Purchased Steam and Hot Water Emissions
Electricity arrives at your meter in one honest unit — the kilowatt-hour — and a single factor turns it into carbon. Purchased steam and hot water arrive in pounds, kilograms, gigajoules, or “heat delivered,” and none of those becomes carbon until you have crossed two conversions that electricity never demands: how much energy the thermal medium actually carried, and how efficiently it was generated.
Get the enthalpy and the efficiency right, and purchased heat is a clean Scope 2 line. Skip them, and every steam number is a guess wearing a decimal point.
Purchased steam/hot-water emissions = metered thermal energy ÷ generation efficiency × the fuel or supplier heat factor, reported under Scope 2. The two steps electricity skips — converting the metered unit to energy, and dividing by generation efficiency — are where most errors live.
Citation-grade reference for measuring greenhouse-gas emissions from purchased steam, hot water, and district heating and cooling under GHG Protocol Scope 2 Guidance. Defines the metering-unit problem, the four-tier factor hierarchy (supplier-specific, fuel-based reconstruction, default district-heat factor, and electric-proxy), the metered-quantity-to-CO2e calculation chain, CHP/cogeneration allocation methods, district-cooling accounting, and market-based thermal claims. Uses live DEFRA 2026 district-heat factors and EUROSTAT renewable-share data via the MasterBrain, with three hardcoded audit-record worked examples verifiable in the Purchased Steam & Hot Water Calculator. GWP basis: IPCC AR6 GWP-100 per IPCC AR6.
The metering-unit problem: steam is not electricity
Named concept · Citable definition
The metering-unit problem
Purchased thermal energy is billed in whatever unit the supply contract specifies — mass of steam (lb or kg), volume of hot water, delivered energy (kWh, GJ, or MMBtu), or heat delivered at the interface — but an emission factor is defined per unit of energy. Converting a mass or volume reading into an energy quantity requires the thermal medium’s enthalpy (its actual heat content at the supplied temperature and pressure), and converting delivered heat into emitted carbon requires the generation efficiency of the plant that produced it. Neither step exists in electricity accounting, and both are where purchased-heat inventories most often go wrong.
⚐ The central editorial tension this page resolves
A steam meter reading of “500 tonnes” is not an energy quantity — it is a mass, and its energy content depends on the steam’s pressure and temperature. Saturated steam at low pressure carries less usable heat than superheated steam at high pressure, so the same mass reading can represent materially different energy, and therefore different emissions. Worse, the carbon emitted to generate that steam depends on the boiler or CHP efficiency, which the buyer rarely sees. The metering-unit problem is the structural reason a purchased-heat figure cannot be read from a bill alone — it must be paired with an enthalpy basis and a generation-efficiency assumption, and those two assumptions dominate the uncertainty of the whole line. The enthalpy and efficiency constants a rigorous engine needs are named in this methodology as steam.framework.*; they are supplied from the contract or from published steam tables, not read blindly from a factor library.
Mass, energy, and heat-delivered — three ways heat is billed
The first task in any purchased-heat inventory is to resolve the billing unit to a common energy basis. The reference conversions below are exact SI relationships (unit conversions, not emission factors), and are the bridge every subsequent step depends on.
| Billed as | Convert to energy via | Reference relationship |
|---|---|---|
| Mass of steam (kg, lb, tonne) | × enthalpy at supplied P/T, − feedwater enthalpy | Useful heat (kJ) = mass (kg) × (hsteam − hfeedwater) kJ/kg |
| Volume of hot water (m³, litre) | × density × specific heat × ΔT | Heat (kJ) = volume × ρ × cp × (Tflow − Treturn) |
| Delivered energy (kWh, GJ, MMBtu) | Direct — already energy | 1 GJ = 277.78 kWh; 1 MMBtu = 293.07 kWh; 1 kWh = 3,600 kJ |
| Heat delivered at interface | Meter reads useful heat directly | Apply the delivered-heat factor with no efficiency step |
Enthalpy values are read from published steam tables (IAPWS-IF97) at the supplied pressure and temperature — for saturated steam they typically fall in the range 2,675–2,800 kJ/kg, but the correct figure is the one for your actual supply conditions, and the feedwater return enthalpy is subtracted because you only pay carbon for the heat that was added. These are engineering constants from a steam table, not emission factors; supply them from your contract or the table, never from memory.
Provider directory
Now it has to hold up all year.
