Monday, December 30, 2013

Testing Observations at the Wood Stove Decathlon

By Norbert Senf

The Wood Stove Decathlon was a historic event. It was the first ever attempt to gather a collection of stoves in the field (literally, in this case) and test them for particulate matter (PM) emissions.
Norbert Senf, right, with Neils Wittus,
center, and John Ackerly
For something untried until now, the side-by-side field-testing can claim several firsts. 

It successfully compared stoves within a surprising range of categories including masonry heaters as well as retrofit kits. The project was a success not only as a media event but also in advancing the real world testing of wood burning stoves.

Cordwood is an extremely complicated fuel to get repeatable data with because it is so inherently variable. To add to the challenge, PM is particularly difficult to measure, even in a laboratory. While the test results from the Decathlon were not sufficient to provide PM numbers that allowed comparison with EPA numbers, they did allow a ranking of the stoves against each other. This is a substantial achievement in itself.

PM is the wood fuel pollutant of greatest interest since it causes the most public health concerns. Carbon monoxide (CO) is another pollutant.  It is created by incomplete combustion like PM, but it is much easier to measure. It is generally not considered a health hazard in low atmospheric concentrations outside of densely trafficked urban areas, and eventually oxidizes to CO2 on its own.
Due to new wood burning emissions regulations in Germany, two new portable instruments for measuring PM in the field were recently developed there. Fortunately, this happened just in time for the Decathlon to try them out. The instruments are limited to the 15-minute test cycle that is mandated in the German regulation, and therefore can only measure what happens during a portion of the burn. Measuring an entire test cycle will certainly be a goal for future Decathlons.

Common wisdom holds that low carbon monoxide (CO) emissions, which are easy to measure, will also ensure low PM, which is difficult to measure. The contest results did not bear this out. The stove with the lowest PM had the second highest CO. For the stove with the lowest CO, there were 4 stoves with lower PM. To be sure, the data set is limited. The 15-minute test window did not allow for average values to be measured over the burn of an entire fuel load.

Repeatability is one of the most important measures of data quality. Since each stove in the Decathlon received two (in theory) identical test runs, we can get a brief glimpse here, as well. Discarding obvious outliers, we see a coefficient of variation (CV) in repeat runs of 43% on PM, 40% on CO and 7% on efficiency. This compares favorably with EPA inter-laboratory repeatability studies, where the two stoves with the largest data sets both came in at 44% CV on PM. For masonry heaters, an MHA (Masonry Heater Association) laboratory study on repeatability with dimensional lumber fuel cribs yielded 10% CV on PM, 1.5% on CO, and 0.26% on efficiency.

The repeatability metric provides a useful baseline for judging data quality in future decathlons. There is an ongoing fueling protocol debate in the testing community between the repeatability achievable with fuel cribs, and the real world randomness of cordwood. EPA testing is currently done with cribs. To get a repeatable EPA cordwood number may require running a large number of (expensive) laboratory test runs and taking an average. To date, very little work has been done to provide data for either side of the debate.
All in all, the Wood Stove Decathlon was a great effort towards advancing our knowledge about how wood stove emissions compare in the real world. This was particularly valuable to see for different classes of appliances with no commonly defined EPA testing methods.
Valuable lessons and insights were had for designing a future challenge. Seeing the complex testing issues play out in real life was a unique educational opportunity for contestants, organizers, judges, regulators and the testing community itself.

Norbert Senf was one of the ten judges at the Wood Stove Decathlon. He joined early efforts to write codes and standards, and was a founding member of the Masonry Heater Association of North America (MHA). He currently chairs the MHA Technical Committee.


  1. Very interesting the variability results ..

    Ive been trying to find out information on the Testo ..Does it differentiate between PM2.5/10 or larger ? What particulate size was monitored during the decathalon tests ?


    1. The particulates on the Testo are collected on a gold impactor plate. There is no data so far on what sizes, or how it compares with collecting particulates on a filter, as the Wohler does.

      At the Decathlon, the two runs on each appliances were alternated between the Testo and the Wohler, and the data is currently being analyzed at Brookhaven National Laboratory. My guess would be that any variability between the Testo and the Wohler would not show up given the large variability in the cordwood firing, but we will have to wait and see.

      Recently, some comparison testing was done at Brookhaven, running the two analyzers simultaneously, and that data is also currently being analyzed.

      From previous studies, 90% of the particles in wood smoke are smaller than PM 1 (1 micron). I doubt if there is much difference between PM 10 and PM 2.5 in wood smoke.

      During dirtier smoldering combustion (blue smoke from the chimney) the particles will be tar droplets. On pellet stoves and during flaming combustion on regular stoves, the particulate emissions will be much lower, and will be mostly soot particles (EC, or elemental carbon). On very clean stoves, such as commercial scale pellet boilers, the particulates will be mostly minerals and salts (fly ash).

      What is often lost in the crossfire is the fact that the dirtiest appliances (outdoor boilers) are 100 times dirtier than appliances such as pellet stoves. Not only that, but the emissions themselves are more hazardous because tar contains PAH's (polycyclic aromatic hydrocarbons), which are carcinogens. If we could avoid smoldering fires by outlawing outdoor boilers, mandating the sale of dry wood only (as Washington state does) and teaching proper stove operation, 90% of the pollution problem would be solved.

  2. Here is a bit more detail about the Testo and also the Wohler: