A cohort that haa received the full nine years
•xpoaure (level of a
of exposure years)
    W   exposure (level of shading Indicates the number
                           f
Population growth la reflected In larger alze of new cohorts over time
                                                                                       tMO»-1

-------
                                             30

       EJ.4 Extrapolating to the National Level

       The Agency characterized aggregate national human health risk by estimated risks at a
sample of real facilities and extrapolating to the national level.  Specifically, the Agency
characterized individual and population risks at the RIA sample facilities that trigger corrective
action. These facilities belong to a stratified random sample of RCRA facilities.  Each facility in
the sample has a "facility weighting factor" that reflects the relative frequency with which that
type of facility occurs in the U.S. For example, a facility with a weighting factor of 3.3 represents
3.3 similar facilities that exist in the U.S, and the sum of all facility weighting factors for the
facilities  equals the number of facilities in the U.S. that would trigger corrective action.  EPA
used facility weighting factors to extrapolate individual and population risks from  the facility level
to the national level." The Agency used slightly  different approaches to extrapolate population
risks and to extrapolate individual risks; these approaches are described below.

       Extrapolating Population Risk

       To extrapolate population risks, the Agency multiplied the estimated the number of
people living near each facility that bear cancer or noncancer effects by the facility weighting
factors.   This extrapolation of facility-level population risks results in the total number of people
affected  at all facilities in the U.S. over the modeling period.  Thus, EPA added the extrapolated
facility-specific risks for all sample  facilities to estimate the number of people in the U.S.
expected to have cancer or non-carcinogenic effects (i.e., have hazard indices above 1) at  all
RCRA facilities that trigger corrective action.

       Extrapolating Individual Risk

       Because individual and population risks are measured differently, the Agency used a
different approach to extrapolate individual risks to the national level than it used to extrapolate
population risks.

       EPA expressed national-level  individual risks by determining how many facilities in the
U.S. would cause individuals to incur the various levels of cancer risks and noncancer effects.
The Agency calculated the distributions of facilities  at various risk levels by adding the facility
weighting factors of facilities with similar risk levels or hazard indices.  For example, a facility
that has  an individual  cancer risk of 2.3 x 10'5 and a facility weighting factor of 3.3 is counted as
3.3 facilities in the U.S. that have (or cause) individual cancer risks at the 10's level. This
approach produced a distribution that shows how many facilities at the national-level would have
each cancer risk or noncancer effect level.

       E3J5 Characterizing Effects from Lead Exposure

        EPA characterized  effects from lead exposure separately from cancer risks and
noncancer effects due to other chemicals.  The Agency used the statistical function within the
    11 See Chapter 3 for a discussion of the weighting factors.

                               ••• Draft - March 23,1993 •••

-------
                                          31

UBK LEAD program to create graphs of the "Distribution Probability Percent Around a Mean
Value" based on the calculated blood concentrations for the target population at each facility.
These graphs show the mean blood lead concentrations, and also report the percent of the target
population estimated to have blood lead levels above the 10 ug/dL cutoff point.  The calculated
geometric mean blood lead concentration when the input variables are left as program defaults is
3.23 ug/dL (based on a normal distribution).

       EPA determined incremental effects due to contamination from the facility by comparing
the predicted geometric mean blood concentration of lead with the threshold blood level of 10
ug/dL.
                              *•* Draft - March 23, 1993 •••

-------
                                      APPENDIX F

           ECOLOGICAL BENEFITS:  METHODOLOGIES AND CASE STUDIES

       This Appendix provides additional information on both the methods and results of the
ecological analyses described in Chapter 8.  Specifically, the methodology for identifying lower
and higher-threat facilities based on the proximity analysis (section F.I), the method for deriving
screening ecological benchmark levels for surface waters and soils (section F.2), and the
methodology  for estimating the length of a river or stream contaminated at levels above
ecological benchmarks (section F3) are described.  More detailed data for the results of each
analyses  are provided in sections F.4 through F.8.

F.I    Methodology for Proximity Analysis

       The proximity analysis described in sections 8.1.1 and 8.2.1 addressed the potential of
releases of hazardous substances from RCRA-regulated facilities to reach valuable or vulnerable
ecological resources.  The first step in dividing facilities into  potentially higher- and lower-threat
groups on the basis of proximity  to RCRA-regulated facilities was to sort the facilities according
to total acreage of habitat types considered more valuable and more vulnerable to releases of
hazardous substances.  Two of the land-use categories (i.e., surface water, terrestrial
environments) were considered more valuable than the other three categories (i.e., agricultural,
residential, industrial/other) because the former areas generally are more likely to include
sensitive, diverse and productive ecosystems than the latter areas.  It is recognized that  specific
surface water bodies and/or terrestrial environments in the vicinity of the sample facilities may be
disturbed or  polluted significantly; moreover,  this disturbance or pollution may not be related to
the facility. For this analysis, however, surface water and terrestrial environments were generally
assumed to be more valuable than the other habitat types. Of these two categories, surface
waters were considered more vulnerable to contamination because substances released  to surface
water can quickly spread to contaminate large areas and expose all members of an aquatic
community.  Releases to land tend to contaminate confined  areas and expose only a portion of
the community.

        Facilities were sorted initially by surface water acreage to reflect an emphasis on areal
extent of vulnerable habitat types in evaluating relative threats. The sorted facility data indicated
that a group of about 30 percent of the sample facilities had relatively small areas of surface
water nearby (i.e., less than 100 acres) compared with the acreage of nearby surface waters for
remaining facilities  (i.e., greater than 100 acres). Those facilities with smaller areas of surface
water nearby were categorized as lower-threat and those with larger areas of nearby surface
water as higher-threat.  Other characteristics  of  the environmental setting  (e.g., sensitive
environments and the proportion of the surrounding area  that consisted of relatively undisturbed
terrestrial ecosystems) also were considered,  and a few facilities recategorized as described in
 section 8.2.1.

 F.2    Methodology for Deriving Screening Ecological Benchmark Levels

        For the concentration-based screening analysis, benchmark levels were derived  for the
 protection of aquatic life as described  in section F.2.1.  For the case study analyses, three
 additional types of ecological benchmarks were  estimated:  ambient levels for sediments  (benthic

                             *  * * DRAFT - March 24,1993 • * •

-------
                                             F-2
 invertebrates; section F.2.2), drinking water or dietary intake levels (birds and mammals; section
 F.2.3), and ambient levels for soil (plants).

        F.2.1  Benchmarks for Aquatic Life

        The primary objectives of estimating benchmarks for the protection of aquatic life were:
 (1) to identify threshold concentrations above which adverse impacts to aquatic organisms are
 likely, and  (2) to provide a reasonably uniform basis for evaluating the relative ecological threats
 associated with each of the sample facilities. Criteria and extrapolation factors used to derive
 benchmark levels are discussed first, followed by data sources for toxicity and to estimate
 bioconcentration factors (BCFs).

        Criteria and Extrapolation Factors

        The starting point was EPA's chronic ambient water quality criteria (AWQC), which are
 intended to be protective of 95 percent of the species in an aquatic ecosystem.  When a chronic
 AWQC was not  available for a particular hazardous substance, an ecological benchmark level
 that was roughly equivalent to a chronic AWQC was derived by applying one or more
 extrapolation factors to toxicity values from the readily available literature. To summarize, the
 following criteria and toxicity values (in descending order of preference) were  used to derive
 ecological benchmark levels:

   •    EPA chronic Ambient Water Quality Criteria (AWQC);

   •    Lowest observed effect levels (LOELs), no observed effect levels (NOELs), maximum
        acceptable toxicant concentrations (MATCs), or other chronic effect or no-effect values
        from AWQC documents or the literature;

   •    EPA acute AWQC; and

   •    LCjo values from AWQC documents or the literature.

For values other than chronic AWQC, one or more of the following extrapolation factors was
applied to derive an ecological benchmark value that is roughly equivalent to a chronic AWQC:

   •   A factor of 5 to account for variation in species sensitivity;

   •   A factor of 10  to extrapolate from an acute to chronic value; and

   •   A factor of 10  to account for high bioaccumulation potential.

       Exhibit F-l presents a summary of the criteria or toxicity values and extrapolation factors
used to derive screening ecological benchmark levels.
                            * * • DRAFT -- March 24, 1993 • • •

-------
                                           F-3

                                      EXHIBIT F-l
        CRITERIA, TOXICITY VALUES, AND EXTRAPOLATION FACTORS USED TO
                DERIVE SCREENING ECOLOGICAL BENCHMARK LEVELS
                    Type of Criterion or Toxicity Value
Extrapolation Factor
      EPA Chronic AWQC
      LOEL, NOEL, MATC, or other effect/no effect value from AWQC
             document or other literature
      EPA Acute AWQC
      LCjo from AWQC document or other literature
       - substance with log K^ less than 3.5 or BCF less than 300
       - substance with log K^ £ 3.5 or BCF 2 300
        1
        10*

        50^
       500*'
         - A factor of 5 for variation in species sensitivity
         - A factor of 10 to extrapolate from an acute to a chronic value
         - A factor of 5 for variation in species sensitivity and a factor of 10 to extrapolate from an
           acute to chronic value
         * A factor of 5 for variation in species sensitivity, a factor of 10 to extrapolate from an
           acute to a chronic value, and a factor of 10 to account for high bioaccumulation potential
       Data Sources

       The following sources (in descending order of preference) were used to obtain chronic or
acute aquatic toxicity values from the literature:

       •      A database of chronic freshwater toxicity values for approximately 190 air toxics
              compiled in June 1991 by ORD's Environmental Research Laboratory —
              Duluth.1  All toxicity values in this database were subject to  quality assurance
              (QA) reviews by ERL—Duluth staff.  Toxicity values were based on data in
              EPA's Aquatic Information Retrieval  (AQUIRE) database,  ERL—Duluth's
              quantitative structure-activity relationship (QSAR) algorithms, and professional
              judgment.
    1 ICF Incorporated (ICF). Focus Chemicals for the Clean Air Act Amendments Great
Waters Study. August 15, 1991 draft report prepared for the U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards, Research Triangle Park, NC, 1991.
                           * * * DRAFT -- March 24, 1993
                                                         * * •

-------
                                           F-4

       •     A database of acute toxicity values (freshwater, saltwater) for 330 chemicals
             compiled in April 1991 for the Superfund Hazard Ranking System.2  All toxicity
             values in this database were subject to QA reviews by EPA's Office of Emergency
             and Remedial Response. The primary data source for toxicity values was
             AQUIRE.

       •     The lowest available toxicity value (acute, freshwater for any aquatic organism)
             from compilation of freshwater toxicity values for 410 chemicals.3

       The following sources (in descending order of preference) were used to obtain log K^4
and BCF values from the available literature:

       •     A database of BCF values for approximately 190 air toxics compiled in June 1991
             by ORD's Environmental Research Laboratory — Duluth.5 All BCF values in
             this database were subject to quality assurance (QA) reviews by ERL—Duluth
             staff. BCF values were based on data in EPA's Aquatic Information Retrieval
             (AQUIRE) database, ERL—Duluth's QSAR algorithms, and professional
             judgment.

       •     A database of BCF values (freshwater, saltwater) for 330 chemicals compiled in
             April 1991 for the Superfund  Hazard Ranking System.6 All BCF values in this
             database were subject to QA reviews by EPA's Office of Emergency and
             Remedial Response.  The primary data source for BCF values was AQUIRE.
    2  U.S. Environmental Protection Agency (EPA).  Superfund Chemical Data Matrix.  Data
file compiled in April 1991 by the Office of Emergency and Remedial Response, Site Evaluation
Division, Site Assessment Branch, 1991.

    3  Mayer, F.L., Jr., and Ellersieck, M.R. Manual of Acute Toxicitv:  Interpretation and Data
Base for 410 Chemicals and 66 Species of Freshwater Animals. Washington, D.C.: U.S.
Department of Interior, Fish and Wildlife Service, 1986.

    4  K^ is defined as the octanol-water partitioning coefficient of a substance. Log K^ values
correlate with log BCF values over a certain range of K^ values.

    5  ICF Incorporated (ICF). Focus Chemicals for the Clean Air Act Amendments Great
Waters Study. August 15, 1991 draft report prepared for the U.S. Environmental  Protection
Agency, Office of Air Quality Planning and Standards, Research Triangle Park, NC, 1991.

    6  U.S. Environmental Protection Agency (EPA).  Superfund Chemical Data Matrix.  Data
file compiled in April 1991 by the Office of Emergency and Remedial Response, Site Evaluation
Division, Site Assessment Branch, 1991.
                            • • * DRAFT « March 24, 1993 •
                                                          * *

-------
                                           F-5

       F.2.2 Benchmarks for Sediments and Benthic Communities

       Sediment criteria for the protection of benthic invertebrates are derived based on the
equilibrium partitioning (EP) approach suggested by EPA's Office of Water:7

                                     SB = K,, x SWB,

       where  SB     =     sediment benchmark level (mg/kg);
              Kj     =     sediment partition coefficient; and
              SWB   =     surface water benchmark (mg/1).

The sediment partition coefficient (K^) for a non-ionic organic compound is calculated from its
organic carbonrwater partition coefficient (K^)  and the site-specific fraction of organic carbon
found in the sediment
For this analysis, K^ values were obtained from the literature, and f,,,. was assumed to be 0.01.

       F.2.3  Benchmarks for Drinking Water and Diet for Wildlife

       Drinking water or dietary ecological benchmark levels for terrestrial wildlife were derived
based on no-observed-adverse-effect levels or other low-observed-effect levels for oral exposures
reported in the literature.  Where appropriate, one or more extrapolation factors were used to
account for uncertainty in applying available toxicity values to the receptors of concern. For
toxicity values reported as dietary concentrations, an additional factor based on reported or
assumed daily intake levels and body mass was used in order to express benchmark levels in

       Benchmark Levels for Terrestrial Wildlife

       The toxicity of contaminants ingested by wildlife species were derived based on dietary
no-observed-effect levels (NOELs), lowest-observed-effect levels (LOELs), LCjo values, or LDjo
values reported in the literature; NOELs and, secondarily, LOELs are preferred and were used
when available. Likewise, data on chronic studies are preferred and were used when available,
but most benchmarks are based on subchronic or acute data.  Where appropriate, one or more
extrapolation factors were used to account for uncertainty in applying available toxicity values to
the receptors of concern.

   •    A factor of 500 was applied to LCjo and LDM values based on:
              A factor of 5 to extrapolate from 50 percent mortality to a low percent mortality
              level (EPA 1986);
    7  U.S. Environmental Protection Agency. Interim Sediment Criteria Values for Nonpolar
Hvdrophobic Organic Compounds. Washington, D.C.: Office of Water, 1988.

                            * • * DRAFT -- March 24, 1993 * * •

-------
                                             F-6

               A factor of 10 to extrapolate from acute to chronic exposures; and

               A factor of 10 to account for variation in species sensitivity.

       A factor of 50 was applied to chronic LOELs and other effect levels based on:

               A factor of 5 to extrapolate from a LOEL to a NOEL; and

               A factor of 10 to account for variation in species sensitivity.

       A factor of 10 was applied to a subchronic NOEL to account for variation in species
       sensitivity.
For toxicity values reported as dietary concentrations (e.g., LCjo values), an additional factor
based on reported or assumed daily intake levels and body mass was used in order to express
benchmark levels in appropriate units (mg/kg body mass).

       Drinking Water Intake

       The following formula was used to estimate daily drinking water intake for each receptor
of concern:

                                       D = C x I x P,

       where D =  daily dose (mg/kg body mass);
              C =  concentration in surface  water (mg/1);
              I  =  daily water intake for receptor (I/kg body mass); and
              P =  proportion of daily water intake obtained from surface water body being
                    evaluated.

For this analysis, P was assumed to be 1 for each assessment area, and I was estimated separately
for deer, small mammals, and birds from the literature.

