Thursday, August 29, 2019

Records reveal successes and challenges in laboratory wood heater testing

Stove and boiler regulations appear to have survived Trump’s first term 

As the wood stove industry nears the May 15, 2020 deadline for meeting stricter emission standards, EPA records show a steady stream of stoves being certified but do not show which models, if any, failed. Manufacturers are required to notify the EPA of stoves they are testing 30 days in advance of the test and to report results after 60 days. Labs are also required to provide emission data to the EPA within 60 days, even if a test is suspended.
Percent of certification tests
scheduled for larger EPA-
approved test labs.  

Data from stove certification tests is a core resource for understanding and improving the effectiveness of regulations. A trove of documents just became available that helps us better understand how stove testing works and how a declining number of EPA staff is trying to oversee detailed regulations that partially rely on an honor system within the industry they are regulating.
Data recently released by EPA shows the pipeline of what is being tested by which lab, what has not been certified and other trends. The 2015 performance standards for residential wood heaters require manufacturers to notify the EPA at least 30 days in advance of the model’s certification test in an EPA-approved laboratory.

The EPA does not maintain a public database of 30-day notices, but the Northeastern interstate air quality organization NESCAUM requested copies of these reports through the Freedom of Information Act (FOIA). The main function of the 30-day notice is to enable the EPA and other agencies to witness testing, which otherwise occurs behind closed doors.

There is nothing proprietary in the 30-day notices and the EPA has no grounds to keep them confidential. Once anyone files a FOIA, the information is made public on the US government FOIA online site. To find the underlying data from these requests, type in “NESCAUM” in the FOIA online site or use these record locators: EPA-HQ-2019-006560, EPA-HQ-2019-000324 and EPA-HQ-2018-006770.

The EPA provided 30-day notices filed between 2015 and 2019 for 143 distinct heaters and 41 duplicate notices.  Matching the 30-day notices to the end result on the EPA's certified heater database indicates that approximately two-thirds of them are certified to 2020 standards.

We could not tell, for example, if crib tested stoves had a higher or lower passage rate then cordwood tested stoves, or which fuel led to more suspensions. The rapid rise in popularity of the cordwood test may indicate the stove manufacturer community’s confidence that they will not be more likely to suspend or fail a test.

More than 10 stoves or central heaters may have been certified since 2015 but only achieved Step 1 standards. The reason for this is unclear. It could be that these stoves already had Step 1 status but failed to achieve Step 2 emission limits. It may also indicate that manufacturers thought the EPA regulations would be struck down or EPA would grant a sell-through, neither of which appear to be happening at this point.

The EPA’s inaction to date on revising  2015 NSPS, and HPBA's strategy to seek repeated delays in their litigation, means that Step 1 stoves will likely be illegal to sell in the US after May 15, 2020. For companies that certified to Step 1 standards since May 2015, the expenses of testing the stove, and potentially modifying it, resulted in a stove that can only be sold for 2 – 4 years on the US market.

Approximately twenty-three stoves with 30-day notices do not appear on the EPA database of certified heaters. It is possible that some of the unlisted stoves have been tested very recently and are still in the certification process, while others might have failed certification testing, did not make it to market for a separate reason, or were discontinued by the manufacturer. 

Test labs

Two labs, Polytest and Omni, test a majority of stoves for the American market. Polytest is based in Montreal and Omni in Portland, Oregon. PFS acquired Dirigo, so it is now one lab, also based in Portland Oregon. Dirigo used to qualify or certify many outdoor wood boilers. Myren Labs is now only an R&D shop and no longer conducts certification tests, but EPA-approved labs can certify at Myren's lab. Two relatively new European EPA labs – RISE (formally SP) in Sweden and the Danish Technology Institute (DTI) have done a few certification tests.  Only the SZU lab in Czech Republic has apparently not yet done certification testing for any heaters for the US.
Four labs account for the great
majority of certification testing.

Polytest and Omni are also the labs that certify the most stoves with the ASTM cordwood protocol, though all labs now have experience with cordwood for either stoves or boilers.

Many stoves certification tests are done in manufacturers’ or R&D labs by EPA-approved lab technicians who travel there to undertake the testing. Using the same lab that the stove was developed in may be one way to enhance the repeatability of emissions testing and help ensure the stoves passes.