See who does this work. Every listing names the standards it works to, and paid placements are labelled. Including KERAMIDA and The Carbon Trust.
Browse 6 carbon accounting & inventory providers →Do this work? A listing is US$390 a year. Get listed →
GHG Protocol placement and the dual-reporting parallel
Purchased steam, hot water, and district heating are named indirect energy purchases and belong in Scope 2 alongside purchased electricity, per the GHG Protocol Scope 2 Guidance. The same architecture that governs electricity applies: a location-based figure using an average or default heat factor, and — where the supplier can substantiate a specific generation mix — a market-based figure. The location-based and market-based distinction is not an electricity-only concept; it maps directly onto purchased heat.
| Reporter situation | Emission | Placement |
|---|---|---|
| Buys steam / hot water / district heat from a third-party network | Emissions embodied in the purchased heat | Scope 2 purchased heat/steam |
| Buys district cooling from a third-party plant | Emissions of the electricity/heat driving the chillers | Scope 2 purchased cooling |
| Upstream fuel extraction/processing of the fuel burned to make the heat | Well-to-tank emissions of the generation fuel | Scope 3 Cat 3 — the WTT companion |
| Generates its own steam on-site by burning fuel it owns | Combustion of the owned fuel | Scope 1 — not purchased heat; see natural gas combustion |
The last row is the boundary that most often trips reporters: heat you generate by burning your own fuel is Scope 1 combustion, not Scope 2 purchased heat. Only heat that crosses an ownership boundary — bought from a supplier — is Scope 2. And exactly as with purchased electricity, the well-to-tank emissions of the generation fuel sit in Scope 3 Category 3, reported as a companion line, never folded into the Scope 2 total.
The four-tier factor hierarchy
Named concept · Citable definition
The purchased-heat factor hierarchy
The ordered preference for selecting an emission factor for purchased thermal energy, in descending order of specificity: Tier 1 supplier-specific factor (the network’s own disclosed emission rate), Tier 2 fuel-based reconstruction (the generation fuel factor divided by plant efficiency), Tier 3 default district-heat factor (a published average such as DEFRA’s), and Tier 4 electric-proxy (grid factor for electrically generated heat where no heat-specific data exists). Apply the highest tier for which you hold qualifying evidence.
Unlike the metering conversions, this hierarchy is ranked by quality — a supplier-specific factor beats a reconstruction, which beats a default. The reporter uses the most specific tier the evidence supports, and documents why a lower tier was necessary where one is used.
delivered heat ÷ efficiency × fuel factor. The fuel factor comes live from [gc_fuel] (e.g. natural gas at 2.02633 kg CO₂e/kWh) [GreenCalculus fuels.gbr.natural_gas.m3 · DEFRA 2026 'Fuels'!D40 · v2026.193]. Efficiency is a steam.framework.* engineering input, supplied from the contract or a typical band (boilers 80–90%, per GHG Protocol Scope 2 Guidance).
The calculation chain
Whatever tier is chosen, the physical calculation runs through the same five stages. The tier only changes stage four (which factor); stages one to three are the metering-and-efficiency work that electricity never requires.
useful heat is the enthalpy-resolved energy delivered, efficiency converts delivered heat to fuel input (skip it if the factor is already a delivered-heat factor, e.g. DEFRA Tier 3), and emission factor is the tier-appropriate value. The efficiency divisor is the single most sensitive assumption in the chain.Read the metered quantity
Mass of steam, volume of hot water, or delivered energy — as billed. Record the metering unit and the supply conditions (pressure, temperature, flow/return ΔT).
Resolve to useful heat (energy)
Apply enthalpy (steam) or ρ·cp·ΔT (hot water) from published steam tables. This is the steam.framework.* enthalpy step — an engineering constant supplied from a steam table, not an emission factor. Delivered-energy meters skip this stage.
Divide by generation efficiency
Convert useful heat to fuel input using the plant efficiency (boilers typically 80–90%). Skip this stage only when the factor is already a delivered-heat factor (Tier 1 supplier-specific or Tier 3 DEFRA default already embed generation and losses).
Apply the tier-appropriate factor
Supplier-specific (Tier 1), fuel factor via [gc_fuel] (Tier 2), DEFRA district-heat via [gc_factor] (Tier 3), or grid factor via [gc_grid] (Tier 4). This is the only stage the tier choice touches.
