0°v/t\ COMMITTEE ON EPA 542-R-01-002
THE CHALLENGES OF January 2001
MODERN SOCIETY www.clu-in.org
www.nato.int/ccms
NATO/CCMS Pilot Study
Evaluation of Demonstrated and
Emerging Technologies for the
Treatment of Contaminated Land
and Groundwater (Phase
2000
SPECIAL SESSION
Decision Support Tools
Number 245
NORTH ATLANTIC TREATY ORGANIZATION
-------
NATO/CCMS Pilot Study
Evaluation of Demonstrated and Emerging
Technologies for the Treatment and Clean Up
of Contaminated Land and Groundwater
(Phase III)
SPECIAL SESSION ON
Decision Support Tools
Wiesbaden
June 26-30, 2000
January 2001
-------
NOTICE
This report was prepared under the auspices of the North Atlantic Treaty Organization's Committee on
the Challenges of Modern Society (NATO/CCMS) as a service to the technical community by the
United States Environmental Protection Agency (U.S. EPA). The report was funded by U.S. EPA's
Technology Innovation Office. The report was produced by Environmental Management Support, Inc.,
of Silver Spring, Maryland, under U.S. EPA contract 68-W-00-084. Mention of trade names or specific
applications does not imply endorsement or acceptance by U.S. EPA.
-------
Decision Support Tools NATO/CCMS Pilot Project Phase III
CONTENTS
Introduction 1
Executive Summary 3
Opening Comments 5
Framework for Decision Support Used in Contaminated Land Management in Europe and
North America 9
Geospatial Decision Frameworks for Remedial Design and Secondary Sampling 31
Decision Support Tools: Applications in Remediation Technology Evaluation and Selection.... 42
Common Factors in Decision Making and Their Implications For Decision Support for
Contaminated Land in a Multiobjective Setting 58
Case Study: Cost Benefit Analysis/Multi-Criteria Analyses for a Remediation Project 69
Modelling the Financial Risks of Remediation 83
Decision Support Using Life Cycle Assessment in Soil Remediation Planning 92
Approaches to Decision Support in the Context of Sustainable Development 100
Summary and Conclusions 113
Country Representatives 125
Attendees List 128
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
INTRODUCTION
The Council of the North Atlantic Treaty Organization (NATO) established the Committee on the
Challenges of Modern Society (CCMS) in 1969. CCMS was charged with developing meaningful
programs to share information among countries on environmental and societal issues that complement
other international endeavors and to provide leadership in solving specific problems of the human
environment. A fundamental precept of CCMS involves the transfer of technological and scientific
solutions among nations with similar environmental challenges.
The management of contaminated land and groundwater is a universal problem among industrialized
countries, requiring the use of existing, emerging, innovative, and cost-effective technologies. This
document summarizes the special session on decision support systems from the third meeting of the Phase
III Pilot Study on the Evaluation of Demonstrated and Emerging Technologies for the Treatment and
Clean Up of Contaminated Land and Groundwater. The United States is the lead country for the Pilot
Study, and Germany and The Netherlands are the Co-Pilot countries. The first phase of the pilot study
was successfully concluded in 1991, and the results were published in three volumes. The second phase,
which expanded to include newly emerging technologies, was concluded in 1997. Final reports
documenting 52 completed projects and the participation of 14 countries were published in June 1998.
Through this pilot study, critical technical information is made available to participating countries and the
world community.
The Phase III study focuses on the technical approaches for treating contaminated land and groundwater.
This includes issues of sustainability, environmental merit, and cost-effectiveness, in addition to
continued emphasis on emerging remediation technologies. The objectives of the study are to critically
evaluate technologies, promote the appropriate use of technologies, use information technology systems
to disseminate the products, and to foster innovative thinking in the area of contaminated land.
The first meeting of the Phase III study was held in Vienna, Austria, on February 23-27, 1998. The
meeting included a special technical session on treatment walls and permeable reactive barriers. The
proceedings of the meeting and of the special technical session were published in May 1998. The second
meeting of the Phase III Pilot Study convened in Angers, France, on May 9-14, 1999, with representatives
of 18 countries attending. A special technical session on monitored natural attenuation was held. This
report and the general proceedings of the 1999 annual meeting were published in October 1999. This third
meeting was held in Wiesbaden, Germany from June 26-30, 2000. The special technical focused on
decision support tools.
This publication is the report from the special session on decision support tools. This session was chaired
by Dr. Paul Bardos from r3 environmental technology Ltd (UK) and Dr. Terry Sullivan from Brookhaven
National Laboratory (US).
This and many of the Pilot Study reports are available online at http: 7/w ww .nato. int/ccms/. General
information on the NATO/CCMS Pilot Study may be obtained from the country representatives listed at
the end of the report. Further information on the presentations in this decision support tools report should
be obtained from the individual authors.
Stephen C. James
Walter W. Kovalick, Jr., Ph.D.
Co-Directors
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
THIS PAGE IS INTENTIONALLY BLANK
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
EXECUTIVE SUMMARY
Environmental management of contaminated lands is a complex process requiring a wide variety of
decisions encompassing different technical, social, and political questions. Decision support for
contaminated land management is an emerging field. Decision support involves integration of expertise
and data, followed by analysis and interpretation of the results to produce outcomes in terms of decision
variables (health risk, cost, suitability, etc.). The decision support can be in the form of guidance that
provides a framework for performing the analysis or software that has codified the expertise to allow
more rapid analysis by many. The magnitude and similarity between contaminated land management
problems has led to development of several decision support tools (DSTs) to address different aspects of
the problem (site characterization, cost-benefit, risks, sustainable development, etc.).
Four major categories of DST use were identified during the special session discussions:
• Written guidance produced, for example, by regulatory bodies,
• Identifying sites on a regional or organizational (e.g., corporate) basis and setting management/
policy goals,
• Prioritization among different sites within a single area of responsibility,
• Using DST for specific tasks at a single site. Examples of these approaches include analysis of
human health risks, remedy selection, site characterization, and cost-benefit analysis. In most
applications, a single decision criterion is evaluated. However, use of multi-criteria analysis
(MCA) and life cycle analysis (LCA) approaches are often found.
The session had a series of invited talks on different aspects of decision support including implementation
of decision support tools. This report contains the following papers:
• Framework for decision support used in contaminated land management in Europe and North
America
• Geospatial decision frameworks for remedial design and secondary sampling
• Decision support tools: applications in remediation technology evaluation and selection
• Common factors in decision-making and their implications for decision support for contaminated
land in a multi-objective setting
• Case Study - Cost benefit analysis/multi-criteria analyses for a remediation project
• Modelling of financial risks of remediation
• Decision support using Life Cycle Assessment in Soil Remediation Planning
• Approaches to decsion support in the context of sustainable development
• Managing environmental data
In addition, two guided discussion sessions were conducted and one set of written questions was prepared
and distributed to the conference participants. Responses from the questions were analyzed and the results
were reported at the meeting. The discussion sections focused on obtaining information on the uses of
decision support tools and the strengths and limitations of these tools. The questionnaire focused on
gathering information on the use of decision support in the different countries participating in the
meeting. These discussions have been summarized in the closing paper of this report: Review of
discussions about decision support issues in Europe and North America at The NATO/CCMS Special
Session, and overall conclusions. The main findings of this discussion are as follows.
The major advantage of using appropriate DST's is that they can ensure the decision making process is
robust, consistent, transparent and reproducible. Specific advantages include:
• Providing a means of relatively easy analysis for multiple scenarios,
• Optimizing the contaminated land management process (leading to lower costs),
3
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
• Incorporating uncertainties into the decision framework to enhance the decision making process.
(This permits the decision to be based on the problem holder's aversion to failure).
• Improving communication between various stakeholder groups.
• Use as an educational tool.
• Improving the transparency of the process through documenting all parameters and assumptions
used in the analysis and explaining the approach used to reach a decision.
However, current DSTs do suffer some limitations, which affect their usefulness.
• Gaining acceptability of a DST with all stakeholders can be difficult.
• Verification/validation of DST performance can be technically challenging
• If the assumptions and data used by DST are not understood, output from the DST may be viewed
as "black box" information and may not be trusted. Proper use of DSTs requires users and
interested stakeholders receive training on the theory, application and limitations of the DST.
Decision support tools must be maintained to keep current, relevant and useful.
• Garbage In - Garbage Out: a decision support tool is only as good as the data and assumptions
used to perform the analysis
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
OPENING COMMENTS TO THE SPECIAL
SESSION ON DECISION SUPPORT TOOLS
Paul Bardos1 and Terry Sullivan2
1. r3 environmental technology Ltld, PO Box 58, Ware, SG12 9UJ, UK
2. Brookhaven National Laboratory, Upton, N.Y., 11973, USA
BACKGROUND
The Council of North Atlantic Treaty Organisation (NATO) established the Committee on the Challenges
of Modern Society (CCMS) in 1969. The CCMS was charged with developing meaningful environmental
and social programmes that complement other international initiatives in solving specific problems of the
human environment. A major activity of the CCMS is the transfer of technological and scientific
solutions and experiences among nations with similar environmental challenges. Further information
about the work of the CCMS is available on www.nalo.inl/ccrns/info.hlrn.
In 1997 the NATO CCMS adopted a proposal from the USA for a Pilot Study on treatment technologies.
It will run from 1998 to 2002, with a final report in 2003 and is under the direction of the USA, the
Netherlands and Germany.
The NATO/CCMS Pilot Study on the "Evaluation of Demonstrated and Emerging Technologies for the
Treatment of Contaminated Land and Groundwater (Phase 3) is the third in a series of Pilot Studies
considering remedial technologies. These Pilot Studies followed a Pilot Study on the problems of
contaminated land directed by the UK and Germany.
The three NATO/CCMS Pilot Studies on remediation technologies has been perhaps the foremost
international forum for the exchange of practical and research experience of remedial technologies. The
series includes:
• Phase 1, 1986 to 1991 (Martin et al, 1997; NATO, 1993; Smith et al 1998, US EPA, 1995 &
1998)
• Phase 2, 1992 to 1997 (Franzius et al, 1996, US EPA, 1998a)
• Phase 3, 1998 to 2003 (U.S. EPA, 1998b ,1998c, 1999a, 1999b, 2000).
The Phase 3 Pilot Study has attracted participation from the following countries across the world. Australia,
Austria, Belgium, Canada, the Czech Republic, Denmark, France, Germany, Greece, Hong Kong, Hungary,
Italy, Japan, the Netherlands, New Zealand, Norway, Poland, Portugal, Romania, the Slovak Republic,
Slovenia, Spain, Sweden, Switzerland, Turkey, the United Kingdom, and the United States have all been
represented at one or more meetings by a project, government representative, CCMS Fellow, or an
individual expert.
The current Pilot Study continues the theme of emerging research and technology demonstration. At each
meeting a special one-day session on a topic of particular interest for the remediation of land contamination
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
is held. In 1998 the special session was on treatment walls (US EPA1998c), and in 1999 it was on
monitored natural attenuation (US EPA 1999b).
In 2000 the topic for the special session was decision support issues. This report presents the papers of that
special session, and a summary paper of the session discussions and conclusions. It has been published by
NATO and the U.S. Environmental Protection Agency as part of an ongoing series of Pilot Study
publications. Other publications in this series are listed in the reference section.
GOALS
The aims of the report on the special session on decision support are to:
1. Provide a general understanding of Decision Support (DS) approaches used in contaminated land
remediation / risk management, their use, their features and their strengths and weaknesses, for all
the NATO delegates whatever their level of knowledge about DS (a wide range of knowledge has
been assumed from poor to expert)
2. Involve the Pilot Study in discussion and to document from this debate:
• perceived needs for and uses of DS from the perspective of end-users
• factors seen are most important in decision making
• evaluation of the strengths and weaknesses of existing DS and their use
• needs for DS development, in particular to take advantages of the opportunities for international
collaboration offered by the Pilot Study
3. Inform both the users and potential users of DS, and also DS developers of the state of the art.
APPROACH
The emphasis of the session was on the use of decision support tools for actual remediation decisions. It
considered two perspectives:
• site-specific decision making for example choosing a particular remediation system;
• remediation in terms of a risk management / risk reduction process as part of a wider process of
site management.
These were addressed both as general topics and as case studies. Case studies were included to provide
information on decision support techniques for specific contamination problems such as remedy selection.
In the case studies, the authors present the general process to provide decision support and then discuss
the application to a specific problem. The intent of this approach is to provide the interested reader with
enough knowledge to determine if the process could be used on their specific set of problems. The general
topics included broader issues that are not directly tied to a specific problem. The general topics included
papers on the role of stakeholders in the decision process and decision support approaches for sustainable
development.
Decision factors were explored from an end-user perspective, rather than what a DS developer would like
them to be. Ultimately, it is the end-user that drives the decision process. There are a range of possible
end-users, including regulators, property developers, local authorities, and specialist users. Furthermore,
national perspectives on the use of DS appear to vary. Eliciting the differences in national perspectives
was obtained through discussion and a set of questions provided to all meeting participants. The session
sought to display the state-of-the-art in decision support for contaminated land management and define
future directions in this area. Important issues pertaining to DS include:
-------
Decision Support Tools NATO/CCMS Pilot Project Phase III
• Are DS tools perceived as being useful?
• How are DS being used?
• What are the advantages and disadvantages to using Decision Support Tools (DST)?
• Are information needs for evaluating contaminated land management options understood?
It is salient to note that DS are a topic for the next call for bids for the EU Framework 5 Programme. The
US EPA "owns" a number of detailed data-sets for testing and validation of DS that may offer an
opportunity for collaboration. There could well be other R&D synergies too.
THE SESSION REPORT
While the selected set of papers is not inclusive of all work being done on decision support, it is
representative of the state-of-the-art approaches to decision support and covers the spectrum of
approaches. The first presentation sets the framework for decision support and defines key terms and
common approaches. The topics covered include data management, site characterisation and sample
optimisation, life-cycle assessment, multi-criteria analysis, evaluating financial risks to land developers,
sustainable development, and stakeholder involvement in the decision process. A range of discussion
activities took place to permit audience participation to define issues in decision support. The other papers
in this session report are as follows.
• Framework for decision support used in contaminated land management in Europe and North
America
• Geospatial decision frameworks for remedial design and secondary sampling
• Decision support tools: applications in remediation technology evaluation and selection
• Common factors in decision-making and their implications for decision support for contaminated
land in a multi-objective setting
• Case Study - Cost benefit analysis/multi-criteria analyses for a remediation project
• Modelling of financial risks of remediation
• Decision support using Life Cycle Assessment in Soil Remediation Planning
• Approaches to decsion support in the context of sustainable development
• Managing environmental data
• Review of discussions about decision support issues in Europe and North America at The
NATO/CCMS Special Session, and overall conclusions
REFERENCES
Franzius, V., Grimski, D., and Stietzel, H-J. (1996) International Pilot Studies on the demonstration of
clean-up / remediation technologies and on the reuse of former military lands. UTA International 1996
Part 1 pp 70-73
Martin, I., Visser, W., and Bardos, P. (1997) Review of policy papers presented to the NATO/CCMS
Pilot Study on Research, Development and Evaluation of Remedial Action Technologies for
Contaminated Soil and Groundwater. Land Contamination Reclamation 5 (1) 11-40.
7
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
NATO Committee on the Challenges of Modern Society (1993) Demonstration of Remedial Action
Technologies for Contaminated Land and Groundwater. Final Report. Published by the United States
Environmental Protection Agency. Report EPA/600/R-93/012 (Parts a, b and c)
Smith, M.A., James, S.C. and Kovalick, W.W. (1998) Comparative international study of soil and
groundwater remediation technologies, pp 9-18 IN Contaminated Soil '98, Proc 6th Intern FZK/TNO
Conference on Contaminated Soil, 17-21 May 1998, Edinburgh, UK. ISBN 07277 26757
Thornton, S.F., Davison, R.M., Lerner, D.N. and Banwart, S.A. (1998). Electron balances in field studies
of intrinsic bioremediation. GQ98 International IAHS conference on Groundwater Quality: Remediation
and Protection, 21-25th September, 1998Tiibingen, Germany.
United States Environmental Protection Agency (1995) NATO/CCMS Pilot Study Evaluation of
Demonstrated and Emerging Technologies for the Treatment and Clean Up of Contaminated Land and
Groundwater (Phase II). EPA Report: EPA/542/R-95/006.
United States Environmental Protection Agency (1998a) NATO Committee on Challenges to Modern
Society: NATO/CCMS Pilot Study Evaluation of Demonstrated and Emerging Technologies for the
Treatment and Clean Up of Contaminated Land and Groundwater. Phase II Overview Report No 219.
EPA Report: 542-R-98-001a,b,c.
United States Environmental Protection Agency (1998b) NATO Committee on Challenges to Modern
Society: NATO/CCMS Pilot Study Evaluation of Demonstrated and Emerging Technologies for the
Treatment and Clean Up of Contaminated Land and Groundwater. Phase III. Annual Report. No 228.
EPA Report: 542-R-98-002.
United States Environmental Protection Agency (1998c) NATO Committee on Challenges to Modern
Society: NATO/CCMS Pilot Study Evaluation of Demonstrated and Emerging Technologies for the
Treatment and Clean Up of Contaminated Land and Groundwater. Phase III 1998 Special Session
Treatment Walls and Reactive Barriers. No 229. EPA Report: 542-R-98-003.
United States Environmental Protection Agency (1999a) NATO Committee on Challenges to Modern
Society: NATO/CCMS Pilot Study Evaluation of Demonstrated and Emerging Technologies for the
Treatment and Clean Up of Contaminated Land and Groundwater. Phase III. Annual Report. No 235.
EPA Report: 542/R-99/007
United States Environmental Protection Agency (1999b) NATO Committee on Challenges to Modern
Society: NATO/CCMS Pilot Study Evaluation of Demonstrated and Emerging Technologies for the
Treatment and Clean Up of Contaminated Land and Groundwater. Phase III 1999 Special Session
Monitored Natural Attenuation. No 236. EPA Report: 542/R-99/008.
United States Environmental Protection Agency (2000) NATO Committee on Challenges to Modern
Society: NATO/CCMS Pilot Study Evaluation of Demonstrated and Emerging Technologies for the
Treatment and Clean Up of Contaminated Land and Groundwater. Phase III. Annual Report. No 244.
EPA Report: 542/R-OO/xxx.
Note: Phase 2 and Phase 3 Pilot Study reports are available on http://www, clu-in. coin and from
The are also available from the National Center for Environmental
Publications and Information in the USA (fax +1 5 13 489 8695).
-------
Decision Support Tools NATO/CCMS Pilot Project Phase III
FRAMEWORK FOR DECISION SUPPORT USED IN
CONTAMINATED LAND MANAGEMENT IN EUROPE AND
NORTH AMERICA
Bardos1, R.P.; Mariotti2, C.; Marot3, F.; and Sullivan4, T.
1. r3 environmental technology Ltd, PO Box 5 8, Ware, SG12 9UJ, UK
2. Aquater SpA (ENI Group), 61043 S. Lorenzo in Campo (PS), Italy
3. ADEME , Direction de I'Industrie, BP 406, 49004 Angers Cedex 01, France
4. Brookhaven National Laboratory, Upton, NY, 11973, USA
SUMMARY
Effective contaminated land management requires a number of decisions addressing a suite of technical,
economic and social concerns. This paper offers a common framework and terminology for describing
decision support approaches, along with an overview of recent applications of decision support tools in
Europe and the USA. A common problem with work on decision support approaches is a lack of a
common framework and terminology to describe the process. These have been proposed in this paper.
1. INTRODUCTION
The NATO/CCMS Pilot Study on Remedial Action Technologies for Contaminated Soil and
Groundwater Phase 3 is a multi-national forum for the exchange of information on emerging remediation
technologies and technology demonstration. The Pilot Study is an activity of NATO Committee on
Challenges for Modern Society (Web site: http://www.nato.int/ccms/info.htm). The Pilot Study has
decided to hold a special session on the subject, which is the third in a series of special sessions. Previous
topics were treatment walls (USEPA, 1998a) and monitored natural attenuation (USEPA, 1999).
This paper has been produced for the NATO/CCMS Pilot Study Special Session on Decision Support
(June 2000). The session was organized by Brookhaven National Laboratory (USA) and r3 Environmental
Technology Ltd. (UK) on behalf of the US Environmental Protection Agency and the Environment
Agency of England and Wales, respectively.
This paper also draws upon work carried out by CLARINET, the Contaminated Land Rehabilitation
Network for Environmental Technologies in Europe. CLARINET is a Concerted Action within the
Environment & Climate Program of the European Commission DGXII (web site: www.clannet.af).
CLARINET is a research network for soil and groundwater protection; risk assessment; remedial
technologies; and decision support issues including socio-economic and political aspects. CLARINET
includes a Working Group (WG2) specifically addressing decision support issues. WG2 has conducted an
extensive survey of CLARINET countries to review both key factors for decision support and risk
management, and to identify decision support approaches, which it is cataloguing in a Microsoft Access
database. CLARINET is also developing a range of decision support concepts and plans a web based
contaminated land information system, if funding can be secured.
2. BACKGROUND
Several billion EURO are spent in the EU, as are several billions of dollars in the USA each year on
remediation of land affected by contamination. Decision making, in the face of uncertainty and multiple
and often conflicting objectives, is a vital and challenging role in environmental management that affects
a significant economic activity. Although each environmental remediation problem is unique and will
require a site-specific analysis, many of the key decisions are similar in structure. This has led many
countries to attempt to develop standard approaches. As part of the standardization process, attempts have
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
been made to codify specialist expertise into decision support tools. This activity is intended to facilitate
reproducible and transparent decision making. The process of codifying procedures has also been found
to be a useful activity for establishing and rationalizing management processes.
The uses envisaged or desired for decision support include:
• Identifying realistic management choices;
• Integrating information into a coherent framework for analysis and decision making, discerning
key information that impacts decision making from more basic information;
• Providing a framework for transparency (i.e., all parameters, assumption, and data used to reach
the decision should be clearly documented) and ensuring that the decision making process itself is
documented.
Decision making for environmental contamination problems involves integration of knowledge from
many disciplines. There is also a range of contexts in which decisions have to be made, for example
compliance with a regulatory need, enabling redevelopment, reducing liabilities, registering and mapping
sites, and/or prioritizing use of resources. Each has its own suite of decisions. For example, consider the
suite of decisions that have to be made when considering remediation as part of a redevelopment process
for a particular site.
• In a typical analysis, the first step in the process is to collect information about the site such as
location of spills or disposal areas, the type of contamination that can be expected and the amount
of contamination (area, volume, or concentrations). Based on this information, decisions
pertaining to collection of site-specific data on the nature and extent of contamination must be
made. These types of decisions include the number, frequency, and location of samples balanced
against the cost of collecting and analyzing the samples and the value of additional data in
arriving at a more robust decision.
• Based on the initial site characterization data, interpolation, extrapolation, and other modeling
techniques are often used to estimate the contamination levels between measured data locations.
This information is often used in human health risk assessments to guide decisions on the need
for remedial action (including monitored natural attenuation). If remedial action is required,
decisions pertaining to what regions to treat and what level of remediation is technically and
financially achievable must be addressed.
• Projections of contamination levels often have a high degree of uncertainty (i.e., only a few data
points are available for estimating contamination over large regions). This uncertainty requires a
decision on whether more data is needed to better define the region requiring remediation or to
improve the remedy selection or remedy design.
• After remedial actions are complete, monitoring is often required to demonstrate the effectiveness
of the remediation. This requires further decisions on what and where to monitor, and the
duration of monitoring. A similar list of questions could be generated for other management
processes or functions, such as prioritizing development of several contaminated sites or
assessing financial risks for sustainable development.
It is unlikely that any single person will have the knowledge to perform all of the analyses required in
supporting all of the decisions pertaining to the management of land contamination. Typically, a number
of people with different areas of expertise are involved in interpreting basic information and providing it
in a form useful for others with less expertise in a given area. It is also apparent that there are many
specialist underpinning decisions (e.g., what risk levels are acceptable, what to sample, when to sample,
what technologies should be used, etc) that need to be made before general decisions on the reuse of
contaminated land can be made. Table 1 lists some of the supporting secondary decisions that need to be
made to make the overarching decision on contaminated land management. Table 1 is meant to be
illustrative rather than exhaustive.
10
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
The range of decisions and their inter-relationships lead to a great variety of decision support approaches.
CLARINET WG21 has found that these address different management problems, different segments of
each problem, and that they operate on a variety of scales and complexities, using a variety of analysis
and techniques. The broad range of decision support tools available in the USA has been reviewed by
Sullivan et al. (1997, 1999-2000), and new methods are regularly announced on the US Environmental
Protection Agency's (US EPA) "TechDirect" service2. The language used to describe decision support
methods has not been found to be consistent by these studies. A common terminology (as far as such a
thing is possible), and a general conceptual framework for describing decision support methods, would
greatly assist comparisons of methods and their applications, particularly in an international context.
Table 1. Example issues to be addressed in evaluating
remedial requirements and technologies for a site. (Bardos et al 2000)
Category
Example Issues
Risk Management
What risks may be posed by the
contamination now and in the future
(considering the sources, pathways and
receptors and the significance of any
linkages found)?
What risks may result to workers as
part of the remediation effort?
For affected aquifers: their use and
importance
How can the risks best be managed?
What are the regulatory criteria?
What are the success criteria for the
proposed remediation?
Fate of contaminants
Is there contamination entering the site
from outside?
Technical Suitability / Feasibility
What specific contamination properties
need to be addressed (e.g., free-phase
organics, concentration ranges,
speciation, sorption, toxic by-products,
etc.)?
How will remediation performance be
measured?
The availability and suitability of
existing information for the site
What time-scale is appropriate for
remediation? What is the site
availability for remediation works?
What is the size of the site? What space
is available for remediation operations?
What are the current uses of the site?
Ground conditions (materials, surface
conditions, geology)
1 Publications on this subject are forthcoming from CLARINET in the next 12 months and will be announced on its
web site: www.clarinci.at
2 Information on TechDirect is available at www.cln-in.org
11
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Technical Suitability / Feasibility (cont'd)
Stakeholders' / Third Parties' views
Sustainable Development
Costs
Does the remediation need to cope with
underground structures and/or work
under buildings?
Hydrogeology and groundwater
monitoring
Site access, security, services and
facilities
What are the adjacent properties, who
owns them and how are they affected?
How will stakeholder communication
be managed?
What impact will the remediation have
on site occupants and neighbors?
Restrictions: e.g., planning, covenants,
other contract terms, confidentialities
What impact will remediation have on
other environmental compartments and
are these acceptable (wider
environmental value)?
Wider economic value
Wider social value
Use of resources, including land
resources, for example: what in relation
to the long-term use of the site and how
this is to change
Capital and operating costs
Balance of costs to benefits / cost-
effectiveness
Funding
Restrictions: insurances, liabilities,
securities
3. WHAT CONSTITUTES DECISION SUPPORT - TERMINOLOGY
The dictionary definition of "decision" is: "the act or result of deciding; the determination of a trial,
contest or question". The dictionary definition of "support" includes, amongst other things: "to furnish
with necessaries, to provide for, to give assistance to, to advocate, to defend, to substantiate, to
corroborate". So for the purpose of providing clarity "decision support" can be defined as: the assistance
for, substantiation and corroboration of, an act or result of deciding; typically this deciding will be a
determination of an optimal or best approach. Although obvious, it is important to point out that decision
support is NOT the same as taking a decision. The actual deciding has to remain the shared responsibility
of those with a legitimate stake in the outcome of the decision, i.e., the stakeholders. Stakeholders
typically include any individuals or groups that may be affected by the environmental contamination.
Stakeholders include federal, state, and local regulators, local businesses, citizens, citizen groups, problem
holders, environmental industry, and public health officials (PCCRARM, 1997; SNIFFER, 1999).
Another important point pertaining to decision support is that it can come in the form of written guidance
or in the form of software. Written guidance is frequently provided by regulatory agencies as a means of
obtaining a standardized, reproducible approach to reaching a decision. Most regulatory agencies view
written guidance as an essential part of the approach to contaminated land management. In many cases,
12
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
this guidance is translated into computer software to assist in the calculations (e.g., risk assessment).
Software tools are also developed to assist in the decision process for computationally intensive analysis,
e.g., flow and transport, geostatistical modeling, and multi-criteria analysis.
The following words are often used in the context of decision support for contaminated land management:
map, technique, tool, tree or system, e.g., "decision support tool", "decision support system". This list is
not necessarily exhaustive, and in general, the current usage outlined in Table 2 is useful and efficient.
Table 2. Terms Used in Decision Support
Term
Contemporary Usage
Dictionary Definitions (UK)
Map
A figurative illustration of decision
processes, the route taken for a
decision
A delineation: To arrange or plan in
detail.
Roadmap
A diagram showing the major steps
in reaching a decision.
Colloquial: A detailed plan for
achieving specified objectives.
Technique
A principal, series of operations
used to assist decision making
A mode of artistic performance or
execution, a mechanical skill in art,
craft etc
Tool
A document or software produced
with the aim of supporting decision
making, i.e., something that carries
out a process in decision support
Includes anything used as an instrument
or apparatus in one's occupation or
profession
Tree
A logical progression of decision
making steps
A diagram with branching lines
System
Variable: for some people "system"
is synonymous with "tool" above,
for others "system" conveys the
entire approach to decision making,
including all its components. For
them this totality is the decision
support system, and something that
deals with just a component part
would be a "tool" rather than a
"system"
Co-ordinated arrangement; organized
combination; method; a co-ordinated
body of principles facts, theories
doctrines etc; a logical grouping; an
organized combination of things
working together performing a
particular function; any complex and
co-ordinated whole
"System" is a particularly problematic word, in that it is used to refer to both a component part of the
overarching set of decisions necessary, or the whole, both of which are in line with the dictionary
definition. However, for the purposes of clarity, it is necessary to select just one of the two alternative
meanings for "system", even although this is more limiting than English language usage. Thus, "system "
conveys the entire decision making approach, including all its components. The reasons for this selection
are that: (1) "tool" already conveys the component part definition, and (2) there are those who believe that
general rules can be drawn up for the overarching system, and not just its component parts.
4. THE PROCESS OF DECISION SUPPORT
Decision support methods codify expert knowledge and know-how into a "stored" method or process. The
"stored" process could be written guidance on how to address a problem or software that helps to analyze
the problem. When addressing a contaminated land management problem, the decision support methods
13
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
use problem specific information; with the aim of providing a concise representation of the key decision
making issues for that particular problem. Hence, decision support integrates information to produce
usable knowledge, as illustrated in Figure 1. For example, consider the decision to select between two
different remedial alternatives. The analyst would start with knowledge about the nature and extent of
contamination. This information would be used to estimate the volume requiring treatment based on the
"stored" knowledge (e.g., best practice, regulatory limits, cost data, data management and analysis
techniques including interpolation, etc.). This information could then be used as the basis for the selection
and/or design of the remedial options. For example, "stored" information on typical remediation costs
could be used to estimate likely project costs. Other knowledge such as the degree of uncertainty in the
volume requiring remediation and the reliability of the different remedial options could also be evaluated.
The decision maker would then be presented with information on costs, probability of success, and what
is being treated for the money spent to support the decision on a course of action.
problem ^. /stored general \ ^. decision
specific \expertise / knowledge
information
Figure 1. Illustration of Decision Support
Decision support methods help to make the decision making process transparent, documented,
reproducible, (hopefully) robust and provide a coherent framework to explore the options available.
Figure 2 illustrates the stages used to arrive at decision support knowledge for a typical site.
The starting point is to define the objectives for contaminated land management and the constraints on
how to manage the land. For a single site, the objective may be to remediate the land to a levels that is
acceptable for residential use. For a series of contaminated sites, the objective may be to prioritize which
sites to remediate first to minimize risks while maximizing the amount of land available for use. In both
cases, the constraints could be time, budget, technical feasibility, and public acceptability. Decision
support can then assist the identification of the optimal way to meet the objectives within the constraints.
The stages of the decision support process are confined within the dotted lines of Figure 2. Taking the
decision is illustrated as being supported by the process. The first stage in the decision support process is
to use experience and site-specific information (for example relating to the source terms, pathways and
receptors) and site-specific data (for example, soil properties and hydrology). The second stage uses this
information to develop simple conceptual models of the site behavior. The conceptual model is the basis
for the analysis (third stage in the process), which combines information on the technology being
proposed (if any) and the information used to form the conceptual model. Often all of this information is
processed in computer software. There are several reasons for the use of software. First, the sheer amount
of data in many problems favors electronic storage and manipulation. Second, the complexity of the
analysis (e.g., geostatistics, groundwater flow, and transport, human health risk assessment) requires
many calculations, which can easily be done on a computer. Third, the use of computers permits rapid
evaluation of the effects of changing parameters or scenarios. This may permit uncertainties to be
addressed. One perceived limitation of computers is that people tend to accept computer output as being
correct and therefore not examine the underlying assumptions. A caveat applies to all computer-generated
output; the output is only as good as the data and modeling assumptions used by the software.
14
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Determine Objective! suit
OMUiraliiti for Contaminated Lmd
Experience and Problem
Specific Knowledge
Figure 2. Flow chart containing the key steps in the decision support process
For example to determine the effectiveness of different remedial options, estimates of contaminant
concentrations before and after remediation may be determined through a combination of data,
geostatistical interpolation and flow and transport models. Usually this information has to be interpreted
and analyzed in terms of the decision variable (fourth stage in the process). In this example, the
contaminant concentrations can be compared to regulatory thresholds and the region that exceeds the
threshold can be defined for each remedial option. The computer software may facilitate the interpretation
and analysis, but it is the responsibility of the analyst to insure that the analysis is accurate and the output
is in a form useful for decision making.
The knowledge supplied to the decision makers (fifth stage) should be transparent and readily
understandable by different stakeholders, not just specialists. Indeed, even specialists might struggle with
the sheer volume of detail that arises from many sites, and so require some form of rational abstraction of
information into a more manageable volume and level of detail. These five stages form the basis for
decision support, which uses information abstracted from other (and often more detailed) analyses.
Decision knowledge is supplied to the decision makers, who then evaluate whether all stakeholders agree
that the information provided is sufficient to support a decision. All environmental decisions are made
with some degree of uncertainty. Complete knowledge is never available or attainable. If the stakeholders
15
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
conclude that a decision cannot be made, they may request additional data, improved conceptual models,
consideration of different technologies or refined analysis. The process of providing decision support is
repeated with the new information until a decision can be reached. In some cases, it may not be possible
to get all stakeholders to agree to an approach. When this occurs, the process may be vulnerable to
litigation.
There is an element of choice in which stakeholders to involve, from those possible (outlined in Section
9). However, some, for example, the regulator, will be an obligatory consultee. There is a difficult balance
to be drawn between who to involve and who not to involve. Involving a larger number of stakeholders in
decision making will add to the costs, complexity and duration of decision making. However, there is a
quid pro quo, in that this involvement may save future difficulties that might be caused by the reactions of
aggrieved stakeholders who were not consulted early enough.
Figure 2 also includes the idea that using models is not the same as decision support. Rather using
models, and modeling techniques and software, is a step in information collection that precedes decision
making. It is the integration of model results and their interpretation in terms of the decision variable that
supplies decision support. This is an important distinction and is made on the basis that decision support
implies making usable information available to a variety of stakeholders. A variety of stakeholders may
play a role in contaminated land decision making. For example, land owners/problem holders; regulators
and planners; site users; those with a financial connection to a site; the neighbors to a site including the
local community; the consultants, contractors, researchers and vendors involved in designing and
implementing the remediation. In some cases, advocacy groups and pressure groups may also seek
involvement. Clearly, it would be an unlucky site manager who had to defend his decision making against
all of these stakeholders simultaneously, but any decision made should be clear to them. In particular the
site owner and a busy regulator, dealing with a variety of issues, not just contaminated land, will want
reliable information that can be easily and quickly understood.
Figure 3 shows a conceptual framework for information use in decision making and emphasizes that the
"system" is the totality of the decision process. In this framework, models are not considered as decision
support, but rather as input. Tools, techniques, trees and maps can represent one or more component parts
of the decision making process, whereas a "system" supports the totality of a particular decision making
process.
Decision support input:
problem specific
information / models
Decision support tools,
techniques, maps, trees
Decision support
systems
Figure 3. Decision Support Information, Tools and Systems
Decision support exists within three broad sets of boundaries: the range of technical possibilities; the level
of detail that is appropriate and the legislation and regulations pertinent to the decision. An effective
decision support tool needs to offer options that are both technically and economically feasible and
permitted by regulators, the public and other stakeholders. In a practical sense, it is equally important that
the level of detail is appropriate. The level of detail provided to the decision-makers must be sufficiently
explanatory, but it must also be readily understood (as pointed out above). The implications of excess
detail are not only reducing the helpfulness of the decision support, but also increasing the cost of the
decision support knowledge.
16
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
5. TYPES OF DECISION SUPPORT
Contaminated land management involves a series of decisions, as management for a particular site
progresses. Decision support methods can play a role at each stage of the contaminated land management
process: as a decision support tool, for specific issues and, in the view of some commentators, over the
entirety of a management problem, as a decision support system.
Types of management problems might include: dealing with a contaminated site; prioritizing a number of
contaminated sites; or setting an overall sustainable development strategy for contaminated land
management in a particular region. For each problem-solving role, different functional applications for
decision support can be discerned. For example in managing an individual site, decision support might be
required for: site investigation, risk assessment, risk management, aftercare, monitoring, evaluating wider
impacts (environmental economic etc) and sustainability appraisal. In a broad sense, these are
management steps separated by decision making; for example an appreciation of risk (assessment) leads
to decision making for risk management. Within each management step more detailed information will be
processed by specialists, for example engineers designing and implementing a remedial system; of life
cycle assessment specialists carrying out an appraisal of the wider environmental impacts of competing
remedial systems. Translation of the outputs of their work into decision-making knowledge constitutes
the role of decision support.
6. CATEGORIES OF DECISION SUPPORT
CLARINET has been using four categories to describe decision support tools and other approaches:
• The decision making role of the approach,
• Functional application, i.e., the contaminated land management application
• The analytical techniques used in the decision support approach
• The nature of the decision support product
The decision making role describes the type of decision making being supported, e.g., for managing a
single site, or for prioritizing a number of sites. This deals with the overarching decision being made at
the site.
The functional application to contaminated land management describes whether the decision support is
for risk management, remediation, monitoring and aftercare, sustainable development etc. This deals with
the issues that must be addressed to support the overarching decision.
Several different techniques can be employed to assist environmental decision-making. Pollard et al (1990
identified the following: life cycle analysis (LCA); environmental risk assessment (ERA); environmental
impact assessment (EIA); cost benefit analyses (CBA); multi-criteria analysis (MCA); multi-attribute
analysis (MAT); environmental audit; and sustainability appraisal. In practice, many decision support
tools use several of these techniques, or mixtures of different parts of them
The nature of the product describes whether the tool is written guidance; a "map" of some sort, a series of
procedures or a software based system. In practice, a number of decision support tools (DST) address
multiple decision criteria. For example, software tools might combine risk assessment and cost-benefit
analysis techniques to generate risk maps, cost comparisons between remedial options and other decision
information.
This framework is summarized in Table 3.
17
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Table 3. Categories for Decision Support Tools
Problem
Solving Role
Identification -
of problem sites
Prioritization
Comparison -
of options
Strategy
development
- policy
- site specific
Functional Application
Problem Identification
Site investigation
Risk assessment
Risk Management
Aftercare
Monitoring
Evaluating Wider Impacts
(environmental economic
etc)
Sustainability appraisal
Analyses
Used
Risk
Assessment
Cost benefit
Life Cycle
Multi -criteria
analysis
Nature of the
Product
Written
guidance
Model
procedure
Software
In practice many DST use several analytical techniques, or mixtures of different parts of them. The most
commonly applied technique in contaminated land management is environmental risk assessment (see
Section 8). Cost benefit analysis (CBA) often in conjunction with multi-criteria analysis (MCA) is
increasingly being applied to decision making for remedial option selection once risk based objectives for
a problem site have been decided. MCA is briefly described in Appendix 1.
Interest is growing in Europe in also considering the broader impacts of remediation, in the context of
sustainable development. For example, LCA techniques have been applied to considering wider
environmental impacts in the Dutch "REC" system (NOBIS 1995a; 1995b).
MCA approaches have been considered in the UK for the same purpose. One possible qualitative
approach is to assess "wider environmental value" (WEV) in a way that makes use of the views of
different stakeholders. Three features of this approach are (i) its use of layered sets of choices to remove
potential decision making conflicts, (ii) the recording of these choices as individual rankings which are
combined to provide an overall ranking at the end of the assessment process; and (iii) and consulting
more than one stakeholder to gain a degree of objectivity in the rankings. The general assessment steps
that might be included in such a framework are presented in Table 4 (Bardos et al 2000b).
The involvement of different stakeholders (e.g., Consultant, community, regulator, problem owner) in a
consistent decision making process is increasingly seen as being important (Pollard et al 1999; ESRC
1997, PCCRARM 1997. USEPA 1995, USEPA, 1998b). Decision making also has to encompass an
increasing range of viewpoints and disciplines, not just soil science and environmental engineering but
also economic, political and social aspects. Environmental decision-making is in its infancy as a general
discipline, and so current approaches tend to be fragmented and overlapping (Pollard et al., 1999; Tonn et
al, 1999).
18
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Table 4. An approach to assessing wider environmental value.
Step
Action
Determining the objectives of the assessment
Identifying the stakeholders for consultation
Determining the scope of the assessment (i.e., which components should
be included and their basis for assessment)
Determining the boundaries for the assessment
Making a comparison of WEV for an existing shortlist of remediation
techniques (using an modified MCA approach)
Refining comparisons and testing sensitivity to changes in input values
Interpretation
7. OVERVIEW OF DECISION SUPPORT APPROACHES CURRENTLY IN USE IN EUROPE
AND THE USA
The concern over potential human health effects resulting from poor environmental practices and the
limited amount of clean land in economically desirable areas has led to the growing need to evaluate the
extent of contamination and remediate as necessary. The magnitude of these problems has caused many
countries to examine these problems on a national basis to develop priorities for sustainable development.
The management of contaminated lands must support multiple goals that are often conflicting. That is the
management decisions must be protective of human health while making appropriate use of economic
resources and supporting sustainable development.
The large number of contaminated land problems with similar characteristics has led to several attempts
to develop tools (DST) that support the wide range of decisions related to contaminated land management
and re-use. One objective of development of these tools is to obtain a consistent, reproducible and
transparent approach to supporting decisions. Another objective is to provide a consistent methodology to
compare contamination issues at different sites and serve as a basis for setting priorities.
CLARINET WG2 has found that for evaluation of contamination at a single site, there is a general
commonality of approach that is emerging internationally, albeit with some differences at the operational
level. A similar set of management tasks has been identified for dealing with land contamination, which
typically include:
a. problem identification (including historical assessment and as a result the identification of
potential sites);
b. problem investigation determination of the need for remediation;
c. risk identification (actual and potential);
d. detailed risk evaluation and the identification of the remediation goal;
e. selection and implementation of remedial measures;
f monitoring of sites following remediation.
Although these tasks have been listed sequentially, in practice efficient implementation of the process
often involves feedback and iteration between them. Recently, in the USA, there has been an emphasis on
using a three step process involving systematic planning, dynamic work planning and on-site analysis to
assist technical decision making at a contaminated site (Crumbling, 2000). In this approach, data (for
characterization or monitoring) are analyzed on-site, risk assessments are updated based on the new data,
and the need for additional samples is evaluated and the work plan is altered to reflect the most recently
available data. The approach is intended to provide a more efficient characterization and better technical
support for decision making as compared to following steps a-f in a sequential manner.
19
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
Whilst this forms the broad skeleton of many flow diagrams, the actual flow diagrams are frequently
more complex when applied to specific problems or sites. In fact, DST are often used to support all steps
of the contaminated sites management process (from investigation through remediation and monitoring),
with different DST applied to different steps or groups of steps. A few examples of these types of
applications include:
• providing a visual depiction of the extent of contamination as a means of highlighting areas of
concern (problem and risk identification);
• providing a technical basis for sample selection based on the existing data and the probability of
exceeding a regulatory limit (problem investigation);
• defining the volume of remediation required as a function of the confidence in meeting regulatory
goals (For example, one could remediate only at sample locations that are above the limit. In this
case, one would have little confidence that the entire site is clean. On the other hand, one could
remediate the entire site if any single measured value was above the limit. This would lead to
high confidence that regulatory goals were met, but would be very expensive in most cases).
• providing estimates of current and future human health risks as a function of the amount of
remediation (detailed risk evaluation);
• providing cost-benefit analysis between competing remedial technologies (selection and
implementation of remedial measures); and
Overarching decision support systems include the "Model Procedures", written guidance under
development in the UK (DETR and Environment Agency 2000). Overarching decision support
systems remain the goal of a number of decision support software development teams.
The preceding examples focused on addressing issues at a single site. DSTs are also used to address
problems at multiple sites. For example, life cycle cost analysis tools are useful to examine a range of
problems and to identify the problems with the largest life-cycle costs and the areas that lead to the
greatest costs. This can be used as one basis for identifying areas of opportunity to reduce costs.
DST has also been used to support litigation. Litigation often occurs when the responsible party is
difficult to ascertain due to complex geology or multiple sources. In these cases, DST have been used to
analyze the data using detailed technical models, abstract and interpret the model output to address the
technical questions, and present this information (often through visualization techniques) for use by a
non-technical audience (judge and jury) (Green, 2000).
To some extent, this commonality of approach in contaminated land management should not be
surprising. The nature of the basic steps of evaluation and remediation are determined by the practicalities
of contaminated site management, which of course is not country dependent. Decision making in many
countries is now increasingly seen as seeking a balance between "cost" and "benefits". 'Costs' are
increasingly seen from an environmental as well as an economic perspective. In all countries, resources
are limited so remediation work must show a clear balance of benefits over costs.
8. RISK-BASED DECISION FACTORS
8.1 Human Health
Human health risks that may be caused by contamination are becoming a primary basis for supporting
decisions on remediation throughout the EU and the USA (USEPA, 1989, USEPA, 1996a; USEPA, 1996
b; CLARINET and NICOLE, 1998, Ferguson et al 1998, Ferguson and Kasamas, 1999). In this process,
risk assessment and the subsequent step of risk management are intimately related elements that form the
basis for decisions on the fitness-for-use approach to land affected by contamination. The goal of risk
assessment is to provide an objective, scientific evaluation of the likelihood of unacceptable impacts to
20
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
human health and the environment. The goal of risk management is to support decisions on risk
acceptability for specified land uses and to determine the actions to be taken. It is the process of making
informed decisions on the acceptability of risks posed by contaminants at a site, either before or after
treatment, and how any needed risk reduction can be achieved efficiently and cost effectively (Ferguson
et al 1998, Ferguson and Kasamas 1999). In this way, the over riding needs for the protection of human
health and the environment can be clearly identified and work prioritized accordingly.
The assessment and management of land contamination risks considers three main elements, as illustrated
in Figure 4:
• the source of contamination (e.g., a solvent spill, or buried materials on a redevelopment site)
• the receptor (i.e., a part of the ecosystem that could be adversely affected by the contamination,
such as groundwater, human beings, flora and fauna)
• the pathway (the route by which a receptor could come into contact with the contaminating
substances).
A hazard exists when contamination exists: i.e., a source of toxic substances. A hazard is a situation in
which contamination in the ground has the potential to cause harm (e.g., adverse health effects,
groundwater rendered unfit for use, damage to underground structures, etc.) to a particular receptor. Risk
is commonly defined as the probability that a substance or situation will produce harm under specified
conditions. Risk is a combination of two factors, the probability of exposure multiplied by the
consequence of exposure (PCCRARM, 1997). Risk occurs when all three components are present (a
source, a receptor and a pathway for that receptor to be exposed to the toxic substances from the source).
Thus, if a hazard exists and there is a chance that a receptor will come in contact the hazardous material
through any pathway, a risk exists.
The presence of all three elements is also referred to as a pollutant linkage. Risk assessment involves the
determination and characterization of such a relationship, including for example, delineation of the
source, measurement/modeling of fate and transport processes along the pathway, and the potential effect
and behavior of the receptor. A consideration of risk must also take account of not only the existing
situation but also the likelihood of any changes in the conditions in the future.
Risk management is the art of managing environmental contamination so that the risks posed by
contamination are controlled or reduced to levels agreed upon by the regulators, problem owners, and
other stakeholders. Risks should be assessed on a site-by-site basis to ensure that a site is suitable for its
designated use.
Figure 4. A Pollutant Linkage
Pathway
Source
21
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
8.2 Ecological Risks
In the United States and Europe, there has been a recent trend to include ecological risks as a decision
variable for contaminated land management. The process of ecological risk assessment follows the same
paradigm as human health risk assessment with the exception that the receptors are the plants and animals
that inhabit the site. For example, guidance on which receptors should be considered in ecological risk
assessment (USEPA, 1997, USEPA, 2000) and how to manage ecological risks (USEPA, 1999) has been
published in the USA and the Netherlands (Ferguson et al 1998, Rutgers et al 2000). In Europe the
pollutant linkage paradigm is used to consider human health and risks to other receptors such as
ecosystems, groundwater and even buildings.
9. Other Decision Making Factors
Although human health risk is the most widely used factor to support decision making, there are a number
of other factors that impact the decision process. These include:
• Technical suitability / feasibility
• Stakeholder / Third Party views
• Costs and Benefits
• Sustainable development
Each of these is addressed below.
Technical suitability/feasibility
Suitability is closely entwined with feasibility. Suitability refers to the ability of the technical solution to
meet remedial objectives. Clearly, it is not worthwhile to attempt a remedial approach that is not suitable
for the risk management problem posed. However, a proposed solution may appear to be suitable, but is
not feasible. Factors that might cause concern over feasibility include:
• Track record of the solution for the particular environmental remediation problem ;
• Ability to offer validated performance information for previous projects;
• Expertise of the purveyor;
• Ability to verify the effectiveness of the solution when it is applied;
• Confidence of stakeholders in the solution and in its costing;
• Acceptability of the solution to stakeholders who may have expressed preferences for a favored
solution or have different perceptions and expertise.
Stakeholders
The owner of the site is not the only stakeholder in contaminated land management decisions. The
principal stakeholders in remediation are considered those with an interest in the land, its redevelopment,
and the environmental, social and financial impacts of any risk management activities. Depending on the
size and prominence of the site these stakeholders will include several of the following (Bardos et al
1999):
• Land owners / problem holders;
• Regulatory and planning authorities;
• Site users, workers, visitors;
• Financial community (banks, flinders, lenders, insurers);
• Site neighbors (tenants, dwellers, visitors);
• Advocacy organizations and local pressure groups;
• Consultants, contractors and technology vendors; and possibly
• Researchers (in some circumstances).
22
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
Each will have their own perspective, priorities, concerns and ambitions regarding any particular site. The
most appropriate remedial actions will offer a balance between meeting as many needs as possible,
including also the need to protect the environment, without unfairly disadvantaging any individual
stakeholder. Such actions are more likely to be selected where the decision-making process is open,
balanced, and systematic. Given the range of stakeholder interests, agreement of project objectives and
project constraints such as use of time, money and space, can be a time consuming and expensive process.
A diverse range of stakeholders for example, the site owner, regulators, planners, consultants, contractors,
site neighbors and perhaps others, may need to reach agreement before specific remedial objectives can
be set. Unsurprisingly once these remedial objectives are set, it may be hard to renegotiate them.
Costs and Benefits
The aim of the assessment of costs and benefits is to consider the diverse range of impacts that may differ
from one proposed solution to another such as the effect on human health, the environment, the land use,
and issues of stakeholder concern and acceptability in a common units. Deciding which impacts to
include or exclude from the assessment is likely to vary on a site-by-site basis. In many instances, it is
difficult to attach a strictly monetary value to many effects. Hence, assessments can involve a
combination of qualitative, formal CBA and MCA methods. It is also useful to include a sensitivity
analysis step, particularly where this encourages decision-makers to question their judgements and
assumptions through the eyes of other stakeholders.
Sustainable Development
The concept of sustainable development was first considered at the United Nation's Earth Summit
conference in Rio de Janeiro in 1992 A number of definitions for sustainable development have been
proposed, a widely used definitions is; "...development that meets the needs of the present without
compromising the ability of future generations to meet their own needs" (Brundtland, 1987). At a
strategic level, the remediation of contaminated sites supports the goal of sustainable development by
helping to conserve land as a resource, preventing the spread of pollution to air and water, and reducing
the pressure for development on green field sites.
Interpreting sustainable development in the context of land remediation is a complex issue and requires
guidance on specific components of the decision process, such as the environmental effect of different
types of remedial options as well as overall guidance on the whole risk management process. The
importance of the environmental effects for each option considered will be dependent on the site itself, for
example, nuisance issues (e.g., odors, dust, noise) associated with remedial options for a remote site may
be less important than for one in a city center. In addition, the significance of such effects will vary at a
local, regional and / or national level.
Combination of Decision Factors
Typically risks to human health risk and other receptors are used as a basis for setting remediation goals.
In these cases, other decision factors such as technical feasibility and cost are used to select from amongst
different remedial alternatives. In cases when the desired level of protection for receptors can not be
attained due to costs or technical difficulties in remediating the site, treatment levels are agreed upon by
the stakeholders on a case by case basis. If the risks are viewed to be large enough, extreme measures to
reduce the exposure pathway may be taken (e.g., evacuation). If the risks are only slightly above
regulatory standards cost/benefit analysis may be used to reach consensus on clean-up standards. For
example, in the U.S. there is a screening level for risk such that if the excess human lifetime cancer risk is
less than 1 part in 106, no further efforts need to be made to reduce risks. A case can be made to have risk
cleanup goals exceed 1 part in 106 if it is not technologically or economically feasible to reduce it below
this level. If the risk is too large, for example, if the excess lifetime cancer risk exceeds 1 part in 104
remedial actions are required to reduce risk.
23
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
Depending on the problem, any of these factors may become the overriding basis for making a decision.
For example, even if a technically feasible solution that protects human health and the environment to
within regulatory limits at an acceptable economic cost is available, if the stakeholders do not accept the
solution, remediation should not proceed until a solution agreeable to all parties is found. If remediation
proceeds, it is at the risk of having substantial opposition that may cause the efforts to be stopped or
modified. This can lead to greater program costs. The literature contains several examples where
decisions that were acceptable from a technical and regulatory perspective were not acceptable to all of
the stakeholders. For example, siting of new waste disposal facilities and the use of the incineration as a
treatment option have been prevented because of stakeholder concerns.
10. DIFFERENCES IN THE DECISION MAKING PROCESS BETWEEN COUNTRIES
Although there is a general commonality in approach to contaminated land management, differences in
the decision making process exists between different countries and even within different regions of the
same country. When this occurs, it is, generally because of one or more of the following:
• differences in the applications of general principles (such as which receptors are to be
considered);
• differences in the use of analytical techniques, datasets and assumptions;
• differences in priorities for environmental protection;
• differences in administrative approach;
• regional variation in characterization of land, land use, society and economy.
These differences tend to mean that decision support tools intended for an operational application are not
always directly transferable from country to country. Another important reason that DST are not always
transferable between countries is that unless the tool has received extensive documentation, application,
verification testing and peer review in the country its use is proposed in, the quality of the tool for use
there may be difficult to judge. Table 5 presents the key transferability issues, providing examples in
terms of analysis of soil or groundwater contamination. However, the major issues still apply to other
types of analysis (e.g., Life cycle analysis, multi-criteria analysis, etc). To address the issue of quality of
decision support software tools, the US EPA extensively tested six different tools on existing
environmental contamination problems as part of their Environmental Technology Verification program
(Sullivan, 1999a, b; Sullivan, 2000 a,b,c,d).
Differences in applications of general principles can, for example, include whether or not ecological
impacts are explicitly included in guideline values. Other differences include the characterization and
treatment of uncertainty in the decision process and how end uses are categorized and then considered for
risk assessment tools.
24
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Table 5. Issues in portability of Decision Support software tools
Criteria
Issue for Portability
Documentation of Models
and Assumptions
Are the model assumptions reasonable and appropriate? Analysis
of environmental problems requires conceptualization of the 'real
world' into a construct that permits analysis using a computer.
This conceptualization process involves a number of assumptions.
It is important for the models and assumptions to be thoroughly
documented to permit an evaluation of the models relevancy to
specific problems.
Multiple Lines of Reasoning
Can the model address uncertainty in data and model parameters?
The variability in natural systems makes analysis difficult. Often,
multiple approaches can be used to define the extent of
contamination. Models that can easily provide multiple
realizations of the problem can help address uncertainty issues.
Applications on Similar
Problem
Has the model been successfully used for similar applications?
Successful application of a tool on similar problems can build
confidence in the tool.
Validation/Benchmarking
Has the model been validated or benchmarked? Comparison of
model predictions with analytical solutions (validation) and
predictions of other accepted models (benchmarking) can build
confidence in the model.
Ease of Use
Is the software easy to use? Some software has features that
improve the usability of the product. For example, it is
advantageous to use software that allows data to be imported or
exported in many formats, to write scripts to perform repetitive
tasks, to generate reports to document all model parameters, and
to generate hardcopy graphics and visualizations. Software that is
easy to use is more efficient at using the analyst's time.
Training & Technical
Support
Are training and technical support available? Many of the DS
tools require specialized expertise (i.e., flow and transport
modeling, geostatistics, human health risk). Training and the
availability of technical support to address non-routine issues are
crucial for effective use of many tools.
Efficiency and Range of
Applicability
Is the model flexible enough to handle other problems that you
might encounter in the future? Some DS tools are limited to
specific problems or a narrow range of problems while others can
simulate a wide range of problems. The tool must be applicable to
the set of conditions anticipated for the analysis.
Differences in priorities for environmental protection often underpin the differences in end use
consideration. A major difference between countries is the way in which groundwater not currently in use
is considered as a resource. This can be markedly different for countries depending on their surface water
resources. More generally, while there is considerable awareness of the need to address issues of
sustainability (wider economic, environmental and social effects), these are explicitly considered only in a
limited number of cases.
Differences in regional variations include the extent to which industrialization and industrial change has
occurred, the attitude to accepting risks, differing social priorities, and the financial and technical
resources that are available to deal with any problems. Both economic factors and the attitude of society
to contaminated land problems determine the resources made available.
25
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
11. CONCLUSIONS
Contaminated land management is an important issue throughout Europe and the U.S.A. The need for
developing techniques and approaches to improve the decision making process for reuse and/or
remediation of contaminated lands is widely recognized. As a starting point, to improve communication
on this topic, the following definition is offered. Decision support can be defined as: the assistance for,
and substantiation and corroboration of, an act or result of deciding; typically this deciding will be a
determination of optimal or best approach. The decision support process integrates specific information
about a site and general information such as legislation, guidelines and know-how, to produce decision-
making knowledge with the goal of being transparent, consistent and reproducible. The complexity of
environmental remediation problems necessitates several layers of decision support including technical
decisions on sample collection (how many and where), economic decisions pertaining to are the costs
worth the benefits, and social/political decisions on sustainable land development. Each of these layers
may need to be addressed as part of the overarching decision on land management and many of these
'layers' are interdependent. In all cases, the decision support process takes basic input information
(problem definition); uses decision support tools to integrate, analyze and abstract from the information
and provides knowledge directly relevant to the decision. Approaches to contaminated land management
have been found to follow a similar broad outline independent of the country where the problem is
located.
The large number of contaminated land problems with similar characteristics has led to several attempts
to develop tools (DST) that support the wide range of decisions related to contaminated land management
and re-use. One objective of development of these tools is to obtain a consistent, reproducible and
transparent approach to supporting decisions. Another objective is to provide a consistent methodology to
compare contamination issues at different sites and serve as a basis for setting priorities. DSTs have seen
widespread use in all steps of the contaminated site management process (from investigation through
remediation and monitoring).
Contaminated land management decisions often involve a number of factors. The most widely used
decision factor is protection of human health to regulatory prescribed levels of risk. Other factors such as
technical suitability and feasibility, cost-benefits of remediation, stakeholder concerns, and long-term
sustainability may also be used in the decision process. Often human health risks are used as the basis for
setting remedial objectives. In this case, the decision often becomes what technology can meet the health
risk goals at the lowest cost while meeting stakeholder concerns. The most appropriate remedial actions
will offer a balance between meeting as many needs as possible, including also the need to protect the
environment, without unfairly disadvantaging any individual stakeholder.
Despite the similarities between contaminated land problems throughout the world, there are differences
in the approach to these problems. These include differences in application of general principles (e.g.,
some countries consider ecological risk as one basis for analysis while others do not); differences in
priorities (e.g., groundwater management is more important to countries with limited surface waters);
differences in administrative and regulatory approach; and differences in social attitudes towards risk and
the resources available for land management.
12. ACKNOWLEDGEMENTS
This work has been supported by the Environment Agency and the Department of the Environment
Transport and the Regions (UK), ADEME (France) and Aquater (ENI Group) (Italy).
The authors gratefully acknowledge the views of many from the NATO/CCMS and CLARINET
communities, in particular Ian Martin, Joop Okx and Malcolm Lowe, and the agreement of CLARINET
to use its material.
26
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
The views in this paper are the authors' and are not necessarily those of their employers or sponsoring
organizations.
13. REFERENCES
NATO/CCMS papers are available on: http://www.nato.int/ccms/info.htm. CLARINET papers for this
review are published on its web site: www.clarinet.at.
Bardos R.P. (2000) Source Management - Findings of the May 2000 NICOLE Workshop. To be
published by NICOLE, Summer 2000
Bardos, R.P., Martin, I.D. and Kearney, T. (1999) Framework for Evaluating Remediation Technologies.
Presented at: IBC's 10th Conference. Contaminated Land. July 5th, Royal MarsdenNHS Trust, London.
Pub. IBC Technical Services Ltd, Gilmoora House, 57-61 Mortimer Street, London, WIN 7TD
Bardos R.P., Kearney, T.E., Nathanail, C.P., Weenk, A. and Martin, I.D. (2000b) Assessing the Wider
Environmental Value of Remediating Land Contamination, 7th International FZK/TNO Conference on
Contaminated Soil, 18-22 September 2000, Leipzig, Germany.
Bardos, R.P., Morgan, P. And Swannell, R.P.J. (2000) "Application Of In Situ Remediation Technologies -
1. Contextual Framework." Submitted to Land Contamination and Reclamation
Brundtland. G.H. (1987) Our Common Future. World Commission on Environment and Development.
CLARINET/NICOLE (1998) Clarinet / Nicole Joint Statement: "Better decision making now." October
1998. Available from http://www.nicole.org
Crumbling, D. (2000), "Improving the cost-Effectiveness of Hazardous Waste Site Characterisation and
Monitoring," US EPA Technology Innovation Office, Washington, D.C.
http://clu-in.org/tiopersp/default.htm
Department of the Environment, Transport and the Regions and Environment Agency (2000) Model
Procedures for the Management of Contaminated Land, CLR11, Procedure for Risk Assessment,
DETR/EA, in preparation
ESRC (1997) "The Use of Models in Policy Making: Towards a Comparison and Evaluation of
Experiences." Workshop Report. 10-11 April, 1997, University of Lancaster, UK.
Ferguson, C., Darmendrail, D., Freier, K., Jensen, B.K., Jensen, J., Kasamas, H., Urzelai, A. & Vegter, J.
(1998) "Risk Assessment for Contaminated Sites in Europe. Volume 1. Scientific Basis", LQM Press,
Nottingham.
Ferguson C. C. and Kasamas, H. (1999) "Risk Assessment for Contaminated Sites in Europe. Volume 2.
Policy Frameworks." LQM Press, Nottingham. ISBN 0953 309010.
Green, S., and Delaney, T., (2000). "Show and Tell: Computer animation and modeling of subsurface and
surface conditions can aid expert witnesses testimony during contaminated litigation." Environmental
Protection, April 2000, pp. 50 - 56. (www.eponline.com).
NOBIS -Netherland OnderzoeksprogrammaBiotechnologische In situ Sanering (1995a). "Risk
Reduction, Environmental Merit and Costs." REC-Method, Phase 1. Document 95-1-03, CUR/NOBIS,
Gouda, The Netherlands, tel. +31 182 540680.
27
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
NOBIS -Netherland OnderzoeksprogrammaBiotechnologische In situ Sanering (1995b). "Risk
Reduction, Environmental Merit and Costs. REC-Method, Phase 2: A methodology based on Risk
reduction, Environmental merit and Costs." Document 95-1-03, CUR/NOBIS, Gouda, The Netherlands,
tel.+31 182540680.
PCCRARM, Presidential/Congressional Commission on Risk Assessment and Risk Management (1997)
Framework for Environmental Health Risk Management. Final Report 2 Volumes.
http://www.riskworld.com.
Pollard. S., Brookes A., Twigger-Ross C., and Irwin J., (1999). "Fragmentation. Convergence and
Harmonisation: Where Are We Going with Integrated Decision-Making?" In: facing the New
Millennium, proceedings 9th Annual SRA-Europe Conference, Rotterdam, October 10-13, 1999, SRA-
Europe.
Rutgers, M., Faber, J.H., Postma, J.F. and Eijsackers, H. (2000) Site-specific ecological risks: A basic
approach to the function-specific assessment of soil pollution. Rapporten Programma Geintereerd
Bodemonderzoek Volume 28 ISBN 72370448
Tonn, B., Turner, R., Meckling, J., Fletcher, T. and Barg, S. (1999). "Environmental Decision Making
and Information Technology: Issues Assessment." NCEDR, Knoxville, USA,. Report NECDR/99-01
SNIFFER 1999. "Communicating Understanding of Contaminated Land Risks." SNIFFER Project
SR97(11)F, SEPA Head Office, Erskine Court, The Castle Business Park, Stirling, FK9 4TR.
Sullivan, T.M., P.O. Moskowitz, M. Gitten, and S. Schaffer, "Environmental Decision Support Software,
Identification and Preliminary Review," BNL-64384, Brookhaven National Laboratory, 1997.
Sullivan, T.M., M. Gitten, and P.O. Moskowitz, "Evaluation of Selected Environmental Decision Support
Software, " BNL-64613, Brookhaven National Laboratory, 1997.
Sullivan, T.M., P. Moskowitz, and M. Gitten, "Overview of Environmental Decision Support Software,"
Toxicology in Risk Assessment. Ed. Salem H. and Olajos, E.J. Taylor and Francis, Ann Arbor, MI, 2000.
Sullivan, T., A.Q. Armstrong, A.B. Dindal, R.A. Jenkins, J. Osleeb, and E.N. Koglin (1999a).
Environmental Technology Verification Report. Environmental Decision Support Software.
Environmental Software, SitePro™ Version 3.0, EPA/600/R-99/093. (www.epa.gov/etv/verifrpt.htm)
Sullivan, T., A.Q. Armstrong, A.B. Dindal, R.A. Jenkins, J. Osleeb, and E.N. Koglin (1999b).
Environmental Technology Verification Report. Environmental Decision Support Software. ESRI,
ArcView, EPA/600/R-99/094. (www.epa.gov/etv/verifrpt.htm)
Sullivan, T., A.Q. Armstrong, A.B. Dindal, R.A. Jenkins, J. Osleeb, and E.N. Koglin (2000a).
Environmental Technology Verification Report. Environmental Decision Support Software. University of
Tennessee Research Corporation, Spatial Analysis and Decision Assistance (SADA), EPA/600/R-00/036.
(www.epa.gov/etv/verifrpt.htm)
Sullivan, T., A.Q. Armstrong, A.B. Dindal, R.A. Jenkins, J. Osleeb, and E.N. Koglin (2000b).
Environmental Technology Verification Report. Environmental Decision Support Software. DecisioTtFX,
Inc., GroundwaterFX, EPA/600/R-00/037. (www.epa.gov/etv/verifrpt.htm)
Sullivan, T., A.Q. Armstrong, A.B. Dindal, R.A. Jenkins, J. Osleeb, and E.N. Koglin (2000c).
Environmental Technology Verification Report. Environmental Decision Support Software. DecisionFX,
Inc., SamplingFX, EPA/600/R-00/038. (www.epa.gov/etv/verifrpt.htm)
28
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
Sullivan, T., A.Q. Armstrong, A.B. Dindal, R.A. Jenkins, J. Osleeb, and E.N. Koglin (2000d).
Environmental Technology Verification Report. Environmental Decision Support Software. C Tech
Development Corporation, Environmental Visualization System (EVS), EPA/600/R-00/047.
(www.epa.gov/etv/verifrpt.htm)
Tonn, B., Turner, R., Meckling, J., Fletcher, T. and Barg, S., (1999). Environmental Decision Making and
Information Technology: Issues Assessment. NCEDR, Knoxville, USA,. Report NECDR/99-01
USEPA, 1989. U.S. Environmental Protection Agency, 1989. "Risk Assessment Guidance for Superfund
Volume I Human Health Evaluation Manual (Part A) Interim Final," Office of Emergency and Remedial
Response, Washington, D.C. EPA/540/1-89/002.
USEPA, 1995. United States Environmental Protection Agency, (1995). "Guidance for Community
Advisory Groups at Superfund Sites," EPA-540-K-96-001,
USEPA, 1996a. United States Environmental Protection Agency, 1996. Soil Screening Guidance:
Technical Background Document. Office of Emergency and Remedial Response, Washington, D.C.,
EPA/540/R-95/128. (www.epa.gov/superfund/pubs.htnrfh)
USEPA, 1996 b. United States Environmental Protection Agency, 1996. Soil Screening User's Guide.
Office of Emergency and Remedial Response, Washington D.C., EPA/540/R-96/018. Second Edition.
(www.epa.gov/superrund/pubs.htnrfh)
USEPA, 1997. United States Environmental Protection Agency, 1997. Ecological Risk Assessment
Guidance for Superfund: Process for Designing and Conducting Ecological Risk Assessments. Interim
Final. EPA 540-R97-006, OSWER 92857-25, PB 963211.
(www. epa.gov/oerrpage/superfund/programs/ri sk/ecori sk/ecori sk. htm)
USEPA, 1998a. United States Environmental Protection Agency. (1998). NATO Committee on
Challenges to Modern Society: NATO/CCMS Pilot Study Evaluation of Demonstrated and Emerging
Technologies for the Treatment and Clean Up of Contaminated Land and Groundwater. Phase III 1998
Special Session Treatment Walls and Reactive Barriers. No 229. EPA Report: 542-R-98-003
USEPA, 1998b. United States Environmental Protection Agency, (1998). "Superfund Community
Involvement," EPA-540-R-98-027, OSWER 9205.5-12A, PB98-963235.
(www.epa.gov/superfund/pubs.htnrfh)
USEPA, 1999. United States Environmental Protection Agency, (1999). NATO Committee on Challenges
to Modern Society: NATO/CCMS Pilot Study Evaluation of Demonstrated and Emerging Technologies
for the Treatment and Clean Up of Contaminated Land and Groundwater. Phase III 1999 Special Session,
Monitored Natural Attenuation. No 236. EPA Report: 542-R-99/008.
USEPA, 2000. U.S. Environmental Protection Agency "Ecological Soil Screening Level Guidance
DRAFT, " Office of Emergency and Remedial Response , Washington, DC, July 10, 2000,
oy/suTO^^
29
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
APPENDIX: MULTI-CRITERIA ANALYSIS
Multi-criteria Analysis (MCA) is often used in decision making. MCA is a structured system for ranking
alternatives and making selections and decisions. Considerations are: how great an effect is (score) and
how important it is (weight). A general outline of the MCA method is shown in Figure A.I. MCA goes
one step further than a decision matrix by allowing scores to be combined into overall aggregates and
allowing scores to be weighted. With MCA, ranking and decision making processes can be made very
transparent.
I I
alternatives criteria
scores of
alternatives
on criteria
value
functions
I
weights
of criteria
valued scores
of alternatives
on criteria
total scores of
alternatives
Figure A.I. A General Outline of the MCA Method
Taken from: Bardos, R.P., Nathanail, C.P., and Weenk, A. (1999) "Assessing the Wider Environmental
Value of Remediating Land Contamination." Environment Agency R&D Technical Report P238.