       Dietary Intake

       The following formula was used to estimate daily  food chain  intake for each receptor of
concern:

                                     D = CxIxFxP,

       where D =  daily dose (mg/kg body mass);
              C =  concentration in prey item (mg/kg);
              I   =  daily dietary intake for receptor (kg/kg body mass);
              F =  fraction of diet comprised of prey item; and
              P =  proportion of daily food  intake obtained from assessment area being
                    evaluated.
                            • * *
                                 DRAFT « March 24,1993 • • •

-------
                                           F-7

Values for C were based entirely on estimated residues in earthworms (for robins and shrews)
and shrews (for red-tailed hawks and barred owls). Residues were estimated based on measured
soil concentrations within each exposure area and on literature values for soihearthworm and
soil:shrew bioaccumulation factors (BAFs). Note that most of the other prey eaten by hawks and
owls are herbivorous and would not be expected to bioaccumulate significant amounts of soil
contaminants.  Values of the exposure parameters (I, F, and P) were estimated separately for the
robin, shrew, barred owl, and red-tailed hawk from the literature.
       Aqueous Benchmarks for Piscivorous Wildlife

       EPA's Environmental Research Laboratory — Corvallis developed a procedure for
deriving screening-level surface water criteria protective of wildlife (SLWC) through exposure via
drinking water and consumption of contaminated aquatic prey.* The method involves back-
calculating a surface water concentration based on a NOEL or other appropriate toxicity value
for the chemical and wildlife species of concern, the aquatic life bioconcentration factor (BCF)
for the chemical of concern, various exposure parameters (e.g., daily food and water consumption
rates, body mass) for the species of concern, and a unitless extrapolation factor to account for
differences in species sensitivity to the chemical of concern.  The general  formula for deriving an
SLWC is:

                                  SLWC_NOELxMxSSF
                                          W+(FxBCF)

where: SLWC = screening-level wildlife criterion (mg/L);
       W     = average daily water consumption of animal (L/day);
       F      = average daily food consumption of animal (kg/day);
       BCF   = aquatic life bioconcentration factor (L/kg);
       M     = average body mass of animal (kg);
       SSF    = species sensitivity factor (unitless values from 0.01 to 1); and
       NOEL = no observed  effect level or equivalent toxicity value (mg/kg-day).

       BCF values for contaminants of concern were obtained by taking the lower of two values:
the BCF value for fish listed in the September 1989 draft of the Chemical Data Bases for the
Multimedia Environmental Pollutant Assessment System (MEPAS):  Version I,9 or the higher of
   8 U.S. Environmental Protection Agency. Water Quality Criteria to Protect Wildlife
Resources. Corvallis, OR: Office of Research and Development, Environmental Research
Laboratory, December 1989. EPA Report No. EPA/600/3-89/067.

   9 U.S. Department of Energy. Chemical Data Base for the Multimedia Environmental
Pollutant Assessment System (MEPAS):  Version 1.  Richland, WA: External Review Draft
prepared for the Assistant  Secretary, Office of Environmental Audit by Pacific Northwest
Laboratories, 1989.
                           • * * DRAFT - March 24, 1993 * • *

-------
                                            F-8

available freshwater or saltwater BCF values listed in the April 1991 draft of the Superfund
Chemical Data Matrix (SCDM). These BCFs do not represent measured values for site-specific
fish likely to be eaten by the receptors of concern.  Values in the MEPAS data bases generally
represent whole-body BCFs for freshwater fish calculated from the octanol-water partition
coefficient (organic substances) or by other extrapolation techniques (inorganic substances).
Values in the SCDM represent the highest BCF measured  for any aquatic species (i.e., they are
not limited to fish). Because BCF is a term in the denominator of the SLWC methodology,
using the lower of these two values yields a higher SLWC (i.e., a higher BCF would lower the
ecological benchmark level).

       Fish Contamination Benchmarks for Piscivorous Wildlife

       Dietary concentrations in fish flesh protective of piscivorous wildlife (i.e., birds,
mammals) were  determined following New York State's approach.10

       ₯2.4  Benchmarks for Soils

       Soil ecological benchmark levels for terrestrial plants were derived based on soil
contaminant concentrations associated with LOELs extrapolated from a LOEL to a NOEL.
Phytotoxicity data for other soil contaminants were not found in the available literature. Data
also were available on the acute toxicity of contaminated soils based on earthworm toxicity tests.
Finally, the US Fish and Wildlife  Service's data for evaluating soil contamination also were
used.11
FJ    Methodology for Estimating Extent of Contamination

       Extent of contamination was estimated for surface water contamination at facilities where
non-metals exceeded ecological benchmarks.  The following steady-state decay equation was
used:
       Where:  A   =  distance contaminated above benchmark (m)
                Cb   =  ecological benchmark level (mg/1)
    10 New York State Department of Environmental Conservation. Niagara River Biota
Contamination Project:  Fish Flesh Criteria for Piscivorous Wildlife. Albany, NY: Division of
Fish and Wildlife, Bureau of Environmental Protection, 1987. Technical Report 87-3.

    11 Beyer, W.N.  Evaluating soil contamination. U.S. Fish and Wildlife Service/DOI.
Biological Report Volume 90 (1990), pp.16-17.

                            • • • DRAFT - March 24,1993 • • •

-------
                                            F-9

                C0  = predicted concentration at point of discharge (mg/1)
                V   = stream velocity (m/sec)
                X   = decay coefficient (sec)"1

In this equation, the decline in contaminant concentrations over distance (i.e., the decay
coefficient or X) is determined solely by the degradation rate of the hazardous substance in
surface water (i.e., it does not account for dilution occurring in the surface water body due to
downstream tributary, runoff, and ground water inputs). Degradation parameters used to
estimate X included aerobic and anaerobic oxidation, photolysis, photooxidation, reduction,
volatilization, and hydrolysis; rates for all of these processes were summed.

F.4    Proximity Analysis Results

       Exhibit F-2 summarizes data on estimated total acreage of surface water, terrestrial,
agricultural, residential, and industrial/other environments on-site and within one mile of each
facility in the analysis.  Also presented for each facility are distance to the nearest surface water
and terrestrial habitat, number of "sensitive environment" types on-site and within one mile, and
distance to the nearest sensitive environment.  These results were summarized and discussed  in
section 8.2.1.

F.5 Concentration-based  Screening Analysis Results

       Exhibit F-3 lists the hazard index (HI) estimated for sample facilities at which maximum
concentrations predicted for nearby surface waters by MMSOILS exceeded the benchmarks for
the protection of aquatic life.  Under the central-tendency scenario, the HI for nine of the
sample facilities are predicted to exceed 1.0, for two sample facilities to exceed 10, and for one
sample facility to exceed 100.  Under the high-end scenario, the His for eighteen of the sample
facilities are predicted to exceed 1.0, for fourteen sample facilities to exceed 10,  for six sample
facilities to exceed 100, for three sample facilities to exceed 1,000, and for one sample  facility to
exceed 10,000.

F.6 Time Sequence Results

       For facilities at which surface water concentrations of constituents  exceeded ecological
benchmark levels, predicted concentrations were plotted over the 128-year modeling period.
Results are shown in Exhibit F-4.

F.7 Extent of Contamination Results

       The maximum extent of contamination above benchmark levels (capped at 12 miles,
approximately 63,000 feet) for each facility is presented as distance (ft) in  Exhibit F-3. In cases
where more than one constituent exceeded benchmark levels, only the constituent with the
greatest  estimated extent of contamination is shown in the table.  "NA" in  Exhibit F-3 indicates
that only metals exceed ecological benchmark levels, and extent of contamination could not be
estimated.  Under central tendency assumptions, extent of contamination could be estimated  for
                            • • • DRAFT - March 24,1993 • • •

-------
                                            F-10
two of the seven facilities at which estimated environmental concentrations exceeded ecological
benchmark levels.  Estimated extent of contamination ranged from 290 feet to 12 miles (cap).
                            • • • DRAFT - March 24, 1993 • * *

-------
                                             F-ll



                                         EXHIBIT F-2




                        NEARBY LAND USE AND SENSITIVE ENVIRONMENTS*'
HIGHER-THREAT FACILITIES

Obs.
No.
008
060
162
109
019
081
152
065
018
063
163
104
044 '
Combined Total Area On-sile and Within One Mile of Facility (Acres)*'
Surface
Water
9.100
6,400
3.600
3.300
3.200
2.700
2.100
1,800
UOO
uoo
1,200
UOO
910
Terrestrial
1300
1,000
470
13,000
UOO
3.000
270
4.200
2,200
2,000
590
14,000
360
Agricultural
0
0
560
0
0
2.400
<60
180
0
200
880
380
120
Residential
330
2300
850
130
430
1.900
240
<90
3400
<80
290
2,800
670
Industrial/
Other
2,700
3,700
2,600
600
1,700
1,900
3.900
320
4,500
'.-500
<20
740
2,000
Total*
13,970
13,480
8.040
17.180
6380
11,970
6,500
6,600
11.120
4350
2.940
18,820
4,100
Nearest Distance
(miles)?
Surface
Water
0.00
0.00
0.00
0.00
0.00
000
000
000
0.00
000
025
000
0.00
Terrestrial
000
0.00
0.00
0.00
0.00
0.00
0.00
000
000
0.00
0.00
000
0.00
Sensitive Environments^-
Number
2
1
3
1
1
1
0
1
1
1
1
1
1
Nearest Distance
(miles)!'
0.00
0.00
000
0.04
000
000
—
000
064
003
0.30
0.00
0.47
                                    •»*
                                       March 24,1993 Draft
                                                       *•»

-------
                                                  F-12




                                        EXHIBIT F-2 (continued)
HIGHER-THREAT FACILITIES (continued)

Obs.
No.
093
013
144
041
021
122
168
131
025
057
112
Combined Total Area On-site and Within One Mile of Facility (Acres)6'
Surface
Water
820
710
690
680
670
610
580
450
430
410
310
Terrestrial
2,600
1,600
180
200
2,400
2,100
1300
1,900
2.500
3.100
3,200
Agricultural
6300
5,700
2300
0
<40
1,700
200
960
430
2,500
0
Residential
410
0
100
2,100
1.400
2,900
700
1,400
950
990
3.400
Industrial
/Other
2,700
230
210
470
170
520
730
930
240
1300
2.100 '
Total*
12.720
8,280
3,490
3,440
4.690
7370
4,020
5,710
4,520
8350
9.050
Nearest Distance
(miles)?
Surface
Water
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.00
0.00
0.00
0.00
Terrestrial
0.00
0.00
0.00
0.00
0.00
0.0
0.00
0.00
0.00
0.00
0.00
Sensitive Environments^
Number
0
0
1
1
1
1
0
0
1
1
1
Nearest Distance
(miles'
—
—
0.00
0.00
0.00
0.00
—
—
035
0.00
0.42
                                           March .
93 Draft

-------
                                                  F-13




                                         EXHIBIT F-2 (continued)
HIGHER-THREAT FACILITIES (continued)

Obs.
No.
029
165
114
167
120
136
069
039
084
082
132
046
014
'078
Combined Total Area On-site and Within One Mile of Facility (Acres)?
Surface
Water
290
290
240
220
200
200
200
180
180
140
110
100
<90
<30
Terrestrial
750
1.100
810
630
620
710
1.100
2,100
2,300
3.200
370
1,200
2,900
2,600
Agricultural
0
<50
0
2.900
1,100
0
<90
0
330
570
1.400
1.500
0
160
Residential
1.200
380
4.500
1.700
1,100
1.500
500
260
2,900
160
750
5,500
310
380
Industrial
/Other
1.600
770
180
320
1,100
1,700
2.900
130
350
140
l,00b'
630
1,400
70
Total*
3,790
2.570
5,760
5.730
4.130
4.070
4.840
2,640
6,100
4.220
3,680
8.960
4.710
3,220
Nearest Distance
(miles)?
Surface
Water
0.00
0.00
0.00
0.00
0.28
0.13
0.00
0.00
0.00
0.01
0.00
0.00
0.00
0.00
Terrestrial
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Sensitive Environments-4'
Number
1
1
1
0
1
0
1
1
1
0
1
0
1
1
Nearest Distance
(miles)!'
0.00
0.00
0.00
—
1.00
—
0.00
0.34
0.38
—
0.80
—
0.00
0.00
                                        **•
                                           March 24,1993 Draft
                                                              *•»

-------
                                                                 F-14

                                                     EXHIBIT F-2 (continued)
LOWER-THREAT FACILITIES
Obs.
No.
125
124
002
074
017
048
146
047
080
049
190
153
067
027
Combined Total Area On-site and Within One Mile of Facility (Acres)*'
Surface
Water
180
<90
<80
<50
<50
<30
<30
<30
<20
2
1
1
0
0
Terrestrial
360
3.200
100
160
<50
520
<60
1,000
340
3.000
<70
1300
<50
210
Agricultural
300
460
3,000
50
0
0
0
1,400
0
30
0
200
0
uoo
Residential
90
780
780
1300
UOO
720
1,700
300
1,400
0
uoo
20
2,100
1000
Industrial/
Other
2,100
360
1,200
3,700
920
uoo
1.500
<60
450
3
2.100
<20
480
170
Total*
3,040
4380
5,200
5.790
2,300
3,340
3.270
2,760
2,260
2.990
3.680
2.040
2,620
2350
Nearest Distance
(miles)?
Surface
Water
0.23
0.00
0.00
0.16
0.15
0.34
0.70
0.53
0.00
0.00
0.25
0.09
-
-
Terrestrial
0.00
0.00
0.19
0.06
037
0.00
0.03
0.00
0.14
0.00
0.80
0.00
0.17
0.00
Sensitive Environments"*'
Number
0
0
0
0
1
0
0
0
0
0
0
0
0
0
Nearest Distance
(miles)^
—
—
—
—
0.99
—
—
—
—
—
—
—
—
—
* Facilities are grouped according to professional judgment into higher- and lower-threat categories as explained in sections 8.1.1 and F.I; within each category,
         facilities are sorted by surface water acreage.
" For a definition of land-use categories and sensitive environments, see text.
* Zeros in these columns indicate that  surface waters, terrestrial habitats, or other sensitive environments are present within the facility boundary, dashes in this
         column indicate that none of these environments were present on-site or within one mile.
* The sum of areas in the five categories does not match the total column because values for the five categories were estimated to within two significant digits
         whereas the value for the total was estimated separately to the nearest 10 acres.
- Dashes in  this column indicate that no sensitive environment is  present within one mile of the  facility.
                                                      »**
                                                          March
/93 Draft
                                                                                  **•

-------
                                         F-15

                                    EXHIBIT F-3

FACILITIES AT WHICH MAXIMUM PREDICTED CONCENTRATIONS EXCEEDED
                       ECOLOGICAL BENCHMARK LEVELS
Obs.
No.
013
021
025
041
044
046
049
060
063
069
081
104
112
114
122
124
144
153
167
Hazard Index
CI*
3 .5 x 102
2.2 x 10's
7.8 x 10"'
1.4 x 10°
2.4 x 10 3
1.1 x 10°
1.8x10°
9.1 x 10'1
4.2 x 10°
6.2 x ID"4.
4.1 x 10°
1.7 x 10°
8.2 x lO'2
4.4 x 10+1
8.9 x lO'2
1.2 x 10°
1.7 x ID'1
4.7 x 10°
2.6 x 10-1
HE*'
U x 104
3.2 x 10°
2.6 x 102
5.4 x 10'
4.0 x 10'
5.0 x 10°
4.3 x 10 l
8.4 x 103
2.7 x 102
13 x 10'
2.3 x 10'
5.1 x 10°
13 x 10'
3.0 x 102
23 x 10'
1.2 x 103
1.5 x 101
8.7 x 10'
1.9 x 10°
Constituents'
formaldehyde
mercury
fluorene
acetone
chromium
lead
cadmium
phenol
tetrachloroethyleneA
cyanides
chromium
xylenes
chloroform
metals
pentachlorophenol
selenium
metals
benzene
metals
chromium
Maximum Extent of
Contamination* (feet)
era'
[63,000]
NA
NA
—
NA
NA
NA
NA
200
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
HE*
[63,000]
NA
[63,000]
[63,000]
NA
NA
NA
[63,000]
3,000
NA
17,300
28,400
NA
[63,000]
NA
NA
8,000
NA
NA
   • Central tendency scenario.
   -' High-end scenario.
   - Contaminant for which the calculated extent of contamination was greatest at that facility.
   - Extent of contamination was capped at twelve miles (approximately 63,000 feet)
   NA = not applicable (no organic chemicals exceeded ecological benchmark levels)
   — Unavailable.
                          ***
                              March 25,1993 Draft

-------
                                    F-16

                               EXHIBIT F-4

   SUMMARY OF TIME COURSE OF PREDICTED EXCEEDANCES OF
                  ECOLOGICAL BENCHMARK LEVELS
Obs.
No.
013
021
025
041
044
046
049
060
063
069
081
104
112
114
122
124
144
153
167
Approximate Duration of
Predicted Exceedance Beyond
1992 (years)*
CTS/
>130
—
—
20
.._
>130
>130
—
30
—
—
—
—
—
...
>130
—
>130
—
HE*
>130
80
>130
>130
>130
>130
—
>130
>130
>130
>130
20
>130
>130
>130
>130
>130
>130
60
Overall Pattern of
Predicted
Concentrations-'
Falling
Rising
Rising
Mixed
Rising
Constant
Constant
Rising
- ' Falling
Falling
Mixed
Falling
Rising
Rising
Rising
Mixed
Mixed
Constant
Rising
- For substance with greatest exceedance duration
* Pattern exhibited by majority of substances that exceeded benchmark levels
- Centra] tendency scenario
* High-end scenario
— indicates that no constituent exceeded its ecological benchmark under that scenario
                          March 24,1993 Draft

-------
                                            F-17

Under high-end assumptions, extent of contamination could be estimated for nine of the 19
facilities at which estimated concentrations exceeded benchmark levels.  Estimated extent ranged
from 2,700 feet to 12 miles (cap).

F.8 Qualitative Case Studies

       Qualitative case studies were prepared for three facilities to identify types of risks that
are not accounted for by the proximity and screening concentration-based analyses, as described
in section 8.1.3. The case-study analyses followed the EPA Risk Assessment Forum's framework
for ecological risk assessment.  For each facility, assessment and measurement endpoints were
selected to evaluate  ecological threats.  Assessment endpoints are the actual entities or
environmental characteristics that are to be protected (e.g., population abundance, biodiversity)
or adverse effects that are to be prevented (e.g., extinction, contamination); measurement
endpoints are measurable environmental characteristics that approximate, represent or lead  to
the assessment  endpoint(s) using field or laboratory methods (e.g., the chemical concentration
shown to cause a reduction in survival, growth, or reproduction in a standard laboratory toxicity
test).

       F.8.1  Facility A

       Facility  A is a large facility that  has manufactured a wide variety of chemicals  for several
decades. Roughly one-third of the more than thousand acre facility is developed. The property
boundary is bounded by a forest, a neighboring facility, and a large river. A wide array of
chemicals have been generated as waste during manufacturing operations; the majority of wastes
and residues have been managed, treated, and disposed on-site throughout the site's history.
Numerous areas on-site, once used for the management and disposal of wastes in the past, have
since been closed or replaced by newer  systems.  Elevated levels of chlorinated pesticides, volatile
solvents, and inorganic chemicals have been found in ground  water, surface water, sediments, air,
and soils.

       For terrestrial areas, representative ecological receptors of concern were selected based
on the plant and animal species characteristic of each area, the relative potential for exposure to
DDT and its metabolites (DDTR), and  the availability of toxicity data. Terrestrial receptors of
concern included the robin,  red-tailed hawk, barred owl, shrew, deer, and other small  mammals
and birds.  Earthworms also were selected as receptors of concern for soil toxicity tests used to
evaluate the acute toxicity of contaminated soils. For this analysis, assessment endpoints for
terrestrial biota are focused at the population level; measurement endpoints were focused at the
individual level. Assessment endpoints  for aquatic biota are focused at the
population/community level; measurement endpoints were mortality and stream benthic
community structure.

       The evaluation focused on exposure pathways likely to result in the greatest levels of
exposure for the receptors of concern: soil, surface water, sediment, and terrestrial biota. The
exposure assessment focused on developing estimates of representative and maximum
concentrations and/or exposure levels within each assessment area via the appropriate exposure
pathway(s). Representative concentrations or exposure levels were based on the mean
                              »**
                                  March 24,1993 Draft

-------
                                            F-18

concentration within the exposure area of concern; maximum concentrations or exposure levels
were based on the highest measured concentration.  To estimate daily intake levels for wildlife
from environmental concentration data, diet composition and food and water ingestion rates
were identified from the available literature, including EPA's Wildlife Exposure Factors
Handbook, currently being developed by the Office of Research and Development.  Risk to
receptors of concern was evaluated by comparing average and/or maximum measured
contaminant  concentrations (or estimated  intake levels, as appropriate) to the appropriate
ecological benchmark levels. The results of the risk characterization are  summarized below.

       There appears to be little risk to terrestrial wildlife from drinking contaminated surface
water. With  the exception of diazinon, maximum estimated intake levels are well below
benchmark levels; average intake levels for all contaminants are below benchmark levels. Robins
and shrews ingesting earthworms and other invertebrates from  the contaminated areas may be at
risk for lethal effects from DDTR exposure.  Estimated mean and maximum intake levels for
owls and hawks feeding on small mammals from the contaminated area are well below
benchmark levels, suggesting little risk to these species.  Mean  and maximum concentrations of
arsenic, copper, lead, and DDTR are greater than phytotoxic levels in all exposure areas
evaluated within the floodplain area.  However, benchmark levels for the inorganic chemicals are
below site-specific background levels, so evaluating risk from these contaminants is difficult.
Areas of bare soil coincide with high soil contaminant concentrations, suggesting phytotoxic
effects of these contaminants.

       Maximum aqueous concentrations  of all contaminants are below benchmark levels for the
protection of aquatic life in on-site  surface waters; no contaminants were found above
background off-site.  There thus is little risk to aquatic life from contaminated surface water.
Sediment concentrations, however,  exceed benchmark levels. In an on-site creek, average and
maximum sediment concentrations  of ametryn, diazinon, and metolachlor were greater than
ecological benchmark levels. In the off-site river, average and  maximum sediment concentrations
of diazinon were greater than ecological benchmark levels.  These results suggest that adverse
effects may be occurring in the benthic communities of the creek and river. To the extent that
these communities support part of the entire aquatic community's food chain, indirect adverse
effects on higher trophic levels also might occur.  This conclusion is supported by a survey that
found that the benthic community in the creek is dominated by chironomids, species typically
dominant in  stressed waters. However, such a finding could be due to low dissolved oxygen in
the creek.

       Approximately 60 acres of soil in the floodplain areas (70 percent of suitable open field
habitat) are contaminated above benchmark levels for terrestrial plants, robins, and shrews.  Tens
of acres of the  floodplain are devoid of vegetation.  A significant proportion of the populations
of the robins and shrews utilizing the floodplain  area may be adversely affected. Given  the
tendency of DDTR to bioaccumulate in terrestrial biota, impacts beyond the contaminated  area
are possible, although  estimated intakes for hawks and owls are below benchmark levels.
Sediments in an on-site creek are contaminated above benchmark levels  for a distance of
approximately 6,000 ft (1.2 miles).  It is possible that benthic communities  are altered within this
entire length. Downstream river sediments at sampling points approximately 18,000 ft (3.5 miles)
                                  March 24, 1993 Draft

-------
                                           F-19

apart are contaminated above benchmark level.  Benthic life within the river is possibly adversely
affected for at least 3.5 miles, and the zone of impact may extend farther downstream.

       The available data do not allow for separation of the effects of SWMU releases from
permitted releases under the facility's NPDES outfalls.

       F.8.2  Facility B

       Facility B is a large waste treatment facility sited on approximately several hundred acres
of land bounded on three sides by surface water: two brackish tidal marshes and a brackish tidal
creek.  The watershed of the tidal creek is predominantly agricultural.  Point source discharges
from other major industrial/commercial facilities in the  area may contribute pollutant loadings to
the creek.  Stream loadings of various other non-point source pollutants are also attributable to
runoff from agricultural  activities. The water quality of the creek is generally poor;
concentrations of ammonia,  phosphorus, and copper routinely exceed water quality criteria and
state standards.  Historical concentrations of cadmium,  mercury, and PCBs have occasionally
exceeded water quality criteria and standards.

       Receptors of concern for surface water contamination include fish and other aquatic life
as well as piscivorous wildlife.  Representative piscivorous receptors (i.e., great blue heron,
kingfisher, mink) were selected based on likelihood of presence in the area, position  in the food
chain (i.e., top carnivores), and availability of toxicity data. Receptors of sediment contamination
include benthic invertebrates. Assessment endpoints for aquatic biota (both pelagic and benthic)
are focused at the population/community level; measurement endpoints were mortality and
stream benthic community structure.  Assessment endpoints for piscivorous wildlife are focused
at the population level; measurement endpoints were mortality, reproduction, and sublethal
adverse effects.

       The evaluation focused on exposure pathways likely to result in the greatest levels of
exposure for the receptors of concern:  surface water, sediment,  and aquatic biota. The exposure
assessment focused on developing estimates of representative and maximum concentrations.
Representative concentrations or exposure levels were based on the arithmetic mean of available
concentration data for an exposure area; maximum concentrations or exposure levels were based
on the highest measured concentration.  Risk to receptors of concern was evaluated by
comparing mean and/or maximum measured contaminant concentrations to media-specific
ecological benchmark levels. The results of the risk characterization are presented below.

       In the tidal creek and north marsh, maximum concentrations of arsenic, chromium,
mercury, and possibly PCBs  exceeded chronic benchmark levels. Maximum concentrations of all
except mercury (and possibly PCBs) exceed acute benchmark levels as well. In the north marsh,
average concentrations of many of these substances also exceeded chronic benchmark levels.
Thus there is some risk  to aquatic life from contaminated surface water.

       Maximum sediment concentrations of arsenic, barium, cadmium, chromium, copper,
mercury, PCBs, and zinc all  exceeded benchmark levels in the north marsh and tidal  creek;
average and maximum concentrations of many of these  substances in the north marsh, south


                             *** March 24, 1993 Draft ***

-------
                                            F-20

marsh, and tidal creek also exceeded guidelines for heavily polluted sediments.  These results
suggest that contamination from the facility may be adversely affecting the benthic communities
of the creek and the marshes.  This conclusion is supported by a benthic organism survey in the
tidal creek that found no amphipods near the facility outfall, although many were present both
upstream and 2,000 feet downstream from the facility.

       There appears to be some risk to piscivorous wildlife from eating contaminated aquatic
organisms. Maximum measured surface water concentrations of mercury, zinc, and possibly
PCBs exceed ecological benchmarks for the protection of piscivorous wildlife. However, the
inability to determine average aqueous concentrations  from available data and the uncertainty
involved in developing benchmark levels makes it difficult to characterize risk to these receptors
of concern.

       Approximately 45 acres of tidal wetland sediments in marshes and possibly 4,000 feet of
the tidal creek are contaminated with metals and PCBs above ecological benchmarks.  Water
quality in the tidal creek is generally within water quality standards  for the protection of aquatic
life, except that zinc exceeds both background levels and the ecological benchmarks for the
protection of aquatic life as well as benchmarks for the protection of piscivorous wildlife. The
water-table zone ground water that surfaces in one marsh area is contaminated  with mercury and
possibly PCBs above  benchmark levels  for the protection of aquatic life.  Water contamination in
another marsh area is undetermined.  Moreover, concentrations of most contaminants in
background samples also exceeded ecological benchmark levels, making it difficult to determine
whether observed  contamination can be attributed to the facility.

       The available data  do not allow for separation  of the effects of SWMU  releases from
permitted releases under the facility's NPDES outfalls. However, ground-water flow is the
primary method for migration of contaminants from the site.

        F.8.3  Facility C

        Facility C  is sited in a lightly developed area, and the undeveloped on-site land is either
woodland or field. Off-site land near the facility is predominantly surface water or wetland. The
facility is surrounded by woodland, wetlands, residences, and a lake. Two creeks flowing from
the lake pass the facility.

        Most of the wastes generated at the facility are shipped off-site, but some are treated on-
site.  Ground water is contaminated with volatile organic constituents (VOCs) and heavy metals.
Lead, zinc, mercury, and other metals are  found in the edible portions of fish in the lake.

        Ecological receptors  of concern include fish and other aquatic life as well as piscivorous
wildlife.  Representative piscivorous receptors (i.e., mink, belted kingfisher, and great blue heron)
were selected based on likelihood of presence in the area and availability of toxicity data.
Assessment endpoints for aquatic biota and piscivorous wildlife were focused on the
population/community level. Measurement endpoints  include mortality and
morphological/physiological condition  of individuals.
                                  March 24, 1993 Draft

-------
                                           F-21

       The evaluation focused on exposure pathways likely to result in the greatest levels of
exposure for the receptors of concern. The exposure assessment focused on developing estimates
of representative and maximum concentrations.  Representative concentrations or exposure levels
were based on the arithmetic mean of available concentration data; maximum concentrations or
exposure levels were based on  the highest measured concentrations.  Risk to receptors of
concern was evaluated by comparing mean and maximum contaminant concentrations in surface
water and fish tissue to ecological benchmark levels. The results of the risk characterization are
presented below.

       In the lake near the facility, maximum and average concentrations of cadmium, copper,
lead, and mercury were above ecological benchmark levels for the protection of aquatic life,
indicating that waters of the lake pose a significant threat to these ecosystems. In a nearby
stream/wetland where contaminated ground water discharges, mean and maximum concentrations
of cadmium, copper, lead, and mercury were  above ecological benchmark levels.  Lead, zinc, and
copper showed slight declines in concentration with increasing downstream  distance.  In addition,
mean and maximum concentrations of three volatile organic constituents were above ecological
benchmark levels. Concentrations of VOCs were highest at the suspected ground water seeps,
and they attenuated rapidly with downstream distance due to volatilization and dilution.
Additional evidence that Facility C poses a threat  to aquatic life is the presence of internal and
external parasites and of fin rot (a fungal disease) in fish from the lake. Parasites and fin rot
were most prevalent among fish collected closest to the facility.

       Concentrations of heavy metals were  above dietary benchmark levels for the protection of
piscivorous wildlife in all three fish species collected from the nearby lake.  Two constituents
(i.e., cadmium and mercury) were several orders of magnitude above ecological benchmark levels
in all three fish species collected from the  lake.  Only one constituent (i.e.,  lead) was ever found
below dietary benchmark levels.  Chromium, copper, and zinc were consistently approximately
one order of magnitude above dietary benchmark  levels. Given these results, fish-eating wildlife
in the vicinity of the lake appear to be at risk.

       Because heavy metals were found above surface water ecological benchmark levels for the
protection of aquatic life at all stations (including the background station),  the entire lake is
likely to exhibit an impaired aquatic ecosystem.  Samples of fish tissue clearly suggested that
contaminants enter  the food chain and pose risk to fish-eating wildlife species such as mink,
belted kingfisher, or great blue heron. Contaminated waters exit the lake via two streams.
Heavy metals were above benchmark levels in at least  1,560 feet of one of the streams  and its
seep tributaries in a nearby wetland.  If the surface water flowing from the  lake is assumed to be
contaminated above benchmark levels, then an additional  1,580 feet of the  stream would be
contaminated for a total of 3,140 feet. The wetland may be contaminated above benchmark
levels by ground water discharge and by surface water overflow from the stream.  However, the
extent of this contamination cannot be determined, because wetland  sediments were not sampled
and aqueous samples were taken within water courses only.
                              »**
                                 March 24,1993 Draft

-------
                                      APPENDIX G

                               KEY PARAMETERS MATRIX
       This appendix presents the values of the parameters used in the human health risk
assessment and their sources.  In particular, the parameters are those for the major steps in
EPA's risk assessment methodology:

              Waste characterization;
              Release analysis;
              Fate and transport analysis;
              Exposure analysis;
              Dose-response assessment; and
              Risk characterization.

