Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
21 Although emissions inventory scaling as described in Chap- ter 3 illuminates the changes in primary PM and precursors for secondary PM such as nitrogen oxides and sulfur dioxide, it does not capture the end changes in particulate matter con- centration. Capturing these changes in concentration, which ultimately affect human health, requires a full atmospheric chemistry model. To demonstrate the changes in concentration and changes in emission inventories, the Hartsfield-Jackson Atlanta International Airport (ATL), located in Atlanta, Georgia, was modeled. ATL was chosen for its size, its location in a nonattainment area, and importantly, to leverage previous research efforts on ATL. In addition to being the busiest airport in the world, ATL is also located in both PM2.5 and ozone non- attainment areas (Environmental Protection Agency, 2008). 4.1 Methodology In order to model the changes in pollutant concentrations, a series of three programs was used. First, the EDMS was used to create an emissions inventory for aircraft and GSE. For the air quality modeling, the months of June and July were used. Second, the emission inventories were reformatted in SMOKE (Sparse Matrix Operator Kernel Emissions). The reformatted emissions inventories were then combined with a dispersion model, an atmospheric chemistry model, background inven- tories, and meteorological conditions in CMAQ (Community Multiscale Air Quality modeling system). Finally, the CMAQ output was processed using the EPA program MATS (Modeled Attainment Test Software). MATS adds the particle-bound water to the ionic concentrations computed by CMAQ. EDMS is the required tool for airport emissions inventory compilation. EDMS calculates GSE emissions by using emis- sion factors provided by EPAâs NONROAD model and con- siders the regulations in effect for the year being modeled, engine age, and horsepower. EDMS computes aircraft emis- sions by using a combination of the ICAO Engine Exhaust Emissions Databank, thrust calculations obtained through SAE-AIR-1845, fuel flow rates from Eurocontrolâs Base of Aircraft Data (BADA), and Boeing Fuel Flow Method 2. The model year within EDMS was set to 2011 to force a low sulfur content fuel for the GSE. This choice was made because using an earlier year with high sulfur content fuel would overstate the benefits of an alternative jet fuel, and an alternative jet fuel will not realistically be available in large quantities while the high sulfur content diesel fuel is in use; therefore, the reduc- tions in SOx would be overstated. For the inventory scaling, the full annual inventory for both GSE and aircraft was used. CMAQ is an EPA-developed, three-dimensional Eulerian chemical-transport model. The model has three main com- ponents: a meteorological model system, an emissions model, and a chemistry-transport modeling system. For each time step and grid cell, CMAQ calculates the change in chemical concentration based on advection, diffusion, chemical for- mation, removal of each species, and the given emissions. Previous modeling of ATL provided aircraft emission inventories for scaling and the necessary information for sim- ulating the change in air quality (Arunachalam et al., 2008). This included the EDMS emissions inventories processed by SMOKE. The work by Arunachalam et al. used a four- kilometer grid size to examine the relative impact of aircraft at ATL on the region. Details regarding the analysis are pro- vided in Arunachalam et al. (2008) and Donohoo (2010). 4.2 GSE Vehicle Inventory The GSE emissions inventory was created with EDMS using the ATL aircraft schedule. To assess model accuracy, a comparison was made to a partial GSE inventory from Delta Airlines. For ATL, EDMS models 684 individual GSE, 445 of which are diesel. Because EDMS does not record emissions for individual GSE, the emissions cannot be scaled on a unit- by-unit basis. Instead, it is assumed that the proportion of GSE emissions from diesel GSE is directly related to the num- ber of GSE. Although diesel GSE comprise approximately 66% of the total GSE inventory by fuel type, electric GSE are C H A P T E R 4 Air Quality Assessment for a Selected Airport
not responsible for any emissions at the airport. Therefore, the emissions were divided between the gasoline- and diesel- powered GSE. Of this portion, diesel GSE makes up 78%. Unlike the aircraft flight schedule, which provided actual flights and aircraft used, the GSE vehicle inventory was pro- duced by an internal EDMS algorithm. Although the actual GSE vehicle inventory for all of ATL is unknown, the GSE inventory for Delta was provided for analysis. The Delta GSE vehicle inventory differs from the EDMS inventory both in composition and number. The Delta vehicle inventory con- tains 2,251 individual pieces of equipment compared to the EMDS vehicle inventory, which has 684. The Delta vehicle categories were mapped into the EDMS categories, but 372 vehicles in the Delta GSE fleet did not have a correspon- ding EDMS category. As can be seen in Figure 6, there are also several cate- gories for which vehicles exist in one vehicle inventory but not another. For example, the EDMS inventory includes 51 hydrant trucks while the Delta inventory has none. One of the greatest disparities is that the Delta inventory contains 918 baggage tractors while the EDMS inventory contains 61. Within the baggage tractor category, the Delta inventory indi- cates that 58% of baggage tractors are diesel powered while the EDMS inventory assumes that all are gasoline. For some vehi- cle categories missing from the Delta inventory, such as cater- ing trucks, it is likely that Delta outsources the task to an outside company and thus does not own or track the vehicles. Although the gross number of vehicles varies dramatically, it is difficult to compare the two inventories because the man- ner in which the vehicles operate is unknown. For example, the vehicles in the EDMS inventory could be modeled as oper- ating continuously throughout the day while the Delta inven- tory could contain units that are no longer operated or are only operated sporadically. Nonetheless, the differences in types of units indicate that the EDMS default modeling may not accu- rately capture the GSE population at ATL. 22 0 25 50 75 100 125 150 175 200 Air Conditioner Air Start Aircraft Tractor Baggage Tractor Belt Loader Cabin Service Truck Cargo Loader Cargo Tractor Cart Catering Truck Deicer Fork lift Fuel Truck Ground Power Unit Hydrant Truck Lavatory Truck Lift Passanger Stand Service Truck Sweeper Water Service Cab SVC Truck ATLDelta ATL EDMS 918 Figure 6. Delta and EDMS GSE vehicle inventories. The number of baggage tractors in the Delta inventory (918) exceeds the range covered in the chart [from Donohoo (2010) with permission].
