Heat Pump vs Gas Boiler — Emissions Methodology and Calculation Approach
Every “heat pump vs gas boiler” comparison lives or dies on one hidden decision: whether the two devices are measured on the same basis. Fuel burned in a boiler and electricity drawn by a heat pump are not interchangeable units — one is chemical energy at the meter, the other is delivered heat after a coefficient of performance does the work.
Get the common basis right and the answer is unambiguous; get it wrong and you can make either technology look better than it is.
Compare on delivered heat: gas emits 2.02633 ÷ boiler efficiency; a heat pump emits 0.131 ÷ SPF (DEFRA 2026, kg CO₂e/kWh). On the current UK grid the heat pump wins above an SPF of ~0.65.
The Comparison Problem: Why You Can’t Compare kWh to kWh
A gas boiler and an electric heat pump both answer the same question — how much carbon is released to keep a building warm — but they answer it in different currencies. A boiler consumes natural gas measured in kWh of chemical energy at the meter and converts most of it to useful heat, losing the rest up the flue. A heat pump consumes electricity and moves several times that energy in as heat from the outside air or ground. Placing the two side by side on a raw-kWh basis compares a fuel input against an electricity input and silently ignores the two conversion steps that actually determine emissions.
The correct comparison unit is delivered useful heat — one kWh of warmth actually supplied to the building — with each device’s conversion efficiency in the denominator. Everything else on this page follows from that single normalisation.
Fuel-input energy vs delivered useful heat
Building heat demand is normally quantified as useful heat: the energy that leaves the emitters (radiators, underfloor loops, hot-water cylinder) and enters the space. To supply that heat, a boiler must burn more than one kWh of gas per kWh delivered, because combustion and flue losses are real. A heat pump supplies that same heat while drawing less than one kWh of electricity, because it harvests ambient heat rather than generating it. The ratio between input energy and delivered heat is the efficiency term — boiler seasonal efficiency for the gas branch, seasonal performance factor for the heat-pump branch — and it belongs in every calculation from the first line.
The gross-CV vs net-CV trap
The UK DEFRA natural-gas factor used here is the gross calorific value (higher heating value) one. DEFRA publishes both bases, and its methodology makes gross the default for gas in kWh because UK suppliers bill on gross. Much European and engineering literature quotes boiler efficiencies and heat demand on a net CV basis, where efficiencies above 100% are possible. Mixing a gross-CV emission factor with a net-CV efficiency overstates or understates gas emissions by roughly 10%. Keep both terms on the gross-CV basis when using DEFRA factors, or convert consistently — never blend.
This is the most common silent error in consumer-facing comparisons. Because condensing boilers recover latent heat from water vapour in the exhaust, a modern boiler quoted at “over 90% efficient” on a net-CV basis is closer to the low-90s on a gross-CV basis — and it is the gross-CV number that pairs correctly with the DEFRA gross-CV gas factor. The methodology on this page uses gross CV throughout, consistent with the natural-gas combustion methodology and UK DEFRA conversion factors.
Provider directory
Need help keeping this up all year?
See who does this work. Every listing names the standards it works to, and paid placements are labelled. Including Attain Zero and Small World Consulting.
Browse 6 carbon accounting & inventory providers →Do this work? A listing is US$390 a year. Get listed →
The Common-Basis Emissions Formula
Both technologies reduce to the same shape: emissions per unit of delivered heat equal an emission factor divided by a conversion efficiency. Normalising to delivered heat makes the two branches directly subtractable.
Emissions per kWh of delivered heat
Gas boiler: EFgas ÷ ηboiler
Heat pump: EFgrid ÷ SPF
where EF is the relevant DEFRA 2026 factor (kg CO₂e/kWh, gross CV for gas), η is seasonal boiler efficiency (gross CV, 0–1), and SPF is the seasonal performance factor.
Delivered-heat normalisation
Given an annual useful-heat demand Q (kWh/yr), total annual emissions are:
| Branch | Input energy | Annual emissions (kg CO₂e) |
|---|---|---|
| Gas boiler | Q ÷ ηboiler | (Q ÷ ηboiler) × EFgas |
| Heat pump | Q ÷ SPF | (Q ÷ SPF) × EFgrid |
The gas-boiler branch and the heat-pump branch
The gas branch is a Scope 1 combustion calculation: fuel input times a fuel emission factor, identical in structure to any stationary-combustion inventory line under the GHG Protocol Scope 1 stationary-combustion guidance. The heat-pump branch is a Scope 2 electricity calculation: electricity input times a grid emission factor, following the Scope 2 electricity methodology. That the two branches live in different scopes is not a footnote — it changes how the switch appears in a corporate inventory, addressed in the scope-classification section below.
