No Perfect Heater: New Framework Puts Health at the Center of Home Heating Choices

When winter arrives, most of us simply want to stay warm. But the way we heat our homes may be quietly shaping our health in ways that energy statistics alone cannot capture. A new study published in the journal Clean Technologies and Environmental Policy by Maria Antuna-Casal and Juan Carlos Ríos-Fernández of the University of Oviedo in Spain has developed a health-centered, screening-level multicriteria framework for evaluating residential and small-building heating technologies. Its central and somewhat provocative conclusion is that no heating technology is universally preferable. The best choice depends on a web of interacting factors: the local electricity mix, seasonal performance of equipment, combustion exposure, delivery design, fuel availability, and the values of the people making the decision.

The framework addresses a long-standing methodological blind spot by explicitly separating the thermal source of heat from the mode of heat delivery. Boilers, biomass appliances, direct electric resistance heaters, air-source heat pumps, ground-source heat pumps, and solar thermal systems all determine source-related energy use, emissions, operating costs, and indirect health burdens. By contrast, forced-air systems, fan coils, hydronic radiators, radiant floors, and radiant ceilings determine much of the local comfort and direct occupant exposure profile, influencing air velocity, mean radiant temperature, pollutant dispersion, ventilation interaction, acoustic disturbance, and control stability. This distinction matters because the same heat source can produce very different comfort outcomes depending on how the warmth reaches the occupant, and the same delivery mode can have very different environmental impacts depending on what powers it.

To compare heating configurations systematically, the researchers implemented a multicriteria decision analysis based on PROMETHEE II, an outranking method that compares alternatives pairwise for every criterion using preference functions and aggregates these into net outranking flows. The evaluation criteria included greenhouse-gas emissions, indirect health burden, direct indoor exposure, levelized cost of heat, regional availability, and regulatory compatibility. Criteria weights were obtained through a Delphi-based consultation with ten interdisciplinary experts, including thermal engineers, environmental-health specialists, and energy-economics researchers. Crucially, the authors treated these weights as an expert-informed base case rather than universal preferences, recalculating rankings under alternative weight sets that emphasized climate impact, occupant health, local air quality, cost, and energy security.

The electricity dimension of the analysis is particularly revealing. For electric heating technologies, operational emissions per unit of useful heat were calculated by dividing the regional electricity emission factor by the seasonal coefficient of performance. In an illustrative Spain-based application, the 2024 annual-average electricity factor was calculated at 0.103 kg CO2-eq per kilowatt-hour, based on national system data showing that renewables supplied 56.8 percent of generation. A 20-year policy-scenario average of 0.055 kg CO2-eq per kilowatt-hour was derived by scaling the 2024 factor according to Spain’s target of 81 percent renewable electricity by 2030. A winter-peak marginal sensitivity factor of 0.370 kg CO2-eq per kilowatt-hour, corresponding to combined-cycle natural gas generation, was also applied, because electric heating demand during cold snaps may coincide with fossil marginal generation rather than the annual-average mix.

The results show how dramatically these assumptions change the picture. Direct electric resistance systems convert electricity to heat with a coefficient of performance of roughly one, so their emissions per unit of useful heat simply mirror the grid. Heat pumps, with assumed seasonal coefficients of performance of 3.0 for air-to-air units, 3.5 for air-to-water systems, and 4.5 for ground-source machines, dilute those emissions considerably. Natural gas boilers were calculated using an emission factor of 0.182 kg CO2-eq per kilowatt-hour of fuel and a seasonal efficiency of 92 percent, while fuel-oil boilers used 0.267 kg CO2-eq per kilowatt-hour at 88 percent efficiency. Biomass and solar thermal systems were deliberately not assigned single universal values, because their performance depends on appliance type, fuel quality, biogenic-carbon accounting, pollutant emissions, solar fraction, storage losses, auxiliary electricity, and backup sources.

The health evidence underpinning the framework is sobering. A recent epidemiological study in China with 14,068 participants found significant associations between domestic solid-fuel use and elevated risks of hypertension, coronary heart disease, and diabetes, with odds ratios of 1.70, 2.58, and 2.48 respectively. The pathophysiological mechanisms include oxidative stress, systemic inflammation with cytokine release, and alterations in blood coagulation. However, the authors caution that such studies aggregate heterogeneous appliances and fuels, often without reporting the proportion of manually fed log stoves, certified pellet stoves, fuel moisture, or flue condition. They therefore support concern about solid-fuel exposure but do not justify assigning a single risk score to all biomass heating technologies. The framework accordingly distinguishes manually fed log stoves, pellet stoves, chip boilers, and externally located biomass boilers, penalizing log stoves for higher direct indoor exposure uncertainty.

