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 A-1 APPENDIX A: LITERATURE REVIEW This appendix discusses the following: • Airport sustainability plans, air quality and noise management reports • Monitoring of fine particulate matter with a diameter of less than 2.5 micrometers (PM2.5) in the vicinity of airports • Calculation of emissions at airports • Aircraft and auxiliary power unit (APU) emissions • Aircraft alternative fuels • Ground support equipment (GSE) emissions • GSE alternative fuels • Road vehicle emissions • Road vehicle alternative fuels • Other emission sources • Dispersion modeling at airports AIRPORT SUSTAINABILITY PLANS, AIR QUALITY AND NOISE MANAGEMENT REPORTS In response to airport environmental stewardship initiatives or recent requests from regulatory agencies and the general public, airports are increasingly preparing environmental sustainability plans and air quality management plans on a voluntary basis. These plans generally include air emission inventories, with PM2.5 emissions among the pollutants evaluated. Airports may also be driven to assess PM2.5 For example, Boston Logan International Airport prepares an annual environmental data report in which PM emissions to improve public relations, further sustainability initiatives, cut operational costs or adopt a proactive stance on air pollution. Similarly, noise management reports, sustainability plans and the data collected to support them can also be used as a basis for calculating airport emissions inventories. This is especially the case with the ongoing development of the U.S. Aviation Environmental Design Tool (AEDT) combined noise and air pollution model, which will replace the Federal Aviation Administration’s (FAA) Emissions and Dispersion Modeling System (EDMS) in the future. 2.5 emissions are computed for all airport-related sources (e.g., aircraft, APUs, GSE, stationary sources and road vehicles) (Massport, 2008). Under state law, the Rhode Island Airport Corporation (RIAC) also prepares an air emissions inventory for T.F. Green Airport, which includes an annual inventory of greenhouse gases (GHGs) and criteria air pollutants (RIAC, 2009). Similarly, in 2009, the San Diego County Regional Airport Authority (SDCRAA) prepared its first GHG and criteria pollutant emissions inventory for San Diego International Airport (SAN) as part of a memorandum of understanding with the California Attorney General’s Office (MOU, 2008).

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-2 MONITORING OF PM2.5 Several PM IN THE VICINITY OF AIRPORTS 2.5 air monitoring campaigns have been carried out at airports across the U.S. (SCAQMD, 2000a, 2000b; Fanning et al., 2007; Westerdahl et al., 2008; Hu et al., 2009; Massport, 2010; ENVIRON, 2008; RI DEM, 2008; Dodson et al., 2009 and BCAA, 2006), with the overarching goals of apportioning airport contributions to PM2.5 and assessing the potential impact of PM2.5 CALCULATION OF EMISSIONS AT AIRPORTS on nearby residences and public areas. To summarize, the overall findings appear to indicate that while airports may be contributing to emissions, their overall impact (ambient concentration) away from key sources diminishes rapidly and that other key sources such as road vehicles may have more of an impact on local air quality, depending on the relative locations of the sources of the emissions and the location of interest (i.e., ambient monitoring location). The following sections outline the methodologies typically used to estimate airport PM2.5 emissions, primarily in the U.S. context, but also with reference to other key studies. Each section deals with a particular source sector such as aircraft, APU, GSE, roadways, parking lots, and other ancillary sources such as stationary sources. In addition, each section also discusses methodologies relevant to the estimation of airport-related PM2.5 AIRCRAFT AND APU EMISSIONS from alternative fuels. An aircraft’s engine is its main propulsion unit and comes in four main types—turbofan, jet turbine, turboprop (including turboshaft), and piston. • Turbofan engines combine a gas turbine with a ducted fan and are primarily used on executive jets and larger, high-altitude passenger aircraft. They burn aviation kerosene, known as Jet-A in the U.S. • Jet turbine engines are similar to turbofans, but do not have a fan. They also burn Jet-A. • Turboprop engines combine a gas turbine with a propeller and are primarily used on short-range, medium-altitude executive aircraft, airliners, and helicopters. This kind of engine primarily burns Jet-A. • Piston-engines are reciprocating internal combustion engines used to power small, short- range, low-altitude general aviation aircraft. This kind of aircraft engine primarily burns leaded aviation fuel called AvGas. The most common grade is 100LL (low lead). APUs are small gas turbines that typically burn Jet-A. They are usually mounted at the rear of larger executive aircraft and airliners. These units supply electricity to operate electrical, hydraulic and air-conditioning systems when the main aircraft engines are not running. APUs are also used for main engine startup. The majority of aircraft use turbine technology and burn Jet-A (or similar) fuel. Whitefield et al. (2008) provides the following typical diameters of types of particulate matter emitted from aircraft turbine engines: • Non-volatile carbonaceous particulate matter (soot or black carbon) ranging from 0.02 µm to 0.06 µm • Volatile particulate matter ranging from 0.001 µm to 0.015 µm

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-3 Both of these types of particulate matter are typically classified as PM10, PM2.5, and PM0.1 Jet Turbines and Turbofans . The development of methods to estimate the emissions of particulate matter from aircraft engines is still at a relatively early stage when compared with other modes of transport. It relies primarily on the International Civil Aviation Organization (ICAO) regulatory compliance measurement of smoke emissions from aircraft engines for engine certification. Similar certification data are also available for hydrocarbon (HC), carbon monoxide (CO), and oxides of nitrogen (NOx Therefore, the ICAO’s Committee on Aviation Environmental Protection (CAEP) tasked the Society of Automotive Engineers’ (SAE) E-31 Aircraft Exhaust Emissions Measurement Committee (which comprises research institutes, engine manufacturers, and regulators) with developing a methodology. ) emission rates. Smoke number (SN) measurement is an imprecise, 50-year-old method that does not provide any data on particle size, type or size distribution and does not represent all of the particulate matter that have an impact on human health and the environment. Furthermore, the smallest particulate matter found in aircraft engine exhausts can penetrate the smoke filter used to measure SN, so the SN is more likely to be representative of larger smoke particles. In April 2002, the E-31 Committee issued a position paper (SAE, 2004) calling for the development of a set of aircraft engine particulate matter measurement recommendations covering: • Measurement at the engine exit • Characterization of non-volatile particulate matter • Exclusion of the characterization of volatile particulate matter In April 2003, this led to the publication of Aerospace Information Report (AIR) 5892, Nonvolatile Exhaust Particle Measurement Techniques, with AIR 5892 Revision A following in July 2004 (SAE, 2004). Over the past decade, a series of research projects in Europe and the U.S., including PartEmis (Petzold et al., 2005), Aircraft Particle Emissions eXperiment (APEX) (Kinsey, 2009) and SAMPLE (Petzold et al., 2009), have been completed, with additional testing in progress (e.g., SAMPLE II (European Aviation Safety Agency, 2009)). These experiments have helped to gain a better understanding of emissions at the engine exit plane and in the exhaust plume. This work has also been used to evaluate existing estimation methodologies. Also in 2003, a literature review of particulate matter estimation methodologies was conducted (Wasyon et al., 2003), leading to the development of the current particulate matter estimation methodology, the first-order approximation (FOA), based on the three most widely recognized studies at the time (Champagne, 1971; Whyte, 1982; Hurley, 1993). The FOA has evolved over time, with the current international version, FOA3, developed by the ICAO’s CAEP (ICAO 2007 and 2011). FOA3 is applicable to certified commercial aircraft engines above 26.7 kN of thrust. The current U.S. version is FOA3a, which is incorporated in the U.S. regulatory model for airport air quality, EDMS (U.S. FAA, 2009). EDMS assumes that all PM10 is PM2.5 and uses FOA3a only when appropriate SN data are present.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-4 The general consensus is that the FOA methodology is not sufficiently accurate and that work in the U.S. (including the U.S. Environmental Protection Agency (EPA), FAA, Volpe, and SAE E- 31 Committee) and Europe is ongoing to improve it (Volpe, 2010). Ultimately, it is intended that the FOA methodology will be replaced by a database of verified engine emissions with an aerospace recommended practice (ARP) for aircraft non-volatile particulate matter issued. This is expected to happen by December 2011 (Whitefield, in progress). FOA3a FOA3 was completed in late 2006. It was accepted in February 2007 for international use by CAEP and first published in the Airport Air Quality Guidance Manual (ICAO, 2007) and subsequently in 2011 (ICAO, 2011). It has been supplemented by FOA3a in the U.S. However, both methodologies are in use and have been incorporated into EDMS, with U.S. airports using the EPA’s approved FOA3a and non-U.S. airports using the ICAO/CAEP approved FOA3 (CSSI, Inc., 2009). The FOA3 methodology, including assumptions and derived equations, is discussed in detail in the publication Methodology to Estimate Particulate Matter Emissions from Certified Commercial Aircraft Engines (Wayson et al., 2009). FOA3a builds on the FOA3 methodology and was developed under PARTNER Project 15 Aircraft Impacts on U.S. Local and Regional Air Quality (Ratliff et al., 2009). It was completed in 2009. To summarize, the total particulate matter emissions from an engine are calculated by summing the volatile and non-volatile contributions: PMtotal = PMvol + PM The methodology also identifies the three main components of volatile particulate matter in aircraft engine exhausts as: nvol PMvols Based on the limited data available at the time, volatile particulate matter driven by nitrates was considered to be a small contributor to the total particulate matter and, as the residency time is short, they were not incorporated into the FOA3 methodology. = F(fuel sulfur content) + F(fuel organics) + F(lubrication oil organics) Non-volatile Contribution to Particulate Matter The link between non-volatile particulate matter emissions and SN is well established and the FOA3 and FOA3a (Wayson et al., 2009; Ratliff et al., 2009) methodologies reflect this by deriving estimates for PMnvol In FOA3a, this approach is applied to all aircraft engines (Wayson et al., 2009; Ratliff et al., 2009), but it is strictly only a suitable assumption for engines where the core flow and bypass flow are mixed before the engine exit plane. The majority of aircraft engines flying today mix from the SNs listed in the ICAO Engine Emissions DataBank (2010). In the non-volatile component, FOA3 and FOA3a allow for instances when the SN is measured with bypass air (the air that passes through the engine, but does not pass through the engine core). The engine bypass ratio, β, is used as a multiplier, in the form (1 + β), to estimate the exhaust volume.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-5 the flows after the exit plane. In engines with external mix, this multiplication factor is conservative as it increases the non-volatile primary particulate matter component. Fuel Sulfur Content Contribution to PM The quantity of sulfates produced in an engine exhaust is directly linked to the sulfur content of the fuel. Under ASTM International’s D1655 specification, aviation fuel (Jet-A) can contain up to 0.3% sulfur by mass (ASTM, 2010). Typically, the sulfur content in aviation fuel is considerably lower than this (as low as 0.05% (ICAO, 2007 and 2011)). vol FOA3a assumes a conservative fuel sulfur content of 0.068% by mass (the value listed in the American Gasolineeum Institute’s Handbook of Aviation Fuel Properties) and a sulfur dioxide (SO2) to sulfuric acid (H2SO4 Organic Contribution to PM ) conversion efficiency factor of 5%. Organic particulate matter comes about as a result of incompletely combusted fuel and species formed through pyrolysis in the engine’s combustion chamber. As with fuel sulfur content contributions, residence time and atmospheric conditions are important determinants of the organic contribution. vol Data gathered from testing a single CFM56-2-C1 engine during the Aircraft Particulate Emissions eXperiment (APEX) 1 were used to inform the calculation of the organic contribution to particulate matter. Two key assumptions regarding the APEX 1 results were made to permit the estimation of fuel organic particulate matter emissions for all engines. The first is that the test data gathered from one engine are representative of all engines in the ICAO Engine Emissions DataBank (2010). The second is that fuel organic particulate matter is proportional to total HC emission indexes, which are measured to achieve engine certification. Lubrication Oil Contribution to PM A development of FOA3a over FOA3 included an estimate for the contribution of engine lubrication oil to particulate matter emissions. Data were scarce at the time, so engineering judgments were made based on engine manufacturer data. vol The conclusion was that around 1.4 grams (Wayson et al., 2009; Ratliff et al., 2009) of volatile organic particulate matter is released per landing and takeoff (LTO) cycle, and this was added to the contribution from fuel organics to arrive at a total organic volatile component. Turboprop and Turboshaft Aircraft EDMS includes emission factors for turboprop and turboshaft aircraft for some pollutants, but it does not include emission factors related to particulate matter, as there are no FAA accepted emission factors for these aircraft. Very little data exist on turboprop and turboshaft aircraft particulate matter emissions. For the ACRP 02-23 project, two key sources of particulate matter emission factors were reviewed for suitable data with regard to turboprop and turboshaft aircraft, one from the U.S. Air Force (2002) and a supplement to the EPA AP-42 (U.S. EPA, 1980). It

