MHA Data, Lopez Labs, OMNI Study, EPA Exemption, Calculator Software & Cross-Jurisdictional Comparison (US / EU / UK)
A masonry heater is a high-mass (typically 900–3,600 kg / 2,000–8,000 lbs), site-built solid fuel appliance that stores heat in its masonry thermal mass and radiates it slowly over 18–24 hours after a single, rapid burn cycle of 1.5–2 hours. A single charge of wood (typically 25–30 kg / 55–65 lbs) is burned rapidly at a high burn rate (10–15 kg/hr), then the stored heat radiates with zero emissions during the heat release period. This is the fundamental distinction from a woodstove, which burns continuously and releases heat in real time through thin metal walls.
Why this matters for emissions: The high burn rate (10–15 kg/hr, compared to 1–2 kg/hr in a damped-down woodstove) means combustion occurs in air-rich, high-temperature conditions. Smoldering — the primary source of PM in residential wood combustion — never occurs in a properly operated masonry heater. The operator cannot "damp down" the fire because there are no air controls to restrict; combustion air flows freely through the firebox.
Why this matters for regulation: All three major regulatory jurisdictions (US, EU, UK) have developed emission frameworks primarily for continuous-burn appliances. Masonry heaters fit awkwardly — or not at all — into these frameworks.
Lopez Labs (now the MHA Research Laboratory) maintains the largest masonry heater performance database in North America. Key milestones:
| Parameter | Instrument / Method | Notes |
|---|---|---|
| PM (particulate matter) | Oregon Method 41 (Condar portable dilution tunnel) | Constant-flow sampling; captures total PM (filterable + condensable) on glass fibre filter. Result reported in g/kg. |
| CO (carbon monoxide) | Testo 330-2 commercial combustion analyser | Widely used in European testing; auto-calibration routine |
| O₂ | Testo 330-2 | Manufacturer's stated accuracy ±0.2 Vol.% |
| CO₂ | Calculated from O₂ and CO | Standard combustion stoichiometry |
| Efficiency | Calculated via Oregon Method 41 (Barnett, 1985) | Uses CO and PM emissions, stack dilution, stack temperature, wood type, and moisture content |
| Temperature | Stack thermocouples, firebox surface sensors | Multiple locations; internal pressure sensors added for calculator calibration |
| Opacity | Added to data acquisition from 2018 | Real-time optical measurement of flue gas opacity |
What was NOT measured: NOx, OGC (organic gaseous carbon), SO₂, individual PAH species, PM2.5 as a separate fraction. The Condar captures total PM including PM2.5 but does not fractionate.
| Dataset | # Runs | Average PM | Range | Notes |
|---|---|---|---|---|
| All HeatKit runs | 87 | 1.29 g/kg | 0.48–4.61 | Wide range of conditions including worst-case |
| "Normal" conditioned HeatKit | 52 | 0.97 g/kg | — | Warm heater, top ignition, cordwood, excludes cold starts and cribs |
| Double Bell (Chernov) | 12 (cordwood) | 1.01 g/kg | 0.45–1.67 | All maple, various firebox/air configurations |
| Double Bell wet cribs | 2 | 3.43 g/kg | 3.05–3.80 | Intentional worst case: 35% moisture pieces |
| All 103 tests (2010 summary) | 103 | 1.27 g/kg | 0.48–4.61 | Combined HeatKit + Double Bell |
| 2019 cold firebox series (eco-firebox) | 13 | ~1.09 g/kg (typical) | 0.64–2.01 | Austrian eco-firebox, fired every 2nd day |
The 52 "normal" HeatKit runs averaged 11.81 g/hr during the burn cycle. However, this rate applies only during the 2-hour burn. Over the full 24-hour heating cycle:
This is the critical point that Paul Tiegs of OMNI Environmental Services made in his 1994/1995 paper: the g/hr emission rate of a masonry heater during its short burn cycle is misleadingly high, because it masks the fact that total emissions per unit of heat delivered are very low.
