New green mandates force portfolio planners to re-think their models.

Fortnightly Magazine - November 2008

In the past year, six states either have passed mandatory renewable portfolio standards (RPS) or added RPS through voluntary utility commitments, bringing the total number of states with RPS to 33—or two-thirds of all states. The U.S. Senate and House have introduced legislation to create a national RPS, although to date nothing has materialized. These initiatives, combined with a general evolution towards a “green” society, are having a significant impact on energy companies and their resource-planning processes.

Quantifying the impacts of RPS on utility integrated resource plans (IRP) sounds straight forward—just add more wind, solar, hydro, biomass, etc., to the plan and everything should be good to go. The reality is not quite so simple.

Green Planning

To understand the impacts of RPS on a utility’s IRP process, resource planners need to go back to the basics and make sure they understand the IRP process itself.

IRP came into vogue in the 1980s, went out of fashion in the late 1990s (market competition was supposed to take care of all of the industry’s capacity needs), and now is back again with several states requiring their utilities to file IRPs. Some states never abandoned the requirement.

However, between losing expertise from an aging workforce and the previous lack of a firm requirement to file an IRP, a number of utilities are out of practice and need to revisit their IRP processes to make sure they align with current best practices.

Traditional integrated resource planning focuses on developing a series of plans that allow the utility to meet its future energy requirements. Planners are seeking to balance risk and return by selecting the least-cost plan with the level of risk that fits the company’s risk profile—usually the least-cost, lowest-risk plan. Plan options typically include new generating-unit construction, long-term purchased-power agreements, demand response, and transmission expansion. Reliability is a key constraint and is expressed as a minimum reserve margin, a maximum un-served energy level, or a loss-of-load hours.

These days, virtually all IRPs involve environmental compliance planning. This adds additional constraints and complexities, such as increasing plan options to consider renewable energy, retrofits of existing units, and retirement or replacement. Hard emissions limits on minimum renewable energy or maximum emissions released may be imposed, and these must be examined as part of the IRP process. Evolving markets for purchasing credits for renewable energy and emission allowances also need to be considered.

Regulatory uncertainties around environmental issues—such as a recent D.C. Court of Appeals ruling that invalidated the EPA’s Clean Air Interstate Rule (CAIR)—along with ongoing changes in RPS rules and other regulatory and market factors need to be examined and factored into final IRP planning decisions.

Evaluating the numerous plans, their countless variations (i.e., what and when to build or retire, where to build, how much to build, etc.), and factoring in different price, demand, and regulatory scenarios involves modeling complex utility operations and optimizing new resource additions. The final output of the IRP is the optimal resource plan. Software modeling techniques can help resource planners with these optimal plans, but the different mathematical algorithms deployed to optimize the utility portfolio present certain strengths and weaknesses. Dynamic programming techniques, for example, show all of the feasible capacity plans to meet demand given constraints, rather than a single optimal plan, and sometimes the preferred plan is not actually the optimal plan. These models, however, generally have less detailed dispatch and transmission representations than do other solutions. On the other hand, linear programming or mixed integer programming techniques provide only a single optimal plan, but these models typically excel at providing a more realistic dispatch, transmission representation, and constraint treatment.

For an IRP that must include RPS, the planning process—and the computational models it uses—must incorporate the different renewable portfolio limits as a percentage of energy:

 Renewable limits are treated similar to operating reserve (as a constraint);
 Existing and potential generators must be characterized in terms of contributing renewable energy credits (REC) or production tax credits (PTC);
 Once the RPS requirement is set, only renewable generation can contribute towards the minimum renewable percentage requirement; and
 For areas that don’t meet the RPS requirements, renewable resources must be added.

