AGH Comments to the EPA on New Stove Data

The EPA is in the process of posting everyone's comments.  As of Monday, August 4, they have posted comments by Rick Kurkeet from Intertek Labs, Stuart Clark from Washington Department of Ecology and several anonymous ones.  They can be found here and ours are posted below.

Alliance for Green Heat

Comments on the

July 1, 2014 Notice of Data Availability

Regarding EPA’s Proposed Standards of Performance for New Residential Wood Heaters, New Residential Hydronic Heaters and Forced-Air Furnaces, and New Residential Masonry Heaters

Docket ID No. EPA-HQ-OAR-2009-0734

July 31, 2014

The Alliance for Green Heat (Alliance), appreciates the opportunity to comment on EPA’s July 1, 2014 Notice of Data Availability (NODA) regarding the agency’s proposed New Source Performance Standards (NSPS) for wood heating devices.[1]  The Alliance is an independent non-profit organization that works with environmental and forestry organizations, air quality experts, the wood and pellet stove industry, and others in the wood burning community to promote high-efficiency wood combustion as a low-carbon, sustainable, local and affordable heating solution.  The Clean Air Act requires EPA to review and revise, if appropriate, the NSPS at least every eight years.  As we explain in the original comments we submitted on the Proposed Rule, the Alliance strongly supports EPA’s decision to update the standards for wood stoves and to require a number of previously unregulated wood heating devices to reduce their emissions.  We also believe that the new standards, which reflect significant improvements in wood heating technology, are both appropriate and long overdue.  We offer the following additional comments in response to EPA’s July 1, 2014 NODA.

First and most importantly, it is imperative that, going forward, all stoves be required to complete a test at the lowest possible burn rate setting.  The data released in the NODA provided a vital insight that few people know: 80% of non-catalytic stoves cannot be successfully tested at the Category 1 burn rate (0.8 kg/h or less).  Under the current EPA test method, Method 28, these stoves have been allowed to test at the higher Category 2 burn rate, as long as they can successfully complete a test at 1.0 kg/h or less in Category 2.

The possibility of avoiding testing a stove at the lowest burn rate—which is when emissions of PM and other toxics are often the highest—has apparently created a loophole that has incentivized some manufacturers to design their stoves so that they fail to hold a flame for a sufficient amount of time in Category 1, so that they can be tested at the easier-to-meet 1.0 kg/h burn rate in Category 2. 

If EPA finalizes its proposal to certify stoves based on only the highest and lowest burn rates[2], this potential loophole could become even more important.  EPA should clarify in the final rule that all stoves certified under the new NSPS must be capable of being tested at their lowest burn rate, and that stoves that cannot successfully complete a test burn at their lowest rate will no longer be eligible for certification.

This will still allow stoves to be tested at 1 kg/h but assures that the consumer cannot operate the stove at a lower burn rate.  We are not opposed to allowing stoves to use a minimum burn rate of up to 1.15 for cordwood testing, but again, they must be successfully tested at their lowest burn rate.  Otherwise, the proposal to raise the minimum to 1.15 kg/h could just enlarge the loophole of stoves testing at higher burn rates than consumers could operate them.

Second, it has become abundantly clear in recent months that more and more manufacturers are using the K list to make minor changes in their stoves to ensure 5 more years of sales life before retesting.  While the K list provides an important function, it is now being used to significantly delay retesting of what could be a majority of EPA certified stoves.  One of the results of this is that manufacturers can avoid testing at their lowest burn rate and avoid releasing their efficiency data until 2019 for many or most or in some cases, all of their stoves. For this reason, we again urge the EPA to require that B415.1 efficiency numbers using HHV be submitted to the EPA within 6 months of promulgation. 

The data released in the NODA confirmed previous data sets, that the efficiency range of pellet stoves is far wider than wood stoves.  These certified pellet stoves ranged from 62 to 80% efficient with a standard deviation of 8.5, double the standard deviation of non-cats and four times the deviation of cat stoves.  Because of this wide variation of efficiencies, it is in the public interest to require efficiencies be released as soon as practicable.  Manufacturers already have efficiency data for most or all of their stoves so there would not be any significant burden to them.

In addition, we note that the data provided in the NODA appears to establish that an emission rate of 1.3 g/h is already being achieved by some sources in the source category using cord wood.  The Alliance continues to support EPA’s proposed approach of allowing stoves to be certified with either crib wood or cord wood during the first phase of the NSPS, and to require certification with cordwood thereafter.  In addition, the Alliance recommends that EPA continue to gather test data on cord wood emission rates for non-catalytic stoves over the next five years and consider revisiting the emission limits in the NSPS if additional data suggest that the NSPS should be adjusted for non-catalytic stoves. 

