Environmental Risk Management for Radiological Accidents:Integrating Risk Assessment and Decision Analysis forRemediation at Different Spatial ScalesBoris Yatsalo,y Terrence Sullivan,z Vladimir Didenko,y and Igor Linkov*§yObninsk Institute for Nuclear Power Engineering - branch of National Research Nuclear University, Moscow Engineering PhysicsInstitute (OINPE NRNU MEPhI), Obninsk, RussiazBrookhaven National Laboratory (BNL), Upton, New York, USA§Environmental Laboratory, US Army Engineer Research and Development Center, 696 Virginia Road, Concord,Massachusetts 01742, USA

(Submitted 14 April 2011; Returned for Revision 3 May 2011; Accepted 16 May 2011)

ABSTRACTThe consequences of the Tohuku earthquake and subsequent tsunami inMarch 2011 caused a loss of power at the Fukushima

Daiichi nuclear power plant, in Japan, and led to the release of radioactive materials into the environment. Although the full

extent of the contamination is not currently known, the highly complex nature of the environmental contamination

(radionuclides in water, soil, and agricultural produce) typical of nuclear accidents requires a detailed geospatial analysis of

information with the ability to extrapolate across different scales with applications to risk assessment models and decision

making support. This article briefly summarizes the approach used to inform risk-based land management and remediation

decision making after the Chernobyl, Soviet Ukraine, accident in 1986. Integr Environ Assess Manag 2011;7:393–395.

� 2011 SETAC

Keywords: Remedial decisions Spatially integrated risk assessment Decision analysis

INTRODUCTIONOn 11 March 2011, the Tohuku earthquake and the

subsequent tsunami caused a loss of power at Japan’sFukushima Daiichi nuclear power plant. This led to a seriesof equipment failures, a loss of fuel cladding integrity, and therelease of radioactivity into the environment. Initial estimatesbased on aerial surveys around Fukushima suggest thatagricultural monitoring and possible intervention will berequired for several hundred square kilometers surroundingthe site (USDOE 2011). Location and treatment methodsapplied to agricultural land and produce will require carefulevaluation. These decisions will need to balance benefits suchas reduced radiation exposure and the sociopolitical benefits ofremediation against costs associated with the remediation,along with the risks associated with remedial worker exposure.

As a historical precedent to the nuclear concerns surround-ing Fukushima, the lessons learned from the Chernobylaccident of 1986 can provide a guide as to what to expect interms of remediation needs. One of the main lessons learnedfrom Chernobyl is that integrated risk management and theoptimization of resources are both required to developremedial and abatement policies (Karaoglou et al. 1996).High spatial variability in environmental contamination (e.g.,concentration of radionuclides in water, soil, and agricultural

produce) is typical of nuclear accidents, and requires adetailed geospatial analysis of information with the abilityto extrapolate across different scales, including the correct useof spatially distributed data for model assessments anddecision making support on local and regional levels.

Significant uncertainty associated with our knowledge onthe environmental fate and transport of radionuclides andtheir associated risks requires an integration of geospatial andtemporal scales with the ability to support a wide range ofdecisions, which in turn requires computerized decisionsupport systems. After the Chernobyl accident, severalmodels and computer systems were developed to estimatethe large-scale short-term and long-term consequences of anuclear accident (Karaoglou et al. 1996; Baverstam et al.1997), but not all were generalized enough to be applicableacross the various problem-specific and site-specific condi-tions, and only the RODOS (Real-time On-line DecisionSupport) (Erhard and Sershakov 1996) system has integrateddecision support capabilities.

DECERNS (Decision Evaluation in Complex RiskNetwork Systems) is one example of a general frameworkthat addresses a wide class of problems on remediation ofcontaminated sites, risk-based land management and land-useplanning, protection of the population and environment, andother problems connected with spatial information, alter-native choices, and decision support (Sullivan et al. 2009).Integration of methods for Multi-Criteria Decision Analysis(MCDA functionality) and tools for spatial data processing,analysis, and presentation (geographical information systemfunctionality) are necessary to provide an effective approach

Integrated Environmental Assessment and Management — Volume 7, Number 3—pp. 393–395� 2011 SETAC 393

EDITOR’S NOTE

This is 1 of 17 invited commentaries in the series ‘‘Challenges Posed by Radiation and Radionuclides in the Environment.’’These peer-reviewed commentaries were prepared to address some of the environmental issues raised by the March 2011nuclear power plant accident in Japan.

* To whom correspondence may be addressed: Igor.Linkov@usace.army.mil

Published online 23 May 2011 in Wiley Online Library

(wileyonlinelibrary.com).

DOI: 10.1002/ieam.229

Invite

dCommenta

ry

to support risk management and evaluation of alternatives(Linkov and Moberg 2011).

As an example on how to conduct this type of analysis forradiologically contaminated sites, DECERNS was used toexamine 6 remedial alternatives for treatment of agriculturallands and products in the Novozybkov district, Russia,which had substantial Cs-137 contamination as a result ofthe Chernobyl accident (Sullivan et al. 2009; Yatsalo et al.2010).