Available from: Environment Agency R&D Dissemination Centre, c/o WRC, Frankland Road, Swindon,
Wilts SNF 8YF.
30
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
GEOSPATIAL DECISION FRAMEWORKS
FOR REMEDIAL DESIGN AND
SECONDARY SAMPLING
Robert Stewart, University of Tennessee, USA
University of Tennessee, 804 Volunteer Blvd, Knoxville, TN 37996
INTRODUCTION
This paper provides an overview of geospatially-based decision frameworks for remedial and secondary
sampling design strategies. These methods were generated or implemented during the construction of
Spatial Analysis and Decision Assistance (SADA), a freeware package for Windows. SADA is supported
by the Environmental Protection Agency and the Department of Energy. For more information on SADA
or on the methods described here, see http://www.sis.utk.edu/cis/sada/.
Although the remedial design frameworks are quite straightforward, they rely on geospatial and human
health risk modeling results that are beyond the scope of this paper. Therefore, this paper only presents
how output from the modeling practices can be used explicitly in a decision-making process.
Additionally, a general overview of geospatial analysis provides context for the methods. Three remedial
design approaches are presented: block scale, site scale, and site-block scale. Each framework has
important implications for both risk assessment and remedial design, and in practice each has better
defined the area of concern and ultimately saved valuable resources during the remedial process.
Similarly, the sampling designs in this paper rely on the same geospatial models. These secondary
sampling designs assume that a round of sampling has already occurred, a geospatial model has been
chosen, and a goal for taking another round of samples has been decided. Five distinct sampling strategies
are presented, each with a separate goal.
OVERVIEW OF GEOSPATIAL ANALYSIS
For this paper, geospatial analysis refers to the modeling of concentration values or uncertainty at points
that have not yet been sampled. Geospatial models estimate or predict the value of a contaminant at an
unsampled point based on nearby sample values, spatial correlation, and a number of other possible
parameters depending on the method chosen. These models are often called contouring algorithms and
include well-known methods such as inverse distance weighting and minimum tension gridding. Other
increasingly popular methods for environmental characterization include ordinary kriging, indicator
kriging, and co-kriging. These geostatistical methods contour concentration values and also provide a
model of uncertainty about those predicted concentration values through the use of spatial covariance
models. For more information on this subject, see Applied Geostatistics3, GSLIB4, or Geostatistics for
Natural Resources Evaluation5.
In most cases, a geostatistical analysis begins by defining a grid over the site.
3 Issaks and Srivasta, Applied Geostatistics, Oxford University Press, 1990.
4 Deutsch and Journal, GSLIB, Oxford University Press, 1997.
5 Goovaerts, Geostatistics for Natural Resources Evaluation, Oxford University Press, 1997.
31
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
%3°
"2604
6
Q2J7
st
2139
12.00
b.oo
10.00
i.OO
i.OO
;.oo
5.00
5.00
1-4.00
Figure 1. Defining a grid over a sampled site.
The blocks formed by this grid become the basis for remedial design and for secondary sampling
strategies later. Once the grid is in place, a spatial model is run and the site is contoured.
32
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
lit Ac-225 Slid Overlay
~^X
r>
r
~
\
-
h,
J
70
--^
n{
I,
<
'',•
XH
^
1,
(
X
^
,i
Jl
^
I
*;v
tfl
7"
I
|i
'in
u
\
t»,
^
\
f
1
1
—
''< 1
•><
?
N
— ,
<
h
i5
i •
'n
5 *
*.™j
n
xL
^f
n
«~
n
<-
«,
t
nr
4
f i ',
y
,4"
s
\
\
\
Y
30
»4.30
to.oo
36.00
32.00
28.00
!0.00
16.00
12.00
100
3.80
lli.Ac-225 Eslimates
03.80
03.90
14.30
(0.00
36.00
i2.00
!8.00
M.OO
20.00
16.00
1200
S.OO
S.80
Figure 2. Contouring a gridded site with a geospatial model.
REMEDIAL DECISION FRAMEWORKS
Given a spatial model of concentration values, three frameworks are available for determining the
remedial design: block scale, site scale, and the site-block framework. Each framework has a separate
objective for remediation, can give a significantly different result depending on the spatial distribution of
contamination, and is connected to a decision criteria. This criteria may include a cleanup concentration
goal, a human or ecological risk goal, or a state or federal guideline for maximum concentration values.
Depending on the geospatial model utilized, other goals may be a part of the criteria, including a
confidence level about the remedial design. Once a criteria is available, it is straightforward to implement
the following design strategies.
Block Scale
In the block scale framework, the decision criteria is applied to each block. In other words, each block
must pass the acceptable criteria or be remediated. Choices for the block size include the exposure unit
and the remedial unit size.
Site Scale
The site scale framework requires a region or subset of blocks to meet the decision criteria. In this case,
the blocks may be equal in size to the remedial unit if the region is itself the exposure unit. In particular, if
a representative statistical value of the blocks (e.g., average value or mean) fails to pass the acceptable
criteria, then remedial action must be taken on the region. Under this system, the blocks are ordered from
most to least contaminated. The blocks are then remediated from worst to best until the selected statistical
value is below the acceptable criteria. This is a powerful approach that operates nicely under the concept
of exposure unit within risk assessment. Under this concept, only the worst blocks are removed until the
risk to an individual or species exposed to the entire site or exposure unit area is below a target risk level.
33
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
This framework, however, may result in individual blocks exceeding the target risk value. This issue is
addressed in the site-block framework.
Site-Block Scale
In this approach, there are two decision criteria. The first is the acceptable site value and the second is the
acceptable block concentration. First, the site scale is applied to reduce the site wide exposure level to a
suitable value. Next, a review of remaining block values is performed to determine if any single block
value exceeds the maximum concentration value. If so, the block scale framework is applied until the
maximum remaining block value is less than the second constraint. From a risk perspective, this may be
the most appealing framework because the exposure unit risk is acceptable and unacceptable hot spots are
removed.
Ultimately, this framework is reduced to either the site scale or block scale framework in practice. The
site-block framework is effectively the block scale framework when the site scale fails to remediate far
enough to meet block scale requirements. Conversely the site-block framework is equal to the site scale
framework when enough blocks are remediated such that the maximum contamination remaining also
satisfies the block scale framework.
Example
Consider the following site contaminated with Arsenic. The human health risk assessment has established
that for an industrial landuse scenario, the target carcinogenic risk will be set at 1E-6, corresponding to an
exposure value 3.5pCi/g. The following figure shows the location of samples. Those enclosed by boxes
exceed the target risk limit.
Hi Arsenic, Inorganic
o
• o
o
X
Figure 3. Location of sample points exceeding acceptable risk level
Through stakeholder discussions, the value 3.5 pCi/g becomes a cleanup goal. A geospatial analysis is
performed on the site, yielding the following contour map.
34
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
O
o
O
o
p.oo
a i.oo
10.00
J.OO
5.00
7.00
3.00
5.00
t.00
Figure 4. Geospatial contour of Arsenic concentrations across the site.
At this point, any of the decision frameworks may be applied. The framework must be chosen with a
cleanup goal in mind.
In the block scale framework, any block exceeding 3.5pCi/g will be remediated. The following figure shows
the remedial design for this framework. The areas shaded in gray must be remediated.
..Arsenic, Inorganic Remediation Zones Block Scale Concentra
y
o
o
12.00
11.00
10.00
J.OO
5.00
7.00
5.00
5.00
t.OO
Figure 5. Remedial zone for the block scale framework.
Using the site scale framework, the worst areas are remediated until the site average concentration drops
below the target risk value. This corresponds to the site-wide risk dropping below 1E-6.
35
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
III Arsenic. Inorganic Remediation Zones bite Scale Concentration Basis
•5.32
%09
©5.43-
Figure 6. Remedial zone for the site scale framework.
For the site/block framework, stakeholders decided that the site-wide risk must be less than 1E-6 risk
(which corresponds to a site wide average concentration of less than 3.5pCi/g) and the contamination for
each block must be less than twice this concentration value. Thus our site scale goal is 3.5 pCi/g and our
block scale goal is 7pCi/g. The site/block framework results in the following design.
ili Arsenic, Inorganic Remediation Zones
v
/ / °
"^vL /f • ©
"^^
v /
1 C/@ o
o x>, /
V |
-,"--- O i i €>
^~^-x . !_ •
*~""~~~~~~, \ X
° '^^--xT^^,
/ X
,---
/ _
lp I2-00
\ Bl.OO
o \ rlo.oo
\~,
;
f-
0
I
o
=
=
-
J.OO
S.OO
7.00
5.00
5.00
t.OO
j
1
Figure 7. Remedial zone for the site-block scale framework.
By comparing Figure 7 with the site scale result for 3.5 pCi/g (Figure 6), the site/block framework is
reduced to the block scale with a cleanup goal of 7pCi/g.
SECONDARY SAMPLING STRATEGIES
Geospatial modeling routines open avenues into other decision frameworks in sampling design. In a
geospatially-based design, the contouring methods provide a model of concentration values or uncertainty
at each point across the site. With this result, a suite of sampling strategies is available with unique goals
for taking additional samples.
In an ideal situation, the sampling design strategies described here would select a location for a new sample,
the sample would be taken, and the result would be analyzed and put back into the model to produce the
next optimal sample location. Under this ideal situation, the following sampling strategies could drive a
36
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
sampling effort in real time. This is possible for sampling devices with quick turn around capabilities and
has been a useful option on some sites already.
I
•* Lab
-^ H '•
Figure 8. Flow for chart for using geospatial analysis to drive a sampling design in real time.
Without quick response time, the method must be able to predict the optimal location of several samples
at once. This is achieved with simulated sampling. In other words, if multiple new samples are requested,
the most optimal location is chosen first, and a modeled sample value is taken at that point and treated as
if it where a true sample. The model is rerun, and the next optimal location is chosen for the second
sample point. This is repeated until the number of samples requested is generated. Although the accuracy
of each additional request is reduced as more and more dependency is placed on modeled values, this
sampling is a valuable alternative when faced with producing a plan for multiple samples at once.
Five sampling strategies are demonstrated on the following sample site. The suggested new sample
locations are highlighted with circles. Each block center becomes a candidate for a new sample location.
lit Ac-225 Slid Overlay
1
. 1
.—--
r>
-•
r
~
-
r*
j
7n
"
'-v
fl{
\
<
j f
XH
*v
1,
(
X
^_
,i
Jl
I
»;1.
tfl
7"
I
|i
'in
u
\
•,
y
\
/
1
1
—
^ 1
Jl
y
N
— ,
<
h
^
i •
X
'n
5 *
n
i"t
^f
n
n
<-
«,
!X
«i
^n
~>
S(
,J
nf
^1
t
nr
M
^--zr>^
y
'r
s
s
\
\
Y
f
30
»4.30
to.oo
36.00
32.00
28.00
M.OO
!0.00
16.00
12.00
100
3.80
Figure 9. Gridded site used for building a secondary sampling design.
Adaptive Fill
The goal of this strategy is to fill spatial data gaps by sampling in those areas where data are far apart
relative to the rest of the set. It is easy to implement this strategy because it is not dependent on a
contouring method.
37
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Ik New Arsenic, Inorganic Samples (Adaptive Fill)
O
o
Figure 10. Adaptive fill results.
Estimate Rank
The goal of estimate rank is to place new sample locations where the highest concentrations are predicted
to be. This corresponds to a confirmation type sampling design. The result is a design that will determine
if the area has relatively high concentration values - hot spot confirmation.
Ordinary Kriging Results
Estimate Rank
(jf)
Figure 11. Concentration Contours and estimate rank results.
Variance Rank
This sampling strategy is based on the ordinary kriging method, where model variances are produced
along side concentration estimates. These two quantities define a distribution of possible concentration
values at each point. This sampling strategy will place new samples where the model variances are the
highest and will result in a sampling design that reduces modeled variability.
38
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Model Variance Map
Variance Rank Method
<£>
Figure 12. Model variances and variance rank sample design.
Percentile Rank
Geostatistical routines, such as ordinary and indicator kriging, will provide a distribution of possible
concentration values at any unsampled point. This distribution describes the magnitude and variability in
the modeled response to the sample data. In percentile rank, the goal is to identify those points with the
potential to have extremely high value due to the span of the distribution rather than those points that have
the highest predicted value (usually corresponds to the mean type value for geostatistics). This approach
is useful to sample for potential hotspots. The following figure is based on the 90th percentile.
Unfortunately, a graph of the 90th percentile map is not available.
Percentile Rank
Figure 13. Percentile rank results
Uncertainty Rank
The uncertainty rank, like variance and percentile rank, assumes a geostatistical approach to contouring
has been utilized. Uncertainty rank differs from the former methods in that it is directly connected to a
decision criteria. In uncertainty rank, new samples are placed where the model is most uncertain about
exceeding the specified criteria (i.e., Prob > Criteria ~ .50). This is primarily useful for delineating the
boundaries of the area of concern.
39
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Probability of Exceeding Industrial Ingestion Scenario Uncertainty Rank - Industrial Ingestion Scenario
C)
-
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
In the following exercise, the concept of "value of information" is demonstrated with a simple geospatial
analysis. The modeling results are shown for each round of sampling. As the number of samples
increases, the model changes less.
Iterative Sampling: Modeling Subsequently Higher resolution data.
20 Samples 30 Samples
40 Samples
50 Samples
Figure 16. Subsequent analysis result with each round of new samples.
With each new sample, less new information is provided to the process. In fact, upon visual inspection,
one can see that the difference between 40 and 50 samples if very little. The implications of using this
approach are great for remedial design. For sampling designs that intend to refine the remedial process
this approach is directly applicable.
CONCLUSION
Geospatial analyses can have an explicit and influential impact on environmental decision making
processes. Delineating information within a spatial context aids in the definition of many decision
processes and the quantification of their impact on cost and risk reduction. These impacts can result in
enormous cost savings over traditional approaches. The methods presented in this paper represent the
basic approaches in this area; more methods are being developed. All the described methods are available
in the SADA software package, which can be freely downloaded at http://www.sis.utk.edu/cis/sada/.
41
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
DECISION SUPPORT TOOLS: APPLICATIONS
IN REMEDIATION TECHNOLOGY
EVALUATION AND SELECTION
S. Vranes, E. Gonzalez-Valencia, A. Lodolo, S. Miertus, International Centre
for Science and High Technology
United Nations Industrial Development Organization (ICS-UNIDO) Padriciano
99, 34012 Trieste, Italy
ABSTRACT
Soil remediation is a difficult, time-consuming and expensive operation. A variety of mature and
emerging soil remediation technologies is available and future trends in remediation will include
continued competition among environmental service companies and technology developers, which will
definitely result in further increase in the clean-up options. Consequently, the demand has enhanced
developing decision support tools that could help the decision makers to select the most appropriate
technology for the specific contaminated site, before the costly remedial actions are taken. Therefore, a
Decision Aid for Remediation Technology Selection (DARTS) is currently being developed with the aim
of closely working with human decision makers (site owners, local community representatives,
environmentalists, regulators, etc.) to assess the available technologies and preliminarily select the
preferred remedial options. The analysis for the identification of the best remedial options is based on
technical, financial, environmental, and social criteria. These criteria are ranked by all involved parties to
determine their relative importance for a particular project.
Key words: decision support tools, multicriteria optimization, soil remediation
INTRODUCTION
Remediation of contaminated soils is a field of technology that has developed and grown recently.
Development and use of remediation technologies has progressed and a large number of clean-up
alternatives have evolved and improved over the past decade. In addition, the technology developers and
environmental service companies have sprung up in the hope of secure a place for their process in the
market. Therefore, there has been a remarkable decrease in unit cost for land treatment options.
Remediation has become affordable, allowing owners of small- and medium-sized contaminated sites to
undertake soil clean-up programs in a more cost-effective manner. However, both site owners and
environmental managers confront the challenge of making decisions to select and deploy the most
suitable soil remediation technologies to address a variety of problems and, in some cases, satisfy a
number of conflicting criteria.
These choices are increasingly more complex because a greater variety of contamination problems are
being defined and innovative technologies are becoming available every day as potential (sometimes
cheaper and/or more effective) alternatives to existing technologies (1-3). Innovative remediation
technologies, which lack a long history of full-scale applications, do not have, in some cases, the
extensive documentation necessary to make them a standard choice in the engineering/scientific
community. However, many innovative technologies have been successfully used at contaminated sites in
the United States, Canada, and Europe despite incomplete verification of their overall performances.
Some of the technologies were developed in response to hazardous waste problems and some have been
adapted from other industrial uses. Only after a technology has been used at many different types of sites,
and the results fully documented and assessed, it is commonly considered as a well-established
technology.
42
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
Decision makers are also asked to integrate information about remedial options and are required to
balance information about technology performance with limited budgetary resources and regulatory
constraints. In addition, information about the concerns of stakeholders, as well as their meaningful
involvement in the larger decision process, influences the ultimate technology selection and deployment
decision. Therefore, all involved parties (environmentalist, policy makers, local community
representatives, site owners, other stakeholders) need some tools to help them assembling and
synthesizing information to respond to these challenges and conflicting issues.
Therefore, this problem has been chosen to develop DARTS in order to perform evaluation and
comparison of technologies for environmental remediation. DARTS provides a set of criteria for
evaluating technologies to address site-specific clean-up activities, and would accomplish the following
tasks:
• To enable users to identify and systematically compare information about innovative and
conventional technologies to meet remediation goals, highlighting their strengths and weaknesses
• To establish a structure of technology evaluation and selection process, which simplifies the
decision-making and streamlines the variety of factors involved in the remediation process
• To define consistent, qualitative and quantitative indicators for key technical, environmental,
economic, and legal criteria that influence selection and deployment of technologies
• To provide documented, reproducible evaluation which can be updated as needed information
becomes available
• To provide flexible, multicriteria optimization approach allowing trade-offs among criteria on the
basis of contaminant type and site-specific needs
• To enhance communications and help focus dialogue between local community, environmental
managers and stakeholders, including regulators and policy makers
• To enable explanation and justification of the choice by offering evidence on the advantages and
disadvantage of the possible choices in a concise and consistent way.
• To fasten development of feasibility study of the remedial options
• To provide site owners, environmental managers and other stakeholders with the opportunity to
explore alternative options, etc.
Before a treatment technology can be selected for a contaminated site, detailed information about the site
and contaminants characteristics must be collected. DARTS uses this information to determine which of
the possible remedies will be capable of meeting the clean-up standards set by its users, respecting the
previously mentioned constraints (technical, economical, legal, etc.). The following section will further
clarify the role of DARTS in supporting remedial actions.
DARTS functionality
Remedial actions usually involve these main tasks:
• Site discovery, preliminary assessment, and site inspection, conducted to quickly determine if
there is a contamination problem.
• Site assessment that determines the type and extent of contamination
• Evaluation of clean-up alternatives, and selection of remediation technology, based on the type
and extent of contamination, clean-up time required, physical, and geological site characteristics,
available technologies, resource requirements, social acceptance, and compliance with federal
and state laws, etc.
• Site clean-up, application of selected remediation technique, and
• Site closure and compliance monitoring, ensuring that the identified contamination problems
have been adequately addressed
43
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
DARTS aims at providing a decision support for the evaluation of clean-up alternatives and selection of
the best available technology for the site concerned, simultaneously taking into account a number of
different technical, economic, social, legal and environmental criteria. Each criterion could be weighted
by the panel of experts, environmental managers, technology providers, policy makers, local community
representatives, to capture its relative importance in the overall balance.
There are many actors in the decision-making process, which have their own interests and perspectives.
For instance, the person(s) responsible for contamination and the owner of the polluted plot of the land
can be mainly interested in the financial issues, while the user of the plot and the direct environment,
including nearby residents, are mainly interested in health and environmental issues. Another group is
composed by representatives of public bodies. They should consider the overall interests, which includes
socio-economic and environmental issues.
The aim of DARTS is working with all mentioned actors to perform a preliminary selection of the most
efficient remediation technologies, by analyzing simultaneously some key criteria of available remedial
techniques. These criteria can be ranked by all involved parties to determine their relative importance for
a particular project.
Multicriteria analysis of all these factors determines whether a remediation strategy is a feasible, effective
and efficient solution and whether it satisfies all criteria and constraints defined by the user. Depending
on the context in which remediation technology assessment and selection are performed, the users can
tailor decision strategy balancing out various effectiveness and efficiency parameters, other criteria and
constraints. From the user's point of view, a general algorithm for DARTS analysis is described in Figure
1.
44
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
The user selects one ore more target contaminants and
defines
The program displays the classified tree of available
technologies for the specific contaminants and soils,
and the classified tree of criteria
The user rates each criterion depending how critical it is
considered (weight ranges between 0 and 100%)
From the tree of technologies the user removes
technologies or groups of technologies that doesn't want to
Multicriteria analysis is performed and recommended
technologies are shown and sorted (together with ranked
alternatives^
Yes
The user selects a remediation technology
* Soil type option in current implementation
An example of practical approach using DARTS is when applying a soil remediation decision tree (4)
(Figure 2); in this case, the user is able to analyze and compare subgroups of technologies, or to compare
technologies within the same subgroup. The technologies available in the current stage of DARTS are
indicated in Figure 2. This approach also applies for groundwater remediation programs, which can be
integrated in the remediation decision tree supported with DARTS.
45
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
SOIL REMEDIATION
EXCAVATE?
No
1
[N-SITU APPROACH
CONTAINMENT
»Landli11 capping *
»Siurry wall / barriers
TREATMENT
\
| - :
Shallow groundwafer
Yes
IN-SITU TREATMENT IN-SITU TREATMENT
»Solidifieation/9tabilization * »Bioremediation (Air
?>Soil vapour extraction * plus ex , [3 or <(> sParKmg)
»Bioremediation* »Passive treatment walls
»Phytoremediation* »Steam stripping
»Soil flushing* »Soil Hushing* plus a., [J
»Eleclrokmcticsll!
» Thermal enhanced S.V.E.* plus a . p or § OAJvTS
DARTS
Two-technologies design
{Separation and post -treatment)
» Soil Vapour Extraction*
» Soil washing*
» Solvent extraction*
» Thermal desorption*
DARTS analysis
Replace on site
Contaminated
• i i
residuaJs
DARTS
analysis
* Technologies included in trie current stage of DARTS.
= SEPARATION SYSTEMS
»Ultramembi-ane
»Micro filtration
»Ion exchange
»Carbon absorption
»EI ectro coagul ati on
KX-SITU APPROACH
~l
$ - OFF-SITE
TRliATMliNT
»High temp, incineration*
»Landfiil disposal
»Cement kilns
One-technology design i>pyrolvsis
» B io rein ediat ion
* Land farming*
* Bioreactors/microbal filters*
« Biopiles/ composting
» Dehoiogenation
» Stabilization/solidification* r\
» Chemical reduction/oxidation*
Automated segregation (radioactive
oil)*
Other innovative technologies:
Vitrification
Supercritical water oxidation
Molten metal
Plasma ARC Systems
»plasma ARC Systems
analysis
A
DARTS,
analysis
Separated waste
Figure 2. Soil remediation decision tree supported with DARTS.
CRITERIA TO ASSESS AND SELECT REMEDIATION TECHNOLOGIES
DARTS' user selects a subset of technologies in which is interested, or uses a full set of technologies and
ranked criteria (6, 7); selects the criteria, preference functions (or use default functions chosen by the
DARTS developers) and corresponding weighting factors, defines the contaminant (or a group of
contaminants), soil type and then performs a multicriteria analysis.
The criteria included in the current stage of DARTS prototype are applicability, overall cost, minimum
achievable concentration, clean-up time required, reliability and maintenance and public acceptability
(that varies depending on the country and site location); two more criteria are currently being
implemented: by-products/wastestreams post treatment required and decontaminated soil quality. A
numerical rating of 1 (= better), 2 (= average) or 3 ( = worse) is given to each technology in each category
(6). These categories are taken from the ratings reported in UN-ECE Compendiums of soil remediation
technologies (6, 7), and are mainly based on the US-EPA's evaluations. The categories are briefly
explained below.
Overall cost. Includes design, construction, operation and maintenance costs of the core process that
defines each technology. It does not include site preparation or post-treatment costs. Excavation costs of
$55/metric ton are assumed for ex situ technologies.
46
-------
Decision Support Tools NATO/CCMS Pilot Project Phase III
Ratings: 1 = Less than $ 110/metric ton;
2=$110-$330/metricton;
3= More than $330/metric ton
Minimum achievable concentration. Refers to the minimum pollutant concentration achievable by the
technology.
Ratings: 1= Less than 5 mg/kg soil;
2= 5-50 mg/kg soil; and
3= More than 50 mg/kg soil
Clean-up time required. This refers to a "standard" site of .41 hectare and 3.04 m depth. The soil mass is
18,200 metric tons.
Ratings for ex situ techniques: Ratings for in situ techniques:
1= Less than 6 months; 1= Less than 1 year;
2= 6 months -1 year; and 2= 1-3 years; and
3= More than 1 year. 3= More than 3 years.
Reliability and maintenance. Refers to the level of complexity of the system or technology, and how easy
it is to maintain.
Ratings: 1= High reliability and low maintenance;
2= Average reliability and maintenance; and
3= Low reliability and high maintenance
Public Acceptability
Degree to which the technology is acceptable to the public. This category can, of course, vary widely
depending on the country and the level of community involvement.
Ratings: 1= Minimal opposition from the community is likely;
2= Public involvement usually occurs, but the technology is generally accepted;
3= Serious public involvement is likely and the outcome is uncertain.
Other ratable criteria that will be included in the system prototype are: data needs/characterization (refers
to the extent of pre-remediation investigations) and safety requirements (refers to the measures required to
ensure safety of workers, public and environment).
MULTICRITERIA ANALYSIS ALGORITHM
A multicriteria analysis performed by DARTS is the process during which the relative merits of the
remediation alternatives are compared to each other and the most appropriate is selected from among
them for site clean-up implementation.
47
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
There are a number of fundamental problems when there are multiple objectives. For instance, consider
the case where there are a number of decision makers, each with a preference ordering over a number of
alternatives. The goal is to choose the "fair" alternative that aggregates the preferences of the decision
makers. This is an example of multiple criteria decision making (each decision-maker represents one
criterion), and those objectives need to balance in a fair way. The situation is even more complicated
when there are also multiple and even conflicting criteria like in the DARTS (where for instance,
minimizing cost and clean-up time could be conflicting requirements). The decision-maker is asked to
specify goals and relative weightings for the different criteria. Relative weightings are used to find most
preferred solutions. The weighting can be changed to assess sensitivity of solution or to reflect different
opinions.
The explicit consideration of multiple, even conflicting objectives in a decision model has made the area
of multiple criteria decision-making (MCDM) very popular among researchers during the last two
decades. It is quite conceivable that certain modifications in the existing MCDM procedures provide the
long awaited bridge between the important fields of Operations Research and Decision Support Systems.
In order to support the decision maker that must solve multicriteria problems, three kinds of methods
were essentially considered - aggregation methods using utility functions, interactive methods and
outranking methods. In our work, we adopted the last ones, actually a special outranking method, based
on extensions of the notion of criterion (5) (PROMETHEE I, providing a partial preorder, and
PROMETHEE II, providing a total preorder on the set of possible decisions).
These extended criteria can be easily defined by the decision maker, because they represent the natural
notion of intensity of preference, and the parameters to be fixed (maximum 2) have a real world meaning.
The extension is based on the introduction of a preference function, giving the preference of the decision
maker for an action a with regard to b. This function is defined separately for each criterion, where its
value is between 0 and 1 (meaning a range between 0 and 100%), within the same defined criterion. The
smaller the function, the greater the indifference of the decision maker; the closer to one, the greater his
preference. In case of strict preference, the preference function is 1. Numerous practical applications of
the PROMETHEE method have shown that it is very easily accepted and understood by the practitioners,
being the easiest approach for solving a multicriteria problem by considering simultaneously extended
criteria and outranking relations.
A preference function, Ph(a, b) , is usually presented by a function p(x) :
p(x): x -» [0, 1] and x = f (a) - / (b) ,
where /(a) and f (b) represent the values of a particular criterion, h, for actions a and b respectively.
Using a preference index, n(a, b) , we can determine the preference for a with regard to b over all
criteria:
n(a, b} =
h=l
where k represents the number of criteria, W/j is a weight for the criterion h, and Ph(a, b) is the
preference function for the criterion h.
A valued outranking graph consists of nodes represented by actions and arcs, where each arc (a, b) has a
value n(a, b) . When obtained, the valued outranking graph offers a decision -maker means for
determining a partial preorder (PROMETHEE I), or a total preorder (PROMETHEE II).
48
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
In order to rank the actions by a partial preorder, we must evaluate the outgoing flow:
where K is the set of all actions, and the incoming flow:
xeK
The outgoing flow 0+(a) describes the degree to which a dominates the other actions in K, while the
incoming flow 0~(0+(£),
aP~b if ~a>~fc;
al+b if 0+(a) = 0+(fe),
a/~fc if ~a = ~fc.
Then the partial preorder (PW, XI), ^?) can be determined by considering their intersections:
a outranks b (a P('b)
if l
a P+b and a P~ b,
aP+bandar b,
ap- b,
a is indifferent to b (a I(1) b) if al+bandal b,
a and b are incomparable (a R b) otherwise.
The net-flow: 0(a) = 0+(a) - 0 (a)
is used to rank the alternatives by a total preorder
a outranks b (a P(2) £)
[a is indifferent to & (a 7(2)
if 0(a) >
if a =
DARTS EXPERIMENTAL PROTOTYPE
A laboratory prototype of DARTS has been developed as JAVA application, using Symantec Visual Cafe
dbDE development environment.
49
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
The DARTS presents its users with a variety of configuration and input parameters from which to choose.
Several are mandatory (such as identifying technologies to be evaluated), but there are many that the user
can choose to leave blank or use the supplied default values. This way, the user decides how to tailor the
analysis to satisfy specific needs.
Prototype configuration and data entry process involves several tasks:
• Entering available technologies and their descriptions
• Entering criteria to be considered simultaneously
• Setting values of chosen criteria and selecting the type of preference function for each criterion
The application's main window (Figure 3) consists of the current state of configuration, and a few dialogs
for data entry purposes. It is connected to the database that contains previously entered information on
available technologies and selection criteria; database should be registered by the user and/or software
administrator. An application window consists of the following sections:
• Technologies tree structure
• Buttons for manipulating nodes of the technologies tree
• Criteria tree structure
• Buttons for manipulating nodes of the criteria tree
• Button for setting values of the selected criteria
• Button for starting multicriteria decision making process
• Button for selecting contaminants for multicriteria decision making process
A dialogue box requesting the user to select the technologies to be simultaneously evaluated and
compared with one another is shown in Figure 3.
50
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
1°
1°
io
SOIL CLEAN-UP TECHNOLOGIES
Thermal Treatment
Physico-Chemical Treatment
Biological Treatment
A
SOIL CLEAN-UP CRITERIA
io Contaminant type
i@
< Value >
Analyze
Select
Global criteria I
Overall cost I
Minimum Achievable Concetration I
IIP Clean-up Time Required I
Mr Reliability and Maintenance I
Public Acceptability I
Mr By products post treatment requirec(
Decontaminated soil quality I
Figure 3. Main application window
A dialogue box Technology Properties (Figure 4) is used for entering and updating information on
particular technology. It consists from a few text fields and standard OK and Cancel buttons. Main fields
are for technology identification code, name and description. Three other text fields are disabled, and they
are used for presentation of multicriteria analysis results. Radio buttons On and Off are used for
including/excluding selected technology in multicriteria analysis.
A dialogue box Criterion Properties (Figure 5) is used for entering and updating attributes of particular
criterion. It consists of several text fields, four radio buttons, and standard OK and Cancel buttons. Main
fields are for criterion identification code, name, weighting factor, function ID, i.e., he identification code
of the selected preference function for the criterion, and its parameters. Radio buttons On and Off are used
for including/excluding selected criterion in multicriteria analysis. Min and Max radio buttons show
whether selected criterion is maximized or minimized.
51
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Technology Properties
111
OK
jSoil Vapour Extract!
Cancel
On
IThis method involves the use of
lextraction wells to apply a vacuum
land create a pressure gradient in
[the unsaturated (vadose) zone. The
|vapour Is collected and processed
[by separate equipment.
Off
4.2479338842975;
2.2382920110192E
-2.0096418732782
Figure 4. Technology properties dialogue box.
Criterion Properties
® On
OMax
Public Acceptability
Ooff
• Min
70
ID;
3.0
0.0
Figure 5. Criterion properties dialogue box.
52
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
A dialogue box Criterion-Technology Properties (Figure 6) is used for entering and updating a value of
the specific criterion for selected technology.
Criterion-Technology Properties
Soil Vapour Extraction \ OK
Public Acceptability \ Cancel
11.0
Figure 6. Criterion-Technology properties dialog box.
A window Technology Criteria Overview (Figure 7) is used to overview the values of all selected criteria
for particular technology.
Technology Attributes
Soil Vapour Extraction
Criterion
[Overall cosC
! Minimum Achievable Concetration
11.0
2.0
CP Value
| Clean-up Time Required
I Reliability and Maintenance
I Public Acceptability
| By products post treatment required
I Decontaminated soil quality
2.0
1.0
2.0
1.0
1.0
Figure 7. Complete Technology criteria overview
A window Multicriteria analysis results (Figure 8) is used for the presentation of the results of
multicriteria analysis process. The best technology (with maximum net flow) is emphasized using red
color. Here, the decision has been made upon the arbitrary choice of input parameters, so the results must
not be taken seriously.
53
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Multictiteiia analysis results
__Subgroug
11 P hys i c o- C hemic a I Treatm ent IN-SITU
;2 Physicg-Chemical Treatment EX-SITU
j 3 Physico:Chemic a I Treatment EX-SITU
;4 Biological Treatment EX-SITU
;5 Biological Treatment EX-SITU
|6 Physico-Chemical Treatment IN-SITU
;7 Thermal Treatment EX-SITU
J[echnology_Narne
S o 1 1 Vap oy r Ext r act i on
Soil washing
S o i I Va poyr_ Ext r a cti o n
Land farming
Bioreactors
Soil flushing including complexation
Incineration
: 1 . 2 5 9
0.926
0.926
0.593
0.259
-0.407
-0.741
iPhysicp-Chemical Treatment EX-SITU
i Physico-Chemical Treatment IN-SITU
Solvent Extraction -1.407
Containment Systems, Barriers -1.407
"ike
Sell
OK
Figure 8. Multi-Criteria analysis results
Please note that the above results are obtained for arbitrary set of contaminants, selection criteria and their
values and preference functions. Soil Vapor Extraction Technology has been recommended as the best
choice for this random selection of input parameters. We deliberately avoided presentation of the real
world, interactive decision-making session with DARTS.
DARTS TESTING
Several tests have been made in order to verify the accuracy of DARTS results against reported real cases.
Some criteria considered by Brownfields Technology Support Center (8) for selecting and recommending
remediation technologies for the Union Pacific Railroad Site (UPRS), Clinton, Iowa, are presented in
Table 1.
In the first step, the conditions measured in the study case were translated to parameters in DARTS,
which are presented in Table 2. In order to make the DARTS analysis results comparable to those of
Brownfields, the analysis was separately performed for each group of contaminants (VOCs and heavy
metals).
54
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Table 1. Parameters obtained from Environmental Assessments performed on
UPR Site, Clinton, Iowa.
Criteria / Parameter
Applicability
Risk-based clean-up level
*
Clean-up time required
Cleaned soil availability
Description
VOCs and Petroleum
hydrocarbons in soil and
groundwater
22 ppm (benzene in soil)
0.36 ppb (benzene in gw.)
< 1 year
Arsenic in soil
Arsenic and lead in groundwater
3.8 ppm (arsenic in soil)
0.045 ppm (arsenic in gw)
50 ppm (lead in gw)
< 1 year
To be used as a light industrial and/or commercial retail area
Table 2. Criteria translated to parameters in DARTS.