Report CO₂e on a stated GWP basis, plus the WTT companion
Aggregate to tCO₂e on the AR6 basis (methane fossil = 29.8 [GreenCalculus gwp.CH4_fossil.ar6_100 · IPCC AR6 WGI Ch 7 Table 7.SM.7 (2021) — AR6 GWP-100], N₂O = 273 [GreenCalculus gwp.N2O.ar6_100 · IPCC AR6 WGI Ch 7 Table 7.SM.7 (2021) — AR6 GWP-100]). Report the Scope 3 Cat 3 well-to-tank companion separately — the DEFRA heat WTT row is live via district_heating.gbr.district_wtt.
The generation-efficiency divisor dominates the result. A steam figure computed at 90% assumed efficiency versus 75% differs by 20% before any other input changes. When the supplier does not disclose efficiency, state the assumed band and treat the resulting range as the reported uncertainty — a single-point steam figure with an undisclosed efficiency overstates its own precision.
CHP / cogeneration allocation
When purchased heat comes from a combined heat and power (CHP) plant, the plant’s fuel emissions must be split between the electricity and the heat it co-produces — and the split method changes the heat factor materially. This is the purchased-heat analogue of the Scope 2 instrument-choice problem: the arithmetic is defensible under more than one method, and the reporting requirement is to disclose which was used. For the standalone treatment see the CHP cogeneration allocation methodology.
| Allocation method | Rule | Assigns heat… | Best use |
|---|---|---|---|
| Energy (proportional) | Split fuel emissions by each output’s share of total energy output | Its share of raw MJ out | Simple, transparent; the common default |
| Efficiency (work-potential) | Weight each output by the fuel a separate plant would need to make it | Its reference-plant fuel-equivalent share | Rewards the thermodynamic quality difference between heat and power |
| Exergy | Split by the useful work potential (exergy) of each output | Its exergy share (heat < electricity per MJ) | Most thermodynamically rigorous; least common in practice |
The method choice is not cosmetic. In worked example C below, the same CHP plant produces a heat footprint of 297 tCO₂e under the energy method and 209 tCO₂e under the efficiency method — the efficiency figure is roughly 30% lower, from the allocation rule alone. Because heat carries less work potential than electricity per unit energy, the efficiency and exergy methods assign it less carbon than the energy method does. Disclose the method, or the heat figure is unfalsifiable.
District heating and cooling as purchased thermal
District networks are purchased heat (or cooling) at community scale, and they introduce two accounting refinements beyond a single-supplier steam contract: a network renewable-share that lowers the effective factor, and — for cooling — a metering basis in tons of refrigeration rather than heat delivered. The dedicated treatment is the district heating and cooling methodology; the essentials for a purchased-heat inventory follow.
Renewable-share adjustment (EUROSTAT)
A district network’s carbon intensity depends on how much of its input is renewable or recovered waste heat. EUROSTAT publishes a per-country renewable-plus-waste-heat share of district-heat final energy, which adjusts the effective factor downward for cleaner networks. The spread across Europe is wide.
| Country | District-heat renewable + waste-heat share | Implication |
|---|---|---|
| Denmark | 72.0 % | Among Europe’s cleanest heat networks — a low effective factor |
| Spain | 82.0 % | High renewable share; effective factor well below fossil baseline |
| EU-27 aggregate | 34.5 % | Roughly a third renewable/waste-heat network-wide |
| Albania | 3.4 % | Low renewable share; effective factor near the fossil baseline |
Shares render live from MasterBrain (EUROSTAT) via [gc_factor]; the section carries ~30 countries under district_heating.<iso3>.eurostat.renewable_share. Use the renewable share to adjust a fossil-baseline factor only where the supplier has not already disclosed a network-specific factor (which would supersede it as Tier 1 evidence).
District cooling
District cooling is metered in tons of refrigeration and driven mostly by electricity, so its footprint is the plant’s electrical input intensity (kW per ton delivered) multiplied by the grid factor. MasterBrain carries ASHRAE plant-efficiency benchmarks: a typical-plant benchmark of 1 kW/ton [GreenCalculus district_cooling.framework.plant_efficiency_benchmark · ASHRAE DC 2013], against which real plants range from roughly 0.89 kW/ton (efficient) upward. Multiply the delivered cooling by the plant’s kW/ton and then by 0.131 kg CO₂e/kWh [GreenCalculus grid.gbr.electricity.location_based] (or the relevant grid) to reach CO₂e.
