NSPS Supplemental Proposal for Supplemental Regulatory Impact Analysis (RIA) for “Standards of Performance for New Residential Wood Heaters, New Residential Hydronic Heaters and Forced-Air Furnaces


Memorandum

TO: Docket for rulemaking: "Standards of Performance for New Residential Wood Heaters, New
Residential Hydronic Heaters and Forced-Air Furnaces (EPA-HQ-OAR-2018-0195)"

DATE: November 20, 2018

SUBJECT: Supplemental Regulatory Impact Analysis (RIA) for "Standards of Performance for New
Residential Wood Heaters, New Residential Hydronic Heaters and Forced-Air Furnaces" Introduction

1. Introduction

The U.S. Environmental Protection Agency (EPA) promulgated revisions in February 2015 to new source
performance standards (NSPS) for residential wood stoves, and promulgated NSPS for other wood
heating appliances such as pellet stoves, forced air furnaces, single burn rate stoves, and hydronic
heaters. The EPA conducted this rulemaking under the authority of section 111 of the Clean Air Act
(CAA), "Standards of Performance for New Stationary Sources," under which the EPA establishes federal
standards of performance for new sources within source categories which cause or contribute
significantly to air pollution, which may reasonably be anticipated to endanger public health or welfare.
The 2015 NSPS broadened the applicability of the wood heaters regulation beyond adjustable burn rate
heaters (i.e., "stoves", the focus of the original regulation) to specifically include single burn rate
heaters, pellet stoves, hydronic heaters, and forced-air furnaces. Heaters/stoves and model lines
manufactured after the compliance dates would be required to meet fine particulate matter (PM2.5)
standards. Since that action was economically significant according to Executive Order 12866 (i.e., either
costs or benefits of $100 million or greater in any one year), the EPA prepared a regulatory impact
analysis (RIA) of the potential costs and benefits associated with the 2015 NSPS, "Standards of
Performance for New Residential Wood Heaters, New Residential Hydronic Heaters and Forced-Air
Furnaces," (80 FR 13671). The March 2015 NSPS RIA is available at Docket ID No. EPA-HQ-OAR-2009-
0734. EPA now proposes to allow an additional two years for sell-through of certain wood heaters - new
residential hydronic heaters and forced-air furnaces, or central heaters, as defined in subpart QQQQ of
this NSPS - beyond the compliance date of May 15, 2020 for Step 2 requirements contained within the
2015 NSPS. A sell-through is an allowance to manufacturers and retailers to sell for some period of time
non-Step 2-compliant wood heaters past the compliance date of May 15, 2020. In addition, the proposal
includes a request for comment on allowing an additional two years for sell-through of new wood stoves
covered by the 2015 NSPS (sources covered under subpart AAA). Analysis and results for wood stoves is
included in Appendix A of this supplemental RIA. The Agency also seeks comment on whether and what
type of small business relief may be appropriate in place of the extended sell-through period.

Wood heater manufacturers and retailers informed the EPA in early 2018 that a substantial number of
retailers are already reducing or even ending their purchases of Step 1 certified wood heating devices
from manufacturers. The retailers are concerned they will not be able to sell these devices before the
May 2020 Step 2 compliance date and will be left with unsaleable inventory. The Step 1 requirements
became effective on May 15, 2015 and are emissions limits that reflect demonstrated wood heater
technologies at that time. Step 2 requirements are more stringent emissions limits that manufacturers
were expected to achieve by the May 15, 2020 compliance date. Meanwhile, as stated in the proposed

1

-------
preamble and in the public docket for this rulemaking, manufacturers may need until May 2020 to
develop, test, and certify wood heating devices to meet the 2020 Step 2 standards. Thus, some
manufacturers may potentially face revenue losses in the near term to the extent that retailers are not
willing to purchase the manufacturers' Step 1-certified models and the manufacturer does not yet have
Step 2-certified models to sell. Further, as May 2020 approaches, it is reasonable to assume that
retailers may become increasingly reluctant to purchase non-Step 2-compliant wood heating devices
which they will not be able to sell after May 2020, resulting in stranded capital. EPA is aware that there
are a small number of Step 2 certified models currently available to purchase. However, the price
difference between the Step 2 models and Step 1 models may dampen demand for these heaters.

To address these issues, EPA is proposing to amend the 2015 NSPS, 40 CFR part 60, subpart QQQQ
requirements to create a two-year sell-through period for Step 2 that is comparable to the Step 1 sell-
through period. EPA is proposing an amendment that will allow Step 1-compliant hydronic heaters and
forced-air furnaces manufactured or imported before May 15, 2020, to be sold at retail through May 15,
2022. Under this proposal, after May 15, 2022, only Step 2-compliant hydronic heaters and forced-air
furnaces can be sold at retail.

EPA is seeking comment on this two-year sell-through period for Step 2, including both the need for a
Step 2 sell-through period and, if a sell-through period is added, what length of sell-through period is
reasonable, and why. EPA is also seeking comment on whether and what type of small business relief
might accomplish the same goal. Further details on the proposal can be found in the proposal preamble.

This memorandum presents EPA's estimates of the cost savings and forgone health benefits that may
result from the proposed additional two-year sell-through of residential hydronic heaters and forced-air
furnaces.1 This memorandum is a supplement to the RIA prepared for the 2015 NSPS; it does not replace
the 2015 NSPS RIA but analyzes a proposed action that affects only a portion of the 2015 NSPS. The cost
and emissions data from the 2015 NSPS RIA are used in this supplemental RIA. The baseline for this RIA
is the 2015 NSPS, since the requirements of this standard remain in effect as of this proposal. In
addition, this memorandum recognizes that consumers, in the absence of this rule, may behave largely
in one of two ways. Because these heaters are durable goods, they may delay their purchase of new
wood heaters if their existing wood heaters are still working, even if near the end of the replacement
cycle. In the other extreme, despite the price difference, consumers may purchase Step-2 wood heaters
because their existing heaters are not functioning. We do not contemplate in this analysis consumers
switching to heaters that use other fuel types such as natural gas. We seek comment on the likely
consumer response to the regulatory proposals and the market conditions driving them.

In addition, EPA is using a different method to calculate the estimates of forgone emission reduction for
this analysis. We are presenting information on truncated emission reduction, instead of more typical
annualized emission reduction estimates. For the final rule, the agency will provide annualized emission
reduction estimates.

1 To help commenters understand the estimated effects of an additional two-year sell through of new
wood stoves, the cost savings, forgone emissions reductions and health benefits associated with
including new wood stoves in this proposed sell through are included as an appendix.

-------
Within the RIA, impacts will be estimated for appliances (or units, these terms are interchangeable in
this RIA) covered under subpart QQQQ under three scenarios:

1. Hydronic heaters and forced-air furnace manufacturers covered under this proposal
produce no additional appliances beyond the production estimates in the 2015 NSPS RIA
and have no stranded Step 1-compliant models;

2. Manufacturers produce hydronic heaters and forced-air furnaces in the same quantities
estimated for the 2015 NSPS RIA. However, non-Step 2-compliant hydronic heaters and
forced-air furnaces, that would otherwise have been stranded, will now be available for sale
for two years beyond the May 15, 2020 compliance date; and

3. Manufacturers take advantage of the additional two years of sales of non-Step 2-compliant
hydronic heaters and forced-air furnaces by increasing production of these appliances.

Newly produced non-Step 2 appliances will be available for sale in addition to those that
would have otherwise been stranded.

