Research Background and Methodology
Catalogue
Emission inventory
Health impact assessment
References
Emission inventory
The first step in assessing the health impacts from coal-fired power plant emissions is obtaining information on how much is emitted and where the emissions take place. For the purposes of this project, a database of over 2000 coal-fired power plants. As Chinese government and companies, unlike their counterparts in e.g. Europe and the U.S. do not report plant-level emission data, the emission data for the power plants had to be estimated based on national total emissions, reported fleet-level emission rates of large utilities, available plant-specific information, and national regulation on power plant emissions. The resulting estimates are robust on the national and company level, to the extent that reported emission data is accurate, but there are additional uncertainties involved in estimating the emissions from individual power plants.
Emissions from operating power plants were estimated for year 2011. The operating data of the power plants is obtained from the China Electricity Council (CEC) yearbook 2012. The publication has data on installed capacity, operating hours, and thermal efficiency. This data is also used to establish size-dependent average values for those power plants for which data is missing.
Capacity, more than (MW) / Thermal efficiency (LHV net) / Operating hours (h/a)0 / 28.4% / 3761
20 / 30.2% / 3793
50 / 31.4% / 4302
100 / 33.3% / 5055
300 / 35.2% / 4644
500 / 36.8% / 5322
1,000 / 38.2% / 5537
2,000 / 39.4% / 5928
Table 1.Average operating parameters used for operating power plants lacking data.
Plant type / Thermal efficiency (LHV net)Subcritical / 39.0%
Supercritical / 42.0%
Ultrasupercritical / 44.0%
Under construction, steam condition unknown / 41.4%
Planned, steam condition unknown / 41.8%
Table 2.Thermal efficiencies assumed and estimated for new power plants in the Platts database and the WRI Global Coal Risk Assessment report.
The locations of the power plants were mapped by Greenpeace, up to district or county level, and when possible, exact coordinates where used. Information on the ownership of the power plants was obtained from Platts World Electric Power Plants database.
Information on pollution controls installed at the power plants is from Ministry of Environmental Protection, which maintains a list of all power plants with FGD and de-NOx equipment installed. This data also has the year of operation for the power plants, which helps establish the emission limit values applying to each unit at the power plant. However, the power plant listings in the CEC and MEP data do not completely match each other, and average penetration rates for each province and power plant size class were applied to those power plants that could not be matched between the two databases.
Power plant efficiency was based on steam conditions (subcritical, supercritical or ultrasupercritical) reported in the WEPP database. All power plants commissioned after 2011, and those still in the pipeline, were assumed to have both FGD and de-NOx equipment installed, and to meet the new 2011 emission standards. This is a conservative assumption, given that the existing power plant fleet still does not meet the old 2003 standard.
Data on coal quality, namely flue gas volume and mercury content, comes from USGS World Coal Quality Inventory. First, average values of all thermal coal samples were calculated from the database for each province. Second, the average values for traded coal were estimated by taking average of values for each province weighted by their coal exports. Lastly, the average values for coal burned in each province were estimated by calculating the average of the values for the province’s domestic coal and traded coal weighted by the percent of coal that the province imports. Flue gas volume per energy input (Nm3/GJ) was calculated by first converting the energy content given in the database from Higher Heating Value (HHV) to Lower Heating Value (LHV), using an empirical formula provided by World Coal Institute(2007):
LHV = HHV - 0.212H- 0.0245M- 0.0008O,
whereLHV and HHV are given in MJ/kg; M is percent moisture, H is percent hydrogen and O is percent oxygen (from ultimate analysis on net as received basis). Flue gas volume per kg of fuel is calculated on the basis of the empirical formula in European Standard EN 12952-12.
Province / Flue gas volume[1] (Nm3/GJ) / Mercury content (mg/GJ)Anhui / 344.9 / 7.8
Beijing / 356.6 / 7.4
Chongqing / 349.5 / 4.9
Fujian / 359.4 / 3.8
Gansu / 348.5 / 2.2
Guangdong / 354.5 / 3.5
Guangxi / 354.4 / 5.1
Guizhou / 347.8 / 9.1
Hebei / 350.9 / 4.5
Heilongjiang / 345.1 / 2.9
Henan / 347.6 / 7.9
Hubei / 354.5 / 3.9
Hunan / 353.2 / 5.2
Inner Mongolia / 345.8 / 10.5
Jiangsu / 353.6 / 4.6
Jiangxi / 352.1 / 7.9
Jilin / 346.3 / 4.2
Liaoning / 349.9 / 6.1
Ningxia / 348.2 / 11.8
Qinghai / 348.6 / 2.8
Shaanxi / 342.4 / 7.5
Shandong / 350.3 / 4.4
Shanxi / 347.3 / 6.4
Sichuan / 348.4 / 4.0
Xinjiang / 347.1 / 1.4
Yunnan / 345.6 / 5.7
Table 3.Average properties estimated for the coal burned in each province.
