MEMORANDUM

DATE: 	January 30, 2015

SUBJECT:	Residential Wood Heater Cost Effectiveness Analysis

FROM:	Jill Mozier, EC/R Inc.

TO:		David Cole, EPA/OAQPS
		Gil Wood, EPA/OAQPS
		Amanda Aldridge, EPA/OAQPS

	The purpose of this memorandum is to present our cost effectiveness (CE) analyses for the PM emission reductions estimated to occur under the final New Source Performance Standards (NSPS) for Residential Wood Heaters. The final stepped standards and their underlying emission and cost assumptions have already been summarized in the projected emissions, manufacturer cost impacts  and unit cost memos. This memo will explain how we applied the derived cost and PM emission estimates (based on our already-documented assumptions) in our CE analyses.

   I. Application of Cost Estimates to Cost Effectiveness Analysis
      
      A. 	NSPS-Related R&D, Certification, and Reporting & Recordkeeping Costs
      As noted in our manufacturer cost impacts memo, one round of R&D is assumed  -  beginning in 2015, the rule's effective date, and ending in 2020, the year of the final Step 2 PM standard. We made assumptions regarding the number of model lines that would undergo R&D and assume the related costs. In cases where we assumed less than 100% of the models would require R&D (e.g., wood stoves, pellet stoves and hydronic heaters), this was based on the percentage of models estimated to already meet the 2020 Step 2 PM limit, as explained in both the manufacturer cost impacts memo and the projected emissions memo. It is important to note that manufacturers will likely consolidate their model lines, focusing on models able to achieve the limit, rather than incur R&D costs for every currently non-compliant model line. Therefore fewer models than we assumed in our cost estimates will likely undergo R&D, which will decrease cost impacts, potentially significantly. But since we do not have data to suggest the degree of consolidation which may occur, we assumed all non-compliant models would remain in existence and incur R&D costs, in order to avoid underestimating costs. Thus, in our CE analyses we assumed the following percentages and models would incur NSPS-related R&D costs beginning in 2015:

 74% of wood stove manufacturers (with an estimated 125 models representing catalytic, noncatalytic and hybrid designs);
 30% of pellet stove manufacturers (with an estimated 125 models);
 100% of single burn rate stove manufacturers (with an estimated 20 models) would face a doubling of the R&D engineering cost portion of total model development costs in 2015 and 2016 to represent intensification of R&D efforts for this previously unregulated appliance type;
 100% of forced air furnace manufacturers (with an estimated 50 models) would face a doubling of the R&D engineering cost portion of total model development costs in 2015 and 2016 to represent intensification of R&D efforts for this previously unregulated appliance type; and
 82% of hydronic heater manufacturers (with an estimated 120 models).
      
      As noted in the manufacturer cost impacts memo, all manufacturers, except for wood stove manufacturers who are subject to the current 1988 NSPS (and therefore already incur certification costs), will face ongoing certification costs (every 5 years per model line) above baseline conditions. However, in the 2015 to 2020 timeframe, we have incorporated these costs as part of the overall R&D expenditures. 
      
      After 2020, our CE analysis tracks the ongoing annual certification and annual reporting and recordkeeping costs faced by manufacturers. We based reporting and recordkeeping (R&R) costs on the annual average costs derived from development of the Information Collection Request (ICR) supporting statements.[,] These are annual estimates of the ongoing R&R burden to manufacturers associated with final rule compliance. 
      
      B.	Model Design Lifespan
	For proposal, we assumed a 20-year model design lifespan. We made this assumption to best characterize the actual model design lifespan given that many models developed for the 1988 NSPS are still being sold (after 25 years), and many "new" models still have the same internal working parts with merely exterior cosmetic changes. Respectful of comments on the proposed rule, however, in which some industry representatives commented that a shorter model lifespan is more accurate, we reduced our assumed model design lifespan to 10 years for this analysis. Therefore our CE analysis tracks shipments through year 2029 (assuming a 10-year design life for a model meeting the Step 2 limit in year 2020).


   I. Application of Emission Estimates to Cost Effectiveness Analysis
      
      A.	Baseline and NSPS emission estimates
      As noted in our emissions memo, we multiplied our tons/appliance PM emission estimate by the estimated number of shipments in each year to calculate total PM emissions from each appliance category per year. We made this calculation under baseline conditions (i.e., in the absence of this final NSPS) and under the final NSPS. Furthermore, in order to not overstate emission reductions under the NSPS, baseline emissions were discounted by the estimated percentage of appliances already meeting the Step 2 PM limit (i.e., 26% of wood stoves, 70% of pellet stoves, and 18% of hydronic heaters are estimated to already meet Step 2, as noted in Section I.A).
      