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-6 should be noted that these data relate to particulate matter in general and not to PM2.5, though it has been assumed that all particulate matter is PM2.5 Piston-engine Aircraft . In addition, military aircraft may not be using Jet-A specification fuel, but will be using a similar military specification (e.g., JP-8). EDMS includes emission factors for piston-engine aircraft for some pollutants, but there is no inclusion of particulate matter from piston-engine aircraft (ACRP, 2008). In general, data on particulate matter emissions from piston-engine aircraft are scarce. However, the Swiss have undertaken a test program to measured CO, HC and NOx for 17 piston-engine aircraft. Particulate matter emissions were also measured in a parallel project that looked at the impact of using different fuels and additives on particulate matter emissions (FOCA, 2007a). The Swiss study suggests the soot emission factors shown in Table 8 (all in the 2.5 micron range), with 91/96UL being the best alternative fuel from a particulate matter perspective. Tab le 8 – Soot Emis s ion Ind ices (mg/kg fue l) Fuel Taxi Approach Climb Takeoff AvGas 100LL (leaded) 50 40 70 100 AvGas 91/96UL (unleaded) 1 1 2 3 Source: FOCA, 2007a. Auxiliary Power Unit Particulate Matter Emissions Estimation The EDMS database contains emission factors for PM10 and PM2.5 ICAO’s Airport Quality Guidance Manual (ICAO, 2007 and 2011) outlines three APU estimation methodologies that focus on NOx, HC, CO and PM for APUs. The emission factors used in EDMS were sourced from FAA and EPA documentation and industry correspondence. The FAA reviewed the information available in 2000 by getting the principal manufacturer (Honeywell) to comment on the datasets the FAA was recommending at the time. The resulting set of APU emission indices have been widely used in compiling airport emissions inventories. 10 The Airport Cooperative Research Program (ACRP) Project 02-06 report (Webb et al., 2008) discusses potential needed research in the context of airports and particulate matter. From that report, there is a recently commissioned study (Missouri University of Science and Technology, in progress) that should provide a better basis for data on APU emissions and that could be incorporated into EDMS in the future. Ideally, the ACRP 02-23 project would have incorporated that data. However, that was not feasible due to timescales, but it is recommended that it be included in future studies. emissions – a simple approach, an advanced approach and a sophisticated approach. Each of these requires an increasing resolution of data and offers an increasing level of accuracy of output. The simple approach would appear to be similar to that used in EDMS. The advanced approach, in principle, is based on work for the Project for the Sustainable Development of Heathrow (PSDH) study, where British Airways (BA) derived data from detailed manufacturer’s data (Underwood, 2007). The sophisticated approach is only appropriate where very detailed data can be obtained.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-7 Brake and Tire Wear No brake and tire wear is included in EDMS for aircraft or, typically, in U.S. airport emissions inventories in general. However, estimates of brake and tire wear for a number of airport emissions inventories, since and prior to the PSDH study (Underwood, 2007), have been included for UK airports. Without undertaking aircraft specific calculations outside EDMS, and given that brake and tire wear emissions are not directly affected by the use of alternative fuels, these have not been considered further in the ACRP 02-23 project. AIRCRAFT AND APU ALTERNATIVE FUELS Most of the research on alternative fuel emissions for aircraft has been related to jet fuel, which is used in turbine powered aircraft. As discussed previously, there is relatively little information available in the context of PM2.5 The high capital cost of airport infrastructure, distribution systems, and replacement of engines and supporting aircraft systems makes drop-in fuels economically necessary. To assure quick and widespread adoption of an alternative to jet fuel, the commercial airlines and the military require that the fuel be a drop-in fuel. The primary domestic fuel currently used in commercial aircraft turbine engines is a petroleum-derived Jet-A (ASTM D1655). Jet A-1 is the international standard for commercial jet fuel and JP-8 is the U.S. military’s jet fuel; both are derived from petroleum. emissions for piston-engine aircraft fueled by AvGas or diesel. In addition, the majority of aviation fuel consumed in the U.S. is jet fuel as opposed to AvGas. As of July 2011, the average petroleum products supplied, per day, in the U.S. are: aviation gasoline 15 thousand barrels; and kerosene-type jet fuel 1418 thousand barrels (U.S. Energy Information Administration, 2011). Therefore, the ACRP 02-23 project has concentrated on jet fuel alternatives. According to the Commercial Aviation Alternative Fuels Initiative’s (CAAFI) website glossary (CAAFI, 2010), a drop-in fuel: • May be used “as-is” on existing aircraft. • “Is completely interchangeable and compatible with conventional jet fuel when blended with conventional jet fuel”. • Requires no changes to the aviation fuel distribution system or aircraft or engine fuel system. There are many types of feedstock being considered for use as a substitute for jet fuel. The feedstocks must be able to produce a sufficient quantity to satisfy the growing demand for aviation fuel, estimated in 2011 as 14 million barrels per day for jet fuel and 14 thousand barrels per day for aviation gasoline in the U.S. (U.S. Energy Information Administration, 2011). To be adopted, the jet fuel developed must meet all of the specifications of jet fuel standards (e.g., freezing point, viscosity, flash point, density, and sulfur content). The International Air Transport Association’s (IATA) Fact Sheet on Alternative Fuels summarizes the requirements of jet fuel as having a freezing point below -40oC for Jet-A and -47oC for Jet A-1, not forming deposits in the engine in high-temperature locations and having an energy content of at least 42.8 MJ/kg (IATA, 2010). The two major categories of alternative jet fuels are alternative fossil fuels and biomass

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-8 fuels. Newer alternative fuels are frequently referred to by their source feedstock (coal, natural gas, or a specific plant or animal biomass) and their chemical processing methods, such as Fischer-Tropsch (FT) or hydroprocessed renewable jet (HRJ). Examples of alternative jet fuels derived from fossil fuels include natural gas (gas-to-liquid (GTL)) and coal (coal-to-liquid (CTL)). Examples of biomass-to-liquid (BTL) fuels include those derived from animal fats or from plants such as sorghum, switchgrass, jatropha, algae, and camelina. In 2011, the ASTM D7566 – 11a Standard Specification for Aviation Turbine Fuel Containing Synthesized Hydrocarbons was approved. This standard “covers the manufacture of aviation turbine fuel that consists of conventional and synthetic blending components” (ASTM, 2011). “Aviation turbine fuel manufactured, certified and released to all the requirements of this specification, meets the requirements of Specification D1655 and shall be regarded as Specification D1655 turbine fuel”, (ASTM, 2011). As a result of this newest revision of ASTM D7566, alternative fuels from both FT (BTL, CTL, GTL) and HRJ (described as Hydroprocessed Esters and Fatty Acids (HEFA) fuel derived from biomass feedstocks) produced according to D7566 are to be regarded as D1655 turbine fuels. Careful attention is necessary when attempting to compare results of studies when different percentages of full throttle are used or different engine power setting referencing methods are used (such as percent thrust, percent maximum continuous power or other engine parameters). Engine thrust settings influence emissions, particle size and chemical composition. Therefore, it is assumed in the ACRP 02-23 project that the AAFEX results have been normalized to negate the effects of temperature and pressure changes. The ICAO LTO default specifies thrust setting as a percentage of full throttle and the duration in minutes. EDMS uses the following default thrust settings, though the timing may be altered by the user: • 7% for 26 minutes, representing idle • 100% for 0.7 minutes, representing takeoff • 85% for 2.2 minutes, representing climb-out • 30% for 4.0 minutes, representing approach The three categories of alternative fuel types explored fully in the RAND 2009 (Hileman et al., 2009) report are: • Production from oil sands and oil shale • Fischer-Tropsch (FT) synthesis of natural gas, coal, and biomass • Refining oil products from biomass into synthetic paraffinic kerosene (SPK) fuel HRJ and FT fuels have a similar chemical structure. The emissions of particulate matter and secondary particulate matter from sulfur are expected to be reduced by more than 10% compared with the baseline fuel (Jet-A) according to the RAND 2009 report. Similarly, using camelina HRJ as a drop-in fuel could reduce carbon dioxide (CO2 Large reductions in particulate matter emissions are possible using FT fuels (Hileman et al., 2009). As the percentage of FT fuel increases, the reduction in particulate matter mass also ) emissions by over 80% during the life- cycle from the field to the wake (Goodrich, 2009). Ultra-low-sulfur jet fuel (ULS) is suggested in the RAND 2009 as a more quickly realizable method to reduce primary and secondary particulate matter caused by aviation. “ULS conventional” is jet fuel produced with lower acceptable sulfur levels (i.e., between 10 and 100 ppm).

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-9 increases, as found in U.S. Air Force tests comparing jet fuel with varying percentages of FT fuel. Recent tests in the Alternative Aviation Fuel Experiment (AAFEX) program were designed to run combinations of FT (two separate fuels, GTL and CTL), biomass fuels, and Jet-A in a CFM56 (a modern, high-bypass turbofan engine) operating on the ground on a DC-8 (Whitefield et al., 2008). Using low-sulfur, low-aromatic alternative fuels, such as those created from FT synthesis, may reduce primary particulate matter and appears to provide consistent particulate matter reductions across a variety of types and ages of gas turbine engines (Hileman, 2008). Military JP-8 and FT fuels tests on turboshaft engines, similar to turboprop engines, revealed that total particulate matter carbon, or non-volatile, emissions and diameters increased with increased engine power settings at idle, 75% of maximum continuous and 100% of maximum continuous power (Cheng et al., 2008). The engines were T700 and T701C GE engines, typically used in helicopters. Neat FT fuel (not blended with any other fuel) had reduced elemental carbon emissions, attributed to the lack of aromatics, which are soot precursors. Neat FT fuel had reduced organic carbon emissions at idle power, but not at higher engine power settings. Formation of volatile particulate matter emissions is negligible in neat FT fuel due to the lack of aromatics and sulfur. Tests show that elemental carbon (soot) emissions for engines running FT and JP-8 were dramatically higher at maximum continuous power than other power settings— 130 g/m3 and 30 g/m3, respectively. Particulate matter mass emission indices ranged from 0.2 to 1.4 g/kg fuel for the T700 and 0.2 to 0.6 g/kg fuel for the T701C (Corporan and Cheng, 2010). The entire fleet of U.S. Air Force (USAF) aircraft is expected to be certified to use blended alternative fuels by 2016 For the FT fuel, elemental carbon at idle (not specified further) and 75% of maximum continuous power was negligible. Organic carbon (non-soot) emissions for the JP-8 and FT engines were reported as statistically identical. . The U.S. Air Force is currently certifying aircraft to operate with a 50/50 blend by volume of FT and JP-8 fuel. The emissions of the T701C engine were compared while using JP-8 fuel and a neat FT fuel (Syntroleum Corporation’s GTL from natural gas). FT fuels have smaller particle number emission indices (EI) relative to fuel flow at all power settings (less than 1.0 x 1014 at idle) compared to convention fuel, typically, with reductions of between 40% and 97% in particle number emission indices (PN-EI) and the highest reductions at idle. Particle size distributions for FT are dramatic, with average reductions of 25% in mean particle diameter observed at all power settings. Smoke numbers for FT fuel at the three power settings were dramatically lower – an average of 65%. Smoke number (SN) trends are consistent with PN-EI trends. All engines produced higher CO and lower NOx emissions at the lower power settings. NOx AAFEX was conducted in 2009 at NASA’s aircraft facility in Palmdale, California in the Dryden Flight Research Center DC-8. NASA acquired and burned JP-8, FT GTL (FT-1) and FT CTL (FT-2) fuels to assess changes in performance and emissions in the two inboard CFM-56 main engines (AAFEX, 2010). The key results, in the context of the ACRP 02-23 project, are shown in emissions for FT were negligibly different. CO emissions were reduced by between 5% and 10% using FT. Formaldehyde (HCHO) is the primary aldehyde produced and FT fuel had minimal impact on production, except at the 75% of maximum continuous power setting. Table 9. In addition to the studies on the CFM-56 main engines, AAFEX also reported the effects of alternative fuels on the APU. The Garrett AiResearch GTCP85-98 CK APU onboard a DC-8 parked at Palmdale, California, was studied in January and February 2009 (Beyersdorf and

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-10 Anderson, 2009) using two fuels – JP-8 and FT CTL. Running JP-8, the black carbon and organic compound emissions from the APU were measured as exhaust gas temperature (EGT) increased. As EGT increased from around 365oC to 610oC, the emissions dropped for black carbon and organic compounds. Using approximations from Beyersdorf and Anderson (2009), the black carbon emissions at around 365oC were nearly 500 mg/kg of fuel burned, and at around 610oC were around 200 mg/kg of fuel burned. APU emissions were reported as around 20 times the emissions from the DC-8’s CFM56 engine at idle. Using approximations from the study, the organic compound emissions at around 365oC were between 6 and 7 mg/kg of fuel burned, and at around 610o Running FT CTL in the same AAFEX study, the black carbon and organic compound emissions from the APU were measured as EGT increased. As the EGT increased from around 365 C were around 200 mg/kg fuel burned. oC to 610oC, emissions dropped for black carbon and organic compounds. Using approximations from the study, the black carbon emissions at around 365oC were about 40 mg/kg of fuel burned and at around 610oC were about 5 mg/kg of fuel burned. Using approximations, from Figure 5 in the AAFEX study, the organic compound emissions at around 365oC were about 1 mg/kg of fuel burned and at around 610o The Partnership for Air Transportation Noise and Emissions Reduction (PARTNER) is an FAA Center of Excellence, sponsored by the FAA, NASA, Transport Canada, the U.S. Department of Defense and the EPA. PARTNER Project 20, Emissions Characteristics of Alternative Aviation Fuels (Missouri University of Science and Technology, 2011a), is working with the aviation community to gather accurate data on emissions from candidate alternative fuels and to compare these emission characteristics with those of conventional aviation fuel types being gathered in PARTNER Project 9, Measurement of Emissions (Missouri University of Science and Technology, 2011b). These data will provide the essential information for PARTNER Project 17, Alternative Fuels (Missouri University of Science and Technology, 2011c) and to the aviation community at large as it charts a course for environmental sustainability in an uncertain energy future. The planned outcome is the creation of a database of particulate matter and hazardous air pollutant emissions from engines burning Jet-A/JP-8 and alternative fuels, such as FT synthetic fuel. C were about 0.5 mg/kg of fuel burned. Analysis of data from AAFEX II experiments using HRJ fuels is underway, and readers should consider those results when they become available. The data found from the literature review have been used to generate emission factors for alternative fuels as summarized in Table 9. Tab le 9 – Data to Support Alte rna tive Airc raft Fue l Emis s ion Fac tors Turbine Engine Fuel Type Engine Setting SOx SN (mg/kg fuel) EI particles/kg fuel Black Carbon mg/kg fuel HC EI g/kg fuel Source T-63 FT GTL blended with JP-8 (0% up to 100% FT) Idle. 0.40 kg/minute fuel flow Decrease linearly with increase in FT% 0-100% FT 6.6 to <1 Corporan et al., 2007

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-11 Turbine Engine Fuel Type Engine Setting SOx SN (mg/kg fuel) EI particles/kg fuel Black Carbon mg/kg fuel HC EI g/kg fuel Source Cruise. 1.32kg/minute fuel flow Decrease linearly with increase in FT% 0-100% FT 29.7 to 3.8 Corporan et al., 2007 FT GTL 50/50 blend with JP-8 Low PM EI 0.53 of JP-8 Corporan et al., 2007 High PM EI 0.46 of JP-8 Corporan et al., 2007 JP-8 and Methyl Ester Biofuel Blend (80/20) Ground idle 7.5 Corporan et al., 2007 Cruise 31.4 Corporan et al., 2007 Takeoff 35.3 Corporan et al., 2007 T700- GE-701C FT GTL Idle 65% reduction using FT over JP- 8 on same engine. Negl. Corporan and Cheng 2010 75% MCP 2 Corporan and Cheng 2010 100% MCP 12 Corporan and Cheng 2010 CFM56 FT GTL neat Low power (4%-45% max rated power) [4% 1,000 lbs/hour] <0.3 0.1 JP-8 (90% reduction compared with JP-8) Beyersdorf and Anderson, 2009 7% max rated thrust 0.41 JP- 8 Anderson et al., 2011 85% max rated thrust 0.16 JP-8 0.41 JP- 8 Anderson et al., 2011 85% max rated thrust <0.3 Miake-Lye et al., 2009 High 1 Bulzan et al., 2010 100% max rated thrust [7,600 lbs/hour] <0.3 0.4 JP-8 Beyersdorf and Anderson, 2009

