United States
           Environmental Protection
           Agency
              Office of
              Radiation Programs
              Washington. DC 20460
EPA 520/1-89-028
October 1989
           Radiation
vvEPA
Comparing Risks from Low-Level
Radioactive Waste Disposal on Land
and in the Ocean:  A Review
of Agreements/Statutes, Scenarios,
Processing/Packaging/Disposal
Technologies, Models, and Decision
Analysis Methods

-------
 COMPARING RISKS FROM LOW-LEVEL RADIOACTIVE WASTE
    DISPOSAL ON LAND AND IN THE OCEAN:  A REVIEW OF
             AGREEMENTS/STATUS.SCENARIOS,
PROCESSING/PACKAGING/DISPOSAL TECHNOLOGIES, MODELS,
             AND DECISION ANALYSIS METHODS
            P. D. Moskowitz, P. D. Kalb, S. C. Morris, M. D. Rowe,
                M. Marietta,  L. Anspaugh, and T. McKone
             Sandia National Laboratories, Albuquerque, New Mexico
          Lawrence Livermore National Laboratory, Livermore, California
                           October 1989
          Prepared for the Office of Solid Waste and Emergency Response
  Office of Radiation Programs, U. S. Environmental Protection Agency, Washington, DC
         BIOMEDICAL AND ENVIRONMENTAL ASSESSMENT GROUP
                  DEPARTMENT OF APPLIED SCIENCE
                BROOKHAVEN NATIONAL LABORATORY
                   ASSOCIATED UNIVERSITIES, INC.
              Under Contract No. DE-AC02-76CH00016 with the
                      U. S. Department of Energy

-------
COMPARING RISKS FROM LOW-LEVEL RADIOACTIVE WASTE DISPOSAL
ON LAND AND IN THE OCEAN: A REVIEW OF AGREEMENTS/STATUTES,
SCENARIOS, PROCESSING/PACKAGING/DISPOSAL TECHNOLOGIES, MODELS,
AND DECISION ANALYSIS METHODS
P.D. Moskowitz, P. D. Ka1~, s.c. Morris, M.D. Rowe,
M. Marietta, L. Anspaugh and T. McKone2
October 1989
PREPARED FOR THE
OFFICE OF SOLID WASTE AND EMERGENCY RESPONSE
OFFICE OF RADIATION PROGRAMS
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, DC
;sandia National Laboratories, Albuquerque, New Mexico
Lawrence Livermore National Laboratory, Livermore, California
BIOMEDICAL AND ENVIRONMENTAL ASSESSMENT DIVISION
DEPARTMENT OF APPLIED SCIENCE
BROOKHAVEN NATIONAL LABORATORY
UPTON, NY 11973

-------
DISCLAIMER
This report was prepared as an account of work sponsored by an agency of the United
States Government. Neither the United States Government nor any agency thereof.
nor any of their employees. nor any of their contractors. subcontractors, or their
employees, makes any warranty, express or implied. or assumes any legal liability or
responsibility for the accuracy. completeness. or usefulness of any information.
apparatus. product. or process disclosed. or represents that its use would not infringe
privately owned righta. Reference herein to any specific commercial product. process.
or service by trade name. trademark. manufacturer, or otherwise, does not necessarily
constitute or imply its endorsement. recommendation, or favoring by the United States
Government or any agency, contractor or subcontractor thereof. The views and
opinions of authors expressed herein do not necessarily state or reflect those of the
United States Government or any agency. contractor or subcontractor thereof.

-------
COMPARING RISKS FROM LOW-LEVEL RADIOACTIVE WASTE
DISPOSAL ON LAND AND IN THE OCEAN: A REVIEW OF
AGREEMENTS /ST A TUS,SCENARIOS,
PROCESSING/P ACKAGING/DISPOSAL TECHNOLOGIES, MODELS,
AND DECISION ANALYSIS METHODS
P. D. Moskowitz, r D. Kalb, S. ~. Morris, M. D. !owe,
M. Marietta, L. Anspaugh, and T. McKone
1 Sandia National Laboratories, Albuquerque, New Mexico
2Lawrence Livermore National Laboratory, Livermore, California
October 1989
Prepared for the Office of Solid Waste and Emergency Response
Office of Radiation Programs, U. S. Environmental Protection Agency, Washington, DC
BIOMEDICAL AND ENVIRONMENTAL ASSESSMENT GROUP
DEP AR TMENT OF APPLIED SCIENCE
BROOKHAVEN NATIONAL LABORATORY
ASSOCIA TED UNIVERSITIES, INC.
Under Contract No. DE-AC02-76CHOOOI6 with the
. U. S. Department of Energy

-------
COMPARING RISKS FROM LOW-LEVEL RADIOACTIVE WASTE
DISPOSAL ON LAND AND IN THE OCEAN: A REVIEW OF
AGREEMENTS/ST A TUS.SCENARIOS.
PROCESSING/PACKAGING/DISPOSAL TECHNOLOGIES. MODELS.
AND DECISION ANAL VSIS METHODS
P. D. Moskowitz, r D. Kalb, S. s:. Morris, M. D. ~owe,
M. Marietta, L. Anspaugh, and T. McKone
ISandia National Laboratories, Albuquerque, New Mexico
2Lawrence Livermore National Laboratory, Livermore, California
October 1989
Prepared for the Office of Solid Waste and Emergency Response
Office of Radiation Programs, U. S. Environmental Protection Agency, Washington, DC
BIOMEDICAL AND ENVIRONMENTAL ASSESSMENT GROUP
DEPARTMENT OF APPLIED SCIENCE
BROOKHAVEN NATIONAL LABORATORY
ASSOCIATED UNIVERSITIES, INC.
Under Contract No. DE-AC02-76CHOOOI6 with the
U. S. Department of Energy

-------
ABSTRACT
Large
volumes
of
low-level
radioactive
waste
(LLW)
are
either
in
temporary storage or unexcavated from Superfund and U.S. Department of Energy
Defense Facility sites. This waste must be isolated from the public for very
long periods of time because of its long half-life. Because of the need to
manage this waste, questions have been raised about the hazards associated
with the disposal of these materials on land or in the ocean. Similarly, the
U. S. as a signatory to the London Dumping Convention is now engaged in
international discussion about the comparative risk of disposing of LLW on
land and in the ocean. In support of these national and international
activities, the U.S. Environmental Protection Agency is examining scientific
and technical questions related to the comparative health and environmental
risks of ocean- and land-based disposal of LLW.
In support of these efforts,
this report gives background information
on:
The history
of LLW disposal
in
the U. S . ;
agreements,
statutes
and
regulations for the disposal of LLW; disposal scenarios and alternative
treatment options for LLW; methods and models which could be used to assess
and compare risks associated with land and ocean options for LLW disposal;
technical
and
methodological
different
options;
and,
roles
issues associated with comparing risks of
of decision making approaches in comparing
risks across media.
In any assessment effort, the minimal set of health risks. to be examined
should include routine and accidental radiation exposures to workers and the
public from. transport (land and ocean), disposal, and post-closure releases
of radionuclides. Potentially important contributors to health and
environmental risks which have not been fully evaluated include waste
processing and event probabilities.
Since waste processing,
disposal
requirements
for
land
and
ocean
disposal
can
packaging and
differ, their
contribution to the overall
risk can vary by medium.
Human and natural
processes that precipitate accident- initiated releases
are also
important
determinants of risk and must be evaluated.
Ideally, models selected for
analysis should have completed some validation exercises and be capable of
producing explicit estimates of model uncertainty. In practice, however,
most models have not undergone complete validation efforts and are not now
able to produce explicit estimates of output uncertainty.
iii

-------
It is clear, that the general public is concerned about a much broader
range of characteristics and potential risks of the LLW disposal options than
just health risks. These especially include socio-economic risks and the
equity of the distribution of risks and benefits in space and time. Most LLW
risk assessment models do not normally include these important
considerations, and few of the models reviewed could be modified to include
any of them with modest effort. This is a serious deficiency in current
capabilities that will undermine the credibility of results in the eyes of
the general public. When this information becomes available, there are
decision aiding methods for incorporating a broad range of characteristics,
potential risks, and stakeholder values into quantitative comparisons of
alternatives. Development of necessary data and application of these
approaches
decisions.
can pe rmi t
decision-makers
to make
better
and more
informed
iv

-------
ACKNOWLEDGMENTS
This work was supported by the Office of Radiation Programs and the
Office of Solid Waste and Emergency Response, U.S. Environmental Protection
Agency, Washington, DC. We thank Robert E. Dyer, C. Elliot Foutes, Larry J.
Zaragoza, and Marilyn E. Varela for their guidance and helpful discussions.
We are indebted to Arlean Vanslyke for her help in typing and organizing this
report.
v

-------
CONTENTS
1
1
INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2
HISTORICAL BACKGROUND. . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3
3
REGUlATORY BACKGROUND. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5
4
RISK ASSESSMENT ISSUES................................................. 11

4.1 Nature of the Problem............................................. 11

4.2 Scenario CharacterizatiQn......................................... 11

4.2.1 Scenario Definition........................................ 11

4.2. 2 Inventory/Source Term...................................... 14
4.2.3 Waste Processing Technology............. """ ............. 15
4.2.4 Waste Disposal Options - Land.... ........ ........ .... ...... 22
4.2.5 Waste Disposal Options - Ocean... ..... ..,... ............... 27
4.3 Consequence Modeling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ~ . . . . . . . . . . . .. 33
5
MODEL EVALUATION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 45

5.1 Model Identification, Selection and Evaluation.................... 45
5.2 Model Reviews - Land.............................................. 48

5 . 2 . 1 PRESTO[[[ 48

5.2.2 BARRIER[[[ 49

5.2 . 3 IMPACTS[[[ 50

5 . 2 .4 GEOTOX[[[ 51

5.3 Model Reviews - Ocean............................................. 51

5.3.1 Bryan-Semtner-Cox.......................................... 51

5.3.2 SANDIA Ocean Modeling System............................... 52

5.3.3 Holland[[[ 52

5 . 3 . 4 Harvard[[[ 53

5.3.5 NRPB9l[[[ 53

5 . 3 . 6 MARINRAD[[[ 53

5.3.7 MARK A. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 54
6
COMPARING RISKS OF THE lAND AND OCEAN DISPOSAL OPTIONS................. 55
6.1 Characteristics of Concern........................................ 60
6.2 Quantitative Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 72
6.3 Qualitative Results.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 73
6.4 Gaps in Available Quantitative Information......... ............... 74
6.4. 1 Health Risks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 74
6.4.2 Environmental Impacts....... ............................... 75
6 .4 . 3 Economic Impac ts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 75

6.4.4 Social Impacts............................................. 76
6.5 Methods of Filling Gaps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 77
6.6 Relative Risk Evaluation.......................................... 78

6.6.1 Values[[[ 79


-------
8 REFERENCES.. 0 0 0 0 . . . . . 0 0 . . 0 . . 0 0 0 . 0 0 . 0 . 0 0 . 0 . 0 . . 0 . . . .' 93
.. ... .. .... . . .. . . . ..
APPENDIX A:
APPENDIX B:
APPENDIX C:
APPENDIX D:
AGREEMENTS, STATUTES, AND REGULATIONS FOR DISPOSING

OF LLW.. . . . . . . . 0 0 000 0 . 0 . . . . . 0 . . . . . . . 0 0 0 . 0 . . 0 . . 0 . 0 0 0 0 0 . 0 . . . . .. 99


MODEL REVIEWS - LAND........................................ .125
MODEL REVIEWS - OCEAN........... 0 . 0 . . . . 0 0 0 . 0 . . . . 0 0 . . . . . 0 . . . . .185
DECISION ANALYSIS APPROACHES FOR LLW DISPOSAL....o o. 0 o. 0'" ..227
vii

-------
TABLES
1.
Phenomena potentially relevant to scenario analysis for
shallow ground repositories. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 13
2.
Quantities of LLW generated in the U.S. by source type,

1984-1987. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 16
3.
Activity levels of LLW generated in the U.S. by

source type, 1984-1987............................................... 17
4.
Typical radionuc1ides in low-level waste. ... .......... ... ............ 18
5.
Potential LLW treatment options... ............ ..;.... ... .... ... ... ... 20
6.
Potential LLW solidification and packaging alternatives. ....... ...... 21
7.
Potential alternatives to conventional shallow land disposal

of LLW. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 23
8.
Status of five LLW alternative disposal methods...................... 25
9.
NEA design basis for ocean disposal packages......................... 32
10.
Proposed waste package performance criteria for ocean disposal

of LLW. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 34
11.
Pathways and modes of exposure for the land disposal option.......... 38
12.
Pathways and modes of exposure for the ocean disposal option......... 41
13.
Models for evaluating health hazards associated with 1and-

and ocean-based disposal............................................. 46
14.
Model evaluation criteria for performance assessment models to
compare health hazards from land and ocean disposal of low-level .

radioactive wastes.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 47
15.
Important characteristics determining individual evaluation of

seriousness of risks........ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 61
16.
Impacts, characteristics, and risks of concern for comparing
land versus ocean disposal of LLW.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 64
17.
Impacts, characteristics, and risks of concern for comparing
land versus ocean disposal of LLW: Summary and Recommendations....... 70
18.
Rationales for selecting characteristics and impacts for inclusion
in the multimedia comparison. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 89
viii

-------
FIGURES
1.
Time-line indicating significant international milestones and
agreements for ocean disposal of low-level radioactive waste. . . . . . . . .
6
2.
Time-line highlighting past U.S. ocean disposal practices and
U.S. laws/regulations governing such practices.... ............. ......

Modules used for land-based risk assessment purposes........ ......... 36
8
3.
4.
Integrated environmental/socioeconmic impact assessment model. . . . . . .. 56
5.
Hypothetical distributions of risks of human exposure, one with
only routine low-level risks and another with only higher-level

risks of accidents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 85
ix

-------
1
INTRODUCTION
Large
volumes
of
low-level
radioactive
waste
(LLW)
are
either
in
temporary storage or unexcavated from Superfund and U.S. Department of Energy
Defense Facility sites. This waste must be isolated from the public for very
long periods of time because of its long half-life. Because of the need to
manage this waste, questions have been raised about the hazards associated
with the disposal of this material on land or in the ocean.
Similarly, the
u. S. as a signatory to the London Dumping Convention is now engaged in
international discussion about the comparative risk of disposing of LLW on
land and in the ocean. In support of these national and international
activities,
the
U.S.
Environmental
Protection Agency
(EPA)
is
examining
scientific and technical questions
related to the comparative health and
environmental risks of the ocean- and land-based disposal of LLW.
In support
of these efforts, this report gives background information on technical and
methodological issues associated with comparing risks among the different
options. Results presented are based on an exploratory effort to identify
the current state-of-knowledge and important existing gaps which should be
evaluated before comprehensive and equitable multimedia risk assessments are
prepared. In this context, this report includes:
(1)
A brief review of the history of LLW disposal in the U.S.
(2)
A review of agreements, statutes and regulations for the disposal
of LLW.
(3)
An identification of disposal scenarios and alternative treatment
options for LLW.
(4)
An identification,
characterization and analysis of methods and
models which could be used to assess and compare risks associated
with land based and ocean options for LLW disposal.
(5)
Rol~ of decision making approaches in comparing risks across media.
The information presented in this report on models for risk assessment
purposes is based on model documentation, publications, and previous
experiences of the authors. Time and resources allocated for this effort did
- 1 -

-------
not permit full evaluation of the computer codes, nor were test runs of the
models made.
- 2 -

-------
2
HISTORICAL BACKGROUND a
LLW is generated during the routine operation of nuclear power plants;
from biomedical applications such as nuclear medicine therapy and diagnostic
techniques at hospitals and medical facilities; scientific research at
universities and other research and development institutions; industrial
small
amounts
of
radioactive
materials
mixed
in
It usually contains
larger volumes of
isotope production; and from defense-related operations.
nonradioactive materials.
During the early days of research using radioactive materials (from just
after World War I to World War II), LLW was either burned, buried in shallow
trenches,
or
diluted
and
released
to
sewer
systems
at
the
point
of
Limited knowledge of the health effects of radiation and
environmental transport of pollutants contributed to the lack of attention
paid to waste disposal issues. Beginning with the Manhattan proj ect to
develop nuclear weapons, the volume of radioactive materials and diversity of
isotopes generated as waste by-products grew tremendously. As a result, the
Atomic Energy Commission (AEC) created new disposal sites at federal
generation.
facilities located in Hanford, Washington;
Idaho Falls, Idaho; Los Alamos,
New Mexico; Oak Ridge, Tennessee; Savannah River, Georgia; and at the Nevada
Test Site, Nevada. Disposal techniques consisted mainly of shallow land
burial in which long, shallow trenches were dug, filled with waste, and
backfilled with soil.
These facilities were primarily developed to handle
defense-related wastes, but as the nuclear power industry began to expand,
commercial wastes were also accepted at some sites.
This continued until
1962 when the AEC licensed private companies to operate LLW disposal sites.
The first commercial sites licensed were at Beatty, Nevada, and shortly
thereafter at Maxey Flats, Kentucky.
In addition to the expanded use of shallow-land burial techniques, the
AEC initiated disposal of LLW at sea in 1946. Originally carried out by the
U.S. Navy, waste containers (55-gallon drums) were filled with LLW materials
and with concrete to provide a cap and increase density to ensure sinking;
approximately 107,000 waste containers comprising 4 x 1015 Becquerels (Bq)
were placed into the Atlantic and Pacific Oceans. Ocean disposal was
a Information presented in this chapter is extracted from Burns and Briner,
1980; Holcomb, 1982; and Parker, 1988.
- 3 -

-------
extensively pursued until 1962, when the introduction of commercial 1and-
based disposal facilities made it economically unfeasible; operations were
co~p1ete1y curtailed in 1970. European countries have also used the ocean
for waste disposal. Since 1949, they have disposed of an estimated total
mass of 100,000 metric tons containing 3 x 1016 Bq in-the Northwest Atlantic
Ocean.
In the U.S., LLW is currently being disposed of on land only. In 1982,
over 75,000 m3 of LLW was disposed of in commercially licensed shallow-land
burial disposal sites.
Six such sites have been in operation.
In addition
to Beatty and Maxey Flats, they include: Barnwell, South Carolina; Hanford,
Washington; Sheffield, Illinois; and West Valley, New York. Maxey Flats,
West Valley and Sheffield are now closed - the first two due to technical and
environmental problems and the latter because it was
filled to capacity.
Because the number of commercial sites is so limited and the volume of LLW
has increased greatly in the last decade, there is a need for new commercial
sites. In addition, dwindling disposal capacity at existing sites has made
existing host states increasingly reluctant to accept wastes generated
outside their own borders.
- 4 -

-------
3
REGULATORY BACKGROUND
Since current LLW disposal practices date back more than 40 years, both
disposal techniques and concerns over environmental effects have varied
considerably. Consequently, laws and regulations governing the disposal of
LLW have also changed. Consideration of current and pending legislative
constraints may playa key role in the selection of new LLW disposal options.
Applicable international agreements and federal statutes/regulations for
ocean dumping, as well as federal and state statutes/regulations governing
land-based disposal are summarized in this section.
A more complete review
of agreements, statutes and regulations for the disposal of LLW is included
in Appendix A.
Since the onset of ocean disposal in 1946, growing worldwide concern for
preserving the condition of the seas and newly acquired knowledge on the
effects of ionizing radiation on human health and safety have led to a
dynamic regulatory environment. International concern was initially focused
at the first U.N. Conference on the Law of the Sea (UNCLOS I), held in Geneva
in
1958.
This conference enacted the first international law of the sea,
which is still in force today. In 1972 the London Dumping Convention (LDC)
developed the most comprehensive set of international regulations for marine
pollution by dumping. The requ1ations were ratified by some 50 countries I
including the major maritime nations. Due to concern over uncertainties in
environmental
impacts,
and widespread social protests,
ocean disposal was
temporarily suspended in 1983. Resolution (28): 10 to the LDC, approved in
1986, called for a voluntary moratorium on the use of the ocean for disposal
of LLW' until certain scientific and technical matters,
as well as social,
political
and
economic
issues
are
addressed.
An
LDC
sponsored
Intergovernmental Panel of Experts on Radioactive Waste Disposal at Sea has
established two working groups that are currently investigating these issues.
Figure
1
traces
significant
international
milestones
and agreements
for
disposal of LLW' at sea.
In addition to its support of international ocean disposal agreements,
the U. S. has developed a parallel set of laws and regulations.
Following
passage of the National Environmental Policy Act (NEPA) in 1970 which set
forth general environmental policy, Congress enacted a series of specific
environmental laws including the Marine Protection, Research and Sanctuaries
- 5 -

-------
------------
OCEAN DISPOSAL OF LLW
International Involvement/Agreements
UNCLOS I
(1958)
U.S. begins ocean
disposal 01 LLW
1 (1948)


I I I
1946 t 1951 1956

U.K. begins ocean
disposal 01 LLW
(1949)
London Dumping Convention
(1972)
IAEA Updates
Guidelines
(1978)
Bryneilason Report
(1961)

\NEA/OECD coordinates
international disposal
(19f7)

I I I
1961 1966 t1971

U.S. curtails
ocean dumping
(1970)
Convention on
Law 01 the Sea
(1982)
t
I I
1981 t 1986

European nations
curtail dumping
(1983) -
I
1976
Figure 1.
Time-line indicating significant international milestones
agreements for ocean disposal of low-level radioactive waste.
and
- 6 -

-------
Act (MPRSA) of 1972. This act empowered EPA to oversee adherence to the
policy guidelines of the MPRSA and issue regulatory criteria for ocean
disposal permits. In January 1983, an amendment to MPRSA imposed a two-year
moratorium on ocean disposal of LLW and a more stringent, supplementary set
of permit requirements following the moratorium. For a LLW ocean disposal
permit, it must be demonstrated that:
.
the proposed dumping is necessary to conduct research either i)
on new technology related to ocean dumping, or ii) to determine
the degree to which the dumping of such substance will degrade
the marine environment;
.
the scale of the proposed dumping is limited to the smallest
amount of such material and the shortest duration of time that
is necessary to fulfill the purposes of the research, such that
the dumping will have minimal adverse impact upon human health,
welfare, and amenities, and the marine environment, ecological
systems, economic potentialities, and other legitimate uses;
.
after consultation with the Secretary of Commerce, the
potential benefits of such research will outweigh any such
adverse impact; and
.
the proposed dumping will be preceded by appropriate baseline
monitoring studies of the proposed dump site and its
surrounding environment.
Other requirements under the 1983 amendment include submission of a
Radioactive Material Disposal Impact Assessment and final approval by a joint
resolution of Congress.
Past u.S. ocean disposal practices and U.S. laws and
regulations governing such practices are highlighted in Figure 2.
Policy and regulatory responsibilities for land-based disposal of LLW
are covered by the Atomic Energy Act of 1954 (AEA) and the LLRWPA of 1980 and
their amendments. The former establishes the Nuclear Regulatory Commission
(NRC) as the licensing and regulatory body for commercial land-based
radioactive waste disposal and provides states regulatory power under
authority granted by NRC. Defense-related LLW generated by federal
facilities and their contractors are the responsibility of the u.S.
- 7 -

-------
------
OCEAN DISPOSAL OF LLW
U.s. LAWS/REGULATIONS
40 CFR 220-228
(1977)
1946
1951
AEC moratorium
(1962)
,I

1961
NEPA enacted
1 (1970)

Marine Protection
Research & Sanctuar.ies
~ct of ~972


1966 t1971 1976 1981 t 1986

U.S. curtails Amendment to MPRSA
ocean disposal (1983)
(1970)
1
U.S. begins
LLW ocean disposal
(1946)
1
1956
Figure 2.
Time-l~ne highlights past U.S. ocean disposal practices and U.S
laws/regulations governing such practices.
- 8 -

-------
Department of Energy (DOE).
and safety resulting from
defense-related disposal
EPA is responsible for protecting public health
exposure to radiation from either commercial or
operations.
NRC
has
issued
a
comprehensive
regulation [NRC, 1982] governing licensing procedures, performance
objectives, and technical requirements for land-based LLW disposal. DOE has
promulgated an internal order [U. S. Department of Energy, 1988] covering
disposal operations under its jurisdiction. EPA has issued an Advanced
Notice of Proposed Rulemaking for LLW disposal standards that would establish
allowable exposure limits, define a "below regulatory concern - BRC" waste
classification, and set groundwater protection standa~ds.[U.S. Environmental
Protection Agency. 1983] Recognizing that the combination of these factors
could lead to a severe disposal crisis in the near future, Congress mandated
that states assume Il"esponsibility for disposal of their oWn LLW. The Low-
Level Radioactive Waste Policy Act of 1980 (LLRWPA) as amended in 1985 [U.S.
Congress, 1980] required that individual states or groups of states that form
specific compacts must develop new disposal sites for their LLW by 1993 or
face stiff financial and other penalties. In spite of these Congressional
efforts, progress towards implementation of the LLRWPA has been slow. The
primary reasons for the 1985 amendments to the LLRWPA were to extend the
original deadline for implementation by five years, p!ovide incentives for
those states that do comply, and penalties for those that do not.
Siting of
new disposal facilities continues to be a serious problem for the states and
regional compacts.
areas
that
meet.
In addition to the difficulties of locating potential
technical and environmental criteria (e.g., climate,
hydrology, demographics, location), public acceptance of waste disposal sites
within local communities has been hard to attain.
- 9 -

-------
10

-------
4
RISK ASSESSMENT ISSUES
4.1
Nature of the Problem
TIle
objective
of
risk
assessment modeling
of radioactivity in
is to predict future
different environment~l
quanti ties or
concentrations
media, and to estimate from these the final dose to man from which health
hazards can be calculated.
TIle process of assessing the risks of different
LLY disposal options can be divided into two basic steps:
(1)
Scenario characterization;
(2)
Consequence modeling.
TIle following sections expand on these steps.
4.2
Scenario Characterization
TIle first task establishes the conceptual bounds of the analysis. It
involves identification and quantification of phenomena which could initiate
release of radionuclides and or influence rates at which releases occur.
In
developing ~isk estimates of LLY disposal options, careful consideration must
be given to setting technical and natural boundaries of the system to be
characterized. If the boundaries are too small, important contributors to
risk may be missed. Similarly, if the boundaries are too large, excessive
time
and
effort could be spent evaluating issues of only
Several major issues associated with the setting of
for analysis are discussed below.
secondary
appropriate
importance.
boundaries
4.2.1
Scenario Definition
In scenario definition, identification of the activities of interest is
of primary concern.
Activities
are
defined here
as
a
set
of
actions
associated with the handling, processing, transport and disposal of waste.
In this context, boundaries for analysis of risk generally must include all
activities ranging from waste generation to the post-closure stage at the
disposal site. To the extent that these options differ among the
alternatives, a complete life-cycle analysis is essential if risk estimates
are to be comparable.
- 11 -

-------
Events may occur for each activity that could result in the release of
radioactive materials. Table 1 gives a list of sample phenomena which are
potentially relevant to scenario analysis for land disposal of LLW. These
phenomena or events can be divided into three major categories; human
activities (e. g., construction, drilling for mineral resources). natural
processes (e.g., erosion and flooding), and waste and disposal site processes
(gas generation or mechanical disturbance of soils or rocks at the disposal
site. (International Atomic Energy Agency, 1984]
Sets of combined activities and events (phenomena) that could contribute
to release of radionuclides from a disposal system and result in human
exposure can be defined as a scenario. Although system characterization is
very important to developing conceptual models for risk assessment, scenario
development is equally important for characterizing the different hazards
associated with the disposal operation as well as for showing regulatory
compliance.
Selection of appropriate time scales is important because of the long
half-lives of many radionuclides as well as the time scales in which various
events may occur.
include:
Time scales of relevance
to the LLW disposal option
(1)
The assumed duration of dumping.
(2)
The half-lives of the radioisotopes.
(3)
The time-scales associated with physical transport processes.
(4)
The time-scales associated with biological processes.
(5)
The time-scales over which events occur.
(6)
The integration times for assessing dose
annual and lifetime.
to man,
including both
Selection of appropriate spatial resolution for analysis should be based
on such important considerations as the media into which materials are
released and the time-scales of interest. The geographic scales of interest
for routine releases from the ocean disposal option will be orders of
magnitude larger than that of the land-based option. For accidents, however,
- 12 -

-------
Table 1. Phenomena potentially relevant to scenario analysis for shallow
ground repositories [International Atomic Energy Agency. 1984a].
HUMAN ACTIVITIES
Improper design and operation

Chemical liquid waste disposal
Draining system obstruction
Improper waste equipment
Top cover failure

Future intrusion
Construction activities
Farming
Groundwater exploitation
Habitation
Salvage
Re-use of disposed material
Archaeology
NATURAL PROCESSES AND EVENTS
Biological

Animals
Plants

Faulting seismicity
intrusion
Fluid interactions
Erosion
Flooding
Fluctuations in the
Groundwater flow
Seepage water

Weathering

Deterioration with time
Freezing/thawing
Wetting/drying
water-table
WASTE AND REPOSITORY PROCESSES
Gas generation
Waste and soil compaction
Waste soil interaction
- 13 -

-------
the differences may not always be as pronounced, especially if releases to
the atmosphere from the land-based option must be modeled.
Careful consideration must be given to modeling exposures via different
pathways (see Section 4.5) to critical population groups (CPG) and to the
committed dose. In each case, such estimates must be prepared for public and
occupational health impact assessment. Finally, occupational health impacts
arising from non-radiologic events must also be evaluated.
4.2.2
Inventory/Source Term
LLW is defined as radioactive waste containing source materials, special
nuclear materials, or by-product materials that are acceptable for disposal
in an NRC licensed land facility. [U.S. Nuclear Regulatory Commission, 1982]
Specifically excluded in this definition are high-level waste, spent nuclear
fuel, by-product material specified as uranium or thorium tailings and waste,
and transuranic (TRU) wastes [waste material contaminated with alpha-emitting
radionuclides with atomic numbers greater than 92 and half-lives greater than
20 years in concentrations exceeding 100 nCi/g (3700 Bq/g)]. However,
retrievable inventories of TRU stored since 1970 are currently being sampled
to characterize these wastes. It is estimated that 38 percent (35,800 m3)
may be design~ted as LLW.[Oak Ridge National Laboratory, 1988]
Commercial nuclear power plants generate about 50% by volume of the
waste shipped to commercial disposal sites; another 10% is produced by
commercial nuclear fuel processors (UF 6 conversion, uranium enrichment and
fuel fabrication plants). Routine wastes consist of spent resins, evaporator
bottoms, filter sludges, dry compressible waste and contaminated plant
hardware. Periodic disposal of small amounts of high-activity irradiated
components is considered non-routine waste. Utilities account for about 50%
of NRC Class A commercial LLW and about 90% of the comm"ercial Class Band
Class C waste.
The remainder of commercial LLW is produced by industry and
institutions, including research laboratories, hospitals and medical
laboratories, non-DOE governmental facilities, and universities. Industrial
and institutional wastes may be biomedical (animal carcasses, tissue samples)
or non-biological (compacted trash, absorbed liquids, contaminated hardware).
- 14 -

-------
More than 17,000 licenses authorizing handling and use
materials have been issued by the NRC or by Agreement States.
of
radioactive
Low-level waste is also generated by DOE facilities through
uranium
enrichment operations, the naval nuclear propulsion program, and defense and
research and development activities. By volume, DOE LLW consists mainly of
dry solids and decontamination debris; it is disposed of at DOE sites.
Tables 2 and 3 show volumes and activity of LLW from the commercial and
DOE sectors from 1984-1987. Table 4 lists the principal radionuc1ides in the
waste produced by each of these sectors. [EG&G, 1987; Oak Ridge National
Laboratory, 1988]
LLW also results from environmental remedial action projects.
Programs
at DOE sites include the Formerly Utilized Sites Remedial Action proj ect
(FUSRAP)
and
the
DOE
Environmental
Restoration
(ER)
and
Defense
Decontamination and Decommissioning (D&D) Programs.
Former DOE sites now
classified and administered as civilian projects are the responsibility of
the Surplus Facilities Management Project (SFMP). Projected volumes and
characteristics of waste from these activities are currently being estimated.
In general they are expected to be high volume and low specific activity LLW.
4.2.3
Waste Processing Technology
Both physical
and chemical
properties
of
the
waste
are
important
criteria used in the selection of appropriate waste processing technology.
Basic physical properties of the as-generated waste stream (e.g., solid vs.
liquid) influence the decision on which treatment option is chosen and the
stability of the waste after disposal. For example, combustible solids that
are incinerated with resultant ash solidified in cement or other binder will
be more stable than similar waste that is simply compacted and shipped for
disposal. Chemical properties also affect selection of treatment options
(e.g., compatibility of waste and binder materials) and can have a
significant impact on disposal performance parameters including leachability
of radionuclides from the waste form and sorption characteristics of the
trench, unsaturated and saturated zone soils. Chemical composition and ionic
form can affect waste solubility and precipitation reactions in groundwater.
Presence of chelating agents used in cleaning and decontamination processes
can alter sorption interactions between waste and soiL
Volume reduction
- 15 -

-------
Table 2. Quantities (m3) of LLW generated in the U. S. by source type,
1984-1987 [Oak Ridge National Laboratory, 1988].  
Source 1984 1985 1986 1987
Commercial 69374 75909 51112 52233
Utility 42787 43260 29301 26602
Institutional 4398 4712 3780 5187
Industrial 22189 27937 18031 20444
DOE/Defense 90600 121200 97000 99500
- 16 -

-------
Table 3. Activity levels (103 Ci)l of LLW generated in the u.s. by
source type, 1984-1987 [Oak Ridge National Laboratory. 1988].
Source 1984 1985 1986 1987
Commercial 600.9 749.0 233.7 269.6"
Utility 441.3 582.5 170.5 219.8
Institutional  8.1 5.0 7.1
Industrial  158.3 58.2 42.7
DOE/Defense 2053.0 1009.0 772.0 2750.0
1 1 Ci - 3.7x1010 Bq.   
- 17 -

-------
Table 4. Typical radionuclides in low-level waste [Oak Ridge National
Laboratory, 1988].
Source
*
Radionuclides
Commercial
Utility
Institutional
Co-58. Cr-51. Mn-54. Cs-l34, Zn-65. Cs-13?, Co-60,
H-3. Ni-63, Fe-55, 1~131

H-3. C-14. 1-125, P-32. 5-35. 1-131, Cs-137, B-137m,
Cr-51, U-238, Co-60, Mo-90, Fe-55, Ir-192, Na-22
Industrial
H-3, P-32. Cs-137, Ba-137m, 5-35, Co-60, U-238,
Th-232. Ta-182, C-14. Ir-192, Sr-90, Y-90
DOE/Defense
U-238. Th-234, Pa-234m, Pu-241, Co-58, Mn-54,
Cs-137. Ba-137m, Ce-144, Pr-l44
*
Listed in order by concentraticn or contribution to total activity.
Sources: ORNL 1988, EG&G 1987.
- 18 -

-------
techniques can affect physical and chemical composition of the waste and
increase activity concentrations to the extent that waste classifications may
be subj ect to change. For example, large volumes of low concentration
contaminated liquids can be reduced by evaporation leaving a more highly
concentrated sludge residue (evaporator bottoms). The advantages inherent in
volume reduction must be weighed against potential waste-binder compatibility
problems and increased exposures associated with higher radioactivity levels,
introduced by the secondary waste stream.
Characteristics
of
treated
waste
must
also
be
considered
in
the
performance assessment.. For example, if wastes are encapsulated in a solid
matrix, does the process physically bind the waste (as in thermoplastic
materials such as bitumen or thermosetting polymers such as vinyl-ester
styrene); or is a chemical bond formed (as in the hydration reaction of
cement) that reduces radionuc1ide mobility? Leaching properties of
cementitious waste forms are highly isotope- and species-dependent, whereas
those of solidification materials
such as
bitumen are not.
For
those
isotopes that are chemically bound, to what extent and under what conditions
are the reactions reversible? In addition to decreasing radionuc1ide
mobility, solidification of LLY improve~_disposa1 site stability by reducing
the potential for slumping of backfill materials caused by biodegradation.
Thus, solidified waste form performance is an important p~~ameter in
projecting potential health impacts from land disposal of LLY. Properties
that can impact waste form structural stability include compressive strength,
and resistance to freeze-thaw cycling and biodegradation damage. Treated
and/or solidified waste .is generally contained in a waste package prior to
disposal. Package behavior (i.e., corrosion resistance) can also affect
overall disposal site performance. However, the most common packaging
material (mild steel drums) corrode relatively quickly, and consequently some
assessments do not. take credit for the expected useful life of packaging
containers.
Many types of LLY treatment and packaging.options are currently in use.
Potential treatment options for primary waste streams that are either
currently in use or are being developed are summarized in Table 5.[Trig1~o,
1981] Solidification and packaging options for primary and secondary waste
streams are included in Table 6.
- 19 -

-------
Table 5.
Potential LLY treatment options [Triglio, 1981].
Primary
waste stream
Treatment process
Secondary waste stream product
SOLIDS
Combustibles:
Controlled air incineration
Cyclone incineration
Fluidized-bed incineration
Rotary kiln incineration
Agitated hearth incineration
Controlled pyrolysis
Molten salt combustion
Acid digestion
Non-Combustibles:
Compaction
Shredding

Melting-casting
equipment)

Dissolution

Decontamination (contaminated
equipment)
(discarded
LIQutDS
Fil tration
Centrifugation
Ion exchange
Membrane technology
Evaporation
Calcination
non-combustible, highly refractory oxide
non-combustible, highly refractory oxide
non-combustible, refractory oxide 
non-combustible, refractory oxide 
non-combustible, refractory oxide 
non-combustible, refractory oxide 
non-combustible salt-ash or an oxide if
sal t is leached    
non-combustible sulfates and oxides
volume reduced solids
volume reduced solids
radioactive ingot and process slag
acidic slurries

decon solutions (e.g., alkaline
permanganate, mineral and organic acids,
detergents, and chelates)
sludge, spent filtration media (e.g.,
sand, diatomaceous earth, carbon,
cartridges, pre-coat cartridges)

concentrated sludge

spent ion exchange resins (powdered or
bead)
moderately concentrated blowdown
effluents

highly concentrated condensate slurries

dry solid, non-fused oxides
- 20 -

-------
Table 6.
Potential LLW solidification and packaging alternatives.
Technology/Material
Description
Solidification:
Cement
hydration reaction with
waste forms solid matrix
aqueous
Bitumen
thermoplastic polymer; melted,
mixed with waste to form
homogeneous mixture; cooled to
form solid
Polymer Concrete
(e.g., vinyl-ester styrene)
liquid monomer emulsified with
waste, polymerized by chemical
initiators to form solid
Glass
high temperature process to
calcine and/or incinerate waste
and incorporate in glass matrix
(currently R & D for HLW only)
Polymer-modified Gypsum Cement
hydrated modified CaS04 cement
mixed with waste to form solid
Polyethylene
thermoplastic polymer; melted,
mixed with waste to from
homogeneous mixture; cooled to
form solid
Packaging:
Mild steel
55 gal. drums
liners; suitable
compacted waste;
life «50 yrs)
or large volume
for solidified or
limited expected
Wood/cardboard boxes
Suitable only for low
(Class A). dry solid
biodegradable, subject to
degradation irt soil
activity
waste;
rapid
High-integrity containers
Specially designed alloy or
polymeric containers as per NRC
specifications; minimum lifetime
des~gn goal of 300 yrs
- 21 -

-------
4.2.4
Waste Disposal Options - Land
Passage of the LLRWPAA. implementation of NRC's disposal site
performance criteria and the development of new regional and state disposal
facilities have created interest in alternative [(i.e.. other than
conventional shallow land burial (SLB)] disposal technologies. [U.S. Nuclear
Regulatory Commission, 1982; u.s. Congress. 1986] Since many regional
compacts and individual states have already banned the use of SLB. it is
clear that alternative disposal methods will playa major role in the
planning of future land-based disposal of LLW. The ability to address
alternative disposal. options is therefore an important element in selecting a
performance assessment model.
The LLRWPAA required the NRC, in consultation with the States and-othar
interested parties. to identify methods for the disposal of LLW other than
SLB (Table 7). and to establish and publish technical guidance regarding
licensing of facilities that use such methods. In their Branch Technical
Position Statement on Licensing Alternative Methods of Disposal for LLW. NRC
defined alternative disposal methods as "disposal facility designs or
disposal concepts which incorporate engineered barriers or structures, or
otherwise differ from the past and present methods of near-surface land
disposal of LLW by shallow land burial." [U. S. Nuclear Regulatory Commission,
1986]
SLB disposal facilities generally consist of long. unlined trenches
about 15 m wide and 10 m deep. After excavation. trenches are filled with
waste containers (either stacked or randomly placed), backfilled to the
surface, capped with a mound of soil, and seeded to prevent erosion.
Radionuclide migration is slowed only by natural processes such as the action
of the sloped cap to divert excessive precipitation and the sorptive capacity
of indigenous elements in the unsaturated and saturated zones. Several
modifications to basic SLB technology have been proposed that also rely on
natural barriers to isolate radionuclides from the accessible environment.
These include small, unlined trenches; unlined augers and slit trenches which
provide a large ratio of length to diameter or width to reduce surface area
exposed to - water infiltration and decrease potential for plant, animal or
human intrusion; and intermediate depth disposal (15 - 30 m) which features a
thicker cap to reduce permeability, erosion, and the
possibility of
- 12 -

-------
Table 7.
LLY.
Potential alternatives to conventional shallow land disposal of
Modified shallow land burial:

Natural barriers
. Small, unlined
. Unlined auger
. Slit trench
Engineered barriers
. Concrete-lined
. Concrete-lined
trench
trench
slit trench
Alternative methods:
.
.
.
.
.
.
.
.
.
.
Above-ground vault disposal
Below-ground vault disposal
Modular concrete canister disposal
Intermediate depth disposal
Hydrofracture
Deep well injection
Deep geological disposal (mined cavity)
Earth-mounded engineered bunkers
Lined shafts or boreholes
Caissons or pipes
- 23 -

-------
intrusion. Other SLB modifications have been suggested that incorporate
engineered barriers such as improved cap designs using less permeable
materials (e.g., clay or treated concrete); backfill materials with high
sorptive capacity (e.g., natural zeolites) and trench liners of concrete or
other materials. During the operational phase, active controls such as
leachate collection and treatment systems may be used, but after closure it
is assumed that only passive systems are operable.
Alternative disposal technologies rely on engineered structures or on a
combination of natural and engineered barriers. They are designed to reduce
contaminant migratio~, provide increased isolation of waste, and improve
long-term stability of the disposal site. Numerous alternative concepts have
been proposed including: above - ground vaults, below- ground vaults, modular
concrete canisters, earth-mounded concrete bunkers, lined shafts or
boreholes, caissons or pipes, concrete walled trenches, hydrofracture, deep
well injection, and shallow land burial disposal (e.g., mined cavities).
Many of these concepts- are based on disposal techniques currently used or
planned in other countries. Some designs have been proposed or are in use
for temporary storage of LLTJ by agencies in the U. S. and elsewhere. The
current status of five alternative methods is summarized in Table 8.
Below-ground vaults can be constructed of masonry blocks, reinforced
formed or sprayed concrete, fabricated metal or polymers molded in situ.
This design provides a barrier to both unauthorized and inadvertent
intrusion, good structural integrity, and protection against exposure of the
waste due to erosion.
Above-ground vaults can be constructed from the same materials as below-
ground vaults and differ principally by the fact that they are located at or
above grade. Advantages include potential retrievability for site
remediation, reduced susceptibility to flooding and a greater degree of
freedom in siting (since site geology is less critical). Disadvantages
include increased vulnerability to unauthorized access and greater exposure
to erosion.
Modular concrete cannisters consist of large metal or concrete
over-
packs that are filled with LLTJ packages, void spaces filled with cement grout
or sorptive material and then sealed. Canisters can be placed in
- 24 -

-------
Table 8.
1984] .
Status of five LLW alternative disposal methods [Bennett et al.,
Disposal method
Statusa
Location
Below-ground Vaults
S
D

S
S
Above-ground Vaults
S
S
Earth-Mounded Concrete S
Bunkers  
  S
  S
  D
Mined Cavity  D
  R
  R
  R
  S,D
  S,D,
Augered Holes  R
Disposal Test  
  R
  R
  S
  S
  S
  S
Canada, Chalk River National Laboratory
(CRNL). Ontario; shallow vaults
UK, Drigg; below ground, shallow vault

Canada, WNRE, Manitoba; shallow vaults

USA, Oak Ridge National Laboratory (ORNL) ,
Tennessee; shallow vaults
Canada, Ontario Hydro, Bruce Site, Ontario

Canada, New Brunswick Electric Power
Commission Pt Leprau Site, New Brunswick

Canada, Hydro Quebec Gentilly Site,
Quebec
Canada, CRNL, Ontario
Canada, WNRE, Mani toba
France, Centre de la Manche Site

Sweden, Low-level and Intermediate-Level
radioactive wastes
W.Germany, Gorlebon; boreholes in bedded
salt mine floor
USA, Department of Energy (DOE)

USA, Tennessee Valley Authority (TVA)

W.Germany, Asse Salt Mine (radwaste
facility)

W.Germany, Herrfa-Neurode Potassium mine
(hazardous waste facility)
USA, DOE, Nevada, Greater Confinement
W.Germany. Gorlebon, boreholes in bedded
salt mine floor

Canada, AECL, boreholes in glacial till
USA, ORNL, Tennessee

USA, Los AlamQs National Laboratory, New
Mexico
Canada, Ontario Hydro, Bruce Site, Ontario;
"tileholes"
Canada, CRNL, Ontario; "tileholes"
aR - Research; S - Storage; D - Disposal.
- 25 -

-------
conventional SLB sites, modified SLB sites or in combination with any other
engineered disposal sys~em as part of a multi-barrier approach.
Earth-mounded concrete bunkers currently used in France for LLW' and
intermediate-level waste disposal are a hybrid design using both above- and
below-ground construction together with an earthen cap similar to SLB
designs. Higher activity waste is placed in the concrete below-ground vault
aud back-filled, while lower activity waste is stacked above grade and is
covered with a low permeability earthen cap after internal voids are filled.
Lined shafts/boreholes, caissons and pipes are modifications of the
auger disposal concept described above. Addition of engineered materials
provides improved structural integrity and reduced waste-soil interaction.
Concrete walled trenches are
similar
to below- ground vaults but do not
incorporate engineered materials
components resemble SLB design).
for
floor
or
cap
construction
(these
Hydrofracture is a system that was used for LLW disposal by DOE at Oak
Ridge, Tennessee. A waste-cementitious grout mixture was injected under high
pressure into shale, causing it to ~,-acture and provide locations within the
host rock for the mixture to solidify. Deep-well injection involves pumping
liquid waste into favorable geological media, isolated from pathways to the
accessible environment. Due to the highly site-specific geological
requirements associated with these methods and technical difficulties that
have been experienced with pilot scale facilities, these methods are not
considered likely candidates for LLW disposal. Existing limestone or bedded
salt mine cavities have been used for low-level, high-level, and hazardous
waste
disposal
in West
Ge!"many,
Canada,
Sweden,
and,
for
research
and
development purposes, in the U. S .
Deep geological isolation of waste has
many advantages (e. g. ,
long-term stability).
isolation,
reduced likelihood of intrusion,
proven
Performance assessment considerations affected by disposal technology
include selection of appropriate transport and exposure pathways, influence
on infiltration and leach rates, prediction of containment failure time and
extent, and impact on intruder scenarios. With the exception of modeling the
failure of engineered structures and impacts on intruder scenarios, these
additional considerations can be based on engineering experience and
empirical data. Assessment of long-term performance of engineered barriers
- Z6 -

-------
in the disposal environment and resulting impact on waste immobilization,
however, requires analysis of interaction between materials, waste
components, and soil geochemistry and how these factors affect structural
integrity, water infiltration and leachate exfiltration rates. Most disposal
site performance assessment models do not handle these analyses.
Selection of LLW treatment options is closely linked to available
disposal options. Historically, the initial impetus for LLW volume reduction
can be traced to rising disposal costs driven by the limited capacity of
existing disposal facilities and the difficulties associated with siting and
licensing new facilities. More recently, in an attempt to improve the
environmental performance of land-based disposal sites, federal regulations
imposed waste acceptance criteria on site operators. [U.S. Nuclear Regulatory
Commission, 1982] These criteria encourage the practice of segregating waste
according to hazard level by implementing regulations based on waste
classification, and require generators to stabilize more hazardous waste
streams by a combination of treatment and packaging.
Since treatment and packaging of waste can have a major impact on the
mobility of radionuclides in the environment, these parameters must be
examined in the context of overall disposal site performance. The extent to
which treatment/packaging parameters are addressed and the manner in which
they are handled in performance assessment models can affect the accuracy and
level of uncertainty of exposure projections. It should be noted that many
of the available LLW treatment options result in secondary waste streams that
r~quire further treatment, solidification, or packaging prior to disposal.
Properties of waste in its pre-disposal form can vary significantly from
those of as-generated waste. Accordingly, the source terms used in disposal
site performance assessment models should reflect the properties of treated,
packaged waste, as delivered for disposal.
4.2.5
Waste Disposal Options -- Ocean
The LDC represented the first comprehensive international effort to
regulate marine pollution caused by dumping. The LDC classified wastes in
three categories: (i) those prohibited from ocean disposal, including high-
level radioactive waste; (ii) those that may be disposed under controlled
conditions, requiring issuance of special permits (including LLW); and, (iii)
- 27 -

-------
those that may be disposed under general permit requirements. Operational
control requirements were developed, with a strong emphasis placed on siting
issues. Specific site selection criteria [International Atomic Energy
Agency. 1978] include:
.
The chance of recovering the waste by processes such as trawling
shall be minimized;
.
Dumping shall be restricted to those areas of the oceans between
latitudes SOoN and sOoS. The area shall have an average water depth
greater than 4000 m. Recognizing that variations in sea-bed
topography do exist, this restriction should not be interpreted to
exclude those sites within which there are localized areas with water
depths of 3600 m;
.
Sites should be located clear of continental margins and open sea
islands, and not in marginal or inland seas. Nor should they be
situated in known areas of natural phenomena, for example volcanic
activity. that would make the site unsuitable for dumping;
.
The area must be free from known undersea cables currently in use;
.
Areas shall be avoided that have potential seabed resources which may
be exploited either directly by mining or by the harvest of marine
products, or indirectly (e.g., spawning) as feeding grounds for
marine organisms important to man;
.
The number of disposal sites should be strictly limited;
.
The area must be suitable for the convenient conduct of the dumping
operation and so far as possible shall be chosen to avoid the risk of
collision with other traffic during maneuvering and undue
navigational difficulties. The area chosen should be covered by
electronic navigational aids.
.
The dumping site shall be defined by precise coordinates.
to ensure a reasonable operational flexibility, it should
area as small as practicable, but no larger than 104 km2.
have an
In order
Many of the siting requirements described above (e.g., minimum depth,
limiting number of sites) represented improvements over past practices. By
- 28 -

-------
developing a coordinated international ocean disposal siting strategy and
implementing controls on the location and characteristics of new disposal
sites, the LDC aimed to minimize health and environmental consequences from
future sea disposal operations.
In support of the objectives of the LDC, the Organization for Economic
Cooperation and Development (OECD) Council established a Multilateral
Consultation and Surveillance Mechanism for Sea Dumping of Radioactive Waste.
One of the responsibilities of this advisory body is to assess the
suitability of new and existing ocean disposal sites. According to Article
2(a)(iii) of the OECD Council Decision, site assessments must be conducted at
least
every
five
years. [Organization for Economic Cooperation and
As part of this effort, the OECD Nuclear Energy Agency
review of the Northeast Atlantic site which: (i)
Development, 1977]
(NEA) published
a
characterizes its physical, geological and biological aspects,
( 11) reviews
operational factors such as inventories, waste composition, packaging,
etc. ,
(iii) uses the generic International Atomic Energy Agency (IAEA) ocean
disposal model to estimate International Council for Radiologic Protection
(ICRP) exposure doses, and (iv) reviews compliance with the "provisions
established?y the LDC.[Nuclear Energy Agency, 1980]
Under authority granted by the MPRSA [U.S. Congress, 1972], EPA has also
published a set of ocean disposal site selection criteria. [U.S. Environmental
Protection Agency, 1977] General criteria restrict dumping of materials into
the ocean to only those areas that minimize interference with other
activities in the marine environment such as existing fisheries or
shellfisheries,
located beyond
and regions of heavy navigation.
the continental shelf in areas
Sites
such
should be:
(i)
that
temporary
perturbations in water quality and environmental effects return to normal
ambient conditions before reaching any beach, shoreline, marine sanctuary, or
known fishery/shellfishery.
and
(11)
be
11mi ted
in size
to
facilitate
monitoring and identification of adverse effects. Additional, specific site
selection criteria require consideration of the following factors:
.
Geographical position, depth of bottom water, bottom topography. and
distance from coast;
- 29 -

-------
.
Location in relation to breeding, spawning, nursery, feeding,
passage areas of living resources in adult or juvenile phases;
or
.
Location in relation to beaches and other amenity areas;
.
Types and quantities of wastes proposed to be disposed of, and
proposed methods of release, including methods of packing the waste,
if any;
.
Feasibility of surveillance and monitoring;
Dispersal, horizontal transport and vertical mixing characteristics
of the area, including prevailing current direction, and velocity, if
.
any;
.
Existence and effects of current and previous discharges and dumping
in the area (including cumulative effects);
.
Interference with shipping, fishing, recreation, mineral extraction,
desalination, fish and shellfish culture, areas of special scientific
importance and other legitimate uses of the ocean;
.
The existing water quality and ecology of the site as determined by
available data or by trend assessment or baseline surveys;
.
Potentiality for the development or recruitment of nuisance species
in the disposal site;
.
Existence at or in close proximity to the 'site of any significant
natural or cultural features of historical importance.
Another significant change in ocean disposal policy introduced by the
LDC was in the area of design and performance criteria for disposal
packaging. Previous policy assumed that dispersion and dilution of
radioactive
contaminants
is
sufficient
to
adverse
prevent
health
and
environmental impacts. Current policy emphasizes containment of
radionuclides to the extent practically achievable. In their publication on
guidelines for LLY ocean disposal packaging, NEA recommends: [Nuclear Energy
Agency. 1979]
The packages should be designed to ensure containment of the
waste during their handling, transportation, dumping at a water
depth corresponding to the dumping site in use but at any rate
- 30 -

-------
not less than 4000 m, descent to and impact upon the sea floor,
and to minimize to the extent reasonably achievable subsequent
release of radionuclides.
Specifically. NEA guidelines require radioactive waste be either in a
solid or a solidified form that is not permanently buoyant with a specific
gravity of at least 1. 2 to ensure waste descends to and remains on the
seabed. NEA design basis recommendations for ocean disposal packages that
address containers, package contents, the package as a whole, and pressure
equalization devices are summarized in Table 9.
U. S. regulations issued by EPA [U. S. Environmental Protection Agency,
1977] go beyond NEA Guidelines in requiring containment of radioactive
contaminants. For example, 40 CFR 227.7 requires all LLW' be containerized
and meet the following conditions:
.
the materials to be disposed of decay, decompose or radiodecay to
environmentally innocuous materials within the life expectancy of the
containers and/or their inert matrix;
.
materials to be dumped are present in such quantities and are of such
nature
that only short-term
localized adverse
effects will occur
should the containers rupture at any time;
.
containers are dumped at depths and locations where they will cause
no threat to navigation, fishing, shor~lines, or beaches.
performance
criteria
in
support
of
number of specific waste package
existing regulations. [Colombo and
A recent
EPA report
recommends a
Fuhrmann,
1988]
These
packaging
criteria
are based on a multibarrier
approach consisting of components to contain and isolate radioactive elements
from the accessible environment.
The containment system consists of the
solidified waste form and its container. Together these components make up
the waste package, and each component contributes to its overall performance.
For
example,
the
container
protects
the
waste
form
from
erosion
and
degradation while the waste form provides structural integrity for the
package under the hydrostatic pressure of disposal. A container lifetime. of
200 years or 10 half-lives of the longest lived nuclide (whichever is less)
is proposed.
The isolation system includes the waste package in conjunction
- 31 -

-------
Table 9.
NEA design basis for ocean disposal packages [Nuclear Energy Agency, 1979].
Container

Containers should be designed to meet the following functions:
. act as a receptacle for receiving conditioned radioactive waste at the time of
preparation of package;
. provide physical integrity against impact and/or damage during handling and
transport, taking into account the severity of expected conditions at sea;
. provide a barrier to prevent the spread of radioactive contamination;
. contribute to shielding and .reduction of radiation levels at outer surface of

package;
. facilitate handling operations by its shape and/or
Container should be of good quality material, (e.g.,
suitable size, taking into account weight limitations.
provide strength, shielding or protection from corrosive
configuration.
metal or concrete) and of
Containers may be lined to
materials.
Content

Conditioned radioactive waste when placed in the container should become an integral
part of the sea dumping package. In this way, if the containers are damaged or
deteriorate, release and spreading of radioactive materials would be slowed. Waste
may be solidified with or without internal reinforcement, or may be packaged as an
assembly of several components if void spaces are minimized.

Packa~ing

Minimum packaging requirements include:
. Specific gravity of ~ 1. 2. If materials with lower specific gravity are
contained, they should be conditioned to prevent floating to the surface if
container integrity is breached;
. Ability to withstand external sea pressure exerted during descent to sea floor.
If not strong to resist deformation, pressure equalization device should be
installed; .
. Container should be closed with a suitable cap of proper material and
dimensions that is an integral part of the package and contributes to
structural stability;
. Package should provide inherent shielding so that radiation levels are kept
within acceptable limits; .
. Package should be strong enough to resist damage from handling and transport
operations. Should be compatible with lifting and handling equipment.
. Package should be strong enough to remain intact upon impact with seabed and
for a period of time thereafter to minimize to the extent reasonably achievable
the radioactivity that might be released.
Pressure Eaualization

Pressure equalization devices may consist of discs, plugs, seals, one-way or no.
return valves or other devices which are activated or ruptured by a pressure
differential across them, so long as they permit no escape of material from the
container nor result in increased radiation exposure levels.

Tubes, valves and rupture devices should be designed to prevent ingress of water
during storage and transport. They should be positioned to prevent them from being
damaged during storage, handling, transport and dumping. Vent tubes should be
oriente? to minimize ~xternalbradiation. If voids are not interconnected, a means of
connect1on or appropr1ate num er of pressure equalization devices should be provided.
- 32 -

-------
with natural barriers such as sediments and the residence time of water at
the depth of disposal. For example, a recent study of the adsorption
capacity of sediment collected from the Atlantic 3,800 m disposal site
indicated that 90 % of the Cs-137 is adsorbed under well-mixed conditions at
SoC. In addition, modeling studies of dissolved tracers released -at the
seabed indicate that residence time for the western basin of the North
Atlantic at 5000 m is about 110 years. Waste package performance criteria
proposed for use by EPA in evaluating permit applications for ocean disposal
of radioactive waste are presented in Table 10.
4.3
Consequence Modeling
The second task consists of developing or applying mathematical ~ode1s
to calculate
environmental
transport,
human dosimetry
and response,
and
integrating each scenario' s consequence models into an overall, coherent
analysis. These models provide a basis for rough assessment which can be
expressed
as
mathematical
equations
that
can
be
solved
directly
by
conventional mathematical or analytical methods. Models are, 'however, very
idealized representations of the natural environment which cannot include all
processes that are important. Thus, their value is that they can identify
distances and times over which concentrations vary, they may be useful in
establishing the largest concentrations that could occur, and they may be
useful guides in finding the most important prcesses by comparing the results
of including one or another in the ca1cu1ations.[Brookhaven National
Laboratory, 1986]
Because of the complexity of these evaluations, the calculations are
often done in a modularized manner, where each module describes a unique
process, event, or environment.
The modular approach to risk assessment is
shown in Figure 3. In -this
included for land disposal:
context,
the following modules are usually
.
Source-term (i.e., inventory and radionuclide release);
.
Groundwater regional and local flow;
.
Radionuclide transport in geosphere and biosphere;
.
Human dose commitments, and effects.
- 33 -

-------
Table 10. EPA proposed waste package performance criteria for ocean disposal
of LLW [Colombo and Fuhramm, 1988].
Criterion

Package should have adequate density
to ensure sinking to the seabed.
Package shouid be designed to remain
intact upon impact with sea surface
and seabed.
Container should be capable of main-
taining its contents until nuclides
environment decay to acceptable limits.
Liquid radioactive waste should be
immobilized by suitable solidification
gents.
Buoyant material should be excluded or
treated to preclude its movement or sep-
aration from waste form.
Waste package should be able to with-
stand hydrostatic pressure encountered
during and after descent to seabed.
Leach rate of waste form should be as
low as reasonably achievable.
Particulate waste should be rendered
non-dipersible.
Free radioactive gases should be
prohibited from ocean disposal
- 34 -
Snecification

Specific gravity of waste package
should not be < 1.2.
Package should maintain integrity
on impact with sea and ocean floor
at velocities of 10 m/s
Waste container should have expect-
ed life-time in deep-sea
of 200 yrs or 10 half-lives of
longest lived nuclide, whichever is
less.

