R290 vs R32 in 2026: the brand split that now defines Europe’s heat-pump market

The 2026 split: how much of EPREL is now R290 vs R32

Europe’s EPREL heat-pump catalog is still overwhelmingly an R32 market in 2026: 13,935 listed models use R32 versus just 537 using R290, or 22.85% against 0.88% of the 60,989-model universe (market_index_snapshot). That is the core market fact behind the refrigerant debate: propane is visible, but still numerically small in the database.

The rest of the listed universe is fragmented. R410A still appears on 1,896 models, equal to 3.11% of all listings, while all other declared refrigerants combined account for 14 models if counted strictly outside R32, R290 and R410A, or 14,621 models if grouped as “everything not R290” once R32 and its two propane spelling variants are excluded (market_index_snapshot). The data does not support a neat middle camp between R290 and R32; it supports one dominant mainstream refrigerant, one small but watched natural-refrigerant niche, and a long tail.

That matters for how buyers should read brand strategy. A market where R290 listings remain below 1% of EPREL is not yet a market where propane has won on breadth. It is a market where propane can be used as a signal. For the wider baseline, the market index snapshot and refrigerants reference show just how uneven that split remains.

Which brands have gone hardest into propane, and which still anchor on R32

The surprise is that the supplied 2026 corpus does not include manufacturer-level refrigerant splits, so it cannot answer which brands have the highest share of R290 models or the highest share of R32 models in their own catalogs. That is the central limitation here.

What the corpus does show is which manufacturers matter most in the overall EPREL universe. Daikin Europe N.V. leads by a wide margin with 14,668 models, or 24.05% of all listings, followed by Mitsubishi Electric Europe B.V. with 5,575 models and 9.14%, and Johnson Controls Hitachi Air Conditioning Europe with 5,207 models and 8.54% (brand_share). Bosch Thermotechnik GmbH is fourth with 3,602 models and a notably high average SCOP of 4.69, ahead of Ariston SpA on 2,618 models and average SCOP 4.66 (brand_share).

Among the rest of the top 15, Vaillant GmbH has 1,195 models, BDR Thermea 925, Panasonic 894, Ferroli 685, Gree 639, Toshiba Carrier 497, Riello 464 and Gorenje 440 (brand_share). Those brand sizes matter because any shift from R32 to R290 by a top-five incumbent would have a much larger effect on the market mix than a similar shift by a smaller specialist.

So the article’s proposed “top R290 adopters by brand share” and “top R32 holdouts by brand share” cannot be ranked from this corpus. The database snapshot supports the market split, but not the manufacturer-by-refrigerant league table.

What the brand catalogs say about scale and efficiency

Although the corpus cannot split each catalog by refrigerant, it does show a clear divide in scale across Europe’s big manufacturers. The top three brands alone account for 25,450 models — 41.73% of the full EPREL catalog — and none of them posts the strongest average SCOP among the top 15 (brand_share).

Daikin’s average SCOP is 4.44, Mitsubishi Electric’s is 4.51, and Johnson Controls Hitachi’s is 4.18 (brand_share). By contrast, several smaller top-15 manufacturers score higher on average efficiency: Bosch reaches 4.69, Gorenje 4.67, Ariston 4.66, Gree 4.65, Ferroli 4.64 and Riello 4.60 (brand_share). That does not prove refrigerant causality, but it does fit the broader commercial pattern implied by the angle: scale leaders tend to optimise for breadth, while some smaller or more heating-focused players can concentrate around higher-performing lines.

The market-wide average SCOP is 4.55 across all 60,989 models, so Daikin sits 0.11 points below the market average, Mitsubishi Electric is 0.04 below, and Johnson Controls Hitachi is 0.37 below (market_index_snapshot). Bosch is 0.14 points above the market average; Ariston is 0.11 above (market_index_snapshot; brand_share). For readers comparing manufacturer profiles or scanning the full heat-pump catalog, that is the more robust 2026 brand signal available from the supplied data: catalog scale and average efficiency do not move together in a simple way.

The requested comparison between “R290-heavy” and “R32-heavy” brands by average catalog size and average SCOP cannot be calculated from the corpus because the brand-level refrigerant mix is missing.

Are R290 models clustering at the top of the performance table?

The strongest evidence supplied for premium positioning is indirect. Across the full EPREL market, 23,466 models carry A+++, equal to 38.48% of all listings, while the average SCOP is 4.55 (market_index_snapshot). That establishes where the top tier begins, but it does not by itself assign that tier to R290.

The corpus then fails to return any entries for the “top 20 R290 models by SCOP” query: the result is an empty array (top_models). That means the supplied data does not prove that R290 models are overrepresented among the highest-SCOP products, nor does it quantify how often propane units reach the very top of the top SCOP leaderboard or the air-to-water SCOP leaderboard. It may be true in the underlying market, but it is not demonstrated here.

What can be said is narrower. Natural refrigerants account for 3.27% of the market by the dataset’s own “natural refrigerant share” field, while explicit R290 declarations account for 537 models, plus three additional propane spelling variants, or 540 total propane-coded entries if those variants are grouped together (market_index_snapshot). Because that footprint is so small, any concentration in top-efficiency classes would be meaningful if confirmed. But this corpus does not confirm it.

What this means for pricing power and market positioning in 2026

The 2026 signal is less “propane has taken over” than “propane remains scarce enough to function as positioning.” With just 537 R290 models against 13,935 R32 models, R290 is still a minority technology in EPREL by count, not a mainstream default (market_index_snapshot). That supports the idea of R290 as a selective brand and product-line choice rather than a continent-wide volume strategy.

At the same time, the largest incumbents still dominate the market through catalog breadth. Daikin alone accounts for nearly one in four listed models, and the top four brands together reach 47.64% of all EPREL entries (brand_share). That is what “protecting breadth and mass-market coverage” looks like in the data: scale sits with multi-thousand-model catalogs, not with the tiny propane footprint.

The missing piece is price. The supplied corpus contains no price field and no manufacturer-level refrigerant mix, so it cannot directly show an R290 price premium or tie that premium to specific brands. The best available proxy would have been top-efficiency concentration, but the R290 top-model extract returned no rows (top_models). So the stronger pricing claim cannot be made from this evidence alone.

What can be said, and said cleanly, is that Europe’s 2026 EPREL market still runs on R32 volume, while the higher-efficiency brand story is more mixed and does not map automatically onto scale leaders (market_index_snapshot; brand_share). For anyone tracking where the next shift could show up, the places to watch are the R290 catalog slice, the brand leaderboard hub, and individual large-brand pages such as Bosch Thermotechnik GmbH and Daikin Europe N.V.. If those big catalogs start moving materially, the market split will stop being a niche signal and start becoming a volume story.

Sources

market_index_snapshot — Househeating Pulse · Market Index v1, computed from EPREL Public API. Snapshot: 2026-06-25.
refrigerant_universe — IPCC AR6 GWP table; EU Reg. 2024/573 phase-out schedule; EPREL declared codes. Snapshot: 2026-06-25.
brand_share — EPREL Public API · brand-share aggregation. Snapshot: 2026-06-25.
top_models — EPREL Public API via Househeating Pulse catalog. Snapshot: 2026-06-25.