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-12 Turbine Engine Fuel Type Engine Setting SOx SN (mg/kg fuel) EI particles/kg fuel Black Carbon mg/kg fuel HC EI g/kg fuel Source FT CTL neat Low power (4%-45% max rated power) [4% 1,000 lbs/hour] <0.3 0.1 JP-8 Beyersdorf and Anderson, 2009 7% max rated thrust 0.53 JP- 8 Anderson et al., 2011 85% max rated thrust 0.29 JP-8 0.59 JP- 8 Anderson et al., 2011 85% max rated thrust <0.3 Miake-Lye et al., 2009 High 1 Bulzan et al., 2010 100% max rated thrust [7,600 lbs/hour] <0.3 0.4 JP-8 Beyersdorf and Anderson, 2009 50/50 JP- 8/FT GTL Low power (4%-45% max rated power) [4% 1,000 lbs/hour] 0.5 JP-8 Beyersdorf and Anderson, 2009 100% max rated thrust [7,600 lbs/hour] 0.7 JP-8 Beyersdorf and Anderson, 2009 50/50 JP- 8/FT CTL Low power (4%-45% max rated power) [4% 1,000 lbs/hour] <0.5 0.5 JP-8 Beyersdorf and Anderson, 2009 100% max rated thrust [7,600 lbs/hour] 0.7 JP-8 Beyersdorf and Anderson, 2009 All fuels in test Beyersdorf and Anderson, 2009 Garrett APU GTCP85- 98CK JP-8 200-500 Beyersdorf and Anderson, 2009 100% FT/JP-8 CTL Low power 0.16 JP-8 0.16 JP-8 Anderson et al., 2011 100% FT/JP-8 CTL High power 0.13 JP-8 0.11 JP-8 Anderson et al., 2011 50/50 JP- 8/FT CTL 10-50 Beyersdorf and Anderson, 2009

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-13 GROUND SUPPORT EQUIPMENT (GSE) EMISSIONS GSE emissions tend to refer to the airside emissions from aircraft support equipment, such as mobile generators, air-conditioning units, baggage, fuel, food and cargo trucks, and loaders and tugs. It can also be used to refer to buses used airside to transport passengers between remote aircraft and terminals and cargo trucks. Road vehicles are dealt with separately under road vehicles, and only GSE are discussed further in this section, although there are many parallels with road vehicles. Conventional Fuels U.S. federal standards for off-road diesel engines have evolved over a period of time (Dieselnet, 2010). Tier 1 standards were phased in from 1996 to 2000, Tier 2 and Tier 3 standards were phased in from 2000 to 2008, Tier 3 standards for particulate matter were never adopted and Tier 4 standards are to be phased in from 2008 to 2015. The Tier 4 standards require control technologies that include advanced exhaust gas after-treatment. In addition, nonroad diesel will need to have lower sulfur content in the future. Fuel refiners began to produce low-sulfur nonroad diesel in June 2007 (U.S. EPA, 2009a). This will be further lowered in the future (U.S. EPA, 2004a), which should reduce the particulate matter emissions, as sulfur acts as a substrate for secondary particulate matter formation. In the U.S., emissions for GSE are typically calculated using the inbuilt EDMS model emission factors (refer to the section Conventional Fuels under Road Vehicle Emissions for trucks and buses). The emission factors were generated in EPA’s NONROAD2005 emission factor model in EDMS version 5.1.2 (U.S. FAA, 2009) because the engines used by GSE manufacturers are those typically used elsewhere in other equipment (due to market size). A more recent version of the NONROAD model is available (U.S. EPA, 2008). The NONROAD model for off-road vehicles covers compressed natural gas (CNG), liquefied petroleum gas (LPG), 2- and 4-stroke gasoline, and diesel fuel. Internally, NONROAD develops this information based on available engine testing data, such as from certification. The NONROAD emission factors are only available for total PM2.5 for GSE within EDMS. However, the derived emission factors incorporate deterioration factors. The emission factors derived from NONROAD PM2.5 The ICAO/CAEP guidance (ICAO, 2007 and 2011) suggests two simplified approaches, based on aircraft movements (multiplied by an appropriate average GSE emission factor) or total fuel use and an average GSE emission factor. The more advanced methodology that relates to the PSDH developments in that emissions are calculated on a time-use basis for each piece of GSE and includes degradation and a load factor. The ICAO/CAEP guidance only discusses direct GSE emissions and does not include discussion of brake and tire wear, re-suspension or secondary emissions in the context of GSE. Comparison of the application of a very similar also incorporate some volatile emissions (i.e., from sulfur) (U.S. EPA, 2005). However, other volatile emissions from other pollutant interactions are not specifically included (although some will be accounted for as primary exhaust emissions are based on certification data). Similarly, brake and tire wear or re-suspended solids, such as dust, are not included. Ideally, these sources would be included in this ACRP 02-23 project; however, developing the EDMS model and detailed emission factors for these sources is beyond the scope of the ACRP 02-23 project other than in the context of alternative fuels.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-14 methodology (the same, but without the degradation factor) for GSE and ground power unit (GPU) emissions can be found in the reports from Zurich Airport (Unique, 2004 and 2006). The ACRP Project 02-06 Report Research Needs Associated with Particulate Emissions at Airports (Webb et al., 2008), discussed potential needed research in the context of airports and particulate matter. As a result of this report, there is a recently commissioned study (CDM, in progress) where one of the aims is to develop a “representative inventory of powered GSE at airports to help the industry assess the contribution of GSE to air quality impacts at airports.” Unfortunately, given the ACRP 02-23 project’s timescale, the related data and information could not be used to supplement this study. It may be feasible for airports in general to obtain estimates of total airside fuel use for a particular airport, which could be used as a mechanism to check estimates of GSE fuel use and, therefore, indicate the validity of the emissions. Brake and Tire Wear Ideally, brake and tire wear would be included in the ACRP 02-23 project, although most airport studies in the world do not include brake and tire wear as there is not a defined methodology. However, a recent study for London Heathrow Airport to compile an emissions inventory for a base year of 2008/09 (Underwood et al., 2010) included estimates for brake dust, tire wear and re-suspended road dusts for GSE based on the UK methodology for road vehicles (described in this document under road vehicles). For GSE, it was assumed that the equivalent emissions in terms of g/kg fuel can be applied to small GSE as for cars, medium GSE as medium road vehicles and large GSE as large road vehicles. However, as discussed in the Brake and Tire Wear section under Aircraft and APU emissions of this document, the ACRP 02-23 project is concerned with the impact of alternative fuels, which is not directly affected by brake and tire wear emissions. GSE ALTERNATIVE FUELS The Voluntary Airport Low Emission (VALE) program (U.S. FAA, 2010a) is focused on helping airports to improve air quality. It provides funding from the FAA to commercial airports in areas where air quality standards are currently not attainable. In terms of the VALE process, via designated Department of Energy (DOE) and EPA guidelines, eligible alternative fuels are: • Electricity (including photovoltaic) • Natural gas and liquid fuels domestically produced from natural gas (CNG or liquefied natural gas (LNG)) • LPG/propane • Mixtures containing 85% or more by volume of alcohol fuel with gasoline, including denatured ethanol (E85) and methanol (M85) (i.e., biogas) • Hydrogen • Coal-derived liquid fuels • Biodiesel (B85 to B100) • P-series fuels

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-15 Many of these fuels have very limited availability or are still at the research and development stage (e.g., P-series and hydrogen). In terms of air quality emissions, hydrogen and electricity produce zero particulate matter direct emissions. Examples of the types of electric vehicles and equipment tried at airports include St Paul’s Airport, Minneapolis (Energy Efficiency News, 2009); London’s Heathrow Airport (Smith Electric Vehicles, 2010); Tokyo’s Haneda Airport (TreeHugger, 2008); and many others (U.S. FAA, 2006b and 2010b). In EDMS, GSE emission factors are available for several fuels (diesel, gasoline, electric, CNG, and LPG) (U.S. FAA, 2009). The CNG and LPG EDMS emission factors appear to be based on an EPA study, which found that particulate matter emissions from 4-stroke, spark-ignition engines running on LPG and CNG were 0.05 g/hp-hour (U.S. EPA, 2004b). The Inherently Low Emission Airport Vehicle (ILEAV) program (U.S. FAA, 2006b), VALE’s predecessor, also looked at non-electric alternative fuels replacements. Although two key alternative fuels reported in the ILEAV program, CNG and LPG, are a replacement for gasoline, because the engines provide a spark, these fuels are not a replacement for diesel fuel in compression-ignition (CI) engines. An engine can be adapted to use these fuels, but it involves major engineering work to change the compression ratio of the engine and add an ignition system. CNG and LPG can be used in a dual-fuel engine, together with diesel; the diesel is needed to ignite the air-fuel mixture in a CI engine. In terms of other alternative fuels, many existing diesel GSE could theoretically be run using low-sulfur diesel and biodiesel blends without specialist (non-standard) engines designed to run on specific alternative fuels. Biodiesel typically has lower sulfur and aromatic content than standard diesel, which, therefore, generally acts to reduce particulate matter emissions. However, there is a limit to what can be achieved. Most engines are not designed for use with biofuels, especially GSE, where there is a limited market, and many manufacturers will not guarantee existing equipment on higher-biofuel blends. In addition, biofuels are typically thought to improve particulate matter emissions (e.g., compared with diesel) (Lapuerta et al., 2005; Krahl et al., 2009). However, some sources suggest an increase of particulate matter for biodiesel relative to standard diesel (Gaffney and Marley, 2009), although this would appear to contradict the general consensus, and it seems to be related to the hydrocarbon content of the fuel. Further research is needed to quantify the impact that specific types of biofuel (by feedstock, blend and engine type) will have on primary and volatile (i.e., “secondary”) particulate matter emissions. In terms of biofuels, it is also worth considering that many will solidify at cool temperatures, block fuel filters and generally create the need for higher maintenance. Similarly, high humidity can cause microbial growth in biofuels, again resulting in higher levels of maintenance. In terms of low-sulfur diesel, nonroad diesel will need to have lower sulfur content in the future. Fuel refiners began to produce low-sulfur nonroad diesel in June 2007 (U.S. EPA, 2009a). This will be reduced further in the future (U.S. EPA, 2004a), which should act to reduce the particulate matter emissions. Resulting from this, there is a briefing document that accompanies the NONROAD model that suggests sulfur contents to use for future years, will reduce from 2,284 ppm to 11 ppm by 2015 (U.S. EPA, 2009d).