Market-based thermal claims
As with electricity, a supplier may offer a market-based thermal product — heat certified as renewable, or waste heat recovered from an industrial process. The market-based claim is only as good as its substantiation, and the methodology’s posture mirrors the Scope 2 market-based discipline documented in the market-based method.
- Certified renewable heat requires the same instrument rigour as a renewable electricity claim — a substantiated certificate, not a marketing statement. An unsubstantiated “green heat” tariff defaults to the location-based factor.
- Recovered waste heat is often near-zero-carbon at the point of delivery because its generation emissions are allocated to the primary process — but this must be documented, not assumed, and double-counting between the primary process and the heat buyer must be excluded.
- Dual reporting still applies. A market-based thermal figure does not replace the location-based figure — report both, exactly as with electricity under GHG Protocol Scope 2 Guidance.
Worked examples — audit records
Three audit-record snapshots at this page’s review date (2026-07-08). All numerics are hardcoded (worked examples are audit records — they must reconcile to their stated inputs regardless of future MasterBrain refreshes). Thermodynamic constants (enthalpy, efficiency) are illustrative engineering inputs from published steam tables, cited as such — a reader supplies their own supply-condition values. Each is replicable in the Purchased Steam & Hot Water Calculator.
Scenario: a factory buys 500,000 kg of steam per year, supplied at conditions giving a net useful enthalpy of 2,750 kJ/kg (illustrative steam-table value); the supplier discloses natural-gas generation at 85% boiler efficiency but no finished factor. Reconstruct under Tier 2.
Enthalpy (2,750 kJ/kg) and efficiency (85%) are illustrative engineering inputs from a steam table and the supplier disclosure — not MasterBrain values. Natural-gas factor is a hardcoded review-date snapshot of the live 2.02633 value (0.18231 kg/kWh, DEFRA 2026). Report the Scope 3 Cat 3 well-to-tank companion on the same fuel input separately.
Scenario: an office buys 1,200,000 kWh of district heat per year from a UK network with no bespoke disclosure, applying the DEFRA default district-heat factor.
Factor 0.170 and WTT 0.030 are hardcoded illustrative review-date snapshots of the live district_heating.gbr.district and district_heating.gbr.district_wtt rows — the live values render at the top of this page and in the Tier 3 card. No efficiency step: the DEFRA default is already a delivered-heat factor with generation and losses embedded. The 36 t WTT sits in Scope 3 Cat 3, never in the Scope 2 total.
Scenario: a CHP plant burns 10,000 GJ of natural gas (56.1 kg CO₂e/GJ, illustrative) and co-produces 4,000 GJ of electricity and 4,500 GJ of heat. What heat footprint does the buyer report under each allocation method?
Fuel EF (56.1 kg/GJ), reference boiler efficiency (85%) and reference power efficiency (45%) are illustrative engineering inputs, not MasterBrain values. The reporting lesson: the heat buyer’s footprint depends as much on the plant’s disclosed allocation method as on its fuel. Require the method in the supplier disclosure; do not assume the energy method silently.
Edge cases and failure modes
Condensate return
If condensate is returned to the supplier, the feedwater enthalpy the buyer paid for is lower — subtract the return enthalpy from the supply enthalpy so the buyer is charged carbon only for the net heat added, not the full steam enthalpy.
Undisclosed efficiency
When the supplier will not disclose generation efficiency, do not default to 100%. State an assumed band (e.g. 80–90%), report the resulting range, and prefer a Tier 3 delivered-heat factor that sidesteps the efficiency step entirely.
Waste-heat networks
Heat recovered from an industrial process may be near-zero at delivery because its emissions are allocated upstream — but only if that allocation is documented. An undocumented waste-heat claim defaults to the network’s fossil baseline, not to zero.
Mixed steam and hot-water supply
A site drawing both high-pressure steam and low-temperature hot water from one network cannot use one enthalpy for both. Resolve each stream to energy on its own supply conditions before summing — a single blended enthalpy misstates the split.
Audit checklist
- Metering unit and supply conditions recorded. Every steam/hot-water line states its billed unit and the pressure/temperature (or flow/return ΔT) used to resolve it to energy.
- Enthalpy source cited. The steam-table basis (IAPWS-IF97 or supplier data) is named, with feedwater/return enthalpy subtracted where condensate is returned.
- Factor tier justified. The tier applied (supplier-specific, fuel-based, DEFRA default, electric-proxy) is documented, with a note on why a higher tier’s evidence was unavailable where a lower tier is used.