Finally, the Alliance has published a blog post that addresses a range of issues related to the data released in the NODA.  We have appended this post to these comments and request that EPA consider it along with our other comments as it finalizes the Proposed Rule.

                                                                                    Sincerely,

                                                                                    /s/ John Ackerly

                                                                                    Alliance for Green Heat

---

[1] Standards of Performance for New Residential Wood Heaters, New Residential Hydronic Heaters and Forced-Air Furnaces, and New Residential Masonry Heaters, Notice of Data Availability, 79 Fed. Reg. 37,259 (July 1, 2014) [hereinafter “NODA”].

[2] See Proposed Rule, 79 Fed. Reg. 6,329, 6,367 (Feb. 3, 2014).

EPA bolsters legal case for NSPS with new data

Alliance for Green Heat, July 15, 2014 - On July 1st, the EPA released a NODA – a Notice of Data Availability – consisting of data from 3 stoves tested with cordwood and 106 stoves tested with crib wood. 

Vermont Castings combustion engineer
Doug Fongeallaz uses cord wood to test
a stove instead of  Doug fir cribwood,
according to the Valley News.

The NODA is a vehicle for letting people know the EPA is relying on this evidence to make their final rule, so parties won’t be able to say that the agency surprised them, in violation of procedure for issuing new rules, if the EPA cites it.  The NODA also provides stakeholders a chance to dispute the additional evidence since it was not available at the time the rule was proposed.  The fact that EPA has published this data and the NODA in the federal register now allows EPA to rely on this data for the final rule.  It is likely that HPBA will publicly respond and a number of manufacturers and agencies are also planning on responding.

This NODA lays a better legal basis to withstand a potential lawsuit from the hearth industry and also gives more insight into what final emission standard the EPA is able to set.  Many in industry claim that the EPA does not have sufficient data to set an emission standard based on cordwood.  And some manufacturers have data but are not sharing it with the EPA in an attempt to prevent the EPA from establishing a cordwood emission standard.

Overview

This new cordwood data shows that some catalytic stoves can already meet a 1.3 grams per hour standard with cordwood.  However, the non-catalytic stove performed far worse with cordwood than it did with crib wood, with grams per hour of 4.2 on the Category IV burn, 11.7 on Category III and a whopping 39.9 on Category II.  The stove was not able to burn at Category I, the lowest air setting, either with cordwood or in the crib wood certification test.  Among the 63 non-cat stove tests released by the EPA in the NODA, only 8 of them could be tested at Category 1, the low burn rate.  It is is unclear how many of these 63 stoves can be operated by the consumer at a lower burn rate than they were tested at in the lab. The HPBA campaign to raise the minimum burn rate from 1 kg/h to 1.15 kg/h could have significantly added to the problem of stoves not being tested the way consumers often use them.

The data partially confirms what industry has been arguing – that it's still too early to set a cordwood standard for non-cats.  Firstly, there is too little publicly available data, and, even if the standard remained at 4.5 grams per hour (but with cordwood), manufacturers may not be able to make non-cat stoves that qualify. The incredibly high emissions from this non-cat stove could be a factor in leading the EPA to postpone a cordwood certification standard.  Presumably the EPA would require cordwood testing over the next 8 years and possibly make cordwood certification optional, while continuing to rely on crib certification.

The data also partially confirms what many air agencies have been saying: that some stoves can readily meet a 1.3 grams per hour standard, which is all that is legally required by the EPA to meet the best available technology test.

At the core of this issue may be the decision by EPA, not contested by HPBA, to set a single standard for cat, non-cat and pellet stoves.  There is data to show that pellet and cat stoves can meet a 1.3 standard now, but without separating them, the non-cat stoves may delay reaping the low-emission benefits of those technologies. Some manufacturers, such as Jotul, initially argued that cat and non-cat standards should be delinked, but others were against delinking, including the Catalytic Hearth Coalition.  The Alliance for Green Heat and a few others argued for a separate category for pellet stoves.  The internal politics and public rift within HPBA over this issue contributed to the industry association remaining silent on the issue of single or separate standards for cat and non-cat stoves.  Only time will tell if this is a successful legal strategy.

An attorney who specializes in NSPS law and is familiar with this case says that "the EPA could, in theory, set a legally defensible emission limit for both cat and non-cat stoves that relies on emission rates that are achievable by catalytic stoves but that may not be achievable by non-cats. EPA can’t require a specific technology under section 111, but it has the discretion to set the emission standard at a level that has only been achieved by one type of technology.  I suspect EPA would probably win on this point if challenged, however at this early stage, such predictions are still very speculative.  Ideally, there will be more data points than just 2 stoves, but I don’t think that’s going to be fatal to EPA, especially if industry isn’t willing or able to produce cordwood test data contradicting EPA’s data."