| Dataset | Typical Range | Notes |
|---|---|---|
| HeatKit "normal" runs | 4–20 g/kg | Varies with moisture, stacking, ignition method |
| 2019 eco-firebox series | 4.23–19.86 g/kg | Cold firebox series; CO higher on cold starts |
| Double Bell | 36–75 g/kg | Higher CO typical of bell heater designs |
| Dataset | Typical Range | Notes |
|---|---|---|
| HeatKit "normal" | 66–79% | Overall (including stack losses) |
| 2019 eco-firebox | 72–78% | Cold firebox series |
| Double Bell | 73–79% | |
| EN 15544 guaranteed | >78% LHV | Austrian calculation method guarantee (UmweltPlus firebox / UZ37 Ecolabel) |
| Best single run | 99.21% combustion efficiency | MHA-1911 (0.83 g/kg PM, 4.23 g/kg CO) |
From the 185+ test dataset, the following variables were identified as significant:
The Condar portable dilution tunnel (Oregon Method 41) is not an EPA-accredited method. MHA has conducted extensive correlation testing:
Bottom line: The Condar method appears reasonably well-correlated with EPA M5G for the PM range typical of masonry heaters (0.5–3 g/kg), but is not EPA-accredited and has been criticised for its constant-flow (rather than proportional-flow) sampling.
16 runs designed to test whether heat exchanger variations on the same firebox produce comparable emissions: - PM range: 0.60–2.30 g/kg (red oak, 62 lbs, ~21–33% moisture) - CO range: 3.09–29.73 g/kg - Efficiency: 66–79% - Key finding: firebox wall starting temperature significantly affects PM - Purpose: support "model line" certification approach — if firebox is tested, heat exchanger variations should not require retesting
Five commercially available masonry heaters tested in homes in western Oregon and Washington by OMNI Environmental Services (Stockton Barnett), using the AWES (Automated Woodstove Emissions Sampler) system — the same methodology used for all published EPA field studies of stoves, fireplaces, and pellet stoves. Each heater was operated by the homeowner in their normal fashion over a week-long test period.
Heaters tested: Biofire, Grundofen (cf. 850° Handwerklicher Grundofen e.V.), Royal Crown 2000, Contraflow, Tulikivi KTU 2100.
| Metric | Value |
|---|---|
| Average PM (5 heaters) | 3.2 g/kg |
| Average daily PM rate | 1.8 g/hr |
| Normalized daily PM rate | 3.2 g/hr (at 1 kg/hr reference) |
| Average CO | 74 g/kg |
| Average net delivered efficiency | 58% |
| Average heat output | 7,425 BTU/hr |
| Average burn rate | 0.68 dry kg/hr |
| Appliance | PM (g/kg) | Emission Reduction Credit |
|---|---|---|
| Conventional woodstoves | 14.9 (AP-42) | — (baseline) |
| Phase II noncatalytic stoves | 7.0 (AP-42) | 64% |
| 5 masonry heaters average | 3.2 | 81% |
| 3 overfire air masonry heaters | ~1.7 | 91% |
| Certified pellet stoves | 1.7 | 91% |
Key finding: the three overfire-air masonry heaters performed equivalently to certified pellet stoves. The two underfire-air heaters pulled the average up.
When EPA developed its NSPS for residential wood heaters in 1988, masonry heaters were deliberately excluded using a weight-based criterion: appliances over 800 kg (1,764 lbs) were exempt. Three factors drove this:
In its 2014 proposed NSPS revision, EPA proposed creating a new Subpart RRRR specifically for masonry heaters, including PM limits, software certification approach, and accommodation for small-volume custom builders.
EPA reversed course, not finalizing Subpart RRRR, stating it was allowing "additional time for the Masonry Heater Association to finish their efforts to develop revised test methods, an emissions calculation program and an alternative dimensioning standard."
EPA justified this by noting the small scale: fewer than ~1,000 masonry heaters per year, less than 10 tons PM2.5 annually, vs 200,000+ wood stoves emitting 11,000+ tons PM2.5.
MHA's negotiating position rested on five pillars:
In November 2018, EPA issued an Advance Notice of Proposed Rulemaking. MHA submitted detailed comments (February 2019) requesting EPA work with them to develop appropriate accommodation, allow certification based on software simulation, and ease compliance costs.
Current status (2026): Masonry heaters remain completely exempt from EPA NSPS certification requirements. Permittable under IRC R1002.2 using ASTM E1602 (custom-built) or UL 1482 (manufactured) — no EPA emissions certification required. In September 2025, EPA agreed to a consent decree to review the NSPS, but the scope for masonry heater inclusion remains unclear.
