Valuing risk reduction for renewables and DSM.
Resource planners are faced with complex choices for developing cost-effective and robust energy supply portfolios. These choices are complicated by uncertainties inherent in future fuel and emissions costs. In the summer of 2008, retail energy providers with supply primarily from wind generation had a substantial cost advantage over gas-fired generation. In the summer of 2009, though, gas prices plummeted in the wake of the recession. Reversing the previous trend, this shift causes wind generation to appear more costly relative to gas-fired generation.
Dramatic swings in market prices are inevitable in today’s competitive energy markets. Recognizing that consumers of energy are risk averse, there exists a real and measurable value to resource choices that reduce the uncertainty in energy supply costs. By applying advanced analytics that achieve higher levels of prudent portfolio management, the value of the risk avoided through integration of renewable energy sources and demand-side management (DSM) can be directly quantified.
DSM and renewable energy sources have unrealized and unmeasured value to reduce energy supply risk. The inherent absence of fuel requirements and emissions of these resources removes exposure to market prices and provides a latent, yet quantifiable, value of avoided market risks to ratepayers. This risk reduction value (RRV) can be measured using techniques similar to those used by insurance companies to develop actuarially derived premiums. Quantifying the RRV for DSM and renewable energy, while integrating this value into the resource planning process, appropriately credits the value derived from reduced market risk exposure. For DSM programs, RRV provides a directly measureable net benefit, along with more easily measured avoided fuel, capacity, emissions, transmission and distribution expenditures.
Numerous factors give rise to uncertainty in power supply costs. These risks primarily manifest themselves as volatility in market prices for power, fuel and emissions allowances. Further, market risks are magnified by volumetric uncertainty. Various unforeseeable, stochastic factors, such as weather and unit outages, as well as long-term factors such as economic growth and transmission congestion, contribute to volumetric uncertainty.
DSM measures and renewable energy technologies by their very nature tend to mitigate market risks and volumetric uncertainties. Once a DSM measure is in place, market exposure decreases proportional to the respective avoided energy. Renewable energy mitigates risk by producing power without consuming fuels that typically experience high degrees of market price volatility.
In order to remove the inconsistencies and potential biases from the resource selection process, it’s best to apply a portfolio planning methodology that explicitly measures and accounts for the RRV for DSM and renewable energy. Direct incorporation of the RRV into the resource valuation process levels the playing field with respect to traditional supply resources. Therefore, resource planners should:
• Make use of leading analytic practices to determine the latent RRV inherent in DSM and renewable energy alternatives; and
• Include RRV as a measureable program benefit for DSM and renewables.
In making these decisions, it’s important to note that for DSM, RRV becomes part of the avoided-cost calculation alongside fuel, emissions, generation, and transmission. Utilities have an additional incentive to consider RRV as part of their net-benefit calculation because they share directly in the net benefits accrued. Also, for renewable energy, RRV factors directly into the portfolio analysis to assess these resources versus traditional fossil generation options.
RRV is measured from the expected distribution of energy supply costs. RRV is determined by quantifying the value of the risk premium (RP) for each supply portfolio in terms of the areas from the mean to the upper tail of the cost distribution. In Figure 1, RP is shown as the orange dotted area corresponding to the upper half of the cost distribution. Integrating the upper half of the cost distribution determines the value of risk similar to the functions employed to value options based on the uncertainty in commodity prices or the calculation of the value of insurance premiums.
The RRV corresponds to the difference in RP of the portfolio with DSM measures and renewable energy (i.e., the green portfolio) from the base portfolio (see Figure 2). That illustrates energy supply cost distributions for two respective energy supply portfolios: 1) a base case without DSM and renewable energy (shown in orange in Figure 2); and 2) a second case including DSM and renewable energy (shown in blue). These two distributions of costs correspond to an integrated electric utility with a peak demand of approximately 2,000 MW. The expected (mean) annual cost of supply for the base portfolio is $605 million, whereas the green portfolio costs increase by $15 million in place of new traditional generation of the base portfolio. Although the expected cost of supply for the green portfolio exceeds that of the base portfolio, the potential for cost increases are substantially mitigated by the green portfolio’s reduced exposure to market price volatility. The 95th percentile of cost for the green portfolio is $830 million versus the base portfolio cost of $900 million.
