                                                                               
MEMORANDUM


DATE: 	February 21, 2013

SUBJECT:	Unit Cost Estimates of Residential Wood Heating Appliances

FROM:	Jill Mozier, EC/R Inc.
      Beth Friedman, EC/R Inc.

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


	The purpose of this memorandum is to present an estimate of the unit costs of residential wood heating appliances under both the proposed NSPS ("Proposal") and an alternative approach being considered ("Alternative").  We will present the major variables and assumptions underlying the analysis.  Major variables include the estimated number of units shipped per year, the costs to develop new models, baseline costs of models, and the schedule for implementing the revised New Source Performance Standards (NSPS) under both the Proposal and Alternative.  Note that the majority of cost and shipment data used in this analysis was provided by industry sources. 
 
   I. Estimated Annual Shipped Units  
	We used data in the Frost & Sullivan Market (F&S) report on 2008 shipments by product category, and F&S revenue forecasts which incorporated the weak economy in years 2009 and 2010, to calculate the reduced number of shipments in years 2009 and 2010.  Forced air furnaces were outside the scope of the F&S report.  Instead, we used manufacturer estimates of total industry sales in 2008 and applied the F&S market factors to estimate shipments through 2010.  The F&S wood stove numbers included both certified and non-certified (single burn rate) stoves, so we have estimated numbers of non-certified stove shipments included in the total reported wood stove category (i.e., 40,000 units shipped in 2008).[2]  These shipments were then deleted from the adjusted total wood stove category shipments.  We expanded the 2008 single burn rate estimate using the F&S factors.  For years 2011 through 2038 (for the Proposal) and 2011 through 2041 (for the Alternative) estimated shipments are based on a forecasted revenue growth rate of 2.0%, in keeping with the average annual growth in real GDP predicted by the US Bureau of Economic Analysis.  There is not a perfect correlation between shipments and revenue (for example, because of their higher unit cost, pellet stoves generate more absolute revenue than wood stoves), but we think the overall trend in the projection is reasonable in the absence of specific shipment projections. See Table 1, below for a truncated summary of the shipment data through 2023 for all appliances except masonry heaters (i.e., for proposed Subparts AAA and QQQQ). 
	Masonry heaters manufacturing cost impacts vary by the type of producer and the type of certification method.  According to one manufacturer, the masonry heater industry in the U.S. is dominated by the Finnish firm Tulikivi, which manufactures and imports about half of the U.S. masonry heater units installed yearly through its network of installing distributors.  The same manufacturer said that the second largest producer is a Canadian firm, Temp-Cast.  The remainder of the industry is "dozens" of small producers, with probably fewer than 100 (or at least fewer than 200) generating any masonry revenue at all.  Some commercial operations sell core units and/or design kits based on their own design, and other sell units they license from other U.S. or foreign companies.  Finally, some units are custom built.  Based on this information, we assumed that 50 percent of masonry heaters sold per year in the U.S. are Tulikivi models and 35 percent are sold by other manufacturers.  The remaining 15 percent of units are sold by independent contractors.  

      Using the assumptions described above, we subcategorized the masonry heater segment into three groups:  50 percent of stoves are sold by the largest manufacturer, Tulikivi, 35 percent of stoves are sold by other large manufacturers, and 15 percent of stoves are custom built models sold by small manufacturers.  See Table 2, below for a truncated summary of the masonry heater shipment data through 2023 (i.e, for proposed Subpart RRRR).

                                       
Table 1.  Estimated Annual Shipped Units  -  Subparts AAA & QQQQ
                                       
                                       
Table 2.  Estimated Annual Shipped Units  -  Subpart RRRR, Masonry Heater

   II. Cost Model Variables
         
      A. 	Model Development Costs
	We have heard various estimates of the costs to bring a wood heater from concept to completion, from $300,000 for a single model to $1,360,000 for a 4-firebox model line.  A Hearth and Home article estimated the total cost to bring a model from conception to market as $645,000 to $750,000 for steel stoves and over $1 million for cast-iron, enameled wood stoves.  The authors indicated that costs would decrease for separate models in the same line by up to 25%.  Based on this information, we estimate that a 4-model steel line would cost up to $328,125 per model to develop.  These costs include marketing, design, developing first generation, second generation and prototype units; NSPS and safety testing, equipment tooling, etc.  The manufacturer supplying these figures for the article estimated that the NSPS and safety testing component of these costs would constitute $40,000 per model.  

	Two other manufacturers also provided estimated development costs for a 4-box model line, which are summarized in Table 3.

