P. 1
da7

da7

|Views: 1|Likes:
Published by mihai37
mit
mit

More info:

Published by: mihai37 on Jul 03, 2013
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

07/03/2013

pdf

text

original

Engineering Risk Benefit Analysis

1.155, 2.943, 3.577, 6.938, 10.816, 13.621, 16.862, 22.82, ESD.72

DA 7. Decision Analysis in Practice
George E. Apostolakis Massachusetts Institute of Technology

Spring 2007
DA 7 Decision Analysis in Practice 1

The Problem
• Recommend a technology to remediate a contaminated site. • The recommendation should reflect the views of a number of stakeholders. • Multi-Attribute Utility Analysis (MAUA) and deliberation are the bases for the recommendation.
DA 7 Decision Analysis in Practice 2

Involving the Stakeholders •
Risk assessment can and should be used to involve stakeholders and provide a mechanism for the consideration of their cultural, socioeconomic, historical, and religious values, in addition to the risks to human health and the environment associated with the contamination of DOE facilities and their remediation.
National Research Council, Building Consensus, 1994

DA 7 Decision Analysis in Practice

3

replicable methods. National Research Council. 1996. and law . evaluated under the agreed protocols of an expert community .to arrive at answers to factual questions. logic. Understanding Risk.The Analytic-Deliberative Process • Analysis uses rigorous. social. • Deliberation is any formal or informal process for communication and collective consideration of issues. as well as mathematics. DA 7 Decision Analysis in Practice 4 . or decision sciences.such as those of disciplines in the natural.

Analytic-Deliberative Decision Making S te p 1 F o r m u la te d e cisio n a lter n a tiv e s S te p 2 Is ea c h d ec isio n a lte r n a tiv e a cc ep ta b le? No R e m o v e fr o m fu rth er c o n sid er a tio n Y es S te p 3 R a n k a n d a n a ly z e d e cisio n a lte r n a tiv es fo r ea c h sta k e h o ld e r (P R A . M A U T . o th er ) S tep 4 D elib e ra te a n d c h o o se th e b est a lte rn a tiv e DA 7 Decision Analysis in Practice 5 .

ebris 4 miles to nearest drinking . lab trash. metal acids.9 Acres Disposal 1962 to 1981 Solvents.The Site • • • • • • • 1.water well 3 miles to nearest spring 480 feet to water table Network of vapor extraction wells to reduce TCE vapor plume • Landfill-wide excavation to top 15 ft to remove shallow primary sources of potential contamination DA 7 Decision Analysis in Practice 6 . PCBs. misc.

DA 7 Decision Analysis in Practice 7 . George E. and Pickett.” Risk Analysis 18 (1998): 621-634. Please see Table 1 in Apostolakis. Susan E. “Deliberation: Integrating Analytical Results into Environmental Decisions Involving Multiple Stakeholders.The Stakeholders Image removed due to copyright restrictions..

. Susan E. George E. DA 7 Decision Analysis in Practice 8 .” Risk Analysis 18 (1998): 621-634.Decision Options Image removed due to copyright restrictions. Please see Table 2 in Apostolakis. and Pickett. “Deliberation: Integrating Analytical Results into Environmental Decisions Involving Multiple Stakeholders.

Fundamental and Means Objectives • A distinction is made between “those things that participants fundamentally care about. 1994. 40: 1035-1048.” Management Science. DA 7 Decision Analysis in Practice 9 . such as environmental quality. “Creating policy alternatives using stakeholder values. and those that matter only through their effect on these fundamental concerns. such as waste disposal…”* *Gregory & Keeney.

• Examples: ¾ Core damage in nuclear reactors ¾ Water contamination • The distinction is important.Means Objectives • Means objectives are not important in and by themselves. They help to achieve the fundamental objectives. We should consider only fundamental objectives (things we really care about). DA 7 Decision Analysis in Practice 10 .

