and project managers are responsible for the risk (or may be deciding whether to become responsible), then ways of reducing the risk need to be developed. Identification of risk treatment options is carried out concurrently with assess- ment of the potential benefits and costs of reducing the exposure to risk to ac- ceptable levels. These actions are important steps in the process of formulating a risk treatment strategy, which is discussed in Chapter 7. W ATER U TILITY E XAMPLE The following water utility example demonstrates application of the steps in the risk analysis stage. The event tree of Figure 6.8 shows that the annual frequency of a sunny-day failure caused by earthquake-induced cracking of the embankment is estimated to be 9 × 10 –8 per year (or around 1 in 11 million years). The estimated frequency of a sunny-day failure caused by earthquake-induced slippage of the embankment is estimated to be 9 × 10 –7 per year (or around 1 in 1.1 million years). The total likelihood of an earthquake triggering a sunny-day failure was esti- mated, by summing the above frequencies, to be 9.9 × 10 –7 per year (or around 1 in 1 million years). The event tree shows that earthquake-induced slippage of the embankment is the predominant contributor to the total likelihood of earthquake- induced sunny-day failure. Figure 6.8 indicates that the likelihood of sunny-day failure of the embankment due to all trigger events is estimated to be 1.01 × 10 –4 per year (or almost 1 in 10,000 years). The likelihood of embankment instability during full storage con- ditions (9.08 × 10 –5 per year) is by far the greatest contributor to the overall like- lihood of sunny-day failure. Figure 6.9 indicates the forecast distribution chart of consequential cost as- suming that a sunny-day failure occurs. It also includes a table of forecast cost for increasing confidence level intervals of 5 percent. The chart shows that the esti- mated cost distribution is highly skewed toward the high cost end. A central esti- mate for the consequential cost of a sunny-day failure would be approximately $350 million. The table shows that the lowest cost estimate computed during the 2,000 trials was around $120 million; the highest calculated cost derived in the tri- als was approximately $5 billion. The Utility decided that for the assessment, a conservative yet reasonable cost that could be used for planning purposes would be defined as the 80 percent confidence-level cost. The Utility selected the 50 percent confidence-level cost as representing an optimistic position and the 95 percent confidence-level cost as pessimistic. In the sunny-day failure example, the estimated optimistic cost, plan- ning cost, and pessimistic cost values were approximately $340 million, $560 million, and $1 billion, respectively. Table 6.3 shows the calculated risk quotient, using the conservative estimate of consequence cost, for each dam failure type. It also shows the total risk (the sum of the risk quotients of all failure types) posed by the dam. 84 / Stage 3: Analyze the Risk 3672 P-06 5/3/01 2:24 PM Page 84 9.00E-08 9.00E-07 9.90E-07 9.00E-05 8.10E-07 9.08E-05 9.00E-06 9.00E-06 SUNNY-DAY FAILURE 1.01E-04 Combined Probabilities (Frequency per year) RI 5 Earthquake: Breach of Main Embankment SUNNY-DAY FAILURE TRIGGER EVENTS Earthquake Embankment cracks Piping failure Main embankment breach Sunny-day failure (high volume, very high rate) 0.00009 1.0 0.1 0.01 Batter slip Deformation Main embankment breach Sunny-day failure (high volume, very high rate) 0.1 0.1 1.0 Annual Frequency Probability Probability Probability RI 6 Geotech: Embankment Instability, Breach of Main Embankment Storage full Residual strength U/S stability fail Main embankment breach Sunny-day failure (high volume, very high rate) 0.9 0.1 1.0 0.001 Softened strength U/S stability fail Main embankment breach Sunny-day failure (high volume, very high rate) 0.9 0.001 0.001 RI 7 Geotech: Piping, Breach of Main Embankment Earthquake: breach of main embankment Geotechnical failure: embankment instability, breach Geotechnical failure: piping, embankment breach Storage full Cracks below FSL Piping failure Main embankment breach Sunny-day failure (high volume, very high rate) 0.9 0.1 0.01 0.01 Figure 6.8 Sunny-day failure event tree showing the combined probabilities of events that could lead to a sunny-day failure. 