Househeating Pulse
EU Heat-Pump Market Intelligence

Comparison · 10 min read · Updated 2026-05-28

2026 EU heat-pump price vs efficiency: how R290 and R32 differ in the EPREL data

Using EPREL listings, this article compares purchase prices and efficiency ratings for R290 and R32 heat pumps across the EU, showing where the refrigerant premium buys better performance and where it does not.

R290 vs R32 in EPREL: how many models, brands, and listings are actually in the sample

The EPREL snapshot used here covers 60,989 heat-pump models from 777 manufacturers as of 2026-05-28 (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API). Within that universe, R32 appears on 13,935 listings while R290 appears on 537 listings (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API).

That makes R32 a very large segment at 22.85% of all listed heat pumps, versus 0.88% for R290, calculated directly from the same listing counts and total model count of 60,989 (market_index_snapshot / Househeating Pulse · Househeating Pulse · Market Index v1, computed from EPREL Public API). Put differently, R32 outnumbers R290 by roughly 26 to 1 in this registry snapshot, based on 13,935 versus 537 listed models (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API).

For readers wanting to inspect the raw listing universe, the live EPREL catalog can be filtered directly to R32 models and R290 models. The wider market context sits in the 2026 market index, where the average SCOP across the full sample is 4.55 and the average outdoor sound power is 61.3 dB (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API).

The corpus does not provide a refrigerant-by-brand cross-tab, so the registry snapshot here cannot say how many brands sell R290 versus R32 specifically. It can say only that the overall market includes 777 manufacturers, led by Daikin Europe N.V. with 14,668 listings or 24.05% of the full catalog, Mitsubishi Electric Europe B.V. with 5,575 listings or 9.14%, and Bosch Thermotechnik GmbH with 3,602 listings or 5.91% (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API). Other large groups in the same market-wide ranking include Ariston SpA, ATLANTIC SOC FRANCAISE DEVELOP THERMIQUE, Vaillant GmbH, BDR Thermea Group B.V., and JOHNSON CONTROLS HITACHI AIR CONDITIONING EUROPE SAS, SUCURSAL EN ESPAÑA (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API).

Efficiency first: the SCOP gap between refrigerants, overall and by product type

The article brief asks for the average SCOP difference between R290 and R32 overall and by product type. The corpus, however, does not contain a refrigerant-by-SCOP aggregation. It includes overall type averages, but not type averages split into R290 versus R32. So the registry does not record, in the supplied dataset, the average SCOP gap between R290 and R32 either across the whole market or within each major heat-pump type.

What can be stated is the type baseline against which any refrigerant comparison would have to be judged. Across all refrigerants, average SCOP is 4.77 for ground-water units, 4.54 for air-water units, and 6.15 for water-water units (type_efficiency / EPREL Public API · type aggregation). For air-air and heat-pump water heaters, the supplied type aggregation has no average SCOP value recorded, so no type-level SCOP benchmark can be reported there (type_efficiency / EPREL Public API · type aggregation).

That matters because product mix is likely to affect any refrigerant comparison. In the provided top-model slices, all listed low-capacity R290 examples are air-air units, and the same is true for the low-capacity R32 slice (top_models power_asc refrigerant=R290 / EPREL Public API via Househeating Pulse catalog; top_models power_asc refrigerant=R32 / EPREL Public API via Househeating Pulse catalog). Without a direct refrigerant-by-type efficiency table, broad claims that “R290 is X points better on SCOP” would be stronger than the evidence allows.

For readers tracking the highest-efficiency listings more generally, the top SCOP leaderboards, plus the type-specific air-to-water SCOP ranking and ground-source SCOP ranking, are the right next stop. But from this corpus alone, the SCOP gap question remains unanswerable numerically.

Price versus performance: where the R290 premium buys more efficiency, and where it does not

The brief also asks for the average purchase-price gap between R290 and R32 overall and by type. The supplied corpus contains no price field at all. EPREL can include commercial information in some contexts, but this dataset does not expose purchase prices, price averages, or price-ranked model slices. As a result, the registry snapshot here cannot quantify any R290 premium, any R32 discount, or whether a price gap persists after splitting by type.

