# Press release — The reckoning on the electricity conversion factor (OneDpe)

---

**PRESS RELEASE — Paris, 25 August 2026**

*Quotations from French orders, consultation documents and official statistical publications below
are OneDpe's translations; the French originals of the load-bearing ones are given alongside, and
all are identified in the method note.*

## Nearly 400,000 EPC-assessed homes have left the energy-sieve bracket with no works done — and 600,000 that nothing will take out of it

**On 1 January 2026 the primary-energy conversion factor for electricity used by France's energy performance certificate — the A-to-G rating required whenever a home is sold or let — moved from 2.3 to 1.9 (order of 13 August 2025). No reckoning on that reform has been published since. We have done it, and we did not model it: 392,948 EPC-assessed homes left the energy-sieve bracket, the F and G classes, through that change alone — without a kilowatt-hour saved and without a euro of works. ADEME, the French environment agency, kept both ratings, the old one and the new one; we counted them, one by one, across 14,938,757 certificates. That is more than one EPC-assessed energy sieve in five: 22.86%. A draft order would move the same factor to 1.7 on 1 January 2027; if it were published as it stands, a further 161,479 EPC-assessed homes would leave the bracket, on a target perimeter of 9,625,591. And 604,669 that **no factor will ever take out of it**: it is not their consumption that rates them, it is their boiler.**

### The count nobody had done

The primary-energy conversion factor for electricity is the number by which the kilowatt-hours of electricity a home consumes are multiplied to obtain the primary energy that decides its rating. On 1 January 2026 it moved from 2.3 to 1.9. Not a wall was insulated, not a boiler replaced: the stated consumption of an electrically heated home fell overnight, in the exact ratio of 2.3 to 1.9. We measure that ratio in the data: it sits at **1.211 in mass** — exactly 2.3 / 1.9, the arithmetical signature of the reform.

We did not estimate what this produced: we counted it. For **14,938,757 certificates** — 82% of the 18,203,808 rows in the dataset, and **99.50%** of live certificates issued between July 2021 and December 2025 — ADEME kept the old rating and the new one side by side. We counted the EPC-assessed homes rated F or G under 2.3 that are no longer rated F or G under 1.9: there are **392,948**. Within that population, the stock of energy sieves falls from **1,718,766 to 1,325,818**.

The rate depends on no perimeter, no deduplication and no reconstruction: 392,948 out of 1,718,766, that is **22.86%**. More than one EPC-assessed energy sieve in five. Restricted to certificates issued from 1 July 2024 onwards — born under the small-floor-area scale, and therefore uncontaminable by construction — it rises to **24.12%** (135,325 exits out of 561,043 energy sieves): the cleanest population is the most responsive, not the least.

Two reforms followed one another and conflating them is the commonest mistake on this subject. The **small-floor-area** reform, in July 2024, took at least **69,049** EPC-assessed homes out of the bracket before the conversion factor came into play — a floor, measured on the 1,506,571 certificates covered by both sets of columns. The **conversion-factor** reform, on 1 January 2026, took 392,948 out — counted across the 14,938,757 certificates for which ADEME kept both ratings. The two counts do not rest on the same base. We know how to separate them, and we do separate them.

### The check that could have failed

No downgrade is observed: no home enters the energy-sieve bracket because of the factor. That is what arithmetic requires — primary energy can only fall when the factor falls, and emissions do not move — and what the data confirm without exception. So it is not a check, and we do not present it as one.

The check that could have failed is elsewhere. A home is an energy sieve through its energy, through its emissions, or through both: the rating retains the worse of the two classes. A home already rated F or G on the carbon axis therefore cannot leave the bracket through a factor that touches only energy. Of the 1,718,766 energy sieves under the 2.3 factor, **714,457 are in that position** — 523,010 in carbon class F, 191,447 in G. We count **zero exits** among them. Exactly zero. Had our measurement chain been wrong, that number would have been positive: it is a figure that denounces itself.

Two further checks hold the headline figure. It is identical to the unit whether it is measured against the published rating or against the arrival column ADEME kept — **0 divergence across 14,938,757 rows**. And the gap between the two published consumptions bears on electric kilowatt-hours only: the extra primary energy is exactly **0.9000** times the electricity consumed — that is, 1.9 minus 1 — at the 5th, 50th and 95th percentiles, across 603,948 homes.

### What no factor will do: 604,669 EPC-assessed homes

**A conversion factor does not decarbonise a boiler.**

For **604,669 EPC-assessed homes**, the conversion-factor reform will change nothing: it is not their consumption that rates them energy sieves, it is their boiler. **95.46%** heat with gas or oil (577,205 homes: 308,528 on gas, 268,677 on oil), close to three in four exceed their carbon threshold by one class only, and for one in four carbon is the **sole** cause of the rating.