Paul Tiegs of OMNI Environmental Services wrote the definitive analysis (1994, revised 1995).
EPA chose g/hr because woodstoves have adjustable burn rates, heat production and release happen simultaneously through thin metal walls, and within narrow NSPS constraints, g/hr provides reasonable ranking between models. But g/hr does not estimate field performance. The EPA M28A test is "hot-to-hot" and misses 33% of total emissions (Shelton, 1986).
Tiegs concluded that the most useful unit would be grams of emissions per unit of useful heat produced (g/MJ or gr/BTU). This was rejected in the 1980s because verified efficiency measurement methods were not available. MHA has since developed this capability.
The Cold Climate Housing Research Center tested three appliances under realistic daily use patterns:
| Device | Average g/hr (cycle) | Emissions per 100,000 BTU | Daily fuel (kg) |
|---|---|---|---|
| Masonry heater | 3.5 | 16.4 | 41.3 |
| EPA certified woodstove (2.6 g/hr rated) | 6.3–9.0 | 34.7–48.5 | 14.6–28.5 |
| Multi-fuel stove (pellets) | 5.6 | 12.7 | 93.2 |
The EPA-certified woodstove, rated at 2.6 g/hr, tested at 14.5 g/hr at low (damped) burn rate and 2.3 g/hr at high burn rate. When averaged over realistic daily use, it emitted 2–3× more PM per unit of heat than the masonry heater.
EN 15544 (published 2009) is a European Standard providing a calculation method for dimensioning site-built Austrian-style masonry heaters (Kachelofen). It is approximately 10 pages of original content. The principle: calculate and match pressure losses in the heater against chimney draft.
| Parameter | EN 15544 Guaranteed Level |
|---|---|
| Efficiency | >78% LHV |
| CO | <1,000 mg/MJ |
| NO₂ | <150 mg/MJ |
| OGC | <80 mg/MJ |
| Dust | <60 mg/MJ |
Key design parameters derived from maximum fuel load: - Firebox surface area = max fuel load (kg) × 900 cm² - Rate of burn = max fuel load × 0.78 (results in fixed 77-minute burn) - Air factor (average) = 2.95 (excess air) - Firebox temperature (average) = 700°C - Heat exchanger entrance temp = 550°C - Channel air speed = 1.2–6.0 m/s
The central equation: At rated heat output, all rising forces (buoyancy/stationary pressure) must equal ±5% all resistances (friction + direction changes).
In Austria, the Tile Stove Association (Kachelofenverband) developed proprietary software based on EN 15544. This was submitted to the Combustion and Emissions Certification Laboratory at the Technical University of Vienna, along with test reports from stoves calculated using the software then physically tested. The lab verified software predictions matched measured performance. It is now widely used to certify one-off stoves without lab testing each one.
The problem: EN 15544 has a very limited scope — only Austrian-style channelled heaters. It does not work for: - Contraflow heaters (most common North American type): Parallel downdrafting ducts with high aspect ratio. Buoyancy effects not accounted for. - Double Bell heaters: Low-speed, low-friction systems where EN 15544 predicts flow is too slow for good heat transfer — but they actually transfer heat extremely well.
Since January 2014, MHA has worked with Damien Lehmann, a French engineer who developed an open-source masonry heater simulator based on EN 15544 and EN 13384. Progress: - 2014: Four masonry heaters instrumented with internal pressure and temperature sensors - Initial results: Model predicts well for Austrian heaters; discrepancies for contraflow and bell types - Solution proposed: Add a buoyancy coefficient (Bernoulli's principle) to pressure terms - 2023: Latest update by Lehmann presented at MHA
Current status: The calculator has not been formally submitted to or accepted by EPA. It remains a work in progress.