These scenarios reflect the uncertainty within the data available to EPA in the proportion of appliances
affected by this proposal that could comply with the Step 2 requirements in the 2015 NSPS by May 15,
2020. The scenarios also reflect the uncertainty in the reaction of manufacturers and retailers to this
proposal. These same three scenarios will also be applied to residential wood stoves in this
supplemental RIA; these impacts will be presented separately in Appendix A.

For scenario 1, it is assumed that manufacturers do not produce additional wood heaters beyond those
projected in the 2015 NSPS RIA, and further, that they are able to sell all step 1-compliant appliances in
advance of the 2020 compliance date. Therefore, there are no cost savings and no forgone emission
reductions. As part of the rationale for this scenario, manufacturers are presumed to act with perfect
foresight and consistent with the modeling of wood heater production estimates included in the 2015
NSPS RIA.

For scenario 2, units that are non-Step 2-compliant are assumed to be available for sale in 2021 and
2022. This scenario assumes that manufacturers produced more units than were needed for purchase
by retailers prior to the sell-through. Hence, impacts of this proposal will occur beyond the 2015 NSPS
compliance date. Retailers will receive additional revenues from the sale of non-Step 2-compliant units,
and forgone emission reductions will occur not only during the sell through period, but over the life of
the non-Step 2-compliant units.

For scenario 3, manufacturers affected by this proposal are assumed to produce additional non-Step 2-
compliant units beyond those projected in the 2015 NSPS RIA in anticipation of the sell through period.
Cost savings are experienced by manufacturers since the cost of producing non-Step 2-compliant models
is estimated to be lower than that of Step 2-compliant models, and, as in scenario 2, manufacturers are
able to sell units that would otherwise have been non-saleable after the compliance date for the 2015
NSPS. As in scenario 2, forgone emission reductions would be experienced during the sell through period
and over the expected life of the non-Step 2-compliant units. As part of the rationale for this scenario,
manufacturers are assumed to increase production in anticipation of the additional 2-year sell-through
and retailers are assumed to be willing to carry and sell these units. Under this scenario, impacts begin
in 2019 since manufacturers increase production prior to 2020.

-------
To provide background to the estimation of cost savings and forgone emission reductions, we show in
Table 1 the number of appliances projected to be sold (or shipments) from 2019-2022 according to the
analyses done for the final 2015 NSPS. The number of wood stoves projected to be sold in those years
can be found in Table A-l in Appendix A. These projections of appliance shipments are the basis for the
analyses of impacts under the three scenarios.

Table 1. Projected Appliance Sales from 2019-2022 for Hydronic Heaters and Forced Air Furnaces*

2019

2020

2021

2022

Forced-Air Furnaces

30,058

30,659

31,272

31,898

Hydronic Heaters

15,048

15,349

15,656

15,969

* All estimates from the RIA for Residential Wood Heaters NSPS Revision. February 2015. Chapter 4, and U.S.
Environmental Protection Agency (U.S. EPA), 2015. "Estimated Emissions from Residential Wood Heaters." January
30, 2015. Prepared by EC/R. p. 11.

Impacts presented in section 2 of this supplemental RIA include the estimated increase in revenues
(hereafter called the "cost savings" for this proposal) to manufacturers that produce appliances affected
by this sell-through and retailers of these appliances. The estimated forgone PM2.5 emission reductions
associated with this proposal, truncated at the close of the sell through period and covering years 2019-
2022 are presented in section 3. Forgone emission reductions of volatile organic compounds (VOC) and
carbon monoxide (CO) are also estimated for the years 2019-2022, but any forgone health and
environmental impacts associated with these forgone emission reductions are not quantified or
monetized due to lack of data, resources, and methodological limitations. Other impacts, such as for
employment, energy, and small businesses, are in section 4. The estimated monetized forgone benefits
associated with the truncated forgone PM2.5 emission reductions are discussed in section 5. Finally, an
assessment of net benefits (the cost savings minus the forgone monetized benefits) is presented in
section 6. All estimates are presented in 2016 dollars and are relative to the baseline, which is the 2015
NSPS. Estimates are for 2019-2022 for scenario 2 and 3, though there are no impacts in 2019 under
scenario 2. All cost savings and forgone monetized benefits covering years 2019-2022 only are
presented in annual and present value (PV) terms. Scenario 2 is treated as the primary, or most
representative, scenario analyzed in this RIA.

2. Cost Savings

For this analysis, EPA assumes the Step 2 levels of compliance through 2020 that are consistent with the
2015 NSPS RIA.2 Because the compliance date for the 2015 NSPS is not changed by this proposal, all
model development costs prior to 2020 are unaffected by this proposal. The cost savings estimated in
this proposal are the increase in revenues to manufacturers and retailers from selling appliances
covered under subpart QQQQthat are compliant with Step 1 (but not Step 2) two years beyond the May
15, 2020 compliance date.

2 U.S. Environmental Protection Agency. Regulatory Impact Analysis for Residential Wood Heaters NSPS
Revision. EPA-452/R-15-001. February 2015. p. 4-14. Available at

https://www3.epa.gov/ttn/ecas/docs/ria/wood-heaters ria final-nsps-revision 2015-02.pdf.

-------
The cost savings estimates for each scenario are based on the expected number of wood heaters
produced that are Step 1-compliant in 2019. For each scenario, we estimate that 85 percent of the wood
heaters produced to meet projected sales in 2019 are Step 1-compliant. This estimate of Step 1
compliance (and, therefore, non-Step 2 compliance) is consistent with wood heater model data taken as
of summer 2018 from EPA's lists of certified wood stoves, certified hydronic heaters, and certified
forced-air furnaces.3 We also estimate that 20 percent of these Step 1-compliant units will not be sold
until the sell-through period begins on May 15, 2020. For 2020, all wood heaters produced are expected
to be Step 2-compliant, and therefore no Step 1-compliant wood heaters will be produced after 2019.
We then project the sales of Step 1-compliant wood heaters for years 2020 through 2022.

For scenario 3, in addition to the leftover units sold under scenario 2, we estimate that manufacturers
increase production of Step 1-compliant wood heaters by 10% in 2019 over projections reported in the
2015 NSPS RIA. The impacts associated with scenario 3 will be greater than those estimated for scenario

2 since a larger number of Step 1-compliant units are expected to be sold during the sell through
extension.

The estimates of affected wood heaters under scenarios 2 and 3 are found in Tables 2 and 3,
respectively. As shown in Table 2, we estimate that 5,110 forced air furnaces and 2,558 hydronic heaters
that are Step 1-compliant would be sold during the two-year sell through period under scenario 2. Table

3 shows that 7,665 (5,110 + 2,555) forced air furnaces and 3,837 (2,558 + 1,279) hydronic heaters that
are Step 1-compliant would be sold during the two-year sell through period under scenario 3. In
recognition of the diminishing stock of Step 1-compliant units over the sell through period as well as the
potential for variation in demand for Step 1-compliant units, especially lower demand as the end of the
sell through period approaches, we estimate that 70 percent of the Step 1-compliant units will be sold in
2020, 25 percent in 2021, and 5 percent in 2022.

These calculations are also available for wood stoves, and can be found in Table A-2 in Appendix A.

Table 2. Estimated Appliance Sales from 2019-2022 for Hydronic Heaters and Forced Air Furnaces Under
Scenario 2 *

2019

2020

2021

2022

Forced Air Furnaces

30,058

30,659

31,272

31,898

Step 1

20,439

3,577

1,278

256

Step 1- Stranded units not sold in 2019**

5,110

Step 2

4,509

27,082

29,994

31,642

Hydronic Heaters

15,048

15,349

15,656

15,969

Step 1

10,233

1,791

640

128

Step 1 - Stranded units not sold in 2019**

2,558

Step 2

2,257

13,558

15,016

15841

* All estimates taken from or based on the RIA for Residential Wood Heaters NSPS Revision. February 2015.
Chapter 4, and U.S. Environmental Protection Agency (U.S. EPA), 2015. "Estimated Emissions from Residential

3 U.S Environmental Protection Agency. List of EPA Certified Wood Stoves, List of EPA Certified Hydronic
Heaters, and List of Certified Forced-Air Furnaces. Retrieved in June 2018 from

https://epa.gov/compliance/list-epa-certified-wood-stoves. https://epa.gov/compliance/list-epa-
certified-hvdronic-heaters and http://epa.gov/compliance/list-epa-certified-forced-air-furnaces.