Criteria / Parameter
Applicability
Minimum achievable
concentration
Clean-up time required
Description
Weight ( % )
VOCs / Hydrocarbons
100%
5-50 mg/kg (benzene)+
>50 mg/kg (other hydrocarbons)
30 - 50%
< 1 year
50%
Heavy metals
100%
< 5 mg/kg (arsenic)
100%
< 1 year
50%
+Benzene concentrations detected during Environmental Assessments were always below 22 ppm.
55
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Table 3. Comparison of recommendations made by Brownfields and
those obtained with DARTS.
Brownfields recommendation
Air sparging
Bioremediation (ex-situ)
Bioslurry (ex-situ) f
Bioremediation gw (in-situ)
Bioventing f
Chemical treatment
Dual phase extraction f
Soil flushing
Soil vapor extraction
Thermal desorption
Chemical treatment
Phytoremediation
Soil flushing
Solidification/stabilization
Solvent extraction
DARTS Multicriteria Analysis results*
VOCs /Hydrocarbons
Thermal desorption
Chemical treatment
Thermally enhanced soil vapor
extraction
Soil Vapor Extraction/ Air sparging (in-
situ)
Bioreactors
Soil Vapor Extraction (ex-situ)
Solvent extraction
Land farming
Bioremediation (in-situ)
Soil flushing
Heavy metals
Chemical treatment
Phytoremediation
Solidification/stabilization (ex or in-situ)
Solvent extraction
Containment systems / Barriers
Soil flushing
DARTS
ranked list
1.302
1.036
1.036
0.770
0.770
0.770
0.504
0.504
0.504
0.238
1.232
0.700
0.700
0.700
0.168
0.168
* "Overall cost" not included as criteria, f Technologies not available in current DARTS prototype.
DARTS results are presented and ranked in Table 3, and are compared against recommendations made by
Brownfields. Most of the technologies proposed by DARTS are included in Brownfields
recommendations, only some variations of Bioremediation (Landfarming and Bioreactors) do not match
since these technologies are grouped in DARTS as Bioremediation. Soil flushing is ranked on the last
place, because its low ability to clean until acceptable levels which DARTS considers as more than 50
mg/kg soil. UPRS case requires between 5-50 mg/kg for benzene and less than 5 mg/kg.
Some biological remediation technologies (bioremediation in-situ and landfarming) are also classified on
the last places because of their high times required to complete the clean-up (usually more than 1 year). In
the UPRS case, it was proposed a restriction of time: less than one year. If the overall cost were
considered, the bioremediation technologies would increase the ranked level because of its lower cost
compared to other thermal or physical-chemical technologies.
Coming to the end, the selection of the remediation technology is a matter of balancing out environmental
achievements against reasonable cost. Different technologies have different performance, and this holds
for technical and financial aspects as well as for environmental aspects. The aim of DARTS is to help
integrating all these aspects and make a comparative analysis of the best available technologies, taking
into account site-specific requirements, and various criteria set by environmental managers, policy
makers, site owners and other stakeholders.
Internet accessible version of DARTS is currently under construction. The client-server architecture
adopted for Internet version, assumes that all the analysis and database administration is done at the
server side, while a light client (i.e., a distant user) needs only a standard Web browser and proper
authorization to access and use the DARTS.
56
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
ACKNOWLEDGMENTS
The authors wish to express their gratitude to Vladimir Simeunovic for his precious contribution during
the development of the experimental prototype of the software. The authors also thank Branislav Opacic
for his support during the project.
SUPPORT INFORMATION AVAILABLE
Information about conventional and innovative remediation technologies is available free of charge via
the Internet at http://www.environment.gov.au/epg/swm/swtt/contents.html, http://www.frtr.gov and
http://www.epareachit.com. The Compendium of soil clean-up technologies and soil remediation
companies, edited by ICS-UNIDO and UNECE, offers three sections: soil clean-up technologies and
criteria to assess the options, list of web sites describing remediation technologies and a worldwide
directory of companies dealing with soil remediation.
Additional information about Multicriteria Decision Making (MCDM) Methodologies can be found in:
• Buchanan, John T., Erez J. Henig and Mordecai I. Henig (1998), "Objectivity and subjectivity in
the decision making process", Annals of Operations Research (Issue on Preference
Modelling),80, 333-345.
• Buchanan, John T. and James L. Corner (1997), "The effects of anchoring in interactive MCDM
solution methods", Computers and Operations Research, 24(10), 907-918.
• Corner, James L. and John T. Buchanan (1997), "Capturing decision maker preference:
Experimental comparison of decision analysis and MCDM techniques", European Journal of
Operational Research, 98(1), 85-97.
• Henig, Mordechai I. and John Buchanan (1997), "Tradeoff directions in multiobjective
optimization", Mathematical Programming, 78(3),357-374.
• Buchanan, John (1997), "A naive approach for solving MCDM problems: The GUESS method",
Journal of the Operational Research Society, 48(2), 202-206.
• Henig, Mordechai I. and John T. Buchanan (1996), "Solving MCDM problems: Process
concepts", Journal of Multi Criteria Decision Analysis, 5(1), pp. 3-12.
LITERATURE CITED
(1) USEPA, Remediation and Characterization Innovative Technologies (REACHIT), EPA's
Technology Information Office, 2000.
(2) FRTR, Remediation Technologies Screening Matrix and Reference Guide, Third edition, Federal
Remediation Technology Roundtable, 1997.
(3) CMPS&F Environment Australia, Appropriate Technologies for the Treatment of Scheduled
Wastes, Technology Reviews, Review Report Num. 4, 1997.
(4) Freeman, H.; Hazardous waste treatment and disposal, Castaldi, F., McGraw Hill, 1997.
(5) Brans, J.P. and Vincke, P., A Preference Ranking Organisation Method, Management Science.
1985, 31, 647-656.
(6) UNECE, Compendium of soil clean-up technologies and soil remediation companies, first
edition, New York and Geneva, 1997.
(7) ICS-UNIDO and UNECE, Compendium of soil clean-up technologies and soil remediation
companies, second edition, New York and Geneva, 2000.
(8) Brownfields Technology Support Center, Technical Assistance for the Union Pacific Railroad
Site, Clinton, Iowa, 1999.
57
-------
Decision Support Tools NATO/CCMS Pilot Project Phase III
COMMON FACTORS IN DECISION MAKING AND
THEIR IMPLICATIONS FOR DECISION SUPPORT FOR
CONTAMINATED LAND IN A MULTIOBJECTIVE SETTING
J.P. Okx
Tauw BV, Research & Development,
P.O. Box 133, 7400 AC Deventer, The Netherlands
E-mail: jpo@tauw.nl
ABSTRACT
Arriving at the best soil remediation alternative involves a decision process. Tools can support some of
the routines within the decision process. Support, however, is not the same as taking a decision. The
actual deciding is not provided by a tool, but remains the shared responsibility of the stakeholders.
Decision processes can be seen as goal-oriented systems. The informationless paradigm - giving it a try -
is the least powerful in achieving a high performance. The feedback paradigm performs better since it
allows learning from experience. The feedforward paradigm allows anticipation. The full-information
paradigm combines the benefits of both the feedback and the feedforward paradigm. Although we like to
focus on the decision process our attention should be stretched to the resulting remedial actions and
feedbacks and feedforwards should consequently enable those involved to anticipate and learn from the
results of the actions. Different stakeholders often have different objectives and, thus, their preference for
remedial alternatives may differ. Nevertheless multiattribute models are useful tools when trying to make
a decision. Three decision support tools are discussed in terms of their role in the decision process, the
way they allow for feedback and feedforward, and, the way they support decision making in a
multiobjective setting.
ONCE UPON A TIME ...
Quite a few years ago I had to visit one of our offices in Germany. When I arrived from the Netherlands I
found a rather depressed team. They just received some very negative comments on what they considered
as an almost perfect report on a chlorinated hydrocarbon problem. Still unaware of the nature of the
comments I started reading the report, and after a few hours I had to admit that it was an excellent report.
It gave very complete and detailed insight in the situation, and I could not detect the flaws. I did some
more talking with the guys involved, to discover what could be wrong, but it was not very helpful. After
that I called the problem owner, made an instant appointment, took my bike, and cycled across the Rhine-
bridge to pay him a visit. The problem owner - an experienced manager - told me that it was the first
time that he was confronted with contaminated land issues. After a half-hour the problem was a lot
clearer: the problem owner expected a report listing the complete consequences for his company, and all
he got was an expensive report that stated that his site was contaminated! It took me another hour to fill in
the communication gap completely. At the end of our session we made another appointment to discuss the
remedial options, and I managed to convince him to involve the (city) authorities. Back at the office I
explained what went wrong, and they were happy that things had been solved. Since I promised the
problem owner to invite the authorities, the next thing I did was making a phone-call to the city hall. Very
soon after I got the responsible guy on the telephone I knew that I was talking to a fundamentalist:
"Multifunctionality! No compromises to the environment!" I started hating myself to involve the
authorities as early as I did. We had one week to prepare the meeting. My team members worked out a
number of alternatives: the geohydrologist had worked out a geohydrological containment option, our
civil engineer had worked out something that would put 'his' Caterpillars and Komatsus at work, and our
chemical engineer came up with an innovative in situ alternative in which phenol was proposed as a co-
substrate for chloroethene degrading bacteria. We just had success with such an alternative in the
Netherlands, and so I asked our chemical engineer to prepare for the meeting.
58
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
Finally, we all met to discuss the options. Everybody had spelled every letter of the proposed alternatives.
Within three minutes the situation was clear: the problem owner made his choice for the containment, the
cheapest solution; the authorities made their choice for the excavation, the safest but most expensive
solution; and our chemical engineer expressed his preferences for the smart and elegant in situ alternative.
The authorities made it perfectly clear that infiltration of things like phenol were forbidden. By law! I
tried to start a discussion, but after half an hour everybody had disappeared in his own trench. Within the
hour I had managed to succeed in three things: the problem owner - our client - was confronted with the
highest remediation costs possible, the local authorities would never accept any Tauw alternative for the
next decade, and my colleagues would never take any communication lesson from me again. This time
depression was mine ...
Three weeks later we had the next meeting. By that time I had spend many hours doing my homework,
and lost my day and night rhythm completely. For each of the alternatives I had figured out the effects
from every possible perspective. My reasoning, which was based on a newly developed method for the
comparison of remedial alternatives, pointed out that the cheapest alternative - the geohydrological
containment - was save, required a lot of energy and renewable sources, but solved little. Moreover, the
alternative could give problems in case of a future take-over of the site. The safest option - excavation -
was very expensive and required a huge amount of energy and renewable resources. In terms of
environmental merits the alternative had a negative score. The in situ alternative was not the cheapest, but
still a lot closer to the price of the containment, than to that of the excavation. It would not yield quite the
same reduction of pollutant concentration as the excavation, but it would be very close to that. In terms of
environmental merit it was outperforming the other alternatives completely: low energy use, the use of
renewable resources was negligible, and so on. Last but not least: with the predicted results, the owner
could sell the site without any problem, and phenol as co-substrate could be replaced by something more
acceptable. To make a long story short: we were asked to work out the in situ alternative; it was
implemented; it took one year longer than predicted; and after four years we all celebrated the completion
of the project, and we all agreed to publish our bloopers.
The story above never happened to me as a whole. It was not in Germany. I never did meet such ugly
caricatures, and I was most certainly not always the hero I liked to be! Yet, I lived through most of the
scenes that I have combined above: they are scenes of the profession collected all over the place, and I
feel that the whole story could have happened to me. Fortunately, caricatures such as described do not
exist, and they never met each other, but some people do at least remind me of these caricatures, and they
could have met.
There are patterns in decision making, in the way we anticipate, in the way we learn, and in the way
people interact when defending their own interest. Such patterns are described within alien sciences such
as systems science and management. Having the made-up story at hand makes it easier for me to explain,
and that is exactly why I started as a storyteller.
DECISION MAKING PROCESSES
Arriving at the best soil remediation technique involves a decision process. A general model for decision
processes (Figure 1) is given in Mintzberg et al. (1976). The seven central routines in the figure can be
linked to the three main phases of decision-making: problem identification, development of problem solving
alternatives and selection of the best alternative. The identification phase consists of the central routines:
recognition, in which the problem is recognised and evokes decisional activity and diagnosis, in which the
decision makers seek to comprehend the evoking stimuli and determine the cause-effect relations for the
decision situation. In our made-up story the problem owner recognised that he was having a problem, and
asked a number of experts to work it out. The experts mistook this request for a diagnostic one. They used
all their skills and tools such as geographical information systems, geohydrological models, laboratory
experiments to provide the initially unwanted diagnosis. The development phase contains a search routine to
find ready-made solutions and a design routine to develop tailor-made solutions. This was what the problem
owner expected, but did not get at first. The selection phase contains a screen routine to reduce the number
59
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
of generated ready-made solutions, an evaluation/choice routine, which operates in three different modes -
judgment, bargaining and analysis - and an authorisation routine to obtain approval. This is what happened
after the first disastrous meeting. We used an analytical tool to compare the different remedial alternatives,
and bargaining and personal expert judgment did the rest. These phases and routines can also easily be
identified in most guidelines for contaminated soil (Gotoh and Udoguchi, 1993; Dreschmann, 1992;
Eikelboom and Von Meijenfeldt, 1985).
Interrupts - act so as to prevent from proceeding continuously - may occur in the process, originating from
the decision environment, and can delay, accelerate, stop or restart the decision process. Interrupts are
caused by disagreement on the need to make a strategic decision. Internal and external interrupts are
common in soil remediation and are related to the nature of the strategic decision. In our made-up project we
faced an internal interrupt when the negative comments of the problem owner could not be processed by our
experts. We faced an external interrupt when the stakeholders could not agree on a solution, and
consequently could not make a decision. New option interrupts occurring when the decision scope is
suddenly broadened by technological development or changing policy are less common, but may occur in
cases of considerable timelag between authorisation and the final realisation of a project.
Figure 1. A general model for decision processes (Mintzberg et al., 1976)
Seven types of decision processes according the path taken through the Mintzberg's model are identified
(Mintzberg et al., 1976; Nutt, 1984; Janssen, 1992). Only two of these will be referred to in this paper:
1. Modified search decision processes consisting of finding and modifying ready-made solutions.
2. Dynamic design decision processes, involving complex search and design cycles. These are the
most complex decision processes.
For most remediation problems ready-made solutions do not exist. Therefore, the search and screen
routines are always followed by a design routine in which ready-made in situ concepts are modified into
solutions. Thus, selection of a remedial technology then should be considered as a modified search
decision process, characterised by a development routine in which in situ concepts are modified into
tailor-made solutions.
In some cases the development routine involves complex search and design cycles and encounters
multiple interrupts. This corresponds to a so-called dynamic design decision process.
60
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Tools can support some of the routines within the decision process. This kind of support is defined by
Bardos et al. (2000) as: the assistance for, and substantiation and corroboration of, an act or result of
deciding; typically this deciding will be a determination of optimal or best approach. More important
than the definition is remark that decision support is not the same as taking a decision. The actual
deciding remains the shared responsibility of those with a legitimate stake in the outcome of the decision,
i.e., the stakeholders (Bardos et al., 2000).
1. INFORMATION PARADIGMS
Decision processes can be seen as goal-oriented systems (Klir, 1991). The most primitive paradigms of
goal-oriented systems are usually conceived as structure systems with two elements. One of the elements
is a system in terms of which the goal is defined. It is usually called a goal-implementing element A.
You could call it the planned action. The other element, which is called the goal-seeking element B,
generates states of a goal-seeking variable. It is equal to the actual experience. By using experience as an
additional input to the goal-implementing element, its performance with respect to the goal increases.
Block diagrams of four paradigms are displayed in Figure 2.
In formationless
paradigm
Feedback
paradigm
ALHOB ALN-DB A
Feedforward
paradigm
Full-in formation
paradigm
Figure 2. Four information paradigms
The informationless paradigm - act and hope for the best - is the least powerful in achieving a high
performance. An example: people trying to sell standard solutions are often working this way. Whatever
the situation is: they do not want to know about the specific situation, because they cannot anticipate and
change the solutions they sell. And, as long as they cannot change the product, the performance of their
solution is of no importance to them.
The feedback paradigm is less restrictive since it allows utilising information about the output variable.
In plain English: it allows you to learn. Methods based on this paradigm are well developed. In the made-
up story the experience from another project was used to develop the 'winning' alternative. Thus, the
paradigm allows to learn and react, but does not provide for anticipation.
The feedforward paradigm provides the possibility to anticipate future events. In our made-up story we
should have anticipated the reaction of the stakeholders involved.
However, both feedback and feedforward paradigms are inferior to the full-information paradigm. The
full-information paradigm is the combination of the feedback and the feedforward paradigm.
Note: anticipation as well as learning can be applied to the reaction of stakeholders, as well as to physical
phenomena such as the reaction of physical the contaminated soil to remedial actions. Watzlawick et al.
61
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
(1967) - a classical work for psychologist - give some fine examples of patterns/models related to human
interaction.
If A is a good and deterministic model of the corresponding real phenomenon, then the feedforward
mechanism may give better results than the feedback mechanism. If A is not a good model, then the
feedback mechanism may lead to a higher performance (Klir, 1991). Thus, the feedback and the
feedforward paradigms are not comparable and their performance depends on the circumstances.
So much for systems science, let us talk business again. Within the development phase of decision
process we can try - for instance by involving the stakeholders - to anticipate to what might happen in the
selection phase. Moreover, in an adaptive decision process such as the modified search or the dynamic
design decision process each iteration allows including the 'lessons learned' from the selection phase.
Thus, the decision process allows for the full-information paradigm. Although this is a promising start,
we should not be satisfied. Our target - or goal-seeking element - is not the decision process, but the
effect and the efficiency of the resulting action. After all in our made-up project we would not have been
satisfied by an authorised failure! This implies that the decision process is only a part of the goal-oriented
system. For full optimisation the system should stretch out to the resulting actions and feedbacks and
feedforwards should consequently enable those involved to anticipate and react to the results of the
remedial actions. This requires remedial designs that can be operated adaptively. If we do not provide this
type of adaptive designs, then we cannot change the outcomes of our action, and the only option is to use
the experience for some future project.
2. THE MULTIOBJECTIVE SETTING
The majority of decision situations in soil remediation share important similarities. First, stakeholders
evaluate a set of remedial alternatives, which represent the possible choices. The objectives to be
achieved drive the design (or screening) of alternatives and determine their overall evaluation. Clearly the
stakeholders in the made-up project do have different objectives. Attributes are the measurement rods -
the decision makers ruler - for the objectives and specify the degree to which each remedial alternative
matches the objectives. We have used these attributes in the second stakeholder meeting of the made-up
project to show that the in situ alternative came very close to the objectives. Finally, factual information
and value judgments jointly establish the overall merits of each option and highlight the best compromise
solution (Beinat, 1997). Figure 3 summarises the information that plays a role in a multiattribute model.
The information items are the multiattribute profiles (Ah..,Am) allowing measurement of the achievements
of the (remedial) alternatives, the value functions (v,, i=l,..,n) representing human judgments, the weights
(Wi, i=l,..,ri), and the multiattribute value function that associates an overall value with each alternative
value functions weights
Multiattribute profiles
additive
representation
Am-(Xlm>--»Xnm) '
overall values
Figure 3. Information items in a multiattribute model (Beinat, 1997)
62
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
In this example, the overall merit of a decision alternative is computed as a weighted sum of single-
attribute performances regarding all attributes. Although this evaluation scheme is very common and
widely used, it is important to stress that it can be applied only under very precise conditions. Without
going into this topic (see Beinat 1997 for an overview), it is sufficient to say that the additive rule can be
applied only if independence conditions across attributes are met. This, in turn, calls for a careful
structuring of the decision problems and a careful choice of the attributes.
Simple enough! Unfortunately different stakeholders often have different objectives and, thus, their
multiattribute profiles, their functions and their weights are not identical. In our made-up project the
stakeholders had different objectives, and their multiattribute profiles, functions and weights were
different. Calculating an overall value for one stakeholder is possible, calculating one overall value for
more than one stakeholder is only possible if objectives, profiles, functions and weights are identical.
Consequently, what you normally get is an overall value or a score for a proposed alternative seen
through the eyes of a particular stakeholder. Agreement in our project is not only reached by the fact that
the objective attribute scores were more identical than expected, but also because we managed to unify
the multiattribute profiles a little. In other words: the stakeholders gradually agreed upon the selection
criteria.
Not all of the attributes can be expressed in the form of some number and often these attributes will not be
explicit. In other words, the overall quality of an alternative within a specific decision context is a
function of the explicit and implicit attributes. This function can in turn be explicitly or implicitly used for
negotiations between actors. In formula:
v(Aj) = / (wlvl (xl}),.., wnvn (xn]), implicit values)
Where v(A) is the quality of alternative A; wiVi (x^), .., wnvn (xnj) the explicit values for that alternative;
and/is the decision rule, explicitly or implicitly used. On this basis, there are several possible approaches
to the decision on the basis of the multiattribute model outcomes. They can be broadly classified as shown
in Table 1.
Table 1. Possible uses of the multiattribute model outcomes
Multiattribute model outcomes are
sufficient to make a decision
Other factors contribute to the
decision
The decision rule is
explicitly known as a
function of
attributes.
The decision rule is not
made explicit.
(1) The alternatives are evaluated by
applying the decision rule and are
ranked from best to worst and the
decision is reached through analysis
and evaluation.
(3) The evaluation of alternatives is
based on the multiattribute model
outputs, but the pros and cons are
discussed between stakeholders and
the decision is reached through expert
judgment or bargaining and
negotiation.
(2) The alternatives are evaluated by
applying the decision rule and are
ranked from best to worst. The
decision rule can than be extended to
include other aspects and the
decision is reached through analysis
and evaluation.
(4) The evaluation of alternatives is
based on the multiattribute outputs
and on additional attributes, but the
pros and cons are discussed between
stakeholders and the decision is
reached through expert judgment or
bargaining and negotiation.
63
-------
Decision Support Tools NATO/CCMS Pilot Project Phase III
4. TOOL 1: BEST AVAILABLE TECHNIQUES
4.1. SEARCH AND SCREEN (AND ANALYSIS/EVALUATION) SUPPORT
BAT (Best Available Techniques) is an MS-Access/Virtual Basic product for Windows. Tauw and VITO
(Flemish Institute for Technology Development) have developed it. BAT is a decision support tool that is
aimed at supporting soil remediation experts, policy makers and environmental managers (Gevaerts et al.,
1998).
The knowledge base of the tool consists of a large number of factsheets related to remedial concepts and
techniques. As input BAT requires four types of information: soil characteristics, contamination
characteristics, other characteristics such as presence of buildings and infrastructure and, finally, remedial
targets and duration. As soon as the input is entered the tool starts to compare the input data with the
factsheet data and the output is a list of suitable concepts and techniques, unsuitable concepts and
techniques, concepts and techniques for which a decision requires additional data and, finally, concepts
and techniques which are not relevant to the problem described.
For soil remediation experts the tool supports during the search and screen routines of a modified search
or dynamic design decision process and as such it is related to the development phase of the decision
process. Its goal is to give the user an overview of remedial alternatives that could be worked out. For policy
makers and environmental managers the tool enables to check the work of environmental experts. As check
it is probably more in line with the screen and analysis/evaluation routines and thus with the selection
phase of the decision process. Note that in case of second opinion soil remediation experts can use the tool
to check the work of their colleagues.
4.2. FEEDBACK AND FEEDFORWARD
The selection of a number of possible remedial alternatives is an important stage in the process. Although
trying to anticipate to the preferences of the stakeholders by presenting a specific subset of the outcomes
of the BAT model is possible, we feel that the total set of outcomes of the BAT model should be
discussed with the stakeholders. This may tempt the stakeholders to change their preference profile, and,
as a designer you can never be accused of having a narrow scope. Thus, when using the BAT model
within the decision process we recommend not using the feedforward paradigm. After the results have
been discussed in the selection phase of adaptive decision processes such as the modified search or the
dynamic design decision process the 'lessons learned' should be included in the next iteration.
As has been discussed before our target is not the decision process, but the effect and the efficiency of the
resulting action. If we stretch our attention to the remedial actions, then we can make use of the
experiences with the remedial alternatives. These experiences should be included in the knowledge base
of the tool, i.e., in the factsheets. This kind of feedback requires clear guidelines for the maintenance of
the model. If the outcomes of certain remedial alternatives are more uncertain than those of others, then
this should be communicated. This kind of feedforward/anticipation (see Okx and Stein, 2000) should
avoid disappointment.
4.3. THE MULTIOBJECTIVE SETTING
The BAT models goal is to give the user an overview of remedial alternatives that could be worked out.
In this stage the decision rules are not made explicit yet, and other factors than those used by the model
do contribute to the decision. Thus, the evaluation of alternatives is based on the model outputs and on
additional attributes, but the pros and cons are discussed between stakeholders and the decision is reached
through bargaining and negotiation.
Our observation is that in this stage the legal authorities are seldom involved, and discussions are held at a
technical level. This should be considered as a missed opportunity. Discussions like 'what if that and that
64
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
happens' are most valuable and could result in modified solutions capable of handling less likely
situations as well as the expected situation.
5. TOOL 2: IN SITU AIR SPARGING
5.1. DESIGN SUPPORT
In situ air sparging - A technical guide, Version 1.1 has been developed by Tauw and GeoDelft. We
gratefully acknowledge the views and comments of many people, in particular Terry Walden of BP Oil
Europe UK and Rick Johnson of Oregon Graduate Institute in the United States. The technical guide is a
tool for consultants/designers to help plan the remediation process and to design and operate an in situ air
sparging (IAS)/soil vapour extraction (SVE) system. The guide should be able to support decisions for the
clean-up of different types of sites ranging from small sites with permeable soil to large industrial estates
with strongly heterogeneous and stratified soils (Pijls et al., 2000). Its predecessor - Version 0.1- had a
totally different appearance. It was structured around a large number of flowcharts and meant to lead the
expert via a series of questions to an optimal design. Its structure was ideal for the envisaged software
implementation. During the test period, however, the future users rejected the product for two reasons:
• They preferred a reference guide rather than some kind of workflow oriented tool; and
• They were questioning the systems ability for special cases.
Although I am still not convinced whether it was the only option, we decided to capitulate and developed
the classical technical guide. We were reported that in the same period a similar project in the United
States faced a similar fate (Leeson et al., 1998).
The present tool is organised in a number of chapters covering: the theoretical background, remediation
concepts, feasibility studies, design, hardware/equipment, installation of the IAS/SVE system, operation
of the IAS/SVE system, shutdown and postclosure measures and costs. Thus, it follows more or less the
designers' logical steps. However, if the designer has a particular question on blowers and compressors,
then he can skip everything except the section on injection equipment In Version 0.1 skipping sections
was less easy.
For soil remediation experts the tool supports during the design routine of a modified search or dynamic
design decision process and as such it is related to the development phase of the decision process. Its output
is a detailed IAS/SVE design.
5.2. FEEDBACK AND FEEDFORWARD
IAS/SVE design takes place after the selection of a number of remedial alternatives by the stakeholders.
Again I feel that all possible options in IAS/SVE should be discussed, but, as long as technical choices do
not lead to different results, not all of the stakeholders need to be involved. The processes induced by
IAS/SVE are rather complicated, and, in order to discuss items such as blowers your discussion partner
should have some knowledge about lateral rotary blowers, rotating lobe blowers, rotary sliding vane
blowers, and so on. Selection is based on expert judgment rather than on analysis or bargaining. Thus,
when using the technical guide in a modified search or the dynamic design decision process the expert
judgment or 'lessons learned' should be included in the next iteration.
If we stretch our attention to the remedial actions, as we did with the BAT model, then we can make use
of the experiences with IAS/SVE concepts, feasibility studies, design features, hardware/equipment,
installation, operation, postclosure measures, and costs. These detailed experiences should become part of
the next versions of the technical guide. This kind of detailed feedback requires even more clear
guidelines for the maintenance of the guide. Again negative and positive experiences allow for
anticipation.
65
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
5.3. THE MULTIOBJECTIVE SETTING
The IAS/SVE guide supports IAS/SVE design. In this stage the decision rules are connected to the
implicit knowledge of experts, and other factors than those used by the guide do practically not contribute
to the decision. Thus, the evaluation of alternatives is based on the model outputs, but the pros and cons
are discussed and the decision is reached through expert judgment. Our observation is that legal
authorities are seldom involved in this stage, and discussions are held at a technical level. As long as
technical choices are unlikely to change the effect and efficiency of the clean-up operation, it should not
be considered as a missed opportunity.
6 TOOL 3: COMPARISON OF REMEDIAL ALTERNATIVES
6.1. Analysis/evaluation support
REC is an Excel/Virtual Basic Decision Support Tool for Windows for the analysis and evaluation of
possible clean-up strategies for a contaminated site. REC was developed by Tauw, the Institute for
Environmental Studies of the Free University of Amsterdam and Berenschot Osborne of Utrecht. The aim
of REC is to support the choice of the most effective and efficient strategy for soil remediation in terms of
risk reduction, environmental merit and costs for the site concerned (Okx et al., 1998; Okx, 1999).
The core of the model consists of a database comparable to those of Life Cycle Analysis (LCA) tools and
a large number of formulas able to calculate risk, reduction, environmental merit, and costs. As input
REC requires three types of detailed design data of the chosen remedial alternatives: risk characteristics
such as type of contamination, risk profile, land use, the area involved, environmental characteristics such
as expected amount of clean ground and ground water, use of resources like clean ground, ground water
and energy, air emissions, pollution of surface water, production of final waste and spatial occupation,
and, finally cost characteristics such as preparation costs, demolition costs, remedial costs, replacement
costs and discount rate. The output of REC is a set of three indices for each clean-up alternative: the risk
reduction index, the environmental merit index and the costs index. Together, these indices summarise the
overall performances of each option (see Figure 4).
Risk reduction
0.4'
Environmental merit
2.i) 1
0.0 +
Costs
I II III
I II III
III
Figure 4. R-, E- and C-indices of three remedial alternatives I, II and III
Thus, the tool supports the analysis/evaluation routine and, thus, is linked with the selection phase of the
decision process.
6.2. FEEDBACK AND FEEDFORWARD
The selection of the remedial action is seen as the most important stage in the process. In general the
decision rules of the different stakeholders are known. Three strategies are quite common: a focus on
66
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
effectiveness, which aims for the selection of the most effective option provided the budget available is
sufficiently high; a focus on costs, which aims for the selection of the cheapest solution provided some
significant risk reduction is achieved; and, finally, a focus on efficiency, which aims for the selection of
the solution that gives the best ratio between risk reduction, environmental merits and costs.
We saw in practice that the designers who worked with REC developed - by using the models feedback -
a feeling for the relation between design features and the R-, E- and C-indices and they gradually
developed into designers able to anticipate the results of the REC model. Their designs became smarter:
an advantage of the use of the full-information paradigm.
Stretching our attention to the remedial action means that we will be able to improve the LCA-like
database of the tool, and that the model will improve steadily. Again we need to organise this kind of
feedback. If we don't we will not benefit from our experiences. If the outcomes of certain remedial
alternatives in terms of the R-, E- and C-indices prove to be uncertain, then this should be communicated.
The present model presents the cost uncertainties.
6.3. MULTIOBJECTIVE SETTING
The REC model gives the stakeholders user an overview of the consequences of remedial actions in terms of
risk reduction, environmental merit, and costs. In this stage the stakeholders have their decision rules set, but
are not always willing to expose them. Other factors than those used by the model do contribute to the
decision. Thus, the evaluation of alternatives is based on the model outputs and on additional attributes,
but the pros and cons are discussed between stakeholders and the decision is reached through bargaining
and negotiation.
Our observation is that in this kind of analysis/evaluation routines the legal authorities are nearly always
involved. Although the multiattribute profiles and specially the weights attached to the attributes differ for
the stakeholders involved, our experience is that quite frequently one alternative outclasses the others
regardless of profiles, weights and values. Such an alternative is easily accepted by all of the stakeholders,
which gives them almost a co-responsibility for the chosen alternative. Note that co-responsibility is not
the same as co-liability! Discussions like 'what if that and that happens' are again a common feature and
frequently result in modified solutions capable of handling less likely situations as well as the expected
situation.
HAPPILY EVER AFTER ...
There are many tools, which supports the decision process. In this article we have shown only a few of
them: a tool that lists the best available techniques for the problem at hand, a tool that enables IAS/SVE
design, and, finally a tool that enables to select the best design on the basis of risk reduction,
environmental merit and costs. Presently the tools are used independently of each other, but signs of
integration by using feedbacks and feedforwards are observed. The case of the REC using designers give
rise to a project trying to formalise the design rules which were implicitly used by our designers. Once
these rules are explicit, they should be fed to tools such as BAT or the IAS/SVE guide. The reduction of
the designers' degrees of freedom, however, should be avoided. This can be realised by simply enabling
the users of the tool to switch the rules off or on. In my opinion this kind of development does not mean
that we should aim for an all-including integrated decision support tool which does it all. Instead we
should develop a number of smaller tools that are almost invisibly interlinked by feedbacks and
feedforwards, but they have to be used independently of each other. The boundary of the decision process
as described by Mintzberg does not include the actual actions, and this inhibits a gradually improving
decision quality. Feedback and feedforward should include the actions. If not, then a proper evaluation of
our decisions will remain impossible. Involving the stakeholders in the decision process is necessary and
beneficial. An alternative accepted by all of the stakeholders, is the best alternative.
67
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
REFERENCES
BARDOS, R.P., MARIOTTI, C., MAROT, F. AND SULLIVAN, T., 2000. Decision support for contaminated
land in Europe and North America. In: Proceedings of the Seventh International FZK/TNO Conference on
Contaminated Soil 2000,Leipzig.
BEINAT, E., 1997. Value functions for environmental management. Kluwer Academic Publishers,
Dordrecht
DRESCHMANN, P., (1992). Altlastensanierung in den USA am Beispiel des RI/FS-Process. Altlasten-
Spektrum, 1/92
ElKELBOOM, R.T. AND VON MEIJENFELDT, H., 1985. The soil clean-up operation in the Netherlands.
Further developments after five years of experience. In : Proceedings of First International TNO
Conference on Contaminated Soil '85, Assink, J.W. and Brink, W.J. van de (eds.). Martinus Nijhoff
Publishers, Dordrecht.
GEVAERTS, W., LAURYSSEN, K., PIJLS, C.G.J.M., DIJKMANS, R. AND DRIES, V., 1998. Best available
techniques for contaminated sites. In: Proceedings of the Sixth International FZK/TNO Conference on
Contaminated Soil 1998, Edinburgh.
GOTOH, S. AND UDOGUCHI, A., 1993. Japan's policy on soil environment protection - history and present
status. In : Proceedings of the Fourth International KfK/TNO Conference on Contaminated Soil, 3-7 May
1993, Berlin.
JANSSEN, R., 1991. Multiobjective decision support for environmental problems. Thesis of the Free
University of Amsterdam, Amsterdam.
KLIR, G.J., 1991, Facets of systems science, IFSR International Series on Systems Science and Engineering,
Volume 7, Plenum Publishing Corporation, New York
LEESON, A., JOHNSON, P.L., JOHNSON, R.L., HINCHEE, RE. AND MCWORTHER, D.B., 1998. Air sparging
design paradigm. Battelle, Columbus, Ohio.
MlNTZBERG, H., RAISINGHANI, D. AND THEORET, A., 1976. The structure of unstructured decision
processes. Administrative Science Quarterly, 21.
NUTT, P.C., 1984. Types of organizational decision processes. Administrative Science Quarterly, 29.
OKX, J.P., BEINAT, E., VANDRUNEN, M.A. ANDNIJBOER, M.H., 1998. The REC-framework: integral
risk management for contaminated soil. In: Proceedings of the CARACAS-Workshop, Berlin
OKX, J.P., 1999. The REC decision support tool for comparing soil remediation alternatives (in Japanese).
In: Proceedings of the 2nd International Workshop of Geo-Environmental Restoration, Yokohama.
OKX, J.P. AND STEIN, A., 2000. Use of decision trees to value investigation strategies for soil pollution
problems. Environmetrics, Volume 11, Issue 3
PIJLS, C.G.J.M., BOELSMA, F., MARNETTE, E.C.L., VAN REE, C.C.D.F., VREEKEN, C., 2000. In situ air
sparging-A technical guide, Version 1.1. CUR/NOBIS report 95-1-13, Gouda.