Live Emission Factors (DEFRA 2026)
Two factors drive the entire comparison for a UK building. Both are DEFRA 2026 values, resolved live from the factor library so the page never drifts from the current dataset.
Both factors are published by DEFRA on an IPCC AR5 GWP-100 basis. This is the DEFRA convention and part of the factors’ definition. Do not convert them to AR6, and do not sum them into a corporate total that otherwise uses AR6 factors — mixing GWP bases within one total is a reconciliation error. See the IPCC AR6 reference and the global warming potential glossary entry for the AR5/AR6 distinction.
The grid factor’s 26% year-on-year fall (from the prior 0.177 to the current 0.131 kg CO₂e/kWh [GreenCalculus grid.gbr.electricity.location_based · DEFRA 2026 'UK electricity'!E25 · v2026.284], reflecting the cleaner 2024 UK generation mix) is not a rounding detail — it materially widens the heat-pump advantage, and it means any comparison must state which dataset vintage it used. The grid electricity emission factors dataset and DEFRA emission factor dataset hold the full series.
For a location-based inventory, use the grid factor above. If the building’s electricity is on a supplier tariff with contractual instruments, a market-based heat-pump figure can differ from the location-based figure. The two methods are dual-reported — see the location-based vs market-based comparison. This page uses location-based as the default comparison basis.
Efficiency Parameters: Boiler Gross Efficiency and Heat-Pump SPF
The emission factors are library values; the efficiencies are assumptions the analyst must state and defend. There is no single “correct” boiler efficiency or heat-pump SPF — both depend on equipment, installation, climate, and operation. The methodology requires that each be declared explicitly and sourced, never buried as a hidden constant.
Boiler seasonal efficiency
Seasonal efficiency captures real-world part-load and cycling behaviour across a heating season, not the peak bench-test figure. On a gross-CV basis, typical UK values fall in these bands:
| Boiler type | Gross-CV seasonal efficiency (typical) | Notes |
|---|---|---|
| New condensing (well-installed, low-temperature system) | ~90–92% | Achieves condensing mode most of the season |
| New condensing (high return temperatures, poor commissioning) | ~85–88% | Rarely reaches full condensing benefit |
| Older non-condensing | ~70–80% | Flue losses dominate; no latent recovery |
Efficiency bands are typical UK field values on a gross-CV basis, consistent with condensing-boiler performance literature and UK product energy-labelling conventions. Treat as declared assumptions, not library factors.
What the calculator uses as its default
The calculator’s default boiler efficiency is not one of the bands above. It is the seasonal efficiency SAP 10.2 (the UK Standard Assessment Procedure, 17-12-2021) states in Table 4b, page 162, and it is read live from the data layer: a regular condensing gas boiler with automatic ignition at 0.84 [GreenCalculus heating_systems.sap_table_4b.gas_regular_condensing.winter · SAP 10], a non-condensing one at 0.74 [GreenCalculus heating_systems.sap_table_4b.gas_regular_non_condensing.winter · SAP 10], a condensing oil boiler at 0.84 [GreenCalculus heating_systems.sap_table_4b.oil_condensing.winter · SAP 10] and a standard oil boiler from 1998 or later at 0.8 [GreenCalculus heating_systems.sap_table_4b.oil_standard_1998_or_later.winter · SAP 10], all as gross-CV fractions for the winter (space-heating) season. SAP heads its gas block “mains gas, LPG and biogas”, so LPG takes the gas figures.
These are SAP’s fallback values: Table 4b’s first footnote says the table “is to be used only for gas and liquid fuel boilers which cannot be located in the database”. A boiler listed in the Product Characteristics Database carries its own, usually higher, figure, which is why the calculator invites you to enter your boiler’s efficiency if you know it. Until 22 September 2026 the default was a round figure GreenCalculus had chosen; it is now labelled as such in the data layer and no longer used here. The worked examples below state their own efficiencies as explicit inputs and are unchanged. SAP 10.2 carries Crown copyright and states no licence; it is served under a provisional position while the publisher is asked.