85 3672 P-06 5/3/01 2:24 PM Page 85 Confidence Limit 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 55% 60% 65% 70% 75% 80% 85% 90% 95% 100% Calculated Cost ($million) 122.5 195.2 216.7 231.9 245.6 257.1 270.9 286.8 305.2 322.1 338.4 360.1 385.1 416.4 457.8 504.3 559.2 643.3 766.5 1,042.5 5,067.9 Frequency Chart .000 .019 .037 .056 .074 0 37 74 111 148 0.0 437.5 875.0 1312.5 1750.0 2,000 Trials 33 Outliers Forecast: Set 3 Cost Probability Frequency Figure 6.9 Sunny-day failure forecast distribution of consequential cost and tabulated forecast values at 5 percentile confidence levels. 86 3672 P-06 5/3/01 2:24 PM Page 86 Figure 6.10 shows a ranked risk profile of the failure types applicable to the dam. The risk profile is a bar chart and shows that the riskiest failure type (sunny- day failure with a calculated risk quotient of $30,000 per year) is approximately twice as risky as the second riskiest failure type (vehicle accident, $17,000 per year). In turn, the second riskiest event presents approximately 1.5 times more risk Water Utility Example / 87 Table 6.3 Calculated Risk Quotients Failure Type Financial Risk Quotient ($ × million per year) F1 Flood & Main Embankment Breach 6.08E–05 F2 Flood & Secondary Embankment Breach 6.07E–05 F3 Sunny-Day Failure 2.96E–02 F4 Outlet Works—short-term outage 3.37E–03 F5 Outlet Works—long-term outage 1.21E–02 F6 Vehicle-Accident 1.71E-02 Total Dam Financial Risk Quotient 6.23E–02 Figure 6.10 Financial risk profile showing a comparison of the risk of each failure and the total risk presented by all failure types. 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 F3 Sunny- Day Failure F6 Vehicle Accident F5 Outlet Works Long-Term Outage F4 Outlet Works Short-Term Outage F1 Flood & Main Embankment Breach F2 Flood & Secondary Embankment Breach Total Dam Financial Risk Quotient Risk Quotient ($million per year) 3672 P-06 5/3/01 2:24 PM Page 87 than the third most risky event (outlet works failure and long-term outage, $12,000 per year). The risk profile is quite typical, in that it shows that a relative handful of events accounts for the vast majority of the total risk amount. Total risk is the sum of the calculated risk quotients for all relevant risk events, shown in Figure 6.10 as the total dam financial risk quotient bar. Figure 6.11 shows an example of a proportional risk profile. It shows the events ranked in order of decreasing risk quotient. The height of each bar indicates the proportion (percentage) contribution of each event to total risk. The bar chart in- dicates, for example, that the riskiest event, sunny-day failure, contributes to around 48 percent of the total risk and that the second most risky event (vehicle ac- cident) is responsible for around 28 percent of the total risk. The line plot of Figure 6.11 shows an alternative way of expressing the contri- bution of each additional risk event to total risk. The line plot shows the progres- sive contribution to total risk, from the most risky event to the least risky event. The line plot shows that proceeding on from the riskiest event, the two most risky 88 / Stage 3: Analyze the Risk Figure 6.11 Contribution of failure types to total dam risk showing the cumulative increase in risk for failure types ranked in order of decreasing risk quotient. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Contribution to Total Dam Risk Contribution to dam risk (%) Cumulative Risk (%) F3 Sunny-Day Failure F6 Vehicle Accident F5 Outlet Works Long-Term Outage F4 Outlet Works Short-Term Outage F1 Flood & Main Embankment Breach F2 Flood & Secondary Embankment Breach 3672 P-06 5/3/01 2:24 PM Page 88 events contribute to approximately 75 percent of the total risk and the three most risky events contribute to around 95 percent of the total risk. The plot also clearly indicates that the two least risky events (both events related to flooding) provide a negligible contribution to total risk. Figure 6.12 shows the exposure profile for the example water storage. The risk events are ranked in order of decreasing risk