EPA Analysis of Alternative SO 2 and NO x Caps for Senator Carper July 31, 2009

1. U.S. Environmental Protection Agency Office of Air and Radiation EPA Analysis of Alternative SO2 and NOx Caps for Senator CarperJuly 31, 2009

2. Request for Analysis

3. This document represents EPA’s analysis of the six scenarios for SO2 and NOx caps requested by Senator Carper. The analysis was conducted by EPA’s Office of Air and Radiation.

4. Analytic Approach • This analysis estimates the emissions reductions, costs and benefits that would occur under six requested legislative scenarios, relative to currently implemented regulations including the Clean Air Interstate Rule (CAIR). This analysis does not compare the scenarios with full implementation of the Clean Air Act including National Ambient Air Quality Standards for PM (2006) and Ozone (2008). • Under the Act, EPA is required to promulgate several rules that will achieve emissions reductions within the time frame of this analysis. We do not estimate the emissions reductions and benefits that would occur under continued implementation of the CAA, or compare those with emissions reductions and benefits of potential legislation. • This was not feasible within the time frame for this analysis and would have required assumptions about the requirements of rules not yet proposed (Utility MACT, rule to replace CAIR in response to court remand, etc.) • Therefore the results of this analysis can not be portrayed as the costs and benefits of the scenarios compared to what might happen under current law. It only allows comparison of the scenarios modeled with rules issued so far. Reference case for this analysis • The reference case for this analysis includes major federal and state rules for EGUs that are on the books and applicable to sources, as well as controls required in consent decrees. The baseline does not include any requirements beyond those on the books (climate policy, Utility MACT standard, BART limits, etc.) • EPA must issue regulations to replace CAIR pursuant to a court decision. Those regulations have not yet been proposed. For analytical purposes we assume full implementation of CAIR rather than attempting to prejudge future rulemakings. The Clean Air Mercury Rule (CAMR) was not included in the modeling. • The reference case is essentially the same as the one used for the recent Waxman-Markey climate bill analysis. • All emissions reductions, costs and benefits in this analysis are relative to this reference case.

5. Clean Air Interstate Rule States Covered in Clean Air Interstate Rule (CAIR) for SO2 and NOX Source: EPA, 2007 Note: On February 8, 2008, the U.S. Court of Appeals issued a decision vacating the Clean Air Mercury Rules (CAMR) and thereby suspending the program that allowed mercury emissions trading.

6. Analytical Scenarios The analysis focuses on six different power sector cap & trade scenarios for SO2 and NOx. Control Scenario 1: Annual Emissions Caps Control Scenario 2: SO2 cap same as #1 in 2012, then 1.5 million tons in 2015; NOx caps same as #1 Control Scenario 3: SO2 same as #2; NOx caps same as #1 in 2012, then 1 million ton NOx cap for Eastern (CAIR) and 0.25 million ton NOx cap for Western (non-CAIR) region in 2015 Control Scenario 4: SO2 cap same as #1 in 2012, then 1 million ton cap in 2015; NOx caps same as #1 Control Scenario 5: SO2 caps same as #2; national NOx caps equal to sum of regional NOx caps in #1; no regional NOx caps Control Scenario 6: SO2 caps same as #2; existing NOx requirements until 2015 (no new 2012 caps), then same as #2 for 2015 and beyond Eastern region for this analysis includes ME, VT and NH in addition to the original 28 CAIR states and DC. Currently, power sector NOx emissions are more than 3 million tons annually, of which 2.4 million tons are in the Eastern region and 0.67 million tons are in the Western region. Power sector SO2 emissions are approximately 7.6 million tons nationally.

7. Detailed Electricity Sector Modeling Results

8. Detailed Electricity Sector Modeling with IPM • Motivation for Using the Integrated Planning Model (IPM): • EPA has employed the Integrated Planning Model (IPM) to project the near-term impact of alternative SO2 and NOx caps on the electricity sector. IPM has also been used for modeling conventional pollutant trading programs in the past such as CAIR, multi-pollutant legislative proposals, and climate policies such as the Waxman-Markey bill. • Power Sector Modeling (IPM 2009 ARRA Ref. Case): • The model has been updated to include assumptions from the revised Energy Information Administration's Annual Energy Outlook 2009, taking into account the impacts of the American Recovery and Reinvestment Act (ARRA) of 2009. This update changes the reference case forecast for renewable energy considerably. • This version of IPM was used to analyze the Waxman-Markey energy and climate bill. The reference case assumes the continuation of CAIR for the entire modeling period. • This version of the model incorporates key updates related to technology costs, state rules, current and planned emissions controls; carbon capture and storage technology for new and existing coal plants; and technology penetration constraints on emission control retrofits and new generation capacity. • No demand response was modeled in these scenarios. Note: For more detail on the assumptions used in EPA’s application of IPM, please see more detailed documentation for IPM at http://www.epa.gov/airmarkets/progsregs/epa-ipm/index.html.