-------
Wood Heaters." January 30, 2015. Prepared by EC/R. p. 11.

** Estimates are for appliances produced in 2019 but not sold until the sell-through period begins.

Table 3. Estimated Appliance Sales from 2019-2022 for Hydronic Heaters and Forced Air Furnaces Under
Scenario 3*

2019

2020

2021

2022

Forced Air Furnaces

30,058

30,659

31,272

31,898

Step 1

20,439

5,366

1,916

383

Step 1 - Produced but not sold in 2019**

5,110

Step 1 - Additional units produced for sell through

2,555

Step 2

4,509

25,293

29,356

31,515

Hydronic Heaters

15,048

15,349

15,656

15,969

Step 1

10,233

2,686

959

192

Step 1 - Produced but not sold in 2019**

2,558

Step 1 - Additional units produced for sell through

1,279

Step 2

2,257

12,663

14,697

15,777

* All estimates taken from or based on the RIA for Residential Wood Heaters NSPS Revision. February 2015.

Chapter 4, and U.S. Environmental Protection Agency (U.S. EPA), 2015. "Estimated Emissions from Residential
Wood Heaters." January 30, 2015. Prepared by EC/R. p. 11. Estimates for 2019 reflect a 10% increase in production
of Step 1-compliant heaters only based on sales projections from the 2015 NSPS RIA and an 85% compliance rate
pre-2020.

** Estimates are for appliances produced in 2019 but not sold until the sell-through period begins.

The cost savings are estimated by assuming that the increase in revenue for each covered appliance
during the two years of additional sell-through can be approximated by the avoided manufacturing cost
associated with producing Step 2-compliant units. For scenario 2, in the absence of the proposal,
manufacturers would be required to manufacture Step 2-compliant models to meet demand in 2020
and beyond. We assume under this scenario that total sales as projected in the 2015 NSPS RIA for these
wood heater appliance categories will not change. Under the proposal, manufacturers are able to sell
left over, Step 1-compliant wood heaters, displacing the sale of (and need to manufacture) a portion of
the Step 2-compliant units. The per unit cost savings, under scenario 2, is therefore the total
manufacturing cost of the displaced Step 2-compliant appliance (additional manufacturing cost + the
base manufacturing cost), or $2,674 per forced-air furnace and $8,123 per hydronic heaters as shown in
the unit cost memo prepared for the 2015 NSPS.4 The other costs of manufacture such as research and
development (R&D) are sunk and still required by the NSPS, and thus are not included in the cost savings
estimate for this scenario.

Scenario 3 builds on scenario 2 in that, in addition to the stock of left-over step 1 units, manufacturers
choose to make additional Step 1-compliant units, increasing production of these appliances by 10% in
2019, in anticipation of the 2-year sell-through extension. The per unit cost savings for the newly
manufactured additional units are the incremental manufacturing costs associated with Step 2 appliance
production, $1,700 per forced air furnace and $3,200 per hydronic heater, since manufacturers are
choosing to produce additional Step 1-compliant units in 2019 in lieu of producing Step 2-compliant

4 U.S. Environmental Protection Agency (U.S. EPA), 2015. "Unit Cost Estimates of Residential Wood
Heating Appliances." January 30, 2015. Prepared by EC/R. p. 5.

-------
appliances in 2020. The cost savings for the leftover Step 1-compliant models under this scenario are as
reported under scenario 2 ($2,674 per forced air furnace and $8,123 per hydronic heater). As before,
the R&D costs for developing the Step 2-compliant models are still required by the NSPS and are not
affected by this proposal.

These appliance costs are then applied to the estimated number of Step 1 appliances that will be
affected under each scenario using the estimates of appliances shipped annually for each category as
found in Chapter 4 of the 2015 NSPS RIA and other calculations under scenarios 2 and 3 as shown
earlier. Thus, the cost savings reflect the estimates derived in Tables 2 and 3, respectively, of the
number of Step 1-compliant units that would be sold during the sell-through period. Finally, all the cost
savings estimates are escalated to 2016 dollars from the original 2013 dollars to conform to Executive
Order 13771 requirements by using the annual Gross Domestic Product (GDP) implicit price deflator.5
Estimates of increased revenues to manufacturers and retailers are calculated for each year of the
additional two-year sell-through. These total estimated cost savings for forced air furnaces and hydronic
heaters under scenarios 2 and 3 are presented in Table 4 below. (Estimates for wood stoves are
presented in the Appendix (Table A-2).)

Table 4. Estimated Cost Savings of the Proposal (2016$, millions)

2019

2020

2021

2022

Scenario 1

Forced Air Furnaces

Hydronic Heaters

Annual Totals

Scenario 2

Forced Air Furnaces

$7.2

$3.7

$0.7

(primary)

Hydronic Heaters

15.1

5.6

1.0

Annual Totals

22.3

9.3

1.7

Scenario 3

Forced Air Furnaces

$10.5

$4.8

$0.9

Hydronic Heaters

18.1

6.7

1.2

Annual Totals

$28.6

$11.5

$2.1

To be consistent with the presentation of impacts in the 2015 NSPS RIA, we calculate the annual average
of the cost savings over the period 2019-2022. For the primary scenario, this annual average is $8.3
million (2016 dollars). For scenario 3, the annual average is $10.6 million (2016 dollars). For wood stoves
only, the annual average cost savings is $5.7 million for both scenarios 2 and 3. The cost savings for
wood stoves is the same for each scenario because there is no difference in production costs between
the Step 1 and Step 2-compliant models as stated in the unit cost memo. Thus, manufacturers only incur
a cost savings for the sale of the leftover models as determined in scenario 2. Documentation for these
calculations can be found in Appendix B for this supplemental RIA and in the spreadsheet that provides
the estimates of forgone net benefits (including cost savings and forgone benefits) for this proposed
rule.

Since this proposal is economically significant under Executive Order 12866, and this action is subject to
Executive Order 13771, we provide estimates of the cost savings in present value (PV) terms for the

5 For more information on the GDP implicit price deflator, see

https://fred.stlouisfed.org/series/A191RD3A086NBEA. The 2016 annual GDP implicit price deflator is
1.04107 higher than the 2013 value (110.221/105.873). Accessed on November 5, 2018.

-------
primary scenario. As can be seen in Table 5, the estimated total PV of the cost savings is $25 million (7%
discount rate) and $29 million (3% discount rate). The equivalent annualized value (EAV) of the cost
savings are $9 million (7% discount rate) and $10 million per year (3% discount rate). Present values are
for 2016, and EAVs are consistent with the PVs. All estimates are in 2016 dollars. These estimates are
presented in the spreadsheet that provides the estimates of forgone net benefits (including cost savings
and forgone benefits) for this proposal.6

Table 5. Present Value of the Cost Savings for the Proposed Primary Scenario 2, discounted to 2016 (2016$,
millions)

7% Discount Rate

3% Discount Rate

2019

2020

2021

2022

Present Value in 2016

Equivalent Annualized Value

3. Forgone Emission Reductions

In this section, we present estimates of the forgone emissions reductions associated with each of the
three scenarios. The estimated forgone emission reductions (as well as estimates of forgone benefits)
are made under the assumptions discussed in section 1. The estimates do not account for any changes
in baseline conditions since the analysis for the 2015 NSPS was conducted (e.g., technological change or
other factors affecting compliance with the 2015 NSPS) nor do they account for any foregone trailing
emissions reductions beyond 2022 that are associated with the installation of cleaner appliances in the
years of the analysis. This includes the emissions factors estimated for Step 1 and Step 2 for each wood
heater category included in this analysis, which are the same emissions factors included in the 2015
NSPS RIA.

