Comprehensive Electricity Competition Act: Supporting Analysis

The POEMS analysis examines the economic and environmental impacts of a transition to retail competition. However, the analysis does not attempt to explicitly account for state actions that are already beginning the transition to competition or reflect the timing of future actions that states might take to implement competition consistent with the Administration’s proposed "flexible mandate" for retail competition. From an analytical perspective, it is difficult to isolate the economic and environmental effects of the Administration’s proposed legislation from the effects of state actions alone. Moreover, the Administration’s proposal will benefit consumers even in states where the transition to competitive markets is already underway by providing additional authority to help assure that potential gains from competition are actually realized.

The focus of the POEMS analysis is the impact of full retail competition nationally compared to a continuation of cost of service regulation that includes wholesale competition. The main results of the analysis are as follows:

The POEMS analysis was informed by an effort to identify potential cost savings from restructuring, a summary of which is provided as Appendix C. Using information from reports filed by investor-owned and public utilities, quantifiable potential cost reductions resulting from competition in operations and maintenance costs, administrative and general costs, more efficient use of the transmission and distribution system, and capital cost savings at existing facilities are estimated to exceed $20 billion annually. This estimate does not include: the savings that would result from a reduction in the need for new capacity due to more efficient pricing, the benefit to consumers of avoiding the costs of any future mistakes with respect to capacity planning, technology choice, or project management that have in the past raised the cost of power to consumers; or the greater economic value to consumers of new products and services that will be created in a competitive environment.

The remainder of this document summarizes the POEMS analysis. Section 2 presents the results. Section 3 presents key assumptions used in the analysis and the rationale for the scenario formulation. Section 4 discusses future direction of the modeling and analytic efforts. Appendix A provides supplementary figures and graphs. Appendix B provides a summary table of results. Appendix C summarizes an analysis of costs at existing plants that was used to derive projected efficiency improvements for the Competitive Scenario. Appendix D provides documentation for the POEMS.

The introduction of retail competition is projected to lead to lower electricity rates for customers in all regions of the country. Figure 1 illustrates the projected average electricity prices in 2010. These prices are the average across all customer classes and include generation, transmission, and distribution costs, and stranded cost recovery.

Both the Reference and Competitive scenarios include projections of declining electricity rates. In 1995, the national average delivered price was 6.9 cents per kilowatt-hour (kWh). With continued cost of service regulation, the price in 2010 is projected to be 5.9 cents per kWh. The decline is the result of capital cost depreciation of existing high cost plants, as well as the continued entrance of more efficient capacity. With retail competition, greater efficiencies and marginal cost pricing lead to larger price reductions, and the national average price is projected to be about 12 percent lower, or 5.2 cents per kWh in 2010.

As shown in Figure 1, there is a wide range of projected prices with continued cost of service regulation, ranging from 4.3 cents/kWh in the Pacific Northwest to 8.6 cents/kWh in New York. With competition, the variation in price is likely to be smaller, ranging from 4.2 cents/kWh in the Pacific Northwest to 7.5 cents/kWh in New York¹ . In general, the regions projected to have the highest prices under cost of service regulation are those that are likely to see the largest decrease in prices under competition. The remaining regional variation is due to differences in fuel prices, operating costs, transmission costs, transmission constraints, distribution costs, and stranded cost recovery. Differences in transmission and distribution costs are the most important factors.

Consumer savings are projected for all classes of customers: residential, commercial, and industrial. A comparison of 2010 national cost of service and competitive rates by customer class is shown in Figure 2. Residential customers are projected to see the largest price decreases in 2010 with competition. In part, this is because historical capacity costs were allocated to customer classes based on their contribution to peak demand. Residential customer demand tends to have more variation by time of day and season than industrial and commercial; therefore, it receives a relatively greater share of the costs under peak demand allocation than if costs were allocated based on sales. Although the energy costs associated with peak demands are generally high, the capital costs for peaking turbines is relatively low compared to baseload units. In the competitive market, higher marginal costs at peak periods will lead to higher average generation prices for residential customers than for customer classes with flatter load profiles, but with less of an average premium than under the previous capital cost allocation.

The competitive prices include the recovery of a projected $85 billion in stranded costs for existing and productive generating assets. When stranded costs for existing and productive generating assets are recovered over a 10-year period as described below, the national average additional charge to the average electricity price is 0.2 cents per kWh in 2010. On a regional average basis, stranded cost recovery factors are projected to range from 0.03 to 0.5 cents per kWh. Within regions, stranded cost recovery factors will vary across individual utilities due to differences in generating asset portfolios and price differences across power control areas.

