Methods for incorporating uncertainty into dispersion and deposition models

✍ Scribed by Williamson, Derek G. ;Graettinger, Andrew J. ;Yegnan, Anand

• Publisher: American Institute of Chemical Engineers
• Year: 2004
• Tongue: English
• Weight: 254 KB
• Volume: 23
• Category: Article
• ISSN: 0278-4491
• DOI: 10.1002/ep.10016

✦ Synopsis

Abstract

Dispersion and deposition models are critical components of risk assessments for airborne discharge from industrial facilities. The probabilistic nature of risk assessment, the complexities of dispersion and deposition models, and the natural variability of input parameters all support the need for probabilistic output from dispersion/deposition models. However, difficulty in implementation and data availability has prevented the routine use of uncertainty in dispersion/deposition modeling. A series of relatively simple case studies demonstrate that the Taylor series is a flexible and computationally efficient method for incorporating input uncertainty into a widely used air transport model (ISCST). Relatively simple dispersion and dispersion with deposition modeling scenarios using real input data demonstrate that the Taylor series: (1) provides similar estimates of uncertainty compared to traditional Monte Carlo simulation, (2) is far more computationally efficient, and (3) in most cases provides conservative estimates of output concentration. More over, an evaluation the computational time needed for traditional Monte Carlo simulation when considering several input variables simultaneously demonstrates the potential of the Taylor series technique to estimate uncertainty in far more complicated modeling scenarios. © 2004 American Institute of Chemical Engineers Environ Prog, 23: 158–167, 2004