Liquid wastes should be solidified
to form a homogeneous, monolithic,
free-standing solid containing no
more than 0.5% free or unbound
liquid by volume of waste form.

Buoyant materials should be treated
to form a homogeneous free-standing
mono-lithic solid having a spec.
gravity ~ 1. 2.
Triaxial compressive strength (or 4
times uniaxial compressive
strength) should be 25% greater
than pressure encounte~ed at dis-
posal depth (125 kg/cm uniaxial
compo strength for disposal at
4000m)
Leach rate
guidelines
Leach Test
seawater.
should be $ regulatory
as measured by ANS 16.1
for leaching in
Particulate waste (ash, powders,
etc) should be immobilized by a
suitable solidification agent
to form a homogeneous monolithic
free-standing solid.
No radioactive gaseous waste should
be accepted for ocean disposal
unless they have been immobilized
into stable waste forms such that
waste package pressure does not
exceed atmospheric pressure.

-------
Table 10.
cont.
Mixed wastes, which contain hazardous
constituent should not be disposed of
at a LLW ocean disposal site.
Waste should be physically and chem-
ically compatible with the solidifi=
cation agent.
Wastes that
prohibited,
nants in 40
disposed at
site.
contain constituents
as other than contami-
CFR 227.6 should not be
a LLW ocean disposal
Waste forms should retain their
structural stability after
immersion in seawater for 180 days.
- 35 -

-------
IUIOOO 10
w..,.
~rDP8r'I.-
r---------
,
I
I
I
,
I
I
I
I
I
,
I
I
I
I
I
I
I
I
I
I
I
I
,
I
I
I
I
,
I
"
- - - - F..elb.cll ""'Iy

. "I r.elIDnucllcl. -_c..
In ,". .......
EPA Aul.
.OCAFI".13
Conlalnm.nt
EPA Aul.
.OCAF1".1.
Ground.al.r
Proleclion
(1000 yr)
NRC Aule
10CFA60.113
Ground..'er
Tr....' Time
(1000 yr)
EPA Rule
.OCFA1".'5
Indh,1do81
Prolect.
(1000 p)
NRC Aul.
10CFAIO.113
PaclI.ge Lil.
Ae"..e Aal.
(10.000 yr)
Figure 3.
Modules used for land-based risk assessment purposes.
- 36 -
DI!;POSAL
SYS1[M & A[GIOMAl
CHAAACTtAIZA'ION


+
SCENAmo
D£VIELOPIltNT

I
1
CONSEOUENCE
ANALYSIS


I
.
SENSITlV'TY &
UNCEATAINTY
ANAL VSIS

I

AEGULATORY
COMPLIANCE
ASSESSMENT
t

-------
Similarly. for ocean disposal options, these modules are:
.
Source-term (i.e., inventory and radionuclide release);
.
Basin-scale, ocean circulation;
.
Regional-scale ocean circulation;
.
Radionuclide transport;
.
Human dose commitments, and effects.
Because of EPA and NRC regulatory criteria, much of the compliance
evaluation efforts use results generated before the last module, human dose
commitments.
For example,
the containment requirements included in EPA' s
proposed 40 CFR 191.13 use results of radionuc1ide transport in the geosphere
to' calculate discharge to the accessible environment. For this reason, a
modular approach to radionuclide transport for the geosphere and biosphere is
convenient.
In land-based option analyses, the primary pathways (Table 11) that can
move the material from its buried position to some altered position or state
where it is accessible to man are water and wind.
Most of the attempts to
model this situation have given a greater emphasis to waterborne movement.
This is perhaps due to the earlier attempts to model ~ovement from the deep
sites suitable for the disposal of high-level waste. For that case,
waterborne movement is the only real possibility, if we neglect volcanic
activity.
In this context, regional groundwater flow modules simulate groundwater
flow fields in a large region about the controlled area of the repository.
The region should extend to natural boundaries or far enough that boundary
conditions do not effect transport calculations on the time scale of
interest. The regional flow fields then determine boundary conditions for
the smaller scale model, the local groundwater flow model which then
simulates flow in the repository-controlled area. The local model determines
boundary conditions for the land disposal model if groundwater enters the
site. Compliance with the NRC engineered barrier requirements can then be
evaluated.
If the scenario involves transient flow conditions, a coupled
groundwater flow and transport module may be required.
Then compliance with
the proposed EPA containment and groundwater protection requirements can be
- 37 -

-------
Table 11. Pathways and modes of exposure from the land disposal option
[International Atomic Energy Agency, 1984a].
TRANSPORT PATHWAYS
Hydrologic
Leaching
Deep seepagea
Groundwater
Surface water
Direct human consequences (ingestion, immersion)
Indirect consequences (through fool chain /ingestion)
Atmospheric
Trench erosion
Surface contamination
Suspension
Deposition
Direct human consequences {inhalation, immersion)
Indirect consequences (through food chain/ingestion)
Food
Chain
Drinking water
Crops
Meat
Milk
Aquatic foods
EXPOSURE SCENARIOS
Operations Phase
Routine release
Accidental release
post-Closure Phase
routine release
Accidental releaseb
Groundwater
Leachate accumulation
Intruder
Drilling
Construction
Discovery
Agriculture
Exposed waste
aCovers exposure to the public, not disposal site workers.
bRefers to post-closure impacts resulting from operational spills.
- 38 -

-------
evaluated. To evaluate compliance with the praposed EPA individual
protection requirements requires the last two modules, biosphere transport
and dosimetry.
A great deal of effort has been spent in considering the rate at which
various radionuclides can be leached from land-disposed wastes, and then
transferred to ground water. Once part of the groundwater system,
radionuclides
are
transported
food
crops
or
the
indirect
at various rates depending upon their
and solid phases. Eventually, dose can be
contaminated potable water, the use of
with the subsequent direct contamination of
contamination of food crops via soil
partitioning between the aqueous
delivered to man by drinking
contaminated water for irrigation
contamination, or the consumption of aquatic foods that have accumulated the
radioactive contaminants.
Another pathway of great significance is that of surface erosion. Over
geologic time, the erosion of soil by this process is dramatic and its
effects are easily seen in undisturbed landscapes. After any overburden is
removed, the radioactive contaminants can be removed as well, and the dose-
to-man pathways would be the same as those above.
In this case, however, the
source of the contaminated water would be surface rather than ground. The
amount of material transported by this water erosional pathway is strongly
dependent upon episodic events, such as severe rainstorms.
Despite such well known occurrences as the "dust bowl" of the 1930s, the
erosional power of the wind is not often recognized and treated in models.
In areas of the western United States with low rainfall, the erosional
effects of wind are readily seen. Often, it is not possible to separate the
erosional effects of wind and rain, but both can be effective in soil
removal, especially during episodic events. Material that is removed by wind
will be deposited, either on the ocean surface, directly on food crops, or on
soil. In the latter case, subsequent contamination of foodstuffs may occur.
While the material is suspended, it can also be inhaled by man. And, when
the material is deposited, it will be on the surface and can deliver dose via
the external gamma-exposure pathway.
In
the ocean, human dose commitments are usually the performance
but it is still convenient to separate radionuclide transport from
measure,
human dose commitment and effects modules.
In the ocean, the basin, e.g., N.
- 39 -

-------
Atlantic, replaces the regional scale groundwater flow model. The
continental shelf is the natural boundary. The northern and southern
boundaries must be open to allow Norwegian Sea Water to feed the deep water
formation in the northeast at the Iceland overflow and to simulate the
connection with the equatorial circulation and the Antarctic.
to
resulting
options involve linking codes of
of a release from a deep bottom
dose commitments and effects.
Risk assessments for ocean disposal
different scales to trace the evolution
source
and
estimate
human
Required codes
include
ocean circulation
modules
for
several
scales,
a
simplified transport module
such as
a box model
that uses
a
simulated
circulation as input, biological pathways, human dose commitment, and effects
models. Risk assessment of ocean disposal options requires a systems
approach to assimilate observed data and known features of the ocean into
model simulations of the dispersal of radionuclides from deep bottom sources
to local, basin or global scales. For ocean disposal, an example system
would comprise source, regional, basin, box model (transport/particle), and
pathways/dose/effects modules.
Risk assessments for ocean options are site
specific so meshes for this hierarchical series of model scales must be
generated for each case. Given a specific site and a mesh for the bottom
boundary layer, regional, and basin scales for that site, the ocean
circulation can be simulated. A source is introduced and dispersal through
at least the basin and perhaps the global ocean must be computed. Particle
and .radionuclide transport equations can be coupled with hydrodynamic
equations, but this approach is computationally difficult.. To make the
problem tractable, previous risk assessment programs have used so-called box
models as an interim fix while better numerical approaches were developed. A
box model is a highly coarsened version of the basin finite-difference grid
with nested boxes over the site to account for the bottom boundary layer and
regional scales. Box models require ocean circulation as an input. The
credibility of the resulting radionuclide concentrations
completely dependent on the input circulation.
in each box
is
In the ocean, the exposure pathways to man (Table 12) may include:
consumption of surface water fish, mid-depth water fish and deepwater fish;
consumption of seaweed, molluscs, crustacea, and plankton; consumption of
salt and desalinated seawater; inhalation of suspended airborne sediments and
- 40 -

-------
Table 12. Pathways and modes of exposure for ocean disposal
[International Atomic Energy Agen~y, 1984b).
Pathway
Mode of exposure
Actual pathways
Surface fish consumption
Mid-depth fish consumption
Crustacea consumption
Mollusc consumption
Seaweed consumption
Salt consumption
Desalinated sea water consumption
Ingestion
Suspended airborne sediments
Marine aerosols
Inhalation
Boating
Swimming
Beach sediments
Deep-sea mining
External irradiation
External irradiation/inhalation
Hypothetical pathways
Deep-sea consumption
Plankton consumption
Ingestion
- 41 -

-------
marine aerosols; external irradiation during sailing, swimming and sunbathing
at a beach; external radiation as a result of mining of minerals from the
seabed. Some of these pathways are regarded as unimportant because they do
not exist now or are unlikely to result in meaureab1e doses.
In the process of estimating the risks associated with different
options, there are other important analytic issues which must be also
examined. These include explicit description and estimation of model
uncertainty and matching of model sophistication to the decision problem and
to the natural attributes of the system to be modeled. Since any estimate
produced by a model contains inherent uncertainty due to various sources
including misspecification or oversimplification of models and simple lack of
knowledge, characterization of the size and distributional form of the
es~imate uncertainty is critical. Of the appropriate techniques for
producing uncertainty estimates, Monte Carlo analysis has been widely used
for three major reasons. First, Monte Carlo analysis creates a mapping from
input to output that can be studied by a variety of techniques (e.g.,
scatterp1ots, distribution functions, regression analysis, partial
correlation analysis) £or essentially any input or output variable. Unlike
differential analysis and response surface methodology, this mapping does not
smooth
and
obscure
discontinuities
and
transitions
between
regimes
of
behavior, i.e., modules. Second, Monte Carlo analysis can accommodate large
uncertainties and discontinuities that occur between linked codes. Although
the analysis
is
complicated by
these
factors,
it
is
superior
to
other
techniques when such complications exist. [Helton et a1., 1988]
Third, Monte
Carlo sampling can include variables with wide ranges and incorporate
correlations between variables. The ability to adapt Monte Carlo algorithms
to the larger land- and ocean-based models requires further exploration.
Finally questions about model complexity must be addressed.
There is
much debate surrounding this
from the NRC state:
issue.
In this context, Starmer et a1.
[1988]
"Modeling must be defensible. The most suitable model will be that
which is consistent with the modeling objectives and easiest to use
considering the complexity of the system and the data which can
practically be obtained from the site and intended facility. The
model should be verified for appropriateness of 1.
app l.cation,
- 42. -

-------
validi ty of assumptions, accuracy of algori thms , and
representativeness of input data. A critical consideration for the
user will be the adequacy of the data available and uncertainty
associated with the data.
Generally, more complex models require
more abundant and detailed data, while less sophisticated models
rely more on simplifying assumptions and more generalized data.
Where data is inadequate for complex models, there may be a
temptation to use approximations based on assumptions. In this
case, a more complex model provides no more support than does a
simple systems model."
- 43 -

-------
5
MODEL EVALUATION
5.1
Model Identification, Selection and Evaluation
Mathematical models are used to characterize health risks associated
with LLW disposal on land and in the ocean.
Calculating such estimates
requires evaluation of package contents, decay and release ; transport and
dispersion through the envir~nment; human exposures via food chains;
calculation of doses to workers and the public; and, estimation of health
effects.
In this - context,
a wide range of models have been developed to
cover individual components of this process, for example, air transport and
dispersion, groundwater flow or surface transport. Others integrate two or
more areas,
e.g. ,
air or water
transport combined with food chain, human
dosimetry and response models.
Eventually, a compatible set of pathway models must be used to evaluate
health hazards from LLW.
Since
there is a very large number of single
component and integrated models used for radiation-related risk assessments
(see for example [EG&G, 1985]), the first task of this effort was to identify
a subset of models to be reviewed.
Actual selection of models was based on
the combined judgment and experience of the research team and the EPA project
staff. Strong candidates were those models that have been used previously by
EPA, DOE, NRC, NEA/OECD and IAEA. Models selected for review by this process
are listed in Table 13.
The model evaluation process itself, began with establishment of formal
evaluation criteria (Table 14).
The criteria can be divided into three major
groups:
include
Administrative, Technical, and Scientific.
Administrative criteria
such
points
as
availability
of
documentation
and
computer
requirements.
Technical issues of concern focus on such points
as peer
review and availability of results from sensitivity analyses. Finally, the
last criterion Scientific includes such important considerations as
validation of model outputs against actual field data. The criteria are
applicable to all models and were developed to provide a -standardized
evaluation approach across models and media.. The criteria are qualitative
and are not meant ~or quantitative scoring of models.
The
reviews
were
based
on
model
documentation,
publications,
and
previous experience with these or similar models.
Time and resources did not
- 45 -

-------
Table 13. Models for evaluating health hazards associated with land- and
ocean-based disposal.
Media Model 
Land PRESTO 
Land IMPACTS 
Land BARRIER 
Land GEOTOX 
Ocean MARK A 
Ocean MARINRAD 
Ocean Bryan-Semtner-Cox 
Ocean Sandia Ocean Modeling System
Ocean Holland 
Ocean Harvard 
Ocean NRPB9l-Box 
Sponsor
U.S. Environmental Protection Agency
U.S. Nuclear Regulatory Commission
Electric Power Research Institute
U.S. Department of Defense
Nuclear Energy Agency
Nuclear Energy Agency
National Oceanographic and Atmospheric
Administration
U.S. Department of Energy
National Science Foundation and Office
of Naval Research
Office of Naval Research
U.K. National Radiological Protection
Board
- 46 -

-------
Table 14. Model evaluation criteria for performance assessment models to compare
'health hazards from land and ocean disposal of low-level radioactive wastes.
Administrative:

Documentation - Is documentation available?
for implementing and reviewing code?

Hardware Requirements Is model designed to run on a mainframe or desktop
computer?

Application - How long does it take to run a single problem? Can it be run in a
batch-oriented mode?
Does it include sufficient information
Level of Expertise Required - What level of expertise is needed to implement the
model?

Technical:
Peer Review - Has model ~dergone independent peer review?

Verification - Have models undergone testing to verify mathematical computations?

Uncertainty - Does the model propagate input parameter uncertainty to calculate
resulting uncertainty in the calculated outputs?

Sensitivity - Have sensitivity analyses been performed to determine the relative
importance of individual input parameters on the overall assessment?

Required Input - Are input data readily available? Are data generic or site
specific? Can model parameters be modified by the user? Is code structured in
modular subroutines so that pieces of the code can be updated if improvements
become available?

Output - How are output data presented? Does the model estimate human health
effects or exposed dose; cumulative or maximum effects? What time frame is
covered; what incremental time steps are used?
Source Term - Does the model accommodate a representative inventory of waste types
and isotopes? Are data fixed or supplied by user input?
Scenarios - Are both accident and routine release scenarios included?
scenarios sufficiently define potential event?

Relationship to Regulatory Standards - Are model outputs directly relevant to
existing regulations?
Do the
Scientific

Theory - Are theoretical bases for each model component based on state-of-the-art
information?
Validation - Has model been validated,
compared with actual disposal site data?

Treatment of Radioactive Decay Products Is the production of radioactive
daughter products considered in the source term?

Underlying Assumptions Are assumptions explicitly stated, complete (1. e. ,
adequately define the problem), and credible?

Pathways - Does the model adequately represent all credible pathways to human
exposure?

Dose Conversion and Dose Response - Does the model incorporate dose conversion and
dose response algorithms? What models are used to estimate human exposures and
resJ)onse?
i. e., have performance predictions been
- 47 -

-------
permit full evaluation of the computer codes, nor were test runs of the
models made. Presented below are brief summaries of the completed model
reviews. More complete details on each model are contained in Appendixes B
(Land) and C (Ocean).
5.2
Model Reviews - Land
5.2.1
PRESTO
PRESTO-EPA
(Prediction
of
Effects
from
Trench
Shallow
Radiation
Operations) [Fields et al., 1987] is a suite of computer models developed for
EPA to evaluate possible health effects from shallow land burial disposal of
LLW. It's original purpose was assessment of impacts from varying shallow
land burial disposal scenarios to assist in the development of environmental
standards. Individual models in the PRESTO-EPA family include:
PRESTO-EPA-POP
PRESTO-EPA-DEEP
PRESTO-EPA-CPG
PRESTO-EPA-BRC
PATHRAE-EPA
Estimates cumulative population health effects to local
and regional basin populations from land disposal of
LLW by shallow methods; long-term analyses are modeled
(generally to 10,000 years) [Fields et al., 1987].
Estimates cumulative population health effects to local
and regional basin populations from land disposal of
LLW by deep methods [Rogers and Hung, 1987].
Estimates maximum annual whole-body dose to a critical
population group (CPG) from land disposal of LLW by
shallow or deep methods; dose in maximum year is
determined [Rogers and Hung, 1987b].
Estimates cumulative population health effects to local
and regional basin populations fro.m less restrictive
disposal of BRC wastes by sanitary landfill and
incineration methods [Rogers and Hung, 1987c].

Estimates annual whole-body doses to a critical
population group from less restrictive disposal of BRC
wastes by sanitary landfill and incineration method
[Rogers and Hung, 1987d].
- 48 -

-------
Since PRESTO-EPA-POP was developed first and serves as the basis for
other codes in the PRESTO family, it was the focus of this review. PRESTO-
EPA-POP assesses radionuc1ide ~ransport, resultant exposures, and health
impacts of a LLW disposal site on a static local population for 1000 years
after closure and on the general population residing in the downstream
regional basin for an additional 9000 years. The model simulates leaching of
nuclides from a waste form, hydrological, hydrogeological, and biological
transport, resultant human exposures, and finally assessment of potential
human health effects. Exposure scenarios treated by the model include normal
release (leaching, spills during operations), human intrusion, and site
farming/reclamation. Environmental pathways considered include ground water
transport,
over-land
water
flow,
erosion,
surface
water
dilution,
resuspension, atmospheric transport, deposition, inhalation, and ingestion of
contaminated foods and water. Individual and population doses are
calculated, as well as doses to intruders and farmers. Cumulative health
effects (deaths from cancer) are calculated for the population over the 1000
year period using a life-table approach. Model performance predictions,
however, have not been compared with a~utua1 shallow land disposal site data.
5.2.2
BARRIER
BARRIER [Shuman et a1.,1988] was developed for the Electric Power
Research Institute (EPRI) to assist the nuclear power industry meet federal
and state disposal site performance aSSE:ssment requirements and expedite
licensing of new disposal facilities. It is an integrated model that
estimates: groundwater flow through a disposal facility; radionuc1ide
release; long-term performance and degradation of concrete barriers used in
various engineered storage disposal designs (e.g., below-ground vault, above-
ground vault, modular concrete canister disposal and earth mounded concrete
bunker); transport through an aquifer to the accessible environment;
doses to the CPG. Projected performance of engineered disposal options
compared to a shallow land burial base-case. Inclusion of a module
and
were
that
simulates the performance of concrete structures is a unique feature of this
code. Prediction of engineered barrier performance is a significant addition
to the capabilities of performance assessment modeling, especially in light
of the increasing emphasis on alternative disposal technologies. At the same
time, however, lack of empirical data on concrete behavior over long time
- 49 -

-------
periods introduces new uncertainties in the calculation that must be
addressed. Barrier has not undergone formal peer review. Model performance
has not been compared with actual disposal site performance data.
5.2.3
IMPACTS
IMPACTS [Oztunali et al., 1986] was initially developed for the NRC, to
assist in preparation of shallow land disposal regulations for low-level
radioactive waste. [U.S. Nuclear Regulatory Commission, 1982] It is an
integrated performance, assessment model for comparing potential health
impacts from various land disposal options on a generic basis. Users are
cautioned 'against using the methodology for a site-specific application,
where site-specific models, inventories, disposal options, and environmental
parameters would be required to accurately simulate conditions. Further
caution is advised in interpreting the absolute magnitude of results.
Rather, the model was intended to provide a relative estimate for comparing
potential benefits and costs of a number of potential disposal options.
Users provide information on combination of waste streams to be considered
and regions where they are generated and then select specific waste
processing scenarios, the environmental setting of disposal site, and the
particular combination of disposal technologies to be used. The model
calculates effective dose equivalents (mrems/yr) for 9 organs plus effective
whole body equivalent for each exposure scenario and waste classification.
For chronic exposure scenarios, estimates are given in varying time
increments from 20 to 20,000 years. In addition to IMPACTS, users may select
the following subroutines: INVERSE calculates acceptable nuclide total
activity and/or concentration limits for disposal; ECONOMY calculates
transportation and
routine
operational radiological impacts as well
INTRUDE analyzes radiological impacts to
function of time; and VOLUMES calculates
as
disposal
cost
estimates;
an
inadvertent intruder as a
and
updates region and waste stream dependent annual volume projections. A
separate code (CLASIFY) is used to classify the waste streams and organize
input data for use by IMPACTS. Model performance predictions have not been
compared with actual disposal site performance data.
- 50 -

-------
5.2.4
GEOTOX
GEOTOX [McKone, 1981; McKone and Layton, 1986] is a set of programs used
to calculate time-varying chemical concentrations in multiple environmental
media (e.g., soil and ground water) and to estimate potential human
exposures. The chemical transport component of this model uses landscape
data and physiochemical properties to determine the distribution and
concentration of chemicals among compartments such as air, water and soil.
Environmental concentrations are linked to human exposures and health effects
using an exposure model that accounts for intake through inhalation,
consumption of food and water, and dermal absorption. GEOTOX is intended for
use in public health and environmental risk assessment and risk management
efforts, especially for screening and ranking chemicals according to their
potential risks. In this context, GEOTOX was originally developed for
ranking potential health risks associated with toxic metals and radionuclides
in the global environment.
Recently. the model has been extended to handle
organic chemicals. The GEOTOX program was tested and debugged as part of the
development program. This included verification of the mathematical
computations. GEOTOX offers the user the option of performing a Monte Carlo
analysis, but this option is not available in Version 1. 2. Sensitivity
analyses can be performed. Model performance predictions have not been
compared with actual disposal site performance data.
5.3
Model Reviews - Ocean
Ocean disposal risk assessment requires implementation of regional
(1. e., Harvard, Holland, and Sandia Ocean Modeling Systems), basin (1. e. ,
Byran-Semtner-Cox and Sandia Ocean Modeling System). and box (i.e., MARK A,
NRPB, and MARINRAD) circulation models.
Human dose commitment and effects
models are usually independent of the ocean circulation codes and can be
.often linked with any suitable set of codes. Thus, circulation models which
are critical to any ocean disposal risk assessment are discussed below.
5.3.1
Byran-Semtner-Cox
The Bryan-Semtner-Cox (BSC) model [Cox, 1984] is acclaimed as "the
principal tool for modeling ocean circulation in irregular domains having
realistic
coastlines
and bottom
topography". [Semtner
and
Chervin,
1988]
However, the BSC has been used by English, French, and the German scientists
- 51 -

-------
in connection with the SWG-POTG project, and all but the Germans dropped it
in favor of other models, and the others questioned the validity of the
German results. The BSC does not address boundary layer phenomena, which are
important in environmenta~ questions. It is fully coupled to the "free-
stream" submodel, which models the overlying ocean. The Bryan-Semtner-Cox
model has not been shown to converge as resolution increases. Such
demonstration is a necessary part of verification.
5.3.2
Sandia Ocean Modeling System
Sandia Ocean Modeling System (SOMS) [Marrietta and Simmons, 1988] was
ini dally developed for the U. S . DOE, under the Subseabed Was te Disposal
Program,
to
help
evaluate
deep
ocean
circulations.
SOMS
is
designed
specifically to address bottom boundary layer flows over realistic
topography, but is applicable to the whole range of geophysical flows, from
small lakes to large oceans. SOMS has been applied to both high-level
subseabed [Marrietta and Simmons, 1988] and low-level sea dumping [Nyffeler
and Simmons, 1989]. SOMS has also been applied to the circulation of the
North Atlantic Ocean, continental shelf phenomena, and flow around a
seamount. SOMS is the only deep ocean model designed to address in detail
the dynamics of boundary layers over realistic topography; is far less
dissipative, more accurate, and runs ten times faster for a given resolution
than the world standard BSC model; and is the first ocean model to clearly
demonstrate convergence in a realistic geophysical prototype problem, which
is a fundamental requirement in verifying a model.
5.3.3
Holland
The Holland model has had some noted success in modeling ocean
flows. [Schmitz and Holland, 1986; Holland and Schmitz, 1985] It is limited
to free-ocean calculations. Due to its vorticity-streamfunction formulation,
it is inapplicable to flow over realistic topography, which must be addressed
by primitive equations, or to turbulent boundary layer flows. The main
motivation for using the Holland model is that its numerical efficiency is
aided by not having to resolve internal waves in time. However, this is
largely compensated by needing to solve Poisson equations, one for each layer
at each time step. Important data on model convergence used to determine
model validation have not been published.
- 52 -

-------
5.3.4
Harvard
The Harvard model [Miller et al.., 1983] is part of the ocean prediction
forecasting system presently being put into operation by the U.S. Navy. It's
use is limited to free-ocean calculations with strongly stratified thermal
structures.
Due
to
its
vorticity-streamfunction
formulation,
it
is
inapplicable to flow over realistic topography, which must be addressed by
primitive equations, or to turbulent boundary layer flows. The Harvard model
has some sound theory behind it, and can be used as a teaching and research
tool in spite of its practical limitations (although its value as a research
tool is limited). The Harvard model was the first model used for ocean
forecasting and as such has compared favorably with observed field data by
several investigators.
5.3.5
NRPB9l
NRPB9l has been used for risk assessments of high-level waste disposal
in the subseabed and assessments of low-level sea dumping at the North-East
Atlantic NEA site. NRPB9l is a coarse-grid box model which calculates
radionuclide transport based on an ocean circulation that was subjectively
assembled from an extensive literature review. It is not coupled with a
dynamical circulation model although like all box models, it could be driven
by simulated circulations.
The compartment model covers the area of the
Atlantic Ocean from 50 S to 65 N and uses observed isopycnal surfaces to
define the vertical box structure since mixing and movement in the ocean are
believed to occur principally along isopycnal surfaces. Exchanges with other
oceans are also included since radionuclides entering the Atlantic Ocean wili
eventually disperse throughout the worlds oceans. Box models like NRPB9l
have not been able to adequately reproduce ocean tracer, heat and salinity
distributions.
5.3.6
MARINRAD
MARINRAD was a participant code in a model comparison study conducted by
the NEA CRESP Modeling Task Group. [Mobbs et al., 1986] It has has been used
for risk assessments of high-level waste disposal in the subseabed of both
the Atlantic and Pacific and assessments of low-level sea dumping at the
North-East Atlantic NEA site. MARINRAD is a coarse-grid box model which
calculates radionuclide
transport based on an ocean circulation that was
- 53 -

-------
subjectively assembled from historical literature. It is not coupled with a
dynamical circulation model although like all box models, it could be driven
by simulated circulations. If simulated circulations are used, interbox
transports must be hand-calcuated and entered. Interbox transports for
nested boxes which must be site-specific have to be obtained from subjective
expert estimates. Like other box models, MARINRAD outputs have not compared
favorably with any ocean tracer distribution.
5.3.7
MARK A
Mark A [Robinson and Marietta, 1985; Marrietta and Simmons, 1988; de
Marsily et al., 1989] has been used for risk assessments of high-level waste
disposal in the subseabed for two different locations. The Mark A code was
developed by an NEA working group comprised of national exp~rts from the CEC,
FRG, France, Switzerland, U.K, and the u.S. It has been exercised and
intercompared at Sandia, Ecoles des Mines de Paris, and the CEC-Ispra. The
Mark-A is a coarse-grid box model which calculates radionuclide transport
using an ocean circul'ation as input. Interbox transports must be obtained
via a subsidiary calculation from a high resolution simulation of the basin
circulation. Interbox transports for nested boxes which must be site-
specific have to be otained from local eddy-resolving simulations of the site
region. Like other box models, Mark A has not been shown to compare
favorably with
ocean
tracer distributions.
Further work is required to
reproduce both natural and introduced tracers in the ocean.
- 54 -

-------
6
COMPARING RISKS OF THE LAND AND OCEAN DISPOSAL OPTIONS
Comparing the relative risks of land and ocean disposal of LLW is
exceedingly complicated, not only because there are many important
considerations having high uncertainty. but also because they are extremely
different in their characteristics and impacts. This chapter outlines a
framework for bridging the gap between available information and assessment
capabilities, and needed information and capabilities. It focuses on ways to
expand risk assessments to include a broader range of risks and
characteristics that are important for comparing the land and ocean disposal
options. It describes how quantitative estimates of different risks can be
combined in a common metric by which the options can be compared directly in
spite of their differences in characteristics. It includes discussion of the
impacts that are addressed in some way by existing analytical models,
identifies gaps in the available knowledge and modeling capabilities, and
discusses how capabilities could be expanded. The intent of the chapter is
to provide the necessary background for planning the next stage in the
comparison of the LLW disposal options.
6.1
Characteristics of Concern
Branch, et al. describe the setting in which the relative risks of LLW
disposal options must be evaluated (Figure 4) and the kinds of
characteristics and risks that should be included in the evaluation.[Branch
at a!., 1987]
The basic elements are:
.
The physical environment,
.
Natural ecosystems,
.
Human systems,
.
The evaluation process and characteristics of the implementation, and
.
Outsid~ influences.
The authors envision that Figure 4 is recursive, essentially rolled into
a cylinder matched at the box for the physic-al environment,
evolving risks and changing responses of the system over time.
to represent
- 55 -

-------
~'f'ICA"
IIMIIONIIIft
..
UNO
."11-
CUl&1I
Figure 4.
..~
."".
""1114 8IUI'I\
. ,......
.,..
.~
.......
., ..-'
.......
.,--~
.....
...-
.~
OUTIU DECISIONS
AND "ESOU"CtS
~CT
ouaaen...TICI
.--
......--..
.......-
.~-
. ........ U8
.--'-we
.a---
Integrated environmental/socioeconomic
[Branch et al., 1986].
- 56 -
"~IL
."".
"""",11ft)
.~
.,..
.~
....
.11818
.--
IUI&II
."".
IICOOICICMOI8CII

....-
..... -.-........
.-.
..--.......
. .....
...-
.a.a.w
impact
assessment
""'IICAI.
IN",lIQHllllft
..
LMIt
...".
CI."""
model

-------
Historically, risk assessments of LLW disposal have emphasized human
health and safety. But it is now becoming increasingly clear that fears
about radiation health and safety can produce ripples that span a much
broader range of possible impacts, and the magnitudes of the effects of fear
are nearly independent of estimated levels of health risks from radiation
exposure.
In particular, fear of consequences of radiation exposure causes
people to change their behavior with respect to potential sources of
exposure, and the economic and social impacts of that change in behavior can
be profound. So in many cases, it may be that the change in behavior
produced by fear of health risk is important, not the risk, itself.
As a result, potential radiation exposure has two kinds of impacts,
which we call risk- induced and perception- induced. Risk- induced inipac ts
arise directly out of "real" risks of "real" exposure (a finite probability
of an effective exposure); these include health effects, costs of mitigation
measures, and costs of dealing with accidents. Perception- induced impacts
are indirect, arising from peoples' behavior in response to fea~ of exposure;
these mostly include a broad range of avoidance behaviors, with associated
costs, that are not justified by "real" risks of exposure. Note that some
costs and some behaviors are justified by "real" risks of exposure; how much
and what behaviors
is not clear.
The division between risk- induced and
perception-induced impacts depends on individual values and attitudes toward
risk,
which are discussed in more
detail below.
Both risk- induced and
perception-induced impacts are important in risk assessment.
At least two kinds of importance must be considered when comparing risks
of alternatives:
.
Importance to relative risks.
A characteristic is important if it is
sufficiently different among alternatives, and of sufficient concern,
that leaving it out would change the risk ranking of alternatives;
and
.
Importance to the completness of the risk evaluation.
An important
function of a comparison of risks is to quantify their absolute
magnitudes .to determine whether or not they are acceptable. So an
impact or characteristic can be important if leaving it out would
make the assessment incomplete,
regardless of its contribution to
- 57 -

-------
differences in risks of alternatives.
This creates an impression of
ignoring or hiding something bad, which creates stress and conflict.
An example of the second might be something such as transport of LLY by
truck, the overall risks of which could be about the same for land and ocean
disposal. Seen from the national perspective, all persons along all routes
are equal, 80 it might not seem important to the comparison. Seen from an
individual perspective, each person will be along a transport route or not --
a large difference that matters to those involved. Ignoring risks of
transportation in such a comparison is sure to create hostility, regardless
of the magnitudes and distributions of differences among the alternatives.
Slovic et al. identify 19 characteristics by which people judge the
relative seriousness of risks. [Slovic et al., 1978] They are listed in Table
15 with a desc~iption of differences in the land and ocean disposal options
with respect to these characteristics. Basically, people perceive risks to
be greater, and fear them more, if they are involuntary, catastrophic, not
personally controllable, inequitable in distribution, unfamiliar, highly
complex, have delayed and uncertain effects, cause fatalities, and arise from
unnecessary technologies. Slovic et al. combine these characteristic into
two risk-perception "factors" -- dreadness and unfamiliarity. [Slovic et a1.,
1978]
Risks
of
nuclear
technologies
have
most
of
these
frightening
characteristics and have extremely high levels of dreadness and
unfamiliarity. As a result, the general public believes risks of. nuclear
technologies to be much greater than are estimated by knowledgeable
professionals. [Slovic et al., 1978; Slovic et al., 1980; Covello, 1983] Even
when they understand the levels of risk, people often fear them more than
similar levels of risk from other, nonradiological sources; they think that
it is in some way more "serious."
Three
important
reasons
for
the
large
differences with
respect
to
perception of risks of radiation exposure are:
.
Differences
in
KnowledEe.
Unfamiliarity with the details of
lack of information on normal levels of
radiation health effects,
radiation
exposure,
and
miscellaneous
misinformation
and
disinformation cause many to
over-estimate
the magnitudes of the
- 58 -

-------
risks.
risks
This is exacerbated by the special attention radiation health
receive in the news media relative to risks from other
sources. [Greenburg et al., 1989] Most people, including experts, are
over-confident in the quality of their estimates of risks and are
reluctant to change their minds, even in the face of contradictory
evidence. [Slovic and Fischhoff, 1980]
.
Distrust. The public lacks confidence in the government and its
hired authorities, which causes them to assume that the "official"
estimates
of
experts
are
self-serving
and
excessively
optimistic. [National Academy of Sciences, 1984]
.
Fear. There is a high level of aversion to the effects of radiation
exposure, even when the levels of risk are well understood. [Slovic et
a1., 1980]
These three reasons are linked in that the unknown can produce a high
level of fear and decrease confidence in the assurances of authorities, or
distrust can reduce attempts to obtain knowledge from "untrustworthy" sources
or reduce understanding by deflecting the distrustful to "more trustworthy"
misinformation or disinformation.
It is common for risk experts to attribute differences between their
assessments of risks and public perceptions of
"irrational" evaluations on the part of the public.
risks
to
ill-informed,
Risk assessors attempt
to make their evaluations scientifically rational, which they believe to be
"correct" for society at large.
And they assume that the public would agree
with their evaluations if only they understood them. But this is equivalent
to stating, "Anyone who disagrees with me is ignorant," a judgment that is
seldom correct. It is now becoming increasingly clear that public
evaluations of risk are not necessarily irrational; rather they are made from
a different perspective on the significance of hazards and the role that risk
should play in individual and public decision making. [Freudenburg, 1987]
While it is true that the general public may often be uninformed or
misinformed, they also often disagree with experts when properly informed.
The disagreement arises from differences in values. The significance of
values to risk perception and decision making is discussed in more detail
below.
- 59 -

-------
6.2
Quantitative Results
To understand how risks of land and sea disposal of LLW differ, we must
be able to quantify the characteristics of concern. The first step in
quantification is to find a unit of measure. For example, if we wish to
characterize length quantitatively, we use feet or meters as the measure. If
we wish to characterize human health effects, however, we find difficulties.
Simple measures like number of cases of disease or person-days of disability
may be insufficient to characterize important differences. There are many
gradations of disability, and different diseases have different implications
in cost, human suff~ring, and social consequences. A more detailed, multi-
faceted measure seems needed, yet the ability to distinguish differences or
to even understand the meaning of the measure can be lost in the morass of
detail.
Choosing an adequate method to quantify effects of concern is thus an
important part of the process of measuring relative risk of land and sea
disposal. To clarify this process we distinguish between the characteristics
of concern, discussed in the previous section, and the measures used to
quantify each characteristic. These are tabulated in Table 15, along with
comments describing short-comings, possible additions, sources of
information, and other clarifying information. Given the complexity and
subjective nature of many of the important characteristics, success in
quantifying them varies. Table 16 includes a ranking of the state of the
measure, in terms of current capabilities to estimate risks of land and ocean
disposal options, from "currently quantified" to "no current way to
quantify." Table 17 summarizes the current capabilities to estimate risks of
the land and ocean disposal options by four major categories of impact, and
provides recommendations on how these can be expanded.
The quantitative measures available for comparing the
disposal alternatives arise from:
land and ocean
.
Site plans and engineering design calculations describing the size,
operating conditions, and requirements of the necessary facilities
and transportation systems,
.
Information on average occupational illness and accidents per person-
hour in related industries,
- 60 -

-------
Table 15. Important characteristics determining individual evaluation of
seriousness of risks.
Characteristic
Comments
Vo1untariness of
risk
Immediacy of
effect
Individual knowledge
ab~ut risk
Level of
scientific
knowledge
Level of personal
control
Old or new risk
Chronic or
catastrophic risk
Common or dread
risk
Exposures associated with land and ocean disposal are
both involuntary.
An accidental release in transport or packaging of
waste could lead to immediate doses to the public for
both land disposal and the land or near-land
component of ocean disposal. Once the waste is in
place, leakage from the waste package in the oceans
would not reach man for hundreds of thousands of
years. For land disposal, depending on local
conditions, exposure could begin in tens of years.
People especially fear radiation because they cannot
tell when they are being exposed. This applies to
land and ocean disposal, but exposure from ocean
disposal might be from seafood anywhere in the world.
Specific wells around a contaminated land disposal
site could be identified and monitored, making risks
more visible.
In general, the health risks of radiation doses are
well understood by specialists, but not by the
public.
People have little control oyer risks from either
option, although the more easily identified area at
risk in the land option makes monitoring more
feasible. Also, people can participate in the site
selection process for the land operation, which may
make them feel they have more control.
People consider radiation to be a new risk, and
therefore more serious. Length of experience with
land and ocean disposal of LLY is about equal (40
years).
In neither option are the risks catastrophic. Even
accidents leading to exposure of large numbers of
persons would not lead to catastrophic health
effects, since the levels of activity of the waste
are so low and the effects would be diluted in time
and space.
Radiation is clearly not a risk people have learned
to think about reasonably and calmly. It is a dread
risk.
- 61 -

-------
Table 15.
cont.
Characteristic
Comments
Fatality
The low probability of exposure and the low level of
exposure were there an accid~nt are such that the
likelihood of significant consequences is small.
Cancer is the consequence of special concern, and the
fatality rate for cancers is currently fairly high.
Since the probability of exposure extends over
hundreds or even thousands of years, we can expect
that the fatality rate for cancers will decrease
significantly, perhaps even approaching zero.
Preventability
of risk
The chance of an accident can be reduced considerably
by many control measures, but it can never be reduced
to zero. In this context, the differences between
processing and transportation requirements for the
land and ocean options may produce different types of
accidents with different characteristics with respect
to preventability.
Controllability
of damage
In some cases (e.g., a massive fire in a waste
packaging or processing plant), control may be
difficult. Since these kinds of accidents occur in.
the early s:~ges of the waste system, retrievability
of waste packages may be required. It is much easier
to retrieve waste packages from land sites.
Number of persons
at risk
Many more persons are at risk of contamination in the
ocean option because of potential worldwide exposure
through the marine food chain. But dilution in the
ocean is so great that the individual level of risk
is exceedingly small, much less than that in the land
option.
Threat to future
generations
Both options threaten future generations.
Personal threat
The degree of feeling personally threatened is much
greater for those living near a land disposal site.
A similar, although perhaps lower level of personal
threat will exist along transport routes for both
options, and in the area around the port for the
ocean option. It is doubtful that any rational
person would feel a personal threat from contaminated
seafood at the time a decision to resume ocean
disposal is made. If elevated radiation levels were
reported in seafood some time later, then individuals
might feel personally threatened.
- 62 -

-------
Table 15.
cont.
Characteristic
Comments
Equitability of
distribution of
benefits
Benefits include military defense, electric power,
and medical diagnostics and care. In general, people
near a land disposal site feel they are taking the
risks while others are gaining most, if not a11 of
the benefits. Equitably is better defined for ocean
disposal as that among nations, rather than
individuals. Less-developed countries having no
nuclear weapons, no nuclear power, and only limited
access to nuclear medicine feel they get none of the
benefits and are either placed at some small risk or
suffer some loss of their share of ocean resources.
Threat of global
catastrophe
Nei ther land nor ocean disposal of.
threatens global catastrophe. Some
case for the ocean, but no argument
land.
low-level waste
might argue the
can be made for
Observability of
damage-producing
processes
Monitoring is relatively easy on land and may provide
early warning of contamination. Monitoring at sea is
technically more complex and much more expensive.
Increasing or
decreasing risk
Since ocean disposal has stopped temporarily, that
risk is static. The risk from. land is increasing as
more waste is disposed on land and new regional and
state disposal facilities are planned throughout the
country. The moratorium on sea disposal, combined
with increasing difficulty in obtaining approval for
land sites, continues to build pressure in the
system.
Reducibility of
risk
Technology exists for reducing risk to virtually
nothing for both options, but costs would be
unreasonable compared with reductions in risk
achieved. The costs would be equally prohibitive for
both options.
- 63 -

-------
Table 16.
Characteristics
Impacts" .characteristics, and risks of concern for comparing land versus ocean disposal of LLW.
Measures
S ta te
Comments
Public Health and Safety

Death, illness, injury, stress
Radiological,
nonradiological
Magnitude, uncertainty
(]'o
~
Location
Timing
Occupational Health and Safety

Death, illness, injury
Radiological,
nonradiological
Magnitude, uncertainty
Location
HEALTH RISKS
Person-rem exposure,
cancers
Cases
I
I
2
Traffic accidents could be added from data in
site plans
Expected value,
variance,
1
4
The uncertainty of most modelling results is
essentially unknown
maximum individual,
worst case,
1
1
Local,
national,
international
1
2
I
Routine, accidental,
current and future
1
I
Person-rem exposure,
cancers
Cases
1
1
1
Expected value,
variance,
maximum individual,
worst case
1
'2
2
I
Variance is included in available statistics
Local
1
Production of materials and equipment is not
included.

-------
Table 16.
cont.
Characteristics
Measures
Comments
State
Timing
Current routine and
accidental
1
1
ENVIRONMENTAL IMPACTS
Air, water, land
Local, national,
international
Concentration
Managed ecosystems
Productivity,
robustness
Reproduction,
productivity,
stability
Natural ecosystems
'"
lJ1
2
2
4
4
3
4
ECONOMIC IMPACTS
Risk-Induced Costs

To U.S. government for
normal operations
Dollars,
opportunity cost,
conservatisms
Reduction of magnitude of
potential impacts
Cost of recovery
Cleanup,
compensation
accident mitigation
recover ability
accidents
To local area .
Neighborhood/Town for
normal operations
Dollars
accidents
Cost of infrastructure,
need for facilities,
emergency preparedness
Emergency response
2
4
4
2
3
3
2
2
4
4
The significance of concentration is often not
well understood
land use effects
land use effects
Cost of alternatives
Difficult to separate
Difficult, what' base?
Needs are related to public perception

-------
Table 16.
cont.
Characteristics
Measures
Comments
State
To individuals for
normal operation and
accidents
Home
Dollars
Real estate value,
taxes
Business,
investments,
income
Prices,
taxes,
medical care
4
4
4
4
4
4
4
4
Li ve 1 ihood
Cost of living
To other countries
Dollars
4
0".
0'
Perception-Induced Costs

To U.S. government for
normal operations
accidents
3
3
4
Dollars
Conservatisms
Excess cleanup,
compensation
To region, state
Stigma
Dollars
4
To local area for
normal operations,
Dollars
Excess emergency
preparedness
Excess emergency
response,
loss of business,
loss of products,
loss of markets
Loss of markets,
outmigration
4
4
4
4
4
4
accidents
4
stigma
Much research is required in this area to
separate differences between facilities with
and without radiation risks and the associated
perception-induced effects
Even with significant amounts of research,
only a few effects of accidents can be
expected to be quantified and only with high
uncertainty
Public perception has a huge influence here;
some details are essentially unpredictable

-------
Table 16.
cont.
Characteristics
Measures
Comments
State
Dollars
To individuals for
normal operation,
accidents, stigma
Home
Livelihood
Cost of living
Real estate value, taxes 4
Business and income
Prices,
taxes,
medical care (stress)
To other countries
Dollars
Asthetic Impacts (linked to property value)
Appearance,
Intrusion on important
views
Decibels
Vehicles per day
Concentration
:r-
......
Noise,
Traffic
Pollution
Benefits
To U.S.
Opportunity cost of
waste disposal ($)
To local area
Jobs, $ business,
taxes, compensation
in dollars and
infrastructure
To individuals
Compensation in dollars
infrastructure, and
taxes
To other countries
?
4
4
4
4
2
4
4
4
4
2
2
2
2
2
2
2
2
2
2
4
Even with significant amounts of research,
only a few effects of accidents can be
expected to be quantified and only with high
uncertainty
Public perception has a huge influence here;
some details are essentially unpredictable
From site plans
From site plans

-------
Table 16.
cont.
Characteristics
Measures
State
Comments
National
Government programs
International politics
Regional
Employment
Archaeology
~
Local
Culture and lifestyle
Social system and
infrastructure
(boom town)
Power structure
Stigma
SOCIAL IMPACTS
Number, dollars
Number of governments,
see Equity
1
2
In-migration
1
4
Number of outsiders
2
In-migration,
new facilities
2
2
5
4
Out-migration,
community pride
and identity
5
Equity - Distribution of Costs and Benefits
Benefits and costs
Local: long-range,
national and
international
Producers and nonproducers
of LLW
Current and future
generations
Ratios and distributions
2
2
3
3
2
See economic impacts
From site plans
From site plans
Useful methods of quantifying equity must
be developed

-------
Table 16.
cont.
Characteristics
Measures
State
Comments
Desirable Characteristics of the Decision Process
Acceptance and involvement of
public as legitimate partner
Planning a~d. performance
evaluation
Listening to public concerns
No measures other than
presence or absence
4
Perhaps presence of desirable characteristics
is sufficient
Honesty, frankness, openness
Coordination and collaboration
0'
-c
Meeting needs of news media
Communication skill
Svrnbolic Meaninsz
Radioactive garbage dump
Inviolability of oceans
Local control - NIMBY
No measures
5
5
5
Atomic bombs, cancer,
mutants
Same for both
alternatives
5
Information Available and Individual Know1edsze
DOE response is similar for both alternatives.
Success in reaching individuals is different.
KEY: (1) Currently quantified; (2) Quantify with small additional effort; (3) Quantify with large modelling
effort; (4) Quantify with more research; (5) No current way to quantify.

-------
Table 17. Impacts, characteristics, and risks of concern for comparing land versus ocean disposal of LLW:
Summary and Recommendations.
Impacts and
Characteristics
Quantity of measurement
Recommendations
~ealth risks
Environmental
impacts
~
o
Economic
impacts
Social
impacts
Human exposure to radionuclides is
well covered with respect to magni-
tude, distribution, and timing.
Occupational accidents are the
only nonradiological health risks
included.
Non-radiological risks to the public could be
included. The international distribution
of radiological risks could be specified in more
detail.
Current models do not include
quantitative estimates of
environmental impacts. Some
general information might be
provided in site plans.
There are currently no methods for making
quantitative estimates of environmental
impacts at the required level except for
simple measures like acres consumed.
The only information on economic
impacts currently provided
is in site plans.
More research and modeling is required to
facilitate estimation of economic impacts other.
than costs. Health-related system costs could
be broken out and displayed separately to
increase understanding of their contribution to
the whole.
The only information on social
impacts currently provided (if
any) is in site plans. Mostly
this is related to estimates
of employment.
Measures of equity of distributions
of costs and benefits could be added. More
detailed estimates require more research and
modeling. Characteristics of decision
processes that are important to public
acceptance ~f risky facilities should be
included.