The Competitive Scenario also provides for recovery of regulatory assets and decommissioning costs. The pace of recovery in these categories for both scenarios reflects recent state-level practices, and is assumed to be similar to projected recovery in the Reference Scenario. Provision for recovery of regulatory assets and decommissioning costs adds 0.1 cents per kWh to the estimated national average price of electricity in 2010.

The underlying electricity demand forecast, which is taken from the EIA Annual Energy Outlook, projects electricity growth of 1.5 percent per year from 1995 to 2010. There are several elements of the Act that will affect this demand. Lower electricity prices resulting from competition are likely to stimulate additional demand for electricity. This is dampened to some extent by lower projected natural gas prices that result from overall lower gas demand. The proposed provision for a Public Benefits Fund and the expectation that competition will spur efforts to package energy efficiency (and other energy service products) with power sales, also are projected to reduce demands. The net result is slightly higher electricity demand in the near term and lower demand as the Act comes into full effect. By 2010, projected electricity demand is 2.4 percent lower with the Act than in the Reference Scenario.

Most of the generation capacity used in 2010 will be capacity that exists today. For example, in the Reference Scenario in 2010 only 20 percent of the projected total capacity are new additions. Currently, coal plants account for 41 percent of all capacity. Other major types of capacity are: other fossil (oil and gas) steam at 19 percent, nuclear at 14 percent, hydroelectric at 13 percent, and combustion turbines at 8 percent. Gas-fired combined cycle plants and all other renewables have relatively small shares, at 4 and 1 percent respectively. In the future, this mix is projected to shift more towards gas technologies and away from coal and nuclear plants.

In both the Reference and Competitive scenarios, the greatest share of new construction is projected to be gas-fired combined cycle plants. This result reflects the combined effect of high efficiencies, short construction periods, modularity, and modest projected increases in natural gas prices. In the Reference Scenario, they are projected to represent 59 percent of the cumulative new plants from 1995 to 2010, with gas-fired combustion turbines that serve peaking requirements capturing an additional 27 percent of total capacity additions.

The dominant change in the mix of future generating capacity in the Competitive Scenario is the increase in the share of renewable capacity (see Figure 3) that results from the Renewable Portfolio Standard included in the Administration's proposal and consumers’ interest in green power. Less capacity of other types of plants are needed as a result. However, much of additional renewable capacity is intermittent. For example, generation from wind power is dependent on when the wind is blowing and therefore operates less than a plant that can be run 24 hours a day. Because of this, much of the renewable capacity added receives only a partial credit towards meeting capacity requirements. This results in more installed capacity under the Administration’s proposal than in the Reference Scenario, even though demand is slightly lower.

The differences in generation by fuel type across the Reference and Competitive scenarios (see Figure 4) do not directly track the differences in capacity additions outlined above. Because competition will provide strong incentives to run low-cost plants more efficiently and to shorten scheduled outage periods, existing coal and nuclear plants are run more often in the Competitive Scenario. As a result, even though there is slightly less coal capacity, coal generation in the Competitive Scenario is actually slightly above the Reference Scenario level. The change in non-hydroelectric renewable generation under the Administration’s proposal is significant, as shown in Figure 5. Because of the proposed Renewable Portfolio Standard and consumers’ interest in green power, non-hydro renewables generation in 2010 is projected to be 6.3 percent of total generation in the Competitive Scenario, approximately twice its share in the Reference Scenario. All of other types of generation are below the Reference Scenario levels in the Competitive Scenario.

Gas Fired Combined Cycle

Coal

Gas-Fired Combustion Turbine

Other Fossil

Nuclear

Renewable

Carbon emissions from electricity generation are projected to increase by 166 million metric tons of carbon equivalent (MMTCE) between 1995 and 2010 in the Reference Scenario. In the Competitive Scenario, carbon emissions are projected to be 39 MMTCE lower than the Reference Scenario in 2010.

Recognizing the inherent uncertainty of future market developments, the Administration estimates that its proposal will lead to emissions reduction of between 25 and 40 MMTCE in 2010. This approach parallels the conservative approach used in evaluating economic benefits, which recognizes that the impacts of the Administration’s proposed legislation and those of competition itself are not easily separated. Emissions reductions in this range are likely to be achieved even if most of the uncertainties discussed below are ultimately resolved in a direction that tends to increase emissions beyond the modeled level.

