EVALUATION OF DISPERSION MODEL PERFORMANCE IN PREDICTING SO2 CONCENTRATIONS FROM PETROLEUM REFINERY COMPLEX
Keywords:
AERMOD, CALPUFF, Sulfur dioxide, Multi criteria attribute, ThailandAbstract
The AERMOD and CALPUFF air dispersion models are tested for their performance in
predicting ground level concentration of sulfur dioxide in Thailand. Emission data used in this study are
obtained from petroleum refinery complex. Predicted results are compared with those measured data using the
year 2012 as a reference year. A set of statistical parameters are employed to evaluate model performance.
Overall results indicated that both AERMOD and CALPUFF can provide good results. However, AERMOD
can perform better in predicting of extreme end of the concentration distribution at the receptor sites. The
maximum ground level concentrations of sulfur dioxide within the modeling domain are about 359 and 456
µg/m3 for AERMOD and CALPUFF simulations, respectively. This result indicates that CALPUFF provides
more conservative of maximum result than predicted data from AERMOD. The decision to select an
appropriate dispersion model in the study is accomplish by using the Multi-Criteria Attribute (MCA) analysis.
Result from MCA supports that AERMOD is more appropriate to be applied for study of air dispersion in this
area than CALPUFF system.