Estimating the forgone emission reductions closely follows the procedure laid out above for estimating
the increases in revenues to manufacturers and retailers. This is because the annual growth assumption
and other necessary adjustments to the emissions from the 2015 NSPS RIA are similar. The one major
difference in our procedure is to apply the estimated percent of emission reductions that will take place
beginning in May 2022. This estimate will include the percent emission reductions applied for each
appliance category in 2020 (when the Step 2 2015 NSPS requirements become effective) but adjusted to
account for the increase in shipments and emissions between 2019 and 2022. The percent reduction in
emissions is derived from emissions data for each pollutant presented in Chapter 4 of the 2015 NSPS
RIA.

The forgone emission reductions occur as a result of the Step 1-compliant hydronic heaters and forced-
air furnaces that are estimated to be stranded in 2020 under scenarios 2 and 3. These appliances are
then available for sale during the two-year sell through period. The additional emissions generated by

6 PV_EAVs-forgonebenefits+costsavings_RWHproposalfinal.xls. Available in the public docket for this
proposal.

-------
these appliances as compared to Step 2 appliances are calculated using the procedure described above.
These emissions now occur later than projected under the emission analyses done for the 2015 NSPS
RIA and continue through the life of the Step 1-compliant appliance. The forgone PM2.5 emission
reductions, truncated at the close of the sell through period, serve as an input to the calculation of
forgone benefits, which are accounted for in Section 5.

Table 6 presents annual forgone PM2.5 emission reductions, truncated at year 2022, for each scenario.
(The forgone emission reductions for VOC and CO, also truncated, are presented in Table 5 and the
truncated forgone emission reductions of each pollutant for each scenario for wood stoves are
presented in Table A-5.)

Table 6. Estimated Annual Forgone PM2.5 Emission Reductions(tons)7

2019

2020

2021

2022

Scenario 1

Forced Air
Furnaces

Hydronic Heaters

Annual Totals

Scenario 2
(primary)

Forced Air
Furnaces

214

290

305

Hydronic Heaters

Annual Totals

271

368

387

Scenario 3

Forced Air
Furnaces

320

434

458

Hydronic Heaters

118

124

Annual Totals

406

552

582

Note: Estimates do not reflect foregone emissions reductions beyond 2022.

For purposes of this supplemental RIA, and to be consistent with the presentation of impacts in the 2015
NSPS RIA, we calculate the annual average of the forgone emission reductions over the period of the
additional sell-through.8 These estimates are provided in Table 7. (Similar estimates are provided for
wood stoves in Table A-6.) For Scenario 2 (the primary or proposal scenario), the annual average
estimate of forgone emission reductions for the years 2019 to 2022 for PM2.5 is 257 tons. This annual
average forgone emission reduction estimate is about 2% of the average annual estimate of PM2.5
emission reductions expected from the 2015 NSPS for the years 2019 to 2022 (12,638 tons).9 We use
this forgone emission reduction estimate as an input to the monetized benefits analysis presented in
section 5. The procedure for estimating forgone emission reductions is found in Appendix B of this

7 We note that the annual forgone emissions of PM2.5 are 387 tons for Scenario 2 and would be 7,740
tons over the 20-year model life typical of hydronic heaters and forced-air furnaces. For Scenario 3,
these estimates are 582 tons and 11,640 tons, respectively.

8 The costs and benefits in the 2015 NSPS RIA were calculated as the annual average of impacts from
2015 to 2020 (inclusive). This presentation of impacts allowed for inclusion of impacts in the first year of
implementation of Step 1 emissions limits (2015) and the first year of implementation of Step 2
emissions limits (2020). For more information on how impacts are presented in that RIA, refer to p. 1-3
of the 2015 NSPS RIA.

9 Memo to David Cole, U.S. EPA, Gil Wood, U.S. EPA, and Amanda Aldridge, U.S. EPA from Ec/R, Inc.
Residential Wood Heaters Cost Effectiveness Analysis. January 30, 2015. Calculated from data on p. A-24
of this memo.

-------
supplemental RIA and in the cost and emissions spreadsheet for this proposal. The annual average of
forgone PM2.5 emission reductions for scenario 3 can also be found in Table 7, and so can the annual
averages of forgone VOC and CO emission reductions for each scenario. More details on the annual
forgone VOC and CO emission reductions can be found in Tables A-4 and A-5, respectively, in Appendix
A.

Table 7. Annual Average of Forgone Emission Reductions from Hydronic Heaters and Forced-Air Furnaces at Each
Scenario for Each Pollutant for the 2019-2022 Period (tons)

PM2.5

VOC

Scenario 1

Scenario 2 (primary)

257

271

1,444

Scenario 3

385

551

2,162

Note: Values reported are not annualized and do not reflect foregone emissions reductions beyond 2022..

4. Other Impacts

Energy and Employment Impacts

These proposed actions are anticipated to have negligible impacts on energy costs or usage. To the
extent that Step 1-compliant hydronic heaters and forced-air furnaces continue to be sold for an
additional two years, it is difficult to determine the precise energy impacts that might result from this
proposed action. Wood-fueled appliances compete with other biomass forms for residential heating as
well as more traditional energy sources such as oil, electricity and natural gas. There is also a lack of data
to determine the potential for affected consumers to choose other types of fuels and their associated
appliances, nor the potential impacts to employment by affected manufacturers. Employment may
increase among affected manufacturers because of being allowed to continue sell-through of wood
burning appliances for additional two years. For a complete discussion of employment impacts, see the
2015 NSPS RIA.

Small Entity Impacts

We prepared a small entity impacts analysis for the final 2015 NSPS that is included in the 2015 NSPS
RIA. As part of that analysis, a profile of the industries subject to the 2015 NSPS indicated that a great
majority (more than 90%) of affected firms were small businesses as defined by the Small Business
Administration (SBA).10 Industries that manufacture hydronic heaters and forced air furnaces will not be
adversely impacted by this proposal. The small businesses in these industries will experience a net
positive impact. We believe that the estimate of the percentage of firms affected by this proposal that
are small is about the same as that for the 2015 NSPS. EPA guidance for implementing SBREFA (Small
Business Regulatory Enforcement Fairness Act) indicates that no small business analysis is required as
part of a rulemaking since affected small businesses will experience a positive net economic effect,
though EPA has the discretion to prepare one in such an instance.11

10 U.S. Environmental Protection Agency. Regulatory Impact Analysis for Residential Wood Heaters NSPS
Revision. EPA-452/R-15-001. February 2015. p. 6-13. Available at

https://www3.epa.gov/ttn/ecas/docs/ria/wood-heaters ria final-nsps-revision 2015-02.pdf.

