Skip to content
Industrial Unit Converter

Engineering Guide

Boiler Efficiency: Direct Method vs Indirect Heat Loss Method

Published July 3, 2026 · by Industrial Unit Converter Editorial Team

Boiler Efficiency: Direct Method vs Indirect Heat Loss Method

A 50,000 lb/hr firetube boiler running on natural gas at 150 psig burns through 58,100 scf of gas every hour. At $8/MMBTU, that single unit costs the plant $4.2 million per year in fuel. A 1% efficiency improvement saves $50,000/year. A 5% gap between spec sheet and field reality costs $210,000/year, for the life of the equipment.

ASME PTC 4-2017 defines two methods: direct (input-output) and indirect (heat loss). Both should agree within 2% on a well-instrumented unit.

This guide covers both methods, with a 50,000 lb/hr case study, the seven loss categories, HHV versus LHV, and a field audit procedure.

Background: why boiler efficiency varies in the field

A new Cleaver-Brooks CBLE-200 firetube ships with an 84% efficiency guarantee. Three years later, with fouling and a drifted O2 setpoint, the same unit is closer to 79%. The nameplate is a laboratory number on test oil. Real operation includes load swings, fouling, and scale.

Boiler efficiency is the largest operating cost variable in a steam plant. The Boiler Efficiency Calculator handles both methods. The sister guides cover the rest: Steam Quality Calculator, Steam Quality Explained, Steam Boiler vs Hot Water Boiler, and Steam Tables IAPWS-IF97 Deep Dive.

The direct method (input-output)

The direct method divides useful steam energy by fuel energy:

η_direct = (m_steam × Δh) / (m_fuel × HHV) × 100%

where:

  • m_steam = steam mass flow (lb/hr)
  • Δh = h_steam - h_feedwater (BTU/lb)
  • m_fuel = fuel mass flow (lb/hr) or gas volume flow (scf/hr)
  • HHV = higher heating value (BTU/lb for solid or liquid, BTU/scf for gas)

Worked example: 50,000 lb/hr, 150 psig, natural gas

At 150 psig (164.7 psia), saturated steam: h_g = 1194.0 BTU/lb. Feedwater at 220°F: h = 188.5 BTU/lb. So Δh = 1005.5 BTU/lb.

Steam output: 50,000 × 1005.5 = 50.28 MMBTU/hr.

Assume 84% HHV efficiency. Fuel input: 50.28 / 0.84 = 59.85 MMBTU/hr. At 1,030 BTU/scf, that is 58,100 scf/hr.

Annual fuel cost at $8/MMBTU: 59.85 × 8,760 × $8 = $4,195,000/year. A 1% drop adds $50,000/year.

What the direct method requires

Steam flow is measured with an orifice plate, Venturi, or vortex shedding meter. Each has 2-3% uncertainty in clean saturated steam. Fuel flow uses a turbine or Coriolis meter with 1-2% uncertainty. The combined uncertainty is 3-4%, which is why ASME PTC 4-2017 calls for calibrated instrumentation and an 8-hour steady-state run.

At a glance: direct vs indirect

The direct method requires steam and fuel flow meters. The indirect method needs a flue gas analyzer, a thermocouple, and ambient temperature. The direct method is more accurate in principle; the indirect method is faster, cheaper, and the only practical option for a routine audit. Run both once, then run the indirect method quarterly.

The indirect method (heat loss)

The indirect method adds up every place the heat goes that is not the steam:

η_indirect = 100% - Σ(losses)

Losses are a percent of the fuel HHV input. The seven categories from ASME PTC 4-2017 are below.

  1. Dry flue gas sensible heat loss (L_dry): heat in N2, CO2, O2. Largest single loss. 8-12% natural gas, 10-14% #2 oil, 5-8% bituminous coal.

  2. Moisture from combustion (L_H2O): latent heat of water vapor from burning hydrogen. 10-15% for natural gas because CH4 + 2O2 produces 2 moles of H2O per mole of CH4. 8-12% #2 oil. 2-4% coal.

  3. Moisture in fuel (L_fuel_H2O): heat in water entering with the fuel. Negligible for natural gas. 0-1% #2 oil. 2-5% coal.