      Another manufacturer estimated costs of new product development, including design, prototype development, testing, tooling equipment and other manufacturing changes, marketing support, materials, training, and education to be in excess of $300,000 over an 8- to 12-month schedule for a relatively uncomplicated product.  Costs will increase for products that have more sophisticated controls.  One other manufacturer estimated that their typical model development costs are around $200,000/model.
	
	Based on this information, we estimated average costs to develop a new model line, including testing with both crib and cord wood, of $356,250 for certified wood stoves and pellet stoves, as outlined in Table 3.  We also assumed $356,250 for single burn rate stoves, which may be somewhat high if they are as clean as the manufacturers claim.  We likewise assumed a development cost of $356,250 for forced air furnaces and hydronic heaters.   
      
      According to information provided by one masonry heater manufacturer, R&D costs for a masonry heater may be estimated at $250,000.  For our cost analysis, we assumed R&D costs were the same as for other wood heater appliances ($356,250).  For facilities conducting R&D, these costs include the costs for certification testing.  We estimate that the cost of testing a heater design in an EPA accredited lab to be approximately $10,000 for masonry heaters.  


Table 3.  Example of Manufacturers' Estimates of Costs to Develop Model Line (4 Fireboxes)
         
                                Cost Component
                                Estimated Costs
                                     Notes
Salaries
$850,000
Using 2-full time experienced employees to bring the products to market, salaries and benefits are estimated at $160,000 per year for at least 5.3 years. Tasks include design, prototyping, testing, production-line integration, and marketing.
Laboratory Equipment
$50,000
In order to accelerate R&D and avoid validating each result with independent testing labs (too costly for most manufacturers), new testing equipment will need to be purchased in order to sample flue gases, measure test load weight loss, record data automatically, and analyze flue gas composition.
Prototypes
$25,000
Numerous prototypes will be needed until the final product can be approved. For each firebox, an estimated 8 prototypes will be needed, at a cost of $700 each. Numerous samples of various components will also have to be purchased from vendors. 
Test Fuel*
$45,000*
Each test costs at least $50 in fuel (assuming cribs are used), including waste.  And estimated 150 tests will have to be conducted for each firebox for a total of $7,500, or $30,000 for a 4-firebox model line based on crib testing.* 
Testing Services*
$150,000*
Testing services for emissions, efficiency, and safety are estimated to last approximately 3 weeks for each firebox. At an average of $1,500 per day plus travel expenses, this amounts to approximately $25,000 for each firebox, or $100,000 for a 4-firebox model line based on crib testing.*
Outside Consultants
$160,000
The average manufacturer will need outside help for design and testing. Testing equipment, knowledge of the test standard, and general guidance is normally offered by outside consultants (not necessarily certified EPA test labs). The average manufacturer will need approximately 300 hours of consulting services per year ($40,000) for 4 years.
Re-tooling
$120,000
For each firebox, new molds and jigs will need to be purchased or produced. Re-tooling charges will reach an estimated $30,000 per firebox, or $120,000 for a 4-firebox model line.
Marketing
$25,000
New pictures will need to be taken and all the current marketing material, including web sites and owner's manuals, will have to be updated. 
Total
$1,425,000
Equal to $356,250/model
*Note: The costs originally provided by industry for this table were presumed to be based on crib wood testing, not both crib wood and cord wood testing. Therefore we increased the industry-based "Test Fuel" cost by 50% (to the $45,000 shown above) as well as the industry-based "Testing Services" cost by 50% (to the $150,000 shown above) in order to estimate the additional cost to test with both crib wood and cord wood.
	
      
      
      
      
      The masonry heater cost analysis also makes use of a unique software package based on a European masonry heater design standard.  This standard has been verified in the laboratory and under field conditions to produce masonry heaters that would meet the proposed NSPS emission limits.  The software produces for printout a certification for a given design application and the design definition documents, as well as operating instructions customized to the given design.  Thus the software verification and certification record is created for and attached to the design, and the resulting documents can be submitted as part of the certification application.  The cost of this software to the user is approximately 1,000 Euros (approximately $1,500) for the package with a 300 Euro (approximately $450) annual fee that commences in the second year following purchase.  
      
      We assumed that the large manufacturers will use the 6-year period starting at proposal of the draft standards, i.e., 2013 to 2018, to recover the costs of R&D.  The emission testing and software certification costs will be incurred on a real-time basis, as they occur.  We also assume that the small manufacturers will purchase the software package in the year the final rule is promulgated (2014) and continue to renew their license in the following years, in anticipation of the 2019 small manufacturer compliance date.	
      