11 U.S. Environmental Protection Agency. Final Guidance EPA Rulewriters: Regulatory Flexibility Act as
Amended by the Small Business and Regulatory Enforcement Fairness Act. November 2006. p. 13,

-------
5. Forgone Benefits

We quantify the value of the forgone PM2.5 emission reductions using an approach that is
methodologically identical to one we employed in the 2015 analysis. For Scenario 2, using the truncated
foregone emissions reported above, we estimate the yearly average total forgone monetized benefits
for this proposal to be $100 million to $230 million at a 3% discount rate and $94 million to $210 million
at a 7% discount rate for the period from 2019 to 2022. For scenario 3, we estimate $160 million to $350
million at a 3% discount rate and $140 million to $320 million at a 7% discount rate for the period from
2019 to 2022. All estimates reported in 2016$ (Table 8). Estimated values for wood stoves may be found
in Table A-6. These estimated benefits reflect the monetized value of the reduction in incidence of
morbidity and premature death projected under the 2015 NSPS that might no longer occur if this
proposal is finalized as proposed. Data, resources, and methodological limitations prevented EPA from
monetizing the forgone benefits from several important benefit categories. In the interest of brevity and
consistency with the 2015 NSPS RIA, we refer readers interested in learning more about the methods
employed to quantify the number and economic value of the PM-related impacts to the 2015 NSPS RIA
(EPA, 2015).

Included among the non-monetized forgone benefits are those associated with reduced exposure to
about 271 tons per year of VOCs. VOCs are also precursors to ozone formation, and therefore also
associated with reducing health impacts due to ozone exposure. Further, this proposal is estimated to
result in forgone reduction of 1,444 tons of CO each year, as well as benefits from pollutants whose
emissions changes we did not quantify, such as black carbon emissions and several HAP emissions like
benzene, formaldehyde, and dioxin. This proposal is also estimated to result in forgone reductions of
ecosystem effects as well as visibility impairment attributable to the forgone PM emission reductions.
Further, consistent with the analysis conducted for the 2015 NSPS RIA, the analysis presented here does
not reflect the trailing emissions (those occurring beyond 2022) from appliances installed during the sell
through period.

Table 8. Summary of Annual Average Monetized PIVh.s-Related Health Forgone Benefits Estimates for the New
Residential Wood Heaters, New Residential Hydronic Heaters, and Forced-Air Furnaces NSPS Proposal, 2019-
2022 (2016$)a'b

Emissions
Changes (tons)

Benefit per ton
(Krewski, 3%)

Benefit per ton
(Lepeule, 3%)

Benefit per ton
(Krewski, 7%)

Benefit per ton
(Lepeule, 7%)

Scenario 1

$400,000

$910,000

$360,000

$820,000

Scenario 2

257

Scenario 3

385

Total Monetized Forgone Benefits
(millions $2016 at 3%)

Total Monetized Forgone Benefits
(millions $2016 at 7%)

Scenario 1

Scenario 2

$100

$230

$94

$210

Scenario 3

$160

$350

$140

$320

Footnote 14. Available at https://www.epa.gov/sites/production/files/2015-06/documents/guidance-
regflexact.pdf.

-------
a All monetized forgone benefits are annual average estimates that reflect the average of annual forgone emission
reductions expected to occur between 2019 and 2022 (inclusive) resulting from rulemaking implementation. All
estimates are rounded to two significant figures so numbers may not sum across columns. The forgone
monetized benefits do not account for effects from exposure to NO2 or ozone, ecosystem effects, or impaired
visibility, or trailing foregone emissions beyond 2022 for appliances installed during the sell through period. All
fine particles are assumed to have equivalent health effects regardless of chemical composition, but the benefit
per ton estimates vary depending on the location and magnitude of their impact on PM2.5 levels, which in turn
affect the number of individuals exposed. The forgone monetized benefits incorporate the conversion from
precursor emissions to ambient fine particles. The benefit per ton method prevents us from reporting
confidence intervals.

b Estimates of forgone VOC and CO health benefits are currently not monetized and are addressed only
qualitatively.

The forgone benefits estimates presented in Table 8 are based on average annual estimates of forgone
emission reductions realized between 2019 and 2022. Tables 9 and 10 present the PV of these forgone
benefits and their equivalent annualized values for scenarios 2 and 3.

Table 9. Present Value of the Forgone Benefits of Scenario 2 discounted to 2016 (2016$, millions)*

7% Discount Rate

3% Discount Rate

2019

$0 to $0

2020

$90 to $210

$110 to $240

2021

$120 to $260

$140 to $320

2022

$120 to $260

$140 to $320

Present Value in 2016

$250 to $560

$350 to $780

Equivalent Annualized Value

$70 to $160

$90 to $210

There are no forgone emission reductions in 2019 from Scenario 2, thus there are no forgone benefits for that
year. Values presented do not reflect foregone benefits associated with trailing foregone emissions beyond 2022.

Table 10. Present Value of the Forgone Benefits of Scenario 3, discounted to 2016 (2016$, millions)41

7% Discount Rate

3% Discount Rate

2019

$0 to $0

2020

$100 to $230

$120 to $260

2021

$130 to $290

$150 to $350

2022

$130 to $290

$160 to $360

Present Value in 2016

$270 to $610

$380 to $860

Equivalent Annualized Value

$80 to $180

$100 to $230

Characterization of Uncertainty in the Monetized Forgone PM?.s Benefits
A complex analysis such as this one includes many data sources as inputs, including emission
inventories, air quality data from models (with their associated parameters and inputs), population data,
population estimates, health effect estimates from epidemiology studies, economic data for monetizing
benefits, and assumptions regarding the future state of the world (i.e., regulations, technology, and
human behavior). Each input parameter is subject to uncertainty, which in turn affect the estimated
forgone benefits estimate. When the uncertainties from each stage of the analysis are compounded,
even small uncertainties can have large effects on the total forgone quantified benefits. Therefore, the
estimates of annual forgone benefits should be viewed as representative of the magnitude of annual
forgone benefits expected, rather than the actual forgone benefits that would occur every year.

-------
This supplemental RIA does not include the type of detailed uncertainty assessment found in the PM
NAAQS RIA (U.S. EPA, 2012). However, the results of the uncertainty analyses presented in the PM
NAAQS RIA can provide some information regarding the uncertainty inherent in the forgone benefits
results presented in this analysis. Sensitivity analyses conducted for the PM NAAQS RIA indicate that
alternate cessation lag assumptions could change the PM2.5-related mortality benefits discounted at 3%
by between 10% and -27% and that alternate income growth adjustments could change the PM2.5-
related mortality benefits by between 33% and -14%.12

We did not simulate the change in the level and location of air quality changes associated with this
rulemaking. As such, it is not feasible to estimate the proportion of benefits occurring in different
locations, such as PM2.5 nonattainment areas. Instead, we applied benefit-per-ton estimates, which
reflect specific geographic patterns of emissions reductions and specific air quality and benefits
modeling assumptions for each sector (US EPA, 2013). For more information and updated methodology,
see the TSD describing the calculation of these benefit-per-ton estimates (U.S. EPA, 2018). We are
unable to estimate the percentage of premature mortality associated with this specific action's forgone
emission reductions at each PM2.5 level. However, we believe that it is still important to characterize the
distribution of exposure to baseline air quality levels. As a surrogate measure of mortality impacts, we
provide the percentage of the population exposed at each PM2.5 level in the baseline of the source
apportionment modeling used to calculate the benefit-per-ton estimates for this sector. Baseline
exposure is only one parameter in the health impact function, along with baseline incidence rates
population, and change in air quality. Thus, the percentage of the population exposed to air pollution
below the lowest measured level (LML) is not the same as the percentage of the population
experiencing health impacts because of a specific emission reduction policy. Therefore, caution is
warranted when interpreting the LML assessment for this action because these results are not
consistent with results from rules that had air quality modeling.

