UC Berkeley

Air pollution is one of the world’s greatest environmental health risks, responsible for over 7 million premature deaths annually. Around half of these premature mortalities are linked to biomass cooking fires that release harmful air pollutants into people’s homes, such as particulate matter (PM). Although nearly 3 billion people worldwide depend on biomass cooking fuels, relatively little scientific research exists on mitigating the smoke they generate. One promising approach for reducing emissions is the injection of secondary air into the biomass cookstove’s combustion chamber. However, cold secondary air can also quench the combustion process when improperly injected, and so many secondary air injection cookstove designs do not actually reduce harmful smoke emissions relative to a traditional three stone fire (TSF).

Since wood is a common cooking fuel throughout the world, this dissertation presents an experimental
 wood-burning cookstove platform, dubbed the ‘Modular (MOD) stove’, to identify and optimize secondary air 
injection parameters that reduce the emission of harmful pollutants. The MOD stove enables systematic, repeatable experiments in which various secondary air injection design features, such as flow rate and location, can be quickly and easily adjusted. Over 130 experimental trials were conducted, demonstrating that wood combustion is highly sensitive to small changes in the secondary air injection parameters. Using a systematic experimental approach, an optimal design configuration was identified that reduces mass emissions of PM2.5, carbon monoxide (CO), and black carbon (BC) by ~90% relative to a traditional TSF, while also improving thermal efficiency.

Using an updated version of the MOD stove, an additional 111 performance tests were conducted to quantify the practical design requirements (e.g., secondary air pressure and temperature) to achieve ≥ 90% mass emission reductions relative to a TSF. Using this experimental data, I demonstrate that low-cost (<$10) fans and blowers are currently available to drive the secondary flow, and this hardware can be independently powered using an inexpensive thermoelectric generator mounted nearby. Furthermore, size-resolved PM measurements demonstrate that secondary air injection effectively inhibits particle growth, but the total number of particles generated remains relatively unaffected. I investigate the potential impacts for human health and explore methods to mitigate the PM formation mechanisms that persist. As a whole, the MOD stove platform demonstrates that secondary air injection is a practical, effective, and potentially economical method for meaningfully reducing smoke emissions from biomass cookstoves. However, designs should be experimentally validated and optimized, and further research is needed to eliminate the persistent formation of ultrafine particles that are particularly harmful to human health.

Ambient air pollution is also widespread, and linked to significant adverse health outcomes. Over 90% of the world’s population lives in areas where ambient air pollution concentrations exceed World Health Organization recommendations, resulting in ~4 million premature deaths annually. The health impacts of ambient air pollution are particularly acute in urban settings. In 2010, premature deaths due to ambient air pollution were about 50% more common in urban than in rural environments, and this could increase to nearly 90% by 2050. Although ground-based air quality measurements are needed to address this growing health crisis, traditional regulatory monitoring networks do not provide sufficient spatial coverage and resolution to adequately assess air pollution exposures in urban environments. Distributed networks of low-cost air quality sensors are emerging to fill this gap.

Black carbon (BC) is an important component of PM pollution, strongly linked to adverse human health outcomes and climate change, but low-cost sensors for monitoring this critical pollutant are lacking. This dissertation presents the Aerosol Black Carbon Detector (ABCD), specifically designed for distributed air quality monitoring networks. As such, the ABCD integrates a compact weatherproof enclosure, solar-powered rechargeable battery, and cellular communication to enable long-term, outdoor deployments. Most importantly, the ABCD incorporates a number of novel design features to provide uniquely accurate BC concentration measurements in tough operating environments that debilitate existing commercial instruments. Over 100 ABCDs were operated outdoors, and their measurement performance was comparable to that of a commercial BC instrument collocated inside a regulatory monitoring station.

The validated fleet of ABCDs was deployed to 100 sampling sites in West Oakland, California – a neighborhood disproportionately affected by air pollution associated with the nearby Port of Oakland and surrounding highways. Over 100 days, the wireless sensor network successfully collected 84.0% of the 240,000 hourly BC concentration measurements desired (100 sampling sites × 2,400 hours). The widespread failure of miniature vacuum pumps was responsible for most missing measurements. The resulting BC concentration maps demonstrate that concentrations vary sharply over short distances (~100 m) and timespans (~1 hour), and generally depend on surrounding land use, traffic patterns, and location relative to prevailing winds. BC concentrations at each sampling site are highly repeatable over the diurnal and weekly cycles, and periodic trends are analyzed throughout the community. Using these trends as a reference, unusually polluted locations are detected, and likely emissions sources nearby are identified. In this way, the 100x100 Network demonstrates the value of low-cost sensor networks to accurately characterize urban air pollution distributions, and provide regulatory agencies, governments, and community stakeholders with actionable insights to mitigate the sources.

Air pollution is a pervasive and persistent health threat that can only be tackled if we work to both mitigate and monitor emission sources. As such, the MOD stove studies presented here are targeted towards abating the world’s deadliest polluters. However, even the best of emissions reduction efforts are left blind without accurate measurements of the resulting air quality. The ABCD meets this need, providing accurate BC monitoring capabilities in a uniquely practical and economical package. Together, these complementary technologies will help underpin more comprehensive, data-driven efforts to quantifiably reduce harmful air pollution exposure throughout the world, and ultimately prevent millions of premature deaths.