| Parameter | US EPA NSPS (Stoves) | EU Ecodesign (Closed Stoves) | EU EN 15544 (Site-Built Masonry) | UK BS 3841 (Fuels) | UK Defra Exempt (PD 6434) | UK 2026 Proposed |
|---|---|---|---|---|---|---|
| PM | ≤2.0 g/hr (crib) / ≤2.5 g/hr (cord) | ≤40 mg/m³ @13% O₂ | ≤60 mg/MJ (dust) | ≤5 g/hr smoke | ≤5 g/hr + 0.1g/0.3kW | ≤1 g/hr + 0.1g/0.3kW |
| CO | Reported, no limit | ≤1,500 mg/m³ | ≤1,000 mg/MJ | NOT MEASURED | Recorded, not pass/fail | Not proposed |
| NOx | Not regulated | ≤200 mg/m³ | ≤150 mg/MJ | NOT MEASURED | NOT MEASURED | Not proposed |
| OGC | Not regulated | ≤120 mg/m³ | ≤80 mg/MJ | NOT MEASURED | Not pass/fail | Not proposed |
| Efficiency | Reported, no minimum | ≥65% | ≥78% LHV | NOT REQUIRED | NOT REQUIRED | Not proposed |
| MH status | EXEMPT (>800 kg) | EN 15250 (manufactured) or exempt (site-built via EN 15544) | Calculation-certified | N/A (tests fuel) | Not addressed | Not addressed |
| PM unit | g/hr | mg/m³ @13% O₂ | mg/MJ | g/hr | g/hr | g/hr |
| PM method | EPA M28/M28A dilution tunnel (hot-to-hot) | Heated filter (filterable only) OR dilution tunnel | Calculation-based | Dilution tunnel/ESP (total PM) | Dilution tunnel/ESP (total PM) | TBD |
United States: No federal emission limit. Completely exempt from NSPS. Permitted under building code (IRC R1002.2 / ASTM E1602). EPA justified on practical grounds (testing difficulty) and materiality (<10 tons PM2.5/yr nationally). Some local jurisdictions have imposed restrictions.
European Union: Subject to Ecodesign since January 2022. Tested under EN 15250 (manufactured/kit) or EN 15544 calculation method (site-built Austrian heaters). Must meet: PM 40 mg/m³, CO 1,500 mg/m³, NOx 200 mg/m³, OGC 120 mg/m³, efficiency ≥65%. Significant concern about steady-state test conditions not being representative of real-life operation.
United Kingdom: Masonry heaters are not explicitly mentioned in any UK emission regulation. They fall between the three regimes. A site-built masonry heater in a Smoke Control Area would need to burn authorised fuel (BS 3841) or be Defra-exempt (PD 6434 — designed for stoves). The 2026 consultation does not address masonry heaters.
| Unit | Meaning | Used by | Comparable with |
|---|---|---|---|
| g/hr | Emission rate | US EPA, UK BS 3841, UK Defra | Only other g/hr from same test conditions |
| mg/m³ @13% O₂ | Flue gas concentration | EU Ecodesign | Only other mg/m³ at same reference O₂ |
| g/kg | Fuel mass factor | MHA/Lopez Labs, OMNI, academic research | Only other g/kg from comparable conditions |
| mg/MJ | Energy output factor | EN 15544, some EU research | Only other mg/MJ values |
Conversion between these requires flue gas volumetric flow rate, burn rate, excess air ratio, fuel calorific value, and test duration — none standardised across regimes.
| Method | Captures | Used by |
|---|---|---|
| Dilution tunnel (total PM) | Filterable + condensable PM | US EPA M5G/5H, UK BS 3841, MHA Condar |
| Electrostatic precipitator | Total PM | UK BS 3841 (alternative) |
| Heated filter (filterable only) | Only particles solid at filter temperature — misses condensable fraction | EU Ecodesign primary method |
Critical finding from Ricardo RDE002 (2022): EU Ecodesign heated filter method "is likely to understate appliance PM2.5 emissions" because it misses the condensable fraction. The older BS 3841 dilution tunnel and MHA Condar both capture total PM. The methods are not only in different units — they measure different things.
The same incompatibility identified in the UK dossier (g/hr vs mg/m³) exists globally. US: g/hr. EU: mg/m³ and mg/MJ. UK: g/hr and mg/m³. No jurisdiction has a single, unified emissions unit enabling direct comparison.
Ricardo's finding that EU Ecodesign's heated filter understates PM2.5 is confirmed by MHA experience. The Condar (portable dilution tunnel) and BS 3841 both capture total PM including condensable fraction — mirroring exactly what Ricardo RDE002 (2022) found.
