Triple Bottom Line Risk Management Enhancing Profit Environmental Performance and Community Benefits_8 pot

The client needed to understand the potential liability that itcould assume in order to: • Structure the bid to minimize exposure to risk events • Adjust the bid price as appropriate to

This case study examines:

• Including risk in asset valuations

• Including risk in setting a bid price

• Restructuring bids to account for risk

B ACKGROUND

The client is a large U.S.-based independent power producer At the time of thecase study, the client was engaged in a phase of aggressive acquisition of powergeneration assets, both within the United States and globally As a matter of fidu-ciary responsibility, the client’s senior management maintained a policy of risksensitivity (particularly environmental and regulatory risk) in relation to acquisi-tions Until a year or so before this case study, accumulation of the client’s powergeneration assets had traditionally been concentrated on “green field” sites ordemonstrably “clean” sites

However, as opportunities for “clean” assets diminished, the client needed toconsider acquisition of older, operating sites in order to maintain planned growth.Along with these facilities came the potential of inheriting environmental and reg-ulatory liabilities from past and, potentially, future activities Through application

of the RISQUE method, the client felt more able to bid for these older power eration assets and adopted the method of risk profiling during the due diligenceprocess as the key determinant of potential environmental and regulatory liability.The client also considered that use of the method allowed the company to gainsome competitive advantage in the marketplace

gen-163

S ETTING

The client was engaged in a bid (that was ultimately successful) for the purchase

of a large, 2,000 megawatt, coal-fired power generation facility located in theeastern half of the United States The bid process was performed under conditions

of strict confidentiality

Due to past and/or present activities, the client considered that there was somepotential for environmental liability at the site Under draft contract conditions, thepurchaser was required to indemnify the vendor against claims due to environ-mental impairment The client needed to understand the potential liability that itcould assume in order to:

• Structure the bid to minimize exposure to risk events

• Adjust the bid price as appropriate to reflect the extent of exposure to liability

In addition, the client needed to understand the cost implications of future vironmental regulatory changes as they might affect plant operations Thesechanges, too, could impact the bid price

en-Project Objectives

The purpose of the risk profiling was to identify and quantify the potential cial liability from environmental and regulatory risk events In the event that thebid was successful, the risk profiling findings would provide a useful basis forlater development of an environmental risk management strategy

finan-The specific objectives of the risk assessment were to:

• Identify the significant environmental and regulatory risk events associatedwith the past, ongoing, and future operation and management of the generationfacilities

• Quantify the frequency of occurrence and the financial consequences of theidentified issues that may occur during the life of the project

• Identify the riskiest issues and the cost (risk cost) if these issues were to occurover the life of the project

• Provide schedules of potential risk expenditure that could be directly rated into the comprehensive financial model

incorpo-• Present the results in an easy-to-understand, graphical format

Risk Assessment Structure

The approach used in the risk assessment was to consider the following two types

of cost that could reasonably be incurred over the life of the project: (1) base costand (2) risk cost

The base cost for the project was the cost of purchase, operation, and ment plus any associated costs to which the client would be committed (e.g.,planned plant upgrades, taxes) The risk cost for the project was considered to bethe cost of the consequences of those risk events that may occur over the life of thescheme (e.g., the cost of soil remediation due to past spills at a site) For the pur-poses of this assessment, the project life was set at 15 years.

manage-By the end of the 15-year project life, the project’s cost would be equal to thebase cost of the project plus the consequential costs of the risk events that actuallyoccurred The base cost is relatively easy to determine in advance using conven-tional financial modeling However, the cost component associated with riskevents (risk cost) is very difficult to predict due to uncertainty

The risk assessment method involved determination of the two key financialmeasures of risk: (1) risk quotient and (2) risk cost (described above)

The risk quotient is the product of probability and cost and is widely referred

to elsewhere in the literature as “expected cost.” The risk quotient was used torank the identified risk events according to this measure of risk and then determinethe riskiest issues