quotient, as shown by the line graph of “financial risk.” Exposure profiles reflect the exposure to residual risk resultant from ongoing business risk management. Consequently, exposure profiles ideally show that the least risky events present the greatest exposure. The exposure pro- file of Figure 6.12 is unusual (and undesirable) because the riskiest event (sunny- day failure) presents the greatest financial exposure of all the identified risk events. Figure 6.12 shows that a conservative estimate (at the 80 percent confidence level) of the potential cost if a sunny-day failure were to occur would be approx- imately $560 million. An optimistic estimate of the exposure would be approxi- mately $340 million; a pessimistic estimate would be over $1 billion. In contrast, Water Utility Example / 89 Figure 6.12 Financial risk and exposure profile showing the estimated costs if the failure types occur, with failure types ranked in order of decreasing risk. 0 200 400 600 800 1000 1200 Risk Cost ($million) 0.00 0.01 0.02 0.03 0.04 0.05 0.06 Financial Risk ($million) CL 95% Financial Risk Cost ($m) CL 80% Financial Risk Cost ($m) CL 50% Financial Risk Cost ($m) Financial Risk ($m per year) F3 Sunny-Day Failure F6 Vehicle Accident F5 Outlet Works Long-Term Outage F4 Outlet Works Short-Term Outage F1 Flood & Main Embankment Breach F2 Flood & Secondary Embankment Breach 3672 P-06 5/3/01 2:24 PM Page 89 90 / Stage 3: Analyze the Risk the combined pessimistic cost of the remaining three events that contribute to most of the total risk would be approximately $65 million. Figure 6.12 also allows comparison of uncertainty associated with the cost esti- mates. The large range of optimistic and pessimistic cost estimates ($340 million to $1.05 billion) for sunny-day failure clearly shows that there is considerable uncer- tainty in the estimate of cost of that event. Compared with sunny-day failure, the second lowest risk event (flooding leading to a breach of the main embankment) presents a similar optimistic estimate of exposure ($320 million). However, the cost range (uncertainty) is substantially less than sunny-day failure, as shown by the pessimistic cost estimate of approximately $770 million for the flooding event. Figure 6.13 shows a risk map of the potential dam failure modes. For compar- ison, the map shows plots of the house and car risk parameters from the household insurance example discussed earlier and illustrated in Figure 6.6. Using the lines of equal risk in Figure 6.13 as a guide, sunny-day failure is clearly the riskiest event, with a relatively low likelihood (approximately one in 10,000 years) but a very high cost if it occurs (average around $300 million). In comparison, the vehicle accident event poses almost as much risk as sunny-day Figure 6.13 Risk map of dam failure modes showing the risk relationship of each failure type and a comparison with the household insurance example. F1 Flood & Main Embankment Breach F2 Flood & Secondary Embankment Breach F3 Sunny-Day Failure F4 Outlet Works Short- Term Outage F5 Outlet Works Long-Term Outage F6 Vehicle Accident Home Car 0.0000001 0.000001 0.00001 0.0001 0.001 0.01 0.1 1 10,000 100,000 1,000,000 10,000,000 100,000,000 1,000,000,000 Annual Frequency Annual Probability of Failure Typical Household Insurance Risk Conditions Average Cost ($) Risk Quotient = $10 per year Risk Quotient = $100 per year Risk Quotient = $1,000 per year Risk Quotient = $10,000 per year 3672 P-06 5/3/01 2:24 PM Page 90 failure, which is achieved because the accident has a relatively high chance of oc- curring (around one in 250 years) but has a relatively low financial consequence ($4 million). In contrast to sunny-day failure, the lowest risk event shown on Figure 6.13 is a major flood leading to failure of the main embankment. This event has the same consequences as sunny-day failure but has almost three orders of magnitude less likelihood of occurring (likelihood is approximately one in 4.5 million years). The project manager and the risk analyst evaluated the ranked financial risk profile (Figure 6.10) and the proportional risk profile (Figure 6.11) and selected the risk threshold for calculation of the risk cost at $10,000 per