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-16 Interestingly, EDMS automatically assumes lower sulfur content when calculating future GSE emissions for SOx, but not other pollutants (U.S. FAA, 2009). It is feasible that the sulfur content can be altered and that an additional set of low-sulfur diesel emission factors can be generated for use in EDMS using the equations within NONROAD’s supporting documentation (U.S. EPA, 2010e). This methodology is basically the same in the 2005 version of NONROAD supporting documentation (U.S. EPA, 2004b). However, specific emission factors for E85, M85, B85 and B100 are more difficult to estimate, though Table 3 under Road Vehicle Emissions in Chapter 2 gives some broad factors that could be used to convert emission factors from either gasoline or diesel to bioethanol or biodiesel. While retrofit technology is not the subject of this report, it could be advantageous to fit equipment (e.g., particulate matter traps) to existing GSE diesel engines given the uncertainties of particulate matter emissions and to be cost-effective. Where vehicle replacement is an option, electric GSE is better when compared with other alternative fuels in terms of reducing directly emitted particulate matter (U.S. FAA, 2010a). Around the world, electric vehicles are available as replacements for baggage tugs and belt loaders. A few other specialist airside electric vehicles have been trialed, and there are a few makes of electric aircraft push-back tugs. However, their relatively modest capacity suggests they would not be very flexible and unable to deal with larger aircraft. ROAD VEHICLE EMISSIONS Conventional Fuels A detailed study by the EPA found that emissions of particulate matter deteriorate exponentially with the age of the vehicle, but remain constant after about 20 years (Beardsley, 2006). The study also found that particulate matter emissions increase exponentially with vehicle power (or road or engine load). The EPA found that emission data for heavy-duty vehicles are not sufficient to permit stratification according to engine size, vehicle weight, and injection type (direct and indirect) (U.S. EPA, 2009b). Emissions from road vehicles are calculated using the emission factors defined by the EPA under the MOBILE program. The approach adopted in the European Environment Agency (EEA)/United Nations Economic Commission for Europe (UNECE) Cooperative program for monitoring and evaluating the long- range transmission of air pollutants in Europe (known as EMEP) (EEA, 2009, 2007 and 2005) considers passenger cars, light duty vehicles, heavy-duty vehicles, and motorcycles and mopeds. It covers gasoline, diesel, LPG, and natural gas fuels. The EMEP system assumes that all particulate matter is PM2.5 When estimating airport-related air emissions resulting from surface traffic and other road vehicles emissions, EDMS calls upon the EPA’s MOBILE6.2 emission factor model at the national default level. This provides emission factors in grams per vehicle miles traveled (VMT) for gasoline and diesel-fueled road cars, trucks, buses, motorcycles and other vehicles. It should be noted that while EDMS continues to use MOBILE, this model has now been replaced outside EDMS by a newer regulatory model, the Motor Vehicle Emission Simulator (MOVES) (U.S. EPA, 2010c). However, for consistency with the EDMS model, the ACRP 02-23 project used MOBILE 6.2. The current version of the MOVES model includes a number of alternative fuel , as it is assumed that the coarse fraction is negligible in vehicle exhaust. This is consistent with the findings of Ristovski et al., (1998) who found that the mean particle diameter in emissions from gasoline-fueled vehicles was below 1µm.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-17 emission factors, such as gasoline, diesel, CNG, LPG and electricity (U.S. EPA, 2010f). It is intended that MOVES will eventually include other fuels, such as ethanol (E85), methanol (M85), gaseous hydrogen, and liquid hydrogen (U.S. EPA, 2009c), but it did not at the time this report was written. MOBILE6.2 can also be run independently of EDMS, the results of which can then be re- incorporated into EDMS, to account for area specific parameters (e.g., local registration data, VMT data, emissions control program parameters, meteorological data and sulfur fuel content) that may have been established by state air quality agencies in non-attainment areas, or other considerations. In addition, MOBILE6.2 (U.S. EPA, 2003a and 2003b) assumes that the total exhaust particulate matter is made up of three factors for recent years (assumes no lead): organic derived particulate matter, elemental particulate matter, and sulfur-derived particulate matter. For gasoline, the first two factors are combined (due to lack of separate data). However, for diesel, all three factors are separate. MOBILE also calculates brake and tire wear particulate matter. MOBILE6.2 also includes estimated particulate matter from natural gas vehicles (NGVs) by assuming the particulate matter emissions are, in essence, the same as those for very low-sulfur gasoline. Therefore, it is possible to alter the diesel sulfur content in the input file for MOBILE (this must be done outside EDMS) and then use the new output emission factors in their composite form in EDMS. It is also possible to alter the assumed market shares of ether and ethanol blends in MOBILE. Similarly, the output files from MOBILE can be used to estimate the different components of the composite particulate matter emission factor used in EDMS. Therefore, some volatile emissions can be split out, although it will not include the organic volatile emissions from gasoline vehicles due to the combined nature of gasoline organic and elemental particulate matter. Brake and Tire Wear Research carried out by the EPA indicates that 10% of brake wear particulate matter is PM2.5 (Nam and Srivastava, 2006). The EPA data suggest that the MOBILE model used by EDMS is likely to underestimate brake and tire wear emissions. For example, a simple study, using default factors within EDMS, generated the emission factors in MOBILE where brake and tire wear accounted for around 17% of PM2.5 emissions. A much smaller proportion of PM10 The emission factors (in g/km) for brake and tire wear used in the UK National Atmospheric Emissions Inventory (NAEI) are described in the Air Quality Expert Group (AQEG) report on particulate matter (AQEG, 2005). The methodology draws on a review of brake and tire wear carried out for UNECE, which has informed the methodology included in the recent versions of the European EMEP/CORINAIR Emission Inventory Guidebook (EEA, 2005). These emission factors indicate that the UK and European methodology assumes much higher factors for tire wear than that assumed in the U.S. (less than 0.1%) was reported in a study of tire wear from motorcycles and small cars traveling at constant speeds on a concrete surface (Aatmeeyata and Kaul, 2009). The reason for this discrepancy is not clear, although this could be due to the sizes of the vehicles and the roadway surface.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-18 ROAD VEHICLE ALTERNATIVE FUELS The U.S. DOE defines the following alternative fuels for vehicles under the Energy Policy Act (1992) (U.S. DOE, 2010): biodiesel, electricity, ethanol, hydrogen, methanol, natural gas, and propane. Several emerging fuels are currently under development and are also regarded by DOE as alternative fuels. These include biobutanol, biogas, BTL, CTL, FT diesel, GTL, hydrogenation derived renewable diesel, P-Series, and ultra-low-sulfur diesel. The EPA carried out a measurement survey that confirmed the beneficial effect of biodiesel mix on emissions of particulate matter from diesel-fueled vehicles (U.S. EPA, 2002b). The effect on particulate matter and other emissions is given by the following equation: Change in emissions = e-0.06384×%B The European EMEP inventory (EEA, 2009, 2007 and 2005) provides guidance on the effect of biodiesel blends on emissions of particulate matter. For older diesel technologies with no advanced combustion concepts and after-treatment systems, biodiesel may lead to a higher proportional reduction in emissions of particulate matter because the presence of a carbon- oxygen chemical bond reduces the particulate matter formation by intervening in the chemical formation process. where B is between 0 and 100 (percent biodiesel) For more recent technologies with ultra-high-pressure combustion and after-treatment, the biodiesel effect is difficult to predict because of changes in physical properties of the fuel. The European EEA estimates that biodiesel blends B10 and B20 reduce vehicle particulate matter emissions by between 10% and 20% (Table 3-104: EEA, 2009, 2007 and 2005). This is a slightly greater decrease in emissions than that reported by the EPA. For heavy-duty vehicles, the estimated reduction in emissions for B100 is 47%, identical to that reported by the EPA. The DOE suggests that pure biodiesel (B100) greatly reduces emissions other than NOx The Argonne National Laboratory found that vehicles that, effectively, have zero tailpipe emissions could have relatively high or relatively low particulate matter emissions when considered on a life-cycle basis (Argonne, 2005). The use of renewable versus non-renewable sources of electricity was found to be an important factor. Liquid hydrogen fuel-cell vehicles performed relatively well, whereas a gaseous hydrogen internal combustion engine performed relatively poorly. and that B100 could potentially be used advantageously by professional fleets with appropriately equipped maintenance departments. A literature review carried out for the Dutch government found a mixed picture in terms of the effects of biofuels on emissions of particulate matter (TNO, 2004). Ethanol, FT diesel, and bioDME (dimethyl ether) were found to result in reduced emissions of particulate matter. Biogas was found to result in low emissions, but with a risk of higher emissions if product quality is variable. The picture for biodiesel is not straightforward. The low sulfur content of biodiesel, FT diesel and bioDME would be favorable for the use of catalytic converters, if used. Using biofuels can result in operational difficulties with associated emissions issues (e.g., ethanol can act as a solvent for past gasoline deposits), but such issues can generally be overcome. The effects of biofuels on emissions of other non-regulated pollutants (e.g., individual potentially hazardous volatile organic compounds (VOC)) are favorable overall.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-19 A review carried out for the European Commission also investigated the effects of biofuels on emissions (JNC, 2006). Using pure biodiesel was found to have a mixed effect on emissions from heavy-duty vehicles, with reductions up to 80% and increases up to 40%. Increases were observed for biodiesels with a higher soluble organic fraction. Biodiesel blends generally resulted in a reduction of particulate matter of up to 50%. Using vegetable oils in heavy-duty vehicles was found to have a mixed effect on emissions of particulate matter. The European Commission study also reported research which indicated that alternative fuels may reduce particulate matter emissions from light duty vehicles under fuel-rich driving conditions, such as heavy accelerations. Emissions of particulate matter increased in order from LPG, CNG, 85% ethanol with 15% gasoline, 85% methanol with 15% gasoline, to the highest emissions from reformulated gasoline. A detailed review of the effects of biodiesel on diesel engine emissions found that 95% of publications report a decrease in particulate matter emissions with biofuel compared with diesel and 3% report an increase (Lapuerta et al., 2008). The study found that most authors have reported increases in the number of small particles with the use of biodiesel, with most particles smaller than PM0.1 A UK study reviewed a wide range of emissions studies and provided an assessment of the effect on particulate matter emissions compared with a reference fuel (gasoline or diesel) for use in emissions inventory compilation (AEA, 2008). No correction for sulfur content is provided. The estimated factors are set out in . For example, Wang et al., found that B35 biodiesel resulted in a 25% reduction in emissions of particulate matter, consistent with the above EPA formula (Wang et al., 2000). These authors considered that the reduction was due mainly to the oxygen content of biodiesel, and also to the lower sulfur and aromatic content of biodiesel. Table 3 in Chapter 2 of this report and are comparable to those set out in the above equation. Reductions in particulate matter emissions were confirmed in a study carried out for the World Bank (Kojima and Johnson, 2005). A study of biomass-to-liquid fuels provided a comparison of emissions reduction from BTL and GTL (i.e., FT) diesel compared with oil derived diesel (Kavalov and Peteves, 2005). This indicated that biomass-derived diesel delivers about a 25% to 65% improvement in PM2.5 emissions, and FT diesel delivers a 26% to 50% reduction. This study indicates that bioDME can deliver up to 90% reductions in emissions of NOx Use of methanol, ethanol, and methyl tertiary butyl ether as fuels can lead to increases in secondary particulate matter due to the formation of peroxy acetyl nitric acid (PAN) (Gaffney and Marley, 2009). and particulate matter. This is supported by measurement data, including data from the U.S. (Norton et al., 1998; Muncrief et al., 2007). The data discussed and presented above could be used to generate proxy alternative fuel emission factors for road vehicles and GSE. The data for the use of biodiesel blends are summarized in Figure 6 in Chapter 2 of this report.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-20 OTHER EMISSION SOURCES “Other” emissions refers to emissions from on-airport sources other than the major sources discussed above (aircraft, APU, GSE, and road vehicles). This category includes emissions from: • Heating plant • Training fires • Aircraft maintenance activities • Fugitive emissions from fuel handling (aircraft and vehicular) • Construction activities The range of sources is wide if fugitive VOC emissions are considered in the context of particulate matter emissions. However, heating plant and training fires are typically the main sources of “Other” (i.e., not aircraft, GSE or road vehicles) particulate matter. Therefore, these sources are the principal focus here. However, construction emissions are also an important source of particulate matter, though, by their nature, they tend to be limited to the period of construction. Therefore, they are not addressed further in the ACRP 02-23 project. It may also be convenient to identify source categories, such as the additional emissions arising from cold starts in airport parking lots, queuing taxis, and idling buses and coaches, that require a different emissions methodology from the one used for road vehicle emissions on the landside roadway network. Nevertheless, it is assumed here that these sources are included in the principal road vehicles source category discussed above. Types of other sources included in EDMS include heating and power raising (boilers and incinerators fueled by coal, oil, gas, LPG or general waste), emergency generators (fueled by gas, oils or LPG), aircraft engine testing (in essence, covered in the aircraft section), deicing, fuel tanks and solvent use. Deicing, fuel tanks and solvent use are not directly related to particulate matter, although the fugitive VOC emissions from these sources could potentially cause PM2.5 emissions. However, in EDMS only VOC is included as a pollutant for these sources, and it should also be considered that their contribution to total PM2.5 is likely to be relatively small. Stockpiles of things such as salt and sand are included in EDMS. Including these sources does result in PM2.5 Conventional Fuels – Heating Plant Emissions emissions, though these are not impacted by alternative fuels. The term “airport heating plant” is used as shorthand for an on-airport plant using local combustion of fuel to produce heating and/or electrical energy. While electricity provided to the airport from the grid also creates emissions, they are assumed to be non-local to the airport. Emissions data for heating plant stationary sources are provided by the EPA in its Compilation of Air Pollutant Emission Factors (U.S. EPA, 2010g), and that data are used as a basis for the EDMS emission factors. Traditionally, commercial and industrial boilers have been used to supply space heating and hot water to passenger terminals, commercial buildings, and maintenance hangars, and are fired by gas or fuel oil (either distillate oil, sometimes called “gasoil” or heavy fuel oil). Particularly when fired by liquid fuels, such plant may constitute one of the largest sources of annual particulate matter emissions at an airport. However, stack design usually ensures that those emissions do not make a major contribution to off-airport airborne ambient particulate matter concentrations. However, if there are residential population areas

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-21 close to the airport perimeter, the details of stack efflux characteristics may play a critical role in ensuring that the contribution from this source is minor. Natural gas, which, as a fuel, has relatively few associated particulate matter emissions, is currently used at a large number of U.S. airports (Lau et al., 2010). Besides conventional boilers, an alternative type of plant used at airports for heat and energy generation is combined heat and power (CHP) plant. Various types of CHP plant have been used or considered for airports, such as conventional turbines fueled by gas and/or gasoil, converted aircraft engine turbines fueled by kerosene and large diesel engines. It should be noted that representative (default) particulate matter emission factors for turbines (in g/MJ) are usually somewhat higher than those for conventional boilers. However, in practice, the emission factors vary widely with details of the plant design and the type of control technology implemented. Conventional Fuels – Training Fire Emissions Major airports must have on-airport facilities for firefighting and rescue, in accordance with ICAO requirements, and must make provision for fire training. The emissions associated with fire training are commonly included in airport emissions inventories for completeness. However, in annual terms, they comprise a very small fraction of the total near ground particulate matter emissions on the airport. Conventionally, kerosene has been used to create training fires to ensure realistic fire temperatures and smoke densities but, more recently, kerosene has been replaced by other fuels at some airports. Even with kerosene fuel, the estimated annual particulate matter emissions from fire training exercises are a small fraction of the total on-airport particulate matter emissions. Of course, the emissions derive from a relatively small number (typically tens) of training exercises in the year, so the chief health concern may relate to short-term concentrations during the exercises rather than the contribution to long-term exposure. Nevertheless, 24-hour 98th percentile off-airport particulate matter exposures on fire training days are not likely to be demonstrably higher than those experienced on other days. Alternative Fuels – Heating and Power As discussed previously, the AP-42 includes emission factors and methodologies for calculating a wide variety of heating plant emissions for different fuel types, many of which are incorporated in EDMS. However, although heating plant emissions are unlikely to make a major contribution to off-airport ground level airborne particulate matter concentrations, there may still be an interest in reducing emissions per se. This is particularly the case if there are targets and limits on the overall emissions burden of the airport in addition to limits on ambient airborne concentrations. A change of fuel may arise as a consequence of a complete change of heating plant type (e.g., from boilers to CHP or from large diesel engine CHP to gas turbine CHP). This type of replacement may be driven primarily by economic considerations, but could have a beneficial impact on particulate matter emissions if the plant is chosen carefully. An example of a simple change to a less conventional fuel type would be a switch to LPG in boilers or turbines. The use of LPG could yield a high reduction in particulate matter emissions over fuel oil, so it could provide an alternative in situations where a network gas supply is not available.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-22 There are other motivations for changing the type of fuel and/or type of plant to supply heating and energy on an airport in addition to economic considerations and the desire to reduce local air quality pollutant emissions. In recent years, an important driver of change has been the desire to reduce the airport’s carbon footprint, and biomass-fueled plant are increasingly being considered as a replacement for existing plants or as a supplement to meet the requirements of airport expansion. However, a modern wood burning plant is likely to produce much higher particulate matter emissions than those produced by a gas fired boiler or turbine, so there is likely to be a trade-off between carbon footprint and air pollutant emissions. Similarly, airports create large volumes of mixed waste, and energy from waste (EfW) plants is an attractive way to reduce waste volume while supplying some of the energy requirements of the airport. However, once again, they could increase particulate matter emissions where the increase depends on the type and size of plant, and on the sophistication of the emissions control technology implemented. Furthermore, if the energy generated replaces electricity from the grid, all of the emissions produced represent an addition to the local emissions inventory. There are options for meeting an airport’s heating requirements that do not depend on combustion so, in principle, they generate virtually no local air pollutant emissions in operational use. Examples include solar photovoltaic at Phoenix Sky Harbor International (PHX) and Fresno Yosemite International (FAT) airports, solar thermal heating at Dallas Fort Worth International Airport (DFW), and wind turbines at Minneapolis-St. Paul International Airport (MSP). Other options also include geothermal and ground source heat pumps, co-generation, and thermal storage (peak shifting) (Lau et al., 2010). However, the practicability of such options depends on location and economic viability. Of course, general best practices for energy management, insulation, etcetera, which are not the subject of this report, will reduce energy consumption and the resulting particulate matter emissions. Alternative Fuels – Training Fires The smoke pollution caused by the open burning of kerosene has led some airports to turn to alternative fuels, although there are no statistical data on how widespread the switch from kerosene has been. At the major London airports in the UK (Heathrow, Gatwick, and Stansted), LPG burners have been used for fire training, although small amounts of kerosene may still be burned for specific training exercises. Although particulate matter emission factors for the open burning of kerosene and for LPG burners of the type used for training fires are not well characterized, LPG fueling clearly produces much less visible smoke and can be assumed to generate much lower particulate matter emissions. EDMS includes emission factors for JP-4, JP-5, JP-8, propane (LPG), and tekflame. Tekflame and LPG theoretically produce fewer particulate matter emissions than the other more conventional fuels, according to the EDMS model. DISPERSION MODELING AT AIRPORTS According to FAA Order 1050.1E Change 1, dispersion models prepared for the evaluation of airport air quality impacts must be prepared using EDMS (version 5.1.2.). EDMS typically invokes EPA’s AERMOD dispersion model to translate the emissions inventories it calculates into predicted concentrations of air pollutants in the study domain. AERMOD/EDMS incorporates information on the spatial arrangement and emission characteristics of airport sources, terrain and elevation, meteorological variables, and other physical considerations when

