chiptec in vermont makes several models of chip boilers,its not the cost of the boiler but the conditioning and delivery system of the chip that makes it out of residential reach. they use cleaver-brooks controls with a ametek in situ flue gas analyser. my experimentation so far with cordwood and 02 control is a '' compromise because of the burning stages of cordwood , start-up, with a time period of 02 deficit , next stage a optimum combustion with a 2to 4 % 02 reading, next stage about 2/3 through burn, o2% will gradually rise to about 13% about 75% coals and a few small logs left. the last third basically just coals will rise from 13 to 20.9% end of burn. in the garn when it puffs the o2 reading is at 0%, so in a nut shell with a static fan speed , under aired for first third of burn a good o2 range for second third of burn and over aired for last third of burn. this cycle can be manipulated with size of splits, frequency of reload and moisture content. i suspect most '' lambda controlled'' cordwood boilers are designed to try to limit theses '' compromises''by carefully designing fire box capacities and reccomending small splits, primary/ secondary air ratios and volumes to try to minimize the cordwood burn cycle swing. I am in the process of defining and testing programable runs in the allen bradley plc for different moisture content wood. one button for a program that will controll dry wood,another for average wood ,and a third for less than optimum wood., these programs basically differ in initial static fan speed start-up time, also looking to end burn with a declining flue temp for a given length of time, and of course a overide toggle to run manually. this experiment, ongoing is leading me to believe that a continious fuel delivered system, pelletts or chips for example is more successfully controlled with o2 control. i would appreciate any comments or ideas that you think might be incorporated, thanks
tom