| Abstract | This study focuses on the dispersion of air pollutants in the Laem Chabang Industrial
Estate (LCB I.E.), Chonburi, Thailand, during 15 to 31December2003 using the Industrial
Sources Complex Short Term model (ISCST3) and the California Mesoscale Puff Model
(CALPUFF). The major goal of this study is to select and evaluate dispersion models that
are applicable to the LCB I.E.. The pollutants considered include sulfur dioxide (S02),
nitrogen oxides (NOx), and particulate matter size less than 10 μm (PM10).
The S02 emission data were taken from fuel consumption reports of all industries in LCB
LE, while NOx and PM10 emission data were estimated from their Environmental Impact
Assessment reports (EIA). Totally, emissions from twenty three stacks were used for
simulation, which account for 86, 87, and 55% of total emission reported in EIA for the
whole LCB I.E. of the S02, NOx, and PM1o, respectively. Meteorological data were
gathered from the both Pollution Control Department and the Thailand Meteorological
Department.
Two different case studies were done in CALPUFF modeling, which are fresh water and
seawater cases. The fresh water case, with no effect function in coastal area, is used for
comparison the performance between CALPUFF and ISC. The outputs were set to display
the hourly average ambient concentrations at the Cartesian receptor grid and four discrete
points corresponding to four monitoring stations. The 1-hr maximum concentrations of
S02, NOx, and PM10 output from ISC are 2132, 2065, and 165 μg/m3
, respectively. For the
CALPUFF case 1 (fresh water), the 1-hr maximum concentration of S02, NOx, and PM10
are 1437, 611, and 243 μg/m3
, respectively, the corresponding values for case 2 (seawater)
are 1198, 603, and 203 μg/m3
, respectively. Both models indicated that these maximum
concentrations occurred at the western part of the Laem Chabang area during the modeling
period.
Model performance was evaluated by comparison between model outputs of S02, NOx,
and PM10 and the corresponding monitoring data at the four monitoring stations. The
performance criterion is R-square between observed and output data. Between 50
maximum 1-hour average concentrations and observed data, good correlation was
obtained. However, poor correlation was obtained for all hourly data were considered.
Correlation between the ISC and CALPUFF model outputs was also good between the 50
maximum concentrations. According to time series analysis, the model outputs are not
agreed well with the monitoring data, i.e. S02 and NOx are overestimated, whereas PM10 is
underestimated. The models were also run to identify the emission level to meet the
ambient air quality standard at LCB I.E .. From the model outputs, it was found that at 75%
of S02 reduction of emission at considered sources was necessary. |