Vermont Heat Research
An Experimental Wood Chip Furnace
the heat exchanger, above. Note the creosote-free white ash coating of the chamber walls.
from page 1
If we were to build an ideal wood fire, as opposed to a natural one, we would have the embers on top. The radiant embers would gasify the wood underneath. Cool air and fuel gas would mix well in some porous insulating medium at this lower level and move upwards until they reached the embers where ignition would take place. If we could just push air through the substance of wood from underneath we'd have nearly complete combustion, a hot transparent flame, and little or no smoke.
Well, wood isn't a porous spongy insulator. But a bed of wood chips in a container is a close approximation. We can filter air up through an insulating pile of chips. If we can move embers to the top, the fuel gases should burn, and burn cleanly. This is the method used in the VTHR furnace.
While it is a simple concept, the main design challenges in developing the burner were maintaining stratified temperatures for clean even combustion, fuel to air ratio, thermal stresses of the furnace wall, heat exchanger design, fuel loading method, prevention of gas leakage, and compact footprint and form factor.
Why green wood?
Water in fuel can either stop or slow combustion, depending on the amount, the temperature, and the location. Slowing combustion isn't always a bad thing, however. In fact we slow combustion on purpose in high compression auto and aircraft engines.
High octane fuel is actually fuel that burns slower than low octane fuel. Some racing engines have used water injection to mimic high octane fuel and slow combustion. In an engine this results in a longer period of steady expansion in a cylinder, rather than a sudden damaging explosion of fuel. We call explosion in an engine "ping" or detonation and it is detrimental and power robbing.
In the experimental furnace I'm studying, combustion is aided by being slowed down. Damp fuel means that the air supply can be filtered through the chips to adequately supply the needs of the slow burning upper hot layer. Faster combustion would yield insufficient air supply for the upper layer and a fuel-rich condition.
What this means is, in the VTHR furnace, if we use dry fuel, we get smoke. If we use green wood it burns cleanly. This is just one of the many non-intuitive aspects of this design.
Another thing that happens with moisture in the experimental furnace is that we get a phenomenon called the "water gas reaction". This basically causes the incandescent charcoal in the hot layer to combine with superheated steam from below to produce two good fuel gases, carbon monoxide and hydrogen, which recombine cleanly with oxygen only above the charcoal where the temperature is a little cooler.
Another reaction which results from the movement upwards of fuel gases from cool to superheated zones is that tars and creosote are broken down by the intense heat before combustion to simpler fuel gases in a process called "cracking". Like water gas, these cracked fuel gases, well mixed with the right amount of oxygen, burn cleanly only as they move past the dissociating incandescent charcoal layer to a slightly cooler area.
The result of all this is an unusual looking form of clean combustion. The flames in the VTHR furnace have reduced incandescence and turbulence (due to low gas speed and good prior mixing) and look different than those in a conventional wood fire. The appearance is a little like photos of the surface of the sun with a glowing orange crust and long thin flowing translucent blue and orange laminar gas flares rising above it. Of course this is a chemical process, not fusion, but the image is noticeable and evocative.
What about green wood stack losses?
As the exhaust from the experimental furnace rises above the combustion zone it contains all of the water in the original wood chips (about 50% by weight) plus the water created when we burn any fuel completely. In a 50 pound charge of green wood we have to evaporate over 25 pounds of water and send it up the chimney.
The heat contained by this steam is lost. This fact has sometimes been mentioned as a criticism of green chip combustion. A common mistake is to think that 50% moisture content by weight means 50% less heat. We recently lost a Vermont EPSCoR research grant competition because a single reviewer made a similar intuitive mistake without calculating the amount of stack loss.
While it's true heat is given up to steam, it is an acceptably small loss compared to the heat produced.
To understand this we need a little math. Here it is:
Sensible heat is the heat needed to bring the moisture in the chips up to 212 degrees, plus any additional heat after stream is formed.
Latent heat is the heat needed to create steam from 212 degree F. moisture.
For 1 pound of 50% moisture chips, water raised from 60 to 350 degrees:
Sensible heat = (350 - 60 degrees) x 0.5 lbs = 290 x .5 = 145 Btu
Latent heat = (970 Btu/lb) x 0.5 lbs. = 485 Btu
Total lost heat = (145 Btu + 485 Btu) = 630 Btu
Fuel Heat value = .5 lbs. x 8600 Btu/lb. = 4300 Btu
Heat loss = (630 Btu/ 4300 Btu * 100%) = 14.6%
We appear to lose only about 14.6% of our total available heat by burning green wood. But even that's not the whole picture. Because so called "dry" wood and pellets also contain some water (about 12%) they also lose heat up the stack as steam. Dried wood loses about 3.5% of it's heat value to stack evaporation. This means that the penalty for using green wood in a batch combustor is only about 11% compared to dry wood.