These 604,669 homes are the ones that would remain energy sieves even under the 1.7 factor, held there by their emissions alone: **58.52%** of the 1,033,290 energy sieves that would then survive on our target perimeter of 9,625,591 EPC-assessed homes. **95.44%** were built before 1975, 48.95% before 1948, 582 after 2000. **59.34%** are houses, against 41.9% of the perimeter. Their median emissions are 87.9 kg CO₂/m²/year and their median distance to the threshold that applies to them is 17.0 kg. **432,306** of them — 71.89% of the 601,313 for which that distance is measurable — exceed their threshold by one class only. And **156,067**, one in four (25.81%), are energy sieves on carbon alone: on the energy axis they are already E or better, **against the statutory threshold that applies to them**. Moving to 1.7 would take **1.54 kWh/m²/year** of primary energy off them at the median — measured on the **578,969** homes of the cohort for which both consumptions are available. That is the third and last denominator in this section, after the 604,669 of the cohort and the 601,313 whose distance to threshold is measurable; we label them rather than let them blur together.

It will be objected that electrifying is enough. At constant final energy, replacing heating alone with an electric vector would bring **589,593** of these homes below their carbon threshold — 98.05% of the 601,313 — and their median emissions from 87.9 to 32.9 kg. But the same substitution, read on the energy axis, would take the median primary-energy consumption of those 601,497 homes from **374.2 to 656.3 kWh/m²/year**: **99.873%** would remain blocked by energy (600,736 out of 601,497, excluding the 3,172 already electrically heated). Changing vector without touching efficiency lifts **both** locks for **750 homes out of 600,924**. This finding argues for no technology: it says that at constant efficiency you move the lock, you do not open it.

### The mechanism, measured inside one and the same pool

Who leaves? Those who consume electricity, and only they. Measured on the counted subset alone — the 8,871,468 certificates of the target perimeter for which ADEME kept both states — the exit rate of the first reform rises monotonically with the share of electricity in final consumption: **2.31%** below 25% electricity, 25.70% from 25 to 50%, 32.57% from 50 to 75%, 36.16% from 75 to 99%, **45.04%** at all-electric. Nearly one all-electric energy sieve in two leaves the bracket; one low-electricity energy sieve in forty-three. A ratio of 19.5.

And among those low-electricity homes, **85.6%** of the energy sieves that remain are held there by their carbon. The mechanism and the carbon ceiling are not two results: they are the same fact, seen twice.

One qualification that cuts against us, and that has to be given: the axis of this table is built on a quantity that **overstates** the electricity actually carrying the factor. The bias pushes low-electricity homes up the bands and dilutes the top band. The 45.04% is therefore a **floor**, and the real gradient is steeper than the one we publish. These rates, on the other hand, do not recompose into a national rate: they are measured on the counted subset alone.

### The news: the draft 1.7 order

A draft order (NOR VLOL2617030A), put out to consultation from 9 July to 6 August 2026 and examined by the Conseil supérieur de l'énergie, France's higher council for energy, on 23 July 2026 — item 18 on the agenda; only the examination is established, no vote is public — would move the factor from 1.9 to 1.7 on 1 January 2027. As at 25 August 2026, **it has not been published in the *Journal officiel***, France's official gazette: the version of the order of 31 March 2021 that Légifrance serves as at 1 January 2027 still carries "+ 1.9 for electricity".

If the text were published as it stands, a further **161,479** EPC-assessed homes would leave the bracket, that is **13.52%** of the 1,194,769 energy sieves that survive under the 1.9 factor on our target perimeter of 9,625,591 EPC-assessed homes. **147,117** of those exits are **counted** on the two consumptions published by ADEME; 14,362 are **simulated**, on the 7.8% of certificates for which ADEME never produced a state at 2.3.

That 13.52% does not read against the 22.86% of the first reform: its denominator is a stock already skimmed by it. On the same target perimeter, the first reform takes 24.78% of the energy sieves out of the 2.3 factor; the second takes 13.52% of those that remain.