Stoves vs. boilers, crib vs. cordwood

Currently, there are 130 appliances on the EPA's central heater database, however we only have 14 30-day notices in for central heaters (boilers and furnaces), indicating a drastic cut back on testing central heaters to 2020 standards since 2015. Moreover, the testing of those units only resulted in 3 certified heaters: 2 European pellet boilers and 1 domestic cordwood furnace. Possibly the most notable manufacturer that does not have a 2020 certified unit is Central Boiler, who led the charge to stave off the EPA’s 2020 emission standards (litigation still outstanding). Omni labs filed a notice to test a new Central Boiler unit in the winter of 2017, but as this unit does not appear on the EPA list, it is unclear if the testing was suspended or what the results were.

The majority of the 30-day notices obtained were for stoves or room heaters, a category which includes wood, pellet, and multi-fuel stoves. Of these 129 unique room heaters, 36% listed crib wood, 33% pellets, 25% cord wood, and 2% coal/wood and >1% densified logs as the test fuel on the notice. A few 30-day notices did not have a fuel type selected. Twenty-two heaters were listed as having a catalyst, and at least five stoves were single burn rate stoves, a feature that is sometimes not advertised to consumers.

Posting non-confidential lab reports

A new provision of the 2015 EPA stove regulations requires manufacturers to publish their certification testing reports. These lab reports include all emission data, efficiency, grams per hour for each test run and the averaged final numbers. They also include photos of the stove and the crib loads, how the stove was loaded and other data once routinely kept private by manufacturers. These lab reports are usually too technical for the average consumer, but often read by other manufacturers, regulators, retailers and stove hobbyists.

However, these reports are sometimes hard to find on company websites and sometimes appear not to be posted at all by a few smaller manufacturers. After extensive searching, AGH was able to find more than 90% of the 2020 certified heaters’ certification reports online, as well as most Step 1 reports. To be listed as a certified heater by the EPA, manufacturers must provide the EPA with a link to their online report.  While major manufacturers are posting their reports, some remain hidden or difficult to find. And, since most stoves were grandfathered into the 2015 Step emission category, many lab reports detail the original certification testing which happened 15 – 30 years ago. (Once a stove was certified, it received a 5-year certificate of certification, which could be renewed without any additional testing every five years. Some currently certified stoves models are using tests conducting in the early 1990s, having received 5 EPA renewal approvals.)

Lab reports are often more than 200 pages long, but some companies limit what they disclose. Some are only 20 pages, and black out things such as the name of the company representative who witnessed the test. One very old report, from 1992, was only 2 pages long. Manufacturers are only required to post test reports for units tested after 2015, but the overwhelming majority posted test reports for all their certified units, providing testing transparency for the first time and access to third-party efficiency data, which consumers did not have access to before.  A few companies, including Central Boiler, only make reports public for tests done after 2015 and not for all of its certified units. (Those tests reports could likely be obtained through a FOIA request.)

Most companies maintain a single, centralized page for all their lab reports, making it easy to find them with a single click. They are called a variety for names from test report, to EPA data to website report, to certification certificate. Innovative Hearth Products that owns Astria, Ironstrike and Superior have some of the hardest pages to find as they are simply titled “Wood Heaters” and tech support told us that the company did not post those reports. Other manufacturers have the lab test report on the individual product page. At least 4 companies fixed broken links and reposted their test reports after we contacted them, notifying them that we could not locate them.

Links to non-CBI test lab reports

This is not an exhaustive list of manufacturers but includes most of the larger stove companies. (We will periodically update these links.)

Lab reports on central page: 509 Fabrications, Arada Stoves, Astria, Blaze King, Buck Stoves, Central Boiler, DroletEnerzone, Enerco Group (Mr. Heater), England's Stove Works, Enviro, Fire Chief Industries, Foyers Supreme, FPI Regency, Harman, Hearthstone, Iron Strike, JA Roby, Kuma, Morso, Napoleon, Osburn, Pacific Energy, Quadrafire, Regency, STUV, Superior, Travis/Lopi, US Stove, Vermont Castings, Woodstock Soapstone.

Lab reports on individual product pages: Froling, Jotul, MF Fire, RSF, Sierra Products, Thelin, Thermorossi, Timberwolf, Wittus (links go to a sample report).

Lab reports not found: Invicta, Laminox (we sent multiple emails and will post links if we get responses.)