As shown in Figure 6, carbon emissions may rise slightly in the early years of competition compared to the Reference Scenario with continued regulation. However, as existing coal plants become fully utilized, the Renewable Portfolio Standard requirements increase, and additional energy efficiency investments take place, emissions grow more slowly. Another factor leading to lower emissions is the improved efficiency in power generation, as generators faced with competition have a direct financial incentive to reduce their input costs.

Improvement in heat rates and capacity availability. There is considerable uncertainty regarding the extent of heat rate or capacity availability improvements likely to occur at existing plants in a competitive scenario. Heat rate improvements and capacity availability improvements at fossil-fired plants work in opposing directions, with the former tending to reduce carbon emissions and the latter tending to increase them. Should either of these improvements in a competitive environment diverge from the estimates used in the Competitive Scenario, carbon emissions would be directly affected.

Other state-level and private decisions under competition. Decisions that could result in lower-than- modeled emissions include higher energy-efficiency spending due to either competition among retail electricity suppliers or the operation of the Public Benefits Fund (the results presented here assume an incremental $2 billion annually in efficiency spending due to the public benefit fund), more consumer interest in green power, fewer nuclear retirements due to competition, or lower than assumed availabilities for coal-fired power plants. Forces that could result in higher-than- modeled emissions include additional nuclear retirements attributable to competition, greater responsiveness of demand to price reductions than provided for in the demand modules of the National Energy Modeling System, or a binding Renewable Portfolio Standard cost cap that is not offset by increased green power demand.

The Administration’s proposal includes provisions that clarify Environmental Protection Agency (EPA) authority to require a cost-effective interstate trading system for nitrogen oxide (NOx) pollutant reductions addressing the regional transport contributions needed to attain and maintain the ozone ambient air quality standard. However, no change is proposed to existing EPA authority to determine geographic coverage or level of reductions required in addressing regional transport contributions.

Consistent with these provisions, the projected level of nitrogen oxide emissions will primarily be determined by past, pending and future actions taken by EPA under its existing regulatory authorities. For example, the emissions level in 2000 and beyond are significantly below the 1995 level in both the Reference and Competitive scenarios due to the Phase 2 Clean Air Act NOx standards, which were included in both cases. Starting from existing emissions rates as modified by the Phase 2 standards in 2000, both annual and summer season NOx emissions in the Competitive Scenario are projected to be roughly 4 percent below projected levels in the Reference Scenario in 2010. The projected reduction in NOx due to competition generally results from the same set of factors that provides the reduction in carbon dioxide emissions discussed in the previous section.

These POEMS model results do not reflect the Environmental Protection Agency's pending proposal to establish absolute caps on ozone season NOx emissions for 22 States and the District of Colombia or the reductions currently being undertaken within the eastern States comprising the Ozone Transport Commission. As noted above, such actions under existing regulatory authority will be the primary determinant of future emissions levels. For example, in regions and seasons where regulatory actions take the form of an absolute cap on the level of emissions, the cap itself would determine the level of NOx emissions independent of the price of electricity or the structure of electricity markets for those regions and seasons in which it was applicable.

This observation may be applied directly in the case of sulfur dioxide emissions, since an annual nationwide cap on sulfur dioxide emissions from the electric utility sector has already been established pursuant to the 1990 Clean Air Act Amendments. For this reason, emissions of this pollutant are projected to be the same in both scenarios.

In order to measure the impacts of the retail competition, a baseline must first be established. As with any forecasting exercise, this Reference Scenario is meant to be a reasonable expectation of possible future events and not a statement of what will happen. For this analysis, the Reference Scenario assumes a continuation of existing forms of utility regulation and cost-of-service pricing. The Competitive Scenario represents implementation of retail competition as envisioned in the CECA.

The Reference Scenario assumes wholesale competition as achieved through open transmission access under FERC Order 888. All new capacity is assumed to be built by exempt wholesale generators and not included in the rate base of utilities. In addition, transmission fees are pancaked. In other words if power is wheeled across two transmission systems, each will charge a separate fee for providing the transmission service.

The Reference and Competitive scenarios share the same underlying macroeconomic and energy sector assumptions, which are taken from EIA's Annual Energy Outlook 1997. However, the electricity demand and fuel price projections are not identical between scenarios, because of the feedback with electricity prices, fuel demands of the electric sector, and other sectors fuel prices due to the transition to competition and specific provisions of the CECA.

The Administration’s proposal was represented in the POEMS model through a range of assumptions that represent a vision of how a full retail competition market might emerge as a result of the Administration's proposed policies. Table 1 outlines the major set of assumptions and provides a comparison with the Reference Scenario. Each of these assumptions is discussed below in the context of the Act.