**The official figure, and why we do not compare ours to it.** The post-consultation summary, put online on 13 August 2026, states that the change will lead "to a certain number of homes leaving energy-sieve status (estimated at 300,000 homes), with no real physical modification of the buildings and no immediate reduction in their consumption" (*« à la sortie d'un certain nombre de logements du statut de passoires énergétiques (estimé à 300 000 logements), sans modification physique réelle des bâtiments ni réduction immédiate de leurs consommations »*). Neither the draft order nor its explanatory memorandum carries a number of homes, a perimeter or a method — we checked that on the documents themselves, not on the page hosting them. We do not compare our count to that figure: theirs is an **ex ante estimate on a modelled stock of principal residences**, ours a **count of certificates**, after the fact, overseas included. The two are not comparable in headcount, and we will not compare them.

What does compare is **rates**. The official simulation of the first reform takes energy sieves among principal residences from 3.9 to 3.2 million, that is **17.7%** of exits; we count **22.86%**. That gap is not a difference in responsiveness, and we have corrected our own reading of it: **at equal main heating energy our exit rates are slightly lower than the official ones** — 43.42% against 45.40% on electricity, and the gap has the same sign on all four energies. It comes **entirely from composition**: our energy sieves are more electric than those of the national stock. It would therefore be wrong to present our figure as "smaller, therefore more cautious" — and just as wrong to scale it to the national stock by a rule of three, which would measure nothing but our selection bias.

One precedent, finally, on how solid ex ante estimates are. The estimate for the 1.9 reform was 850,000 homes; the current estimate is around 700,000. That is not our reading, it is that of the ministry's statistical service, whose footnote 1 states that "the ex ante evaluation of the reform had been estimated on the basis of the stock at 1 January 2023, which gave around 850,000 fewer energy sieves". On that same 1.9 reform, the Conseil supérieur de la construction et de l'efficacité énergétique, France's higher council for construction and energy efficiency, had returned an **unfavourable** opinion on 24 July 2025 — 15 against, 4 in favour, 10 abstentions — noting "the absence of a socio-economic impact assessment".

### What our rates give on the national stock, once tested

Our counts bear on certificates, never on the French housing stock. There is nevertheless a way of making our rates speak at the scale of that stock without grossing up our own sample: **borrow the official denominator and lend nothing but our rates**, one main heating energy at a time. That transposition can be tested — and we tested it before applying it, on the 2.3 → 1.9 step, **already in force and whose result the administration has published**. Applied to the energy-sieve headcounts published by the ministry's statistical service, our four exit rates (electricity **43.42%**, wood and other 7.37%, gas 4.12%, oil 1.13% — all **below** the official ones, and that is the check that matters: applied to **their** heating-energy structure, our rates recompose to **16.60%** against their 17.73%. At constant structure our apparent edge in responsiveness **changes sign**: it was composition, not measurement) give **◇ 650,000 mainland principal residences** against the official **694,000**, that is **−6.3%** — where a plain rule of three on our overall rate misses the same target by **+29%**, and a stratification by dwelling type by **+26%**. The criterion authorising this publication was written **before** the measurement and was not reopened. Applied to the step the draft 1.7 order would produce if it were published as it stands, the same transposition gives **◇ 275,000** principal residences, within a range of **248,000 to 294,000**; the public-consultation summary writes that the measure would lead to « **la sortie d'un certain nombre de logements du statut de passoires énergétiques (estimé à 300 000 logements)** » — a number of homes leaving the energy-sieve bracket, put at 300,000 — **with no published method**. ⚠️ That 300,000 sits **above our upper bound**, by 2%: we do not reproduce it exactly, we reproduce its **order of magnitude**. That is the nuance that matters, and we state it rather than let it be used against us. **We corroborate the official order of magnitude — and we publish how we get there; nobody has published how they get theirs.**

### The map, and the segment

Across the full span 2.3 → 1.7 — the reform in force, plus the draft order if it were published as it stands — **555,068** EPC-assessed homes leave the bracket on the target perimeter of 9,625,591, that is **34.95%** of the 1,588,358 energy sieves that perimeter held under the 2.3 factor. The rate varies by a factor of **4.3** between *départements*: in the **Hérault**, two EPC-assessed energy sieves in three leave the bracket (**66.61%**); in the **Haute-Saône**, one in six and a half (**15.40%**). Behind them, the Aude (62.09%), the Bouches-du-Rhône (59.94%) and the Var (59.24%) on one side; the Creuse (18.19%), the Moselle (18.37%) and the Ardennes (19.22%) on the other.

**A word on populations, because the addition does not come out even.** The 392,948 of the assessment and the 555,068 of the full span do not rest on the same population: the first is counted across the 14,938,757 certificates for which ADEME kept both labels, the second across the target perimeter of 9,625,591, which is narrower — it sets aside superseded certificates and those generated from a whole-building EPC. On that target perimeter, the same first step yields **393,589** exits, or **24.78%**, and the addition then comes out to the unit: 393,589 + 161,479 = 555,068. The direction of the gap is worth noting: applying the stricter perimeter to the headline figure would **raise** it — by 641 EPC-assessed homes and by close to two points of rate. We publish the lower of the two.

