Document 0LwVOD8XL6x6Qa6OeOeRe7MBm

AR226-2622 \ \ 0-I Draft: Prepared at the Request of Council AR226-2622 This report was not issued as shown, but was incorporated into a larger Chambers Works report under the consent order Section 1 Chambers Works ISC3 Modeling Methodology and Results Emission Source Information The ISC3 model was used to calculate ambient ground-level air concentrations for emissions ofC8 from the Chambers Works site. Table 1 shows the stack parameters and emission rates used in the model for each emission point. Section 3 of this report contains the basis for the emission estimates ofC8. Modeling Methodology Dispersion and deposition modeling was performed using the U.S. EPA's Industrial Source Complex 3 Model (ISC3), version 00101, provided by Trinity Consultants. All modeling was done in accordance with the procedures in EPA's Guideline on Air Quality Models (40 CFR Part 51, Appendix W). The EPA regulatory default options and rural dispersion coefficients were used in the model. The C8 emission sources were evaluated for downwash effects from surrounding buildings. The Building Profile and Input Program (BPIP) provided by Trinity consultants was used to provide wind direction specific building parameters. All buildings on the site were evaluated to determine if they could potentially impact the stack by causing building downwash effects. Table 2 lists the buildings included in the model and their heights. A 100-meter receptor grid extending out 600 meters from the plant fenceline was used. In addition, discrete receptors with 50-meter spacing were placed on the plant property line. Since the area surrounding the plant site is flat, no terrain elevations were used. A plot of the receptor grid used is shown in Figure 1. Five years of meteorological data (1989-1993) were analyzed. The surface data is from New Castle County Airport (Wilmington, DE) and the upper air data is from Washington Dulles Airport (Sterling, VA). An anemometer height of 6.1 meters was used for the modeling. Modeling Results An averaging time of one year was used to determine the annual average ground-level concentrations over the entire receptor grid. The modeling results for each year are shown in Table 3. A contour plot of the annual average concentrations for the year 1990 is shown in Figure 2. The maximum off-site value predicted by the model was 0.00364 |J,g/m3for the year 1990. This value was located at a receptor on the plant property line along highway US 130. 1/11 Section 1 11/12/04 Draft: Prepared at the Request of Council Table 1 Stack Parameters and Emission Rates DDE Building 1163 Stack Telomer A DMA Roof#l IZFAN Cmde 1156 Building Hotwell 185 Alcohol Drying 185 Alcohol Drying 185 Building Hotwell TS-45 Tank Vent D Building D Building D Building Roof D Building Jets 0.033 6.06E-06 4.45E-05 2.94E-06 1.62E-06 1.87E-05 2.04E-06 25ft 70ft 12ft Oft 25ft 55ft 88ft EO Center Hotwell and Stack *185 building hotwell stack points at the ground 2.19E-05 85ft 31" 92ft/s 194F 2" 17acfin 150F 2" 0.033 acfin 65C 6" 0 acfin* 3" lOft/s 40C 70C 2" 1 acfin ambient 36" 10,000 ambient acfin 4" 80 acfin ambient Table 2 Building Heights Naaie '--y^' -; - yiyf1'^ l J26 1089 1094 T3 589 745 888 1183 1182 669 1205 60 43 38 1050 15.5 1163 30 115 East 45 115 West 15 656 50 1402 60 185 60 1156 60 234 40 788 31 16 22 13.5 45 30 30 25 68 45 30 40 2/11 Section 1 11/12/04 Draft: Prepared at the Request of Council Table 3 Modeling Results 0.00309 0.00364 0.00305 0.00267 0.00300 plant fenceline along US 130 jplantfenceline along DE River 3/11 Section 1 11/12/04 Draft: Prepared at the Request of Council ...: .:.. .:.. :... :... :... :... :...:...:...:...:... .,.-......, < .... ...? . .'vt"^ .'"^-..-,;^\.......7.................................................... :::::::::::::::::: .......................^.../^' .................................... :::::::::::: 'J /:::::::::::::::::: ..............'. .-....^-. ..'" ...f..."..................................... f ............ .^ . ^,.<............................................ ... ... ................... . . . .. .../ ..., - . ^ . ^ .................................................... ... ... ... ... ... ... ... . . . -1000 1000 2000 3000 4000 Figure 1 Receptor Grid 5000 6000 7000 4/11 Section I 11/12/04 Draft: Prepared at the Request of Council Figure 1 Annual Average Concentrations - 1990 Meteorological Data 5/11 Section 1 11/12/04 Draft: Prepared at the Request of Council Section 2 Washington Works Screen3 Modeling Methodology and Results Data and Modeling Procedures Dispersion modeling was performed using the U.S. EPA's Screen3 model. This model is a screening tool which gives predictions of ambient ground level concentrations for a single stack. Due to the assumptions made in the model, particularly those regarding meteorological conditions, these predictions will always be more conservative than predictions made by the ISC3 model. Although there are several different sources ofC8 emissions in the Telomers area at the Washington Works site, these emission points are all headered together into a common stack. The following parameters for this stack were used in the model: Stack Height: 85 feet Diameter: 1.5 feet Temperature: 68F Velocity: 47 ft/s C8 Emission Rate: 0.16 Ib/yr (1.83xl0"5 Ib/hr) The C8 emission rate is based on the maximum production rate for the Telomers area. The basis for the emission calculations is shown in the attached document titled "Calculation of Emissions from the Telomers Process." To determine the impacts of building downwash on dispersion from the stack, the Screen3 model uses only the dominant downwash structure as an input to the model. The dominant downwash structure can be found by first locating all buildings which have an area of influence encompassing the stack. The area of influence is defined by EPA as a distance of five times the lesser of the height or the maximum projected width of the building. Once all of the potential downwash structures are located, the dominant downwash structure is determined by calculating the GEP (Good Engineering Practice) stack height for each building. The building with the greatest GEP stack height is the building that should be included in the Screen3 model. The GEP stack height is calculated according to the following formula: H=h+1.5L where H = GEP stack height h = building height L = lesser of the building height or maximum projected width 6/11 Section 2 11/12/04 Draft: Prepared at the Request of Council The following table shows the buildings that were included in the downwash analysis and their respective GEP's: "This building has the highest GEP and was used in the model. Modeling Results Using the above data the Screen3 model predicted a maximum off-site ground level concentration of 0.0023 p-g/m3at a distance of 289 feet from the source. This concentration is based on a 1-hour averaging time. To convert a Screen3 model result to an annual average, EPA guidance directs the user to multiply the predicted Screen3 concentration by a value of 0.05. This gives an annual average concentration of 1.15x10~4 jig/m3. This concentration is four orders of magnitude below the recommended ambient air concentration level in the community. A copy of the Screen3 model output is attached. 7/11 Section 2 11/12/04 Draft: Prepared at the Request of Council 08/01/03 *** SCREEN3r MODEL RUN *** *** VERSION DATED 96043 *** 13:17:07 WASHINGTON WORKS TELOMERS EMISSIONS ** 0 SIMPLE TERRAIN INPUTS: SOURCE TYPE EMISSION RATE (G/S) STACK HEIGHT (M) STK INSIDE DIAM (M) STK EXIT VELOCITY (M/S)= STK GAS EXIT TEMP (K) = AMBIENT AIR TEMP (K) RECEPTOR HEIGHT (M) URBAN/RURAL OPTION BUILDING HEIGHT (M) MIN HORIZ BLDG DIM (M) = MAX HORIZ BLDG DIM (M) = POINT .230576E-05 25.9080 .4572 14.3256 293.1500 293.0000 .0000 RURAL 29.2608 36.5760 59.4360 THE REGULATORY (DEFAULT) MIXING HEIGHT OPTION WAS SELECTED. THE REGULATORY (DEFAULT) ANEMOMETER HEIGHT OF 10.0 METERS WAS ENTERED. BUOY. FLUX = .004 M**4/S**3; MOM. FLUX = *** FULL METEOROLOGY *** ********************************** *** SCREEN AUTOMATED DISTANCES *** 10.719 M**4/S**2. *** TERRAIN HEIGHT OF *** 0. M ABOVE STACK BASE USED FOR FOLLOWING DISTANCES DIST (M) CONC (UG/M**3) STAB U1 OM (M/ S) us TK (M/ S) MIX :HT (M ) PL1UME HT (M) SH3MA Y (M) SI'GMA Z (M) DWASH 10. .0000 0 .0 .0 .0 .00 .00 .00 NA 100. 