"Our boiler is 85% efficient." Measured when? At what load? On which method — and on gross or net calorific value? Boiler efficiency is the most quoted and least examined number in the boiler house. This lesson gives you the two ways to measure it, the five places the missing percent go, and the two knobs that move it.

Two ways to the number

The direct method divides output by input: steam × (hg − feed heat) ÷ fuel × GCV. One line of arithmetic, but it needs honest metering of steam and fuel — weighed husk, not "lorry-loads" — and it tells you what the efficiency is, never why.

The indirect method starts at 100% and subtracts each measured loss. More work, but it hands you the repair list: you finish knowing where every missing percent went. Serious audits use both — direct for the headline, indirect for the diagnosis, and agreement between them as the credibility check.

Where the percent go

useful heat into steam ≈ 75% dry flue gas ≈ 13% — the big one moisture + hydrogen in fuel ≈ 8% radiation ≈ 1.5% · blowdown ≈ 1.5% unburnt in ash ≈ 1% Typical bands for a husk-fired shell boiler in fair condition — your stack analyser gives the real split.
One hundred kilograms of husk, audited. The stack owns most of the loss — which is good news, because the stack is measurable and adjustable.
  • Dry flue gas — hot gas leaving the chimney. Grows with stack temperature and with excess air. The main handle, detailed below.
  • Moisture and hydrogen — water in the fuel, plus water formed by burning the fuel's hydrogen, all evaporated in the furnace and gone up the stack as vapour. For biomass this is structural: fresh husk at 10% moisture is honest fuel; monsoon husk at 15%+ burns your efficiency before it burns. (It is also why a quoted efficiency must say GCV or NCV — the same boiler scores several points "better" on NCV arithmetic with nothing changed.)
  • Unburnt fuel — black ash is money in the ash pit; grate condition and air distribution.
  • Radiation and convection — the hot casing; roughly constant in kcal, so its percentage doubles at half load. Boilers idling at low fire score badly.
  • Blowdown — hot water deliberately drained to control dissolved solids. Often left out of the efficiency number entirely; its own lesson covers rates and heat recovery.

Excess air — the main handle

Combustion needs more air than the chemistry minimum, or smoke and unburnt losses explode. But every kilogram of extra air is cold nitrogen heated to stack temperature and thrown away. The flue-gas oxygen reading is the dial: 21 ÷ (21 − O₂%) gives the air factor.

O₂ = 6% → excess air ≈ 40% (a fair target for grate-fired husk) O₂ = 10% → excess air ≈ 90% (heating the sky — common on neglected dampers)

Two working rules of thumb, both approximate but directionally reliable: every ~20 °C cut in stack temperature ≈ 1 percentage point of efficiency, and trimming excess air typically buys a further point or two — because less air also lowers the stack temperature. The two knobs turn together.

The numbers — three points on the practice plant
Today: 75% efficient · 29 t husk/day (from lesson A1) Found: O₂ 9% → excess air ~75% · stack 220 °C Fixed: damper trim + secondary-air balance + economiser cleaned → O₂ 6%, stack 180 °C → efficiency ≈ 78% Fuel at 78%: 29 × 75 ÷ 78 ≈ 27.9 t/day → saving ≈ 1.1 t/day ≈ ₹2,750/day ≈ ₹10 lakh/year — for damper adjustments and a brush.

Bands, honestly stated: what a tune-up yields depends entirely on how far out the boiler was. Measure first — a ₹15,000 flue-gas survey before believing any promised percentage, including this one.

At site
  • Log daily: stack temperature, flue O₂ (or CO₂), feedwater temperature, steam pressure, fuel consumed. Five numbers, one line, the whole story over a month.
  • Stack temperature creeping up at constant load = fouling — soot side or scale side. It is the earliest cheap warning the boiler gives.
  • Check efficiency at the load the boiler actually runs, not at the test load. Radiation loss punishes chronic low-fire running; right-sizing beats tuning.
  • Ask the O₂-trim question when the burner modulates often: automatic trim holds the setpoint that manual dampers hold only on the day they were set.
Pin this
  • Direct method = the headline; indirect method = the repair list; both together = credible.
  • The stack owns the loss: dry flue gas ~13% typical, driven by stack temperature × excess air.
  • O₂ is the dial: 21 ÷ (21 − O₂). Target the band your fuel and firing can honestly hold.
  • ~20 °C off the stack ≈ 1 point of efficiency. Every point ≈ ₹3–4 lakh/year at practice-plant scale.
  • Always ask: GCV or NCV, and at what load?
Steam stories

The flow diagram with fat and thin arrows that every energy auditor draws has a boiler pedigree: Captain Matthew Sankey drew the first one in 1898 — for a paper comparing the energy balance of a real steam plant against an ideal one. The diagram style got his name; the subject it was invented for was exactly this lesson.