“ Building Strong “ Delivering Integrated, Sustainable, Water Resources Solutions Risk-Informed Decision Making Charles Yoe, PhD [email protected] Institute.

“ Building Strong “

Delivering Integrated, Sustainable, Water Resources Solutions

Risk-Informed Decision Making

Charles Yoe, [email protected]

Institute for Water Resources2010



In a Nutshell

• Decision making is going to change if we do risk assessment

• Making decisions under uncertainty• Using the available information

– Considering more than expected value



Risk-Informed Decision Making

• Confluence of risk management and risk assessment– Assessors convey significance of uncertainty– Managers take it explicitly into account in

decision making

• Develops and uses risk information to aid decisions made under uncertainty– Decision metrics and rules developed by

decision makers– Use risk assessment approach to develop

science-based values of risk metrics



The Job

• Success in risk management is defined by practical and useful solutions for dealing with uncertainty

• It is the risk assessors’ job to address the uncertainty and variability in inputs

• It is the risk managers’ job to address the uncertainty and variability in the decision criteria– And to manage risks that are not acceptable



Key Focal Points

• Decision criteria/metrics• Acceptability/tolerability• Principles for choosing an option• Uncertainty• Strategies for managing risk



Metrics

• Decision criteria will reflect– Risk metrics– Other risk management objectives

• These are the usual impacts of interest to decision makers

– Costs, benefits, compliance with laws, budget impacts, capability, and so on



Possible Risk Metrics

• Life safety risk– Number of lives at risk, social vulnerability

• Economic risk– Net economic benefits, financial risks

• Engineering risk and reliability– PUP

• Residual risk• New, transformed and transferred risk• New metrics like partitioned risk



First, Understand

• …assessors have to become more clever and help risk managers know what is possible

• …conditional probabilities can help– …given that something happens then how bad

is it?



Consider the Hydroeconomic Risk Model

Existing

Existing

Levee

Levee



Expected Annual Damage

$12,411,000

$2,151,000

Existing Risk

Residual Risk



Refining Residual Damage

Existing flood risk (EAD) = $12,411,000Residual flood risk (EAD) = $2,151,000Flood risk reduction (EAD) = $10,260,000

What are the EAD given that a flow with an exceedence frequency of 0.002 (.2 on left) or less occurs? This conditional probability requires a risk partition.

Rescale the partition

90.0% 5.0%

0.8418 0.9644

0.8

20.8

40.8

60.8

80.9

00.9

20.9

40.9

60.9

81.0

01.0

2

Values in Millions

0.0

0.2

0.4

0.6

0.8

1.0

New CDF / INTERVAL

Cumul(840130.3798,999999.9783,{},{})

Minimum 840130.3798Maximum 999999.9783Mean 872658.5740Std Dev 36000.3467

Integrate this rescaled partition and the expected damage is now$872,659,000. Given that a ‘bad’ flood occurs, it will be very bad.



Extreme Event Conditional Expected Values

• With condition residual = $2,151,000• Benefits = $10,260,000• Partitioned residual = $872,659,000• A channel or nonstructural project could

have lower benefits and a much reduced partitioned risk



What Kinds of Decisions

• Is the status quo acceptable? Can we accept this risk?

• If the existing risk is not acceptable, what level of risk is tolerable?

• What is the best way to achieve a tolerable level of risk?



Two Important Ideas

• Acceptable Risk-A risk whose probability of occurrence is small, whose consequences are so slight, or whose benefits (perceived or real) are so great, that individuals or groups in society are willing to take or be subjected to the risk (or that the event might occur).

• Tolerable Risk-A tolerable risk is a non-negligible risk that has not yet been reduced to an acceptable level, but the risk is tolerated due to the inability to reduce it further, the cost of doing so or the magnitude of the benefits associated with the risky activity.



Principles for Choosing Management Option

• Policy• Zero Risk• Weight-of-evidence• Precautionary Principle • ALARA Principle• ALOP Principle• Reasonable Relationship• Safety Standards



Policy

• Some decisions have been made for the risk manager, e.g., the national economic development plan

• Corps guidance often restricts what can be done

• Working with EPA means dealing with their policy restrictions and requirements

• International treaties and agreements may identify solutions or limit options



© Charles Yoe, Ph.D.

Zero Risk• The ban or taboo—any action that involves any

risk at all is forbidden—ignorance made it possible to imagine 0 risk

• “De Minimis”—numerical value of risk too small to be bothered about

• 1-in-a-million has captured the imagination in some contexts– 1-in-a-million annually is 70 or 80 times more than 1-

in-a-million lifetime– “Practically” zero

• PMF, 100-year, 500-year have all been used by the Corps



Weight-Of-Evidence

• We like once and for all resolution on the basis of compelling scientific evidence

• Risk managers assess credibility of conflicting evidence about hazards and risks in a systematic and objective manner

• Diverse group of scientists examine the evidence to reach consensus views

• Weight-of-evidence is an ongoing activity attempting to balance positive and negative evidence of harmful effects based on the best available evidence