WATZLAWICK, P., BEAVIN, J.H. AND JACKSON, D.D., 1967. Pragmatics of Human Communications.
W.W. Norton & Company, Inc., New York.
68
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
CASE STUDY: COST BENEFIT ANALYSIS /
MULTI -CRITERIA ANALYSES FOR A
REMEDIATION PROJECT
Dieter Weth
Mull und Partner Ingenieurgesellschaft, Germany
Osteriede 5, 30827 Garbsen, Germany
E-mail: weth@mullundrjartaerde
SUMMARY
Transparent planning processes are necessary to increase acceptance of remediation projects with the
general public and those affected. Compared to simple cost estimates, an evaluation of remediation
schemes, compatible with the space, that is carried out using an economic and ecologic assessment of a
remediation concept can ensure a uniform comparison of various remediation measures. Further planning
may then be carried out on that basis.
The following instruments are used when preparing an overall remediation concept that is compatible
with the site in the planning phase of a remediation scheme:
1. A dynamic efficiency calculation considering also - expressed in monetary terms - the loss of use
of a contaminated site and the duration of the remediation measure, and
2. A value in use analysis of the ecological effects of a remediation scheme.
This results in a scope for decision, which ensures optimal planning results for the remediation concept.
This scope relates to identifying the optimal economic and ecologic remediation solution in each case for
the remediation process to be used in a particular case of damage. The economic costs and the ecological
benefits of a remediation concept need to describe the effectiveness and/or compatibility with the site of a
remediation scheme and thus allow justifiable decisions to be made.
1. INTRODUCTION INTO THE ECONOMIC AND ECOLOGIC OPTIMISATION WHEN
SELECTING REMEDIATION PROCESSES
When following a line of action in the remediation planning that is compatible with the site the selection
of the remediation process or method used is optimised by the technical, economic and ecologic
requirements in an iterative process. Alternative possible uses are considered to come to a solution that is
economically and ecologically viable. As a rule, the owner of the land is free to define the target
parameters for the remediation of contaminated sites within the given general statutory and technical
conditions (see figure 1). In this, steps to avoid dangers for a non-sensitive use represent the lower limit
(minimal concept). In enhancement of the ecological use of a remediation can even lead to the realization
of a maximum concept that would allow universal subsequent use under even the most sensitive
requirements.
69
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Scope for Dedsions
Economic expenditure
Which of the remediation concepts is in
regard of
• economic aspects favourable?
* ecological aspects favourable?
Which decisions lead at additional low
economic expenses to high ecological
benefit?
Minimal concept
Maximum expenditure
Minimal expenditure
*• Ecological benefit
Maximum concept
Figure 1. Scope for Decisions
Taking into consideration the avoidance of danger the following criteria need to be considered when
selecting the remediation process and/or a combination of processes in addition to the costs:
Precedence of the destruction of harmful substances over separation
Minimizing masses and mass flows
Waste avoidance, minimizing residuals
Substance recycling
Resource conservation
To evaluate the remediation concept in terms of its economic and ecological aspects, all technically
feasible remediation processes together with their costs and uses in the form of ratios need to be included
as alternatives in the decision making process, in order to provide as wide a base a possible for arriving at
a decision (Gehrke 1993).
2.
ECONOMIC VALUE LEVELS
The remediation processes that could be used need to be assessed in respect of their economic value
levels:
70
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Table 1. Economic value levels
Value level
Cost level
Expense level
Investment
level
Description
Costs of the remediation process
Benefit due to increased value of
decontaminated site
Costs due to different lengths of
time needed for remediation
Process
Comparative cost
method
Comparative profit
method
Capital value method
Efficiency
calculation
approach
static
static
dynamic
(according to Gehrke 1993)
The costs of a remediation process, however, depend primarily on the amounts to be treated. In addition
the various fix costs for site setting up and safety at work need to be taken into account (see figure 2).
Cost comparison calculation
m, X k.
Sum of the quantity dependent method costs
Index for different remediation methods
Contaminated soil and groundwater quantity
Method specific quantity dependent cost attempt [$/m3]
Figure 2. Cost comparison calculation
On the other hand, varying revenue is obtained as a consequence of a remediation, in particular revenue
derived from subsequent use of the remediated site, but also the avoidance of higher remediation costs at
a later date, the enhanced image of the landowner or the avoidance of claims for damages put forward by
the owners of adjacent land that would be affected by the pollution spreading. The revenue from any
subsequent use is important and can be assessed from a business costing point of view and it needs to be
compared to the costs of carrying out the remediation.
The alternative remediation options can differ widely as regards the duration of their use. A significant
influence on the economic results, which is calculated by the capital value method is brought to bear by
the
• loss of interest due to having to pay procedural costs
• loss of interest due to lost monetary use of the site
(Figure 3). The capital value is composed of the following elements:
1. Costs of the procedure
2. Loss of monetary use due to a permanent limitation of the use of the building plot on completion
of the remediation
3. Loss of interest due to varying durations of the measures, and
4. Loss of interest due to the varying length of occupation of building plots (Gehrke, 1993)
71
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Capital value method
Capital value: - Method costs
- Loss of monetary due to a permanent limitation of the use of the building plot on completion of the
remediation
- Loss of interest due to varying durations of the measures
- Loss of interest due to the varying length of occupation of building plots
Calculation of
the capitel value:
K(t)
K(t) :
r :
n :
K(t=0) X [ 1 + r ]n
Capital value at the time of (t)
Internal interest rate
Period
Figure 3. Capital value method
The economically most advantageous process is that which yields the maximum final sum in relation to
the alternatives. The criteria of advantage for a comparison of the alternatives is therefore:
• An investment I is more advantageous that an alternative II, if its negative capital value is smaller
than the negative capital value of alternative II (Gehrke 1993).
3 ECOLOGICAL VALUE LEVELS
Soil and groundwater remediation is carried out to change the state of the contaminated site to such an
extent that they do not pose a risk in respect of the use of the property/assets to be protected. Assets to be
protected are primarily people's health. Carrying out a remediation project, however, also involves other
affects. Two value levels should therefore be distinguished for an evaluation (see figure 4):
Ecological deciding factors:
Al~*~"^ ^—-A n
B
Primary effects:
Effects on the location
Ecological benefit
Secondary effects:
Effects of technical activities
on the environment
(material and energy flow)
Ecological rucksack
^HP^"'
&
..'&
Figure 4. Ecological value levels
Primary effects = effects caused by the measurements on the environment of the site (figure 5)
• Quality and structure of the soil
• Biotope quality
• Groundwater quality and recharge
• Topography/relief
• Climate regulation potential
72
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Ecological deciding factors:
A
— Ecological quality of the location
Primary effects:
Groundwater
recharge
Quality change
Groundwater
quality
Fresh air
^emergence.
(^Soil structure?)
Topography/
Relief
Figure 5. Primary effects
The individual primary ecologic effects (Figure 6) are ranged and weighted in relation to the local
conditions.
1 . Secondary effects = ecologic rucksack (supra-regional effects through the technical activities of
the use of the process on the environment) (see Figure 7):
• Substance streams
• Energy streams
Damage to the environment and health
Ecological deciding factors
Assessment of the results of remediation measures
Criteria
Condition improvement in the concerned media by
remediation
Soil quality
Soil structure
Topography/relief
Groundwater recharge
Groundwater quality
Fresh air emergence
Air exchange
Biotope quality
Level of the method conditional remaining loads
Risks by the remaining loads
Durability of the success of the remediation measure
Degree of the
attainment of
targets
Weighting
Result
Ecology value 1
Figure 6. Ranging and weighting the primary effects
73
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Ecological deciding factors
B
Secondary effects:
Ecological rucksack
Effect of technical activities on the environment
Technical activities
Energy flow
Environmental and
health damages
Figure 7. Secondary effects
With the secondary ecological effects too a weighting is done before both levels are combined into one
ecologic value (Figure 8).
Ecological deciding factors
Assessment of the execution of remediation measures
Criteria
Waste and sewage issues
Waste disposal
Sewage elimination
Protection of the human health
Air emissions of contaminants
Dust issues
Water emissions of contaminants
Noise issues
Resource saving
Energy
Additional chemical substances
Natural raw materials
Degree of the
attainment of
targets
Weighting
Result
Ecology value II
Figure 8. Ranging and weighting the secondary effects
The procedure followed when developing the ratios is a follows in the context of the comparison of
procedures:
1. Defining an objective for the problem to be investigated
2. Translating an objective into evaluation criteria
3. Subjective weighting of the criteria according to certain comprehensible and controllable rules
4. Evaluation of alternatives based on profit targets
The effects that can be felt vary in the extent to which they occur both as regards quality as well as
quantity. By using the 'value in use analysis' the effects on the environment are represented in the form of
ratios and can thus be compared.
74
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
4.
CONCLUSION
The model makes clear the connection between the economic effort and the ecological benefit of a
remediation concept (figure 9). The ecologically best overall remediation concepts consists of the process
alternatives with the smallest capital values with the benefit as a result of an increase in value of the
rehabilitated site and the costs due to the varying timescales of the remediation also being taken into
account in addition to the costs for the remediation scheme itself. The ecologically best overall
remediation concept is expressed by the highest values in use. These result on the one hand from the
extent and the sustainability of the improvement of the state in respect of the exposition of human beings
and the environment by the remediation measure and on the other from process related environmental
pollution by the remediation project itself, which have supra-regional ecologic effects.
Business management
consequences
Primary effect on the
environment
(Ecological benefit)
Secondary effect on the
environment
("Ecological rucksack")
Cost benefit analysis
Assessment of the
results of remediation
measures
Economy value
I
Weighting
Assessment of the
execution of
remediation measures
Weighting
Treatment optimized economically and ecologically
Ecology value
ecologically
i
r
Ecology value II
Weighting
Economic profit
FD
Ecological
damage
Economic loss
O !
......................... 4D...T....Eco|°9ica|
o '•• benefit
•A
Ecology
Figure 9. Treatment optimized economically and ecologically
The combination of the ratio pairs of the investigated remediation concepts is illustrated by the two
extremes - the ecologically best concept that does not take into account economic considerations (Figure
9; case D) and the economically most advantageous concept, which does not consider the ecological
benefit (Figure 9; case F). Where the economically best remediation concept varies from the ecologically
best scheme, a scope for decision results. Purely formally, that concept will show the best "economic-
ecological effect" by which the ration of increase in ecological benefit to the increase in the capital value
is the largest (Figure 9; case E).
When calculating the economic and ecological ratios there are naturally uncertainties. With the economic
ratios these are largely the result of the risk of changes in the volume, for the ecological ratios they relate
mainly to the previously relatively subjective setting of the objectives and the weighting that needs to be
carried out. Those ratio pairs are decisive in the comparison of the results of several remediation concepts
which, as regards the identified bandwidth are on "the safe side", i.e., the - in respect of the amount -
smallest ecological benefit with the largest economic expenditure.
To be able to carry out the effectiveness of remediation measures in the context of the preplanning fast
and transparently, we developed the DV supported effectiveness analysis model "WILMA". The
75
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
economic and ecological standards can be described and covered with the corresponding code numbers so
that the economic and ecological effectiveness of a remediation is determined.
With WILMA different use scenarios can be simulated in connection with location specific factors. The
use scenarios of the redeveloped area depend on toxicologic examining results, in which we distinguish
in:
Children's playground
Living use, general
Park/public green space
Fallow area
Industry and trade area
At the moment we are able to carry the calculation with WILMA for the approx. 20 most important
remediation methods. By the flexible conception of the model the standard details of the database can be
adapted to regional conditions. An individual case obtained use of the planning instrument gets possible
so. Prerequisite is an exact knowledge of the damage and this one for the location specific conditions
essential for remediation and use.
5.
SALZWEDEL CASE STUDY
Using WILMA a remediation involving a site contaminated with petroleum-derived hydrocarbons and
BTEX (benzene, toluene, xylene) on the former helicopter port of the East German border troops at
"Salzwedel - Fuchsberg" that has already been completed was subsequently calculated for calibration and
testing purposes (see figure 10).
Case study: "Salzwedel"
Helicopter base of the former GDR - border troops
Contaminations of soil and groundwater in the area of a gas station by kerosene,
diesel oil and gasoline
Soil:
Initial concentrations: MKW up to 17.800 mg/kg
BTEXupto2QOmg/kg
Remediation aim:
Volume to be
redeveloped:
Soil type:
MKW: 150 mg/kg
BTEX: 2,5 mg/kg
approx. 1.100 m3
fine - middle grained sand
Groundwater:
Initial concentrations: BTEX up to 50 mg/l
Remediation aim: BTEX: 20 //g/l
Figure 10. Case study 'Salzwedel' - Basic information
76
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
5.1
BRIEF DESCRIPTION OF THE SALZWEDEL CASES
On the former filling station site mean petroleum-derived hydrocarbon concentration of 2,000 mg/kg
(max. 17,800 mg/kg) and/or BTEX levels of approx. 200 mg/kg were found in the soil. The aromatic
hydrocarbons had already entered the groundwater, which had BTEX levels of up to 50 mg/1. The soil
contamination affected approx. 500 m2, groundwater contamination affected approx. 5000 m2. The
volume of soil to be rehabilitated was approx. 1,100 m.
The site is partly used for residential purposes (renovation of the old barracks building), and part of it is
derelict (still owned by the federal government) and used by a school.
5.2
RESULTS
As an initial step using the WILMA model a selection of the processes that could be used for the
contaminants identified was made based on the details regarding the type and extent of contamination.
After entering the site-specific soil properties and the available resources the program narrowed down the
applicable processes. Resources are those site conditions that have a bearing on the construction or setting
up of the remediation plant such as the necessary space (for the facility or for intermediate storage of soil,
etc.), water, power, etc. Figure 11 shows the results of this pre-selection.
Case study: "Salzwedel"
Possible remediation methods
(MKW/soil):
R
Soil clean-up off-site
Microbiological treatment on-site
Microbiological treatment off-site
1C
Pyrolysis off-site
Incineration off-site
:.-
.:>
E-
Exclusion of the method by:
Possible remediation methods
(BTEX/soil):
R clean-up
Soil clean-up off-site
Microbiological treatment on-site
Microbiological treatment off-site
1C
Pyrolysis off-site
Incineration off-site
Soil venting with active coal
Soil venting with catalytic oxidation
£• Slot wall
£ • wa 11
B
Pt: Lack of location specific resources (available areas, water, energy etc.)
1C: Location specific criteria (e.g. grain size, too low groundwater level referred to ground elevation)
B: Editor
Figure 11. Case study 'Salzwedel' - Possible remediation methods
After calculating the economic costs of the processes left and the ecological benefits, and after combining
the individual processes the result shown in Figure 12 was arrived at.
77
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Case study: "Salzwedel"
i
o nnn nnn
„ 2.500.000
i
Q
" 2.000.000
0)
1/1
0) 1.500.000
Q.
X
0)
O 1.000.000
E
c 500.000
8
n
BTEX: Soil \
(cata
MKW: Micrc
'entir
lytic
iblol
BTEX; Soil venting
(active coal)
MKW; Pyrolysis off-site
BTEX: Soil clean-up on-siti
»g
oxidation)
ogy on-site v-
^
I *
BTEX; Soil venting *
(active coal) /
MKW: Microbiology on-site
jj
BTEX: Soilve
(cataly
MKW: Microt
_ |/
nting
ic oxidation)
iology off -site
BTEX: Soil venting
(active coal)
^ MKW: Microbiology off-site
, 4^
BTEX: So!
__b__ ,
•7*— (ad
^ MKW: Inc
venting
ive coal)
neration off-site
0 2 4 6 8 10
Ecological value [-]
Figure 12. Case study 'Salzwedel' - Results
It became apparent that the combination of soil air extraction with cleaning of the extracted air by
activated charcoal for eliminating the volatile BTEX from the unsaturated soil zone with a
microbiological on-site process (clamp) - and when ignoring the ecological benefits - involved the lowest
economic costs. On the other hand additional ecological benefits would only have resulted - involving a
significant increase in the economic costs -from the combination of "soil air extraction with catalytic
oxidation" with "microbiology off-site." This combination of processes showed the highest eco benefits
when ignoring the economic costs. Cleaning the soil from the petroleum-derived hydrocarbons by means
of a thermal process (off-site incineration) too would have meant a significant increase of economic costs
at only small increases in ecological benefits due to the transport costs involved.
6. LIFE CYCLE ASSESSMENTS AS METHODS IN THE QUALITATIVE AND
QUANTITATIVE EVALUATION OF ENVIRONMENTAL EFFECTS IN THE
REMEDIATION OF CONTAMINATED SITES
Apart from the value in use analysis shown a further method for assessing the effects of the
environmental pollution through the remediation project itself can be calculated by means of life cycle
assessments.
In the Value in use analysis' a methodological comparison of processes is made. Similar to the
methodological steps of the environmental impact assessment under the UVP Act this involves the
assessment of the ecological benefit and all the effects of the remediation scheme on the environment,
which are described, forecast regarding both quality and quantity, assessed and summarized as a ratio, in
order to obtain a comparison and be able to arrive at a decision as regards the selection of the method.
The "life cycle assessment' too should take into account as far as possible ecological criteria when
specifying possible remediation processes in order to achieve long-term ecological optimisation. For this
it is necessary to know what effects the methods considered may have on the environment. The Federal
Environment Office suggests the following definition of the term:
78
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
The life cycle assessment is a comparison, which is as comprehensive as
possible, of the environmental effects of two or more different products, product
groups, systems, process or behaviours. It serves to point out weak points, to
improve the environmental properties of products, the decision making process
in procurement and purchasing, to promote environmentally friendly products
and processes, to compare alternative courses of action and to back up
recommendations for action to be taken. Depending on the underlying question
this comparison is supplemented by further aspects, e.g., an assessment of the
environmental protection efficiency of funds, (from Schmidt-Bleek 1993)
The Federal Environment Office suggest to standardize the method as follows:
1. step: Definition of the objective of the report
2. step: Operation balance: setting up a data basis by analysing all environmental influences on
the lifecycle of a product starting with raw material procurement to disposing of the waste
in life periods, the modules (vertical analysis). Considering the entries and exits of the
modules in respect of the use of primary energy, raw materials, water and the emission
into the air, the wastewater and solid waste (horizontal analysis). Examining the link
between the modules (lifecycle criteria). Selection of the data.
3. step: Impact balance: Description of the impact of the substance streams covered in the
operation balance on the environment.
4. step: Balance assessment
The concept centres on the method prepared on behalf of the LfU Baden-Wurtemberg, "Environmental
balances of the remediation of contaminated sites" by C.A.U. GmbH.
Using this method it is possible, as has already been shown, to calculate remediation caused
environmental effects as a result of an operating balance in which the potential effects, differentiated by
certain impact criteria, has been calculated (Figure 13). In practice, however, the problem will often have
to be solved as part of the necessary assessment of the balance, to weigh advantages and disadvantages of
the individual remediation measures, in order to identify "the best possible method". In cases where not
one single method combines all the advantages over the other methods the person dealing with it at the
building authorities is faced with difficult questions:
• Which disadvantages for the environment are so serious that any measures that include those
disadvantages must be excluded as a matter of principle?
• Which environmental benefits balance out which disadvantages and to what extent?
79
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Ecological deciding factors
Life cycle assessment of the secondary effects
Balance sheet of masses und energy
Balance sheet of the effects
Effect categories j Results of the effects
• Use of energy
• Use of area
• Greenhouse effect
XT GJ
x2 m2
x3 kg C02
Assessment
"Umweltbilanzierung von Altlasten-
sanierungsverfahren" (LfU / C.A.U.)
"Environmental balance of
remediation measures for
contaminated sites" (LfU /C.A.U.)
Figure 13. Ecological deciding factors-Life cycle assessment of a measure
A part objective is therefore to be able to calculate and overall index from the many indices calculated by
means of the "CAU method", which allows a statement to be made on the best ecological remediation
concept.
To do so it is necessary to obtain cost factors that are as objective as possible for the individual impact
categories of the CAU method which, by multiplying them with the result of the impact balance, yield the
external costs of that particular remediation scheme (Figure 14).
So the key aspect of the model presented here is "monetarisation". Calculating monetary values for all
relevant decision criteria allows a direct link and comparison of economy and ecology. Simply summing
of the economic value, the ecological values I and II result in a quantitative basis for selecting the optimal
remediation method from business and economic aspects. If it is possible to minimize existing
uncertainties in the monetarisation no further weighting of the above mentioned elements is necessary,
such as would be the case in an assessment in which verbal arguments are put forward to evaluate
ecologic concerns.
In a further step standards have to be developed for comparing the primary and secondary environmental
impacts that present the local effects of the scheme at the site in an objective way. This can be achieve by
monetarising the primary effects:
Quality and structure of the soil
Biotope quality
Groundwater quality and recharge
Topography/relief
Climate regulation potential
As a result a ratio for the ecological effectiveness of a remediation scheme is obtained that, again together
with the economic ratio from the cost/benefit analysis, delivers an objective basis for the decision making
process for the most effective remediation scheme.
80
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Assessment by comparison of the
monetary consequences of the effect categories
Effect category
Category result
Use of energy:
Greenhouse effect:
x2 kg CO2
X
2.80 S/GJ
0.022 $/kg CO2
Ecology value I
Economic damage by the
external effects of the measure
(external costs)
Figure 14. Ecological deciding factors—Assessment by comparison of the
monetary consequences of the effect categories
7.
SUMMARY AND CONCLUSION - ASSESSMENT OF THE RESULTS SO FAR
The results obtained by means of a value in use analysis in respect of the evaluation of any impact on the
environment of the remediation of contaminated sites are well suited to a praxis-based comparison of
alternative remediation methods. However, the weighting of the individual elements contained therein
also reflects a subjective judgement. Contrary to that, a more "objective" from of evaluation could be
carried out by a life cycle assessment plus subsequent monetarisation of the impact categories.
However, the results obtained from the various studies so far are based on different mathematical
approaches (avoidance costs, damage costs, etc.), something which asks for criticism of aggregating the
individual categories. The current state of environment economic research, however, makes it necessary
to accept this mixture of methodologies. The problem of a "limited comparability" cannot be solved at the
moment. For almost all impact categories taken into account there exist substantial gaps in research and
that is the reason why the costs identified in many cases are not more than "rough estimates".
For these reasons we work with a value in use analysis to quantify the ecological consequences of
remediation measures. One receives a two parametric result representation which serves as basis for the
decision-making process.
8.
REFERENCES
Bundesministerium fur Raumordnung, Bauwesen und Stadtebau, Bundesministerium der Verteidigung
1996: Arbeitshilfen Altlasten zur Anwendung der baufachlichen "Richtlinien fur die Planung und
Ausfiihrung der Sicherung und Sanierung belasteter Boden" des BMBau fur Liegenschaften des Bundes.
Hannover: OFD Hannover -Hausdruckerei-
C.A.U. GmbH 1996: Umweltbilanz von Altlastensanierungsverfahren. - Bericht fur die Landesanstalt fur
Umwelt Baden-Wurrtemberg. - Dreieich.
81
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
Commission of the European Communities 1995: Towards Fair and Efficient Pricing in Transport. -
Policy options for internalising the external costs of transport the European Union. Green Paper. -
Briissel.
DOETSCH, P. u. A. RUPKE 1998: Revitalisierung von Altstandorten versus Inanspruchnahme von
Naturflachen. - ,,Texte 15/97" des Umweltbundesamtes. - Berlin.
Gehrke, D. 1993: Die Entwicklung von Sanierungskonzepten fur Altstandorte. Fortschritt-Berichte VDI,
Reihe 4: Bauingenieurwesen, Nr. 122, Diss. Diisseldorf
F. SCHMIDT-BLEEK 1993: Wieviel Umwelt braucht der Mensch? MIPS - Das MaB fur okologisches
Wirtschaften. - Berlin.
M&P Ingenieurgesellschaft mbH 1997: M&P report; WILMA - Wirksamkeitsanalysemodell fur
Altlastensanierung. Ulrich Eggert GWK, Hannover
Weth, D. 1999: Nachhaltigkeitsbetrachtungen bei der Sanierungsplanung von Altlasten. DEA-A-94-GE
1520 Environmental Technology, 12™ General Meeting 1999, Proceedings. Bundesamt fur Wehrtechnik
und Beschaffung (BWB), Koblenz
82
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
MODELLING THE FINANCIAL RISKS OF REMEDIATION
Jim Finnamore, Environmental Finance Unit, WSP Environmental,
Buchanan House, 20-30 Holborn, London, EC IN 2HS, UK
Tel: +44 (0) 207 314 5000; Fax +44 (0) 207 314 5005
E-Mail: jim.finnamore@wspgroiip.com
ABSTRACT
The performance of remediation represents a significant source of financial risk, which, if ignored or
mismanaged, can have a serious effect on the commercial success of a project or business. Risk
management relies heavily on accurate forecasts of the probability that remediation will fail to meet its
objectives as well as the associated financial implications - typically expressed as a reduction in net
present value or internal rate of return. This paper discusses an analytical approach and methodology,
based on stochastic modelling, to translate the technical risks of remediation into monetary expressions of
risk. This approach provides an invaluable management tool that can generate real business benefits. Most
importantly, commercial decisions - which inevitably require a company to take a risk - are made with
greater confidence and certainty. There are then opportunities to optimise the management of risk by
contractual mechanisms and structuring of project finance.
Key Words: commercial risk, financial forecasting, modelling, internal rate of return, net present value,
expenditure, revenue, asset value
INTRODUCTION
For many businesses, remediation of soil and groundwater has assumed greater importance through its
potential to influence liquidity, solvency and overall financial performance. This trend is likely to
continue throughout Europe as pressure mounts to re-use previously developed 'brownfield' land, a
proportion of which will inevitably be found to be contaminated. Some companies view the role of
remediation as simply protecting and maintaining asset (property) value whilst avoiding legal liabilities.
Others recognise the commercial opportunities that remediation can generate in terms of enhancing the
value of brownfield sites. Whatever the business case, there are corresponding financial risks relating to a
company's ability to meet its corporate and project objectives. These risks can have favourable and
unfavourable effects depending on whether there is a downward or upward variation from the expected
costs, revenue and asset value. The positive and negative variation can in turn cause a company to
perform better or worse than expected.
The precise nature and extent of financial risk depends on the context in which remediation is undertaken.
Where remediation forms part of an investment project, such as brownfield site reclamation, the
underlying financial risk surrounds the internal rate of return (IRR) or Net Present Value (NPV). The
IRR, which represents the return that can be earned on the capital invested in a project, can be greatly
reduced to a point at which a project becomes non-viable commercially. The IRR reflects the volatility in
the risk - the two factors tending to show a positive correlation (see Figure 1). NPV represents the present
day cost of some action taken at some time in the future; in essence the present day value of that distant
cost is discounted by the applicable interest rate over that period of time.
Remediation performance increasingly features as one of the principal sources of project risk and
uncertainty to which organisations involved in brownfield development and reclamation are exposed.
Thus, lenders providing finance on built projects (i.e., which do not involve remediation) will generally
set a floor IRR (i.e., a minimum rate of return) that is lower than the rate for a brownfield reclamation,
other factors being equal (see Figure 1). Interestingly, lenders in the UK do not necessarily set different
rates of return for similar projects on brownfield sites and land that has not been previously developed
83
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
(Finnamore et al, 2000). Instead, borrowers are normally required to demonstrate increased levels of due
diligence and risk management where contamination is a known or perceived issue.
Where remediation forms part of a defence plan to avoid liabilities, the principal risks surround escalation
in costs and the realisation of liabilities if the remediation fails to meet the risk management objectives.
Of particular concern are potential third party liabilities for bodily injury and property damage, which can
cause unlimited financial impact. There are also knock-on effects in terms of reduced confidence amongst
a company's stakeholders, which can translate directly into reduced share value or indirectly into
increased cost of future financing and insurance. Independent research carried out on behalf of Citibank
suggests that around 70% of a company's net worth is determined by the markets' perception of the
company.
Investment risk
Brownfield
reclamation
Greenfield
development
Built projects
Government lending
e.g. Treasury bonds
(essentially risk-free
investment)
IRR > floor IRR
(project/investment acceptable)
Floor IRR
IRR < floor IRR
(investment
unacceptable)
IRR = -7%
Internal rate of return
Figure 1. Illustration of the relationship between IRR, project risk and project acceptability
The effectiveness of remediation can be judged financially in terms of:
meeting, or preferably beating, projected cost estimates (project costs)
ensuring timely release of the property asset for income generating use (revenue)
maintaining, or preferably, enhancing the value of the property (asset value)
increasing the liquidity of the asset (liquidity)
reducing/avoiding existing liabilities whilst avoiding creating new liabilities (risk management)
The financial risk associated with remediation stems from an inability, or perceived inability, to forecast
its effectiveness in meeting these project objectives. The difficulty in forecasting this risk stems from the
inherent volatility in, and complexity of, project costs and revenues, asset value, liquidity and risk
management. The volatility is due to variation in remediation-specific factors and other external technical,
scientific, regulatory, financial and economic factors. Certain variables are well understood and defined.
Others show considerable natural variation that is poorly defined. Understanding these variables and the
relationships that link them underpins any approach that seeks to forecast the financial risk of
remediation.
Generally, the process of estimating the financial risks of remediation is improving, and will continue to
do so. This is due in large part to growing experience of using various remedial technologies.
Nevertheless, considerable uncertainty persists, whilst the levels of confidence that can be attached to
financial forecasts show significant variation according to, amongst other factors, the technology under
84
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
consideration. This reflects the different uptake, application and, hence, experience of working with some
technologies.
FINANCIAL RISK ANALYSIS
The process of financial risk analysis for remediation is no different to the appraisal techniques applied to
other project risks, and can be broken down into the following steps (Institution of Civil Engineers and
the Faculty and Institute of Actuaries (1998).
Risk Identification
The main objective of the risk identification stage is to identify events, which, if they arose, could
threaten the achievement of the project's objectives. Such events are summarised in the form of a risk
matrix, which outlines the risks that exist throughout the various stages of a project, along with the
underlying causes. Each event can be triggered by one of more causes and can generate a number of
financial outcomes, as illustrated in Figure 2.
Event
1. of
c¥eiit
2. of
(If
Figure 2. The likelihood of possible outcomes
(Institution of Civil Engineers and the Faculty and Institute of Actuaries, 1998).
1. RISK EVALUATION
The risk evaluation stage generates forecasts of the impact of remediation risk on the financial
performance of a project. The impact is analysed using bespoke financial models, which describe the
financial performance of remedial technologies under multiple scenarios. The basis of each model is a
series of algorithms. These describe mathematically the relationship between the variables which
determine the probability of a risk event arising, as well as the probability and impact of the associated
financial consequences (if the event occurs). Each model runs stochastically (probabilistically). Using this
approach, discrete input values for variables are replaced by probability distribution functions (PDF's)
which are sampled randomly many times by Monte Carlo simulation to build up a probability distribution
of possible financial outcomes. The stochastic approach can avoid 'creeping' conservatism of
deterministic models caused by sequential selection of 'worst-case' input values. However, stochastic
models may suffer from systematic error caused by inaccuracies in constructing the model and defining
PDFs for each variable or in generating the random numbers used to sample the PDFs.
85
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
The process of generating a financial model is illustrated in Figure 3.
Scenario x
Scenario y
Scenario z etc
1
(1) Identify contaminated land scenario /
risk events e.g. aborted development, failure
to secure tenant
(2) Assess company's exposure to each
scenario
Cost of environmental risk
i) Identify the variables which
affect the risk
ii) Assign values to each variable
NB. Values are provided in the form of
probability density functions (PDFs)
rather than discrete values
iii) Generate the equation which links the
variables
iv) Run Monte Carlo or Latin
Hypercube simulation
to generate aggregate
financial risk profile
(3) Discount financial impact to present
value
Figure 3. Simplified illustration of the modelling process
The structure and composition of the algorithms depends on the type of remediation and site/project specific
characteristics. Algorithms are constructed by 'process mapping' remedial technologies and applying
deductive methods. Alternatively, empirical or data-based methods may be used.
Process mapping breaks the remediation process down into various components, addressing inputs and
outputs such as energy, waste etc. PDF's are assigned to variables based on:
• empiric data which relies on site-specific measurements and observations;
• default data from similar projects; or
• informed j udgement
For remediation techniques that are relatively well understood, such as excavation and disposal, process
mapping is relatively straightforward. In such cases, constructing the cost model algorithm and assigning
PDF's is also relatively simple. For more complex remediation projects, or those involving innovative,
untested process-based technologies, the process can be more complex and systematic errors are likely to
increase. The use of probabilistic models in valuing contaminated land has been described by Kennedy, et
al. (1996). In their work Kennedy et al built in uncertainty regarding remediation cost and duration in to a
spreadsheet of NPV for the redevelopment of a fictitious site. They used Latin Hypercube Sampling
(LHS) - a more computationally efficient variant of Monte Carlo (Nathanail 1994) for sampling the
PDFs.
Critical model variables in terms of the importance of obtaining site-specific data include those relating to
the nature and extent (lateral and vertical) of contamination. Interpolating from point information (such as
trial pits or boreholes) to model the spatial distribution of contamination is frequently carried out using
deterministic interpolators such as triangulation or inverse distance squared or by manual methods. Such
approaches fail to adequately recognise the uncertainty in interpolation. Geostatistical techniques such as
kriging do recognise this uncertainty and provide a qualitative indicator of uncertainty in the interpolation
86
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
through the kriging variance. However kriging, in common with other weighted average interpolators,
tends to smooth the raw data - effectively underestimating in areas of high values and overestimating in
areas of low values. Geostatistical conditional simulations on the other hand preserve the variability in the
original measurements and can be used to produce estimates of the probability of exceeding threshold
values - such as risk based levels. Conditional simulations may therefore be used to inform cost models
of the uncertainty in the extent and volume of contamination to be 'tackled - and therefore the cost of
remediation (Nathanail et al. 1998; Nathanail & Rosenbaum 1991).
Models are constructed in three sections - 'front end', processing and data presentation - reflecting the
three functions of the model. The front end comprises a system for downloading information relating to
the risk being analysed. The processing section of the model contains the algorithms that compute and
'score' the risks on the basis of the entered information. The data presentation component contains the
macros that enable interrogation of the model.
1. FINANCIAL RISK MANAGEMENT
The findings of the risk analysis can be used to formulate an appropriate risk management strategy, typically
based on a combination of risk control (reduction and mitigation) financing and transfer techniques. Two
techniques are discussed below:
a. Contractual risk transfer
In the example shown in Figure 4, the remediation costs of development are expected to be £10 million. The
contract incorporates an excess layer of insurance, which covers cost overruns of up to £20 million for a
premium of £1 million. The £1.5 million deductible is covered in full by the contractor, which puts all of its
profits at risk. The problem holder bears the unlikely risk that costs exceed the insurance cover (i.e., any
project costs in excess of £30 million, whilst sharing any cost savings (if the project costs less than £10
million) with the contractor according to a predetermined formula.
Cost cap:
£18.5 million
insurance limit
£1 million
insurance premium
\
£10 million
estimated —
development
cost to owner
\
£1.5 million insurance
deductible
y, . , Potential Potential
Expected ,.,.,. ,-,-,•
,• .. liability liability not
remediation , ,
covered by covered by
insurance insurance
cost
Owner remains responsible for
development costs exceeding £30 million
Insurer responsible for development
costs in excess of £11.5 million and
less than £30 million
Contractor responsible for development
costs in excess of £10 million and less
than £11.5 million
If actual costs are below expected
£10 million, owner and contractor share
savings
Figure 4. Illustration of a risk sharing remediation contract involving insurance
(Merkl, A and Robinson, H (1997)
87
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
b. Insurance
Clean up cost cap policies indemnify the insured for remediation costs, as defined in the remedial
strategy, that exceed a cost estimate agreed with the insurer. This value may be defined by modelling the
remedial plan in order to express technical risks in financial terms. The policy attaches over a prescribed
self-insured retention (SIR), which is generally equal to the expected costs of remediation plus a buffer
layer.
