During the 1980s and early 1990s, integrated resource planning (IRP) was a required practice for many utilities. Then competitive wholesale markets, merchant generation, and restructuring initiatives led many utilities to abandon IRP.
While wholesale competition generally has been successful, the regulatory process changes it brought were less so. And utilities now are getting back into long-term resource planning studies to provide decision support for their “back to basics” business strategies.
IRP, originally designed as a regulatory means of ensuring that a utility’s expansion plans included a broad array of alternatives and were transparent,1 was regarded as redundant in late 1990s amidst restructuring developments in many parts of the country. Many utilities decided there was no need for regulatory mandated IRPs since market discipline in a competitive environment would ensure that the least-cost plan would be chosen. Under the circumstance, there would be no need to force market players to choose the least-cost option; they would be doing it on their own. In many cases, state regulators either explicitly agreed or acquiesced by not insisting that utilities file updated IRPs.2
California, which once engaged in elaborate resource planning exercises involving multiple utilities, disbanded its practice after passage of its ill-fated restructuring law in 1996. A number of other states followed California’s lead. In the process, California’s investor-owned utilities (IOUs) disbanded entire departments once engaged in demand forecasting and resource planning.
California’s electricity crisis of 2000-2001 proved them wrong. Not only did California’s ratepayers and utilities suffer from the meltdown of the dysfunctional market, run-away prices, and abuse of market power by generators, but the crisis spilled over to neighboring states. Utilities as far away as Oregon and Washington were affected adversely by capacity shortages and volatile wholesale prices that spread across the entire region. Consumers in Seattle, for example, ended up with higher retail rates while their local utilities faced financial difficulties emanating from California’s electricity crisis.
The ensuing public outcry in the Pacific Northwest led to a debate on how to avoid similar recurrences in the future, resulting in a number of important decisions by regulators in the region who ordered IOUs under their jurisdiction to conduct long-term resource planning exercises where supply-and-demand options would compete on an even playing field.3 The old-fashioned, forgotten IRP was reborn. A number of other regulatory commissions followed suit, mandating IRPs by load-serving entities (LSEs) or utilities in their jurisdiction.
California regulators reached a similar conclusion in 2003, ordering the IOUs in the state to undertake comprehensive IRPs. California also has mandated resource adequacy standards to ensure that utilities have secured adequate near-term resources. In fact, the popularity of IRP has spread beyond America’s borders, to countries as far away as South Africa.
Two popular concepts of the regulatory era, demand-side management (DSM) and IRP, also called least-cost planning (LCP), are back in vogue. DSM has been reinvigorated with the addition of demand-response (DR) programs, where consumers with flexible needs are financially rewarded to curtail usage during peak demand periods.
Although the regulatory requirements and the level of detail varies, a typical IRP process is focused on determining how best to meet future energy needs given the available resources. The “objective function” is no longer mere cost minimization, although costs are an important variable. The “best” portfolio of resources is one that meets the demand for power at minimum cost while providing a measure of supply security, risk minimization, resource diversity, and so on. Other criteria typically include environmental factors (e.g., greenhouse-gas emissions), resource adequacy, service reliability, and, increasingly, the inclusion of mandated renewables and possibly other considerations.
This is a complex set of parameters to consider and balance, with many conflicting objectives. Follow an aggressive strategy to add too many renewables, for example, and the emissions decrease while the costs may increase. Heavy reliance on coal may be the cheapest short-term option, but it could make the utility vulnerable should there be a significant increase in price of coal or a future requirement to reduce CO2 emissions. Having an extra cushion of capacity adds to supply reliability but also to costs. Investing in extra transmission capacity increases opportunities for importing lower-cost generation from distant generators, but is costly to build. A heavy emphasis on energy conservation and peak-load reduction reduces the need for supply-side resources, but at a cost. How far should one option be pursued versus another? What is the right mix? What is the optimal portfolio of resources?
Ultimately, the goal of the IRP is to help regulators, policymakers, consumers, environmental and renewable energy advocates, investors, LSEs, generators, and the private sector agree on what is the best combination of resources.
While there are no right or wrong ways to conduct an IRP study, the following steps normally would be followed:
1. Initial Process Validation
2. Demand Forecasting
3. Resource Characterization
4. Risk Characterization
5. Strategy Building—Creating Alternative Resource Portfolios
6. Strategy Evaluation
7. Strategy Insights and Implementation
Often the responsibility for these steps is decentralized among departments at a utility, or outsourced to consultants specializing in certain technologies or processes.4 Since many jurisdictions require that the process be participatory and transparent, stakeholders or oversight committees often are involved in the IRP study. In such cases, it is critical to engage and educate the stakeholders during the process so that there is “buy-in” and support for the results at the end of the day. Each step is briefly outlined below.
