Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports (2012)

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.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports 30 CHAPTER 5: METHODOLOGY This chapter describes the methodology developed to assess the principal sources of PM2.5 emissions and the contributions of these sources to local PM2.5 • Emissions Inventories – The purpose of an emissions inventory is to quantify the amounts (i.e., total mass) of air emissions by emission source. The emissions in the ACRP 02-23 project are reported in kilograms or tons. The emissions inventories were separated for the base case by aircraft operating mode, aircraft type, non-volatile versus volatile emissions, GSE type and fuel, and other sources, as a means of comparing case study airports and evaluating alternative fuels. The sources of emissions included in the inventories are identified as aircraft, GSE, APUs, on/off-airport road vehicle operations and airport-related stationary sources (e.g., boilers, generators). concentrations, for each of the five case study airports, for the base case and the alternative fuel scenarios. Further details of the methodology are contained in Appendix D. The principal components of the methodology comprise emissions inventories, atmospheric dispersion modeling, and air quality monitoring data. • Atmospheric Dispersion Modeling – The purpose of the atmospheric dispersion analysis is to convert the emissions inventory results to ambient (i.e., outdoor) concentrations of PM2.5 at locations (i.e., receptors) located at airport public access points, along the airport perimeter and in the airport vicinity. Dispersion models are used to calculate the movement of the emissions due to meteorological conditions (e.g., wind speed and direction) and the resultant ambient concentrations at receptors. The estimated concentrations in the ACRP 02-23 project are reported in micrograms per cubic meter of air (µg/m3 • Ambient Monitoring Data – The purpose of the ambient monitoring data is to determine background concentrations, to account for the contributions from non-airport regional and natural sources of PM ). 2.5 The following list contains the key reference documents on which the base case analyses were based, followed by the analysis year used. Further details are outlined in . The background concentrations are added to the atmospheric dispersion modeling results to obtain a total air pollution concentration. Table 12 in Appendix D and in the following sections. • Hartsfield-Jackson Atlanta International Airport 2008 Air Emissions Inventory (ATL, 2010) – analysis year 2008 • Southern Nevada Supplemental Airport Environmental Impact Statement Air Quality Technical Report (LAS, 2010) – analysis year September 2007 to August 2008, plus updated 2009 stationary source data (LAS, 2011) • Manchester-Boston Regional Airport 2007 Air Emissions Inventory (MHT, 2009) – analysis year 2007 • Philadelphia International Airport Capacity Enhancement Program Air Quality Technical Report (PHL, 2010) – analysis year 2004 • San Diego International Airport Air Quality Management Plan Baseline Emissions Inventory (SAN, 2009) – analysis year 2008

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports 31 EMISSIONS INVENTORY FOR BASE CASE Where available, airport-specific data were used to develop the emissions inventory and dispersion modeling analyses for the base case. Data included aircraft fleet mix, airfield taxi times, GSE fleet mix and operating times, APU usage rates, runway use, airfield operational profiles (quarter hourly, daily, and monthly), atmospheric mixing heights, meteorological conditions, receptor locations, and background PM2.5 Within each analysis year and for each airport, all the data elements were kept consistent. For example, the analysis year for ATL was 2008, which includes 2008 data for aircraft, road vehicles, meteorological conditions, and background PM concentrations. 