Risk Analysis

1
Risk Analysis
2
Risk

• Risk is the likelihood and severity of adverse
  outcomes
• Often measured as potential loss-of-life,
  property damage, or undersired ecosystem changes
• Often, likelihood and outcome are to some degree
  uncertain

3
Risk Points

• Risk is everywhere
• Some risks more serious than others
• Zero risk is not an option
• Risk is unavoidable
• We need to. . .
• Identify describe risks (Risk Assessment)
• talk about them (Risk Communication)
• do something about them (Risk Management)

4
Why Do Risk Analysis?

• To improve decisions
• To ensure safe infrastructure
• To ensure reliable production of water resource
  outputs
• To improve cost control
• To make informed resource allocation choices
  under uncertain conditions
• Because we have to (ER 1105-2-100 others)

5
Historical Response to Risk

• Risk presumed to be managed by engineering
  design
• Freeboard on levees and dams
• Cost contingencies
• Underkeel clearance design criteria
• Channel width design criteria
• Probable maximum flood spillways

6
Traditional Risk Management

• Increased certainty of engineering performance
• Presumption that only added to costs
• Little or No addition to benefit estimates
• Viewed as necessary to achieve claimed benefits

7
Traditional Risk Management

• Risk management was implied attempt to provide
  some level of assurance of project performance
• Value of assurance not quantified

8
Risk Analysis
Risk Assessment
Risk Management
Risk Communication
9
Risk Analysis

• Risk analysis is the process of separating the
  whole of risk into its component parts by
  assessment of the risk and related uncertainties
  for the purpose of efficacious management of the
  risk,facilitated by effective communication about
  the risks.

10
Risk Assessment

• What can go wrong?
• How can it happen?
• How likely is it?
• What are the consequences?

11
Risk Management

• What questions do we want risk assessment to
  answer?
• What can be done to reduce the impact of the risk
  described?
• What can be done to reduce the likelihood of the
  risk described?
• What are the trade-offs of the available options?
• What is the best way to address the described
  risk?

12
Risk Communication

• With whom do you communicate?
• What do people know? How do they know it? What
  do they want to know?
• How do you get both the information you need and
  the information others have?
• How do you convey the information you want to
  communicate?
• When do you communicate?

13
Risk Exercise I

• Name the one riskiest thing you do
• Name a risk free thing you do

14
Uncertainty in Civil Works Assessment

• Knowledge Uncertainty
• Model Uncertainty
• Quantity Uncertainty

15
Knowledge Uncertainty

• Things are unknown
• Learn
• Experts
• Things are unknowable
• Knowledge will come
• Knowledge will never come

16
Model Uncertainty

• The way we think about a problem
• Assumptions
• Model structure
• Difficult to address
• Important but rarely investigated

17
Sources of Uncertainty in Empirical Quantities

• Random error statistical variation
• Sample error
• Systematic error subjective judgment
• Calibration of equipment, feels like an 8
• Linguistic imprecision
• Fill to -1
• Variability
• Tons unloaded by ships, some fish die

18
Sources of Uncertainty in Empirical Quantities

• Randomness unpredictability
• Casualties
• Disagreement
• Partners, experts

19
Tools of Risk Analysis

• Probability and Statistics
• Event Trees/Fault Trees
• Monte Carlo Simulation

20
Examples of Risk in Evaluations and Decisions

• Flood damage Urban Flood Damage Reduction
• Vessel safety
• Dam safety
• Major rehabilitation

21
Risk in Navigation Analysis

• Vessel safety
• Lock/Dam component and channel reliabilities
• Ecosystem vulnerability and resilience
• Economic and financial loss

22
Key Uncertainties in Navigation

• Commodity demand uncertainty
• Future with and without
• Foreign/Domestic
• Origin/destination
• Characteristics
• Model uncertainty
• Technology

23
Vessel Safety

• Risk quantification
• Probability of casualty
• Collision, allision, grounding
• Consequences
• Property loss
• Delay
• Environmental damage (spill)
• Personnel Injury

24
Vessel Safety

• Navigation improvement
• Channel widening
• Reduces collision risk in vessel passes
• Channel deepening
• Reduces grounding risk in vessel transits

25
Vessel Safety

• Typical Evaluation
• Expected value of reduced property losses and
  delays as benefit
• Environmental effects subsumed into frequency of
  spill and spill size distributions
• Injury ignored

26
Vessel Safety

• Navigation improvement
• May initiate response that pushes risk level back
  to pre-project level
• Wider vessels
• Vessel managers convert risk reduction into
  profit so there are no benefits from risk
  reduction
• Only from transportation cost savings

27
Multiple Scenarios AnalysisEx.--Mt. St. Helens

• Problem
• Sediment deposited in rivers will increase
  residual flood risk
• Choose least cost method to maintain level of
  flood protection
• Key uncertain variable
• Forecast volume and timing of eroded sediment
• The Sediment Budget

28
Multiple Scenarios AnalysisEx.--Mount St. Helens

• Alternatives Considered
• Single and multiple sediment retention dams
• Dredging
• Combinations

29
Multiple Scenarios AnalysisEx.--Mount St. Helens

• Approach
• Evaluate each alternative for each scenario based
  on
• economic cost
• environmental values and impacts
• risk
• uncertainty in sediment yield relative to
  breakeven yield

30
Mount St. Helens Decision Matrix
31
Mount St. Helens Decision Matrix cont.
32
Forecast vs Actual Sediment
33
Questions in Risk Analysis

• Quantification of uncertainties?
• Distributions
• Scenarios
• Use of Monte Carlo simulation?
• What to simulate?
• Inclusion of Correlation?

34
Conclusions

• Risk and associated uncertainty inherent in water
  resources
• Key quantities frequently exhibit both
  variability and uncertainty
• Uncertainty arises from things like
• Lack of knowledge
• Small samples
• Use of models

35
Conclusions

• Probabilistic risk analysis results can be very
  powerful
• Expose strong from weak
• May result in ambiguity about alternatives
• No clear winner

36
Conclusions

• Probabilistic risk assessment may not always be
  credible
• Lack of knowledge
• Lack of models
• Decisions should explicitly acknowledge
  uncertainty
• Influence on recommendation