Step One: Initial Process Validation.
As in any business-planning exercise, having the correct and current input data and making use of the best available input assumptions is critical to the ultimate success of the study and the usefulness of the results. Use of inaccurate, out of date, or biased data will lead to suspect results. Step One should focus on defining the broader goals, objectives, and metrics to be used in the IRP, and can serve as a consensus building step within the organization.
Step Two: Demand Forecasting.
A critical step in any resource planning exercise is to determine, and agree on, a forecast of the load to be served. There is a large body of literature on the electric demand forecasting process.5
In sum, it is a process that uses both econometric and end-use (or engineering) models to determine energy consumption on a disaggregated or sectoral level. For example, separate forecasts typically are prepared for residential, commercial, and industrial customer groups, and aggregated to produce the overall growth forecast.6
Step Three: Resource Character-ization.
The next step in the IRP exercise is to define a comprehensive list of supply and demand-side resources and their physical and cost characteristics (see Table 1). This is a critical step since it defines the basic assumptions and the input that go into the remaining steps. If a major resource is not considered or its characteristics are ill-defined, the results of the IRP exercise are suspect.
Environmental considerations, renewable energy mandates, requirements about adequate reserve margins, cost or penalties for resource diversity, risks, and other variables typically are defined in this stage.
Step Four: Risk Characterization.
Everyone involved in the IRP process undoubtedly will be concerned with risks of various types. In the past, IRP risk analysis often has been limited to a few sensitivity case analyses. Newer analytical techniques allow for a much improved approach to assessing risk explicitly within the IRP framework.
Broadly speaking, risks can be categorized into three categories, stochastic, scenario, or paradigm risks (see Table 2).
Step Five: Strategy Building—Creating Alternative Resource Portfolios.
The strategy-building process typically begins with an examination of the so-called “risk responsive” resource plans using an optimization model to see what resource plan is “picked” for each of the scenarios from Step Four. These risk-responsive strategies are built to be the lowest-cost solution to each scenario and sensitivity case in the IRP. At this point, cost minimization is the only criteria—risks will be incorporated in Step Six.
An additional strategy building approach is to analyze structured strategies with pre-determined components that allow an examination of the relative costs and measure their cost/risk trade-off. Table 4 illustrates how six hypothetical strategies (“Status Quo Economic,” etc.) can be built from a number of key “strategy elements.”
Ways to construct alternative portfolios are limited only by one’s imagination and the availability of time and resources. At a minimum, the alternatives considered should include the obvious options advocated by particular interest groups (e.g., the renewable energy advocates, pro- or anti-nuclear advocates, etc.) and options that test the limit of reasonableness on variables such as fuel diversity and capital investment. In some cases, the regulators, consumer advocates, environmentalists, or other interveners may insist that certain portfolios (e.g., phase out of coal or nuclear) be considered in the analysis.
Environmental considerations play a significant role in many IRP studies. For example, the effect of alternative portfolios that meet certain emission goals (e.g., limits on CO2 emissions) may be considered. What would be the effect of phase-out of “dirty” coal, to be substituted by “clean” coal gasification technology? These types of policy questions are legitimate in IRP studies and provide policy makers with the answers they need to make important decisions.
Step Six: Strategy Evaluation.
This is the final analytical step, wherein the results of the previous efforts are loaded as input to a planning model and the numbers are “crunched” to produce results. In simple terms, a typical model would optimize meeting the forecasted load subject to all sorts of constraints (e.g., generation capacity, reliability, system stability, transmission congestion, etc.) while minimizing costs, emissions, risks, and/or other parameters.
Typically, software tools appropriate for such analysis are employed.7 Multiple iterations are necessary to determine how alternative resource portfolios fare on important attributes such as capital investment costs, net present value of revenue requirements, cost of power supplied to end-use customers, emissions, and other criteria. Many portfolios would be eliminated during this process as impractical, overly costly, or otherwise not feasible. The initial results tend to provide interesting insights about what options and portfolios are robust and practical, and those that are not.
Step Seven: Strategy Insights and Implementation.