(The Alliance for Green Heat does not support a cordwood standard such as 1.3 grams per hour that some cat stoves appear to be able to meet, but regularly consults legal experts to see what EPA can do.  The danger may be that regardless of the recommendations from the EPA's stove experts based in North Carolina, far more senior EPA officials in Washington may take a much harder line.)

Summary of Woodstock Soapstone stove

Tom Morrissey, back row, 2d from
right and the team that built the
Ideal Steel Hybrid that emits less
than 1.3 grams an hour with
cordwood and cribs.

Woodstock Soapstone voluntarily provided full cord and crib wood test results to the EPA, something that perhaps no other stove manufacturer did during the NSPS process.  Woodstock Soapstone is one of the few stove manufacturers that is not a member of HPBA and thus does not follow their advice and guidance on sharing information with the EPA and what positions to take on contentious NSPS issues.

Woodstock Soapstone's stove showed a very notable consistency in emissions of grams per hour between cord and crib wood.  Certification tests using crib wood produced between 0.3 and 1.3 grams per hour for the four burn rates.  R&D testing in the manufacturer's lab with crib wood produced 0.4 and 0.5 grams per hour and between 0.5 and 0.8 grams using cordwood. 

Similarly, the stove showed high levels of consistency in efficiency between cord and crib wood and the certification lab and the manufacturer’s lab. Crib tests were between 74 and 84.5% efficiency HHV, and between 74.2 and 82.3% for cordwood for all the burn rates.

Summary of catalytic stove #2

A second, unidentified manufacturer voluntarily provided test results for 2 catalytic stoves.  A summary note from Gil Wood said, “the test data show that these two EPA-certified catalytic wood stoves when tested using cordwood - and making no design changes to adjust for crib wood versus cordwood in the tests - have similar emissions as when they were tested for the official EPA certification tests using crib wood. That is, the test data show particulate emission rates (g/hr) using cord wood that are equal to or less than the corresponding test data using crib wood for Category 1 (minimum burn rate) and Category 4 (maximum burn rate).”

The stove produced between 0.3 and 0.8 grams per hour for cord wood for Category I and IV.  A second Category IV test came in at 0.8 as well.

Summary of non-cat stove

A popular non-cat stove was purchased for testing at Brookhaven National Lab that was under contract with EPA to test cordwood in non-cats.  According to the Brookhaven test report, “the test cord wood was guided by draft cord wood specifications and procedures under active development by ASTM.  Testing was successfully completed in Method 28 Categories IV, III, and II.  It was not found possible to achieve the Cat. I burn rate with this fuel when the air control damper was fully closed. In this case Method 28 provides a method for determining average emissions based on weighting the results of the other three categories in which the stove was tested. It should be noted that in the earlier certification testing with crib wood, Category I operation was also not achieved.”

One notable result is that this popular non-cat had three very reproducible emission rates over a 3-day period.  On Cat IV, the stove produced 4.2 grams per hour on May 19th, 4.1 on May 20th and 4.3 on May 21th for an average of 4.2.  Reproducibility fell apart on Category III, with burns over 5 days producing a range from 6.4 to 17.4 grams an hour, for an average of 11.7.  Category II had slightly better reproducibility when the stove was smoldering at very high rates of particulate matter.  It produced 36-48 grams per hour over 4 days for an average of 39 grams per hour.  The high emission rates may partially be the result of fuel where the core moisture content was sometimes above 30%, although an average of core and shell was always below 25%.

The stove was also tested with very high moisture content fuel of 48.4% moisture content on a wet basis. Brookhaven found that particulate emissions were extremely high: 50.6 g/hr over the run, and 11.8 times higher than the average emission rate with the drier test fuel in Category IV.

The Brookhaven report concluded: “For optimal performance of this stove on cord wood, some rebalancing of the primary air / secondary air ratio may be required. It is possible as well that the details of the air damper setting and the procedure for loading and the timing of the startup operations contributed to differences between cord wood and reported crib data.”

Summary of 106 stoves certified by EPA since 2009

The data on the 106 stoves released in this NODA consisted of EPA approved lab certification test reports conducted since 2009.  None of them contained any data about cord wood, but this data set may be extremely important if the EPA decides to continue using cribwood to certify wood stoves in the 2015 NSPS.  Analysis of this data set could form the basis of Step 2 emission standard that is considerably below 4.5 grams per hour.  Some experts expect that the EPA may ultimately designate a standard between 2.0-2.5 grams per hour that would likely take effect in 2020. 