-------
.
Models to estimate radiation exposure-dose relationships, and
.
Computer models designed to estimate human exposure to radionuc1ides
released from the facilities under routine conditions and in
accidents.
Thus potential characteristics related to human health are well covered
(Tables 15, 16). The quality of the information, however, is not uniform.
In general, the uncertainty of the estimates increases with distance and time
from the source of radionuc1ides released to the environment, mostly because
of random variability of
incorporating it in models.
natural
processes
and
relative
success
in
Measures that could be generated with relatively small changes in data
requirements or computer codes are mostly:
.
Those that can be extracted directly or with modest effort from site
plans,
engineering
designs
(provided
they
exist) ,
and
readily
available statistics; and
.
Those that are related to intermediate physical changes estim~ted by
existing computer models but not currently displayed as part of the
results.
The first includes such things as traffic volume and accidents, local
employment and business, in-migration, needs for additional ~apacity in the
infrastructure (schools, roads, police, hospitals, etc.). These are either
required for engineering design calculations or can be estimated using
statistics on unit requirements appropriate to the area (e.g., students per
employee, teachers per student, classroom space per student).
The second type of measure includes anything related to environmental
concentrations of radionuc1ides in air, water and soil, which are calculated
internally in order to estimate human exposure, but not necessarily displayed
in the results. Different spatial disaggregations of human exposure, for
example, can be produced by apportioning results with appropriate
coefficients or adjusting the spatial units of the models. Impacts on
natural environments can be estimated to the extent that they are known to be
proportional to levels of anything that is included in the models.
- 71 -

-------
Given the extremely low concentrations of radionuclides that are
expected, it is possible that one of the largest impacts on the natural
environment will be from physical disruption during construction and
operation of facilities (not related to concentrations). and to changes in
human behavior with respect to low levels of radionuclide concentrations (not
directly proportional to concentrations). For example, changes in magnitude,
timing, and spatial distribution of fishing effort as a result of fears of
contamination could affect survival patterns and reproductive success in
international marine fisheries. The effect could be positive or negative and
would be specific to the actions taken and the characteristics of the
fisheries, especially the distribution and timing of changes with respect to
the breeding cycle. Although the responses of fish populations might be
predictable using available information, the responses of people are not.
Distributional equity is a special case among characteristics that could
be measured with small additional effort. On a coarse scale, equity can be
described as ratios between things people care about. We might calculate,
for example, the spatial distribution of the benefit: cost ratio, ratios of
local effects to long- range effects, current effects to future effects, or
exposures to producing nations and nonproducing nations. The difficulty in
such a case is not how to measure equity, but what is the meaning of the
measure provided. In part, this is a matter of experience and interpretation
- - understanding the relationship of the measure to the real world, and in
part is a matter of values -. the relative importance of specific levels of
the measure to the comparison.
6.3
Qualitative Results
Some impacts and characteristics are readily available in descriptive
form, but are not easily quantified for inclusion in an analytical
comparison. Aesthetic characteristics are of this type, as are many kinds of
impacts on natural and managed environmental systems. Characterization of
the process is another, although if all could agree on the ideal
charac~eristics and the amounts of each, then it would be relatively easy to
construct an index of success in meeting these ideals.
Often, however, qualitative results are provided not because it is
impossible to create quantitative measures, but because the effort required
- 72 -

-------
is excessive compared to the available
significance of the results obtained.
resources
and
the
quality
and
Lack of quantification is only a problem for those characteristics that
are important to an analytical comparison -- that is, those that will make a
difference in the ranking of alternatives, or those that must be included to
demonstrate the acceptability of risks. All of the quantitative methods for
making comparisons (Appendix D) require that the attributes of the
alternatives be expressed as numbers of some kind. They need not necessarily
be such straightforward numbers as persons or dollars or picocuries per
liter; they can be indices that express the magnitudes of the attributes
indirectly. But they must at a minimum be the numbers 0 and 1, expressing
presence or absence of a specified amount (unacceptable, harmful, illegal,.
etc.) of an attribute. Methods of quantifying attributes are discussed in
more detail below.
6.4
Gaps in Available Quantitative Information
Gaps in quantitative information are apparent in the inability of the
measures listed in Table 16 to adequately describe the characteristics of
concern; in some cases there is no ability to quantitatively characterize the
concern. Table 16 thus comments on the current state-of-the-art of
quantifying and assessing specific risks of the land and ocean disposal
options. The list of risks is long and the gaps in available information and
capabilities are large. Many of these gaps could be filled with additional
research and modeling effort. Gaps in the four major categories of impact
are discussed in greater detail in the four subsections below.
In considering information gaps, one must keep in mind that there will
always be gaps between what we know, and what we would like to know. Human
resources are never sufficient to satisfy all needs. The key is to rank the
relative importance of the gaps and focus on filling only the most important
gaps. A form of quantitative analysis, usually termed "value of future
information," can aid in determining the ranking, but the principle factors
in ranking are perceptions of the relative importance of various kinds of
risk and impact. This varies among different population groups and often
between analysts and the general public. Treatment of these value
considerations is covered in Section 4.6.
- 73 -

-------
6.4.1
Health Risks
Human exposures to radionuc1ides are well covered in the models with
re!!pect to magnitude, distribution, and timing. More work to reduce and
characterize the uncertainty of these estimates would be useful.
An argument could be made that more detail should be provided on the
international distribution of exposures from ocean disposal. But a parallel
argument can be made that current modeling capability is inadequate to
support more detailed estimates because the uncertainties are simply too
high.
The models now estimate total global exposure.
Wi th effort, they
might be d.isaggregated to estimate exposure at a smaller scale, perhaps at
the level of continents or even nations, but only at the expense of a huge
decrease in the accuracy and precision of the smaller- scale results. The
relative merits of smaller-scaled information of especially high uncertainty
are not clear. Some would argue that it can increase concerns without
providing more useful information for evaluating relative risks.
Occupational accidents are the only nonradiological health effects that
have been included in LLW risk a'~oassments. They are not included in
existing models. Risks of transportation accidents could be added easily,
but their only real significance is in helping to maintain perspective;
transportation risks are normally much larger than radiological risks. No
other nonradiological health effects are likely to be of significance.
6.4.2
Environmental Impacts
The only quantitative information on environmental impacts or risks
produced by current models is physical contamination of groundwater, surface
water, and selected terrestrial and aquatic foodchains. Extension to the
environmental or ecological significance of that contamination is confined to
the resulting exposure to humans.
Some general land-use measures like acres required and fuels consumed
are available in site plans. As a first-order analysis, we can assume that
the ecosystems of areas covered by land- or ocean-based facilities will
effectively be replaced by other organisms. What might replace them and the
significance of that replacement to the surrounding environmental system is
site-specific. Capability exists to model some impacts at the local scale;
mostly these are expected to be too small to cause concern, so long as no
- 74 -

-------
endangered
species,
special
enviranments,
ar ather
unique,
site-specific
features are inva1ved.
6.4.3
Ecanamic Impacts
The mast significant gaps with respect to. camparing risks af land and
acean dispasa1 af LLW are in patentia1 ecanamic impacts, especially thase
that directly affect individuals -- hame and live1ihaad (real estate value,
emp1ayment, sales af 1aca1 praducts, etc.). LLW assessment made1s in use
praduce no. infarmatian an ecanamic impacts. Same infarmatian may be pravided
in site plans and enviranmenta1 assessments an emp1ayment in canstructian and
aperatian af facilities that is either useful directly ar can be used to.
estimate direct ecanamic impacts. Little infarmatian is available related
specifically to. facilities inva1ving radiaactive materials. as distinct fram
thase inva1ving ather hazardaus materials ar undesirable characteristics.
Generating useful
infarmatian an
ecanamic
impacts af accidents will
require same research and made1ing.
Few data are available an ecanamic
effects af accidents inva1ving radiatian, because few such accidents have
occurred.
Far LLW, the accidents that have occurred have been re.1ative1y
minor, with few praducing release af radianuc1ides to. the enviranment, and
those being such small releases that ecanamic
insignificant. [U.S. Caunci1 far Energy Awareness, 1988]
impacts
were
The risk-induced impacts af an accident can be expected to. be similar to.
thase af ather accidents inva1ving hazardaus substances requiring special
care, such as chemical spills ar fires, far which some data are
available. [Lee et a1., 1989] Mast1y these are casts af cleanup, evacuation,
and lass af business and incame.
No. LLW accidents have inva1ved mare than
costs of cleanup and repair af equipment.[U~S. Caunci1 far Energy Awareness,
1988] Camputer made1s are available that estimate direct, risk-induced casts
of a reactar accident, but these are nat applicable to. LLW.[Lee et al., 1989]
Almast nathing is available an perceptian-induced ecanamic impacts af a
radiatian accident. The an1y radiatian accident in the United States far
which ecanamic impact data are available is the Three Mile Island nuclear
pawer plant accident, the nature and scale af which is camp1ete1y different
fram any accident passib1e in the LLW dispasa1 system, with the passib1e
exceptian af a large- scale fire. Devastating perceptian- induced ecanamic
- 75 -

-------
impacts are reported for the accident in Goiania, Brazil, in which cesium-137
found during a metal-scrapping operation was spread about as children played
with it. [Patterson et a1., 1988]
We
can
expect
that
the
perception-induced
economic
impacts
of
an
accident in the LLY disposal system will not be significantly related to
actual levels of risk, but instead will be linked to the time required for
cleanup and the handling of the accident by the news media. Mostly this is
related to the length of time it is treated as a potentially serious problem
by the media, which is also related to cleanup time. The attention span of
the news media is normally short, even for serious problems. [Greenburg et
a1., 1989]
6.4.4
Social Impacts
The only information on social impacts of the LLW disposal system is
that related to employment during construction and operation of facilities.
Land-based facilities are not large enough to produce "boom-town" effects,
but a large port facility might be if nearby towns lack sufficient capacity
to accommodate additional workers and their families without strain.
Information on social impacts of radiation accidents is mostly confined
to general discussion and qualitative descriptions; those from the Three Mile
Island accident are an example. [Lee et a1., 1989] There are few data.
Although current models do not deal with the question of equity of
distributions of costs and benefits, they could produce useful information
with modest additional effort. Mostly this would be in the form of maps of
costs and benefits and ratios of good and bad characteristics, which might
include:
.
Local:national:international radiation exposures,
.
Exposures to producers and nonproducers of LLW,
.
Exposures to current and future generations,
.
Compensation to receiving communities,
.
Costs to sending communities, or
.
Economic cost:benefit ratios - local, national, and international.
- 76 -

-------
Table 16 lists characteristics of the planning process that are
important to public perception of the process and the government, and that
can produce important social impacts related to public responses to
government actions and associated turmoil. [Covello and Allen, 1988] These
characteristics are not easily quantified other than by presence or absence,
but that is probably enough. If planners deliberately seek to incorporate
these characteristics into the process, then the important differences will
be related to the local versus international scale of the land and ocean
disposal options and the differences in approach that would be required for
each.
6.5
Methods of Filling Gaps
The methods available for filling gaps in information include:
.
Modification of computer codes to produce intermediate calculations
that are not now included in results,
.
Addition of coefficients to modify existing measures in models or
computer codes,
.
Creation
of
surrogate
measures
or
indices
that
capture,
quantitatively, relative magnitudes of differences that cannot be
quantified directly.
.
Research to produce information required for existing models
computer codes,
and
.
New modeling based on existing information,
.
Research on which to base new models, and
.
Elicitation of expert opinions.
Some gaps can be filled relatively easily once it is determined that
they are important. But in general, gaps are results of inadequacies of some
kind. Often it is no more than an a perception on the part of.ana1ysts that
the cost of filling the gap exceeds the direct importance of the information
to the comparison. In such cases, careful attention must be given to the
importance of the information to the process to ensure that analysts are not
accused of hiding important problems.
- 77 -

-------
Ability to modify models or add coefficients to existing computer codes
depends mostly on an individual's familiarity with the code which, in turn,
is affected by the quality of the code. Unless a code is well organized and
documented, it is extremely difficult for someone other than its creator to
make changes with confidence that they will work as intended and not
adversely affect other parts of the code. For models of even modest
complexity that are not well organized and documented, it is often considered
easier to start anew than to try to understand the code well enough to make
necessary changes. In many cases, modifications to codes are best left to
persons already familiar with them for other reasons.
Expert opinion is used to clarify data of low quality or in the absence
of applicable data. [Morgan et al., 1981] Expert opinion may also be u~ed to
guide the design of studies to obtain necessary data. Elicitation of expert
judgment is a more formal approach than what is called "engineering
judgment." The basic difference between expert judgment and elicitation of
expert opinion is in the specificity of the results. In elicitations, the
experts are identified and qualified, an attempt is made to use experts who
span the range of accepted opinion, and elicitation methods are highly
formalized, with great pains taken to obtain an understanding of the causes
and magnitudes of uncertainties in their estimates. Results are usually
expressed as subjective probability distributions over the range of values
the experts consider possible. Because of its emphasis on quantifying
uncertainty, elicitation of expert opinion can often be better than use of
averages of unspecified uncertainty. which tend to make small differences
among alternatives seem more important than they really are.
6.6
Relative Risk Evaluation
The decision-aiding theories and models appropriate for making the kind
of complex, multiattribute analyses needed to compare risks of the land and
ocean disposal alternatives are discussed in Appendix D. These methods
require calculation of a n goodness (or badness) score" that combines the
necessary information on risks into a single measure that can be used for
direct comparisons of alternatives with different characteristics. Most
involve the sum of the products of measured levels of attributes times value
coefficients expressing the relative importance of each attribute to the
comparison. In theory, this application is straightforward; in practice, it
- 78 -

-------
is not.
Problems can arise from decisions on selection of attributes to be
included, determining how to measure them, the quality of the resulting
measure, determining whose values should be represented in the coefficients,
and measuring those values.
to such a comparison.
This section stresses the significance of values
6.6.1
Values
In the sense used here, values are measures of the relative importance
of characteristics and the tradeoffs that can be made among them. They can
be explicit or implicit, and they can be for individuals or representative of
groups.
Explicit values are stated subjective relationships used in an analysis.
Since we are concerned with quantitative comparisons, explicit values must be
quantitative and refer to measured levels of attributes.
Implicit
values
are
those
incorporated
in
the
(usually)
unstated
assumptions used in selecting the objectives of an assessment, the attributes
to be quantified, ways of quantifying or describing them, and ways of
combining them. Selecting for analysis only those measurements that are
,currently available, for example, contaIns an implicit value judgment that
the value of the time and cost of obtaining more information exceeds the
value of the contribution of that information to the analysis. But this
implicit judgment is normally made by default by someone -"doing the best he
can with the available information," without much evaluation of its
implications.
judgments.
Any quantitative analysis contains many such implicit value
';"'4- ..~ c
If attributes have a recognized common metric, such as dollar cost, then
no separate values need be placed on them; their dollar costs can be compared
directly. This is done in cost:benefit analysis (see Appendix D). But if
the attributes are measured in incommensurable units -- those that cannot He
compared directly with a common metric then analysts must use value
judgments about how im~ortant each is relative -to the others and how much of
one should be given up in order to get some of ~nother. This is the familiar
"apples and oranges" problem. Of all the components of a quantitative
comparison of characteristics of the land and ocean disposal alternatives I
determination of values will be the most troublesome.
- 79 -

-------
What is required is a numerical "weight" for each attribute expressing
how important a difference of one unit of measurement of that attribute
should be and the tradeoffs analysts can make among the alternatives. These
weights cannot be established in isolation; they are specific to the problem,
to the units in which the attributes are measured. and to the person or
persons whose values are involved. One cannot, for example, simply state
that "human health is more important than environmental quality." This
implies that any health impact is worse than any environmental impact, which
at the extreme is clearly not the case. Instead, what is intended by that
statement is, "To me, the human health effects I expect are more important
than the environmental impacts
I expect, II where the expected amounts are
specific enough to support the judgment. To establish a weight, analysts
must be able to go a step further and specify that X specific human health
effects are equal in importance to Y specific environmental impacts.
more difficult, they must specify that a given probability of X
human health impacts is equal to Y specific environmental
Mechanisms for assisting in making these difficult judgments of
importance are outlined in Appendix D.
Or even
specific
impac ts .
relative
A large part of the uncertainty involved in comparing land and ocean
disposal of LLW arises from a need to evaluate the relative importance of
many different potential impacts that are incommensurable (Table 16). These
two alternatives are different in the kinds and magnitudes of potential
impacts they entail. Land-based approaches mostly threaten local water
supplies and ocean-based approaches mostly threaten international fisheries,
for example. One especially important difference is the distribution of
risks of contamination between those who benefit directly from the activities
producing the LLW and those who do not, a question of equity. Risks to
producing nations should clearly be evaluated differently from risks to
nonproducing nations, even in cases where they have a common metric (e. g. ,
human exposure).
The uncertainty from incommensurables becomes very large when the full
range of characteristics of concern is included and the special concerns of
all persons affected are accounted for. When analysts are faced with dealing
with a large number of incommensurables intuitively, the normal response is
to eliminate those that are seen as relatively less significant and evaluate
- 80 -

-------
only a manageable subset that is considered unavoidably significant. If the
full range of incommensurab1es is to be included in a comparison, then it is
necessary to deal separately with each and to devise ways of quantifying
their relative importance by some common metric. It is specifically this
application for which analytical decisionmaking methods were ~evised
(Appendix D).
6.6.2
Individual Values
For any particular person, there is clearly a hierarchy of values for
various kinds of characteristics, but it is not always clear what that
hierarchy is. In general, people care more about things that are close to
them in space, time, and personal relationship:
.
Here, there, somewhere.
.
Home, neighborhood, town, country.
.
Now, soon, sometime, a long time.
.
Me and mine, you, them, someone.
.
Food, clothing, shelter, health,
income,
lifestyle, etc.
But even this generalization varies depending on the person and the
characteristic evaluated. Most people would place their chi1drens' welfare
above their own, especially with respect to health issues, but not all and
not in all things (e.g., conflicts between children and career are relatively
common). 50 the specifics of individual value systems are highly personal
and not readily predictable.
They are, however, quantifiable. The two basic approaches to
quantifying these values are through questions designed to elicit expressed
preferences (e.g., "How much of this are you willing to trade for that?") and
statistical analyses designed to quantify preferences implied by observed
behaviors (e.g., average housing prices are lower close to an undesirable
facility) .
The situation is further complicated by differences in perspectives
between individuals protecting things they value personally, and perhaps
engaging in strategic bias and gamesmanship in an attempt to influence
outcomes
in
their
favor,
and
government
analysts
and
decisionmakers
- 81 -

-------
representing society, but who also have personal values that influence their
views of what societal values should be.
6.6.3
Values in Conflict
Whose values should be used in social decision making (e.g., Congress,
environmental groups or others) to make better and more informed decisions?
If all agree on appropriate weights for attributes, then there would be no
conflict in evaluating alternatives. But it is clear that there are large
differences in values, not only among individuals, but also between
individuals (including decision makers) and the society as a whole. This is
most particularly true for risk of exposure to radiation, which has a
special, and especially sensitive, place in individu~l perceptions. [Slovic et
a1., 1978] Large differences produce conflict.
Payne
individual
and. Williams identify three sources
values related to radioactive waste
of conflict affecting
management: [Payne and
Williams, 1985]
.
The social history of development of nuclear technology,
.
Differing value orientations, and
.
Differing perceptions of acceptable risks.
The social history is important because of fear associated with nuclear
weapons, which is transferred in the minds of lay persons to all nuclear
technologies. Most people incorrectly think about all nuclear technologies
as though they were potential bombs. In addition, people do not readily
change these attitudes, especially when presented with conflicting
information, but also when presented with convincirig information that is
contrary to strongly-held beliefs.
Differing value orientations arise from four basic areas:
.
Participatory democracy,
.
Stewardship,
.
Environmentalism, and
.
Equity.
- 82 -

-------
People, groups, and communities want to maintain control over their own
destinies. Conflict arises when they feel excluded from decisionmaking
processes affecting them directly. They also feel that common resources
(land, water, air) under government stewardship should be used carefully and
productively for the common benefit. Environmentalism, in particular,
emphasizes conservation and preservation of common natural resources that
should be maintained without harm for the common benefit, present and future.
This, in part, arises from a sense of a need for equity in the distribution
of costs and benefits in all things, but especially in use of common
resources. Because of differences in orientation with respect to these
values between decisionmakers at the national level, who think in terms of
benefits to the society as a whole, and local communities directly affected
by their decisions, who think in terms of their own lives, potential for
conflict is high. Neither is necessarily wrong or selfishly motivated. They
simply have different perspectives on what is important.
6.6.4
Social Amplification of Risk and Effects on Values
Differences in risk perception between "experts" and the general public
are well documented. They arise from differences in knowledge and experience
"that produce different assumptions yielding different conclusions. In
general, people overestimate the frequency of catastrophic and sensational
hazards with which they have little experience, and they especially fear
hazards that are involuntary,
not under personal
control,
inequitable in
distribution
of
costs
and benefits
(not
fair),
and highly
complex
and
unfamiliar.
The result of these differences is that the general public believes the
risks of exposure to radiation to be greater and more serious than those of
other, more familiar risks, and they overestimate the likelihood of accidents
producing exposures., so they do not accept the numbers
"experts" that are in conflict with their preformed beliefs.
calculated
by
Kasperson, et al. outline the sources 'of special public concern about
the risks of radiation exposure, which they call "social amplification of
risk. " [Kasperson et al., 1987] The roots of social amplification of risk lie
in experience, either direct or indirect through information from friends and
the media. Direct experience with risky activities can be reassuring, as in
- 83 -

-------
driving automobiles, or
e~perience with dramatic
heightens perception of
greater perspective and
accidents.
alarming, as in tornados or floods. Limited
accidents tends to make them more memorable, which
risk.
But repeated experience
can also
afford
capability for avoiding risk,
as
in
occupational
The role of the news media in amplification of risk is critical. The
key attributes of information that cause amplification are volume, level of
dispute, and extent of exaggeration. All of these are under the control of
the news media and are used to maximum advantage to generate sales. [Greenburg
et a1., 1989] Misinformation and distortion are common. [Combs and Slovic,
1978; Freimuth et a1., 1984] High volumes of information mobilize latent
fears about a particular risk. Debates among ftexpertsft increase uncertainty.
Erroneous information, usually exaggeration to increase their value to the
news, increases memorability and fear without basis in fact.
Public amplification of the relative importance of catastrophic
accidents means that decision makers must make some conscious judgment about
how to deal with their significance in a decision.
Consider, ~or example,
the comparison shown in Figure 5, in which two alternatives differ in the
probability distribution of human exposure to radionuclides. The
distributions have the same expected value, but one alternative has risks
concentrated in the medium range (perhaps routine emissions with no
likelihood of a severe accident) while the other has risks spread over a wide
range, mostly at very low levels, but including a reasonable possibility of a
catastrophic accident. Yith respect to expected exposure, these alternatives
are the same. Yith respect to the worst possible outcome, the first
alternative is superior. Yith respect to the most likely outcome, the second
alternative is superior-. How should these different measures of impact be
valued?
Mathematically, if the two distributions have the same expected value
and the more catastrophic outcomes are considered more important per unit
exposure, then a distribution having a higher probability of catastrophic
accidents will always be more important than one having a lower probability.
But if they do not have the same expected value, then a judgment must be made
of the relative importance per unit exposure of small, frequent exposures to
many versus large, infrequent exposures to a few. This judgment can be an
- 84 -

-------
Figure 5.
 IOf)  
~   
" ~ C' ~Acc;j~~1
.......
~ 
.......   
to   
'*   ~ Rou{;AJ~
~  
~  
 0  
 0 CO}JSEQlI £NC.£
Hypothetical distributions of risks of human exposure. one with
only routine low-level risks and another with only higher-level
risks of accidents.
- 85 -

-------
important source
perspective, for
of
conflict
between
decision
makers
with
a
nationa~
whom
all
exposures
are
much
the
same
per
unit
of
measurement,
and the general public,
for whom catastrophic accidents are
especially feared.
6.6.5
Incorporating Stake-holder Values
Establishing values is an important stage in the comparison of the land
and ocean disposal options. These include not only specific values for
measured attributes, but also implicit and explicit values, judgments, and.
tradeoffs imbedded in decisions on what attributes should be included in the
analysis and what methods should be used to evaluate them.
important here is the question of whose values should be used.
Exceedingly
Appendix D discusses some theories of social dec~sion making and
problems with applying them in the real world. Treatments of values are
fundamentally different in these approaches. Cost-benefit analysis uses
monetary value as established by market forces or other estimates of
"willingness to pay,n.which are assumed to represent aggregate social value
(no decision maker). Decision analysis uses the values of one person (or a
group of like mind) representing himself, his organization, or some larger
group of stakeholders, including society as a whole (one decision maker).
Social welfare theory attempts to provide a rational synthesize of the
preferences of all stakeholders through some aggregate social welfare
function (many decision makers).
If the outcome of the comparison is mostly of political significance,
then it is useful to maintain separate value functions for the different
groups of stakeholders involved, either by measuring them for small groups or
selecting a representative to measure for larger groups. The results
obtained using such value functions reveal the significance of differences of
opinion on the outcome of a comparison. In fact, the weights can even be
constructs representing hypothetical points of view with respect to important
attributes.(Radioactive Waste Division of the Department of Environment,
1986] This approach has the advantage of not requiring specific measurements
of value functions to complete a comparison, and the disadvantage that the
results do not represent the values of real people, which may not be captured
well by the constructs.
- 86 -

-------
6.7
Designing a Quantitative Comparison
All analytical methods for comparing alternatives involve some basic
steps that separate planning functions, information on characteristics of
alternatives, and the values and preferences used to evaluate them. The
steps usually include most of the following, in whatever order is convenient.
.
Define objectives
.
Identify alternatives and structure the analysis
.
Define performance measures
.
Identify important characteristics
.
Quantify characteristics
.
Specify values, preferences, and permissible tradeoffs
.
Evaluate alternatives
..
Evaluate sensitivity and value of additional information.
The objective of the land versus ocean comparison is predetermined --
quantify the relative risks to provide an information base for
decisionmaking.
Alternatives to be evaluated must be determined by analysts and decision
makers.
The land option can be represented as a system of existing and
planned sites,
a representative site, or a hypothetical generic site, with
alternative or representative treatment and disposal techniques as required.
Only one ocean site is. currently in use, but alternative treatment and
disposal techniques can be included. Needs to analyze a broad range of
alternatives within each option depend, in part, on the nature and magnitude
of the differences between them. In a general comparison of land and ocean
disposal, if the differences between alternatives within options are small
compared
to
the
differences
between options,
then
little
is
gained by
expanding the list of alternatives considered. This can occur if there are
large differences between the options with respect to relative magnitudes of
risks or with respect to the relative importance (value) of risks of
different kinds. The more the options tend to overlap with respect to
magni tudes and values of risks,
the greater is the need to
include all
available alternatives within them.
- 87 -

-------
Analysts should keep in mind that risks of alternatives are variable
functions of cost. The more money is spent on protective measures, the
smaller will be the magnitudes of the risks. This distinction is universally
ignored in risk assessments, which are normally based on fixed engineering
designs specifically intended to meet some predetermined set of criteria, and
which are thereafter taken to be immutable. We forget that we can always
spend more money to reduce specific risks if it is justified.
Important
characteristics
and
impacts
that might be
included
in a
comparison are outlined in Table 16. Constraints on time
restrict the comparison to some subset of the total that
and resources
includes
the
information currently available plus whatever other information is considered
sufficiently important to justify additional expenditures of time and
resources.
Importance should be evaluated not only with respect to the
comparison,
itself, but also with respect to the process
and effects on
public opinion. This comparison involves tradeoffs under a finite budget
between relative impor~ance of information, the cost of getting it, and its
likely effect on the outcome. Table 18 outlines some of the objectives and
criteria that might be used in the selection.
This stage of an analytical comparison is seldom quantified, and
selections are mostly based on direct experience, knowledge of other, similar
problems, and intuition. Many implicit value judgments must be made in the
absence of specific, quantitative information related to the alternatives
under evaluation. It would be useful with respect to public opinion. if
analysts could provide some kind of rationale for the selections made at this
stage, particularly if something the general public cares a lot about is left
out (e. g., economic impacts). There is a tendency, however, to be as
unspecific as possible at this stage to avoid providing ammunition to critics
early in the process.
Specifying values poses two problems -- whose values are to be used and
how to measure them. Because of the political sensitivity of the land versus
ocean comparison, it would appear that maintaining separate value functions
for the various stakeholder groups would provide the most useful information.
Not only might it demonstrate the overall effect of different value systems
on the rankings of alternatives and options, but also it might assist in
- 88 -

-------
Table 18. Rationales for selecting characteristics and impacts for inclusion
in the multimedia comparison,
Objective
Criterion
Maximize feasibility
Potential for producing obstructive
responses
Minimize fear 
Minimize health risk
Minimize cost 
Public perception of relative risks
Potential Health effects
Cost-related characteristics, including
cost of obstructive behavior caused by
objections to leaving out other
important information
Minimize total impact
Experts estimates of magnitudes and
importance of impacts
Social optimization
Societal preferences for general impacts
Promote acceptance of a
particular alternative
Favorability ratio
Provide appearance of objectivity
Level of controversy
Support organizational goals
What are the organizational goals?
- 89 -

-------
showing the specific differences in risks and values that actually affect the
comparison. These indicate candidates for more detailed attention.
- 90 -

-------
7
CONCLUSIONS
Methods
and models
are
available
to
compare
some health
and
some
environmental risks associated with the ocean- and land-based disposal. of
LLW. If only health risk estimates are prepared, the minimal set of risks to
be examined should include routine and accidental radiation exposures to
workers and the public from transport (land and ocean), disposal, and post-
closure releases of radionuclides; health consequences for maximally exposed
or critical population groups, and the estimated collective dose to workers
and the public for each defined scenario. Since waste processing, packaging
and disposal requirements for land and ocean disposal can differ, their
contribution to the overall risk can vary by medium; these have not yet been
fully evaulated.
Human and natural processes that precipitate accident-initiated releases
are also important determinants of risk and must be examined. Some of these,
such as human/animal intruder scenarios or climatic events (e.g., a 100 year
flood) may only relate to one medium, while others, such as earthquake
activity, apply to both media. Differences in exposure scenarios can have a
large impact on the overall risk.
On land, the geographic scales of interest for routine releases will
generally be confined within the boundaries of the ground water system. In
contrast, the ocean option will require analysis at more global scales (i.e"
oceans) because of the nature of the circulation of the deep ocean. Because
of extensive computational requirements to model ocean circulations at such a
wide scale,
all available dynamical models
are too
large to be directly
incorporated into a systems analysis model for full uncertainty analysis.
Thus, dynamical models should be used in performing "most realistic" ocean
transport
calculations
for
risk.
For
efficiency in systems studies,
albeit crudely, by box model
calculations
such
as
these
are
replaced,
calculations. It is tempting, in box modeling, to increase the number of
grid points so as to produce greater resolution and/or accuracy, and to some
extent, this can be done. However, it is a dubious practice because errors
in the underlying model parameters usually increase as the resolution of
problems increase.
- 91 -

-------
In principle, the models selected for analysis should have completed
some validation exercises and be capable of producing explicit estimates of
model uncertainty. In practice, however, it appears that most available
pathway models have not undergone rigorous validation e.fforts or are not
capable of producing explicit estimates of output uncertainty. Similarly,
while estimates produced by models containing complex descriptions of
physical phenomena may appear to be more credible or accurate, this is not
always true. In any case, continued efforts to validate the predictions of
performance assessment models against actual field data are critical to
reduce uncertainty.
It is clear that the general public is concerned about a much broader
range of characteristics and potential risks of the LLW disposal options than
just health risks. These especially include socio-economic risks and the
equity of the distribution of risks and benefits in space and time. None 'of
the available LLW risk assessment models include these important
considerations, and few of the models could be modified to include any of
them with modest effort. This is a serious deficiency in current
capabilities that will undermine the credibility of results in the eyes of
the general public. When this information becomes available, there are
decision aiding methods for incorporating a broad range of characteristics,
potential risks, and stakeholder values into quantitative comparisons of
alternatives. Development of necessary data and application of these
approaches can permit decision-makers to make better and more informed
decisions.
- 92 -

-------
8
REFERENCES
Bennett, R.D., et a1., 1984. Alternative Methods for DisDosa1 of Low-Level
Radioactive Wastes, NUREG/CR- 33774, Vo1.1, Geotechnical Laboratory, U.S.
Army Engineer Waterways Experiment Station, Vicksburg, MS.
Branch, K.M., C. Cluett, and T.L. Page, 1987. "Integrating Environmental and
Socioeconomic Assessment," Vol. 2, pp. 85-90, in R.G. Post, ed., Waste
Management '87 Proceedings of the Symnosium on Waste Management. Tucson.
AZ. March 1-5. 1986, American Nuclear Society, LaGrange Park, IL.
Brookhaven National Laboratory - LLW Deep Sea Group, 1986. Technologies for
and ImDacts of Low-Level Radioactive Waste in the DeeD Sea, Draft report
prepared for the Office of Technology Asses~ment, U.S. Congress,
Washington, DC.
Burns, M.E., and W.H. Briner, 1988. "Setting the Stage," in M.E. Burns, ed.,
Low-Level Radioactive Waste Regulation - Science. Politics. and Fear,
Lewis Publishers, Inc., Chelsea, MI.
Colombo, P. and M. Fuhrmann, 1988. Waste Packa~e Performance Criteria for
DeeDsea Disposal of Low-Level Radioactive Wastes, EPA 520/1-88-009. Office
of Radiation Programs, U. S. Environmental Protection Agency, Washington
DC.
Combs, B., and P. Slovic, 1978. "Newspaper Coverage of Causes of Death,"
Journalism Ouarter1y 56:837-843.
Covello, V., and F. Allen, 1988. Seven Cardinal Rules of Risk Communication.
U.S. Environmental Protection Agency, Washington, DC.
Covello, V. T, 1983. "The Perceptions of Technological Risks: A Literature
Review," Techno1. Forecast. Social Change. 23:285.
Cox,M.D., 1984. A Primitive Eauation 3-Dimensiona1 Model of the Ocean,
Technical Report No.1, Geophysical Fluid Dynamics Laboratory. Princeton,
NJ.
de Marsi1y, G. et a1., 1989. Feasibility of Dis~osa1 of
Radioactive Wastes into the Seabed. Volume 2. Radio1o~ica1
Nuclear Energy Agency, Paris.
High-Level
Assessment,
EG&G, 1985. Low- Level Waste DisDosal Performance Assessment Model Review,
EGG-LLW-8083, Idaho National Engineering Laboratory. Idaho Falls, Idaho.
EG&G, 1987. The 1986 State-by-State Assessment of Low-Level Radioactive
Wastes Received at Commercial Disposal Sites, DOE/LLW-66T, Idaho National
Engineering Laboratory, Idaho Falls, Idaho.
Fields, D.E., et a1, 1987. PRESTO-EPA-POP: A Low-Level Radioactive Waste
Environmental Transport and Risk Assessment Code, EPA 520/1-87 -024 (2
Volumes), Office of Radiation Programs, U.S. Environmental Protection
Agency, Washington DC.
- 93 -

-------
Freimuth, V.S., R.H. Greenburg, J. DeWitt, and R. Romano, 1984. "Covering
Cancer: Newspapers and the Public Interest," Journal of Communications
.J!,:62-73.
Freudenburg. W.R, 1987. "Rationality and Irrationality in Estimating the
Risks of Nuclear Waste Disposal," in Waste Management '87. ProceedinEs of
the SvmDosium on Waste Management. Tucson. AZ. March 1-5. 1987. 2:109-115,
American Nuclear Society, LaGr~nge Park, IL.
Greenburg, M.R., D.B. Sachsman, P.M.
. "Network Evening News Coverage of
i:119-126.
Sandman, and K.L. Salomone, 1989.
Environmental Risk," Risk Analysis
Helton et a1., 1988. "Integration of NUREG-l150 Analyses: Calculation of
Risk and Propagation of Uncertainties," in Proc. Fifteenth Water Reactor
Safety ~esearch Information Meeting.
Holcomb, W. F., 1982. "A History of Ocean Disposal of Packaged Low-Level
Radioactive Waste," Nuclear Safety 22(2).
Holland, W.R. and W.J. Schmitz, 1985. "Zonal Penetration Scale of Model
Mid1atitude Jets.," J. Phys. Oceanogr. 1859-1875.
International Atomic Energy Agency, 1978. Convention on the Prevention of
Marine Pollution by DumDing of Wastes and Other Matter. The IAEA Revised
Definition and Recommendations of 1978 Concerning Radioactive Wastes and.
Other Radioactive Matter. Referrs'.i to Annex I and II of the Convention.
Information Circular INFCIRC/205/Add.1/Rev.1, Vienna, Austria.
International Atomic Energy Agency, 1984a.
Analysis MethodoloEies for Radioactive
Ground, Vienna, Austria.
Safety Series No.
Waste Re~ositories
64 - Safety
in Shallow
International Atomic Energy Agency, 1984b. Safetv Series No. 66 The
OceanograDhic and Radiological Basis for the Definition of HiEh-Level
Wastes Unsuitable for DumDing at Sea, Vienna, Austria.
Kasperson, R.E., J. Ernel, R. Goble" C. Hohenemser, J.X. Kasperson, and R.
Renn, 1987. "Radioactive Wastes and the Social Amplification of Risk," in
R.G. Post, ed., Waste Management' 87. ProceedinEs of the SvmDosium on
Waste ManaEement. Tucson. AZ. March 1-5. 1986, 2:85-90, American Nuclear
Society. LaGrange Park, IL.
Lee, J.C., M.D. Rowe, C.A. Grimshaw, P.D. Ka1b, and C.G. Miles, 1989.
Economic ImDacts of Accidents in the High-Level Nuclear Waste DisDosal
System: A Review of Methods and Data. Draft report, Brookhaven National
Laboratory. Upton, NY.
Marietta M.G. and W.F. Simmons, 1988. Feasibility of DisDosal of High-Level
Radioactive Wastes into the Seabed: DisDersal of Radionuc1ides in the
Oceans: Models. Data Sets. and Regional DescriDtions. SAND87-0753 Sandia
National Laboratories, Albuquerque, New Mexico. '
- 94 -

-------
McKone, T.E. and D.W. Layton, 1986. "Screening the Potential Risks of Toxic
Substances Using a Multimedia Compartment Model: Estimation of Human
Exposure". Re~lato~ Toxico1o~y and Pharmaco1o£v 6:359-380.
McKone, T. E., 1981. Chemical Cycles and Health Risks of Some Crustal
Nuclides. Ph.D. Dissertation, University of California, Los Angeles.
Miller, R.N., A.R. Robinson, and D.B
Quasigeostrophic Open Ocean Model," J
Haidvoge1, 1983.
Como. Phys. 50:38-70.
"A
Baroc1inic
Mobbs S., et a1., 1986. A Pre1iminarv Comnarison of Models for the
Disnersion of Radionuc1ides Released into the Deen Ocean. Report No.
NRPB-R194, Chilton, Didcot, Oxon, United Kingdom.
Morgan, M.G., M. Henrion, and S.C. Morris, 1981. Exnert Jud~ents for Policy
Analysis, BNL-51358, Brookhaven National Laboratory, Upton, NY.
National Academy of Sciences Panel on Social
Radioactive Waste Management, 1984. Social
Radioactive Waste Disnosa1: Considerations for
National Academy Press, Washington, DC.'
and Economic Aspects of
and Economic Asnects of
Institutional Management,
Nuclear Energy Agency, 1979. Guidelines for Sea Dumnin~ Packages of
Radioactive Waste. Nuclear Ener~ Agency. Or~anization for Economic
Coo~eration and Deve1onment, Paris, France.
Nuclear Energy Agency, 1980. Review of the
Dumning Site for Radioactive Waste in the
Energv Agency, Organization for Ecc'nomic
Paris, France.
Continued Suitability of the
North-East Atlantic. Nuclear
Cooperation and Development,
Oak Ridge National Laboratory, 1988. Inte~rated Data Base for 1988: Spent
Fuel and Radioactive Waste Inventories. Projections. and Characteristics,
DOE/RW-0006, Rev. 4, Oak Ridge National Laboratory. Oak'Ridge, TN.
Organization for Economic Cooperation and Development, 1977.
OECD Council of 22nd July. 1977. Paris, France.
Decision of the
Oztuna1i, 0.1., et a1., 1986.
Methodology, NUREG/CR-4370,
Commission, Washington, DC.
Update of
(2 Volumes)
Part
U.S.
61: Imnacts Analysis
Nuclear 'Regulatory
Parker, F.L., 1988. "Low-Level Radioactive Waste Disposal," in M.E. Burns,
ed., Low-Level Radioactive Waste Regulation - Science. Politics. and Fear,
Lewis Publishers, Inc., Chelsea, MI. .
Patterson, J.S, 1988. "From Perception to Reality: The Goiania Socioeconomic
Impact Model," In R.G. Post, ed., Waste Management '88. Proceedings of
the Symnosium on Waste Management. Tucson. AZ. Februa~ 26 to March 3.
1988, American Nuclear Society, laGrange Park, IL.
- 95 -

-------
Payne, B.A., and R.G. Williams, 1985. "Conflict, Public Communication, and
Radioactive Waste Management," In R.G. Post, ed., Waste Mana~ement '85.
Proceedings of the Symnosium on Waste Management. Tucson. AZ. March 24-28.
1985. 3:103-106, American Nuclear Society, laGrange Park, IL.
Radioactive Waste (Professional) Division, 1986. Assessment of Best
Practicable Environmental Ontions (BPEO's) for Management of Low- and
Intermediate-level Solid Radioactive Wastes, Department of Environment,
United Kingdom.
Robinson, A.R. and M.G. Marietta, 1985. Research. Progress and the Mark-A
Box Model for Physical. Biological and Chemical Transnorts, SAND84-0646,
Sandia National Laboratories, Albuquerque, New Mexico.
Rogers, V., and C. Hung, 1987. PATHRAE-EPA:
Environmental Transnort and Risk Assessment
of Radiation Programs, U. S. Environmental
DC.
A Low-Level Radioactive Waste
Code, EPA 520/1-87-028, Office
Protection Agency, Washington
Rogers, V., and C. Hung, 1987. PRESTO-EPA-CPG: A Low-Level Radioactive Yaste
Environmental Transnort and Risk Assessment Code, EPA 520/1-87-026, Office
of Radiation Programs, U. S. Environmental Protection Agency, Washington
DC.
Rogers, V., and C. Hung, 1987. PRESTO-EPA-DEEP: A Low-Level Radioactive
Yaste Environmental Transnort and Risk Assessment Code, EPA 520/1-87-025,
Office of Radiation Programs, U.S. Environmental Protection Agency,
Washington DC.
Rogers, V., and C. Hung, 1987. PRESTO-EPA-BRC: A Low-Level Radioactive Waste
Environmental Transnort and Risk Assessment Code, EPA 520/1-87-027, Office
of Radiation Programs, U. S. Environmental Protection Agency, Washington
DC.
Schmitz, W.J. and W.R. Holland, 1986. "Observed and Modeled Mesoscale
Variability Near the Gulf Stream and Kuroshio Extension," J. Geonhvs. Res.
21:9624-9638.
Semtner, A.J. and R.M. Chervin, 1988. "A Simulation of the Global Ocean
Circulation with Resolved Eddies," J. Geonhvs. Res. 93:15502-15522
Shuman, R., et a1., 1988. Performance Assessment for Low-Level Waste
Disnosal Facilities, NP-5745, Prepared for Electric Power Research
Institute, by Rogers and Associates Engineering Corp., Salt Lake City, UT.

Slovic, P. and B. Fischhoff, 1980. "How Safe is Safe Enough." In R.C.
Schwing and Y.A. Albers, eds., Risk Assessment: How Safe Is Safe Enough?,
pp. 181-216, Plenum Press, NY.
Slovic, P., B. Fischhoff, S. Lichtenstein, S. Read, and B. Bombs, 1978. "How
Safe Is Safe Enough? A Psychometric Study of Attitudes Toward
Technological Risks and Benefits," Policv Sci. 9:127.
- 96 -

-------
Slovic, P., B. Fishhoff, and S. Lichtenstein, 1980. "Facts and Fears:
Understanding Perceived Risk." In R.C. Schwing and W.A. Albers, eds. Risk
Assessment: How Safe is Safe Enou~h? pp. 181-216 Plenum Press, NY.
Starmer R.J., L.G. Deering and M.F. Weber, 1988. "Performance Assessment
Strategy for Low-Level Waste Disposal Sites," Tenth Annual DOE Low-Level
Waste ManaEement Conference. August 30-Sentember 1, Denver, Colorado.
Triglio, G., 1981. Volume Reduction Techniaues in Low-Level Radioactive
Waste Management, Prepared. for U.S. Nuclear Regulatory Commission,
Division of Waste Management, by Teknekron, Inc., Berkeley, CA.
U.S. Congress, 1972. Marine Protection. Research. and Sanctuaries Act of
l2Z2, Enacted by P.L.92-532, 86 Stat. 1052, 33 U.S.C. 1401 et seq., and 16
U.S.C. 1431 et seq..
U.S. Congress, 1980. Low-Level Radioactive Waste Policy Act of 1980, P.L.
96-573, as amended by P.L. 99-240.
U.S. Congress, 1986. Low-Level Radioactive Waste Policy Act Amendments of
1985, P.L. 99-240 (H.R. 1083).
U.S. Council for Energy Awareness, 1988. Transnorting Low Level Radioactive
Waste: Ouestions and Answers, Washington, DC. October, 1988.
U.S. Department of Energy, 1988. Order 5820.2A. Chanter III. Management of
Low-Level Waste, September 26, 1988.
U.S. Environmental Protection Agency, 1977. "40 CFR 227 - Criteria for the
Evaluation of Permit Applications for Ocean Disposal of Materials," 42 FR
2476, January 11, 1977.
U. S. Environmental Protection Agency, 1977. "40 CFR 228, Cri teria for the
Management of Disposal Sites for Ocean Dumping," 42 FR 2482, January 11,
1977 .
U.S. Environmental Protection Agency, 1983. "Environmental Radiation
Protection Standards for Low-Level Radioactive Waste Disposal: Advanced
Notice of Proposed Ru1emaking," 48 FR 39563, August 31, 1983.
U.S. Nuclear Regulatory Commission, 1982. "10
Requirements for Land Disposal of Radioactive
December 27, 1982.
CFR 61
Waste," 47
Licensing
FR 57446,
U.S. Nuclear Regulatory Commission, 1986. "10 CFR 61 Branch Technical
Position Statement on Licensing of Alternative Methods of Disposal for
Low-Level Radioactive Waste," 51 FR 7806, March 6, 1986.
- 97 -

-------
- 98

-------
APPENDIX A
A REVIEW OF AGREEMENTS, STATUES, AND REGULATIONS
FOR THE DISPOSAL OF LOW-LEVEL RADIOACTIVE WASTE
- 99 -

-------
1
INTRODUCTION
Low-level radioactive waste (LLW) disposal practices date back more than
forty years, over which time both techniques-and concerns over environmental
effects have varied considerably. Consequently, laws and regulations
governing the disposal of LLW have changed. Consideration of current and
pending legislative constraints may playa key role in the selection of new
LLW disposal options. Applicable international agreements and federal
statutes/regulations for ocean dumping, as well as federal and state
statutes/regulations governing land-based disposal are reviewed in this
report.
2
REGULATIONS FOR OCEAN DISPOSAL OF LOW-LEVEL RADIOACTIVE WASTE
2.1
Background
Disposal of low-level radioactive wast~ (LLW) in the ocean was first
initiated by the United States in 1946. Between 1946 and 1970, the U. ~
placed approximately 107,000 waste containers comprising about 4.3 x 10 ~
Becquerels (Bq) in the Atlantic and Pacific Oceans. Most of this waste was
dumped prior to 1962; an Atomic Energy Commission moratorium on ocean
disposal licenses significantly reduced the practice (only about 350 packages
estimated at 8.5 x 10 2 Bq were dumped between 1963-1970) until it was
curtailed completely in 1970. The United Kingdom began ocean disposal of LLW
in 1949, and was joined by other European nations in 1967 when a cooperative
program was organized by the Nuclear Energy Agency (NEA). Since 1949, an
estimated tottl mass of 100,000 metric tons of packaged LLW accounting for
about 3 x 10 Bq has been disposed by European nations in the Northeast
Atlantic Ocean. Due to concerns over uncertainties iri environmental impacts
and widespread social protests, European ocean disposal was suspended in
1983.
Increased awareness of environmental pollution in general, and concern
over potential hazards associated with radioactive waste in particular, has
led to the promulgation of federal laws and regulations and international
agreements pertaining to the disposal of LLW at sea. Since the onset of LLW
ocean disposal, the regulatory environment has been a dynamic one, influenced
by: (i) a growing scientific understanding of the nature of waste package
performance, radionuclide transport mechanisms, and biological/environmental
effects and, (11) a changing social and political climate in the areas of
nuclear power and environmental affairs.
2.2
International Laws/Agreements
International concern over potential environmental impacts resulting
from ocean disposal of radioactive wastes was initially addressed at the
first U.N. Conference on the Law of the Sea (UNCLOS I), held in Geneva in
1958. This conference provided the first codification of the international
law of the sea, still in force today. Although unanimity among participating
nations over the advisability of ocean disposal was not reached, several
- 100 -

-------
general principles were incorporated in Article 25 of the Geneva Convention
on the High Seas that laid the groundwork for future accords [1]:

Every State shall take measures to prevent pollution of the
seas from the dumping of radioactive wastes, taking into
account any standards and regulations which may be formulated
by the competent international organizations.
.
.
All States shall cooperate with the competent international
organizations in taking measures for the prevention of
pollution of the seas or air space above, resulting from any
activities with radioactive materials or other harmful agents.
The need for international research, supervision and control of waste
disposal was further recognized in the recommendation to the International
Atomic Energy Agency (IAEA) and other organizations to [2]:
.pursue whatever studies and take whatever action is
necessary to assist States in controlling the discharge or
release of radioactive materials to the sea, in promulgating
standards; and in drawing up internationally' acceptable
regulations to prevent pollution of the sea by radioactive
materials in amounts which would adversely affect man and his
marine resources.
Following the recommendations of UNCLOS I, the IAEA assembled a panel of
experts on ocean disposal of radioactive wastes which issued a report of its
findings in 1961. The resultant document (sometimes referred to as the
BrYnielsson Report) [3] recommended a ban on high-level radioactive waste
disposal at sea and that low-level waste disposal be conducted under
controlled and specified conditions, on a site-specific basis. It also
proposed that wastes be certified and internationally registered, that dump-
sites be designated, and that operational procedures for disposal be
formulated.
In 1967 the NEA of the Organization for Economic Cooperation and
Development (OECD) agreed to coordinate international disposal of LLW at sea.
It set up the Multilateral Consultation and Surveillance Mechanism for Sea
Dumping of Radioactive Waste to provide a disposal system at a Northeast
Atlantic site. Guidelines on packaging and disposal methods were issued in
conj unc tion with the IAEA recommendations. [4,5] The NEA also called for a
one-year notification requirement prior to disposal operations, an
internationally supervised surveillance program, and a system of permanent
records to be kept by the International Maritime Organization (IMO). A
coordinated research program to carry out a thorough radiological assessment
of the site was initiated in 1980 at the suggestion of the U. S. This
Coordinated Research and Environmental Surveillance Program (CRESP), still
ongoing, contributed significantly to the 1985 site suitability review.
These reviews are required every five years.
The general obligation of nations to preserve
reaffirmed at the United Nations Conference on
Stockholm, Ju~e 1972. The Stockholm conference
Convention on the Prevention of Marine Pollution
the marine environment was
the Human Environment in
was the impe tus for the
by Dumping of Wastes and
- 101 -

-------
Other Matter (better known as the London Dumping Convention, LDC) in November
of 1972. The LDC is recognized as the most comprehensive international
regulation of marine pollution by dumping undertaken to date [6]. Some fifty
nations, including the major maritime countries, are now Contracting Parties
to the Convention. Three categories of waste are defined under the LDC: {i)
materials prohibited from dumping, including high-level wastes as defined by
the IAEA, and listed in Annex I, (ii) materials requiring a special permit
for dumping, including low-level wastes, listed in Annex II, and, (iii) all
other materials requiring a prior general permit. Permits are issued by
appropriate national authorities following the guidance provided in Annex
III. Factors to be considered for issuance of permits include waste
characteristics (e.g., type and quantity of contained activity, radioactive
half-life, toxicity, bioaccumulation, dose-response); disposal site
characteristics (e.g., location, depth); disposal methods (e.g., waste
packaging, disposal density); effects on other uses of the ocean (e.g.,
fishing, navigation, other industrial uses, amenities); and availability of
alternative land-based disposal options. The LDC also authorized the IAEA to
formulate recommendations on low-level waste dumping and requires the
reporting of permits (issued by national authorities) to the.' IMO.
Responding to the LDC, the IAEA first formulated its recommendations in
1974 and revised them in 1978 [7]. These recommendations call for detailed
environmental and ecological assessments to be submitted to IMO with each
permit application. The assessment and permit procedure are designed to
ensure that ocean disposal of low-level waste will involve no unacceptable
degree of hazard to human health, harm to living resources and marine life,
damage to amenities, or interference with other legitimate uses of the sea.
To this end, limits on allowable releases of radioactivity at a given site
and for a finite ocean basin are specified (summarized in Table I) and the
need for suitable packaging to optimize containment is emphasized. Specific
site selection criteria are given as follows:
.
Sites should be located between 500S and SOoN latitudes to
avoid sources of bottom water (characterized by strong vertical
mixing) and areas of high biological productivity.
.
Depth should be ~ 4000 m (biological, chemical, physical and
topographical gradients are generally low, bottom water
circulation is slow, and organic carbon in the sediments tends
to be low).
.
Sites should be away from continental margins (to avoid regions
of high biological productivity. and active resource
exploration/exploitation), and areas where geologic hazards
(e.g., submarine slides, volcanos and earthquakes) reduce
stability.
.
Areas encompassing potential seabed resources, trans-oceanic
cables, commercial fishing trawlers, or areas that are
difficult to navigate, should be avoided.
.
Designated sites ~hould be as small as possible with a maximum
size of 10,000 km. Precise site coordinates are required and
navigational aids for relocation are desirable.
- 102 -

-------
.
The number of disposal sites should be strictly limited.
.
Features such as submarine canyons that can adversely impact
the rate of exchange between deep and surface waters near the
continental shelf should be avoided.
.
Sites should be selected for convenience of disposal operations
and to minimize navigational traffic.
.
Bottom current shear stress should not exceed critical
erosional shear stress to prevent high rates of resuspension
and erosion of sediments.
.
Monitoring of the area near the disposal site should be carried
out to examine waste package and site performance.
Another international treaty that could affect ocean disposal of
radioactive waste is the Convention on the Law of the Sea (LOS). The U.S.
has thus far declined to sign the agreement, but it has been signed by 132
nations and ratified by 11. If 60 nations ratify the document, the LOS treaty
will govern participants' international maritime affairs. Although the LOS
convention does not directly address the issue of radioactive waste disposal,
it does strongly emphasize protection of the marine environment. Article 194
declares, for example, that "States shall take ...all measures.. .necessary to
prevent, reduce and control pollution of the marine environment from any
source..." Articles 232 and 235 hold a nation financially responsible for
damages caused by pollution from its sources.
The U.S. policy as negotiated by the Department of State (DOS) has been
that any substantive decisions with respect to ocean disposal of low-level
radioactive waste must be based on sound technical and scientific grounds.
In 1986, Resolution (28): 10 to the LDC called for a voluntary moratorium
until certain scientific/technical, as well as social, political and economic
issues were addressed. To accomplish this, in 1987, an Intergovernmental
Panel of Experts on Radioactive Waste Disposal at Sea was formed, consisting
of two working groups: Working Group I is addressing social, political and
economic questions, and Working Group II is addressing scientific and
technical issues.
The U.S. is participating in the deliberations of both working groups.
However, particular focus has been given to assisting the international
organizations charged with resolving the scientific and technical questions
identified in Working Group II. A key question to be addressed
internationally is the scientific and technical issues related to the
comparative ocean and land-based options for disposal of LLW and the
associated costs and risks. The IAEA was recently asked to address these
issues pursuant to its status as the internationally recognized cognizant
authority. Additional details of current IAEA activities and U.S.
involvement in comparing risks of land-based and ocean disposal options are
included in Appendix A.
- 103 -

-------
2.3
Federal Laws(Regu1ations
In the U. S., the
1970 was designed to
projects and encourage
Its stated purpose was
National Environmental Policy Act (NEPA) enacted in
focus attention on environmental impacts of federal
greater public participation in environmental affairs.
to:
declare a national policy which will encourage productive and
enj oyable harmony between man and his environment; to promote
efforts which will prevent or eliminate damage to the environment
and biosphere and stimulate the health and welfare of man; to
enrich the understanding of the ecological systems and natural
resources important to the Nation.
In keeping with environmental policy set forth by NEPA, a number of specific
environmental laws were passed that cover air quality (Clean Air Act), water
quality (Federal Yater Pollution Control Act, Safe Drinking Yater Act, and
Marine Protection, Research and Sanctuaries Act), and toxic materials/waste
disposal (Toxic Substances Control Act, Resource Conservation and Recovery
Act, and Comprehensive Environmental Response, Compensation and Liability
Act). .
The Marine Protection, Research and Sanctuaries Act of 1972 (MPRSA)
recognized that unregulated dumping of radioactive and non-radioactive
materials into the ocean can endanger human health, welfare, and amenities,
and the marine environment, ecosystems, and resources [8]. It establishes as
u.S. policy to:
regulate the dumping of all types of materials into ocean waters
and to prevent or strictly limit the dumping into ocean waters of
any material which would adversely affect human health, welfare,
or amenities, or the marine environment, ecological systems or
economic penalties.
Dumping of high-level radioactive waste and radiological, chemical, or
biological warfare agents are prohibited. The Environmental Protection
Agency (EPA) is empowered to issue waste dumping permits within the policy
guidelines of the Act. Minimum criteria for evaluation of dumping permit
applications as specified in Section 102 of MPRSA are listed in Table A-2.
A number of regulations
Ocean Dumping Regulations of
these regulations and define
pursuant to MPRSA have been issued by EPA. EPA
1977, found in 40 CFR 220 establish the scope of
five categories of ocean dumping permits [9]:
.
General permits are issued for dumping of certain materials
that have minimal adverse environmental impact and are
generally disposed of in small quantities or for specific
classes of materials that must be disposed of .in emergency
situations.
.
Special permits, valid for three years, are issued
materials that meet 40 CFR 227 criteria (see below).
for
- 104 -

-------
.
Emergency permits are issued for materials specified in 40 CFR
227.6 that are present in concentrations above trace levels,
provided that an imminent threat to human health and safety is
demonstrated and no feasible alternative options exist.
.
Interim permits were issued for some wastes prior to April,
1978 during the regulatory phase-in period.
.
Research permits are issued on a short-term basis (18 months)
for disposal of materials associated with a research. project
whose merits outweigh potential environmental or other damage.
EPA permit application requirements and evaluation criteria are contained in
40 CFR 221 and 227, respectively [10,11]. These requirements include
demonstrating compliance with environmental impact criteria (Subpart B),
absence of. acceptable alternatives (Subpart C), and lack of adverse effects
on esthetic, recreational, or economic resources (Subparts D and E). LLW
must be containerized (40 CFR 227.7) and meet the following conditions (40
CFR 227.11):
.
the materials to be disposed of decay, decompose or radiodecay
to environmentally innocuous materials within the life
expectancy of the containers and/or their inert matrix;
.
materials to be dumped are present in such quantities and are
of such nature that only short-term localized adverse effects
will occur should the conta:~'ers rupture at any time;
.
containers are dumped at depths and locations where they will
cause no threat to navigation, fishing, shorelines, or beaches.
Criteria for managing waste disposal sites in the ocean are contained in 40
CFR 228 [12].
A January 1983 amendment to MPRSA imposed a two-year moratorium on ocean
disposal of low-level radioactive wastes and a more stringent, supplementary
set of permit requirements following the moratorium [13]. For a LLW ocean
disposal permit, it must be demonstrated that:
.
the proposed dumping is necessary to conduct research either i)
on new technology related to ocean dumping, or ii) to determine
the degree to which the dumping of such substance will degrade
the marine environment;
.
the scale of the proposed dumping is limited to the smallest
amount of such material and the shortest duration of time that
is necessary to fulfill the purposes of the research, such that
the dumping will have minimal adverse impact upon human health,
welfare, and amenities, and the marine environment, ecological
systems, economic potentialities, and other legitimate uses;
.
after consultation with the Secretary of Commerce,
potential benefits of such research will outweigh any
adverse impact; and
the
such
- 105 -

-------
.
the proposed dumping will be preceded by appropriate baseline
monitoring studies of the proposed dump site and its
surrounding environment.
In addition to complying with other requirements, a Radioactive Material
Disposal Impact Assessment (RMDIA) must be included with the permit
application. Information to be included in the assessment is listed in Table
A-3. Under the amended law, final approval for LLW ocean disposal permits is
also required by a joint resolution of Congress. EPA is currently developing
updated criteria for ocean disposal of LLW to be included in the Agency's
revisions to the 1977 Ocean Disposal Regulations.
Another recently enacted amendment to MPRSA known as the "Ocean Dumping
Ban Act of 1988" bans dumping of sewage sludge and industrial waste into
ocean waters after December 31, 1991. The amendment defines industrial waste
as: "any solid, semisolid, or liquid waste generated by a manufacturing or
processing plant, other than an excluded material." It is not yet clear
whether this definition will be interpreted to include low-level radioactive
waste. Previous congressional actions involving LLW, however, contained
explicit reference to it as such, indicating that the ban on dumping
industrial waste was probably not intended to cover LLW.
3
REGULATIONS FOR LAND-BASED DISPOSAL OF LLW
3.1
Federal Laws and Regulations
Regulations and responsibilities t'br1and-based disposal of LLW' are
outlined in two federal laws: (i) the Atomic Energy Act of 1954, as amended,
42 U.S.C. 2011 et seq., and (ii) the Low-Level Radioactive Waste Policy Act
of 1980 (P.L. 96-573) as amended by the Low-Level Radioactive Waste Policy
Act Amendments of 1985 (P.L. 99-240).
The Atomic Energy Act of 1954 (AEA}, as amended by the Energy
Reorganization Act of 1974, empowers the Nuclear Regulatory Commission (NRC)
to license and regulate commercial LLW disposal and grants the Department of
Energy responsibility for LLW generated by defense-related activities. It
also provides states the opportunity to license and regulate disposal
practices by authorizing special agreements between NRC and individual
states. Under Section 274.b of the Atomic Energy Act, a state can assume
licensing and regulatory authority and become an "Agreement State" if it
agrees to enact and uphold statutes comparable to existing NRC regulations.
State involvement in LLW disposal policy was expanded by the Low-Level
Radioactive Waste Policy Act of 1980 (LLRWPA) and its amendments of 1985
(LLRW'PAA), which mandated state responsibility for developing new land-based
disposal capacity. Under the LLRWPA, individual states or groups of states
that form compacts are responsible for disposal of all LLW generated within
their borders, except for waste produced by federal facilities. The LLRWPAA,
which superceded the 1980 law, went even further in spelling out specific
incentives for implementation. After a phase-in period ending in 1992,
states or compacts can refuse to accept waste generated elsewhere, or may
impose severe economic penalties if they choose to accept waste from other
- 106 -

-------
regions/states. Thus the maj or thrust of the LLRWPAA was to decentralize
shallow land burial and "encourage" the development of additional disposal
capacity. In addition to its main provisions the LLRWPAA:

allows the NRC to intercede and grant emergency access to a
regional disposal facility if an immediate and serious threat
to public health and safety or security occurs that cannot be
mitigated by alternative means;
.
.
requires generators of LLW to implement volume reduction to the
maximum extent practicable or face economic penalties;
.
directs the Department of Energy (DOE) to: (i) provide
states/compacts with technical assistance in the areas of site
selection, alternative technologies for disposal, volume
reduction and management techniques to maximize disposal
capacity, transportation issues, health and safety
considerations for storage, shipment, and disposal, and
development of a computerized data base to monitor LLW disposal
management; and (ii) to provide financial assistance through
1993 to implement the Act;
.
requires the DOE Secretary to prepare yearly progress reports
to Congress on status of LLW disposal;
.
requires NRC, in consultation with the States and .other
interested parties, to: (i) identify alternative methods for
LLW disposal, (ii) develop application requirements for
alternative methods, and (Hi) publish technical criteria for
licensing of alternative methods;
.
directs NRC to establish criteria for evaluating LLW that
contains radionuclides in sufficiently low concentrations to be
considered "below regulatory concern".
Under authority of the Atomic Energy Act of 1954, NRC implemented a
comprehensive set of LLW shallow land disposal regulations in 1983 (10 CFR
61) aimed both at waste generators and burial site operators [14]. Major
portions of these requirements are reviewed below.
Responding to needs and requests of the public, Congress, industry, NRC
and other federal agencies, the purpose of 10 CFR 61 was to provide licensing
procedures, performance obj ectives and technical requirements for licensing
of LLW disposal sites on a nationally consistent basis. It specifically
addresses procedural and technical requirements of near-surface disposal
(i.e., ~ 30 m) and expressly does not apply to other methods such as burial
below 30 m or ocean dumping. Objectives of 10 CFR 61 include insuring
protection of the public from radioactive releases, protection of individuals
from inadvertent intrusion at the site after closure, radiological protection
for operators, and long-term site stability.
The regulatory approach taken to achieve these objectives is to maintain
sta~ility. of both waste packages and disposal site. Since leaching of
rad1.onucl1.des by natural precipitation is the main environmental pathway
- 107 -

-------
leading to potential human exposures (via ground water contamination),
stability is critical. For waste packages, physical integrity is important
to prevent slumping or collapse of the disposal trench that can lead to water
intrusion, and to maintain minimum surface areas subject to leaching;
chemical stability reduces potential for interactions that can enhance
leachability. Degradation of site stability (e.g., by subsidence) can
seriously alter important site characteristics such as permeability and
drainage. Recognizing that radioactive hazards are both isotope- and
concentration-specific, NRC developed a waste classification scheme that
establishes three categories of waste:
.
Class A: Minimally contaminated waste which because of its
physical form (e.g., trash and other bio-degradables) is
generally segregated from other waste at disposal;
.
Class B: Contain higher concentrations of short-lived
radionuclides, must meet more rigorous stability requirements;
.
Class C: Contain significant concentrations of long-lived
radionuclides, thus must meet stability criteria and additional
measures against inadvertent intrusion.
Wastes are classified based on consideration of specific concentrations
of long- and short-lived isotopes, as shown in Table A-4. If concentrations
exceed the maximum specified for Class C, the waste is considered unsuitable
for disposal by shallow-land burial. For mixtures of radionuclides,
classification is determined by summing ratios of concentrations/limits for
each isotope as described in 10 CFR 61.55 (a) (7). Wastes containing
radionuclides other than those listed in Table A-4 are considered Class A.
Disposal requirements vary by classification, but all wastes must meet the
following minimum criteria designed to facilitate safe handling:
.
Waste must not be
fiberboard boxes.
disposal
or
cardboard
packaged
for
in
.
Liquid waste must be solidified or, if Class A, may be packaged
in sufficient absorbent material to absorb twice the volume of
the liquid.
.
Solid waste containing liquid shall contain as little free
standing and noncorrosive liquid as is reasonable achievable,
but in no case shall the liquid exceed 1 % of the volume.
.
Waste must not be readily capable of detonation or of explosive
decomposition or reaction at normal pressures and temperatures,
or of explosive reaction with water.
.
Waste must not contain, or be capable of generating, quantities
of toxic gases, vapors, or fumes harmful to persons
transporting, handling, or disposing of the waste. This does
not apply td radioactive gases packaged as described below.
- 108 -

-------
.
Waste must not be pyrophoric.
in waste shall be treated,
nonflammable.
Pyrophoric materials contained
prepared, and packaged to be
.
Waste in a gaseous form must be packaged at a pressure that
does not exceed 1.5 atm at 200C. Total activity must not
exceed 100 Ci per container.
.
Waste containing hazardous, biological pathogenic, or
infectious material must be treated to reduce to the maximum
extent practicable the potential hazard from the non-
radiological materials.
In addition to these minimum requirements, Class Band C wastes must
also meet three criteria intended to maintain waste and site stability and
reduce exposure to an inadvertent intruder by ensuring waste forms are still
identifiable. These are:
.
Waste must maintain structural stability und~r expected
disposal conditions such as weight of overburden and compaction
equipment, presence of moisture, and microbial activity, and
internal factors such as radiation effects and chemical
changes. Structural stability can be provided by the waste
itself, processing to a stable form, or placing the waste in a
disposal container or structure that provides stability after
disposal.
.
Class Band C liquid wastes must be converted into a form that
contains as little free standing and noncorrosive liquid as is
reasonable achievable, but in no case can the liquid exceed 1%
of the volume of the waste when in a disposal container
designed to ensure stability, or 0.5% of the volume of the
waste for waste processed to a stable form.
.
Void spaces within waste and between waste and its package must
be reduce4 to the extent practicable.
Further clarification on stability requirements was issued by NRC in the
form of a Branch Technical Position on Waste Form [15], which identified a
series of waste form performance tests that could be used to demonstrate
long-term stability.
In addition to waste form standards, 10 CFR 61 establb;es a set of
technical requirements for land disposal facilities. These include
specifications for: (i) preliminary site suitability for shallow land burial
that cover geology, hydrology, population, etc.; (ii) site design, covering
features to protect against water intrusion and ensure stability after
closure; (iii) operation and closure, such as segregation of Class A wastes,
placement of Class C wastes to guard against inadvertent intrusion, filling
of yoids, etc.; and (iv) environmental monitoring before, during, and after
disposal operations are complete. Radiation exposure standards for the
general public included in 10 CFR 61.41 limit releases to the environment by
all pathways, such that annual average dose equivalents do not exceed 25 mrem
to the whole body, 75 mrem to the thyroid, and 25 mrem to any other organ.
- 109 -

-------
Department of Energy policy and regulations concerning LLW treatment and
disposal are covered in the recently amended DOE Order 5820.2A. .[16] Similar
to 10 CFR 61, this order establishes a maximum effective dose equivalent of
25 mrem/yr to any member of the public from external exposure to radioactive
materials released from DOE LLW disposal facilities, while "reasonable
e.fforts" must be made to maintain releases as low as reasonably achievable
(ALARA) . Atmospheric releases must meet requirements for hazardous air
emissions in 40 CFR 61. Specific requirements for waste performance
assessment, generation, minimization, characterization, acceptance criteria,
treatment, storage, shipment, and disposal are included. In addition,
criteria for disposal site selection, design, operations, closure, and
monitoring are discussed.
In August 1983, EPA issued an Advanced Notice of Proposed Ru1emaking
(ANPRM) stating the agency's intention to develop generally applicable
standards for LLW disposal and requesting public comment on its form and
content. [17,18]. The ANPRM consists 6f two parts: (i) 40 CFR 193
encompasses LLW treatment and disposal at DOE and commercial facilities, and
(11) 40 CFR 764 which deals with NARM (naturally occurring and accelerator
produced radioactive waste) under authority granted by the Toxic Substances
Control Act. NARM wastes are not covered by existing NRC or DOE rules.
40 CFR 193 consists of three subparts that address allowable exposures
to the public from LLW management, storage, and disposal facilities, and
extend groundwater protection standards to encompass these activities.
Subpart A establishes environmental standards for management, processing and
storage and Subpart B presents environmental standards for land disposal.
Both require that activities be conducted in such a manner that combined, no
member of the public in the general environment shall receive an annual
effective dose equivalent of more than 25 mrem from all routes of exposure.
Subpart A also defines the category of "below regulatory concern" (BRC) for
LLY that contain sufficiently low concentrations of radioactivity that their
disposal, in combination with all other BRC waste streams, would not expose
any member of the public to an annual effective dose equiva1e~t of more than
4 mrem in any year. These wastes are exempt from disposal requirements in 40
CFR 193. Subpart B requires: (i) credit for active institutional controls
can only be taken for 1.00 years after disposal in order to meet exposure
standards ;' (ii) danger warnings must be posted in the most permanent way
practicable to discourage inadvertent intrusion; (i11) sites may not be
located in areas where there is a reasonable potential for future exploration
of resources; and (iv) monitoring after disposal is required without
jeopardizing integrity of isolation. Subpart C proposes groundwater
protection requirements for facilities regulated under Subparts A or B. Two
options are proposed and differ only with respect to Class II groundwater.
The first option stipulates that disposal of radioactive waste cannot result
in: (i) any increase in levels of radioactivity for all Class I groundwaters,
or (ii) any increase in the levels of radioactivity for Class II
groundwaters from high yield aquifers such that an individual can receive
more than 4 mrem annual effective dose equivalent by drinking two liters per
day of affected groundwater, while the remainder of Class II groundwaters are
protected within the 25 mrem/yr limits of Subparts A and B. Class III
groundwaters are also protected within the 25 mrem/yr limits of Subparts A
and B unless such groundwater is highly interconnected to a higher class of
groundwater, in which case the more restrictive standards are extended to the
- 110 -

-------
Class
would
Class
III groundwater. The second groundwater protection option proposed
apply the 4 mrem/yr limit to all Class II groundwater while Class I and
III groundwater would be protected as in the first option.
Standards for management and disposal of NARK wastes are contained under
Subchapter R (Toxic Substances Control Act regulations), 40 CFR 764.
Appendix A of the regulation contains a classification system based on the
approach used in Appendix A of 10 CFR 61. NARK wastes greater than 2 nCi/gm
(with the exception of certain consumer items) would be disposed in an AEA
regulated LLY disposal facility and meet shipping manifest requirements given
in Appendix B of the regulation.
3.2
State Regulations
By statutory authority of the Agreement State Amendment to the AEA [19],
states may assume certain regulatory authority from NRC over radioactive by.
products, source materials, and small quantities of special nuclear
materials. Under this agreement, a state regulates institutional and
industrial radioactive waste generators and commercial LLY disposal sites,
but does not have jurisdiction over nuclear power plants or federal
facilities. Alternatively, an NRC policy statement issued in 1981 allows
states to license and regulate LLW disposal facilities only, by seeking to
become a Limited Agre~ment State [20]. In either case, to be eligible a
~tate must pass enabling legislation and have an adequate program (compatible
with NRC regulations) to protect the public health, safety and the
environment. States are also required to have staff with training and
expertise in the areas of radiation protection, State law, regulations and
procedures. Twenty-nine states are currently classified as Agreement States.
These are listed in Table A-5. .
LLY disposal activities that can be regulated by states include site
selection, leasing/contracting of site to an operator, facility licensing,
operations and closure, and long-term care following closure. State
responsibilities vary depending on whether the state is a host state, i.e.,
provides disposal capabilities .within its borders, and whether it is an
Agreement State, a limited Agreement State, or a non-agreement state. Host
state responsibilities are summarized in Table A-6 by disposal activity and
type/status of state agreement [21].
The AEA as amended in 1959 requires agreement states to meet or exceed
federal standards in regulating LLW disposal. [22] Some st,:;.tes have gone
beyond federal standards by expressly prohibiting certain disposal options in
the relevant enabling legislation. For example, six of the nine regional
compacts (representing 28 states) and four unaffiliated states have all
passed measures banning the use of shallow-land burial for LLW disposal [23].
The New York State Low-Level Radioactive Waste Management Act includes the
following passage concerning allowable disposal options [24]:

...the Department [of Environmental Conservation] shall publish
d:aft regulat~ons. which specify the criteria for siting permanent
d1SP?s~l fac11iues and shall promulgate final regulations...
spec1f1c to the types of disposal methods which may be employed
at a permanent disposal site and shall include criteria for: (a)
- 111 -

-------
above ground, engineered, monitored disposal; (b) underground
mined repository disposal; and (c) where practicable, other
disposal methods for which there are applicable regulations but
in no event including shallow land burial [emphasis added].
In this case, shallow land burial is specifically defined as
"emplacement of low-level radioactive waste in or within the upper 30 meters
of the surface of the earth in trenches, holes, or other excavations in which
only soil provides structural integrity, a barrier to migration ot low-level
radioactive waste from or subsurface water into such excavation, or in a
manner that fails to allow during the institutional control period for
monitoring and control of releases of radioactivity." In other words, near-
surface disposal will be allowed only if certain engineering controls are
incorporated to maintain structural integrity, minimize water intrusion and
leaching, and monitor site performance.
- 112 -

-------
------------------------------------------------------------------------
Table A-l.
Radioactive waste release limits for ocean disposal
------------------------------------------------------------------------
Radioactive
Waste Type
Release Rate Limits (Ba/yr)
Single Dumpsite Finite Ocean Basin
------------------------------------------------------------------------
alpha emittersa

beta and gammab
(t1/2 ~ 0.5 yr)
3.7 x 1015
3 . 7 x 1017
3.7 x 1015
3.7 x 1018
beta and gammaC
(t1/2 ~ 0.5 yr)

------------------------------------------------------------------------

a Limited to 3.7 x 1014 Bq/yr for 226Ra and supported 210po.
b Excluding 3H. Includes beta and gamma emitters of unknown t1/2'
c Including 3H.
Source: Reference 5
3.7 x 1021
3.7 x 1022
------------------------------------------------------------------------
- 113 -

-------
------------------------------------------------------------------------
Minimum criteria for evaluation of ocean disposal permits a
Table A- 2.
------------------------------------------------------------------------
.
Need for proposed dumping.
.
Effect of dumping on human health and
economic, esthetic, and recreational values.
including
welfare,
.
Effect of dumping on fisheries resources,
shellfish, wildlife, shore lines and beaches.
fish,
plankton,
.
Effect of dumping on marine ecosystems, particularly with respect
to: i) the transfer, concentration, and dispersion of such
material and its by-products through biological, physical and
chemical processes, ii) potential changes in marine ecosystem
diversity, productivity, and stability, and iii) species and
community population dynamics.
.
Persistence and permanence of the effects of dumping.
.
Effect of
materials.
concentrations
of
volumes
particular
dumping
and
.
Appropriate locations and methods of disposal or recycling,
including land-based alternatives and probable impact of
requiring use of alternate locations or methods upon
considerations affecting the public interest.
.
Effect on alternate uses of oceans such as scientific study.
fishing, and other living resource exploitation, and nonliving
resource exploitation.
.
In designating recommended sites, EPA shall utilize wherever
feasible, locations beyond the Continental Shelf.
------------------------------------------------------------------------
a Adapted from Reference 6
------------------------------------------------------------------------
- 114 -

-------
------------------------------------------------------------------------
Table A-3. Minimum requirements for Radioactive Material Disposal
Impact Assessment a
------------------------------------------------------------------------
.
Listing of all radioactive materials
disposed, the number of containers to
diagrams of each container, the number
in each container, and the exposure
inside and outside of each container;
in each container to be
be dumped, the structural
of curies of each material
levels (in rems) at the
.
An analysis of the environmental impact of the proposed action on
human health and welfare and marine- life, at the site where the
applicant desires to dispose of the material;
.
Any adverse environmental effects at the site which cannot be
avoided should the proposal be implemented;
.
An analysis of the resulting environmental and economic-
conditions if the containers fail to contain the. radioactive
waste material when initially deposited at the specific site;
.
A plan for the removal of the disposed nuclear material if the
container leaks or decomposes;
.
A determination by each affected State whether the proposed
action is consistent with its approved Coastal Zone Management
Program;
.
An analysis of the
resources;
impact upon other users of marine
economic
.
Alternatives to the proposed action;
.
Comments and results of consultation with State officials and
public hearings held in the coastal States that are nearest to
the affected areas;
.
A comprehensive monitoring plan to be carried out by the
applicant to determine the full effect of the disposal on the
marine environment, living resources, or human health. The plan
shall include, but not be limited to, exterior monitoring of
container radiation, taking of water, sediment, fish and benthic
animal samples, and the acquisition of such other information as
the [EPA] Administrator may require;
Any other information the [EPA] Administrator may require
order to determine the full effects of such disposal.

;-----------------------------------------------------------------------

Source: Reference 11
.
in
-----------"-------------------------------------------------------------
- 115 -

-------
-----------------------------------------------------------------------------
Table A-4.
NRC waste classification scheme contained in 10 CFR 61.55 a
-----------------------------------------~~;~~-~~~~~~~;~~;~~~:-~;i~3-------
----------------------------------
Radionuc1ide
Class A
Class B
Class C
-----------------------------------------------------------------------------
Long-lived isotones
C-14
C-14
Ni-59
Nb-94
Tc-99
1-129
Alpha TRUs b
Pu-24l
Cm-242
(in activated metal)
(in activated metal)
(in activated metal)
(in activated metal)
0.8
8.
22.
0.02
0.3
0.008
10.c
350.c
2,OOO.c
8.
80.
220.
0.2
3.
0.08
100.c
3500.c
20,OOO.c
Short-lived isotones
Total of all nuclides with
Tl/2 < 5 yrs
H-3
Co-60
Ni-63
Ni-63 (in activated metal)
Sr-90
Cs-l37
700.
40.
700.
3.5
35.
0.04
1.
d
d
d
70.0
700.0
150.
33.
d
d
d
700.
7,000.
7,000.
4,600.
-----------------------------------------------------------------------------
a Based on Tables 1 and 2 and Section 61.~5 in Ref. 12.
'b Alpha emitting transuranic elements with T1/2 > 5 yrs.
c Concentration in nCi/g.
d No limits established for these nuclides in Class B or C wastes. Practical
considerations (e.g., external radiation, heat generation) will limit
concentrations.
These wastes are Class B unless concentrations of other nuclides in table
make them Class C.
-----------------------------------------------------------------------------
- 116 -

-------
------------------------------------------------------------------------
Table A-5.
Agreement States as of October 1988 a
------------------------------------------------------------------------
Alabama
Arizona
Arkansas
California
Colorado
Florida
Georgia
Idaho
Illinois
Iowa
Kansas
Kentucky
Louisiana
Maryland
Mississippi
Nebraska
Nevada
New Hampshire
New Mexico
New York
North Carolina
North Dakota
Oregon
Rhode Island
South Carolina
Tennessee
Texas
Utah
Washington

~-----------------------------------------------------------------------

Source: Reference 25
--------------------------------~---------------------------------------
- 117 -

-------
------------------------------------------------------------------------
Table A-G.
State regulatory responsibilities for LLW disposal
------------------------------------------------------------------------
Site Selection -
All Host States can:
. influence siting through zoning laws,
criteria and procedures.
. approve or reject a new disposal site.
land
use
and
siting
~ase/Contract for DisDosal Facility -
All Host States can:
. select operator.
. require operator conform to license; enforce compliance.
. set additional requirements for records, reporting systems,
security, buffer zone, and closure etc.
. as landlord, exercise oversight and management responsibility,.
within radiological and safety guidelines.
. require payment of local business taxes, payment in lieu of State
taxes, incentives to State and local communities, fees and
surcharges, financial arrangements for closure, extended care,
insurance, surety (consistent with and in addition to 10 CFR 61).
Licensin2 of Facilities -
Agreement States can:
. license treatment, storage and disposal facilities that are not
part of facilities licensed only by NRC (reactors and federal
facilities) .
. regulate operations, monitoring, radiation control, performance
objectives, site design, assurances of adequate financing for
operations, closure, and stabilization (compatible with NRC
regulations) .
. when provided in State law, conduct environmental reviews
necessary for licensing.
Limited Agreement States can:
. license disposal facilities that are not part of those licensed
only by NRC. (NRC licenses treatment and storage facilities)
. regulate disposal site operations, monitoring, radiation control,
performance objectives, site design, assurance of adequate
financing for operations, closure, and stabilization.
. when provided in State law, conduct environmental reviews
necessary for licensing.
Non-Agreement States can:
. provide review and input to NRC licensing process particularly
under NEPA, and can be party to NRC licence hearings if held;
conduct State and local hearings. NRC is responsible for
licensing and regulation of disposal facilities (specifies site
operations, monitoring, radiation control, performance objectives
etc. )
------------------------------------------------------------------------
- 118 -

-------
------------------------------------------------------------------------
Table A-6.
cont.
------------------------------------------------------------------------
Operations and Closure of Facilities -
Agreement States can:
. regulate, monitor, inspect and enforce regulations for treatment,
storage and disposal facilities.
Limited Agreement States can:
. regulate, monitor, inspect and enforce regulations for disposal
facilities.
. pursuant to agreement with NRC (under section 274i of the AEA,
inspect NRC licensed facilities, notify NRC of violations, and
enforce requirements applicable under state law. NRC regulates,
monitors, inspects, and enforces regulations for treatmen~ and
storage facilities.
Non-Agreement States can:
. pursuant to agreement with NRC (under section 274i of AEA),
inspect NRC licensed facilities, notify NRC of violations, and
enforce requirements applicable under state law. NRC regulates,
monitors, inspects ,and enforces regulations for treatment and
storage facilities.
Lon2-Term Care -
All Host States can:
. carry out long-term institutional care responsibilities under
terms of license and operator's lease or contract; or
. request consideration by the federal government to carry out
long-term care of the facility.
------------------------------------------------------------------------
- 119 -

-------
REFERENCES
1.
United Nations Conference on the Law of the Sea (UNCLOS), Convention on
the High Seas, 13 U.S.T. 2312, TIAS No. 5200, 450 U.N.T.S. 82, entered
into force September 30, 1962, U.N. Conference on the Law of the Sea,
Geneva, 1958.
2.
United Nations Conference on the Law of the Sea (UNCLOS), Resolutions
Adopted by the U.N. Conference on the Law of the Sea, Pollution of the
High Seas by Radioactive Materials, 450 U.N.T.S. 58, U.N. Conference on
the Law of the Sea, April 27, 1958.
3.
International Atomic Energy Agency, Radioactive Waste Disposal into the
Sea, Safety Series No.5, International Atomic Energy Agency, Vienna,
1961.
4.
Nuclear Energy Agency, Guidelines for Sea Dumping
Radioactive Waste, Revised Version, NEA, Organization
Cooperation and Development, Paris, April 1979.
Packages of
for Economic
5.
Nuclear Energy Agency, Recommended Operational Procedures for Sea
Dumping of Radioactive Waste, NEA, Organization for Economic Cooperation
and Development, Paris, April 1979.
6.
Nuclear Energy Agency. Seabed Disposal of High-Level Radioactive Waste -
A Status Report on the NEA Coordinated Research Programme, NEA,
Organization for Economic Cooperation and Development, Paris, 1984.
7.
International Atomic Energy Agency, Convention on the Prevention of
Marine Pollution by Dumping of Wastes and Other Matter. The IAEA Revised
Definition and Recommendations of 1978 Concerning Radioactive Wastes and
Other Radioactive Matter, Referred to Annex I and II of the Convention.
Information Circular INFCIRC/205/Add.l/Rev.1, International Atomic
Energy Agency, Vienna, 1978.
8.
Marine Protection, Research, and Sanctuaries Act of 1972, Enacted by
P.L.92-532, 86 Stat. 1052, 33 U.S.C. 1401 et seq., and 16 U.S.C. 1431 et
seq., October 23,1972.
9.
40 CFR 220.3, Categories of Permits, 42 FR 2468, January 11,1977.
10.
40 CFR 221, Applications for Ocean Dumping Permits Under Section 102 of
the Act, 42 FR 2470, January 11, 1977.
11.
40 CFR 227, Criteria for the Evaluation of Permit Applications for Ocean
Disposal of Materials, 42 FR 2476, January 11, 1977.
12.
40 CFR 228, Criteria for the Management of Disposal Sites for Ocean
Dumping, 42 FR 2482, January 11, 1977.
13.
Amendment to MPRSA contained in P.L. 97-424, January 6, 1983.
14.
10 CFR 61, Licensing Requirements for Land Disposal of Radioactive
Waste, 47 FR 57446, December 27, 1982.
- 120 -

-------
15.
16.
u.s. Nuclear Regulatory Commission, Branch Technical Position on Waste
Form, U.S. NRC, Washington, DC, May 1983.
U. S. Department of Energy, Order 5820. 2A, Chapter III, Management of
Low-Level Waste, September 26, 1988.
17.
U.S. Environmental Protection Agency, Environmental Radiation Protection
Standards for Low-Level Radioactive Waste Disposal: Advanced Notice of
Proposed Rulemaking, 48 FR 39563, August 31, 1983.
18.
Gruhlke, J.M., F.L. Galpin, W.F. Holcomb, and M.S. Bandrowski, USEPA's
Proposed Environmental Standards for the Management and Land Disposal of
LLW and NARM Waste, Presented at Waste Management '89, Tucson, AZ,
February 26 - March 2, 1989.
19.
Agreement State Amendment to Atomic Energy Act, contained in P.L. 86-
373.
20.
U.s. Nuclear Regulatory Commission, Criteria for Guidance of States and
NRC in Discontinuance of NRC Regulatory Authority anc \ssumption Thereof
by States Through Agreement, 46 FR 7540, January 23,1~8l.
21.
Brenneman, F.N., and S.N. Salomon, The Role of the
Regulation of Low-Level Radioactive Waste, NUREG-0962,
Regulatory Commission, Washington, DC, March, 1983.
in the
Nuclear
State
U.S.
22.
Burns, M.E., and W.H Briner, Setting the Stage, Chapter in Low-Level
Radioactive Waste Regulation: Science, Politics, and ;- car, M.E. Burns,
ed., Lewis Publishers, Chelsea, MI, 1988.
23.
Salomon, 'S.N., U.S. NRC Office of State Programs, Internal NRC Memo Re:
Current Status of Each State in Providing Disposal of LLW, August 16,
1988.
24.
New York State Low-Level Radioactive Waste Management Act, Laws of New
York, Chapter 673, July 26, 1986.
25.
Salomon, S.N. , U. S. NRC Office
communication, October 12,1988.
Programs,
personal
of
State
- 121 -

-------
Summary of Current IAEA Activities Comparing Land-Based
and Ocean Disposal of LL W
- 122 -

-------
The LDC Intergovernmental Panel of Experts on Radioactive Waste Disposal at Sea
(IGPRAD) is currently investigating scientific/technical, as well as social, political and
economic issues relating to ocean disposal of LL W.
The IAEA w~ asked, as the
internationally recognized cognizant authority, to develop a framework for comparison
of risks from land and ocean disposal.
Current IAEA activities/schedule and U.S.
involvement are reviewed in this appendix.
IAEA Safety Series 65 will be used as the main reference in reviewing the
comparative assessments made to date.
The following studies will also be considered,
two of which have been provided by the United States:
1. Assessment of best practicable environmental options (BPEO's) for management of
low- and intermediate-level solid/solidified radioactive wastes. DOE-UK/March
1986.
2. Long-term management of the existing radioactive wastes and residues at the
Niagara Falls storage site. DOE- USAf April 1986.
3. Disposal of decommissioned defueled naval submarine reactor plants. Department
of Navy-USA/May 1984.
4. Comparative environmental and safety assessment of four generic disposal options
for the Surrey low-level radioactive wastes. AECL-Canada/ April 1986.
5. Studie naar de rnogelijkheden voor de verwijdering von uit Nederland ofkomstig
laag - en middelactief vast afval anders dan door storten in de Atlantische
Oceaan. CHVRA-NetherIJnds/March 1983.
The second and third documents were submitted for consideration by the U.S. in
1987 at the first meeting of the LDC IGPRAD.
Two additional documents were
submitted for consideration in September 1988, at the second IGPRAD panel meeting,


one of which refers to the study currently being conducted by EPA. These are:
- 123 -

-------
1. A U.S. study of risk and ullcertainty comparisons associated with the disposal of
materials containing elevated levels of naturally occurring radionuclides.
(The
first phase of this study should be available in April 1989.)
2. A comparative assessment by Sweden of land disposal and ocean dumping of
nuclear power wastes.
The projected IAEA schedule for completion of tasks is as follows:
Comoletion Date
AugUst - September 1988
Task
Collection and analysis of documents.
Preliminary review -. IAEA Secretariat.
March 1969
Consultants' meeting in Vienna to prepare
working document (2-3 exp~i'ts).
July or September 1989
Consultant's meeting (or possibly Advisory
Group meeting) in Vienna to review and
improve working document (6-8 experts).
Autumn 1989
Technical editing by Secretariat.
December 1989
Submit for publication (IAEA Technical
Report Series or Technical Document).
September 1990
Submit report to the London Dumping
Convention.
The U.S. plans to provide expertise to the consultant/advisory group deliberations.
- 124 -

-------
APPENDIX B
HODEL REVIEWS - LAND
- 125 -

-------
1
IDENTIFICATION
1.1
Name:
PRESTO-EPA-POP
1.2 Prepared By:
Oak Ridge National Laboratory, U. S. Environmental
Protection Agency, Rogers and Associates Engineering
Co.
1.3
Prepared For:
U.S. Environmental
Radiation Programs
Protection Agency,
Office
of
1.4 Report Title:
PRESTO-EPA-POP: A Low-Level Radioactive Waste
Environmental Transport and Risk Assessment Code
(Vol. 1: Methodology Manual; Vol. 2: Users Manual)
1.5 Report Number:
EPA 520/1-87-024-1 and EPA 520/1-87-024-2
1.6 Report Date:
December 1987
1.7 Availability:
U.S. Environmental Protection Agency
1.8 Purpose and Scope:
Evaluate possible health effects from shallow land
burial disposal of LLW to assist in the development
of environmental standards
2
SUMMARY OF FINDINGS
PRESTO-EPA (Prediction of Radiat:on Effects from Shallow Trench
Operations) is a suite of computer models developed for EPA to evaluate
possible health effects from shallow land burial disposal of LLW. It's
original obj ective was the assessment of impacts from varying shallow land
burial disposal scenarios to assist in the development of environmental
standards. Since PRESTO-EPA-POP was developed first and serves as the basis
for other codes in the PRESTO family. it is the focus of this review.
PRESTO-EPA-POP assesses radionuclide transport, resultant exposures, and
health impacts of a LLW disposal site on a static local population for 1000
years after closure and on the general population residing in the downstream
regional basin for an additional 9000 years. The model simulates leaching of
nuclides from a waste form, hydrological, hydrogeological, and biological
transport, resultant human exposures, and finally assessment of potential
human health effects. Exposure scenarios treated by the model include normal
release (leaching, spills during operations), human intrusion, and site
farming/reclamation. Environmental pathways considered include ground wat~r
transport, over-land water flow, erosion, surface water dilution,
resuspension, atmospheric transport, deposition, inhalation, and ingestion of
contaminated foods and water. Individual. and population doses are
calculated, as well as doses to intruders and farmers. Cumulative health
effects (deaths from cancer) are calculated for the population over the 1000
year period using a life-table approach.
- 126 -

-------
3
ADMINISTRATIVE CRITERIA
3.1
Documentation
Separate documentation is available for each of the individual models in
the PRESTO suite of models. The two volumes issued for PRESTO-EPA-POP cover
theory and background (Volume 1) and specific implementation data along with
sample input and output and a listing of the code (Volume 2). Additional
information for all of the PRESTO models is contained in EPA' s Background
Information Document for the. Draft Environmental Impact Statement for
Proposed Rules [EPA 520/1-87-012-1].
3.2
Hardware Requirements
PRESTO-EPA-POP is written in FORTRAN IV for implementation on an IBM
3081 computer or comparable system, using 850K bytes of memory. Users of
non-IBM systems may have to modify the job control language, NAMELIST inputs,
and other program segments where character manipulations are used.
3.3
Application
PRESTO is an integrated program designed to run in batch mode. Time
required to execute depends on the scope of the problem. For example,
evaluation of local and regional basin health effects considering a total of
31. nuclides over a period of 10,000 years takes approximately 7 minutes I
whereas the same problem for only 10 nuclides requires less than 2 minutes.
3.4
Level of Expertise Required
In spite of its size and complex calculational abilities, execution of
this program appears relatively straightforward. The well documented Users
Manual which contains an overview of the code, sections on input and output,
an example problem with printout, and a listing of the code, is especially
helpful in this regard. Organization of output data is also designed to
assist inexperienced users.
4
TECHNICAL CRITERIA
4.1
Peer Review
The Radiation Advisory Committee of EPA's Science Advisory Board (SAB)
conducted a peer review of the Background Information Document to the Draft
Environmental Impact Statement for Proposed Rules (3/13/85). This document
contains a description of the theory and methodolgy used in PRESTO to predict
disposal site performance. The SAB found it to be "a reasonable presentation
of the potential sources and risks associated with the disposal of low-level
radioactive wastes." Deficiencies were cited in the explanation of how the
assessment will be used and in descriptions of data and calculational
uncertainties. The SAB felt that the 10,000 year time frame was
unrea~istically long and disagreed with some .of the methods and presentation
used ~n the dose response section. Recommendations for additional research
were presented. EPA's latest version of the Background Information Document
- 127 -

-------
(6/88) incorporates some of these comments, but a response to the peer review
has not been published.
4.2
Verification
External verification was performed for an April 1983 version of PRESTO-
EPA by Inter Systems, Inc. (151). A quality assurance audit was prep$red to
check the coding of equations, compare the logic flow with that presented in
the documentation, and verify the correctness of calculations and results.
All parts of PRESTO-EPA were checked except for the dose conversion/response
calculations contained in DARTAB and its subroutines. Discrepencies between
the code and documentation and problems with unit conversions, logic flow,
and mass balance were identified by 151. Appropriate changes were
incoporated in the code and approved by EPA so that PRESTO-EPA "now performs
the calculations as described in the documentation package including ISI' s
revis ions. "
4.3
Uncertainty
The issue of uncertainty in risk estimates is not addressed in the model
documentation, but is discussed for the PRESTO suite of models in the
Background Information Document. Uncertainties are divided into 5
components: (i) source term concentration, (ii) nuclide geosphere transport,
(iii) food chain tranport, (iv) human organ dosimetry, and (v) health effects
conversion factors. Due to limited time and budget, quantitative analysis of
uncertainty is limited to nuclide transport in the geosphere (uncertainties
in the other 4 components are discussed qualitatively). Instead of Monte
Carlo or other simulation method, uncertainty is estimated using a method
proposed by Hung which calculates joint probability density distribution for
two successive random variables. A number of simplifying assumptions are
made including: (i) a quasi-steady-state approximation that assumes cap
failure reaches maximum at the start of the analysis and remains constant,
(ii) conversion of a dynamic, numerical ground water model to a steady-state
model that can be solved analytically, and (iii) consideration of a humid
permeable site only (worst case assumption.) Values for an arbitrary
distribution of probability density for each of the parameters in the
anlytical ground water equation are estimated based on engineeering judgment.
Input parameters that are varied include: trench material, soil, and aquifer
Kds, degree of trench cap failure, trench to aquifer distance, site to river
basin distance, ground water velocity, and percolation velocity in host soil.
Some of the input parameters were not assigned probability density
distributions because the .expected standard deviations were small enough to
treat them deterministically, or the probability density distribution was
calculated from other random input parameters.
Uncertainties calculated for geosphere transport were combined with
qualitative uncertainty estimates for the other four components using the
Theorem of Variance for the joint distribution of random variables. Assuming
each component probability density distribution varies log-normally, this
method estimates total uncertainty as the sum of the standard deviations of
the components.
- 128 -

-------
4.4
Sensitivity
Discussion of sensitivity analyses is also included in the Background
Information Document. Sensitivity analyses were performed to test the
relative impact of single parameters and scenarios consisting of a set of
assumed variables. In total, 54 sensitivity test runs, analyzing 30 input
parameters were conducted for PRESTO-EPA-POP. An additional 10 test runs
were performed to evaluate the health effect conversion factor used by most
versions of PRESTO-EPA.' Because the measure of impact for PRESTO-EPA-POP is
cumulative health effects integrated over a long period of time, few
parameters were found to have a significant impact beyond short-term effects
when varied. A number of input parameters affected near-term (i.e., <1000
yrs) health effects to the local population, but these effects are but a
small fraction of the total cumulative health effects. Input parameters that
were judged to be relatively sensitive were those that affected structural
stability of the trench cap and leaching out of the trench. The isotope-
specific health effects conversion factors were found to be very sensitive to
river-flow-to-population ratio, fish consumption rate, and fish
bioaccumulation factors.
4.5
Required Input
Four sets of input and control data are required by PRESTO-EPA-POP: (i)
site- and nuclide-specific data to calculate nuclide concentrations in the
environmental transport parts of the program, (ii) subroutine DARTAB control
options for processing exposure data, dosimetric data and tabulation of
output, (iii) hydrogeologic and meteorologic d~:a for subroutine
INFILtration, and (iv) dosimetric and health effects data used by DARTAB.
The first three sets of input are supplied by the user; the fourth is
generated by the program RADRISK and a copy of this data file is needed for
execution of PRESTO-EPA-POP. Site- and nuclide-specific data consht of
about 150 variables that describe the physical and hydrogeological
characteristics of the dispoal site, data for the biological pathways, and
the radionuclide inventory, as well as define the release/exposure scenario
for each run. A partial sample of input requirements is included in Table
B-1.
4.6
Output
Program output is intended to be self-explanatory and does not assume
that the user is familiar with code structure. Definitions and descriptive
comments are included, where appropriate. Intermediate tabulations are given
in addition to final resul ts . Output is organized into 12 sub - sec tions as
summarized below:
1.
Summary of Input Data - With the exception of subroutine INFIL,
input data are printed as they are read.
2.
Input Data Organization
according to data type,
facilitate review.
Data are organized
transport sub-system
and summarized
or pathway to
3.
Radionuclide Summary Tables - Three summary tables are presented
that specify initial inventories (in the trench, surrounding soil,
- 129 -

-------
4.7
stream, and atmosphere); decay constants; solubility constants;
chemical distribution coefficients (Kds) for the surface soil,
trench, underlying soil and aquifer; and nuclide-specific food
chain parameters.
4.
INFIL Input/Output -
INFIL.
Input data and results for the subroutine
5.
Unit Response Calulations - Nuclide-specific annual transport and
soil loss calculations used for each simulation year.
6.
Annual Summary Tables - Hydrologic and tranport variables such as
status of trench cap, water depth in trench, and trench inventory.
as well as nuclide concentrations in key pathways are reported for
user-specified years.
7.
Radionuclide Concentration Tables- Average concentration, maximum
concentration, and year of maximum concentration are given for well
and stream.
8.
Radionuclide Exposure Tables Population intake of
percent attributable to drinking water, nuclide-specific
maximum dose by exposure route, and year of maximum
included.
nuclides,
exposure,
dose are
9.
DARTAB Control Information -
conversion) control data such
nuclide uptake and clearance
etc.
Included are subroutine DARTAB (dose
as critical organs for consideration,
data, organ dose weighting factors,
10.
DARTAB Dose Tables Individual and collective
radiation type, organ,and nuclide are presented.
rates
by
dose
11.
DARTAB Fatal Cancer Risk Tables - Individual and collective fatal
cancer risk and genetic risks are summarized.
12.
Residual Radioactivity Released and Health Effects - Presents the
quantity of nuclides released to the basin each millenium during
the 10,000 year simultation, aggregated total release of each
nuclide, and basin population health effects by nuclide.
Sample output data from PRESTO-EPA-POP are given in Table B-2.
Source Term
PRESTO-EPA-POP can handle up to 40 user-specified radionuclides.
Physical and chemical properties of individual waste streams are not
considered directly. Rather, these properties are treated generically for
the total waste inventory when the user chooses one of five possible leaching
options.
- 130 -

-------
4.8
Scenarios
Scenarios that are considered include: normal disposal site operations,
spillage of waste during operations (with residual contamination remaining on
soil surface at closure), resident intruder on site after ~losure, farming of
the site, and eroded trench cap with subsequent atmospheric contamination via
suspension of waste/soil mixture. Normal operations cover post-closure
impacts to local population assuming ingestion of off-site water and foods
and inhalation of downwind air. User can modify scenario by specifying
parameters such as population size, location and distance to well, degree of
trench cap failure, resuspension rate, etc. Two scenarios that are not
included in analysis are: maj or changes in meteorology or mining of trench
contents. Neither of these are routinely considered in assessment of LLW
disposal (they are considered in HLW repository performance projections)
since probability of occurrence is low especially in the relatively short
time periods of interest. The code is structured to consider one
scenario/run.
4.9
Relationship to Regulatory Standards
Model output includes an estimate of annual mean effective dose
equivalents which is compatible with the proposed standards contained in 40
CFR 193. The current standard contained in NRC's 10 CFR 61 does not use the
organ weighting factor associated with effective dose equivalent. Federal
regulations also specify allowable radionuc1ide concentrations in air and
water. Radionuclide concentration data for surface soil, surface water,
atmosphere at spill, atmosphere downwind, and well water are presented for
each radionuclide at user-specified time intervals.
5
SCIENTIFIC CRITERIA
5.1
Theory
The assessment is performed in two phases: the initial stage calculates
healtl~ effects to the local population for up to 1000 years and the second
stage calculates health effects to the regional basin population over 10,000
years. After 1000 years the local community is assumed to be incorporated in
regional basin community, simplifying calculations. The analysis begins
after closure. Specified inventories are adjusted by the program to account
for decay during operations phase. Time to container failure is user
specified. .
. Co~p1ex physica~ and chemical interactions between nuclides and geologic
medla. ln . the leac~lng and groundwater subroutines are treated using the
distrlbutlon coefflcient (Kd)' Different Kd values are applied for soil,
soil/w~ste mixture, sub-trench soil, and aquifer.
Leac~i~g options a~e calculated by submodule LEAOPT using one of five
user-speclfle~ ca1c~latlon methods: Kd with waste in total contact with
water or partlally lmmersed (wetted fraction is calculated as the ratio of
maximum water depth to trench depth); solubility factor to estimate maximum
concentrations of nuclides in water, also as either total contact or wetted
- 131 -

-------
fraction; and finally, a user-specified average annual fractional release.
Fraction release is used for solidified waste. Solubility limits are used
when Kd values are not known. Average annual fractional release of the total
inventory is specified and the program calculates adsorption effects inside
and outside of the waste form before release from trench.
Container effects are accounted for by multiplying concentration in the
trench water by a time dependent container fracture factor (CFF). Initially
CFF - 0 when conatainers are intact and then incr~ases over time to unity
when all containers have failed. Apparantly CFF is the fraction of all
containers that have completely failed, not a measure of extent of failure.
Groundwater flow below the trench is vertical to the aquifer and
horizontal through the aquifer to a well. Groundwater in the vertical reach
is assumed to be saturated or partially saturated. Degree of saturation
determines water velocity and retardation factors. Nuclide transport in the
aquifer is evaluated using Hung's groundwater transport model which uses an
approximate solution to the basic transport equation to avoid extensive
computer time needed for numerical solutions. The groundwater model
initially neglects the effect of radionuclide transport by dispersion and
subsequently compensates for this effect by applying a health effects
conversion factor. Calculated concentrations in well water are averaged over
the length of the simulation and are used by food chain and human exposure
modules of the code for drinking water and cattle feed pathways. Trench
waters that overflow can result in overland flow to nearby surface water
bodies, percolation down to the aquifer, or sorption on the soil leading to
potential resuspension and air transport or delayed downward migration.
Atmospheric transport is handled by a "simplified, compact algorithm
suitable for those sites where the population is concentrated into a single!
small community." Gaussian transport module called D~ is incorporated as
a subroutine. For complex population distributions, an externally computed
X/Q ratio (e. g., using AIRDOS-EPA) should be used to explicitly specify
population and wind rose distribution. Atmospheric source strength is the
sum of wind-driven suspension and mechanical disturbance during user
specified time interval.
Average concentrations of each nuclide over the assessment period in
environmental media, such as well water or the atmoshpere, are used to
calculate radionuclide concentrations in foodstuffs. Foodstuff
concentrations and average human ingestion and breathing rates are utilized
to calculate annual average radionuclide intake/individual in the local
population by ingestion and inhalation. These intake data are then used to
estimate dose rate and health effects.
5.2
Validation
Model performance predictions have not been compared with actual shallow
land disposal site data.
5.3
Treatment of Radioactive Decay Products
To maintain simplicity and reduce computer time and expense, PRESTO does
not calculate in-growth of daughter products. In cases where the major dose
- 132 -

-------
contribution is from external exposure to short-lived progeny in equilibrium
with parent radionuc1ides and accounting for daughters can significantly
impact results, daughter product activity can be externally estimated and
added to the initial trench inventory for appropriate nuclides.
5.4
Underlying Assumptions
As in any comparative ris1:t model, many simplifying assumptions are
required to perform the assessment. Because of the number of a~sumptions and
the magnitude of uncertainties inherent in such an analysis, EPA strongly
cautions the user against misuse or misinterpretation of results. For
example, results will not be valid if modeling assumptions are not
appropriate or significant input parameters such as inventories, site
meteorology, hydrology, geology, or demographics, are improperly ;assigned.
The underlying assumptions in each of these areas are explicitly described in
the documentation.
5.5
Pathways
Hydrologic, atmospheric,. and food chain transport pathways are
considered. Hydrologic tranport is divided into two main pathways: (i)
precipitation initiated leaching carries radionuclides through trench,
vertical soil column, and aquifer to a well or stream, and (ii) operational
spillage and/or overflow from trench contaminates surface soil, and
radionuclides are carried either to an aquifer via deep seepage where
ultimately they enter a well or stream, or to a surface water body via.
leaching. Resultant exposure' pa:~ways are either direct via drinking
contaminated water, or indirect via dthe food chain (irrigation, plant
uptake, and ingestion of contaminated plants, animals, and milk) .
Atmospheric transport is triggered by suspension of contaminated surface soil
or trench soil that has been exposed by erosion. Resultant exposure pathways
include direct exposure by immersion and inhalation of respirable
contamination, or indirect exposure via the food chain (deposited
cont~ination and ingestion of crops contaminated with deposited
radionuclides). Hydrologic and atmospheric environmental transport pathways
are illustrated in Figures B-1 and B-2. Other exposure pathways considered
include an inadvertent intruder living at the site after closure and off-site
exposures due to suspension of particles from on-site farming activities.
5.6
Dose Conversion and Dose Response
Dose and health effect calculations are made within the subroutine
DARTAB. Three calculations and tabulations for dose rate and dose are
performed: (i) dose rate to an individual at a selected location, (ii) dose
rate to a mean or average individual, and (Hi) collective population dose
rate. A weighting factor for the various organs (developed by EPA) was used
to combine dose rates in a manner similar to the approach recommended by ICRP
30.
- 13 3 -

-------
Table 8-1.
.............-.-.-.---------------------.-..--.-.--------.-------------------
Sample input data for PRESTO-EPA-POP
...---...-------.-.-.---------------------------..-.-----..----------------..
. ..all1.206., ...1
I

.
'II"~ - a RODIL FOI '1,llcll.O '"I "1,la'llI OF la.loac"vl wa.111
riD" lMALLOW I .CII au'laL 'II
I'" aliA ..... '0' IUM "., IACIL.lf LIIAIAt'll
.a. LUlII~L 'I.OlftAIIOI ."
fill IUI AL .111 I' LOCAIII af I11I alia
~"I ~1"ULafIOl VILL IUI rOI ..,. 'IAII ... WILL IICLUDI 33 IUCLIIII
latll 10 ° 10. IIU"'I, 2 YILL f U fl
II f~al '.20 ~f III ca' U L AS5U"ID '0 'AIL
11111 YllL COIIIINUE UIIIL '.3' MAl ,alLII ,. 'IAI 2..
fa, .A' ALS~ rAJL II JUlrAtt 1I011ON
fA~9l=I'&IIV':,ll:~0IA,KI't~~ 01 vlLL II 111 10 II' IIIICII '0 IOUlrll .I.IAIICI
GII~.AL 'O'ULAI'DI II.OIUII WILL I' UII. II CALCULlII llAL1I .rrlCtI
l:t81 81 JIII:~1~aDcA¥rll'DI11~I=fLI03tf~ IIDeo¥l~ rlOM VILL
1.000 or .1 .. NO VA'II fDI IIU.ANS WILL I OD'I II FIOft VILL
0.000 OF 11111all.. VA'II VILL II 801111 '10" 1111A.
0.000 O. DII"KI"O YA'~I IDI AII.AL. WILL II GO"" '10. 5111A"
0.000 OF DIIIIK III YA'II 01 MUIIAI' vlLL I IDIIIN 'OU. I' IAII
." 'IIICII 1.'OI..A'IOI "~A
.
'"I IIIICII IIA~ .. AlIA or '.50001+00 IOVAII .1111. All a DIPIH If '.2'001+'1 .11111
'lelCH 'UluSII' II '.25
ANNUAL IN'ILIIIIIO. rOI 'MI IAtllSNID II ,.4300 "11111
"I A,ulr~1 IlrOI"'IIDN 'AA
'H' GkOUIID YATII HAS-' V~LDCIIY Of 27.100 .IIIIS 'II fIAt
TtlNCH '0 aOUlfEI DI$IINCI IS 7.~ "ETEIS
IIEII'H 'D WELL 1151ANCE IS .610.00 .EI.IS
wELL 10 S'IIA" DISIANCE IS 3~~'100.IIEIS
IHi AUUI'~I IHILKlliSS IS 30.~0 11£ £15
TH£ AOUlfEI IIS'.ISIUM ANGLI IS 0.3000 I.DIAI&
'Oka~I" Of 'N' A~Ulfll I'GIOI IS 0.39000
tDla~.rY I'NlaIH 'HI tlENCH IS 0.3~000
'II"'A"LI" 1'"IAI11 I"' 'IIM'N IS 2.200 .11.,5/'1.1
"~A atIlOS'HIIIC II.DI.AIIOI '"
SOUICI HEIIH' IS I.' IIIII.S
VILOCII' Of IlavlIAIIOI 'ALL IS 0.01 .IIII./SICOND
YIN~ VELOCIII IS 2.01 HEIIIS/5ECOII
DltO~I'IOII VELDt 111 I~ 0.10 H~III&I'ECDND
"AUGE IISIIN[e rlOIl SOUICE IS 17700.00 111111.
LIII H'16HI IS ~O.OO H'IIIS
HOSKU 10UGHIIISS FACI'.I 1$ '.01
I'" or ~IAaILIII faa"ULAIIDM II I
SUIII.ln UA~S II 4
rl.LIIUN or IIHI WIIID ILaws lowal. POPUL.lro. IS 0.414000
IESU~rEIISIOII FAerOI '..AHETUI 0.10001-0:' -0.1500'." 0.10001'10
F.O.. Y"I . 10 YEA. 0 IHI I'SUS'INSION 1.11 IVI 10 HICHANICAL DISlua.AIICES WILL II
IHI' WILL OCCUI DUliNG 2.40 0. 'ACM '1'1
A"~ IIUtLIIIE IlIiOIHAtlON A"
NUCLaIlI
H-3
t-14
CI-51
"N-:l4
F£-55
CD-50
111-:19
CO-60
111-63
51-90
""-~4
tc-"
tU-IOe.
58-I ~:!
1-129
CS-U4
CS-l;':!
u:m
£U-I:i4
U-234
U-~3S
"-:!37
U-2311
tU-231
rU-:!;'9
tU-240
rU-:!4I
AK-HI
rU-24~
AH-24:1
CH-243
CH-244
AKI 1M nENCH
CI
6.10171-01
9."00£-01
5.~2:1UoOO
9.42121-02
<1.111331-01
1.3;'01£-01
9.9000£-01
3.7:1"7£-01
'I.!lOOOI-OI
7.'J438t:-01
'1.9000£-01
't.!l000~-01
1.03411:-01
2.:::l4U-01
'1.'10001-01
1.74811-01
'1.'.1000£- O.
8.0204.:-01
8."'684£-02
.. ""8E-01
9.'10001-01
9.9000£-01
'1.9000£-01
'.900,)L-Ol
9.9000£-01
9.9000£-01
9.9000£-01
1o.2'J:I,)[-01
9.9000£-01
').9000.-01
').'1000£-01
8.11 9JE-01
'."?Ol-Ol
" ILUIiE
tI
'.1017£-1:1
9.'1000£-03
5.52:lJE-0;!
9.UI:ZE-04
2.183:1£-03
1.3:tItU-03
'.9000£-03
3.7227£-03
9.90001-03
'1.94:111£-03
9.'J000£-OJ
9.9000£-0:1
1.0341£-03
2.234:1"-03
'.'10001'-03
1.7488£-03
'.9000£-03
V.0104£-03
I. ')61141-04
'.7630£-03
'.'J000"-03
'.9000£-0:)
'J.'00~1-03
'J.9000l'-03
'.'JOOO£-OJ
'.9000£-03
'.9000£-03
6.2'I1'J£-03
'.90001-03
9.90001'-03
9.9000£-03
1.11931-03
'.9I?Oa,;-o;,
SltUII AliI
CI
0.0000£.00
0.00001000
0.0000£000
0.00001+00
O.OOOOtooo
0.0000.:'00
0.0000£000
0.0000£<00
0.0000£0011
0.0000"000
0.0000£000
0.0000£<00
0.0000£000
0.00001000
0.0000£+90
0.0000£+00
O.OOOOE+OO
0.0000£+00
0.00001000
O.OOOOUOO
0.0000£+00
0.0000"00
0.0000£000
0.0000[+00
0.00001'00
0.0000£+00
0.0000£000
0.0000£000
0.0000£000
0.0000"000
0.0000£000
O.OOOOE+OO
O.OOOO~+"
4.00001-04
IltOkftA'ION OM IMBIVIIIUAL IUCLIDES
All CUNC~. IIIC.f tOIS' "LuaILII' COISI DICAY CO.IICIIOI raCIOI
CI/.AAa IIY 6/11L
0.00001+00 5.'500£-02 0.0000£+00 '.1'33E-01
0.0000£000 1.1100£-04 0.00001+00 1.00001.00
0.0000"000 9.1400£+00 0.00001.00 5.571IE+00
0.0000£+00 ..0900£-01 0.00001+00 '.~164E-02
O.OOOOE+OO 2.5700£-01 0.0000.+00 2.2054"-01
0.0000"<00 3.55001'00 1.0000£+02 1.:1415£-01
0.0000£'00 1.6'00£-0' 0.00001.00 1.0000£000
0.00001.00 1.3~00£-01 1.0000£002 3.'603£-01
0.0000..00 7.5300£-03 0.0000£+00 1.0000£000
0.00001.00 2.4;'001-02 0.0000£000 1.0:141£-01
0.00001000 a.4700£-OS 0.0000£'00 1.0000£+00
O.OOOOE'OO 3.25001-06 O.OOOOEoOO 1.00001+00
0.00001+00 '.3'J00£-01 0.00001+00 1.0445E-01
0.0000"'00 2.5000£.01 1.0000Eoo. 2.~5'91-01
0.0000£.00 4.08001-08 .00001+00 1.0000£+00
0.0000£+00 3.JbOO£-01 3.0000E+0~ 1.""51-01
0.0000£.00 2.3000E-07 0.0000£+00 1.00001+00
0.0000£'00 i.atooc-02 3.0000£+02 8.1014£-01
0.0000£000 .9000"-01 0.00001+00 '.0~'01-02
O.OOOOE+OO 4.3300E-02 0.0000£+00 '.8372E-01
0.0000£000 2.030'~-06 0.0000£000 1.0000£000
0.0000£'00 ~.d5001-10 0.0000£000 1.0000£000
0.0000£000 3.3000t-07 0.00001+00 1.0000£.00
O.OOOOttOO 1.5500"-10 0.0000'000 1.0000£000
O.OOOOt.OO 7.'1000£-03 0.0000£+00 1.0000£000
0.0000£000 2.8?00E-05 0.00001+00 1.0000£.00
0.0000£'00 1.0600£-04 0.0000£.00 1.0000£'00
O.OOOOE
-------
-------.------.-.-.-..-..-.----------------------

------~-i---(---~-;--~~~~i~-~nput data for PRESTO-EPA-POP

Table -. con. ------------------------------------------
- - - - - -.-.-. - - - -.- - --- - - - -. - - - - - - ---
  DISIII'UIIOM COErflCIEMTS IIL/O
IIUCL lilt SU.,An tltMeM .,UlltliL AOU If"
.-1 1.00N' 1.0n-oJ 1:ln:U 1.000-oJ
c- 4 .on-02 .00£-0 .001-0
et-n ~.OOlt02 ~.00It03 .,.oott03 3.00lt02
IIH-!!4 I . )OU02 .:'0...02 1.:5ott02 I. )01003
non '.OOt.o:! ..00tt03 ..00ttO:l '.00100:1
CO-i- :5.,OltO\ :5.:5ottO\ :5.:50It" '.:50ttOI
III- . 1.)OItU 1.)01102 1.)OU03 1.:'0It03
co-roo 5.:lOtolll '.'011" :s.50EtOI '.501tOi
111-63 1.)OltO:l 1.50\:to:;l 1.51f' o:z 1.)O.:t02
U-':!\' 1.)0~00: 1.50£tO;: 1.:SOloO;: 2.001001
""-~4 '.:"II:t02 :I.)nto:l :I. sGt.. 02  :'.!!Ott02
te-'t.) 5.00e-01 '.OOt-OI '.00&-01 ).oot-Ol
'¥-1~6 2.:01'02 2.20\:t02 2.:101:'32 2'1,,:t01
1 - _:S 4.:101-01 4.:5011" 4.5Oto I 4. 0100
-1.'1 :I.oot-tOO J.ontOO 3.00£100 ). n- 0
CI-IJ4 I.Otll-OJ 1.00l'OJ 1.00ltO;l 5.00110
es-u!! 1.00Ktn 1.001003 1.00E'03 ).00It02
AA' SUlrACt IlIiOI"AtIOIl "~A
'AIAIIITIIS FOI UNlv~aIAL LOSI EOUAIIOM
aAIHrALL 250.00
EIO~I'ILltT 0.23
STE~'NE~S-~LU'E 0.27
COVEl 0.30
EIO$10H CONtauL 0.30
~ELIVEIT IAflU 1.00
SOIL tUIOSltT II 0.39000
SOIL IULK OENSI Y IS 1.60000 GICC
IUNOFt .IAtIION IS 0.2'000
SIIEA" fLOW IAtl 1$ 3.)7001005 tUIIC "It II' 'II T1AI
CIOS~ SLO'I: EIIINI 0. S'ILLAGI IS 1003.00 "EIEIS
ACflVI ~OIL OEtlN IS 0.10 "tUIS
AVI:IA6E DOVN SLOtl DISIANCI: 10 5111." IS 100.00 "ElliS
,.. All-fOODCMAIN INfOI"AtION ."
AGIILULTUIAL taODUCTIVllY rOI GlASS 0.67 kG/"AA2
Ar,IICULIUIAL 'IO~UCIIVlfY rOI VeGEIAtlON 0.'5 KG/R"2
SUI.ACE DENSIIY rOI SOIL 240.00 KG/""2
VE.fNEI DECAY CONSTANI 0.00 IIHOUIS
tEIIUU 'ASTUIE GlASS EX'OSUIE alOUING SIASOM 720.00 NOUIS
'IIIO~ CIO,/vEGEIAtION EX'OSUIE GIO~ING 51.50. 1440.00 HOUI.
~E.IKP 'ITW1~N HAIVESI 'ASIOIE 614155 AND INGESIION ay ANI"AL 0.00 HOUIS
,~tIO= tEI~fl= ~1~t~rlrEI:f;N3E~~¥flfl3NA:1 1=~~:~IOM1~tO"::I=~Y~I.) 24.00 HOUIS
'EIIOD IETWEEN NA.vESt 'IODUCE AND INGEstION IT MANIM.I.I.I 1440.00 HOUIS
'ElIUI1 tElVEEII HAlVESt LEAn VEG AIID IIIGESTION IT MANI6.'.I.1 336.00 NOUIS
'EI 00 IE WEEN HAlVES 'IODUCE AND INGE~TIOM IT "ANIG.I.I.) 33'.00 NOUIS
F.ACIION OF YEAI ANIMALS GlAZE ON 'AStUIE 1.00
FIACTION OF DAILY FleD tHAt 13 rIE~N 6'41'1 0.13
AHOUHT Of .E~D CON~UHED IY CAtlLE 50.00 KG
AHOUNt or FEED CONSUMED IY 60A1S '.00 KG
TIANS~Olr liME fEED-"ILL-klCE'IOI fOI M.I.I. 48.00 HOUIS
IIANStOlt 11"1 fEED-"ILL-IECE'IOI rOI G.'.E. ".00 HOUIS
IIHt rlOH SLAUGNtEI Of MEAt TO CONSUM'IIO. 480.00 HOUIS
A~SOLurE HUHIDlfT 01 tNE Ar"U~'HE.I ,.~o 0/""3
FIAttlONAL IOUILI.kIU" 1411/0 101 C-14 1.00
"~A WATEI-FOOOCNAIN INfOIMAtlON A"
FIACII011 Of Y1AI CIO'S 41.1 IIIIGAIID 0.40
1IIIGAriON IATt 0.01~ L/IM"2-")
AHOUNT or WArEI CONSUMED IT COVS 60.00 LII
'MOUNT Of WAt~1 tONSUIIED IY GOAII ..00 LID
AHOUNt Of WAtEI CON~U"ED ay IEEf CAltLE 50.00 LID
AA' NUMAN IN~EStION AND INHALAtiON 14111 INfOI"ArION ."
ANNUAL INtAKE Of L1AfY VEG 2.00 KILOGIA"S 'EI YEAI
ANNUAL INIAKI OF ~IODUCE 11.00 KILOGIAMS ~II YIAI
ANNUAL INIAKE OF COY'S "ILK 11.00 Llt~IS PII YEAI
ANNUAL INfAKE of 60At.S "ILK 0.00 LJfll~ lEI TEAl
ANNUAL INtAKE Ot M1AI 9.00 KILOGIAMS tEl YE..
ANHUAL INTAKI Of OIINKING VAIEI 370.00 Lltll~ 'EI YIAI
ANNUAL INHALAtiON lATE Of All 1000.00 CU'lt "ElliS PI:I YEAI
A 'O'ULATION OF 25. WILL IE CONSIDEIED
- 135 -