2.5 EDMS is the FAA required model for assessing airport-related air quality impacts. Therefore, to estimate airport-related emissions, the data discussed above were used with FAA’s EDMS (Version 5.1.2) (U.S. FAA, 2009) and its internal databases. This model version was the most recent available at the initiation of the ACRP 02-23 project. Where airport-specific data were not available, standard EDMS defaults or professional judgment were used. Appendix D discusses the general EDMS and AERMOD control options, APU operating time, choice of geographic locations for calculating the impact (i.e., receptors), terrain data used, meteorological data used, mixing height, and monitoring data used. concentrations. These analysis years and datasets were used for the base case and for the alternative fuels scenarios. To enable more detailed emission factors to be generated and used (i.e., for the separation of non-volatile and volatile emissions) and to enable comparison with previous studies at the case study airports, EPA’s MOBILE6.2 model (all case study airports, except SAN) (U.S. EPA 2003b) and the California Air Resources Board (CARB) (2006a) Motor Vehicle Emission Factor Model (EMFAC2007) model (used for SAN) were used outside EDMS to generate emission factors for road vehicles. The use of MOBILE6.2 outside of EDMS allowed region specific vehicle mixes to be used (i.e., the proportion of different types of vehicles such as automobiles, trucks, and vans). These emission factors were used as inputs to EDMS to enable generation of emissions for roadways and parking facilities. It should be noted that the definition of “road vehicles” varies between airports. For the purposes of the ACRP 02-23 project, the term refers to vehicles that access the public roadways within the on-airport road network and external road network. Therefore, this includes some airport-related rolling stock. The boundary for road vehicles was dependent on the extent that the data were available. In most cases, the original data used had been compiled for analysis in compliance with NEPA for an airport development project and so conformed to the relevant boundary issues related to that particular airport (i.e., the spatial extent to which roadway sources were included). To enable cross comparison between different airports, road vehicle emissions are presented in Appendix E in grams per vehicle mile. NONROAD (U.S. EPA, 2008) and OFFROAD2007 (CARB, 2006b) models were used for SAN to generate GSE emission factors. These models do not incorporate any separation of non- volatile and volatile emissions, so no separation of non-volatile and volatile GSE emissions was undertaken.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports 32 Non-Volatile and Volatile PM2.5 The emissions inventories in the ACRP 02-23 project report PM Emissions 2.5 For PM emissions as non-volatile and volatile, where information is available to allow that distinction (i.e., for aircraft main engines and road vehicles). 2.5 The FOA3 and FOA3a methodologies use smoke number to estimate non-volatile emissions, and use hydrocarbon engine emission factors and sulfur fuel content to estimate volatile particulate emissions. In EDMS, the emissions are not reported separately for non-volatile and volatile PM emissions from jet aircraft, the FAA has developed (with assistance from others and EPA concurrence) the FOA3a methodology, based on the ICAO agreed FOA3 (ICAO 2007 and 2011). Both the FOA3 and FOA3a methodologies are incorporated directly into EDMS, where FOA3a is used for U.S. airports. 2.5. Therefore, the FOA3a equations were used to develop a methodology for separating out volatile PM2.5 (i.e., into PM2.5 Road vehicle PM originating from sulfur, hydrocarbons and lubricating oil) outside of EDMS (discussed further in Appendix D). This also allowed separate scaling factors to be applied to the non-volatile, hydrocarbon and sulfur components of the jet aircraft emissions when calculating the alternative fuel emissions. 2.5 emissions were partitioned into non-volatile and volatile particulate matter, based on the available information developed in the MOBILE6.2 and EMFAC emission factor models. Non-volatile and volatile are not separated out for gasoline vehicles in MOBILE. Therefore, volatile PM2.5 emissions from road vehicles include organic carbon and sulfates (SO4 EMISSIONS INVENTORY FOR ALTERNATIVE FUEL SCENARIOS ) from diesel. The justification for the alternative fuel and source combination scenarios are discussed in Chapter 4 and Appendix C. HRJ fuels have not been analyzed separately in the ACRP 02-23 project due to lack of data at the time of study (Whitefield et al., 2011) as well as the similarity of the chemical structure of HRJ fuels to FT fuels, which are included. To calculate the alternative fuel emissions for each fuel and source combination, scaling factors were applied. These scaling factors were related to: • The ratio of the scenario emissions to the base case emissions per source and fuel combination. • The alternative source fuel penetration (i.e., whether a 100% of that source use the fuel). In terms of road vehicles, the alternative fuel scenarios were