  4. Unburned CO (L_CO): heat in unburned CO. 0.1-0.5% natural gas, 0.2-1% #2 oil, 0.5-2% coal.

  5. Unburned hydrocarbons (L_HC): heat in unburned fuel that escaped. Below 0.1% natural gas, 0.1-0.3% oil, 0.5-1% coal.

  6. Radiation and convection (L_rad): heat lost from the boiler's outer surface. 0.5-2%. A 50,000 lb/hr Cleaver-Brooks firetube runs around 0.7%. A 5,000 lb/hr Bryan boiler runs closer to 1.5%.

  7. Unaccounted losses (L_other): blowdown heat not recovered, sootblower steam leaks, casing air infiltration. 0.5-1.5%.

Loss breakdown by fuel type (ASME PTC 4-2017)

Loss Category Natural Gas #2 Oil Bituminous Coal
Dry flue gas sensible 8-12% 10-14% 5-8%
Moisture from combustion 10-15% 8-12% 2-4%
Moisture in fuel 0% 0-1% 2-5%
Unburned CO 0.1-0.5% 0.2-1% 0.5-2%
Unburned hydrocarbons <0.1% 0.1-0.3% 0.5-1%
Radiation and convection 0.5-2% 0.5-2% 1-3%
Unaccounted 0.5-1.5% 0.5-1.5% 1-2%
Total losses 20-30% 20-30% 15-25%
Efficiency (typical) 70-80% 70-80% 75-85%

Worked example: 84% efficient natural gas firetube

For the 50,000 lb/hr Cleaver-Brooks from earlier, running at 2.5% O2, 320°F flue, 30 ppm CO, 70°F ambient:

  • L_dry = 8.5%
  • L_H2O = 10.4%
  • L_CO = 0.2%
  • L_rad = 0.7%
  • L_other = 0.2%
  • Sum = 20.0%, so η = 80.0% (HHV)

The 4-point gap between this 80% indirect result and the 84% direct result is unaccounted radiation, blowdown heat, and measurement uncertainty. ASME PTC 4-2017 allows up to 2% disagreement, so 4% means the test is suspect.

HHV vs LHV: the 10% difference that confuses everyone

HHV (Higher Heating Value) is the total heat released when fuel is burned and the water vapor in the flue gas is condensed back to liquid. LHV (Lower Heating Value) assumes the water vapor stays as vapor.

For pure methane, LHV is about 12% lower than HHV (1,012/1,152 = 0.879). For typical pipeline gas at 90-95% methane, the difference is closer to 10%.

The same boiler reports 84% on HHV (ASME PTC 4, US) and 92% on LHV (EN 12953-11, European). The 8-point gap is the most common source of cross-border confusion. A Cleaver-Brooks spec sheet in the US is HHV. A Viessmann or Bosch catalog in Germany is LHV. Normalize before comparing.

Condensing boilers recover the latent heat from flue gas water vapor, reaching 92-98% on HHV (Weil-McLain SlimFit, Burnham Alpine).

Field audit procedure

A credible indirect-method audit takes one shift and four tools: a portable flue gas analyzer (Testo 340, Bacharach Insight Plus, or MSA 5Star), a Type K thermocouple, an ambient air thermometer, and the IAPWS-IF97 steam tables.

  1. Steady the boiler: hold the firing rate constant for at least 30 minutes. The most common error is auditing during a load swing.

  2. Measure O2 and CO in the flue gas: insert the probe past the last heat recovery section (after the economizer if there is one, not before). Record at 1-minute intervals for 10 minutes and average.

  3. Measure flue gas temperature: thermocouple in the same plane as the gas analyzer probe. Should be 50-100°F above T_sat. At 150 psig, T_sat is 366°F, so 420-470°F is normal.

  4. Measure ambient temperature: 5 ft from the boiler, shielded from the casing.

  5. Look up steam enthalpies: h_steam and h_feedwater at the measured drum pressure and feedwater temperature. IAPWS-IF97 values are reliable to ±0.1%.

  6. Calculate the seven losses using ASME PTC 4-2017 equations or the Boiler Efficiency Calculator.

  7. Compare to baseline: deviation more than 3% from the manufacturer's spec sheet indicates fouled tubes, misaligned burner, leaky air damper, or carryover.

