MEMORANDUM
To:		Gil Wood, USEPA
		David Cole, USEPA
From:		Jill Mozier, EC/R Inc.
Date:		November 27, 2012
Subject:	Notes from the November 9, 2012 Synthesis Discussion of the National Educational
      Forum on the Residential Wood Heater NSPS in Minneapolis, MN
____________________________________________________________________________________
Synopsis of Discussion Focus (topics discussed most):
 
Certification Test Procedures:

The Multi-Jurisdictional Organizations (MJOs) and state stakeholders present (together "the states") recommended a two-phase approach to certification test improvements which addresses short-term fixes for this NSPS revision while collecting data for longer-term improvements to be made in the next NSPS revision. The states noted that their primary concern is protection of human health and air quality, and that these concerns are motivating their suggested changes. Among other  changes, the states support: a move to cord wood while allowing either cord or crib wood during a transition period; limiting the time the door is open during start-up  to ensure capture of start-up emissions; requiring more filter pulls to better characterize variations in emissions and reduce loss of semi-volatiles; and eliminating Category 2 and 3 burn rates in favor of 3 runs at the higher emission rate of Category 1 or 4 (for a total of 4 emission measurements with no averaging between Category 1 and 4, and at similar cost) in an effort to better characterize maximum emission rates.

While unanimously supporting a switch to cord wood, the Hearth, Patio, and Barbecue Association (HPBA) industry delegation present ("HPBA delegation") was concerned that there is not enough data to determine the risk to the industry in the near-term and that some states might require cord wood testing rather than just allowing it as an option (a.k.a. "the cord wood off-ramp"). The HPBA delegation was generally opposed to limiting the start-up time during the test, stating that longer start-up burn times are necessary in the real world to start successful fires and EPA should consider longer times for better replication in the lab (or at the very least that 5 minutes already represents worst-case and a shorter time is not practical), and that start-up emissions are already being captured in any event. The HPBA delegation was opposed to frequent filter pulls, claiming that the testing cost would greatly increase and that the imprecision of the test method would be amplified with each additional filter weighing. The HPBA delegation, with some exception, was also generally opposed to the changes in Category burn rates due to a claimed lack of data to gauge risk to the industry in the near-term, and generally opposed to testing stoves in worst case Category 1 or 4 conditions rather than at "sweet spot" Category 2 and 3 burn rates more typical in the real world. The HPBA delegation expressed support for collecting CO measurements, although not for having a CO emission limit in the NSPS proposal (with some exception). All industry stakeholders noted the importance of a bridge to cord wood and also expressed support for ASTM methods, including state participation in the ASTM process. The HPBA delegation noted their concern about the imprecision of the test method and questioned its accuracy relative to the real world (related especially to the use of crib wood and the variable nature of wood burning) and that therefore tightening of the standard based on Method 28 is not defensible in their view and will unnecessarily keep some good stoves off the market. 

Overall, while there was agreement that greater precision and a transition to cord wood is needed; there was not agreement on how to best improve the precision or how/when to make the switch to cord wood. Also, there was disagreement regarding Category burn rates and start-up time changes. The group agreed to continue discussing a way forward in a work group.
 
Level and form of the standards:

The states presented a straw proposal which included combining wood burning devices into 3 categories based on function (room heaters, central heaters, and thermal storage devices), and employing 2 mandatory Bins (with Bin 1 sunsetting into Bin 2 at as yet unspecified years) and an aspirational Bin 3.  All 3 categories in each of the 3 Bins include proposed PM emission standards and caps, as well as proposed efficiency minimums, and to-be-determined CO limits. 

The straw proposal was favorably received by the pellet boiler and catalytic device manufacturers but the HPBA delegation claimed that the emission limits were arbitrary due to their belief that the test method is not precise enough to distinguish 3 Bins and that such lumping and emission limits would shut down some of the industry (e.g., hydronic heater manufacturers). The HPBA counsel argued that the lumping together of functions was not in keeping with the CAA but rather that EPA must group by device commonalities and best demonstrated technologies, that the cost effectiveness analysis would fail for several of the proposed standards, and that the industry would litigate such issues and push for a rigorous defense by EPA. Some industry members opposed efficiency standards, claiming the data does not exist, a full third of the models would be excluded by an efficiency standard (a point disputed by Washington State), and that combustion efficiency should merely be reported on the hang tag which would then allow market driven forces to select the most efficient models over time. 

Notwithstanding the general opposition by the HPBA delegation, some other industry members and citizen group representatives expressed support for the technology driving nature of the states' straw proposal, including the device grouping and even the most stringent Bin 3 option, which some claimed could be a boost to the industry by offering EPA certification of better performing stoves. Regarding categories, some industry members expressed support for the states grouping catalytic with non-catalytic and pellet with wood. Most industry representatives suggested splitting the central heaters group into air-based versus water-based. Overall, there was industry agreement on the concept of using a time-lined binning approach, but disagreement regarding the specific device categories, most of the states' emission numbers, and even the existence of an aspirational Bin 3. The states noted that they were not surprised by industry's reaction to their straw proposal, that these numbers would change if based on cord wood, and that industry had not suggested an alternative to the 50% best performing stove metric (used in Bin 2).  

In-use durability:

Oregon proposed some options for the NSPS and/or for state programs including warranties, incentives for proper consumer upkeep of stoves including modular/replaceable parts, tests by manufacturers to simulate aging, decreasing the ability (via stove design) for consumers to operate the stove improperly, and perhaps an auto-shutdown of the stove if it is out of specification (e.g., O2 sensors). In general, Oregon proposed that EPA/states and the manufacturers work together on this issue of shared concern. The HPBA delegation's general response was that in-use durability regulations were not necessary because manufacturers were already taking such steps to protect their own interests (against warranty claims and bad reputations), and that in any event the states were in a better position to control consumer behavior in this regard. The states expressed an interest in pursuing the durability issues further in a work group.