      B. 	Number of Models
	
	We need to reconcile the data on shipments of individual units with the certification costs associated with model lines.  Therefore, we estimated the average number of models produced by manufacturers of each product type.  For certified wood stoves and fireplace inserts, we used information provided by the Hearth, Patio and Barbecue Association (HPBA) that currently 125 models are being produced on EPA's list of certified stoves.  Because pellet stoves are produced at levels comparable to wood stoves, we estimated that there are up to 125 models being currently produced.  

	The single burn rate estimate of 20 different models is based on an internet search of major manufacturers (United States Stove and Vogelzang).  The actual estimate could be higher if there are a large number of importers of foreign produced units.  However, a search of the Northern Tool and Equipment, Tractor Supply Company, Lowes, Home Depot and Walmart Websites did not identify any other major producers of single burn rate stoves that are sold in the U.S.

	The HPBA provided a list of manufacturers by product type.  Using this as a starting point, we conducted additional internet searches to quantify the numbers of models offered by manufacturer.  We further refined the estimate of hydronic heater models based on EPA's experience with the voluntary program.  We understand that a number of manufacturers list products on their website that are no longer manufactured, only sold in limited quantities, and/or are sold under license to another manufacturer.  Furthermore, one manufacturer dominates industry sales, so on a sales-weighted basis there are fewer numbers of models produced than a strict interpretation of web sites would indicate.  Based on this process, we estimate the following number of models by product type:

      * Forced air furnaces  -   50
      * Hydronic heating systems (outdoor and indoor)  -  120 
         
For masonry heaters, there are three scenarios for potential cost impacts.  In the case of Tulikivi and some U.S. firms, e.g., Timely Construction, these companies have already invested in R&D in order to gain access to U.S. markets which restrict sales (e.g., Colorado) of uncertified units.  These companies will face testing costs only, with an assumed total of nine tests conducted (i.e., a total of nine models certified through testing). We assume that these costs will be spread out over 50 percent of the Tulikivi models and 50 percent of the "other manufacturer" models.  For purposes of our cost analysis, we also assumed that two additional companies will conduct R&D to develop two new models each to meet the proposed NSPS.  These costs will be spread over 50 percent of the "other manufacturer" models.  Finally, we have been told that Tulikivi will use a software certification approach to certify up to eight additional models, which will be spread out over the remaining 50 percent of Tulikivi models.  

We also project that the remaining 15 percent of custom built units will use the software certification approach to demonstrate they comply with the proposed NSPS.  Our estimate of the number of custom built models is based on the average number of models sold per year in the 15 percent model category (i.e., 85 per year). We assumed each custom manufacturer would sell 2 models per year, for a total of 42 manufacturers participating in the software certification option.

      C. 	Model Design Lifespan
	
We assume a 20-year model design lifespan as well as a 20-year use/emitting appliance lifespan.  These assumptions were made to best characterize the actual model design and use lifespans given that many models developed for the 1988 NSPS are still being sold (after 25 years), many "new" models still have the same internal working parts with merely exterior cosmetic changes, and most stoves in consumer homes emit for at least 20 years and often much longer.  Therefore our analysis tracks shipments and costs through year 2038 for the Proposal and through year 2041 for the Alternative (i.e., assuming a 20 year design life for a model meeting the standard in 2019 under the Proposal, and in 2022 under the Alternative approach, as described in more detail in Section IV).
      D. 	Certification Testing
	The costs of certification testing are included in the model development (R&D) costs.  Models that already meet the specified limit will merely need to certify emissions and thus face certification costs only (e.g., pellet stoves facing the 2014 NSPS limit, described in Section IV).  However, for purposes of unit cost increase, the majority of costs are faced by models that require new R&D and increased costs to manufacture, and these costs are modeled in this memorandum for all appliances that need to do more than merely certify to meet an NSPS year limit.

      E. 	Additional Manufacturing Costs
         
	We estimated the following additional manufacturing costs per unit based on appliance type:

      * Certified wood stoves and pellet stoves represent a well-developed technology and we could not identify price differences between models with lower emission levels compared to models with higher emission levels.  Therefore, we have assumed no additional manufacturing costs.
      * One manufacturer estimated that it will cost an average of $100 more to manufacture a lower emitting single burn rate product (exclusive of recovering R&D cost).
      * We have limited information that indicates the cost to produce a forced-air furnace that complies with a Step 1 emission standard (see discussion below in section IV) is approximately 75% more, or $1,600 (exclusive of recovering R&D cost). 
      * We have seen a range of estimates for additional costs to manufacture a hydronic heater capable of meeting the emission standards associated with the EPA voluntary program, with an average being approximately $3,000.  We assume that the additional costs to manufacture an NSPS certified furnace will be comparable, i.e., $3,000 (exclusive of recovering R&D cost).
         
