Let’s be completely honest with each other. Nobody, and I do mean absolutely nobody, does dispersion modeling for fun. It is not as physically demanding as nailing down roofing singles in the summertime heat, but both, roofing and dispersion modeling, are those things you do only when you have to and do no more than is necessary.
Some state agencies and local air quality programs regulate air toxics in addition to the federally regulated criteria pollutants. For some industry sections, analysis of dozens or even over a hundred air contaminants may be required in order to be issued an air quality permit.
Just because dozens of air contaminants need to be modeled, it does not mean dozens or even hundreds of modeling runs need to be executed. A lot of time and effort can be save if a simple modeling technique is used: unitized emission or generic modeling.
What Is Unitized Emission Modeling?
As the name implies, unitized emission modeling means each source is represented in the model input file with a unit emission rate, either 1 lb/hr or 1 g/sec. (Hint: It is good practice to use a value that is of the same order of magnitude as the largest air contaminant emission rate for any source.) Since models like AERMOD calculate results by defined source groups, the other element of unitized emission modeling is to create separate source groups for each source and have the results reported for each source. The final result is to multiply the unitized concentrations for each source by the air contaminant emission rate and then add all the single source results together for a project concentration.
This technique can be used any time you perform dispersion modeling (except for NO2 when using the ARM2, PVMRM, or OLM options) but it is extremely useful when you have at least a dozen or so sources and a dozen or so air contaminants to evaluate.
Different Levels of Unitized Emission Modeling
Unitized emission modeling is primarily an screening technique but it can be used to get results equivalent to using air contaminant specific emission rates. We will refer to the levels as 1-D, 2-D, and 3-D unitized emission modeling.
1-D Unitized Emission Modeling
The simplest, and probably most widely used level of unitized emission modeling, is when only the predicted maximum ground-level concentration (GLCmax) for each source in the analysis is used. The temporal and spatial variations in the predicted concentrations for each source are ignored making this result independent of time and space. The result is very over-predictive.
The results are typically reported in a spreadsheet format, with one sheet listing the unitized GLCmax for each source, the second sheet listing the air contaminant emission rates for each source, and the third sheet listing the products of the unitized GLCmax values and the emission rates for each source. The sum of the products for each air contaminant are typically listed on the same page.
One of the reasons this level of the technique is so widely used is that the result could come from any dispersion model, such as SCREEN3, AERSCREEN, ISC-PRIME, or AERMOD.
2-D Unitized Emission Modeling
Use of the 2-D level requires the use of refined dispersion model such as AERMOD or ISC-PRIME. This technique does take into consideration the spatial distribution of predicted concentrations from each source but does not consider the hour when those concentrations were estimated to occur (independent of time). The results are still over-predictive from air contaminant specific modeling, but not to the extent as the 1-D technique.
To put the 2-D technique in to action, in the model run file, create a source group for each source as shown below:
Then, in the output portion of the run file, report the maximum predicted concentration for each receptor using the PLOTFILE option as show below:
You unitized model results will look something like this:
When you report the project results, you have a choice of using a spreadsheet or a database program. With the spreadsheet, you will need to show the unitized results for each source, the emission rates for each source, the products of the unitized concentration values and the emission rates for each source at each receptor, the sum of all the products, and then the maximum value.
Using a spreadsheet, you will need to use a spreadsheet workbook with the number of air contaminants plus two sheets. For example, if you are evaluating 12 air contaminants, you will need 14 pages in your spreadsheet workbook to display the results.
My personal preference is to use a database program, like MS Access or SQL Server. With a database, you need one table with all of the PLOTFILE results, one table with all the emission rates for the all the sources, and a query that performs the calculation of applying the unitized concentration at each receptor to each emission rate, creating the sum for all sources at a receptor, and then taking the maximum for all receptors by air contaminant.
This query can be done in about two dozen lines of code with only three lines (applying the emission rate, creating the sum, and taking the maximum) being important. The spreadsheet method entails (the number of receptors) X (number of sources) X (the number of air contaminants) calculations to review.
