Build carbonization or buy char: a three-year cash comparison

The conclusion first: run the ratio, then talk capex

Divide the delivered char price by the wet shell price. Below 3.3, do not build carbonization. Above 3.9, a kiln usually pays back inside a year. This is the arithmetic for a 1 t/h finished-carbon line; the larger the line, the more it favours building.

The ratio is not a guess. One tonne of finished carbon needs 1.68 t of char, working back from 62% activation yield and 4% screening loss. That 1.68 t of char needs 6.4 t of shell at 12% moisture when carbonization yield is 30%. Divide 6.4 by 1.68 and you get 3.81: the raw material cost of one tonne of char equals the price of 3.81 t of wet shell. Raise yield to 34% and the number falls to 3.34; drop to 28% and it rises to 4.06.

So the real variables in a three-year cash comparison are the wet shell price, the carbonization yield and the delivered char price, not whether the kiln costs 300,000 more or less. A 1 t/h finished-carbon line consumes about 10,000 t of char a year; a USD-equivalent capex of roughly 170,000-280,000 spread over that tonnage is only 40-65 per tonne of char. On the same line, a yield slippage from 30% to 28% burns shell worth hundreds of thousands a year. Equipment scope and mass balance are on the carbonization furnace page.

The cash structure, the sensitivity table and the four commonly omitted costs below can all be recalculated with your own quotations plugged in.

Compress three numbers into one ratio

Step one, the ratio: delivered char price divided by delivered wet shell price. The threshold moves with yield - 4.06 at 28%, 3.81 at 30%, 3.34 at 34%. Below that line you are paying more for raw material than the market charges for finished char, and no amount of capex optimisation fixes it.

Step two, add processing cost. Power, labour, maintenance and chemicals run 80-140 per tonne of char above 1 t/h, which adds roughly 0.06-0.10 to the ratio at a 1,350 shell price. Small kilns are worse: at the 300 kg/h scale, processing can exceed 200 per tonne, which is exactly why small lines should buy char.

Step three, size the annual gap. Above 1 t/h you need about 10,000 t of char a year; at 0.5 t/h, about 5,000 t. Below roughly 3,000 t a year, fixed labour and environmental costs cannot be absorbed, and the apparent saving on goods is eaten by fixed cost.

The calculation itself is four steps: convert your finished tonnage into annual char demand; multiply by the delivered char price for the buy case; build the make case from shell price, yield and processing cost; divide capex by the annual difference. If payback exceeds three years, ownership is not an argument.

Four cash items that get left out

• Permitting and continuous emission monitoring: a new off-gas train plus CEMS brings year-round consumables and service cost. Buying char hides this inside the supplier price, but it does not remove it.
• Start-up fuel: the kiln needs external fuel for the first 3-4 hours, every time it restarts. Six restarts a year is six fuel bills.
• Operator shifts: one person per shift, three shifts, unless the kiln can genuinely share a crew with the activation line - which most layouts cannot.
• Refractory and downtime: the lining is 8-15% of equipment value, replaced every 3-5 years on a rotary kiln with 10-15 days offline, or 5-8 years on a shaft furnace with 15-20 days. During those days the activation furnace waits, and the loss is measured in finished-product margin, not char price.