Participants & Observers of Synthesis Discussion:
                               Participant Name
                                 Organization
Andy Ginsberg
Oregon Department of Environmental Quality
Arthur	Marin
NESCAUM
Bob Ferguson
Ferguson, Andors & Company (consultant to HPBA)
Carol Cenci
PSCAA
Chris Neufeld
Blaze King/Travis Industries
Dan Johnson
WESTAR
David Thornton
MPCA
David Menotti
Crowell & Moring (counsel to HPBA)
Dick Valentinetti
Air Quality, DEC, Vermont
Dutch Dresser
Maine Energy Systems
Eric Merchant
Montana DEQ
Gil Wood
US EPA
Gregory Green
US EPA
Jim Hodina
Linn County Air Quality Division, Iowa
John Ackerly
Alliance for Green Heat
John Crouch
HPBA
John Dupree
USEPA
Larry Sorrels
USEPA
Lisa Rector
NESCAUM
Martin Lunde
Garn Inc.
Nathan Russell
NYSERDA
Rick Curkeet
Intertek Testing Services NA, Inc.
Rod Tinnemore
Washington State Department of Ecology
Tom Morrissey
Woodstock Soapstone Company
Tomas Carbonell
Environmental Defense Fund
                                 Observer Name
                                 Organization
Beth Friedman
EC/R Inc
Bob Lebens
WESTAR
Jill Mozier
EC/R Inc
Lisa Herschberger
Minnesota Pollution Control Agency
Nancy Kruger
National Association of Clean Air Agencies (NACAA)
Rachel Sakata
Oregon Department of Environmental Quality
Rob Kaleel
LADCO
Susan Wierman
MARAMA

Discussion Details:

Goals and Process for the Meeting:

Dan Johnson of WESTAR opened the synthesis meeting by expressing that representatives from the various state Multi-Jurisdictional Organizations (MJOs) were at the forum to learn so that the representatives could better inform their respective MJOs on the complexity of the issues involved in the woodstove NSPS. He hoped that by the end of the forum the MJOs would understand everyone's position  -  consensus/agreement would be ideal  -  but realistically consensus may not be possible. In response to EPA confirming that a draft proposal to EPA management by the end of the calendar year was the goal, Dan said that such a date would be "light speed" for the MJOs but he would nonetheless make recommendations to his members, and furthermore requested comments in the on-line "Basecamp" discussion by December 9, 2012.  One week after December 9, the MJOs would craft a letter to EPA. 

EPA clarified that they would like to complete by mid-January those things in the current draft woodstove NSPS that could be changed quickly, in order to get a new draft proposal to OMB by February or March.

Note: ** below indicates an attempt at summarizing points of agreement and/or areas where more detailed discussion to reach an understanding is needed.

Certification Test Procedures:

Rod Tinnemore of Washington State Department of Ecology (ECY) explained that the states recommend a two-phase approach to certification test improvements involving a short-term fix for this NSPS with additional revisions for the next NSPS revisions (hopefully no more than 8 years away). The goals are:
 Improve precision;
 Limit "gaming" the test;
 Eliminate elements that don't add value;
 Characterize maximum emission rate; and
 Obtain information to inform new test method development and potentially new emission limits.

Regarding efficiency, there seems to be general agreement to better define which pieces of CSA B415 to use (e.g., high heating value (hhv) versus low heating value (lhv)) and that the lack of low burn rate in B415 needs to be addressed.

Regarding the test method, the states propose:
 Requiring a filter pull every 60 minutes;
 Requiring that filter weight be measured and reported at 12 hours, 24 hours, and 48 hours due to concern about potential loss of VOCs during "drying."

Regarding device operation changes, the states propose:
 Move to cord wood or allow both crib- and cord wood during a transition period;
 Report density of wood instead of wood species;
 Limit start-up time so that start-up emissions are not missed;
 Base coal bed weight on % of fuel load with +-1% deviation;
 Eliminate Category 2 and 3 burn rates and require only Categories 1 and 4, perform 3 runs at the highest emission rate (in Cat 1 or Cat 4), and set emission standard for each Category with no averaging between Categories.
 For hydronic heaters, determine actual max heat output by test; and
 For woodstoves, willing to consider change from max of <1 kg/hr to < 1.15 kg/hr for low burn rate.

David Menotti representing HPBA commented that the (above) basic ideas are sound regarding a short-term versus long-term fix, but cautioned that tweaks can't be so substantial as to remove EPA from the available database  -  that some of these elements raise concern in this respect. Gil Wood of EPA clarified that this is a 2-phase process being proposed and that the additional data (e.g., from the 60 minute filter pulls) are not for this NSPS revision. HPBA is concerned about dropping Categories 2 and 3 and replicating test runs on Categories 1 and 4, because limited data is there yet  -  therefore the category revision should be a 2[nd] tier change not 1[st] tier. When EPA asked in what way the data was missing, HPBA responded that they haven't tested the stoves with the new proposed approach, with replicate runs and without weighting Category 1 most heavily  -  that while they do have Category 1 and Category 4 data, industry does not know what would result from the 3 proposed replicate test runs ("which puts all the risk on industry").

Chris Neufeld of Blaze King asked that if we're trying to represent real-world burning and if the consumer uses stove in low or medium-low burn rate 80% of time, why eliminate this "sweet spot" representing most of the real-world use?

Andy Ginsburg of the Oregon Department of Environmental Quality (DEQ) responded that from a public health perspective there is concern about short-term high emissions that will occur. So from a public health perspective, stoves should be designed so that they're engineered to be controlled under the worst operating mode  -  hence the focus on Categories 1 and 4. Minimizing emissions in the worst mode may also offer clues to oversizing.

Rod Tinnemore commented that this would also put all manufacturers at an "equal disadvantage," so for example catalytic and non-catalytic stoves are treated equally with no unfair advantage. 

Rick Curkeet of Intertek Testing commented that emissions in g/hr are different than in g/heat-delivered  -  that actually the highest burn rate is not the cleanest but has the highest g/hr  -  so units need to be discussed.

Arthur Marin and Lisa Rector of NESCAUM responded that they don't assume the highest burn rate is the cleanest, but they parsed data by lb/MMbtu and by g/hr and hence propose testing Categories 1 and 4 where worst emissions happen.  Furthermore, they commented that the NSPS is not about averaging but rather supposed to be computing maximum emission rates.

Dan Henry of 5G3 Consulting commented that it's a mistake to assume that if stove is controlled under Categories 1 and 4 then Categories 2 and 3 will fall into place.  Tuning the stove to air:fuel ratios is complex and manufacturers would have to go back and ensure that Categories 2 and 3 were not messed up by the design focused on Categories 1 and 4. So EPA shouldn't just look at Cats 1 and 4 and leave out most typical burn rates. Furthermore, regarding right sizing, the average consumer has been educated that bigger isn't better and stove manuals focus on right sizing and climate.

Andy Ginsburg asked industry if, with the proposal for separate Category 1 and 4 standards (no averaging between Cat 1 and Cat 4), would they also like separate Category 2 and 3 standards?

David Menotti responded that the discussion was mixing standard and test method issues, and it may be better to average and cap emissions. John Crouch of HPBA further commented that filter pulls every 60 minutes will increase testing cost.  Bob Ferguson representing HPBA commented that having more than 1 test run is probably good for increasing precision, but HPBA needs to discuss weighting impact on level of standard. Bob Ferguson commented that if three runs are done in 1 category, then the standard should not be a weighted average, as it would be too heavily weighted on low end.