That limitation should shape the buying takeaway. The current evidence supports a statement about availability and product mix, not about purchase economics. R32 is deeply mainstream in the registry at 13,935 listings, while R290 remains niche at 537 (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API). Buyers comparing quotations therefore face a much larger searchable R32 catalog segment than R290 segment, which alone can influence the practical chance of finding a close substitute.

The corpus also cannot answer whether an R32 model “matches or beats” an R290 model on SCOP while staying cheaper, because neither top-SCOP slices nor price data are included. The two supplied top_models probes sorted by scop_desc for R290 and R32 are empty arrays in this research package, so there is no evidence here on top-SCOP outliers by refrigerant (top_models scop_desc refrigerant=R290 / EPREL Public API via Househeating Pulse catalog; top_models scop_desc refrigerant=R32 / EPREL Public API via Househeating Pulse catalog).

Top-end models and outliers: which refrigerant dominates the best SCOP rankings

On the specific question of the top SCOP rankings, the supplied research corpus simply does not provide the list. Both top-25 SCOP queries by refrigerant return no records in this package (top_models scop_desc refrigerant=R290 / EPREL Public API via Househeating Pulse catalog; top_models scop_desc refrigerant=R32 / EPREL Public API via Househeating Pulse catalog). That means the following requested measures cannot be answered from the evidence provided:

  • how many of the top SCOP models are R290 versus R32;
  • what SCOP threshold separates the best-performing models from the rest;
  • how many notable R32 price-performance outliers appear in the top model slice.

What the corpus does allow is a read on energy-class concentration, which is a rougher efficiency proxy than SCOP. Across the whole market, 23,466 models carry A+++ heating class, equal to 38.48% of all 60,989 listed models (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API). Another 8,924 are A++ or 14.63%, and 16,845 are A+ or 27.62% (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API).

But the corpus does not split energy class by refrigerant either. So it cannot support the requested ranking of A+++ share for R290 versus R32, nor the percentage-point difference between them. Readers can still inspect the A+++ segment of the catalog and compare it with the refrigerants reference, but that step goes beyond the quantitative evidence supplied for this article.

Capacity and noise: whether the efficiency premium comes with a different hardware profile

Here the corpus is more useful, though still only partially. It does not provide average capacity and average noise for all R290 versus all R32 listings. It does, however, provide a concrete low-capacity slice for each refrigerant, both sorted by minimum power ascending.

In the smallest-capacity R290 top-25 slice, the minimum listed capacity ranges from 0.7 kW to 3.2 kW, with the smallest example being Swegon Germany GmbH GAW 30 ECO at 0.7 kW and rank 1 (top_models power_asc refrigerant=R290 / EPREL Public API via Househeating Pulse catalog). In the equivalent R32 slice, the range runs from 1.0 kW to 1.9 kW, with GESTION INTEGRAL DE ALMACENES, S.L. SNWS026VE2 at 1.0 kW in rank 1 (top_models power_asc refrigerant=R32 / EPREL Public API via Househeating Pulse catalog). So the smallest listed R290 unit in this slice is 0.3 kW smaller than the smallest listed R32 unit, based on 0.7 kW versus 1.0 kW (top_models power_asc refrigerant=R290 / EPREL Public API via Househeating Pulse catalog; top_models power_asc refrigerant=R32 / EPREL Public API via Househeating Pulse catalog).

Noise is more mixed. In the R290 low-capacity slice, several entries report 0 dB outdoor noise, which is almost certainly a placeholder or missing-value artifact rather than a physical result: for example, Klimabrands A/S Quick Install and Klimabrands A/S Quick Install V2 both show 0 dB, as does the smallest Swegon Germany GmbH GAW 30 ECO entry (top_models power_asc refrigerant=R290 / EPREL Public API via Househeating Pulse catalog). Excluding interpretation, the observed non-zero R290 values in this slice run from 50 dB to 65 dB, including 50 dB for SC Trade & Services GmbH KBO12-R290 and 65 dB for INFINITON WORLD ELECTRONIC SL PACSF25A (top_models power_asc refrigerant=R290 / EPREL Public API via Househeating Pulse catalog).