Suspension of certification tests

From a review of 30-day notices of lab tests, and resulting certification, it would appear that virtually no stove fails a certification test. That is because if initial test runs are not going well, the certification test is suspended. There is no point continuing costly tests if the average grams per hour for all the tests is not going to be below 2 or 2.5 grams an hour. However, the EPA’s 2015 stove heater regulators require labs to “agree to immediately notify the Administrator of any suspended tests through email and in writing, giving the date suspended, the reason(s) why, and the projected date for restarting. The laboratory must submit the operation and test data obtained, even if the test is not completed.”

This data enables the EPA, state regulators and industry experts to understand how rigorous – or easy – a test protocol is for stoves of various firebox sizes and designs depending on whether crib or cordwood is used, for example.

In addition, the 2015 wood heater regulations stipulates, “Within 60 days after the date of completing each performance test, e.g., initial certification test, tests conducted for quality assurance, and tests for renewal or recertification, each manufacturer must submit the performance test data” to the EPA. Thus whether the tests resulted in certification, or were suspended or failed, the data still goes to the EPA and becomes part of the public record.

However, most labs appear to not always comply with this part of the NSPS and the EPA appears to not be strictly enforcing compliance (NESCAUM as submitted a FOIA request (EPA-HQ-2019-008306) for suspended test reports). This undermines the process of developing better test methods that could result in genuinely cleaner stoves.  In an unusual move, the EPA made a memo to test labs public that detailed lapses in lab test reports.  Filing reports from suspended tests was not among the issues the EPA raised to labs.

The high certification success rate of stoves scheduled to be tested may appear to undermine industry claims that the 2020 standards are impossible or crippling. However, stoves are intensively tested in private or internal labs before they are scheduled for expensive certification testing, as manufacturers need to know that they have a high chance of success to justify the expense. Manufacturers need to be able to repeat the prescribed test protocol relatively consistently to avoid a level of randomness between practice tests and certification tests that could result in repeatedly suspending certification tests. This issue of repeatability is at the core of the entire concept that wood heater testing is a valid and useful way to tell which heater is clean enough to put into homes and communities across America.

The fact that a very high percentage of certification tests are successful may undermine some industry claims that the variability of testing stoves under 2 grams an hour is far too high to render the test useful. Manufacturers have reams of data from in-house testing that could show how close their certification test results were to their practice tests. Some communities, often outside of the HPBA core, are far more transparent in sharing testing data, such as the Masonry Heater Association and various parts of the international cook stove community.

It is unclear if all manufacturers are following the same guidance about filing 30-day notices, as the EPA has become laxer about when wood heaters can be tested. Heaters often face delays in testing due to a variety of reasons, technically requiring a new 30-day notice. One stove AGH found has seven 30-days notices filed for it. Others had only one, but they may have delayed the test multiple times.

Conflicts of interest

An equally thorny issue is the conflict-ridden economic relationship between the labs and manufacturers. Labs want their clients to have confidence that they can get their stoves or boilers to pass the certification test. There is a process leading up to a certification test where labs come to know the strengths and weaknesses of a heater and match those with the areas of flexibility in a test protocol.

Labs guard their privacy and confidentiality on behalf of their clients. Curtains are often used to ensure that visitors cannot see whose heaters are being tested, and when and how. The 30-day notification rule was supposed to enable state or federal regulators to periodically witness certification - announced or unannounced. Witnessing testing allows regulators to understand the complexities, nuances and loopholes involved. Without that understanding, regulators cannot gain the level of expertise necessary to enforce existing rules, much less write better ones.

But here is the rub: AGH is not aware of any federal or state regulator who has witnessed a certification test in the last 5 years.  Rod Tinnemore, a Washington State regulator was one of the few who did witness tests and he gained the expertise, trust and respect of many stakeholders.  After his retirement, Washington appears to have backed off its multifaceted wood heater roles. Certification tests often last 4 – 7 days and can be very boring. Labs exist in only a few states and agencies do not have the time, resources – and sometimes even the authority – to witness testing. Most regulators do not even have the expertise to fully understand what they are witnessing and what nuanced lab practices are allowed or not allowed. The result is nearly a classic catch-22.

