Thanks for the data-based info. I don't have a means to measure temperature in the secondary chamber of my Tarm, but wish I did. Where exactly did you place the sensor?
Instantaneous combustion temperature, I believe, may be considerably higher than temperature a short distance from the point of combustion, as the gases cool quickly as they mix with air in the secondary chamber. Since I have no data to back this up, FWIW, I would surmise that sustained secondary temperatures, except during the very beginning and end of the burn, should be well above 1000F. Your chart shows some spikes, but the average shows pretty close to 1000F, so maybe my thinking is incorrect. Perhaps this relates to the actual measuring location. Does anyone else have any data to compare with this?
From 1000-2000F, especially into the higher range, is where gasifiers really perform their magic. The highly combustible CO, H and formaldehyde are vigorously generated in this range, as well as methane formation/combustion and consumption of carbon. CO spontaneously ignites at about 1200F, H at about 1100F, formaldehyde at about 800F, and methane up to 1400F, while carbon is consumed at 1800F and above.
An efficiency calculation has some key assumptions and variables, but that's often as close as most of us can get, unless we follow laboratory procedures. Those include assumed moisture content, as well as calculated home and tank heat loss in your situation. What are your specifics on these heat loss calculations? Outside temperature, wind, sun, and much more impact home heat loss. I wouldn't know how to calculate actual heat loss from a structure (except from a known BTU source, like a storage tank, which was drawn down to maintain a constant structure temperature), and that heat loss calculation would change with any change in the influencing factors. I also don't know how to measure heat loss of a tank, except in a static situation, because again it would change with any change in the influencing factors.
What types of sensors did you use?
So far I have found only two ways to reduce the variables and keep the assumptions to a minimum. 1) do a controlled burn with boiler output only to a storage tank, no other heat draw on the system, and then measure the BTU gain in the storage tank over the burn period; or 2) do a controlled burn and measure supply/return temperature and gpm flow, and then calculate the BTUh over the burn period. The second method reduces heat losses variables, but determining actual gpm flow can quite difficult, and small errors can result in large changes in the final calculations.
Are you able to do shut down all draws on your system, do a controlled burn and only charge the tank? When I did this with my Tarm Solo 40, I calculated 86% efficiency based only on temperature rise of the tank, which is quite well insulated. That did not include heat loss from the tank, boiler and piping, which if I could have measured would have yielded a higher efficiency.
For those of us who try to rely on data, we struggle with actual BTU content of the wood. That too is a big assumption, if high accuracy in the data is important. I have used 6,050 BTU/lb at 20% MC and 400F flue temp. Regardless, an assumption has to be made, and at least the data has some relative accuracy, even if it lacks absolute accuracy.
Keep up the work. It helps all of us.