Payment is triggered by increased projects costs caused, for example, by the discovery of unknown
contamination or greater amounts of known contamination. Cover is also provided for extra costs to
change the remediation works, in the event that the enforcing authority requires these changes. Cover may
also be available for additional site investigation and any legal costs that could not have been reasonably
anticipated under the remediation plan.
The buffer that is set will reflect the insurer's confidence in the insured's initial cost estimates and
contingencies. The premium will normally depend on the cost and type of remedial work, the
comprehensiveness of the remedial strategy and the amount of self-insurance.
CASE STUDY: A Site in East London, UK
Background
This case study is based on a site in East London, which was undergoing development for a combination of
residential and light industrial purposes. A comprehensive site investigation had identified widespread
contamination for which excavation and disposal had been put forward as an appropriate remedial strategy.
A quantitative risk assessment had defined site-specific target levels (SSTLs) for remediation.
Using an investment model, the expected value of the project was £6 million based on future earnings from
renting the commercial premises and from sales of the residential plots. The costs associated with
development of the site were estimated at £4.5 million, including the purchase price of the land and
construction costs, but excluding remediation costs. Thus the maximum potential net present value for the
project was £1.5 million.
Financial analysis
The potential impact of remediation costs on the project NPV was simulated using a separate cost model.
Based on site investigation data, the lateral and vertical extent of contamination was modelled using
geostatistics. The outputs comprised a series of 3-d contour plots of contamination exceeding the SSTLs
at various degrees of confidence. For each plot, the corresponding volume of soil (as measured in the
ground) was calculated, creating an empirical probability distribution of contaminated soil volumes. In
this case, the data approximated to a lognormal distribution, which was incorporated into the cost model.
The re-use of clean, inert excavated material as backfill for other areas of the site was also incorporated
into the model.
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Examples of the variables on which the model was constructed are illustrated in Table 1.
Table 1. Examples of the variables for the cost model
Variable
A. Quantity of soil requiring disposal
Volume of contamination exceeding SSTL (m3)
Excavation bulking factor (excavation)
B. Cost of excavation and disposal of contaminated soils
Distance to landfill site (miles)
Trimming and excavation charges
Volume of waste per lorry load (m3)
Disposal charge (landfill) (£ per m3)
Special waste consignment charges (£ per load)
Haulage rates (£ per m3)
Importation of clean fill material (£ per m3)
PDF
Lognormal
Triangular
Triangular
Uniform
Triangular
Triangular
Discrete
Uniform
Triangular
Data source
Empirical (based on geostatistics)
Default
Site- specific data
Default
Default
Default
Default
Default
Default
C
Number of
Monte Carlo
trials
Forecast: TOTAL COSTS
30,000 Trials^) Cumulative Chart (^40 Outliers,
/ \ .OOIT
.750-
3 .500-
n
.0
o
CL -250-
.000-
O.qo 437,500.00 875,0
1 Certainty is 94.92% fron
HO II II i ii ii ii II II II Mil
-
-
-
oo.oo 1,312,500.00(^1,759,00
n -Infinity to 268,333.33 j "
Frequency
D
D
1
- 0
ooj)
Cost forecasts
lying outside
/ range
Display range
Maximum cost in range
Figure 5. Cost forecast chart
89
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
Table 2. Remediation cost forecasts and associated limits of confidence
Limit of confidence in cost forecast (%) Maximum remediation costs (£)
99.99 -4,800,000
99.9 -2,000,000
99.6 1,500,000
99.0 702,000
95.0 270,000
90.0 160,000
80.0 86,000
70.0 55,000
60.0 40,000
50.0 26,000
40.0 18,000
30.0 12,000
20.0 8,000
The results of the financial analysis of remediation costs were compared with the NPV generated by the
investment model. The results of the analysis confirmed that at a level of confidence of 99.6%
(representing a remediation cost forecast of £1.5 million), the remediation costs would not impact on the
projected NPV to an extent that would make the project non-viable.
CONCLUSIONS
Probabilistic modelling of remediation costs can be used to analyse the financial risks arising from land
contamination and determining where the largest uncertainty and potential impacts on company or project
performance are to be found. In this way remediation related risks can be managed in a way best suited to
the organisation facing the risk and unforeseen consequences arising from land contamination can be
minimised.
Detailed guidance tailored to the UK situation is forthcoming (Finnamore et al. 2000).
ACKNOWLDEGEMENTS
The author gratefully acknowledges review comments received from Dr Paul Nathanail (Land Quality
Management, UK) during the preparation of this paper.
REFERENCES
Finnamore, et al. (2000) Land contamination: management of financial risk. CIRIA Report, CIRIA,
London. (In publishing)
Institution of Civil Engineers and the Faculty and Institute of Acutaries (1998) Risk Analysis and
Management for Projects, Thomas Telford Ltd, London. ISBN: 0 7277 2697 8
Kennedy, P.J., Nathanail, C.P., Abbachi, A., and Martin, I.D. 1996. Impact of spatial distribution of land
contaminants and property investment appraisal. In: Proc. AGI '96. Association for the Geographic
Industry, London 25 Sept 1996
Merkl, A. and Robinson, H (1997) Environmental risk management: take it back from the lawyers and
engineers" The McKinsey Quarterly, 1997, No.3
Nathanail C P, Ferguson C C, Browne M J & Hooker P J. 1998. A geostatistical approach to spatial risk
Pn
90
assessment of lead in urban soils to assist planners. Proc 8th Int Congress IAEG, Balkema, 369-376
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
Nathanail, C. P. & Rosenbaum, M. S. (1991) Spatial interpolation in GIS for environmental studies.
Proceedings of the Third National Conference of the Association of Geographical Information,
Birmingham. Association of Geographical Information, Great George Street, London, pp 2.20.1 - 2.20.7.
Nathanail, C. P. (1994) Systematic modelling and analysis of digital data for slope and foundation
engineering. PhD Thesis, London University (unpublished).
91
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
DECISION SUPPORT USING LIFE CYCLE ASSESSMENT IN
SOIL REMEDIATION PLANNING
S. Volkwein, C.A.U. GmbH, Germany
Daimler Sir. 23, 63303 Dreieich, Germany, ww\vicaiiK)nlincidc
Key words: Decision support tool, environmental balancing, environmental management, life cycle
assessment, planning, soil remediation, sustainability, "Umweltbilanzierung von
Altlastensanierungsverfahren"
ABSTRACT
Sustainable soil remediation shall be based on decision support systems that cover the evaluation of the
environmental burdens caused by the remediation itself. The basic tool life cycle assessment can fulfill
this task. Since the year 2000, all four elements of the life cycle assessment method are described by
international standards (ISO 14040, ISO 14041, ISO 14042, ISO 14043). Since the year 1999, a life cycle
assessment based tool is publicly available as software, which can help soil remediation planners to
evaluate the environmental burdens of remedial actions itself. Today, a state of the art soil remediation
planning tool-box includes life cycle assessment based tools.
1 INTRODUCTION
The requirement of sustainable development restricts the right to development (principle 3 of the Rio
declaration on environment and development; UNEP, 1992). The right to development must be fulfilled
so as to equitably meet developmental and environmental needs of present and future generations. One
important issue in sustainable development is the management of soil. Soil management has often to
tackle with contaminated soil. Many sites have been contaminated due to incompetent or criminal
industrial or governmental management. Some sites are additionally or primary contaminated due to
military industrial activities or military actions. Table 1 lists some contaminated sites and some soil
pollutants.
Table 1. Some contaminated sites made by NATO or NATO member states
Contaminated sites made in peace
Kelly Airforce Base, San Antonio,
Texas, USA: hydrocarbons (BMBF,
1996)
Building 360 at Naval Air Station
Alameda (California, USA,
Department of Defense): chloroethene
compounds (NATO, 1999)
Carswell Air Force Base, Fort Worth
(Texas, USA): Chlorinated solvents
(NATO, 1998)
Kleingotz near Giinzburg, Germany,
air base, ammunition testing:
Contaminated sites made in war (UNEP, 1999)
Pancevo (Serbia) "HIP Azotara" fertilizer plant and
"HIP Petrohemija Pancevo" petrochemical plant,
waste water canal to Danube and Yugoslavian,
Romanian and Bulgarian Danube (4 April - 7 June
1999): 2,1 gigagram 1,2-dichloroethane (EDC), 8
Megagram mercury, oil, chlorinated solvents, depleted
uranium
Kragujevac (Serbia) Zastava car plant (9-12 April
1999): 1 Megagram poly chlorinated biphenyls (PCB)
and polychlorinated dibenzodioxins (dioxins)
Novi Sad (Serbia) oil refinery (5 April - 2 May 1999):
oil products, depleted uranium
Bor (Serbia) capacitors (15 May 1999):
polychlorinated biphenyls (PCB)
92
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Contaminated sites made in peace
trinitrotoluol (TNT), heavy metals,
polycyclic aromatic hydrocarbons
(PAH), 1,1,1 -trichloro-2,2-bis(4-
chlorophenyl)ethane (DDT)
(Heuschneider et al., 2000)
Contaminated sites made in war (UNEP, 1999)
The requirement of sustainable development demands the prevention of contamination of soil.
Furthermore, the requirement of sustainable development often can only be fulfilled if the contaminated
sites are remediated. The way of cleaning-up contaminated soil is also relevant for sustainable
development.
Since 1993, the German state Baden-Wurttemberg requires by law (VwV, 1993) the consideration of the
environmental burdens of the remediation action itself. Six years later, the legislation for whole Germany
(BBodSchV, 1999) followed Baden-Wurttemberg (table 2). In the mean time, Baden-Wurttemberg
directed the development of a method for the evaluation the potential environmental impacts of soil
remediation options (Volkwein 2000a, 2000b, 2000c; Volkwein et al., 1999, Bender et al. 1998,
Volkwein et al. 1998). Finally in 1999, Baden-Wurttemberg released a software tool (LFU, 1999)
"Umweltbilanzierung von Altlastensanierungsverfahren" ("Environmental balancing of soil remediation
measures") which should be easily used by soil remediation planners in the planning process.
Table 2. Foundation for looking at environmental burdens caused by soil remediation measures itself
Foundation of looking at
environmental burdens caused by soil
remediation measures itself
International legislation or conventions
National legislation
Local legislation
International standards
Example
Rio Declaration, sustainability principle (UNEP,
1992)
German legislation about soil remediation
(BBodSchV, 1999)
Baden-Wurttemberg legislation about soil
remediation (VwV, 1993)
Environmental management systems for
organizations: continuous improvement of overall
environmental performance of activities (ISO
14001:1996)
Between 1996 and 2000 several international standards about environmental management emerged. The
starting point is the standard ISO 14001:1996. ISO 14001 addresses to organizations. An environmental
management system of an organization must include a commitment for a continuous improvement of the
overall environmental performance of the activities of the organization. The overall environmental
performance of an activity (service) can be measure with the management tool life cycle assessment (ISO
14040:1997, ISO 14041:1998, ISO 14042:2000, ISO 14043:2000). Due to the standardization, life cycle
assessment is an internationally accepted tool. No other tool for the determination of the overall
environmental performance of services has such detailed description of the methodology in ISO
standards. Life cycle assessment is applied since 30 years under different names in industrial and other
organizations (Hunt and Franklin, 1996). Life cycle assessment is the first choice for the evaluation of the
overall environmental performance of a service.
The overall environmental performance of a site clean-up can be explored using life cycle assessment
based tools. The decision support tool "Umweltbilanzierung von Altlastensanierungsverfahren"
("Environmental balancing of soil remediation measures") allows the fulfillment of the
93
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
• sustainable development principle for soil remediation and
• the organizational requirements of ISO 14001 for soil remediation organizations.
2 LIFE CYCLE ASSESSMENT FRAMEWORK
The life cycle assessment (LCA) method is still under development, but practitioners can refer to four
international standards (ISO 14040:1997, ISO 14041:1998, ISO 14042:2000, ISO 14043:2000).
According to the ISO standards, every life cycle assessment consists of four parts (Figure 1):
goal and scope definition
life cycle inventory
life cycle impact assessment
life cycle interpretation
LCA framework
f "N
Goal and
scope
definition
It
S N
Inventory
analysis
V J
It
C "^
Impact
assessment
V J
\_
,. 2
^ 2
^ 2
r A
Inter-
pretation
V J
^s
Direct applications:
•,,Umweltbilanzie-
rung von Altlasten-
sanierungs-
verfahren"
(Environmental
balancing of soil
remediation
measures)
Figure 1. Life cycle assessment (LCA) framework adapted from ISO 14040:1997 with
special direct applications for soil remediation planning
Life cycle assessment can analyse whole life cycles of services (products). The primary application of
LCA is the analysis of the use phase of a product (service). Also important is the application in the
manufacturing phase (design phase, eco-design). Another field for the application of LCA is the end of
life analysis or end of pipe service analysis.
3 SOIL REMEDIATION PROCESSES
The soil remediation planning often starts with a historical reconnaissance of the contaminated site. A
more detailed investigation of the pollutants, the amount of the pollutants and the distribution of the
pollutants follows. Preliminary site remediation options are developed. At this point, the evaluation of the
environmental burdens of the different site remediation options can start. The soil remediation planner
can then use the result of this environmental evaluation together with the results of financial, legal, social
and risk assessments to select the finally applied soil remediation option.
94
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
The future regular evaluation of the environmental burdens of the contaminated sites remediation has the
following requirements:
• soil remediation planner should be able to make this evaluation himself
• low demand of time for this evaluation
Therefore, a tool for the evaluation of the environmental burdens should speak the "language" of the soil
remediation planner. This language includes phrases like the following:
mass of soil to be treated
volume of soil to be treated
transport distances for soil transport
time of the clean-up
type of the applied technologies for soil remediation
The tool for the evaluation of the environmental burdens should have predefined processes. Such a feature
can support a quick application of the tool. The process names should be those which the soil remediation
planner uses. The input data for the processes should be in the "language" of the soil remediation planner.
Table 3 lists several processes that might be necessary to make a full soil remediation option.
Table 3. Processes often used in soil remediation measures
General type of
process
Decontamination
(clean-up)
Ensuring
technologies
Secondary
technologies
Group of
processes
Soil washing
Microbiological
soil treatment
Thermal
treatment
Pneumatic tech.
Groundwater
treatment in-situ
Immobilization
Sealing walls,
leak proof walls
Surface sealing
Civil
engineering
Processes, detailed for soil remediation option selection
Soil washing - mobile facility
Soil washing - semi-mobile facility
Soil washing - stationary facility
Microbiological soil treatment - turning bed
Microbiological soil treatment - rotting/composting
Microbiological soil treatment - reactor
Microbiological soil treatment - near to the surface zone
in-situ
Thermal treatment - Herne
Thermal treatment - Deutzen
Vacuum distillation
Pneumatic techniques
Microbiological groundwater treatment in-situ
Reactive walls
Immobilization
Sheet-pile wall
Narrow wall
High-pressure injection wall
Capillary break
Surface covering
Asphalt covering
Foundation, floor plate
Consolidating, compacting
Material consumption - processed earth materials
Material consumption - plastics and concrete
Material transport on-site
Distribution with bulldozer
Wells
95
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
General type of
process
Group of
processes
Waste air
cleaning
Ground-water
cleaning
Transport
Processes, detailed for soil remediation option selection
Excavation work
Soil treatment: sieving, crushing
Adsorptive waste air cleaning
Catalytic waste air cleaning
Biological waste air cleaning
Extraction of ground-water
Ground-water cleaning - stripping
Ground-water cleaning - sedimentation
Ground-water cleaning - precipitation, flocculation
Ground-water cleaning - chemical oxidation
Ground-water cleaning - adsorption
Ground-water cleaning - ion exchange
Ground-water cleaning - dewatering
Mobilization, demobilization
Soil transport street lorry
Soil transport river ship
Soil transport rail train
Transport of persons by car
The processes listed in table 3 are some of the processes included in the software tool "Environmental
balancing of soil remediation measures" (LFU, 1999). The software tool "Environmental balancing of soil
remediation measures" links these process data to generic life cycle assessment data and to a life cycle
impact assessment model. The calculation of the life cycle impact assessment is automated. Results of the
life cycle impact assessment and the life cycle inventory are transformed to an easy to interprete
disadvantage factor table. An example is given in the following section.
4 CASE STUDY "FORMER COMPANY REINIG IN SINSHEIM"
The example "former company Reinig in Sinsheim" includes the comparison of three remedial
alternatives. The contaminated site of the former company Reinig in Sinsheim has an area of 20000
square meter. Mineral oil contaminates 530 cubic meter, polycyclic aromatics (PAH) 750 cubic meter,
and chromium 530 cubic meter soil.
"On-site ensuring" means the excavation and on-site redumping of the contaminated soil. The second
remedial option ("soil sealing") is the simple sealing of the surface by asphalt. The "decontamination"
option requires the excavation and three different treatments. 75 m3 of contaminated concrete is included
in the "on-site ensuring" option, but not in the other two options "soil sealing" and "decontamination". In
the option "on-site ensuring", only 50 % of the PAH contaminated soil is excavated. The other 50 % of
the PAH contaminated soil is under the clamp of the redeposited contaminated soil.
The details of the environmental balancing of the case study "former company Reinig in Sinsheim" are
described in LFU (1999) and Volkwein et al. (1999). The disadvantage factortable 4 shows one important
result. A disadvantage factor 1 means that the parameter value for the remediation option is the lowest
among the compared remediation options. The cumulative energy demand (one of several parameters) is
for the "on-site ensuring" lower than for "soil sealing" and "decontamination". Therefore, the
disadvantage factor for the cumulative energy demand of "on-site ensuring" is one. The "soil sealing"
requires a 20 times higher cumulative energy demand than on-site ensuring.
96
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Table 4. Disadvantage factors for the case study "former company Reinig in Sinsheim"
Impact categories and energy and waste
Cumulative energy demand
Waste total
Waste from contaminated site to landfill
Fossil resources
Water
Land
Global warming
Acidification
Photo-oxidant formation
Toxicity air - remote emissions
Toxicity water
Toxicity soil
Odor - remote emissions
Toxicity air - near emissions
Odor - near emissions
Noise immission 60 dB(A)
Noise immission 66 dB(A)
Sum of disadvantage factors
On-site ensuring
1
1
2
Soil
sealing
20
2
30
5
7
5
5
20
3
30
30
5
1
1
1
Decontamination
4
40
1
4
5
1
5
3
4
4
4
4
3
1
1
1
NOT ALLOWED
A "!" indicates that the other options have a parameter value "0". Soil washing in the option
"decontamination" results in a certain amount of waste from the contaminated site. The options "on-site
ensuring" and "soil sealing" have no "waste from the contaminated site" (parameter value = 0). This is the
reason for the disadvantage factor "!" for the "decontamination" option.
One among several conclusions of this case study is, that the option "soil sealing" is in all analyzed
parameters in table 4 equal or worse than the "on-site ensuring" option. A more detailed discussion of the
results can be found in Volkwein et al. (1999) and LFU (1999).
5 CONCLUSIONS
Life cycle assessment is in use for analyzing whole life cycles of services (products). The primary
application of LCA is the analysis of the use phase of a product (service). Also important is the
application in the manufacturing phase (design phase). Another field for the application of LCA is the end
of life analysis or end of pipe service analysis. The remediation of contaminated sites is a end of pipe
service (repairing service for upstream industrial processes).
A necessity exists for knowing the environmental burdens of remedial actions itself if the sustainable
development principle is applied or if compliance with ISO 14001 is desired. Life cycle assessment is the
tool with the biggest international acceptance for evaluating environmental burdens of services
(products). There are 30 years of industrial experience with life cycle assessment. There are three years of
experience with an international standard about the basics of life cycle assessment (ISO 14040:1997).
There is one year of experience with a publicly available software tool based on life cycle assessment. A
state-of-the-art soil remediation tool-box shall include life cycle assessment based tools.
97
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
6 REFERENCES
BBodSchV: Bundes-Bodenschutz- und Altlastenverordnung (BBodSchV). Bundesministerium fur
Umwelt, Naturschutz und Reaktorsicherheit, Berlin 16.6.1999. TerraTech 1999, 4, 22 - 36
Bender, A.; Volkwein, S.; Battermann, G.; Klopffer, W.; Hurtig, H.-W.; Kohler, W.: Life cycle
assessment for remedial action techniques: methodology and application. ConSoil '98, Sixth international
FZK/TNO Conference on contaminated soil. Thomas Telford, London. 367 - 376
BMBF: Forschungsvorhaben bilaterale Zusammenarbeit BMBF - U. S. EPA. Vorbereitung,
Koordinierung, Durchfuhrung und Auswertung der im Rahmen der deutsch-amerikanischen
Zusammenarbeit geplanten zusatzlichen Untersuchungen an ausgewahlten Altlastensanierungsfallen.
SchluBbericht fur den Bearbeitungszeitraum 1991 bis 1996. Band 3: Im den USA durchgefiihrte
Technikdemonstrationen. Umweltbundesamt, Projekttrager Abfallwirtschaft und Altlastensanierung des
BMBF (Bundesministerium fur Bildung, Wissenschaft, Forschung und Technologie). 1 - 320
Heuschneider, P.; Beyer, J.; Kesselheim, J.; Abstein, G.: Konversion einer Riistungsaltlast zum
Freizeitpark. altlasten spektrum (Berlin) 2000, 2, 115 - 124
Hunt, R. G.; Franklin, W.: LCA - How it came about. Personal reflections on the origin and the development
of LCA in the USA. Int. J. LCA 1, 1996, 4-7
ISO 14001:1996: Environmental management systems. Specification with guidance for use. International
Standards Organization, Geneva 1996 (www.iso.ch)
ISO 14040:1997: Environmental Management. Life cycle assessment. Principles and framework.
International Standards Organization, Geneva 1997 (www.iso.ch)
ISO 14041:1998: Environmental Management. Life cycle assessment. Goal and scope definition and
inventory analysis. International Standards Organization, Geneva 1998 (www.iso.ch)
ISO 14042:2000: Environmental Management. Life cycle assessment. Life cycle impact assessment.
International Standards Organization, Geneva 2000 (www.iso.ch)
ISO 14043:2000: Environmental Management. Life cycle assessment. Life cycle interpretation.
International Standards Organization, Geneva 2000 (www.iso.ch)
LFU: Umweltbilanzierung von Altlastensanierungsverfahren. Version 1.0 Rev. 16. 1999 CDROM
including description of methodology. Landesanstalt fur Umweltschutz Baden-Wurttemberg (LFU).
Available from AHK Gesellschaft fur Angewandte Kartographie mbH, RehlingstraBe 9, 79100 Freiburg,
Germany, FAX +49 - 7 61 - 7 05 22 - 20, www.ahk-freiburg.de
NATO: NATO/CCMS Pilot study. Evaluation of demonstrated and emerging technologies for the
treatment of contaminated land and groundwater (Phase III). 1999 Annual report. Number 235. North
Atlantic Treaty Organization (NATO), Committee on the challenges of modern society (CCMS). 1-187.
www.clu-in.org
NATO: NATO/CCMS Pilot study. Evaluation of demonstrated and emerging technologies for the
treatment of contaminated land and groundwater (Phase III). 1998 Annual report. Number 228. North
Atlantic Treaty Organization (NATO), Committee on the challenges of modern society (CCMS). 1 - 135.
www.clu-in.org
UNEP: The Kosovo conflict. Consequences for the environment & human health settlements. United
Nations Environment Programme (UNEP), Nairobi. United Nations Centre for Human Settlements
(UNCHS), Nairobi. 1999, 1 - 107. www.unep.org
98
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
UNEP: Rio declaration on environment and development. The United Nations conference on environment
and development, Rio de Janeiro from 3 to 14 June 1992. United Nations Environment Programme
(UNEP), Nairobi. www.unep.org/Documents
Volkwein, S.: Governmental policy for soil remediation regarding overall environmental performance.
Proceedings of ConSoil 2000a. 7th International FZK/TNO Conference on Contaminated Soil. Thomas
Telford, London. Submitted
Volkwein, S.: Life cycle assessment of soil remediations with a software tool. Proceedings of ConSoil
2000b. 7th International FZK/TNO Conference on Contaminated Soil. Thomas Telford, London.
Submitted
Volkwein, S.: Comparison of software tools: "REC" and "Umweltbilanzierung von
Altlastensanierungsverfahren". Proceedings of ConSoil 2'
Contaminated Soil. Thomas Telford, London. Submitted
Altlastensanierungsverfahren". Proceedings of ConSoil 2000c. 7th International FZK/TNO Conference on
Volkwein, S.; Bender, A.; Klopffer, W.; Hurtig, H.-W.; Battermann, G.; Kohler, W.: Life cycle
assessment method for remediation of contaminated sites. ConSoil '98, Sixth International FZK/TNO
Conference on Contaminated Soil. Thomas Telford, London. 1069 - 1070
Volkwein, S., Hurtig, H.-W., Klopffer, W.: Life Cycle Assessment of Contaminated Sites Remediation.
Int. J.LCA 1999, 4, 263-274
VwV: Gemeinsame Verwaltungsvorschrift (VwV) des Umweltministeriums und des Sozialministeriums
liber Orientierungswerte fur die Bearbeitung von Altlasten und Schadensfallen. Umwelt- und
Sozialministerium Baden-Wurttemberg. 1115 - 1123. Gemeinsames Amtsblatt ISSN 0939-2726. 41 (33)
30.11.1993.
99
-------
Decision Support Tools NATO/CCMS Pilot Project Phase
APPROACHES TO DECISION SUPPORT IN
THE CONTEXT OF SUSTAINABLE
DEVELOPMENT
Simon J T Pollard, Jonathan Fisher, Clare Twigger-Ross and Andrew
Brookes, National Centre for Risk Analysis and Options Appraisal,
Environment Agency, UK
Steel House, 11 Tothill Street, London, SW1H 9NF, UK
ABSTRACT
Decisions on the management of risks from contaminated land and groundwater have much in common
with many modern environmental decisions. The desire to integrate technical, socio-political and
economic factors during risk management has resulted in the development of decision "frameworks" that
allow an holistic approach within a context of sustainable development1. The application of integrated
decision-making is in its relative infancy, however, and only recently have practitioners in contaminated
land management considered its use. Here, we explore some of the decision tools that are available and, in
the context of contaminated land remediation, explore the practical challenges with respect to integrated
decision-making. Our aim is to set out some of the support tools that might be applied to contaminated
site remediation, particularly at the site level.
INTRODUCTION
Environmental decision-makers have at their disposal a vast array of support tools that have been
developed over the last 30 years2. These tools are typically applied to assist with screening environmental
impacts, for the assessment of risk / benefit trade-offs, for engaging a wider stakeholder community
within decision-making processes and to the integration of technical, socio-political and economic factors
that inform decisions on environmental management . Decision support tools for environmental appraisal
are being used increasingly at the policy, programme, plan and project level across a spectrum of
environmental issues4. This presents opportunities for the cross-fertilisation of expertise and experience
between disciplines and decision-making contexts (Table 1).
Over a similar period, policy makers and practitioners in contaminated land management have been
developing decision-making frameworks of their own5'6, centering on well-established processes of risk
assessment, risk management and risk communication (e.g., Figure 1). These frameworks embody many
of the tools and component processes used by decision-makers elsewhere, including, but not restricted to:
• brainstorming techniques (for hazard identification and conceptual model development for
example);
• scoping and screening (in qualitative risk assessment; in remedial technology selection);
• environmental fate and transport modelling (as a component of exposure assessment);
• sensitivity and uncertainty analysis (in quantitative risk assessment);
• economic appraisal of costs and benefit (in comparing remedial approaches, including the costs of
'do nothing' and the assessment of appropriate times for intervention)
• data analysis (in site investigation); and
• the collection of opinions and lay-perspectives on risk (for risk communication).
100
-------
cs
H
Contaminated site context
a
<0
03
JS
I
a •—
*3 &
03
"3 ^
S o
S 2
d -C
o a
b a>
*p« ?>-
"• a
S §
w 8
V
X
1
H
S '>
Context well established; identification and
analysis of hazards and potential environme]
harm from contaminated sites; use of
probabilistic techniques for dealing with
uncertainties in exposure assessment; sensiti
analysis to assist in remedial technology
selection.
^
^ is "R
«•> e ^
3 |£tf
1 N §
!.§- £
H > '?
C3 H «
a S a s
r, • •> (11 .
^£"1
•sisjgp
H " 2 u -a
O 5 O *= *3
5 JC § «
t3 ft b" '3 f*-.
Hip
"=3 "S § : « ^
O O QH ^J w
*3 di^ V0 S,
W O
"g =H 0
8| 8
•sll
a> «f.s
S ^H 1«
'•3 2 S
s ts -^
^a¥
I1!
i-H . TO
OH W) .
13^.^
1 1 s,
c "S Qj .^2
8 ^ ^ JJL
*5 ^j H '^>
w ^ 1
o J3
'3 fo -So
Environmental, social and econon
impacts of proposed development
summarised in non-technical
language; participatory approach
advocated; can also be applied at
policy or sectoral level as strategi
environmental assessment (SEA).
1
« -g
a ^ §
8 03 |
'i & «
W -^ 03
0 0
On more complex sites, SIA could be used t
analyse the social context of a site or
regeneration programme and to identify key
pressures and benefits / possible constraints
remediation
applied in UK, moR
loping countries, but
s for assessing the
le development.
'et conventionally
in USA and deve
rovide useful basi
part of sustainabl
•^^ ^ o. ^
a a a -a
^ s -a o
5 8 -a S
^ <)H B ^
T3
§ *0 (S
Assessment of the impacts of plar
developments on the social fabric
communities, includes equity issu
and impacts on social processes.
o
ft S
s s
13 %
'§ "
03 03
Nested within human health risk assessment
DH -3 c§
-------
CS
o
1
«= .9 S 'Sb
•g S X> O ^ TO1
. S3 o " a
1 8g*
8"B n .« "3
•4— »
-, -i-H
-*-> CH (D , ,
lj£l
1-SlSf
•f 1 | |.§
O Qi £? ~^ >-
Assessment of risk management
consistent appraisal of environm
water quality improvements by v
as to prioritise these measures in
cost-effectiveness of securing en
benefits.
•a -a i .
.a ^ § g ^
_§ ° to ^ rt
•S ^ .a g &
^
1 % '3 S
^1 | i V
a wj .| u H
?3 -9 & B E« B
Improvement plans; setting insui
corporate environmental account
statements; liability (mergers, ao
divestitures); regulatory complia
of environmental management s>
due diligence; waste minimisatio
^ -g ^ 2 to
. 2 u -S ^ i S 3
0 B g . „ B > Q£
•^3 o &o'-*:3'a3 ° ^
8 '^ "n g •§ ^ g =3
Illicit
l|| 8>-S 's a 1
1 M! |"t||^
O'^firJ 3 ^ 5)
1 "i ^ -| .0 1 i ^ o
OSnPjrtt^38K>HH
^1 p^^-i 11 I
1
r^
-^
& g
W S
T3
The design of sustainable remediation
technologies and programmes; extension beyon
simple capital and operational expenditure to
issues of energy consumption and the secondary
issues of residual products, off-gases and
leachates and decommissioning, for example.
See Volkwein et al.9
•g a
§ ]g
S 0
Manufacture; product replaceme
substitution; supply and product
management.
a
C3
"c3 "2
•^ a
to P
to ^3
§ 8
"H 's
?, ^
^p s
w bo
W 2
ua
f
73
^*
o
4n
'-1
-------
o
9)
3
CS
H
ri3 -4— »
a S y
Om ^
.,_. CM o
"3 • _Q
g 1 2
t+H g JB
"i so
'§8§
ft H/-3
& « a
lit
•s g s
2 S y
B y. •£
.2 o S
•S 22 o
•-"Hi
"3, re 6
ft— B •
re -3 ti to
a c £
"c3
to
•g '§
5 p
^ a
TO (j
"ft —
1*
o ^
• rH QJ
Sb-ri,
22 §>
t+H C3
0 £j
3 -a
•1 a
S-H O
ft -X
ft 5o
< 22
Developing area. Appraisal
framework using sustainability
objectives; potential application at all
levels; flexible iterative process;
indicators to monitor process;
matrix/checklist approach.
.1?
rj
•re -3
B to
+3 S
S2 &i
B ft
&0 c?
O "3 <->
bo ^ B ifl
B ^ a ro -rj
^ to
•3 ^ 2 a ^ c
a -s
w c3 0 .2
S B
-H ^ y
re T ^H to ^H r2
O ft re ft B
Approach to analyzing the
stakeholders involved in the decision
process. Informal to formal
approaches available; enables
systematic consideration of all
relevant parties and the relationships
between those parties. Can provide an
audit trail.
^
•S
<•>
•5 >
y x>
c^ Js
5s re
-i T3
03 s
M
-------
0>
OJO
I
*?
e
TS
cs
Q
TS
a
a
&
a
-*^
a
o
O
o
"S
I
eio
cs
s
cs
o
0>
cs
8JO
5 OS
> 'o
•S -a
S 8
a a
M
_l
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
Recent international reports and conference proceedings have pointed to a need to develop and adopt a
broader range of decision support tools within the community of contaminated land practitioners7'8. Here
we examine how some of the existing decision support tools might be applied to the challenges of
contaminated land within a context of sustainable development.
SUSTAINABLE DEVELOPMENT: ISSUES IN INTEGRATED DECISION-MAKING
A widely used international definition of sustainable development11 is 'development which meets the
needs of the present without compromising the ability of future generations to meet their own needs'. The
UK strategy for sustainable development12, which sets about establishing a better quality of life for
present and future generations, establishes as its key objectives:
• social progress which recognises the need of everyone (equity within and among generations);
• effective protection of the environment (proactive approach to limiting environmental damage);
• prudent use of natural resources (clean and efficient use of non-renewable and renewable
resources); and
• maintenance of high and stable levels of economic growth and employment (improved living
standards for all; quality goods and services; education and skills).
This agenda gives weight to the integration of technical, socio-political and economic factors that inform
decisions across Government, decisions that include, for example, about when and how to remediate
contaminated sites. For practitioners, integration offers an opportunity to bring together appraisal and
decision support tools historically used in isolation. Practical application, however, involves some
considerable challenges, not least in resolving the differences in terminology, philosophy and output that
are associated with individual tools (Table 2). The interface between the sustainable development agenda
and the issue of land contamination is focussing on issues both relating to historic land contamination and
those associated with the wider aspects of soil quality, including but not restricted to13"15:
• bringing land back into early beneficial use;
• the efficient use of national resources to tackle issues of highest risk at priority sites;
• reducing pressure on greenfield sites, thus conserving agricultural land and natural habitats;
• adoption of a suitable-for-use approach towards land remediation;
• prioritisation of remedial action so as to address the worst risks first in relation to the use of the
land concerned;
• distribution of impacts on communities;
• the application of sustainable remediation technologies that conserve land and resources;
• the consideration of point and diffuse sources of soil pollution over the long term;
• the development and maintenance of new partnerships and fora among key stakeholders with
agreements on a common research agenda; and
• the development of monitoring systems that allow early detection of adverse soil changes.
Addressing these issues requires a combination of policy, regulatory and technological responses that in
themselves may require application of integrated decision support tools for a variety of policies, plans,
programmes and projects. For example, the prioritisation of remedial measures in terms of which sites to
act on first, the technology to be used and the appropriate times for intervention are increasingly subject
to economic appraisal alongside issues of risk and technical feasibility16. As one contribution to the
appraisal of sustainable remediation, economic appraisal brings new considerations to the decision-
making process. We use this example below to illustrate some of the integration issues identified in Table
2. A similar discussion can be had for social and environmental components.
105
-------
6*
•s
O
1/1
"o
o
1/1
o>
73
ex
2
ex
CS
H
•fcj
X
0>
•S
q
o
cfi
V
'S
T3
•fcj
«
a
'§
A
•*j
O
O
General commentary
V
ei
a
0>
13
O
^H
"S T3 c2
i— H r™ ^"^
3 3 -5 *"
&-S o S3
QJ (J f X OJ
O C O
lilt
ill!
-H O CO *^
*S3 O co ^H
2 -.
tp w
tj
"H o
> > % ^ ~& M
SS -9:2 38
1^1 5^^
^^asjog
Nil &!