6. Net Benefits

In Table 11, we present one perspective on the costs and benefits of this proposal by offering a
comparison of the forgone benefits from the targeted pollutant - PM2.5 (the costs of the proposal) -
truncated at the close of the sell through period, with the cost savings or increased revenues to
manufacturers (the benefits of the proposal). Table 11 contains net benefits for the primary scenario
(the proposal scenario). Note that in reporting the benefits, costs, and net benefits of this proposal in
this table, we modify the relevant terminology to be more consistent with traditional net benefit/cost
analysis. Net benefits, then, equals the benefits minus the costs (or, in the terminology applied
elsewhere in this report, the cost savings minus the forgone benefits). These net benefits are presented
using the annual average cost savings and forgone benefits as a basis.

Table 11. Cost Savings, Forgone Benefits, and Net Benefits of the Scenario 2 (Primary) Associated with the
Targeted Pollutant, as an Annual Average for the 2019-2022 Time Frame (billions of 2016$)a

3% Discount

7% Discount Rate

Rate

Cost: Forgone Benefits'5

($0.10) to

($0.09) to ($0.21)

($0.23)

12 http://www.epa.gov/ttn/ecas/regdata/RIAs/finalria.pdf (pp 6-16).

-------
Benefit: Cost Savings from
Increased Manufacturers' and
Retailers' Revenues



$0.01



Net Benefits:

($0.09) to
($0.22)



($0.08) to ($0.20)

aAII estimates in this table are rounded to one decimal point, so figures may not sum due to independent rounding.
b Forgone benefits were calculated using a benefit-per-ton estimate corresponding to each of three regions of the
U.S. The forgone health benefits reflect the forgone PM2.5 benefits for the sell through period and reflect the range
based on adult mortality functions (e.g., from Krewski et al. (2009) with Bell et al. (2004) to Lepeule et al. (2012)
with Levy et al. (2005)). The forgone monetized health benefits do not account for forgone emissions of directly
emitted VOC, CO, and hazardous air pollutants; trailing forgone PM2.5 emissions beyond 2022, ecosystem effects;
or visibility impairment.

In Table 12, we present the benefits, costs, and net benefits of this proposal's present values and show
the present value of net benefits. The PV of the benefits (cost savings) are $0.01 billion (7% discount
rate) and $0.01 billion (3% discount rate). The PV of the costs (forgone benefits) are $0.09 to $0.21
billion (7% discount rate) and $0.10 to $0.23 billion (3% discount rate). Hence, the PV of the net benefits
are $0.07 to $0.19 billion (7% discount rate) and $0.07 to $0.20 billion (3% discount rate). The
monetized present values of net benefits (or net present values) presented in Table 11 are negative as
denoted by the values being in parentheses, meaning that the estimated costs (forgone benefits) of the
proposal are greater than the estimated benefits (cost savings). All estimates are discounted to 2016
and are in 2016 dollars.

Table 12. Estimated Present Value of the Net Benefits of Scenario 2 (Primary) (billions of 2016$)



7% Discount Rate

3% Discount Rate

PV of Benefits1

$0,025

$0,029

PV of Costs2

($0.09) to ($0.21)

($0.10) to ($0.23)

NPV of Net Benefits

($0.07) to ($0.19)

($0.07) to ($0.20)

1 The PV of benefits are the cost savings (increased manufacturer and retailer revenue) as presented in Section 2.

2 The PV of costs are calculated from the forgone monetized benefits as presented in Section 5. The PV of forgone
benefits are the forgone benefits at 7 percent (3 percent) discounted to 2016 using a 7 percent (3 percent)
discount rate. Results are rounded to two significant figures. Totals may not sum due to rounding.

Table 13 shows the equivalent annualized values of the net benefits for the proposal discounted at 7
and 3 percent. The equivalent annualized values (EAV) are the annualized present values, or the
levelized flow of the present values, over the four years affected by the proposal. The EAV of the net
benefits are $0.06 billion to $0.15 billion per year (7% discount rate) and $0.08 billion to $0.20 billion
per year (3% discount rate). The negative values indicate that the EAV of the estimated benefits (cost
savings) of the proposal are smaller than the EAV of estimated costs (forgone benefits). In addition, this
calculation of net benefits does not capture forgone net benefits associated with the operation of units
beyond 2022 that are non-Step 2-compliant, units that have a likely operational life of 20 years as
mentioned earlier in Section 3.

14

-------
Table 13. Estimated Equivalent Annualized Value of the Net Benefits of Scenario 2 (Primary) (billions of 2016$)

7% Discount Rate

3% Discount Rate

EAV of Benefits1

$0.01

EAV of Costs2

$0.07 to $0.16

$0.09 to 0.21

EAV of Net Benefits

($0.06) to ($0.15)

($0.08) to ($0.20)

1 The EAV of benefits are the EAV of the cost savings presented in Section 2.

2 The EAV of costs are calculated from the PV of the forgone monetized benefits as presented in Section 5. Results
are rounded to two significant figures. Totals may not sum due to rounding.

References

Bell, M.L., McDermott, A., Zeger, S.L., Samet, J.M., Dominici, F., 2004. Ozone and short-term mortality in
95 US urban communities, 1987-2000. JAMA 292, 2372-8. Available at:

https://doi.org/10.10Ql/iama.292.19.2372.

Krewski D, Jerrett M, Burnett RT, Ma R, Hughes E, Shi, Y, et al. 2009. Extended follow-up and spatial
analysis of the American Cancer Society study linking particulate air pollution and mortality. HEI
Research Report, 140, Health Effects Institute, Boston, MA.

Lepeule J, Laden F, Dockery D, Schwartz J 2012. "Chronic Exposure to Fine Particles and Mortality: An
Extended Follow-Up of the Harvard Six Cities Study from 1974 to 2009." Environ Health Perspect.
Jul;120(7):965-70.

Levy, J.I., Chemerynski, S.M., Sarnat, J.A., 2005. Ozone exposure and mortality: an empiric bayes
metaregression analysis. Epidemiology 16, 458-68. Sisler, J.F. 1996.

U.S. Environmental Protection Agency (U.S. EPA). 2012. Regulatory Impact Analysis for the Final

Revisions to the National Ambient Air Quality Standards for Particulate Matter. EPA-452/R-12-003.
Office of Air Quality Planning and Standards, Health and Environmental Impacts Division. December.
Available at http://www.epa.gov/pm/2012/finalria.pdf.

U.S. Environmental Protection Agency, 2018. Quality Assuring BenMAP-CE Demographic and Economic
Input Data. Research Triangle Park, NC. Available at:

https://www.epa.gov/sites/production/files/2018-02/documents/benmap vl.3 qa memo.pdf.

-------
APPENDIX A

Table A-l. Projected Appliance Sales from 2019-2022 for Wood Stoves*

2019

2020

2021

2022

Wood Stoves

102,000

104,040

106,121

108,243

* All estimates from RIA for Residential Wood Heaters NSPS Revision. February 2015. Chapter 4, and U.S.
Environmental Protection Agency (U.S. EPA), 2015. "Estimated Emissions from Residential Wood Heaters." January
30, 2015. Prepared by EC/R. p. 11.