Based on these data, air pollution emissions for each power plant were first calculated assuming that all power plants meet the national emission standards applying to them. After this, the emission rates were adjusted so that the total modeled emissions from all power plants and from each company's power plants match the reported total. Total emissions of acid gases and particulate matter from the power sector were taken from China Environment Statistical Yearbook 2012 (National Bureau of Statistics 2013). Information on the emissions of large power companies is compiled from the companies' CSR reports. It was also ensured that the total power plant emissions make up a reasonable share of the reported total emissions of each province.
Power plant commissioning date / Pollutant stack emission limits (mg/Nm3)SO2 / NOx / TSP
2004 or later / 400 / 450 / 50
before 2004 / 400 / 650 / 50
before 1997 / 1200 / 1100 / 200
2012 or later / 100 / 100 / 30
new power plants in key regions / 50 / 100 / 20
Table 4.Stack emission concentration limits applying to operating power plants in 2011, and to new power plants.[2]
Power plant commissioning date / mg/Nm3SO2 / TSP
2004 or later / 1200 / 100
before 2004 / 1200 / 100
before 1997 / 1200 / 200
Table 5.Exceptions to general emission limits appying to power plants burning domestic low-sulphur coal in Chongqing, Sichuan, Guizhou, Yunnan, Tibet, Shaanxi, Gansu, Qinghai, Ningxia, Xinjiang, Guangxi and Inner Mongolia.
TSP emissions were converted to primary PM10 and PM2.5 emissions using the emission factors in U.S. EPA AP-42:
TSP to PM10 / 0.675TSP to PM2.5 / 0.300
Table 6.Ratios between different size ranges of particulate matter.
Company / Emission rate (g/kWh)SO2 / NOx / TSP
China Power Investment / 2.36 / 3.23 / 0.3
Huadian / 2.4 / 3.0 / 0.3
Datang International Power / 0.38 / 1.33 / 0.12
Datang total / 1.92 / 3.17 / 0.29
Huaneng Power International / 0.57 / 1.55 / N/R
Huaneng total / N/R / N/R / N/R
Guodian / 2.14 / N/R / N/R
China Resources Power / 0.56 / 1.35 / 0.38
Guangdong Yudean / 0.44 / 1.36 / 0.07
Shenhua / 0.21 / 0.87 / 0.1
Table 7.Air pollution emission rates as reported in key power companies' CSR reports for 2011 (N/R=not reported).
Mercury emissions were estimated based on the average mercury content of coal as shown above, and removal rates associated with different pollution controls according to Wu et al (2009).Estimated mercury emissions were 20% lower than in the earlier estimatefor 2005 by Streets et al (2008), which is in line with the increased penetration of FGD equipment and increased coal consumption in the power sector.
ESP / 29.4%ESP+FGD / 69.0%
coal washing / 30.0%
Table 8. Mercury removal rates of different technologies according to Wu et al. (2009).
Stack parameters (stack height and diameter, flue gas temperature and velocity) are required for estimating how high the flue gases rise initially, which influences their dispersion. Actual stack parameters were compiled for a few power plants in Beijing and Shanghai, but for the vast majority, this information was not readily available. Zhou et al (2006) argue that Chinese power plants are built to very similar engineering standards and most power plants will conform with the guideline values. Furthermore, their results show that the total health impacts are not particularly sensitive to varying the assumed stack parameters within feasible range. Recommended values for the stack parameters were taken from Lan et al (2011), except for flue gas temperature, typical European values from Pregger & Friedrich (2009) were used.
Capacity, up to (MWe) / Stack height (m) / Flue gas temperature (°C) / Exit velocity (m/s) / Diameter (m)Existing power plant / New power plant
25 / 80 / 80 / 140 / 14 / 4
50 / 100 / 100 / 140 / 23 / 4
200 / 120 / 150 / 140 / 20 / 4
300 / 150 / 180 / 140 / 20 / 4
800 / 180 / 240 / 110 / 30 / 7
1,200 / 210 / 240 / 100 / 30 / 7
8,000 / 240 / 240 / 100 / 30 / 7
Table 9.Stack parameters used for modeling when plant-specific data is not available (most modeled sources).
Selection of the directly modeled sources was done by first dividing the modeling domain into a 0.5x0.5° grid, and selecting the largest source within each grid cell that contained at least 1200MW of coal-fired capacity. Additional sources were selected to maximize the share of total emission inventory that is modeled directly, to maximize spatial coverage and to maximize coverage of key regions (Beijing-Tianjin-Hebei, Yangtze River and Pearl River deltas). The directly modeled sources cover 50% of coal-fired capacity, 43% of estimated SO2 emissions in 2011, and 41% , 40% and 48% of NOx, TSP and mercury emissions, respectively.
