Developing a Predictive Emission Model for a Diesel Fuel-fired Power Plant
Terence P. Tumolva
Thesis (M.S. Chemical Engineering)--University of the Philippines, Diliman.-2006
Previous studies have demonstrated the feasibility of Predictive Emission Monitoring Systems (PEMS) as an alternative to the Continuous Emission Monitoring Systems (CEMS) in monitoring emissions from industrial plant operations. PEMS is more cost effective compared to CEMS and has the added feature of diagnostic and trending capabilities. In view of its potential applicability in the country, a study was conducted in on the use of PEMS in a privately owned diesel power plant in the Philippines. A computer-based "first principles" PEMS model was developed to estimate emission concentrations levels for four criteria air pollutants, namely sulfur oxides (as SO2), nitrogen oxides (as NO2), particulate matter (PM) and carbon monoxide (CO). Using plant operations data as input, the model simulates the diesel engine operations and predict the resulting emission characteristics using stoichiometric and thermodynamic principles. The model was validated by comparing emission concentrations calculated using the model with actual emissions data measured. The model was calibrated using historical plant operation data to increase the PEMS accuracy. Additional refinements un the computer model, such as emission data recording and storage, were also done to increase the model's handiness and practicability in operation. The study showed that the PEMS model developed for the diesel power plant could effectively predict PM emissions.