The RP for the two energy portfolios are $100 million and $75 million for the base case and green portfolios, respectively. The RRV of the green portfolio corresponds to the difference in the risk premium of $25 million, which approximately corresponds to the difference between the orange hatched area and the blue hatched area to the right of the mean in Figure 2. By directly including the RRV as a credit to the green portfolio, the initial preference for the lower expected cost-based portfolio is reversed, with the green portfolio providing a net $10 million benefit of value (i.e., derived by subtracting the $15 million difference in expected costs from the $25 million benefit from the reduced RP).
Calculation of a portfolio’s RP and RRV moves beyond traditional risk metrics to provide a discrete risk value that can be used to directly modify calculated portfolio costs. The incremental savings represented by the RRV can be credited against portfolio costs. The analytic value of the RRV over traditional risk metrics follows from its direct and additive monetization of risk under well-known practices applied by insurance companies and option traders. The benefit of being able to directly subtract the RRV from a portfolio’s costs simplifies the resource-selection process by making risk an explicit component of total resource costs.
The inclusion of market and volumetric uncertainty can yield tremendous insight into portfolio risks when done well. However, proper valuation requires an underlying modeling framework that meets rigorous criteria for simulation of future market states. The current leading practices in energy resource modeling apply an integrated framework that incorporates observed market dynamics with system fundamentals to capture critical covariate relationships:
• Implied heat rate distribution;
• Weather > load > DSM > price relationships;
• Weather > wind and solar generation > price relationships; and
• Correlations between commodity prices.
Model validation requires meeting rigorous benchmarks that demonstrate the ability to reflect previous patterns of market and volumetric uncertainty and evolve these relationships forward based on current expectations. The integrated simulation framework serves as the analytical foundation to rendering portfolio valuations.
Resource valuation then follows an hourly time-step process that reflects the coincident value of market prices and resource characteristics. While simplification into larger time steps can be computationally attractive, critical attributes of resource flexibility, such as interruptible load or coincidence of static generation with market prices, become undervalued with simplified analytical approaches. Getting resource valuation right requires adhering to the higher levels of prudent portfolio management by utilizing all available information from the application of leading modeling techniques.
RRV should thus play a critical role in valuing the risk reduction effectiveness of alternative DSM programs. Including RRV supports a more balanced assessment of cost effectiveness of DSM programs by accounting for the direct value attributed to the reduced risk. The RRV for specific DSM measures can be expressed as a percentage of each program’s total cost (see Figure 3). RRV for DSM ranges broadly, from approximately 5 to 35 percent, depending on the utility supply portfolio.
Alternatively, programs can be valued in terms of net benefit per megawatt hour (MWh) of energy conserved. The RRV becomes a direct credit to the avoided cost function as follows (see Figure 4): Net Benefits = Avoided Costs ƒ (generation, fuel, T&D, emissions, and risk) – Program Costs.
The eventual goal of resource analysis is to determine the cost not of a single resource, but the cost of a portfolio of resources. Incremental analysis, in which individual resources are added to a portfolio in order to ascertain their respective economic impacts on the whole, also is a common approach to resource analysis. Figure 5 incorporates DSM and renewable energy RRV into incremental analysis. In this case, four scenarios are presented, representing the various combinations of base case, DSM, and renewable energy resources. The height of each bar represents the nominal total expected cost of each scenario; the orange portion of each column represents the cost of each scenario, net of the RRV created by DSM and renewable resources. The total and net costs can result in completely different interpretations of the value of each respective resource plan.
Green resource options have substantial latent value over traditional supply resources to reduce risk. Converting risk to an economic measure that can be directly included into the evaluation of future resource options removes a common bias of traditional cost analysis. Further, by quantifying the value of risk reduction as part of the supply costs, resource valuation adheres to the higher standards of prudency by utilizing all available information along with the application of current modeling techniques. The analytic foundation necessary to adequately calculate the value of risk requires detailed modeling of market dynamics, load, and resource characteristics to reflect the full economic attributes of each resource. By getting the details right, resource evaluation can remove systemic biases that preclude a balanced valuation of energy resource options.