only considered for on-airport roadways and parking (i.e., those under airport control and ownership). The base case and scenario emission calculations include some road vehicle emissions from vehicles not owned by the airport as it was not possible to separate the data for airport-owned and other road vehicles (e.g., passenger travel to airport). Therefore, emission results were separated spatially according to whether the roadways are on-airport or off-airport. This approach maintains consistency with how airports are typically preparing their emissions inventories (for criteria pollutants and greenhouse gases), as some airports may have the ability to control vehicular use on some on- airport roadways. It is worth noting that, while not calculated due to lack of suitable data, road

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports 33 vehicles that drive on-airport are also likely to drive off-airport. Therefore changes in fuels used by these road vehicles would effect on-airport and off-airport emissions. Ratio of Scenario Source Type Emissions to Base Case The ratio of emissions of the alternative fuel versus the base fuel for each relevant source and fuel type were calculated. These ratios were used to scale the base case emissions to the alternative fuel scenario emissions, assuming all relevant sources use the alternative fuel as indicated in the equation below. The actual ratios used are shown in Table 13, Table 14, and Table 15 in Appendix D, though for some fuel and source combinations the alternative fuel emissions have been calculated using either EDMS databases (e.g., GSE) or MOBILE6.2 (e.g., for road vehicles) rather than using a specific ratio. AFE = Base x AF Where: AFE = Alternative fuel emissions Base = Base case emissions for source and fuel type AF = Alternative fuel ratio Alternative Fuel Penetration It is not always feasible for all emission sources of a particular type to use one particular alternative fuel (e.g., not all diesel-fueled road vehicles will use B100). Therefore, a penetration factor, P (ranging from 0 to 1) was applied to scale the emissions for each alternative fuel and source type. The remainder (1–P) of the non-penetrated sources’ emissions were assumed to be as per the base case calculations. For example, if only 2.6% of diesel-fueled road vehicles use B100, then the penetration factor (P) applied is 0.026 for the B100 road vehicle emission calculation, and 0.974 (1–P) for the base case road vehicle emission calculation: Scenario = (AFE x P) + ((1–P) x Base) Where: Scenario = Scenario emissions per source and fuel type Base = Base case emissions for source and fuel type AFE = Alternative fuel emissions for source type P = Penetration factor For each of the source and fuel combinations, a number of penetration options have been considered. For drop-in fuels that can be used in existing aircraft, such as FT (natural gas) or FT (coal), it is assumed that only one type of jet fuel would be available, as airports are unlikely to have multiple jet fuels available. Therefore, it has been assumed that 100% of the aircraft fleet operating at these airports would be refueling on the alternative jet drop-in fuel, and a penetration factor of 1 has been assigned. The 91/96UL AvGas fuel can, in theory, be used as a drop-in fuel in many piston-engine aircraft (and is the main fuel in many countries). However, not all piston- engine aircraft in the U.S. are certified to use it. As 91/96UL was only considered as an extension to the sensitivity analysis, a hypothetical penetration factor of 1 was assumed.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports 34 The use of the EDMS databases to recalculate the emissions for some GSE scenarios allowed only the GSE with a relevant replacement to be considered (e.g., if no LPG alternative exists, it is not replaced). Therefore, a 100% uptake of the related GSE scenarios has been assumed to be feasible. Similarly, it is feasible that all gates at an airport could be fitted with pre-conditioned air and electric ground power, so the reduction in APU time was assigned a hypothetical 100% uptake. For non-drop-in fuels (e.g., natural gas for road transport), a much lower penetration has been considered, based on different datasets (e.g., U.S. Energy Information Administration (2011) fuel use projections in 2020) and expert knowledge. Other penetration factors for the different scenarios and the sources and assumptions made are outlined in Table 16 in Appendix D. ATMOSPHERIC DISPERSION MODELING ANALYSIS For the purpose of evaluating