The generation price in the Competitive Scenario is composed of the marginal generation cost, ancillary charges, a renewable portfolio standard (RPS) premium (if applicable), and stranded cost recovery charges. Marginal generation cost is established in each power control area (PCA) based on the bid price of the last unit running in each time/season period. The last unit could be native to the PCA or determined through trade with other PCAs. In accord with the standard economic model of perfect competition, the bid price for each unit is assumed to be its marginal cost -- the sum of fuel cost and the variable portion of operating and maintenance (O&M) costs.

As stated in the Comprehensive Electricity Competition Plan (announced March 25, 1998), the Administration "endorses the principle that utilities should be able to recover prudently incurred, legitimate and verifiable retail stranded costs that cannot be reasonably mitigated". The Competitive Scenario assumes that stranded costs associated with productive generating assets are recovered over a ten-year period following the introduction of competition. Recovery of regulatory assets and decommissioning costs in the Competitive Scenario is assumed to be similar to that in the Reference Scenario, with the pace of recovery in these categories for both scenarios reflecting recent state-level practices.

Retirement of plants is economically driven. The economic retirement decision for all generating plants is based on both short-term and long-term criteria. The short-term requirement is that plants can cover their "going-forward" costs, which includes all fixed and variable O&M costs as well as recovery of the annualized value of new capital additions. If a plant cannot cover these costs, it becomes a candidate for early retirement. The second consideration is the cost of building new generating capacity. In the capacity-planning module, all existing units must pay their going-forward costs if the capacity is to be used over the full planning horizon. Thus the planning module has the opportunity to "decide" to economically retire any or all of the existing units and instead build new capacity. If the planning module does decide to economically retire a unit and this same unit did not cover its variable costs in the last forecast year, it is retired. A plant must be uneconomic in both the short-term and long-term to be retired.

As the electric power industry is transformed into a more competitive, market-based industry, the historical level of costs are expected to be reduced due to the pressures of competition. In the generation segment of the industry, costs per unit of output will decrease due to the changing mix of capacity, i.e., more expensive generating units are replaced by new, more efficient generating technology (typically natural gas fueled). In the Competitive Scenario, competitive pressures are expected to lower costs at existing generating facilities as well, as they begin to compete with other existing and new facilities. Competitive pressures are assumed to also spill over into the regulated segment of the industry. Transmission productivity improves by 0.75 percent per year and distribution productivity improves by 1.5 percent per year to 2010 due to the introduction of performance incentives to improve productivity in these functions.

Under competition, electricity generators will not be guaranteed a fixed rate of return on their investments. As a result, plant owners will demand a greater expected return to compensate for the risk associated with their revenues. They will also need to finance their investments with less reliance on debt and more on equity. The weighted cost of capital is 12.8 percent in the Reference Scenario and 14 percent in the Competitive Scenario.

Historically, utilities were not rewarded for reducing their fuel costs. In fact, in many states, fuel costs were directly passed through to consumer bills. As long as a utility was acting "prudently," regulators would provide little pressure to reduce these costs. This is likely to explain in part why there is currently a very wide range of heat rates in power plants of the same type, size, and age. With intense competitive pressures, generator owners are likely to make cost-effective improvements and change their operations to improve the efficiencies of these existing plants. Any improvements will lead either directly to increased profits or allow the plant to continue to operate where it might otherwise be priced out of the market. Based on an analysis of existing heat rates, the Competitive Scenario assumes that existing plants will make significant strides towards achieving heat rates closer to those of the top 25 percent of comparable plants.

Competition will also give generators an incentive to maximize the availability of their facilities because they will only receive revenue when they are operating. In the Reference Scenario, fossil-fuel- fired steam units are assumed to have availabilities of 85 percent. Nuclear availabilities vary by region and are based on EIA’s AEO98 assumptions. In the Competitive Scenario, the steam units are assumed to have 90 percent availabilities, which is equivalent to a one-third reduction in the outage times. A one-third improvement was applied to the nuclear units as well, with a maximum of 90 percent availability. For most regions, the resulting nuclear availabilities fall below the 90 percent cap.

The Federal Energy Regulatory Commission, in its Order No. 888, required that "…seller(s) (and each of its affiliates) must not have, or must have mitigated, market power in generation and transmission and not control other barriers to entry." ² For transmission owning utilities, this meant that the utility must have on file with the Commission an open access tariff for the provision of comparable service. The Competitive Scenario assumes that all transmission owners have an open access tariff over which retail and wholesale sales can occur. The Competitive Scenario, as in the Reference Scenario, assumes that exempt wholesale generators (EWGs) will provide all new generation capacity. The Competitive Scenario assumes that all consumers (i.e., residential, commercial and industrial) will have equal access to the power exchanges.³ To assure that consumers of all types have the necessary information to make these informed decisions, the Administration’s proposal provides for a uniform label information on price, terms and conditions of service.