The client employed the RISQUE method because it was a systematic, sible, and transparent methodology The following summary steps explain the riskassessment procedure that was followed during the due diligence process toachieve the risk assessment objectives

defen-Step 1: Identification of Risk Events This step entails identification, by a diverse

panel of selected experts, of the significant environmental, regulatory, ing, and other risk events that potentially could occur over the project life.The panel judged costs associated with the occurrence of some environmentaland regulatory events initially identified as risk events to be base costs rather thanrisk costs The client decided, as part of the bid structure, to implement actions thataddressed events that were considered to be almost certain to occur (probability ofoccurrence was close to 100 percent) Events assigned to the base cost were indi-cated in the risk register Base costs were not included in the risk model but, in-stead, were transferred into the comprehensive, in-house financial model

engineer-Step 2: Quantification of Risk Events In this step the expert panel quantified the

frequency of occurrence of potential risk events and the consequences (as costs)

of relevant risk events

Step 3: Modeling of Risk Costs Here the risk analyst determined the risk quotient

for each risk event, determined the riskiest issues, and quantified a reasonable timate of the liability (risk cost) associated with the riskiest issues for the powerplant

es-Step 4: Development of Risk Management Strategies In this step the client

iden-tified risk reduction options and the risk analyst assessed impacts on model resultsand the restructured project

Setting / 165

Step 5: Selection of Bid Structure In the final step the client selected the ultimate

risk management actions and incorporated the calculated risk cost into the bid

R ISK I DENTIFICATION

Expert Panel

Identification and quantification of the risk events associated with the plant sitewas performed by a qualified expert panel composed of representatives fromenvironmental and regulatory consultants and senior client staff with expertise inoperations, finance, legal, regulatory, and environmental fields

Nature of Risk Events

The process of risk event identification and quantification was begun at a teamstrategy meeting During the meeting, environmental and regulatory issues werediscussed in terms of air- and water-related issues and contamination issues foreach facility

Following this meeting, the environmental consultant prepared issue sheets foreach facility These sheets formed the basis for the formal risk register The pur-pose of creating the risk register was to:

• Enable those involved in the risk identification process to review the issuesraised to ensure that the list was complete

• Specify reasons for excluding particular risk events in the model

• Document and justify the risk events

The frequency and cost of each risk event was based on the consensus of expertpanels and historical data (if available)

Frequency was expressed as chance of occurring over the next 15 years of theproject life Probability was expressed as the likelihood of a consequence occur-ring, assuming the initiating issue has occurred Due to uncertainty of magnitude

of costs, all costs were provided by the experts in current dollars, to represent “thelikely cost if the issue occurs,” or estimated mean cost and “the cost unlikely to

be exceeded more than 5 percent of times,” or 95 percent confidence limit.For each event, the above cost pairs were assumed to represent the respectivecontrol points on a log normal distribution, unless otherwise stated The spread be-tween the mean and 95 percent values provided a relative comparison of uncer-tainty The larger the spread of values, the larger the uncertainty contained in thoseestimates

Net present value (NPV) costs of risk events were more difficult to estimate cause assumptions of timing of risk event occurrence needed to be made Al-

be-though it was not possible to predict when a risk event would occur, the panel wasable to estimate the earliest year in which each risk event could most likely occur.Assuming that the costs were incurred at the earliest time enabled calculation of aconservative estimate of NPV risk cost for each issue.

Several sets of risk calculation were performed throughout the risk assessmentprocess, and input costs and frequencies of occurrence were adjusted as the level

of understanding of events increased or management plans were modified times in order to mitigate unacceptable risk events) The values that were pre-sented in the risk register were those finally agreed on and adopted for the riskmodel

(some-Panel Conclusions

Table 11.1 lists all of the environmental risk events identified by the expert panel

in relation to the plant It includes a site code, which was derived from the ation sheets, and a brief description of each event The category refers to whetherthe event was classified as a base cost (included directly in the comprehensive, in-house financial model) or a risk event for inclusion in the risk model The right-hand column shows whether the issue was included in the risk model Some issueswere excluded either due to low potential liability or because the risk was coveredelsewhere in the financial assessment

evalu-R ISK A NALYSIS

Risk Modeling

The risk modeling process consisted of the following steps:

Step 1 Input Issues and consequences were input in event-tree format.