Lisa Rector commented that there is only 1 data point for 4 different burn rates in the current test method. She believes that the test method ideally needs to go to N of 3 for each burn category, but it is unrealistic to ask for 12 runs (that is, 3 runs for each of 4 tests). So the states' proposal is to improve the precision but not increase the cost, and no longer use a weighted average for the standard.

Dutch Dresser of Maine Energy Systems commented that for pellet boilers he likes the different standards based on different burn rates, but pellet stoves heaters should not be tested down in a dirty range where they aren't designed to perform anyway.

** For group, it was noted that there appears to be consensus that greater precision is needed but all don't necessarily agree on burn rates. There is some possible agreement regarding different standards for different burn rates.

Dan Henry commented that the ASTM method approach of 40% low/ 40% medium/ 20% high for weighted average of 3 runs is comparable for most EPA-certified stoves (very little difference with current weighting of 4 runs), so it is possible to do something with the spare 4[th] run. 

Bob Ferguson commented that the impacts of a change like this would need to be analyzed, because it would be getting away from the current database. Changing something that's been done this way for 25 years carries significant risk to industry. Instead, all stakeholders should share the risk. 

Andy Ginsburg commented that, regarding short-term versus long-term goals, he recognizes that an incremental process is required to improve test precision to help discuss more stringent standards  -  agrees there is a tension here  -  but that there is risk of better test method to states and industry alike. Nonetheless, the test method needs improvement. 

Bob Ferguson agreed with this, but commented that it's too late to change the test method "5 minutes before EPA goes to OMB".

John Ackerly of Green Heat Alliance expressed support for comments by Dutch Dresser regarding woodstoves using oxygen sensors and that test methods should accommodate this. There needs to be an alternative test method so there's no barrier to innovation.

Tom Morrissey of Woodstock Soapstone commented that it would be less radical a change to have 4 tests: high, low, another test at whatever produced worst emissions, then last test at combined Categories 2 and 3. Weighted averages are meaningless to consumers, so he likes new approach. Need to publish burn rate emissions in a transparent way  -  that is, 3 columns of numbers not just 1 weighted average. Then consumer can decide which column is most applicable to them and on which to base purchasing decision. This will trigger innovation.

Gil Wood commented that EPA had previously considered using approach of 3 runs plus extra run at "worst case", similar to the fireplace program. Lisa Rector commented that N of 3 is crucial. Gil Wood further commented that 3 runs at each burn rate is being considered now.

Rick Curkeet commented that product performance precision is what is being discussed, but that there is already a provision to throw out outliers and re-run. So what would happen to outlier provision? What would count and what wouldn't? Burning wood is unpredictable. EPA asked how many times outlier provision was used, but the response from Rick and from Bob Ferguson was that this data is not available.

** Dan Johnson commented that regarding the burn category discussion, the states have straw ideas, not a firm proposal. The discussion should not be closed, if there are better ideas. There is agreement on need for a more precise test method and that this new method not be much more expensive. Rod Tinnemore and Lisa Rector have proposed a more precise test. If HPBA disagrees with this test, then what should the test look like?  Greg Green of EPA added that EPA wants to address test method precision and would like HPBA/industry to offer thoughts.

Tom Morrissey commented that labs never run test without multiple filters  -  so it's not a multiple filter issue, but rather about number of runs. It can already be determined where max emissions occur.

Andy Ginsburg confirmed for John Crouch that yes, the states are suggesting doing away with weighted averages.

Dan Henry commented that the 5G emission measurement method has been agreed upon unanimously and that his experience when running 5G dual train filters is that it's impossible to capture enough on filters to weigh catch. A dual filter dual sample train already has audit features. Imprecision will be increased by additional weighing and will require someone weighing non-stop (including tare weights and sample weights). This will double or triple the cost. 

Rick Curkeet added that this is not specific to woodstoves but also applies to hydronic heaters. The first catch is significant and uncertainty increases when trying to weigh multiple filters with smaller catches on each. He recommended looking at CO when trying to determine peak emissions, rather than doing multiple runs.

Lisa Rector asked if testing was needed over full charge, and if a shorter test was possible in next round (not this round). Not only should precision be improved, but data should be collected for future test.

Rick Curkeet commented that a full test is needed for efficiency anyway. CO and CO2 should be measured, and that to partition particulate one could look at area under CO curve (e.g., what % of area comes in first (1/2) hour) so that we can get to what we're looking for in a less expensive way, with data already available. Dan Henry further commented that filters are not run during thousands of product development tests  -  rather industry knows that opacity-free is achieved when the ratio of O2:CO hits 10:1. The higher that ratio is, the cleaner the burn. If that occurs in 15 to 30 minutes, industry knows it's a successful test and will run to the end to determine burn rate. A ratio of 200:1 is achieved near end because CO drops off so much.

Bob Ferguson commented that all stakeholders want data and generating data for the future is a good thing but there are no available short cuts in certification labs. Multiple runs are not necessary, as the point of max emission is well known (e.g., in 1[st] hour for woodstoves, when transitioning from on to off for hydronic heaters). Should not deal with dozens of sets of filters; more pragmatic solution exists.

Dick Valentinetti of Air Quality, Vermont DEC, asked if industry was suggesting no particulate standard but only a CO standard. Dan Henry responded that if the data were mined, one would find that CO and PM track perfectly. Dick noted that he's most concerned with consistency and if CO is a better way, then it should be considered. David Menotti commented that the precision issue is most profound due to the inherent variability of burning wood, and that Dick's CO idea is too bold a change at this point.

Lisa Rector noted the states' concern about using filters too long (potential loss of semi-volatiles). HPBA responded that too many filters are needed for once-per-hour pulls. Lisa responded that perhaps labs could do several pulls in the first hour and then at max duration. Bob Ferguson noted that they aren't concerned about the primary filter because there is a secondary filter, but in any event it's been shown that the primary filter catches all. Bob further noted that filter pulls and weighing need to be minimized due to increase of imprecision (he likes single filter catch). Lisa noted that some of PM evaporates and back-up filter doesn't catch that loss  -  some semi-volatiles will be lost on filter so 12-hr, 24-hr, 48-hr measurements will help determine that loss (without that accounting cleaner units will be penalized because semi-volatiles come off dirty units). Bob noted that the method starts with a dry (desiccated) filter  -  that a wet filter catch is not possible since much of the weight in burning is from moisture. Lisa recommended continuing this discussion with George Allen.

David Thornton of Minnesota PCA noted that CO data is not available now, so maybe use for Phase 2.

Andy Ginsburg commented that enough short-term samples should be collected for the future standard, but that such data should also be used for the current NSPS in some way  -  that a combination approach is needed. Hydronic heaters are a concern and a test method is needed that captures variability (likewise for pellet stoves). Gil Wood noted that this is a proposal, not a final rule being crafted and EPA will explain all issues and illustrate potential multiple options in the proposed rule to allow fine tuning for the final rule. 

