THWRC Awarded Proposal 515LUB0040H

Project Number:       515LUB0040H

Title:                             Proactive Abnormal Emission Identification via Modeling, Optimization, and

                                     Design of Air-quality Monitoring Network

Lead PI:                     Qiang Xu

Awarded Amount:    $30,000



Chemical facilities, where large amounts of chemicals and fuels are housed, processed, and produced, have high risks to originate air emission events. Such emission events can be caused by planned operations or uncertainties like equipment failure, false operation, natural disaster, and terrorist attack. Chemical plant emission events can be very dangerous to both local communities and their surrounding environments. Generally, the plant personnel should document and report emission details when an emission event occurs, so that valuable information like the hazardous release rate, transportation speed and direction, and potential harmful impacts on exposed populations and ecosystems can be evaluated to support right decisions. Such decisions are very critical. Thus, another independent information channel: the real-time measurement from a local air-quality monitoring network (AQMN) is also vitally needed. This is especially important for industrial regions heavily populated by various chemical facilities. Based on the AQMN, it is possible to detect emission sources responsible for serious emission events, so as to timely and effectively support diagnostic and prognostic decisions for all stake holders, including government agencies, chemical plants, and residential communities.

 From the literature survey, the inverse modeling for emission source identification was initially employed with atmospheric dispersion models, which were used in forward modeling problems to determine downwind contamination concentrations. Even though inverse modeling methods based on Gaussian plume models have been reported, they are generally used to estimate average emission rates of point sources over a long-time period. That means their emissions are assumed under steady-state conditions and emission rates are treated as constants. However, abnormal emission identifications are quite different from those under normal emissions because of their unsteady-state emissions and fast response requirements. Hitherto, there is still a lack of studies on the reverse modeling for abnormal emission identifications with considerations of unsteady-state emission rates and process root cause analysis

 In this project, a systematic methodology for abnormal emission identifications will be developed. It includes four stages of modeling and optimization work: (i) determination of background normal emission rates from multiple emission sources; (ii) identification and quantification of abnormal emission sources based on the current AQMN; (iii) identification of chemical plant root causes from the abnormal emission observation and identification; and (iv) improving emission event detection ability via the optimal design/retrofit of AQMN. This project, for the first time, proposes a systematic methodology coupling abnormal emission observation, point source identification and quantification, process root cause analysis, and AQMN design/retrofit optimization. This study will not only determine emission source location, starting time, and time duration responsible for an observed emission event, but also reversely estimate dynamic emission rates and the total emission amount from accidental emission sources. Additionally, it can provide valuable information for accident investigations and root cause analysis for those chemical plant emission events. If successful, the project achievements will benefit all stake holders and environmental quality in Texas. This project will take two years to complete the study. The total budget requested is $65,646.