Values input into the model were event frequencies and a schedule

of potential event costs over time

Step 2 Calculations Calculation of cost (NPV) assuming each risk event

occurs and the risk quotient for each risk event

Step 3 Determine riskiest issues The riskiest issues were defined as those

issues that contributed to 95 percent of the total project risk The 95percent mark was selected because at around this point, additionalissues did not add significantly to the total project risk

Step 4 Run the risk model Perform Monte Carlo simulation for 2,000

trials

Step 5 Generate output Produce graphical risk profiles and risk

relationships

Risk Analysis / 167

168

Model Results

The primary aim of the first run of the risk model was to assess whether the ronmental and regulatory risk cost of the project, as perceived at the time, was ma-terial to the bid If so, the model results were to form the basis of development ofthe most cost-effective strategy to reduce the risk cost to the lowest reasonableamount After several iterations to progressively reduce the risk cost, the clientwould be in a position to judge whether the exposure was material and, if neces-sary, incorporate the risk cost and risk reduction measures into the bid

envi-Initial Risk Cost Table 11.2 shows the identified risk events tabulated in order of

decreasing risk quotient for Run 1 The risk quotient was estimated at three levels

of confidence selected by the client to represent optimistic, planning, and simistic cost estimates The selected confidence levels were 50, 75, and 95 per-cent, respectively

pes-Preliminary inspection of the results of Table 11.2 was carried out to determinethe riskiest events, that is, the risk events for which the costs of occurrence should

be included in the calculation of risk cost The client considered that the costs sociated with the three lowest-risk events were not material; therefore, neither wasthe risk The cost of the fourth lowest-risk event (Mine bond) was around $6 mil-lion (planning confidence level) and was considered by the client to be material.However, the client considered that the risk quotient of $0.05 million per year waslow (due to a low likelihood of occurrence) and that the risk was probably ac-ceptable It was decided that the occurrence cost of all risk events with a calculatedrisk quotient of greater than $0.05 million would be included in the risk cost.Those events excluded from contributing to risk cost are shaded in Table 11.2.Figure 11.1 shows a bar chart of the initial estimated risk cost The estimatedrisk cost at the planning level of confidence for contingent environmental and reg-ulatory events was $115 million NPV The client decided that the risk cost was

as-Risk Analysis / 169

Table 11.2 Modeling Results

Abbreviated Cost ($ × million) Cumulative Event Risk Optimistic Planning Pessimistic percent of Name Quotient (CL 50%) (CL 75%) (CL 95%) risk

clearly material and that the estimated risk component of the project was too high.Therefore, that the client required the risk cost be reduced, if possible.

The client also evaluated the inherent uncertainty contained within the estimate

of risk cost Uncertainty of cost is independent of uncertainty of occurrence In thiscase, the uncertainty of cost was defined as the cost range between the optimisticand pessimistic cost estimates The client considered that the cost range ($75 mil-lion) between the optimistic estimate of risk cost ($90 million) and the pessimisticestimate ($165 million) was very large, particularly as a proportion of the esti-mated planning cost, and should also be reduced if possible

Initial Risk and Exposure Profile Figure 11.2 shows a combined risk and

expo-sure profile, ranked from highest- to lowest-risk quotient The line graph showingthe risk quotient for each risk event indicates that the four most risky events (gen-eration performance standards, global warming, mercury emissions, and waste-water discharge) posed the greatest risk These four risk events accounted for over

99 percent of the total risk (as shown in Table 11.2) The bar graphs of the rence costs of the risk events clearly show that the three highest-risk events eachpresented substantial exposure to financial liability

occur-The third ranked risk event (mercury emissions) presented by far the greatestexposure The planning cost estimate of the mercury emission risk event was $60

Figure 11.1 Estimated risk cost of power station acquisition: Run 1 results.

0 20 40 60 80 100 120 140 160 180

Run 1 Risk Cost

Pessimistic Planning Optimistic

million NPV, which was almost three times the estimated planning cost of the twohigher-risk events The uncertainty of mercury emission cost was also muchgreater than for the other risk events.

The exposure profile of Figure 11.2 indicates that the estimated cost of each ofthe remaining three risk events that contribute to the risk cost are either marginallymaterial (arsenic concentrations in wastewater discharge) or clearly not material(coal pile liner and mining contamination of groundwater studies)

Interpretation of Figure 11.2 led the client to conclude that any strategy oped to reduce the risk cost of the project should concentrate on reducing the riskposed by the top four risk events

devel-B ID S TRUCTURE

The ultimate structure of the client’s bid depended on many factors While tion of environmental and regulatory risks was only one component that contributed

mitiga-to the final structure, risk reduction was a key input mitiga-to restructuring the final bid

Risk Reduction Options

Several options, which could be drawn together in various combinations to reducerisk, were available to the client The general options that could be adopted to ad-dress each substantial risk event were to:

Pessimistic Cost (CL 95%) Planning Cost (CL 75%) Optimistic Cost (CL 50%) Cut-off Risk Quotient Risk Quotient