- Generation efficiency stated with sensitivity. The efficiency divisor is disclosed and its uncertainty band reported; no single-point steam figure without a stated efficiency.
- CHP allocation method disclosed. Where heat comes from cogeneration, the allocation method (energy, efficiency, or exergy) is named — not assumed.
- WTT companion reported separately. The Scope 3 Category 3 well-to-tank line for the generation fuel is present and outside the Scope 2 total.
What the calculator handles vs what you decide
The Purchased Steam & Hot Water Calculator automates the calculation chain. This methodology covers the upstream decisions the calculator cannot make for you.
Calculator handles automatically
Metering-unit conversion (mass/volume/energy to a common basis); enthalpy lookup at supplied conditions; efficiency division; tier-appropriate factor application with live DEFRA and EUROSTAT values; dual-method (location + market-based) display; the Scope 3 Cat 3 WTT companion; unit conversion to tCO₂e on the AR6 basis. See also the district heating & cooling calculator and the CHP cogeneration calculator.
You must decide first
The supply conditions (pressure, temperature, ΔT) that set the enthalpy; which factor tier your evidence supports; the generation efficiency where the supplier does not disclose it; the CHP allocation method; whether any market-based heat claim is substantiated; and whether recovered waste heat is documented as allocated upstream.
Resolve any metering unit to CO₂e with the four-tier factor hierarchy, CHP allocation, and the WTT companion built in.
Frequently asked questions
Because a steam bill is usually in mass or volume, not energy, and an emission factor is defined per unit of energy. You first convert the metered mass to useful heat using the steam’s enthalpy at its supply pressure and temperature, then divide by generation efficiency to get fuel input, and only then apply the factor. Electricity skips both steps because the kilowatt-hour is already energy and the grid factor already embeds generation. The one exception is a delivered-heat factor such as DEFRA’s district-heat value — that already includes generation and losses, so you apply it directly to delivered kWh with no efficiency step.
Purchased steam, hot water, and district heating are Scope 2, alongside purchased electricity, per the GHG Protocol Scope 2 Guidance. If instead you burn your own fuel on-site to make the steam, that is Scope 1 combustion, not purchased heat. And the well-to-tank emissions of the generation fuel sit in Scope 3 Category 3 as a companion line, never folded into the Scope 2 total.
The one for your actual supply conditions, read from a published steam table (IAPWS-IF97) at the supplied pressure and temperature. Saturated steam typically falls in the 2,675–2,800 kJ/kg range, but superheated steam at high pressure carries more, and you must subtract the feedwater or returned-condensate enthalpy so you count only the heat that was added. These are engineering constants from a steam table, not emission factors — supply them from your contract or the table, and never estimate from memory. The worked examples on this page use 2,750 kJ/kg purely as an illustration.
Materially. In worked example C, the same cogeneration plant gives a heat footprint of 297 tCO₂e under the energy method and 209 tCO₂e under the efficiency method — the efficiency figure roughly 30% lower, from the allocation rule alone. Because heat carries less thermodynamic work potential than electricity per unit of energy, the efficiency and exergy methods assign heat less carbon than the simple energy split does. The requirement is to disclose which method the supplier used; a heat factor quoted without its allocation basis is unfalsifiable. The CHP allocation methodology covers the full treatment.
District cooling is metered in tons of refrigeration and driven mostly by electricity, so its footprint is the plant’s electrical intensity (kW per ton delivered) multiplied by the grid factor. MasterBrain carries ASHRAE benchmarks — a typical plant around 1 kW/ton, efficient plants near 0.89 kW/ton. Multiply delivered cooling by the plant’s kW/ton, then by the grid factor, to reach CO₂e. It is still Scope 2 purchased energy; the district heating and cooling methodology gives the detail.
Only with the same substantiation a renewable electricity claim requires. A market-based renewable-heat claim needs a substantiated certificate, not a tariff name; an unsubstantiated “green heat” product defaults to the location-based factor. Recovered waste heat can be near-zero at delivery because its emissions are allocated to the primary process — but that allocation must be documented, and double-counting between the primary process and the heat buyer must be excluded. And dual reporting still applies: a market-based figure never replaces the location-based one. The country-level renewable share EUROSTAT publishes (Denmark and Spain high, Albania low) informs the network baseline but is not itself a substantiated per-customer claim.