Environmental Protection

This Year's Model

The U.S. Environmental Protection Agency (EPA) recently promulgated a new modeling system, the AERMIC Model (AERMOD), for modeling air emissions from industrial sources. AERMOD replaces the venerable ISC (Industrial Source Complex) model that has been used for many years to demonstrate compliance with air quality regulations. AERMOD was designed to use much of the same input information as ISC, however, the meteorological input files and the air dispersion modeling equations in AERMOD are far more complex. This article describes the AERMOD system and provides an example of its use for industrial source modeling.

Before making certain modifications to an existing industrial facility, or building a new facility, dispersion modeling is often necessary to assess the potential air quality impacts as part of the permitting process. The Industrial Source Complex (ISC) model has been the workhorse model for this purpose for several decades. The model has roots in air dispersion research that was conducted in the 1960s. For some time, a comprehensive overhaul of EPA's basic regulatory models has been needed. The two primary catalysts for such an overhaul have been advances in computing capabilities that allow for increasingly sophisticated modeling schemes, without a concomitant increase in computing time, and improved understanding of atmospheric processes near the earth's surface. In response to these developments, a group of scientists from the American Meteorological Society and EPA formed the AMS/EPA Regulatory Model Improvement Committee (AERMIC). The group focused on regulatory models designed for estimating near-field impacts from a variety of industrial source types. As a result of their work, the AERMIC Model (AERMOD) was created as a replacement to ISC.

In designing AERMOD, a few outstanding issues were addressed by AERMIC. First, a model was needed that could provide reasonable concentration estimates under a wide variety of conditions. The model had to be user-friendly, requiring reasonable input data and computer resources. Finally, a robust dispersion model should also accommodate modifications as science evolves to be tractable for long-term use. As new advances in boundary layer modeling become available, for example, they should be readily implemented into the AERMOD code in a fairly seamless manner.

The Modeling System
The AERMOD modeling system consists of the air dispersion model, and two preprocessors -- AERMET and AERMAP. AERMET accepts surface weather observations, upper air soundings, and on-site data from instrumented towers. The program then processes the meteorological data and subsequently generates planetary boundary layer (PBL) parameters. These parameters include physical contributions to turbulence, mixing heights, and heat inputs into the boundary layer itself. Turbulence levels can be very sensitive to the character of the underlying surface. Therefore, AERMET was designed to incorporate types of land use, including urban areas, forest, and grassland. Emerging guidance on use of AERMET is expected to require specification of land use at the meteorological station. However, AERMET can also be used as a research tool to examine the effects of different land uses on air dispersion.

AERMAP generates terrain-height data required in AERMOD. This reflects a physical relationship between surface features and plume behavior and allows AERMOD to produce concentration data. AERMAP allows AERMOD to simulate the effects of airflow "splitting" around or over terrain features not previously possible with the ISC model. This capability produces more accurate concentration calculations.

AERMOD was recently modified to include an algorithm for modeling deposition and removal. This algorithm handles dry deposition and wet depletion for both particles and gases. Dry deposition processes are handled by two methods: one for particle diameters in excess of 10 microns and the other for particle size distributions that are not well known. For the second method, 90 percent of the particles are assumed to be less than 10 microns in diameter.

For regulatory purposes, the SCREEN3 model has been traditionally used to determine the need for refined ISC modeling. However, with the promulgation of AERMOD, a new screening model was needed as well. To meet this need, AERSCREEN is being developed. AERSCREEN uses a standard set of worst-case meteorological data designed to produce more conservative concentration estimates. AERSCREEN is still in development.

An Example Application
To illustrate AERMOD's functionality, we will look at a typical application at an industrial facility. For this illustration, EPA's AERMOD Fortran codes are accessed through a commercial product, BREEZE AERMOD GIS Pro. This product provides a Windows® Graphical User Interface (GUI) to easily enter parameters into the relatively cumbersome Fortran code interfaces.

An analyst typically starts by importing a facility plot plan that contains buildings, pollutant sources, roads, and the property boundary (fenceline). These maps are often in AutoCAD® .dxf (Drawing eXchange Format) format. The next step is to create sources of emissions and receptors. To simplify the creation of receptors, a receptor "grid" -- a set of points at which AERMOD will calculate concentrations beyond the facility, can be automatically generated. GUI and GIS (geographical information system) editing tools allow the receptor grid, as well as fenceline receptors along the facility boundary, to be easily specified.

After specifying a property boundary and a receptor grid, the user is ready to create buildings and other windflow obstructions within the GUI. Although these buildings may not be sources, they are of prime importance as plume dispersion is sensitive to building downwash -- the effect of turbulence generated by the buildings. After adding the buildings, AERMAP is run to import terrain elevations to all modeling objects. This can be verified by using the BREEZE product's 3D view features. An example of a 3D view is given in Figure 3. The vertical scale is exaggerated in the figure to illustrate the terrain relief.

Once the terrain elevations are imported, we are ready to run AERMOD in order to predict the pollutant concentrations on the receptor grid and any discrete receptors. AERMOD output can be viewed in tabular or graphical format. The tabular file contains pollutant concentrations, among other parameters, for all receptors, listed with their coordinates. The concentration data can also be displayed in graphical format, which is particularly useful for reports and presentations.

AERMOD can accommodate multiple sources and source types that originate from different heights over different areas. The sources at this example facility include an ore pile, conveyor belt, roof monitor, and a stack. AERMOD is capable of modeling this facility using a combination of volume, circular area, and point sources of differing heights. Note the complex terrain in this figure. This gives decision makers and analysts a good picture of the underlying terrain, which is useful when interpreting results from AERMOD.

Summary
Although similarities, such as the input and output computer architecture, exist between the ISC and AERMOD, advances in the understanding and simulation of turbulence and plume rise are integrated into the AERMOD system. In addition, enhanced land use parameterization allows for greater flexibility in modeling sites that do not fit neatly into the urban or rural characterization. Furthermore, well-designed interface products allow analysts to easily construct AERMOD simulations of air pollution sources, visualize their modeling domain in high-impact 3D projections, and perform the post-processing of the concentration results often needed for a variety of planning and permitting scenarios.

This editorial originally appeared in the April 2006 issue of Environmental Protection, Vol. 17, No. 3

This article originally appeared in the 04/01/2006 issue of Environmental Protection.

comments powered by Disqus

Free e-News Subscription

I agree to this site's Privacy Policy