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-23 predicting concentrations. However, it should be noted that AERMOD is a short-range dispersion model used to assess local air quality impacts and does not include chemical interactions that result in the formation of secondary atmospheric particulate matter. When using EDMS/AERMOD, secondary particulate matter can only be accounted for by adding a background particulate matter component. This section outlines some recent and pertinent environmental studies at airports that included dispersion modeling as part of a National Ambient Air Quality Standards (NAAQS) assessment required by the National Environmental Policy Act of 1969 (NEPA) or were recommended due to agency and/or public concerns. Hartsfield-Jackson Atlanta International Airport (ATL) Unal et al., (2005) studied air quality impacts in the Atlanta ozone (O3) and particulate matter non-attainment area in relation to ATL aircraft and GSE operations. Emissions were calculated using an older version of the FOA (not FOA3) methodology within EDMS version 4.01, whereby PM2.5 emissions are computed as a function of an aircraft engine’s SN and fuel flow rates. The analysis tested two approaches with respect to SN, one of which applied a “characteristic” SN for each engine, while the other applied a mode specific SN to account for differences in engine power applied during flight procedures. Finally, the emissions were applied to a dispersion model to apportion the results relative to other sources operating in the non- attainment area, as well as to discern what sort of impact ATL has on ambient PM2.5 The dispersion model indicates that ATL aircraft contribute up to 0.13% of the total PM concentrations in its vicinity. 2.5 emissions burden in the area, non-airport area sources comprise over 90%, nonroad equipment (besides GSE) contributes 4.5%, and GSE contributes only 0.05%. Moreover, the dispersion model indicates that when using the “characteristic” SNs, the airport contributes 25 µg/m3 to the modeled concentrations at the receptor of maximum impact, although predicted concentrations are typically highly variable depending on receptor location. When applying the mode specific SNs, the impact of ATL is reduced to around 1 µg/m3 at the receptor of maximum impact. GSE also contributes an additional maximum 9 µg/m3 of PM2.5 O’Hare International Airport Modernization Program (ORD) to the modeled concentrations within 16 km of the airport property in both scenarios. PM2.5 dispersion modeling was conducted in support of the ORD Modernization Program environmental impact statement (EIS) using EDMS (U.S. EPA, 1999). In this analysis, background concentrations were developed using monitoring data available from the Illinois Environmental Protection Agency (IEPA), corresponding to 35.2 µg/m3 for evaluation against the 24-hour standard and 13.3 µg/m3 Philadelphia International Airport Capacity Enhancement Program (PHL CEP) for comparison against the annual NAAQS. Fifty-three discrete receptors were placed around the airport property line and along the terminal areas. Dispersion modeling was conducted for all development alternatives under consideration as required by NEPA, with some minor variations resulting from different development scenarios. As part of the PHL CEP EIS, dispersion modeling was conducted at PHL using EDMS to ascertain whether the planned improvements associated with PHL CEP would cause or contribute to existing or additional infractions of the PM2.5 NAAQS (U.S. FAA, 2010c). Thirty- two discrete receptors were placed around the airport property at assumed areas of maximum

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports A-24 impact on sensitive populations, including at terminal curbsides, runway ends, and in surrounding nearby residential or public use areas. For evaluation against the 24-hour PM2.5 NAAQS, a background concentration of 36.9 µg/m3 was applied based on available monitoring data for evaluation against the annual PM2.5 NAAQS, a background concentration of 15.0 µg/m3 was used. The results indicated that the point of maximum impact under the no action alternative was located at the receptor placed along the Terminal B and C curbside, returning predicted concentrations of 47.1 µg/m3 and 17.8 µg/m3 relative to the 24-hour and annual standards respectively (background inclusive). Under the preferred development option, the point of maximum impact in the build-out year 2025 shifted to the receptor located at the general aviation (GA) tarmac, with predicted concentrations of 43.5 µg/m3 and 16.6 µg/m3 relative to the 24-hour and annual standards respectively (background inclusive). With respect to the preferred development option in the build-out year (2025), the points of maximum impact shifted to the receptor located at the Centralized Headhouse surrounding Terminals 3 and 4, with predicted concentrations of 41.8 µg/m3 and 16.2 µg/m3 It was concluded that, under the preferred alternatives, emission sources at PHL would contribute between 1.3 µg/m relative to the 24-hour and annual standards respectively (background inclusive). 3 and 1.6 µg/m3 to the annual average concentration of PM2.5 Providence T.F. Green Airport (PVD) (or about 8% to 9% of the total concentration), while the remaining concentrations were attributed to background sources. In a similar way to the PHL CEP EIS, the PVD EIS (published in 2010) sought to evaluate the air quality impacts of the planned developments at the airport using a dispersion model prepared using EDMS (U.S. FAA and RIAC, 2010). The background concentrations used in the analysis were reportedly 31.1 µg/m3 and 10.6 µg/m3, respectively, for the 24-hour and annual PM2.5 NAAQS. Overall, the analysis concluded that in the build-out year 2025, the point of maximum impact would occur proximal to the main terminal building, with predicted concentrations equaling 34 µg/m3 for the 24-hour standard and 12 µg/m3 when considering the annual standard.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports B-1 APPENDIX B: CASE STUDY AIRPORTS This Appendix discusses the detailed methodology used to determine the five case study airports for the ACRP 02-23 project. These five airports were considered to offer the best opportunities to produce meaningful results for the ACRP 02-23 project. The underlying data tables are included towards the end of this Appendix. EVALUATION CRITERIA The evaluation criteria used in identifying, evaluating and selecting the case study airports were initially identified in the Proposal and restated in the Working Plan approved by the Project Panel for the ACRP 02-23 project (PPC, 2010). As such, these criteria and their application are considered to be among the most important in evaluating the impacts of airport-related fine particulate matter with a diameter of less than 2.5 micrometers (PM2.5 AIRPORT ACTIVITY LEVELS ) emissions on local air quality and assessing the potential benefits of alternative fuels on these conditions. Although it is recognized that other evaluation criteria may exist, they were not viewed as crucial. The Federal Aviation Administration (FAA) classifies U.S. airports that provide scheduled passenger services and have over 10,000 annual passenger boardings (i.e., enplanements) per year as primary airports. According to the FAA, there are 388 primary airports in the U.S. As a means of reducing the size of this list of potential candidates and more effectively applying the evaluation criteria, the median enplanement and operational levels of this group served as the threshold for this assessment. In other words, airports with enplanement and operational levels greater than 135,000 and 58,000, respectively, were included in the initial list of candidate airports. Below this level of enplanements it was judged unlikely that the airport would be contributing significantly or measurably to ambient PM2.5 From this initial screening, 138 airports met the median enplanement and operational criteria. These airports range from Hartsfield-Jackson Atlanta International Airport (ATL) with nearly one million aircraft operations to Lincoln Airport (LNK) in Nebraska with about 70,000 operations. These encompass airports of all hub sizes, operational types, geographic locations and meteorological conditions throughout the U.S. concentrations. For the purpose of the ACRP 02-23 project, these 138 airports were identified as “first-order airports” and were subjected to the remaining evaluation criteria. AVAILABILITY AND APPROPRIATENESS OF DATA Sources of Airport PM2.5 Among the “key” elements of the ACRP 02-23 project is the assessment of airports’ contribution to PM Emissions 2.5 levels, by emission source type. However, most airports comprise a varied and unique assortment of emissions sources, each with its distinctive set of PM2.5 emission rates, PM2.5 formation mechanisms and PM2.5 transport characteristics. Therefore, to properly account for the various sources of PM2.5 emissions associated with airports, the emissions and operational data for the following are required:

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports B-2 • Aircraft • Auxiliary power units (APUs) • Ground support equipment (GSE) (e.g., belt loaders, baggage tugs, aircraft tugs) • Road vehicles (e.g., private automobiles, shuttle vans, taxis, buses) • Stationary sources (e.g., boilers, cooling towers, emergency generators, fire training, incinerators) Types of Airport PM2.5 There are also three categories (or types) of “assessments” that are considered necessary for quantifying the local impacts of PM Assessments 2.5 • Air emissions inventory data – used to quantify the total amount of emissions (referred to as mass, because it considers the molecular weight of the quantity of pollutants measured) of individual sources in a defined study area (commonly expressed in tons/year, tons/day, pounds/hour). from these airport-related emission sources. Each data set is used in different ways, and has its own particular applications and limitations, which include: • Atmospheric dispersion modeling data – used to estimate the pollutant concentrations in the ambient (i.e., outdoor) air. Concentrations refer to pollutant levels that an individual would be exposed to at a specific location in the study area (commonly expressed as micrograms/cubic meter (µg/m3 • Air quality monitoring data – actual measurements of ambient pollutant concentrations at a specific location (again, commonly expressed as µg/m ). 3 Airport PM ). Although useful, the data do not readily enable the apportionment of the concentration by source. 2.5 To develop these assessments, specific sets of data are necessary, often involving extensive data gathering and development efforts such as traffic surveys, airfield simulation modeling, and on- airport surveys. Given the financial resources and timescales to obtain these data (i.e., months and years), this was considered beyond the scope of the ACRP 02-23 project. Therefore, relative to the objectives and design of the ACRP 02-23 project, the most important data needs included the following: Assessment Data Needs • Aircraft fleet mix, aircraft taxi and delay times, taxiway and runway configurations, primary taxi paths (arrival runway end to terminal to departure runway end), airfield coordinates, runway use, and temporal (i.e., hourly, daily, and monthly) operational profiles • Information on gate power and/or pre-conditioned air • GSE fleet, fuel type, equipment size, operating conditions, time of operation, and location of aircraft servicing (often by terminal area) • Road vehicle traffic volumes and operating characteristics (i.e., roadway, parking lot and curbside configurations, emission factors) • Stationary source use, equipment size, fuel type, exhaust release parameters and location • Meteorological data (e.g., wind speed, direction) Again, given the manpower, time and other resources required to obtain or develop these airport- specific “source” and “assessment” data, the availability, age, and comprehensiveness of any existing data were considered to be among the most important factors for selecting candidate

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports B-3 airports. For this reason, a search was conducted to determine which of the first-order airports had conducted air quality studies that were reasonably recent and had datasets that were potentially useful to the ACRP 02-23 project. This information was identified as essential, as the ACRP 02-23 project was not scoped to generate this information for the requisite candidate airports. Prepared for environmental impact evaluations under the National Environmental Policy Act (NEPA) or similar state level programs, as part of a State Implementation Plan (SIP) or in support of other airport-specific environmental initiatives (e.g., air quality management plans), these air quality studies were combined with expert knowledge of other potential sources of information and data. From this research, it was determined that 30 airports had air quality information in terms of emissions inventories, dispersion analyses and local background information that could be of some potential use to the ACRP 02-23 project. For the purposes of this assessment, these 30 airports were called “second-order airports.” DATA RATING INDEX To better define the value of the data, the second-order airports were assigned a data rating index (DRI) ranging from A through E representing the data type (i.e., emissions inventory, dispersion modeling and/or ambient monitoring), the availability of data and the timeliness of the data. Developed specifically for this project, Table 10 presents a description of the DRI. Many of these data elements are related to specialized studies, and it was not expected that all airports would have the information. Rather, the rating used here was designed to help show which airports already had information necessary and was not intended to be a critique of the analysis completed for any airport. By way of example, Providence T.F. Green Airport (PVD) recently completed a comprehensive emissions inventory and dispersion modeling analysis for airport sources, operates a number of PM2.5 Tab le 10 – Data Rating Index (DRI) monitoring stations near the airport and data were readily available. Thus, PVD received a DRI of “A” with respect to the ACRP 02-23 project. By comparison, Minneapolis-St Paul International Airport (MSP) conducted an ambient monitoring study in 2002, but no further publicly available airport emissions inventory and dispersion modeling analyses were found. Thus, MSP received a DRI of “E.” Rating Description A Data are available in two or three of the desired categories (i.e., emissions inventory, dispersion modeling and air monitoring). Data are recent, contain airport-specific information, and are readily available. B Data are available in one or two of the desired categories. Data are recent, contain airport-specific information, and are readily available. C Data are available in one or two of the desired categories. Data are either not recent, do not contain airport-specific information, and/or are not readily available. D Data are available in one of the three categories. Data are not recent, do not contain airport-specific information, and are not readily available. E Data are not available or limited in desired categories. Data are not recent, do not contain airport-specific information, and are not readily available.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports B-4 With respect to atmospheric dispersion modeling data, airport air quality assessments that also contain airfield simulation modeling results from models such as the Total Airspace and Airport Modeler (TAAM) or Airport and Airspace Simulation Model (SIMMOD), combined with airport-specific GSE and APU use data, and surface traffic were considered more desirable for the ACRP 02-23 project due to the higher level of accuracy. Based on data availability and the DRI outcomes, a total of 16 airports with DRIs of A, B or C were identified and designated as “third-order airports” (Figure 16). These airports were considered to be good candidates for the ACRP 02-23 project and were further evaluated as part of the screening process as discussed in the following sections. Figure 16 – Candida te Cas e Stud y Airpo rts

