When the kiln was originally fabricated the prior year by a shop near San Jose, a series of miscues and lack of qualified engineering oversight resulted in a number of defects and deficiencies. These were noted last August when Stephen Joseph was on hand to commission the kiln. Stepehen determined that further operation would rapidly result in degradation and eventual mechanical failure, so he left the Costa Rican engineers with a fix-it list. They prepared some engineering drawings, but a breakdown in communications between the NGO administering the project, the organization charged with building the unit, and their engineers and machine shops resulted in a stalemate. Six months later, the kiln was still in the same sorry state. When funding was approved for the 2010 season, we (Biocombustibles de Costa Rica--BCR) were given the nod as mechanical contractors, with on-site engineering oversight by the kiln's original designer, Nik Foidl of Austria. That's when we settled on the big TLUD to prime the kiln (see prior post). 
When we took delivery of the kiln, a number of additional materials and mechanical faults were noted, and more still when Nik finally arrived. There was a flurry of work, with lots of improvisation to substitute for materials we were unable to track down, or specialty machine shop work that would have resulted in excessive delays. Throughout this period we experienced exceptionally high rainfall. Since the kiln and TLUD were out in the open, the rains significantly limited our testing time. In addition to abundant rains, here in the humid tropics the equilibrium moisture content percentage of seasoned wood hangs in the mid- to high-20's. Driving out all the water significantly increases the amount of time and fuel needed to prime and operate the kiln. It also adds so much water vapor to the exhaust stream that the combustible gases of pyrolysis can be impossible to ignite for much of the firing cycle, resulting in unacceptably high emissions. A drying oven--a converted shipping container powered by the waste exhaust from the kiln--had been designed into the facility, but this would be one of our last mechanical tasks once we were under cover. 
So there we were, out in the mud, dodging rainstorms, cobbling together hardware and complaining about wet wood. We did one run with high moisture content wood, quite memorable for the quantity of smoke produced that day! After that, we used the exhaust from my diesel truck to pre-dry the sawdust, and a half-load of sawdust in the TLUD to pre-dry the kiln load--incredibly inefficient, but better than smothering the planet with smoke. 
Finally, we had a load of dry sawdust in the TLUD, and a reasonably dry load of wood in the kiln basket. Ready to fire! The TLUD was connected to the kiln through three 10cm automotive steel flex tubes, wrapped with insulation, and connected to injection ports in the kiln base. Though we'd already run and tested the TLUD, optimizing heat transfer to the kiln proved tricky. At one point the bonnet of the TLUD got so hot that the automotive flex tubes failed, resulting in an all-hands-on-deck fire drill with welding gloves and baling wire. We eventually figured out that the TLUD temperature was less important than the delta T between TLUD and injection port, and maximum air mass transfer was key. Finally, the kiln load crept up to 300C and could sustain pyrolysis without the TLUD. We switched off the blower, sat back, and watched the show. Click here for an annotated online slide show of the big burn.
Next up, building a roof structure to house the kiln and TLUD, integrating the drying oven, building a new much larger BMC reactor to replace last years mechanical disaster, and a specialty cross-draft pyrolyzer for African oil palm waste (pending approval of funding).