SPF, SCOP, and what drags them down
The seasonal performance factor (SPF) is the ratio of heat delivered over a full season to electricity consumed, measured in the field. It is the honest analogue of a boiler’s seasonal efficiency and the correct term for an annual emissions calculation. SCOP (seasonal coefficient of performance) is the standardised lab-rated version; a datasheet SCOP typically overstates the SPF a real installation achieves, because it excludes several loss mechanisms.
| Factor | Effect on SPF | Why |
|---|---|---|
| Low outdoor temperature | Reduces | Larger lift between source and flow temperature |
| High flow temperature (retrofit radiators) | Reduces | Each extra degree of flow temperature costs performance |
| Auxiliary/defrost and immersion backup | Reduces | Direct electric resistance runs at effective COP of 1 |
| Oversized emitters / low-temperature design | Increases | Lower flow temperature raises COP |
| Ground source vs air source | Increases | Warmer, more stable source temperature in winter |
Typical UK field SPFs: air-source ~2.5–3.2, ground-source ~3.0–4.0, with well-designed low-temperature systems at the upper end and poor retrofits below. A planning default of ~3.0 for air-source is defensible; state the value used. Values are declared assumptions, not MasterBrain factors.
Prefer a field-measured or metered SPF over a datasheet SCOP whenever available. If only a SCOP is on hand, apply a conservative haircut (a common convention is to assume the achieved SPF sits below the rated SCOP) and disclose the adjustment. An emissions figure built on an unadjusted datasheet SCOP is optimistic by construction.
The Carbon Crossover: Break-Even SPF
The single most useful output of this methodology is the break-even SPF — the heat-pump performance at which its emissions exactly equal the gas boiler’s. Above it, the heat pump wins on carbon; below it, the boiler does.
The break-even inequality
The heat pump emits less per unit of delivered heat when:
EFgrid ÷ SPF < EFgas ÷ ηboiler
Rearranged, the heat pump wins whenever:
SPF > (EFgrid × ηboiler) ÷ EFgas
Substituting the current DEFRA 2026 factors (grid 0.131, gas 0.18231 kg CO₂e/kWh, gross CV) against a new condensing boiler at 90% gross efficiency:
A heat pump beats a modern gas boiler on carbon whenever its SPF exceeds roughly 0.65. Since a broken heat pump running as a pure electric resistance heater still achieves an SPF near 1.0, and any functioning unit delivers 2.5 or more, the conclusion on the current grid is unambiguous: a working heat pump always wins on operational carbon against gas in the UK. The margin is wide, not marginal.
Per kWh of delivered heat, DEFRA 2026 factors, gross CV. Bar widths scaled to the old-boiler value.
Sensitivity to the grid factor
The result is not a fixed law of physics — it moves with the grid. The break-even SPF is proportional to the grid factor, so as the UK grid decarbonises, the bar a heat pump must clear falls. Historically this mattered: on a mid-2010s UK grid the break-even SPF against a 90% boiler was above 2.0, meaning a poorly performing air-source unit could genuinely lose on carbon. Today it is 0.65.
| Grid factor (kg CO₂e/kWh) | Era | Break-even SPF vs 90% boiler | Implication |
|---|---|---|---|
| ~0.46 | Mid-2010s UK grid | ~2.3 | A weak heat pump could lose |
| ~0.23 | ~2020 UK grid | ~1.1 | Almost any heat pump wins |
| 0.177 | DEFRA 2025 dataset | ~0.87 | Heat pump wins comfortably |
| 0.131 | DEFRA 2026 dataset | ~0.65 | Heat pump wins decisively |
Historical grid factors are indicative era values for the sensitivity illustration; the 0.177 and 0.131 figures are the DEFRA 2025 and 2026 published UK location-based factors. Gas factor and boiler efficiency held constant.
Two consequences follow. First, a comparison is only as current as its grid factor — cite the dataset vintage. Second, the direction is one-way: on a decarbonising grid the heat pump’s advantage grows every year, while the boiler’s emissions are fixed by the carbon content of gas. This asymmetry is the core reason the operational-carbon case for electrification strengthens over an asset’s lifetime, and it links directly to the renewable carbon-payback methodology for capital-embodied trade-offs.