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports B-5 PRIMARY EVALUATION CRITERIA The following were considered to be among the most important characteristics (i.e., primary criteria) for a case study airport: • Availability of existing, recent and appropriate air quality assessment data • Willingness to participate in the ACRP 02-23 project • Current or planned alternative fuel program • Representative of other airports based on size, location, climate, etc. • PM2.5 Airport Willingness to Participate non-attainment status While every airport that was contacted expressed support for this project, some did not have the staff resources or capability to support this research or the required data. The ability to provide assistance was then considered for purposes of this study “willingness to participate” and represented an important criterion to enable the ACRP 02-23 project to be completed on time and on budget. Without this first-hand involvement and cooperation, many of the other factors (e.g., availability of data, attainment/non-attainment status, activity levels, meteorological conditions) were considered of reduced value to the ACRP 02-23 project. The reasons most likely to motivate an airport to serve as a case study airport included the following: • Contribute to and help advance environmental research. • Improve agency and public relations. • Obtain information about PM2.5 • Assess the potential benefits of an existing or planned alternative fuel program. data at the airport. Promises of cooperation were received from the following airports: • Hartsfield-Jackson Atlanta International Airport • Las Vegas McCarran International Airport • Manchester-Boston Regional Airport • Philadelphia International Airport • San Diego International Airport Alternative Fuels Programs, Plans, and Interests Consistent with the principal aim of the ACRP 02-23 project, airports actively considering or implementing alternative fuel programs were identified. Based on expert knowledge, the following airports were identified as being representative (but not inclusive) of this group: • Hartsfield-Jackson Atlanta International Airport – strong interest from Delta Airlines and State of Georgia in an alternative fuels project • Detroit Metropolitan Airport – alternative fuels and feedstock study underway • Denver International Airport – solar panel projects in place and underway • Los Angeles International – state-mandated conversion program to convert GSE to no- and low-emitting fuels • Port Authority of New York and New Jersey – has launched a study to implement a municipal solid waste (MSW) to liquid fuel project • Seattle-Tacoma International Airport – actively participating in alternative fuel projects

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports B-6 PM2.5 The PM Attainment/Non-attainment Areas 2.5 Figure 4 National Ambient Air quality Standards (NAAQS) attainment/non-attainment status of an area is important in the context of federal and state air quality regulations, SIP requirements and timetables, and the potential eligibility for the funding of alternative fuel initiatives such as the FAA Voluntary Airport Low Emissions (VALE) program for non-aircraft sources. and Figure 5 in Chapter 1 display areas of the U.S. currently in violation of the annual and 24-hour PM2.5 Of the 16 third-order airports, the following were assessed as being located in PM standards, respectively, based on recent air monitoring data. As shown, non- attainment areas are generally located in California, mid-Atlantic, Midwest, Utah, and southeastern states. 2.5 • Hartsfield-Jackson Atlanta International Airport non- attainment areas: • Chicago O’Hare International Airport • Philadelphia International Airport • Sacramento International Airport • Westchester County Airport Therefore, these locations were more likely to consider the value of different PM2.5 SECONDARY EVALUATION CRITERIA emissions reduction actions in the future, possibly including alternative fuels. In addition to the primary evaluation criteria, the following secondary criteria were also considered when evaluating potential case study airports: • Meteorology, climate and geography • Airport operational parameters • Demographics and land use • PM2.5 Meteorology, Climate and Geography ambient monitoring data Meteorological conditions (e.g., wind speed, wind direction, temperature, relative humidity, atmospheric mixing height, precipitation, sunlight) play important roles in the formation and dispersion of air pollutants (including PM2.5 For the ACRP 02-23 project, these parameters were generally categorized as “cold,” “temperate,” or “warm” based on an airport’s annual average temperature compared with the nationwide annual average. These were defined relative to the average temperature within continental U.S. (53.1ºF or 11.72 ºC) during 2009. Temperate was defined as within 2.5ºF (1.34 ºC) of the average, cold at or below 50.6ºF (10.33 ºC), and warm at or above 55.6ºF (13.11 ºC). Other meteorological data such as the number of days with temperatures greater than 90°F, the ), both regionally and locally. Climatic (e.g., continental, oceanic, mountainous) and geographic (e.g., latitude, elevation) conditions can also have an effect on fuel combustion, fuel type and fuel use, and can influence the feasibility of an alternative fuel.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports B-7 number of days with temperatures less than 32°F, average wind speed, the number of days with measureable precipitation, the average annual relative humidity, the percent of time the sun shines, and the number of heating and cooling degree days (HDD/CDD) were also considered. This information was used to identify candidate case study airports that would be representative of other airports that have similar meteorological, climatic, and geographic characteristics. Airport Operational Parameters Beyond the selection of potential case study airports based on operational levels discussed previously, the range and variation of the airport’s operational parameters are also considered important when trying to account for the application and transferability of the ACRP 02-23 project to other airports. Activity levels (e.g., aircraft landing and takeoff (LTO) cycles or enplanements) can vary widely among different commercial airports. These operational levels can also vary significantly at the same airport both temporally and spatially based on the season, runway layout, meteorology, and noise abatement procedures. As such, these factors are considered to be important when assessing the effects of PM2.5 For this assessment, airport operational and enplanement data were used as an indicator of airport activity levels (U.S. FAA, 2010d and 2010e). Similarly, an airport’s hub size (i.e., large, medium or small) was used to further categorize these airports. The types of aircraft at an airport were also considered to be an important factor, and range from commercial, commuter, air taxi, general aviation (GA), and military. on local air quality. Additionally, because aircraft PM2.5 This information was used to identify candidate case study airports with a variety of operational levels and aircraft categories that would be representative of the airports in the nationwide airport system. emissions are most notable in the taxi operating modes, data related to an airport’s taxi and ground delay time on arrival and departure (U.S. FAA, 2010f) were considered important. PM2.5 • Airports with activity levels sufficient enough to generate “measurable” air quality impacts in the vicinity of most commercial airports occurs at low levels and the particles are nearly indistinguishable from those that are associated with non-airport sources. Therefore, the assessment of the airport operational parameters mainly focused on the following: • Airports with activity levels that best represent the range of facilities that will benefit from this research (i.e., large, medium, or small) • Airports with representative aircraft GSE types • Airports with representative operational characteristics (i.e., taxi and ground delay time) Notably, even though GA airports are not specifically included, some of the candidate airports have significant GA fleets.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports B-8 Demographics and Land Use Notwithstanding land use regulations that aim to guide compatible development around U.S. airports, population densities adjacent to many of these facilities are increasing, especially near some of the oldest (e.g., Chicago-Midway, Providence T.F. Green, Dallas Love Field) and newest (e.g., Denver) facilities. Pollutant exposure to airport-related emissions is second only to noise as the principal health concern among people that live and work near airports. This is especially relevant to airport operators who must now address emerging concerns about soot, hazardous air pollutants, and PM2.5 Consequently, local population density, distribution and composition were considered. Population density (i.e., population per square mile) for cities with 100,000 or more people were determined based on available data (Census, 2000). These data were used to gauge the potential significance of the population exposures to airport-related PM , particularly among the old, very young, and infirm. 2.5 PM . The data also include the FAA Region, latitude and longitude, and elevation of the evaluated airports. 2.5 State and local environmental agencies conduct air quality monitoring in their jurisdictions on a regular and continuous basis. These monitors are typically designed to determine regional air pollution conditions while a select number are designed to measure ambient background conditions or specific air pollution sources. Ambient Monitoring Data For example, the distances from several airports to nearest air monitoring stations are as follows: • Hartsfield-Jackson Atlanta International Airport – 3.0 miles • San Diego International Airport – 2.7 miles • Philadelphia International Airport – 4.0 miles • Manchester-Boston Regional Airport – 5.0 miles • Las Vegas McCarran International Airport – 57 miles Airport-specific PM2.5 • Boston Logan International Airport – Massport is undertaking a two year air quality monitoring program at the airport. Initiated in 2007 and to be completed in late 2011, the program is intended to evaluate the effects (if any) of a new center-field taxiway on air quality (including PM air monitoring campaigns have also been carried out at several U.S. airports. The following provides a summary of the available ambient monitoring studies at these airports, at the time of writing. 2.5 • Los Angeles International Airport – since 2000, numerous air monitoring studies have been conducted around the airport in an effort to assess the air quality impacts (including PM ) in the adjoining neighborhoods. 2.5 • Provident T.F. Green Airport – from 2006 to 2007, the Rhode Island Department of Environmental Management (RI DEM) conducted an air quality monitoring program in the vicinity of the airport. Measurements of PM ) of airport operations on surrounding neighborhoods, as well as the impacts of other emission sources in the same area (i.e., surface roadways and stationary sources). 2.5, ultra-fine particulate matter, various organic compounds and meteorological data were collected.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports C-1 APPENDIX C: CASE STUDY ALTERNATIVE FUELS This Appendix discusses each of the criteria used to assess the alternative fuels described in Chapter 4. Table 11, which follows the discussions of the criterion at the end of this Appendix, presents the detailed assessment of each fuel and source combination and supports the selection of the final case study alternative fuels, as described in Chapter 4. When a particular airport is assessing whether a particular alternative fuel should be taken forward, they should not necessarily use the weightings used in the ACRP 02-23 project. Instead, they should consider their own business priorities to determine the most appropriate weightings for their own context. The “pre-weighted” information is provided for airports’ use at the end of this Appendix, and separately in the Guidance Document. CHANGE IN PM2.5 This is the most important criterion and was classed as high priority, with a weighting of 45%. The decrease in emissions was taken from the sources cited. EMISSIONS Jet-fueled Aircraft The primary domestic fuel used in commercial aircraft turbine engines is Jet-A (ASTM D1655), a fuel derived from oil. Jet A-1 is the international standard for commercial jet fuel and JP-8 is the U.S. military’s jet fuel—both are derived from oil. To be adopted, any alternative jet fuel that is developed must meet all of the specifications of jet fuel standards (e.g., freezing point, viscosity, flash point, density and sulfur content). The International Air Transport Association’s (IATA) Fact Sheet on Alternative Fuels (IATA, 2010) summarizes the requirements of jet fuel as having a freezing point below -40°C for Jet-A and -47°C for Jet A-1, not forming deposits in the engine in high-temperature locations and having an energy content of at least 42.8 MJ/kg. The two major categories of alternative jet fuels are alternative fossil fuels and biomass fuels. Examples of alternative jet fuels derived from fossil fuels include Fischer-Tropsch (FT) derived gas-to-liquid (GTL) and coal-to-liquid (CTL) fuels. Examples of biomass fuels derived using either FT or hydroprocessed renewable jet (HRJ) fuel processes include fuels derived from animal fats or plants such as sorghum, switchgrass, and camelina. ASTM International has approved an alternative jet fuel specification in annexes to ASTM D7566 for FT and HRJ fuels blended with at least 50% conventional jet fuel. In published literature, measurements for particulate matter emissions for engines are often recorded using inconsistent metrics by which to establish “thrust” setting and conversions to standard day (standard temperature and pressure) are not always stated. This creates difficulty when analyzing the results and comparing them with, for example, the emission estimates produced by the Federal Aviation Administration’s (FAA) Emissions and Dispersion Modeling System (EDMS) during a standard landing and takeoff (LTO) cycle (specified thrust settings as prescribed by the International Civil Aviation Organization (ICAO) Annex 16, Volume II, Aircraft Engines Emissions). To address this difficulty, the ACRP 02-23 project has grouped results from different studies as either low power/thrust (i.e., up to 50% thrust) or high power/thrust (i.e., over 50% thrust). This allows the relative changes in emissions (between standard fuel and an alternative fuel) at low and high thrust to be separately quantified, while still allowing some comparability with EDMS (i.e., taxi and approach would be classed as low thrust, and takeoff and climb as high thrust).

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports C-2 In addition, to quantify primary and secondary particulate matter emissions separately, it is necessary to have data not just on the changes in black carbon mass (g/kg) for estimating primary emissions, but also on sulfur and hydrocarbon (HC) emissions. This enables secondary emissions to be approximated by applying the relative change to the related secondary emissions components estimated for the base case. The main data that were available for the ACRP 02-23 project have been derived from the Alternative Aviation Fuel Experiment (AAFEX) tests conducted in 2009 at Palmdale, California (Beyersdorf and Anderson, 2009 and Bulzan et al., 2010). Emission indices were available for JP-8, FT GTL, FT CTL, 50/50 JP-8 with FT GTL and 50/50 JP-8 with FT CTL for the CFM56- 2C1 engine (i.e., a jet engine). The results of these studies generally indicate that a high reduction in emissions of particulate matter (primary emissions) is possible for neat FT fuels and medium reductions for 50/50 blends. Since neat FT fuel is reported as containing no sulfur, the sulfur emissions are assumed to be negligible and are reported to be below the range of detection by experiments conducted (Cheng, 2009; Corporan, 2007). No renewables data were available at the time of writing in a suitable format. However, as FT and HRJ fuels have similar structures, it is likely that the relative changes in emissions will be of a similar order of magnitude. Additionally, as the FOA3a equation discussed in Chapter 5 and Appendix A is used to derive particulate matter emissions, it is feasible, in theory, to reduce the sulfur content in the equation and thereby derive a relative-change emission factor for particulate matter for low-sulfur fuels. Turboprop and Turboshaft Aircraft The existing data discussed above are related to jet engines and, therefore, are not directly applicable to turboprop and turboshaft engine aircraft, given that they burn fuels differently. A few studies have been undertaken for turboshaft aircraft (Corporan and Cheng, 2010, Corporan et al., 2007, Cheng, 2009) that consider emissions of particulate matter for JP-8 and FT GTL. The results from these studies indicate a high particulate matter emission reduction for neat FT fuels and medium reductions for 50/50 blends compared with standard fuel. However, EDMS does not calculate particulate matter emissions for turboprop and turboshaft aircraft, and turboprop and turboshaft aircraft emissions were calculated as part of the sensitivity study only (refer to Chapters 5 and 6 and to Appendices D and E). Therefore FT GTL was considered on an emissions only basis as part of the sensitivity analysis for turboprop and turboshaft aircraft. Auxiliary Power Units (APUs) Very few data exist for APUs. The data that do exist (Bulzan et al., 2010) suggest that the changes in emissions are similar to those that occur for aircraft main engines at “high” thrust. AvGas Aircraft Appendix A discussed the limitations of EDMS with regard to its lack of inclusion of particulate matter emission data for piston-engine aircraft. A separate sensitivity analysis was undertaken to estimate the potential additional emissions from piston-engine aircraft using 100LL and the potential emissions reductions from 91/96UL. The emission estimates for both 100LL and 91/96UL were taken from limited data published by the Swiss Federal Office of Civil Aviation (FOCA, 2007b), which suggested soot emission factors for piston-engine aircraft (all in the