. 1996 • Sometimes this hierarchy is called a value tree. California: Duxbury Press. R. Clemen. 2nd Edition.Structuring the Fundamental Objectives • The set of fundamental objectives should be: ¾Complete ¾As small as possible ¾Not redundant • Fundamental objectives can be organized into a hierarchy in which the lower levels explain what is meant by the higher levels.T. Belmont. Making Hard Decisions. DA 7 Decision Analysis in Practice 11 .

Please see Fig. and selection of environmental remediation alternatives.. J. E.” Journal of Hazardous Materials 71 (2000): 35-57.The Value Tree: Impact Categories Image removed due to copyright restrictions. ranking. “Application of risk assessment and decision analysis to the evaluation. DA 7 Decision Analysis in Practice 12 . et al. 5 in Bonano.

ranking. Please see Fig. J. “Application of risk assessment and decision analysis to the evaluation. et al. and selection of environmental remediation alternatives. DA 7 Decision Analysis in Practice 13 ..” Journal of Hazardous Materials 71 (2000): 35-57.The Value Tree: Objectives Image removed due to copyright restrictions. 5 in Bonano. E.

“Application of risk assessment and decision analysis to the evaluation. 5 in Bonano. Please see Fig.The Value Tree: Performance Measures (Attributes) Image removed due to copyright restrictions. ranking. et al. and selection of environmental remediation alternatives.” Journal of Hazardous Materials 71 (2000): 35-57. DA 7 Decision Analysis in Practice 14 .. J. E.

Stakeholder Changes • Most agreed with the tree on slide 14.” DA 7 Decision Analysis in Practice 15 . • Some stakeholders placed long-term public risks under the category “environment.

. DA 7 Decision Analysis in Practice 16 . and Pickett. The decision is not made in general but for the specific problem.Performance Measure Units and Ranges Image removed due to copyright restrictions. Please see Table 3 in Apostolakis. George E. “Deliberation: Integrating Analytical Results into Environmental Decisions Involving Multiple Stakeholders. Susan E.” Risk Analysis 18 (1998): 621-634.

so that the expected utility (Performance Index) of the jth decision option is PI j = ∑ wi uij PM N i =1 where ∑ wi = 1 1 DA 7 Decision Analysis in Practice 17 NPM .Additive Independence (1) • The Performance Measures are assumed to be additive independent.

• “Even if used only as an approximation.” (Clemen) • Deliberation will follow the analysis. the additive utility function takes us a long way toward understanding our preferences and resolving a difficult situation. DA 7 Decision Analysis in Practice 18 . additive independence represents reasonably well people’s utilities. • Interaction among the attributes is not allowed. • For cases with no or little uncertainty. • For complex problems. it should not matter what the other attribute’s level is. it could be a useful first-cut approximation.Additive Independence (2) • When we assess the utility of one attribute.

The Weights • Recall that PI j = ∑ wi uij PM N i =1 • The weights are scaling factors that sum to unity ∑ wi = 1 1 NPM • They represent trade-offs between PMs. (Clemen) DA 7 Decision Analysis in Practice 19 . They can be assessed directly or using structured approaches.

2000.The Analytic Hierarchy Process – Relative rankings of the objectives are determined with respect to an overall goal – Pairwise comparisons are used to derive weights representative of decision maker concerns T.L. DA 7 Decision Analysis in Practice 20 . Saaty. RWS Publications. Pittsburgh. Fundamentals of Decision Making and Priority Theory with The Analytic Hierarchy Process.