Gil Wood asked Rick Curkeet whether all labs collect CO. Rick responded that his lab has CO data, but not necessarily all labs do since CO is not currently required.

David Menotti commented that there is not a robust CO data set, so HPBA is not ready to switch to CO standard. Bob Ferguson further noted that CO and particulate don't track perfectly, although there are strong trends.

John Ackerly asked if EPA could be transparent during this time before sending to OMB. Gil noted that EPA will be as transparent as possible and explained about 90 day public comment period.

** There appears to be general agreement about a two-tiered approach to get to bolder steps eventually and that a work group will continue to discuss details and areas of disagreement. John Crouch commented that tweaks should not add to cost. Andy Ginsburg commented that with this two-tier approach tweaks could be made in the short term that will improve precision and obtain short-term emission data (taking into account good points made about pulling more samples) for the next NSPS  -  thus tweaks should inform the current NSPS revision while gathering data for future NSPS revisions.

Lisa Rector asked for feedback on low burn rate. David Menotti commented that the most important issue for industry is making a bridge to cord wood. Andy Ginsburg commented that limiting start-up time is important. Bob Ferguson commented that all start-up emissions are captured in testing.

Regarding low burn rate, Dan Henry commented that when you force a stove to be built such that it can operate at < 1 kg/hr under non-catalytic technology, then by design you are giving consumers license to smolder. The 1988 NSPS requires a non-removable stop at its lowest burn rate, e.g., at the 1 kg/hr position. Raising the minimum burn rate would reduce consumer-caused stove smoldering, allow arrival at the sweet spot quicker, and also address stove forgiveness for wet wood. If a model can burn clean at a lower setting, e.g., many catalytic stoves, then still allow it for that model.

Martin Lunde of GARN noted that movement to cord wood is the single best way to replicate the real world.

David Menotti gave history of crib wood testing noting that industry agreed to crib because the thought was there'd be better precision; but there isn't. So industry has been designing stoves around crib wood, not real-world cord wood  -  and SIP planners are not even using crib test data, but rather are using cord wood data. The risk is too high to industry to switch immediately after so many years of designing around crib wood, so a certification off-ramp is needed  -  that is, manufacturers should be given the option (not the requirement) to test with cord- instead of crib wood to determine the passing grade. There's currently an ASTM test method for cordwood but it needs more refinement.

** There is general agreement about transitioning from crib- to cord wood, but HPBA comments that the devil is in the details especially with limited database. There is agreement to discuss the transition from crib- to cord wood more off-line in a work group.

David Menotti comments further regarding the risk to industry of switching to cord wood  -  that is, that some states will see the optional off-ramp and make this their more stringent standard. Industry approached EPA for a consistent standard in 1988 because industry could not tolerate different state standards. Thus industry would like to negotiate with states now and ensure cord wood testing is not a deal breaker. Arthur Marin noted that they cannot commit that every state/local government won't use the authority granted to them in the CAA to be more stringent. David responded that industry is not looking for a sue-able contract but is rather looking for good faith understanding. Dick Valentinetti noted that half of the states cannot be more stringent than the federal government/EPA anyway, and although he cannot speak for New York, he thinks he can bring the other states to a good faith understanding. David noted again that industry doesn't want to be sandbagged.

Greg Green commented that perhaps the first step is to agree on a range of options for proposal  -  that the options don't need to be nailed down yet, rather give EPA options.

Dan Henry commented that at every conference European technology is extolled and yet Europe doesn't use cribs nor require testing at minimum burn rates and their tests are run exactly as burned in a typical home. Consequently European stoves work in lab just like they work in the home. If the US goes to cord wood, we need to replicate the real world  -  stoves operate at different burn rates  -  it is what it is. That's what wrong with the US standard today.  

Andy Ginsberg noted that Austria has database and NYSERDA has paid for the database. Can we use this database? Bob Ferguson responded that European models often have to be re-designed to meet US test method standards  -  the test methods are not comparable. Gil Wood noted that the best European hydronic heaters are typically designed to avoid operation at the low burn rates and poor combustion conditions that have been so common and so problematic for US hydronic heaters.

Regarding start-up time and Paul Tiegs' (now deceased) contentions, Bob Ferguson noted that stoves should be run for a full hour on pre-test load before loaded for the test so that the coal bed is fully charcoalized. Bob Ferguson noted that no emissions should be missing (i.e., not captured by testing) during the start-up.

Andy Ginsburg commented that the issue is how consumers, who are not necessarily as skilled as manufacturers, will use stove (e.g., all about getting hot fast).  Too much optimization will measure too sophisticated a start-up than is typically practiced by consumers.

Bob Ferguson commented that the manuals tell wood stove consumers to burn for >15 minutes for start-up, so 5 minutes is a reasonable worst-case. Very short start-up times by consumers are not wanted by industry. In addition, limiting the start-up time further (less than 5 minutes) is practically difficult to accomplish in terms of how long it takes to load. Finally, Bob doesn't see improvements from changing the 5 minute start-up time as this is overwhelmed by the random nature of burning wood. Bob and Rod Tinnemore agree to discuss this topic more off-line.

Chris Neufeld commented that he discussed with Paul Tiegs the point that how long the door is open affects the interior temperature, and that limiting how long the door is open during fuel loading decreases the interior temperature. Chris himself keeps his personal wood stove door open < 60 seconds but nonetheless agrees that 5 minutes is a good worst case real world estimate.

Rick Curkeet commented that Europe assumes consumers follow the manual. The door is one issue, but air control is more important  -  operators need to let the start-up wood burn for 10 to 15 minutes before adjusting the air control.  Five minutes is barely on the edge of getting a fire going; many stoves will not even light in 5 minutes. This is a totally arbitrary number; start-up time needs to be increased.

** There is an agreement to move start-up time discussion off-line.

Greg Green commented that everyone present has probably gotten an appreciation of industry's knowledge regarding the test methods. Thinking outside of the box for discussion purposes, what if EPA were to propose an X kg/hr operating standard that 8 out of 10 stoves must meet and then allow industry to determine test method? (In other words, start with the standard and develop test method around the standard to ensure device operates at that level  -  this would take pressure off test method.) David Menotti responded that this turns standard setting on its head and is potentially a suicide pact for industry  -  data from testing is needed to set standard, in his opinion.

** Dan Johnson concluded that stakeholders need to re-visit real-world discussion. The states are being informed today and a work group needs to drill down further on specific test method issues. Dan asked if there were any other issues regarding test methods. David Menotti responded that the work group should be open-ended so that issues may be brought back to it, and also that state participation is needed in the ASTM process. Industry and ASTM have created 5 different test methods, and states' lack of participation needs resolution. Rod Tinnemore and Lisa Rector confirmed for Dan that this request by industry for state participation in the ASTM process is not a new issue.