So if the RPS functions just like any other portfolio constraint (such as operating reserve, for example) then it should be easy simply to incorporate these parameters into the traditional IRP process. Unfortunately, it’s not quite that simple. Each state has its own requirements (see Figure 1). There also are regional and national criteria that need to be met. Regulatory uncertainty is a factor, as not every state has a requirement, and even states that do have requirements may change them over time. For example, in 2001 the state of Nevada implemented a requirement of 15 percent renewable energy capacity by 2013 and then, in 2005, amended this to 20 percent by 2015. Resource planners also need to consider factors such as renewable energy credits (REC), green tags, white tags, and other evolving market mechanisms.

Multi-State RPS Modeling

While traditional IRP is portfolio-based, regional organizations such as independent system operators (ISO) and regional transmission organizations (RTO) also perform resource planning to monitor and assess the effects on transmission-expansion planning. For example, the Midwest ISO (MISO), PJM, and other ISO/RTOs regularly perform transmission-expansion assessments for their regions. Companies also might own assets across multiple states. For example, AEP, PacifiCorp, Entergy, Exelon, Duke, Progress, and others have multi-state service territories. Transmission upgrades increasingly are a resource option, and transmission availability or adequacy presents a major constraint for some utilities. In these cases, the portfolio must be expanded to include some level of regional impacts.

For example, in PacifiCorp’s 2004 IRP, the company’s plan modeled two zones (West and East) with an intertie between them. The analysis included thermal generating, wind, demand- side management, and contract options. However, the 2006 IRP included 23 zones and 34 interties. There were 62 existing resources with 96 existing contracts. PacifiCorp considered more than 80 potential resources, over 40 potential contract purchases, and seven transmission expansion options (see

Utilities that operate in multiple states or regions face added complexities when evaluating where to build resources—particularly renewable resources such as wind and solar, which require extensive site analysis.

A typical RPS requirement might specify 15 percent of a utility’s generated power must come from renewable resources. However, some RPS may require specific technologies. For example, Colorado, Nevada, Delaware, and Pennsylvania all state specific percentages of the total generating capacity that must be met by solar technologies. Other states may mandate a specific renewable capacity target (for example, Texas defines a specific capacity of 5,880 MW). Massachusetts, on the other hand, specifies the required renewable percentage as only applying to new resources, while Minnesota specifies a separate percentage that applies just to Xcel Energy. All of these variations increase the scope and complexity of the IRP.

In preparing a multi-state IRP, resource planners must be able to accurately capture and analyze complex variations such as:

 The total renewable percentage requirement by state or area;
 Secondary requirements for specific technologies (e.g., percent wind or solar);
 The ability to specify a particular required capacity (MW) of renewable resources (either in aggregate or by technology);
 Separate percentage requirements for a specific company;
 Setting generator contributions only on new generation; and
 Recognizing that only renewable generation contributes towards the minimum RPS percentage requirement; even though additional resources may be required to meet the overall IRP goals.

Resource planners also need to keep in mind the prospects for national RPS requirements. In December 2007, the Senate removed a federal RPS provision from the Renewable Fuels, Consumer Protection, and Energy Efficiency Act of 2007. While the Edison Electric Institute (EEI) and Southern utilities were against national RPS, other stakeholders have been pushing to include national RPS in proposed bills.

In 2009, the United States will have a new president. Whoever wins the election strongly will influence the development of any national RPS. Sen. John McCain voted against a 10 percent national RPS standard in 2005, and Sen. Barack Obama supports achieving a 25 percent level by 2025.

If a national RPS moves forward, it could require every provider in the nation to source between 15 and 20 percent of its power from renewables. A key uncertainty is whether a national RPS would contain any trading mechanisms to help parties achieve their targets.

The effect of a national RPS on the IRP process is that it would supersede some state requirements, but other states still would have more stringent requirements (e.g., Oregon and Illinois). IRP processes within those states must continue to account for area and regional-level requirements.

RPS Uncertainties

So how do some of these regulatory realities and uncertainties affect IRP processes? First, the timing of state and national RPS plans will influence when utilities build, or retire, generating resources. The makeup of the specific state or national plans—the percentage of renewable resources required, specific technology requirements, and even the definition of “renewable”—will influence what gets built or retired.

