1、 第 1 页，共 59 页 翻译 原文 Comparison of CALPUFF and ISCST3 models for predicting downwind odor and source emission rates Abstract CALPUFF model and ISCST3 Gaussian dispersion models were evaluated for predicting downwind odor concentrations and back-calculating area source odor emission rates. The compari
2、son between the predicted and fieldsampled downwind concentrations indicates that the CALPUFF model could fairly well predict average downwind odor concentrations. However, ISCST3 tended to under predict downwind odor concentrations as compared to the measured concentrations. Both the CALPUFF and IS
3、CST3 models failed to predict peak odor concentrations using the constant average emission rate. Odor emission rates obtained by back-calculating fluxes using CALPUFF and ISC models with the same field measurements of downwind odor concentrations are significantly different. It indicates that back-c
4、alculated emission rates are model specific. The modeled emission rates tended to be higher than flux chamber source sampling results. The flux chamber protocol may under-estimate odor emission rates. Keywords: Odor modeling； CALPUFF； ISCST3； Odor emission rate； Odor flux 第 2 页，共 59 页 1. Introductio
5、n Odorous gas emission from large confined animal feeding operations has been of increasing concern in the USA. This concern accentuates the need for additional study of odor mitigation and modeling. Currently, atmospheric dispersion models are used as a tool to predict downwind pollutant concentrat
6、ions, or to back-calculate average pollutant emission rates from downwind concentration measurements (Gassman, 1992； Chen, et al., 1998； Jacobson, et al.,2000； Zhu, et al., 2000； Hoff and Bundy, 2003). Industrial Source Complex-Short Term, Version 3 (ISCST3) is the US Environmental Protection Agency
7、 (EPA) approved and recommended dispersion modeling program that is being used by most State Air Pollution Regulatory Agencies (SAPRAs)in the USA to estimate downwind concentrations of pollutants. ISCST3 includes a set of Gaussian plume-based models that can be used to predict downwind concentration
8、s from point, line, and area sources. As pointed out by Smith (1993), there are a number of important factors that cause difficulties when using Gaussian plume model technique to predict downwind odor intensities. A considerable amount of research has been conducted to improve ISCST3 model accuracy
9、for downwind particulate matter (PM) concentration predictions and back calculated PM emission rates. However, use of ISCST3 to predict odor emission from large animal feeding operations remains a challenge. It has been reported (Wang et al., 2004) that ISCST3 can be used to predict average downwind
10、 odor concentrations, but for peak odor concentrations. ISCST3 also had difficulty predicting downwind concentrations at wind speeds higher than 6m/s. CALPUFF dispersion model, and other similar models and programs were developed by Sigma Research Corporation as a generalized non-steady state air em
11、ission modeling system for regulatory use (Earth Tech and Inc., 2000). The original development of the CALPUFF system was sponsored by the California Air Resources Board. The US EPA has proposed to use the CALPUFF modeling system as a guideline model for regulatory applications involving long range
12、transport and on a case-by-case basis for near-field applications where non-steady-state effects may be 第 3 页，共 59 页 important. Evaluation of a model performance for odor emission prediction requires a large amount of fieldwork. This paper reports a quantitative examination of the performance of ISC
13、ST3 and CALPUFF for fugitive odor emission using field sampling data. The ultimate goal of this research is to address the problems associated with these two dispersion-modeling systems in application for odor modeling. 1.1. ISCST3 Gaussian plume models ISCST3 Gaussian plume models for predicting do
14、wnwind odor concentrations from point, line and area sources can be described by the following equations: 2 222e x p e x p22pyz yzC uQ y H (for point source） （ 1） 222 e x p2 2Lz zC uQ H ( for line source ) (2) 21e x p22AxyyzVC d y d xu yyQ (for area source ) (3) where C is the downwind odor concentr
15、ation in odor units (OU), QP the point source odor emission rate(OUm3s-1), QL is line source odor emission rate (OUm2 s-1), QA the area source odor emission rate (OUms-1), y, zthe PasquillGifford plume spread parameters based on stability class, u the average wind speed at pollutant release height (
16、ms-1), H the effective height above ground of emission source (m), V the vertical term used to describe vertical distribution of the plume, x the upwind direction (m), and y the cross wind direction (m). 1.2. CALPUFF modeling system As described in A Users Guide for the CALPUFF Dispersion Model (Ear
17、th Tech and Inc.,2000), Puff models represent a continuous plume as a number of discrete packets of pollutant. The puff model evaluates the contribution of a puff to the concentration at a receptor by a snapshot approach.Each puff is frozen at particulate time intervals (sampling steps). The concentration due to the frozen