A well-run indirect audit on a 50,000 lb/hr boiler takes about 2 hours of instrument time plus 1 hour of calculation. The savings, when the audit finds a problem, are an order of magnitude higher. The full system is covered in the Steam System Design, Headers, Condensate, and Traps guide.

Why 2-3% O2 is the design target

Most natural gas burners are tuned for 2-3% O2 at design load. Below 2%, CO and unburned hydrocarbons rise. Above 3%, the dry flue loss rises because more nitrogen is heated and sent up the stack. A 1% O2 increase costs 0.5-0.8% efficiency.

Typical efficiencies by boiler type (HHV basis, US)

Numbers below are HHV basis at design load, 70°F ambient, 2-3% O2, for a well-maintained unit in the field.

Boiler Type HHV Efficiency Year of Mfr
Old cast iron hot water (1960s-1980) 65-72% 1960-1980
Modern cast iron hot water (1990s+) 78-85% 1990-present
Firetube steam (1990s+) 80-86% 1990-present
Watertube steam (2000s+) 85-90% 2000-present
Condensing hot water (2000s+) 92-98% 2000-present
HRSG (with gas turbine) 80-85% varies
Electric boiler 99-100% n/a

A 95% Weil-McLain in a US catalog is 95% HHV. A 109% Bosch in a German catalog is 109% LHV. Consistent only on the same basis.

HHV by fuel (US standard conditions)

Fuel HHV Notes
Natural gas (pipeline) 1,030 BTU/scf 1,020-1,050 with composition
Propane 2,516 BTU/scf 91,500 BTU/gal liquid
#2 fuel oil (distillate) 138,500 BTU/gal 19,200 BTU/lb
#6 fuel oil (residual) 150,000 BTU/gal 18,500 BTU/lb
Bituminous coal (US) 12,500 BTU/lb avg 10,500-14,000 range
Wood pellets 8,000 BTU/lb 7,500-8,500 range

Common mistakes that cost money

The six errors below are the ones a plant engineer is most likely to make. Dollar numbers are based on the 50,000 lb/hr firetube from earlier (annual fuel cost ~$4.2M).

Cost per percentage point

On a 50,000 lb/hr, 150 psig natural gas boiler at 84% efficiency, 1% = $50,000/year in fuel at $8/MMBTU. 5% = $250,000/year. For a 200,000 lb/hr watertube at 88%, 1% = $200,000/year, 5% = $1M/year.

1. Reporting efficiency on LHV instead of HHV. A European vendor reports 92% LHV. A US plant engineer treats it as 92% HHV. The same boiler on the same fuel reads 84% on HHV. The 8-point gap is a basis mismatch, not a malfunction. The consequence: $336,000/year of overstated savings. Always state the basis in writing on every report.

2. Measuring O2 before the economizer. The O2 reading is not affected by heat exchange, but the flue temperature at that location is higher than the post-economizer value, inflating the dry flue loss. Reported efficiency comes in 2-3% lower than actual. On the 50,000 lb/hr case, that is $84,000-126,000/year of phantom savings. Fix: insert the probe after the last heat recovery section.

3. Ignoring surface radiation loss. A plant engineer sees L_dry + L_H2O + L_CO = 19% and declares efficiency 81% on the spot, skipping L_rad. ASME PTC 4-2017 requires radiation loss estimation based on boiler surface area and casing temperature. Skipping it is non-compliant and overstates efficiency by 0.5-2%. On the case boiler, 1% overstatement = $50,000/year of phantom savings.

4. Running a one-time audit and assuming it is permanent. Boiler efficiency drifts. Fouling on the fire side adds 1-3% to the dry flue loss over 12-18 months. Burner wear shifts O2 by 1-2 points. A single audit at one load is a snapshot, not a baseline. ASME recommends quarterly audits on critical units. Skipping the trending costs a 50,000 lb/hr unit $30,000-80,000/year of undetected drift.

5. Using a nameplate efficiency for a 10-year-old boiler. The Cleaver-Brooks nameplate was measured on test oil at sea-level conditions on commissioning day. A 10-year-old firetube runs 4-8% below the nameplate. A $5M/year fuel budget based on the nameplate is short by $200,000-400,000/year.