Scope Classification and the Reporting Asymmetry
Switching from a boiler to a heat pump does more than change a number — it moves the emission from one scope to another, which reshapes how the change appears in a corporate inventory even before any net reduction is counted.
| Attribute | Gas boiler | Heat pump |
|---|---|---|
| GHG Protocol scope | Scope 1 (direct combustion) | Scope 2 (purchased electricity) |
| Emission factor source | DEFRA fuel factor | Grid electricity factor |
| Location vs market-based | Single factor | Dual-reported (location + market) |
| Upstream fuel/energy | Well-to-tank (Scope 3 Cat 3) | T&D losses + WTT (Scope 3 Cat 3) |
| Direct on-site refrigerant risk | None | F-gas leakage (Scope 1) |
A heat pump reduces total emissions but raises Scope 2 while eliminating a Scope 1 line. An organisation with a Scope-1-only target, or one that reports scopes separately without a combined view, can misread the switch. Always compare the combined Scope 1 + Scope 2 total, not a single scope in isolation. A heat pump also introduces a small new Scope 1 line — refrigerant leakage — that a boiler never had.
Three second-order reporting effects deserve attention. First, the heat-pump electricity carries well-to-tank and transmission-and-distribution losses in Scope 3 Category 3, just as gas carries its own well-to-tank upstream factor; a like-for-like comparison should either include both upstream tails or exclude both. Second, under market-based accounting a heat pump on a genuine renewable tariff can report a lower Scope 2 figure than the location-based number, but the location-based figure remains the honest physical comparison and is what this methodology defaults to. Third, the refrigerant charge in a heat pump is a direct emission risk governed by the EU F-gas Regulation — negligible when the system is sealed and correctly decommissioned, but real if it leaks.
Worked Examples — Audit Records
Three snapshots at DEFRA 2026 factors. All inputs, formulas, and results reconcile and are fixed as audit records: they must reproduce a year from now regardless of future dataset changes. Each is replicable in the Heat Pump vs Gas Boiler Calculator.
A — Typical home, like-for-like replacement
12,000 kWh useful heat, new condensing boiler at 90% vs air-source heat pump at SPF 3.0. The mainstream retrofit case.
B — Cold-climate stress test
40,000 kWh, an old 80% boiler vs a heat pump managing only SPF 2.2. Even a poor heat pump against a dirty boiler.
C — Small dwelling, modest SPF
12,000 kWh, an efficient new 92% boiler vs a heat pump at SPF 2.4. The efficient-boiler benchmark.
Example A — 12,000 kWh, new boiler (90%) vs heat pump (SPF 3.0)
| Step | Gas boiler | Heat pump |
|---|---|---|
| Useful heat demand Q | 12,000 kWh | 12,000 kWh |
| Efficiency term | η = 0.90 | SPF = 3.0 |
| Input energy (Q ÷ eff) | 13,333.3 kWh gas | 4,000.0 kWh elec |
| Emission factor (DEFRA 2026) | × 0.18231 | × 0.131 |
| Annual emissions | 2,430.80 kg = 2.4308 tCO₂e | 524.00 kg = 0.5240 tCO₂e |
Delta: −1,906.80 kg (−78.4%) — the heat pump cuts operational heating emissions by roughly four-fifths.
Example B — 40,000 kWh, old boiler (80%) vs heat pump (SPF 2.2)
| Step | Old gas boiler | Heat pump (poor SPF) |
|---|---|---|
| Useful heat demand Q | 40,000 kWh | 40,000 kWh |
| Efficiency term | η = 0.80 | SPF = 2.2 |
| Input energy (Q ÷ eff) | 50,000.0 kWh gas | 18,181.8 kWh elec |
| Emission factor (DEFRA 2026) | × 0.18231 | × 0.131 |
| Annual emissions | 9,115.50 kg = 9.1155 tCO₂e | 2,381.82 kg = 2.3818 tCO₂e |
Delta: −6,733.68 kg (−73.9%) — even a heat pump performing badly (SPF 2.2) against an inefficient boiler wins decisively, because both efficiency and grid cleanliness favour the heat pump.