Conflicts are also potentially abundant in stove testing protocols like the ASTM E3053-17, which was developed and paid for by many of the same industry players whose stoves will be certified by those methods. Developing a protocol is an expensive and lengthy process, and there is no guarantee that it will be approved for use by the EPA and the EPA can unilaterally modify its use for certification testing. The EPA, however, may not have the resources and data to fully understand the nuances of a new alternative protocol until years later, when enough stoves have used the method. Further handicapping their oversight role is not consistently getting data from manufacturers for suspended tests.

Ultimately, audit testing of stoves in the same and different labs that certification testing was performing will likely be needed to bring attention and clarity that many of these issues deserve.  Audit testing by the EPA and/or states may help all labs ensure that testing parameters are consistently applied.

Similar, but more entrenched conflicts of interest in the European community stove testing community have led to what many experts regard as weak and ineffectual testing regimes. Some fear a convergence of cultures in US and European test labs if the EPA loses more staff and resources, or has its hands tied by appointed officials who favor industry concerns.

Future Challenges

Corporate, religious and educational institutions across America are coming to terms with legacies where people felt they had conflicts of interest and didn’t speak up about infractions and abuses. The VW auto testing scandal is a perfect example because it was in plain sight all along but everyone involved kept quiet. The issues and the stakes appear to be far smaller in the wood heater community, but we are dealing with devices that combust at extraordinarily hot temperatures in our living rooms and basements. When those temperatures are not present, the chimney may be emitting excessive particulates into our neighborhoods. Like the European auto testing community, the wood heater industry is also a relatively small group of seasoned experts, many of whom are older, known each for decades and abide by spoken and unspoken agreements not to publicly criticize others by name or expose issues to authorities.

In the 1970s, many industry insiders had a compelling mission - helping hundreds of thousands of families avoid incredibly high prices and live a simpler life that was more connected to nature and their own hard work. That mission is still a prominent driver for much of the wood heater industry, but instead of struggling to weld together enduring, affordable stoves, they are now filing 30-day notices, 60-day notices and a dizzying array of other paperwork.

One result is consolidation and a banding together to fend off stricter regulations and more oversight. Many manufacturers feel that they need to keep as much data as possible away from the government to survive. When outdoor wood boilers took off, just like when sales volumes of exempt single burn rate stoves soared, most industry insiders stayed quiet publicly, even when they privately fumed over the damage that these unregulated devices were causing to airsheds and the industry’s reputation. Top industry leaders also stayed quiet while highly exaggerated efficiency values proliferated and deployed their lawyers to ensure all stoves could be treated as 75% efficient or higher for IRS purposes.

Issues surrounding 30-day notices and suspension reports are not at the core of the struggle for the future of how well wood heaters work in the hands of homeowners, but they illuminate a slice of the struggles facing industry, EPA, state regulators and the air quality community as our country decides which renewable technologies will power our future. Once represented by the Washington DC based trade group “Wood Heat Alliance” in the 1980s, wood heaters are now represented by the HPBA that also represents more profitable gas and propane appliances and struggles to fit in to the renewable sector that thrives on greater transparency.


1. The 30-day notifications sent to the EPA are currently only being made through a Freedom of Information Act request, rendering them far less useful than if the EPA maintained a publicly available database. This could also provide transparency of key information that is not available elsewhere.

2. If state regulators who have labs in their state visited labs periodically to witness even part of a certification test, it could provide valuable insight and potentially oversight into the nature of testing procedures for states, the EPA and others.

· Oregon is by far the most important state since Omni and PFS-TECO are located there, and has a unique and longstanding role in stove regulation that predates the EPA’s role. The state also has areas where excessive wintertime wood smoke is a serious problem.

· NESCAUM has risen to become the most active and informed entity in the wood heating testing community. Its staff and member states can use their expertise to put pressure on the EPA and assist them in developing future test methods.

· California often plays a large role in national air quality policy, the state has taken a back seat for stationary wood heaters so far. Their engagement could help.

3. Manufacturers should start to routinely comply with the rule that they send a report with data on suspended tests within the required 60-day period. EPA also needs to enforce this regulation and clarify what those reports should look like. HPBA could take a leadership role in highlighting this regulation for labs and manufacturers.

4. The EPA and/or states should start conducting audit testing of stoves to better understand the application of crib and cordwood test methods.  This would assist all stakeholders to focus on vital issues about lab testing during a time when wood and pellet heating is under public scrutiny as a renewable energy solution.