EN 15544 provides: multi-pollutant limits (CO, NOx, OGC, dust), an efficiency guarantee (>78%), and a calculation-based certification path for site-built appliances. The UK fuel authorisation test (BS 3841) measures none of these — no CO, no NOx, no OGC, no efficiency.
EPA exempted masonry heaters because they are "inherently clean-burning" due to high burn rate and inability to smolder. The UK permits authorised fuel in open fireplaces — the exact opposite. The EPA's reasoning for exempting masonry heaters is the same reasoning that should prompt the UK to restrict open fireplace use.
Neither the US nor the EU has anything equivalent. In the US, masonry heaters are exempt but they are enclosed, high-efficiency appliances. In the EU, open fireplaces are subject to Ecodesign with weaker limits and a required warning label: "This product is not suitable for primary heating purposes." The UK alone permits burning authorised fuel in an open fireplace with no emission control.
Neither Defra nor HETAS has proposed a calculation-based certification method for any appliance or fuel. The EN 15544 approach — allowing site-specific verification without expensive lab testing — represents a regulatory model the UK has not explored.
Defra has already granted a Smoke Control Area exemption to a masonry heater: the Jansen Design stove (Kakkelovnsmakeriet / Camina), a Swedish 5-channel system (Exemption No. 554, instruction manual ref CSC-JDS-01, dated 21 October 1991). This proves the technology works and the UK government has already approved it. The precedent needs extending to all EN 15544-compliant site-built heaters.
The Austrian UmweltPlus firebox (UZ37 Ecolabel) represents the pinnacle of aerodynamic firebox engineering. The German 850° Handwerklicher Grundofen e.V. position paper demonstrates that artisanal Grundöfen operating at >850°C eliminate user error. Traditional Kachelofen craftsmanship was inscribed on the UNESCO Intangible Cultural Heritage list in Germany (June 2023).
| Question | Answer |
|---|---|
| What does MHA test data show? | PM ~1 g/kg (normal cordwood), 0.5–4.6 g/kg range. Comparable to pellet stoves. Efficiency 58–79%. CO highly variable (4–75 g/kg). NOx and OGC not routinely measured. |
| What did the OMNI in-home study find? | 5 heaters averaged 3.2 g/kg PM, 81% emission reduction vs conventional stoves. Three overfire-air heaters matched pellet stoves at ~1.7 g/kg. |
| How did MHA get the EPA exemption? | Clean inherent performance, impracticality of lab-testing site-built heaters, inappropriate g/hr metric, and small market size (<10 tons PM2.5/yr nationally). |
| What is the calculator software? | Open-source simulator based on EN 15544, developed by Damien Lehmann (AFPMA) with MHA since 2014. Works for Austrian heaters but NOT for most common North American types. Still in development. |
| Current EPA status? | Masonry heaters remain completely exempt from NSPS. Subpart RRRR was never finalized. No action as of 2026. |
| How do US, EU, UK compare? | All use different units, methods, pollutant suites. EU most comprehensive (5 parameters). US measures PM only. UK has three incompatible regimes plus an open fireplace loophole. Masonry heaters: exempt in US, regulated in EU, invisible in UK. |
| What's the single most important finding? | The unit problem is universal. Every jurisdiction has ended up with incompatible measurement frameworks. The UK's three-regime problem is a symptom of a global failure to agree on how to measure solid fuel emissions. |
| Feature | US (EPA) | EU (Ecodesign + EN 15544) | UK (Current) | UK (2026 Proposed) |
|---|---|---|---|---|
| MH regulation | Exempt (>800 kg) | EN 15250 (manufactured); EN 15544 (site-built) | Not addressed | Not addressed |
| Woodstove PM limit | 2.0–2.5 g/hr | 40 mg/m³ @13% O₂ | 5 g/hr (Defra) | 1 g/hr (proposed) |
| Multi-pollutant limits | PM only | PM, CO, NOx, OGC | PM only | PM only |
| Efficiency requirement | Reported, no min | 65–79% by type | None | Not proposed |
| Open fireplace treatment | Not federally regulated | Member state; warning label | Authorised fuels permitted | Not addressed |
| Calculation-based cert. | Under development | Established (Austria) | Does not exist | Not proposed |
| PM captures condensable? | Yes (dilution tunnel) | Depends on method | Yes (BS 3841) | TBD |