This is why we chose to design the VTHR furnace for green chip combustion.
It is costly to dry wood -- whether chips or cord wood or pellets. The process of drying wood always requires the input of at least this same 11% of heat -- generally it takes considerably more when using artificial drying methods. In fact wood pellet slurry contains even more than 50% water, and consumes much more energy to dry than green wood chips lose at the stack.
Wood Chip Research
While much of the modern research into bio-fuels centers around fluid fuel conversions, comparatively little work has been done to learn how to cleanly burn so-called low grade solid fuels. At least not on a home heating scale. Most solid fuel experimentation that I'm aware of has focused on either small scale (cooking and camp stoves) or at the other extreme, commercial scale projects..
There is a large amount of Federal funding for high tech (and energy intensive) fluid conversion fuel technologies, but almost none for basic solid fuel combustion research. A recent example of this is the 2006 Department of Energy SBIR grant topic announcement which provides small business funding for research in hundreds of areas of synthetic liquid fuel research, nuclear research, hydrogen and fuel cell research, but not one grant for solid fuel combustion studies.
When we talk about renewable energy we must realize that biomass is essentially solid. Why not use it efficiently in that form to produce simple heat, where feasible? We seem to be convinced wood must somehow be converted into some other form to provide useful clean and efficient energy.
The VTHR Furnace
The VTHR green chip furnace was built to explorepractical methods of utilizing solid biomass fuel. Several smaller experimental furnaces were tested in February and March of 2006, and the larger prototype was built by July. Initial outdoor tests in summer conditions showed promise and verified the concept as viable, but did not test it in service under real winter conditions connected to a dwelling.
To begin a true test process, the furnace was installed in its own building in October 2006 along with a 2500 pound hot water storage tank and connected to the home heating system. It provided full house heat until March of 2007. Preliminary estimates were that it produced about 175,000 Btu of hot water per burn at about 80% efficiency.
Problems and Conclusions
Initial tests were made with home-grown chips, and these appeared to work satisfactorily for several months. However a changeover to purchased red oak chips of much larger size, stored outdoors under a tarp revealed problems with the furnace design. While these problems were relatively inferequent in appearance, it was decided that they made the furnace impractical as a commercially developable project.
I found that green chips without surface moisture behaved differently in the furnace than chips that had been wet by rain or thawing snow. There is a difference in a 50% moisture content chip and a wet 50% moisture content chip.
Storage of chip piles under poly tarps provided to be inadequate in our Vermont winter, and the chips were penetrated by rain and snow thaws. Wet chips are difficult to dry in a pile, and even more so in winter.
In burning the chips, surface moisture tended to make for inconsistent combustion. Inconsistent combustion led to the creation of holes in the surface burn. These led to a phenomenon we dubbed "tunneling" where combustion moved under the extinguished chips, leading to a very smoky burn which was difficult and messy to extinguish.
This was a relatively infrequent occurence -- it probably averaged once out of 50 to 100 burn tests over the winter. Nevertheless, it was serious enough to render the furnace impractical as a commercial product. In practice, the furnace had to be monitored frequently to correct or prevent the start of any tunneling.
In an attempt to reduce the liklihood of tunneling several design modifications were tried. One of these was a "combustion follower," a form of heat reflector which descended with the burning surface of the fuel batch. The purpose was to focus enough heat to keep the surface ignited. This reduced the frequency of tunneling, but did not eliminate it. It had tghe adverse effect of increasing furnace wall temperatures to the point where the wall was visibly distorted over time. The temperature differentials in this kind of furnace are exaggerated by the combustion follower.
One aspect of the problem which couldn't adequately be explored was the effect of different types and sizes of chips on the tunneling problem. It seems likely that if chip size quality and material was consistent and dry, that tunneling might not have occurred. However in real world usage, a product based on this furnace might be subjected to fuels of all kinds and qualities introduced by users of varying degrees of understanding and vigilance.
Finally, in March, very cold temperatures for several days combined with high winds to freeze our underground heat transfer pipes at a location that had been inadequately covered by soil in the Fall. We decided to drain the system at this point to avoid further freezing damage.
Three issues of a new publication, VTHR Journal were produced while the furnace was in operation to describe the experiment. The hope had been to continue it as the furnace developed. However the dead-end reached by the experiment also ended the subject and basis of the publication. I've decided to add the three issues to the website here, since the information is still I believe useful. I'll be converting the pdf files over to HTML for the web over the next week (11/19/07)