Parameters are presented in matrix format; the matrix includes values of the key parameters used
to model the four fundamental risk descriptors:1

       •      Central tendency individual risk;
       •      High-end individual risk;
       •      Risk to highly-exposed subpopulations (i.e., subsistence fishermen, subsistence
              farmers, and pica children); and
       •      Population risk.

In addition, the parameter matrix includes a qualitative description of the relative uncertainty of
each parameter.

       Although this Appendix primarily supplements Chapter 7 and Appendix E, which cover
the human health risk assessment,  the parameters presented in  the matrix are relevant to
Chapters 3, 4, 8, 9, 10, and 12 as well.

       To the extent possible,  the Agency's selection of key parameters is consistent with recent
EPA guidance on risk characterization, (i.e., Guidance on Risk Characterization for Risk
Managers and Risk Assessors,  dated February 26, 1992). Sources of parameter values are
identified in endnotes to the parameter matrix.
   1 See Chapter 7 for a detailed discussion of these risk descriptors.

                              •** Draft -- March 23,1993 ***

-------

Parameter
WASTE CHARACTERIZATION
Waste Volume
Constituent Concentration
Distribution Coefficient (Kd)
Non-karst facilities
Solubility Limit
Organics
Inorganics
Degradation Rate
Individui
Central Tendency
0)
\
Best estimate (e.g.,
based on site-specific
SWMU dimensions)
Midpoint of SWMU-
specific data or TSDR
values
pH-dependent mid-point
value specified by ORD'
Solubility limit values
from ORD'
Central tendency
solubility limit values from
ORD'
pH-dependent best
estimate of hydrolysis
rates from ORD'
il Risk
High-End
(2)

High-end estimate
based on site-specific
SWMU dimensions
High-end of SWMU-
specific data or TSDR
values
pH-dependent low-end
value specified by
ORD'

High-end solubility
limit values from ORD'


Population Risk
(3)

Same
Same
Same
Same as (1)
Same
Same as (1)
Individual Risk to
Highly-Exposed
Subpopulations
(4)

as(1)
as(1)
as(1)

as(1)

Parameter
Uncertainty

Medium-
High
Medium-
High
Low
Low
Low
Medium
*** Draft -- March 23,1993 ***
              1

-------
Parameter
Individual Risk
Central Tendency
0)
High-End
(2)
Population Risk
(3)
RELEASE ANALYSIS
Containment System
Failure
Leachate Concentration
Surface
Impoundments and
Tanks
Landfills
Mass Balance
Volatilization
Particulate Emissions
Soil Erosion
(Dissolved/
Adsorbed)
Leachate
Site-specific best
estimate based on
SWMU characteristics
Site-specific best
estimate of pond/tank
concentration
Organics: Derived from
initial concentrations and
solubilities from ORD,
using organic leachate
model (OLM);b
Inorganics: Used ORD
central tendency
solubility values'
Non-steady state release
Steady-state release until
mass is depleted0
Steady-state release until
mass is depleted0
Non-steady state release
Individual Risk to
Highly-Exposed
Subpopulations
(4)
Parameter
Uncertainty

Same as (1)
Site-specific high-end
estimate of pond/tank
concentration
Organics: Equal to
high-end solubility
limit;'
Inorganics: Used ORD
high-end solubility
values'
Same
Same
as(1)
as(1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Medium
Medium
Medium
Low
Low
Low
Low
***
    Draft
h 23, 1993 ***

-------
Parameter
Individual Risk Individual Risk to
Hinhlw-FvnncoH
Central Tendency
(1)
High-End Population Risk Subpopulations
(2) (3) (4)
Pare, .dr
Uncertainty
FATE AND TRANSPORT ANALYSIS
Recharge
Depth of soil layer
Depth of root zone
Percent Organic
Matter
Field capacity
Wilting point
Saturated
conductivity
Saturated water
content
Exponent "b" for
moisture curve
Ground Water Pathway
Hydraulic
conductivity
Hydraulic gradient
Porosity
(Non-karst facilities)
Dispersivity
Site-specific best
estimate
10 cm (constant
assumption)0
Site-specific best
estimate
Site-specific best
estimate
Site-specific best
estimate
Site-specific best
estimate
Site-specific best
estimate
Site-specific best
estimate
Site-specific best
estimate
Site-specific best
estimate
Site-specific best
estimate
Scale-dependent
Same as (1)
Same as (1)
Site-specific high-end Same as (1)
estimate
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Order-of-magnitude Same as (1)
higher than best
estimate
Same as (1)
Same as (1)
Same as (1)
Low
Low
Low
Low
Low
Low
Low
Low
Medium
Low-
Medium
Low
Medium
*** Draft - March 23,1993 ***
             3

-------
Parameter
Non-Karst Facilities
Advection
Dispersion
Karst Facilities
Retardation
in ground-
water
pathway
Advection
Dispersion
Individual Risk
Central Tendency
(D
Derived from central
tendency estimates of
gradient, conductivity,
porosity
Derived from central
tendency estimates of
advection, aquifer
thickness, and distance
from unit
Assumes no contaminant
retardation, simulated by
setting Kd = 0 for
inorganics and by setting
aquifer 1^ = 0 for
organics
Derived from central
tendency estimates of
gradient, conductivity,
and porosity set to 0.01
Derived from central
tendency estimates of
advection, aquifer
thickness, and distance
from unit
High-End
(2)
Derived from high-end
estimates of
conductivity
Derived from central
tendency estimates of
aquifer thickness and
distance from unit and
high-end estimates of
advection
Retardation is not
applicable given
assumption of
advection and
dispersion
Advection is assumed
to be instantaneous;
downgradient
exposures at any point
in any year are
assumed to be at
concentrations equal
to those exiting the
unsaturated zone in
that year
Dispersion is assumed
not to apply
Individual Risk to
Highly-Exposed
Population Risk Subpopulations
(3) (4)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Parameter
Uncertainty
Medium-
High
Medium-
High
Medium-
High
Medium-
High
Medium-
high
*** Draft
?h 23,1993 ***

-------
Parameter
Distance to wells,
surface water
Well Depth
Non-Aqueous Phase
Liquids (NAPLs)
Air Pathway
Mixing depth of site
soil
Fraction of vegetative
cover
Vehicle traffic
parameters
Stability array data
Distance to
receptors, off-site
fields, and off-site
agricultural fields
Surface Water Pathway
Flow rate in river
Individual Risk
Central Tendency
(1)
Site-specific best
estimate
Site-specific best
estimate
NAPLs modelled using
same
advection/dispersion
approach as for other
contaminants (i.e., there
was no modelling of two-
phase flow for light or
dense NAPLs)
15 cm (constant
assumption)6
Site-specific best
estimate
Site-specific best
estimate
Wind velocity = 3 m/s,
stability class E, 30%
frequency (per ORD)
Site-specific best
estimate
Annual avg. or based on
stream order
High-End
(2)
Population Risk
(3)
Same as (1)
Screened at surface of
aquifer
Same as (1)
Same as (1)
Individual Risk to
Highly-Exposed
Subpopulations
(4)

N/A

Same as (1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Lowest monthly avg.
or based on next
smallest stream order
Same
as(1)
Pe»._..ieter
Uncertainty
Low
Medium
High
Low
Medium
Medium
Medium
Low
Medium
*** Draft - March 23,1993 ***
             5

-------
Parameter
Sediment dilution
ratio (lakes only)
Distance to surface
water intakes and
recreational uses
(swimming, fishing)
Soil Pathway
Sediment delivery
fraction
Area of field
Distance to field
Foodchaln Pathway
Sediment delivery
fraction
Area of field
Fraction of Organic
Carbon in Field
Plant uptake rates
Bioconcentration and
uptake
Individual Risk
Central Tendency
(D
Best estimate
Site-specific, nearest
downstream use-point on
topo maps or default of
10 meters downstream
from discharge point
Site-specific best
estimate
Site-specific best
estimate
Site-specific best
estimate, or default
distance (at facility
boundary)
Site-specific best
estimate
Site-specific best
estimate
Site-specific best
estimate
Chemical-specific best
estimate from literature"
Chemical-specific best
estimate from literature*
High-End
(2)
High-end estimate
Population Risk
(3)
N/A
Individual Risk to
Highly-Exposed
Subpopulations
(4)
N/A
Same as (1)
Site-specific high-end
estimate
Same as (1)
Same as (1)
Site-specific high-end
estimate
Same as (1)
Site-specific high-end
estimate
Same as (1)
Same as (1)
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Parameter
Uncertainty
Medium
Low
Medium
Low
Low
Medium
Low
Low
Medium-
High
Medium-
High
*** Draft -
23,1993 ***

-------
Parameter
Distance to off-site
agricultural fields and
fishing locations
Individual Risk
Central Tendency
(1)
Site-specific. Fishing
location: nearest
downstream use-point on
topo maps or default of
10 meters downstream
from discharge point.
Agricultural field: closest
agricultural field identified
on topo map.
High-End
(2)
Same as (1)
Population Risk
(3)
N/A
Individual Risk to
Highly-Exposed
Subpopulations
(4)
Same as (1)
Parameter
Uncertainty
Low-
Medium
*** Draft - March 23,1993 •••
             7

-------
Parameter
Individual Risk
Central Tendency
0)
High-End
(2)
EXPOSURE ANALYSIS
General Assumptions
(apply across all exposure
pathways, except where
noted)
Body Weight
Exposure Duration
Exposure Frequency
Ground Water Pathway
Maximum Distance
Modeled (from facility
boundary)
Ingestion of
Contaminated
Drinking Water
Contact Rate
Dermal Absorption of
Contaminants while
Showering
Exposure
Time
Surface Area
Exposed
while
Showering
70 kg adult (average)'
9 years (median
residence time)'
350 days/year"
5 miles (or less if site-
specific data indicate
plume intercepted by
surface water)
1. 4 L/day (average)'"
7 minutes/da/"
19,400 square cm'
Population Risk '
(3)
Individual Risk to
Highly-Exposed
Subpopulations
(4)
Parameter
Uncertainty

Same as (1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
N/A
N/A
N/A
N/A
Low
Low
Low
N/A
Low
Low
Low
*** Draft -
23,1993 ***

-------
Parameter
Inhalation of
Contaminants in
Indoor Air from
Contaminated
Ground-water
Contact Rate
Exposure
Location
Atr Pathway
Maximum Distance
Modeled (from facility
boundary)
Inhalation of
Contaminants from
Air
Contact Rate
Exposure
Location
Soil Pathway
Maximum Distance
Modeled (from facility
boundary)
Exposure Duration
Individual Risk
Central Tendency
(1)
15 m3/day (average)" "
Exposure assumed
throughout home to
average concentration in
indoor air.
10 kilometers
20 m3/day (average)'
Exposure assumed
throughout home to
same concentration as in
outdoor air at that
location
Site-specific, depending
on topographic and
meteorological
characteristics of the area
Adult: 9 years'
Child: 5 years"
High-End
(2)
Population Risk
(3)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
N/A
N/A
Individual Risk to
Highly-Exposed
Subpopulations
(4)
N/A
N/A
N/A
N/A
N/A
Same as (1)
Child: 5 years
Paiaineter
Uncertainty
Low
Low
N/A
Low
Low
N/A
Low
*** Draft - March 23,1993 ***
            9

-------

Parameter
Exposure Frequency
Ingestion of Soil
Contact Rate
Body Weight
Dermal Absorption
from Soil
Surface Area
Exposed for
Dermal
Absorption
Exposure Location
Individual Risk
Central Tendency
(1)
Adult: 350 days/year9
Child: 350 days/year9
Adult: 100mg/day
(average)1
Child: 200mg/day
(average)"1
Adult: 70 kg (average)'
Child: 16 kg (median)"
Adult: 5,000 square cm"
Child: 2,500 square cm"
Site specific estimates;
exposure at off-site fields
(e.g., playgrounds,
athletic fields, and school
yards) and off-site
agricultural fields (e.g.,
field crops, pastures, row
crops, orchards,
vineyards, and large
suburban vegetable
gardens) assumed to
constituent concentration
calculated as an average
over entire field
High-End
(2)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Population Risk
(3)
N/A
N/A
N/A
N/A
N/A
Individual Risk to
Highly-Exposed
Subpopulations
(4)
Child: 365 days/year
Child: 800 mg/day
(upper range)'
Child: Same as (1)
Child: Same as (1)
N/A
Parameter
Uncertainty
Low
High
Low
Low
Low
***
   Draft
h 23,1993

-------
Parameter
•H^M^MHi^HM
Surface Water Pathway
Maximum Distance
Modeled (from facility
boundary)
Ingestion of
Contaminated
Drinking Water
Dermal Absorption of
Contaminants while
Showering
Inhalation of
Contaminants in
Indoor Air from
Contaminated
Surface Water
Dermal Absorption
while Swimming
Exposure
Frequency;
Exposure
Time
Surface Area
Exposed
Incidental Ingestion
of Water while
Swimming
Exposure
Frequency;
Exposure
Time
Individual Risk
Central Tendency
0)
Site-specific (specified
uses within 15 miles)
Exposure due to
household use of surface
water is evaluated using
the same assumptions as
for the corresponding
ground water exposure
routes
26 days/year1'1
2.6 hours/day*
19,400 square cm*
26 days/year"
2.6 hours/day1
High-End
(2)

Population Risk
(3)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
N/A
N/A
N/A
Individual Risk to
Highly-Exposed
Subpopulations
(4)
N/A
N/A
N/A
N/A
N/A
hc,.^«neter
Uncertainty
N/A
N/A
Low
Low
Low
*** Draft - March 23,1993 ***
             11

-------
Parameter
Contact Rate
Exposure location
when swimming
Foodchaln Pathway
Maximum Distance
Modeled (from facility
boundary)
Exposure Duration
Exposure Frequency
Exposure
Rate
Leaf
Veg.
Root
Veg.
Beef
Dairy
Products
Individual Risk
Central Tendency
(1)
0.05 L/hour1
Exposure at recreational
use point to constituent
concentration in surface
water
Site-specific
9 years'
350 days/year0
65 g/day (based on
assumption that 25% of
consumption is from
contaminated source)*1'"1
46 g/day (25% from
contaminated source)""
44 g/day (44% from
contaminated source)"1"
160 g/day (40% from
contaminated source)"0
High-End
(2)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
Same as (1)
t
Same as (1)
Same as (1)
Same as (1)
Population Risk
(3)
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Individual Risk to
Highly-Exposed
Subpopulations
(4)
N/A
N/A
Same as (1)
30 years (90th
percentile residence
time) for subsistence
fishermen and 40
years for subsistence
farmers'
365 days/year1
103 g/day (40% from
contaminated
source)"1"1
74 g/day (40% from
contaminated
source)"1"1
75 g/day (75% from
contaminated
source)"1"
300 g/day (75% from
contaminated
source)"-0
Parameter
Uncertainty
Low
Low
N/A
Low
Low
Low
Low
Low
Low
***
    Draft
h 23,1993 ***

-------




Parameter




Exposure to Lead











Future Population


Fish














Growth

Individual Risk

Central Tendency
0)
7.6 g/day (20% from
contaminated source)''""

N/A











N/A

High-End
(2)
Same as (1)


N/A











N/A



Population Risk
(3)
N/A


N/A











County-level growth
rate'
Individual Risk to
Highly-Exposed
Subpopulations
W
99 g/day (75% from
contaminated
source)'*"
Risk to sensitive sub-
population (children)
assessed using Lead
UBK model; calculates
blood lead levels in
children 0-7 years in
age; assumes lead
intake from maternal
blood, inhalation,
water, soil, indoor
dust, and food
ingestion
N/A


Parameter
Uncertainty

Medium-
High

Low











High

*** Draft - March 23,1993 ***
            13

-------
Parameter
Individual Risk
Central Tendency
(1)
High-End
(2)
Population Risk
(3)
Individual Risk to
Highly-Exposed
Subpopulations
(4)
Parameter
Uncertainty
DOSE-RESPONSE ASSESSMENT
Slope Factors
RfDs
Based on upper 95%
confidence limit*
Based on most sensitive
endpoinf
Same as (1)
Same as (1)
Low-
High
Low-
High
*** Draft -
i 23,1993 ***

-------
Parameter
RISK CHARACTERIZATION
Cancer Risk
Non-Cancer Risk
Risk from Exposure to Lead
Addrtlvlty of Risk
Individual Risk
Central Tendency
(1)

Lifetime excess cancer
risk for average individual
Hazard index for average
individual
N/A
Risks added across
chemicals and exposure
routes within a given
pathway (e.g., ingestion
and dermal absorption
via swimming in surface
water pathway)
High-End
(2)
^
Lifetime excess cancer
risk for highly-exposed
individual
Hazard index for
highly-exposed
individual
N/A
Population Risk
(3)

Statistical number of
cancer cases in
population
Number of people
with hazard index > 1
N/A
Individual Risk to
Highly-Exposed
Subpopulations
(4)

Lifetime excess cancer
risk for individual of
highly-exposed
subpopulation
Hazard index for
individual of highly-
exposed
subpopulation
Comparison of
exposure
concentration for
sensitive
subpopulation with
threshold blood lead
level of 10
micrograms/deciliter
Same as (1)
Parameter
Uncertainty

N/A
N/A
N/A
N/A
N/A    Not Applicable



SOURCES:
(Note that the sources listed below may In turn refer to secondary documents as the original source for some of the parameter values.)