**We do not explain this map by climate, and we publish no correlation.** We give the second series beside the first, and the reader makes of it what they will: the share of all-electric homes among energy sieves under the 2.3 factor is **73.67%** in the Hérault and **14.53%** in the Haute-Saône. Two series set side by side, not linked — the mechanism is demonstrated inside one and the same pool, never between territories.

The split by segment says the same thing the other way round. On our population of certificates, 65.1% of the exits across the span are flats and 34.8% houses — a split the ministry's statistical service does not find, because it works on a modelled stock of principal residences, and that has to be said. But the cohort held by carbon flips: **59.34% houses**. The factor takes flats out; what stays blocked is a house.

The *département* map is **mainland** and carries **96 *départements***: we publish no rate below 1,000 energy sieves under the 2.3 factor, which sets aside the eight overseas codes — 138 exits in all. To that is added a mass that is not a territory and that we keep out of every ranking: 156,262 certificates with no geocoding, carrying 8,023 exits. The 96 *départements* cover **98.53%** of the span, not all of it.

### What it is worth — on houses, and on houses only

A rating has a price. Across **825,719 house sales** matched to their certificate (BDNB data published by CSTB, France's building science centre, 2025-07-a vintage, already published on our EPC discount tool), the median price per square metre of an E-rated house exceeds that of an F-rated house by **€190.00**. The **114,738 houses** that leave the bracket at the first step — counted subset, house segment — therefore carry a **price gap associated with the rating** of **€1.95 billion**, for an average floor area of 89.4 m². That is a price difference **observed** between two classes, applied to homes of which nothing has changed — neither new wealth, nor a forecast of transactions. Our gradient is not a controlled regression: it says at what median price per square metre the houses in each class have sold, it does not say what a change of rating would produce.

**We publish no amount on flats, nor on all segments together.** Our own source forbids it: outside Paris and the inner suburbs, a G-rated flat sells per square metre **6.2% above** a D-rated one — the gradient flips. Applying a house gradient to a flat would not be an approximation, it would be a sign error. Three amounts, calculated and correct, were withdrawn from this file on that ground alone.

The shortest argument holds in two numbers, both on houses. The price gap between a G-rated house and a D-rated house is **€470/m²** (€2,119 against €1,649). Taking that same house from G to D through works costs of the order of **€600/m²** before grants. **The discount does not cover the works — and the factor takes the rating away for nothing at all.**

### What changes in law, and what does not

A home that leaves the energy-sieve bracket through the factor **becomes lettable again when it is next let**. The regime remains that of the Climate and Resilience Act (*loi Climat et Résilience*), which phases out the letting of the worst-rated homes: the letting ban and the rent freeze (article 17-1 of the Law of 6 July 1989) attach to tenancy agreements **entered into, renewed or tacitly extended** after each class's milestone — G on 1 January 2025, F on 1 January 2028, E on 1 January 2034. A reclassification — through the attestation that the draft order reserves for certificates in force — lifts the freeze **for the future only**: it opens **no catch-up** on past rents, and the draft is silent on running tenancies.

On validity, the question will not bite: **no** certificate in our base expires before 1 January 2027, the earliest falling on 29 June 2031, across 14,379,621 certificates tested.