^ § 8 a « y
r_^ ii-H QJ Cu QJ .
1^3 S^-a
•s o a § s 2
^ I § .§ & &
*i ?5 Si T5 f 3
g .y -e Sa -g o
" a g S g "
o r fa S3 5 S
0 s o g u a
is ^-* . ia d> a
mill
H O .2 §5 ^ c§
>^ UH £2
These need to be established and agreed both for consistenc
and for transparency and to make the appraisal practical. Fo
example, there is a growing interest in EIA approaches
encapsulating the entire life cycle of a project, ranging from
the winning of raw materials to the final decommissioning a
disposal stages. But there is a danger of becoming
overwhelmed in excessive and unnecessary detail, so the
analysis should concentrate on issues of significance to the
environmental, economic and social impacts.
CO
.v
o3 co
T3 'co
S £>
^ §
•° C3
60 n-i
?• a
S g
00 eg
a|
3 ts
CO CO g
8 a al
H x a
S .S3 u 2
en co r- a
|t|§
•§§&!
0 S 'to S
f |-H CU ^ QJ
'§ atl
^ a a co
a ^ a o
co a CD co
:s ^
*rt o O
a ° "
•^H CO >,
60 a H
rj 3 c3 S
.3 O"-rt «
o -a S ^
ua '-S
vi c2
11
a
° §
ua
o o"
a 'I
Ig
a S
•S u
o "c3
s .9
2 _§ — •
1 B 1
£H -M '?3
fa ^-^ 2
o to a
u y a
ft .3 03
if!
0^-5
43
It is essential to apply rigorous professional standards in eac
discipline's contribution, while still enabling their
combination for the development of integrated techniques.
H 03
-------
o
9)
3
<§ S -H T3 a _
S ^3 u o 3
Clearest example is of using risk assessment outp
remedial design and selection. These processes st
viewed as distinct, however. There is a greater ne
practice early iteration of risk management decisi
following screening level risk assessments, so as
detailed risk assessment work.
This is difficult because of the specific boundaries that surround
techniques and the form in which each technique produces its findings.
For example, risk assessment may report on the significance of a
contaminant exceeding a threshold, whereas options appraisal
techniques also need an analysis of risk reduction and residual risk to
allow decisions to be made. Often it will be necessary to combine
qualitative and quantitative information. Furthermore, models on which
techniques are based may be incompatible.
^H
-G
S D a
£P & 8
s -a a
•£ .s &
S 7j ra
3 S S3
3 §1
•5b.ar8l *
2 ~ fe §3 8
S §£ a.8
•a £ bfi ., C^
A common problem across the field of risk. Term
differences can be overcome by focusing attentio
fundamental questions for which answers are sou
example: hazard identification - what hazards are
site? How might the consequences be realised? V\
is at risk from these consequences? etc.
The above challenges are compounded by differences in the language
and jargon customarily used by different disciplines and the
interpretations attached to specific terms.
'! ^ §>
03 o .5,
S'S?
g &1 !
•g ,2 a g
a o 53 H
s-l s s
a B -a t~
£llt
£H
0
t3 .s
?
si &
^j ^ M •- OS
3 a -C 2c t« S
a (U IH .t« (U *J
f^ a 8^
Increasing interest in risk assessment design and
Mechanisms for participatory risk assessment not
established. Approach to date has been on technic
assessment followed by 'risk communication' exe
discuss output. US experience suggests limited si
regard and many commentators now point to a ne
engagement and participatory approaches.
Current calls for increased public participation. There is considerable
experience in EIA but less so in risk assessment, CB A or technology
assessment. Structured and focused approaches are necessary, together
with an examination of institutional structures and their capacity for
meaningful public involvement. Monitoring and validation of these
mechanisms is critical.
•8 .a
a 3
0 3 -B
!S|
£ tn S
3 c§ >
Sao
5.8!
C3
•S
£
T3
2
Likely to have a complex range of values associa
contaminated site.
Deliberation is one way of uncovering people's values, but there are
challenges as to how those values are incorporated meaningfully into
decision-making and how representative groups are, such that
deliberative analysis of small groups views may need to be
supplemented by surveys of a larger sample of the relevant population.
w
1
•rt
&
0
O
8S|
:a °
a a
g 22
1 &
S .S
Local effects attributed to sites by individuals / cc
Difficulties of establishing causality and reconcil
observations alongside exposure assessments.
Understanding the perspectives from which lay people address
problems and valuing their local knowledge - challenge as to how local
knowledge sits alongside expert knowledge.
•g
g &
8 ,2
a-§
-.
ti £
-------
o
9)
3
+_* ^H T3
t« t« P C
~ £ "-S
r« ^ a
•i.a o3^
2 | 8 §
*P§ S'3
Integral to most environmental decision-makil
of blight and financial risk ensure this is probl
contaminated sites. Emphasis on quality assur
with respect to consulting advice, site investig
practice and remedial monitoring.
>. !?
1 (8 & S
t^ S -jn W bfi
Recognition that in addition to uncertainty, there may be areas ol
ignorance where, in the presence of significant risk, the precautk
principle may be appropriate. To reflect the prevailing uncertain!
results of appraisals need to be shown as ranges rather than discr
numbers. Uncertainty and sensitivity analysis can assist in highli;
the significance of ranges.
t
1
0 OJ "3 ^ 0 T!
^ l-i ^ C5 <-)
o^« a 1 §
8,-li-S'il
S? P ^ « 0 S3 _
P w P 2 s >i g
•-1 y J3 S-a a
Use of experts at public inquiry especially wit
remedial design and exposure assessments. Iss
to human health effects and toxicological data
tend to be addressed with reference to the deli
authoritative expert panels and committees. A
expert elicitation techniques is common place
performance assessment for elicitation of rece
characteristics and representative 'futures'.
S S
*-H O
The appraisal community is now taking on board participatory
methods, combining scientists and non-experts. This raises impo
challenges and methods of eliciting and sharing technical inform
need to be developed and interrogated.
t/3
&
'o
^^ H 3
'$ m 1 i
£ S3
to •§
? a
bo 03 ^
S C a
^ .2 'S
2 ^8
H 0 03
w
%
t«
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
THE EXAMPLE OF ECONOMIC APPRAISAL
Economic appraisal is an important part of the appraisal of risk management options. It builds on the
findings of the risk assessment and might typically involve the following generic steps:
(i). determine whether there are any existing binding statutory requirements or remedial objectives set
by a higher authority (e.g., Government department such as DETR or the EC), already subject to
their own economic appraisal. If so, then the economic appraisal by regulatory agencies should
comprise examining how to achieve these objectives as cost-effectively as possible. It may be that
the statutory requirements stipulate some caveats regarding the stringency with which Agencies
should apply them such as 'unless excessively costly' or 'unless there are overriding public
interests' (see EC's Habitats Directive). In such cases, the analysis would have to consider whether
such caveats apply;
(ii). where there are no existing binding requirements, it will be necessary to identify the alternative
remediation and risk management approaches and strategies that are available. This should include
issues of the timing of the remediation;
(iii). it is then necessary to appraise the environmental, economic and social impacts of these options to
determine an appropriate remedial objective. This appraisal should include the costs of the options
and their environmental impacts. The environmental impacts might include some impacts that could
readily be assessed in monetary terms (e.g., impacts on local properties). There are likely to be other
important intangible impacts that are difficult to value in monetary terms (eg impacts on human
health and ecosystems). The appraisal should therefore set out fully as possible the level, nature and
significance of these intangible impacts. This assessment should build on the scientific and risk
assessments of the reductions in the likelihood, level and nature of the environmental impacts or
risks that the remediation options could achieve;
(iv). once a remedial objective has been defined, then the appraisal should assess the cost-effectiveness
of the available alternative remedial technologies for achieving it.
The purpose of an economic appraisal is not only to estimate the level of the costs and benefits of the
options, but also to identify the key factors determining them so as to seek out and refine better options
for all concerned with lower costs and greater environmental benefits. An economic appraisal brings the
following essential considerations to the discussion and decision-making processes regarding risk
management options.
(a) Market Failure. Environmental economics focuses on efficiently addressing market failures - which
are external impacts that, in the absence of government action, private producers and consumers do
not take into account. For example, market failure might covers cases whereby the costs to a
developer for remediating a site exceed any increase in the value (gain) of the site as a result of the
remediation so that, in the absence of any government action, the developer would not remediate
the site (to a necessary standard).
(b) Costs. The economic appraisal assesses the costs of the options. These are often (in economics) called
the 'opportunity' costs of the options because the options use resources that could be used for other
beneficial opportunities or purposes.
(c) Law of Diminishing marginal Returns. This is the economist's equivalent of the law of
thermodynamics. It basically means that operators can face increasing constraints (eg extra
resources and time required) as they reduce further the contaminants on a site and achieve extra
environmental benefit. The law of diminishing returns has significant implications for the economic
appraisal of remedial options because the trade offs between reducing the environmental damage
from the contamination and the costs of this reduction become more significant as remediation
progresses.
109
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
(d) Valuation. Finally economic appraisal entails the rigorous and consistent valuation of diverse impacts
(apples and oranges) so that, as far as possible, they can be readily aggregated (without double
counting) in terms of a single commensurate unit (usually money). This can then aid the
comparison of the impacts of the options.
PRACTICAL DECISION-MAKING
A recurrent theme of the application of structured decision support tools to complex decisions is that of
practicality and the associated issues of cost and quality of the final decision. Alongside the current
demands for transparency in decision-making are issues of cost-effectiveness that extend beyond the
remediation technologies employed to the costs of the project cycle as a whole. For example, pragmatists
may argue that the incremental costs associated with the environmental appraisal of remediation projects
can not be justified when viewed in terms of the final outcome, which could have been arrived at through
professional judgement without structured analysis17. One response to this is to view not only the cost-
effectiveness but also the uptake of a decision by stakeholders (including the risk takers) as a critical
indicator of a successful outcome. It is also important also not to regard an appraisal tool as an 'add on'
and, therefore, a burden. Appraisal should be integral to the decision-making process and regarded as an
iterative process.
Apart from the cost issue there is also the question of availability of an adequate skills base to carry out
the appraisal. This is particularly the case for the application of specialised economic, social and
environmental appraisals at the policy18, plan and programme levels. Linkages between different levels of
decision-making need further elucidation and development. In practice, many appraisals can be completed
using inexpensive, but transparent screening techniques (rapid appraisal3) with more sophisticated tools
(technical appraisal) being reserved for complex, higher priority projects. Screening is an accepted
methodology in EIA and is becoming recognised (in the UK) as a means of targeting resources at the
most deserving issues at more strategic levels. Checklist approaches, although with recognised
limitations, have been used widely. The distinction between different 'tiers' of analysis is familiar to
contaminated land practitioners in the application of risk assessment techniques. Furthermore, many
appraisals will have core (fixed) and non-core (variable) aspects to their analysis19'20 and decision-makers
can streamline their appraisal efforts by identifying core issues at a screening stage and focusing any
additional effort on decision critical aspects of the analysis. This 'tiered' philosophy of approach is a
common feature of most site investigations and familiar to practitioners in contaminated land assessment.
CONCLUDING REMARKS
Risk management frameworks and their decision support tools have historically offered a systems
approach to addressing environmental risk problems. The sustainable development agenda requires a
more holistic approach, often with the integration of qualitative judgements alongside quantitative
information1. At present, the debate amongst contaminated land professionals as to applicability of these
tools has extended as far as the valuation of the intangible benefits of land remediation, but will need to
extend to address explicitly the social impacts in order to embrace fully the objectives of sustainable
development. Integrated decision support tools are required to assist this. In applying them however, it
will be critical not to lose sight of the practicalities of application, the need for a transparency of approach
alongside the defensibility of the technique and the over-riding objective, which is to make quality
decisions on bringing contaminated land back into beneficial use.
110
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
ACKNOWLEDGEMENTS
The authors gratefully acknowledge review comments received from Ian Martin (Environment
Agency), Paul Bardos (r3 Environmental Technology Limited) and Malcolm Lowe (DETR)
during the preparation of this paper.
BIBLIOGRAPHY
1. See: Bell, S. and Morse, S. (1999) Chapter 4. In: Sustainability Indicators, Earthscan, London, pp. 77-
103.
2. Pollard, S.I.T., Brookes, A., Twigger-Ross, C. and Irwin, I. (1999) In: Risk Analysis: Facing the New
Millennium, (ed. L.H.I. Goossens), Proceedings 9th Annual Conference, Rotterdam, October 10-13,
1999, Delft University Press: 559-563
3. Department of the Environment, Transport and the Regions (1998) Review of Technical Guidance on
Environmental Appraisal, A Report by EFTEC, Department of the Environment, Transport and the
Regions, London, 86pp (ISBN 1 85112 130 7)
4. USEPA (1999) Integrated Decision-Making in the 21st Century, Peer Review Draft, May 1999;
available at: wwwjrgajfoWscj^
5. Canadian Council of Ministers of the Environment (1991) National Guidelines for Decommissioning
Industrial Sites, CCME, CCME Secretariat, Winnipeg, Publication CCME-TS/WM-TR013E
6. Department of the Environment, Transport and the Regions and Environment Agency (2000) Model
Procedures for the Management of Contaminated Land, CLR11, Procedure for Risk Assessment,
DETR/EA, in preparation
7. Ferguson, C. et al. (eds.) (1998) Risk Assessment for Contaminated Sites in Europe, Volume 1.
Scientific Basis, LQM Press, Nottingham, pp. 144-146
8. See, for example: Nijboer, M.N. et al. (1998) REC: A Decision Support System for Comparing Soil
Remediation Options based on Risk Reduction, Environmental Merit and Costs. In: Contaminated
Soil '98, Thomas Telford, London, pp.1173-1174.
9. Volkwein, S. et al. (1998) Life Cycle Assessment Method for Remediation of Contaminated Sites. In:
Contaminated Soil '98, Thomas Telford, London, pp. 1069-1070.
10. For a more general discussion, see for example: House of Lords (2000) Science and Society, 3rd
Report of the Select Committee on Science and Technology, HL Paper 38, The Stationery Office,
London, pp.37-44.
11. World Commission on Environment and Development (1987) Our Common Future (The Brundtland
Report), Oxford University Press (ISBN 0019 282080 X)
12. Cm4345, A Better Quality of Life: A Strategy for Sustainable Development for the United Kingdom,
96pp., The Stationery Office, London, 1999
13. Circular 2/2000, Environmental Protection Act 1990 - Part IIA - Contaminated Land, Department of
Environment, Transport and the Regions, Annex I, paras. 6-7
14. Pollard, S.I.T. and Herbert, S.M. (1998) Contaminated Land Regulation in the UK: The Role of the
Environment Agency (EA) and the Scottish Environment Protection Agency. In: Contaminated Soil
'98, Thomas Telford, London, pp.33-42.
15. Puri, G. and Gordon, I.E. (1998) Soils and Sustainability - A Natural Heritage Perspective. In:
Contaminated Soil '98, Thomas Telford, London, ppl-5.
16. Environment Agency (2000) Costs and Benefits Associated with Remediation of Contaminated
Groundwater: A Framework of Assessment, R&D Technical Report P279, Environment Agency
R&D Dissemination Centre, WRc, Swindon, SN5 8YF, UK
111
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
17. For a discussion of views offered, see: Bardos, P., Nathanail, P. and Weenk, A. (1999) Added
Environmental Value: A Tool to Help Understand the Effects of Remediation within the Context of
Sustainable Development. Summary of a Workshop held at University of Nottingham, R&D Project
P5-23/01, Environment Agency, Olton Court, Solihull, UK, B92 7HX
18. Pollard, V. (2000) Integrated Appraisal of Environment Agency Policies, National Centre for Risk
Analysis and Options Appraisal, Guidance Note 35, Environment Agency, London.
19. Bardos, R.P., Kearney, T.E., Nathanail, C.P., Weenk, A. and Martin, I.D. (2000) Assessing the Wider
Environmental Value of Remediating Land Contamination, Proceedings 7th International FZK/TNO
Conference on Contaminated Soil, in press
20. Gray, P.C.R. and Wiedermann, P.M. (1999) Risk Management and Sustainable Development: Mutual
Lessons from Approaches to the use of Indicators, /. Risk Research, 2(3), 201-218.
112
-------
Decision Support Tools NATO/CCMS Pilot Project Phase III
REVIEW OF DISCUSSIONS ABOUT DECISION SUPPORT
ISSUES IN EUROPE AND NORTH AMERICA AT THE
NATO/CCMS SPECIAL SESSION, AND OVERALL
CONCLUSIONS
Sullivan1, T.; van Veen2 H.J.; Davidson3, L.; and Bardos4, R.P
Brookhaven National Laboratory, Upton, N.Y., 11973, USA
1. TNO MEP, Postbus 342, 7300 AH Apeldoorn, The Netherlands,
2. EarthFX, Inc., Ottawa Ontario Canada, K1V 8J5
3. r3 environmental technology Ltd, PO Box 58, Ware, SG12 9UJ, UK
1. INTRODUCTION
Environmental management of contaminated lands is a complex process requiring a wide variety of
decisions encompassing different technical, social, and political questions. Decision support for
contaminated land management is an emerging field. Currently, a consensus for the best approach for
using decision support does not exist. A special session on decision support was conducted at the
NATO/CCMS meeting held in Wiesbaden Germany in June 2000. The NATO/CCMS Pilot Study on
Remedial Action Technologies For Contaminated Soil and Groundwater Phase 3 is a multi-national
forum for the exchange of information on emerging remediation technologies and technology
demonstration. The Pilot Study is an activity of NATO Committee on Challenges for Modern Society
(Web site: http://www.nato.int/ccms/info.htm).
During the special session two guided discussion sessions were conducted and one set of questions to the
conference participants was prepared. The discussion sections focused on obtaining information on the
uses of decision support tools and the strengths and limitations of these tools. The questionnaire focused
on gathering information on the use of decision support in the different countries participating in the
meeting. This paper summarizes the findings of this information gathering exercise.
2. TECHNICAL BACKGROUND TO THE DECISION SUPPORT SPECIAL SESSION
Environmental management of contaminated lands is a complex process requiring a wide variety of
decisions encompassing different technical, social, and political questions. The scope of contaminated
land management problems range from minor contamination of a single site with a single contaminant, to
multiple sources of different contaminants on a single site, to management of numerous contaminated
sites in terms of sustainable development. The types of decisions that have to be made include:
• Identification / registration of problem sites
• Overarching decisions involving technical and social criteria (e.g., setting contaminated land
policies)
• Setting management goals in a regional planning context (or corporate planning context)
• Prioritization of actions between sites
• Determining a course of action for a particular site
• Determinations within the individual steps of risk assessment / management for a particular site
(e.g., how many samples are needed to support decisions on where to remediate).
The breadth in scope and sheer number of decisions required for contaminated land management has led
to confusion as to what constitutes decision support. In this discussion decision support is taken to be: the
assistance for, substantiation and corroboration of, an act or result of deciding; typically this deciding
will be a determination of an optimal or best approach (Bardos et al ibid). Although obvious, it is
important to point out that decision support is NOT the same as making a decision. Decision support is
113
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
the process of taking experience, data, and problem specific knowledge and the analysis and integration
of this information to produce knowledge that assists the decision maker(s).
Decision support is one component of several in the decision making system. The others are:
information/data, the management of that information/data, means of modeling / visualization of
complicated information in a way that facilitates its interpretation, and gray matter. Gray matter means
the human intellectual input that: sets out the technical approach to the decision making process;
interprets decision making knowledge and reaches the decision. Figure 1 presents these components in a
simple schematic. Figure 1 emphasizes the interdependence and feedback between different aspects of the
problems through the two-way arrows. Eventually, the information is used in the decision making
process.
An example of a decision making process might be the determination of which remedial options to use for
a particular site. In this scenario, the problem begins with definition of a technical approach to the
problem. Data are collected and managed. The data includes any information used to assess the problem
including measurements of contamination and soil and groundwater properties, technical performance of
remedial options, and costs of remedial options. The data are utilized directly for decision support in some
cases. In most cases, the data are used in models that further analyze the data to provide information
necessary for supporting decisions. The outputs from the modeling require interpretation on issues such as
are the proper models and parameters being used for the analysis. The decision support variables also
have to be interpreted in terms of their adequacy in supporting decisions (e.g., what uncertainties are there
in the variables and will these uncertainties possibly lead to a different decision).
Figure 1 highlights the need for detailed thinking about the problem using gray-shaded boxes that use the
term 'gray matter.' Decision support tools and techniques can supplement the decision process but cannot
replace critical thinking, analysis, and judgment.
Technical Approach; Gray
DATA MANAGEMENT
MODELS
VISUALIZATION
DECISION
SUPPORT
Interpretation:
Gray Matter
Gray
Figure 1. Flow diagram of the decision making process.
114
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
A number of tools are possible to support the decision maker. This discussion paper takes "decision
support tool" to be anything used as an instrument or apparatus in one's occupation or profession
(Bardos et al, ibid.). Thus, a decision support tool (DST) is some kind of a product, which has the aim of
supporting decision making.
In all cases contaminated land problems are resolved as a result of a series of inter-related decisions. A
DST typically facilitates one or more of these decisions, as illustrated in Figure 2.
Decisions
Tools
System Boundaries
Figure 2. Schematic representation of the relationship between decision tools and decision making.
A DST can be written guidance on how to assemble and analyze information needed to support a decision
(e.g., regulatory guidance on risk assessment, sustainable development, cost-benefit analysis, etc.).
Alternatively, it can be a software tool that facilitates the data analysis and produces decision knowledge
(e.g., costs, risks, etc.). In some cases, the software tools have
codified the regulatory guidance to permit relatively easy and more consistent application of the guidance.
Figure 2 also shows that several decision support tools may be used in addressing contaminated land
management. The entirety of the decision steps is the decision making system. No current single tool
addresses the entire process. This is an important distinction, as many people would like a single tool (a
decision support system) that could address all of the decisions. This would increase transparency (i.e.,
clarity of the process to all stakeholders) and reproducibility of the decision making process. However,
because of the breadth and scope of decisions that need to be made this is not practical.
The system boundaries represent the constraints to addressing the problem and include regulations, time,
money, and other limitations. Decision tools work within the system boundaries to provide information
that supports the decisions. As shown in the figure, some tools will address a single decision (e.g., what
region needs to be remediated to reduce human health risks to an acceptable level), while others will
address multiple decision variables (e.g., selection of a remedial approach based on economic costs,
protection of human health, technical feasibility of the approach, and stakeholder concerns).
115
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
In general, the use of decision support tools and techniques is an emerging field in contaminated land
management. While some principles such as the use of human health risk assessment in decision making
are widely accepted approaches for decision making, many areas such as ecological risk, multi-criteria
analysis, life-cycle analysis, and financial risk analysis are only emerging as decision support tools. Even
for human health risk assessment where guidance has been published in many countries, there is still
much debate over the best approach (e.g., should specialized risk assessments be done for the young and
old who may be more susceptible to exposure from contamination) to perform the analysis.
3. OUTLINE OF DISCUSSION SECTIONS
Two guided discussions took place during the special session, reviewing the papers presented (and
included elsewhere in this report) and bringing to bear the delegates' own range of experiences from many
countries. In addition, many delegates also provided written feedback over the course of the meeting. A
list of delegates who attended is presented as an Annex to this report.
Ing Johan Van Veen led the first discussion section and focused on addressing the following questions
1) Are decision support tools useful?
2) How are DST being used?
3) What is the role of stakeholders in the decision process?
4) What common factors emerge between decision support tools?
Mr. Laurence Davidson led the second discussion section with the intent of determining the advantages
and disadvantages of using DST.
The list of questions provided to the participants were:
1. How is DS considered in your country as a discipline or technique?
2. How is DS for remediation used in your country (e.g., types of applications, frequency of use? -
Always, sometimes, almost never)?
3. In your view how well are information needs for decision making about remediation understood?
4. What is your view of the usefulness of Decision Support for selection of remedial options / risk
management? Is DS used to support technology selection?
Participants from Austria, Belgium, Canada, the Czech Republic, Germany, Greece, Italy, Japan, the
Netherlands, Norway, Switzerland, Turkey, the United Kingdom, and the United States supplied answers
to these questions.
The following summarizes the results of the discussions and responses to the questions. In several cases,
there was an overlap between the different discussions and questions. The following reports the findings
as they occurred. No attempt was made to consolidate the different thoughts into a more concise manner.
4 FIRST DISCUSSION SECTION: APPLICATIONS OF DECISION SUPPORT TOOLS
4.1 ARE DECISION SUPPORT TOOLS USEFUL?
There was a consensus that DST can be useful not only in facilitating decision making, but also in helping
to ensure consistency and transparency across decisions. However, this was strongly dependent on the
DST approach. Unintelligent use of DST was perceived as counterproductive.
Written guidance on how to provide decision support knowledge was felt particularly useful. An example
of these types of tools include written guidance on the approach and parameters to be used in human
health risk assessment. Several people felt that these guidance types of tools were essential and in some
cases adhering to the guidance is required by national laws.
116
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
There was less agreement on how useful software tools were in supporting the different decisions in the
contaminated land management process. For example, most delegates agreed that human health risk
assessment and cost-benefit software tools were valuable and widely used. Delegates could also see the
usefulness of sample selection based on geostatistical analysis - yet these types of approaches are not
widely used. However, while a number felt that DST could be useful for remedy selection, others felt that
the use of software DST for remedy selection was not particularly useful, given the site-specific
complexity of contamination problems and the absence of reliable general cost data.
A number of concerns were raised about the use of software DSTs in general. Often these tools use
specific datasets and extensive assumptions. While the data and conceptual model are, in reality, the
technical foundation of decision support, if it is unclear what the datasets and assumptions are, their
relevance to the problem in question is unclear, and misuse of the tool a strong possibility. One delegate
went further. He felt that even where a DST made transparent use of data, knowledge and assumptions,
the mere availability of easy to use DST software presented risks of decision making being undertaken by
inadequately skilled individuals.
These criticisms do not reflect a meeting consensus, but rather part of the range of views expressed. Other
delegates felt that the way in which DST could improve the accessibility of data, analysis, and
interpretation beyond those with expertise in the field was fundamentally a good thing. It allowed many
stakeholders to actually "have" their stake in decision making. Those ultimately paying for or approving
remediation decisions, and many of those wishing to influence decision making, are not necessarily
contaminated land specialists.
However, it was suggested that the use of the tools still requires training and expertise in the different
aspects of the decision making process and the analyses used by particular tools. The training should
include guidance on the range of conditions over which the tool is applicable. This supports the notion
that the tools can not be used to replace expertise, but only to enhance it.
The majority of delegates agreed with aspirations for decision support to help to make the decision
making process transparent, documented, reproducible, (hopefully) robust and provide a coherent
framework to explore the options available (Bardos et al ibid). However, not all DST match up to these
aspirations, and indeed the supporting datasets and assumptions of some DST are questionable for many
applications.
4.2 WHAT IS THE ROLE OF THE STAKEHOLDERS IN THE DECISION PROCESS?
A stakeholder is any individual or group that has an interest in the particular contaminated land
management problem. Stakeholders can include problem holders, environmental service providers,
federal, state, and local regulators and public health officials, local businesses, citizens, and citizen
groups. (PCCRARM, 1997; SNIFFER, 1999). The different perspectives held by stakeholders often leads
to conflict in determining an approach to contaminated land management. In most countries, the problem
holder or their consultant(s) analyzes the problem and suggests a remedy to the regulatory body.
Typically, the public and other stakeholders are often informed of these recommendations at a later stage,
often when decisions in principle have already been taken.
Many delegates felt that early stakeholder involvement is beneficial both to avoid later delay and
costs from subsequent arguments with unconsulted stakeholders and for reasons of open
"governance". Inclusivity in decision making is a part of sustainable development, which is an important
policy driver in many countries. However, concern was expressed by several delegates that this
inclusivity could lengthen the time taken to make a decision and in some cases be counterproductive. On
the other hand, failure to include stakeholder viewpoints can often lead to more severe management
problems later. Several suggested that stakeholders must be made part of the decision making process, but
they should not be given control of the decision making process. Strong leadership and communication
117
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
skills were identified as being crucial to dealing with all of the interested stakeholders, but maintaining an
ability to actually make decisions.
4.3 HOW ARE DST BEING USED?
A number of applications of DST were mentioned during the discussions. Four major categories of use
were identified.
• The first is written guidance produced, for example, by regulatory bodies. The guidance
approach is used in a number of countries to enable a more consistent approach to contaminated
land management.
• The second category is use in identifying sites on a regional or organizational (e.g.,
corporate) basis and setting management / policy goals, Activities supported include the
identification of suspect sites, cataloguing suspect sites and setting broad "policy" objectives,
which may be linked to a variety of spatial planning considerations, for example zoning of
development and regional economic policy such as attracting inward investment.
• The third category is the use of DST for prioritization among different sites within a single
area of responsibility. This activity is necessary where a number of suspect sites have been
identified. Resources are not available to treat all simultaneously so the most urgent must be
treated first.
• The fourth category, which is the most commonly recognized application, is use of DST for
specific tasks at a single site. Examples of these type of approaches include analysis of human
health risks, remedy selection, site characterization, and cost-benefit analysis. In most
applications, a single decision criterion is evaluated. However, use of multi-criteria analysis
(MCA) and life cycle analysis (LCA) approaches are often found.
Other important findings from the discussion were:
• Human health risk tools are the most widely used of any DST.
• For the most part, implementation of the tools is in the hands of the consultants and other
technical specialists. Regulatory staff use them to a much lesser extent and the public and other
stakeholders rarely use DST.
• When DST are used they tend to be only a small part of the decision process.
4.4 WHAT ARE THE COMMON FACTORS FOR DECISION SUPPORT?
Many decisions are required for contaminated land management. The decisions range from site and
problem-specific questions that are largely based on technical and economic concerns (e.g., what is the
best remedy to clean the site) to national questions that are largely based on societal concerns (e.g.,
prioritization of resources for the management of contaminated land to permit sustainable development).
Although the emphasis on the decision variables may differ between different problems, they are
interrelated. Site-specific problems can be influenced by societal concerns (e.g., neighbors may object to a
technically viable solution such as incineration of wastes because they are concerned over airborne
releases).
Decision support tools integrate data and report results in terms of a simplified but representative
decision information. For example, assume that human health risk is one decision parameter for deciding
if monitored natural attenuation is acceptable, or if a more aggressive remediation scheme is required.
Many software programs predict the groundwater flow path and rate. While this information is required to
analyze a contaminated aquifer, it alone does not address the consequence of the contamination and,
hence, it is not a decision support tool. A decision support tool would take the information from the
groundwater flow simulation and integrate it with information on the source strength and duration,
contaminant transport processes (for example, removal by biodegradation), and exposure pathways and
118
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
parameters (e.g., receptor location and use of contaminated water) to estimate human health risks over
time.
Stakeholder involvement is an important aspect of the decision process and helps to achieve a
solution for contaminated land management that is acceptable to all. Stakeholders may not always
agree on an approach for contaminated land management. In this case, the regulators are often the
mediators between the different stakeholders.
Risk management decision support tools are the most commonly used decision support tools. A number
of delegates also identified cost-benefit decision support tools as having widespread application.
5. SECOND DISCUSSION SECTION: ADVANTAGES AND DISADVANTAGES OF DST AND
GENERAL ISSUES ARISING FROM THEIR USE
5.1 WHAT ARE THE ADVANTAGES OF USING DECISION SUPPORT TOOLS?
The major advantage of using appropriate DST's is in helping to ensure the decision making process is
robust, consistent, transparent and reproducible. Specific advantages of DST include:
• DSTs provide a method to analyze multiple scenarios. Consideration of a range of scenarios can
increase the confidence when making a decision.
• DST can be used to optimize contaminated land management (leading to lower costs).
• Some DSTs can incorporate uncertainties into the decision framework. Decisions in contaminated
land management are always made with some degree of uncertainty. Addressing this directly can
enhance the decision making process. For example, DST can estimate the volume and costs of
remediation required as a function of the degree of certainty in achieving human health risk goals
(Stewart, 2000) or financial risks (Finnamore, 2000). This permits the decision to be based on the
problem holder's aversion to failure.
• DSTs can provide means to document all parameters and assumptions used in the analysis for a
particular decision (see subsequent discussion of data management systems).
• DST can improve communication between various stakeholder groups.
• DST can be used as an educational tool. For example, the effects of changing parameters on the
decision variable can be demonstrated.
• DST can improve the transparency of the process through documenting assumptions and
explaining the approach used to reach a decision.
5.2 WHAT ARE THE DISADVANTAGES TO USING DST?
• Gaining acceptability of the tool with all stakeholders is often difficult. It takes time and effort to
educate other stakeholders on the use of a tool. If the tool is perceived to be a 'black box'
stakeholders not involved in the application of the tool will not trust the results.
• A common approach to DST is to provide output in the form of a single set of decision variables,
and in some cases a single variable or index. In reporting only the decision variable the rationale
behind its algorithms, supporting data and assumptions may not be understood. The effect of this
reporting approach may be to perpetuate a lack of trust of the analysis, which may be viewed as
"black box" information. This is likely to be a particular problem where DST are used or
interpreted by "non-experts". It also flags the need for clarity and good supporting information on
the part of the system designer AND user.
• Decision support tools must be maintained to keep current. For example, for remedial options as
new cost data are obtained they must be incorporated into the appropriate database for use in the
analysis. In addition, human health risk decision support tools often have a database for risk
parameters. These parameters are continually being updated to reflect the latest scientific
findings.
119
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
• Garbage In - Garbage Out. A decision support tool is only as good as the data and assumptions
used to perform the analysis. The assumptions include not only those used to develop the DST,
but also those used in the conceptual model of how to represent the problem. Therefore, the
analyst should be trained in the use of the tool and in the approach to represent the contamination
problem. (See also Section 4.1).
5.3 WHAT ARE THE ISSUES IN THE USE OF DST?
During the discussion it became apparent that there were many issues that could not be claimed to be an
advantage or disadvantage. For example, ease of use of the decision support tools was cited as an issue.
Many people wanted tools that were easy to use, while others were concerned that without proper training
the easy to use tools could be prone to misuse. For this reason, a third category, issues in using DST was
added and the following issues identified.
• The use of many types of DSTs is in its infancy. In general, DSTs need to gain acceptance from
all of the stakeholders, provide training on how to effectively use them and guidance on when
they would be useful.
• The value added by using DSTs needs to be demonstrated. Purchasing a DST, learning how to
properly operate a DST and getting other stakeholders to agree that the DST is appropriate for the
problem can be expensive and time consuming. If all of this work does not lead to a better
decision or more efficient process to reach the decision, use of the DST could be considered
inappropriate. Anecdotal evidence was presented at the meeting indicating that in one case, use of
a DST saved several million dollars on the remediation project. Situations like this need to be
thoroughly documented and subjected to independent peer review.
• Contaminated land management requires good data management practice. It was suggested that a
data management system is not a DST but it is an adjunct that supports the quality of DST
analysis. As such, the data management system should be independent of individual DST or
visualization tools. An ideal situation might be where a single data management system was used
both to store basic data from its various sources and the interpretation of that data provided by
visualization tools and DST. Indeed the data management package might be handed on across
organizations on a CD-ROM to ensure that source and interpreted data is kept secure and well
referenced. Providing everyone with the same data will allow independent analysis by other
stakeholders using the same data. Maintaining a centralized data management system can also
lead to better quality control of the data as all changes to the database will go through the data
administrator. This will help insure that all data analyses will be performed with a common data
set.
• There are gaps between the latest developments in decision theory and their implementation in
DST. This is to be expected because the development of the theory generally precedes the
implementation in DST. However, it highlights the need to continually maintain and update the
DST, as new information becomes available.
• Validation/Verification of a DST is required, but difficult to perform. Validation refers to the
demonstration that the DST performs as expected. Validation can be achieved by comparison of
DST results with known solutions or with results from other accepted DST. Verification refers to
the demonstration that the DST can accurately predict the behavior of the system. Due to the
natural variability in contaminated land problems, lack of data, and the need for simplifying
assumptions to represent the actual conditions it is generally not possible to verify the DST.