9. Key Model Updates and Major Power Sector Assumptions Modeled in IPM Updates to IPM 2009 ARRA Ref. Case: • Electricity Demand Growth: Calibrated to AEO 2009 ARRA update (issued in April). • Cost of New Power Technologies: Consistent with AEO 2009 ARRA update. • Biomass: Supply curves and non-electricity demand for biomass are calibrated to AEO 2009 ARRA update. • Cost of Carbon: An increase to the capital charge rate for new coal plants (consistent with AEO 2009). • State RPS and Climate Programs: Calibrated to AEO 2009 with finalized regulations like RGGI. • CCS in Baseline: Reflecting updated financial incentives including ARRA, 2 GW of CCS capacity are projected for 2015 in the baseline. Other Key Assumptions • Starting Bank: Calculated using 2008 SO2 allowance bank as a starting value and projected 2012 emissions in the reference case. Used interpolation for 2009 – 2011 emissions. Assumed current vintage year allowances would be retired before pre-2010 allowances. • Use of Banked Allowances: Allowances banked prior to 2012 (or 2015) were allowed to be used in the new trading programs. • Limitation on FGD and SCR retrofits in 2012: Used historical retrofit data to limit number of new controls that could be added in 2012 due to time/resource constraints for coal-fired units. Due to time limitations, EPA made a simple assumption about feasibility based on recent construction rates; more thorough consideration might lead EPA to use different assumptions in future. • State Programs: As they have done historically, States will initially rely on the cap and trade control scenarios to lower SO2 and NOx emissions and plan additional controls on power plants in local areas where warranted by local circumstances. For instance, in the case of the recent NAAQS for Ozone and Nitrogen Dioxide, the regulated community will not need to comply until after 2015 (when the second phase caps of the control scenarios go into effect). We do not specifically account for these actions, which in themselves lead to more expensive compliance by the regulated community. On the other hand, the control scenarios become cheaper when localities directly control power plant emissions. Note: See Appendix B for more detail on updates to IPM and key assumptions. For more detail on the all of the assumptions used in EPA’s application of IPM, please see more detailed documentation for IPM at http://www.epa.gov/airmarkets/progsregs/epa-ipm/index.html.

10. Other Considerations and Analyses • EPA has used the existing CAIR program as part of base case in conjunction with all other EPA and state regulations and enforcement settlement agreements between power plants and EPA and the states that are in effect. Without a CAIR program, States would still have an obligation for attaining the NAAQS for fine particles and ozone. The Agency believes that the direct control regimes they would likely adopt would be more costly in providing the same broad regional reductions of SO2 and NOx emissions. • In the last seven years, EPA and EIA have analyzed various versions of multi-pollutant control legislation, as well as regulations (consideration of CAIR, the Clean Air Mercury Rule, and the Clean Air Visibility Rule) for the power sector. These results provide a useful backdrop in considering the results of the analysis. EPA's analysis may be found at www.epa.gov/airmarkets/progsregs/cair/multi.html. EIA's analysis can be found at www.eia.doe.gov/oiaf/service_rpts.htm. • With additional time, EPA would have tested the sensitivity of the results to key assumptions such as electricity demand, (physical) capital costs, fuel prices, and other factors. Some sense of the implications of these factors can be found by referencing previous EPA analysis cited above. In summary, these comparisons identified the following: • Higher natural gas prices tended to result in less fuel switching and, therefore, more generation from coal. • In addition, allowance prices were higher relative to the main scenarios when higher gas prices were modeled. • With demand response, lower electricity demand reduced overall costs to the power sector.

11. Nationwide SO2 Emissions and Allowance Prices • Scenarios #2, 3, 5, and 6 have nearly the same results for SO2 emissions and allowance prices and are displayed as overlapping lines in the graph. • For the reference case (with CAIR), the price actually represents the total cost a company would pay for the allowances needed to cover one ton of SO2 (or the allowance price per ton of emissions). • Current allowance prices may be lower than modeled prices due to regulatory uncertainty concerning the replacement rule for CAIR. In the reference case, the model reflects the assumption of full implementation of CAIR as EPA originally finalized the rules. • State level emissions data is available in Appendix A for more detailed comparisons of emissions changes.