NSPS Proficiency Test Data and Certification Values versus Real World Data

Presented by Carole Cenci of the Puget Sound Clean Air Agency (PSCAA), who noted that HBPA gave data to PSCAA.

Carole noted with help of graph that proficiency data is not normally distributed based on fact that regression was not linear (her presentation graph shows burn rate weighted emission in g/hr on y axis versus EPA certification value in g/hr on x axis). Carole further noted that lab deviation seems consistent across all stoves and years indicating that there is variability but not bias  -  i.e., test results are not random and variability not linked to stove or year. Deviation from lab mean is +-1.8 g/hr per test.

David Menotti commented that this is the first time he is seeing this graph, but it seems to support a +- 3 g/hr deviation. Lisa Rector responded however that it's unclear which data is based on Category 1, 2, 3 and 4 burn rates and so these individual test runs are attempting to compare apples and oranges. David commented that standards are not based on means but rather on 95% confidence intervals.

Gil Wood agrees that discussion of this graph is too technical for purposes of today's meeting and will follow up with EPA statistician. 

Dan Henry commented the point is that from one test to the next, the variability has a wide spread. David Menotti commented that the graph indicates the random nature of burning wood and that there is not much sense in tightening standard based on Method 28. Rick Curkeet commented that the spread in any one of 4 data points is +-3 g/hr. 

Carole Cenci continued presentation and commented that analysis shows that real world data versus EPA certified values are non-normally distributed (not linear) but rather are lognormally distributed. Thus, lognormal statistical tools are needed rather than normal statistical tools.

David Menotti commented regarding two types of data: in-home and (Dennis Jasman's) lab which simulated in-home conditions in lab but didn't use Method 28. The burn rates of real world data don't conform with Method 28 and are binned into low, medium, and high certification values with linear emission factors. David noted that this analysis points to a seminal issue.

Lisa Rector responded that there are huge issues with such field studies to the extent that the comparison is akin to comparing apples and cars (e.g., whole host of methods used in generating real world data).

Carole Cenci concluded that the analysis shows labs can differentiate between good, better, and best.

Level and Form of Standards

Overview presented by Lisa Rector who noted that uniform standards are needed for devices with common purposes, and that variable timelines can be added using a binned approach (of good, better, best bins). Thus:
 Collapse all devices into 3 basic categories for emission standards:
         o Room heaters
         o Central heating devices
         o Thermal storage devices
 Employ a "bin" concept for emission standards
 CO emissions
         o Modeling analysis using EPA emission rate data shows that Phase 2 HHs can create exceedances of CO standard (analysis using AERMOD showed CO NAAQS was exceeded 80% of time, 50 meters away from HH stack).
            
States' Straw Proposal for emission levels:
                      Bin Concept for Emission Standards
(emission rates based on current weighted averages, will change if just Cat 1 and Cat4)
                              Category Standards
                                     Bin 1
Transitional effective upon promulgation with sunset dates consistent with current draft rule
                                     Bin 2
              Based on average emission rate of top 50% of units
                                     Bin 3
            Alternative average emission rate for top 12% of units
Room Heater
 PM

 CO
 Efficiency

 < 4.0 g/hr; no run > 7.5 g/hr
 CO TBD
 70%

 < 2.5 g/hr; no run > 4.5 g/hr
 CO TBD  -  Cat 4
 75%

 < 1.3 g/hr; no run > 2.5 g/hr
 CO TBD 
 75%
Central Heating
 PM


 CO
 Efficiency

 < 0.32 lb/MMbtu; no run > 18 g/hr
 CO TBD
 70%

 < 0.15 lb/MMbtu; no run > 7.5 g/hr
 CO TBD
 75%

 < 0.06 lb/MMbtu; no run > 3.0 g/hr
 CO TBD
 75%
Thermal Storage
 PM


 CO
 Efficiency

 < 0.32 lb/MMbtu; no run > 18 g/hr
 CO TBD
 70%

 < 0.15 lb/MMbtu; no run > 7.5 g/hr
 CO TBD
 75%

 < 0.06 lb/MMbtu; no run > 3.0 g/hr
 CO TBD
 75%
Note that Bins 1 and 2 (crib or cord) are mandatory; Bin 3 (cord) is aspirational. 

David Menotti commented that category standards are not the basis according to CAA, rather supposed to look at best technologies and start with commonalities of devices. Including catalytic, non-catalytic, pellet and utility / single burn rate stoves (which are new and have no data basis) is akin to including apples, oranges and pears. The numbers are being force-fed.

Gil Wood noted that EPA has always had leeway / discretion with categories and sub-categorization. EPA can legally approach it this way. David Menotti responded that justification for sub-categorization in past has been different best demonstrated technology (BDT), not functionality.

Andy Ginsburg commented that for the states the fundamental function of being a room heater is what's important and industry can compete to make that solution happen. The distinction is the function  -  all are using wood and all are heating rooms, so all should compete on an equal footing. The need is to drive device technology based on standard. 

Chris Neufeld commented that catalytic industry wants equal playing field so catalytic industry favors elimination of separate categories.

Martin Lunde confirmed with Lisa Rector that it was based on loadside heat output measurements and crib data. Martin commented that since it was done with crib wood the data is questionable.

David Menotti commented that it's arbitrary to achieve these numbers given the precision issues, so numbers will eliminate some better performing devices in real world unnecessarily. If optimization of combustion efficiency is wanted, then choose CO or PM and not also efficiency. David is completely opposed to efficiency standards because data doesn't exist for this; i.e., it's been demonstrated that 1/3 of wood stove models would be excluded by this efficiency standard and the cost effectiveness analysis doesn't support requiring new designs, according to David.

Rod Tinnemore noted that HPBA must be using a different dataset because his data has 67 data points and those units that met 70% efficiency didn't exclude a third. David Menotti responded that only EPA or CSA weighted average reference method data counts.

Lisa Rector noted that based on CO data for central heaters there is not a 1:1 correlation for PM reduced and CO reduced. States cannot have significant CO emissions in background (while regulating PM).

John Ackerly commented that these standards could be good for industry and engender confidence in consumers. He can see Bins 2 and 3 as akin to Energy Star ratings. It could be great for industry to let consumers know that EPA certifies these units at these better performance levels.

Martin Lunde commented that 3 bins are not problematic, but the emission standards proposed are problematic. He also noted that he does not trust crib data compared to cord wood data. Lisa Rector responded that if you have alternative data that can be binned, the states are open to using it. Martin wondered if a bad protocol is better than no protocol.

Dan Henry commented that with these proposed emission standards "the industry would be wiped out."

David Menotti commented that if EPA goes with Bin 3 he will insist on rigorous justification for cost effectiveness and demonstrated technologies.