On the economic side, the study reformulated the levelized cost of heat as a parameterized method rather than a universal cost interval. Using a 20-year analysis period, a 3 percent real discount rate, and an annual useful heating demand of 15,000 kilowatt-hours, the illustrative Spanish calculation employed energy prices of 0.2408 euros per kilowatt-hour for electricity, 0.0901 euros for natural gas, 0.0871 euros for fuel oil, and 0.0711 euros for pellets. The analysis highlights a familiar economic tension: direct electric radiant heating has low installation complexity but high operating cost, whereas ground-source heat pumps carry substantially higher initial costs due to drilling and ground-loop installation, yet their high seasonal performance can yield lower levelized costs when annual heat demand is sufficient. In low-demand dwellings or sites with prohibitive drilling costs, simpler alternatives may remain economically preferable.

The framework also draws on a growing body of epidemiological research linking thermal conditions to cardiovascular outcomes. A case-crossover study in Germany using data from 380 districts between 2000 and 2016 found that elevated temperatures significantly increased the risk of total cardiovascular mortality, with a relative risk of 1.24, and respiratory mortality, with a relative risk of 1.34. A meta-analysis of more than 400 laboratory studies documented median elevations in body temperature of 0.8 degrees Celsius and heart rate of 27 beats per minute during heat exposure. Experiments on older adults exposed to indoor temperatures of 31 and 36 degrees Celsius for eight hours showed core body temperature increases of 0.7 and 0.9 degrees respectively, linked to cognitive decline, reduced concentration, and mood disturbances. Vulnerability is unevenly distributed: older adults with altered thermoregulation, children with high metabolic rates, and low-income communities with substandard insulation all face disproportionate risks.

Perhaps the most practically useful insight concerns radiant and hydronic delivery. Radiant floors and ceilings can provide sensible heating with low air velocities, potentially reducing draft and limiting airflow-driven particle resuspension compared with poorly designed forced-air systems. Yet the authors are careful to note that particle resuspension depends on air velocity, airflow pattern, deposited dust load, surface type, and occupant activity, and that radiant delivery alone does not guarantee indoor air quality. Adequate ventilation and filtration remain necessary. Similarly, an air-to-air split unit may provide acceptable comfort when properly sized, modulated, and installed, but may generate draft, acoustic disturbance, short cycling, or uneven temperatures in poorly insulated buildings. Comfort, in other words, is a property of the whole system and building, not of the heat source alone.

The authors are explicit about the limits of their work. The framework is intended for scenario-specific screening, not as a substitute for full building-level life cycle assessment, dynamic simulation, project-specific cost analysis, or local health-impact assessment. DALY-based health damage was not calculated in this screening-level assessment and is identified as a recommended endpoint for future full life cycle impact assessment studies where regional characterization factors and inventory data are available. Low operational carbon intensity should not be read as zero life cycle health impact, since electricity systems carry upstream burdens including mining, fuel processing, waste management, and grid-integration impacts. Cooling-season operation was deliberately excluded, as it would require additional assumptions on cooling loads, condensation risk, and urban heat-island effects. The study’s strongest value, the authors argue, is preliminary screening: it helps separate source-related emissions and costs from delivery-related comfort and exposure, highlights when electric systems depend on the grid, identifies when biomass risk depends on appliance type and operation, and clarifies the trade-offs that decision-makers must confront. Future work should validate the framework using monitored buildings, measured indoor air quality, occupant comfort surveys, and region-specific data, bringing the science of staying warm closer to the science of staying well.

Subject of Research: Health-centered multicriteria evaluation of residential heating technologies

Article Title: Health-centered multicriteria evaluation of heating systems for sustainable indoor comfort

Article References: Antuna-Casal, M., & Ríos-Fernández, J. C. (2026). Health-centered multicriteria evaluation of heating systems for sustainable indoor comfort. Clean Technologies and Environmental Policy, 28(10), Article 270. https://doi.org/10.1007/s10098-026-03631-4

Image Credits: AI Generated

DOI: 10.1007/s10098-026-03631-4

Keywords: heating systems, heat pumps, indoor air quality, multicriteria decision analysis, PROMETHEE II, life cycle assessment, biomass, thermal comfort, electricity mix, public health, levelized cost of heat, sustainability