Table A-2. Estimated Appliance Sales from 2019-2022 for Wood Stoves Under Scenario 2 and 3 *

Scenario 2

2019

2020

2021

2022

Wood Stoves

102,000

104,040

106,121

108,243

Step 1

69,360

12,138

4,335

867

Step 1- Produced but not sold in 2019**

17,340

Step 2

15,300

91,902

101,786

107,376

Scenario 3

Wood Stoves

110,670

104,040

106,121

108,243

Step 1

69,360

18,207

6,502

1,301

Step 1- Produced but not sold in 2019**

17,340

Step 1 - Additional units produced for sell through

8,670

Step 2

15,300

85,833

99,619

106,942

** Estimates are for appliances produced in 2019 but not sold until the sell-through period begins.

Table A-3. Estimated Cost Savings at Each Scenario for Wood Stoves* (2016$, millions)

2019

2020

2021

2022

Scenario 1

Scenario 2 (primary)

$15.9

$5.7

$1.1

Scenario 3

$15.9

$5.7

$1.1

Since there is no difference in production cost between Step 1-compliant and Step 2-compliant wood
stoves, there is no cost savings to manufacturers beyond those occurring from selling leftover stoves
during the sell through period. Thus, the cost savings are the same under each scenario.

Table A-4. Estimated Annual Forgone VOC Emission Reductions (tons)

2019

2020

2021

2022

Scenario 1

Forced Air
Furnaces

Hydronic Heaters

Annual Totals

Scenario 2 (primary)

Forced Air
Furnaces

225

306

322

-------


Hydronic Heaters 0

61

83

87



Annual Totals 0

286

389

409

Scenario 3

Forced Air 0
Furnaces

3458

623

647



Hydronic Heaters 0

124

169

181



Annual Totals 0

582

792

828

Note: Estimates do not reflect foregone emissions reductions beyond 2022.





Table A-5. Estimated Annual Forgone CO Emission Reductions (tons)







2019 2020

2021

2022

Scenario 1

Forced Air Furnaces 0

0

0

0



Hydronic Heaters 0

0

0

0



Annual Totals 0

0

0

0

Scenario 2

Forced Air Furnaces 0

1,203

1,633

1,719

(primary)

Hydronic Heaters 0

323

438

461



Annual Totals 0

1,526

2,071

2,180

Scenario 3

Forced Air Furnaces 0

1,800

2,444

2,573



Hydronic Heaters 0

483

656

690



Annual Totals 0

2,283

3,100

3,263

Note: Estimates do not reflect foregone emissions reductions beyond 2022.





Table A-6. Estimated Annual Forgone Emission Reductions for Each Pollutant at Each Scenario for Wood Stoves
(tons)



2019

2020

2021

2022

PM2.5











Scenario 2 0
(primary)

93

126

133



Scenario 3 0

140

190

208

voc











Scenario 2 0
(primary)

129

175

184



Scenario 3 0

192

243

275

CO











Scenario 2 0
(primary)

1,506

2,044

2,152



Scenario 3 0

2,253

3,065

4,168

Note: Estimates do not reflect trailing foregone emissions reductions beyond 2022.

Table A-7. Annual Average of Forgone Emission Reductions from Wood Stoves at Each Scenario for Each
Pollutant for the 2019-2022 Period (tons)	



PM2.5

VOC

CO

Scenario 1

0

0

0

Scenario 2 (primary)

88

122

1,426

Scenario 3

135

178

2,372

Note: Values are not annualized and do not reflect foregone emissions reductions beyond 2022.

17

-------
Table A-8. Summary of Annual Monetized PM2.5-Related Health Forgone Benefits Estimates for Wood Stoves -
2015 NSPS proposal in the 2019-2022 Time Frame (2016$)a-b	

Scenario

Emissions
Changes
(tons)

Benefit per

ton
(Krewski,
3%)

Benefit
per ton
(LePeule,
3%)

Benefit
per ton
(Krewski,
7%)

Benefit per
ton (LePeule,
7%)

Scenario 2

88

$404,274

$910,182

$364,715

$820,743

Scenario 3

135

$404,274

$910,182

$364,715

$820,743















Total Monetized
Forgone Benefits
(millions $2016 at 3%)

Total Monetized
Forgone Benefits
(millions $2016 at 7%)

Scenario 2

$36

to

O
00
¦uy

$32

to

$72

Scenario 3

$54

to

$122

$49

to

$110

aAII monetized forgone benefits are annual estimates that reflect the average of annual forgone emission
reductions expected to occur between 2020 and 2022 (inclusive) resulting from rulemaking implementation. All
estimates are rounded to two significant figures so numbers may not sum across columns. The forgone monetized
benefits do not account for effects from exposure to NO2 or ozone, ecosystem effects, or impaired visibility. All fine
particles are assumed to have equivalent health effects regardless of chemical composition, but the benefit per ton
estimates vary depending on the location and magnitude of their impact on PM2.5 levels, which in turn affect the
number of individuals exposed. The forgone monetized benefits incorporate the conversion from precursor
emissions to ambient fine particles. The benefit per ton method prevents us from reporting confidence intervals.
b Estimates of forgone VOC and CO health benefits are currently not monetized and will be addressed only
qualitatively.

18

-------
APPENDIX B

Calculation of Cost Savings and Forgone Emission Reductions

Calculation Procedure for Cost Savings (Increased Revenues to Manufacturers and Retailers)

The calculation of the cost savings for this proposal is the increased revenues hydronic heaters and
forced air furnace manufacturers and retailers receive as a result of the 2-year sell through extension for
these wood heaters. As mentioned earlier in this supplemental RIA, the impacts of this proposal are only
estimated for the years 2019, 2020, 2021, and 2022. Impacts for 2022 are only for that portion of the
year up to May 15, 2022, the date that marks the end of the sell through extension. An average of the
cost savings estimated each year is also presented in the supplemental RIA.

The calculation of the cost savings uses the shipments data for hydronic heater and forced air furnace
manufacturers for 2019, 2020, 2021, and 2022 that is in Chapter 4 of the 2015 NSPS RIA found at
https://www3.epa.gov/ttn/ecas/docs/ria/wood-heaters ria final-nsps-revision 2015-02.pdf. In
addition, the unit (or per wood heater) compliance cost estimates for each type of wood heater found in
Chapter 5 of the 2015 NSPS RIA and shown in Section 2 of this supplemental RIA are also used in the
calculation. We also present these estimates for wood stoves, a wood heater category not included in
the proposal but for which comment is requested as to whether it should. A manufacturer that can sell a
Step 1-compliant heater rather than a Step 2-compliant heater will experience an increase in revenue
equal to the cost of compliance for that wood heater as shown in Chapter 5 of the 2015 NSPS RIA. For
scenario 2, for hydronic heaters, this amount is $8,231(2013 dollars); for forced-air furnaces, this
amount is $2,674 (2013 dollars). Under scenario 3, for hydronic heaters, this amount is $3,200 (2013
dollars); for forced-air furnaces, this amount is $1,700 (2013 dollars). For wood stoves, this amount is
$1,295 (2013 dollars) under either scenario. Our analysis includes three scenarios. These scenarios are
described earlier in the supplemental RIA. For scenario 1, the cost savings is zero, and the rationale for
zero impacts is mentioned earlier in the supplemental RIA. For scenario 2, the primary scenario that is
most representative of the impacts of the sell-through extension according to EPA, we calculate the
increase in revenue for manufacturers and retailers using the following procedure:

1) Take the shipment data for the specific wood heater category for a specific year;

2) Apply an 85% estimate of non-Step 2 compliance for units in each affected wood heater
category (reflecting the best Agency estimate based on available wood heater compliance data);

3) Estimate that twenty percent of the Step 1-compliant units will not be sold until after the sell-
through period begins on May 15, 2020. For 2020, all wood heaters produced are expected to
be Step 2-compliant, and therefore no Step 1-compliant wood heaters will be produced after
2019.