12. Nationwide NOx Emissions and Allowance Prices • Scenarios #1, 2, 4, and 5 have nearly the same results for NOx emissions and allowance prices and are displayed as overlapping lines in the graph. • Although the NOx emissions are nearly the same for Scenarios 1, 2, 4 and 5, the allowance price is higher in the East and lower in the West than it would have been when a national cap is modeled instead of the regional caps. However, the regional caps affect the location of emissions reductions. • State level emissions data is available in Appendix A for more detailed comparisons of emissions changes.

13. Coal-Fired Scrubber (FGD) and SCR Retrofits Scenarios #2, 3, 5, and 6 have nearly the same results for FGD additions and are displayed as overlapping lines in the graph. Scenarios #1, 2, 4, 5, and 6 have nearly the same amount of SCR retrofits. In 2008, there was also more than 5 GW of fluidized bed combustion (FBC) generation, which burns coal and achieves approximately 90% SO2 capture.

14. Annual Incremental Costs Although the incremental cost in 2025 appears a little higher under a national cap (#5) than under the regional caps (#2), the incremental cost over the entire time period is, in fact, lower. IPM has found the cheapest operation of the grid under each scenario, which leads to different compliance patterns for installing controls and fuel switching between options.

15. Coal Production and Use in the Power Sector

16. Generation Mix

17. New Capacity Additions and Retirements • Natural gas new capacity additions in 2025 in all scenarios are a function of electricity demand, low gas prices, and its cost competitiveness relative to other technologies when allowances are factored in. • The mix of new renewables capacity is roughly 75% wind, over 15% biomass, and less than 10% other (landfill gas, geothermal, solar). • In reality, uneconomic units may be “mothballed,” retired, or kept running to ensure generation reliability. The model is unable to distinguish among these potential outcomes. • Most uneconomic units are part of larger plants that are expected to continue generating. Currently, there are roughly 120 GW of oil/gas steam capacity and 320 GW of coal capacity.

18. Retail Electricity Prices Electricity prices rise slightly relative to the reference case– 1.5 to 2.5%. Prices are highest in Scenario #3 with the tightest NOx cap. Historical price from AEO2009.

19. Natural Gas Use and Prices in the Power Sector Natural gas prices increase by a small amount relative to the reference case. The increase for all scenarios is about 2% in 2020 and is even smaller by 2025. This is the delivered fuel price for the power sector only. There will be slight price effects in other parts of the economy that are not modeled and, on a percentage increase basis, they would be smaller than in the power sector. Historical consumption and price from AEO2009.

20. Health Benefits Modeling Results

21. Incremental Human Health Benefits • This analysis uses a simplified methodology for assessing human health impacts and associated economic benefits of reducing EGU SO2 and NOx emissions by applying benefit per ton estimates to IPM modeled emission reductions. • PM2.5 benefit per ton • PM2.5 benefit per ton factors are derived from a peer reviewed journal article containing benefit per ton estimates for power sector SO2 and NOx emissions (Fann et al. 2009). The factors were updated using the Value of a Statistical Life (VSL) from EPA’s Guidelines for Preparing Economic Analyses – $7.4M 2006$. • Two concentration response functions estimating incidences of adult premature mortality are presented • Pope et al. 2002 (American Cancer Society cohort) – historically used by EPA as the central estimate of premature mortality • Laden et al. 2006 (Harvard Six Cities study) – a more recent study used by EPA in recent health assessments and regulatory impact analyses • Ground-level ozone benefits • Ozone benefit per ton factors for the total value of all health endpoints are derived from previous multipollutant analyses (Methods for Projection of Health Benefits for EPA’s Multi-Pollutant Analyses of 2005). Ozone benefit per ton factors for incidences of health endpoints avoided could not be generated in the timeframe of this analysis. • These per ton estimates reflect national average impacts associated with regional reductions in SO2 and NOx, and do not represent the specific benefits associated with a particular spatial pattern of emissions reductions. • As such, while application of the per ton estimates to the estimates of nationwide changes in SO2 and NOx emissions provided by IPM gives an approximation of benefits, the specific benefits associated with the geographic pattern of emissions reductions estimated by IPM would be different. • In general, the benefit per ton of regional SO2 and NOx reductions are greater in the east than the west. Because a majority of SO2 emission reductions under the modeled scenarios are expected to take place in the east, applying national benefit per ton numbers likely underestimates total benefits. Note: All references for health benefits modeling can be found in Appendix C.