Health impact assessment
The health impacts resulting from the exposure to PM2.5 were estimated using concentration-response functions adapted from the WHO Global Burden of Disease 2010 project (Lim et al 2012). The study is the most up-to-date and authoritative look into preliminary deaths caused by PM2.5 in China and globally, and developed a new risk model with emphasis on applicability at high average concentrations. The risk functions in the model level off at high concentrations, taking into account the findings showing that risk for the same concentration increase is higher at low concentrations. Total mortality is evaluated as a sum of four cause-specific mortality risks: stroke, lung cancer, Ischemic Heart Disease (IHD), and Chronic Obstructive Pulmonary Disease (COPD). These four causes are responsible for 45% of total deaths in China. The cause-specific approach provides better transferability from one country to another than earlier approaches that used all-cause mortality as the indicator, and provides a breakdown of the causes of the preliminary deaths attributed to PM2.5 from coal-fired power plants.
If the concave risk functions from Global Burden of Disease 2010 were used directly to attribute impacts on different sectors, the sum total of impacts attributed to all sectors would be smaller than the actual total impacts. For this reason, based on a recommendation from the report authors (Burnett&Cohen 2013), average impacts for a 10µg/m3 increase over the observed concentration range were used for attribution. The average risk ratio was calculated for each mortality risk as
,
where is the ratio of mortality risk at concentration to the risk at a counterfactual no-harm concentration, and is the population-weighted average PM2.5 concentration, taken to be 60 µg/m3 (the average concentration estimated for China for Global Burden of Disease 2010 by Brauer et al (2012) was 55 µg/m3). The summation is started from 15 µg/m3, because this represents the no-harm concentration in the risk model (5 µg/m3) plus the concentration increase for which is calculated (10 µg/m3).
Non-fatal health impacts were evaluated by using concentration-response functions recommended by Kan et al (2005) for health impact assessment in China, when available. The response functions were applied conservatively, using the factor for PM10 health effects for exposure to PM2.5. The Kan et al functions were complemented with functions for infant mortality, lost working days and sickness days from literature, following WHO recommendations. Recent epidemiological evidence on the link between PM2.5 and risk of low birth weight in babies was used from a new nine-country study. While overall mortality is estimated on the basis of all-cause mortality, cause-specific factors are used to complement the analysis and provide a breakdown of causes of death.
Application of these response functions requires data on the age structure of the Chinese population, and on baseline incidence of the different health conditions. These were obtained from official statistics, with the exception that World Bank data on low birth weight, and data from academic studies done in China on asthma, were used.
Health impact / Concentration-response functionPollutant / Age group / Increase per 10µg/m3 / Reference
Stroke mortality / PM2.5 / 30- / 12.2% (3.2%-14.8%) / Lim et al 2012; Burnett&Cohen 2013
Lung cancer mortality / PM2.5 / 30- / 5.6% (1.7%-7.4%)
COPD mortality / PM2.5 / 30- / 4.1% (1.9%-5.7%)
Ischemic heart disease mortality / PM2.5 / 30- / 5.5% (3.9%-9.0%)
Infant mortality / PM10 / 1-12 months / 4% (2%–7%) / Woodruff et al 1997 (in Hurley et al 2005)
Low birth weight / PM2.5 / newborns / 10% (3%–18%) / Dadvand et al 2013
Asthma, children / PM10 / 0-15 / 6.95% / Kan et al 2005
Asthma, adults / PM10 / 16- / 0.4% (0.0%–0.8%)
Chronic Bronchitis / PM10 / all / 4.6% (1.5%–7.7%)
Respiratory Hospital Admission / PM10 / all / 1.3% (0.1%–2.5%)
Cardiovascular Hospital Admission / PM10 / all / 0.95% (0.6%–1.3%)
Outpatient Visits (internal medicine) / PM10 / all / 0.34% (0.19%–0.49%)
Outpatient Visits (pediatrics) / PM10 / all / 0.39% (0.14%–0.64%)
Sick leave days / PM2.5 / 15-64 / 4.6% (3.9%–5.3%) / Ostro 1987 (in Hurley et al 2005)
Restricted activity days / PM2.5 / 18-64 / 4.8% (4.2%–5.3%)
Table 10.Concentration-response relationships used to estimate health impacts of particulate matter exposure.