Take the total manufacturing cost estimate for the specific wood heater. Under the proposal,
manufacturers are able to sell left over wood heaters, displacing the sale of (and need to
manufacture) a portion of the Step 2-compliant wood heaters. The cost savings is therefore
calculated by applying the total manufacturing cost of the displaced Step 2-compliant wood

-------
heaters (i.e., additional manufacturing cost + the base manufacturing cost). The R&D costs are
sunk and still required by the NSPS;

4)	Multiply the shipment data in 1) by the total manufacturing cost estimate in 2);

In recognition of the potential for variation in demand for Step 1-compliant units over the sell
through period, especially lower demand for such units compared to Step 2-compliant units by
the end of the sell through period, we estimate that 70 percent of the Step 1-compliant units
will be sold in 2020, 25 percent in 2021, and 5 percent in 2022.

5)	Since the additional manufacturing costs for each wood heater are in 2013 dollars, adjust costs
to 2016 dollars by using the GDP implicit price deflator to generate results for each scenario.
Applying the GDP implicit price deflator, costs in 2016 dollars are 1.04107 greater than costs in
2013 dollars.13

For Scenario 3, the calculation procedure is as follows:

1)	Take the shipment data for the specific wood heater category for a specific year;

2)	Begin calculations using 2019 shipment data. This scenario presumes manufacturers will begin
increasing production in 2019 and not wait until 2020;

3)	Adjust the 2019 shipment data to reflect a 10% economic growth rate rather than the 2.0%
economic growth rate used in the 2015 NSPS RIA. This increase reflects the additional output of
affected wood heaters in response to the sell-through extension in the context of this scenario.

4)	Apply an 85% estimate of non-Step 2 compliance for units in each affected wood heater
category (reflecting the best Agency estimate based on available wood heater compliance data)

5)	Estimate that twenty percent of the Step 1-compliant units will not be sold until after the sell-
through period begins on May 15, 2020. For 2020, all wood heaters produced are expected to
be Step 2-compliant, and therefore no Step 1-compliant wood heaters will be produced after
2019.

6)	Presume that 70 percent of the Step 1-compliant units will be sold in 2020, 25 percent in 2021,
and 5 percent in 2022. In recognition of the potential for variation in demand for Step 1-
compliant units over the sell through period, especially lower demand for such units compared
to Step 2 compliant units by the end of the sell through period,

7)	Take the additional manufacturing cost estimate for the specific wood heater. For this scenario,
the cost is the additional manufacturing cost only because manufacturers have a choice to
produce additional Step 1-compliant or Step 2-compliant wood heaters;

8)	Multiply the shipment data in 1) by the additional manufacturing cost estimate for each wood
heater category;

9)	Since the additional manufacturing costs for each wood heater are in 2013 dollars, adjust costs
to 2016 dollars by using the GDP implicit price deflator to generate results for each scenario.

13 For more information on the GDP implicit price deflator, see

https://fred.stlouisfed.org/series/A191RD3A086NBEA. The 2016 annual GDP implicit price deflator is
1.04107 higher than the 2013 value (110.221/105.873). Accessed on November 5, 2018.

-------
Applying the GDP implicit price deflator, costs in 2016 dollars are 1.04107 greater than costs in
2013 dollars.

Since the procedure for scenario 3 is built off of scenario 2, the analytical procedures for these two
scenarios are quite similar. Results for each scenario are found in the cost and emissions spreadsheet for
this supplemental RIA, which is in the public docket for this rulemaking.

Calculation Procedure for Forgone Emission Reductions

The calculation of the forgone emission reductions for this proposal is the increased revenues hydronic
heaters and forced air furnace manufacturers as a result of the 2-year sell through extension for these
wood heaters. By "forgone emission reductions/' we mean those emission reductions that would have
taken place in the absence of the sell through extension. As mentioned earlier in this supplemental RIA,
the impacts of this proposal are only estimated for the years 2019, 2020, 2021, and 2022. Impacts for
2022 are only for that portion of the year up to May 15, 2022, the date that marks the end of the sell
through extension. An average of the forgone emission reductions estimated for each year is also
presented in the supplemental RIA. We also present these estimates for wood stoves, a wood heater
category not included in the proposal but for which comment is requested as to whether it should.

The calculation of the forgone emission reductions of PM2.5, VOC, and CO uses the emissions data for
emissions of these pollutants from hydronic heater and forced air furnace manufacturers for 2020,
2021, and 2022 that is in Chapter 4 of the 2015 NSPS RIA for scenario 2 and including 2019 emissions
data for scenario 3. We estimate the forgone emission reductions using the emissions from the 2015
NSPS RIA and 2019 emissions, also from the 2015 NSPS RIA, as the baseline. Our analysis that generates
these forgone emission reductions includes Scenarios 2 and 3 (since there are no impacts for scenario 1,
as stated previously in the supplemental RIA).

We calculate the forgone emission reductions using the following procedure for scenario 2:

1)	Take the emissions data for the specific wood heater category for a specific year;

2)	Calculate the difference between the 2020 and 2019 emission reductions for each affected
category;

3)	Apply the factor of 0.85 to represent those emissions from non-Step 2-compliant units with the
difference in emission reductions between 2020 and 2019;

4)	Estimate that twenty percent of the Step 1-compliant units will not be sold until after the sell-
through period begins on May 15, 2020. For 2020, all wood heaters produced are expected to
be Step 2-compliant, and therefore no Step 1-compliant wood heaters will be produced after
2019;

5)	Presume that 70 percent of the Step 1-compliant units with be sold in 2020, 25 percent in 2021,
and 5 percent in 2020. These percentages of sales per year are included in the analysis in
recognition of the potential for variation in demand for Step 1-compliant units over the sell
through period, especially lower demand for such units compared to Step 2-compliant units by
the end of the sell through period;

21

-------
6) Present the cumulative forgone emission reductions for each year of the sell through period.
That is, forgone emission reductions in 2021 are the sum of the forgone emission reductions
from Step 1 units sold in 2020 and those sold in 2021. This calculation can then be applied in
2022 to present the forgone emission reductions for that year.

We calculate the forgone emission reductions using the following procedure for scenario 3:

1) Presume manufacturers will begin increasing production in 2019.

2) Apply the factor of 0.85 to represent those emissions from non-Step 2-compliant units with the
difference in emission reductions between 2020 and 2019;

3) Presume a 10 percent increase in affected wood heater production over the amounts projected
in the 2015 NSPS RIA, beginning in 2019. The higher rate of economic growth is a proxy for the
increase of production of wood heater appliance shipments presumed to occur in this scenario.

4) Estimate that twenty percent of the Step 1-compliant units will not be sold until after the sell-
through period begins on May 15, 2020. For 2020, all wood heaters produced are expected to
be Step 2-compliant, and therefore no Step 1-compliant wood heaters will be produced after
2019.

5) Presume that 70 percent of the Step 1-compliant units with be sold in 2020, 25 percent in 2021,
and 5 percent in 2020. These percentages of sales per year are included in the analysis in
recognition of the potential for variation in demand for Step 1-compliant units over the sell
through period, especially lower demand for such units compared to Step 2-compliant units by
the end of the sell through period;

6) Present the cumulative forgone emission reductions for each year of the sell through period.
That is, forgone emission reductions in 2021 are the sum of the forgone emission reductions
from Step 1 units sold in 2020 and those sold in 2021. This calculation can then be applied in
2022 to present the forgone emission reductions for that year.

We note that there are no forgone emissions for wood stoves under Scenario 3 in 2019 since there are
no emission reductions from the 2015 NSPS in 2019. Under the 2015 NSPS, emission reductions begin in
2020 since all wood stoves are expected to meet Step 1 requirements as of 2015 and the Step 2
requirements are not effective until May 15, 2020. Results for each scenario are found in the
spreadsheet for forgone net benefits (including cost savings and forgone benefits) for this supplemental
RIA, which is in the public docket for this rulemaking.

-------