The last step in typical resource planning exercise is to make sense out of the multitude of model runs, various strategies examined, various scenarios considered, risk and sensitivity analysis, and the myriad of results that are generated. This final step generally leads to:
• A broad awareness of the interdependencies of input assumptions, data, and results on the one hand and the intricacies and complexities of a large interconnected network;
• A better appreciation of which resource options make sense and which strategies are robust in the sense that they perform reasonably well under a number of scenarios; and
• The painful but necessary tradeoffs that must ultimately be made.
If done properly and transparently, with meaningful participation and input by competent stakeholders, the process can lead to strategies that have broad public support and backing. The process, assuming that it is well-documented, also provides an element of insurance against subsequent “Monday-morning quarterbacks” who would inevitably claim that “you should have known better” and “how could you not see x or y coming?”
The risk of a particular strategy can be portrayed as the “revenue requirement at risk” (also called “cost at risk”) (see Figure 1, p. 16). The relative risk of various strategies can be portrayed in a tradeoff diagram, as in Figure 2 (see p. 18).
On the issue of tradeoffs, participants typically gain a better appreciation of what must be given up to gain something else, as they realize that they cannot have it all. For example, advocates of renewable energy technologies, which reduce vulnerability to fuel price increases, learn the potential impact of higher levels of renewable technology investment in retail electricity prices. Those who are in favor of reducing the reliance on coal can see the cost impacts.
Regardless of the ultimate aim and audience for the IRP effort, many useful insights usually are gained, which may not be obvious a priori, nor easily quantified. A very common example of this type of result is that the utility management may have a bias in favor of (or opposed to) a particular technology or fuel type. But the IRP result will demonstrate that within a range of reasonable possible outcomes, the choice of such a technology may entail unexpected added costs or risks, compelling management to re-think its position.
With its origin shrouded in a regulatory past, and resurrected due to recent market malfunctions, IRP carries a lot of old baggage. It conjures images of bureaucratic, protracted paper exercises that drag on and produce little of value to the participants, other than meeting a regulatory mandate.
This is unfortunate. Perhaps a new name and fresh image are needed. Let’s call IRP “resource acquisition and risk mitigation,” or RARM. With the increased complexities of electricity markets, multiplicity of players with divergent interests and short-term focus, rising fuel prices, and concerns about resource adequacy and adequacy of investments in generation and transmission, it would be foolish to avoid this essential process due to the legacy associated with a name.
Another strong feature of the RARM process is that it need not necessarily focus on a single utility, state, region, or even a country. With interconnected networks spanning multiple states, provinces, and countries, RARM is suitable for analyzing regional and continental issues. Nor is the process limited to looking at traditional supply and demand resources. RARM can be applied to looking at regional and global policy issues including:
• Incentives or tax credits for renewable energy;
• Phase-out of nuclear or coal;
• Revitalization of nuclear or clean coal; and
• Various strategies to combat global climate change including CO2 caps and limits, carbon taxes, carbon trading, and sequestration.
Finally, the strength of this process, regardless of what it is called, is that it provides an analytical framework for enterprise-wide decision-making. Using the same tools and data employed for IRP/RARM, a utility can analyze a wide range of decisions for issues varying from environmental compliance strategies to fuel and power hedging. In short, RARM is not your father’s IRP.
2. In many states, regulatory agencies operate on very limited budgets and typically are thinly spread over many regulated industries, including natural gas, telephone service, and water utilities. Allowing IRP activities to lapse was considered prudent.
3. Among the first to do so were Oregon regulators, who required PacifiCorp to conduct an IRP.
4. In the U.S. context, most IRP studies are mandated by the regulatory agency, to be conducted by the local regulated utility(ies) under a prescribed set of rules and conditions in the public domain. But IRP exercises may be conducted by consultants under the direction of a regulatory or oversight agency. IRP-type studies may also be performed by a group of states or countries.
5. For example, see What’s Driving the Demand for Forecasting, by F. P. Sioshansi in Risk and Flexibility in Electricity, Risk Books, London UK, 2003.
6. On a broad level, forecasting can be done using a bottom-up or top-down approach. The former, typically done when the utility has detailed historical sectoral data, is more labor-intensive and can produce a highly accurate pattern of future load growth. The latter is often used when detailed historical sectoral data is not available or is of poor quality.
7. A discussion of the capabilities of various software tools commonly used in such a process goes beyond the scope of the present article. There are a number of firms offering products and tools suitable for IRP studies, including Global Energy Decisions.