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports C-3 2.5 µm range) as reproduced in Table 8 in Appendix A. These data depict a high reduction in particulate matter emissions. It should be noted that few piston-engine aircraft are certified to run on unleaded fuel in the U.S. (although, technically, 91/96 UL AvGas grade is not completely unleaded) and there is limited availability. Ground Support Equipment (GSE) GSE tends to run primarily on diesel. However, as noted in Appendix A, there are a number of alternative-fueled equivalent models, many of which are incorporated in EDMS. These include electric, liquefied propane gas (LPG), and compressed natural gas (CNG). Not all GSE models have a relevant alternative-fueled counterpart in EDMS or generally available, so only those models with an alternative fuel equivalent were used in the ACRP 02-23 project. In terms of the alternative fuels, electric has zero particulate matter direct emissions. Relative to gasoline, LPG results in a small reduction of PM2.5 The sulfur content of fuel contributes a relatively small proportion of the total particulate matter emission formation. Therefore, low-sulfur diesel is likely to have little impact on particulate matter emissions compared with standard diesel. In addition, the legal limit of sulfur content in off-road fuel is being lowered in the future, which means that any gains of using low-sulfur diesel in GSE would only be short-term as the industry moves towards low-sulfur diesel anyway. Particulate matter emission reductions for ethanol in the ACRP 02-23 project are based on a previous literature review of data for road vehicles (AEA, 2008). Low percentage ethanol blends (up to about E10) can be used in gasoline-fueled GSE. Higher blends require some limited conversion of the vehicle, which may invalidate the vehicle’s warranty. For ethanol, limited data were found to be available with E5 and E15 data derived by scaling the relative change for E10 data. The approach for E85 used a worst case approach (i.e., the E5 scaled results), resulting in E85 appearing to have little impact on particulate matter emissions compared with E10. , while CNG performs slightly better than LPG. Diesel GSE produces the highest particulate matter emissions, typically having an emission factor more than ten times greater than that for 4-stroke gasoline engines. Therefore, replacing the fuel used in GSE from diesel to LPG or CNG will produce much higher particulate matter emissions savings compared with replacing gasoline GSE. Similarly, low percentage biodiesel blends (such as B5) may be used in diesel-fueled GSE without significant equipment concerns. Higher-percentage blends may invalidate the equipment’s warranty as discussed below under Road Vehicles. The U.S. Environmental Protection Agency (EPA) (2002) particulate matter emission reductions for road vehicles are estimated at around 6% for B10, 12% for B20 and 47% for B100. Assuming these percentages can be applied to GSE, they are smaller reductions compared with replacing diesel GSE with gasoline, LPG or CNG (where the reduction could be between 90% and 95%). In addition to changes in particulate matter emissions, practical limitations may become an issue such as higher-biodiesel blends gelling, depending on feedstock, in cold weather. Similarly, a biocide may need to be added to higher-percentage blends to prevent microbial growth and subsequent blocking of fuel filters.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports C-4 Road Vehicles Smaller road vehicles tend to run on gasoline and larger vehicles (e.g., trucks) on diesel. In terms of alternative fuels, electric road vehicles produce zero particulate matter direct emissions. The MOBILE model, which EDMS uses, incorporates emission factors for “natural gas” road vehicles as the equivalent “catch-all” for liquefied natural gas (LNG) and CNG. These emission factors can be used to generate emissions for “natural gas” for calculations prior to the 2004 model year. For MOBILE calculations after 2004, the MOBILE “natural gas” emission factors are actually higher than the corresponding emission factors for Tier 2 gasoline vehicles, which are unlikely to be realistic and, therefore, add uncertainty to these emission factors. As noted for GSE, CNG has negligibly less particulate matter emissions relative to gasoline and, therefore, replacing gasoline vehicles will have little impact. However, benefits would be seen by replacing diesel vehicles with those running on gasoline or “natural gas.” Finally, low-sulfur road diesel is already in use in the U.S. and, therefore, this cannot be classed as an alternative fuel. AVAILABILITY OF FUEL This criterion reflects the current availability of the alternative fuel. In time, this will change, especially with regard to aviation fuels, where many are only at the certification stage and not yet commercially produced. Electricity, LPG, CNG, low percentage ethanol and biodiesel blends are readily available in many U.S. states and are, therefore, classed as “high” in terms of fuel availability. Higher-percentage blends of ethanol and biodiesel should only be used in converted or new vehicles and, as such, the level of demand for these fuels is lower. Therefore, availability is more limited. As older vehicles are replaced with newer ones, where manufacturers have tested alternative fuels, demand and availability can be expected to increase. This criterion was classed as low priority with a weighting of 5%. AVAILABILITY OF NEW VEHICLES This criterion reflects the current availability of new vehicles that can use the alternative fuel in question. Again, new models will be developed over time and vehicles specifically designed for alternative fuels will become cheaper and more widespread. As discussed above, for GSE there are limitations in terms of availability and applicability – only low power electric GSE (i.e., small push-back tugs as oppose to large push-back tugs) are currently available. This criterion was classed as low priority with a weighting of 5%. COST TO CONVERT EXISTING VEHICLES In many instances, it is more cost-effective to convert an existing vehicle to run on a new or modified fuel (e.g., high-percentage blends of ethanol and biodiesel) compared with the cost of buying new – unless replacement is already under consideration for other reasons. Therefore, the cost to convert vehicles has been included as a criterion so that it can be compared against costs for new vehicles. This criterion has been classed as low priority as different airports and airside operators will have different priorities and these costs will change in future years. It has a weighting of 5%.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports C-5 DROP-IN FUEL FOR EXISTING VEHICLES If the fuel is a drop-in fuel where no new vehicles or modifications are necessary, then the only cost involved is likely to be in terms of additional fuel costs and infrastructure. Therefore, this is an important consideration. It should be noted that all aircraft fuels considered are drop-in fuels. This criterion has been classed as low priority as different airports, and airside operators will have different priorities. It has a weighting of 5%. GREENHOUSE GAS (GHG) LIFE-CYCLE EMISSIONS GHG life-cycle emissions can be difficult to quantify because they require consideration of the emissions incorporated in extraction, processing, and transmission processes as well as emissions associated with burning the fuel, sometimes referred to as “well to wheel.” Some “green” fuels, such as biodiesel, have higher upstream emissions and are, therefore, classed as having GHGs almost comparable to conventional fuels, unless high-percentage blends are used. In terms of electricity, the method used to generate the electricity can be highly variable from airport to airport and, therefore, electricity has been classed as variable. In terms of aviation biofuels, as most certified or near-certification fuels are around a 50% blend, the emissions may be comparable to conventional fuels in a similar manner to that of biodiesel. It is only when high- percentage blends are used that real savings are seen. This criterion was classed as low priority, with a weighting of 5%. EMISSION DATA SOURCE RELIABILITY This category relates to the emission data that are available for assessing the changes in particulate matter emissions. U.S. government and academic peer reviewed data are classed as high-quality data, especially if they are widely accepted and have been verified by similar studies across the world with similar results. Some government and academic peer reviewed literature relate to a limited sample such as one aircraft engine type and, therefore, are classed as medium quality data. Other data are classed as low quality because they are not specific to the source (e.g., relative changes for road vehicles applied to GSE) even though the original data may have been medium or high quality. Where no appropriate data are available, this column is classed as “N/A” (not applicable). The weighting of this criterion was classed as medium priority because it was considered to be more important than other criteria, but still less significant than the primary criterion being considered, “Change in PM2.5 FUEL COST RELATIVE TO CONVENTIONAL Emissions.” It has a weighting of 10%. This criterion reflects the current price of fuel relative to conventional fuel and is mainly based on the U.S. DOE fuel prices report (U.S. DOE, 2011). It should be noted that this will change over time, especially with regard to aviation fuels, where many are only at the certification stage and not commercially produced and are, therefore, relatively expensive. CNG, low percentage ethanol and biodiesel blends are readily available in many states and are, therefore, not excessively expensive. However, higher-percentage blends of ethanol and biodiesel have limited current commercial use and are, therefore, more expensive. In general, as demand for alternative fuels increases, it is likely that prices will decrease. This criterion was classed as low priority, with a weighting of 5%.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports C-6 VEHICLE COST COMPARED WITH CONVENTIONAL The current availability of some alternative-fueled vehicles is limited and, therefore, the cost of these vehicles is often high compared with standard vehicles of the same type. However, other conventional vehicles are now being manufactured to allow the use of higher-percentage blends of ethanol and biodiesel (sometimes called “flex-fuel” vehicles). Therefore, the price of ethanol and biodiesel compatible vehicles is not particularly high compared with other vehicles. Electric vehicles are relatively expensive, primarily due to the cost of batteries. As demand for alternative-fueled vehicles increases, prices are likely to reduce. This criterion was classed as low priority, with a weighting of 5%. ADDITIONAL INFRASTRUCTURE NEEDED This criterion allows for some consideration of the need for additional infrastructure. It is assumed that there is an existing, nearby supply of electricity, diesel, gasoline, and standard aviation fuel as appropriate. For drop-in fuels, additional infrastructure would be limited to additional storage tanks and fueling facilities. For CNG and LPG, more specialized equipment would be needed. LPG is a gas at normal temperature and pressure, but liquefies at pressures of around 10 atmospheres at 38°C (dependent on its exact composition). Hence, refueling involves pumping a liquid under pressure in a pressurized system. This requires more complex equipment than is required when filling a vehicle with gasoline or diesel. CNG is a compressed gas and will not liquefy above -82°C. Hence, refueling involves moving a highly compressed gas, typically at a pressure of around 300 atmospheres, from a high pressure storage tank. Often, CNG is delivered by a low pressure grid. Therefore, a compressor to pump the CNG into the tank and equipment to dry it are also required. This requires specialist refueling equipment that is quite different to that used for liquid hydrocarbons (HCs). An alternative refueling method is the natural gas equivalent of electrical trickle charging, where gas is slowly pumped into the vehicle (e.g., overnight). This does not require the high pressure storage tank, as gas can be supplied directly from the compressor. Recharging electric vehicles, assuming that trickle charging can be used and spare capacity already exists, requires little infrastructure development. However, issues may arise from parking vehicles and equipment for long periods and with further capacity potentially being required. Gate electricity and pre-conditioned air supply for reducing APU use are likely to require more infrastructure development than vehicle recharging points. Alternative AvGas will need separate tanks and fueling as not all piston-engine aircraft can use alternative blends. This criterion has been classed as low priority as different airports and airside operators will need to consider infrastructure differently. It has a weighting of 5%.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports C-7 WARRANTY VALIDITY ISSUE Some fuels, such as high-percentage blends of biodiesel, may be used in existing vehicles without modification, but may invalidate the vehicle’s warranty. This is an issue that vehicle owners will need to consider where warranties are still in date. Similarly, retrofitting existing vehicles for LPG or CNG use is likely to invalidate the warranty. This criterion has been classed as low priority as different airports and airside operators will have different priorities. It has a weighting of 5%. Table 11 – Alte rna tive Fu el Matrix C ri te ri on C ha ng e PM 2. 5 (H , M , L ) e m iss io ns A va ila bi lit y of fu el (H , M , L ) A va ila bi lit y of n ew v eh ic le s (H , M , L ) C os t t o co nv er t e xi st in g ve hi cl es (H ,M ,L ) D ro p- in fu el fo r ex ist in g ve hi cl e? (Y /N o r N /A ) G H G li fe -c yc le e m iss io ns (H , M , L ) Em is sio n da ta so ur ce r el ia bi lit y (H , M , L ) C os t o f f ue l c om pa re d w ith co nv en tio na l ( H , E , L ) C os t o f v eh ic le s c om pa re d w ith co nv en tio na l ( H , M , L ) A dd iti on al in fr as tr uc tu re n ee de d (H , M , L ) W ar ra nt y va lid ity is su e (Y /N o r N /A ) Low-sulfur Jet-A for aircraft L L H N/A Y H L E N/A L N FT (natural gas) aircraft M L H N/A Y H M H N/A L N FT (coal) aircraft M L H N/A Y H M H N/A L N FT (biomass) aircraft M L H N/A Y H L H N/A L N HRJ (biomass) aircraft M L H N/A Y H L H N/A L N 91/96UL AvGas for piston-engine aircraft H L L L Y H M E N/A M Y FT (natural gas) APU M L H N/A Y H L H N/A L N FT (coal) APU M L H N/A Y H M H N/A L N FT (biomass) APU M L H N/A Y H L H N/A L N HRJ (biomass) APU M L H N/A Y H L H N/A L Y Low-sulfur Jet-A for APU L L H L Y H L E N/A L N Electricity to replace some APU use M H H N/A Y V H V N/A H N Electric GSE H H H N/A N V H V H L N LPG GSE replacing gasoline GSE L H L L N H H H M M Y LPG GSE replacing diesel GSE H H L H N H H H M M Y CNG GSE replacing gasoline GSE M H L M N H H L M H Y CNG GSE replacing diesel GSE H H L H N H H L M H Y

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports C-8 C ri te ri on C ha ng e PM 2. 5 (H , M , L ) e m iss io ns A va ila bi lit y of fu el (H , M , L ) A va ila bi lit y of n ew v eh ic le s (H , M , L ) C os t t o co nv er t e xi st in g ve hi cl es (H ,M ,L ) D ro p- in fu el fo r ex ist in g ve hi cl e? (Y /N o r N /A ) G H G li fe -c yc le e m iss io ns (H , M , L ) Em is sio n da ta so ur ce r el ia bi lit y (H , M , L ) C os t o f f ue l c om pa re d w ith co nv en tio na l ( H , E , L ) C os t o f v eh ic le s c om pa re d w ith co nv en tio na l ( H , M , L ) A dd iti on al in fr as tr uc tu re n ee de d (H , M , L ) W ar ra nt y va lid ity is su e (Y /N o r N /A ) Low-sulfur diesel GSE L H H N/A Y H L E N/A L N E5 in gasoline-fueled GSE L H M N/A Y H L E N/A N/A N E10 in gasoline-fueled GSE M H M N/A Y H L E N/A N/A N E15 in gasoline-fueled GSE M L L L N H L N/A L L Y E85 in gasoline-fueled GSE L M L L N M L H L L Y B5 in diesel-fueled GSE L H H N/A Y H L E N/A N/A N B10 in diesel-fueled GSE L L M L N H L N/A L M Y B15 in diesel-fueled GSE L L M L N H L N/A L M Y B20 in diesel-fueled GSE L M M L N H L E L L Y B100 in diesel-fueled GSE M M L M N M L H L L Y Low-sulfur diesel road vehicles N/A N/A N/A N/A Y N/A N/A E N/A N/A N Natural gas road vehicles to replace diesel H H M M N H L L M H N Electric road vehicles H H L N/A N V H V H L N E5 in gasoline-fueled road vehicles L H H N/A Y H L E N/A N/A N E10 in gasoline-fueled road vehicles M H H N/A Y H M E N/A N/A N E15 in gasoline-fueled road vehicles M L H L N H L N/A L L Y E85 in gasoline-fueled road vehicles L M H L N M L H L L Y B5 in diesel-fueled road vehicles L H H N/A Y H H E N/A N/A N B10 in diesel-fueled road vehicles L L H L N H H N/A N/A M Y B15 in diesel-fueled road vehicles L L H L N H H N/A N/A M Y B20 in diesel-fueled road vehicles L M H L N H H E N/A L Y