-------
Table B- 2.
........-----........-----------------------------------.------------.-.------
Sample output data for PRESTO-EPA-POP
....-----------------------.-.--------------.-.------------------------------
NEAL1N EllltlS IESuLtlNa 'Iaft I.IIDUAL
1.~IPAl'IVII' "LiA5iO IN 1000 'EA~'
NUCLIDE I.SIIUAL CQNVE'~IUN NEALtN
ACUllln 'ACU'I! "nctS
2.UOU-.0) 0.00001:-'0 '.OOO"tO'
'.6049£-01 '.00001-00 '.00001-00
~.7~4~£.0~ O.OOOOttOO 0.0000'-00
3::Zi~f:lt 1:8888f:11 ':'8':1:88
1.~0~'£-06 0.00001-00 0.0000£-00
4.1$'01-03 O.OOOOltOO 0.00001-00
,.'022£-04 0.00001400 0.00001-00
.3JIOI-OJ 0.00001-00 1.00001-00
.~4~L-03 0.00001-00 0.0000'-00
4.14311-03 O.OOOO~-OO 0.00001-00
1.2446a:-01 1.00001-00 1.00001tOO
!:!UU::t :1811f:= :8:::1:=
4.00101:-02 0.00001:-00 O.OOOOltOO
1.$40'~-0' 0.00001-00 0.00001-08
4.112~I-oa 0.00001:-00 O.OOOO£tO
~.'07~1-04 O.ooooa:tOO 0.0000'-00
l.t240t-I' 1'loOOltOI O.OOOOltOO
7. J401- S . OOOltO 0 'OOOltOO
4. 41.1:- a . OOOl:tO 0: 000'-00
4.143'I-OJ 0.00001-00 0.00001-00
4.199'1:-03 O.OOOOt-OO 0.0000'-00
t.143'I-OJ O.OOOOE.OO '.00001-00
1.'¥'7E-04 0.'0001:-00 '.0000.-00
).0~7)I-Oa 0.0000£-00 O.OOOO'tOO
2.93M61:-0a O.OOOOl:tOO 1.0000£tOO
t.40IO~-05 0.0000£-00 0.00001-00
.2~74£-04 0.0000£-00 '.0000a:-00
3.0~'6t:-oa 0.0000£-00 O.OOOO£tOO
2.~~66I-04 O.OUOOL-OO 0.0000£-00
2.0718~-04 0.0000£-00 0.0000£-00
1.00 4L-04 0.0000£-00 O.OOOOL-OO
N-3
C-"
Ct-~I
""-54
n-~~
CO-~It
NI-$'
co-"
111- i3
St-'o
Nt-'4
li:lIt
I-U'
ES-134
5- :.~
(S-Ia.,
il-144
u- 54
-::;,4
U-235
N"-23)
U-230
'u-2al
'U-2a4f
'u- 240
'U-241
AII-241
'U-242
AII-~4a
,"-~43
Cit".:..
tAJLE t~t Ar.COUNIIMG ftOD~L -- YEA. 1000
Ct'LUII" I: lllI'IVII'IiAL (;ANCn USK (UtI,TUSSI
C')LUIIN %: tOJ'ULAtlOli l'OSt CPER:;(lN-tI;IIIIEAtI
r.nLUIIN 3: COLL~ttIVl CANCI. .ISK (~LAIHS/Y£AII
COLUN" 4: GENEtiC .ISK (£iFCcrS/Y£Atl
CULUII" !I: .-/ot1IVITY tU"tE(t OUT TH£ MELL 1St 1000 'EAIS CCiI
COLUN" 6: I-ACTIVlry 'UIIP£ft our rNE Stl£A" 1St 1000 '~ARS Ctl)
CPLUIIN i: I-ACTIvity IEL£AS£D 10 At"OS"N£I£ 1ST 1000 YLAtS (CII
COLUNN 8: I-ACtlvllY I£LEA!>ED DUUN!>'."" 151 1000 rLAt~ (crl
CO LUll" ,: t-ACTIVIT' '£LEAStD DowN~rlEAN LASt .000 Y£AIS (CII
COLUIIH 10: R-/oCTIVI11 IH IIENCN Afl'l 10000 YEAIS A5SUNING '£RF£CT CONIAINIIINI cel)
HC-] !.a663E-IZ 6.7~061:-0' 1..'611-12 7.17331-13 1.66861-06 0.0000£-00 1.34'2£-07 %.4601£-03 3.41161-27 0.0000£'00
-I' 6.8240£-08 0.5'45£-0$ 2.4111£-08 4.4872£-0. 1.186B£-03 1.0000£-00 2.3101£-07 '.604'~-01 '.2"6£-04 2.~$~~£-01
Ct ~I 2.63.,~t-24 ~.3~141:-21 '.aI76£-2~ 3.'~32E-25 O.OOOO£tOO 1.0000£tOO 6.5'0'£-01 5.7'43£-08 0.0000£'00 0.0000£'00
N".54 1.66:8£-182.0'42£-155.874'1-1' 2.5655£-1' 0.0000£'00 O.OOOO£tOO 5.3645£-06 '.4'bJ£.OL O.OOO~£'OO O.OO~~£'OO
l~-~, 2.i43:£-:0 3.4~49£-17 3.6.23£-21 i.$168£-21 0.0000£-00 1.0000~tOO 2.59~4t:-o$ 2.4~04£-06 0.0000£-00 0.0000£-00
.1:5. ~:i~n~:U ~:~nU:l: 2:inn:U s:3~ZU:n ::ggUI:gg :::ggg~:= 1:2nn:g~ l:U~n'.g~ t:~~t ~~:t ~::~sn~:Y
CO-60 3.182~£-16 4.007'E-13 1.1244£-16 4.1'141-17 O.OOOOt:¥OO O.oooot.OO 4_6'1'£-05 $.60221-04 O.OOO~£.OO 0.0600£'00
5"11-.603 1.252$£-19 1.5775£-1' 4.4r~55£-20 1.26'4£-200.0000£-000.0000£'00 1.~3B4£-04 3.3200£'03 ~.~~OJC-I3 1.'646£-33
- 2.8~8.,t-11 3.~004E-15 1.0 00£-18 .4442£-20 0.00001tOO I.OOOOL-OO .~784£-04 1.1456£-03 2.5130£-27 O.OOOO~OO
.t-94 5.3753£-14 6.76"£-11 1.8'92£-14 1.29'0£-1$ 0.0000£-00 O.OO~OL'OO 1.7103£-04 4.1431£-03 1.408'£-01 6.~"4£-Ol
TC-" 1.11'4ft-08 1.022)~-04 2.U690£-01 1.35'2£-0' 1.13~5£-03 0.0000£-00 1.0136£-0$ 5.2446£-~1 4.7142£-01 '.5134£-or
tU-l06 2.70'7£-1' 3.4127£-16 9.573'1-20 3.17611-21 0.0000£-00 O.OOOOltOO 6.8611£-06 9.0312£-06 o.OOOOC-OO O.OOOOC
-------
Figure B-l.
SPILLAGE
OVERFLOW
BASIN
STREAM
BASIN
POPULATION
OCEAN
SINK
SOIL
SURFACE
LEACHING
TRENCH
LEACHN;
VERTICAL
SOL
COLUMN
SEEPAGE
AQUFER
GAOUNDWA TER TRAN.SPOAT
SURFACE
WATER
BODY
WELL
IRRIGATION

~

PLANT UPTAKE
CROPS
AND
ANIMALS
DRINKING INGESTION DRINKING

~HU~
RAE-102201
Hydrologic environmental pathways in PRESTO-EPA-POP
- 137 -

-------
SURFACE
CONT AMINA TION
ERODED
TRENCH
SUSPENSION
   AIR  
 INHALATION  DEPOSITION 
 NMERSION   
    I CROPS
HUMANS IRRADIA TION FROM GROUND AND
(Local Population) -    GROUND
  INGESTION 
     -
RAE 102084
Figure B-2. Atmospheric environmental pathways in PRESTO-EPA-POP
- 138 -

-------
1.1 Name:
1.2 Prepared By:
1.3 Prepared For:
1.4 Report Title:
1.5 Report Number:
1.6 Report Date:
1.7 Availability:
1.8 Purpose and Scope:
1
IDENTIFICATION
BARRIER
Rogers and Associates Engineering Corp.
Electric Power Research Insti~ute
Performance Assessment for Low-Level Waste Disposal
Facilities
EPRI NP-5745M; EPRI NP-5745SP
,.,;>ril 1988
EPRI Research Reports Center, Box 50490, Palo Alto,
CA, 94303
Estimate the long- term performance and degradation
of concrete barriers used in various engineered
storage disposal designs compared to a shallow land
burial base-case.
2
SUMMARY OF FINDINGS
BARRIER was developed for the !lectric Power Research Institute (EPRI)
to assist the nuclear power industry meet federal and state disposal site
performance assessment requirements and expedite licensing of new disposal
facilities. It is an integrated model that estimates: groundwater flow
through a disposal facility; radionuclide release; long-term performance and
degradation of concrete barriers used in various engineered storage disposal
designs (e.g., below-ground vault, above-ground vault, modular concrete
canis~er disposal and earth mounded concrete bunker); transport through an
aquifer to the accessible environment; and doses to the critical population
group (CPG). Projected performance of engineered disposal options were
compared to a shallow land burial base-case.
Inclusion of a module that simulates the performance of concrete
structures is a unique feature of this code. Prediction of engineered
barrier performance is a significant addition to the capabilities of
performance assessment modeling, especially in light of the increasing
emphasis on alternative disposal technologies. At the same time, however,
lack of empirical data on concrete behavior over long periods of time
introduces new uncertainties in the calculation that need to be addressed.
3.1
Documentation
3
ADMINISTRATIVE CRITERIA
Documentation for BARRIER is contained in the interim report published
by EPRI. Two versions have been issued: NP-5745M presents a qualitative
summary of the assessment model and its findings; NP-5745SP is intended for
- 139 -

-------
limited distribution and provides additional detailed data and information on
concrete degradation/failure mechanisms. Neither document contains a code
listing or the information typically found in a user's guide. If BARRIER
becomes available for general use, additional documentation would be needed
to implement.
3.2
Hardware Requirements
The code occupies about 4000 lines of FORTRAN-code and was written for
implementation on a Micro Vax II computer.
3.3
Application
BARRIER is an integrated code that contains subroutines to model
groundwater unsaturated flow, radionuclide leaching and transport, well-water
concentration, food chain and uptake, human dose, and engineered barrier
performance. An inital data set contains required data and parameters to run
the code to completion. Total time for execution is dependent on complexity
of the unsaturated flow calculations, ranging from about 5 minutes for a
si~plified problem to several hours for a complex one.
3.4
Level of Expertise Required
Not discussed in the documentation.
4
TECHNICAL CRITERIA
4.1
Peer Review
BARRIER has not undergone a formal peer review.
4.2
Verification
EG & G Idaho has a licensing agreement with EPRI to implement BARRIER
and has provided comments that have resulted in code modifications.
4.3
Uncertainty
uncertainty analysis was performed to quantify the impact of parameter
variability on model output. Selected parameters relating to the performance
of concrete structures were examined using a statistical variation
methodology originally designed for radionuclide transport from disposal
facilities. RANDIS (Random Distribution) code, which uses a Monte Carlo
method of statistical sampling was modified for this purpose. Probability
distributions for proj ected structural failure of engineered barriers were
generated by specifying a range for each of the relevant variables, along
with the type of distribution (uniform, normal, log uniform, or log normal).
Concrete structural characteristics such as compressive strength, percentage
of calcium aluminate, and bulk modulus, as well as groundwater concentrations
of potentially destructive species such as chloride, magnesium, sulfate, and
oxygen were examined. Although not included in this study, application of
uncertainty methodology could be extended to examine parameter uncertainty in
- 140 -

-------
the other modules of the code (e.g., groundwater flow, contaminant leaching
and transport, food chain, dose assessment).
4.4
Sensitivity
Sensitivity analyses were performed on selected aspects of concrete
structural properties (compressive strength, water-to-cement ratio),
diffusion coefficients of chemical species (CaOH, 02' S04) , and environmental
characteristics (freeze-thaw cycles, maximum thermal gradient). Relative
impact on the time to onset of structural cracking and time to total cracking
failure were measured by varying each parameter within a prescribed range
(e.g., j: 10%) while keeping other factors constant. Sensitivity of these
parameters was examined for each of the alternative disposal options at both
humid and arid sites.
4.5
Required Input
Since the available documentation is not a user's guide, specific input
~equirements such as particular data files and user specified parameters are
not discussed. A general discussion of the types of input used by BARRIER is
included, however. Four disposal options are examined: i) shallow land
burial (SLB), ii) modular concrete canister disposal (KCC), iii) below-ground
vault disposal (BGV) , and iv) above-ground vault (AGV). Each set of options
is examined at both a humid site (data from Charleston, SC) and an arid site
(data from El Paso, TX).
Radioactive waste inventory used in the model includes 27 radionuclides
(listed in Table B-3). Wastes in Class A are handled separately from Classes
Band C wastes, but further characterization of waste composition (such as
physical or chemical composition) is not included. For example, the code
uses a single value of Kd for each isotope, ignoring possible differences
based on waste composition or solidification medium. All solidified waste is
assumed to be in a concrete matrix. Although separate input data for each
disposal site are presented for some parameters, only one set of soil Kds is
reported. It is not clear whether these correspond to the South Carolina
site (silty loam soil) or the Texas site (sandy clay soil). Since soil
distribution coefficients are isotope and soil specific, separate data are
required for each site. Data on aquifer velocities, depths, and
dispersivities were taken from humid permeable site characterization in the
EPA LLW Draft Background Information Document. Eight freeze-thaw cycles are
assumed for humid site, "none for arid. Freeze-thaw cycling is a significant
parameter in the concrete degradation simulation.
4.6
Output
Calculated output from BARRIER includes peak dose, year peak dose
occurs, dominant nuclide contributing to peak dose, and lifetime dose to
members of CPG.
4.7
Source Term
Waste
Conceptual
Quantities
inventories are the same as for RAE study for DOE on LLW
Designs. 27 isotopes are included, 11 of which are TRU.
of each isotope are specified in four inventory categories: Class
- 141 -

-------
A packaged in steel drums, Class A packaged in steel liners, Class B & C
packaged in steel drums, and Class B & C packaged in steel liners. No
distinction is made to account for waste type or physicochemical properties
other than Class A vs. Classes B & C (NRC categories).
4.8
. Scenarios
A simple exposure scenario is used that assumes a well, located at the
site boundary, is used for direct consumption by humans and animals, and for
irrigation of crops. Doses to the CPG result from ingestion of contaminated
water, crops, meat, and milk. Intruder and exposed waste scenarios are not
considered. Although engineered barriers would probably reduce the
likelihood of occurrence for both of these exposure scenarios in the near-
term, after eventual failure of concrete structures the probability of
occurrence would be comparable to a shallow land burial site. The example
problem does not address worker exposures but a separate program called
Worker Occupational Radiation Dose Assessment Code for the Disposal of Low-
Level Waste (WORDRAE) is included in the appendix.
4.9
Relationship to Regulatory Standards
Output of projec.ted effective dose equivalents (in terms of
millirem/year) is consistent with EPA's proposed exposure standards for the
public contained in 40 CFR 193. The current standard contained in NRC's 10
CFR 61 does not use the organ weighting factor associated with effective dose
equivalent. Federal regulations also specify allowable radionuclide
concentrations in air and water. These are calculated and used in the dose
calculations, but are not reported in BARRIER output.
5
SCIENTIFIC CRITERIA
5.1
Theory
The methodology used in this integrated performance assessment code is
organized into four categories:
.
Unsaturated groundwater flow analysis
.
Concrete degradation and failure analysis
.
Contaminant leaching and transport
.
Dose projections
UNSAT-H is used to simulate one-dimensional water flow in unsaturated
soils and sediments. This mechanistic model based on Darcy's Law, can
simulate movement of water in liquid and vapor phases in response to
precipitation and irrigation, plant water extraction, and deep drainage.
Flow through cracked concrete is described by an equation that is similar to
that characterizing bulk flow, with fracture spacing hydraulically equivalent
to granular porous media.
- 142 -

-------
Concrete degradation and cracking over time are simulated by subroutines
that estimate: (i) thermal and static loading moments, (ii) Ca(OH)2 leaching,
(iii) S04 attack, (iv) freeze-thaw degradation, (v) corrosion of reinforcing
steel, and (vi) cracking moment and ultimate strength.
Contaminant leaching is via one of four leaching options: (i) constant
leach rate, (ii) nuclide-specific leach rate, (iii) diffusion leaching
(migration through a solid waste form), and (iv) dissolution or Kd leaching
(movement of nuclides in unsolidified waste or deteriorated waste forms where
porosity is high). Data for the first two options are user specified; the
latter are dynamic parameters calculated by the program. As solidified waste
form or concrete structures degrade, migration increases due to enhanced
water movement and leaching by dissolution eventually dominates. Similar to
other contaminant groundwater pathway codes (e.g., PATHRAE), effective
groundwater migration speed is described as the quotient of the groundwater
flow velocity and contaminent retardation factor As the facility degrades
and cracking develops, hydraulic resistance decreases and flow is enhanced.
For transport through fractured media, it is assumed that contaminants travel
to external surface in one year, then proceed through bul~ media at normal
rates.
Dose conversion is discussed in Section 5.6.
5.2
Validation
Model performance predictions
disposal site performance.
have
not
been
compared
with
actual
5.3
Treatment or Radioactive Decay Products
Treatment of radioactive decay products isnot considered in BARRIER.
5.4
Underlying Assumptions
Assumptions used in the analysis including disposal facility design and
performance characteristics; climate, geology, and hydrology of the generic
sites; and waste and package parameters are clearly stated.
5.5
Pathways
Transport/exposure pathways are all linked by groundwater and" surface
water flow. Unsaturated groundwater flow is m,; jeled (using UNSAT, described
above) as the driving force for disposal site degradation and radionuclide
release. Vertical flow of leachate is expressed as the quotient of
groundwater velocity and a contaminant retardation factor that takes into
consideration soil density, porosity and distribution coefficient (Kd)'
Radionuclide transport through engineered barriers is either via flow in
fractured porous media (upon barrier failure) or advection through intact
concrete matrix. Horizontal transport through the saturated zone to the
accessible environment is handled using the same algorithm used in the PRESTO
f:unily of. codes, ac~ounti~g for aquifer velocity, porosity, and depth,
dlsposal slte area, dlsperslon angle of contaminated plume, and distance to
well. Exposure is by ingestion of contaminated well water, crops, milk, and
meat. Other transport/exposure pathways (e. g. , direct exposure to
- 143 -

-------
inadvertent intruder, resuspension via erosion or flooding with concomitant
air and/or surface water transport). are not addressed.
5.6 Dose Conversion and Dose Response
Dose conversion methodology is the same as that used in PRESTO family of
performance assessment codes (e.g., PATHRAE and PRESTO-EPA). PRESTO
calculates effective whole body dose equivalent as per ICRP 30. Calculations
assume individual: (i) is situated 100 m from site boundary and use"s
contaminated well water for drinking, irrigation of food crops, and as a
water source for animals, and (ii) consumes contaminated quantities of
vegetables, milk, and meat. Dose to the CPG is calculated as annual dose
rate and cumulative lifetime dose. Dose response is not calculated.
- 144 -

-------
----------.--------.------------------------------------.-...---------------.
Table B.3.
Radioactive waste inventories used in BARRIER
.-.....---.-.--..-.-.-....-.-..-------------....-.-.-.-----.-..----..--------
 Class A Inventory Class B & C Inventory
 till  (Cn 
Nuclide Steel Drumsa Steel Liners Steel Drumsa Steel Liners
H-3 5.1E +02 2. 8E +03 3.8E+05 1.4E+03
C-14 2.8E+Ol 1. 5E +02 3. 3E +02 9.1E+Ol
Na-22 2. 3E +00 3.0E+03 2. OE -03 1.3E+03
Fe-55 1.5E+Ol 1.5E+04 1.5E-Ol 2.9E+05
Co-58 4.0E+00 8.6E+03 O.OE+OO 1.2E+05
Co-60 5.8£+01 1.1E+04 9.4E-06 2.2E+05
Ni-59 6.2E-03 1.2E+Ol O.OE+OO 1.8E+02
N1-63 1.0E+00 1.8E+03 2.4E-03 2.6E+04
Sr-90 1.9E+Ol 6.8E+Ol 3.2E-Ol 1.6E+03
Nb-94 3.6E-05 2.2E-Ol O.OE+OO 1.6E+00
Tc-99 4.5E-05 2.1E-Ol 8.1E-07 8.8E-Ol
1-129 2.1E-08 5.8E-02 6.7E-I0 2.3E-Ol
Cs-134 5.5E-13 5.4E+03 O.OE+OO 2. 2E +04
Cs-135 2.5E-05 2.1E-Ol 8.4E-07 8.8E-Ol
Cs-137 2.8£+01 5.7E+03 1.5E-02 2. 3E +04
Ba-137m 2.8E+Ol 5. 7E +03 1.5E-02 2.3E+04
U-235 4.6E-Ol 5.2E-02 5.1E-02 3.6E-Ol
U-238 2.1E+00 2.2E-Ol 2.9E-Ol 1.1E-Ol
Np-237 3.1E-14 1.2E-06 9.9E-16 2. 7E -06
Pu-238 1.1E-05 6.5E+00 3.7E-07 1.4E+Ol
Pu-239 3.2E-06 4.0E+00 1.0E-07 8.3E+00
Pu-241 4.1E-04 1.8E+02 1.3E-05 3.8E+02
Pu-242 5.5E-09 8.8E-03 1.8E-I0 1.8E-02
Am-241 9.7E-03 9.1E+00 2.0E-08 1.9E+Ol
Am-243 7.2E-08 2.6E-Ol 2. 3E -09 5. OE -01
Cm-243 1.7E-08 3.9E-03 5.3E-I0 7.9E-03
Cm-244 9.5E-06 3.8E+00 3.1E-07 7 . 5E +00
dThese wastes were considered to be unpackaged for the modular concrete canister
disposal facility, reflecting compaction of the drums within the canisters.
- 145 -

-------
1
IDENTIFICATION
1.1
N sme:
Part 61 Impacts Analysis Methodology
1.2
Prepared By:
Envirosphere
Commission
Co. ,
and
u.s.
Nuclear
Regulatory
1.3
Prepared For:
U.S. Nuclear Regulatory Commission
1.4 Report Title:
Update of Part 61 Impacts Analysis Methodology
1.5
Report Number:
NUREG/CR-4370, Vol. 1 and 2
1.6 Report Date:
January 1986
1.7 Availability:
Documentation: U.S. NRC; U.S.
Office; National Technical
Software: Radiation Shielding
Oak Ridge, TN
Government Printing
Information Center
Information Center,
1.8
Purpose and Scope:
Integrated performance assessment code for compara-
tive analysis of various land-based disposal options
for low-level radioactive waste.
Note: The full title of this performance assessment code is
Update of Part 61 Impacts Analysis Methodology. For the purposes
of this discussion it will be referred to simply as IMPACTS.
2
SUMMARY OF FINDINGS
IMPACTS was initially developed for the U. S. NRC, to assist in the
preparation of shallow land disposal regulations for low:-1eve1 radioactive
waste (10 CFR 61). It is an integrated performance assessment model for
comparing potential health impacts from various land disposal options on a
generic basis. The user is cautioned in using the methodology for a site-
specific application, where site-specific models, inventories, disposal
options, and environmental parameters would be required to accurately
simulate conditions. Further caution is advised in interpreting the absolute
magnitude of results. Rather, the model was intended as a: relative
comparison to estimate potential benefits and costs from a number of
potential disposal options.
User provides information on combination of waste streams to ~e
considered and regions where they are generated and then selects specific
waste processing scenarios, the environmental setting of disposal site, and
the particular combination of disposal technologies to be used. Output gives
calculated effective dose equivalents (mrems/yr) for 9 organs plus effective
whole body equivalent for each exposure scena~io and waste classification.
For chronic exposure scenarios, data are given in varying time increments
from 20 to 20.000 years. In addition to IMPACTS, user may select the
following subroutines: INVERSE calculates acceptable nuclide total activity
and/or concentration limits for disposal; ECONOMY calculates transportation
and routine operational radiological impacts as well as disposal cost
- 146 -

-------
estimates; INTRUDE analyzes radiological impacts to an inadvertent intruder
as a function of time; and VOLUMES calculates and updates region and waste
stream dependent annual volume projections. A separate code (CLASIFY) is
used to classify the waste streams and organize input data for use by
IMPACTS. Code is written for implementation on an IBM PC.
3
ADMINISTRATIVE CRITERIA
3.1
Document:at:ion
Comprehensive documentation for IMPACTS is available in Volumes 1 and 2
of NUREG/CR-4370. Volume I presents an overview of the method, describes the
theory and algorithms for implementation, and provides background in the
areas of waste inventories, processing and disposal options, and pathway dose
conversion factors. Volume 2 includes a listing of the codes and data files
used by the model and presents several example problems. In ad~ition, the
software package includes a README file which contains information on which
input files are needed to run each subroutine, along with expected run times.
3.2
Hardware Requirement:s
IBM PC (or compatible) with 640 KB, hard disk drive,
coprocessor.
and 8087 math
3.3
Applicat:ion
Six separate subroutines, listed below along with expected run times,
are executed independently. CLASIFY and INVERSE are operated in batch mode;
others are interactive.
CLASIFY:
IMPACTS:
INVERSE:
ECONOMY:
INTRUDE:
VOLUMES:
Classifies waste streams in four
Various impact measures
Activity or concentration limits
Costs of disposal
Impacts of an intruder
Waste stream annual volume
classes
(-10 min)
(-23 min)
(-20 min)
(-12 min)
(-26 min)
not given)
(time
3.4
Level of Expert:ise Required
Although written for implementation on a personal computer, IMPACTS is
not "user friendly". Several characteristics of the program structure
combine to complicate its use. First, almost all of the parameters that
define the problem to be examined (e. g., disposal region and technology,
waste characterization, type of waste processing, waste form behavior) are
assigned index values. The user must therefore frequently refer back to the
series of tables that define index acronyms and provide numerical index
values that correspond with appropriate responses in order to supply needed
information. These index values are contained in input control data files.
More significantly, the programs do not contain interactive subroutines to
create control data files, so user must either: (i) edit the example control
data files supplied with the software to alter the problem, or, (ii) create
programs to write new control data files. The former can be accomplished
- 147 -

-------
without a great deal of computing experience,
consuming, especially if multiple runs are needed.
but the process
is
time
4
TECHNICAL CRITERIA
4.1
Peer Review
The original Part 61 Impacts Analysis Methodology (NRC 1981) written for
the Environmental Impact Statement on 10 CFR 61 was included in the Low-Level
Waste Disposal Performance Assessment Model Review (EG&G 1985), which
compared exposure scenarios and pathways, transport pathways and parameters,
and dosimetry methods, among several land-based disposal performance models.
Model capabilities and parameters were identified, but this report does not
constitute a comprehensive peer review. In addition, many significant
changes in the updated methodology reduce the usefulness of EG&G's review of
IMPACTS. No other independent peer reviews were found.
4.2
Verification
The code was tested and de-bugged at the Radiation Shielding Information
Center, but verification of mathematical computations was apparently not
performed.
4.3
Uncertainty
No uncertainty analysis is reported in the documentation.
4.4
Sensitivity
No sensitivity analysis is reported in the documentation.
4.5
Required Input
IMPACTS requires up to nine input files (summarized below). These
include data base files and output files from several codes that must be run
before IMPACTS (CLASIFY and VOLUMES) to consolidate some of the input data.
CLASIFY classifies waste into regulatory categories based on radionuc1ide
characteristics and activities and requires the following data files:
LIMITS.DAT which contains information representing specific
nuclide/solubility option; WASCAR.DAT which contains user-specified records
representing specific waste streams; CLACON. DAT which specifies manner in
which waste streams are considered, and whether disposal technology can
stabilize waste streams. CLASIFY combines the multitude of alternatives
presented by these data files into one file (CLAOUT.DAT) for use by IMPACTS.
VOLUMES projects future waste volumes based on specific generating
characteristics (from VRATES.DAT input file) and information on nuclear power
reactors (from REACTR.DAT input file) and supplies the output file VOLUME.DAT
for IMPACTS. Because of this modular structure, data subject to change such
as reactor and waste generation data can be easily updated.
IMPCON.DAT
Input command file which specifies problem to be considered
including facility, schedule and disposal configuration
parameters.
- 148 -

-------
FUNDCF.DAT
Fundamental dose conversion factors and other radiological
specific information such as isotope solubility and half.
life, pathway uptake coefficients, etc.
ENVIRO.DAT
Environmental parameters of 4 reference sites and region-
dependent pathway uptake parameters.
INPUTS.DAT
Specifies waste streams to be considered.
ClAOUT . DAT
Output from ClASIFY (see discussion above).
DISTEC.DAT
Disposal technology parameters.
LIMITS.DAT
Nuclide concentration and activity limits.
VOLUME.DAT
Output from VOLUMES (see discussion above).
METALS.DAT
Activated metals information, where applicable.
Necessary data files are provided with the code. Disposal site data are
generic - "typical" values are provided for four regional disposal sites
(Northeast, Southeast, Midwest, and Southwest). A summary of input data used
in IMPACTS is given in Table B-4.
4.6
Output
Output of IMPACTS is stored in IMPOUT.DAT. which provides a summary of
specified parameters used in calculations and calculated effective dose
equivalents (mrems/yr) for 9 organs plus whole body equivalent for each
exposure scenario (intruder, exposed waste, leachate accumulation,
operational accidents, and groundwater) and waste classification. For
chronic exposure scenarios, data are given in varying time increments from 20
to 20,000 years. Information on individual radionuclides (e. g. ,
concentrations, relative contribution to organ and total dose) is not
provided. Human health effects are not calculated. Sample output from
IMPACTS illustrating the groundwater scenario (boundary well) for Classes A,
B, and C waste is included in Table B-5.
4.7
Source Term
IMPACTS accepts a wide range of radionuclides (100 radionuclide and
solubility combinations) and waste types (148 waste streams). Radionuclides
are listed in Table B-6. Waste streams include those from: nuclear power
plants, fuel fabrication, reprocessing and decommissioning activities, other
fuel cycle facilities, institutional and industrial generators, and other
non- fuel cycle generators. Source term characterization in the current
version of IMPACTS represents an improvement over the original program which
consid~red 23 radionuclides and a total of 37 waste types.
Consideration of such a large number of isotopes and waste types can be
extremely unwieldy for a program to process. Thus, the range of potential
options was bounded by consolidation into six waste spectra that denote the
collective volume and other properties of all waste streams after processing
by a set of selected waste treatment options. Each spectrum corresponds to a
- 149 -

-------
general level of waste performance (e.g., structural stability. resistance to
dispersion) that result from application of processing options.
4.8
Scenarios
Exposure scenarios are considered for operations and post-closure
phases. During operations, exposures resulting from routine operations and
two accident scenarios (drop and breakage of waste package; release by fire)
are included. Package drop scenario assumes airborne material may be
transported offsite and calculates offsite exposures. Onsite impacts are not
calculated but algorithm is provided so user can modify program. Similar
approach is used for fire scenario, assuming waste is either combustible or
is mixed with combustible waste material. Two approaches are used to examine
routine exposures: one based primarily on number of shipments received, and a
more involved method that accounts for different types of waste packaging and
emplacement modes.
Post-disposal impacts are considered in four categories:
(1)
Intruder scenarios
disposal area.
inadvertent
human
intrusion
into
waste
(2)
Groundwater scenarios nuclide
exposures via groundwater pathway
migration
resulting
in
human
(3)
Leachate accumulation - leachate percolation and accumulation with
disposal cells
(4)
Exposed waste scenarios
dispersed via air or water.
waste
is
uncovered
at
surface
and
Intruder scenarios (and pre-closure operations) are acute dose exposure
events and are termed "concentration scenarios" because they "depend upon the
concentrations of the radionuclides within the waste streams considered".
These are contrasted with the remaining three chronic exposure categories
which are termed "total activity scenarios" since they "depend upon total
radionuclide inventory a~d waste volume disposed at the disposal site."
Intruder scenarios include: (i) drilling (prior to construction); (ii)
construction (basement is dug for a dwelling); (iii) discovery (same as
construction except that intruder discovers waste and abandons site); (iv)
agricultural (lives in house constructed on site and consumes food grown in
contaminated soil).
Groundwater scenarios trace nuclide migration to four biota access
locations downstream: (i) well at boundary of disposal area, (ii) well
outside disposal buffer zone, (iii) well located between disposal facility
and surface hydrologic boundary, and (iv) stream at surface hydrologic
boundary. Annual dose rate (mrem/yr) during the 50th year of exposure is
calculated as the product of radionuclide concentration, interaction factors
and pathway dose conversion factors.
If the disposal site contains a liner or soil of low permeability
beneath the waste, groundwater migration calculations are complicated due to
- 150 -

-------
potential accumulation of leachate. This "bathtub" effect which increases
leach rate has been observed at some existing sites. Three leachate
accumulation scenarios are addressed: (i) accumulated leachate is
continuously pumped and treated prior to release, (ii) accumulated leachate
overflows to nearby stream and enters accessible environment, and (i11)
accumulated leachate is collected (batch mode) and treated by evaporation.
However, the user must determine if the leachate ac.cumulation scenario is
appropriate by estimating whethe~ leachate will exit the disposal cell at a
rate less than the water infiltrating in. If this is not 1;he case, then
leachate flows through the trench and the groundwater migration scenario
should be selected instead.
Exposure of earthen cover by inadvertent intruder or wind/surface water
erosion lead to four exposed waste scenarios: (i) intruder-air, (ii)
intruder-water, (iii) erosion-air, and (iv) erosion-water. Erosion
scenarios assume a conservative time delay of 2000 years prior to exposure.
Wind transport of eroded material use separate airborne mobilization rates
for each of the four generic sites. A single, conservative surface water
mobilization rate (greater than annual rate for Appalachian region) is used.
In addition, a conservative dil~tion factor is used.
4.9
Relationship to Regulatory Standards
Output of calculated whole-body effective dose equivalent in terms of
millirem/year is consistent with proposed fec~ral radiation standards
contained in 40 CFR 193. The current standard contained in NRC's 10 CFR 61.
does not use the organ weighting - factor associated with effective dose
equivalent. Federal regulations also specify allowable radionuclide
concentrations in air and water. These are calculated and used in the dose
calculations, but are not reported in output from IMPACTS.
5
SCIENTIFIC CRITERIA
5.1
Theory
Given a number of waste streams with specific radionuclide
concentrations and specified disposal and environmental conditions, IMPACTS
integrates release/transport/exposure pathways to calculate radiological
impacts, i. e., annual exposures. This is accomplished using the following
general equation:
Hij -
nCijn x Iijn PDCFn
where Hij is the dose rate to the individual (mrem/yr) resulting from waste
stream (i) in waste class (j), where the impacts are summed over all
individual radionuclides (n) in the waste, and where:
Cijn
- the concentration of radionuclide (n) in the (i)th waste stream
in the (j)th waste class considered;
- 15 1 -

-------
Iijn
- an interaction factor relating the concentration of the (n)th
radionuc1ide in the (i)th waste stream in the (j)th waste class
to the concentration of the radionuc1ide at the biota access
location; and
PDCFn - the path!ay dose conversion factor for that radionuc1ide (mrem/yr
per Ci/m
The interaction factor, Ii1n' represents the fraction of the original
nuclide source term that is tra~sported through all environmental transport
media. It, ~n turn, is the product of factors that account for radionuc1ide
decay, disposal site design characteristics, physical and chemical
characteristics of the waste, and environmental parameters. Each of these
transport factors is determined through a series of scenario-specific
calculations. For example, the interaction factor for groundwater scenarios
considers the effects of: (i) rainwater percolation, (ii) radionuc1ide
release to 1eachate~ and (iii) migration reduction (including effects of site
geometry, dilution, nuclide-soil interaction, groundwater travel time etc.).
The one-dimensional groundwater transport model used in IMPACTS assumes
flow through unsaturated and saturated zones, each of which is stationary,
homogeneous and isotropic, and fluid moving through these zones is
incompressible and of constant viscosity. Migration through the unsaturated
zone is treated separately from the saturated zone (these factors were
combined in the original version of IMPACTS).
For unsaturated flow, the migration reduction factor accounts for
radionuc1ide decay, distance between the waste and the saturated zone, flow
speed, and retardation coefficient (dispersion is neglected to simplify
calculations). Retardation coefficients reflect the ion exchange capability
of soils and represent the ratio of the radionuc1ide v~locity in the soil to
the groundwater velocity. Five nuclide-specific retardation coefficient
options are available that span the range of values for diverse types of
geochemical environments, from sandy soils with low cation exchange capacity
to high-clay soils with high cation exchange capacity. Default values for
retardation coefficients are assigned for the regional generic sites, but
these can be modified if specific conditions warrant (e.g., presence of
chelating agents that reduce retardation effects.) Unsaturated zone
thickness and groundwater travel speeds for each site are also given.
For saturated flow, the migration reduction factor accounts for the time
lapse from initiation of the scenario to exposure, groundwater travel time,
retardation coefficient of the aquifer, source duration time (treated as a
fixed "slug" of activity that decreases linearly until depletion), and waste
stream characteristics. To simplify the calculation, the disposal site is
divided into 10 equal sectors and it is assumed that each sector releases
1/10 of the total activity as a point source. Impacts from radionuclide
migration from each sector are then summed. Reference groundwater parameters
such as travel speeds, dispersivity, and distance to surface water for each
of the four generic sites are provided.
Pathway dose conversion factors (PDCFn) are the product of a pathway
usage factor (specific to the combination of nuclide, pathway, and scenario)
- 152 -

-------
and the fundamental dose conversion factor (specific to the combination of
nuclide, pathway, and organ that receives the dose).
5.2
Validation
Model performance predictions
disposal site performance.
have
not
been
compared
with
actual
5.3
Treatment or Radioactive Decay Products
Production of radionuclide chain daughter products is considered in the
program. Concentrations of daughter products from four natural decay chains
(U-238, U-235, Th-232, and Np-237) and artificial nuclides that merge with
natural chains after several decay steps (e.g., Am-24l, Pu-239) are
calculated. These effects are automatically applied to calculations
involving exposures to individual inadvertent intruders, off-site exposures
to populations due to intrusion, and hypothetical impacts from erosion.
~ractional ingrowth of daughter products are multiplied by respective pathway
dose conversion factors and added to the pathway dose conversion factor for
the parent nuclide. Since it is a gas, effects of radon-222 ingrowth are
considered separately for inadvertent intruder scenarios.
5.4
Underlying Assumptions
As in any performance assessment model that attempts to simulate actual
conditions and events, IMPACTS contains a number of significant simplifying
assumptions. The documentation appears to do a good job alerting the reader
to these underlying assumptions throughout the text. A separate section
devoted to model limitations is included. The major limiting factor inherent
in the analysis to which the user is duly cautioned, is its generic approach.
Simplifying assumptions necessary for a comparative analysis of alternatives
are not appropriate for analyses of a specific waste type, site or disposal
technology. For example, despite the wide range of waste types and disposal
options considered in the program, the combination of specific
characteristics of a given waste stream, treatment/solidification technology,
disposal site and environmental parameters, etc. is not necessarily
encompassed. In the context of a comparative assessment, however, such
limitations are not a significant liability.
5.5
Pathways
A number of exposure/transport pathway combinations
based on the particular scenario simulated. Transport
below, grouped by exposure modes:
are used by IMPACTS,
pathways are listed
Ground and surface water modes: bioaccumulation,
zone groundwater transport (including hydrospheric
and sorption), and ingestion of water.
saturated and unsaturated
dispersion due to mixing,
Air exposure modes: atmospheric
resuspension, and inhalation.
dispersion,
atmospheric
deposition,
- 15 3 -

-------
Bio1o~ica1 eXDosure modes: atmospheric deposition, irrigation, plant uptake,
weathering by wind and/or water erosion, ingestion of contaminated
vegetation, ingestion of contaminated animals.
5.6 Dose Conversion and Dose Response
Five sets of dose conversion factors are incorporated that cover
internal exposure by ingestion and inhalation (50-yr committed dose, mrem/pCi
ingested or inhaled), and external exposure from a volume source, area
source, or immersion in uniformly contaminated air. Each set is
radionuc1ide-specific (data for 53 radionuclides are included) and are given
in terms of nine organs (whole body, red bone marrow, bone surface, liver,
thyroid, kidney, lung, stomach wall, lower large intestine) plus the
effective whole body equivalent, calculated using ICRP-26 and ICRP-30
methodology. Internal exposure dose conversion factors are based on data
from ICRP-30; external dose conversion factors are based on data from Kocher
(NUREG/CR-1918, August 1981). Dose response is not calculated.
- 154 -

-------
Table B-4.
--------_.----------~--------------------_..--------------------------------.
Sample input data for IMPACTS
--------------------------------------------------------------------------.-.
IMPACTS RUN OF REGION 4 . SOUMCES
IA . 4 OYFl. 0 II5UF. 0 NBAN- 0 18[$- 0
ICLS- 1 lOBS- !» IINS- 100
IBEG- 1991 IENO- 2020 IlFE- 30
COMBINATION INOICES AAE:
4
1
4
4
4
MINIMUM DEPTHS ARE:
2.D
2.0
5.0 10.0 10.0
2.0
DISPOSAL CONFIGURATION
NO ID IU IT
1 4 1 1
2 3 2 2
3 3 2 2
4 3 2 2
5 3 2 2
6 10 6 6
IC IE IB IX IS EFF SEF OPT DTIC VOlE AREA
1 1 1 3 1 5.2!»E+OO 6.9OE-Ol 2.00£.00 7.00E+OO 3.53E+03 6.00E+02
2 1 2 2 1 1.14E+Ol 8.80£-01 2.00£.00 1.30E+Ol 6.55£+04 5.4UE+03
2 1 2 2 1 1.14E+Ol 8.80E-ol 2.00£.00 1.30E+Ol 6.5!»E+04 !».40E+03
2 1 2 2 1 1.14E+Ol 8.80£-01 2.ooE.oo 1.30E+Ol 6.55E+04 5.40E+03
2 1 2 2 1 1.14E+Ol 8.8OE-ol 2.00E+OO 1.30E+Ol 6.55E+04 !».40E+03
1 2 3 1 1 5.70E+OO 4.40£-01 2.ooE+OO 5.70E+OO '5.03E+03 8.6OE+02
INPUTS STREAMS: NAME. REGN. BKlG. FRAC - OlD NSTR - 21S
P-IIRESIN 4 0 1.00 P-COHCLIQ 4 0 1.00 P-FSLUDGE 4 0 1.00 P-FCARTRG
B-IXRESIN 4 0 1.00 B-COMeLIQ 4 0 1.00 B-FSlUOGE 4 0 1.00 P-COTRASH
P-NCTRASH 4 0 1.00 B-COTRASH 4 0 1.00 B-NClRASH 4 0 1.00 l-NFRCOMP
L-OECONRS 4 0 1.00 F-PROCESS 4 0 1.00 F-COTaASH 4 0 1.00 F-NCTRASH
U-PROCESS 4 0 1.00 L-PUOECON 4 0 1.00 l-BURHUPS 4 0 1.00 I- )TRASH
I+COTRASH 4 0 1.00 l-ABSLIQD 4 0 1.00 I+ABSLIQD 4 0 1.00 1-~lQSCYL
I+LIQSCYl 4 0 1.00 I-BIOWAST 4 0 1.00 I+BIOWAST 4 0 1.00 N-SSTRASH
N+SSTRASH 4 0 1".00 N-SSWASTE 4 0 1.00 N-lOTRASH 4 0 1.00 N+LOTRASH
N-lOWASTE 4 0 1.00 N-ISOPROO 4 0 1.00 N-ISOTRSH 4 0 1.00 N-SORMFGI
N-SORHfG2 4 0 1.00 N-SORMFG3 4 0 1.00 N-SORMFG4 4 0 1.00 N-NECOlRA
N-NEABl1Q 4 0 1.00 N-NESOLlQ 4 0 1.00 N-NEYIAlS 4 0 1.00 N-NENCGlS
N-NEWOTAl 4 0 1.00 N-NETRGAS 4 0 1.00 N-NETR III 4 0 1.00 N-"ECARLI
N-MWTRASH 4 0 1.00 N-MWABLIO 4 0 1.00 N-MWSOLlQ 4 0 1.00 N-MWASTE
N-TRIPlAT 4 0 1.00 N-TRJTGAS 4 0 1.00 N-TRISCNT 4 0 1.00 N-TRllJOD
M-lRITRSH 4 0 1.00 N-TRIFOIL 4 0 1.00 N-HIGHACT 40 1.00 N-TRITSOR
M-WBSOR 4 0 1.00 N-COBlSOR 4 0 1.00 N-HICICSOR 4 0 1.00 N-STROSOR
N-CESJSOR 4 0 1.00 N-PlU8SOR 4 0 1.00 N-PlU9SOR 4 0 i .00 N-AMERSOR
N-PUBESOR 4 0 1.00 N-AMBESOR 4 0 1.00 N-RANEEOS 4 0 1.00 N-aACEllS
N-RAPlAOU 4 0 1.00 N-RANPAPP 4 0 1.00 N-RABESOR 4 0 1.00 N-RAMISCl
N-CARBSOR . 1 0 1.00 N-COBlSOR 1 0 1.00 N-NICICSOR 1 0 1.00 N-STROSIJI
N-CES I SOR 1 0 1.00 N-PlU8SOR 1 0 1.00 N-PLU9SOR 1 0 1.00 N-AMERSat
N-PUBESOR 1 0 1.00 N-AMBESIJI 1 0 1.00 N-RANEEDS 1 0 1.00 N-RACEllS
N-RAPLAQU 1 0 1.00 N-RANPAPP 1 0 1.00 N-RABESOR 1 0 1.00 N-RAMJSCl
N-CARBSOR 2 0 I.QO N-COBlSIJI 2 0 1.00 N-NICICSOR 2 0 1.00 N-STROSIJI
N-CESJSOR. 2 0 1.00 N-PlUBSOR 2 0 1.00 N-PlU9SOR 2 0 1.00 'N-AMERSOR
N-PUB£SOR 2 0 1.00 N-AMSESOR 2 0 1.00 N-RANEEDS 2 0 1.00 N-AACEllS
N-RAPlAQU 2 0 1.00 N-RANPAPP 2 0 1.00 N-AABESOR 2 0 1.00 N-RAMISCL
N-CARBSOR 3 0 1.00 N-COBLSOR 3 0 1.00 N-NICKSOR 3 0 1.00 N-STROSat
N-CESISOR 3 0 1.00 N-PLU8SOR 3 0 1.00 N-PlU9SOR 3 0 1.00 N-AMERSOR
N-PUBESOR 3 0 1.00 N-AMBESOR 3 0 1.00 It-RANEEDS 3 0 1.00 N-RACELlS
N-RAPlAOU 3 0 1.00 N-RANPAPP 3 0 1.00 N-RABESOR 3 0 1.00 N-RAMISCl
N-RARESIN 4 0 1.00 M-NAVYWET 4 0 1.00 M-NAYYORY 4 0 1.00 P-OECORES
P-DEACJNT 4 0 1.00 P-DEACYES 4 0 1.00 P-OEACTCO 4 0 1.00 P-OECO"'E
P-OECONtO 4 0 1.00 P-OETRASH . 0 1.00 P-OERESIN 4 0 1.00 P-DEFIlCR
P-DEEVAPB 4 0 1.00 B-OECORES 4 0 1.00 B-DEACINT 4 0 1.00 B-DEACYES
B-DEACTCO 4 0 1.00 B-O£CO"'E 4 0 1.00 B-oECOfCO 4 0 1.00 B-DETRASH
B-OERESIN 4 0 1.00 B-DEEVAPB 4 0 1.00 N-TRJTS~ 1 0 1.00 N-TRlTSOR
N-TRJTSOR 3 0 1.00 0 0 .00 0 0 .00
- 155 -
4 0 1.00
4 0 1.00
4 0 1.00
4 0 1.00
4 0 1.00
4 0 1.00
4 0 1.00
4 0 1.00
4 0 1.00
4 0 1.00
4 0 1.00
4 0 1 .00
4 0 1.00
4 0 1.00
4 0 1.00
4 0 1 .00
4 0 1.00
4 0 1.00
4 0 1.00
1 0 1.00
1 0 1.00
1 0 1.00
1 0 1.00
2 0 1.00
2 0 1.00
2 0 1.00
2 0 1.00
3 0 1.00
3 0 1.00
3 0 1.00
3 0 1.00
4 0 1 .00
4 0 1.00
4 0 1.00
4 0 1.00
4 0 1.00
2 0 1 .00
o 0 .00

-------
Table B-S.
.----------------------------------------------------------------------------
Key to symbols and acronyms used in IMPACTS.
------------------------------------~----------------------------------------
Symbol
AREA
BKLG
DPT
DTK
EFF
FRAC
IBEG
IBUF
IC
ICLS
ID
IE
lEND
IINS
ILFE
lOBS
IR
IS
IT
IU
IX
NBES
NBRN
NSTR
OVFL
REGN
SEF
VOLE
Property
Disposal cell area
Years of backlog
Disposal cell depth
Disposal cell thickness
Volumetric disposal efficiency
Fraction of waste volume shipped for
First year of operation
Disposal facility buffer zone index
Cover index
Closure period
Disposal technology
Emplacement index
Last year of operation
Active institutional control period
Operations period
Observation period
Region index
Chemical segregation index
Topmost waste
Utilization index
Compaction index
Scenario index
Barnwell classification index
Number of waste streams
Overflow scenario index
Region
Surface disposal efficiency
Disposal cell volume
disposal
------------------------------------.----------------------------------------
- 156 -

-------
---------------------------------------------------------------------------.-
Table B-6. Sample output from IMPACTS groundwater scenario (boundary well)
by waste type.

-----------------------------------------------------------------------------
CLASS. A
TIME LUNGS S. WALL LLl WAll T. BOOY KJDMYS LIV£R R£D MAR BON£ THYROID leRP
20 Ytt 7.2SE-I0 9.40£-10 1.25£-09 1.23£-10 7.45£-10 7.21£-10 7.19£-10 5.7lE-l0 7.21£-IU 7.82(-IU
40 YR 2.77£-07 1.96£-07 5.80£-07 8.53£-07 2.85E-07 3.02£-07 2.95£-07 3.22E-07 1.91E-03 5.74E-0~
60 VA 3.11E-07 2.20£-07 6.5lE-07 9.58£-07 3.20£-07 3.39£-07 3.3l£-07 3.61E-07 2.14(-03 6."£-05
80 YR 1.22£-05 1.69£-05 2.12£-052.80£-05 1.51£-05 1.14£-05 4.67E-05 9.67£-05 4.26£-03 1.'9[-04
100 YR 1.36£-05 1.89£-05 2.35£-05 3.12£-05 1.6BE-05 1.94£-05 5.22£-05 1.08£-04 4.3U£-o3 1.52£-0'
120 YR 1.3SE-05 1.91£-05 2.42£-05 3.21E-05 1.11E-05 1.97E-05 5.24E-05 1.0BE-04 6.41£-03 2.16£-0'
160 VA 2.67£-05 3.72E-05 4.65£-05 6.15£-05 3.31£-05 3.83E-05 1.03£-04 2.13£-04 8.55E-03 3.03[-0.
200 YR 2.71£-05 3.76E-05 4.72£-05 6.26£-05 3.35£-05 3.88£-05 1.04£-04 2.14£-04 1.07£-02 3.67E-o'
400 YR 6.52£-05 9.07E-05 1.13£-04 1.50£-04 8.08£-05 9.34£-05 2.5lE-04 5.20£-04 2.14E-02 7.5!)E-o'
600 YR 8.84£-05 1.24£-04 1.53£-04 2.0JE-04 1.10£-04 1.27£-04 3.43£-04 7.12£-04 2.1SE-02 7.99£-04
800 VA 1.22E-04 1.71£-04 2.10£-042.78£-04 1.51£-04 1.75£-04 4.76£-04 9.89£-04 2.16£-02 8.60£-0.
11( YR 1.19£-04 1.6t1E-04 2.06£-04 2.72£-04 1.48E-04 1.72£-04 4.66£-04 9.68£-04 2.16£-02 8.55£-04
SIC YR 7.47£-05 1.04£-04 1.29£-04 1.71£-04 9.25£-05 1.07£-04 2.88(-04 5.98£-04 2.15£-02 7.14£-04
101C YR 4.22£-05 5.79£-05 7.35£-05 9.80£-05 5.20£-05 6.00£-05 1.59£-04 3.2&-04 2.15£-02 7.15£-04
201t Ytt 1.47£-05 1.87£-05 2.62£-05 3.59£-05 1.17£-05 2.02£-05 4.97£-05 1.00£-04 2.14E-02 6.65E-04
CLASS. 8
TIM£ LUNGS S. WALL LLI WALL T. 800Y KIDN£YS LIV£R RED MAR 80"£ THYROID ICRP
20 YR 1.59£-11 2.05£-11 2.71£-11 1.58£-11 1.62£-11 1.57£-11 1.57£-11 1.25£-11 1.57£-11 1.70£-11
40 YR 1.54£-09 1.13£-09 3.30£-09 4.68£-09 1.59£-oy 1.69E-09 1.64£-09 1.79£-09 1.04£-05 3.13£-01
60 YR 1.72£-09 1.26£-09 3.70£-09 5.25£-09 1.78£-09 1.89£-09 1.84£-09 2.00£-09 1.17£-05 3.52£-01
80 YR 6.23£-096.51£-09 1.22£-08 1.68£-08 7.04£-09 7.83£-09 1.49£-082.74£-08 2.32E-05 7.02E n
100 YR 6.60£-09 7.01£-09 1.28£-08 1. 76£-08 7.49£-09 8.35£-09 1.62£-08 3.02£-oS 2.34£-05 7 .O'i 1
120 YR 8.28£-09 8.23£-09 1.64£-08 2.28£-08 9.22E-09 1.02£-08 1.8O£-OS 3.22£-08 3.4S£-o5 1.0:it~6
160 YR 1.31£-08 1.39£-08 2.54£-oS 3.50£-08 1.48£-08 1.65£-08 3.21£-08 5.91£-08 4.65£-05 1.41£-06
200 YR 1.48E-08 1..51£-08 2.91£-08 4.02£-08 1.66£-oS 1.84E-oS 3.40E-08 6.19£-08 5.S1£-05 1.76£-06
400 YR 3.23£-08 3.41£-08 6.28£-08 8.66£-08 3.66£-08 4.0SE-08 1.87£-08 1.46£-07 1.16£-04 3.53£-06
600 YR 3.80£-084.22£-08 7.26£-08 9.95£-OS 4.37£-OS 4.9OE-oS 1.01£-01 1.93£-07 1.17£-04 3.55E-06
800 YR 4.61£-08 5.38£-08 8.66£-08 l.lSE-07 5.39E-01t 6.08£-o1t 1.34£-07 2.61£-07 1.11£-04 3.56£-U6
lit YR 4.55E-08 5.29£-08 8.55£-oS 1.17£-07 5.31£-08 5.99£-08 1.31£-07 2.56E-07 1.17£-04 3.56£-06
5K VA 3.46£-08 3.13£-08 6.63£-08 9.19E-08 3.95£-08 4.41£-08 8.19£-08 1.65£-07 1.17£-04 3.54£-06
101C YR 2.66E-08 2.58£-08 5.21£-08 7.39E-08 2.95£-08 3.25£-08 5.62£-08 9.91£-08 1.17£-04 3.53£-06
201t VA 1.98E-08 r.59E-08 3.95E-08 5.S6£-o8 2.1u£-o8 "2.25£-08 2.94E-01t 4.33£-oS 1.17£-04 3.51£-06

CLASS. C
TIME LUNGS S. WAll LLI WALL T. BOOY KIDNEYS LIVER RED MAR BONE THYROID ICRP
20 YR 6.34E-14 8.19£-14 1.08E-13 6.29E~14 6.49E..14 6.28£-14 6.26E-14 4.97£-14 6.28£-14 6.81E-14
40 YR 2.43E-I0 1.72E-I0 5.09£-10 7.50£-10 2.50£-10 2.65E-I0 2.59£-10 2.82E-10 1.68£-06 5.05E-08
60 YR 2.88E-10 2.04£-10 6.05£-10 8.89£-10 2.97£-10 3.15£-10 3.01£-10 3.35£-10 1.99£-06 5.9'lE-01!
80 YR 1.49E-09 1.74£-09 2.76£-09 3.8OE-09 1.74£-09 1.96£-09 4.36£-09 8.54E-09 3.69E-06 1.13E-01
100 YR 1.64£-09 1.92£-09 3.03£-09 4.17£-09 1.92£-09 2.17E-09 4.86E-09 9.52E-09 3.94£-06 1.20£-07
120 YR 1.87£-09 2.09E-09 3.52£-09 4.90£-09 2.16£-09 2.42£-09 5.10£-09 9.79£-09 5.57E-06 1.70£-07
160 VA 3.21£-09 3.78£-09 5.93E-09 8.16£-09 3.16£-09 4.25£-09 9.55E-09 1.8S£-08 7.62£-06 2.33£-01
200 YR 3.48£-09 3.98£-09 6.50E-09 9.00E-09 4.05E-09 4.55E-09 9.86£-09 1.91£-08 9.47£-06 2.89E-07
400 YM 7.82E-09 9.22E-09 1.45£-oS 1.99£-08 9.18£-09 1.04E-08 2.33£-OB 4.58£-OS 1.86£-05 5.68E-07
600 YR 9.79£-09 1.20E-08 1.7S£-08 2.44£-OS 1.16£-08 1.32£-083.11£-08 6.20E-OS 1.87£-05 5.75E-07
800 YR 1.26E-08 1.60£-08 2.26E-08 3.07£-OS 1.5lE-08 1.73£-08 4.22E-08 8.51E-0~ 1.8SE-05 5.82E-07
lit YR 1.24E-08 1.57E-08 2.23E-08 3.03E-08 1.49E-OS 1.70£-08 4.15E-08 8.36E-08 1.88E-05 S.8lE-07
SK YR 8.69E-09 1.04£-08 1.59E-08 2.19E-08 1.03£-08 1.16£-08 2.67£-OS 5.29£-08 1.87£-05 5.74E-07
10K YR 5.96E-09 6.52E-09 1.11£-08 1.58E-08 6.85£-09 7.66E-09 1.59E-08 3.03£-08 1.87E-05 5.69E-01
20K YA 3.65£-09 3.20E-09 7.02£~9 1.05E-OS 3.96£-09 4.3IE-0'l 6.74£-09 1.12E-08 1.87£-05 S.6!)E-07
- 15 7 -

-------
.......--...----.------------......------..-------------------.---------.-.--
Table B-7. Radionuclide-solubility combinations considered by IMPACTS.
....._-----_.._-._-._._._--------_..__.._------~------------.-----------------
  Solubi- Hal f-Ufe   Sol ubi-  Half-life
Nuclide 1ities (Years)  Nuclide lities (Years)
H-3 .1 1. 23E+012  Th-228 W,Y 1.91E+00
C-14 . 5.73E+03  Th-229 W,Y 7.34E+03
Na-22 ° 2.62E+00  Th-230 W,V 8.00E+04
Cl-36 O,W 3.08E+05  Th-232 W,V 1.41E+I0
Fe-55 W;Y 2.60E+00  Pa-231 W,V 3.25E+04
Co-60 W,V 5.26E+00  U-232 O,W,Y 7. 20E+Ol.
Ni-59 O,W 8.00E+04  U-233 O,W,V 1. 62E+05
Ni-63 O,W 9.20E+01  U-234 O,W,Y 2.47E+05
Sr-90 O,Y 2.81E+Ol  U-235 O,W,Y 7.10E+08
Nb-94 W,Y 2.00E+04  U-236 O,W,Y 2.39E+07
Tc-99 O,W 2.12E+05  U-238 O,W,Y 4.51E+09
Ru-l06 V 1.01E+00  Np-237 W,V 2.14E+06
Ag-108m O,W,V 1. 27E+02  Pu-236 W,V 2.85E+00
Cd-109 O,W,Y 1.24E+00  Pu-238 W,V 8.64E+Ol
Sn-126 O,W 1.05E+05  Pu-239 W,V 2.44E+04
Sb-125 O,W 2.71E+00  Pu-240 W,V 6.58E+03
1-129 ° 1. 17E+07  Pu-241 W,Y 1. 32E+Ol
Cs-134 ° 2.05E+00  Pu-242 W,Y 3.79E+05
Cs-135 ° 3.00E+06  Pu-244 W,V 7.60E+07
Cs-137 ° 3.00E+01  Am-241 W,V 4.58E+02
Eu-152 W 1.27E+Ol - Am-243 W,V 7.95E+03
Eu-154 W 1. 60E+01  Cm-242 W,V 4.45E-01
Pb-210 W 2.04E+01  Cm-243 W,V 3.20E+Ol
Rn-222 . 1.05E-02  Cm-244 W,V 1.76E+Ol
Ra-226 W 1.60E+03  Cm-248 W,V 4.70E+05
Ra-228 W 6.70E+00  Cf-252 W,V 2.65E+00
Ac-227 W,Y 2.16E+Ol    
(1) Solubility:. - Not applicable, D-Day, Wl- Week, V - Year
(2) Exponential Notation: 1.23E+Ol - 1.23 x 10 
(3) Radiation types: a - Alpha, b - Beta, 9 - Gamma 
- 158 -