Example C — 12,000 kWh, efficient new boiler (92%) vs heat pump (SPF 2.4)
| Step | Efficient new boiler | Heat pump (modest SPF) |
|---|---|---|
| Useful heat demand Q | 12,000 kWh | 12,000 kWh |
| Efficiency term | η = 0.92 | SPF = 2.4 |
| Input energy (Q ÷ eff) | 13,043.5 kWh gas | 5,000.0 kWh elec |
| Emission factor (DEFRA 2026) | × 0.18231 | × 0.131 |
| Annual emissions | 2,377.96 kg = 2.3780 tCO₂e | 655.00 kg = 0.6550 tCO₂e |
Delta: −1,722.96 kg (−72.5%). The break-even SPF against this 92% boiler is 0.66; at SPF 2.4 the heat pump clears it nearly four times over.
Edge Cases and Boundary Conditions
The base methodology assumes a single heat source meeting a single annual demand. Real installations introduce boundary conditions that change the calculation or its interpretation.
- Hybrid systems. A boiler-plus-heat-pump hybrid splits the annual heat demand between the two branches by a heat-share fraction. Emissions are the weighted sum: apply the heat-pump branch to its share and the gas branch to the remainder. The blended figure sits between the two pure cases.
- Cold-snap COP collapse. A single seasonal SPF hides the fact that COP falls at low outdoor temperatures, sometimes dipping toward electric-resistance backup. A seasonal SPF already averages this in; do not double-count by also modelling a cold-snap penalty on top of a field-measured SPF. Only model it separately if starting from a rated SCOP.
- Domestic hot water. DHW often needs a higher flow/store temperature than space heating, which lowers the effective COP for that fraction of demand. Where DHW is a large share, a blended SPF or a separate DHW SPF is more accurate than a single space-heating SPF.
- Refrigerant leakage. A heat-pump refrigerant charge is a Scope 1 emission risk under the F-gas Regulation. For a sealed, correctly maintained system the annual leakage is small relative to the operational saving, but it is non-zero and must be disclosed for a complete lifecycle comparison. It is quantified by charge mass × GWP × annual leak rate.
- Green gas and green electricity. A biomethane or green-gas tariff changes the gas branch’s effective factor; a renewable electricity tariff changes the heat-pump branch under market-based accounting. Both are contractual instruments, not physical changes — keep the location-based physical comparison as the primary basis and treat tariff-based figures as a supplementary market-based view.
- Embodied carbon and payback. Operational emissions are only part of the lifecycle. Manufacturing and installation carbon for a heat pump is repaid over time by lower operational emissions; the carbon-payback methodology handles that capital-vs-operational trade-off.
Governance, Data Quality, and Assumption Disclosure
Because the two efficiency terms are analyst assumptions rather than library factors, the credibility of any heat-pump-vs-boiler figure rests on how those assumptions are governed and disclosed. A defensible comparison states each of the following on the face of the result.
| Parameter | Data-quality question | Best practice |
|---|---|---|
| Grid factor | Which dataset vintage and which method? | Name the DEFRA release year and location- vs market-based |
| Gas factor | Gross or net CV? | Use gross CV to match DEFRA; state it |
| Boiler efficiency | Seasonal or peak; gross or net? | Seasonal, gross-CV, sourced to product/label data |
| Heat-pump SPF | Field-measured or datasheet SCOP? | Prefer metered SPF; haircut SCOP and disclose |
| Heat demand Q | Metered or modelled? | Metered useful heat where available |
| GWP basis | AR5 or AR6; consistent? | Keep DEFRA AR5 factors together; don’t blend bases |
The methodology is only as trustworthy as its two efficiency assumptions. Two analysts using the same DEFRA factors but different SPF and boiler-efficiency values will reach different deltas. Publishing the assumptions alongside the result is what makes the comparison auditable rather than persuasive.
Audit Checklist
- Both branches normalised to the same delivered-heat basis (Q identical on both sides).
- Gas factor and boiler efficiency both on gross CV — no net/gross blend.
- Grid factor dataset vintage stated (DEFRA year) and method (location- or market-based) declared.
- SPF sourced — field SPF preferred, datasheet SCOP haircut and disclosed if used.
- Comparison made on combined Scope 1 + Scope 2, not a single scope in isolation.
- GWP basis consistent — DEFRA AR5 factors not mixed into an AR6 total.
- Upstream tails (well-to-tank, T&D) either included on both sides or excluded on both.
- Refrigerant-leakage line acknowledged for the heat-pump branch where a full lifecycle view is claimed.
- Every efficiency and demand assumption disclosed on the face of the result.