year. Thus, the combined occurrence cost of the three riskiest events (which contributed to almost 95 percent of the total risk) was included in the risk cost calculation. Figure 6.14 shows the forecast risk cost distribution for the example dam. The risk cost distribution consists of the combined cost of sunny-day failure, vehicle accident, and outlet works damage leading to long-term outage. The distribution is dominated by the largest cost component (sunny-day failure) and is heavily skewed toward the high cost end of the distribution. The Utility project manager selected the 50 , 80, and 95 percent confidence levels to represent optimistic, plan- ning, and pessimistic estimates of risk cost. The estimated optimistic, planning, and pessimistic risk costs are $360 million, $580 million, and $1.06 billion, respectively. The calculated risk cost at the optimistic, planning, and pessimistic confidence levels is shown graphically in Figure 6.15. The bar chart confirms that there is considerable uncertainty in the estimate of dam risk cost, as indicated by the wide range of cost between the estimated 50 and 95 percent confidence levels. Water Utility Example / 91 Figure 6.14 Forecast risk cost distribution of sunny-day dam failure. .000 .019 .038 .056 .075 0 37.5 75 112.5 150 0.000 437.500 875.000 1312.500 1750.000 Frequency Probability Forecast: Dam Risk Cost Frequency Chart 2,000 Trials 33 Outliers Certainty is 80.00% from Infinity to 577.363 3672 P-06 5/3/01 2:24 PM Page 91 92 / Stage 3: Analyze the Risk Figure 6.15 Calculated dam risk cost at the optimistic, planning, and pessimistic confidence levels, the range of which provides an indication of the uncertainty associated with the cost estimate. 0 200 400 600 800 1,000 1,200 Dam Risk Cost Pessimistic (CL 95%) Planning (CL 80%) Optimistic (CL 50%) Risk Cost ($million NPV) 3672 P-06 5/3/01 2:24 PM Page 92 7 S TAGE 4: F ORMULATE A R ISK T REATMENT S TRATEGY The objective of risk treatment is to identify and implement appropriate manage- ment actions to address targeted risks, with the objective of reducing their likeli- hood of occurrence and/or the severity of the consequences. Development of a risk treatment strategy involves selection of one or more risk treatment options, which collectively will reduce the overall risk exposure to the project or business to an acceptable level. The options that are selected depend on the degree of control that the organization has over the risk and the relative benefits and costs of treating the identified risks. The greater the level of control, the more likely the risks will be able to be treated using proactive measures. The aim of a risk treatment strategy is to progressively reduce risk in a timely and cost-effective manner. This aim recognizes the constraint that most businesses cannot implement all actions that may be required to reduce risk to acceptable lev- els. Some actions may be more costly to perform compared with the consequen- tial reduction in exposure to risk. In addition, most businesses have a limited annual budget allocation for risk reduction actions. Due to these limitations, busi- nesses usually need to establish interim risk reduction goals. In order to develop a risk treatment strategy that satisfies the above aims, it is necessary to develop an appreciation of the: • Level of risk that would be acceptable, in both the short and the long term • Events that should be addressed first • Actions that should be taken to reduce the risk • Projected effectiveness of any actions • Cost of actions • Available budget to perform required actions • Financial benefits of risk treatment actions 93 3672 P-07 5/3/01 2:26 PM Page 93 [...]