Due to the discrepancies in vehicle inventories, a rudimen- tary check was conducted on the total GSE fuel consumption. Although EDMS does not calculate fuel burn, there is a linear relationship between SOx emissions and fuel burn, which can be used to estimate fuel consumption. According to EDMS, the total mass of SOx produced by the GSE at ATL based on the 2002 aircraft schedule is 7,500 grams, which equates to a fuel consumption of approximately 61 million gallons. An alternative estimate of GSE fuel use is 0.25 to 0.30 gal- lons of diesel per enplaned passenger. This would result in 9.6 to 11.6 million gallons fuel based on 36,639,600 enplaned passengers at ATL in 2002 (Hartsfield-Jackson Atlanta Inter- national, 2008). This represents a potential uncertainty factor of 6 in fuel use and emissions between the EDMS methodol- ogy and an independent estimate of GSE fuel use. Due to this uncertainty as well as the uncertainty associated with both the NONROAD model and the scaling factors, GSE were not included in the full air-quality model. 4.3 Emissions Inventory The aircraft and GSE emissions from EDMS were scaled according to the relationships outlined in Chapter 3 (summa- rized in Appendix D) to examine various GSE and aircraft fueling scenarios as outlined in Table 7. The first scenario models the aircraft emissions with Jet A and GSE emissions from ULSD as it will be operating in 2011 when the ULSD standard comes into full effect. The second scenario is a low- sulfur scenario with aircraft using USLJ and GSE using USLD. The third scenario considers a potential single-fuel airport with both GSE and aircraft using a 50-50 blend of SPK and ULSJ. The last scenario considers the lowest emission case pos- sible, with GSE producing no emissions (for example being converted to all electric or fuel cell) and aircraft burning 100% SPK fuel. These scenarios are summarized in Table 7. As shown in Figure 7, EDMS predicts that aircraft are responsible for more of the emissions affecting air quality at ATL than GSE. Across all of the scenarios, GSE produced 6% or less of the NOx emissions. Because scenarios 1 to 3 assume the GSE use a ULS fuel instead of conventional diesel, the GSE SOx emissions are negligible in comparison to aircraft SOx emissions. The contribution of GSE to primary PM emis- sions varies with the scenario; GSE contribute 11% of the PM in the baseline scenario (Scenario 2 with ULSJ/ULSD), and the contribution would increase with the use of 50-50 SPK and ULSD to 16% (Scenario 3). If both GSE and aircraft were operating on a 50-50 blend of SPK and ULSJ, then GSE would comprise 14% of the emissions (SPK/ULSJ). GSE are respon- sible for a larger percentage of CO than PM, SOx, or NOx. With the use of ULSD, GSE are responsible for 37% of all CO emissions, but this is reduced to 30% with the use of the SPK- ULSJ blend. The changes in aircraft NOx and CO emissions were assumed to be only due to changes in fuel burn. The aircraft NOx emis- sions for ULSJ are reduced by the change in fuel burn 0.5%, while the NOx emissions for 100% SPK would be reduced by 2.1% and the 50-50 blend would be reduced by 1.3%. This result is likely conservative since preliminary studies show that NOx is reduced with SPK fuels. The SOx emissions from aircraft were reduced as a result of the change in fuel sulfur content. The 98% reductions shown in Figure 7 are a result of the change from 680 ppm to 15 ppm fuel sulfur content. The aircraft primary PM emissions from Figure 7 have been broken out by type within Figure 8 to explain the variation in reduction with the various fuels. ULSJ and SPK fuels have ultralow sulfur levels and therefore the emissions of PMS are also nearly zero. The PMFO and PMLO were both assumed to be unchanged with fuel composition; the PMFO reduction is likely erroneous since recent preliminary AAFEX results indicate a reduction in volatile gaseous emissions. As discussed in Section 3.2.3, each of these fuels will have varied emissions of non-volatile PM (PMNV). ULSJ is expected to have simi- lar PMNV emissions to conventional fuel, while a neat (i.e., 100%) SPK fuel should have a large PMNV reduction. The 50-50 