Posted by Ed on June 20, 2007, 11:59 pm E. H. Pechan & Associates, Inc. POWER PLANTS across the United States emit hundreds Several severe air pollution events in the mid-1900s U.S. Environmental Protection Agency (EPA) adopted Deck et al.16 demonstrated that it is possible to predict October 2004 [Vol. 59 (No. 10)] Method In May 2002, a detailed emissions inventory was Seventeen of the power plants are to be full-time The output of the CRDM is a matrix of coefficients Total PM consists of primary PM, material emitted, and (3) nitrate; and (4) ammonium. Similar adjustments to those used in the Particulate Statistical analysis. To calculate the amount of g/m3 change in PM2.5 established by Pope et al.17 The where: y0 county level annual premature death rate per e exponent; PM2.5 coefficient of 0.0040182 (derived from Results The 29 proposed power plants were predicted to Archives of Environmental Health
24.34.144.49
RICHARD P. HERMANN JACK O. LANIER
Chesterfield Health District Virginia Commonwealth University
Richmond, Virginia Medical College of Virginia Campus
FRANK DIVITA, Jr. Richmond, Virginia
Springfield, Virginia
ABSTRACT. The authors estimated the number of premature deaths from particulate
matter
less than 2.5 m (PM2.5) that would result from making 29 proposed fossil fuel
power plants
in Virginia operational. We used a U.S. Environmental Protection Agency air
quality model
(Climatological Regional Dispersion model) to calculate changes in ambient
concentrations
of PM2.5 and Cox proportional hazard modeling to calculate the resulting
premature mortality.
The model predicted that if all 29 plants were operational, PM2.5 concentrations
would
rise in 271 counties across 19 states 5 and increased average annual PM2.5
concentrations
would result in a rate of 17 deaths per 37,900,026 people aged 30 yr and older
(0.45 deaths
per million, 95% confidence interval 0.31, 0.59) per year by the end of 2004,
increasing
thereafter. Over a 6 yr period, 104 cumulative excess deaths would occur due to
operations
of these proposed plants. The authors recommend that precautionary principles be
considered
when policy decisions related to energy production from fossil fuels are made.
of thousands of tons of air pollutants into the
atmosphere each year. This pollution consists of directly
emitted particulate matter (PM) less than 10 and 2.5 m
in diameter (PM10 and PM2.5, respectively), sulfur dioxide
(SO2), nitrogen oxides (NOx), and carbon dioxide
(CO2). Over the past decade, numerous studies have
linked PM to hospital admissions and to a wide range of
adverse human health effects, from premature death and
asthma attacks to chronic bronchitis and impairment of
the immune system. Of particular concern is PM2.5,
because it is small enough to reach the deep alveolar
spaces of the lungs. A reduction or elimination of PM2.5
could dramatically improve quality of life and years of
healthy life, thus addressing one of the overarching
goals of Healthy People 2010.1
established the correlation between high concentrations
of PM and mortality.2–4 By the late 1980s and early 1990s,
there was strong epidemiological evidence linking
health effects to even very low concentrations of PM.5 In
1997, based on these and many other studies,6–13 the
new ambient air quality standards imposing regulatory
limits on PM2.5.14 In 2001, the U.S. Supreme Court
upheld these new standards in spite of challenges by
industry.14
Following deregulation of the electric utility industry
in Virginia in the late 1990s, more than 30 proposals
were initiated throughout the state for new power plants
designed to use fossil fuel. The addition of these plants
would nearly double Virginia’s current electricity generating
capacity.15 In 2002, 29 proposals were active.
the effects of point source emissions on ambient air pollution
levels thousands of miles away, and used long-term
epidemiological studies correlating mortality to PM exposure
to show a change in concentration of PM could be
translated into a change in mortality. In 2002, Pope
et al.17 published the results of a study that followed
500,000 adults for 16 yr to establish the long-term effects
of fine particulate air pollution on premature mortality.
They concluded that fine particulate and sulfur oxide-
related pollution were strongly associated with all-cause
death, lung cancer death, and cardiopulmonary death.
