2026 R290 vs R32 across Europe’s top heat-pump brands: price, noise and efficiency

Which brands are actually moving to R290 in 2026?

The short answer is that this corpus does not contain brand-level refrigerant splits for the leading manufacturers, so it cannot rank top brands by share of R290 models. EPREL model counts by manufacturer are available, and total declared refrigerant counts are available, but the registry extract here does not join those two dimensions at brand level. That matters because the central buyer question — which major brands have shifted most strongly from R32 to R290 — cannot be answered cleanly without that cross-tab.

What the data does show is which manufacturers dominate EPREL listings overall. In the Househeating Pulse market view, Daikin Europe N.V. leads with 14,668 listed models and a 24.05% share of the 60,989-model universe (brand_share / EPREL Public API · brand-share aggregation). Mitsubishi Electric Europe B.V. follows with 5,575 models and 9.14% share (brand_share / EPREL Public API · brand-share aggregation), just ahead of JOHNSON CONTROLS HITACHI AIR CONDITIONING EUROPE SAS, SUCURSAL EN ESPAÑA at 5,207 models and 8.54% share (brand_share / EPREL Public API · brand-share aggregation).

Below that, Bosch Thermotechnik GmbH has 3,602 models and 5.91% share (brand_share / EPREL Public API · brand-share aggregation), Ariston SpA has 2,618 models and 4.29% share (brand_share / EPREL Public API · brand-share aggregation), ATLANTIC SOC FRANCAISE DEVELOP THERMIQUE has 1,516 models and 2.49% share (brand_share / EPREL Public API · brand-share aggregation), Vaillant GmbH has 1,195 models and 1.96% share (brand_share / EPREL Public API · brand-share aggregation), and BDR Thermea Group B.V. has 925 models and 1.52% share (brand_share / EPREL Public API · brand-share aggregation).

That list is still useful. It defines where any refrigerant transition would matter most in market visibility terms. If a brand with a quarter of all listings shifts portfolio composition, the signal is material. Readers wanting the broader brand universe can use the manufacturer index, while the live EPREL catalog remains the easiest way to inspect declared refrigerants model by model.

R290 versus R32: how big is each refrigerant in EPREL listings?

At declared-code level, R32 remains far larger than R290 in this EPREL extract. The registry records 13,935 listings with refrigerant declared as R32 (refrigerant_universe / IPCC AR6 GWP table; EU Reg. 2024/573 phase-out schedule; EPREL declared codes). It records 537 listings as R290 (refrigerant_universe / IPCC AR6 GWP table; EU Reg. 2024/573 phase-out schedule; EPREL declared codes).

Using only those two clean codes, R32 accounts for 96.29% of R32-plus-R290 declared listings, while R290 accounts for 3.71% (refrigerant_universe / IPCC AR6 GWP table; EU Reg. 2024/573 phase-out schedule; EPREL declared codes). In absolute terms, there are about 25.95 R32 listings for every R290 listing in the extract (refrigerant_universe / IPCC AR6 GWP table; EU Reg. 2024/573 phase-out schedule; EPREL declared codes).

There are also declaration-quality caveats. EPREL includes spelling variants and miscoded entries such as R290A at 2 listings, R290a at 1 listing, R410A at 1,896 listings, and R410a at 49 listings (refrigerant_universe / IPCC AR6 GWP table; EU Reg. 2024/573 phase-out schedule; EPREL declared codes). That is a reminder to treat raw refrigerant counts as administrative declarations, not laboratory-standard taxonomy. Househeating Pulse’s refrigerants reference is useful here, and the underlying regulatory context sits in the EU’s F-gas Regulation.

On the chemistry itself, the corpus identifies R290 as propane with GWP 0 and flammability class A3 (refrigerant_universe / IPCC AR6 GWP table; EU Reg. 2024/573 phase-out schedule; EPREL declared codes). R32 is difluoromethane with GWP 771 and flammability class A2L, with a phase-out date in this reference table of 2027-01-01 (refrigerant_universe / IPCC AR6 GWP table; EU Reg. 2024/573 phase-out schedule; EPREL declared codes). For buyers, that explains why R290 has strategic weight even though it is still much smaller in the declared listing base.

Brand-by-brand comparison: model counts, refrigerant mix, and average SCOP

The corpus can answer the model-count and average-SCOP part of the comparison for the top brands, but not the refrigerant mix by brand. The failed brand_detail probes also mean there is no fallback brand drilldown in this dataset.

Top listed brands in EPREL

All figures above are from brand_share / EPREL Public API · brand-share aggregation.

Within this leading group, the highest average SCOP belongs to Bosch Thermotechnik GmbH at 4.69 (brand_share / EPREL Public API · brand-share aggregation), followed by Ariston SpA at 4.66 (brand_share / EPREL Public API · brand-share aggregation), Vaillant GmbH at 4.54 (brand_share / EPREL Public API · brand-share aggregation), and Mitsubishi Electric Europe B.V. at 4.51 (brand_share / EPREL Public API · brand-share aggregation). The lowest among the top eight shown is Johnson Controls Hitachi at 4.18 (brand_share / EPREL Public API · brand-share aggregation).

What cannot be done from this corpus is the requested split between “brands with heavy R290 usage” and “brands still dominated by R32”. The registry sample here does not record brand-level R290 shares, so any such grouping would be speculative. For a live comparison workflow, readers can browse R290 listings, R32 listings, and the leaderboard hub side by side.

Noise matters: which brands and models are the quietest?

The strongest noise evidence in this corpus is at model level, not at top-brand average level. That distinction matters. EPREL here can identify some extremely quiet listed models, but it does not provide average outdoor noise by leading brand in the supplied extracts.