This data set also contained efficiency numbers that had not been reported anywhere before.  A great majority of these stoves are presumably still on the market.  The average HHV efficiency for non-cats and pellet stoves were exactly the same: 71%. Cat stoves had an average efficiency of 79%, but the sample size of 3 stoves was very small.  This data set confirmed previous data sets, that the efficiency range of pellet stoves is far wider than wood stoves.  These certified pellet stoves ranged from 62 to 80% efficient with a standard deviation of 8.5, double the standard deviation of non-cats and four times the deviation of cat stoves.  Some pellet stoves on the market have efficiencies as low as 40% and are likely to be uncertified models. The Alliance for Green Heat will be releasing a separate analysis of these 106 stoves that also shows the correlation between efficiency and emissions.

The EPA is soliciting comments on this NODA before July 31. To comment, click here.

Click here for a related story summarizing comments to the EPA about the proposed regulations.

Can the EPA Set Future Emission Limits Using a Cordwood Test?

In the proposed new regulations for residential wood heating devices, known as the New Source Performance Standards or “NSPS”, the EPA is not only trying to reduce the amount of smoke allowed from new stoves, but it is also proposing to switch the fuel used for testing devices in future years.

Traditionally, stoves have been tested using “crib” fuel, meaning Douglas Fir 2 x 4s and 4 x 4s that are stapled together to maximize the consistency of the fuel load.  EPA is proposing to allow manufacturers to use either crib wood or cordwood (which is more similar to the split wood logs that stove users typically burn) to test new stoves during the first phase of the proposed rule.  However, EPA would require that only cordwood be used to test stoves during the later phases of the rule.  Using cordwood for testing makes sense because a test using cordwood would likely come closer to what the real-world emissions from new stoves might be.

EPA is proposing that new stoves would be required to emit no more than 1.3 grams of particulate matter (smoke) an hour during the later phases of the rule (within 5-8 years) as demonstrated using cordwood.  Some observers have pointed out that although 1.3 grams an hour is already achievable today with some stoves tested with crib fuel, it is not clear how many (or if any at all) could pass the test using cordwood.  There is not any way to convert between tests using crib wood and tests using cordwood, and very few stoves have so far been tested using cordwood. 

Industry has voiced concerns, saying EPA cannot legally set a standard where there is insufficient data to back it up.  However, there are a number of ways EPA could still set a standard based on cordwood, even if it doesn’t have a mountain of data right now.  For example, there is legal precedent supporting EPA’s setting a rule based on what it reasonably predicts a technology could achieve in the future even if it doesn’t have hard data when it proposes a new rule.  In addition, there are many options for EPA to obtain more data before it finalizes the NSPS of which the wood stove community should be aware.

First, if EPA learns of new data using cordwood testing before it issues the final rule, it can issue a Notice of Data Availability, or “NODA” that makes the data available and requests additional comment.  Even absent a NODA, the EPA can consider comments and data submitted after May 5, but it is not required to do so.  This approach could allow EPA to collect more data about cordwood testing during the rest of 2014.

Second, EPA could issue an “Information Collection Request” under the Clean Air Act that would require manufacturers to test and turn in data on their stoves’ emission using a cordwood test.  

Third, EPA could re-open and re-consider the NSPS after it issues the final rule.  For instance, the agency could open a discrete part of the rule in 2 or 3 years for the sole purpose of setting or adjusting the emission limits for wood stoves using a cordwood test method.  EPA would have to go through the notice and comment process, but a technical update to the rule that adjusts the test method or emission limit based on new data could be completed relatively quickly, with a shorter comment period and fewer comments for EPA to consider.  

If the EPA ultimately sets very strict limits using cordwood for later phases of the NSPS, there is only a small chance that any lawsuit could stop the regulation from initially going into effect.  To prevent a regulation from going into effect, there is a very high burden of showing irreparable harm.  Also, courts are supposed to give EPA a lot of leeway in interpreting contested scientific or technical information (such as the achievability of the 1.3 gram per hour limit using cordwood).  Since the initial emission limits will be based on cribwood tests and will have little impact on the wood stove industry for the first several years, the wood stove industry will be hard pressed to show immediate, irreparable harm.  Warm air furnaces, outdoor boilers and other exempt appliances may have a better chance at showing irreparable harm as they have not been regulated before and there are manufacturers of those appliances who are more likely to be harmed or go out of business.  But those are the appliances that industry, states and others most want to be regulated, and where the greatest air quality benefits will come from.  However, EPA could also agree to stay the rule if industry is able to persuade it to do so.

As long as EPA itself doesn’t agree to stay the rule, the regulations would likely go into effect immediately upon final rulemaking, and all the initial emission limits would become law and would be effective even if industry sued.  If the industry were successful in suing EPA, a court could strike down certain provisions or remand the rule to the EPA.  Thus, a lawsuit by industry is very unlikely to stop the regulations from taking effect, may not succeed in the end, and would only prolong the uncertainty that industry has had to face to date.