Home / Industries / Industrial Boilers & Process Heat

Industrial boilers, steam plants & process heaters

Excess air is the highest-leverage efficiency variable a steam or process-heat engineer controls — and on most installed boilers it drifts far from optimum across the firing range. Getting it right is worth 2–5% of the fuel bill.

The excess-air balance

Complete combustion in a real furnace requires more than the stoichiometric air quantity — mixing is never ideal. But every unit of air beyond what combustion needs is heated from ambient to stack temperature and thrown away. The balance point is sharp on both sides: insufficient excess air leaves unburned combustibles — CO, soot, unburned hydrocarbons — which is both an efficiency loss and a safety issue; surplus excess air inflates the dry flue-gas loss, the dominant stack loss on a gas-fired unit.

For natural gas firing, well-controlled burners run roughly 5–10% excess air, corresponding to about 1–2% stack O₂; a common optimization target is 10% excess air (≈1.7% O₂). Combustion efficiency is read directly from two stack measurements — flue-gas O₂ (or CO₂) and net stack temperature — which is what makes this problem so tractable: the feedback signal is cheap, continuous and unambiguous.

Where installed boilers actually operate

Most industrial boilers don't hold their commissioning setpoint. Units with single-point positioning control — a mechanical jackshaft linking the fuel valve to the air damper — are tuned at one condition and drift everywhere else: documented surveys show excess-air spreads of 40 to 80 percentage points between high fire and low fire on jackshaft-controlled boilers. Linkage wear, hysteresis, seasonal air-density change and burner fouling push the operating point further from optimum between manual tune-ups. Manual tuning itself is typically set fat — around 3% O₂ (≈15% excess air) — precisely because the tuner knows conditions will drift and errs on the safe, wasteful side.

The control hierarchy

Combustion control on boilers is a well-defined ladder, and each rung recovers fuel:

  • Positioning control (jackshaft) — one actuator, fixed mechanical fuel/air relationship. Cheap, drifts badly, no composition awareness.
  • Parallel positioning (linkageless) — independent servo control of fuel valve and air damper against a programmed curve. Holds the commissioned fuel/air characteristic across the full firing range.
  • O₂ trim — closed-loop correction of the air setpoint from a stack zirconia O₂ probe. Compensates drift from air density, fuel supply and burner wear automatically. State of the art for excess-air control.
  • Cross-limiting (lead-lag) control — air leads fuel on load increases, fuel leads air on decreases, so the mixture can never pass through a fuel-rich state during transients. The safety architecture that lets a calibration run close to optimum.
  • Composition-aware feed-forward — for variable fuels (refinery gas, blended streams, multi-fuel firing), feed-forward from measured or inferred fuel properties repositions the fuel/air curve before the stack O₂ ever sees the excursion.

Variable fuel is where calibration earns its keep

A fixed fuel/air curve assumes a fixed fuel. Refinery fuel gas, blended process streams, hydrogen-enriched gas and multi-fuel boilers violate that assumption continuously — heating value and stoichiometric air demand move with the process upstream. O₂ trim alone chases these excursions after the fact; a calibration built on fuel characterization anticipates them. This is the same thermally calibrated stoichiometry we apply to engines and turbines, pointed at a furnace.

FuelTypical excess airStack O₂Primary risk if untrimmed
Natural gas5–10%1–2%Stack heat loss climbs fastest per point of drift
Fuel oil10–15%2–3%Soot at low excess air, heat loss at high
Refinery / variable gascomposition-dependentmust floatFixed curves are wrong by design — needs feed-forward

What disciplined combustion control is worth

Documented results for excess-air optimization on industrial boilers run 2–5% fuel savings, with 6–7% achievable where drift was severe or burners were out of calibration. A useful rule of thumb: reducing stack O₂ from 4.5% to 2.0% is worth roughly 4% in fuel. On a steam plant burning tens of thousands of MMBtu a year, payback on a calibration program is typically measured in months.

The vocabulary that matters

Excess Air

Air supplied beyond the stoichiometric requirement. Necessary for complete combustion in a real furnace; every surplus point is heated and discarded up the stack. The central control variable of boiler efficiency.

Stack Loss

Heat leaving in the flue gas — the dominant loss on a gas-fired boiler. Read from stack temperature and O₂. Dry flue-gas loss scales with excess air; latent loss with fuel hydrogen content.

O₂ Trim

Closed-loop correction of combustion air from a stack zirconia O₂ probe, holding a load-dependent excess-air setpoint automatically as air density, fuel and burner condition drift.

Single-Point Positioning

Jackshaft control: one actuator, one mechanical fuel/air linkage, tuned at one point. Documented drift of 40–80 percentage points of excess air across the firing range.

Cross-Limiting Control

Lead-lag logic ensuring air always leads fuel upward and fuel leads air downward, so transients can never pass through a fuel-rich, CO-producing state. Enables running close to optimum safely.

Turndown Ratio

The ratio of maximum to minimum stable firing rate. High turndown avoids cycling losses — but excess-air control at low fire is where most boilers waste the most.

What's your stack O₂ at low fire?

If you don't know, that's the answer. A week of stack data and your fuel analysis is enough for a defensible savings estimate.

Start the Conversation

Related industries