The following 6 alternatives were considered forremediation:

A0¼No countermeasuresA1¼Banning local milk consumption in areas with milkcontamination above 100Bq/LA2¼Comprehensive treatment of pastures and hayfieldscontaminated above 5 Ci/km2

A3¼Comprehensive treatment for pastures and hayfieldswhere model estimates of milk contamination exceed100Bq/LA4¼Adding ferrocyne to feed for milk cows during theyear where model estimates of milk contamination exceed100Bq/LA5¼ Implementing countermeasures A3 and A4 combined

The following 5 criteria were suggested by a group ofexperts to compare the different alternatives:

C1¼CostC2¼Avertable collective doseC3¼ Fraction of the local population in each settlementwith mean internal dose above 1 mSv/yC4¼ Fraction of milk produced with contamination above100Bq/LC5¼ Socioeconomic indicators

The following analyses were performed to quantifycountermeasure performance by each criterion before and

after countermeasure implementation with the use ofcorresponding models:

1) Estimation of agricultural produce and wild mushroomand berry contamination over time for each location.

2) Assessment of internal and external doses to the generalpopulation and critical (i.e., most exposed) populationgroups for all the settlements over time.

3) Risk-based ranking of settlements, farms and agriculturalland.

4) Analysis of fields and settlements where countermeasuresshould and could be implemented in accordance with theexisting standards, requirements and restrictions.

5) Assessments over space and time of the reduction indoses and produce contamination from the use of thecountermeasure.

The results of the MCDA comparison of alternativesalong with relevant weight sensitivity analyses are presentedin Figure 1. The analysis indicates that Alternative A5, acombination of A3 and A4, is the preferred choice underthe selected conditions. As compared to A4 (adding ferrocyneto feed for cows to reduce contamination to milk), it is clearthat A5 (adding ferrocyne and comprehensive soil treatment)provided a small dose reduction to the population, but thatthe cost was much greater. Examining the line weights inFigure 1, it is seen that A4 becomes the preferred alternative ifthe weighting value for cost increases from 0.20 to approx-imately 0.35. For this reason, alternatives A4 and A5 wereselected and recommended for subsequent analysis anddecision making. Experts recommend that the decisionmakers should take into account the availability of financialresources for the specific year and territory to choose betweenthese 2 alternatives. The use of the MCDA method provideda defensible framework to understand the decision process interms of the underlying constraints.

In summary, the management of environmental risksresulting from the Fukushima nuclear accident will requirean enormous effort to address the population’s concerns in a

Figure 1. Evaluation of remedial alternatives at radiologically contaminated sites: multi-attribute value theory (MAVT)-based alternative rankings and weight

sensitivity analysis.

394 Integr Environ Assess Manag 7, 2011—B Yatsalo et al.

timely manner. The decision process must balance the desirefor remediation while keeping in mind the constraints of time,money, and manpower. A promising approach to accomplishthis is through the use of decision support systems capable ofintegrating spatial information (e.g., land use, populationdensity, infrastructure such as the location of roads, railroads,and so forth) with multiple decision criteria (e.g., averted dose,maximum dose, cost, social benefits). This approach provides aprocess to include stakeholder input such as the relativeimportance of various decision criteria to the final decision,which should provide a more transparent decision process thatwill help in the decision process. This type of integratedprocess is recommended in the United States (FR 2008) foraddressing widespread release of radioactive contamination.

Disclaimer—The peer-review process for this commentarywas managed by the Editors without involvement of I.Linkov.

REFERENCESBaverstam U, Fraser G, Kelly GN, editors. 1997. Decision making support for off-site

emergency management. Radiat Prot Dosim 73:0iii–xiii.

Erhard J, Shershakov V, editors. 1996. Real-time On-line Decision Support System

(RODOS) for off-site emergency management following a nuclear accident.

Final Report, Joint Study Project 1, European Commission, DG XII, EUR 16533

EN. Brussels (BE): European Commission.

[FR] Federal Register. 2008. Planning guidance for protection and recovery

following radiological dispersal device (RDD) and improvised nuclear device

(IND) incidents. Fed Reg 73:45029–45048.

Karaoglou A, Desmet G, Kelly GN, Menzel HG, editors. 1996. The

radiological consequences of the Chernobyl accident. Proceedings of the

first International Conference; Minsk, Belarus; 18–22 Mar 1996. European

Commission. ISBN 92-827-5248-8. ECSC-EC-EAEC, Brussels, Luxembourg,

1996. 1192 p.

Linkov I, Moberg E. 2011. (Forthcoming) Multi-criteria decision analysis:

environmental applications and case studies. Boca Raton (FL): CRC. 232 p.

Sullivan T, Yatsalo B, Grebenkov A, Linkov I. 2009. Decision Evaluation for Complex

Risk Network Systems (DECERNS) software tool. In: Marcomini A, Suter

GW, Critto A, editors. Chapter 12, Decision support systems for risk-based

management of contaminated sites. New York (NY): Springer Science. p. 257–

274.

[USDOE] US Department of Energy. 2011. Radiological assessment of effects from

Fukushima Daiichi nuclear power plant, April 22, 2011. [cited 2011 April 22].

Available from: http://blog.energy.gov/content/situation-japan/

Yatsalo B, Didenko V, Tkachuk A, Gritsyuk S, Mirzeabasov O, Babutski A, Pichugina I,

Sullivan T, Linkov I. 2010. Multi-criteria spatial decision support system

DECERNS: Application to land use planning. Int J Inform Syst Soc Change

1:11–30.

Environmental Risk Management for Radiological Accidents— Integr Environ Assess Manag 7, 2011 395