6. Confusing combustion efficiency with thermal efficiency. Combustion efficiency is the indirect method (flue gas losses only). Thermal efficiency is the direct method (energy absorbed by water). They are within 1-2% on a well-run boiler but are not the same number. A Cleaver-Brooks spec sheet lists thermal efficiency. A field audit reports combustion efficiency. Comparing them as if they were the same leads to a 2-3% misread, or $100,000/year on a 50,000 lb/hr unit.

Standards and best practices

  • ASME PTC 4-2017 (Fired Steam Generators): US performance test code. Defines direct and indirect methods, instrumentation, run duration, and uncertainty.
  • EN 12953-11 (2018): European standard for shell boiler acceptance tests. Uses LHV. For US engineers reading European specs, normalize by adding roughly 10% to the natural gas efficiency.
  • IAPWS-IF97 (2012 release): international standard for steam properties. Required for accurate enthalpy lookups.
  • DOE/EE-0288 (Steam System Survey Guide, 2014 revision): practical field guide for industrial steam systems.
  • ASME BPVC Section I (2023): rules for power boiler construction.
  • Cleaver-Brooks Boiler Book (current edition): vendor reference for firetube and watertube design and efficiency ratings.
  • North American Combustion Handbook (1988, still in print): deep reference for heat transfer and combustion calculations.
  • ASHRAE Handbook (HVAC Applications, 2023): chapter on boiler plants.

Frequently asked questions

Q1. The direct method gives 84%, the indirect method on the same day gives 80%. Which is right?

Both, within their uncertainty. ASME PTC 4-2017 allows up to 2% disagreement. A 4% gap suggests a miscalibrated steam flow meter, a missed loss category, or a real boiler problem. Re-run, fix the most likely culprit, and confirm the gap closes.

Q2. Can I trust a portable flue gas analyzer for an indirect method audit?

Yes. A Testo 340 or Bacharach Insight Plus reads O2 to ±0.2% and CO to ±10 ppm, well within the precision needed. The bigger source of error is the flue temperature measurement (use a sheathed Type K). For sub-1% precision or emissions compliance, use a laboratory-grade analyzer.

Q3. Is a 95% condensing boiler real, or marketing?

Real, on HHV basis. A Weil-McLain SlimFit or Burnham Alpine at 140°F return water and 2% O2 will hit 95-96% HHV in the field. The condensing heat exchanger drops the flue gas below 130°F, condensing most of the water vapor and recovering its latent heat.

Q4. How does O2 setpoint affect efficiency?

For natural gas, every 1% increase in O2 above the optimum adds about 0.5-0.8% to the dry flue loss. Going from 2% to 4% O2 costs roughly 1.5% efficiency, or $63,000/year on the 50,000 lb/hr case. Most burners run at 2-3% O2 at design. A reading of 4% is the first sign of a burner trim problem or air damper leak.

Q5. Why does the European boiler spec say 109% efficient?

It is 109% LHV. On HHV, the same unit reads about 98%. The "over 100%" appearance is normal for LHV reporting of condensing boilers, because LHV excludes the latent heat the condensing boiler recovers. The boiler is not creating energy. US and European ratings of the same product are consistent once you know the basis.

Q6. How often should I audit a critical boiler?

Quarterly for any boiler that is the primary steam source for a continuous process. Annually for standby units. Each audit takes 2-4 hours and the cost is recovered if it catches a single 1% drift ($30,000-50,000/year on a 50,000 lb/hr unit). The Steam System Design, Headers, Condensate, and Traps guide covers what else to check.

References and further reading

  • ASME PTC 4-2017, Fired Steam Generators (Performance Test Code). ASME, New York, 2017.
  • EN 12953-11:2018, Shell Boilers, Part 11: Acceptance Tests. CEN, Brussels, 2018.
  • IAPWS-IF97, Industrial Formulation 1997 for the Thermodynamic Properties of Water and Steam, 2012 release.
  • US Department of Energy, Steam System Survey Guide (DOE/EE-0288), 2014 revision.
  • ASME, Boiler and Pressure Vessel Code, Section I: Rules for Construction of Power Boilers, 2023 edition.
  • Cleaver-Brooks, Boiler Book, current edition.
  • North American Combustion Handbook, 1988 edition (still in print).
  • ASHRAE, ASHRAE Handbook: HVAC Applications, 2023 chapter on boiler plants.

Related tools and calculators

Share LinkedIn X Email