2026 EU heat-pump refrigerant market share by brand: who leads on R290, R32 and beyond

The 2026 refrigerant map: which gases actually matter in EPREL

The EU heat-pump catalog in 2026 is large, but the refrigerant mix recorded in EPREL is not especially balanced. The registry snapshot contains 60,989 heat-pump models from 777 manufacturers, with an average SCOP of 4.55 (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API). For buyers browsing the live EPREL catalog, the practical refrigerant story is dominated by one gas above all others.

R32 accounts for 13,935 listings, equal to 22.85% of all models (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API). R290 accounts for 537 listings, or 0.88% of all models (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API). R410A accounts for 1,896 listings, or 3.11% of all models, while the lowercase variant R410a adds another 49 listings, taking the combined R410A/R410a total to 1,945, or 3.19% (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API). Everything else is numerically tiny: R134A appears in 2 listings, R23 in 1, and a long tail of single-entry declarations such as R420A, R423A and R425A each account for effectively 0.00% when rounded to two decimals (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API).

That leaves a large remainder of 44,572 models not represented in the listed refrigerant counts above, equal to 73.08% of all EPREL heat-pump entries (derived from brand_share total_models and market_index_snapshot by_refrigerant / EPREL Public API · brand-share aggregation; Househeating Pulse · Market Index v1, computed from EPREL Public API). The corpus does not explain that gap. Put plainly, the registry snapshot available here does not record a complete refrigerant value for every model, so no honest article can assign the missing 73.08% to named gases.

What can be said with confidence is that among explicitly recorded refrigerants, R32 listings dwarf R290 listings, while older HFC options persist only as a much smaller tail. The refrigerant reference table also shows why these names matter commercially: R290 is a natural hydrocarbon with GWP 0, R32 is an HFC with GWP 771, and R410A is an HFC blend with GWP 1,924 (refrigerants reference; refrigerant_universe / IPCC AR6 GWP table; EU Reg. 2024/573 phase-out schedule; EPREL declared codes). For regulatory background, the governing framework is the EU F-gas regulation Regulation (EU) 2024/573.

Who leads overall: the biggest heat-pump manufacturers by listing volume

Across all heat-pump listings, the market is heavily concentrated at the top. Daikin Europe N.V. leads with 14,668 models and a 24.05% share (brand_share / EPREL Public API · brand-share aggregation). Mitsubishi Electric Europe B.V. follows with 5,575 models and 9.14% (brand_share / EPREL Public API · brand-share aggregation), ahead of JOHNSON CONTROLS HITACHI AIR CONDITIONING EUROPE SAS, SUCURSAL EN ESPAÑA at 5,207 models and 8.54% (brand_share / EPREL Public API · brand-share aggregation).

The top five brands together account for 31,670 models, or 51.93% of all listings (derived from brand_share ranking and total_models / EPREL Public API · brand-share aggregation). That is the first important buying signal: overall catalog breadth is not evenly distributed across the manufacturer base.

Top manufacturers overall

On this measure, scale and average efficiency do not line up neatly. The largest brand, Daikin, sits 0.11 SCOP points below the market average of 4.55, while Bosch is 0.14 points above and Ariston is 0.11 points above (brand_share; market_index_snapshot / EPREL Public API · brand-share aggregation; Househeating Pulse · Market Index v1, computed from EPREL Public API). Further down the ranking, STIEBEL ELTRON averages 4.84 and LG averages 4.93, both well above the market mean, but with only 433 and 405 listings respectively (brand_share / EPREL Public API · brand-share aggregation). Readers comparing broad line-ups can track these shifts in the manufacturer index and the leaderboard hub.

R290 leaders: the brands most associated with propane models

This is where the corpus becomes more restrictive. The article brief asks for brand-by-refrigerant concentration and top-five brand counts within R290, but the dataset supplied here does not include a brand-by-refrigerant breakdown. There is no table showing which manufacturers own the 537 R290 entries, no brand ladder within R290, and no per-brand SCOP for the R290 subset. The top_models probe for R290 is empty, so the registry snapshot available here also does not identify the highest-SCOP R290 models (top_models / EPREL Public API via Househeating Pulse catalog).

That means the registry does not record, in the supplied corpus, which brands are the clear leaders on R290 listings specifically. It also does not let this article quantify the gap between the top R290 brand and the next-ranked manufacturers.

Still, two numerically grounded points matter:

• R290 appears in 537 listings, equal to 0.88% of all 60,989 models (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API).
• Natural refrigerants as a group account for 3.27% of all listings (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API).

Since the explicit refrigerant listing count for R290 is 537 and the total natural-refrigerant share is 3.27%, propane is only one part of a still-small natural-refrigerant footprint in EPREL (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API). Buyers specifically shopping propane units can browse the filtered R290 catalog, but the corpus does not support a defensible top-five manufacturer ranking inside that subset.

R32 leaders: the mainstream brand stack and where it differs from R290

R32 is much easier to describe at the refrigerant level than at the brand level. It appears in 13,935 listings, representing 22.85% of all EPREL models (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API). That makes R32 the overwhelmingly dominant explicitly declared refrigerant in the corpus.

But the same data limitation applies here too. The supplied probes do not show which brands dominate R32 specifically, how concentrated the top five are within the R32 subset, or whether the R32 ladder materially overlaps with the R290 ladder. The top_models probe for R32 is also empty, so the registry does not provide the top-SCOP R32 models in this research packet (top_models / EPREL Public API via Househeating Pulse catalog).