   III. Unit Costs of Certified Wood Heaters

	We used the average 2008 base year costs per appliance type, based on F&S where available, and converted these 2008 costs to 2010 dollars ($) to be consistent with the 2010 R&D costs estimated in Table 3. These costs represent the costs based on manufacturer's price (NOT retail).  For units not included in the F&S report (single burn rate stoves and forced air furnaces), we used manufacturer information.,   Our next step was to develop the following incremental Cost Formula:

(Cost of R&D * Number of models) / (Number of shipments * model life of 20 years),
                                       
which equals the incremental cost of developing a new unit, spread over the number of units expected to be sold during the model life (assuming a flat growth rate of shipments for the purposes of estimating "Number of Shipments" in this formula).  Where there are additional manufacturing costs as discussed above, we added these to the unit cost number.  The total incremental cost was added to the base model cost to derive the post-NSPS unit cost.

Table 4a is a summary of the final unit cost estimates under the Proposal, and Table 4b is a summary of the final unit cost estimates under the Alternative, as reflected by a draft NSPS phase-in schedule, described in section IV.

Table 5 is a summary of the masonry heater unit costs, by type of manufacturer and certification assumption.  (For masonry heaters the Proposal and Alternative do not differ.) Because of the significantly lower cost of emission testing and/or software certification, we did not incorporate a cost recovery period into the equation for these unit price estimates for large manufacturers (like Tulikivi) using software certification.  

We can see from the results that the overall low number of additional R&D masonry heater models, the relatively modest cost of the software certification option and the low number of models requiring additional testing result in incremental costs ranging from $88 to $840 for masonry heaters.  The average cost increase per masonry heater unit is approximately $300.

Table 4a.  Unit Costs of Wood Heaters meeting Proposed NSPS (2010$)


Table 4b.  Unit Costs of Wood Heaters meeting Alternative Standards (2010$)




Table 5. Unit Costs of Masonry Heaters meeting Proposed NSPS or Alternative Standards (2010$)



   IV. NSPS Options
      We developed two primary options to evaluate the cost impacts of implementing new or revised NSPS standards for the residential wood heating appliance industry.  The cost analyses assume specified phase-in schedules, which differ for the Proposal versus the Alternative.  The Proposal represents a scenario where all types of appliances (except masonry heaters) would be required to meet a specified Step 1 emission limit upon promulgation in 2014 and then a stricter Step 2 emission limit five years later in 2019. Under this Proposal, large manufacturers of masonry heaters would also be required to meet a specified Step 1 emission limit upon promulgation in 2014, while small custom manufacturers would not be required to meet this Step 1 emission limit until 2019.  There is no Step 2 emission limit for masonry heaters being proposed.  The Alternative approach represents a scenario where all types of appliances (except masonry heaters) would be required to meet a specified Step 1 emission limit upon promulgation in 2014 and then an interim Alternative Step 2 emission limit three years after promulgation in 2017, followed by a stricter Alternative Step 3 emission limit eight years after promulgation in 2022.  Masonry heaters would be regulated the same way under this Alternative approach as under the Proposal, as explained above.  Note that the Proposal Step 2 emission limit is the same as the Alternative Step 3 emission limit, but the compliance dates differ for this stricter limit (i.e., 2019 versus 2022).  Tables 4a, 4b, and 5 show the resulting unit cost impacts of implementing these options.  These implementation periods represent a tentative schedule and are subject to change. 
       
      It is important to note that the costs represented in the NSPS blocks are for a unit required to make changes to comply with the NSPS.  In reality, not all units will require changes and the related R&D costs.  
      
      Following is a summary of the NSPS implementation assumptions for each appliance type, grouped according to probable NSPS Subparts.  This summary describes the specific emission limits under the Proposal and the Alternative approach.  A discussion regarding the basis for these standards as well as how many existing models of each appliance type already meet the specified limits is provided in the "Estimated Emissions from Wood Heaters" memo.

Subpart AAA ("room heaters"): 
 
      * Adjustable burn rate, single burn rate, and pellet stoves:  Proposal: Step 1 limit of 4.5 g/hr upon promulgation in 2014; and Step 2 limit of 1.3 g/hr five years after promulgation in 2019.  Alternative: Step 1 limit of 4.5 g/hr upon promulgation in 2014; Step 2 limit of 2.5 g/hr three years after promulgation in 2017; and Step 3 limit of 1.3 g/hr eight years after promulgation in 2022.
Subpart QQQQ ("central heaters"):

      * Hydronic heaters (both outdoor and indoor):  Proposal: Step 1 limit of 0.32 lb/mm BTU heat output upon promulgation in 2014; and Step 2 limit of 0.06 lb/mm BTU heat output five years after promulgation in 2019.  Alternative: Step 1 limit of 0.32 lb/mm BTU heat output upon promulgation in 2014; Step 2 limit of 0.15 lb/mm BTU heat output three years after promulgation in 2017; and Step 3 limit of 0.06 lb/mm BTU heat output eight years after promulgation in 2022.
         
