Cookstoves have wide-reaching impacts on human health, air quality, and the climate. We measured emissions from uncontrolled cooking in 41 households in China, Honduras, Uganda, and India using a portable sampler. Test sites were chosen to cover a range of stove types (traditional and “improved”), fuels (wood, charcoal and coal), and cooking practices. We report test-integrated fuel-based emission factors (EFs) of fine particulate matter (PM2.5) mass, organic carbon (OC), elemental carbon (EC), as well as real time EFs of carbon monoxide (CO), black carbon (BC), total particle number, and particle size distributions. There was substantial house-to-house variability in emissions; the distribution of EFs were also highly positively skewed by several “superemitter” stoves in China (those with PM2.5 EFs 5–20 times greater than the median value). The highest PM2.5 mass emission factors were measured in China (median:10.3 g/kg-fuel), and the lowest in Uganda (median: 1.7 g/kg-fuel). The median PM2.5 mass EFs in wood-burning stoves in Honduras and India were similar: 3.7 g/kg-fuel and 4.1 g/kg-fuel, respectively. However, Indian stoves had higher EC EFs then Honduran stoves, demonstrating that emissions depend on more that fuel type; regional differences, such as cooking styles and stove design, may influence aerosol properties as well. Coal and charcoal stoves had higher OC:EC than wood stoves. The differences between the CO, PM2.5, and OC:EC ratios of “improved” and traditional stoves in India and Honduras were not statistically significant. To the best of our knowledge, we report the first cookstove source size distributions measurements from uncontrolled in-home cooking. These distributions varied between countries, which will influence local radiative effects. Particle size distributions from stoves tested in China, Honduras, and India were unimodal in the size range measured, with geometric mean diameters (GMDs) of 66 nm, 48 nm, and 76 nm, respectively. The median GMD of particles emitted from Ugandan charcoal stoves was 39, and when all tests are averaged, the resulting distribution appears tri-modal, with modes near 15, 30, and 100 nm. Real-time emissions data reveal high BC and particle number emissions during startup and fuel additions, which can be seen in the positively skewed distributions. Emissions of BC were most skewed, indicating that they were highly event-driven, followed by total particle number. CO emissions were more evenly spread across cooking events.
All Science Journal Classification (ASJC) codes
- Environmental Science(all)
- Atmospheric Science