Executive Summary

On April 15, 1999, Secretary of Energy Bill Richardson forwarded the Administration's proposed Comprehensive Electricity Competition Act (CECA) to the U.S. Congress. The purpose of the CECA is to provide a framework for the restructuring of the U.S. electricity industry. After the CECA was forwarded to Congress, the U.S. Department of Energy (DOE), Office of Policy, released its Supporting Analysis for the Comprehensive Electricity Competition Act. That report analyzed the impact of CECA using the Policy Office Electricity Modeling System (POEMS), a computer model of the U.S. energy system.⁽¹⁾ The Secretary then asked the Energy Information Administration (EIA) to replicate the analysis reported in the DOE's Supporting Analysis, using the National Energy Modeling System (NEMS), and to compare the results with those from POEMS.

This report provides a discussion of the differences between POEMS and NEMS. In the comparative analysis described here, the modeling assumptions documented in DOE's Supporting Analysis were used with NEMS to prepare a noncompetitive case (the CECA Reference case) and a competitive case (the CECA Competitive case). The two NEMS cases were intended to parallel the cases prepared by the Office of Policy.

Comparison of Methodology

In many ways NEMS and POEMS are similar tools. POEMS has adopted all the NEMS modules with the exception of electricity. Even in the case of electricity, some of the models' components are the same. The key differences between the electricity modules are in their regionality and the methodologies used to calculate competitive prices and regional capacity reserve margins. The POEMS electricity component operates with regions approximating the electric Power Control Areas in the United States, a total of 114 regions. The NEMS electricity module, on the other hand, operates at regions based on the North American Electricity Reliability Councils, a total of 13 regions.

Both models base their competitive prices for generation services primarily on what economists refer to as the marginal cost of power--the short run operating costs of the last plant dispatched during each time period.^(2),(3) To this cost POEMS adds a charge if needed to ensure that new plants recover their investment costs. NEMS adds a reliability price adjustment to reflect the value of reserve capacity.⁽⁴⁾ In both models, it is assumed that power from Federal and State facilities will continue be to priced at the average cost of service. Reserve margins are input assumptions to POEMS, while NEMS solves for reserve margins internally by balancing the cost and additional reliability that comes with adding a new plant against the value that consumers place on reliability.

Comparison of Key Results

The key result of this analysis is that, despite model differences, NEMS and POEMS produce similar results when using the same assumptions (Table ES1). The NEMS and POEMS results for electricity sales, carbon emissions from the electricity sector, and electricity prices are similar. The difference in the CECA Reference cases produced by NEMS and POEMS in 2010 are about 0.5 percent for electricity sales and 1.1 and 0.5 percent for carbon emissions and electricity prices, respectively. In the CECA Competitive cases NEMS and POEMS produce electricity sales and electricity prices which differ by 0.2 percent and 0.7 percent, respectively, in 2010. While the electricity sales figures are closest they are not identical; although NEMS and POEMS use the same demand models, the difference in regional detail between the models can result in small differences. The NEMS CECA Competitive case prices are slightly higher than the comparable values from POEMS because NEMS includes sales taxes for generation services in the competitive price, but POEMS does not.

The CECA Competitive case results for carbon emissions in 2010 differ by approximately 2.5 percent (15 million metric tons) between NEMS and POEMS. However, the 2010 differences in carbon emissions between the CECA Reference and Competitive cases are similar--61 million metric tons lower in POEMS and 53 million metric tons lower in NEMS. The difference in carbon emissions appears to result partially from the difference in electricity losses in the two electricity models. NEMS, using transmission and distribution loss factors calibrated to 1997 data, uses slightly more electricity generation than POEMS to meet a similar level of demand.⁽⁵⁾ The loss factors used in POEMS appear to be about 1.0 to 1.5 percentage points lower by 2010 than those used in NEMS.

Comparison of Renewable Portfolio Standard Results

The similarity in the model results extends to their response to the renewable incentives included in CECA Competitive case in the Supporting Analysis. The Supporting Analysis assumed that new wind and biomass plants built between 2000 and 2015 would receive a 1.5-cent tax credit for each kilowatthour produced over the first 10 years of their operation. A 1.0 cent per kilowatthour incentive was assumed for coal plants co-firing biomass. In addition, the 7.5-percent renewable portfolio standard (RPS) for nonhydroelectric renewables in CECA is included in the CECA Competitive cases. In both NEMS and POEMS, the construction of new wind plants and the increased use of biomass in coal plants provide most of the increase in renewable generation stimulated by these incentives. In NEMS, however, the renewable credit price reaches the 1.5 cent per kilowatthour cap set by CECA before the 7.5-percent share is reached. In 2010, both the NEMS and POEMS competitive cases reach a nonhydroelectric renewable share of 7.0 percent, but in 2015 the shares are 7.1 percent in the NEMS case and 7.7 percent in the POEMS case.⁽⁶⁾

Regional Comparison of Results

At the NEMS electricity region level, the results are generally similar. The largest differences occur in the western regions, where NEMS and POEMS show different electricity trade patterns. NEMS shows more electricity generation in the Northwest (16 percent) than does POEMS (and ships the power to California). As in the Supporting Analysis, at the NEMS region level, nearly all regions show lower prices in the NEMS CECA Competitive case than in the NEMS CECA Reference case in 2010. The only region where this does not occur in NEMS is the Northwest, where the competitive price is 1.4 percent higher than the reference price in 2010.

Summary

The overall similarity of the NEMS and POEMS projections under a common set of assumptions demonstrates that DOE's Supporting Analysis results are not dependent on model features. This study did not evaluate the assumptions used in developing the Supporting Analysis. The results using either of the models would be different if other assumptions regarding the impacts of competition on the operation and performance of the electricity supply sector were used.

The differences between results for the CECA Reference and CECA Competitive cases are nearly the same in the two models. Using NEMS, in 2010, CECA Competitive case electricity prices are 13 percent lower, renewable generation is 35 percent higher, and carbon emissions from the electricity sector are 7 percent lower than in the CECA Reference case. Using POEMS, the corresponding differences between the two cases in 2010 are 14 percent lower electricity prices, 38 percent higher renewable generation, and 9 percent lower carbon emissions from the electricity sector.

Executive Summary Notes