(a)      Memo from Gerard Lanlak (EPA/ERL-Athens) to Barnes Johnson (EPA/OSW/CABD), 'Review and Recommendations Related to Chemical Data Used In the Corrective Action Regulatory
        Impact Analysis (CA RIA)". December 1990.
                                                                *** Draft - March 23,1993 ***
                                                                               15

-------
(b)       OLM final in 51 FR 41084.100. November 13,1986.

(c)       MMSOIL model

(e)       Baes. C.F. Ill, Sharp. R.D., Sjoreen. P.L, and Herman, D.W. November 1984. TERRA: A Computer Code tor Simulating the Transport of Environmentally Released Radlonuclides through
         Agriculture.' Oak Ridge National Laboratory.  ORNL 5785.

         US Department of Energy September 1989. Chemical Data Bases for the Multimedia Environmental Pollutant Assessment System (MEPAS): Version 1. Draft for External Review.

         Topp, E.. Scheunert. I. Altar, A. and Korte. F.  1986. Factors affecting uptake of "C-labeled organic chemicals by plants from soil. Ecotoxteol. Environ. Serf. 11:219-228.

         Travis, C.C., and Arms, A.D. 1988. Bloconcentration of organlcs in beef, milk, and vegetation.  Environ. Scl. Techno).  22(3):271-274.

(f)       USEPA 1989. Risk Assessment Guidance for Suuerfund Volume I:  Human Health Evaluation Manual (Part Al. Office of Emergency and Remedial Response. EPA/540/1-89/002.

(g)       USEPA 1991. Human  Health Evaluation Manual. Supplemental Guidance: Standard Default Exposure Factors.  Office of Emergency and Remedial Response. OSWER Directive: 9285.6-
         03.

(h)       USEPA 1989. Exposure Factors Handbook. Office of Health and Environmental Assessment. EPA/600/8-89/043.

(I)       The general equation to calculate Intake via dermal absorption In the water pathway is:

         Intake    =       CW x CF x SA x PC x ET x EF x ED
                                            BWxAT
         where:

                  CW      =       Chemical concentration In water (mg/l)
                  CF      »       Conversion factor (1  liter/1000cm8)
                  SA      =       Skin surface area available for contact (cm'/day)
                                            (19,400 cm* for adults (central tendency, source (f)))
                 PC
                 ET
                 EF
                 ED
                 BW
                 AT
Chemical-specific dermal permeability constant
Exposure time (hours/day)
Exposure frequency (days/year)
Exposure duration (years)
Body weight (kg)
Averaging time (days)
(j)       Internal communication from ORD.

(k)       The general equation to calculate intake via dermal absorption In the soil pathway is:

         Intake    =        CS x CF x SA x AF x ABS x EF x ED
                                              BWxAT
         where:
                  CS      =        Chemical concentration in soil
                  CF      =        Conversion factor (10"* mg/kg)
                  SA      =        Skin surface area available for contact (cm'/day)
                                           (5,000 cm* for adults (central tendency, source (f)), and 2.500 cm1 for children (default value, source (d)))
                  AF      =        Soll-to-skln adherence factor (mg/cm1)
                                           (0.2 mg/cm9 (central tendency, source (f)))


                                                                     *** Draft - '        23,1993 ***

-------
                  ABS     =       Absorption factor (chemical specific constant)
                  EF       =       Exposure frequency (days/year)
                  ED       =       Exposure duration (years)
                  BW      o       Body weight (kg)
                  AT       =       Averaging time (days)

0)       The exposure frequency value for dermal absorption and ingestion of contaminated surface water while swimming was provided by ORD/OHEA. This assumes that an Individual swims 2
         times a week during the summer (3 months) only.

(m)      Exposure rote accounts for the fraction of total vegetables consumed that comes from the contaminated source. The proportion of contaminated vegetables Is assumed to be 25 percent
         for the general population (average, source (c)) and 40 percent for subsistence farmers (reasonable worst case, source (c)).  This Is used to adjust chronic daily intake, e.g., the final intake
         used to calculate risk would be (intake x 0.25) for the average population and (intake x 0.4) for subsistence farmers   For example, the 65 g/day exposure rate for contaminated leafy
         vegetables represents 25 percent of the total daily consumption of leafy vegetables of 260 g.

(n)       Exposure rate accounts for the fraction of total beef consumed that comes from the contaminated source. The proportion of contaminated beef Is assumed to be 44 percent for the general
         population (average, source (c)) and 75 percent for subsistence farmers (reasonable worst case, source (c)).

(o)       Exposure rate accounts for the fraction of total dairy products consumed that comes from the contaminated source   The proportion of contaminated dairy products Is assumed to be 40
         percent for the general population (average, source (c)) and 75 percent for subsistence farmers (reasonable worst case, source (c)).

(p)       USEPA1991.  Internal EPA Memorandum dated August 19.1991.

(q)       The general equation Is modified for the fish Ingestion pathway to Include fraction from contaminated source.*  This parameter Is Included as a multiplicand In the numerator of the equation
         and Is abbreviated "Fl."  Fl was estimated as 20 percent for the general population (average, source (c)) and 75 percent for subsistence fishermen (reasonable worst case, source (c)).

(r)       County-level population growth projections from Woods and Poole Economics. Inc.

(s)       The slope factors and reference doses are taken from EPA's Integrated Risk  Information System (IRIS), or. if not listed In IRIS, from EPA's Health Effects Assessment Summary Tables
         (HEAST. March 1992).
                                                                      *** Draft - March 23,1993 ***
                                                                                       17

-------
APPENDIX H: FACILITY AND SWMU DATA FORMS

-------
FACILITY DATA FORM

-------
                                           FACILITY DATA FORM
       A. TRACKING DATA

       1.  Enter the following information:
PERSON COMPLETING DATA FORM (FULL NAME)



SECTIONS COMPLETED



       2.   Person entering data to the PC data base (full name):

M      3.   Date of the topographic maps used:  	

       B.  GENERAL

       1.   ICF sample number:     	
       2.  EPA facility Identification number:

       3.  Facility name:       	

           Facility address:     	
           Facility city, state, and zip code:
           Facility county:     	
       4.   If applicable,

           4a. Federal facility identification number:
           4b Federal agency and/or department: (e.g., DOE, BLM, etc.):
       5.   SIC code of current industrial activity, if available (use 4 digits even if only 2 or 3 digits are
           available, i.e., fill in appropriately with zeros):
           5a.    Primary:

           SOURCE:
5b.    Secondary:
       6.   If applicable, describe past industrial activity and period of activity, use SIC code if available (use
           4 digits even if only 2 or 3 digits are available, i.e., fill in appropriately with zeros):
SIC
CODE


DESCRIPTION OF ACTIVITY
(IF SIC CODE IS UNAVAILABLE)


TIME PERIOD
(YEAR ACTIVITY STARTED TO YEAR ENDED)


           SOURCE:

-------
                                                   - 2 -                           Facility:	

       7.  Number of employees at the facility (most recent estimate):   	

           SOURCE:  	

       8.  Year industrial activity (i.e., not necessarily waste management activity) began at facility:	

           SOURCE:

       9.  Year Industrial activity ceased at facility (enter 9999 If still ongoing):  	

           SOURCE:  	

       11. Describe the statutory or regulatory authority under which environmental assessments, remedial
           investigations, or corrective actions have occurred (e.g., RCRA 3004(u)):	
           SOURCE:
       12. Permit status:
           	  01  Permitted, operating facility (currently managing RCRA-regulated hazardous waste)
           	  02  Operating facility seeking permit (interim status)
           	  03  Closing facility that needs a post-closure permit
           	  04  Closing facility that does not need a post-closure permit
           	  05  Closed facility with post-closure permit
           	  06  Other, please specify	j	
           SOURCE:
       C. PHYSICAL DATA

M      1.  Facility longitude:   	  Facility latitude:

           SOURCE:
       2.  Facility dimensions:      Length:      	 Width:

           UNITS:     FT, YD, Ml, MT, KM
           SOURCE:
       3.  Total area of the facility:
           UNITS:     AC, HA, SY, SM, M2
           SOURCE:
M      4.  Facility slope (i.e., average grade):

           SOURCE:

-------
                                                   -3-
                                                                            Facility._
M
M
H/M
D. SOIL-SPECIFIC INFORMATION

Variables for Infiltration. Leaching, and Recharge Data File

1.  Field capacity of surface soil (vol/vol):       	

    SOURCE:
2.   Wilting point of surface soil (vol/vol):

    SOURCE:
3.   Number of subsurface soil/rock layers found at the facility
    (i.e., number of distinct soil/rock horizons In the unsaturated zone):

    SOURCE:
H/M   4.   For each soil/rock layer identified above, specify the following:
LAYER
Subsurface Layer 1
Subsurface Layer 2
Subsurface Layer 3
Subsurface Layer 4
Subsurface Layer 5
Subsurface Layer 6
Subsurface Layer 7
UNITS:
Source:
DEPTH OF SOIL
LAYER (length)







CM, IN, FT, MT

TEXTURAL
CLASSIFICATION1









PERCENT ORGANIC MATTER
MIDPOINT (CENTRAL
TENDENCY)

.







LOW (HIGH END)









           1 Select appropriate USDA textural classification:
                            6. sandy clay loam   11. clay
    1. sand
    2. loamy sand
    3. sandy loam
    4. loam

    5. silt loam
                          16. limestone
                            7. clay loam
                            8. silty clay loam
                            9. sandy clay

                            10. silty clay
12. sand & gravel           17. karst limestone
13. sand & gravel w/sig. clay 18. basalt layers
14. sandstone               19. hydraulically connected
                             basalt
15. shale                   20. fractured igneous &
                             metamorphic
                           21. glacial till

-------
                                                 -4 -
                                                                         Facility:_
M
M
M
Variables for Surface Water Pathway Data File

5.  Obtain the following erosion parameters:
           USLE rainfall factor (tonnes-m/ha-hr):
           USLE cover factor:
           SOURCE:
6.  SCS curve number for surface soil:

    SOURCE:      	
E. METEOROLOGICAL INFORMATION

Variables for Infiltration. Leaching, and Recharge Data File

1.  Complete the following meteorological parameters:
MONTH
JANUARY
FEBRUARY
MARCH
APRIL
MAY
JUNE
JULY
AUGUST
SEPTEMBER
OCTOBER
NOVEMBER
DECEMBER
ANNUAL
AVERAGE
AVERAGE MONTHLY
PRECIPITATION
(length)



-









NO. DAYS WITH
PRECIPITATION













AVERAGE
TEMP.

.











AVERAGE MONTHLY
PAN EVAPORATION
(length)













         Units:

         Source:
                 CM, IN, FT
DC. DF     CM, IN, FT

-------
                                                   - 5 -                             Facility:	

       7.  Average wind velocity (length/time):  	  Units: MS, FS, KH, MH

           Source:_	


       F. LAND USE INFORMATION

M     1.  Identify percentage of each of the following land uses or land covers within the facility boundaries
           (total =100%):

           1 a.   Pasture:                               	   	Specify level of grazing
           1 b.   Meadow:                              	           (high, moderate, or low)
           1 c.   Woodland-                            	   	Specify level of grazing
           1 d.   Cropland:                             	           (high, moderate, or low)
           1 e.   Bare soil or dirt roads:                  	
           1f.   Hard-surface roads:                    	

           1 g   Other paved surfaces:                  	
                (including buildings)
           1h.   Water (e.g , ponds,  lakes,
                surface impoundments, wetlands):       	
           11.   Fraction of other vegetative cover:       	
           SOURCE:
M     2   Indicate percent of facility that is located in each of these terrains (total=100%):

           2a.   In 100-year floodplain:            	
           2b   Above floodplain in river bottom:   	
           2c.   All other above floodplain:        	
           SOURCE:
M      3.   Indicate types of environmental settings at facility:

           	01      Seismic impact areas             	02  Karst terrain
           	03      Floodplains                      	04  Poor foundation conditions
           	05      Complex hydrogeology           	06  Areas susceptible to mass movements
           	07      Ground-water vulnerability/resource value

           SOURCE.

-------
                                                    -6-
                                                                            Facihty
M
M
M



M

M
4.  Indicate types of sensitive or valuable environments in vicinity of facility (enter zero for distance, if on-r'
    adjacent)-

01
02
03
04
05
06
07
	 08
TYPE OF SENSITIVE OR VALUABLE ENVIRONMENT
Wetlands
National Park
Areas identified under Coastal Zone Management Act
Scenic areas identified under National Estuary Program
Critical areas identified under Clean Lakes Program
National Seashore Recreational Area
National Lakeshore Recreational Area
National Forest
DISTANCE FROM FACILITY








           UNITS:      FT, YD, Ml, MT, KM
           SOURCE:
5.  Land uses within 1  and 10 miles of the facility boundary [Enter 0 for percent of land area if the percent is
    unknown but the land use is known to occur]:

                                      Percent of Land Area    Nearest Distance
                                  Within  1 Mile   Within 10 miles-'   to Facility

    5a.   Industrial
    5b.   Commercial
    5c.   Residential
    5d.   Agricultural
    5e.   Forest/Field
    5f.    Elementary School
    5g.   Other School
    5h.   Wetlands
    5i.    Park
    5j.    Surface Water
    5k.   Other, specify:	
            UNITS:     FT, YD, Ml, MT, KM
            SOURCE:           	
7.  Based on land use, surface roughness height (length):

    SOURCE:	_^
UNITS:    CM, IN, FT
8.  Starting month for growing season:

9.  Ending month for growing season:

    SOURCE (for 8 and 9):   	

-------
                                                 - 7 -                           Facility:_

       G.  SURFACE WATER INFORMATION

       2.  Receiving water bodies (check one, both, or none, as applicable):

          01	River, stream, creek, etc.            03	Ocean, coastal area, etc.
          02	Lake, pond, etc.                    04	None
          SOURCE-
          If there are no receiving water bodies, skip to Section H (page 8)

H      3.  Does ground water underlying the facility discharge to surface water?