2091E-02 6 1 .0 1 .7 10000 .0 26 .21 4 .07 18 .28 SS 200. .1291E-02 6 1 .0 1 .7 10000 .0 26 .21 7 .73 24 .04 SS 300. .8797E-03 6 1 .0 1 .7 10000 .0 26 .21 11 .23 30 .17 SS 400. .6730E-03 6 1 .0 1 .7 10000 .0- 26 .21 14 .64 30 .53 SS 500. .5466E-03 6 1 .0 1 .7 10000 .0 26 .21 17 .97 30 .89 SS 600. .4609E-03 6 1 .0 1 .7 10000 .0 26 .21 21 .24 31 .25 SS 700. .3988E-03 6 . 1 .0 1 .7 10000 .0 26 .21 24 .46 31 .60 SS 800. .3517E-03 6 1 .0 1 .7 ,10000 .0 26 .21 27 .63 31 .95 SS 900. .3147E-03 6 1 .0 1 .7 10000 .0 26 .21 30 .78 32 .29 SS 1000. .2848E-03 6 1 .0 1 .7 10000 .0 26 .21 33 .88 32 .63 SS 1100. .2601E-03 6 1 .0 1 .7 10000 .0 26 .21 36 .96 32 .96 SS 1200. .2394E-03 6 1 .0 1 .7 10000 .0 26 .21 40 .01 33 .29 SS 1300. .2217E-03 6 1 .0 1 .7 10000 .0 26 .21 43 .04 33 .62 SS 1400. .2065E-03 6 1 .0 1 .7 10000 .0 26 .21 46 .05 33 .94 SS 1500. .1932E-03 6 1 .0 1 .7 10000 .0 26 .21 49 .03 34 .26 SS 1600. -1815E-03 6 1 .0 1 .7 10000 .0 26 .21 51 .99 34 .57 SS 1700. .1711E-03 6 1 .0 1 .7 10000 .0 26 .21 54 .94 34 .89 SS 8/1 1 Secti(3112 11/12/04 Draft: Prepared at the Request of Council 1800. 1900. 2000. .1618E-03 .1534E-03 .1459E-03 1.0 1.7 10000.0 26.21 57.87 35.20 SS 1.0 1.7 10000.0 26.21 60.78 35.50 SS 1.0 1.7 10000.0 26.21 63.68 35.80 SS MAXIMUM 1-HR CONCENTRATION AT OR BEYOND 10. M: 88. -2254E-02 6 1.0 1.7 10000.0 26.21 3.65 17.65 SS DWASH= DWASH=NO DWASH=HS DWASH=SS DWASH=NA MEANS MEANS MEANS MEANS MEANS NO CALC MADE (CONC = 0.0) NO BUILDING DOWNWASH USED HUBER-SNYDER DOWNWASH USED SCHULMAN-SCIRE DOWNWASH USED DOWNWASH NOT APPLICABLE, X<3*LB **************************************** *** REGULATORY (Default) *** PERFORMING CAVITY CALCULATIONS WITH ORIGINAL SCREEN CAVITY MODEL (BRODE, 1988) **************************************** *** CAVITY CALCULATION - CONC (UG/M**3) = CRIT WS @10M (M/S) = CRIT WS @ HS (M/S) = DILUTION WS (M/S) = CAVITY HT (M) = CAVITY LENGTH (M) = ALONGWIND DIM (M) = 1 *** -8839E-03 1.27 1.54 1.00 38.48 67.12 36.58 *** CAVITY CALCULATION - CONC (UG/M**3) = CRIT WS @10M (M/S) = CRIT WS @ HS (M/S) = DILUTION WS (M/S) = CAVITY HT (M) = CAVITY LENGTH (M) = ALONGWIND DIM (M) = 2 *** .1001E-02 2.37 2.87 1.44 32.60 48.77 59.44 **************************************** END OF CAVITY CALCULATIONS ***************^ ^*******i *************************************** *** SUMMARY OF SCREEN MODEL RESULTS *** *************************************** CALCULATION PROCEDURE MAX CONC (UG/M**3) DIST TO MAX (M) TERIIAIN HT (M) SIMPLE TERRAIN 2254E-02 88. 0. BLDG. CAVITY-1 .8839E-03 67. -- (DIST BLDG. CAVITY-2 .1001E-02 49. ---- (DIST CAVITY LENGTH) CAVITY LENGTH) r*********************************** ** REMEMBER TO INCLUDE BACKGROUND CONCENTRATIONS ** *************************************************** Note; The off-site concentration referenced in the report is the concentration shown for simple terrain. This concentration is valid for receptor locations near the fenceline since this area is relatively flat. Although the model predicted a building cavity concentration which is higher than the predicted value for simple terrain, this concentration will occur on site and is not applicable for this analysis. 9/11 Section 2 11/12/04 Draft: Prepared at the Request of Council Section 3 Calculation of Emissions from the Telomers Process Emissions of perfluoro-octanoic acid (PFOA) from processes on the Chambers Works and Washington Works sites were calculated based on analytical data which suggests that this compound is present in the telomers mixture. Analytical data was provided to me at various points in processing of telomer products and this data was used to determine vapor phase emissions from a variety of vessels during process steps that include filling, evacuating, distilling, reacting, homogenizing and packing out. This document is to serve as an explanation of the basis for such calculations, the assumptions made and areas of uncertainty. Basic Physical Property Data The primary property determining vapor phase composition above a mixture containing PFOA is the vapor pressure. Recent measurements of the vapor pressure (2003) were used in all the emission calculations and is best described by the following equation: Ln (VP) in psia = 12.7965 - 3388.046 / (T oK - 130.441) The vapor pressure of PFOA is quite low, and when combined with the extremely low levels of PFOA in the telomers the emission rates will typically be