Particulate Matter Speciation
Portland Cement Kiln PM10
Recommendations for modeling the particulate matter (PM10) speciation for four types of Portland Cement Kilns ( Wet and Dry processes with Fabric Filters, and with ESPs), are contained in the Excel workbooks available in the Highlights box. Instructions for using the workbooks are available below.
The derivation of these values is based on data from AP-42, Section 11.6 (Portland Cement Manufacturing), as described below. Any individual source may deviate from these recommendations with the conditional approval of the Federal Land Manager (FLM) and the Regulatory Authority. Applicants seeking approval of an alternate speciation profile should submit stack testing data or other documentation supporting use of a different profile for the source in question. Where a different speciation profile is approved, the FLMs may request that the Regulatory Authority include emissions testing requirements in the source’s permit to confirm the validity of the alternate profile.
The filterable PM10 represents the emissions captured using the Method 5 "front-half" filter and the condensable PM10 represents the emissions captured using the Method 202 "back half" method.
Filterable PM10 mass is speciated as follows. Filterable PM10 size speciation data are given in AP-42 Tables 11.6-21 and 11.6-5. Filterable mass sized 2.5 microns or less fall into the fine PM10 category when calculating light extinction with the CALPUFF system. Although most of the fine PM10 has a light extinction coefficient (Bext) of 1.0, the FLMs will assume that a nominal portion of the fine PM10 emissions is unburned elemental carbon (Bext of 10) contained in the effluent. The remainder of the filterable PM10 is coarse PM10 which has a light extinction coefficient (Bext) of 0.6.
The breakdown of the remaining condensable PM10 emissions is based on "inorganic" and "organic" AP-42 Table 11.6-2 emission factors. However, the only organic emissions listed in AP-42 for a calcining process like this is in Table 11.17-2 for "coal-fired rotary preheater (lime) kiln with multiclone, water spray, and fabric filter". This represents the best available data in the absence of other information. The "inorganic" condensable PM is listed at 1.1 lb/ton and the "organic" condensable PM is listed at 0.15 lb/ton. So, about 12% of the total condensable PM mass is "organic". This "organic" fraction should be modeled as Secondary Organic Aerosols (SOA) with a light extinction coefficient of 4.0. The "inorganic" condensable PM10 (CPM IOR) is assumed comprised of sulfate (SO4) with a light extinction coefficient of 3.0 * f (RH), which accounts for the hygroscopic growth of sulfate aerosol in the presence of water vapor. This growth in particle size increases the light scattering abilities of the sulfate aerosols. This "organic" fraction should be modeled as Secondary Organic Aerosols (SOA) with a light extinction coefficient of 4.0. The "inorganic" condensable PM10 (CPM IOR) is assumed comprised of sulfate (SO4) with a light extinction coefficient of 3.0 * f (RH), which accounts for the hygroscopic growth of sulfate aerosol in the presence of water vapor. This growth in particle size increases the light-scattering abilities of the sulfate aerosols2.
The recommended PM10 speciation should be applied to the PM10 emission rate estimated for the source. Ideally, both the "filterable" and "condensable" PM10 emissions would be provided. Also, for modeling of visibility impacts and 24-hour PM10 NAAQS and PSD increments, the PM10 emission rate input to CALPUFF should not represent a compliance averaging time of longer than 24-hours.
Using the Workbooks
PLEASE NOTE: These workbooks are not "recyclable". Depending upon how you use them, certain links may be broken that would be essential for a different application. Download the workbook, rename the workbook before beginning any calculations, and use a new workbook for each application.
It should also be noted that these workbooks were developed primarily for application to existing emission units for which there may be no explicit limits on pollutants such as H2SO4. However, for new emission units for which information may be available on multiple PM10 constituents, the workbooks can be modified to incorporate that additional data.
Select the Excel workbook that most closely resembles the process/control in question. (You can ignore the magenta/bold f(RH) value which is included to test the effect of humidity on relative light extinction of the various species.) Except as indicated, the bold values in the body of the spreadsheets represent data that were entered directly and originated in either AP-42 Table 11.6-2, or, in the case of the "% of Fines" value for elemental carbon, is from Table 6 of EPA’s January 2002 DRAFT "Catalog of Global Emissions Inventories and Emission Inventory Tools for Black Carbon"3. Unless entering "custom" values, do not change any value in a cell that is not both bold and colored magenta, orange, or yellow.
Based upon the emissions data available (total or filterable PM10 in lb/hr or lb/ton of clinker), enter the emission rate into the appropriate (orange or yellow) cell with a corresponding dummy value. Corresponding emission rates for filterables and condensibles will show up in the green cells, and emission rates for each species of condensibles will appear in the blue cells–use these values. All condensable PM10 is considered to be submicron.
In order to separate filterable PM10emissions by size, the AP-42 Table 11.6-5 size fractions for the appropriate kiln and controls4 were used. The resulting charts show filterable PM10 emissions in lb/hr for size ranges modeled by CALPUFF. Match the proper species with the correct size range. The assumption is made that coarse PM10 is between 2.5 and 10 micron, and the corresponding emission rates are shown in the blue cells in the "PM Size" table. It is assumed that elemental carbon represents 5% of the filterable PM10 and is all in the smallest size range. The remaining fine soil was assigned to the size range below 2.5 micron, and the results are shown in the blue cells (in g/sec). Please note that the smallest CALPUFF size range has been split to show entries for fine soil and elemental carbon.
Neither the filterable PM10 speciation spreadsheet nor its associated charts will work unless the correct value is calculated for total filterable PM10 emissions (in lb/hr). There are several ways to do this, depending upon the type of emissions data initially input:
- If you enter Total PM10 in lb/hr (into orange cell C27), everything is calculated automatically for you.
- If you enter Total PM10 in lb/ton (into orange cell C34), you need to make sure that you have also entered the production rate into (magenta) cell D5; then, everything is calculated automatically for you.
- If you enter filterable PM10 in lb/hr (into yellow cell E40), the spreadsheet will calculate Total PM10 (in lb/hr) in (green) cell C40. Transfer that value to (orange) cell C27, and everything is calculated automatically for you.
- If you enter filterable PM10 in lb/ton (into yellow cell E46), the spreadsheet will calculate Total PM10 (in green cell C46) in lb/ton. Transfer that value to (orange) cell C34, and everything is calculated automatically for you, provided that you have entered the production rate into (magenta) cell C5.
Or, you can enter the filterable PM10 emission rate (in lb/hr) directly into (yellow) cell E23 of the "PM (Size)" table.
If you have questions, comments, or suggestions, please contact Don Shepherd at the National Park Service, Air Resources Division in Denver at 303-969-2075 or contact us through the Webmaster link at the bottom of the page.
1 For kilns with an ESP, all Filterable PM10 mass is assumed proportional in the same 0.65/0.77 ratio as for wet process kilns with an ESP. For kilns with a Fabric Filter, all Filterable PM10 mass is assumed proportional in the same 0.17/0.20 ratio as for dry process kilns with a Fabric Filter.
2 The values in row 27 of the tables result from an experiment in which we were trying to understand the effect of coal quality and f(RH) on the overall extinction from a given PM speciation profile.
3 The EPA DRAFT document estimates that, for Industrial combustion, .3.7% of fines would be EC for coal combustion, 7.4% for oil, and 6.7% for natural gas. These values appear in cell K22 of their respective spreadsheets.
4 We assumed that the size distributions for wet and dry processes were the same, and distinguished the size distributions by the type of control.