-------
01: Number of landscape cells
08: Number of compartments in each cell
01: Output control, 01 for more detailed output otherwise 00
.. Compartment names tn 10 A6 format on the following record{s) ..
AIR PHAIR BIOTA SOIll SOIl2 GWTR SWTR SOMT lSTC NXTC
.. landscape and problem title on the following record ..
Western Ecoregion
1.00 e04: 1 area in km..2
1.00 e03: 2 height of the air compartment em)
5.00 e-6: 3 humidity (kg/1)
3.20 eOl: 4 precipitation onto land (cm/yr)
0.30 eOO: 5 precipitation onto surface water (cm/yr)
1.26 eOl: 6 total surface water runoff (cm(yr)
1.10 eOl: 7 land surface runoff (cm/yr)
6.15 eOl: 8 atmospheric dust load (micro-gm/m..3)
3.34 e02: 9 deposition velocity of atmospheric particles mid
7.00 e05: 10 biota dry mass inventory (kg/km**2)
9.00 e04: 11 biota dry mass production (kg/km**2/yr)
3.30 e-l: 12 biota dry mass fraction
1.88 eOl: 13 evapotranspiration from soil (cm/yr)
0.16 eOO: 14 evaporation from surface water (cm/yr)
1.60 e-l: 15 thickness of the A soil horizon (m)
1.50 eOO: 16 bulk density of the soil in the A horizon (kg/L)
3.80 e-l: 17 water content of the soil in the A horizon (kg/L)
7.00 e-2: 18 volumetric air content in the A horizon (l/L)
1.19 e06: 19 mechanical erosion rate (kg/km**2/yr)
0.68 eOO: 20 irrigation from ground water (cm/yr)
1.00 eOl: 21 thickness of the B soil horizon (m)
3.00 e-l: 22 water content of the soil in the B horizon (k9/L)
1.70 eOO: 23 bulk density of the soil in the B horizon (kg/L)
5.00 e-2: 24 volumetric air content in the B horizon (L/L)
1.10 el0: 25 groundwater inventory (kg/km*.2)
3.50 e-l: 26 porosity of rock in the ground water zone (L/L)
1.70 eOO: 27 density of rock in the ground water zone (kg/L)
8.42 e-3: 28 fraction of the total surface area in surface water
6.00 eOO: 29 average depth of surface waters em)
8.90 e-3: 30 suspended sediment load in surface water (kg/L)
3.83 e03: 31 deposition rate of suspended sediment (kg/m**2/yr)
5.00 e-2: 32 thickness of the sed.iment layer (m)
1.50 eOO: 33 bulk density of the sediment layer (k9/L)
0.20 eOO: 34 porosity of the sediment zone
3.83 e03: 35 resuspension rate from the sediment layer (kg/m**2/yr)
2.83 e02: 36 ambient environmental temperature (k)
1.00 e-2: 37 boundary layer thickness at air/soil interface em)
1.00 e-2: 38 boundary layer thickness at water/air interface em)
2.00 e-2: 39 boundary layer thickness at sediment/water interface em)
1.20 e-2: 40 fraction organic carbon in the upper soil zone
2.00 e-3: 41 fraction organic carbon in the lower soil zone
2.87 e-4: 42 fraction organic carbon in the groundwater zone
2.00 e-2: 43 fraction organic carbon in the sediment zone
1.00 eOO: 44 wet deposition scavenging efficiency (default = 1)
1.00 eOO: 45 yearly average wind speed (m/s)
Figure B-3. Sample landscape data file (western ecoregion).
- 159 -

-------
arsenic :e1ement or compound name
7.4900e01 :mo1ecu1ar weight
1 :type (1 for element, 0 or blank for organic compound)
1.00 e-6 :henry's law constant (torr/(mo1e/l)
3.171e-17 :diffusion coefficient in air (m**2/s)
3.171e-17 :diffusion coefficient in water (m**2/s) .
1.00 e-1 :part. coeff., ksp (ppm in plant dry mass)/(ppm total soil)
7.50 e01 :bioconcentration factor, bcf (ppm in fish meat)/(ppm water)
4.00 e-2 :part. coeff., xkmdl (mg/kg in meat)/(mg/kg dm in cattle diet)
4.00 e-2 :part. coeff., xkmd2 (mg/kg in milk)/(mg/kg dm in cattle diet)
1.300 e03 :part. coeff., xkds1 (ppm soi1)/(ppm water) in upper soil
1.300 e03 :part. coeff., xkds2 (ppm soi1)/(ppm water) in lower soil
1.300 e03 :part. coeff., xkdgw (ppm rock)/(ppm water) in groundwater
1.300 e03 :part. coeff., xkdsd (ppm solid)/(ppm water) in surface water
*** removal rate constants (l/day) in each compartment ***
1.10 e-4 8 :sdmt
end of list
*** transform rate constants (l/day) in each compartment ***
end of 11 st
*** the source vector (values in mo1/km**2 per day) ***
2.7397e-34 :so;11
end of 1 ist
benzene :e1ement or compound name
7.8120eOl :mo1ecu1ar weight
:type (1 for element, 0 or blank for organic compound)
4.10 e03 :henry's law constant (torr/(mole/1)
4.80 e01 :organic carbon partition coefficient koc
4.98 e-6 :diffusion coefficient in air (m**2/s)
4.98 e-10 :diffusion coefficient in water (m**2/s)
0.00 eOO :part. coeff.,ksp (ppm in plant dry mass)/(ppm total soil)
7.50 eOI :bioconcentration factor, bcf (ppm in fish meat)/(ppm water)
4.00 e-2 :part. coeff., xkfdl (mg/kg in meat fat)/(mg/kg dm in cattle diet)
4.00 e-2 :part. coeff., xkfd2 (mg/kg in milk fat)/(mg/kg dm in cattle diet)
*** removal rate constants (l/day) in each compartment ***
2 . 08 e - 2 1 : air
2.08 e-2 2 :pmair
1.10 e-4 8 : sdmt
end of list
*** transform rate constants (l/day) in each compartment ***
end of 1 i st
*** the source vector (values in mo1/km**2 per day) ***
2.7397e-3 4 :soill
end of list
Figure B-4.
Sample Chemical data file (arsenic and benzene).
- 160 -

-------
- -.--...-
T.O E01: 1
1. 70 EO 1 : 2
1.80 EOO: 3
0.68 EOO: 4
2.20 E04: 5
1. 00 E04: 6
2 .0 EOO: 7
1.0 EOO: 8
0.13 EOO: 9
0.15 EOO: 10
0.30 EOO: 11
0.50 EOO: 12
0.30 EOO: 13
0.10 EOO: 14
6 . SO E - 3: 15
2 . 00 E - 4: 16
5 . 80 E - 5 : 17
1.00 E-4: 18
AMASS THE BODY MASS OF AN ADULT IN KG
CMASS THE BODY MASS OF A CHILD IN KG
AAREA THE SURFACE AREA OF AN ADULT IN M**2
CAREA THE SURFACE AREA OF A CHILD IN H**2
AINHL THE INHALATION BY AN ADULT IN L/DAY
CINHL THE INHALATION BY A CHILD IN L/DAY
ADRNK THE INGESTION Of WATER BY AN ADULT IN L/DAY
CDRNK THE INGESTION Of WATER BY A CHILD IN L/DAY
AVEG THE INGESTION Of VEGETATION BY AN ADULT IN KG(DM)/DAY
CVEG THE INGESTION OF VEGETATION BY A CHILD IN KG(DM)/DAY
AMILK THE INGESTION Of MILK BY AN ADULT IN L/DAY
CMILK THE INGESTION Of MILK BY A CHILD IN L/DAY
AMEAT THE INGESTION OF MEAT BY AN ADULT IN KG/DAY
CMEAT THE INGESTION Of MEAT BY A CHILD IN KG/DAY
AFISH THE INGESTION OF FISH BY AN ADULT IN KG/DAY
CFISH THE INGESTION OF FISH BY ~ CHILD IN KG/DAY
ASOIL THE INGESTION OF SOIL BY AN ADULT IN KG/DAY
CSOIL THE INGESTION OF SOIL BY A CHILD IN KG/DAY
Figure B- 5.
Sample input data file for the program EXPOSE.
- 161 -

-------
01: Control variable. When set to 00, initial inventories in all
* compartments are set to 0.0 . When set to 01, initial inventories
* are set at the steady-state values obtained from GEOTOX A.
21: Total number of time steps (must be < or . 50)
** Time step values (in days)
i*. listed in F11.1 format (values between colons are not
. 000000.0: - 1 -: 10.0: - 2 -: 20.0: - 3 -:
100.0:- 5 -: 150.0:- 6 -: 200.0:- 7-:
300.0:- 9 -: 350.0:- 10 -: 400.0:- 11 -:
500.0:- 13 -: 550.0:- 14 -: 600.0:- 15 -:
700.0:- 17 -: 750.0:- 18 -: 800.0:- 19 -:
1000.0:- 21 -:
*'The source data is used to construct a source of the following form:
. * SOURCE(i,t). [(A + 8*H(t,Tl,T2)/(T1 - T2)]*AREA
. . where,
. . SOURCE. the time dependent source term for species 1 at
; * time t in mole/day,
I . A . the time-independent component of the source in mol/km**2 per day,
. 8 . the total amount of material that is introduced over the
* time TI-T2 in mol/km**2,
. H(t,Tl,T2) . function whose value is 1 when t is between Tl and
* T2 and 0 otherwise,
. Tl . time in days when the "8" component of the source begins,
. T2 . time in days when the "8" component of the source ends, and
. . AREA = the area of the landscape in km**2.
000.0: Value of Tl in days
100.0: Value of T2 in days
. *Source constants in ElO.3 format
* Chemical nAn constants "8" constants
* number (mole/day) (mole)
------>01:-----> 2.74 e-3:-----> 1.37
02 2.74 e-3 1.37
03 2.74 e-3 1.37
04 2.74 e-3 1.37
read by the progr~m)
30.0:- 4.:
250.0:- 8-:
450.0: - 12 -:
650.0:- 16 -:
900.0:- 20 -:
Compartment Compartment
number name
e02:--------->04:-------> SOIlI:
eOO 04 SOIll
e-2 01 AIR
e-2 01 AIR
Figure B-6.
Sample time step and source term data file.
- 162 -

-------
1.1 Name: 
1.2 Prepared by:
1.3 Prepared for:
1:4 Report Title:
1.5 Report Number:
1.6 Repor~ Date:
1.7 Availability:
1.8
Purpose and Scope:
1
IDENTIFICATION
GEOTOX Version 1.2
Thomas E. McKone, Environmental Sciences Division,
Lawrence Livermore National Laboratory
u.S. A~ Biomedical
Laboratory (USABRDL)
Research
Development
and
GEOTOX Multimedia Compartment Model User's Guide
UCRL-159l3
May 1987
Documentation: U.S. Government Printing Office,
National Technical Information Center Software: T.E.
McKone or D.W. Layton, Environmental Sciences
Divisipn, Lawrence Livermore National Laboratory
Multimedia compartment
simulates the transport
environmental contaminants
exposure.
screening model that
and transformation of
and potential human
2
SUMMARY OF FINDINGS
GEOTOX is a set of programs used to calculate time-varying chemical
concentrations in multiple environmental media (e. g., soil, ground water,
etc.) and to estimate potential human exposures. The current version of
GEOTOX performs two major tasks: (1) it predicts the transport and
transformation of chemicals in a multimedia environment, and (2) it estimates
human exposure. The chemical transport model uses landscape data and
physicochemical properties to determine the distribution and concentration of
chemicals among compartments such as air, water, and soil. Environmental
concentrations are linked to human exposures and health effects using an
exposure model that accounts for intake through inhalation, consumption of
food and water, and dermal absorption. GEOTOX is intended for use in public
health and environmental risk assessment and risk management -- particularly
for the screening and ranking of chemicals according to the potential risks
they pose.

The present version of GEOTOX is coded in MicrosoftR FORTRAN and
designed for the IBM Personal Computer (PC) and its family of successors and
compatibles. GEOTOX was originally developed for ranking the potential
health risks associated with toxic metals and radionuclides in the global
environment (McKone, 1981). Recently, at the Lawrence Livermore National
Laboratory, the model has been extended to handle organic chemicals (Layton
et a1., 1986). GEOTOX was originally implemented on mainframe systems.
Because of this, most of the input must be entered in fixed-format data
files. However, comments have been added to the input files to make this
format easy to use.
- 163 -

-------
The U.S. Army Biomedical Research and Development Laboratory (USABRDL)
has been charged with the responsibility of developing environmental criteria
for military materials such as propellants, pyrotechnics, and explosives as
well as by-products associated with their demilitarization. USABRDL has
funded Lawrence Livermore National Laboratory to prepare a health and
environmental effects data-base assessment on technologies for demilitarizing
conventional ordnance and the environmental by-products of those
technologies. In the first phase of the project, explosives and associated
by-products were ranked according to their potential health risks. The
GEOTOX program addressed in this report was used to estimate the equilibrium
partitioning of demilitarization by-products between different environmental
compartments and to determine human exposures to the contaminants via
different pathways (e.g., inhalation, ingestion, etc.).
3
ADMINISTRATIVE CRITERIA
3.1
Documentation
The original version of GEOTOX was developed at the University of
California at Los Angeles for ranking the potential risks of toxic metals and
radionuclides in the global environment. A detailed description of this
version of the model is provided in the dissertation entitled "Chemical
Cycles and Health Risks of Some Toxic Crustal Nuclides" (McKone, 1981); and a
summary and application of the original model is published in the journal
Risk Analysis (McKone et al., 1983). At Lawrence Livermore National
Laboratory, the model was modified to handle organic chemicals and to run on
a personal computer. These modifications are described in the report
"Demilitarization of Conventional Ordnance: Priorities for Data-Base
Assessments of Environmental Contaminants," (Layton et al. ,1986). Thi$
version of the model was referred to as GEOTOX version 1.0. The theory
behind this version of the model was published in a peer review journal
article (McKone and Layton, 1986). A user's guide for GEOTOX was prepared in
1987 (McKone et al., 1987). Additional discussions of the revised GEOTOX
model (Version 1.0) are available in McKone (1986), McKone and Layton
(1986b), McKone and Kastenberg (1986) and Layton and McKone (1988). The
current version of the model is Version 1.2. Differences between Versions
1.0 and 1.2 are described in the user's guide supplement (McKone, 1988).
3.2 Hardware Requirements

GEOTOX requires an IBM PC, XT, AT, or compatible, running under MS-DOSR
(version 2.0 or higher), with at least 150 kilobytes of free RAM. The ANSI
device driver, supplied with DOS, is also requires. a printer with 8.5 inch
or wider paper is recommended. The program does not require an 8087 math
coprocessor, but runs faster if one is present.
3.3
Application
GEOTOX consists of three principal programs: GEOTOX-A, GEOTOX-B, and
EXPOSE. Each of these programs is designed to handle a specific task in the
sequence of calculations that translate the continuous or short-term addition
of a chemical to a compartment (e.g., the upper-soil layer) into an estimate
- 164 -

-------
of human exposure. The GEOTOX batch file coordinates the input, execution,
and the output of the GEOTOX modules.
GEOTOX-A calculates the transfer rate constants and finds the steady-
state inventories of chemicals in an eight-compartment landscape. GEOTOX-B
uses the matrix of transfer and transformation rate constants calculated by
GEOTOX-A and sources defined as input to calculate time-varying inventories
and concentrations in the eight compartments. GEOTOX-B calls a set of
subroutines called GEARB to solve systems of first-order ordinary
differential equations. GEARB selects its own time step size according to an
internal convergence test. It only returns output information for the times
specified by the user. EXPOSE uses the information from GEOTOX-A or GEOTOX-B
to determine human exposure by inhalation, drinking water, biota ingestion,
meat and dairy product ingestion, fish ingestion, soil ingestion, and dermal
absorption. Using EXPOSE with input from GEOTOX-A results in calculations of
steady-state exposure, while using EXPOSE with input from GEOTOX-B results in
time-varying exposures.
3.4
Level of Expertise Required
The set of GEOTOX programs is controlled by a batch file which limits
the need for the user to define input and output sets. Through the use of
function keys this batch file allows the user to execute specific GEOTOX
programs and view the output listings and graphic output of the programs.
There is currently no "user friendly" system for preparing input to the
GEOTOX programs. However, the program comes with several sample input files
that have comments nested with each line to assist the user in tracing and
modifying the inputs. Creating or modifying the input files requires an
editor such EDLIN, but no editor is included with the program.
TECHNICAL CRITERIA
4.1
Peer Review
The GEOTOX model itself has not received extensive peer review, however
the models and theory behind this model have been published in the peer
review literature. A summary review and application of the original GEOTOX
model is published in the journal Risk Analysis (McKone et al., 1982). The
theory behind Version 1.0 was published in Re~ulatory Toxicology and
Pharmaco10~ (McKone and Layton, 1986).
4.2
Verification
The GEOTOX program was tested and debugged as part of the program
development process at Lawrence Livermore National Laboratory including
verification of mathematical computations.
4.3
Uncertainty
GEOTOX offers the user the option of performing a Monte Carlo analysis,
but this option is not available in Version 1.2 and will not execute.
- 165 -

-------
4.4
Sensitivity
The user does have the option of performing a first order differential
sensitivity analysis using GEOTOX A and EXPOSE A by requesting option B from
the list of three options that appears when the program GEOTOX A begins
execution. This sensitivity analysis can only be applied to one chemical at
a time.
The differential sensitivity analysis calculates the normalized partial
derivative of the output chemical concentrations and human exposure with
respect to each input parameter in both the landscape and chemical properties
file. This derivative is calculated as follows:
l:i! - a (F[Xi] - F[Xi*(l - a)])/F
1.
(1)
where I dF /dXi I is the normalized partial derivative of the output F with
respect to the parameter Xi and a is small fraction (0.01). The differential
sensitivity analysis allows one to make a quick estimate of the influence of
small changes in each input parameter on the reference estimates of chemical
concentration and human ~xposure.
4.5 Required Inputs
GEOTOX requires four classes of input data: landscape description,
chemical description, human exposure data and source term data. These data
are accessed by the program through three input files, which are detailed
below.
Landscape Data. The GEOTOX program uses the proto~ol
input files end with the extension ".LND". All files with
listed on the screen when the GEOTOX-A program requests
There are four landscape files available in the data disk:
that all landscape
this extension are
a landscape file.
CALIF.LND, NORTHC.LND, SOEAST.LND, and WEST.LND.
which-represent the California, northeast/central, southeastern, and western
ecoregions of the United States. These files contain six setup records and
45 input parameters. Figure 1 provides a listing of the input file WEST.LND
as an example of a landscape data file. Each entry line is divided into two
sections. The entry value is listed first, followed by a colon and a
description of the variable. This description is provided for the benefit of
the user but not read by the program. Lines that begin with an asterisk are
comment lines, which are not read by the program but must be present.
Chemical Data. GEOTOX uses the protocol that all chemical input files
end with the extension ".CHM". all files with this extension are listed on
the screen when the GEOTOX-A program requests a chemical file. There are
fourteen chemical data files available on the data disk:
ARSENIC.CHM
PCE.CHM
TNT. CHM
DBCP.CHM
TCDD.CHM
VOCS.CHM
DBUTLPHT.CHM
TCE.CHM
BENZENE.CHM
EDB.CHM
TEN.CHM
CADMIUM. CHM
LEAD. CHM
TEST.CHM
- 166 -

-------
All but three of these contain a single chemical corresponding to the name of
the file. The file TEN.CHM contains ten of the listed chemical species; the
file VOCS.CHM contains three volatile compounds; and the file TEST.CHM
contains arsenic and benzene data used for the sample problem. GEOTOX-A can
handle up to 10 chemicals per set. However, GEOTOX-B can only handle 4 per
set. Figure B-4 provides a listing of the input file TEST.CHM as an example
of a chemical data file. The first record provides the name of the first
chemical species in the file. The second record provides the molecular
weight of this element or compound. The third record signals whether the
chemical is an organic (0 or blank), or inorganic species (1). For an
inorganic species, the next eleven records provide information on its
environmental properties. For an organic species, the next eight records
provide environmental data. There must be exactly eleven records for an
inorganic species and exactly eight records for an organic species. The
removal rate constants are stored in an array of length n (where n is the
number of co~partments). Only nonzero values need be specified. The rate
constant (d- ) and the corresponding compartment number are read in the same
line. The removal rate constants are the rate constants for removal 1;>y
radioactive decay and all physical and chemical transformations, inciuding
biodegradation, hydrolysis, etc. The input array is terminated by a record
having ten blank columns at the beginning. The nexl array contains the
transform rate constants, which reflect the rate (d- ) at which a given
chemical transforms to the next chemical on the list. The last ~rray is the
steady state source vector, which specifies the source (mo1e/km -d) and the
compartment number. -There must be a non-zero source term in at least one
compartment for the program to run.
EXDosure Data. The program EXPOSE reads input parameters for the
exposure calculation (i.e. human body mass and ingestion rates) from the file
EXPOSE.DAT, which can be modified by the user. Figure B-S lists the contents
of the file EXPOSE.DAT. In this file as in all GEOTOX input files, anything
to the right of the colon is not used as input but is only for
identification.
GEOTOX-B InDut. The file TIME.DAT is the time step/source term input
file for GEOTOX~. This file tells GEOTOX-B the time steps for the output
file and for the exposure calculation and defines the time-dependent source
term. TIME.DAT can have source constants listed for four chemicals even if
less than four chemicals are being used in the calculation. Only the
required records are read and the others are ignored. The first line of this
file allows the user to accept the steady-state inventories from GEOTOX-A as
the initial inventories or to set all compartment inventories to zero. The
sample input file TIME.DAT is shown in Figure B-6.
4.6
Output
GEOTOX-A OutDut. GTXA.OUT is the output file from GEOTOX-A that is
sent to the printer. The first page of output summarizes the distribution of
mass among the eight compartments of the landscape cell and the exchange of
soLids and. liquids between the compartments. Following this overview; the
movement of each species identified in the chemical input file is allotted
one page of output. A listing of the physical constants and partition
coefficients is followed by a table of the rate constants that define the
movement of the species in the eight-compartment landscape. Three rate
- 107 -

-------
constants mediate the movement of chemicals: alpha (diffusion), beta (liquid
advection). and betas (solid advection). After the rate constants are
calculated, the inflow and outflow of the chemical is listed for each
compartment. Finally, the distribution of the chemical in each compartment
is summarized. The final summary table provides an estimated inventory of
the specific chemical in each of the eight compartments. Figure B-7 shows
GEOTOX-A output for the sample problem.
GEOTOX-B Output. GTXB.OUT is the output file from GEOTOX-B that is
sent to the printer. For each time step specified in TIME. DAT, GTXB. OUT
lists the parameters that define the compartmental inventories for each
chemical being considered. These parameters include the source rate, the
inventory. the concentration, and the loss rate for each compartment.
EXPOSE Output. EXPOSA.OUT is the steady-state exposure file written by
EXPOSE that is based on the compartment inventories produced by GEOTOX-A.
For each chemical being considered, EXPOSA.OUT lists the adult, child, and
lifetime average exposures via seven pathways. EXPOSB. OUT is the time-
varying exposure file written by EXPOSE that is based on the compartment
inventories produced by GEOTOX - B . For each chemical be ing cons idered,
EXPOSA.OUT lists the adult, child, population average, and cumulative
exposures via seven pathways, for each of the time steps specified in
TIME.DAT as input to GEOTOX-B.
Graphic Output. Graphic output displays are available for the programs
GEOTOX A and EXPOSE A. The graphic output for GEOTOX-A lists the
distribution of each contaminant by compartment. Figure B-8 provides a
sample of the graphic output from GEOTOX-A. In the EXPOSE graphic output,
the chemical selected has its exposure distribution summarized by one of two
histograms. The EXPOSE graphic output also provides an option to compare all
chemicals in a given set on the basis of the ratio of the exposure estimated
by GEOTOX relative to a reference exposure entered by the user. Figures B-9
through B-14 show sample graphic output from EXPOSE. GEOTOX-B produces an
output fi~e called GXBOUT. PRN in a format that allows it to be imported to
Lotus 123 R) in order to create graphs of the output or perform other types
of analyses available in the s(Jread-sheet program. Figure B-1S shows a plot
that was made using Lotus 123 R) and the data in GXBOUT.PRN for the sample
problem.
4.7 Source Term
The source S(i,n,t) for the nth chemical in compartment i at time t is
defined by the expression:
S(i,n,t) - (a + b/(t2 - tl) x H(t,t1,t2)}A
(1)
where:
a - source constant for the continuous source, mo1ejkm2-d;

b - source constant for the time-varying source, molejkm2;
t1,t2 - time that define the interval during which a time-varying source
is present, d;
- 168 -

-------
H(t,tl,t2) - a step function that is equal to 1 if t is between tl and
t2' and is equal to 0 otherwise, and

A - area of the landscape, km2
GEOTOX-B uses Equation 1 to obtain the compartment inventories
situations in which the source term varies as a function of time.
for
4.8
Scenarios
GEOTOX is a tool for screening the potential risks of contaminants
released as nonpoint sources to soils, air, ground water, and surface water.
It can handle radionuc1ides, metals, volatile organic compounds, and other
organic compounds in a single execution. GEOTOX serves as a screening model
when it is used to calculate the exposure that results from the addition of
each chemical to environmental compartments within a specified ecoregion.
It is widely recognized that of the rou~ly 70,000 chemicals now in
commercial use, some can cause serious harm at low exposure levels.
Nonetheless, relatively little is known about how most of these species
behave in the environment. The very magnitude of this list precludes a
detailed assessment of every chemical. However, with limited data, one can
perform preliminary screening studies to identify the compounds that
represent the most immediate hazards. This, indeed, is the goal of GEOTOX.
A multimedia compartment model can screen chemicals based on their
inherent toxicity, persistence, and dilution. In this context, it can be
viewed as a tool for comparing potential risks. The concept of risk applies
here, because we are dealing with uncertainties in both exposure and health
effects. McKone and Kastenberg (1986) have identified five criteria that
make a multimedia. model suitable for risk assessment or risk management: (1)
able to handle organic chemicals, trace metals, and radionuc1ides; (2) fast
enough to allow multiple runs for sensitivity studies; (3) able to link
environmental concentrations with human exposure pathways; (4) able to handle
dynamic and steady-state situations; and (5) flexible enough for easy
alteration to simulate different types of environments. No model fully
satisfies all these criteria, but the GEOTOX model was developed with these
objectives in mind.
An important feature of the GEOTOX model is the ability to link
estimated environmental concentrations with exposure pathways in order to
project accumulated lifetime exposures within a population. GOETEX uses
three primary exposure pathways inhalation, ingestion, and dermal
absorption. The ingestion route is divided into water; fruits, grains, and
vegetables; milk and meat; fish; and soiL Exposure is expressed as the
daily contact in mg/kg of body weight of a given chemical with the lungs, gut
wall, or skin surface. Exposure is averaged over three age categories--child
(0 to 15 y), adult (15 to 70 y), and lifetime average.
- 169 -

-------
5
SCIENTIFIC CRITERIA
5.1
Theory
Predicting the movement of a toxic substance in the environment involves
the use of models that describe the partitioning of chemical species among
the various environmental media, including air, water, soils, and sediments.
One approach to this problem is the use of compartment models. Models of
this type have been developed over the last decade for studying the global
fate of toxic elements and radionuclides and for studying the transport and
transformation of organic chemicals. Compartment models are often a
reasonable starting point in environmental assessments because they cover all
the primary media simultaneously. Multimedia models can be structured to
describe complex systems with hundreds of state variables. However, the
amount of information available for most chemicals restricts practical models
to the consideration of one or more atmosphere compartments, two or three
soil zones, and land biota, ground water, and sediment compartments used by
the GEOTOX model.
The transport and transformation equations solved by GEOTOX have the
general form:
dN(i,n,t) J 1 m
dXi - -L(i,n)N(i,n,t) - SUM~ii' [T(i->j,n)N(i,n,t)]-

T(i==>o,n)N(i,n,t) + K(i,n-l)N(i,n-l,t) +
(2)
SUMj=l,m [T(j-->i,n)N(j,n,t)] + S(i,n,t)
j/i
where:
N(i,n,t) = time-varying inventory of species n in compartment i, moles;
L(i,n) - first-order rate constant for removal of species
compartment i by chemical decompositon, etc., lid;
n
from
T(i-->j,n,t) = rate constant for the transfer
compartment i to compartment j, lid;
of
species
n
from
K(i, n-l) - first-order rate constant for the transformation of species
n-l to species n within compartment i,. lid;
T(i-->o,n)
= nite constant for
compartment i to a
system, lid;
the transfer of species
compartment outside of the
n frQm
landscape
S (i, n, t) - source term for the introduction. of
compartment i, mole/d; and
species
n
into
m - total number of compartments within the landscape system.
- 170 -

-------
GEOTOX-A, the first module of GEOTOX, obtains the steady-state solution
to Equation 2. When Equation 2 is written for m compartments and k
chemicals, the following matrix equation is obtained:
~tn(t) - [T]n(t) + ~(t)
(3)
in whichn and ~ are vectors of length M-mxk and [T] is a matr~x
M. A steady-state solution for this system of equations 1S
defining a constant source vector
of size K by
obtained by
Results of the GEOTOX-A and GEOTOX-B modules are used in another module,
termed EXPOSE, which estimates exposure levels for humans from a series of
pathways. Exposure is expressed as daily intake per unit body weight averaged
over the population. The general model used to calculate exposure has the
form:
E - (ajBW) SUMj-l,p Ii(t)
where:
E - lifetime exposure, mgjkg-d;
BW - body weight, kg;
p - number of pathways; and
Ii - the daily intake by pathway i, mg/d based on the environmental
concentrations estimated using GEOTOX-A or -B.
5.2
Validation
Kodel performance predictions
disposal site performance.
have
not
been
compared
with
actual
5.3
Treatment of Decay Products
GEOTOX simulates all decay and transformation processes (radioactive
decay, hydrolysis, photolysis, oxidation, biodegradation, sedimentation, and
advective losses) as first-order removal processes. Each process is governed
by its own decay constant, and is treated as irreversible. Any fraction of
the contaminant which is transformed in a given compartment can be treated as
a source for another contaminant in the simulation. This allows the
treatment of radioactive decay chains or chemical transformation sequences.
5.4
Underlying Assumptions
GEOTOX lumps each component of the environment into a homogeneous
subsystem or compartment that can exchange water, nutrients, and chemical
contaminants with other adjacent compartments. This compartment system is
used to simulate the behavior of a single substance or of two or more
substances linked by transformation, such as a radionuclide and its progeny.
A compartment is described by its total mass, total volume, solid-phase mass,
liquid-phase mass, and gas-phase mass. Mass flows among compartments include
- 171 -

-------
solid-phase flows, such as dust suspension or deposition, and liquid-phase
flows, such as surface run-off and groundwater recharge. The transport of
individual chemical species among compartment occurs by diffusion and
advection at the compartment boundaries. Each chemical species is assumed to
be in chemical equilibrium among the phases within a single compartment.
However, there is no requirement for equilibrium between adj acent
compartments.
For example, in the upper soil layer; which contains solids, liquids,
and gases; an organic chemical added to the soil distributes itself among
these three phases such that it achieves chemical and physical equilibrium.
Among the potential transport pathways from the upper-soil compartment are
liquid advection (s01l water run-off), solid-phase advection (erosion to
surface water or dust stirred up and blown about), and diffusion from the
soil gas phase into the lower atmosphere.
5.5
Exposure Pathways
In GEOTOX, the exposure function E is divided into a series of terms
that relate sources to environmental concentrations and the concentrations to
human contact:
E - SUM C.(S)F6i.
i j 1 J
(i-air,soil,water)
(j-inhalation,ingestion,dermal)
where Ci(S) is the concentration of contaminant in environmental compartment
i associated with the source Sand Fi' is the pathway exposure factor (PEF)
that relates this concentration to a level of human contact through pathway
j. The PEF incorporates information on anatomy, physiology, diet, activity
patterns, and environmental partitioning into a term that translates a unit
concentration into a daily exposure in mgjkg-d via inhalation, ingestion, or
dermal absorption.
5.6
Dose Conversion and Dose-Response
The GEOTOX model calculates lifetime equivalent daily exposure but
not explicitly calculate the risk associated with this exposure. It
provide the option of comparing exposures on the basis of a reference
dose (or exposure).
does
does
safe
- 172 -

-------
6
REFERENCES
Layton, D.W., T.E. McKone, C.H. Hall, M.A. Nelson, and Y.E. Ricker (1986)
Demilitarization of Conventional Ordnance: Priorities for Data Base
Assessments of Environmental Contaminants, Lawrence Livermore National
Laboratory, UCRL-15902.
Layton, D.W. and T.E. McKone (1988) "Multimedia Transport of
Contaminants and Exposure Modelling," Toxicology Letters (in press).
Organic
McKone, T.E. (1981) Chemical Cycles and Health Risks of Some
Nuclides, Ph.D. dissertation, University of California, Los Angeles.
Crustal
McKone, T.E. (1986) ."Dioxin Risk Management at Times Beach, Missouri:
Evaluation," The Environmental Professional ~:13-24.
An
Mckone, T. E. (1988) GEOTOX Multimedia Compartment Model
Supplement, Lawrence Livermore National Laboratory, UCID-21673.
Users'
Guide
McKone, T.E., and D.W. Layton (1986a) "Screening the Potential Risks of Toxic
Substances Using a Multimedia Compartment Model," Reg. Tox. Pharm. £:359-380.
McKone, T.E., and
and Health Risk:
Risk Mana~ement.
Pittsburgh.
D.W. Layton (198Gb) "Chemical Transport, Human exposure,
A Mul timedia Approach," in J. S . Evans (ed), Environmental
Is Analysis Useful?, Air Pollution Control Association,
McKone, T.E., W.E. Kastenberg, and D. Okrent (1983) "The Use of Landscape
Chemical Cycles for Indexing the Health Risks of Toxic Elements and
Radionuclides"" Risk Analysis 1: 189-205.
McKone, T.E., and W.E. Kastenberg (1986) "Application of Multimedia Transport
Models to Risk Analysis," in Y. Cohen (ed), Pollutants in a Multimedia
Environment, Plenum Press, New York.
- 173 -

-------
-stern (r JrI'g Ion
ANDSCAP£ MAS SOl S TR I BUT I ON AND EXCHANG£
COMPARTMENT
AIR
PMAIR
BIOTA
SOILl
SOlL2
GWTR
SWTR
SDMT
GAS MASS +
(kg)
1. 200E+13
O.OOOE+OO
5.120E+02
1.343E+08
6.000E+09
O.OOOE+OO
1. 434E+04
O.OOOE+OO
LIQUID MASS +
(kg)
5.000E+07
O.OOOE+OO
1.421E+I0
6.080E+ll
3.000E+13
1.100E+14
5.052E+ll
8.420E+08
TOTAL MASS
(kg)
1.200E+13
6,150E+05
2.121E+10
3.008E+12
2.000E+14
6.443£+14
5.097E+ll
7. 157E+09
SOLID MASS.
(kg)
O.OOOE+OO
6.150E+05
7.000E+09
2.400E+12
1.700E+14
5.343E+14
4. 496E +09
6.315E+09
VOL lJ4£
nlters)
1.000E+16
1.000E+16
2.121E+I0
1.600E+12
1. 000E+14
3.143E+14
5.052E+ll
4. 210E+09
MISS Exchange of So lids Among the Compartments in kg/d :
Matrix of flows from top compartments to the left-hand compartments:
AIR PHAIR BIOTA SOlll SOll2 GWTR SVTR SOMT lSTC TOn .INFlW
O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO
O.OOOE+OO O.OOOE+OO O.OOOE+OO 3.290E+05 O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO 3.290E+05
O.OOOE+OO O.OOOE+OO O.OOOE+OO 2.466E+06 O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO 2.466E+06
O.OOOE+OO 2.054E+05 2.466E+06 O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO 3.260E+07 3.527E+07
O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO
O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO
O.OOOE+OO O.OOOE+OO O.OOOE+OO 3.260E+07 O.OOOE+OO O.OOOE+OO O.OOOE+OO 8.835E+08 O.OOOE+OO 9.161E+08
O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO 8.835E+08 O.OOOE+OO O.OOOE+OO 8.835E+08
O.OOOE+OO 1.236E+05 O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO 3.260E+07 O.OOOE+OO
AIR
PMAIR
810TA
SOILl
SOlL2
GWTR
SWTR
SDHT
NXTC
TOTl OTFlW O.OOOE+OO 3.290E+05 2.466E+06 3.540E+07 O.OOOE+OO O.OOOE+OO 9.161E+08 8.835E+08
HISS Exchange of Liquid Among the Compartments in kg/d :
Matrix of flows from top compartments to the left-hand compartments:
AIR PHAIR BIOTA SOIL! SOIL2 6IlTR SWTR SOMT LSTC TOTL INFlW
O.OOOE+OO O.OOOE+OO O.OOOE+OO 5.151E+09 O.OOOE+OO O.OOOE+OO 4.384E+07 O.OOOE+OO 3.452E+09 8.647E+09
O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OooE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO 0.000£+00 O.OOOE+OO O.OOOE+OO
O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.ooOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO
8.767E+09 O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO 1.863E+08 O.OOOE+OO O.OOOE+OO O.OOOE+OO 8.953E+09
O.OOOE+OO O.OOOE+OO O.OOOE+OO 7.890E+08 O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO 7.890E+08
O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO 7.890E+08 O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO 7.890E+08
8.219E+07 O.OOOE+OO O.OOOE+OO 3.014E+09 O.OOOE+OO 6.027E+08 O.OOOE+OO O.OOOE+OO O.OOOE+OO 3.699E+09
O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO
O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+OO O.OOOE+oo 3.452E+09 0.000£+00
AIR
PMAIR
BIOTA
SOIL1
SOIL2
GWTR
SWTR
SDMT
NXTC
TOTL OTFlW 8.849E+09 O.OOOE+OO 0.000£+00 8.953E+09 7.890E+08 7.890E+08 3.496E+09 O.ooOE+OO
Figure B-7.
Sample output from GEOTOX A.
- 174 -

-------
C"£MICAL 'ROPERTIES FOR ar..nlc
MOLECULAR WEIGHT... 7.490[+01
AIR DIFFUSION CONSTANT... 3 171[-17 M882/S
PARTITION COEFFICIENTS:
AIR/WAT[R..... ~.663[-ll ~G/L
SOILZ/WATER... 1.300[+03
S[QMT/SFWTR... 1.300E+03
MEAT/DIET..... 4.000[-02
fISH/WATER. ... 7.500[+01
H[NU'S LAW CONSTANT... 1.000l-06 TORR/(MOlE/L)
WATER DIFFUSION CONSTANT... 3.171[-17 M8.2/5
SOIL I/WATER. . .
ROCK/GROWTR. . .
BIOTA/SoILl. . .
MILK/DIET.... .
1.300[-03
1 . 300E +03
1. OooE -01
4.000£-OZ
CALCULATION OF RAT£ CONSTANTS FOR arsenic
ALPHA O/day) + flOW (kg/day) .x BETAl (I/kg) + SflOW (kg/day) x 'BETAS (I/kg) - T(luj) O/day)
3.04ZE-15 8. 767E+09 1:978E-08 o.OOOE+oO O.OOOE+OO 1.734E+02
2.5B3E-17 8.219E+07 1.978E-08 0.000£+00 O.OOOE+OO 1.625E+00
2.010E-Ol 0.000£+00 1.978£-08 0.000£+00 O.OoOE+OO 2.010£-01
0.000£+00 0.000£+00 O.OOOE+OO 2.054£+05 1.626£-06 3.340E-Ol
0.000£+00 0.000£+00 0.000£+00 1.236E+05 1.626£-06 2.010E-Ol
O.OOOE+OO O.OoOE+OO O.OOOE+OO 2.466£+06 1.429£-10 3.523E-04
4.930(-23 5.151£+09 O.OOoE+OO O.OOOE+oO 4.166£-13 4.930£-23
0.000£+00 0.000£+00 3.205E-16 3.290E+05 4.166£-13 1.371E-07
0.000£+00 O.OOOE+OO 3.205E-16 2.466£+06 3.324E-14 8.197£-08
0.000£+00 7.890E+08 3.205E-16 0.000£+00 4.166£-13 2.528E-07
0.000£+00 3.014£+09 3.205£-16 3.260£+07 4.166£-13 1. 455E-05
0.000£+00 7.890£+08 4. 524E-18 O.OOOE+OO 5.882£-15 3. 570E -09
0.000£+00 1.863£+08 - 1.440£-18 O.OOOE+OO 1.871E-15 2.682E-l0
0.000£+00 6.027£+08 1.440£-18 O.OOOE+OO 1.871E-15 8.676E-I0
2.057£-22 4.384£+07 0.000£+00 0.000£+00 2.047£-10 2.057£-22
1.816E-12 0.000£+00 1.575£-13 8.835£+08 2.047E-l0 1.809£-01
0.000£+00 3.452£+09 1. 575E-13 3.260£+07 2.047£-10 7.213£-03
1.405£-12 0.000£+00 1.218£-13 8.835£+08 1. 583E-l0 1. 399E-Ol
AIR -> SOILl
AIR a> SWTR
AIR a>. NXTC
PMAIRa> SOILl
PMAIRa> NXTC
BIOTAa> SOILl
SOILla> AIR
SOILl-> PMAIR
SOILla> BIOTA
SOILla> SOIl2
SOILla> SWTR
SOIL2a> GWTR
GWTR -> SOILl
GWTR -> SWTR
SWTR a> AIR
SWTR -> SoMT
SWTR -> NXTC
SOHT a> SWTR
COMPARTMENT
AIR
PMAIR
BIOTA
SOILl
SOlL2
GWTR
SWTR
SOHT
1------- SOURCES (.ol/day)
direct transform
0.000£+00 O.OooE+OO
O.OOOE+OO O.OOOE+OO
0.000£+00 O.OOOE+oO
2.740£+01 0.000£+00
0.000£+00 0.000£+00
0.000£+00 0.000£+00
O.OOOE+Oo 0.000£+00
0.000£+00 0.000£+00
------------11------- lOSSES (l/day) -------------------------------1

last cell transform reaction + transfer - total
O.OOOE+OO O.OOOE+OO O.oOOE+OO 1.752E+02 1.752£+02
0.000£+00 O.OOOE+OO 0.000£+00 5.350E-Ol 5.350£-01
O.OOOE+OO 0.000£+00 0.000£+00 3.523E-04 3. 523E-04
0.000£+00 0.000£+00 0.000£+00 1.502E-05 1.502£-05
0.000£+00 O.OOOE+OO O.OOOE+OO 3.570E-09 3.570£-09
0.000£+00 0.000£+00 0.000£+00 1.136E-09 1.136£-09
O.OOOE+OO 0.000£+00 0.000£+00 1.881E-Ol 1.881E-Ol
0.000£+00 0.000£+00 1.100E-04 1.399E-Ol 1.400E-Ol
ENVIRONMENTAL DISTRIBUTION FOR arsenic IN LANDSCAPE CELL 1
COMPARTMENT SOURCES INVENTORY CONCENTRTN CONCENTRTN CoNC£NTRTN FRACTIONAL FRACTI ON LoSS£S
 (mo lId) (moles) (mg/kg) totl (mg/L) vol. (mg/L) wtr INVENTORY REMOVAL (me lId!
AIR 0.000£+00 5.255E-19 3.280E-27 3.936E-30 7.784E-22 9.636E-28 3.855E-21 1. 056£-
PMAIR 0.000£+00 4.745E-Ol 5.779£-02 3.554£-12 O.OOOE+OO 8.701E-I0 3.481£-03 9.538£-
BIOTA O.OOOE+OO 4.310E+02 1.522£-03 1.522E-03 0.000£+00 7.903E-07 0.000£+00 0.000£+
SOILl 2. 740E+Ol 1. 852£+06 4.612£-02 8.670E-02 4.445£-05 3.396E-03 O.OOOE+OO 0.000£+
SOlL2 O.OOOE+OO 1.312E+08 4.913£-02 9.826E-02 4.445E-05 2.406E-Ol O.OOOE+OO 0.000£+
GVTR 0.000£+00 4.123£+08 4.793£-02 9.826E-02 4.445E-05 7.56oE-01 0.000£+00 0.000[+
SVTR 0.000£+00 3.71 O£ +03 5.451E-04 5.500£-04 4.375£-05 6.802E-06 9.773£-01 2.677£+
SQMT 0.000£+00 4. 792E+03 5.015E-02 8.526£-02 4.372E-05 8.787E-06 1.924£-02 5.271E-
TOTAlS 2.740£+01 5.453£+08     1.000E+00 1. OOOE +00 2.740£+
Figure B-7. (continued).      
    - 175 -   

-------
(WfIlICA( F'ROI'£RTl£S FOR ben,.n.
M:,(CUlAR W[ IGHT " 7.811[+01
,:~ DIfFUSION CONSUNT... ..980£-06 ""2/S
PA~TIT 1 ON COE rr I C 1£ NTS :
AIR/WATER.... l.3llE-01 KG/l
SOllUWATER... 9.600£-02
SEDHT /SFWTR. .. 9.600£ -01
MEAT-FAT/DIET. ..000E-02
rJSH/WATER. . . . 7. 500E+Ol
HENRY'S LAW CONSTANT. . . .100E+03 T~R/(MOlE/l)
WATER OIFFUSION CONSTANT... ..980£-10...l/S
SOILI/WATER.. .
ROCK/GROWTR. . .
BIOTA/SOlll.. .
"ILK-FAT/DIET.
5.760E-Ol
1.378£-02
8.6Bl£+00
..000£-02
tALCUlATION OF RATE CONSTANTS FOR benzene
ALPHA (l/day) + FLOW (kg/day) x BE TAL (l/kg) + SFLOW (kg/day) x BETAS (l/kg) - T(I->J) (l/day)
4.266E-02 8.767E+09 ..307E-16 O.OOOE+OO O.OooE+OO 4.267E-02 .
7.BOOE-OB 8.219E+07 4.307E-16 O.OOOE+OO O.OOOE+OO 1.134E-07
2.010E-Ol 0.000£+00 ..307E-16 0.000£+00 0.000£+00 2.010E-Ol
0.000£+00 0.000£+00 0.000£+00 2.054E+05 1.626E-06 3.340£-01
O.OOOE+OO 0.000£+00 0.000£+00 1.236E+05 1.626E-06 2.010E-Ol
O.OOOE+OO O.OOOE+OO 0.000£+00 2.466£+06 1.429E-I0 3.523E-04
4.913E+Ol 5. 151E+09 O.OOOE+OO O.OOOE+OO 2.857E-13 4.913E+Ol
O.OOOE+OO O.OOOE+OO 4.959E-13 3.290E+05 2.B57E-13 9.399E-OB
O.OOOE+OO O.OOOE+OO 4.959E-13 2.466E+06 2.886E-12 7. 115E-06
O.OOOE+OO 7.890E+08 4.959E-13 O.OOOE+OO 2.857E-13 3.913E-04
O:OOOE+OO 3.014E+09 4.959E-13 3.260E+07 2.857E-13 1.504E-03
O.OOOE+OO 7.890E+08 1.438E-14 O.OOOE+OO 1.381E-15 1. 135E-05
O.OOOE+OO 1.863E+OB 8.521E-15 O.OOOE+OO 1.174E-16 1.587E-06
O.OOOE+OO 6.027E+08 8.521E-15 O.OOOE+OO 1.174E-16 5. 136E-06
3.554E-04 4.384E+07 O.OOOE+OO O.OOOE+OO 1.884E-12 3.554E-04
3.555E-04 O.OOOE+OO 1.963E-12 8. 835E+08 1.884E-12 2.020E-03
O.OOOE+OO 3.452E+09 1.963E-12 3. 260E+07 1.884E-12 6.837£-03
2.624E-02 O.OOOE+OO 1.448E-I0 8. 835E+08 1.390E-I0 1.491E-Ol
AIR -> SOILI
AIR -> SWTR
AIR a> NXTC
PHAIR-> SOl L1
PAAIRa> NXTC
BIOTAa> SOIL1
SOILl-> AIR
SDILl-> PHAIR
SDlll-> BIOTA
SOlll-> 50lL2
SOllla:> SWTR
SOIL2-> GWTR
GIITR a:> SO I L1
GllTR -:0 SWTR
SIITR -:0 AIR
SIITR -:0 SOMT.
SIITR -> NXTC
SDHT -> SWTR
COMPARTMENT
AIR
PMAIR
BIOTA
SOILl
SOIL2
61lTR
SIlTR
SDHT
1------- SOURCES (mol/day)
direct transform
0.000£+00 0.000£+00
O.OOOE+OO O.OOOE+OO
O.OOOE+OO 0.000£+00
2.740E+Ol O.OOOE+OO
O.OOOE+OO O.OOOE+OO
O.OOOE+OO 0.000£+00
O.OOOE+OO O.OOOE+OO
O.OOOE+OO 0.000£+00
------------11------- LOSSES (l/day) -------------------------------1
last cell transform reaction + transfer - total
O.OOOE+OO O.OOOE+OO 2.080E-02 2.437E-Ol 2.645E-Ol
O.OOOE+OO O.OOOE+OO 2.080E-02 5.350E-Ol 5.558E-Ol
O.OOOE+OO O.OOOE+OO O.OOOE+OO 3.523E-04 3.523E-04
O.OOOE+OO O.OOOE+OO O.OOOE+OO 4.913£+01 4.913E+Ol
O.OOOE+OO O.OOOE+OO O.OOOE+OO 1. 135E-05 1. 135E-05
0.000£+00 O.OOOE+OO O.ooOE+OO 6.723E-06 6.723E-06
O.OOOE+OO O.OOOE+OO O.OOOE+OO 9.212E-03 9.212E-03
O.OOOE+OO O.OOOE+OO 1.100E-04 1.491E-Ol 1.492£-01
ENVIRONMENTAL DISTRIBUTION FOR benzene IN LANDSCAPE CELL 1    
COMPARTMENT SOURCES INVENTORY CONCENTRTN CONCENTRTN CONCENTRTN FRACTIONAL FRACTION LOSSES
 (mol/d) (moles) (mg/kg) totl (mg/L) vo 1. (mg/L) wtr INVENTORY REMOVAL (mol/d)
AIR O.OOOE+OO 1. 235E+02 8.041E-07 9.649E-I0 4.156E-09 6.641E-Ol 1.000E+00 2. 740E+Ol
PHAIR O.OOOE+OO 1. 124E -07 1.428E-08 8.784E-19 O.OOOE+OO 6.046E-I0 9.103E-l0 2.494E-OB
BIOTA O.OOOE+OO 1. 343E -02 4.947E-08 4.947E-OB O.OOOE+OO 7.222E-05 O.OOOE+OO O.OOOE+OO
SOILI 2.740E+Ol 6.650E-Ol 1.727E-08 3.247E-08 2. 576E-08 3.575E-03 O.OOOE+OO O.OOOE+OO
SOIL2 O.OOOE+OO 2.293E+Ol 8.957E-09 1. 791E-08 2.576E-08 1. 233E-Ol O.OOOE+OO O.OOOE+OO
61lTR O.OOOE+OO 3.870E+Ol 4.693E-09 9.620E-09 2.576E-08 2.081E-Ol O.OOOE+OO O.OOOE+OO
SIlTR O.OOOE+OO 1.686E-Ol 2.5B4E-08 2.607E-08 2. 585E-08 9.064E-04 4.207E-05 1. 153E-03
SDMT O.OOOE+OO 2.283E-03 2.492E-08 4.236E-08 2.583E-08 1.227E-05 9.166E-09 2.511E-07
TOTALS 2. 740E+Ol 1. 860E +02    1. OOOE +00 1. OOOE +00 2.740£+01
Figure B-7.
(continued).
- 176 -

-------
, '
. .
..
AIR
9.649E-I0
mg/L
" ,
,',

.;.:.i .:.:
;:;: AIR PARTICLES ,:;:
8.784E-19

v
SURFACE WATER
2 . 584 E - 08 mg/ l
A
j~1 V I V jf:~
:::: LOWER SOIL SEDIMENTS ::::
8.957E-09 mg/kg 2.492E-08 mg/kg


ijWestern Ecoregion I
r OUTPUT FOR benzene VALUES IN BOXES ARE CONCENTRATIONS t
1IIIm~~lli~ili~ili1ili~~j~~iliiill~~i~~!~~jilli~~~!illt~1illj~~illiili~ill1ill1~~i~j~1~ji~i~~1jj~~m~ili~1ili!ili~ill~ilij~j1!jj~~~1illi~l~jj~1i~11ffi111i~~lli~~~1illM1~lm~~~1illjlli1jj~~imili~imm~ill~1ij~~ffi~;~~jllijl~~jj~~~:
mg/kg
mg/L
UPPER SOIL
1.727E-08
mg/kg
I
Y
GROUND-WATER ZONE
2.576E-08 mg/l
Figure B-8.
Chemical distribution of benzene as displayed by the graphic-outp~~re
routine for GEOTOX A.
177

-------
...
..",..
-,..
..
..
..; .

"'S.7E~07., ..'.
..
..
..
.,'.

I ;~j
4.6E-07



'.'.'.',',',
',',',',',',
:.:.:.:.:.:.