22. Incremental Human Health Benefits • Methodology notes • The methodology for estimating these benefits is consistent with EPA's most recent Regulatory Impact Analyses (U.S. EPA, 2006, 2009a, 2009b) and reflects the most recent review of PM science (U.S. EPA, 2009c). • The choice of concentration-response functions linking PM2.5 and premature mortality is consistent with advice from EPA's Science Advisory Board and with the recent state of the art expert elicitation on PM-related mortality sponsored by EPA, and published in the peer-reviewed literature (Roman et al, 2008). • Each of the studies that provide a PM2.5 mortality effect estimate has strengths and weaknesses, and as such, both estimates should be considered in evaluating the potential benefits of SO2 and NOx emissions reductions. • The range of mortality reductions across the effect estimates based on the two epidemiological studies reflects the views of most (10 out of 12) of the experts included in EPA's expert elicitation. • The impact function for fine particles is approximately linear within the range of ambient concentrations under consideration. Thus, the estimates include health benefits from reducing fine particles in areas with varied concentrations of PM, including both regions that are in attainment with fine particle standard and those that do not meet the standard. • Uncertainty in estimating incidences of premature mortality from PM2.5 • Inhalation of fine particles is judged to be causally associated with premature death at concentrations near those experienced by most Americans on a daily basis. Although biological mechanisms for this effect have not been established definitively yet, the weight of the available epidemiological evidence supports an assumption of causality. • All fine particles, regardless of their chemical composition, are assumed to be equally potent in causing premature mortality. This is an important assumption, because PM produced via transported precursors emitted from EGUs may differ significantly from direct PM released from diesel engines and other industrial sources, but no clear scientific grounds exist for supporting differential effects estimates by particle type.

23. Incremental Human Health Benefits • PM2.5 Results • For each assessment year, the range of health benefits estimated for each policy scenario is defined by Control Scenario 1 (minimum) and Control Scenario 4 (maximum). • In 2015 and beyond, there is very little difference between Control Scenarios 2, 5, and 6. Scenario 3 achieves additional benefits due to additional NOx reductions. Note: Different levels of emission reductions may yield equivalent incidences of premature mortality avoided and different levels of incidences of premature mortality avoided may yield equivalent total value due to rounding.

24. Incremental Human Health Benefits • PM2.5 Results Annual incidences of PM2.5 related premature mortality avoided over time using Laden et al.* Annual value** of all quantified PM2.5 health endpoints over time using Laden et al.* (billions, 2006$) *In general results using Pope et al. are 40% of results using Laden et al. ** 3% discount rate

25. Incremental Human Health Benefits • PM2.5 Results • In addition to the significant benefits to adult premature mortality, each scenario assessed provides considerable benefits to infant premature mortality and morbidity endpoints. • Incidences avoided under Control Scenario 1 illustrate the magnitude of morbidity benefits under the different scenarios. • In 2025, Control Scenario 1 is expected to reduce: • 18,000 incidences of acute myocardial infarction • 180,000 incidences of asthma exacerbation • 6,900 incidences of chronic bronchitis • 19,000 incidences of emergency room visits and hospital admissions for respiratory and cardiovascular conditions

26. Incremental Human Health Benefits • Ozone Results • The metric illustrated here is the annual value of all quantified ozone mortality and morbidity endpoints including incidences of premature mortality, school loss days, worker productivity, and emergency room and hospital admissions for respiratory conditions. • For each assessment year, the maximum ozone health benefits are achieved by Control Scenario 3. • In 2015 and beyond, there is very little difference between the remaining Control Scenarios. Annual value of all quantified ozone mortality and morbidity endpoints (billions, 2006$) Note: Ozone benefits do not use a discount rate because there is no time lag for benefits.

27. Examples of Unquantified Benefits • Improvements in visibility in national parks and recreational areas • Improvements in visibility in residential areas • Decreases in sulfur deposition (resulting in reduced acidification of surface waters and damage to forest ecosystems and soils) • Decreases in nitrogen deposition (resulting in reduced acidification of surface waters, damage to forest ecosystems and soils, and coastal eutrophication) • Decreases in mercury deposition, leading to reduced exposure to mercury through fish consumption • Decreases in ozone-related damage to agricultural and forest production