Environment Programmatic vs. Human Health & Safety Environment vs. 2. 10. 3. Socioeconomic / Cultural vs. Socioeconomic / Cultural Human Health & Safety vs. Life Cycle Cost Programmatic vs. 9.An Example of Pairwise Comparisons Relative Importance Assessment Objective categories: Compare the following with respect to the OVERALL DESIRABILITY objective 1. 6. 8. 4. 7. Socioeconomic / Cultural Programmatic vs. Environment Life Cycle Cost vs. Programmatic _______ _______ _______ _______ _______ _______ _______ _______ _______ _______ Key for the evaluation: 1 equally 3 weakly 5 strongly 7 demonstrably or very strongly 9 absolutely Use even numbers to express compromise. DA 7 Decision Analysis in Practice 21 . Life Cycle Cost Socioeconomic / Cultural vs. Human Health & Safety Environment vs. Life Cycle Cost Human Health & Safety vs. 5.

not for consistency among stakeholders. • Redundant information is elicited. DA 7 Decision Analysis in Practice 22 . • The CI is for internal consistency only. • Define the consistency index as λ max − n CI = n −1 • If CI > 0. • The assessor always approves the final weights. identify inconsistencies and inform the assessor.The Practice • People are not consistent in their assessments.2.

STAKEHOLDER RANKINGS AND WEIGHTS Category Stakeholder SH1 SH2 SH3 SH4 SH5 SH6 Programmatic Life Cycle Cost 4 (8) 5 (2) 6 (2) 5 (5) 6 (3) 4(12) 3 (11) 3 (14) 4 (7) 4 (8) 4 (10) 6 (5) Socioeconomic Cultural Environment Human Health & Safety 2 (34) 1 (38) 2 (39) 3 (17) 2 (20) 2 (27) 1 (39) 1 (38) 1 (40) 1 (41) 1 (52) 1 (33) 6 (4) 6 (2) 5 (4) 2 (25) 5 (4) 3 (13) 6 (4) 4 (6) 3 (8) 6 (4) 3 (11) 5 (10) Figure by MIT OCW. DA 7 Decision Analysis in Practice 23 .

1 24 . each PM range is divided into three intervals: good. • In this problem. Good Good 1 Moderate 1/3 Bad 1/6 Moderate Bad 3 6 1 3 1/3 1 DA 7 Decision Analysis in Practice u 0..Utilities • PM scales provide the means to measure the degree of achievement of the objectives. • Utilities are developed using the AHP again.3 0. e.6 0. and bad.g. moderate.

. + w i u ij + . + w k ukj + ... the decision maker should be indifferent between these two levels...Consistency (sanity or reality) Checks • Recall again that PI j = w 1 u1 j + .. + w N u N PM PM j • Suppose that for two PMs i and k the levels l and m on the constructed scales are such that m = wi u wk ukj l ij • Then. DA 7 Decision Analysis in Practice 25 .

DA 7 Decision Analysis in Practice 26 .POLLUTANT PATHWAYS Atmospheric Transport Precipitation Evapotranspiration Run off Resuspension Infiltration Leaching Over flow Deposition Drinking Water & Irrigation Stream Site Boundary Transport to Aquifer at Retarded Velocities Transport Through Aquifer at Retarded Velocities Figure by MIT OCW.

DA 7 Decision Analysis in Practice 27 .

205 (4) .130 (5) .111 (5) .115 (5) .128 (4) . Figure by MIT OCW.205 (1) 3 .120 (4) .082 (5) .071 (6) .258 (1) 2 .132 (5) .181 (1) 4 .094 (6) .155 (3) .183 (5) .089 (4) 6 .194 (1) Performance Indices and RAA rankings for all stakeholders.159 (2) .172 (3) .223 (2) .107 (1) .154 (4) .048 (6) .053 (6) .139 (6) .091 (3) .122 (3) . DA 7 Decision Analysis in Practice 28 .179 (2) .128 (2) 5 .STAKEHOLDERS RAA A B C D E F 1 .216 (3) .177 (3) .114 (4) .185 (2) .135 (1) .091 (2) .050 (6) .

DA 7 Decision Analysis in Practice 29 . •The dominant drivers for the decision choices are also reported. •Preliminary conclusions regarding all stakeholders are drawn.Preparing for Stakeholder Deliberation •Each stakeholder receives his/her numerical results in advance. This concept is borrowed from risk assessment.

