Based on the simplicity and effectiveness of the Experiment 3 trials, a larger reactor was constructed. The outer shell was a 55 gallon drum with no air holes and an open top. The fuel chamber was made by extensive modification of a second 55 gallon drum - with a portion of side wall removed so the bolted-together remnant shell had a diameter approximately 10 cm less than the outer “draft shell”. Part of the tank bottom was modified to make a bottom for this inner chamber. Approximately 30 holes of 1/2-inch diameter (12mm), were made in the bottom, and it was fitted with bolts for “legs” to hold it 4cm above the bottom of the draft shell. A drum lid was modified with a central chimney of light gauge sheet-metal. The height was approximately 90 cm and the diameter was approximately 25 cm.
This design proved to be faulty. The chimney-induced updraft was not sufficiently energetic to draw fresh air from the annular space through the bottom of the fuel chamber. The fuel would cease to burn efficiently after flames reached about 1/3 of the depth of the barrel. Holes were added to the draft shell to improve air flow (first four holes of approximately 25 cm^2, and later four more such holes). But these changes did not significantly improve the performance.
It should be noted that top-lighting of a wood fire is not traditional, and achieving a successful burn is highly dependent on dry fuel and placement of fuel to permit relatively uniform air flow. The results of test burns are affected by many factors and other experimenters might find better success than this failed boy scout.
Ultimately, this experimenter dispensed with the fuel chamber and conducted burns in the now-aerated draft shell barrel with the chimney affixed. A burn on March 20, 2024 using a mix of cardboard debris, lumber scrap, and branches up to 10 cm in diameter burned relatively thoroughly in 30 minutes. Larger pieces of wood seemed to convert to char to a depth of about 2 cm with 30 minutes of high temperature. So some unburned pieces remained in the bottom when the burn was quenched with water.
Making biochar in a TLUD is a balancing act between complete conversion to char, and preservation of produced char. The high temperature in the accumulating coal bed is converting carbon to CO2, not just pyrolyzing wood. The barrel skin in the coal zone exceeds 500C at the “end” of the burn, while the chimney - which was 450C or greater early in the burn, is down to about 350C when the flames in the barrel seem to be subsiding. When the fire is quenched, there is still copious smoke produced.