• DSTs are supposed to enhance transparency of the decision process. However, their development
requires highly specialized knowledge and skills. For example, DST may implement state-of-the-
art models for any or all of the following: geostatistics, subsurface flow and transport, human
health risk assessment, ecological risk assessment, economic analysis, and decision theory. This
highlights the previously identified need to educate and train stakeholders in the use of DST and
the limitations in their use.
120
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
• The results from using DST may receive unwarranted credibility through the cloak of scientific
rigor. The concern expressed was that if a well-accepted DST is used in the analysis, people will
blindly accept the results without critically analyzing the assumptions and parameters. This
highlights the need to remember that although the DST may be quite sophisticated in its analysis
techniques it is just a tool. The decision process should still be based on thinking.
5.4 THE IMPORTANCE OF DATA MANAGEMENT
Decision support can be greatly improved through the use of data management tools that store the
information electronically and permit its use by all stakeholders. A concern was expressed by some of the
participants that if each DST had its own dataset this could lead to inconsistencies. Proper data
management would remove this problem and can lead to improved quality control of data. Ideally, the
data management system would contain all of the data related to the contaminated land management
problem and be the sole source of data for decision support analyses. The different DSTs would access
the database and extract the data needed for their analysis. Use of a centralized data management system
would help improve consistency.
5.5 WHAT ARE THE ISSUES IN MULTI-CRITERIA ANALYSIS (MCA)?
Multi-criteria analysis is a well-established technique for optimizing decision making, however, use of
MCA for decision support of contaminated land management is an emerging technique. In MCA, several
alternatives are ranked against a list of criteria. These criteria can include costs, human and ecological risk
reduction, societal values for the benefits of remediation, technical feasibility, and so on. From the
preceding example, it is clear that each of these criteria will have different measurement scales and may
rely on subjective judgement. Each alternative is evaluated against each criterion and given a score. The
scores are then normalized to a single scale. Often economic cost is used for the scale. Using the
normalized score, each criterion is given a weight to reflect its relative importance to the decision. For
example, meeting societal values may be given a weight of 0.3, while meeting ecological values may be
given a weight of 0.1. Then, for each alternative, the individual scores for meeting each criterion are
multiplied by the weight for the criterion and a total score is obtained. The total scores for each alternative
are then ranked to support the decision on selection of an alternative. As MCA is an emerging practice in
this field, there is little guidance on how to score the different criteria, normalize to a single scale or select
the weights applied to each criterion. This has led to the following questions for the use of MCA.
• Does it make sense to normalize all criteria to a single scale? Often everything is assigned a so-
called monetary value. Is this the best choice?
• What is the best way to integrate more subjective data (e.g., societal values) with more technical
data (e.g., costs or risks)?
• What is the basis for obtaining the criteria weighting factors? Optimally, they would be obtained
by consensus among all of the stakeholders.
• How is transparency in the decision process maintained when weights and scoring are subjective?
• Is the process rigorous and robust when using subjective normalization and weighting?
It is clear that there are major concerns about the process of quantifying subjective data and comparison
of dissimilar criteria. In order for MCA to become an important tool for contaminated land management,
these issues will have to be addressed and general guidance on acceptable approaches is needed.
6. RESPONSES TO THE QUESTIONNAIRE
6.1 HOW IS DECISION SUPPORT USED IN YOUR COUNTRY?
In general, three categories of response to this question were obtained: a) not used at all; b) used in the
form of guidance for best practices; or c) used for site-specific problems. In some countries, DS is not
121
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
widely used. In most countries, DS in the form of regulatory guidance is frequently used and its
application is required by some nations. When DS is being used, human health risk assessment and cost-
benefit analysis were the most frequent applications. Multi-criteria analysis and ecological risk
assessment are emerging uses for DS. LCA is being used on a limited basis for special problems. All
respondents considered DS to be a technique rather than a separate discipline.
The following example applications were supplied in the responses:
• Regulatory guidance for conducting human health risk assessment or best practices for
remediation.
• Prioritization of projects for obtaining state funding, and for social and land-use planning;
• Data management,
• Human and ecological risk assessment,
• As a communication tool for the spatial context for risk and through visualization of data,
• As a method to insure uniform application of regulations,
• To support selection of monitored natural attenuation as a risk management strategy,
• Optimization of remedial technology operation parameters to minimize costs.
6.2 HOW WELL ARE INFORMATION NEEDS FOR DS UNDERSTOOD?
There was a range of perceptions on this issue. Some people believed that information needs were well
understood, while most did not. Most people felt that the needs were understood at the thematic level (i.e.,
contamination data, risk data, etc.), but not at the working level (amount of data required to make a
defensible decision). Most agreed that the information needs were well understood by specialists and
researchers, less understood by project management and regulators and not understood by stakeholders
that are not involved in the analysis process. A few responses identified the following issues in
information needs.
• Several areas of science are not well understood. Improved understanding could lead to better
decision-making. Areas identified include long-term performance and cost data for remedial
techniques, better understanding of subsurface flow and transport, and toxicology data.
• For MCA, using subjective criteria such as the value of remediation to society, approaches to
quantify the value in monetary terms are needed.
• Data quality needs are not well understood. The impact of natural variability and uncertainties in
the data on the decision need to be addressed.
One respondent pointed out that the challenge for decision support tools is to simplify the systems so that
data needs are reasonable in terms of the number of parameters and the cost to collect the data. The
simplifications have to be balanced against the loss of technical accuracy in the results (i.e., does the loss
of technical accuracy and, therefore, increased uncertainty impact the decision?). Accuracy is only one of
several required attributes for decision information. The overarching question being asked is how to best
manage the contaminated land given the problem constraints. For example, in the UK the emphasis is
now on data quality that is fit for purpose - in some circumstances this may imply that a fixed budget is
spent on more information but of lower (but adequate) quality.
6.3 WHAT IS YOUR VIEW OF THE USEFULNESS OF DECISION SUPPORT FOR
SELECTION OF REMEDIAL OPTIONS / RISK MANAGEMENT? IS DS USED TO
SUPPORT TECHNOLOGY SELECTION?
Many respondents felt that DS was useful for initial screening in the selection of remedial options. A few
respondents felt that it was also useful in the final selection of a remedy. Those that did not feel DS was
useful for final remedy selection indicated that the uncertainties in the cost and performance data were too
high for new and emerging remedial technologies to permit use of decision support tools. Most
122
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
respondents agreed that decision support is useful for risk management. In many countries, guidance on
risk assessment is available, and risk assessment is routinely used.
Many respondents generalized the question to express how decision support was most useful in their
country. Most respondents felt that decision support was very useful in the form of regulatory guidance to
obtain a consistent analysis framework. This helped set the stage for dealing with the different
stakeholders in a fair and consistent manner. Other advantages cited for decision support included:
• Improved communication with stakeholders. Visualization of data was acknowledged as an
important method of communication.
• Better management, integration and use of data. The use of an overarching data management
system that managed the data for all decision support tools can improve quality control and
permit greater access to the data.
• Ability to determine key processes and parameters that impact the decision.
• Better transparency to the decision process.
7. CONCLUSIONS AND FUTURE DIRECTIONS
Many decisions are required for contaminated land management. The decisions range from site and
problem-specific questions that are largely based on technical and economic concerns (e.g., what is the
best remedy to clean the site) to national questions that are largely based on societal concerns (e.g.,
prioritization of resources for the management of contaminated land to permit sustainable development).
Although the emphasis on the decision variables may differ between different problems, they are
interrelated. Site-specific problems can be influenced by societal concerns (e.g., neighbors may object to a
technically viable solution such as incineration of wastes because they are concerned over airborne
releases).
Decision Support involves integration of expertise and data, followed by analysis and interpretation of the
results to produce outcomes in terms of decision variables (health risk, cost, suitability, etc.). For
example, assume that human health risk is one decision parameter for deciding if monitored natural
attenuation is acceptable, or if a more aggressive remediation scheme is required. Many software
programs predict the groundwater flow path and rate. While this information is required to analyze a
contaminated aquifer, it alone does not address the consequence of the contamination and, hence, it is not
a decision support tool. A decision support tool would take the information from the groundwater flow
simulation and integrate it with information on the source strength and duration, contaminant transport
processes (for example, removal by biodegradation), and exposure pathways and parameters (e.g.,
receptor location and use of contaminated water) to estimate human health risks over time.
The decision support can be in the form of guidance that provides a framework for performing the
analysis or software that has codified the expertise to allow more rapid analysis by many. The magnitude
and similarity between contaminated land management problems has led to development of several
computer software DSTs to address different aspects of the problem (site characterization, cost-benefit,
risks, sustainable development, etc.).
Regulatory guidance is the most widely used type of decision support. In several countries, adherence to
the guidance is required or strongly recommended. For software based DSTs, human health risk
assessment and cost-benefit are the most commonly used. Ecological risk assessment and multi-criteria
analysis are starting to see more use.
Stakeholder involvement is an important aspect of the decision process and helps to achieve a solution for
contaminated land management that is acceptable to all. Stakeholders may not always agree on an
approach for contaminated land management. In this case, the regulators are the mediators between the
different stakeholders. Effectively integrating the stakeholders into the decision process is a difficult task
requiring strong leadership and good communication skills.
123
-------
Decision Support Tools NATO/CCMS Pilot Project Phase I
The strengths, limitations, and applications of DST have been identified and discussed in this paper. The
major strengths identified were the ability to provide a consistent, reproducible process for decision
making and the ability to enhance communication between different stakeholder groups. The major
disadvantage in using DST was in gaining acceptability of the tool to all stakeholders. This can be a time
consuming process. A secondary disadvantage that was cited involved concerns that making the tools
easy to use could lead to their misuse. Careful review is required for all results that support a decision.
Decision support can be greatly improved through the use of data management tools that store the
information electronically and permit its use by all stakeholders. A concern was expressed that if each
DST had its own dataset this could lead to inconsistencies. Proper data management would remove this
problem and lead to improved quality control of data and would help improve consistency.
A number of unresolved issues pertaining to the use of DST were identified. Based on these findings
several areas for improvement were identified. Some of the more important areas requiring further
development include:
• Improved methods for valuation of criteria and determination of weights for MCA approaches.
This includes the need for improved methods and approaches for handling subjective (soft data).
Work needs to be done to develop a consistent agreed upon approach to using MCA.
• Improved transparency for the concepts behind decision support to all stakeholders. Greater
stakeholder involvement is needed to gain acceptance of DST.
• Improved transparency in the output from DST. Decision support tools often involve abstraction
from multiple sources of data and involve complex technical analysis.
• Improved methods for verification of the performance of DST. This is especially true in
computationally intensive areas that require extensive experience to use correctly and are often
based on data sets that permit multiple interpretations. These areas include flow and transport
calculations, geostatistical modeling and optimization of remedy performance.
• Improved methods for understanding the impacts of natural variability and uncertainty on the
decision process. Some DST address the role of uncertainty in making a decision, but this is an
emerging field that needs further development.
• Critical evaluation of the successes and failures in the use of DSTs. This evaluation would help to
focus future development work.
8. REFERENCES
Bardos, R.P.; Mariotti, C.; Marot, F.; and Sullivan, T., "Framework for Decision Support used in
Contaminated Land Management in Europe and North America, " in NATO Committee on Challenges to
Modern Society: NATO/CCMS Pilot Study Evaluation of Demonstrated and Emerging Technologies for
the Treatment and Clean Up of Contaminated Land and Groundwater. Phase III 2000 Special Session
Decision Support For Contaminated Land Management. Ibid.
Finnamore, J., "Modeling the Financial Risks of Remediation," in NATO Committee on Challenges to
Modern Society: NATO/CCMS Pilot Study Evaluation of Demonstrated and Emerging Technologies for
the Treatment and Clean Up of Contaminated Land and Groundwater. Phase III 2000 Special Session
Decision Support For Contaminated Land Management,, ibid.
Stewart, R. "Geospatial Decision Frameworks for Remedial Design and Secondary Sampling," in NATO
Committee on Challenges to Modern Society: NATO/CCMS Pilot Study Evaluation of Demonstrated and
Emerging Technologies for the Treatment and Clean Up of Contaminated Land and Groundwater. Phase
III 2000 Special Session Decision Support For Contaminated Land Management,, ibid.
124
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
COUNTRY REPRESENTATIVES
Directors
Stephen C. James (Co-Director)
National Risk Management Research Laboratory
U.S. Environmental Protection Agency
26 Martin Luther King Drive
Cincinnati, Ohio 45268
United States
tel: 513-569-7877
fax:513-569-7680
e-mail: james.steve@epa.gov
Walter W. Kovalick, Jr. (Co-Director)
Technology Innovation Office
U.S. Environmental Protection Agency
1300 Pennsylvania Ave, NW (5102G)
Washington, DC 20460
United States
tel: 703-603-9910
fax: 703-603-9135
e-mail: kovalick.walter@epa.gov
Co-Pilot Directors
Volker Franzius
Umweltbundesamt
Bismarckplatz 1
D-14193 Berlin
Germany
tel: 49/30-8903-2496
fax: 49/30-8903-2285 or -2103
e-mail: volker.franziusfi2kiba.de
H. Johan van Veen
TNE/MEP
P.O. Box 342
7800 AN Apeldoorn
The Netherlands
tel: 31/555-493922
fax: 31/555-493921
e-mail: h.j.vanvccn@mcp.tno.nl
Country Representatives
Anahit Aleksandryan
Ministry of Nature Protection
35, Moskovyan Strasse
375002 Yerevan
Armenia
tel: +37/42-538-838
fax:+3 7/42-151-938
e-mail: goga@arminco.com
Nora Meixner
Federal Ministry of Environment, Youth and
Family Affairs
Dept. HI/3
Stubenbastei 5
A-1010 Vienna
Austria
tel: 43/1-515-22-3449
fax: 43/1-513-1679-1008
e-mail:
Jacqueline Miller
Brussels University
Avenue Jeanne 44
1050 Brussels
Belgium
tel: 32/2-650-3183
fax: 32/2-650-3189
e-mail: jmillcr@ulb.ac.bc
Lisa Keller
Environmental Technology Advancement
Directorate
Environment Canda - EPS
12th Floor, Place Vincent Massey
Hull, Quebec K1A OH3
Canada
tel: 819/953-9370
fax: 819/953-0509
e-mail:
125
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Hana Kroova
Czech Ministry of the Environment
Vrsovicka 65
100 10 Prague 10
Czech Republic
tel: 420/2-67 12-1111
fax: 420/2-673 1-0305
Kim Dahlstrem
Danish Environmental Protection Agency
Strandgade 29
DK-1401 Copenhagen K
Denmark
tel: +45/3266-0388
fax: 45/3296-1656
e-mail: kdaigimst.dk
Ari Seppanen
Ministry of Environment
P.O. Box 399
00121 Helsinki
Finland
tel: +358/9-199-197-15
fax: +358/9-199-196-30
e-mail: ari .seppanen@vyh.fi
Andreas Bieber
Federal Ministry for the Environment
Ahrstrasse 20
53 175 Bonn
Germany
tel: 49/228-305-305-3431
fax: 49/228-305-305-2396
e-mail:
Anthimos Xenidis
National Technical University Athens
52 Themidos Street
15 124 Athens
Greece
tel: 30/1-772-2043
fax: 30/1-772-2168
Pal Varga
National Authority for the Environment
F6 u.44
H-10 11 Budapest
Hungary
tel: 36/1-346-8310
fax: 36/1-3 15-08 12
e-mail: yMgifiiffimili!!^
Matthew Crowe
Environmental Management and Planning
Division
Environmental Protection Agency
P.O. Box 3000
Johnstown Castle Estate
County Wexford
Ireland
tel: +353 53 60600
fax: +353 53 60699
e-mail: m.crowc@cpa.ic
Francesca Quercia
ANPA - Agenzia Nazionale per la Protezione
dellAmbiente
ViaV. Brancati48
I -00 144 Rome
Italy
tel. 39/6-5007-2510
fax 3 9/6-5 007-25 31
e-mail: guercia@anpajt
Masaaki Hosomi
Tokyo University of Agriculture and
Technology
2-24-16 Nakamachi
Tokyo 184-8588
Japan
tel: +81-42-388-7070
fax:+81-42-381-4201
e-mail:
Bj0rn Bj0rnstad
Norwegian Pollution Control Authority
P.O. Box 8100 Dep
N-0032 Oslo
Norway
tel: 47/22-257-3664
fax: 47/22-267-6706
e-mail: bjom.bjomstad@sft.tclcmax.no
Marco Estrela
Institute de Soldadura e Qualidade
Centre de Tecnologias Ambientais
Tagus Park
EC Oeiras - 2781-951 Oeiras
Portugal
tel: +35 1/21-422 90 05
fax: +351/21-422 81 04
e-mail: maestrela@i sq . pt
126
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
loan Gherhes
EPA Baia Mare
I/A Iza Street
4800 Baia Mare
Romania
tel: 40/4-62-276-304
fax: 40/4-62-275-222
e-mail: cpa@multinct.ro
Branko Druzina
Institute of Public Health
Trubarjeva 2-Post Box 260
6100 Ljubljana
Slovenia
tel: 386/61-313-276
fax: 386/61-323-955
e-mail: branko.druzina@gov.si
Vitor A.P.M. dos Santos
Spanish National Research Council
Professor Aubareoal
18008 Granada
Spain
tel: 34/958-121-011
fax: 34/958-129-600
e-mail: vasantos@ccz.csis.cs
Ingrid Hasselsten
Swedish Environmental Protection Agency
Blekholmsterrassen 36
S-106 48 Stockholm
Sweden
tel: 46/8-698-1179
fax: 46/8-698-1222
e-mail: mh@cnviron.sc
Bernard Hammer
BUWAL
3003 Bern
Switzerland
tel: 41/3 1-322-9307
fax: 41/3 1-382-1456
e-mail:
Kahraman Unlii
Depratment of Environmental Engineering
Middle East Technical University
Inonii Bulvari
06531 Ankara
Turkey
tel: 90-312-210-1000
fax:90-312-210-1260
e-mail: kj_nlu@mgtu_.c_du-tr
Ian D. Martin
Environment Agency
Olton Court
10 Warwick Road
Olton, West Midlands
United Kingdom
tel: 44/121-71 1-2324
fax: 44/121-71 1-5830
e-mail: ianmartin@environment-agency.gov.uk
127
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
ATTENDEES LIST
Anahit Aleksandryan (c.r.)
Ministry of Nature Protection
35 Moskovyan str.
375002 Yerevan
Republic of Armenia
tel: 37/42-538-838
fax: 37/42151-938
e-mail: goga^anmncacom
P.A. (Arne) Alphenaar
TAUW
P.O. Box 133
7400 AK Deventer
The Netherlands
tel: 31/570699911
fax: 31/570 699 666
e-mail: pah@tauw.nl
Paul Bardos
R? Environmental Technologies Ltd.
P.O. Box 58
Ware-Hertfordshire SG12 9UJ
United Kingdom
tel: 44/1920-484-571
fax: 44/1920-485-607
e-mail: p-bardos@r3-bardos.dcmon.co.uk
Paul M. Beam (c.r.)
U.S. Department of Energy
19901 Germantown Rd.
Germantown, MD 20874-1290
United States
tel: 301-903-8133
fax: 301-903-3877
e-mail: paul.bcaiii@cm.doc.gov
Jorg Becht
Hessisches Ministerium fur Umwelt,
Landwirtschaft und Forsten
Mainzer Str. 98-102
65189 Wiesbaden
Germany
tel: 49/611-815-1380
fax: 48/611-815-1947
e-mail: abteilung.3@mue.hessen.de
Eberhard Bellinger
WCI Umwelttechnik GmbH
Heinrich-Hertz-StraBe 3
63303 Dreieich
Germany
tel: 49-61 03-9 38 90
fax:49-6103-938999
e-mail:
Andreas Bieber (c.r.)
Federal Ministry for the Environment
Ahrstrasse 20
53 175 Bonn
Germany
tel: 49/228-305-305-3431
fax: 49/228-305-305-2396
e-mail: bieber.andreas@bmu.de
Bj0rn Bj0rnstad (c.r.)
Norwegian Pollution Control Authority
P.O. Box 8100 Dep
N-0032 Oslo
Norway
tel: 47/22-257-3664
fax: 47/22-267-6706
e-mail: bjom.bjornstad@telemax.no
Volker Bohmer
Hessische Industriemull GmbH
Bereich Altlastensanierung
Kreuzberger Ring 58
65205 Wiesbaden
Germany
tel: 49/61 1-7149-700
fax: 49/61 1-7149-322
e-mail: volkcr@bochmcr@him.dc
Michael Bosley
International Engineering Center
US Army Corps of Engineers-Europe
Konrad Adennauar Ring 39 Box 20
65 187 Wiesbaden
Germany
tel: 49-61 1-816-2692
e-mail:
MICHAEL.J.BOSLEY@nau02.usace.armv.mil
128
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Harald Burmeier
Fachhochschule North-East Lower Saxony
Department of Civil Engineering
Herbert Meyer Strasse 7
29556 Suderburg
Germany
tel: 49/5103-2000
fax:49/5103-7863
e-mail: Lburmeier@t-oiiline.de
Erol Er^ag
Istanbul University
Dept. of Chemistry
Avcilar Campus, Avcilar 34850
Istanbul
Turkey
tel: 90/212-5911-998
fax: 90/212-5911-997
e-mail: ismailbfiadstanbul.cdu.tr
Laurence Davidson
c/o EarthFx Inc.
2635 Ulster Crescent
K1V 8J5 Ottawa, Ontario
Canada
tel: 613.260.2020
fax: 613.260.252
e-mail: |d@cartlj.fx._com
Branko Druzina (c.r.)
Institute of Public Health
Trubarjeva 2-Post Box 260
6100 Ljubljana
Slovenia
tel: 386/1-313-276
fax: 386/1-323-955
e-mail: br_aiikgAuzina@jvzrrsji
Vitor A.P.M. Dos Santos (c.r.)
Spanish National Research Council
Professor Aubareoal
18008 Granada
Spain
tel: 34/958-121-011
fax: 34/958-129-600
David Edwards
Leader ExSite
VHE Holdings pic.
CEO's Office
Shafton, Barnsley, S72 8SP
United Kingdom
tel: 44/1977-683300
fax: 44/870-1314537
e-mail: exSite@ibtinternet.com
James Finnamore
WSP Environmental
Buchanan House
24-30 Holborn
London EC IN 2HS
United Kingdom
tel: 44/20-73 14-5 000
fax: 44/20-73 14-5 005
e-mail: jim.finnamore@wspgroup.com
Volker Franzius
Umweltbundesamt
Bismarckplatz 1
D- 14 193 Berlin
Germany
tel: 49/30-8903-2496
fax: 49/30-8903-2285 or -2103
e-mail: volker.franzius@uba.de
loan Gherhes (c.r.)
EPA Baia Mare
I/A Iza Street
4800 Baia Mare
Romania
tel: 40/4-62-276-304
fax: 40/4-62-275-222
e-mail: IGherf|ej3@aj)^
Detlef Grimski
Umweltbundesamt
Bismarckplatz 1
14 193 Berlin
Germany
tel: 49/30-8903-2266
fax: 49/30-8903-2103
e-mail: dctlcf.grimski(a;uba.dc
129
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Karl Grundler
Hessisches Ministerium fur Umwelt,
Landwirtschaft und Forsten
Mainzer Str. 98-102
65 189 Wiesbaden
Germany
tel: 49/61 1-815-1373
fax: 49/61 1-81-1947
e-mail: abtrilun.3muc.hcsscn.dc
Bernhard Hammer (c.r.)
Federal Office of the Environment,
Forests & Landscape (BUWAL)
Federal Department of the Interior
Buwal Laupenstrausse 20
3003 Bern
Switzerland
tel: +41/3 1-322-6961
fax: +41/3 1-382-1546
e-mail:
Gregory Harvey
Aeronautical Systems Center
Environmental Safety and Health Division
1801 10th St.
Bldg. 8, Suite 200 - Area B
WPAFB, OH 45433
United States
tel: 937-255-7716 (ext. 302)
fax: 937-255-4155
e-mail: g regory . harvey@wpafb . af . mil
R.A.A. (Rolf) Hetterschijt
P.O. Box 6012
2600 JA Delft
The Netherlands
tel: +31 152696257
fax: +31 152564800
e-mail rjietterjchjjt^nitgjnojil
Howard Hornfeld
Programme Coordinator for the Chemical
Industry
United Nations Economic Commission for
Europe
Palais des Nations 429-3
CH-12 11 Geneva 10
Switzerland
tel.: 41 22 917 3254
fax.: 41 22 917 0178
e-mail: chem@un.ece .org
Masaaki Hosomi (c.r.)
Tokyo University of Agriculture and
Technology
2-24-16 Nakamachi, Koganei
Tokyo 184
Japan
tel: 81/3-423-887-070
fax: 81/3-423-814-201
e-mail: hosomi@cc.tuat.ac.jp
Stephen C. James (Co-Director)
U.S. Environmental Protection Agency
26 Martin Luther King Dr.
Cincinnati, OH 45268
United States
tel: 513-569-7877
fax:513-569-7680
e-mail: jamgs._stcvg@c|)a..gov
Harald Kasamas
EU Concerted Action CLARINET
Breitenfurterstr. 97
A- 11 20 Vienna
Austria
tel: 43/1-804 93 192
fax: 43/1-804 93 194
e-mail:
Lisa Keller
Environmental Technology Advancement
Directorate
Environment Canada - EPS
12th floor, Place Vincent Massey
Hull, Quebec K1A OH3
Canada
tel: 819-953-9370
fax: 819-953-0509
e-mail: Lisa.Kcllcr@cc.gc.ca
Peter Kontny
Probiotec GmbH
SchillingstraBe 33
52355 Duren-Giirzenich
Germany
tel: 49/2421-6909-65
fax: 49/2421-6909-61
e-mail: info@probiotgc_.acj^urcgia,dc
130
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Hans-Peter Koschitzky
Technical director, VEGAS, Research Facility
Chair for Hydraulics and Groundwater
University of Stuttgart
Pfaffenwaldring 61
D - 70550 Stuttgart
Germany
tel: 49/711-686-4717
fax: 49/711-685-7020
Kazuhide Kuzawa
Japan Environment Agency
1-2-2 Kasumigaseki
100-8975 Chiyoda-ku, Tokyo
Japan
tel: 81/3-5521-8322
fax: 81/3-3593-1438
e-mail: KAZUHIDE_KUZAWA@canct.go,ip
Walter W. Kovalick, Jr. (Co-Director)
Technology Innovation Office
U.S. Environmental Protection Agency
1200 Pennsylvania Ave. (5102G)
Washington, DC 20460
United States
tel: 703-603-9910
fax: 703-603-9135
e-mail: kovalick.walter@epa.gov
Hana Kroova (c.r.)
Czech Ministry of the Environment
Vrsovicka 65
100 10 Prague 10
Czech Republic
tel: 420/2-6712-1111
fax: 420/2-6731-0305
Andrea Lodolo
ICS-UNIDO
Pure and Applied Chemistry
Area Science Park Building L2
Padriciano, 99
34012 Trieste
Italy
tel.: 39-040-9228114
fax:39-040-9228115
e-mail: cmanucla.corazzi(rt}ics.tricstc.it
Ian D. Martin (c.r.)
The Environment Agency
Olton Court, 10 Warwick Road
Olton, West Midlands
United Kingdom
tel: 44/121-71 1-2324
fax: 44/121-71 1-5830
e-mail: ia
Nora Meixner (c.r.)
Federal Ministry of Environment, Youth and
Family Affairs
Dept. HI/3
Stubenbastei 5
A- 10 10 Vienna
Austria
tel: 43/1-5 15-22-3449
fax: 43/1-5 13-1679-1008
e-mail: Nora. Meixner@bmii.gv.at
Jochen Michels
DECHEMA
Theodor-Heuss-Allee 25
60486 Frankfurt am Main
Germany
tel: 49-69-75 64-2 35
fax: 49-69-75 64-2 35
e-mail: michels@dechema.de
Jacqueline Miller (c.r.)
Brussels University
Avenue Jeanne 44
1050 Brussels
Belgium
tel: 32/2-650-3183
fax: 32/2-650-3 189
e-mail: jtriillerjSiulb_.ac.be
Walter Mondt
Ecorem n.v.
Zwartzustersvest 22
B-2800 Mechelen
Belgium
tel: 32/15-21 17 35
fax: 32/15-21 65 98
e-mail: ccorcm@glo.bc
131
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Christoph Munz
Env. Eng., Dept. Head Chemical Risk
Assessment
BMG Engineering AG
Ifangstrasse 11
CH-8952 Zurich
Switzerland
tel. 41/1-732-92 77
fax. 41/1-732-9221
e-mail: christoph.miinz^bmgcng.ch
Joop Okx
Tauw Milieu bv
PO Box 133
7400 AC Deventer
The Netherlands
tel: 3 1570 699911
fax: 3 1570 699666
e-mail: jpo@tauw.nl
Johannes Pastor
Bundesministerium fur Umwelt
Naturschutz und Reaktorsicherheit
Postfach 12 06 29
53048 Bonn
Germany
tel: 49-2 28-3 05-34-30
fax: 49-2 28-3 05-23 96
Simon Pollard
The Environment Agency
Steel House
1 1 Tothill Street
London SW1H 9NF
United Kingdom
tel: 44 20 7664 6832
fax: 44 20 7664 6836
e-mail: simon.pollardfficnvironmcnt-
Francesca Quercia (c.r.)
ANPA - Agenzia Nazionale per la Protezione
dellAmbiente
ViaV. Brancati48
I -00 144 Rome
Italy
tel. 39/6-5007-2510
fax 3 9/6-5 007-25 31
e-mail
Charles Reeter
Naval Facilities Engineering Service Center
U.S. Navy
1 100 23rd Avenue, Code 411
Port Hueneme, CA 93043
United States
tel: 805-982-4991
e-mail: reetoiciiSnfo
Mathias Schluep
BMG Engineering AG
Ifangstrasse 11
8952 Schlieren
Switzerland
tel: 41/1-730-6622
fax: 41/1-730-6622
Ari Seppanen (c.r.)
Ministry of Environment
P.O. Box 399
00121 Helsinki
Finland
tel: 358/9-199-197-15
fax: 358/9-199-196-30
e-mail: An_.Sej3pangn@vyh_.fi
Rainer Siebert
Bundesministerium fur Umwelt,
Naturschutz und Reaktorsicherheit
Postfach 12 06 29
53048 Bonn
Germany
tel: 49-228-305-3434
fax: 49-228-305-2396
e-mail: sicbcrt.raincr@bnm.dc
Roberrt Siegrist
Colorado School of Mines
Environ. Science and Eng. Division
1500 Illinois Ave.
Golden, CO 80401-1887
United States
tel: 303-273-3490
fax: 303-273-3413
e-mail: rsiegris@mines.edu
132
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Sjef Staps
TNO Institute of Environment, Energy, and
Process Innovation
Department of Environmental Biotechnology
P.O. Box 342
7300AHApeldoorn
The Netherlands
tel: 31555493351
fax: 31 555493523
e-mail: sjjte^)s@mcj3jncyil
Kai Steffens
PROBIOTEC GmbH
Schillingsstra e 333
D 52355 Duren-Giirzenich
Germany
tel: 49/2421-69090
fax: 49/2421-690961
e-mail: steffens@probiotec.de
Rita Hermanns Stengele
Professor for Environmental Geotechnics
Institute of Geotechnical Engineering
ETH Honggerberg/HIL
CH-8093 Zurich
Switzerland
tel. 41/1-633-2524 or 633-2525 (secreatriat)
fax. 41/1-633-109
e-mail: hcrmanns@igt.baum.cthz.ch
Robert Stewart
University of Tennessee
1060 Commerce Park
Oak Ridge, TN 37830
United States
tel: 865-241-5741
fax: 865-574-0004
e-mail: u47@onil.gov
Terry Sullivan
Environmental Sciences Department
34 North Railroad Street, Building 830
Upton, NY 11973-5000
United States
tel: 631-344-3840
fax:631-344-4486
e-mail: tsul|ivan@bnl.goy
Jan Svoma
Aquatest a.s.
Geologicka 4
152 00 Prague 5
Czech Republic
tel: 420/2-581-83-80
fax: 420/2-5 8 1-77-5 8
e-mail:
Bert-Axel Szelinski
Bundesministerium fur Umwelt
Naturschutz und Reaktorsicherheit
Alexanderplatz 6
11 05 5 Berlin
Germany
tel: 49-30-2 85 50-42 70
fax: 49-30-2 85 50-43 75
e-mail: szgjinsM_.axgl@bmu_._dc
Safieh Taghavi
Vlaamse Instelling voor Technologisch
Onderzoek (Vito)
Environmental Technology Expertise Center
Boeretang 200
B.2400 Mol
Belgium
tel: 32/14-335162
fax: 32/14-580523
e-mail: safiyh .taghavi@vito . be
Georg Teutsch
University of Tubingen
Sigwartstrasse 10
72076 Tubingen
Germany
tel: 49/707-1297-6468
fax: 49/707-150-59
e-mail: gcorg.toutscli@uiii-tucbingcn.dc
Steven Thornton
University of Sheffield
Mappin Street
Sheffield
United Kingdom
tel: 44/1 14-222-5 700
fax: 44/1 14-222-5 700
e-mail: S.F.Thoniton@shcfficld.ac.uk
133
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Kahraman Unlii (c.r.)
Depratment of Environmental Engineering
Middle East Technical University
Inonii Bulvari
06531 Ankara
Turkey
tel: 90/312-210-1000
fax: 90/3 12-2 10- 1260
e-mail: kunlu@mctu.cdu .tr
H. Johan van Veen (c.r.)
TNO/MEP
P.O. Box 342
7800 AN Apeldoorn
The Netherlands
tel: 31/555-493922
fax: 3 1/555-493921
e-mail: HJ.yjmVcgii@mg|)_.tng.nl
Joop Vegter
The Technical Committee on Soil Protection
(TCB)
Postbus 30947
2500 GX The Hague
The Netherlands
tel: 31/70-339-30-34
fax 3 1/70-339-13-42
e-mail: tcb@euronet.nl
John Vijgen
Consultant
Elmevej 14
DK-2840 Holte
Denmark
tel: 45/45 41 03 21
+fax:45/45410904
e-mail:
^^
Stephan Volkwein
C.A.U.
DaimlerstraBe 23
63303 Dreieich
Germany
tel: 49-61 03-9 83-25
fax: 49-61 03-9 83-10
e-mail: c .a.u . (ait-online .dc
Christian Weingran
Hessische Industriemull GmbH
Mullerwegstannen 46
35260 Stadtallendorf
Germany
tel: 49/6428-9235-11
fax: 49/6428-9235-35
e-mail: as
Holger Weiss
UFZ-Umweltforschungszentrum
Leipzig-Halle GmbH
Postfach 2
Germany
tel: 49-3 41-2 35-20 58
fax:49-341-235-2126
e-mail: weiss@pro.ufz.de
Paul Wersin
Geochemist/Project Manager Safety Analysis
NAGRA (National Cooperative for the Disposal
of Radioactive Waste)
Hardstrasse 73
CH-5430 Wettingen
Switzerland
tel. 41/56-437-12 80
fax. 41/56-437-1317
e-mail: .we rsi n @nagra. ch
Dieter Weth
Mull & Partner Ingenieurgesellschaft
Osterlede 5
30827 Garbsen
Germany
tel: 49-5 13-1 46 94-0
fax:49-513-14694-90
e-mail: wethiSniullund^
Uwe Wittmann
Umweltbundesamt
SeecktstraBe 8-10
13581 Berlin
Germany
tel: 49-30-89 03-
fax: 49-30-89 03-38 33
e-mail: uwc .wittmaiin@uba.dc
134
-------
Decision Support Tools
NATO/CCMS Pilot Project Phase I
Anthimos Xenidis
National Technical University Athens
52 Themidos Street
15124 Athens
Greece
tel: 30/1-772-2043
fax: 30/1-772-2168
Mehmet AH Yukselen
Marmara University
Environmental Engineering Department
Goztepe 81040 Istanbul
Turkey
tel: 90/216-348-1369
fax: 90/216-348 -0293
135
------- |