quite low. All other physical property data used were from estimates of the various telomer products and/or intermediates that are normally processed at the two sites. Vapor Phase Composition Calculations The composition of the vapor phase during all processing steps was determined by calculating the partial pressure of the PFOA above the organic solution using Raolt's law and the ideal gas law. The use of Raolt's law implies that the liquid solution is ideal and that the activity coefficient of PFOA out of telomers is 1.0. For similar compounds this would normally be a reasonable first pass assumption, however there is no measured vapor-liquid equilibrium data to verify that this is indeed the case. As a result the partial pressure could be greater (positive deviation systems) or less (negative deviation systems) than calculated by Raolt's law. The use of the ideal gas law implies that all processing steps are at low to medium pressures (below 100 psig). This is indeed the case, as the highest pressure encountered in processing is 55 psig. The calculation of vapor composition also assumes that the vapor is primarily nitrogen such that the molecular weight of the vapor is approximately 28 Ib/lbmole. Based on processing temperatures and available vapor pressure data this is a reasonable assumption. The mole fraction of PFOA in the vapor is therefore calculated as follows: y = x (VP) / Ptot where y and x represent mole fractions of PFOA in the vapor and liquid phases respectively. Total pressure of the system is represented as Ptot and VP represents the vapor pressure of PFOA at the liquid temperature. lo/n Section 3 11/12/04 Draft: Prepared at the Request of Council Liquid Phase Composition Calculations The liquid phase composition was determined by analysis and is altered only when additions are made to the vessel contents. To be conservative, during a distillation or evacuation step (low boiler removal) it was assumed that the composition of any vapor leaving the condenser was based on the partial pressure above the vessel contents at the condenser temperature. Since the PFOA content in the condenser liquid will be lower than in the vessel this assumption may be quite conservative in estimating the emissions. During processing steps in which other materials are added to the vessel the vapor displaced is assumed to be at the partial pressure prior to dilution. Subsequent process steps take credit for any dilution effects provided by addition of other materials. When two liquid phases are present (water additions) there is no credit taken for dilution and the liquid phase composition is assumed to remain unchanged. This results in a conservative estimate for the emissions estimates. During processing steps if there was analytical data for PFOA content at that step, it was this value which was provided as input to the calculations. Vapor Flow Rate Calculations During any filling step it was assumed that the vapor displaced was equal in volume to the liquid charged to the vessel and in equilibrium with the liquid. During steps in which the vessel pressure was changed, such as in pressurization and venting or evacuation, the mass of vapor discharged was calculated using the ideal gas law and the vapor space of the particular vessel as follows: Total Mass of Vapor Discharged = V AP MW / (RT) The mole fraction of PFOA in the vapor was determined at the initial pressure and instantaneous equilibrium assumed. Pressure let-downs were assumed to occur much faster than the mass transfer rate to the vapor during venting and/or evacuations. Total Annual Emission Calculations Emissions were calculated based on the operating procedures for each step of the typical area processes. These total emissions were then divided by the total batch time to arrive at an hourly rate of emission and multiplied by 8760 hours/year to arrive at an annual rate. For those processes that operated on a continuous basis the emission rate calculated was multiplied by 8760 hours/year to arrive at the annual emission rate. The following represents the total emissions calculated for the telomers area of Chambers Works and Washington Works. Chambers Works Washington Works 0.9 Ibs/yr 0.16 Ibs/yr 11/11 Section 3 11/12/04