        The emission estimates assume that the total number of shipped units meet the PM standard in the year the standard is implemented. Because our PM emission estimates assume that models do not come into compliance until the year they are required to, we do not claim PM emission reductions until each compliance year. This simplifying assumption means that we underestimated how soon PM emission reductions will happen under the final NSPS. In reality, some models will meet the NSPS Step 2 PM limit prior to the 2020 compliance year and will therefore be emitting less than baseline levels prior to that year.
      
      Further, based on available data for wood stoves and pellet stoves, we assumed that baseline PM emissions for both of these appliance types are already at the Step 1 PM level. We therefore claimed no PM emission reductions for wood stoves and pellet stoves until the 2020 Step 2 compliance year. It should furthermore be noted that the assumption that at baseline all wood stoves already meet the Step 1 (Washington State) PM standard may slightly underestimate baseline emissions, as not 100% of wood stoves currently sold do meet the Washington state PM standard. Finally, in our CE analyses PM emission reductions for forced-air furnaces are not estimated to begin until the Step 1 PM limit is required in 2016, even though the work practice/operational standard required in 2015 will result in some (un-estimated) PM emissions reduction. We made these simplifying assumptions in order to avoid any potential for overstating PM baseline emissions and PM emissions reduction under the final NSPS.
      
      B.	Appliance Emitting Lifespan
      As noted above, we used a 10-year model design lifespan for estimating costs. We also assumed a 20-year appliance emitting lifespan. This assumption was made to best characterize the actual use lifespan given that most stoves in consumer homes emit for at least 20 years and often much longer. Our CE analysis therefore tracks emission reductions out through 2048 (assuming a 20-year emitting life for an appliance shipped in year 2029). 
      
      Our CE analysis incorporates "trailing emissions" as part of our 20 year lifespan assumptions. For example, our analysis assumes that a stove shipped in 2015 will emit in homes for 20 years  -  or until 2034 (inclusive of both 2015 and 2034). We therefore drop emissions from this stove in our analysis in year 2035. Likewise, we drop emissions for a stove shipped in 2016 in year 2036, and so on.
      
      
   II. Resulting Cost Effectiveness Tables
      
      Tables 1 through 8 show the results of our CE analyses based on PM2.5 reductions under the final NSPS for each appliance, for appliances grouped by subparts AAA ("room heaters") and QQQQ ("central heaters"), and for Subparts AAA and QQQQ combined. Each table shown is based on 2013 dollars (2010 dollars were used at proposal) and a 7% interest rate applied to the amortized costs during the 6-year R&D period. 
      
      We also prepared estimates based on a 3% interest rate and these are presented in Tables 9 through 16. 
      
      In addition, we prepared both sets of tables based on volatile organic compound (VOC) emission reductions and based on carbon monoxide (CO) emission reductions, although these pollutants do not have emission limits under the final NSPS. The VOC and CO cost effectiveness estimates are available in the docket. 
      
      Finally, we prepared a sensitivity analysis in which we varied the GDP-based forecasted growth rate, used as a proxy for growth in shipments, from 2.0% to 2.1%, 2.5% and 3.0%, which caused emission estimates and resulting cost effectiveness to vary. The CE tables associated with this sensitivity analysis are provided in Appendix A of this memo.











                  PM2.5 Cost Effectiveness Tables 1 through 8
                                       
                                       
      These estimates are based on a 7% interest rate for amortized costs
                    and a 2% forecasted growth in shipments

                                       

                                       
                                       

                                       


                                       
                                       

                                       










                      This page left intentionally blank


                                       













                 PM2.5 Cost Effectiveness Tables 9 through 16
                                       
                                       
      These estimates are based on a 3% interest rate for amortized costs
                    and a 2% forecasted growth in shipments
                                       



                                       
                                       

                                       


                                       

                                       
                                       

                                       
                                       

                                       
                                       
                                       

                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                      This page left intentionally blank










                                  Appendix A
                                       
                  Sensitivity Analysis: Growth in shipments 
                                       
      The following cost effectiveness (CE) tables are based on varying the forecasted growth in shipments and therefore emissions. The sensitivity analysis uses growth rates of 2.1%, 2.5% and 3.0%. A growth rate of 2.0% is used for our final emission and CE estimates described in the body of this memo. All sensitivity analysis estimates are based on a 7% interest rate for amortized costs.

 Tables A-1 through A-8 are based on a forecasted growth in shipments and emissions of 2.1%
 Tables A-9 through A-16 are based on a forecasted growth in shipments and emissions of 2.5%
 Tables A-17 through A-24 are based on a forecasted growth in shipments and emissions of 3.0%


                                       
                                       
                                       





                                       

                                       
                                       


                                       
                                       




                                       

                                       


                                       

                                       





                                       

                                       

                                       
                                       






                                       


                                       

                                       
                                       

                                       


                                       

                                       



                                       
                                       

                                       



                                       





                                       
                                       