The objective of our study was to estimate the number of
premature deaths in Virginia and surrounding states that
may be expected if all 29 proposed power plants are constructed
and used for power generation.
obtained for each of the 29 proposed power plants from
the Virginia Department of Environmental Quality.15
The emission inventory data were provided to the
department by the companies proposing the power
plants. Information included estimated annual emissions
of PM2.5, PM10, SO2, NOx, and ash content as well
as the fuel type and capacity of the boilers at each plant
and the associated stack parameters (geographic coordinates,
height, diameter, flow rate, and velocity).
generators and were therefore assumed to be working at
90% capacity. The remaining 12 power plants are
designed to operate to meet peak demand only (i.e.,
summer) and were therefore assumed to be working at
50% capacity. Emissions data for each of the 29 proposed
sites, a total of 108 boilers, modified by these
operation levels, were input into the Climatological
Regional Dispersion Model (CRDM) for air quality,
which was developed by the EPA to estimate the impact
of the Particulate Matter New Ambient Air Quality
Standards released in 1997.13,18 Running the CRDM
produced estimates of annual average increases in concentrations
of various PM-related species due to these
power plants for every county in the contiguous United
States. The PM-related species are: primary PM2.5 and
PM10; nitrates, sulfates; and ammonium, which is a
secondary pollutant formed from the emitted pollutants
in the atmosphere associated with the secondary sulfate
and nitrate. The model includes climatological variables,
particularly the frequency of various wind and
atmospheric conditions, the dispersion and deposition
characteristics of pollutants, and accounts for chemical
oxidation of SO2 to sulfate and NOx to nitrate in the
atmosphere to calculate the frequency of transport of
pollutants to every part of the contiguous United States
CRDM estimates long-term average concentrations of
air pollutants using a method developed by Turner,19
which takes into account the frequency of occurrence of
various wind and atmospheric stability conditions to
calculate the frequency of transport to various areas.
called the source-receptor matrix (S-R matrix). This is a
grid consisting of fixed coefficients that reflect the relationship
between each county’s annual average PM
concentration value and the specific contributing pollution
source. The annual change in ambient air quality is
reported in units of micrograms per cubic meter ( g/m3).
secondary PM, formed from the chemical conversion of
SO2, NOx, and NH3 to PM2.5 in the atmosphere.
Therefore, four S-R matrices were obtained from the
model: (1) primary PM, including primary emissions of
PM10 and PM2.5 from the proposed plants; (2) sulfate;
To assess the county-level change from baseline in the
annual average concentration of PM2.5, we compared the
pollution estimates of two scenarios: (1) a baseline business-
as-usual scenario and (2) the baseline scenario plus
emissions contributed from the 108 boilers at the proposed
29 power plants. The baseline ambient pollution
concentrations for each county were obtained from recent
data by Deck et al.16, who used the same methodology.
Matter National Ambient Air Quality Standards analysis
were applied to the S-R matrix, including a fugitive dust
adjustment factor of 0.25 for primary PM2.5 and PM10
emissions so that the contribution of this pollutant to
total PM2.5 concentrations better matched monitoring
data.20 In addition, corrections were made for emissions
from natural sources. The calculations assumed full
implementation by 2003 of the Acid Rain Program,
Phase II of Title IV of the Clean Air Act Amendments of
1990,21,22 and the EPA Summer Smog rule.23 It also
assumed no explicit adoption of a global warming
climate treaty.
expected premature mortality, a concentration-response
function was constructed using Cox proportional hazards
survival modeling.24 We used the relative risk of
death as 1.041 (1.008, 1.075) for exposure to every 10.0
formula used was the log-linear relationship of PM2.5 to
premature mortality in the form of:
Change in mortality (y0[e – (change in PM2.5) 1])
population,
100,000 persons 30 yr and older;
ln (relative risk)/change in PM2.5, where the reported
relative risk 1.041 is associated with a change in
mean PM2.5 exposure of 10.0 g/m3 and standard error
0.0024338;
change in PM2.5 change in annual mean PM2.5
concentration derived from the S-R matrix; and
population population aged 30 yr and older.
increase PM2.5 in 272 counties across 19 states and
Washington, DC, with a 2004 population of 37,900,026 primarily in the
mid-Atlantic region and northeastern
people age 30 yr and older (see Table 1; Fig. 1). No other U.S. (Connecticut,
Message Thread:
« Back to thread