      * Forced Air Furnaces:  Proposal: Step 1 limit of 0.93 lb/mm BTU heat output upon promulgation in 2014; and Step 2 limit of 0.06 lb/mm BTU heat output five years after promulgation in 2019.  Alternative: Step 1 limit of 0.93 lb/mm BTU heat output upon promulgation in 2014; Step 2 limit of 0.15 lb/mm BTU heat output three years after promulgation in 2017; and Step 3 limit of 0.06 lb/mm BTU heat output eight years after promulgation in 2022.
Subpart RRRR including masonry heaters:

      * Masonry Heaters:  Proposal / Alternative (same): Step 1 limit of 0.32 lb/mm BTU heat output upon promulgation in 2014 for large manufacturers (defined as manufacturers constructing >= 15 masonry heaters per year), with a 5-year (2019) small volume manufacturer compliance extension (for companies constructing < 15 units/year).  No other phased-in limits are being proposed.
         
   V. Other Cost Information

      A. 	Installation Costs
         
	Although we do not expect any changes in installation costs, they are relevant to the total cost of purchasing a residential wood heating device.  Where possible we used information in the F&S report.  The F& S report notes that installation cost only includes the installation of the appliance and does not include the cost of installing a new chimney, the cost of a blower system or the cost of additional accessories such as mantel and other stone exteriors.  An average installation includes installing floor and wall protection, connecting pipes to the unit using the proper material and construction clearances, ensuring proper ventilation, and testing the unit to make sure it is operating properly.  For single burn rate stoves, we assumed that installation costs would be equal to installation of certified wood stoves.  For forced air furnaces, we assumed an installation cost of $2,500, based on manufacturer information.

      * Certified Wood Stoves:  $500  
      * Single Burn Rate Stoves:  $500
      * Pellets Stoves:  $300  
      * Forced Air Furnaces:  $2,500
      * Masonry Heaters:  $6,000
      * Hydronic Heating Systems (Indoor or Outdoor):  $2,000



      B. 	Price Markups
         
	As described above, the unit costs presented are the manufacturer's price, which is the price of one standard unit and does not include installation, shipping charges, or accessories.  As noted in the F&S report, pricing at the manufacturer's level is used to avoid the price fluctuations that occur once the unit enters the different distribution channels used in the market.  Once again, however, we also have an interest in being able to estimate the cost of a unit to the consumer.  We solicited information from manufacturers on average price markups.  Most of this information is confidential and varies by product type and price point.  However, the following information can be used as a rough guide:

Cost to Manufacture: 0.6 
Producer Cost: 1
Direct Retail markup:  2
Major Retailer markup (e.g., Lowest): 1.5 (volume markup)
	We also looked at information in a related industry, residential gas furnaces and boilers, reviewed by the Department of Energy as it evaluated the impacts of implementing increased efficiency standards.  Following is a summary of mark-ups determined by that study.  The report concluded: 

      The Department used the overall markup to estimate the customer price of baseline equipment, given the manufacturer cost of baseline equipment.  Similarly, the Department used the overall incremental markup to estimate changes in the customer price, given changes in the manufacturer cost above the baseline price resulting from a standard to raise equipment efficiency.  For example, if a standard increases the gas furnace manufacturer cost by $100, the Department multiplied this by the overall incremental markup of 2.05 to estimate that the customer price will increase $205.


     Table 6.  Summary of Markups on Residential Furnaces and Boilers[23]
                                       

                              Baseline Model Cost
                               Incremental Cost
Manufacturer 
                                     1.26
                                     1.26
Wholesaler 
                                     1.36
                                     1.11
Contractor (new/replacement)  
                                   1.41/1.62
                                   1.22/1.33
Builder (new construction only)  
                                     1.43
                                     1.33
Sales tax (replacements only) 
                                     1.07
                                     1.07
Total markup (on manufacturing cost)
Non-weatherized gas furnace
                                     3.12
                                     2.07
Weatherized gas furnace
                                     3.12
                                     2.07
Oil-fired furnace
                                     2.97
                                     1.99
Hot-water gas boiler
                                     2.97
                                     1.99
Hot-water oil-fired boiler
                                     2.97
                                     1.99
Mobile home gas furnace 
                                     2.22
                                     2.22


                                          