If I was reviewing the modeling, I would choose the dozen or so lines to review.
3-D Unitized Emission Modeling
This technique does take into consideration the spatial and temporal distribution of predicted concentrations from each source. The final results are identical to air contaminant specific modeling.
To put the 3-D technique in to action, in the model run file, create a source group for each source in the same way as described with the 2-D technique. Then, in the output portion of the run file, report the all predicted concentrations for each receptor for each hour using the MAXIFILE option as show below:
The reason for using the MAXIFILE option is, if you use a low enough threshold value, all the non-zero values will be captured. The POSTFILE option could be used, however, the majority of the values in the POSTFILE are zeros. Both the MAXIFILE and POSTFILE will produce massive files (tens to hundreds of gigabytes in size), but the POSTFILE takes up more space without providing any value.
When using the 3-D technique, especially, your unit rate should be of the same order of magnitude as the largest air contaminant emission rate. Since the MAXIFILE will only capture results greater than the threshold value, 1.0E-06 in this case, you want to make sure capture as many of the non-zero values as is practicable. For example, if the highest emission rate is 9 lb/hr, it is recommended you use 10 lb/hr as the unit rate rather than 1 lb/hr. A better practice is to use 1 g/sec (7.94 lb/hr) or 10 g/sec (79.4 lb/hr).
This techniques will generate a huge result file, so you will need to use a database program like SQL Server that can handle large datasets. MS Access has only a 2 GB limit which would be insufficient.
The unitized concentration results will look something like this:
The results can be processed with a database. You would need one table with all of the MAXIFILE results, one table with all the emission rates for the all the sources, and a query that performs the calculation of applying the unitized concentration at each receptor and hour to each emission rate, creating the sum for all sources at a receptor and hour, and then taking the maximum for all receptors and hours by air contaminant.
To drive the home the point that spreadsheets have their limitations, if these results could be put into a spreadsheet, it would require (number of contaminants) X 8784 + 2 pages in the spreadsheet workbook.
Why Use Unitized Emission Modeling?
The reason to use unitized emission modeling is to save on model run time and the effort of putting together dozens of model run files and processing dozens modeling result sets. Though the 1-D approach is the simplest and most popular, the 2-D techniques is the most efficient for most dispersion modeling projects, provided you are using a database. The 3-D technique is the most robust, but does take special planning just because of the large amount of data produced by the result set.
To give an idea of the time and effort savings, we recently had a project with 40+ sources and 70+ air contaminants. Using the 2-D approach, the unitized emission model run took 3 hours to execute. To process the PLOTFILE with 40+ result sets, import into the database, and run the query to calculate 70+ GLCmax values to 2-3 minutes. If emission rates change, the results were available within one minute!
For another project, when using the 3-D approach, the unitized emission model run took 45 minutes to execute and then 11 seconds to run the query to get results.
If a dispersion modeling project requires dozens of air contaminant specific model runs, the unitized emission modeling techniques can significantly reduce the number of model runs necessary to produce a result in a fraction of the time. There are three levels of refinement that can produce results that are very conservative (over estimating predicted concentrations) to those equivalent to air contaminant specific model runs.
Regardless of the level of refinement, this techniques generates data that can be presented in a spreadsheet program, or, can be used with a database program. Using a database program is preferred since the amount of data generated may be too large for spreadsheet programs to handle.
If you found this article informative, there is more helpful and actionable information for you. Go to http://learn.naviknow.com to see a list of past webinar mini-courses. Every Wednesday (Webinar Wednesday), NaviKnow is offering FREE webinar mini-courses on topics related to air quality dispersion modeling and air quality permitting. If you want to be on our email list, drop me a line at firstname.lastname@example.org.
One of the goals of NaviKnow is to create an air quality professional community to share ideas and helpful hints like those covered in this article. So if you found this article helpful, please share with a colleague.