blend fell in between these. The result is that the pri- mary PM emissions for ULSJ were 37% lower than Jet A, SPK were 72% lower, and the 50-50 blend were 56% lower than conventional jet fuel. 4.4 Ambient Particulate Matter Concentration Due to the uncertainty in the GSE inventories, only the emissions from the aircraft were modeled for air quality. These cases included aircraft emissions for a baseline scenario with Jet A, as well as three additional scenarios using ULSJ, a 50-50 blend of ULSJ and SPK, and a 100% SPK fuel. Thus, the aircraft inventories from the previous section were input to CMAQ with the output being a cell-by-cell concentration of ionic compounds that comprise particulate matter. These compounds consisted of ammonia, sulfates, nitrates, elemen- tal carbon, organic carbon, and crustal material. In addition to showing the region that was modeled within CMAQ, Figure 9 presents the ambient concentrations of PM2.5 without aviation activity. This background level should 23 Fuel Aircraft GSE Scenario 1 Jet A ULSD Scenario 2 ULSJ ULSD Scenario 3 50-50 blend SPK-ULSJ 50-50 blend SPK-ULSJ Scenario 4 100% SPK â Table 7. Fuel scenarios modeled for ATL emissions inventory analysis [from Donohoo (2010) with permission].
24 0 1,000,000 2,000,000 3,000,000 4,000,000 5,000,000 6,000,000 7,000,000 NOx (kg) 0 100,000 200,000 300,000 400,000 500,000 600,000 700,000 SOx (kg) 0 20,000 40,000 60,000 80,000 100,000 120,000 Primary PM 2.5 (kg) 0 2,000,000 4,000,000 6,000,000 8,000,000 10,000,000 CO (kg) Jet A/ULSD ULSJ/ULSD 50/50 SPK ULSJ 100 SPK/ - Aircraft GSE Figure 7. Scaled emissions inventories for GSE and aircraft based on EDMS analysis of ATL [from Donohoo (2010) with permission].
be kept in mind when reviewing aviationâs contribution since the background level is several orders of magnitude larger. The incremental contributions of aircraft emissions in each of the fuel scenarios are shown using an approximate radial distance, which is shown schematically in Figure 10. In this manner, all of the emissions in the grid cell that correspond to each radius number were averaged. This provides an approx- imate distance from ATL that should suffice for comparing the impact of alternative fuel use on ambient PM emissions from aviation. Distances within the circle shown in Figure 9 have been plotted in this manner. As expected, the composition of the ambient PM2.5 from aircraft emissions, shown in Figure 11, changes with fuel composition. The peak PM2.5 level of â¼0.6 ug/m3 that is due to aviation is two orders of magnitude smaller than the back- ground concentration. Each of these charts provides the con- tribution of each PM2.5 component to the total ambient PM2.5. With Jet A, the largest contribution to PM2.5 is due to sulfates. With each of the alternative fuels considered, the ultralow sulfur content of the fuels makes the sulfate contribution negligible. The largest contributor of PM2.5 for all three alter- native fuel scenarios is organics, followed by elemental carbon. The elemental carbon fraction changes as expected with the reduced inventories for the 50-50 blend and straight SPK. Across all scenarios, the nitrate contribution to PM2.5 is negli- gible, which is consistent with prior work (Arunachalam et al., 2008) for ATL; however, this is not uniformly observed. A nationwide analysis of the air quality impacts from aviation suggests nitrates may be the largest source of aircraft-related PM; see, for example, Brunelle-Yeung (2009). 25 Figure 8. Scaled aircraft primary PM emissions by species using EDMS [from Donohoo (2010) with permission]. Figure 9. Ambient concentrations of PM2.5 in the study area. The circle roughly denotes the location of ATL with the size of the circle corresponding to the radii covered in subsequent charts [from Donohoo (2010) with permission].
26 Figure 10. Grid definition for radial plots of ambient PM2.5 concentration [from Donohoo (2010) with permission]. Jet A ULSJ 50-50 Blend SPK Figure 11. Ambient PM2.5 concentration from aircraft for each of the four fuel compositions considered in this work, with the breakout of individual species [from Donohoo (2010) with permission].