.. .. .. j:~::~.~.~.'.~.~,~:::.~.j:j::.;..1,~."
i
...... 111111111111
!::.i.,.i:!:.I:.!.::!:!:.!:!.:i:.:.~ . . . . . .
~ 1. 1 E -1) 7 1:.i::.~.~:!..:I:.~.::~.~...1:.1.:1,::.1:1:',.. ::::
I mmmm. '0'
!~,,',i'r:i:::. mmmm ' . . . . . i:::~:.:::~.;:,,:,!.::::'
~ ~ i:,~.;,~.~:~:!:~:;.~:~:!:.~:~,::~:
~~~~;~;m;~
i O.OE+OO Inhalation Ingestion Dermal Total 1!j!

~iWtW1iMii;glNt4t.iti.irijiI.4~Wt~t~%tl:J
3.4E-07
2.3E-07
Figure B-9.
Total exposure to benzene for the
PC screen by the program EXPOSE.
sample problem as displayed on the
178

-------
. . . . .
3"7E~09
..
..
..
.,
,'.
~ 3.0E-09 ~


I Z.ZE-09 I



:~. .):









I ::::~:o I I I I I


f? Water Milk Meat Gra-in Fruits & Fish Soil Total ~~~:
f Vegetables ~ I
~i INGESTION EXPOSURES mg/kg-d FOR benzene : child ~f~1: adult . lifetime f~j:

!!!lt1ili~ili~ili~ill~~~~~ili~j[{j~~~m~ill~j~j~j~~~~illj~~ill~illi~ill~~~~1jjill~~~illjill~~~lill~~~~~fi1~j~ilif1illjj~ili~f~j~~1~~~~~~jillj~1~~ill1~~m~ill~illji~~1i~illjill1!1jjlli1~1;lill!~~~jfmllim1~1~~1ill~1m~ill1ill1~~illjllij!ij~m1!1!:
Figure B-10.
Ingestion exposure to benzene from the sample problem as
on the PC screen by the program EXPOSE.
displayed
179

-------
{
;1~
)
~

~;.
I ~

~ For each compound, enter the reference safe dose (RSD) (in mg/kg-d) upon ::::
~ which the relative ranking will be based. For example, for a carcinogen, :~~:
",' the reference safe dose corresponding to a lifetime risk of 1 in a million :.::.;.';.:;.:.
is RSD . 1/{10A6 x q) where q is the potency in kg-d/mg.
~ ~~~

I :~~a=1:gt:=e~~;:1::;:en=~::~~ entries, push PgDn to see the figure I


~ To compare relative exposure enter a '1' in each box. ::':

I I
:J..~.~.~. ..~.~.~.9.~.~. . ..JR:ll~:~:.J9:~IA:~1~A:~1:.:.:... .~.c:tP.~.. ..J~t:.:~gq~:.:.:M:~;~;~:.:.:.:.:.:... .~.9.~.~. . ..J9K:.;~:~lR:.:~I~V.}~j
~1~m~m~I~~;~;~;~mm~E~~;~~;mmmmmm~~mmmmmm~~mm~~m~mmm~~m~m~~m~mmmE~~;EmmmEmmmm~~~mmEm~m~~E;~;~;E~~;~~;;~~~;~~~;~~~;~~E~E;~;E~E;~~;~~;E;E;E;~;~~~;~~E~~;~:;:::;:~:~;~~~~;~;~;~;E;~;~;~;~;~;Emmm~H;~EHn;m~mm~~~mEm~~m~mm~
1.000E+OO:arsen1c
1.000E+OO:benzene
Figure .B-ll.
Display used to compare the contaminants based 0n exposure.
- 180 -

-------
..
6.700E-08:arsenic
1.OOOE-03:benzene
For each compound, enter the reference safe dose (RSD) (in mg/kg-d) upon
whtch the relative ranking will be based. For example, for a carcinogen
the reference safe dose corresponding to a lifetime risk of 1 in a million
is RSD . 1/(lOA6 x q) where q ;s the potency in kg-d/mg.
When all the boxes have nonzero entries, push PgDn to see the figure
r comparing their relative ranks. I
~ i


l~;i~¥~~~;:t;~~;~~~;~i~;:~I~~1~;~i~~1i1f,f~ignl~~if1:lt~tlrJ
Figure B-12.
Display used to compare the contaminants based on potential risk,
- 181 -

-------
..
::~
..
i
t ~~~~
-4.0















. benzene 0
i RELATIVE RANKS FOR THE CHEMICAL SPECIES IN THE Western Ecoregion :m
*! RETURN ....~.~.n..J.Q....-P.R~Yl.Q.~.~..J".~N~.... E SC ....~~-lI...J.9....~.~.N..J1~.NV....L...IQ.....P.Rl.NJ...........................................J:
-5.
-6.0
-7.0
-8.0
arsenic
Figure B-13.
Display comparing exposure between the
two contaminants.
132

-------
~tj~\t:::)?:Y{m?;{\\:::::Y::Y::<\/<>::::C::i:;:7t\::::::))::~{:::::U::::?)::::/\HH::<{::)<:\:::U)):?6?/H:::/\/UU)H}{t/:::;:Hrr~:nmw~m~1

-:::: ... :~

~ i
I 1
~ ~
. ~
r;:s. 2.2 if:
~~ .~

~ I
~ ~
I 0.4 ~
i j
.~: :~~
~~ ~





:,:: - 5 . 0 :~:
~ arsenic ~
~ ~
:::: benzene :~::
1 RELATIVE RANKS FOR THE CHEMICAL SPECIES IN THE Western Ecoregion ~
~~~ RETURN ..J.~lI....IQ.....P.R~.Yl9.~.~J"~N~.... E SC _..~.~lI...IQ.....MAl~:.:.MI~:Y:.:.. . .~. . ..JP:.:X~lNl::::::::::::::::::::::::::::::::::::;.:A
mm~mmmmm~mmmmm~mmmmm~m~mmmmm~mmmmm~m~m~m~m~mmm~mmm~mmwmm~m~i~~~~i~~~~~~~~~~~~~~tmm~mmmmm~mmmmm~m~nli~mmmmmmm~m~mmmmm~m~mmm~~um~~~mmm~~~~m~~;~H;~~j;~;~~~:
-1.4
-3.2
Figure B-14.
Disp1~y comparing potential risks between the
two contaminants.
183

-------
Benzene concentrations from GEOTOX B
 lJ 
 1.0 
 0.9 
 O~ 
I -.. 0:' 
i ~~ OJJ 
. ~ IU 
,~  
"..  
h 015 
~~ 
~" 0.4 
 O.~ 
 0.2 
 OJ 
 0.0 
 0
 o Soli
All conc~nt,.atlons In ppm except aIr.
200
-fOO
+
Time In d~s
AIr mg/dL
o
Srf. Water
ngure B-15.
Output from GEOTOX plotted using Lotus l23(R).
- 184 -

-------
~
APPENDIX C
HODEL REVIEWS - OCEAN
- 185 -

-------
IDENTIFICATION
1.1
Name:
Bryan-Semtner-Cox (BSC) Model
1.2
Prepared by:
Geophysical Fluid Dynamics Laboratory, Princeton
1.3
Prepared for:
NOAA
1.4 Report Title:
A Primitive Equation,
Ocean
3-Dimensional Model of the
1.5
Report Number:
GFDL Ocean Group Tech. Rep. No.1
1.6
Report Date:
1984
1.7 Availability:
Available from:
GFDL, Princeton, NJ
1.8
Purpose and Scope:
A 3-dimensional numerical ocean model designed for
studying the most basic aspects of large-scale,
baroclinic ocean circulation. Highly dissipative
and slow running, it was intended to simulate the
upper ocean circulation, not the deep flow. Does
not accurately handle the vertical circulation or
the heat-driven flow.
2
SUMMARY OF FINDINGS
The Bryan-Semtner-Cox (BSC) model (M.D. Cox, "A Primitive Equation 3-
Dimensional Model of the Ocean," GFDL Ocean Group Technical Report No.1,
1984) is acclaimed as "the principal tool for modeling ocean circulation in
irregular domains having realistic coastlines and bottom topography" (A.J.
Semtner and R.M. Chervin, "A Simulation of the Global Ocean Circulation with
Resolved Eddies," J. Geophys. Res., 93, 15502-15522, 1988). However, the BSC
has been used by the English (MAFF) , the French (CEA) , and the German (DHI)
scientists in connnection with the SWG-POTG project, and all but the Germans
dropped it in favor of other models, and the others questioned the validity
of the German results.
Sandia has used the BSC extensively, including a 40-year simulation to
spin up the North Atlantic circulation. Sandia noted that the BSC requires
excessive dissipation parameters, beyond the levels reasonable for ocean
flows, and that this has negative effects on the results. With 200 krn
horizontal resolution, the BSC was found to be unstable when eddy diffusion
coefficients less than 109 cm2/sec were used. SOMS is stable with 106
cm2/sec eddy diffusion coefficients and 200 krn horizontal resolution. We
have also noted that SOMS runs more than ten times faster (per model year)
using the same resolution. -
Recently, the BSC was applied in a 20-year integration of the world
ocean, using 1/2 degree resolution and 20 layers. This required 500 hours of
dedicated time on the Cray-XMP/48 computer, with all vector processors
operating nearly all the time (99.64%). SOMS would require only 150 hours
cpu time on a much slower Cray 1 computer to do a similar calculation, using
- 186 -

-------
a 90 minute time step. This translates to a factor of ten savings based on
the relative speeds of the two machines.
The BSC does not address boundary layer phenomena, which are important
in environmental questions, while SOMS includes a fully three-dimensional
submode1 specifically for addressing the bottom boundary layer. It is fully
coupled to the "free-stream" submodel, which models the overlying ocean.
Swiss scientists (Nyffeler and Zuur) are presently using SOMS to model
the circulation of Lake ': lchate1 (Zuur and Dietrich, 1989), and high
resolution open-ocean simula~ions in the region of the NOAMP 5 cruise. SOMS
is also being used at Florida State Univesity by Weatherly, in studying flow
around a seamount and by Ezer, in his Ph. D. study of continental shelf
phenomena. This is noteworthy. as SOMS was first made available to others
only one year ago.
In summary, SOMS is far more capable than the BSC in addressing the
important waste disposal equations, due to the BSC's large numerical
dissipation, high computing cost, inaccurate vertical velocity fields, and
lack of a good bottom boundary layer turbulence closure model.
3
ADMINISTRATIVE CRITERIA
3.1
Documentation
The BSC is thoroughly documented in the user's manual. A large number
of papers have been published in the literature using this model.
3.2
Hardware Requirements
The BSC is best run on main frame computers.
3.3
Application
The BSC is applicable to the whole range of hydrostatic, incompressible
geoph~sical flows, except boundary layer phenomena.
3.4
Level of Expertise Required
A working knowledge of Fortran, understanding of basic numerical
modeling concepts, and understanding of modeling requirem~nts of geophysical
flows are all highly recommended.
4
TECHNICAL CRITERIA
4.1
P~er Review
The BSC is a well-established model that has been carefully reviewed and
analyzed by several prominent ocean modelers.
- 187 -

-------
4.2
Verification
The BSC has not been shown to converge (at least in the literature) as
resolution increases. Such demonstration is a necessary part of
verification, and has been done for SOMS. SOMS converges for a realistic
geophysical prototype problem with baroclinic instability in a stratified
environment (Dietrich, Roache, and Marietta, submitted for publication).
Even vertical velocity converges to reasonable values, which the NEA POTG
doubted would happen with the BSC using practicable resolution, in view of
the unrealistic results obtained with 200 km resolution. Vertical
overturnings are very important in issues involving ocean circulations.
4.3
Uncertainty
All models of time-dependent flow are uncertain to some degree.
Accuracy depends on many things, initial and boundary conditions, model
resolution, model assumptions, duration of the calculation, and the nature of
the feature to be determined. However, due to its large required numerical
dissipation, the BSC does not achieve much higher certainty or accuracy
except by using very high resolution and .associated high computing cost,
especially compared to SOMS, which runs much faster for a given resolution
and requires much less numerical dissipation. Further, SOMS' special
treatment of the dominant Coriolis and vertical diffusion terms is much more
accurate than the treatment used by the BSC.
4.4
Sensitivity
The sensitivity of the BSC to important model resolution, dissipation,
and thermodynamic driving parameters has not been addressed in detail in the
literature, in spite of its maturity.
4.5
Required Input
The required input is the usual geophysical modeling inputs.
4.6
Outputs
Contour plots and a number of numerical diagnostics are included in the
output.
4.7 Source Term
 NA. 
4.8 Scenarios
NA.
4.9
Relationship to Regulatory Standards
NA.
- 188 -

-------
5
SCIENTIFIC CRITERIA
5.1
Theory
The theory behind the BSC's basic modeling approach is a control-volume
concept based on finite-volume conservation laws.
5.2
Validation
Validation of the BSC has been slow, in spite of its maturity, due to
its large required numerical dissipation, which is not characteristic of
geophysical flows. Recently, a partial validation (at a'cost of hundreds of
thousands of dollars in computer time) was made by applying it to the world
ocean (A.J. Semtner and R.M. Chervin, "A Simulation of the Global Ocean
Circulation with Resolved Eddies," J. Geophys. Res., 93, 15502-15522, 1988).
Several numerical artifices were applied, including biharmonic diffusion and
direct forcing of all but a relatively thin layer around the thermocline
depth toward climatological values, and the dissipation was relaxed toward
more realistic values (but still too large) only during the last two years of
the simulation. Thus, this expensive calculation, although encouraging due
to its exhibiting many observed features of the world ocean circulation, is
at best -)artial validation. In view of the artificial numerical
constrainL it would have been disappointing if the results had not
exhibited many features of the real ocean.
5.3
Treatment of Radioactive Decay Products
Not addressed by the BSC, but could be added.
5.4
Underlying Assumptions
The BSC uses the hydrostatic and Boussinesq approximations.
5.5
~athways
NA.
5.6
Dose Converstion and Dose Response
NA.
- 189 -

-------
1.1
Name:
1.2
Prepared by:
1.3
Prepared for:
1.4 Report Title:
1.5 Report Number:
1.7 Purpose and Scope:
IDENTIFICATION
Sandia Ocean Modeling System (SOMS)
Sandia National Laboratories
U.S. DOE Subseabed
Laboratories
Sandia
National
Program,
Dietrich, D., Marietta, M.G., and P. Roache (1987).
An ocean modeling system with turbulent boundary
layers and topography: numerical description.
Int. J. Numer. Methods in Fluids 7:833-855, 1987.
Multi-scale dynamical ocean circulation code. High-
resolution eddy-resolving basin scale with realistic
topography and attached fully turbulent boundary
layers. Options include very high resolution for
local open-ocean domains, separate open boundary
layer model, and boundary fitted coordinates for
basin model. Code was written for simulating deep
ocean dispersal for waste disposal. A program is
used to calculate interbox transports for driving
the Mark-A box model with SOMS simulations.
2
SUMMARY OF FINDINGS
SOMS was initially developed for the U.S. DOE, under the Subseabed Waste
Disposal Program, to help evaluate deep ocean circulations. SOMS is designed
specifically to address bottom boundary layer flows over realistic
topography, but is applicable to the whole range of geophysical flows, from
small lakes to large oceans. SOMS has been applied to both high level
subseabed (M.G. Marietta and W.F. Simmons, 1988, NEA Final Report) and low
level sea dumping. (Nyffeler and W.F. Simmons, NEA Report, 1989). SOMS has
a~so been applied to the circulation of Lake Neuchatel (E. Zuur and D.E.
Dietrich, 1989) and the North Atlantic Ocean (D.E. Dietrich, M.G. Marietta,
and P.J. Roache, to be published), continental shelf phenomena (T. Ezer,
personal communications, 1989), and flow around a seamount (G. Weatherly,
personal communications, 1989).
Modeling the deep ocean requires much more carefully designed numerical
approaches than are required to model shelf phenomena. Shelf models are not
very sensitive to the numerical methods used due to the strong local
turbulence and associated large local dissipation rates. On the other hand,
deep ocean flows have extremely low real dissipation, and are dominated by
eddies that are 10 to 100 times smaller than the real ocean scale, yet the
eddy-scale based turbulent Reynolds number is very large. No turbulence
closure model has successfully addressed their role in the ocean circulation
(A.J. Semtner and R.M. Chervin, "A Simulation of the Global Ocean Circulation
with Resolved Eddies," J. Geophys. Res., 93, 15502-15522, 1988). Thus, these
eddies must be addressed directly, using the full (hydrostatic) Navier Stokes
equations. It follows that numerical models must have very low total
- 190 -

-------
dissipation on a scale of 50 kIn or less in order to model ocean flows
realistically. This either requires impractically high resolution (10 km or
smaller grid intervals) or using numerical methods that are robust for large
true cell Reynolds numbers.
To achieve the small numerical dissipation required to model deep ocean
circulations, the numerical method used must be fully conservative. This
translated to requiring that incompressibility be satisfied for even the
smallest resolved grid elements, which requires use of an Arakawa "c"
staggered grid (A. Arakawa and V.R. Lamb, "Computational Design of the Basic
DYnamical Processes of the UCLA General Circulation Model," Methods in
Computational Physics, Vol. 17, Academic Press, 174-265, 1977) for certain
critical parts of the calculation. Deep ocean circulation models must also
include full hydrostatic dynamics, including consistent determination of the
barotropic mode. Proper determination of the barotropic mode requires a two-
dimensional elliptic equation, either for the barotropic stream function as
done by the BSC model - - see below) or a top surface pressure, as done by
SOMS. Doing these things right is a non-trivial task that is not required in
modeling shelf phenomena.
The Bryan-Semtner-Cox (BSC) model is "the principal tool for modeling
ocean circulation in irregular domains having realistic coastlines and bottom
topography" (A.J. Semtner and R.M. Chervin, "A Simulation of the Global Ocean
Circulation with Resolved Eddies," J. Geophys. Res., 93, 15502-15522, 1988).
The difficulty of addressing deep ocean phenomena is clearly demonstrated by
recent application of this model to the world ocean circulation. This
required using on the order of 500 hours of dedicated time on a Cray-XMP/48,
with a fully vectorized code and all processors in use almost all of the time
(99.64%) . From our experience with the BSC, and also according to data
reported by Semtner and Chervin, SOMS is at least ten times faster (as noted
by D.E. Dietrich, M.G. Marietta, and P.J. Roache, "An Ocean Modelling System
with Turbulent Boundary Layers and Topography: Numerical Description,"
International Journal for Numerical Methods in Fluids, 7, 833-8555, 1987),
and requires much less numerical dissipation than the BSC.
SOMS' low numerical dissipation, and the need to have such low numerical
dissipation in order to realistic~lly address deep ocean circulations, is
clearly demonstrated in convergence studies recently performed with SOMS
(D.E. Dietrich, P.J. Roache, and M.G. Marietta, submitted for publication).
In this paper, SOMS i~ applied to a prototype geophysical problem with
baroclinic instability in a statically stable environment characteristic of
intense frontal zones near western boundary current separation. Also in this
paper, a new, more accurate treatment of these dominant terms, including the
treatment used by the present world standard BSC, cannot match SOMS' high
accuracy and corresponding low numerical dissipation. Finally, this paper
shows that SOMS' low numerical dissipation leads to features not
realistically addressed by previous models, such as secondary barotropic
gyres, and clearly reveals the strong influence of the aforementioned
thermodynamically driven eddies, which are the primary drivers of the general
circulation of the deep ocean.
. In summary, SOMS: is the only deep. ocean model designed to address in
detail the dynamics of boundary layers over realistic topography; is far less
dissipative, more accurate, and runs ten times faster for a given resolution
- 191 -

-------
than the world standard BSC model; and is the first ocean model to clearly
demonstrate convergence in a realistic geophysical prototype problem (at
least, no such demonstration exists in the literature), which is a
fundamental requirement in verifying a model.
3
ADMINISTRATIVE CRITERIA
3.1
Documentation
SOMS is thoroughly documented in a standard Sandia National Laboratories
users/ manual (D.E. Dietrich, M.G. Marietta, and P.J. Roache, 1989, that has
just been completed.
3.2
Hardware Requirements
Depending on application, SOMS makes effective use of computers ranging
in power from small workstation computers to super computers. For example,
to model 30 days real time in a 600 km X 480 km rectangular region that is
3750 meters deep, under conditions typical of the North Atlantic Ocean, with
20 km horizontal resolution and 7 layers in the vertical, SOMS requires about
3 hours cpu using a 30 minute time step on an Everex Step 386/20 computer
with an 80387 co-processor (retail price about $7,500). For a 1/2 degree
resolution, 20 layer global ocean calculation, SOMS would use a 90 minute
time step, which would require an estimated 150 hours on a Cray 1 computer
(based on calculations performed at Sandia National Laboratories) compared to
the 500 hours required by the BSC on the much faster Cray-XMP/48 computer
(see Section 2 above). This translated to SOMS running ten times faster than
the world standard model, using the same resolution. As noted above (see
Section 2), SOMS is also much more accurate and has much less numerical
dissipation.
3.3
Application
SOMS is applicable to the whole range of hydrostatic, incompressible
geophysical flows, from boundary layer and lake scales to large ocean scales
(see Section 2 above).
3.4
Level of Expertise Required
A working knowledge of FORTRAN, understanding of basic numerical
modeling concepts such as numerical stability limitations, and understanding
of modeling requirements of geophysical flows are all highly recommended
before any attempt to use any general geophysical model such as SOMS. With
such background, the user will find that SOMS is well documented and
reasonably easy to learn to use and, in using SOMS, will find that SOMS is
robust and accurate due to its good numerical design. Learning SOMS is
facilitated by several built-in sample problems, including the circulation of
Lake Neuchatel and the circulation of the North Atlantic Ocean, both using
real topography data sets. Past experience indicates that an excellent
graduate/post doctorate student can learn to use SOMS in about six months.
- 192 -

-------
4
TECHNICAL CRITERIA
4.1
Peer Review
In view of its having been made available only recently, SOMS' varied
applications in international university and industrial environments, noted
in Section 2 above, provided a detailed code review. Users include (see
Section 2 above): Professor Weatherly. Florida State Univesity; Professor
Nyffeler, at the University of Neuchatel in Switzerland; and Dietrich,
Marietta, Roache, and Steinberg in Albuquerque (Roache and Steinberg are
working on implementing an advanced boundary-fitted coordinate scheme into
SOMS).
4.2
Verification
As noted in Section 2 above, SOMS is the first model that has been shown
to converge (as resolution increases) in a geophysical prototype problem,
with baroclinic instability in a statically stable environment, which is a
necessary part of model verification. Other verifications include:
robustness under extreme modeling conditions (low dissipation); symmetry
preservation in symmetric problems; and global conservation checks.
4.3
Uncertainty
All models of time-dependent flow are uncertain to some degree.
Accuracy depends on many things, including initial conditions, resolution,
model assumptions, duration of calculation, the nature of the feature to be
determined (for example, whether it be the temperature at some point in time,
or a long term average temperature in some region makes a great difference in
model certainty), and when applicable, turbulence closure. How to carry out
uncertainty analyses with these models and related box models for
radionuclide transport is still an unsolved problem, but a reasonable
approach was outlined by the POTG/SWG.
Model sensitivity to resolution and dissipation parameters is discussed
in great detail by Dietrich, Roache, and Marietta (1989).
4.4
Required Input
The required input includes the usual geophysical modeling inputs, plus
if the optional turbulence closure scheme is used, input data for the Mellor-
Yamada level 2.5 turbulence closure scheme (Mellor and Yamada, 1982), which
is used to determine vertical eddy diffusivities. Initial conditions for
velocity and density must be specified, although hydrostatic and geostrophic
relations can reduce the required initial data (initial pressure can be
determined diagnostically directly from velocity and density). Boundary
conditions for mass, momentum, and buoyancy flux must be specified at all
times during a calculation. Model dissipation parameters (horizontal
momentum and heat diffusivities, and nominal vertical diffusivities) must
als.o be specified. Model geometry is specified by an input depth array.
Non- homogeneous boundary conditions are all handled by a special subroutine
and are controlled by input flag parameters. All input parameters are
described in the SOMS users' manual (D.E. Dietrich, M.G. Marietta, and P.J.
Roache, 1989).
- 193 -

-------
4.5
Output
Contour plots as well as a number of numerical diagnostics are included
in the output. Horizontal and vertical cross-sections of many different
fields are contoured, and the amount of output is easily controlled by a few
input parameters.
4.6
Source Term
Sources of heat and momentum (wind driving) are included in the top
layer. Optional open boundaries and ports provide additional sources to
drive the flow.
4.7
Scenarios
Scenarios such as sources of species from waste disposal sites have been
studied (Marietta and Simmons, 1988; Nyffeler and Simmons, 1989); sources of
pollutant to Lake Neuchatel are currently under investigation (Zuur and
Dietrich, 1989). Dispersion of radionuclides has not been added to SOMS yet,
and must be reinforced by the companion box model.
5
SCIENTIFIC CRITERIA
5.1
Theory
The theory behind SOMS basic modeling approach is a control-volume
concept (P.J. Roache, Computational Fluid Dynamics, Hermosa Publishers) based
on finite-volume conservation laws.
Other theories
turbulence closure
integration schemes.
are associated
mode 1 and the
with the two-parameter transportive
application of implicit numerical
5.2
Validation
As noted in Section 2 above, SOMS has been successfully applied to real ocean
and lake flows, to continental shelf phenomena, and compares favorably with
both theory (such as wind driven Sverdrup flow and boundary layer effects)
and observations (including the circulation of Lake Neuchatel and western
boundary current measurements in real oceans). Other model verifications
include: accurate internal wave phase and group velocity; dominant eddy size
corresponds to theoretical most unstable wavelength; realistic quasi-
geostrophic flows; realistic turbulent boundary layer structures; a~d
realistic deep ocean vertical velocities (less than one millimeter per
second) .
5.3
Treatment of Radioactive Decay Products
NA.
- 194 -

-------
5.4
Underlying Assumptions
As with nearly all large-scale geophysical hydrodynamic models, SOMS
uses the hydrostatic and Boussinesq approximations; and SOMS ignores
secondary sphericity effects. It should be noted, however, that non-
hydrostatic effects can easily be added (as described by D.E. Dietrich, M.G.
Marietta, and P.J. Roache, 1987).
5.5
Pathways
NA.
5.6
Dose Conversion and Dose Response
NA.
- 195 -

-------
IDENTIFICATION
1.1
Name:
Holland Model
1.2
Prepared by:
National
Boulder.
Center
for Atmospheric Research
(NCAR),
1.3 Prepared for:
1.4 Report Title:
1.S Report Number:
NSF and ONR
None
None
1.6
Report Date:
None
1.7 Availability:
Not publicly available.
1.8 Purpose and Scope:
A 3-dimensional quasi-geostrophic model designed for
studying high-resolution basin-scale eddy- resolving
ocean circulations.
2
SUMMARY OF FINDINGS
The Holland model has had some noted success in modeling ocean flows
(W.J. Schmitz and W.R. Holland, "Observed and Modeled Mesoscale Variability
Near the Gulf Stream and Kuroshio Extension," J. Geophys. Res.. 91, 9624-
9638, 1986; W.R. Holland and W.J. Schmitz, "Zonal Penetration Scale of Model
Midlatitude Jets," J. Phys. Oceanogr., 1859-1875, December 1985). It is
limited to free-ocean calculations. Due to its vorticity-streamfunction
formulation, it is inapplicable to flow over realistic topography, which must
be addressed by primitive equations, or to turbulent boundary layer flows.
A natural question is therefore why not use the more general primitive
equations in the first place? It appears that the main motivation for using
the Holland model is that its numerical efficiency is aided by not having to
resolve internal waves in time. However, this is largely compensated by
needing to solve Poisson equations, one for each layer at each time step.
Primitive equation models (in particular, SOMS) are thus generally
competitive with the Holland model in terms of computational cost for
problems to which both are applicable, while having a much broader range of
applications. They can even avoid the need to resolve internal waves, by
using implicit numerical approaches whose overhead can be less costly than
the multiple Poisson solutions used by the Holland model.
3
ADMINISTRATIVE CRITERIA
3.1
Documentation
The Holland model is not publicly available.
- 196 -

-------
3.2
Hardware Requirements
The Holland model is best run on main frame computers, but can also be
run on mini-computers and supermicros.
3.3
Application
The Holland model is limited to free-ocean calculations with uniformly
stratified thermal structures. It is not applicable to regions of
significant topography, or to boundary layer phenomena. It is also not
applicable to flows in small bains where deviations from non-divergent
horizontal flow can be important.
3.4
Level of Expertise Required
A working knowledge of Fortran, understanding of basic numerical
modeling concepts, and understanding of modeling requirements of geophysical
flows are all highly recommended.
4
TECHNICAL CRITERIA
4.1
Peer Review
The Holland model is a well-established model, whose results have been
reviewed and analyzed by several ocean modelers. The code has not been used
or reviewed outside NCAR.
4.2
Verification
Important convergence tests have not been published.
4.3
Uncertainty
The Holland model has additional uncertainty compared
equations models '. This is associated with the horizontally
assumption and with the horizontal stratification assumption.
to primitive
non-divergent
4.4
Sensitivity
No sensitivity studies have been published to date.
4.5
Required Input
The Holland model requires pressure specification at some level, and
density everywhere for initial conditions. Buoyancy flux must be specified
at inflow boundary points.
4.6
Output
Contour plots and a number of numerical diagnostics are included in the
output.
- 197 -

-------
4.7
Source Term
Internal heat and momentum sources are not included.
4.8
Scenarios
NA.
4.9
Relationship to Regulatory Standards
NA.
5
SCIENTIFIC CRITERIA
5.1
Theory
The Holland model is based on quasi-geostrophic flow theory, and uses a
conservative difference form for the vorticity equation.
5.2
Validation
A limited validation experiment was performed against the measured
distribution of potential and eddy kinetic energy with reasonable results.
5.3
Treatment of Radioactive Decay Products
Not addressed, but could be added.
5.4
Underlying Assumptions
The. Holland model uses the quasi-geostrophic (more generally, quasi-non-
divergent). hydrostatic, and Boussinesq approximations.
5.5
Pathways
NA.
5.6
Dose Conversion and Dose Response
NA.
- 198 -

-------
1.1 N sme: 
1.2 Prepared by:
1.3 Prepared for:
1.4 Report Title:
1.5 Report Number:
1.6 Report Date:
1.7 Availability:
1.8 Purpose and Scope:
IDENTIFICATION
Harvard Model
Harvard University
ONR
A draft user's manual exists, but is not publicly
available.
None, but see 4.2
None, but see 4.2
Available from:
Harvard, Cambridge, MA
A baroclinic regional-scale eddy-resolving open-
ocean numerical model designed for studying quasi-
geostrophic ocean circulations. Theoretically, it
is not applicable around topography although bottom
bathymetry has been added through the bottom heat
distribution and seems to perform well for gentle
slopes. Model is part of the ocean prediction
forecasting system presently being put into
operational phase by the Navy.
2
SUMMARY OF FINDINGS
The Harvard model (Miller, R.N., A.R. Robinson, and D.B Haidvogel, "A
Baroclinic Quasigeostrophic Open Ocean Model," J. Compo Phys., 50, 38-70,
1983) is limited to free-ocean calculations with strongly stratified thermal
structures. Due to its vorticity-streamfunction formulation, it is
inapplicable to flow over realistic topography, which must be addressed by
primitive equations, or to turbulent boundary layer flows.
A natural question is therefore why not use the more general primitive
equations in the first place? It appears that the main motivation for using
the Harvard model is that its numerical efficiency is aided by not having to
resolve internal waves in time. However, this is largely compensated by
needing to solve Poisson equations, one for each layer each time step, and to
transform between real and normal mode space, which can dominate the
calculation if the vertical resolution is greater than 0(10) layers.
Primitive equation models (in particular, SOMS) are thus generally
competitive with the Harvard model in terms of computational cost for
problems to which both are applicable, while having a much broader range of
applications. They can even avoid the need to resolve internal waves, by
using implicit numerical approaches whose overhead can be less costly than
the multiple Poisson solutions and normal mode transformations used by the
Harvard model.
- 199 -

-------
The Harvard model has some nice theory behind it, and can be used as a
teaching and research tool in spite of its practical limitations (although
its value as a research tool is limited).
3
ADMINISTRATIVE CRITERIA
3.1
Documentation
The Harvard model is well doeumented by an internal user's manual.
is not publicly available.
It
3.2
Hardware Requirements
The Harvard model is best run on main frame computers, but can also be
run on mini-computers and supermicros.
3.3
Application
The Harvard model is limited to free-c;>cean calculations with strongly
stratified thermal structures. It is not applicable to regions of
significant topography. or to boundary layer phenomena. It is also not
applicable to flows in lakes where deviations from non-divergent horizontal
flow can be important.
3.4
Level of Expertise Required
A working knowledge of Fortran, understanding of basic numerical
modeling concepts, and understanding of modeling requirements of geophysical
flows are all highly recommended.
4
TECHNICAL CRITERIA
4.1
Peer Review
The Harvard model is a well-established model that has been reviewed and
analyzed by several ocean modelers. It is not considered to have much
practical applicability. but it is used for ocean forecasting by the Navy.
4.2
Verification
The Harvard model is one of the two ocean models which has undergone
extensive numerical and physical parameter studies (the other is SOMS).
These studies are well reported in both the scientific literature (Haidvogel
et al., 1980; Miller et al., 1983) and SDP reports (Marietta and Simmons,
1987).
Haidvogel, D.B., A.R. Robinson, and F.F. Schuman (1980). The accuracy,
efficiency, and stability of three ocean models with application to open
ocean problems. J. Compo Phys., 34:1-53.
Marietta, M.G. and W.F. Simmons (1987). SDP Annnual Report: Modeling
Studies. SAND86-0929. Sandia National Laboratories, Albuquerque, NM.
- 200 -

-------
Miller, R.N., A.R. Robinson, and D.B. Haidvoge1 (1983).
quasigeostrophic open ocean model. J. Compo Phys., 50:38-70.
A baroclinic
Marietta, M.G. and A.R. Robinson (1986). Status and Outlook of Ocean
Modeling, Research, Dispersion and Related Applications. SAND85-2806, Sandia
National Laboratories, Albuquerque, NM.
4.3
Uncertainty
The Harvard model has additional uncertainty compared to primitive
equations models. This is associated with the horizontally non-divergent
assumption and with the horizontal stratification assumption used in the
determination of normal modes.
4.4 Sensitivity
 See 5.2. 
4.5 Required Input
The Harvard model requires pressure specification at some level, and
density everywhere for initial conditions. Pressure must be specified at
some level for all time, which is a major limitation. Buoyancy flux must be
specified at inflow boundary points.
4.6
Output
Contour plots and a number of numerical diagnostics are included in the
output.
4.7
Source Term
Internal heat and momentum sources are not included.
4.8
Scenarios
NA.
4.9
Relationship to Regulatory Standards
5
SCIENTIFIC CRITERIA
5.1
Theory
The Harvard model is based on quasi-geostrophic flow theory. and uses a
conservative difference form for the vorticity equation.
5.2
Validation
The Harvard model was the first model used for ocean forecasting. Early
work was carried out in the California Current System with the Naval
Postgraduate School (Robinson et al., 1984). Ocean data from AXBTs and XBTs
were transmitted to shore and assimilated in near real time into model
- 2U 1 -

-------
simulations that forecast future fields. These forecast fields were compared
to later observed fields for accuracy. This work led to the development of
the Ocean Descriptive Prediction System (ODPS) for forecasting in real time
using shipboard computers (PDP-ll/73s and micro-VAX lIs). The first such at
sea forecasts were carried out on the Nares Abyssal Plain as a joint project
between Harvard and Sandia (Marietta and Simmons, 1987). Validation was only
partially successful. The first real at sea validation was performed by
Sandia using the Harvard model at the NOAMP site working jointly with the
DHI/FRT and Swiss/Neuchatel through the NEA/CRESP. SOMS was also used
successfully during this work (Kupferman et a1., 1986; Marietta and Simmons,
1988). The Harvard model is currently used by the Navy in the ODPS framework
for at sea forecasting for the fleet. The model's limitations are well known
and a number of questionable tricks have been employed to improve model
performance in flow regimes beyond the model's capability defined by
extensive sensitivity studies, e. g., empirically embedding flow features
observed in satellite pictures, including topography through the bottom heat
distribution instead of explicitly. Still this model has had many successes,
primarily because before SOMS there were no primitive equation models that.
performed well in real ocean validations.
Robinson et al., 1984. A real time forecast of ocean synoptic/mesoscale
eddies. Nature, 309(5971). 781-883.
Kupferman et al., 1986. An intense cold-core eddy in the North-East
Atlantic. Nature, 319(6053), 474-477.
5.3
Treatment of Radioactive Decay Products
Not addressed, but could be added.
5.4
Underlying Assumptions
The Harvard model uses the quasi-geostrophic (more generally. quasi-non-
divergent), hydrostatic, and Boussinesq approximations.
5.5
Pathways
NA.
- 202 -

-------
1.1
Name:
1.2
Prepared by:
1.3
Prepared for:
1.4
Report Title:
1.5
Report Number:
1.6
Report Data:
1.7
Availability:
1.8
Purpose and Scope:
3.1
Documentation
See 1. 4
3.2
Hardware Requirements
1
IDENTIFICATION
NRPB 91-Box Model (9IBXM)
National Radiological Protecti!'n Board
Oxfordshire and Ministry of Agriculture,
Fisheries, Lowes toft , U.K.
(NRPB) ,
Food and
NEA Site Suitability Review (SSR) of the Northeast
Atlantic Dumpsite
Review of the Continued Suitability of the Dumping
Site for Radioactive Waste in the North-East
Atlantic
NEA Report, no number
1985
Available from: National Radiological Protection
Board, Chilton, Didcot, Oxon OXII ORQ, England
9lBXM is a coarse-grid box model which calculates
radionuclide transport based u"- an ocean circulation
that was subjectively assembled from an extensive
literature review. It is not coupled with a
dynamical circulation model although like all box
models, it could be driven by simulated
circulations. 9lBXM has been used for risk
assessments of high-level waste disposal in the
subseabed and assessments of low-level sea dumping
at the North-East Atlantic NEA site.
3
ADMINISTRATIVE CRITERIA
9lBXM has been used on the Harwell CDC machines and the NRPB VAX 786.
It uses an ODE solver developed at Harwell which is available at a
significant cost.
3.3
Application
Code is not user friendly. Interbox transports must be obtained either
by a subsidiary calculation from a high resolution simulation of the basin
circulation or from expert descriptive oceanographers. Ocean data are sparse
so considerable subjective judgement is required. If simulated ocean
circulations are used, interbox transports must be hand calculated and
entered. Interbox transports for nested boxes which must be site-specific
- 203 -

-------
have to be obtained from subjective expert estimates.
new site requires considerable recoding.
Problem setup for a
3.4
Level of Expertise
Running the code is simple. Setting up new problems or performing
sensitivity analyses on physical oceanographic parameters wouldn't be worth
the effort. Box model codes should be simple. This one is not, and the code
could probably be run only by one of the authors.
4
TECHNICAL CRITERIA
4.1
Peer Review
As part of the review of the North-East Atlantic Dumpsite, NEA conducted
expert review workshops on the 9lBXM and data bases. Two review sessions
were held at NEA headquarters. The first group of experts rej ected this
model because of a theoretical problem. NRPB was asked to change their
modeling approach from one that relied on an inverse technique to one relying
on a subjectively determined circulation. As a result the coding of. this
model remains messy. Model performance using the subjective circulation was
judged by the second workshop as adequate for the suitability review although
problems still remain, as they must, since the driving circulation is based
on very sparse data in key regions of the ocean.
4.2
Verification
9lBXM was a participant
the NEA CRESP Modeling Task
comparison of models for the
deep ocean. NRPB-Rl94.
code in a model comparison study conducted by
Group. Mobbs et a!. (1986). A preliminary
dispersion of radionuclides released into the
4.3
Uncertainty
No uncertainty analyses have been carried out with 9lBXM although NRPB
has the LISA systems code for this purpose. Uncertainty analyses using a box
model with too many boxes would not be sensible. Consequently a simpler
version, MINIBOX, was developed for uncertainty analysis.
4.4
Sensitivity
Sensitivity studies on various pathway and physical parameters
performed for the NEA workshops and are discussed in the NEA SSR Report.
were
4.5
Input
lnterbox transports (an ocean circulation).
Geochemical parameters.
4.6
Output
Box concentrations
estimates.
of
radionuclides
and
maximum
individual
dose
- 204 -

-------
4.7
Source Term
release
location.
rates
of each radionuclide from bottom source at repository
4.8
Scenarios
Scenarios for ocean disp9sal have been defined and analyzed but require
different model setups.
4.9
Reg. Standards
There are no ocean standards
5
SCIENTIFIC CRITERIA
5.1
Theory
The compartment model covers the area of the Atlantic Ocean from 50 S to
65 N and uses observed isopycnal (density) surfaces to define the vertical
box structure, sin~e mixing and movement in the ocean are believed to occur
primarily along isopycnal surfaces. Exchanges with other oceans are also
included, since radionuclides entering the Atlantic will eventually disperse
throughout the world's oceans. These exchanges are with the Arctic Ocean
through the Norwegian and Greenland Seas to the north, with the Mediterranean.
Sea to the east and with the Pacii"ic and Indian Oceans, via the Antarctic
Circumpolar Current, to the south.
The Atlantic Ocean is divided into eight areas (Figure 6.37). The
central dividing line corresponds to the Mid-Atlantic Ridge and the areas
extend out to the edge of the continental shelf. Areas 1 and 3 are based on
areas chosen by Worthington (1976) on the basis of bottom topography. Area 2
is chosen to represent adequately the current flow patterns in the north.
The southern boundary of Area 4 (20 N) corresponds approximately to the
position of the Cape Verde Islands. Areas 5 and 6 cover the equatorial
waters and their southern boundaries (both at 25 S) and correspond to the Rio
Grande Rise and Walvis Ridge, respectively. Areas 7 and 8 cover the
remainder of the South Atlantic and extend to 50 S, which is approximately
the latitude of the Falkland Islands in the southwest. The present low-level
dump site is in Area 4, relatively close to the northern boundary.
It is possible to identify well-defined water masses at specific depths
throughout the Atlantic Ocean. Therefore, the eight ocean area were divided
into layers so that each compartment would represent a separate water mass.
The following water types were identified: (1) Antarctic Bottom Water
(levels 9 and 10, only in Areas 6, 7 and 8); (2) North Atlantic Deep Water
(levels 7 and 8); (3) North Atlantic Central Water (levels 5 and 6); (4) MODE
Water (levels 3 and 4); (5) surface water (levels 1 and 2). The boundaries
of the compartments were obtained by tracing the neutral (density) surfaces
through the ocean, using the method of Ivers (1975) and data taken from the
GEOSECS survey (Bainbridge, 1972) with additions from the International
Geophysical Year (IGY) surveys (Fuglister, 1960). Figures 6.38 and 6.39 show
these levels for the east and west Atlantic, respectively. Initially, sic
- 205 -

-------
levels were chosen in the vertical, but it was thought that a better
resolution of the water masses would be obtained using up to 10 levels. This
gives rise to 67 boxes in the Atlantic and a further 24 boxes (defined by the
same density levels) for the other oceans, which have much less spatial
resolution, to give a total of 91 boxes (Figure 6.40). Each box is allowed
to communicate with each of its neighbors, unless topography interferes,
resulting in 442 transfer coefficients to be determined. Instead of
describing these transfer coefficients in terms of an advective flow and a
diffusive mixing among boxes, they are given as a transfer from A to Band
another from B to A. The difference between the two transfers is the
advective flow, and a weighted sum of them can be interpreted as the eddy
diffusivity (Fiadeiro, 1975).
The model structure described above assumes that the oceans are z~na11y
well mixed and that it is possible to describe the isopycnal surfaces of the
oceans from meridional sections. To define these surfaces, the method of
Ivers (1975) was used in which the neutral surface is defined to be normal to
the gradient in potential density, referred to the pressure at the point in
question. Ivers used this definition to discuss the behavior of five neutral
surfaces in the North Atlantic. The shallowest outcrops in the Labrador Sea
and descends to 900 m at 30 N, and the deepest outcrops in the Norwegian-
Greenland Sea and descends to 3000 m at 40 N. Ivers found that most features
in salinity could be explained by lateral flow and mixing, except in regions
of sills and outcroppings, where vertical mixing becomes important.
The GEOSECS results (Bainbridge, 1972) represent a good quality data set
which provides meridional sections required for the model description, except
for the northeast Atlantic. In this area, the continuation of eastern
Atlantic GEOSECS section was compiled frem IGY data (Fuglister, 1960) and was
constructed to run as near as possible along the deep water of the northeast
basin and to the west Rockall Bank to the Norwegian Sea.
The neutral surfaces were constructed bv startin~ at the southern end of
each section with and values, and following
them north from station to station on the s~ctions.
These surfaces tilt steeply near the polar ends of each section, while
remaining almost horizontal in the deep central ocean. Some surfaces
intersect the seafloor, whereas others outcrop. The geographical position of
the outcropping varies from season to season (Levitus, 1982).
Similar methods were used to calculate the neutral surface depths for
the Arctic, Antarctic, and Pacific Oceans from the GEOSECS data. However,
data for the Mediterranean Sea are taken from Sankey (1973) and Wust (1961),
with some adjustments made to obtain consistent values for inflow and
outflow. Only two levels are used in the Mediterranean because the shallow
sill depth in ~he Straits of Gibralter effectively isolates any deeper layers
from communicating directly with the Atlantic. Similarly, the Denmark Strait
restricts communication between the Greenland Sea and Labrador basin. The
flow pattern described
-------
was used to calculate gradients needed to produce the transfers, but when one
box was much larger than the other, the smaller dimension was used.
5.2
Validation
Box models including the 92BXM have ~ot been shown to compare favorably
with any ocean tracer distribution. The NEA SSR panel of experts and the NEA
SYG POTG. Following this reversal the NRPBjMAFF joined the POT approach
using dynamical simulations for driving circulations in an attempt to
validate these ocean models against natural and introduced tracer
distributions. A total of eight different tracers were studied: salinity,
temperature C-12, Th-230, R-226, Th-232,R-228, and Th-228.
5.3
Treatment of radioactive decay products
If data are available, simplified chains could be handled.
5-.4
Underlying assumptions
See 5.1
5.5
Pathways
Pathways are the IAEA pathways described under Mark-A
5.6
Dose conversion and dose response
Dose is estimated to critical individuals and the collective
commitment. Dose response is not included. Requires code coupling.
dose
- 207 -

-------
Table C-l. NOMINAL DENSITIES OF
THE TEN LEVELS
Level
Density
1 (top)
2
3
4
5
6
7
8
9
10 (bottom)
26 <
26.75 <
27 <
27.3 <
45.5 <
45.8 <
45.9 <
46 <
<26
<26.75
<27
<27.3
< 45.5
<45.8
<45.9
<46
<46.1
<46.1
- 208 -

-------
Table C-2. YATER TRANSFERS BETWEEN
BOXES IN SVERDRUPS   
AH - 10000000. cm*cm/s and K-1. cm*cm/s
  Transfer
Box A Box B A to B B to A
1 2 .358 4.358
1 9 .060 3.060
1 16 8.581 .581
1 24 .091 .091
1 68 .025 1.025
2 3 .208 3.208
2 10 .587 .587
2 17 3.274 .374
2 25 .654 .654
2 69 .008 4.008
3 4 .362 1. 362
3 11 2.331 2.331
3 18 .772 .772
3 26 1.665 1. 665
-3 70 .081 2.081
4 5 4.066 4.066
4 12 .232 .232
4 19 .098 1. 098
4 27 .198 .198
4 71 .012 .012
5 6 9.701 7.701
5 13 .241 1.241
5 20 .108 1.108
5 28 .243 .243
5 72 .035 .035
6 7 6.739 4.739
6 14 .143 .143
6 21 .139 .139
6 29 .138 .138
7. 8 3.753 22.753
7 15 .112 .112
7 22 .193 3.193
7 30 24.171 .171
8 24 .085 19.085
8 31 .108 .108
9 10 .047 1.047
9 24 .258 1.258
9 68 .010 1.010
10 11 .0631 1. 063
10 25 1.313 .313
10 69 .038 1. 038
- 2.09 -

-------
1.1
Name:
1.2
Prepared by:
1.3
Prepared for:
1.4 Report Title:
1.5
Report Number:
1.6
Report Date:
1.7 Availability:
1.8
Purpose and Scope:
3.1
Documentation
1
IDENTIFICATION
TASC Box Model, MARINRAD
The Analytic Sciences Corporation
from Sandia National Laboratories
under
contract
U. S . DOE Subseabed
Laboratories
Program,
Sandia
National
User's Guide to MARINRAD: Model for Assessing the
Consequences of Release of Radioactive Material into
the Oceans.
SAND83-7104
1984
Available from:
Albuquerque, NM
National
Sandia
Laboratories,
MARINRAD is a coarse-grid box model which calculates
radionuclide transport based on an ocean circulation
that was subjectively assembled from historical
literature. It is not coupled with a dynamical
circulation model although like all box models, it
could be driven by simulated circulations. MARINRAD
has been used for risk assessments of high-level
waste disposal in the subseabed of both the Atlantic
and Pacific and assessments of low-level sea dumping
at the North-East Atlantic NEA site.
3
ADMINISTRATIVE CRITERIA
Koplik, C.M., M.F. Kaplan, J.T. Nalbandian, J.H. Simonson, and P.G.
Clark, (1984). User's Guide to MARINRAD. SAND83-7104. Sandia National
Laboratories, Albuquerque, NM.
Ensminger, D.A.. C.M. Koplik, and R.D. Klett, (1984). Preli1minary Pre-
Emplacement Safety Analysis of the Subseabd Disposal of High-Level Nuclear
Waste. SAND83-710S, Sandia National Laboratories, NM.
Kaplan, M.F., C.M. Koplik, and R.D. Klett, (1984). PreliJIlinary Post-
emplacement Saftey Analysis of the Subseabed Disposal of High-Level Nuclear
Waste. SAND83-7l06, Sandia National Laboratoxies, Albuquerque, NM.
Kaplan, M.F. (1984). Biological and Physical Sensitivity Analyses for
Subseabed Disposal of High-Level Waste. SAND83-7107, Sandia Natina1
Laboratories, Albuquerque, NM.
- 210 -

-------
3.2
Hardware Requirements
MARINRAD has been used on CDC and DEC VAX machines.
3.3
Application
Code is user friendly. Interbox transports must be entered as input
like the other box models, but there are many fewer boxes. Ocean data are
sparse so considerable subjective judgement is still required. If simulated
ocean circulations are used, interbox transports must be hand calculated and
entered. Interbox transports for nested boxes which must be site-spectific
have to be obtained from subj ective expert estimates. Problem setup for a
new site requires considerable work with the descriptive oceanographic
literature but little recoding.
3.4
Level of Expertise
Running the code is simple.
therefore, MARK A.
Setting up new problems
is not.
See,
4
TECHNICAL CRITERIA
4.1
Peer Review
MARINRAD has never been peer reviewed, but it was part of the NEA CRESP
comparison study (see 4.2 Verification).
4.2
Verification
MARINRAD was a participant code in a model comparison study conducted by
the NEA CRESP Modeling Task Group. Mobbs et a1. (1986). A preliminary
comparison of models for the dispersion of radionuclides released into the
deep ocean. Report No. NRPB-R194.
4.3
Uncertainty
No uncertainty analyses have been carried out
they could be if coupled with a dynamical circulation
a better candidate since it is a full basin model,
MARINRAD .
wi th MARINRAD al though
model. Mark-A would be
not a half basin like
4.4
Sensitivity
Sensitivity studies on various pathway and physical
completed (see documentation). The adjoint method was
sensitivity studies with MARINRAD.
parameters were
also used for
4.5
Input
Interbox transports (an ocean circulation).
Geochemical parameters.
- 211 -

-------
4.6
Output
Box concentrations
estimates.
of
radionuclides
and
maximum
individual
dose
4.7
Source Term
Release rates of each radionuclide from bottom source at repository
location.
4.8
Scenarios
Scenarios for ocean disposal have been
MARINRAD but require different model setups.
defined
and
analyzed
with
4.9
Reg. Standards
There are no ocean standards.
5
SCIENTIFIC CRITERIA
5.1
Geometry and Transports for the MPG-s Model
. ;
A hypothetical north-central Pacific (MPG-l) site is located about 10.
lan north of Hawaii in roughly 6000 m of water. The vertical
compartmentalization of the model is in three levels. A southward flowing
middepth water mass, 2000 m in thickness, is formed from a mixture of
northward flowing deep and abyssal waters, 1000 m in thickness, and weakly
southward flowing surface waters, 1000 m in thickness. A lO-cm deep
bioturbated sediment layer underlies all the deep boxes. Near-shore sediment
boxes with higher sedimentation rates underlie the North and South Pacific
surface boxes. They are taken to be 19.5% of the bottom area of their
associated water boxes.
The initially imposed water mass balance is derived from work on South
Pacific Ocean/Circumpolar Current exchanges and deep vertical South Pacific
advection rates. Specifically, 14.5 sv flow into and out of the South
Pacific from the Circumpolar Current. Of that, 12 Sv of Circumpolar Bottom
Water enter the western boundary undercurrent. South Pacific Bottom Water
advects vertically to the Deep Water at a rate of 8 Sv. South Pacific Deep
Water returns to the Circumpolar Current at a rate of 7.5 Sv and also advects
to the surface water at 3 Sv. The remaining circulation is adjusted to
conserve water mass as required by the above transports and to allow for
vertical advection at roughly 3 m/yr.
Within the Bottom North Pacific box is a set of nested process boxes.
The site region is taken as a 100 x 100 km square. It is overlain by a
turbulent bottom mixed layer (BML) of height H 100 m, taken to be the height
of the site box. The vertical eddy diffusivity in the BML is A 100 cm
/sec. Thus, the time scale for vertical mixing in the BML is t - H /4A 3
days. Using a bottom mean velocity of 3 cm/sec, the site box is flushed in
100 km/3 cm/sec 39 days. Thus, the flushing scale is much geater than the
mixing scale and therefore' a box model approach is acceptable.
- 212 -

-------
The eddy box which circumscribes the site box is taken to be 200 x 200
square so that, at most, four close-packed 100-km diameter deep eddies can
occupy it. This is roughly in keeping with observed eddy length scales. The
physical idea here is tht the BML disperses the contaminant to horizontal and
vertical scales which are large enough for the eddies to work on. They in
turn disperse the contaminant to still greater scales which are large enough
to be dispersed efficiently by the general circulation. This idea, although
correct in principle, is stretched to i ts limits of applicability by a 200 x
200 km eddy' mixing box. One would prefer to see at least a 5 x 5 and
preferably a 10 x 10 eddy box. This problem is addressed in the Mark A
model. The time required for a patch to grow to eddy size is roughly 20 days
(i.e., 100 km/5 cm/sec). At this time the patch is eddy-sized, but may be
strea~. It has diffu6eg vertically upward from the BML by an amount (4 x
0.6 cm /sec x 20 days)' 20 m, where the vertical eddy diffusivity in the
deep water A2 - 0.6 cm /sec. Thus, the height of the eddy box is taken as
20 m, i.e., the height of the BML plus the diffusive spreading height.
The next box in the nest is a basin-scale geographical box, whele m~xing
scales come from horizontal diffusivity of the order of AH 5 x 10 em /sec
and flushing scales from the general circulation, 1. e., the volume of the
boxes divided by th bulk flow rates through them. Large differences between
these times are difficult to establish, i.e., the ocean is not well-mixed.
Geometry and Transports for the NAP Site
A hypothetical western North Atlantic site is located about 1000 kIn
south of Bermuda on the Nares Abyssal Plain in roughly 5000 m of water.
Because deep east-west communication is essentially prohibited by the mid-
Atlantic Ridge, the eastern North Atlantic is not separately
compartmentalized but rather is grouped with Remaining Ocean Waters.
Vertical compartmentalization of the western North Atlantic is into two
layers divided roughly at the 40C isotherm, and corresponding to southward
flowing North Atlantic Dep Water and northward flowing North Atlantic Surface
Water. The lower boxes are roughly 2810 m in height. Sediment boxes are
treated as they were in the Pacific site.
The imposed water mass balance allows for a 3 m/yr vertical advection in
the North American and Guiana Basins and absorbs transports to and froi the
Norwegian Sea as an internal circulation in the Labrador Sea box. The
essential features are (i) a 6-Sv transport from the deep North American
Basin carried by the Western Boundary Undercurrent, (ii) a 3-Sv inflow to the
Deep Labrador Basin, (Hi) a 2.5 Sv inflow to the Surface Guiana Basin
(reduced from Worthington's 4-Sv value in order to allow for vertical
advection at 3 m/yr while retaining 4 Sv inflow to the Surface Larador
Basin). The other features are determined by (i) an assumed 3 m/yr vertical
advection and (ii) mass continuity box-by-box.
Process boxes in the Deep North American Basins are divided
geographically to account for the strong gyre circulatin induced by the Gulf
Stream to the north and the strong eddy field to t~e south. As in the
Pacific, the burial site is taken to be a 100 x 100 km area. Since the BM~
height is 50'm and the vertical boundary layer eddy diff~sivity is 10 em
/sec, the internal boundary layer mixing time becomes t - H /4A or 7 days.
- 213 -

-------
Typical boundry layer velocities at the site are 4 cm/sec, so the
flushing time is 100 km/4 cm/s or about 29 days, roughly four times the
mixing time. Thus, scale separation in the Atlantic model is not quite as
strong as in the Pacific model, but it is acceptable. The eddy box is again
a minimally-sized 200 x 200 km square. The scales for (i) mixing ,nd ~ii)
variability are taken as the time for horizontal diffusion (at AH 10 cm Is)
(i) from the site to the southern edge of the Gyre box and (ii) double that
distance. The times work out to 9.6 yrs and 29 yrs, respectively, which,
again, is close but acceptabale. The gyre box has a characteristic
circulation time of about 2.5 yr and a mixing time which is probably less
th,n a month, since the eddy field in this region is quite intense (AH 5 x
10 cm2/sec).
The. height of the eddy and gyre boxes is taken as the sum of the BML
height plus the advective and diffusive (Az 2 cm2/sec). heights acquired while
the contaminant spreads across the two boxes, i. e., during roughly 12 yr.
This works out to 50 m + 36 m + 390 m or 475 m.
Because AH - 5 x 106 cm2/sec in the Remaining Ocean Waters box, there
are three horizontal diffusivities in the model. In the numerical work, the
smaller diffusivity was. used when there was a choice. A brief sensitivity
analysis suggested small dependencies on this value. Sedimentation was
treated as in the Pacific.
5.2
Validation
Box models including MARINRAD have not been shown to compare favorably
with any ocean trracer distribution.
5.3
Treatment of Radioactive Decay Products
If data are available, simplified chains can be handled.
5.4
Underlying Assumptions
See 5.1
5.5
Pathways
Pathways are the IAEA pathways. described under Mark-A.
5.6
Dose Conversion and Dose Response
Dose
included.
is estimated to critical
Requires code coupling.
individuals.
Dose
response
is not
- 214 -

-------
1.1 Name:
1.2 Prepared By:
1.3 Preparet! For:
1.4 Report Title:
1.5
Report Number:
1.6
Report Date:
1.7 Availability:
1.8
Purpose and Scope:
3.1
Documentation
1
IDENTIFICATION
Mark-A Box Model
Physical Oceanography Task Gro\.1.p,
Group, NEA
Seabed Working
NEA
Research, Progress and the Mark A Box Model for
Physical, Biological and Chemical Transports
SAND84-0646
1985
Available from: Eco1es des Mines, Fountaineb1eau,
France; Commission of European Communities; Joint
Research Center, Ispra, Varese, Italy.
The Mark-A is a coarse- grid box model which
calculates radionuclide transport using an ocean
circulation as input. It must be coupled with a
dynamical circulation model. Mark-A has been used
for risk assessments of high-level waste disposal in
the subseabed for two different locations.
3
ADMINISTRATIVE CRITERIA
1. Robinson, A.R. and M.G. Marietta, (1985). Research, Progress and
the Mark-A Box Model for Physical, Biological and Chemical Transports.
SAND84-0646. 2. Marietta, M.G. and W.F. Simmons, (1988). Feasibility of
Disposal of High-Level Radioactive Wastes into the Seabed: Dispersal of
Radio~uclides in the Oceans: Models, Data Sets, and Regional Descriptions.
SAND87 -0753 (Also Feasibility of Disposal of High-Level Radioactive Wastes
into the Seabed. Volume 5. (1989) NEA, Paris). 3. de Marsily, G. et al.
(1989). Feasibility of Disposal of High-Level Radioactive Wastes into the
Seabed. Volume 2. Radiological Assessment. NEA, Paris.
3.2
Hardware Requirements
Mark-a has been used on DEC PDP11's, VAXs, and IBM PCs.
solver, and GEARB is provided with it.
3.3
Application
It uses an ODE
Code is not user friendly. Interbox transports must be obtained via a
subsidiary calculation from a high resolution simulation of the basin
circulation. Interbox transports for nested boxes which must be site-
specific have to be obtained from local eddy-resolving simulations of the
- 215 -

-------
site region. Problem setup for a new site involves understanding
dynamical ocean simulations from which the transports are extracted.
the
3.4
Level of Expertise Required
Running the code is simple. Setting up new problems or performing
sensitivity analyses on physical oceanographic parameters is not. Ocean
modeling experience and a knowledge of descriptive data bases is required.
4
TECHNICAL CRITERIA
4.1
Peer Review
Code was developed by an NEA working group comprised of national experts
from C.E.C., F.R.G., France, Switzerland, U.K., and U.S. It has been
exercised and intercompared at Sandia, Eco1es des Mines de Paris, and C.E.C.
(Ispra). The various experts who served on the working group and the staff
at these institutions have reviewed the model.
4.2
Verification
In addition to peer review comments, Mark-A was a participant code in a
model comparison study conducted by the NEA CRESP Modeling Task Group.
Nyffe1er, F. and Y. F. Simmons, 1989. Interim Oceanographic Description of
the Northeast Atlantic Dumpsite III. NEA' Paris.
4.3
Uncertainty
Uncertainty analyses were carried out by C. E. C. using a Monte Carlo
sampling approach by coupling the Mark-A with a systems called LISA. LISA is
a C.E.C. version of the Canadian systems code, SYVAC which was also used in
an earlier study on the Mark-A. In both studies using LISA and SYVAC
insufficient data were available to do a meaningful uncertainty analysis.
Parameter ranges and distributions that were randomly sampled were guessed in
order to demonstrate the capability for doing uncertainty analysis when
sufficient data become available. Therefore, results of these studies were
not considered realistic.
4.4
Sensitivity
Sensitivity studies on various box configurations
numerical parameters have been carried out by the NEA
contained in the POTG reports.
and physical and
SYG POTG and are
4.5
Input
Interbox transports (an ocean circulation).
Geochemical parameters.
4.6
Output
Box concentrations
estimates.
of
radionuc1ides
and
maximum
individual
dose
- 216 -

-------
4.7
Source Term
Release rates of each radionuc1ides from bottom source at repository
location.
4.8
Scenarios
Scenarios for ocean disposal have been defined and analyzed but require
different model set-ups.
4.9
Regulatory Standards
There are no ocean standards.
5
SCIENTIFIC CRITERIA
5.1
Theory
For the Mark-A box configuration, East Atlantic and West Atlantic boxes
were placed side by side to constitute a four-zone North Atlantic model.
This choice was motivated primarily by basin geometry, underlying topography.
site locations, local mixing times, and anticipated resolution. The east-
west division of the basin was motivated by the Mid-Atlantic ridge, and the
north-south division by the different senses of flow typifying the northe~
and southern parts of the North Atlantic gyre. Simple two-layer vertica
resolution is sufficient to separate-upper water from deep water, but because
the deep water is bounded intermittently by the Mid-Atlantic ridge, three
layers are necessary in the vertical. Boxes for the Arctic and South
Atlantic Oceans are also included, as is a remaining ocean waters box. These
boxes all have two layers in the ver~ica1. This geographic box arrangement,
is thought to be the simplest coarse grid arrangement consistent with
realistic results.
The source basin box includes two nested boxes: a site box over the
source and a larger, eddy-mixing box sized from the local mean eddy diameter.
Their size and configuration take into account the projected mixing times of
the region and desired resolution of the results. Benthic sediment layer
boxes, described in the geochemistry section, are underneath the water boxes.
A collection of shelf/slope boxes is a subsidiary calculation, but does not
have a significant effect on the distribution of tracers in a basin or gyre-
sized box, except for some sidewall scavengers.
The geochemical scavenging model described herein for the Mark-A model
is essentially that developed by GESAMP (1983) which was taken from the ideas
formulated primarily by Cochran (Robinson and Kupferman, 1985; Marietta and
Robinson, 1981). The geochemical boxes are of higher resolution than the
physical boxes. Geochemical boxes were chosen on the basis of their ability
to describe concentration distributions (natural trace fields and releases)
dicatated by the dominant geochemical processes. These differ significantly
from the dominant physical processes and, correspondingly, Mark-A geochemical
box geometry differs from the physical box geometry.
- 217 -

-------
Two types of particles are considered: large, biogenic particles (>50
m) and small, inorganic «50 m) particles, both of which interact with
dissolved contaminants. The large-particle class includes two types of
biogenic particles: fecal pellets (agglomerations of fine particles rich in
organic matter) and hard parts such as the shells of plankton. Both types of
large particles disintegrate during settling. The organic-rich fecal pellets
break down through organic matter oxidation and the hard parts (tests)
dissolve. The inorganic small particles sink slowly and interact with
dissolved contaminants throughout the water column. Small particles can be
transformed into large particles by grazing of bathypelagic organisms.
Particle-solution interactions are considered first-order interactions
with respect to concentration of contaminants on particles and in solution.
The use of a sorption coefficient, Kd (forward reaction ratefbackward
reaction rate), to parameterize these interactions seems appropriate at this
stage of the effort. The sorption coefficient refers only to water column
particles, not sediments.
The bottom sediments are represented by a sell-mixed bioturbated layer
about 10-cm thick in which particle-associated contaminants can reach
sorption equilibrium with the pore water and can diffuse into the overlying
water of the benthic boundary layer. Net removal from the water column and
the sediment mixed layer is by sediment accumulation (i. e., sedimentary
burial) .
5.2
Validation
Box models including the Mark-A hav~ not been shown to compare varoably
with any ocean tracer distribution. Coupled with a robust ocean circulation
model, the Mark-A has been shown to reporude model distributions. Further
work is required with both the dynamical ocean model and related box model to
reproduce both natural and introduced tracers in the ocean (Marietta and
Simmons, 1988).
5.3
Treatment of Radioactive Decay Products -
If data are available, simplified chains could be handled.
5.4
Underlying Assumptions
See 5. 1.
5.5
Pathways
Compliance with the system of dose limitation recommended by the ICRP
and embodied in the IAEA Basic Safety Standards is checked by assessing the
dose to the critical group of members of the public and ensuring that this is
less than the appropri"ate dose limit. In many cases it is not possible to
asses the doses directly, so that a derived lim~t is used. The derived limit
is directly related to the basic dose limit by a model or models: the mod~ls
and parameter values are chosen to maximize the dose, so that compliance with
the derived limit ensures virtual certainty of compliance with the basic dose
limit. The models used are ones describing the redistribution of the
radionuclides in the environment and specific pathways by which man may be
- 218 -

-------
irradiated. Site-specific assessments, applicable to wastes of known
composition, and optimization and/or an apportionment of a fraction of the
dose limit may lead national authorities to limit dumping rates well below
the dumping rate limits.
The choice of the critical pathway is not straightgorward and for
generic assessments of this type it is necessary to postulate a number of
pathways which may be critical. Calculations are then carried out for all
the postulated pathways; the one which leads to the highest doses from a
given radionuc1ide is the critical pathway for the radionuclide.
The pathways selected are listed in Table C-3 and include many which are
known to exist as well as some that could occur in the foreseeable future due
to changes in the use of ocean resources. Other pathways have been
investigated, the existence of which is possible given the current knowledge
of the marine environment, but whose proabi1ity of leading to exposures is
very low. These are listed in Table C-4 as hypothetical pathways. Even
tpough these pathways are viewed as improbable, they are identified so that
due account can be taken in assessing results for the purpose of setting
limits. This use of many general pathways should ensure that the discovery
or postulation of a new pathway will not involve changes to the limit unless
it is clearly not already covered by a pathway. The parameters selected for
the various pathways are intended to be sufficiently general as to cover
critical groups in all areas of the world. In addition to considering each
separate pathway it is also necessary to determine the manner and extent to
which the doses through the various pathways should be summed in determining
overall dumping rate limits. Bearing in mind current known habits from
experience and dietary habits in North American, European and Asiatic
countries, and guided also by effective limitations of protein intake, it was
concluded that a total consumption of 600 g.d-1 of seafood would be
reasonable and should be used in summing doses via actual pathways.
Both of the hypothetical pathways, deep-sea fish and plankton, should be
considered separately and the doses not added to those derived for the actual
seafood pathway in determining dumping rate limits wihout careful
consideration of such a step.
The dumping rate limit for a. combination of pathways
single critical group is obtained conservatively from
irradiating a
1
I'comb - Li
where:
Lcomb is the dumping rate limit or N pathways combined
i
refers to the pathway
N
is the number of pathways leading to exposure of the same critical
group
Li
is the dumping rate limit for pathway i alone.
- 219 -

-------
Conservative assumptions and parameter values are used throughout the
models leading to the critical group dose assessment. This procedure
obviates the need for specific additional safety factors to be applied to the
resultant dumping rate limits. Lower limits might, however, be set by
introducing safety factors, or through the process of optimization, and these
would then correspond to the 'authorized limits' set by national authorities.
See text.
5.7
Ingestion Pathways
The limiting dumping rate for any radionuclide from ingestion is dervied
from the appropriate concentration in sea water as follows
L A Bq.a-l
ij - Kij Gi Qi
where i refers to the pathway and j to the raionuclide
and Kij
is the radionuclide concentration in sea watel correspo~~ing to the
unit dumping rates of the radionuclide (Bq.L- per Bq.a )

is the consumption rate (kg.a-l)
Qi
G.
1
is the concentratioy factor in the ingested material for the
radionuclide (L.kg- )

is the ALl for the radionuclide (Bq.a-l).
A
(1) Seafood pathways. Consumption rates for seafood were chosen on the
basis of experience of high rate values in a number of countries in which
seafood forms a large part of the diet of communi ties which are often
involved with seafood production. These consumption rates and the
appropriate concentration factors are derived primarily from knowledge of
coastal and continental shelf products. The fish pathway covers all mobile
fish species, whether pelagic or benthic. No deep-sea fish pathway exists at
present and thus there is no direct evidence on which to derive consumption
rates. The deep-sea fish pathway is still considered hypothetical and
assessment is retained primarily for completeness.
The molluscan pathway is representative of sessile filter feeders and
grazers and is characterized by relatively high concentration factors for
non-conservative nuclides. Seaweed is eaten directly and may also be used to
produce food additives. The crustacean consumption pathway is intended to
embrace similar organisms and could include krill.
(2) Other ingestion pathways. Certain other ingestion pathways may be
identified. These pathways may be subdivided into two categories: those of
a more primary (direct intake) nature, and those related to indirect intake.
The second group is related to such activities as the use of seaweed and
other marine products as fertilizers, rainfall on to soil and crops and the
use of rain as a source of drinking water. The primary pathways are the
desalination of sea water for use as drinking water and the evaporation of
- 220 -

-------
sea water to produce salt for human consumption.
of beach sediment was considered as a pathway.
In addition, the ingestion
For some groups, desalinated sea water ma1. form the only source of
drinking water and an intake rate of 2 kg.d- has been adopted. The
con~entration factors in this situation are taken as unity for tritium and
10- for a11 other radionuc1ides, except those with compounds of a volatile
.nature in sea water (e.g. C, Sc, I, Po) for which a concentration factor of 4
x 10-3 is assumed. These numbers are based upon at l~ast a two-:rtage
desalination plant where a decontamination factor of 3x10- and 6x10- per
stage exists for non-volatile and volatile radionuc1ides, respectively.