### Key figures

- ⬛ **392,948** EPC-assessed homes left the energy-sieve bracket through the move from 2.3 to 1.9, across the **14,938,757** certificates for which ADEME kept both states — that is **22.86%** of the 1,718,766 energy sieves under the 2.3 factor, which fall to 1,325,818
- ⬛ **0 exits out of 714,457** energy sieves whose published carbon class is F or G: the check that could have failed
- ⬛ **24.12%** exits among certificates issued since 1 July 2024 alone (135,325 out of 561,043) — the uncontaminable population is the most responsive
- ⬛◨ **604,669** EPC-assessed homes that their emissions alone keep in the bracket whatever the factor: **58.52%** of the 1,033,290 energy sieves that would survive under the 1.7 factor on the target perimeter of 9,625,591 — of which **95.46%** on gas or oil and **95.44%** built before 1975
- ◨ **156,067** of them (**25.81%**) are energy sieves on carbon **alone**, against the statutory threshold that applies to them; **432,306** (71.89% of the 601,313 measurable) exceed that threshold by one class only
- ⬛◨ the mechanism, on the counted subset (8,871,468 certificates): **45.04%** exits at all-electric against **2.31%** below 25% electricity — and **85.6%** of the low-electricity energy sieves that remain are blocked by carbon
- ◨ + ⬜ **161,479** further exits if the draft 1.7 order were published as it stands — **13.52%** of the energy sieves remaining under the 1.9 factor, on the target perimeter of 9,625,591; **147,117 counted**, 14,362 simulated
- ⬛◨⬜ **555,068** across the full span 2.3 → 1.7, target perimeter of 9,625,591 — combined rate **34.95%**
- ⬛◨⬜ geography, 96 mainland *départements* retained: **Hérault 66.61%** against **Haute-Saône 15.40%**; ⬛ share of all-electric among their energy sieves, **73.67%** against **14.53%** — set side by side, never correlated
- *(no marker — a valuation is not a count)* **€1.95 billion** of price gap associated with the rating on the **114,738 houses** that left at the first step, at **€190.00/m²**; **€470/m²** of G↔D gap on a house against about **€600/m²** of works before grants
- ⬛ **0** certificates in the base expiring before 1 January 2027, out of 14,379,621 tested

### Quotation

> "A reform took the energy-sieve rating away from nearly 400,000 EPC-assessed homes without a single wall being insulated, and nobody had done the count. We did not model it: ADEME kept both ratings, the old one and the new one, and we counted them one by one. The check that could have proved us wrong is the one on the 714,457 energy sieves that their emissions were blocking: not one left, and had our measurement been wrong, that number would not have been zero. But the most important result is not that one, it is the one that stays: 604,669 EPC-assessed homes that this factor will never release, because it is not their consumption that rates them, it is their boiler. A conversion factor does not decarbonise a boiler. We are not saying whether the reform is good; we are saying what it has done, and what it will not do."
> — Steven Annonziata, co-founder of OneDpe (RIANN SAS)

### Methodology, in brief

Every figure is established on the public database of energy performance certificates published by ADEME, **extract frozen at 2 June 2026**. These are counts of **certificates**, never a grossing-up of our sample onto the French housing stock: we write "EPC-assessed homes" throughout, and our figures are **floors**. The only comparison with the national stock that we publish is a **transposition** — the official denominator borrowed, our rates lent one energy at a time — given in second place and back-tested first on a step whose official result is already known.

Four markers, carried on every figure — the first three qualify **counts**, the fourth does not:

- ⬛ **observed** — a count of two states **kept by ADEME itself**. No reconstruction, no threshold table, no assumption. This covers the 392,948, the 604,669 and the zero out of 714,457.
- ◨ **reclassified** — a count on quantities **published** by ADEME, reclassified against the **statutory threshold table**. One reconstruction only: saying which side of the threshold the value falls. That table reproduces the published carbon class to **99.97%** and the published F/G energy class to **99.98%**.
- ⬜ **simulated** — reserved for the **7.8%** of certificates for which ADEME never produced a state at 2.3. On the remaining 92.2%, the 1.9 → 1.7 step is **counted**, not simulated; where both routes coexist, the simulation over-counts by 0.56%.

- ◇ **transposed** — **this is not a count.** Our exit rates, measured by heating energy, applied to the energy-sieve headcounts **published by the ministry's statistical service**. The denominator is theirs; only the rates are ours. This covers the 650,000 and the 275,000. ⚠️ A ◇ is never worth more than the marker of the rate feeding it: the 1.9 → 1.7 step being ◨, its transposition is too.

None of these four markers applies to an amount in euros.

**Full method note** — opposable perimeter, the seven populations, the transposition to the national stock and its back-test, the checks that could have failed, the limits we declare ourselves, primary sources and consultation dates, and the frozen English lexicon: `2026-08-25_method-note-electricity-primary-energy-factor.md`, attached to this release.

### About OneDpe

OneDpe (published by RIANN SAS, a French company incorporated on 23 February 2026) is an EPC analysis platform: consistency checking, energy-related price discount, summer comfort. **That activity is commercial, and a study showing that a rating can change with no works done serves it**: which is why we publish the perimeter, the checks and the limits, and why we call for a check by a disinterested third party. The data are public; the calculation can be redone without us.

**EPC verification tool**: onedpe.fr/en/verification-dpe
**Energy-related price discount**: onedpe.fr/en/decote-dpe

**Press contact**: Steven Annonziata — presse@onedpe.fr
**On request**: the method note, the *département* table (96 *départements*, no rate served below the publication threshold), the rate definitions and the perimeter variants.