In the R32 low-capacity slice, reported outdoor sound values span from 59 dB to 65 dB where present, with multiple 63 dB entries from Bosch and 65 dB for Toshiba’s RAS-05BAVG-E+RAS-B05BKVG-E (top_models power_asc refrigerant=R32 / EPREL Public API via Househeating Pulse catalog). The broad message is not that one refrigerant is quieter, because this slice is too narrow and too incomplete for that. It is that the smallest-capacity products in both refrigerants cluster within a fairly similar outdoor sound band once obvious zero-value anomalies are set aside.

All 25 models in the R290 low-capacity slice are air-air units, and all 25 in the R32 slice are also air-air units (top_models power_asc refrigerant=R290 / EPREL Public API via Househeating Pulse catalog; top_models power_asc refrigerant=R32 / EPREL Public API via Househeating Pulse catalog). That lines up with the broader catalog, where air-air accounts for 21,065 listings and has an average power of 5.41 kW plus average outdoor noise of 64.1 dB across all refrigerants (type_efficiency / EPREL Public API · type aggregation). Readers comparing physically small systems can also browse the smallest-capacity leaderboard and the quietest heat pumps.

What buyers should infer from the EPREL split: consistent premium, mixed premium, or no clear premium

From this research package alone, the safest answer is no clear premium can be proven on price or SCOP, because the necessary refrigerant-split metrics are missing. The corpus does not contain average prices, refrigerant-level SCOP averages, refrigerant-by-type SCOP splits, refrigerant-level A+++ shares, or populated top-SCOP model slices.

What it does show clearly is a market-structure split. R32 is the established mass-market refrigerant in the current EPREL universe, with 13,935 listings and 22.85% share of all heat pumps, while R290 remains much smaller at 537 listings and 0.88% share (market_index_snapshot / Househeating Pulse · Market Index v1, computed from EPREL Public API). That alone implies different search conditions for buyers and installers: more breadth, more near-duplicates, and probably more substitution options within R32 simply because the listed pool is far larger.

The smaller-capacity slices add one practical point. R290 is not confined to large monobloc heating boxes in this dataset; it also appears in very small air-air products down to 0.7 kW, such as Swegon Germany GmbH GAW 30 ECO, and in portable or compact entries like DWD-Company Verwaltungs GmbH TVKA18000W and HAIER IBERIA, SLU AS35PBBHRA / 1U35YEBGRA (top_models power_asc refrigerant=R290 / EPREL Public API via Househeating Pulse catalog). That is a useful corrective to simplistic segment assumptions.

For buyers, the practical implication is to treat refrigerant as a filter, not as a shortcut for performance. Start with the relevant heat-pump type view or air-air catalog, then compare actual listed metrics model by model. The methodology notes explain where EPREL fields are complete and where they are not, and the official EPREL public portal plus EU energy-labelling framework set the underlying data structure.

The broader editorial angle remains plausible but not fully testable here: R290 may often coincide with stronger efficiency positioning, yet the supplied evidence is not enough to put a number on that claim. For this corpus, the defensible summary is narrower: R32 dominates by count, R290 remains a small but visible segment, and the available slices do not support a blanket claim that one refrigerant systematically delivers better EPREL efficiency at a given price.

Sources

  • Househeating Pulse · Market Index v1, computed from EPREL Public API — snapshot 2026-05-28
  • EPREL Public API · type aggregation — snapshot 2026-05-28
  • EPREL Public API via Househeating Pulse catalog, top_models sorted by scop_desc for refrigerant R290 — snapshot 2026-05-28
  • EPREL Public API via Househeating Pulse catalog, top_models sorted by scop_desc for refrigerant R32 — snapshot 2026-05-28
  • EPREL Public API via Househeating Pulse catalog, top_models sorted by power_asc for refrigerant R290 — snapshot 2026-05-28
  • EPREL Public API via Househeating Pulse catalog, top_models sorted by power_asc for refrigerant R32 — snapshot 2026-05-28

Continue reading