A Wood-Gas Fuel Powered Sawmill

To allow us to relieve the load when starting our mill, we’ve kept the clutch functional within the bellhousing and used a slave cylinder — governed by a lever-controlled master cylinder at the operator’s station — to disengage the gear train. And, because the system is designed to run on either wood gas or petrol, we’ve hooked up separate smoke and conventional-fuel throttles, as well as a manual “fine tuning” air mix adjuster and a choke to facilitate cold-weather gasoline starts.

Two Changes for the Better

Although the gas-generating components perform the same functions as did those we’ve detailed in the past, we have made a few modifications to this newest group in order to reduce maintenance chores and lower construction costs.

The — which actually produces the burnable fuel from wood scraps — remains structurally unchanged, with two exceptions:

1. The 1-inch vapor inlet pipe leading to the condensate tank has been eliminated entirely, and the condensate drain tube now runs directly into that container, terminating in a slash cut near the base of the vessel. Also, a primary cleanout pipe has been welded to the bottom of the canister and extends into it, preventing complete condensate drainage. (The liquid reserve that’s left inhibits the flow of fuel gases from the generator . . . except under excessively high pressure conditions.)
2. In order to eliminate the need to purchase some of the 90 degree pipe elbows and short nipples required, we made an inexpensive tool (see “Build a Bender on a Budget,” at the end of this article) to actually bend one end of each of the three-eighths-inch iron air inlet tubes into a J-shape. These pipes were then inserted into the gasification chamber from the bottom (rather than the sides), and right-angle street ells — with eighth-inch air holes drilled into their upper surfaces — were threaded to the exposed ends of tubes. As a result, the number of both horizontal and vertical jets has been doubled, thus reducing the potential for solid-fuel “bridging” above the hearth zone. (This last modification has shown us that even the ballpark calculations mentioned in our wood-gas update aren’t all that critical, and that there’s considerable leeway in sizing the nozzle-to-hearth ratio.)

A “Cooler” Design

Through experience, we’ve discovered that the effectiveness of the cooler/filter system is probably the single most important factor in any successful wood-gas generator, because the dual-purpose unit must strip soot and tar from the raw gas and cool the smoke to below the dew point of water to remove excess moisture from the fuel.

And although our “old” scrubber was fully capable of performing both of those tasks, its capacity was limited by its dry medium. Furthermore, the new design is less expensive to make and easier to service.

Our latest unit forces the smoke to pass through two chambers, each filled with a mixture of water and a non-foaming wetting agent (such as automatic dishwasher detergent) that interacts with the ash and tar in the smoke, causing those particles to sink. These filtering chambers are separated by a compartment through which cooling fluid is continuously circulated.

By referring to the illustration in the Image Gallery, you can see how the apparatus works: Hot gaseous fuel is drawn into the preliminary bath near the base of the scrubber and bubbles up against the intermediate (cooling) chamber, splashing water as it does so. The smoke is then pulled back through the fluid and up along the outside wall of the cooler (Tank 2). As the gases approach the top of the unit, they’re drawn back down through the space between Tanks 2 and 3, and into the fluid reservoir in the secondary filter chamber. Another direction change then takes the smoke back up along the concave bottom of the inner vessel (Tank 3), where it’s forced to bubble through a field of quarter-inch holes before being released from the unit. (Any excess liquid that accumulates in the center tank-as a result of positive pressure from the inlet ports below-is allowed to restabilize by returning to the secondary filter chamber through four downcomer tubes . . . and the sight glasses in the filter chambers allow the water level to be monitored easily.)

Besides simply cleaning the gases, of course, the two filter chambers cool it, since the liquid mediums within them are in contact with the low-temperature intermediate chamber at all times. This central cooling cell is formed by welding an extra bottom onto the base of Tank 2, then directing the coolant inlet fitting at an angle tangential to the chamber’s periphery. A coupler, located at the center of the well, serves as an outlet.

(The accompanying photographs show, in addition to the generator and the cooler/filter, two other pieces of equipment that we haven’t yet described: a narrow tank wrapped with copper tubing and a PVC column with a sight glass. One of these is an air preheater and the other a supplemental moisture trap, and both — originally designed for our Chevrolet truck — are now being tested in conjunction with the stationary system.)

Coolant Circulation

In order to eliminate the need for auxiliary pumps or heat-dissipating storage tanks, we’ve used the engine’s existing cooling system, coupled to an air-conditioning condenser, to provide the low temperatures required for the coolant to do its job.

The condenser — which, in this case, functions as an air-to-water heat exchanger — was mounted immediately in front of the engine’s normal radiator, so air is drawn through both finned-tube devices by the fan. Coolant is extracted from the engine block’s drain plug (which is under pressure at all times) and routed to the lower inlet of the extra radiator. After passing through that exchanger, it’s directed to the inlet on the filter’s cooling chamber, then returned through the central outlet to the suction side of the engine’s water pump . . . which also serves as the heater return line.

This simple system has performed even beyond our expectations on the stationary unit (as well as on our Chevy truck), and has brought filtered gas temperatures down to below 120 degrees Fahrenheit at times.

Working in the Real World

Since the smoke generator/engine combination was installed, we’ve logged a total of about 50 hours of operation on the mill . . . trimming planks for our woodworking shop, cutting structural members for our cordwood barn, and simply demonstrating the setup to visitors. Though the powerplant does strain when running the largest timbers through the saw, it shows the same symptoms when burning gasoline, so we don’t believe we’ve sacrificed much power by using wood gas. Startup time (the waiting period necessary to produce high-quality fuel after firing the assembly) has been averaging about 13 minutes, and wood consumption (we use scavenged, fist-sized chunks of pine 2-by-4s) works out to about one generator load — equivalent to three-quarters of a 55-gallon drum — every 4¼ hours . . . at our usual 3,000-RPM operating speed.

Should we ever have to duplicate the setup, the only changes we’d likely make would be in the drive, rather than the fuel, system. We used automotive components because they were on hand and thus didn’t cost anything, but better performance could probably be obtained through the use of a step-down belt drive connected between the engine shaft and the saw’s main shaft.

Right now, we’re satisfied with the way the modified components have been working . . . though of course we’re still keeping a constant watch on them.

Build a Bender on a Budget

Our made-from-plumbing-scraps tube bender is an inexpensive — but effective — alternative to the real McCoy. We first cut the domed top from a 16-inch-diameter water heater and spot-welded into the lid of a 55-gallon drum. Then we welded a three-quarter-inch section — cut from the end of a 2-inch coupling — to the inside of a 1-inch length of 3-inch close nipple . . . and fastened that assembly, in turn, to the center of the dome. A flange nut — sliced from a 3-to-4-inch reducing bushing — was then threaded onto the 3-inch pipe, and a quarter-by-1-by-1½-inch flat metal stop was welded to the surface of the dome at a point about 1 inch from that pipe (or far enough away to allow the flange nut to turn freely). Finally, we took a half-by-2-by-18-inch flat bar, tapered one end of it, and welded the other tip to one side of a 1½-by-2-inch close nipple.

A three-eighths-inch hole, drilled through this flat handle at a point 3 inches from the center of the nipple, accommodates a hardened rod or drift punch nicely . . . and the protruding end of that pin pushes the to-be-curved tube around the bend when it’s clamped beneath the flange nut. A long pipe, slipped over the handle, can provide extra leverage for tough jobs, and — by filling the barrel with water – you can guarantee yourself a platform that’ll be sure to stay put!