          01       Yes   02      No
          SOURCE:
       Stream/River Parameters

H      4.  Name of the nearest streams/rivers downgradient that are likely to receive contamination from facility:
                (a)	
                (b)	
          SOURCE.
          For the streams/rivers Identified above (Question G.4), give the following Information:

       5.  Distance to the streams/rivers from the facility boundary (length):   (a)	
                                                                        (b)	
          UNITS:     FT, YD, MT, Ml, KM
          SOURCE:
M      6.  Average flow rate in streams/rivers (vol/time):

                     central tendency          high end
          (a)        	          	
        -  (b)        	          	
           UNITS:     QM, OF, TS, CS, GP
           SOURCE:       	
M     7.   Average stream/river velocity (length/time):                             (a)
                                                                              (b).

           UNITS:      MY, FY, MS, FS

           SOURCE:	

-------
                                                  - 8 -                           Facility:	
 M     7a. If applicable, provide information on the water bodies that either of the streams (identified in G.4»
           into:
                 Name of the water body (either 4a or 4b) flowing into
                 another water body:
                 Name of the receiving water body within 15 miles:
                 Distance to point of confluence (within 15 miles):	  UNITS: MT, FT, Ml
           SOURCE:	
       Lake or Coastal Water Body Parameters
H     8.  Name of the nearest lake or coastal water body downgradient that is likely to receive contamination from
           facility:	
           SOURCE:
           For the lake or coastal water body Identified above (Question G.8), give the following information:
H     9.  Distance to the nearest lake or coastal water body from facility boundary (length):	
           UNITS:     MT, YD. FT, Ml, KM
           SOURCE:   _^	
H/M   10. Surface area  (only for lakes):      	   UNITS: AC, HA, SY, SM, M2
           SOURCE:
H/M   11. Average water depth (length):       	   UNITS:      CM, IN, FT, MT, YD
           SOURCE:
M     12. Sediment dilution ratios (0 to 1) :
           Central tendency:	 High end:	
           SOURCE:

       H. GROUND WATER INFORMATION
H     2.   Upper aquifer  material:       	
           SOURCE:
H     3.   Regional hydraulic gradient (length/length):      	
           UNITS:
           SOURCE:
H      4.   Hydraulic conductivity of upper aquifer (length/time):     	   UNITS:MY, FY, FS
           SOURCE:

-------
                                                 - 9 -                           Facility:	

M     5.  Porosity of upper aquifer material (vol/vol):      	

           SOURCE:	

H     6.  Bulk density of upper aquifer material (mass/vol):      	    UNITS:    GC, PF

           SOURCE:

H     7.  Fraction of organic carbon in upper aquifer material (model default is 0.001):     	

           SOURCE:

H     8.  Saturated thickness of upper aquifer (length):      	   UNITS:    MT, FT, YD

           SOURCE:	

H/M   9.  pH of ground water:	 SOURCE:	
M     14.  If there are ground-water withdrawal wells (i.e., not monitoring wells or boreholes) on-site, please give the
           following information:

           14a.  Number of wells:                            	

           14b.  Number of people served:                    	
           14c.  Average well depth (length):                  	UNITS:    MT, FT, YD

           14e.  Total pumping rate (vol/time):                 	UNITS:    OF, QM, TS, CS, GP

           14f.  Use of water from wells:                      	

          SOURCE:

H      15. Information on lower aquifer

          15a.  Is there a lower aquifer that is used as a drinking water source?
                	01 Yes      	02 No       03 Unknown
          SOURCE:
          15b.  If 15a. is Yes. depth from ground surface to top of lower aquifer (length):

          UNITS:     MT, FT, YD
          SOURCE:	

          15c.  If 15a is Yes, is upper aquifer hydrauhcally connected with the lower aquifer?
                	01 Yes      	02 No   	03 Unknown

          SOURCE:
                                                ******

-------
                                                    - 10-
Facihty
       17. If there are any upgradient monitoring wells that show ground-water contamination, list the
           contaminants and their concentrations below (i.e., list background levels):
CHEMICAL NAME






CAS NUMBER (IF AVAILABLE)






CONCENTRATION






UNITS
ML, GL, PP, PB, PG
ML, GL, PP, PB, PG
ML, GL, PP. PB, PG
ML, GL, PP. PB, PG
ML, GL, PP, PB, PG
ML, GL, PP, PB, PG
           SOURCE:
       I.  EXPOSURE INFORMATION

M     1.  Distance from facility boundary to the location of nearest person (Maximum Exposed Individual (MEI))
           potentially exposed to air releases (i.e., distance from facility boundary to nearest residential point):
           	N         	E      	S        	W

           UNITS:      FT, YD, Ml, MT, KM
           SOURCE:   	

M     2.  Size and location of population potentially subject to air exposure (within 50 km of facility).  Enter population
           for each distance and direction combination:
Direction
NtoNE
NEtoE
EtoSE
SEtoS
StoSW
SWtoW
WtoNW
NWtoN
Distance Categories (in kilometers from facility boundary)
Oto.5








> .5 to 1








> 1 to 2








> 2 to 5








> 5 to 10








       SOURCE:

-------
M
M
M
M
                                                   - 11 -

       5.  Ground-water flow and potential exposures

           5a.    Direction of ground-water flow beneath facility:

           SOURCE:
                                                                             Facility:_
    5b.    Average well screening depth with respect to surface of upper aquifer for

    (i)     residential use wells:        	
    (ii)    public use wells.            	
    (hi)    agricultural use wells:       	
           UNITS: MT, FT     SOURCE:
    5d.    Location of the nearest downgradient residential use well (public or private) in the direction of
          groundwater flow as identified in I.Sa for MEI calculations:
          Direction:	      Distance (length):	   UNITS: FT, Ml, MT
           SOURCE:
    5e.    Private wells within 2 miles of facility potentially subject to ground-water exposure.  Enter number of
          private wells / number of residents served by private wells (e.g., 3 / 8) in the directions downgradient
          from the facility.  Specify direction of ground water flow, which is assumed to be the centerhne of a
          ground-water contamination plume.  (The plume will usually be assumed to spread to 45 degrees, but
          may be broader in some settings.)
Direction
Second 22.5 °
sector to left of
plume centerhne
First 22.5 ° sector
to left of plume
centerhne
First 22.5 ° sector
to right of plume
centerline
Second 22.5 °
sector to right of
plume centerhne
Distance Categories (in miles and fractions of miles from facility boundary)
0 to .25




> .25 to .5




> .5 to .75




> .75 to 1




> 1 to 1.5




> 1.5 to 2




5e1.      Population served by private wells located near the facility boundary:

           0 -  50 meters from the facility boundary:      	
          50 -150 meters from the facility boundary:      	
          150 - 400 meters from the facility boundary:     	
       SOURCE:

-------
                                                     - 12 -                             Facility:	

M      5f.  Distance to and population served by public wells (within 2 mile of the facility) located in the ground'
            flow direction (i.e., within 45° on either side of centerlme of groundwater flow direction) :

                  Direction from Facility        Distance from Facility       Population Served
                  a.
                  b.
                  c.
                  d.
                  e.
                                             UNITS:       Ml, KM, FT, YD, MT
            SOURCE:
M     3.   Nearest agricultural (crop and cattle) field that Is likely to receive maximum (considering predominant
            wind direction and topography) contamination from the site:

            3a. Direction from facility:
           3b. Distance from facility:  	   UNITS:     FT, YD, Ml, MT, KM

           3c. Area of field:          	   UNITS:     AC, HA, SY, SM, M2

           3d. Type of crop:         	
           3e. Sediment delivery fraction values for
                 Central tendency:	         High end:	
           3f. Fraction of organic carbon in surface soil
                 Central tendency:	         High end:
           3g. Bulk density of surface soil:     	      UNITS:  GC, PF

           3h. Depth of irrigation:             	     UNITS:  MY, FY, IY, CR

           3i. Source of irrigation water:
                 	01 surface water-river  	02 surface water-lake  	03 ground water

           3j. Source of cattle water:
                 	01 surface water-river  	02 surface water-lake  	03 ground water
           SOURCE:
M     5c. If the source of irrigation for the agricultural field in I.3 is groundwater, and if this well is located in the
           qroundwater flow direction, give the location of nearest downgradient agricultural use well:

           Direction:
           Distance from facility (length):	 UNITS: MT, FT, YD, Ml, KM

           SOURCE:

-------
                                                  - 13-
                                                                            Facihty:_
       4.   Nearest off-site field for human exposure via direct contact with soil (e.g., playground, garden):
           4a. Direction from facility:
           4b. Distance from facility:
           4c Area of field:
           4d. Sediment delivery fraction values for
                Central tendency:	        High end:
                                            UNITS:   FT, YD, Ml, MT, KM
                                            UNITS:   AC, HA. SY, SM, M2
           SOURCE:
       6.   Potential surface water exposures

M         6a.   Public drinking water supply intakes located within 15 miles downstream of facility on water bodies
                identified in G.4:
                Distance from Facility
                                    Water Body
Population Served
M
                in.
                IV.
           UNITS:     Ml, KM, FY, YD, MT
           SOURCE:
    6b   Other water uses within 15 miles downstream of the facility on water bodies identified in G 4 (e.g.,
         agricultural uses, swimming, and fishing):                •
                Distance from Facility
                                    Water Body
Water Use
                M.
                in.
                IV.
           Units:       Ml, KM, FT, YD, MT
        ..  SOURCE:
M     7.   Number of residences within 1/4 mile of facility:

           SOURCE:
M
M
J.  FOOD CHAIN PATHWAY

2.   Pasture production (mass/area):

    SOURCE:
UNITS:  KS, KA, PA, PM
3.   Vegetable production (mass/area):

    SOURCE.
UNITS:  KS, KA, PA, PM

-------
                                                  -14 -                           Facility,

M      4.  Is there any evidence of:

           4a.    Cattle grazing in the nearby agricultural field (Yes/No):
           4b.    Human consumption of vegetables produced in the nearby agricultural
                 field (Yes/No):	


       K.  RELEASE INFORMATION

       Identify the types of chemical contaminants found at the facility (e.g., from monitoring or sampling data). Similar
       questions are asked about each SWMU at the facility in the SWMU data form.  If you are unable to answer the
       chemical-specific questions at the SWMU-level, please describe the contaminants found at the facility in this
       section of the facility data form.

       1.   Have there been any releases or contamination noted at the facility in the past?
           (Yes/Suspected/No/Unknown):	

           SOURCE:   	

       2.   Describe facility-wide release information:	
       L. FACILITY-WIDE SWMU INFORMATION

       1.  Number of solid waste management units or areas of concern identified at facility:
           	Regulated RCRA Subpart F land disposal SWMUs (i.e.. landfill, surface impoundment, waste p
                 land treatment unit that managed hazardous waste after 7/26/82)

           	Other SWMUs (e.g., solid waste landfills, other hazardous waste management units requiring permit or
                 exempt)
           	Areas of concern (AOC)
           SOURCE:

-------
                                             -15-
2.   Complete the following table for the SWMUs identified:
                       Facility:
                                                                               NUMBER WITH MIXED HAZARDOUS
                                                                               AND RADIOACTIVE WASTES
NUMBER UNDER CERCLA
AUTHORITY
     Land Treatment
     Treatment Surface Impoundments
     Storage Surface Impoundments
     Unspecified Surface Impoundments
     Treatment Tanks
     Unspecified Tanks
     Waste Transfer Stations
     Waste Recycling Operations
     Routine and Systematic Spill Areas
     Other Spill Areas
     Accumulation Areas
     Process Sewers
     Areas of Concern
    SOURCE:
PLEASE COMPLETE OWE SWMU DATA FORM FOR EACH SWMU IDENTIFIED ABOVE.

M. COMMENTS AND FURTHER FACILITY DESCRIPTION (Use this page -- and attach others as necessary -
to present comments and facility description information that might be relevant for the corrective action RIA but
was not included elsewhere in this form.)

-------
SWMU DATA FORM

-------
                                     SWMU DATA FORM
A. TRACKING DATA

2.   Enter the following information:
Person Completing Data Form




Sections Completed




Person entering data from this form to the PC data base:

3.    Is the location of this SWMU identified on available site maps (Y/N):

B. SWMU IDENTIFICATION

1.    ICF number:  	

2.    Facility name: 	
3.   SWMU type:
     	01 Landfill
     	02 Land Treatment
     	03 Waste Pile
     	04 Treatment Surface Impoundment
     	05 Storage Surface Impoundment
     	06 Unspecified Surface Impoundment
     	18 Other SWMU. specify  	
	07 Treatment Tank
	08 Storage Tank
	09 Unspecified Tank
	10 Incinerator
	11 Injection Well
   12 Waste Transfer St
13  Waste Recycling Operation
14  Routine & Systematic Spill Area
15  Other Spill Area
16  Accumulation Area
17  Process Sewer
19  Areas of Concern
     SOURCE:
4.   Sequential SWMU number (start w/ 001 for each facility): 	

5.   SWMU name (e.g., pond #1): 	

C. GENERAL DESCRIPTION

1.   SWMU Regulatory Status (Check one):

     	01   Regulated RCRA Subpart F land disposal SWMU (i.e., landfill, surface impoundment, waste
               pile, or land treatment unit that managed hazardous waste after 7/26/82)
          02   Other SWMU
     SOURCE:
1a.  Is the SWMU subject to other federal cleanup authorities?

     	01  Yes    	02  No

1 b.  If yes, what authorities could apply?
           01  CERCLA (e.g., whole facility or part of
            operable unit is on NPL)
           02  RCRA Subtitle I (product storage tanks)
           03  TSCA (PCB spills)
                    04 CWA (Discharges to surface water)

                    05 AEA (radionuchdes)
                    06 Other:	

-------
                                            - 2 -                              Facility
                                                                              SWMU:
 1 c.  Is the SWMU subject to state cleanup authorities?

     	  01   Yes, enter name of applicable state authorities:
           02   No
     SOURCE (1a, 1b, and 1c):_
2.   Year SWMU was first used:

     SOURCE-
3.   Operating status of SWMU:

     	01  Currently in service (skip to Question 7)
     	02  Currently not in service (go to next question)
     SOURCE:
4.   Complete one of the following If the SWMU Is currently not in service:

          Year waste is no longer added to the SWMU:
          SWMU is temporarily not in service. What year is it expected to resume use?

     SOURCE:	

5.    If SWMU has undergone closure, how was unit closed?

     	01 Unit dismantled and removed          	05 No closure measures taken
     	02 Excavation and decontamination        	06 Impoundment closed with waste in place
     	03 Containment system installed          	07 Other - Describe 	
        _04 Unit capped    .                  	08 Unknown or unavailable
     SOURCE:
7.   Complete the following SWMU dimensions:
|| Area (length * length)
Central tendency
High End
UNITS


HA. AC. M2, SY. SF
Depth (length)


FT. YD, MT, IN, CM
     SOURCE:
8.   Is this a SWMU of concern?

     	  Yes - Continue filling out the form
     	  No  - Stop here, do not complete remainder of the form (e.g., do not include Subpart F units)


                             Continue only If this Is a SWMU of concern.

-------
                                            -3-
D. KNOWN. SUSPECTED. OR POTENTIAL RELEASES
                                                                         Facility.
                                                                         SWMU
I.Has the SWMU released any wastes/constituents? Specify one of the following for each SWMU:
Yes/Suspected/None Documented/No/Unknown
3.
         Ground Water
         Soil
         Other, specify
                             Surface Water
                             Air
     SOURCE.
Were corrective actions taken for these releases7
 	01 Yes    	02 No     	03 Unknown

 If so,

3a.  What year did actions start?

     	01  pre-1984   	02  post-1984     	03  Unknown
     3b.  What media? (check all that apply)
         	01 Soil  	02 Ground water
         	04 Air   	05 Other

     SOURCE:
                                         03 Surface water
                                         06 Unknown
5    What activities has an RFA or similar site investigation suggested for this SWMU? (Check all that apply.)

     	 01   No further action
     	 02   Additional investigation to determine appropriate action.  Describe recommendations*

     	 03   Specific remedial action.  Describe recommended remedial activities, including interim
              measures:  	

     SOURCE:
6    Likelihood of future releases of concern.  Based on information presented in the RFA or other
     documentation, indicate the relative potential for future releases to each medium with one of the following
     codes-
     01 - Release Expected (High)
     02 - Release Probable (Medium)
     03 - Release Unlikely (Low)
                              04 - Release Not Possible
                              05 - Likelihood Unknown
MEDIUM
Ground Water
Surface Water
Soil
Air
POTENTIAL FOR
RELEASE




BASIS/SOURCE





-------
                                           -4 -
Facility:
SWMU.
   This data form Is now complete and you do not need to proceed to the remaining sections. Fill V-
                   following sections only for the SWMUs that have to be modeled.
E. WASTE TYPE AND QUANTITY

1.   List of waste types/codes and quantities contained by the unit:
WASTE TYPE/CODES





QUANTITY (MASS OR VOLUME)




MASS LB. TN. TM. KG
VOLUME ON. LT, CF. M3, CY
2.   At this SWMU (landfills, waste piles, and land treatment units) are air emissions due to:

     vehicle paniculate disturbance possible? (Y/N):	
     spreading or material movement operations possible? (Y/N):

     SOURCE:
     Complete the following table with constituent concentration and physical state of wastes in the
     SWMU:
CONSTITUENT
NAME










CAS NUMBER










CONCENTRATION (mg/kg for solid or
sludge, mg/l for liquid)
CENTRAL TENDENCY










HIGH END










PHYSICAL STATE
(L=hquid, S=Sohd/
Sludge)










SOURCE:

-------
                                          - 5 -                             Facility:	
                                                                          SWMU:	

F. LOCATION OF SWMU

1.   Distance from edge of SWMU to facility boundary:

    1a  in direction of ground-water flow (length): 	

    1 b. in direction of surface water runoff (if river is on-site, measure the
        distance up to river boundary) (length): 	

    1c. in 4 compass directions (length):

         	N          	E            	S                  	W

    UNITS:   FT, YD, MT, Ml, KM

    SOURCE:                                                            	
G. MODELING INFORMATION

3.   Fraction of contaminated area with vegetative cover:

     SOURCE:
7.    Depth of clean cover (i.e., uncontamlnated topsoil or other material):

     UNITS:   FT, YD, MT, IN, CM
     SOURCE:
9. How will this unit be modeled? (Fill out sections noted for each category below)

    	01   Landfill -  Section H
    	02  Surface Impoundment - Section H
    	04  Injection Well - Section J
    	05  Tanks - Section K

    SOURCE:	

10.  Relationship to other SWMUs:

    10a. Wastes/wastewaters removed/discharged from this SWMU to SWMU #.
    10b. Wastes/wastewaters received by this SWMU from SWMU #:
     10c. Other (e.g., describe groups of SWMUs, similar or identical SWMUs in parallel waste management
         trains):	


     SOURCE:
     If the SWMU is a landfill or a surface impoundment or a surface impoundment converted to a
landfill, complete Section H; Injection well - Section J; tank • Section K. (Section I and L are blank.)

-------
                                          -6-
                                                                     Facility.
                                                                     SWMU-
H. LANDFILL/SURFACE IMPOUNDMENT fSIl
1.
2.
Landfill/Si cover type:
	00    Uncovered
	01    Vegetative
SOURCE:
                                        	02 Clay Cover
                                        	03 RCRACap
Landfill/Si liner type:
	01    Unlined
	02    Clay Liner
	03    Single Synthetic
SOURCE:
                                        	04 Double Clay/Synthetic
                                        	05 MTR Double Composite/Synthetic
3.    Characterize each cover/liner component:

Cover
Liner
Layers
Vegetation (including top soil)
Drainage
Synthetic
Clay
Drainage - 1
Drainage - 2
Synthetic -1
Synthetic -2
Clay
Thickness (length)






,


                                 Thickness Units:
     SOURCE:
                                                 FT, YD, MT, IN, CM
4.    Describe integrity of cover, liner, and other containment features:
     SOURCE:
If this is a surface Impoundment complete the following:

10.  If this is a landfill converted from a surface Impoundment,
     year in which It was converted (enter '9999' If still a SI): _
     SOURCE:
13.  Monthly influx to pond (vol):

     SOURCE:
                                          UNITS: M3, GN, LT, CF, CY
                                         ******

-------
                                           - 7 -                              Facility:
                                                                            SWMU
J. INJECTION WELL

1.    Difference between water table and injection water column (length):

     UNITS:    MT, FT, YD
     SOURCE:
2.    Injection rate (vol/time):

     UNITS:    OF, QM, TS, CS
     SOURCE-
3.    Pump head (length):

     UNITS:    MT, PT, YD
     SOURCE-
4.   Type of injection well.
           01 Water flooding (secondary recovery)
           02 Disposal
    SOURCE:
5.   Type of well failure expected:
    SOURCE:
            01 Grout failure
            02 Well casing failure
            03 Conservative approach between grout and well casing failures

-------
                                          - 8 -                            Facility
                                                                         SWMU:
K. TANKS

1.   Tank placement:
     SOURCE:
            01 Aboveground
            02 Surface
            03 Underground
2.    Construction material:
            01 Carbon
            02 Steel
            03 Concrete
            04 Other, specify:
     SOURCE:
3.    Describe construction and integrity of containment features (indicate if no containment features
     are present):	
     SOURCE:
4.   Tank volume:
         Central Tendency:
         High End.

    UNITS:  GN, M3, LT

    SOURCE:
5.    Tank Code:

     SOURCE:
                                         ******

-------
                                          - 9 -                             Facility	
                                                                          SWMU:	

M. COMMENTS AND FURTHER SWMU DESCRIPTION  (Use this space -- and attach other pages
as necessary - to present comments and SWMU description information that might be relevant for the
corrective action RIA but was not included elsewhere in this form.)
                                         ******

-------
       APPENDIX I.  CHARACTERISTICS OF FACILITY AND SWMU POPULATIONS


        This appendix  describes the population of facilities and solid waste management units
 (SWMUs) in the United States subject to corrective action requirements. The stratified random
 sampling methodology EPA used to select the Corrective Action Regulatory Impact Analysis
 (RIA) sample facilities allows the Agency to extrapolate from data collected on these sample
 facilities and characterize the national population of facilities and SWMUs subject to corrective
 action. The national data presented here are  extrapolated from sample facility data using
 weights derived from the RIA's stratified sampling procedure.

        EPA used the  data forms shown  in Appendix H to collect detailed information on all of
 the facilities and SWMUs in the RIA sample. Section I.I presents characteristics of facilities.
 Facility data include general characteristics, size, environmental setting, information about
 surface and ground water potentially affected  by the facilities, and descriptions of populations
 potentially exposed to releases from facilities. Section 1.2 presents SWMU characteristics,
 including numbers and types of SWMUs at facilities and the characteristics of wastes managed in
 the SWMUs.  Throughout the appendix, EPA's characterization of the regulated population is
 based on extrapolation from the sample, rather than compilation of data for the entire
 population.

        For facilities considered most likely to trigger corrective action, the Agency collected
 some information not  collected for other facilities.  For example, EPA did not collect detailed
 information on environmental and waste characteristics for facilities that were not likely to
 require remediation.  In these cases, the exhibits  presented in this section describe the population
 of facilities  likely to trigger  corrective action, rather than the full population of facilities subject
 to corrective action requirements. The data presented in sections 1.1.1,1.1.2, and 1.2.1 apply to
 the full population of facilities  subject to corrective action (N = 5,800), while those presented in
 sections 1.1.3 through 1.1.7 and 1.2.2 apply only to facilities likely to require remediation (N  =
 2,600).


 I.I  Facility Characteristics

       1.1.1  General Characteristics

       As described in Chapter 3 and Appendix A, the population of facilities subject to RCRA
 corrective action authorities consists of 359 federal facilities and about 5,400 non-federal
 facilities, for a total of approximately 5,800 facilities. The nine very large federal facilities in the
 population were excluded from the RIA sample because of their size and complexity, and
 because better data on the facilities are available  from the federal agencies. The population
 characterized in this analysis includes the remaining facilities in the population, 5,400 non-federal
 and 350 federal. Exhibit 1-1 shows the distribution of facilities by EPA Region, derived by
 extrapolating from the sample to the population.

       EPA also collected data on the current industrial  activity of facilities by SIC code.
 Exhibit 1-2 shows the numbers of facilities in different industries at the two-digit SIC code level.
The data show that the largest number of facilities are  in the chemicals  and allied products

                              ••* DRAFT - March 23, 1993 •••

-------
                                                          Exhibit 1-1
                                                Facilities by EPA  Region
                                                            (N = 5,800)
                                                                                                               Region 1
                                                                                                               290 facilities
                                                                                                               5.1%
                                                                                                        Region 2 also includes
                                                                                                        Puerto Rico and the Virgin Islands

                                                                                                       Region 3
                                                                                                       1,300 facilities
                                                                                                       22%
Region 9
Region 9 also includes Guam and American Samoa

-------
                               1-3

                          EXHIBIT 1-2
              INDUSTRIAL ACTIVITY AT FACILITIES
                            (N = 5,800)

SIC Code
28
49
38
36

29
97
24
33
34
20
35
87
42
82
30

51

37
TOTAL

Industry
Chemicals and Allied Products
Electric, Gas and Sanitary Services
Instruments and Related Products
Electronic and Other Electrical
Equipment
Petroleum and Coal Products
National Security and
International Affairs (Federal)
Lumber and Wood Products
Primary Metal Industries
Fabricated Metal Products
Food and Kindred Products
Industrial Machinery and
.Equipment
Engineering and Management
Services
Trucking and Warehousing
Educational Services
Rubber and Miscellaneous Plastics
Products
Wholesale Trade-Nondurable
Goods
Transportation Equipment

Number of
Facilities
1,100
690
640

560
360
350
350
340
340
-210
210
210
130
130

67

63
6.7
5,800
Percent of
Facilities
19
12
11

9.7
6.2
6.0
6.0
5.9
5.9
3.6
3.6
3.6
2.2
2.2

1.2

1.1
0.12
100
Totals may not sum due to rounding.
                 ***
                     DRAFT - March 23, 1993 ***

-------
                                             1-4

industry (SIC Code 28), with significant numbers also in electric, gas and sanitary services (SIC
Code 49), instruments and related products (SIC Code 38), and electronic and other electric
equipment (SIC Code 36).

       The Agency's investigation also provided data on the operating status of facilities,
including the years that industrial activity began and ended at each facility, and whether
industrial activity is still ongoing. Exhibit 1-3 shows the distribution of start dates of industrial
activity for all facilities subject to corrective action.  The oldest facility opened in 1775; the
newest is a "future facility" that has yet to be  built.  Construction was heaviest between 1961 and
1980, when over 100  facilities per year, on average,  began activity.  It was slowest in the period
including the Depression (1921 to 1940), when only about 13 facilities opened annually. The
Agency's estimates show that the majority of facilities are currently active (about 4,300). There
are an estimated 1,200 inactive facilities, and the status of 300 facilities is unknown.

       Subtitle C of  RCRA requires owners or operators of hazardous waste treatment, storage
and disposal facilities (TSDFs) to obtain operating permits.  Facilities built before  November 19,
1980 are allowed to operate under interim status  until EPA makes a permit decision.  Several
kinds of permits exist.  The most common for operating facilities are treatment,  storage and
disposal permits.  Closing facilities which leave wastes in place must obtain a post-closure  permit
describing the necessary post-closure care.  Exhibit  1-4 shows the numbers of facilities in several
categories of permit  status, for all facilities subject to corrective action.

       1.1.2  Physical Data

       EPA's data collection effort determined the dimensions of facilities and calculated their
total areas.   Exhibit 1-5 shows the distribution of facility areas  for all facilities subject  to
corrective action requirements.  The smallest facility is estimated to cover 0.076 acres, the largest
is 8,900 acres, the mean size is 330 acres, and the median size is 20 acres. Almost 30 percent of
facilities are between 10 and  100 acres.

       1.1.3  Soil  Data

       For facilities  likely to trigger corrective action, EPA collected information on surface soil
and on up to seven soil and rock layers below the surface soil.  These subsurface soil/rock layers
constitute the unsaturated zone, where spaces between soil particles contain both air and water
and water cannot be pumped out. For facilities expected to require  remediation, Exhibit 1-6
shows the thickness of the unsaturated zone, summed across all subsurface layers.  For the large
majority of facilities, the unsaturated zone is between one and five meters thick. The mean
thickness is 5.4 meters, and the median, 2.6 meters.

        1.1.4 Sensitive Environmental Settings

        EPA collected information on the occurrence of sensitive environmental settings at
facilities, primarily for those facilities likely to trigger corrective action.   Exhibit 1-7 shows the
results for the approximately 2,600 facilities predicted to be likely to trigger corrective action.
Environmental settings with seven characteristics  were considered: seismic impact  areas,


                               ••• DRAFT - March 23,1993  ***

-------
Minimum = 1775
Median =1963
Mcan= 1957
Maximum = 19X9
                                         Exhibit 1-3
                           Start Dates of Industrial Activity
                                           (N = 5,800)
^,JUU
2,000
'&
. i-^
— H
'O 1,500
*0
i-i
J^ 1,000
1
500
n
^
—
-

•






130
i 1
630

1







250
1






1,600




1





2,100





1




770

210
1 1
            1775-1900     1901-1920
1921-1940      1941-1960      1961-1980
   Year Activity  Began
1981 -present    Not Available

-------
                                            1-6


                                       EXHIBIT 1-4
                            PERMIT STATUS OF FACILITIES
                                        (N = 5,800)
Permit Status
Permitted, operating facility
Operating facility seeking
permit (interim status)
Closing facility that needs a
post-closure permit
Closing facility that does not
need a post-closure permit
Closed facility with post-
closure permit
Other
Not available
TOTAL
Number of
Facilities
1,900
1,600
250
210
67
1,500
270
5,800 . :
Percent of
Facilities
33
28
4.3
3.6
1.2
26
4.7
100
                  Totals may not sura due to rounding.
floodplains, complex hydrogeology, ground-water vulnerability or high resource value, karst
terrain, poor foundation conditions, and susceptibility to mass movements. EPA has considered
special standards for waste management activities at facilities located in or near these sensitive
environments. Each of these settings represents a situation where there is an increased
probability of environmental release, difficulty in monitoring or remediation, or high resource
value.  To evaluate the occurrence of these settings, the Agency used a methodology developed
for a draft  RIA of location standards for hazardous waste management facilities.1  Note that
more than  one of the settings may exist at any facility.  The estimates suggest that the most
common of the environmental settings investigated are  floodplains and seismic impact areas.
Karst terrain is infrequently found.
   'U.S. Environmental Protection Agency. Summary Regulatory Impact Analysis/Background
Information Document for the Development of Subtitle C Location Standards under Section 3004 fo)(7)
of RCRA.  Washington, D.C.: U.S. Environmental Protection Agency, January 26, 1990.
                             *** DRAFT - March 23, 1993 ***

-------
                0-1
Minimum = 0.076 acres
Median = 20 acres
Mean = 330 acres
Maximum = 8,900 acres
                                         Exhibit 1-5
                              Distribution of Facility Areas
                                           (N = 5,800)
Number of Facilities
•T" r* .N
L/i O Lft C
§o o c
o o c
-
—
640


1,500


1.700


1.100

460 43Q


>IO-100       >100-I.OOO    >1,000-10.000    Not Available
   Area (acres)

-------
                                      Exhibit 1-6
                        Thickness of the Unsaturated Zone
                                       (N = 2,600)
2,000
C/5
i— H
*O 1.500
Cti
O
lli
v 1,000
&
|
500




n
1,900
-
—
•

—
•
—

-

67
1 1 1











1







270
i in

130

• III
1 1 III
                0-1
Minimum = 0.0025 meters
Median = 2.6 meters
Mean = 5.4 meters
Maximum = 74 mclcrs
>l-5            >5-10            >10-30
       Thickness (meters)
>30

-------
                                            1-9
                                       EXHIBIT 1-7
                      ENVIRONMENTAL SETTINGS OF FACILITIES
                                         (N = 2,600)
Environmental
Setting
Seismic Impact
Area
Floodplains
Complex
Hydrogeology
Ground-water
Vulnerability or
Resource Value
Karst Terrain
Poor Foundation
Conditions
Susceptibility to
Mass Movements
Number of
Facilities
980
1,200
260

910
20

560
490
Percent of
Facilities
38
46
10

35
0.77

22
' 19
       1.1.5  Surface-Water Information

       The Agency collected data on the kinds of downgradient water bodies which were judged
to be likely to receive erosion or ground-water releases from the facilities.  Water bodies were
classified into several categories: river, stream, creek, etc.; lake, pond, etc.; ocean, coastal area,
etc.; and no receiving water bodies.  Exhibit 1-8 shows the number of facilities with each kind of
water body, for facilities likely to trigger corrective action.  For any given facility, more than one
kind of water body may be present.  For some facilities, there are no receiving water bodies
present.