Andy Ginsberg noted that earlier David Menotti expressed concern that some states will want to be stricter than the standards. An aspirational standard such as Bin 3 will serve industry's purpose of reigning in the states. There is some benefit here in creating demand, but Andy agrees that the data needs to be made more robust.

Tom Morrissey said that conceptually the binning is great and asked if Bin 3 is a fixed/proposed standard. (The answer given by the states was  -  no; it's voluntary.) In response to Tom Morrissey's question regarding how Bin 3 would appear in the rule, Gil Wood explained that it would be listed in the rule and would be described in the preamble that manufacturers could decide at which level to certify their models. 

Gil Wood clarified that rather than "voluntary," the correct terminology is "alternative means of emission limitation." Both Bins 1 and 2 would be mandatory (with Bin 1 sunsetting) and thus would require full analysis. But Bin 3 as an alternative means (non-mandatory) would not need a cost effectiveness analysis since the NSPS would not be making industry meet these numbers. OGC has already reviewed this binning concept and confirmed that it is legal for NSPS.

Andy Ginsburg commented that no organization or state can control what other states will do, and no one is suggesting that Bin 3 be mandatory. But in the interest of full disclosure, there will be a Bin [3] out there which some states would like to see supported with tax credits and change-out money if at all possible.

In response to John Crouch's question, Gil Wood confirmed that EPA is prepared to "mix" forced air furnaces with atmospherically vented designs (i.e., category standards according to function not design).

Andy Ginsburg responded to industry question regarding whether the draft proposed standard was optimizing around crib wood by explaining that initially there would be optimizing around crib wood, but Bin 3 could be certified based on cord wood or a Bin 4 could be added for cord wood. 

Chris Neufeld commented that the format is what the catalytic stove manufacturers are looking for but wanted to confirm that the actual numbers proposed are merely placeholders  -  that the draft NSPS numbers could be different. Lisa Rector responded that yes, the numbers could change, but the format would still be good, better, best. Andy Ginsburg further responded that the state authors have some justification/methodology for the numbers but are open to discussion regarding them. Andy further commented that there is the idea that Bin 1 would transition into Bin 2 in a certain number of years (number of years not pinned down yet).

In response to Martin Lunde's question regarding does industry need to test with crib wood, Lisa Rector commented that Bins are based on crib- or cord wood, although maybe Bin 3 could just be cord wood. Gil Wood expressed staff level agreement with Lisa's response although he was careful to note that he is not the EPA Administrator who will make the ultimate decision.

John Crouch commented that this proposal will have big small business impacts and will have to pass the Small Business Regulatory Enforcement Fairness Act (SBREFA). Gil Wood noted that EPA has already gone through the SBREFA process and received input that is incorporated into EPA's decision making.

In response to Greg Green's clarification question regarding that industry's concern appears to be the categorization rather than the binning concept, David Menotti commented that the concern regards improper lumping of devices and the emission standards in each bin, as well as concern over Bin 3. Industry does not have a problem with phasing-in standards like was done in 1988, and industry does like the idea of 0.32 lb/MMbtu being "Phase 1", which he noted was the original draft proposal, not the later co-proposal with 0.15.

Dan Henry commented that in the Klamath Valley study an EPA stove at 2.1 g/hr was the best performer based on crib wood (which would be ~3.0 g/hr in real world). The 3100F certification values were improved by improved air:fuel ratio, but the stove does more poorly burning cord wood. Dan's point is that technology should drive standards, but these draft proposal numbers are driving industry in the opposite direction [from where the states and EPA should want it to go for reduced real world emissions]. When asked by Andy Ginsburg if this is also true for [numbers if based on] cord wood, Dan responded that a bigger number [standard] is need for cord wood as well as incentives to get to cord wood. A bridge needs to be built because industry cannot get to 2.5 g/hr based on cord wood [without changing air-fuel ratios], and industry needs 7 years for the return-on-investment (ROI), according to Dan.

David Menotti commented that in the central heaters category (of state straw proposal) there is an inappropriate lumping of forced air furnaces with hydronic heaters. There is hardly any data set for forced air furnaces (e.g., value of 0.92 lb/MMbtu not robust). Furthermore, there is no database for hydronic heater efficiency  -  EPA has disqualified database on efficiency, according to David. Currently David knows of one cycling hydronic heater unit that meets the 0.15lb/MMbtu. The hydronic heaters cannot meet this standard and the HH industry will be shut down by these numbers (which is upsetting to David because he himself pushed the industry to do better and they have made remarkable improvements including reducing emissions by 90%). Gil Wood responded that 10% of a huge number is still a huge number, and people are being impacted by HH emissions. 

In response to a question from Gil Wood regarding whether the straw proposal numbers are achievable by forced air furnaces, Chris Neufeld commented that a forced air furnace model was at 3.8 g/hr with 83% or 84% efficiency based on B415.1.  So the technology exists, but he cannot comment precisely how this would translate to lb/MMbtu. (Gil converted number and it's better than 0.32 lb/MMbtu.) 

Bob Ferguson noted that a Canadian woodstove model with an air jacket would not meet 0.32 lb/MMbtu.

In response to industry claim that the data doesn't exist to support the standard, Dick Valentinetti commented that this has been discussed for two years, so he is wondering where industry's data is. Dick further commented that credible personal air monitors are currently in development and will be ready for widespread use in about 3 to 4 years [at which time the pollutant levels near high-emitting devices will be deemed completely unacceptable by public]. Regarding lack of data, Bob Ferguson commented that industry does not have more forced air furnace data because the Canadian Standards Association (CSA) standards went into effect in spring 2010 and manufacturers have been working to bring products into compliance with CSA  -  the products have not been finalized yet, thus there's no data yet.

David Menotti commented that it's okay to regulate forced air furnaces but not with Phase 1 hydronic heaters because the data doesn't exist, since there was no voluntary program for forced air furnaces.

Andy Ginsburg commented that no central heaters or hydronic heaters can meet Oregon standard. These devices are competing against natural gas units which are hundreds or thousands times cleaner. Oregon cannot justify going from clean to dirty. Fifty percent of states have to accept federal standards, but other states may be more stringent (e.g., NY). David Menotti responded that industry knows it needs to accommodate the states, but the states should understand that these proposed numbers will shut down the industry. Andy clarified that Oregon's goal is to protect air quality and to drive technology in this direction  -  as such, Bin 1 will not fly in Oregon.

Dutch Dresser commented that he represents an industry [pellet central heating systems], categorized as central heaters under the states' proposal, that can already meet the Bin 3 standard.