08 0. DA 7 Decision Analysis in Practice 30 .14 0.06 F D C B E A Remedial Action Alternative Figure by MIT OCW.Stakeholder 3 0.1 0.18 0.2 0.16 Performance Index 0.12 0.

DA 7 Decision Analysis in Practice 31 .” Risk Analysis 18 (1998): 621-634. Susan E.Image removed due to copyright restrictions. George E.. Please see Table 5 in Apostolakis. “Deliberation: Integrating Analytical Results into Environmental Decisions Involving Multiple Stakeholders. and Pickett.

” Risk Analysis 18 (1998): 621-634. and Pickett. “Deliberation: Integrating Analytical Results into Environmental Decisions Involving Multiple Stakeholders. Please see Table 6 in Apostolakis.. DA 7 Decision Analysis in Practice 32 .Image removed due to copyright restrictions. Susan E. George E.

Susan E.Deliberation Image removed due to copyright restrictions. George E.” Risk Analysis 18 (1998): 621634.1 in Apostolakis.. and Pickett. DA 7 Decision Analysis in Practice 33 . “Deliberation: Integrating Analytical Results into Environmental Decisions Involving Multiple Stakeholders. Please see Section 6.

and Pickett.2 in Apostolakis. DA 7 Decision Analysis in Practice 34 .Image removed due to copyright restrictions. Please see Section 6.” Risk Analysis 18 (1998): 621634. Susan E. “Deliberation: Integrating Analytical Results into Environmental Decisions Involving Multiple Stakeholders. George E..

Susan E. and Pickett. “Deliberation: Integrating Analytical Results into Environmental Decisions Involving Multiple Stakeholders.Image removed due to copyright restrictions..” Risk Analysis 18 (1998): 621-634. Please see Table 7 in Apostolakis. George E. DA 7 Decision Analysis in Practice 35 .

. and Pickett. Accorsi. “Prioritizing Stakeholder Concerns in Environmental Risk Management. 2000.E. Bonano. E. P. California: Duxbury Press. G.” Journal of Risk Research.T.E. and Selection of Environmental Remediation Alternatives..... Salter. Clemen. “Application of Risk Assessment and Decision Analysis to the Evaluation. S. E. 71:35-57.. G. Apostolakis. 1999. 18:621-634.” Journal of Hazardous Materials.” Risk Analysis. G. Apostolakis.Additional References Accorsi. R.. G. 34:387-411. 1998.” Progress in Nuclear Energy.. Making Hard Decisions. 1999.. A. E. and Zio.. 1996 DA 7 Decision Analysis in Practice 36 . Apostolakis. R. “Deliberation: Integrating Analytical Results into Environmental Decisions Involving Multiple Stakeholders. E. 2nd Edition.. and Apostolakis. Belmont.J. S. Ranking. and Jennings. Ghassemi. R. “Developing Utility Functions for Environmental Decision Making.E... Zio. 2:11-29.F.

H. Susskind. D. 1996. Hughes. 1994. 1988. 1987. L.. W. von Winterfeldt. UK. Gregory... Harvard University Press. "Deriving Utilities Using the Analytic Hierarchy Process. and Cruikshank.. D. and Keeney. “Simplicity in Decision Analysis: An Example and a Discussion.L. 14:107-130.” Risk Analysis.E. Keeney. National Research Council. Raiffa. 20:393-395.L.” Management Science. and A. “Creating Policy Alternatives Using Stakeholder values. 40:1035-1048.C. and von Winterfeldt. J. DA 7 Decision Analysis in Practice 37 .R. W. Bell. R.. MA.” Socio-Econ. Edited by D.References (cont’d) Edwards. R. Sci. Washington. D. Cambridge. 1994. 1986. Tversky. R. and Moody.” in: Decision Making.. “Managing Nuclear Waste from Power Plants. Understanding Risk: Informing Decisions in a Democratic Society.. National Academy Press.. Plann. Cambridge University Press.. Breaking the Impasse.L. D.

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->