Health impact / Baseline incidence or prevalence / Unit / ReferenceStroke mortality / 0.14% / deaths per year / Ministry of Health 2011
Lung cancer mortality / 0.04% / deaths per year / Ministry of Health 2011
COPD mortality / 0.06% / deaths per year / Ministry of Health 2011
Ischemic heart disease mortality / 0.08% / deaths per year / Ministry of Health 2011
Infant mortality / 1.21% / deaths per year / National Bureau of Statistics 2012
Low birth weight / 2.34% / cases per year / World Bank 2012
Asthma, children / 1.97% / cases / Chen 2003
Asthma, adults / 1.42% / cases / To et al 2012
Chronic Bronchitis / 0.69% / cases / Ministry of Health 2011
Respiratory Hospital Admission / 1.02% / cases per year / Ministry of Health 2011
Cardiovascular Hospital Admission / 1.37% / cases per year / Ministry of Health 2011
Outpatient Visits (internal medicine) / 31% / cases per year / Ministry of Health 2011
Outpatient Visits (pediatrics) / 13% / cases per year / Ministry of Health 2011
Sick leave days / 2.34 / workdays per year / Ministry of Health 2011
Restricted activity days / 39.96 / workdays per year / Ministry of Health 2011
Table 11.Baseline incidence of health conditions included in health impact assessment. For asthma and chronic bronchitis, the epidemiological relationship applies to prevalence, not annual incidence of new cases.
References
Brauer et al 2012: Exposure Assessment for Estimation of the Global Burden of Disease Attributable to Outdoor Air Pollution. Environ. Sci. Technol. 46(2):652–660.
Chen YZ 2003: 中国城区儿童哮喘患病率调查. 中华儿科杂志2003年2月第41卷第2期.[ A nationwide survey in China on prevalence of asthma in urban children. Chinese Journal of Pediatrics 2003(41)2.]
Burnett R & Cohen AJ, June 11, 2013. Personal communication.
Dadvand P et al 2013: Maternal Exposure to Particulate Air Pollution and Term Birth Weight: A Multi-Country Evaluation of Effect and Heterogeneity. Environ Health Perspect 121:367–373.
Hurley et al 2005: Methodology for the Cost-Benefit analysis for CAFE: Volume 2: Health Impact Assessment. AEA Technology Environment.
Kan HD, Chen BH, Chen CH, Wang BY & Fu QY 2005: Establishment of exposure-response functions of air particulate matter and adverse health outcomes in China and worldwide. Biomed Environ Sci. 2005 Jun;18(3):159-63.
Krewski et al 2009: Extended Follow-Up and Spatial Analysis of the American Cancer Society Study Linking Particulate Air Pollution and Mortality. Research Report 140. Health Effects Institute, Boston, Massachusetts.
Lan T, Zhang XY, Wu Z, Sun Y & Zhang C 2011: 锅炉烟囱高度设置合理性论证的实例分析. 环境科学与管理36卷第2 期. [Examples of the Boiler Chimney Height Rational Argument.Environmental Science and Management 36(2).
Lim SS et al 2012: A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990—2010: a systematic analysis for the Global Burden of Disease Study 2010. The Lancet 380:2224-2260.
Ministry of Health 2011: 2011中国卫生统计年鉴[”2011 China Health Statistics Yearbook”].
National Bureau of Statistics 2012: China Statistical Yearbook 2012.
National Bureau of Statistics 2013: China Environment Statistics Yearbook 2012. China Statistics Press.
Pope et al 2002: Lung cancer, cardiopulmonary mortality, and long-term exposure to fine particulate air pollution. The Journal of the Americal Medical Association.
Pregger T & Friedrich R 2009: Effective pollutant emission heights for atmospheric transport modelling based on real-world information. Environmental Pollution 157:552–560.
Streets DG, Hao JM, Wang SX, Wu Y 2008: Mercury Emissions from Coal Combustion in China.Presentation.International Conference of the UNEP Global Partnership on Atmospheric Mercury Transport and Fate Research, Rome, Italy, April 7-11, 2008.
To T et al 2012: Global asthma prevalence in adults: findings from the cross-sectional world health survey. BMC Public Health 2012, 12:204.
World Bank 2012: Low-birthweight babies (% of births). Data Catalog.
World Coal Institute 2007: Coal Conversion Facts.
Wu Y, Streets DG, Wang SX & Hao JM 2009: Uncertainties in estimating mercury emissions from coal-fired power plants in China. Atmos. Chem. Phys. Discuss., 9, 23565–23588, 2009.
Yang&Cui 2012: Global Coal Risk Assessment: Data Analysis and Market Research. WRI Working Paper. World Resources Institute, Washington DC.
Zhou Y, Levy JI, Evans JS & Hammitt JK 2006: The influence of geographic location on population exposure to emissions from power plants throughout China. Environment International 32 (2006) 365 – 373.
[1]On dry, normal temperature and pressure and 6% O2 basis, in line with the Chinese emission standards.
[2]GB 13223-2011 Emission standard of air pollution for thermal power plants