Precautionary Principle

• "When an activity raises threats of harm to human health or the environment, precautionary measures should be taken even if some cause and effect relationships are not fully established scientifically. In this context the proponent of an activity, rather than the public, should bear the burden of proof." Wingspread Statement

• In order to protect the environment, the precautionary approach shall be widely applied by States according to their capabilities. Where there are threats of serious or irreversible damage, lack of full scientific certainty shall not be used as a reason for postponing cost-effective measures to prevent environmental degradation. UNCED Rio Declaration Principle 15




ALARA Principle

• As Low As Reasonably Achievable• Minimize known risks, by keeping

exposures as low as reasonably possible, taking into consideration costs, technology, benefits to public health and safety and other societal and economic concerns

• Similar to best available technology WHO Backgrounder




Appropriate Level of Protection

• Risk society is willing to accept or tolerate• Statement of degree of protection that is

to be achieved by the risk management option implemented

• Can be intentional choice• Can be result of what is doable

– Sometimes overlaps with ALARA



Factors in Determining ALOP

• Technical feasibilityTechnical feasibility of prevention and control options

• RisksRisks that may arise from risk management interventions

• Magnitude of benefitsbenefits of an activity and the availability of substitutes for that activity

• Cost of preventionCost of prevention and control versus effectiveness of risk reduction

• PublicPublic risk reduction preferencespreferences, public values

• DistributionDistribution of risks and benefits

• Policy



Reasonable Relationship

• Costs of risk management should bear “a reasonable relationship” to corresponding reductions in risks

• Not a true BCA but balance of benefits and costs should at least approximate each other



Safety Standards

• Safety standards– Zero risk standards– Thresholds– Balancing standards

• Risk-Benefit Trade-Off: Greater benefit => Greater acceptable risk

• Comparative Risk Analysis: Rank risks for seriousness of threat posed

• Benefit-Cost Analysis: Economic efficiency




Safety Standards

• Threshold standards– Non-zero level of risk stipulated as acceptable

• Tolerable level of risk• Appropriate level of protection

• Procedural standards– Agreed upon process– Negotiation– Referendum



Uncertainty

• Consider net NED benefits as a decision criterion

• Which is the best plan?

Net Benefits Plan 1: Red Plan 2: Blue Plan 3: Green

Mean $6,998,000 $6,001,142 $4,000,039

Decision making will becoming multi-dimensional.Can we handle that?



Uncertainty

12.3% 86.2% 1.5%2.6% 97.4% 0.0%

0.00 20.00

-20

-15

-10

-50

51

01

52

02

5

Net Benefits (Millions)Values in Millions

Frequen

cy

Net Benefit Comparison

Five Number Summary Plan 1 Plan 2 Plan 3Plus Mean Red Blue Green

Mean 6,998,000$ 6,001,142$ 4,000,039$ Minimum (10,173,310)$ (4,114,583)$ (15,726,346)$ 1st Quartile 3,076,969$ 3,960,606$ 318,544$ Median 6,978,948$ 5,952,181$ 4,031,267$ 3rd Quartile 10,931,366$ 8,087,706$ 7,682,767$ Maximum 23,635,943$ 16,894,192$ 19,273,900$

P(Negative return)

Which is the best plan now?



Possible Rules for Making Decisions Under Uncertainty

• Maximax—appeals to gamblers (R, G, B)– Choose plan with best upside payoff

• Maximin—appeals to cautious (B,R,G)– Choose plan with best downside payoff

• Laplace—risk neutral (R,B,G)– Choose plan based on expected value payoff

• Hurwicz—neither neutral nor extreme (depends)– Choose plan based on assigned weights

• Regret-make wrong decision and regret it (Red)– Identify the maximum regret associated w/ each choice– Choose plan that minimizes this regret



Regret

States of the world

Min Max

Red $ (10,173,310) $ 23,635,943 Blue $ (4,114,583) $ 16,894,192 Green $ (15,726,346) $ 19,273,900

Regret MatrixMaximum

Red $ 6,058,727 $ - $ 6,058,727

Blue $ - $ 6,741,751 $ 6,741,751

Green $ 11,611,763 $ 4,362,043 $ 11,611,763

Minimax $ 6,058,727

If we choose red our regret will at most be $6.1 million



Trade-Offs

Plan A Plan B Plan C Plan D Plan E Plan F Plan G Plan HFlood risk management ($1000s EAD) $10 $800 $60 $80 $25 $50 $100 $900Annual HU's 46 0 50 90 85 75 116 116Annual user days of urban recreation 700 0 850 2000 1000 800 3000 3000Property value effects 2 3 2 1 3 1 2 3Cost ($Millions) $15 $15 $23 $33 $20 $27 $53 $68



Sensitivity



General Risk Management Strategies

• Accept • Mitigate

– Avoidance– Reduce probability of event (prevent)– Reduce consequence of event (mitigate) – Insurance



Charles Yoe, [email protected]

Questions?

“ Building Strong “ Delivering Integrated, Sustainable, Water Resources Solutions Risk-Informed Decision Making Charles Yoe, PhD [email protected] Institute.

Documents

risk information

risk partition

partitioned risk slide

risk assessment assessors

risk assessors job

risk managers job

risk management objectives

risk assessment approach