Standards Alignment
| Standard | Role in this methodology |
|---|---|
| UK DEFRA emission factors | Source of the natural-gas (gross CV) and UK grid factors. AR5 GWP-100 basis. |
| GHG Protocol Scope 1 stationary combustion | Governs the gas-boiler branch as a direct combustion emission. |
| GHG Protocol Scope 2 Guidance | Governs the heat-pump branch; dual location/market-based reporting. |
| IPCC AR6 | GWP reference; the AR5/AR6 basis distinction for DEFRA factors. |
| EU F-gas Regulation 2024 | Governs the heat-pump refrigerant-leakage edge case. |
What the Calculator Handles vs What You Decide
| The calculator handles | You must decide |
|---|---|
| Delivered-heat normalisation across both branches | The useful-heat demand Q (metered or modelled) |
| Applying live DEFRA gas and grid factors | Boiler seasonal efficiency (gross CV) |
| Break-even SPF against your boiler efficiency | Heat-pump SPF (field vs haircut SCOP) |
| Scope 1 vs Scope 2 split and combined total | Location- vs market-based electricity basis |
| Delta and percentage reduction | Whether to include upstream/refrigerant tails |
Run your own figures with live DEFRA 2026 factors and your building’s heat demand.
Dataset — the full factor set referenced on this page, versioned with source provenance and downloadable as CSV with a citable Zenodo DOI, is published as the UK domestic energy benchmarks (NEED) dataset.
Frequently Asked Questions
Because they measure different things. Gas kWh is chemical energy at the meter before combustion losses; electricity kWh is drawn energy before the heat pump multiplies it by its performance factor. The only fair basis is delivered useful heat, with boiler efficiency and heat-pump SPF in the denominator. Comparing raw input kWh ignores both conversion steps and produces a meaningless result.
On the current UK grid (DEFRA 2026, 0.131 kg CO₂e/kWh) against a 90%-efficient condensing boiler, the break-even SPF is about 0.65. Any functioning heat pump delivers an SPF of 2.5 or more, so in practice a working heat pump always wins. The break-even bar was much higher a decade ago (above 2.0 on the dirtier mid-2010s grid), which is why older comparisons sometimes found gas competitive — those conclusions no longer hold on today’s grid.
DEFRA publishes its fuel factors on both gross and net bases, and its methodology makes gross the default for natural gas in kWh, because UK suppliers calculate kWh on a gross basis. This matters because boiler efficiencies are often quoted on a net-CV basis in engineering literature, where condensing boilers can appear “over 100% efficient.” Mixing a gross-CV factor with a net-CV efficiency skews the gas result by roughly 10%. Keep both terms gross-CV when using DEFRA factors.
Yes — and that is expected. A heat pump eliminates a Scope 1 gas-combustion line and creates a Scope 2 electricity line. The total falls sharply, but the emission moves between scopes. Always compare the combined Scope 1 + Scope 2 figure; judging the switch on Scope 2 alone will make it look worse than it is, and judging on Scope 1 alone will overstate the benefit.
Use a field-measured SPF whenever you have one — it captures defrost cycles, backup heater use, distribution losses, and real flow temperatures that a lab SCOP excludes. A datasheet SCOP typically overstates achieved performance, so if it is all you have, apply a conservative reduction and disclose it. An emissions figure built on an unadjusted SCOP is optimistic by construction.
For a sealed, correctly installed and maintained system, no — annual refrigerant leakage is small relative to the operational saving of several tonnes of CO₂e a year. But it is a genuine Scope 1 emission risk governed by the F-gas Regulation, and a complete lifecycle comparison should quantify and disclose it (charge mass × refrigerant GWP × annual leak rate). It becomes material only under poor maintenance or improper end-of-life decommissioning.
The heat-pump branch’s emissions are proportional to the grid factor, which falls as the grid cleans up — the UK factor dropped 26% between the 2025 and 2026 DEFRA datasets alone. The gas branch is fixed by the carbon content of natural gas. So the heat pump’s advantage widens every year, while the boiler’s does not. This one-way asymmetry is central to the long-run electrification case.
Under market-based accounting, yes: a biomethane tariff lowers the gas branch’s effective factor, and a renewable electricity tariff lowers the heat-pump branch. Both are contractual instruments rather than physical changes to what flows through the meter, so the location-based physical comparison stays the primary basis. Treat tariff-based figures as a supplementary market-based view, and see the location-based vs market-based comparison for the dual-reporting rules.