       The Agency's data collection also provided information on the proximity of water bodies
to facilities, again primarily for facilities likely to trigger corrective action.  Exhibit 1-9 shows the
distribution of distances from these facilities' boundaries to the nearest streams or rivers
downgradient that EPA judged to be likely to receive contamination  from the facilities.  The
mean distance is about 940 meters from the facility, and the minimum is zero meters (where a
river or stream runs  through or adjacent to the facility).  For an estimated 290 of the facilities,
there are no streams or rivers likely to receive contamination.
                              ••• DRAFT - March 23,1993 ***

-------
                                            I- 10
                                       EXHIBIT 1-8
                      FACILITIES WITH RECEIVING WATER BODIES
                                         (N = 2,600)
Receiving Water Body
River, Stream, Creek, etc.
Lake, Pond, etc.
Ocean, Coastal Area, etc.
No Receiving Water Bodies
Number of
Facilities
2,300
680
160
280
Percent of
Facilities
88
26
6.2
11
       The Agency also collected data on the distances from facility boundaries to the nearest
lakes and coastal water bodies downgradient that could potentially receive contamination from
the facilities. Of the facilities likely to trigger corrective action, about 680 have lakes that could
receive contamination.  EPA estimates that  about 360 of these lakes are within a kilometer of
the facility, while about 320 are more than one  kilometer away. About 160 facilities have oceans
or coastal areas potentially receiving contamination; an estimated 20 of these water bodies are
within a kilometer of the facility, while about 140 are more than a kilometer away.

       1.1.6  Ground-Water Information

       EPA's data collection effort provided information on the hydrogeological characteristics
of facilities likely to trigger corrective action. The saturated zone is the  region below the
unsaturated zone, where spaces between soil particles are filled with water that can be pumped
up for consumption (ground water). The saturated zone may be further divided into layers by
relatively impermeable geologic strata.  Exhibit MO shows the  thickness of the upper saturated
zone across the 2,600 facilities. The thickest saturated zone is  about 230 meters; the thinnest, 1.2
meters. The mean thickness is 26 meters, and the median is 11 meters.

       Exhibit  1-11 shows the average linear velocity of ground water in the upper saturated
zone at these facilities.  The mean ground-water velocity is about 95 meters per year. The
minimum velocity observed  is close to zero,  and the maximum  is 1.1 kilometers per year. At 42
percent of facilities, ground water moves at  a rate between one and ten meters per year.

       1.1.7  Exposure Information

       EPA collected data on the distance from facility boundaries to the location of the nearest
person potentially exposed to air releases (the Maximum Exposed Individual, MEI). Exhibit  1-12
shows the results for facilities likely to trigger corrective action. For almost half of these
facilities, the MEI is immediately adjacent to the facility.  The  mean distance is 86 meters, the
median is close to zero meters, and the maximum is about 2.7  kilometers.
                              ••• DRAFT - March 23, 1993 **»

-------
   1,200
   1.000
                                       Exhibit 1-9

              Distribution of Distances to Nearest Stream or River
                                         (N = 2,600)
c/3
.ti  800


1
IX
<4-i  600
 o
 Ui
 1 - 1,000

Distance (meters)
                                                                        > 1.000 -10,000

-------
                                    Exhibit MO
                       Thickness of Upper Saturated Zone
                                      (N = 2,600)
l.iiUU
1,000
on
(D
; 10-20          >20-40          >40-60
       Thickness (meters)
>60

-------
                                     Exhibit Ml
                     Average Linear Velocity of Groundwater
                                        (N = 2,600)
1,/UU
1,000
Cfl
3 800
c£
^M 600
O
Number
.u
3
200
0
-
-
—
-
—
-
h«i







200
1
0-1







1,100






1







720



I




490

I

63
I
>1-10 >IO-IOO >100-l,000 >1,000
Minimum = 0.000017 meters/year
Median = 9.6 meters/year
Mean = 95 meiers/year
Maximum = 1,100 mctcrs/ycar
                                    Velocity (meters/year)

-------
                                    Exhibit 1-12
                    Distance to Maximum Exposed Individual
                                      (N = 2,600)
I.HUU
1.200
g 1.000
• w*
•*-»
• *H
• l-H
OS 800
(X
VM
O
Number
200
n
1,200
—

•
-








970







280
130
13
              Adjacent
Minimum = 0 meters
Median = 0 meters
Mean = 86 meters
Maximum = 2,700 meters
>0-100         > 100-500        >500-1.000
        Distance (meters)
> 1.000

-------
                                            I- IS
       EPA also determined the total size of populations that are located within ten kilometers
of facilities and could potentially be exposed  to air releases from them.  The Agency determined
the population within each of five concentric  rings around the facilities.  For all facilities likely to
trigger corrective action, Exhibit 1-13 shows the average across facilities of the population in each
of these rings.  In addition, the exhibit shows the average of the cumulative populations within
different distances of the facilities. The average across facilities of the total population within 0.5
kilometers is approximately 1,300; within two kilometers, 13,000; within ten kilometers, 200,000.
                                       EXHIBIT 1-13
                   AVERAGE POPULATION POTENTIALLY EXPOSED
                          TO AIR RELEASES FROM FACILITIES
                                         (N o 2,600)
Distance
From Facility
Boundary
(km)
0 to 0.5
>O.S to 1
>1 to 2
>2to5
>5 to 10
Average
Population
1,300
3,000
8,600
48,000
140,000
Cumulative
Distance From
Facility
Boundary (km)
0 to 0.5
Otol
Oto2
Oto5
Oto 10
Cumulative
Population
1,300
4,300
13,000
61,000
200,000
       The Agency also investigated potential exposure through the ground- and surface-water
routes.  For the ground-water route, EPA determined the distance in the direction of ground-
water flow to the nearest public or private residential use well likely to receive contamination.2
For approximately 640 facilities, the nearest well is less than a kilometer away; for about 140
facilities, it is more than a kilometer away. For the remainder of the 2,600 facilities, there are no
downgradient wells judged to be likely to receive contamination.

       For all facilities likely to trigger corrective action, Exhibit 1-14 shows the average across
facilities of the number of private wells that could become contaminated by releases to ground
water.  The data include only wells downgradient from the facility. The exhibit also shows the
   2For questions involving the number of wells near facilities, wells that were on the other side of
groundwater divides (e.g., surface water intercepts) were not included.
                              *** DRAFT - March 23, 1993 •••

-------
                                           1-16
number of residents served by these wells.  Data for each distance interval are given, as well as
cumulative numbers within different distances of the facilities.  The average number of wells
within two miles of facilities is about 16, and the average population served by wells within two
miles is 43.3
                                       EXHIBIT 1-14
           AVERAGE NUMBER OF PRIVATE WELLS POTENTIALLY SUBJECT
                   TO CONTAMINATION AND POPULATIONS SERVED
                                         (N = 2,600)
Distance From
Facility
Boundary
(miles)
0 to 0.25
>0.25 to 0.5
>0.5 to 0.75
>0.75 to 1
>1 to 1.5
> 1.5 to 2
Average
Number of
Wells
0.80
0.39
1.2
2.2
3.3
8.2
Average
Population
Served
2.1
1.0
3.2
5.7
8.8
22
Cumulative
Distance From
Facility Boundary
(miles)
0 to 0.25
0 to 0.5
0 to 0.75
Oto 1
0 to 1.5
Oto 2
Cumulative
Number of
Wells
0.80
1.2
2.4
4.6
7.9
16
Cumulative
Population
Served
2.1
3.1
6.3
12
21
43
       EPA also collected information on public wells near facilities and the populations they
serve.  Of the facilities likely to trigger corrective action, EPA estimates that about 200 have
public wells in the ground-water flow direction that are likely to become contaminated.4 For
these facilities, the mean and median distances to the nearest well are about 1.2 miles, and the
minimum distance is zero miles.  The average across facilities of the number of people served by
each of these wells is 4,200, and the median is about 3,500. The minimum number of people
served by a well is 25, and the maximum is 53,000.
   The calculation of populations served assumes a ratio of 2.63 residents per well, based on 1990
national census data.

   4In general, EPA examined wells within two miles of the facility. However, for facilities for which
modeling results indicated that contamination would spread beyond two miles, the Agency collected data
on wells within five miles.
                              *** DRAFT - March 23, 1993 ***

-------
                                           1-17

       Finally, for the surface-water route, EPA collected data on public drinking water supply
intakes and their users.  Of facilities likely to trigger corrective action, about 870 are estimated to
have public drinking water supply intakes within IS miles downstream on one of the streams or
rivers described  in section 1.1.5.  For these facilities, the mean distance to the closest intake is 8.4
miles, the minimum distance 3.7 miles, and the maximum  distance 15 miles. The average
number of people served by these public water systems is  130,000, the minimum number 300, and
the maximum number 340,000.
1.2 SWMU Characteristics

       1.2.1 SWMU Identification and General Characteristics

       EPA collected data on the characteristics of SWMUs at all facilities subject to corrective
action requirements. Exhibit 1-15 shows the total numbers of SWMUs and the average number
of SWMUs per facility, for federal  and non-federal facilities.5  Exhibit 1-16 shows the same data
for SWMUs in three regulatory status categories:

       •      regulated RCRA Subpart F land disposal SWMUs (i.e. landfill, surface
              impoundment, waste pile, or land treatment unit that managed hazardous waste
              after 7/26/82)

       •      Other SWMUs (e.g., solid waste landfills, other hazardous waste management
              units requiring permit, or exempt units)          !

       •      Areas of concern

Exhibit 1-17 shows the numbers of SWMUs broken down into detailed categories of SWMU
types.  Exhibit 1-18 presents data on the SWMUs in each of these categories that will require
remediation.  The exhibit shows the average across SWMUs of the age of the SWMU, from  the
year it was first used until the present.  It also  shows the average length of use of SWMUs (from
the year use began until the year it ended for inactive SWMUs and from the  year use began until
the present for active SWMUs). Finally, it shows the area  of SWMUs.

       1.2.2 Waste Characteristics

       EPA also  collected data on the constituents of wastes in the SWMUs  at facilities likely to
trigger corrective  action.  Exhibit 1-19 shows the ten most prevalent constituents, and the
numbers of facilities where these constituents are present.  The most common constituents are
chromium, lead, arsenic, phenol and tetrachloroethylene, each present at least a thousand of the
2,600 facilities likely to require  cleanup.
   'Note that the sample excluded the nine largest Federal facilities and thus underestimates the total
number of SWMUs.

                             *** DRAFT - March 23,1993 ***

-------
                        1-18

                   EXHIBIT 1-15
     TOTAL SWMUs AND SWMUs PER FACILITY
   FOR FEDERAL AND NON-FEDERAL FACILITIES
                     (N = 5,800)
SWMU
Type
Federal •
Non-Federal
TOTAL
Number of
SWMUs
9,700
94,000
104,000
Number of
Facilities
350
5,400
5,800
SWMUs
Per Facility
28
17
18
Totals may not sura because of rounding.
* Excludes the nine very large federal facilities.
                   EXHIBIT 1-16
     TOTAL SWMUs AND SWMUs PER FACILITY
        BY REGULATORY STATUS OF SWMU
                     (N • 5,800)
Regulatory Status
of SWMU
Subpart F
Other SWMUs
Areas of Concern
TOTAL
Number of
SWMUs
4,600
97,000
1,700
104,000
SWMUs Per
Facility
0.79
17
0.29
—
   Totals may not sum because of rounding.
           **» DRAFT - March 23, 1993 **«

-------
                         I- 19


                    EXHIBIT 1-17
       NUMBERS OF SWMUS BY SWMU TYPE
                      (N = 5,800)
SWMU Type
Landfills *
Land Treatment *
Waste Piles *
Surface Impoundments "
Tanks
Incinerators
Injection Wells
Waste Transfer Stations
Waste Recycling Operations
Spill Areas
Accumulation Areas
Process Sewers
Other SWMUs
Areas of Concern
TOTAL
Number of
SWMUs
5,400
790
2,800
9,700
30,000
1,600
430
1,600
2,300
5,800
10,000
5,000
26,000
1,700
104,000
Percent of
SWMUs
5.2
0.76
2.7
9.3
29
1.5
0.41
1.5
2.2
5.6
9.6
4.8
25
1.6
100
Totals may not sum because of rounding.
* Includes both Subpart F units and SWMUs.
           *** DRAFT - March 23, 1993 ***

-------
                               1-20
                          EXHIBIT 1-18
                AGE, LENGTH OF USE AND AREA
            FOR SWMUs REQUIRING REMEDIATION
                            (N = 5,800)
SWMUType
Landfills •
Land Treatment *
Waste Piles •
Surface
Impoundments "
Tanks
Incinerators
Injection Wells
Waste Transfer
Stations
Waste Recycling
Operations
Spill Areas
Accumulation
Areas
Process Sewers
Other SWMUs
Areas of Concern
TOTAL
Number of
SWMUs
Remediated
2,000
320
810
3,500
2,800
10
37
3
0
1,100
740
410
2,900
100
15,000
Age
(Years)
40
18
34
28
29
72
13
20
_
26
37
25
41
15
30
Length
of Use
(Years)
23
11
24
20
25
58
N/A
2Q
.^_
21
28
10
27
13
21
Surface
Area (M2)
20,000
25,000
1,700
18,000
9,700
200
N/A
4
—
1,900
900
390
1,800
23
6,200
NA: Data not available.
Totals may not sun because of rounding.
• Includes both Subpart F units and SWMUs.
                 *** DRAFT - March 23, 1993 •**

-------
              1-21
          EXHIBIT 1-19
MOST PREVALENT CONSTITUENTS
           IN SWMUs
            (N m 2,600)
Constituent
Chromium
Lead
Arsenic
Phenol
Tetrachloroethylene
Benzene
Cadmium
Toluene
Xylenes (Mixed)
Methyl Chloroform
Number of
Facilities
Where
Present
1,500
1,500
1,100
1,000
1,000
990
980
890
800
700
Percent of
Facilities
Where
Present
58
58
42
38
38
38
38
34
.- 31
27
   ••• DRAFT - March 23,1993 **•

-------