the potential impacts of airport emissions on local PM2.5 concentrations in the vicinity of the case study airports, the ACRP 02-23 project used EDMS to generate the initial AERMOD input file. This input file was then edited to allow further source separation of the dispersion modeled results (i.e., by terminal area/concourse, aircraft mode, and internal and external roadways) and AERMOD (Version 09292) run outside of EDMS. As a theoretical example, the total concentration at a specific receptor may be 10 µg/m3 and can be separated into individual source contributions of 2 µg/m3 from aircraft engines, 4 µg/m3 from Terminal A and 2 µg/m3 from Terminal B (where terminal sources are related to GSE and APU), 1 µg/m3 from roadways and 1 µg/m3 from stationary sources. Furthermore, aircraft-related concentrations can be separated by operating mode (taxi, approach, takeoff and climb-out), which is related to aircraft thrust. The results were calculated for averaging periods that reflect the National Ambient Air Quality Standards (NAAQS) for PM2.5 (i.e., the annual average and the 98th Development of Impacts for Alternative Fuels percentile of the 24-hour average). Separating out the different source contributions allows different scaling factors to be applied to each different source contribution. Taking the theoretical example above, if Terminal A emissions (combined emissions from APU and GSE) are found to reduce by a factor of 0.5, then the contribution from Terminal A reduces from 4 µg/m3 to 2 µg/m3. If no other change in emissions is assumed, then the total concentration at the same receptor for this scenario is now 8 µg/m3 (i.e., a reduction of 2 µg/m3 Table 19 ). Similarly, for each scenario, the relative change in emissions between the base case and the scenario (scenario/base) was applied to the relevant source contribution’s dispersion modeled results for each receptor point (i.e., individual locations) as indicated in and Table 20 in Appendix D. This enabled a scenario impact to be calculated for each receptor. Note that the GSE and APU emissions were dealt with on a terminal-by-terminal or concourse-by-concourse basis due to the emissions generated by EDMS being associated with more than one source group (i.e., GSE and APU) and to allow incorporation of the spatial distribution of these activities. Therefore, for each terminal/concourse, the scenario emissions (from APU and GSE) were calculated and the change relative the base case applied. This methodology was applied to the base case annual and the eighth-highest 24-hour (i.e., 24- hour 98th percentile) impact results for each receptor. In theory, the impact of the eighth-highest

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports 35 24-hour scenario impact could occur during a different hour than the base case impact. However, for the ACRP 02-23 project it was assumed to be the same hour. The piston-engine and turboprop aircraft scenarios (which include turboshafts) have emission results reported in Chapter 6 and Appendix E only as part of the sensitivity analysis. This is due to EDMS not incorporating the calculations of these two source types for PM2.5 SENSITIVITIES OF ANALYSIS . EDMS does not typically include PM2.5 emission results for piston-engine, turboprop, and turboshaft aircraft as there are no FAA accepted emission factors for these aircraft. Therefore, a number of alternative methodologies were used in the ACRP 02-23 project to estimate emissions for those aircraft for which EDMS does not estimate PM2.5 emissions. A sensitivity analysis on the impacts of these methodologies upon aircraft emissions was conducted and is reported in Chapter 6 and Appendix E. For those aircraft where there is no appropriate alternative methodology to calculate emissions, emissions were scaled based on the average emissions for that aircraft size. This methodology for determining PM2.5 Alternative Fuels emission factors and the sensitivity analysis for aircraft engines not covered by EDMS was reviewed and discussed with the FAA as part of the ACRP 02-23 project. The methodology is described in detail in Appendix D. As part of the sensitivity analysis, the base case emissions from turboprop (including turboshaft) and piston-engine aircraft were scaled in line with the scaling ratios in Table 14 in Appendix D with an assumed 100% penetration. This enables those airports with a high proportion of turboprop, turboshaft, and piston-engine aircraft to access the potential change in emissions associated with the use of alternative fuels for those aircraft types.