** Dan Johnson confirmed with industry that numbers aside, at least a 2-Bin approach (if not 3) makes sense. A remaining question then is how to determine numbers  -  for example, in the Bin 2 straw proposal, does it make sense to look at the average of the current best 50%? David Menotti responded that the precision problem is more fundamental, that without confidence in the numbers the imprecision is being perpetuated. Dan Henry commented that more discussion time is needed  -  for example, we don't have answer regarding how long to get from Bin1 to Bin 2. Dan Johnson responded that Bin 3 will not be mandatory, but that the states accept the angst from industry nonetheless, and also recognize industry's significant concern about grouping. David Menotti commented about the "wild card" nature of the proposal, that there must be a cost effectiveness analysis, and that Bin 2 will not pass that CE analysis. Dan Johnson further questioned that if we can get to cost effectiveness, is the 50% conceptually good? Bob Ferguson cautioned that industry reps at meeting are not able to accurately represent all of their members' interests because the members aren't yet aware of these proposals and haven't all met  -  HPBA will need input from their members [before answering such questions]. Andy Ginsburg responded that that's what the public comment period is for.

Dan Henry commented regarding the voluntary low mass fireplace program having an original limit of 7.1 g/kg which was good for 5 years and then dropped to 5.2 g/kg, but EPA created 3[rd] category at 3.1 g/kg  -  which killed the program. Industry had not enough time from 7.1 to 5.2 to 3[rd] category, and no state adopted it. So Bin 3 will become default for states that want a lower standard.

** Moderator noted that timing of phasing in of Bins needs to be considered.

Regarding categories, Martin Lunde commented that air-based systems versus water-based systems have very basic property and engineering differences, and the proposal should therefore split the central heater category into air- and water-based. The HPBA delegation agreed.  Martin further responded to Lisa Rector that yes, thermal storage should remain in its own category [or at least have separate test method]. Rod Tinnemore noted that states are looking for ways to simplify by lumping devices together  -  they are not lumping because they don't understand the differences in devices.

Dutch Dresser commented that he liked the categories and sees Bin 3 as technology-forcing which is good for [his] industry.

David Menotti commented that efficiency is already written into the standard if using units of lbs/MMbtu, and this gets around lack of efficiency database.

Chris Neufeld commented that, as Dutch Dresser noted earlier [presumably about being lumped in and forced to perform in a way the device is not meant to], water-based and air-based furnaces should be separated.

Rick Curkeet commented that Europe would like to introduce hybrid systems here but there is not a good fit in the standard for these units, so they are not able to enter US market. Lisa Rector wondered if there should be an "other" Bin. Rick responded that the standard should go to emissions per heat output (to which Lisa commented that this happened to be her original proposal years ago). Lisa attempted to confirm that Rick would like handfed versus automatic feed categories and an emission per heat output, which are basically the classification used by Europe. David Menotti interjected that such a switch would have to be for the future since industry does not have enough data on heat output to support the change. Rick commented that some data exists and that some can be calculated. Martin Lunde commented that consumer's number one concern is efficiency; emissions are an afterthought.

In response to question from Gil Wood regarding pellets and non-pellet woodstoves being lumped together, John Ackerly responded that he is okay with the lumping but doesn't like the term "room heater" from a consumer point-of-view.

Regarding thermal storage units, David Menotti commented that masonry heaters cannot be regulated based on lb/MMbtu because there is no overall efficiency test method and no overall efficiency data, so these can only be regulated based on combustion efficiency (g / kg-of-wood-burned). Therefore the thermal storage unit category should be subdivided into (1) partial thermal storage hydronic heaters and (2) masonry heaters. David further commented that there is no [official] data for partial thermal storage hydronic heaters because the test method is still being developed (EN305 won't work because it leaves off cold start). These units look a lot like better-controlled cycling heaters which are coming in at 0.2 to 0.3 lb/MMbtu so Bin 2 (at 0.15 lb/MMbtu) would eliminate these units.

Andy Ginsburg commented that a goal is to build a database for room heaters based on lb/MMbtu. Bob Ferguson commented in response that HPBA has been promoting efficiency data collection to report on the hang tag rather than to use in the standard directly  -  so they are okay with collecting data for future but don't want efficiency requirements in the standard [now].

Arthur Marin asked Gil Wood if voluntary reporting of efficiency precluded EPA from setting a mandatory [efficiency-based] standard within the 8 year NSPS cycle. Gil explained that this would be a regulatory change and the package would need to go through the regulatory process, but no, EPA would not be precluded from doing so. 

Jim Hodina of the Linn County Public Health Department in Iowa asked Gil Wood about the regulatory timeframe and whether the preamble could state the timeframe in some way as the NAAQS does [to avoid the next NSPS taking longer than 8 years]. Gil noted that there is nothing legally preventing EPA from re-visiting the NSPS in less than 8 years.

Andy Ginsburg would like the current NSPS to build-in a transition to the next NSPS  -  e.g., include a deadline (6 or 7 years) or a process for collecting data for creation of new test method. Rick Curkeet suggested using an ASTM method, commenting that an ASTM method must be reviewed every 5 years in an open process  -  that the methodology is a living document that goes away after 8 years if not reviewed. ASTM could set a minimum efficiency, for example. Andy asked if efficiency % would need to be stated by EPA, and Rick responded that ASTM can set the test standard without EPA.

In-Use Durability

Overview was presented by Andy Ginsberg of the Oregon Department of Environmental Quality, who noted that in-use durability is important because of its impact on air quality (i.e., old stoves are a big part of inventory), because of the long-life and cost [of stoves], and because of consumer behavior. Stoves deteriorate with age, older units may emit at higher levels than when new, and an estimated 50% of stoves in Oregon are [older] uncertified models (i.e.., Phase 2 stoves will be around for many years). Furthermore, central heaters are expensive and thus not amenable to change-out programs. So stakeholders need to keep in mind that the stoves that get put in circulation today will likely impact airshed quality for a long time. Therefore Oregon [and presumably some other states] would like a durability standard.

Andy Ginsberg presented the following draft regulatory options for consideration, but was careful to note that these options do not necessarily represent any other state's proposal and that the other states had not yet seen this table:

                      Options for NSPS or State Programs
                                    Design
                                  Maintenance
                                   Education
Components & Warranties
Service contract with initial sale (e.g., business opportunity for chimney sweeps)
Value of maintenance
Test to simulate aging
Warranty void if not maintained per specifications
Manufacturer's instructions
Modular / replaceable parts
Periodic tune-up requirement (by state, not NSPS)
Proper burning techniques
Less consumer control
In-use testing with follow-up as needed (e.g., 10 years after installation)

Auto shut-down if out of specifications (e.g., O2 sensors)



Chris Neufeld commented that if there's a catalyst, the manufacturer must give a 3-year warranty. But the question is who will void the warranty (i.e., is manufacturer off-the-hook if consumer misuses product by for example burning house down after putting diesel fuel in stove?). Andy Ginsberg commented that incentives should be created for consumer to follow warranties (could be carrots and sticks). Arthur Marin commented that it could be thought of as akin to a consumer using wrong fuel in their car and voiding their warranty. Chris further commented that he had previously suggested defining a list of tolerances for manufacturing materials  -  that manufacturers could voluntarily test for this data and the product would consequently be more predictable  -  this would support building of a quality product that would tolerate prescribed use.