Among the 20 quietest models returned by the catalog query, the first 13 entries are almost entirely industrial-looking or specialist machines rather than the mainstream brand set above. The quietest listed entries include WAMAK, s.r.o. TBW 28 EVI at 1 dB outdoor noise with SCOP 5.42 (top_models / EPREL Public API via Househeating Pulse catalog), WAMAK, s.r.o. TWW 110 WHR at 1 dB with SCOP 6.75 (top_models / EPREL Public API via Househeating Pulse catalog), WAMAK, s.r.o. TWW 28 EVI at 1 dB with SCOP 4.30 (top_models / EPREL Public API via Househeating Pulse catalog), WAMAK, s.r.o. TWW 60 EVI at 1 dB with SCOP 6.27 (top_models / EPREL Public API via Househeating Pulse catalog), WAMAK, s.r.o. TWW 48 EVI at 1 dB with SCOP 6.80 (top_models / EPREL Public API via Househeating Pulse catalog), and WAMAK, s.r.o. TWW 85 WHR at 1 dB with SCOP 6.67 (top_models / EPREL Public API via Househeating Pulse catalog).

That does not mean buyers should treat 1 dB as a realistic benchmark for the mass market. The supplied list is simply a ranking of the quietest declared entries, and the brands dominating that list are not the same as the manufacturers dominating overall EPREL share. Panasonic appears once with a 5 dB test entry, “TEST_JOE20190604_2”, with no SCOP value recorded (top_models / EPREL Public API via Househeating Pulse catalog). Outside that, the mainstream leaders are mostly absent from the noise top 20.

The refrigerant alignment is also unresolved. Most of the quietest entries have refrigerant listed as null in the extract (top_models / EPREL Public API via Househeating Pulse catalog), so the corpus does not support a claim that the quietest cluster is R290-led or R32-led. A few quiet air-air entries in the top 20 do show R290, including MH Handel GmbH’s 823-078V73WT at 2 dB (top_models / EPREL Public API via Househeating Pulse catalog), but this is not enough to generalise across top brands.

For readers focused on siting constraints and neighbour impact, the live quietest models leaderboard and the air-to-water catalog slice are more decision-useful than a simple refrigerant narrative.

Is quietness or efficiency coming with a price premium?

The registry does not record price in the supplied corpus, so it cannot quantify a euro premium for quieter or higher-efficiency models. There are no retail, wholesale, or installed-cost fields here, and no brand-level price ranking.

What can be said is narrower. Some of the very quiet models in the corpus also post unusually high SCOP values: for example, WAMAK TWW 48 EVI combines 1 dB and SCOP 6.80 (top_models / EPREL Public API via Househeating Pulse catalog), while WAMAK TWW 110 WHR combines 1 dB and SCOP 6.75 (top_models / EPREL Public API via Househeating Pulse catalog). But these are not tied to any price signal, and they are not part of the top-brand set by listing share. So there is evidence that ultra-quiet and high-efficiency can coincide in EPREL declarations, but not that buyers pay a specific premium for it.

Likewise, the top_models query for SCOP-descending R290 units returned no records (top_models / EPREL Public API via Househeating Pulse catalog). That does not prove R290 lacks high-SCOP models; it only means this particular query produced an empty result in the supplied extract.

For cost work, buyers need to combine EPREL performance data with installer quotations, subsidy rules, and expected operating conditions. Househeating Pulse’s payback calculator, subsidy calculator, country comparison dashboard, and climate-fit tool are the more appropriate next steps.

What buyers and installers should conclude in 2026

Three points are firm.

First, R32 is still the overwhelmingly larger declared refrigerant in this EPREL extract, at 13,935 listings versus 537 for R290 (refrigerant_universe / IPCC AR6 GWP table; EU Reg. 2024/573 phase-out schedule; EPREL declared codes). Anyone describing the 2026 catalog as already R290-dominated would be overstating the evidence.

Second, the leading brands by listing volume are not a tight performance bloc. Among the biggest manufacturers, average SCOP ranges from 4.18 for Johnson Controls Hitachi to 4.69 for Bosch Thermotechnik (brand_share / EPREL Public API · brand-share aggregation). That spread is large enough that brand scale alone says little about heating efficiency.

Third, the corpus does not support a clean brand-level story about who has shifted most aggressively to R290, nor does it support a quantified price premium for either R290 or quietness. Buyers should not confuse “R290 is strategically important” with “every major portfolio is already mostly R290”. The registry extract here simply does not prove that.

The practical approach in 2026 is therefore less ideological and more comparative: shortlist by application and type in the full heat-pump catalog, filter by declared R290 models or by top energy labels such as A+++ listings, then compare noise, SCOP, and climate suitability against local incentives. For definitions and caveats in how these fields are standardised, the methodology page and glossary are worth keeping open in another tab.

Sources

• EPREL Public API · brand-share aggregation — snapshot 2026-07-03
• EPREL Public API via Househeating Pulse catalog — snapshot 2026-07-03
• IPCC AR6 GWP table; EU Reg. 2024/573 phase-out schedule; EPREL declared codes — snapshot 2026-07-03
• probe failed — data unavailable (brand_detail for daikin) — no snapshot date supplied
• probe failed — data unavailable (brand_detail for midea) — no snapshot date supplied

Continue reading

• Heat pump grants and subsidies by country — Map incentives before comparing installed cost.
• How to size a heat pump correctly — Avoid oversizing and make sense of kW figures in EPREL.
• SCOP, SEER and energy labels explained — A quick refresher on the metrics used in EU product listings.
• R290 heat pumps: safety, rules and trade-offs — What propane changes for siting, handling and buyer expectations.