Now a team from the US and Greece has come up with a simplified air-quality model that works fast enough to analyse power-station emissions hour by hour. If the model shows that pollution is going to hit a population centre, the power-station operator could shut down the plant until atmospheric conditions have changed. The system is not yet in use, but a study of the US state of Georgia in January and July in 2004—2011 found it would have avoided human health impacts costing $175.9 million, for an extra electricity generation cost of $83.6 million, largely due to switching to more expensive natural gas.

"This approach could be used to reduce the impacts of air-pollutant emissions, without needing to wait to install air-pollution control equipment," Valerie Thomas of Georgia Institute of Technology told environmentalresearchweb. "This is the first time that the feasibility of this approach has been demonstrated; the electricity industry is just now learning that this is an option."

Thomas and colleagues created the Air Pollutant Optimization Model, a reduced form of the Community Multiscale Air Quality model that provides fast predictions of air pollutant formation over a 12 km × 12 km grid. They looked at two ways of running Georgia’s power stations — one that minimized production costs only, and one that minimized both production costs and monetized health impacts.

"Pricing the externality associated with electricity generation was impossible before because we were not able to run air-quality models fast enough," said Thomas. "Now, with this new approach, the operation of the electricity system can take into consideration the externalities associated with air quality and correctly price each generator. This was an impossible task before our study because the electricity system operates in real time, while air-quality models run over days or weeks."

Potential health impacts were dominated by PM2.5, the team found. These can increase mortality and asthma rates and boost nonfatal heart attacks, emergency room visits and hospital visits. Around one-quarter of the ambient PM2.5 in the US is sulphate-based.

Sulphate emissions tend to form more PM2.5 in daylight and during the summer, when temperatures are higher. Looking at Plant Bowen, a sulphur dioxide-emitting coal plant north-west of Atlanta, showed that in July 2004 including monetized health impacts would have led to replacement of coal with a natural gas plant on 20 of the month’s 31 days. Plant Bowen averages around $17 of monetized health impact per MWh of electricity generation in July.

"The real advantage is in being able to accurately predict air quality on an hourly basis," said Thomas. "This is important because there are substantial differences from hour to hour for air-quality outcomes, as well as for demand. We show in the paper that by using a yearly or daily average — as it is done nowadays — some benefits are left on the table."

Now the team would like to carry out a larger US study. "We would like to include emissions from transportation, evaluate impacts on ecosystems — especially where population density may be low, but [there are] sensitive ecosystems, crops and other receptors sensitive to pollution — and we would like to apply this approach to countries or regions with high, uncontrolled air-quality challenges," said Thomas. "With these extensions, the possibilities to reduce health impacts will increase. We can test the impacts of the electricity system and natural gas system congestion on air quality, better quantify the benefits and costs of renewables and other clean energy alternatives like electric cars. We can quantify the air-quality impacts of climate change policies."

Thomas and colleagues reported their study in PNAS.

Animation: On the left, a power plant in Georgia from July 4 to 11 in 2007, and on the right, how the power plant could have been operated to reduce health impacts. Image credit Georgia Institute of Technology.

Related links

Related stories