For the sea-salt consumption pathway an intake rate of 3 g.d-l has been
taken. The concentration factor for tritium in this pathway is assumed to be
unity. since tritiated water will be evaporated in salt manufacture. Note
that a very small amount of tritium will remain in the water of
crystallization of sea-salt. For non-volatile and volatile radionuc1ides,
the concentration factor is taken to be 3 x 101. It is recognized that some
volatile radionuc1ides will be lost in the evaporation; however, this has
been ignored to retain a conservative approach.
(3) Other pathways investi~ated. An evaluation was made of the
importance of the ingestion of beach sediment, particularly by children, as
an important pathway. Ingestion of such sediments is expected to become more
important than tre inhalation of resuspended sediment if the ingestion rate
exceeds 2.1 g.a-. In addition, the quantity of sediment that would have to
be ingested for the beach ingestion pathway to attain the same radiological
significance as the adult seafood pathway was determined. Sand concentration
factors were deduced from the available sediment Kd values (which relate to
fine sediment) on the basis that the Kd values for sand are generally only 10
of those for fine sediment. The calculations were done for several
radionuc1ides with long radioactive half-lives and emphasis was placed on
those exhibiting high Kd but low concentsationfactor~9 The necessary sand
~~,estion rate varies. as muc~ as from 10 g. d -1 fot: Tc and 700 g. d -1 for
Pu down to 12 g.d 1 for 22Th. the conclusion was that the ingestion of
sand does not ra~ as a pathway comparable to the consumption of seafoods,
with the possible exception of thorium.
A second problem considered was that of an ingestiong pathway for
sediments that may occur when areas of marine sediment are reclaimed for
agriculture; Kd values for fine sediment were adopted for this purpose. A
comparison was therfore derived between the dose at unit concentration in se~
water between the combined seafood pathway and a vegetarian diet of 500 g.d
green vegetables. It was assumed that there was no change in concentration
as marine sediment was converted to soil: this provides a safety margin
which strengthens the conclusion that this pathway ranks much lower than
marine seafood consumption.
Calculations were performed for Pu, Tc and Cs. Consumption rates for
vegetables as opposed to seafoods by factors of at least more than 1000, 200
and 20, respectively, would be required for the two pathways to attain equal
importance.
- 121 -

-------
5.8
Pathways Involving External Exposure
A second group of pathways that sould be considered is that involving
external exposures.
given by
The dumping rate limit based on external exposure is
Lij
D Bq.a-l
KijEijTiFi
where i refers to the pathway and j refers to the radionuclide
and Dj
Kij
is the appropriate dose limit (Sv.a-l)
is the radionuclide concentration in sea water or on the sediment
corresponding to the unit dumping rate of the radionuclide (Bq.L-l
per
Bq.a-l)
Eij
is the dose factor which gives the dose rate as the result of unit
radionuclide concentration in sea water or on the sediment (Sv.h-
per
Bq . L -1)
is the occupancy time (h.a-l)
T.
1
F.
1
is a geometrical modifying factor
The inclusion of a large number of external exposure pathways, all with
different occupancy times and geometric factors, would over-complicate the
calculations. Only one sediment-related external exposure pathway was
defined, the BEACH pathway, for which both whole body external doses and
doses to the skin would be calculated. The following pathways are
pos tula ted: - -
(1)
Swimming in contaminated sea water (SWIM)
For a critical group a total exposure time of 300 h.a-l seems a
reasonable assumption. Calculation of a dumping rate limit should be done
for total immersion in the sea water.
(2)
Sailing on contaminated sea water (BOAT)
The contaminated sea can be considered as an infinte surface source. It
seems prudent to consider enrichment of radionuclides in the surface
microlayer and any consequent increase in the gamma radiation field, although
this is likely to be minor. People working ilt sea may spend considerabli
time on board a ship or on a platform. An occupancy time of 5000 h.a-
should therefore be used. A geometrical modifying factor of 0.2 may be used
- 222 -

-------
to account for the fact that these people are generally well above the
surface source and are shielded by the vessel or platform.
(3)
Sediment-related pathways (BEACH)
The occupancy time selected for the BEACH pathway is 2000 h. a-I, and
concentration factors appropriate to silt are to be used in the dose
calculations. It is unlikely that fishing gear (nets) and other marine tools
such as anchors, drilling pipes, etc. will have a contamination level higher
than the surface level. Therefore, the exposure due to handling these
materials contaminated with gamma-emitting radionuclides can never be a
critical pathway compared with the situation described in item (2). If
sediments are contaminated with radionculides originally dumped into the deep
sea, the handling of these sediments could result in an exposure because of
dredging activities. This exposure will mainly come from the transportation
of dredged spoils contained in the disposal vessel. The spoils can be
regarded as a surface source. Man can consequently be irradiated by standing
at the edge of this surface source. Operations of this kind as well as other
sediment-related pathways such as lying on the beach and living on
contaminated sediment are included in the BEACH pathway.
(4)
Mining of deep-sea sediments (MINE)
It is assumed that the operations would mine for 250 d.al at a depth of
1000 to 4500 meters. Estimates for the size of a mine site are from 18,000
to 55,000 km2 to be used for 20 years; 1.5 to 3 million tonnes of nodules
would be processed per year. The nodules are assumed to be in a wet slurry.
Since manganese nodule mining is expected to \e a highly mechanized and
remote operation, an occupancy time of 500 h. a-is assumed. As with the
deep-sea fish pathway, a lateral distance from the source is needed.
5.9
Inhalation Pathways
Inhalation pathways are to be dealt with on the basis of the following
assumptions:
(1)
The rate of human air respiration is 23 m3.d-1
(2)
(3)
The concentration of atmospherically borne particles
environment is 10 g.m-3

The atmospherically borne paftic1es comprise 0.25 g. m - 3 fine coasta~
sediment particles, 3 . 3 g. m - dried sea - sa1 t particles, and 6. 6 g. m-
particle-associated water
in the
coastal
(4)
The concentration of atmospheric vapour in coastal air is 10 g.m-3. In
order to include inhalation pathways involving each of these atmospheric
constituents, the following approach is used:
(a)
Inhalation of radionuclides associated with sediment particles
should be calculated using Kd values for coastal sediments.
(b)
Inhalation of radionuclides associated with dry sea salt ti 1 t
- par c e
should be calculated using a common enrichment factor of 3 x 10
- 223 -

-------
for all elements. It is clear that for some subpathways this will
introduce some inherent conservatism for volatile elements in sea
water, but the magnitude of this is unlikely to exceed a factor of
2. The enrichment factor should be used with the surface sea water
radionuclide concentration in the far-field model and the average
surface concentration in the near-field model.
(c)
Inhalation of radionuclides associated with the aqueous phase of
atmospheric particles should be ignored. Enrichment factors (EF)
for nuclides would have to exceed 106 for the aerosol water
inhalation pathway to rival the magnitude of the atmospheric vapour
inhalation pathway. Estimated enrichment factors are less than 10.
(d)
Inhalation of atmospheric vapour should be calculated on the basi~
of the enrichment factor for single phase distillation of 3 x 10-
for most elements. Exceptions to the use of this enrichment factor
are the following elements:
(i)
(11)
3H, which should have an EF of 1

l4C, which should have an EF based upon equilibrium between
C02 in sea water and atmospheric C02
(Hi)
Cl, Kr, Ar, Xe, I, Tc, Se, Te, Po and Hg, which should have
EFs corresponding to equilibrium between atmospheric vapour
and sea waer at STP.
In addition to the inhalation pathways considered above, a further
pathway was identified and considered. This involves the resuspension of
biological material (e. g. algae) from the sea surface into the atmosphere
that has been observed in the Marshall Islands. H.owever, transport and
subsequent inhalation of radionuclides by this pathway would be covered by
the resuspension and inhalation of dry sediment particles, with higher
concentration factors for the radionuclides compared to algae. It was
concluded that there was no need to consider this inhalation pathway
separately. assuming that th~ mass concentration of dry sediment particles in
the atmosphere (0.25 g.m-) was sufficiently large to encompass the
additional concentration of algae resuspended from the sea surface.
5.10
Ocean Regions of Water Concentration Calculations
To carry out dumping rate calculations it is necessary to specify the
regions of the ocean in which water concentrations are to be calculated for
each exposure pathway.
5.11
Calculations for Actual Pathways
For actual pathways, with the exceptions of mid-depth fish and mining,
it is appropriate to use only surface water concentrations of radionuclides.
This is because the mollusca, crustacea, seaweed and most of the fish
currently consumed all originate from surface waters. In addition, all the
external exposure pathways and inhalation pathways, except for mining, are
related to man's immediate environment. In the case of mid-depth fish, it is
appropriate to calculate a water concentration for a depth typical of those
- 224 -

-------
in which the fish (which are caught at present) live. A depth of 1500 m was
therefore selected. For the mining pathway, the ocean region needs to be
selected on the basis o( the areas in which mineral resources are located.
Since dumping sites are chosen outside areas with significant resources, it
is not appropriate to calculate extreme near-field (1. e. dumpsite) bottom
water concentrations for this pathway.
5.12
Calculations for Hypothetical Pathways
Plankton is, at present, viewed as a hypothetical pathway only. and no
firm basis could be found for representing it within the assessment of
dumping rate limits. Because of this the results of calculations should be
dealt with separately. If they should indicate a limiting constraint such as
to influence decisions on grouping, further consideration will be necessary
before any binding decision is taken to adopt them. Surface water
concentrations as such suffice for these calculations.
It is clear that for the deep-sea fish pathway, bottom water
concentrations should be used. However, specifying. the lateral distance from
the dumpsite at which the concentrations should be calculated is a more
difficult matter. The minimum region required to supply food to sustain a
yield of one demersal fish per day has been estimated to be of the order of
hundreds of square kiometre of the dee-p-ocean floor. For a sustainable
fishery the area would be very much larger and would be greater than the
dumpsite area assumed in the ocean' models. Thus, while it is not
inconceivable that fish living immediately above a dumpsite could be caught
and consumed, it is extremely unlikely that members of critical groups would
obtain all their annual intake of deep-sea fish from such a region.
5.13
Dose Conversion and Dose Response
Dose
included.
is estimated to critical
Requires code coupling.
individuals.
Dose
response
is
not
- 225 -

-------
TABLE C-3.
PATHWAYS AND MODES OF EXPOSURE
Pathway
Mode of Exposure
Symbol
Surface fish consumption
Mid-depth fish consumption
Crustacea consumption
Mollusc consumption
Seaweed consumption
Salt consumption
Desalinated sea water consumption
Suspended airborne sediments
Marine aerosols
Boating
Swimming
Beach sediments
Deep-sea mining
Actual Pathways
FISH-S
FISH-M
CRUST
MOLL
WEED
SALT
DESAL
SED
EVAP
BOAT
SYIM
BEACH
MINE
Ingestion
Inhalation
External irradiation
External irradiation/inhalation
Hypothetical Pathways
Deep-sea fish consumption
Plankton consumption
FISH-D
PLANK
Ingestion
- 226 -

-------
APPENDIX D
DECISION-AIDING METHODS FOR COMPARING DISPOSAL OPTIONS
- 227 -

-------
1
INTRODUCTION
Formal decision-aiding approaches are orderly, systematic procedures for
making comparisons among complex alternatives. They clarify the structure
and sequence of a decision process or a comparison by disaggregating it into
smaller sub-decisions or evaluations that are easier to make.
Decision-making theory distinguishes between descriptive theory, which
is concerned with how people actually reach decisions, and normative or
prescriptive theory, which describes how people should reach decisions under
assumptions about rational behavior. This section is devoted to the latter,
although some understanding of the former is useful in predicting public
responses (mostly objections) to formal approaches.
Any basic discussion of decision making should include mention of an
important difference between good decisions and good outcomes. Wfiters on
decision-aiding methods make a strong distinction between the two. A good
outcome is a state of the world that we value above others. A good decision
is one that is logically consistent with the alternatives we know about, the
information we have, and our preferences for the things involved. Under
uncertainty. a good decision can lead to a bad outcome, and visa versa.
It is failure to understand this distinction that causes decisions to
become bogged down in "but what if?" types of questions. No decision-aiding
method can assure the quality of an outcome. All we can do is play the odds
and attempt to include in our decisions resilience for responding to
unfavorable outcomes. Holling cautions:
[Analytical decision-aiding techniques] "are typical of the
search for an optimal solution. It usually comes as a shock to
those nurtured in this perspective that complex living systems
have not organized themselves in accordance with this
principle,...ecological systems sacrifice efficiency for
resilience. . . . Evolution shows that their fail-safe strategy is
eminently suited to a world which is inherently unpredictable at
certain times."
Yildavsky extends these thoughts:3
"In the past we relied on resilience. We anticipated the worst
and we relied on a general capacity to be flexible to respond.
By attempting a politics of anticipation, we destroy any
potential we have for growth and for group accommodation."
There can also be large differences between "good" decisions based on
some set of technical criteria and "satisfying" decisions, ones that makes
people feel they have done the right thing.. This arises in part from
fundamental inabilities of formal approaches to deal adequately with
subjective responses. Subjective criteria may be left out because they are
not understood, are thought not to be important or appropriate, or for lack
of a way of including them that is consistent with the formal method. . Or
they may be included, but evaluated in ways that produce the "wrong" answer,
one that fails to capture the proper values -- either because of inadequacies
in the elicitation or because peoples' subjective probabilities or values are
- :£28 -

-------
based on misinformation and lack of understanding, so that the "real" answer
doesn't match their understanding.
There. may also be feelings that some things should not be subjected to
this kind of analysis. Human lives, for example, should not be for sale.
But a comprehensive analysis might show that the cost of saving a human life
in a particular circumstance is unreasonably high given other lives that
could be saved with the money. The formality of the decision process
therefore might force people to face what seems in this context to be a
reasonable judgment that they otherwise think should not be made.
2
DECISION-MAKING THEORIES
Merkofer outlines the three distinct, internally consistent theories of
normative decision making that are most widely accepted: cost-benefit theory,
decision theorY., and social welfare theory. Other, less accepted theories
~re available. 5 The following overview of these theories is abstracted
directly from Merkofer. Major characteristics of the theories are summarized
in Tab,le D-l.
Cost-benefit theory is based on the premise that alternatives should be
selected through a systematic comparisons of advantages (benefits) and
disadvantages (costs) arising from estire,:ed consequences of the choice. The
theory does not include a decision maker with special responsibility for
deciding what is "best." Instead, "best" is defined in terms of an
"efficiency criterion" calculated as a function of total costs and benefits.
Incommensurable costs and benefits are aggregated as dollars through the
concept of individual "willingness to pay" to avoid or incur them. In the
absence of other information, it is assumed that market prices reflect
willingness to pay. Thus cost-benefit theory attempts to maximize social
utility by maximizing the aggregate economic value of goods and services
consumed by all individuals.
Although cost-benefit theory is appealing, Merkofer points out that it
has several troublesome implications. Most important is that use of
aggregate net benefit means that cost-benefit theory considers only total
societal welfare, without regard for the distribution of that welfare within
the society. Thus, through a higher willingness to pay by wealthier
individuals, the theory can produce a redistribution of resources from the
poor to the rich, so long as the rich gain more than the poor lose.
The general concept of willingness to pay also has deficiencies. Market
prices are not always good indicators of willingness to pay, and the prices
reflect only the preferences of the purchasers of the product. In addition,
quantifying willingness to pay for impacts that are not traded in the
marketplace is exceedingly difficult and highly uncertain.
Other miscellaneous assumptions are also troublesome. The
between money and welfare (utility), for example, is assumed to
the same for everyone, and the current distribution of income
be an equitable distribution of power. .
relationships
be linear and
is assumed to
- 229 -

-------
Decision theory describes how individuals should make decisions in the
face of uncertainty. The axioms of decision theory are based on principles
that define "rational behavior" in mathematical terms. These lead to two
results that are central to the method. First, that a decision maker's
preferences can be encoded in terms of a "utility function" representing a
scaling of the values assigned to outcomes, including ca,%turing attitudes
toward accepting risk (see von Neuman and Morganstern, 1947). Second, that
a decision maker's preferences for alternatives can be "measured by the
expected utility of their outcomes (the sum of the utilities of all outcomes
weighted by their probabilities). The preferred alternative must have the
highest expected utility.
Decision theory contains a number of implied assumptions that may be
questionable. Most important of these are:
.
The decision maker accepts the axioms of rationality on which the
theory is based;
.
All possible events and all significant consequences can be
enumerated in advance;
.
Meaningful subjective probabilities and utilities can be obtained and
assigned to them; and
.
Disparate outcomes of concern to a decision maker can somehow be made
comparable.
The first assumption is demonstrably incorrect in many circumstances,
especially when decisions involve large amounts of money or differences among
probabilities are small. Others may be correct under ideal circumstances but
not under more complicated ones.
Finally, decision theory is focused on individual decision making. It
does not lend itself well to collective decisions. How individual decisions
should properly be extended to collective problems remains an open question.
Social welfare theory is based on the premise that societal decisions
should arise from a rational synthesis of the preferences of all affected
individuals. It is therefore concerned with finding decision rules or
procedures by which individual specified preferences can "be incorporated into
the decision process. There are various lines of research in the field; the
relevant line for this .report attempts to derive ways of aggregating
individual preferences into a societal utility function.
lndividual preferences can be expressed cardinally or ordinally.
Cardinal preferences contain quantitative statements about the relative
strength of preference of one outcome over another (e.g., A is twice as good
as B); ordinal preferences contain qualitative statements about the direction
of preference (e. g. A is better than or equal to B). Because cardinal
expressions of preference are especially vulnerable to subversion through
strategic misrepresentation, early efforts focused on whether or not ordinal
expressions of preference can be combined. It has now been shown that it is
not possible to quantify a group's preference structure using ordinal
preferences in a way consistent with some fundamental assumptions of
- 230 -

-------
rationality. And even if it were possible to find an optimal social
decision-making rule, it is not clear that individuals would reveal their
true preferences without strategic bias, so decisions would inevitably be
suboptimal because they would be based on incorrect information.
Researchers have therefore been forced to rely on cardinal preferences,
despite the difficulty in assessing them. These utility functions are of the
kind used in conventional economics, which measure preferences under
certainty, or those used in decision analysis as described above, which
measure preferences under uncertainty.
Several sets of postulated properties for aggregating individual.
preferences have been explored. Most have specifically been aimed at
justifying use of a linear combination of individual preferences. That is,
group functions should be weighted averages of individual functions.
Harsanyi obtains this result by assuming that group preferences satisfy the
same axioms of rationality as individual preferences and that the group w~la
be indifferent between alternatives if all individuals are indifferent. .,
Harsanyi also obtains it by assuming "impersonally situated" rational
decision makers who face equ~l changes of winding up in the shoes of anyone
affected by the decisions. Kirkwood obtains it by adopting Pareto-
efficiency for aggregating individual preferences that is, if one
alternative is better than another for at least one individual and no worf8
for any other, then that alternative must be preferred by the group.
Keeney and Kirkwood fiyd three assumptions required to provide a linear group
utility function are:
.
Group utilities depend only on individual utilities;
.
If everyone has the same utility function, then the gI:'OUp function
must be a scaled version of the individual function; and
.
If all but one of a group are indifferent between alternatives, the
group preference should match the preferences of the individual who
is not indifferent.
When Keeney' and Kirkwood revise the second assumption to state that
group preferences between alternatives should depend only on the preferences
of individuals who are not indifferent, then they obtain a multiplicative
form for the group function.
A maj or problem with social welfare theory is the need to assess, or
assume, individual utility functions. First, utility functions in general
assume that individuals are "rational" in the sense that they try to behave
in accord with a set of axioms  defining rationality. In general, this in not
the case. Even when individuals try to be systematic in their thinking, they
tend to engage in strategic behaviors intended to influence decisions in
their favor. So it is unlikely that individual utility functions can
successfully be measured for the large number of persons required. It is
also unlikely that all persons involved could be identified.
Another problem is aggregation of individual utility functions. The
aggregate function will be some weighted combination of individual functions.
How should the weights be assigned? All opinions are probably not equally
- 231 -

-------
good. For technical problems that are not well understood by all, it is
especially likely that there will be large differences between quality of
opinion (level of information and understanding) and political power. Should
weights then be based on political power alone? This can produce happier
people and technically weak decisions.
Merkoffer presents an example of an application of social welfare theory
in which eight different aggregation and weighting schemes were used to test
the effects of this stage of analysis on results.
3
DECISION-AIDING APPROACHES.
Practical applications of the decision-aiding theories include a wide
variety of procedures for dealing with all aspects of analysis. Most are
sufficiently flexible to accommodate a great many variations as required.
Kerkofer describes several classes or types of applications that are commonly
discussed in the literature.
The two basic types of cost-benefit analyses are a Paretian approach and
a decision-maker approach. The Paretian approach attempts to produce a
purely objective decision model. It is relatively uncompromising. in
maintaining objectivity and using market values for goods and services. Some
practitioners evaluate only those consequences having market values, stating
that results represent purely economic. efficiency and are regarded as only
one of many social welfare inputs necessary for decision making. The
decision-maker approach attempts to include all issues of importance, not all
of which have market prices. Non-market goods and services are included as
personal values and judgments of the decision makers. Although, in theory,
the decision-maker approach should be based on values and judgments of
decision makers, Merkofer states that in practice,most have not involved the
decision makers at all, but have relied on assumed value functions generated
by the analysts.
Decision analysis methods applied to social decisions can diverge
somewhat from the basic theory, which rests on an assumption of a single
decision maker. The simplest application assumes that a single individual,
or a group with consistent views, has responsibility and authority for making
public decisions. In such a case, the group decision reduces to an
individual decision and no modifications are required. A variation on this
is assumption of a "supradecision-maker , " a benevolent dictator who
synthesizes a social utility function from the preferences of the individuals
affected rather than his own preferences. On occasion, one or a few members
of a group are taken as representative of the whole.
An alternative approach is to base the analysis directly on group
judgments and preferences using special schemes for obtaining probabilities
and utilities that represent society as a whole (social decision analysis).
The analyst creates an explicit model of the decision including prefef~nI~s
of different groups, often called "stakeholders" and "players." , .
Stakeholders are persons with a direct interest in the outcome of a decision;
players are persons interested in influencing a decision. Note that players
need not be large stakeholders in the ordinary sense, but may instead have
interests arising out of other, related agendas.
- 232 -

-------
Either the supradecision-maker or the group judgment approach can be
applied as separate. preference models that are combined into a societal
judgment by a weighting function that expresses the relative importance of
the various groups. The weighting function is supplied by the decision
maker, and can be based on any criterion he considers appropriate, including
size, political power, ability to slow the project through opposition, level
of knowledge and skill, etc., depending on the objectives of the analysis.
In practice, in an open decision it may be politically difficult to use
weights based on anything other than group size or group need, although for
many formulations of proj ect obj ectives, other weighting schemes make more
sense. This is particularly true if the decision maker has some mandated
responsibility that conflicts with public desires. Regulatory control of
cigarettes is a particularly good example in which health effects are
relatively clear (compared to other environmental problems). the government's
responsibility to protect public health is clear, yet legislation is
constrained by the political power of smokers and the tobacco industry.
All decision analysis approaches rely heavily on interviews with
decision makers and technical experts to obtain the necessary information on
probabilities and values. An important characteristic if these interviews is
posing of choices in which respondents much indicate preference or
indifference among clearlv defined alternatives. The process tends to be
dynamic, changing with" ~ue nature of the responses, and can require hundreds
of judgments and decisions from each individual.
Because of this reliance on interviewing, the probabilities used on
decision analysis are often subjective. That is, they are individual beliefs
about the likelihood of events rather than observed frequencies. If the
probabilities are elicited from technical experts, they may be observed
probabilities, generalizations of objective probabilities (i.e. mental
syntheses of observations). or best estimates of objective probabilities in
the absence of specific data (guesses; experts call it "engineering
judgment", which means a guess by someone who claims to know more). If
probabilities are elicited"from ~he decision makers, who mayor may not also
be experts, then they represent part of a model of the decision as seen by
the decision makers. Depending on the knowledge and skill of the decision
makers, the two need not necessarily be different.
Social decision analysis tends to rely more on normative arguments and
less on interviewing individuals than classical decision analysis.
Applied social welfare approaches all attempt to identify decisions that
maximize the welfare of a group as a whole through some aggregate social
welfare function. Uncertainty in the consequences of a decision are dealt
with through a centralized or decentralized process. In the centralized
process, individuals provide value functions describing preferences for
different time streams of outcomes or utility functions describing
preferences for specified probability distributions over those time streams
without regard for the actual outcomes of the decision. Estimation of
consequences is done centrally. using models. This separates the value
judgments from estimation of consequences, reducing possibilities for
strategic behavior. In the decentralized process, each individual is asked
to specify his preference for each alternative outcome directly. The social
- 2.33 -

-------
welfare function then aggregates
uncertainty in the consequences.
the
preferences
without
regard
for
Although the basic theories presented above are relative distinct, in
practice the applications borrow heavily from one another. The decision-
maker approach to cost-benefit analysis often uses procedures from decision
analysis or social welfare theory for representing preferences. Social
decision analysis uses willingness-to-pay and efficiency arguments from cost-
benefit theory to establish a social welfare function. Social welfare theory
often use elicitation procedures from decision analysis to encode
probabilities and utilities. This borrowing of procedures tends to blur the
philosophical differences among the methods. Merkofer provides a "map"
showing relationships among theories and overlapping of rationales for
decision-making approaches (Figure 0-1). With the exception of paretian
cost-benefit analysis and clinical decision analysis, all have significant
overlaps of perspective. Only decision-maker cost-benefit analysis has
significant overlap in all three perspectives.
4
APPLICATIONS
All of the decision-aiding methods described above involve steps that
separate information available on the implications of a decision from the
values and preferences that are used for evaluating them (Figure 0-2). The
core of the process is the decision basis, a formal disaggregation of the
pieces of the problem, including alternative choices at all stages, available
information that connects decisions and outcomes, and preferences of the
decision makers with respect to important characteristics of outcomes, all
connected by feedback loops teat redefine the statement of the problem in
light of preliminary results.1 Each part of the decision basis is narrowly
focused and can be completed relatively simply. rhis helps to prevent
decision makers from becoming overwhelmed with the enormity and complexity of
the whole decision problem. And it makes the basis for the resulting
decision transparent; all of the pieces -- the information and the judgments
-- are open to examination.
Alternatives. The most important step in this process is determining
the structure of the problem. Often, careful specification of the structure
of a problem is can yield sufficieni understanding to show the solution.
Little further analysis need be done. 5 The first stage of determining the
structure of the problem is specifying and characterizing alternatives. Note
that the alternatives need not be end decisions or actions; they might start
with selection of an approach to decision making. One can envision cases in
which selection of a specific decision-making method would essentially
complete the decision, in that all following actions and outcomes are
prescribed by the method and the characteristics of the problem. A decision
to select an alternative on the basis of cost as analyzed in a particular way
might often be such an example, especially if there are large and
uncontroversia1 differences in the costs of alternatives. Diversity of the
alternatives included in an analysis depends largely on analysts'
understanding of the problem, the structuring of the problem in general, and
the effort devoted to creation of options.
- 234 -

-------
von Yinterfeld cautions that the inf~ial stages of problem
are critical to the success of the whole. Many problems can be
than one way; some can be viewed many ways. Improper formulation
sophisticated solutions to the wrong problem.
formulation
viewed more
can lead to
In general, because of greater interaction between decision makers and
analysts, the decision analysis method generates a broader range of
alternatives than other techniques. These often include contingency plans
for responseS to future information. Thus decision analysis can include
desirability of one action over another caused by greater flexibility to
respond appropriately to future information.
The alternatives to be considered can be readily apparent or derived as
part of formulation of the problem using special tools developed by
practitioners of decision analysis. The first stage of selecting
alternatives is framing the problem, often the most difficult stage of all.
Framing is difficult because of fundamental characteristics of human nature
that tend to bury or obscure real problems under a protective facade of face-
saving or decision-simplifying pseudo-problems. Also, in the absence of a
systematic approach to developing statements of problems, preconceptions tend
to interfere by generating subproblems based on preliminary, intuitive
evaluations (blinders). So the first stage, deciding what must be decided,
becomes much more complex than might initially be thought. Few formalized
methods are available to assist in this stage. Mostly it requires
understanding of the way people think about decisions.
Decision trees are fundamental to structuring a decision process,
providing understanding of the relationships among decisions and outcomes and
the flow of the process. These describe the sequence of decisions and
possible random outcomes of decisions from beginning to end (Figure D-3).
Note that the structure of a decision tree is not fixed by the problem. It
is determined by decision makers' understanding of the problem and the way it
should be approached -- the framing.
Decision Criteria. Cost-benefit approaches normally attempt to maximize
net present value (the current value of a time stream of benefits minus
costs) . If probabilities are included in the analysis, then the decision
criterion is expected net present value. Closely related criteria are: (1)
maximizing internal rate of return>, which is the value of the discount rate
that makes the net present value equal to zero; and (2) allocating resources
among alternatives so as to equalize marginal rates of return.
All decision analyses attempt to maximize expected utility. Whose
utility is to be maximized depends on the approach; it may be a single
decision maker representing himself or his organization, a supradecision-
maker representing a group, or several supradecision-makers representing
different stakeholder groups.
Social welfare approaches maximize group aggregate utility derived
without reference to a decision maker. Social decision analysis attempts to
integrate the approaches of cost-benefit theory, decision theory. and social
welfare theory by appealing to efficiency arguments to justify net present
value as a re~sonable measure of social value. It is a value function, not a
utility function in the sense of von Neuman and Morganstern, however, so is
- 235 -

-------
incapable of accounting for attitudes towards risk. Occasionally risk
attitudes are included by converting the value function to a utility function
using an assumed functional form and a "risk aversion coefficient."
~,rfiofer describes an application of social welfare theory by Dyer -and
Miles' that is particularly appropriate to the problem of comparing
different views on disposal of low-level wastes. The problem was to select a
pair of trajectories past Jupiter and Saturn for two Voyager spacecraft given
the requirements of 80 scientists in 11 independent experimental teams, each
having different optimal trajectories. Rather than decide on a single
approach, outcomes were compared when the problem was formulated as:
.
A problem of collective choice based on rank sum (ordinal) judgments;
A bargaining problem;
.
.
A social welfare maximization problem; and
.
A decision for a supradecision-maker concerned with maximizing group
preferences.
Rank sums are one of the oldest and most widely used decision criteria
for collective judgments . Each group assigns a preference rank to each
alternative and the ranks are added to produce a total score for each
alternative. This approach cannot satisfy Arrow's axioms for group decision
making (see above), and it does not account for differences in the strengths
of preferences among alternatives and groups.

The bargaining model was based on a formulation by Nash19 and
generalized to n players by Harsanyi. 20 In this formulation players may
trade goods but not make side payments; they must bargain with what they
have. This can be shown to produce a solution that is:
.
Invariant with respect to utility transformations;
.
Pareto optimal;
.
Independent of irrelevant alternatives; and
.
Symmetric with respect to players roles.
Following some
criterion was found
each alternative.
solution and it may
manipulations and simplifying assumptions, the decision
to be to maximize the product of the teams' utilities for
This formulation does not necessarily have a unique
not be symmetric.
The social welfare model used an additive formulation in which total
welfare was quantified as a weighted sum of team utilities for each
alternative. The team utility functions were scaled so that the differences
between the best and worst alternatives were the same for each team. The
weights were then taken to be the relative importance of each experiment
(team). Participants agreed that all experiments were not equally important,
but were not willing to quantify the differences, so weights were all
- 236 -

-------
assigned to be 1.0 and a sensitivity analysis was done by assigning 2.0 to
six experiments that all agreed were more important.
The supradecision-maker model used the preferences of the project
manager who attempted to maximize group utility. A linear utility function
was assumed for the whole, based on the utilities of each team for each
alternative. The manager selected weights representing his evaluation of the
quality (bias) of each team's estimate of utility and their relative
strengths with respect to the best and worst alternatives, and weights
representing the relative importance of each experiment to the project as a
whole.
A comparison of the results of these four models showed substantial
agreement. All models yielded the same three top-ranked alternatives, but
differed with respect to their order. In spite of the agreement of the four
models, selection w~s complicated by the fact that the individual teams
disagreed significantly with the collective preferences. Only one team was
in complete agreement with the collective judgment; one team gave all of the
top three alternatives its lowest possible ranking.
Thus, regardless of the apparent agreement of the collective judgments,
the actual selection remained highly controversial. This is characteristic
of group decisions of this kind and illustrates the importance of using
methods that are able to characterize and display the differences of opinion
and preference that can generate strong political forces. No such method can
produce a "perfect" solution.
The final decision in this case was reached by discussion and
compromise, essentially through a sense of fairness that produced a
willingness of the "high-weighted" winning groups to give up a little in
order that the losing group would not be left with nothing. Such equity
considerations were not included in the original formulation of the problem.
The alternative selected was adjusted to the extent possible to regain some
lost characteristics without giving up important gained characteristics.
Variables. All of the decision-aiding methods included here require
some ~eans of selecting and quantifying magnitudes of the outcomes on which
the decision will be based and on which decision makers must place values.
Special procedures for generating lists of variables have been developed to
ensure they are complete, appropriate, and free of redundancies.

In decision analysis, Keeney and Raiffa,2l for example, recommend that
objectives be established first through questioning decision makers. These
are then organized in a hierarchy that distinguishes their relative
importance and generality. The resulting objectives hierarchy defines
general objectives (ensure human health and safety) in terms of more
specific, lower-level objectives (low incidence of accidents, exposure,
etc.). The lowest level of the hierarchy is then used to establish
quantitative measures of success in meeting objectives (predicted mortality,
etc.) .
In cost-benefit analysis, procedures are much less formal, relying on
general literature and discussion with decision makers. Partly this is
because of problems with placing monetary values on nonmarket costs and
- 237 -

-------
benefits and resulting confusion about what can and cannot be included in the
analysis.
Values. Merkoffer divides the many valuation procedures used in
decision-aiding methods into those based on market values and those based on
utilities. Market-value-based procedures use established market prices or
measures of willingness-to-pay, either expressed in responses to questions or
"revealed" through analyses of behavior. These procedures assume that
efficient markets insure that prices represent aggregate willingness to pay.
If the decision under analysis will change prices, then the costs and
benefits must be adjusted appropriately through estimated changes in
aggregate consumer surplus - - the difference between what people paid on
average and what they would have been willing to pay, but did not have to.
Utility-based procedures rely on direct questioning of individuals about
relative preferences. The method normally uses exacting and time-consuming
procedures involving assigning numbers on a scale of desirability, stating
relative preferences for a series of conditions, or evaluating tradeoffs
between increases and decreases in pairs of conditions. Value functions are
assessed for consequences that are certain; von Neumann-Morgenstern utility
functions are assessed for consequences that are uncertain. Utility
functions include assessment of attitudes toward risk by having decision
makers evaluate preferences between sure outcomes and lotteries having two
outcomes with known probabilities.
The elicitation methods are difficult to design and administer, and the
responses required of decision makers are often difficult to understand 'or to
relate to familiar circumstances. If more than a few conditions (attributes)
must be evaluated at the same time, it is nearly impossible to make all of
the comparative judgments necessary. Analysts must, therefore, assume
independence of preferences among attributes, which requires care in
definition of attributes to ensure that the assumption is reasonable.
Because of these difficulties, there is not uniform agreement that the
results so ~~tained are always meaningful representations of individuals'
preferences. At a minimum, elicitation of preferences must be done with
great care by experts.
Value of Time. Because cost-benefit approaches tend to rely on dollar
values, the value of time is treated in the standard way of discounting by
opportunity cost. But other approaches include impacts lacking market
prices, and the cost-benefit approach can assign dollar values to nonmarket
goods and services for which the time discounting function may be
significantly different. For example, many people are reluctant to transfer
environmental problems to future generations that are not parties to current
decisions. Others assume that future generations will have improved ways of
solving these problems, so society is better off to defer them to the future.
The first implies a negative discount rate; the second a positive rate.
Provisions must be made for dealing with these different perspectives on the
value of time.
- 238 -

-------
5
CRITICISMS
Merkofer describes a broad range of criticisms of these methods that
have appeared in the literature, ranging from fundamental inadequacies of the
basic logic and assumptions to deficiencies of specific applications. The
following swmnarizes those that are specifically related to social rather
than individual decisions.
The most fundamental criticism arises from the impossibil~ty of finding
a socially optimal decision rule. If individual preferences are expressed as
rankings (as in voting, for example) without an indication of the strength of
their preferences, then there is no consistent decision-making rule that is
not eith~~ dictatorial or completely arbitrary (Arrow's impossibility
theorem) . Most approaches attempt to avoid the impossibility theorem by
incorporation measures of strengths of preferences in social utility
functions. These must either be estimated by a supradecision-maker or
measured by asking a "representative" sample. Asking people what they want
may not produce good answers, because what they say does not always agree
with what they do. Also, the process is subject to subversion -- strategic
misrepresentation or "second guessing" in an attempt to influence the
decision in a favorable direction. Such misrepresentations are common. Even
if preferences are measured accurately. there are problems that have
insoluble structures because of an imbalance between willingness to pay for a
good and willingness to accept payment for a bad. People are often not
willing to pay to get what they want enough to compensate others for giving
up what they have.
The basic approach of all decision-aiding methods is to decompose
decision problems into manageable subproblems or subdecisions. Fundamental
to decomposition is separation of information and preference. But there is
no philosophical distinction between facts and values; normally facts shape
values. Also, focusing on the decomposed details of a problem commonly
oversimplifies it and may divert attention from more creative approaches to
problem solving. Certainly decomposition requires relatively unnatural
decisions that are outside most peoples' experience, and people think most
naturally by analogy with past experiences. Some argue that the results are
therefore spurious.
All decision-aiding methods have limited capability to account for
irreversibility. Because the significance of an irreversible outcome depends
on fu~ure events, and it is impossible to foresee and establish probabilities
for all future events, the methods necessarily undervalue irreversibility. A
decision that is good now can become suboptimal in the future as
circumstances change and available information improves. Because of this,
recoverability from mistakes often has relatively high priority among the
general public (see discussion of resilience, above). All methods are weak
in this area.
Sources of errors in application of decision-aiding methods arise from
.
Omissions and inaccuracies;
.
Difficulty of measuring costs and benefits;
- 239 -

-------
.
Bias in assessment, modeling, and analysis;
.
Interference by analysts;
.
Susceptibility to manipulation;
.
Susceptibility to misuse and misinterpretation;
Decision-aiding methods can also have basic incompatibilities with
established institutional, political, and social norms. There is a feeling,
for example, that it is more important for the decision-making process to
conform with democratic principles of visibility, participation, and fairness
that it is to be logically "correct." The American system of regulatory and
judicial review is based on an adversarial process that seeks to resolve
conflict without violence, not to achieve objective truth. A formal approach
that attempts to produce an integrated, comprehensive, balanced presentation
of issues holds little appeal to advocates. Both sides tend to fear that
introduction of such evidence will reduce the strength of their subjective
arguments. In addition, most approaches concentrate power in the person or
persons who provide value or utility functions to an extent that may be seen
as undesirable by advocates.
It is also clear that decision-aiding methods do not account for
institutional structures and existing governmental decision-making processes,
which tend to be based on bargaining and responses to emergencies. Decision-
aiding methods tend to become locked into a predefined "ritual" that lacks
flexibility to adjust to changing circumstances.

Some argue that formal decision-aid~ng approaches are incompatible with
existing social attitudes, preferences, and practices, so regardless of their
potential value, will not gain real acceptance. Central to this attitude is
a belief that society should not quantify environmental and health impacts,
especially in such a subjective and inconclusive way. Although people
routinely make such judgments implicitly by their actions, they do not want
to be faced with the moral dilemma of an explicit judgment. They also
dislike dealing explicitly with uncertainty, preferring, certain statements
that are incorrect to uncertain statements that fail to support specific
decisions. Ethical concerns include those about equity of distribution of
costs and benefits across peoples and generations, and promotion of
anthropocentric values. There is considerable question about who has a
reasonable right to inflict how much on whom, including mankind on the
natural environment.
6
EVALUATION
There is no formal decision-aiding approach for social decisions that is
free from criticism. And the available approaches have different strengths
and weaknesses. Merkofer and others claim t1\at no formal decision-aiding
approach is intended to replace decision makers. Instead, they pres~nt
useful information in ways that can help decision makers. Choosing a
decision-aiding approach, therefore, is a matter of choosing which
presentation will be most helpful.
- 240 -

-------
This view emphasizes the importance of the analyst - a facilitator whose
skill and capabilities are of prime importance in determining the success
achieved. Emphasis is shifted away from results toward the process, and how
development of the structure and characteristics of the decision can help a
decision maker understand the nature of his problem.
Merkofer characterizes the basic decision-aiding approaches with respect
to five characteristics.
Lo~ica1 Soundness. The soundness of the approaches depends on the
persuasiveness of their underlying behavioral assumptions. Cost-benefit
analysis will be less attractive for situations having considerable
differences in the quality of relevant information. Decision analysis is
less useful if there is no decision maker willing to provide or delegate
authority for providing subjective value judgments. Social welfare theory
has little appeal if the strengths of preferences of affected parties is not
a central issue.
ComDleteness. Formal decision-aiding approaches have few restrictive
assump.tions on how problems should be defined, so completeness is largely
determined by the available information and the skill of the decision makers,
experts, and analysts. Cost-benefit analysis is less effective when little
or no data are available for quantifying important uncertainties or outcomes.
Decision analysis is less useful if decision makers define problems in
habitual ways that overlook options or overemphasize certain factors. Social
decision analysis, because of its broader representation of experts, can
provide a more complete accounting of available information. Supradecision
analysis and social welfare theory provide a more complete representation of
social preferences through integration of preferences of stakeholders.
Standard cost-benefit analysis provides no information on distributional
equity. In general, decision analysis approaches allow a more comprehensive
accounting of factors, because they are able to use data when they are
available and subjective judgments when they are not. Social decision
analysis is especially flexible in this respect.
Accuracv. Cost-benefit approaches are most accurate in assessing wel1-
defined projects, but are less useful when there are significant market
distortions or when information or preferences are changing rapidly.
Decision analysis approaches rely heavily on subjective judgments which are
necessarily less accurate than more objective measures. Social welfare
theory and supradecision analysis approaches are subject to inaccuracies from
motivational biases and strategic misrepresentation. Iteration and peer
review, important components of any analysis, can help to increase accuracy.
Practicality. None of the decision-aiding methods discussed here was
originally designed for application to social decisions involving risk. Each
therefore shows some strain when forced to accommodate aspects of risk that
are unlike those of the more traditional problems for which they were
designed. Cost-benefit analysis is best suited to go/no-go decisions with
immediate, predictable consequences and for which responsive markets exist
and decision makers are well informed. When attempts are made to apply cost-
benefit analysis to other kinds of problems, data requirements, especially
placing dollar values on nonmarket goods and services, cause great
difficulty. Decision analysis approaches have information requirements that,
- 241 -

-------
in theory, are more easily obtained, but they assume decision makers willing
and able to represent society. And if such persons can be found,
considerable time and effort is required of them.
There is a tradeoff in all methods between cost and accuracy. The more
the analysis is restricted by cost, the greater reliance must be placed on
the skill of experts and analysts. Consequently the less representative are
the results of overall societal preferences.
Accentability. Critics of- formal decision-aiding methods worry that
they place too much power in the hands of a technical elite. Cost-benefit
analysis is the worst in this respect, providing little opportunity for
stakeholders to contribute other than in defining the problem. It does,
however, avoid a need for subjective judgments, which makes it seem more
value-free, although in fact it is not. Methods that accept subjective
judgments are more amenable to public participation, because anyone's
perspective can be represented. In the end, however, only one person's
perspective is represented, even if he is a supradecision maker representing
the good of the whole. Decision analysis approaches highlight risk and
uncertainty. which may make them less attractive to decision makers reluctant
to admit the subjectivity and uncertainty inherent in their decisions. These
approaches are also subject to suspicions that analysts are playing
sophisticated number games.
7
SELECTION
Based on the above selection criteria it would appear that cost-benefit
analysis is inappropriate to the comparison of land and ocean disposal of LLW
because of its inability to deal easily with nonmarket goods and services and
with large differences of opinion about their values. Classical social
welfare methods are inappropriate because of difficulties in adequately
quantifying group utility functions and inability to deal with large
individual differences of opinion about values. So we are left with some
form of decision analysis or a social welfare formulation that includes
stakeholder groups. The alternatives available are:
.
Classical decision analysis;
.
Supradecision analysis; or
.
Suprastakeho1ders.
The third alternative is intermediate
approach and the social welfare approach.
must be combined by one of the methods for
welfare theory.
between the supradecision-maker
Suprastakeholders' preferences
individuals in classical social
Classical decision analysis is appropriate to the comparison of land and
ocean disposal of - LLW only if there is a decision maker in EPA who can
represent the agency and who's preferences can be quantified. This must be
determined early in the process.
- 242 -

-------
Supradecision makers can be selected to represent the country/world, or
stakeholder groups in the country/world. Selection from among these
alternatives depends on the extent to which EPA decision makers want to
include preferences of others and to understand the political implications of
alternatives in the comparison. But the success a single individual or small
group could have in quantifying the preferences of the broad range of values
represented in this decision problem is unclear.
Much depends on the nature of the alternatives that must be compared.
The more one alternative tends to dominate others with respect to
characteristics the decision makers care about, the less important are
differences among decision makers' preferences. In such a case~4 the solution.
is "robust," remaining best over a broad ran~e of preferences. The larger
the tradeoffs that must be made among cjaracteristics decision makers
consider important to the comparison, the greater the sensitivity of the
solution to differences in preferences. And the greater the need for more
politically sensitive decision-making methods.
For this reason, the suprastakeholder approach has greatest potential to
illuminate the nature of the comparison, not only with respect to the
technical issues involved, but also with respect to the political
implications of differences of opinions and values.
Note that suprastakeL.Jlders not only have different values, they may
also see (and frame) the comparison differently. So using a suprastakeholder
approach may not be just a matter of eliciting preferences of various groups
for a particular set of outcomes. Instead, the comparison should be examined
from beginning to end for each group as represented by its supradecision
maker. The results of each group analysis should then be combined in some
way by an overall decision maker, perhaps through one of the methods from
social welfare theory with supradecision makers substituted for individuals.
A distinction must be made by the overall decision maker among:
.
How stakeholders would respond to a process and its outcome from the
outside;
.
How they would respond to a process as part of the team; and
.
What outcome they would produce were they, alone the decision makers.
Stakeholders who are not part of the process engage in gamesmanship
intended to place political pressure on the decision makers, and so to enter
the process from outside. Their preferences and the decisions they make
under this condition have goals that may be unrelated to the decision problem
under evaluation. They tend to be aimed more at influencing the process than
the outcome. Some care must be exercised in formulating a multiple-
stakeholder decision problem to ensure that the proper perspective is
maintained for each group. What that perspective should be depends on the
overall decision maker's view of their position in the decision co-
decision makers or exogenous political forces to which he must respond.
This stage of the decision process must address the problem of how to
deal with differences in framing by stakeholder groups. If suprastakeholders
- 243 -

-------
are vastly different in their views of the problem, they will probably also
be much different in the way they frame it. One group might, for example,
reject out of hand particular classes of options at the initial stages of
identifying candidates for consideration. Candidate options that might score
well under other framings or value systems (other suprastakeho1ders), would
not be included in the restricted subset. Some care must therefore be
exercised to ensure that the comparisons made by stakeholder groups have a
sufficiently common basis that the results are comparable among them.
Overall decision makers can only compare the relative merits of the
alternatives with respect to stakeholders' values if all of the alternatives
are evaluated by all of the stakeholder groups (or suprastakeho1ders).
Even if the preferences and framings of the stakeholder groups are
irreconcilably different, it nevertheless is important for understanding
their political responses to try to see the problem from their point of view.
An overall decision maker need not accept that point of view as valid (or
valued) in the process, but it helps to know from where the responses come.
- 244 -

-------
REFERENCES
1.
Howard, R.A. Decision analysis:
Science, 34:679-695 (1988).
Management
practice and promise.
2.
Holling, C.S. The curious behavior of complex systems: lessons from
ecology. In H.A. Linstone and W.H.C. Simmonds, eds. Futures Research:
New Directions. Addison Wesley Press, Readding, MA, 1977. Quoted in
Merkoffer, 1987.
3.
Wildavsky, A. Statement of Professor AAron Wildavsky. Department of
Political Science, University of California at Berkeley, Berkeley,
California. In Risk/Benefit Analysis in the Legislative Process. U.S.
Senate Committee on Science and Technology, No. 71. U. S. Government
Printing Office, Washington, DC, 1980. Quoted in Merkoffer, 1987.
4.
Merkofer, M.W. Decision Science and Social Risk Management.
Publishing Co., Boston, MA, 1987.
D. Reidel
5.
Arrow, K.J. Alternative approaches to the theory of choice in risk-
taking situations. Econometrics, 19:404-437 (1951).
6.
von Neumann, J., and O. Morgenstern. Theories of Games
Behavior. Princeton University Press, Princeton, NJ, 1947.
and Economic
7.
Harsanyi, J . C. Cardinal welfare, individualistic ethics, and
interpersonal comparisons of utility. J. Political Economy, 63:309-321
(1955).
8.
social
Theory
and
welfare
functions.
Harsanyi, J.C. Nonlinear
Decision, ~:3ll-322 (1975).
9.
Social
Hirsanyi, J.C. Mortality and the theory of rational behavior.
Research, 44: 623-667 (1977).
10.
Kirkwood, C.W. Pareto optimality and
analysis. IEEE Transactions on Systems.
~:89-91 (1979).
equity in social decision
Man. and Cvbernetics. SMC-
11.
Keeney, R. L., and C. Kirkwood. Group decision making using cardinal
social welfare functions. Management Science, 22:430-437 (1975).
12.
von Winterfeldt, D., and W. Edwards. Decision Analysis and Behavioral
Research. Cambridge University Press, New York, NY, 1986.
13.
Procter, M. Probing the core of controversy: issue management in the
Office of Technology Assessment. Presented at the European Congress on
Technology Assessment, Amsterdam, The Netherlands, February 2-4, 1987.
14. Howard, R.A., 1988.   
15. von Winterfeldt, D., and W. Edwards. 1986.
16. von Winterfeldt, D., and W. Edwards. 1986.
- 245 -

-------
17.
Dyer, J. S., and R. F. Miles, Jr. An actual application of collective
choice theory to the selection of trajectories for the Mariner
Jupiter/Saturn Project. Onerations Research 24:220-244 (1976).
18.
Dyer, J.S., and R.F. Miles, Jr. A Critiaue and Ann1ication of
Collective Choice Theory: Trajectory Selection for the Mariner
Juniter/Satern 1977 Project. JPL Internal Document 900-777. Jet
Propulsion Laboratory, Pasadena, CA. February .1977.
19.
Nash, J.F.
The bargaining problem.
Econometrica, 11:155-162 (1950).
20.
Harsanyi, J. C. A bargaining model for the cooperative n-person game.
In A. W. Tucker and R. D. Luce (Eds.), Contributions to the Theory of
Games, 40:325-355 (1959).
21.
Keeney, R.L., and H. Raiffa.
Preferences and Value Tradeoffs.
Decisions with Mu1tin1e Objectives:
John Wiley & Sons, New York, NY, 1976.
22.
Rowe, M.D., and B.L. Pierce. Sensitivity of the weighting summation
decision method to incorrect application. Socio-Econ. Plan. Sci.,
16:173-177 (1982).
23.
Arrow, 1951.
24.
Rowe, M.D., and B.L. Pierce.
1982.
- 246 -

-------
Table D-l.
Comparison of decision-making theories. a
Intellectual
roots
Conceptual
basis
Method of
analysis
Perspective
on value
View of
uncertainty
THEORY
Cost-benefit
Social welfare
Engineering
economics
Economic
efficiency
Comparison of
aggregate value
of estimated
consequences
Total monetary
equivalent as
determined by
economic actors
in a free
market
Objective
characterization
of environment
Decision
Engineering,
psychology,
management
science,
economics
Axioms of
individual
choice
Determination
of logical
implications of
alternatives,
information, and
preferences of
decision maker
Responsibility
of decision
maker, objective
is consistency
Subjective
beliefs of the
individual
Welfare
economics
Axioms of
social
choice
Derivation of
group decision
from acceptable
mechanisms for
incorporating
individual
preferences
Social preference
derived from
"equitable" syn-
thesis of prefer-
ences of impacted
parties
Product of
individuals coping
with erratic
envirorunent
a Merkofer, M.W. Decision Science and Social Risk Mana~ement.
Publishing Co., Boston, MA, 1987.
D. Reidel
- 247 -

-------
SCT
P~rspecliv~: Social
Equity vs. Efficiency Conflict
Group va. Individual Conflict
CST Objective ys. Subjective Conflict
Perspective: Technical
DT
Perspective: Personal
KEY:
SWT
MAUT -
CBA
DA
Social welfare theory
Multi-attribute utility
Cost-benefit analysis
Decision analysis
theory
Figure D-l. Implicit overlaps of rationales and perspective conflicts
inherent in decision-aiding methods.
(Source: Merk~fer, M.W. Decision Science and Social Risk Management.
Reidel Publishing Co., Boston, MA, 1987.)
D.
- 248 -

-------
K
N
o
W
L
E
D
G
E
V
A
L
U
E
S
Figure D-2. Partitioning of decision-aiding
(Source: Merkofer, M.W. Decision Science
Reidel Publishing Co., Boston, MA, 1987.)
methods.
and Social Risk Management.
D.
- 249 -

-------
f_""
E,,8Iu..
Apprll..
.....'"
Figure D-3. The decision analysis process.
(Source: Howard, R.A. Decision analysis: practice and promise.
Science, 34:679-695 (1988).
Management
- 250 -

-------