Transmission fees were computed using a formula similar to the pro forma tariff described in Order No. 888. In the Reference Scenario, the transmission fees were assumed to be pancaked⁴ . In the Competitive Scenario, the assumption is that regional transmission groups (RTG), tied together by an independent system operator (ISO) would operate the transmission grid(s). The transmission fees in the Competitive Scenario were therefore assumed to be the same for moving power across the entire RTG region (i.e., a postage stamp rate). Since the bulk of the costs associated with the transmission system are allocated to the native load customers, the transmission fees used were discounted. In the Reference Scenario the discount was 20 percent and in the Competitive Scenario, the discount was 50 percent.

A Renewable Portfolio Standard (RPS) was included as a national standard with potential trading of credits. This means that renewable generation can be constructed wherever it is most cost-effective, rather than requiring it to be spread evenly across the nation. The standard was expressed as a percent of sales that must be met with renewables and was assumed to increase gradually over the 2001 to 2010 period. In 2010, the RPS was set at 5.5 percent. All non-hydro renewable generation qualifies to meet the standard, including industrial cogeneration. Because of the ability to trade credits, renewables will command the same price premium nationally, equivalent to the marginal cost. The premium is paid by all customer classes on a cents per kilowatt-hour basis.

The Competitive Scenario assumes that in addition to the RPS standard, 5 percent of residential customers nationwide would be willing to pay for additional Green Power, which is comprised of 50 percent new non-hydro renewables, above what the RPS requires.⁵ The labeling provisions of the CECA will provide consumers with the information that they need to be able to choose their generation suppliers based on price and the environmental factors important to them. Pilot programs in various states, as well as activity in California, support the assumption that a segment of consumers will value these qualities.

The Administration’s proposal calls for the creation of a $3 billion Public Benefits Fund to be matched by states and used for energy efficiency programs, technology research projects, low-income assistance, and consumer education. In addition, with electricity suppliers competing to meet the needs of customers, they are likely to offer a full range of energy services in order to be competitive. Energy efficiency improvements are already being offered in some of the nascent retail competition areas. Together these efficiency improvements are assumed to reduce electricity demand by 2 percent in 2010. The modeled scenario was developed in the context of a $2 billion increment to annual baseline energy-efficiency expenditures over the 2000 to 2010 period. Additional expenditures in energy efficiency would reduce electricity demand further.

This paper is intended to inform discussions of restructuring policy by comparing a generic cost-of- service scenario to a retail competition scenario that is consistent with the main elements of the Administration’s Comprehensive Electricity Competition Act. Further analyses can provide additional insights as these discussions unfold. While some future analyses will be driven by the specific elements of alternative proposals, some issues that have already been identified as potential subjects of future analysis are briefly summarized below:

Transmission constraints: Transmission plays an important role in the modeling analysis of competition, since only in the presence of transmission constraints will prices in adjacent competitive markets differ by more than the transmission fee plus line losses. Electricity flows on the transmission system do not generally follow the contract path, and the available capacity between two market areas may be influenced by power flows throughout the system. For this reason, it is important to verify the POEMS representation of transmission constraints using tools that can follow physical flows. Work in this area is underway in cooperation with the North American Electric Reliability Council and other transmission experts.

The timing and scope of competition: Notwithstanding the difficulty of separating the projected effects of the Administration’s proposal from those attributable to other steps towards competition, sensitivity scenarios addressing this issue could provide useful insight into the likely impact of alternative transition paths.

Nuclear retirements: For both emissions and economic reasons, there is considerable interest in the implications of competition for continued utilization of existing nuclear plants. A further assessment in this area would review the characterization of retirements in the Reference Scenario and the criteria used to drive retirements in the Competitive Scenario. The role of risk relative to expected profitability in driving retirement decisions should plant owners be called upon to make an early declaration regarding their intended retirement decisions also merits further attention. Finally, the implications of a scenario in which competition results in a concentration of nuclear operations in the hands of firms with a proven record of efficient, safe, and economical operation of these facilities should also be considered.

Alternative Allocations of Stranded Costs: In the same manner that rates under cost-of-service regulation are affected by the allocation of new capacity costs across customer classes, prices in the transition to competition will depend on how the costs of uneconomic generating assets are allocated. Sensitivity analyses could scope out the range of possible outcomes.