... a Risk Treatment Strategy Risk analysts and project managers, often in consultation with other stakeholder representatives, develop the risk treatment strategy Formulation of risk reduction strategy utilizes the results of the risk assessment process Figure 7.1 shows a flow chart of the process required to develop a risk reduction strategy Each step of the flow chart is described below Evaluate Risks... Prioritize events Determine Risk Treatment Actions Identify cause of risk events Develop action list for each event Evaluate impact on risk No Risk acceptable? Yes Develop Risk Treatment Strategy Estimate action costs Evaluate benefits and costs Generate schedule of expenditure Approve Strategy Figure 7.1 Procedure for risk strategy development 3672 P-07 5/3/01 2:26 PM Page 95 Evaluate Risks against Performance... implementation of the risk reduction actions 3672 P-07 5/3/01 2:26 PM Page 97 Develop a Risk Treatment Strategy / 97 Evaluate Impact on Risk The creation of a preliminary version of the risk model by risk analysts greatly aids in the process of developing a progressive list of actions and assessing their potential impact on risk The interactive model should readily incorporate the impact of risk treatment... Stage 4: Formulate a Risk Treatment Strategy the individual event risk quotient, the total risk quotient, the reduction in exposure to the risk event, or the reduction in risk cost Outputs that are generated to assist development of strategy are typically some combination of: • A prevention profile, which compares the estimated risk reduction costs with the exposure costs for the ranked risk events; prevention... expenditure that reduces the risk quotient of progressively less risky events is usually more easy to derive and defend than a schedule based on more sophisticated analysis In many cases, the actions that are required to reduce the risk quotients of the riskiest events to target levels also reduce the risk quotients of lower-priority risk events Thus, the actions that address the riskiest issues frequently... therefore an iterative process In most cases, it is appropriate that the initial target risk quotient is set at the risk threshold value which was selected to calculate the risk cost The threshold is appropriate because the aim of the risk reduction process is to reduce the risk quotients to levels where the estimated risk cost would be negligible However, it may not be possible in the short term to perform... progressively reduce the risk quotient to the ultimate target value Develop Action List for Each Event The panel members need to consider each risk event in turn, until a series of actions that can potentially reduce the risk has been identified The actions need to be considered for all substantial risk events (usually those that contribute to risk cost exposure) In many cases a risk event has a low likelihood... useful in evaluating value for money with respect to risk management implementation scenarios • Cumulative plots of risk reduction action cost vs exposure reduction • Cumulative plots of risk reduction action cost vs percentage of total risk quotient reduction • Graphical plots of base cost vs risk cost for options • Graphical plots of base cost vs total risk quotient for options • Graphical plots of benefit-cost... relatively low cost, risk managers may decide either to plan for the expenditure in the project budget or to take the risk that the event occurs and to pay for the consequences from the operational budget The output of this process should be an itemized list of actions that can be carried out for each risk event, the current risk presented by each event, and the expected future risk posed by the risk event,... represent very good risk reduction value Under these circumstances, the risk treatment strategy that offers the best risk reduction value for the money is to perform the identified risk treatment actions that progressively work down the prioritized list of risk events until the accumulated costs meet budget In other cases, however, it may be more prudent initially to spread the reduction in risk over a larger . 286.8 30 5.2 32 2.1 33 8.4 36 0.1 38 5.1 416.4 457.8 504 .3 559.2 6 43. 3 766.5 1,042.5 5,067.9 Frequency Chart .000 .019 . 037 .056 .074 0 37 74 111 148 0.0 437 .5 875.0 131 2.5 1750.0 2,000 Trials 33 Outliers Forecast:. contribution to total risk, from the most risky event to the least risky event. The line plot shows that proceeding on from the riskiest event, the two most risky 88 / Stage 3: Analyze the Risk Figure. outage 3. 37E– 03 F5 Outlet Works—long-term outage 1.21E–02 F6 Vehicle-Accident 1.71E-02 Total Dam Financial Risk Quotient 6.23E–02 Figure 6.10 Financial risk profile showing a comparison of the risk