Norbert,
 
(snip)

 
It stands to reason that a fueling protocol be established but I find that other variables  affect the burn even more so. The design of the heater itself is of primary concern. How much combustion air is supplied and where it is directed are major factors that affect the burn and emissions.
.
 The combinations of combustion air supply in my heater are:

• Overfire air only
• Underfire air only
• A combination of both
• Controlable afterburn air in the throat and in the bake oven

I can alter these source combinations and do ignition from the top or from the bottom of the crib. I can control where my air is sourced and how much is delivered.  I have installed an air mass flow meter in my air supply to measure the combustion air volume and observed the results.  I don’t have exhaust gas monitoring yet or data logging for the air supply so I have to rely on visual observations for results. The results have been antidotal but have demonstrated a definitive difference in burn rates and apparent cleanness of the burn based on combustion air differences.
 
 My goals are:

• To reach combustion temperatures in the complete combustion range of 1200^o degrees F as fast as possible

•  and hold temperatures at 1500º-1800º F during combustion
• To create as much turbulence as possible to promote complete gas combustion

•  To add sufficient afterburning air and observe that all gasses in the bake oven are fully combusted

• To burn enough fuel to raise the outer mass temperature to a desired level determined by BTU requirements of the dwelling based on outdoors air temperature

 
My preliminary observations are:
                       
                        (1) A burn lit from the bottom with air supplied from under the crib (underfire), lights very quickly, creates an early accelerated draft but initially burns cooler (less than 1000º F). A lot of turbulence is observed and a lot of uncombusted gases are swirling around for the first 15 minutes of the burn. Excess unburned gases are observed in the bake oven for approximately 15 minutes. Final Portions of combustion at the coals stage are complete to ash.

 
            (2) A burn lit on the top of the crib with underfire air supplied was very slow to start, progressed slowly but burned cleanly and finished as above.

 
(3) A burn lit at the top with overfire air started cleanly and the fire progressed rapidly with good turbulence and low observed smoke. The coals portion of the burn lacked sufficient air for a fast burn down of the coals.

 
(4) A burn lit on the bottom with overfire air was very slow starting however once the whole crib was ignited the combustion was the same as above

.
(5) The last 30 observed burns in my heater have been lit in the center rack of a three tier crib. I supply overfire air and partial underfire air. Once combustion is underway I open the bake oven air supply which results in an observed turbulent afterfire of any remaining unburned gases. The partial underfire air burns all the coals to ash. My combustion cycles last from one hour thirty minutes to two hours.