What can still be stated, and is relevant to buyers, is that the overall brand hierarchy is likely to shape a large part of the R32 offer simply because the catalog itself is so concentrated. Daikin alone accounts for 24.05% of all heat-pump listings, Mitsubishi Electric 9.14%, and Johnson Controls Hitachi 8.54% (brand_share / EPREL Public API · brand-share aggregation). Together those three manufacturers cover 25,450 models, or 41.73% of the whole EPREL heat-pump catalog (derived from brand_share ranking and total_models / EPREL Public API · brand-share aggregation). If a buyer sees a dense R32 field in the full catalog, the overall manufacturer concentration suggests that a few large OEMs are likely to be highly visible there too, even though this corpus does not permit an exact R32-only ranking.

That gap is worth flagging because it changes how the data should be read. The current corpus supports a robust refrigerant prevalence view and a robust overall brand-share view, but not a joined refrigerant-by-brand matrix. The underlying methodology notes should therefore matter as much as the headline ranking.

Niche refrigerants and small-brand clusters: where the market is still fragmented

Outside R32, R290 and R410A, EPREL’s declared refrigerant field is mostly a residual tail. Combined niche declarations beyond those three amount to 70 listings, or 0.11% of the 60,989-model catalog (derived from market_index_snapshot by_refrigerant and total_models / Househeating Pulse · Market Index v1, computed from EPREL Public API). Within that tail:

• R410a contributes 49 listings, or 0.08% (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API).
• R134A contributes 2 listings, or 0.00% when rounded to two decimals (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API).
• R23, R420A, R423A, R421A, R411A, R35, R422A, R413A, R41OA, R425A, R419A, R33, R417A and R332 each contribute 1 listing, each around 0.0016% of the catalog (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API).

The naming itself is messy. EPREL declarations include duplicate or inconsistent labels such as R410A versus R410a, and R290 versus R290A versus R290a (refrigerant_universe; market_index_snapshot / IPCC AR6 GWP table; EU Reg. 2024/573 phase-out schedule; EPREL declared codes; Househeating Pulse · Market Index v1, computed from EPREL Public API). That matters when readers interpret tiny counts: some fragmentation is commercial, some is simply registry hygiene.

The brief also asks which small set of brands controls these niche gases. The corpus cannot answer that. There is no manufacturer attribution for those niche refrigerants in the supplied data, so no credible top-five brand concentration metric can be published for them.

What buyers should infer: concentration, choice, and the gap between volume leaders and efficiency leaders

Three points survive the data limitations and are directly useful.

First, the market proxy itself is concentrated. The five largest manufacturers account for 51.93% of all EPREL listings, while Daikin alone accounts for 24.05% (brand_share / EPREL Public API · brand-share aggregation). For buyers, that means breadth of range, documentation load and installer familiarity will often cluster around a small number of brands, especially when using the brand-filtered catalog or browsing major suppliers such as Bosch Thermotechnik GmbH and Ariston SpA.

Second, explicit refrigerant declarations are not evenly distributed. R32’s 13,935 listings are about 25.95 times the size of R290’s 537 listings (derived from market_index_snapshot by_refrigerant / Househeating Pulse · Market Index v1, computed from EPREL Public API). So for shoppers who care about comparing many like-for-like products today, R32 is the broad mainstream in the declared-data slice, while propane remains much narrower.

Third, scale does not equal efficiency leadership. The market-average SCOP is 4.55, but the largest brand, Daikin, averages 4.44, while Bosch averages 4.69, Ariston 4.66, STIEBEL ELTRON 4.84 and LG 4.93 (brand_share; market_index_snapshot / EPREL Public API · brand-share aggregation; Househeating Pulse · Market Index v1, computed from EPREL Public API). So buyers looking beyond simple catalog presence should check the top SCOP leaderboards and, for hydronic systems, the dedicated air-to-water SCOP ranking, not just overall brand scale.

What the corpus does not prove is equally important. It does not show which brands lead R290 specifically, which brands lead R32 specifically, how concentrated each refrigerant subgroup is, or whether the best-scoring R290 and R32 models come from the same brands that dominate listing counts. Until a brand-by-refrigerant cross-tab is available, those questions remain open in this dataset.

For now, the strongest evidence-based reading is narrower: the 2026 EU heat-pump catalog is dominated by a few very large manufacturers overall, and the declared refrigerant field is dominated by R32, with R290 present but still comparatively scarce in EPREL. Buyers comparing products now should treat refrigerant choice and manufacturer scale as separate filters, then use performance tools such as the climate-fit analyzer and payback calculator to narrow the shortlist further.

Sources

• EPREL Public API · brand-share aggregation — snapshot 2026-06-07
• Househeating Pulse · Market Index v1, computed from EPREL Public API — snapshot 2026-06-07
• EPREL Public API via Househeating Pulse catalog — snapshot 2026-06-07
• IPCC AR6 GWP table; EU Reg. 2024/573 phase-out schedule; EPREL declared codes — snapshot 2026-06-07

Continue reading

• How to use EPREL heat-pump listings without misreading model counts — A practical guide to what listing volume does and does not tell you.
• R290 vs R32 heat pumps: comparing catalog breadth, not marketing claims — Focuses on where each refrigerant appears in real product databases.
• How to shortlist heat pumps by SCOP, noise and refrigerant — A workflow for turning a large catalog into a workable buyer shortlist.
• Why manufacturer averages can hide strong and weak model ranges — Explains the limits of brand-level averages when choosing a specific unit.