The locations of newly built capacity also will depend on specific RPS plans. Many feasible renewable technologies—wind, solar, geothermal—are location-dependent and have limited location options. This is why wind development is growing in Texas but not in Georgia.

Finally, resource planners need to balance renewable requirements with their demand forecasts and factor in retiring non-renewable resources.

Best practices in IRP require planners to model a variety of uncertainties. Traditionally these include fundamental uncertainties, such as commodity prices (e.g., fuel, purchased power, emissions), basis, transmission, demand (peak and energy), long-term economic variables (escalation), generator energy limits for hydro or renewable resources, and generator availability (forced outage rates). These typically are best handled using Monte Carlo techniques. In addition to the traditional fundamental stochastic analysis, most IRPs will incorporate stress testing to capture extreme moves in fundamentals—for example, $200/barrel crude oil prices or major weather influences.

Regulatory uncertainty does not lend itself to traditional stochastic techniques, so planners generally will use scenario analysis to study and quantify the various regulatory possibilities. Because the IRP process is labor intensive, resource planners only can evaluate a limited number of scenarios. The art therefore is in choosing the most relevant scenarios. First, the planner needs to develop a finite, well-defined, and manageable number of scenarios based on known and possible state and national regulations. For example:

 States A and B both at 15 percent in 2020;
 State A at 15 percent, State B at 25 percent in 2020;
 National standard at 15 percent implemented in 2024; or
 National standard at 20 percent implemented in 2025.

Such scenarios can be combined with sensitivity analysis to measure the change in output relative to change in input (difference between the required 15 percent or 20 percent). Knowing the cross-over points for different technologies (i.e., the points at which various technologies become or cease to be viable) based on the various possible regulations is valuable information. Of course, IRPs typically encompass between 20- and 30-year views, so everything is a bit uncertain.

A key uncertainty involves the evolution of markets for REC trading.

RECs—variously called green tags, renewable energy credits or certificates, tradable renewable certificates, and environmental attributes—represent a finite quantity of green or environmentally friendly energy generation or its equivalent. Buying such RECs can allow load-serving entities to meet their RPS obligations in more cost-effective ways than building or contracting for capacity in a non-optimal location.

Because electricity follows the path of least resistance rather than a direct, assignable route, purchasers of renewable (or green) energy don’t actually receive the renewable power, but they do expect it is being delivered to the grid. RECs provide a mechanism for tracking renewable energy and for decoupling it from energy production (that is, RECs can be purchased separately to offset energy purchases from non-renewable sources).

REC markets can take on a variety of structures. Market rules may allow, or prohibit, credits to be purchased separately from the energy provided. Entities may be allocated REC allowances where the base allowance level represents the REC requirements. Any RECs generated above the allowance base level represent selling excess credits into the market, while any RECs below the allowance base level represent purchasing the necessary credits.

Increasingly, resource planners are finding it necessary to include the analysis of various REC scenarios in the IRPs.

The number of RECs a particular supply resource generates varies widely depending on who is regulating the market. In some markets, supply resources have an REC ratio that is defined as the ratio of renewable energy (GWh) to total energy output (GWh). The ratio varies over time and state regulations may phase in allowable credits for some technologies under different regimes and timetables. The input ratio is applied to the energy output in order to compute the annual REC (GWh) for each resource. Other markets have specific calculations for different supply resources (see Figure 2).

Regulatory Scenarios

Incorporating RPS into an IRP requires a strong knowledge of the traditional IRP process and how RPS constraints will influence and impact the traditional process. It forces resource planners to consider how to frame and evaluate these new RPS constraints and impacts.

IRP processes and computational models need to be flexible in representing RPS constraints and characteristics for modeling regional systems, states with differing mandates, potential national-level constraints, REC markets, and more. Regional IRP development requires more careful consideration of the “where” question due to more limited siting options for renewable resources, as well as transmission requirements. Finally, including RPS in the IRP process requires that planners evaluate a variety of regulatory uncertainty scenarios using well-defined scenario analysis.