Dan Henry commented that warranty issues bankrupt companies and most companies already give "lifetime" warranties (7 years). The hydronic heater industry, for example, already requires 99% of Oregon's proposed durability standards. A warranty rate of >1% is unacceptable to manufacturers. Andy commented that minimum durability levels would help ensure that progress in industry continues.

Bob Ferguson commented that accelerated stress tests have merely shown that if you put your mind to it, you can break anything. Manufacturers will test [without being told to] to ensure product won't have warranty problems. Regulation is not necessary here.

John Ackerly commented that routine maintenance is a good idea but it needs to be data-driven. Missoula did a good job requiring certification technicians to go out and check gaskets every third year. Manufacturers don't have leverage once they sell product.

Rick Curkeet commented that required safety tests already include stress tests and overfiring for a long period of time. But labs don't conduct emission tests after safety test, or vice versa [because NSPS requires certified stove to be sealed for potential NSPS audit tests later]. The bottom line is that all devices get safety tested for way above what consumer would do (i.e., ULs of the world).

Arthur Marin wondered whether any manufacturers have recruited consumer volunteers to give back stove for manufacturer to test. Bob Ferguson responded that this had been done, but he's not sure how common the practice is and that industry didn't learn what they thought they would. Bob noted that the manufacturer will always get stove back for a warranty claim however. Rick Curkeet commented that it's common for employees of manufacturer to bring stove home and real-world-test.

** Moderator noted that some state stakeholders would like minimum [durability] standards to control bad actors.

Bob Ferguson noted that when industry reviewed warranties it was remarkable how very similar warranties were across products. He also noted that "lifetime" has different meaning for different states.

Chris Neufeld commented that consumers quickly communicate any dissatisfaction to each other via the internet and that manufacturers must respond immediately. Consumer dissatisfaction reporting is akin to a fast moving virus and is therefore self-enforcing for industry.

John Ackerly commented that there is nothing stopping towns from requiring chimney sweeps every year, as in Germany (using a testing device called a Moeller). Lisa Rector further commented that chimney sweeps in Germany are like OSHA here in authority to lock-out/tag-out.

Andy Ginsberg commented that research has shown that idiot lights alone don't incentivize people to fix cars, only performance issues incentivize consumer action. Unless stoves are dramatically malfunctioning (i.e., not heating), people will likely not be motivated to fix stoves. Chris Neufeld responded that on the contrary evidence shows that consumers do replace catalysts because they've spent a lot on stove (notwithstanding John Ackerly's comment that his own father had a catalytic stove and didn't know it was catalytic).

Dan Henry commented that manufacturers can put anything you want in their manuals but manufacturers have no ability to control consumer/user behavior once stove is out there. States however do have this power. Andy Ginsberg responded that states do have some power but looking for states and manufacturers to work together (e.g., incentives from manufacturers to encourage stove maintenance by consumer, like discount on product if replaced in time, etc).

Tomas Carbonell of the Environmental Defense Fund asked if service contracts are bundled with warranties. Dan Henry responded that it depends on the dealer (some pellet dealers do, for example), but most manufacturers don't bundle. Gil Wood noted that legally it could be required in the NSPS but such a provision probably could not be enforced effectively.

Tom Morrissey commented that his small company sells direct to consumers and it's only been in the last several years that they've been asked about service contracts. Rather, his company contacts consumers in the summer about catalysts or service kits (he noted that he wasn't sure consumers would contact the company). Tom further noted that it is clear from this meeting that stove lifespans are long. His company does their own change-out campaigns and gets 100-year-old stoves sometimes, and sometimes their own 30-year-old stoves. The company keeps a database of customers as well as a photo library of every component of stoves being serviced to catalog wear and tear, etc.

Martin Lunde commented regarding oversizing of stoves that manufacturers can't deal with this issue  -  rather sizing needs to be put in building code like Minnesota did. Building code is answer, not manufacturer. 

Chris Neufeld suggested a state tax credit to get stove inspected, with DOE review.

John Ackerly commented that manufacturers should not sell equipment, like screens [so doors could be left open], that should not be used. Dan Henry commented that screens could only be legally sold for stoves that passed emission test with screen on.

Summary and Next Steps

The major components of the wrap-up expressed by Dan Johnson of WESTAR were: 

   (1) Test method  -  Stakeholders are looking for ways to improve precision but in cost effective way. There is no answer yet as to timeline. More intense (work group) discussion needs to take place regarding nature and timeline of transition from crib- to cord wood.
   (2) Level and Form of Standard  -  The states are not surprised by industry's reaction to the proposed numbers, but Dan noted that if NSPS were to move to cord wood those numbers would change. There appears to be agreement about moving to two Bins (and for some to Bin 3 even). Regarding next steps, Dan did not hear an alternative suggestion to 50% best performing stove metric, and noted that there is not agreement on this metric.
   (3) In-Use Performance  -  There appears to be some shared interest in emission and warranty claims. The states would like to talk about this issue more in the work group.
   
Arthur Marin asked EPA if there was merit to spending next two weeks on cord wood off-ramp. Gil Wood and Greg Green responded affirmatively that it was worth EPA time and they'd be involved. Gil would work with John Crouch. John commented that HPBA is interested but its members are still thinking about this. Lisa Rector asked industry to put its cards on the table regarding this issue and others. Bob Ferguson commented that manufacturers are still chewing on cord wood idea, but they are burning wood and collecting data  -  just don't want to jump off cliff blind. David Menotti commented that he took this issue to his caucus but didn't get a quick response.
   
Gil Wood asked Greg Green how hard the end-of-year deadline is. Greg responded that EPA will move on a rule to protect public health and that there's a narrow window with which to move since this is not a court-ordered deadline.
   
Rod Tinnemore noted that the state and industry stakeholders are closer today on some issues than they were a year ago.
   
Tomas Carbonell noted appreciation to attend meeting and explained that the Environmental Defense Fund has received numerous complaints from citizens and grassroots groups, which is why EDF is involved and would like to see a protective standard quickly.

Dan Henry commented that the data on certified stoves ranges from 3 g/hr to 30 g/hr burning cord wood on every model multiple times. The president of HHT said he'd switch to cord wood certification testing if it would make stove better and reduce emissions because that's the right thing to do  -  but it will cost the company millions and cannot be done quickly. EPA should set higher real-world numbers  -  that a more lenient standard should be promulgated based on cord wood, rather than a tighter standard based on crib wood.