• Virtually eliminate exposure to a risk event by budgeting a known expenditure

to carry out actions that will erase the source of the problem

• Commit a predetermined expenditure on activities to effectively prevent currence of the event

oc-• Carry out actions that reduce the risk, either by reducing the likelihood of theevent occurring or by reducing the magnitude of financial exposure to a riskevent, assuming it does occur

• Transfer risk through insurance or other financial instruments

• Accept the risk

Risk Reduction Plan

Risk reduction actions were developed for each risk event, in progression alongthe risk profile from the highest- to the lowest-risk event that contributed to therisk cost Mitigation actions were considered for the six highest-risk events at theplant The proposed actions that were finally selected are shown in Table 11.3.Computer modeling was progressively carried out in order to assess the finan-cial benefit of implementing the proposed risk reduction plan The risk model wasrun for the sixth time using the final information and assuming that the above riskreduction plan had been implemented

Benefits of Action Plan

Figure 11.3 shows a bar chart of the calculated risk cost for Run 6 assuming thatthe risk reduction measures would be carried out This figure shows that the esti-

Engineering solutions to carbon dioxide emissions from coal-fired power stations are unavailable The costs of purchasing carbon dioxide allowances would most likely be passed on to the consumer There remains some risk (likelihood reduced by 15 times) that the costs will not be passed on.

Install emission controls (scrubber) to reduce the volume of emissions and also reduce the likelihood of nonacceptable emissions by five times Install reverse osmosis system to remove arsenic from wastewater and reduce the likelihood of the risk event occurring by 25 times There is no cost-effective way to reduce the risk Propose to accept the risk and plan to install a coal pile liner in seven years.

Financial exposure is not material Plan to perform studies prior to action Accept the risk.

Table 11.3 Proposed Risk Mitigation Actions

Risk Event Proposed Action

mated risk cost of $40 million would be substantially less than the initial estimate

of $115 million, which represents a very substantial benefit to the project Thecosts of achieving the risk benefits were input separately to the financial model asproject base costs

Figure 11.3 also shows that the uncertainty associated with the risk cost wouldreduce markedly as a result of implementation of the proposed risk reduction ac-tions The range of uncertainty between the optimistic cost estimate of $36 millionand the pessimistic estimate of $46 million is approximately $10 million, com-pared with the $75 million range of the initial estimate of risk cost

Figure 11.4 shows the combined risk and exposure profile for Run 6 assumingthat the risk reduction measures have been implemented The risk events are pre-sented in the same order (decreasing initial risk) as Run 1 and shown in Figure11.2, so that direct comparison of the profiles can be made Comparison of Run 6with the initial risk and exposure profile of Run 1 shows that the overall risk andfinancial exposure have been substantially reduced

Exposure to the initially most risky event (generation performance standards)would be eliminated, and exposure to the third most risky event (mercury emis-sions) would be greatly reduced, from $60 million to $9 million The profilesshow that financial exposure to the global warming issue would remain the same,but that the risk would reduce by more than one order of magnitude

Bid Structure / 173

Figure 11.3 Estimated risk cost of power station acquisition after project restructuring to reduce risk: Run 6 results.

0 20

Run 6 Risk Cost

Pessimistic Planning Optimistic

Inputs to Financial Model

Figure 11.5 shows a bar chart of the schedule of risk costs that should be budgetedand directly input to the financial model The graph represents costs that could bereasonably expected to arise due to the occurrence of risk events The schedule ofrisk expenditure is in current dollars and represents the estimated costs at the plan-ning confidence level

The client requested the schedule of risk costs because the company needed toaccount for timing and magnitude of future expenditure The client was aware thatthe schedule was, at best, a reasonable estimate of how much (and when) expen-diture might be incurred due to the occurrence of risk events and that the scheduledoes not represent a prediction of the occurrence of future risk events

The figure indicates that around $10 million should be budgeted to cover theoccurrence of risk events during the first two years of operation of the plant Ap-proximately $14 million should be budgeted within the first four years Approxi-mately $15 million (current dollar value) should be budgeted each year after year

2009 to cater for purchase of carbon dioxide allowances due to the global ing issue

warm-Figure 11.4 Exposure profile for power station acquisition after project restructuring to reduce risk with events ranked by Run 1 risk quotient: Run 6 results.

0 20 40 60 80 100

120

0.001 0.010 0.100 1.000 10.000 100.000 1,000.000