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports C-9 C ri te ri on C ha ng e PM 2. 5 (H , M , L ) e m iss io ns A va ila bi lit y of fu el (H , M , L ) A va ila bi lit y of n ew v eh ic le s (H , M , L ) C os t t o co nv er t e xi st in g ve hi cl es (H ,M ,L ) D ro p- in fu el fo r ex ist in g ve hi cl e? (Y /N o r N /A ) G H G li fe -c yc le e m iss io ns (H , M , L ) Em is sio n da ta so ur ce r el ia bi lit y (H , M , L ) C os t o f f ue l c om pa re d w ith co nv en tio na l ( H , E , L ) C os t o f v eh ic le s c om pa re d w ith co nv en tio na l ( H , M , L ) A dd iti on al in fr as tr uc tu re n ee de d (H , M , L ) W ar ra nt y va lid ity is su e (Y /N o r N /A ) B100 in diesel-fueled road vehicles M M L M N M H H M L Y Note: Bold denotes the fuel/source combinations assessed

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports D-1 APPENDIX D: DETAILED METHODOLOGY Chapter 5 describes how the data for the base case and alternative fuel scenarios were generated. This appendix provides further details on the methodology. EMISSIONS INVENTORY FOR BASE CASE The Emissions and Dispersion Modeling System (EDMS) input data, as shown in Table 12, were obtained and organized from a variety of sources including the airport owner/operator, the airport tenants, and the Federal Aviation Administration’s (FAA) databases. This information includes aircraft activity levels, and ground support equipment (GSE) and auxiliary power unit (APU) use, road vehicle, and stationary source characteristics. Appendix A of the EDMS user manual (U.S. FAA, 2009a) provides an overview and screen shots of the data needed to compile an inventory and how to enter it into EDMS. Table 12 – Data Us ed in PM2.5 Source Category Emis s ion s Inven torie s Data Aircraft LTO by aircraft type and engine type Taxi-in, taxi-out, delay times (aircraft time in mode) Profiles of quarter hour, daily and monthly activity levels Runway and taxiway assignments and coordinates Terminal/gate assignments and locations Ground support equipment (GSE) Number and type by aircraft type Fuel type Size and load Operating times Auxiliary power units (APU) Percent of gates with fixed power units Percent of gates with fixed pre-conditioned air Road vehicles Location by segment Vehicle fleet mix by segment Roadway traffic volume by segment Average speed Emission factors (generated using either MOBILE6.2 or EMFAC2007 models) Parking facility Location by parking lot Vehicle fleet mix by parking lot Traffic volume by parking lot Travel distance Idle time Stationary sources Type and location Fuel type and quantity Stack height and diameter Exhaust temperature and velocity

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports D-2 First-Order Approximation (FOA) 3a Methodology The FOA3a methodology accounts for the volatile and non-volatile components of emissions of fine particulate matter with a diameter of less than 2.5 micrometers (PM2.5), where the fuel sulfur content, fuel-based organics, and lubricating oil contribute to volatile PM2.5. Data on the size distribution of aircraft exhaust particulate matter indicates that virtually all of the mass is associated with PM2.5 As indicated in Chapter 5, to determine the volatile PM . 2.5 PM emissions for jet engines, for each mode within the LTO cycle, the following equation was derived from the simplified version of the FOA3a function for each aircraft: sec Where: PM = (0.0085 x HC) + (3 x 1,000 x FSC x ε x F) + (1.4 x LTO) sec Note that the term 1.4 x LTO is jointly applicable to the takeoff and climb modes. = Volatile particulate matter from aircraft engines (grams) HC = Total hydrocarbon emissions from aircraft engines (grams) FSC = Fuel sulfur content, 0.00068 (assume majority is Jet-A) ε = 0.05 (assume FOA3a) F = Total fuel consumption from aircraft engines (kg) LTO = Total number of landing and takeoff (LTO) cycles for aircraft For the alternative fuels, the above three components (i.e., PM2.5 EMISSIONS INVENTORY FOR ALTERNATIVE FUEL SCENARIOS related to hydrocarbon (HC), sulfur and lubricating oil (1.4 x LTO)), in addition to the non-volatile component, have been separately scaled, based on the anticipated change in non-volatile emissions, fuel sulfur content, and HC 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 used to scale the base case emissions to the alternative fuel scenario emissions as outlined in Table 13, Table 14 and Table 15. Sources of information for the alternative fuel emission factors for each of these source types are discussed in Chapters 2, 4, and 5 and in Appendix A and Appendix C. They are also outlined in Table 13, Table 14, and Table 15.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports D-3 Table 13 – Alte rna tive Fu el Em is s ion Fac tors (AF) Ratios fo r J e t Turb ines and Turbofan Airc raft (main eng ines and APU) Fuel and Source type Source Ratio (AF) FT (natural gas) jet aircraft main engines high thrust non-volatile emissions Derived from black carbon emissions for 50/50 blend and JP-8 in a high-bypass turbofan engine, NASA report on AAFEX (Anderson et al., 2011) 0.41 FT (natural gas) jet aircraft main engines all thrust sulfur emissions As neat FT fuels have negligible sulfur, it is assumed that a 50/50 FT blend would have 50% of the base fuel’s sulfur-related emissions 0.50 FT (natural gas) jet aircraft main engines high thrust volatile emissions Derived from HC emissions for 50/50 blend and JP-8 in a high-bypass turbofan engine (Bulzan et al., 2010) 1.00 FT (natural gas) jet aircraft main engines low thrust non-volatile emissions Derived from black carbon emissions for 50/50 blend and JP-8 in a high-bypass turbofan engine, NASA report on AAFEX (Anderson et al., Feb 2011) 0.41 FT (natural gas) jet aircraft main engines low thrust volatile emissions Derived from hydrocarbon emissions for 50/50 blend and JP-8 in a high-bypass turbofan engine (Bulzan et al., 2010) 1.00 FT (coal) jet aircraft main engines high thrust non- volatile emissions Derived from black carbon emissions for 50/50 blend and JP-8 in a high-bypass turbofan engine, NASA report on AAFEX (Anderson et al., 2011) 0.59 FT (coal) jet aircraft main engines all thrust sulfur emissions As neat FT fuels have negligible sulfur, it is assumed that a 50/50 FT blend would have 50% of the base fuels sulfur-related emissions 0.50 FT (coal) jet aircraft main engines high thrust volatile emissions Derived from HC emissions for 50/50 blend and JP-8 in a high-bypass turbofan engine (Bulzan et al., 2010) 1.00 FT (coal) jet aircraft main engines low thrust non- volatile emissions Derived from black carbon emissions for 50/50 blend and JP-8 in a high-bypass turbofan engine, NASA report on AAFEX (Anderson et al., 2011) 0.53 FT (coal) jet aircraft main engines low thrust volatile emissions Derived from HC emissions for 50/50 blend and JP-8 in a high-bypass turbofan engine (Bulzan et al., 2010) 1.00 FT (coal) APU Derived from APU black carbon high load emissions for 100% FT by assuming mixed 50/50 mix and JP-8 in NASA report on AAFEX (Anderson et al., 2011) 0.56 FT (natural gas) APU Assumed as per FT (coal) APU non-volatile emissions above 0.56 Electricity to replace some APU use Assumes all APU run for 7 minutes based on EDMS defaults and FAA guidance (2010) for APU use when the availability of pre-conditioned air and gate power is available Ratio based on 7 minutes/base case time Note: aircraft high thrust is defined as takeoff and climb, and low thrust as idle and approach. For jet turbines and turbofan engine aircraft, the factors were derived from experiments on one type of high-bypass turbofan and one APU. The NASA Alternative Aviation Fuel Experiment (AAFEX) report (Anderson et al., 2011) was the primary source for the main engine and APU data. A more recent PARTNER report (Lobo, 2011), of APU alternative fuel emissions was published too late to be incorporated in the ACRP 02-23 project. Further study is needed to understand the variation that the use of these alternative fuels could have on other types of turbine engine.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports D-4 Table 14 – Alte rna tive Fu el Em is s ion Fac tors (AF) Ratios fo r o ther Aircraft Typ es Fuel and Source type Source Ratio (AF) 91/96UL AvGas for piston-engine aircraft approach Derived from 91/96UL and 100LL emission factors in Swiss Federal Office of Civil Aviation (FOCA, 2007a) 0.025 91/96UL AvGas for piston-engine aircraft idle Derived from 91/96UL and 100LL emission factors in Swiss Federal Office of Civil Aviation (FOCA, 2007a) 0.020 91/96UL AvGas for piston-engine aircraft takeoff Derived from 91/96UL and 100LL emission factors in Swiss Federal Office of Civil Aviation (FOCA, 2007a) 0.030 91/96UL AvGas for piston-engine aircraft climb Derived from 91/96UL and 100LL emission factors in Swiss Federal Office of Civil Aviation (FOCA, 2007a) 0.029 FT (natural gas) Turboprop and turboshaft aircraft high thrust non-volatile emissions Derived from black carbon emissions for 50/50 blend and JP-8 in small turboshaft engine (Corporan et al., 2007) 0.46 FT (natural gas) Turboprop and turboshaft aircraft low thrust non-volatile emissions Derived from black carbon emissions for 50/50 blend and JP-8 in small turboshaft engine (Corporan et al., 2007) 0.53 Note: aircraft high thrust is defined as takeoff and climb, and low thrust as idle and approach. The data in Table 14 have only been used as an extension to the sensitivity analysis described in Chapter 5. Tab le 15 – Alte rna tive Fu el Em is s ion Fac tors (AF) Ratios fo r GSE and Road Vehic les Fuel and Source type Source Ratio (AF) Electric GSE Used EDMS database files to replace GSE with electric equivalent where available Emissions recalculated using EDMS databases LPG GSE replacing diesel GSE Used EDMS database files to replace diesel GSE with LPG equivalent where available Emissions recalculated using EDMS databases CNG GSE replacing gasoline GSE Used EDMS database files to replace gasoline GSE with CNG equivalent where available Emissions recalculated using EDMS databases CNG GSE replacing diesel GSE Used EDMS database files to replace diesel GSE with CNG equivalent where available Emissions recalculated using EDMS databases E10 in gasoline-fueled GSE Based on E10 factor in AEA (2008) for road vehicles 0.6 B20 in diesel-fueled GSE Based on U.S. EPA (2002) exponential equation for biodiesel for road vehicles 0.880 B100 in diesel-fueled GSE Based on U.S. EPA (2002) exponential equation for biodiesel for road vehicles 0.528 Natural gas road vehicles to replace diesel Based on running MOBILE with 100% natural gas for each airport Emissions recalculated using MOBILE Electric road vehicles Electricity use has no direct PM2.5 0 emissions E10 in gasoline-fueled road vehicles Based on E10 factor in AEA (2008) for road vehicles 0.6 B20 in diesel-fueled road vehicles Based on U.S. EPA (2002) exponential equation for biodiesel for road vehicles 0.880 B100 in diesel-fueled road vehicles Based on U.S. EPA (2002) exponential equation for biodiesel for road vehicles 0.528

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports D-5 Note that as shown in Table 15, for some fuel and source combinations, the alternative emissions have been calculated using EDMS databases (e.g., GSE) or MOBILE6.2 (e.g., for road vehicles), rather than using a specific ratio (AF). Alternative Fuel Penetration For those sources that are not considered to be a drop-in fuel, the following penetration factors shown in Table 16 were used. Table 16 – Penetra tion Fac tor (P) Fuel and Source type Source and Assumption Penetration Factor (P) E10 in gasoline- fueled GSE This is a drop-in fuel and, while not all airports will provide it, it is assumed that those that do will be unlikely to have multiple fuels available. Therefore, 100% of the fleet operating at these airports would be refueling on the drop-in fuel 1 B20 in diesel-fueled GSE This is a drop-in fuel and, while not all airports will provide it, it is assumed that those that do will be unlikely to have multiple fuels available. Therefore, 100% of the fleet operating at these airports would be refueling on the drop-in fuel 1 B100 in diesel-fueled GSE This is not classed as a drop-in fuel and, as such, would need either specialist equipment or some engine modifications (though these may be fairly small), which may invalidate the warranty. While it is feasible that all diesel GSE could be modified to use B100, it is unlikely, therefore, two penetration factors have been used 1 0.5 Natural gas road vehicles to replace diesel In the U.S. airport sector, natural gas accounts for about 9% of total vehicular use (Natural Gas Vehicles for America, 2011) 0.09 Resources for the Future (2010) cites a higher scenario of 32% of the heavy-duty truck fleet fueled by natural gas in 2020 0.32 Electric road vehicles Electric Power Research Institute (EPRI, 2007) developed a scenario that, in 2020, plug-in electric hybrid vehicles will cover about 10% of light vehicle miles, with 5% of vehicle miles using electricity only. Therefore, both 10% and 5% have been used. 0.1 0.05 E10 in gasoline- fueled road vehicles U.S. EIA AEO,2011 (p84) 1 B20 in diesel-fueled road vehicles In U.S. EIA AEO, 2011 (Table 11) the ratio of biodiesel to diesel was 0.026. Assume all biodiesel is used in B20, result is 0.128 (i.e., 0.026/0.2) 0.128 U.S. EIA AEO, 2011 (p11): Based on California’s Low Carbon Fuel Standard 1 B100 in diesel-fueled road vehicles In U.S. EIA AEO, 2011 (Table 11) the ratio of biodiesel to diesel was 0.026. Assume all B100. 0.026 Biodiesel potential from soya beans (Hill et al., 2006) 0.06 ATMOSPHERIC DISPERSION MODELING ANALYSIS General EDMS and AERMOD Control Options Default EDMS aircraft engine assignments, where necessary, were based on worldwide or U.S. designations, depending on the geographic domain serviced by each case study airport. For

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports D-6 example, Hartsfield-Jackson Atlanta International Airport (ATL) used worldwide defaults, while Manchester-Boston Regional Airport (MHT) used U.S. defaults based on the service area of the airports. The determination of aircraft PM2.5 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 the sources listed in emissions and pollutant concentrations used the performance based aircraft times-in-mode with the airfield sequence model (simulation of the movement of aircraft within the airfield). Table 19 and Table 20) and AERMOD (Version 09292) run outside of EDMS. AERMOD default regulatory options (stack-tip downwash, buoyancy-induced dispersion, and final plume rise), default wind-speed profile categories and default potential temperature gradients were used. No pollutant decay was also assumed. The selection of the appropriate dispersion coefficients, accounting for terrain and atmospheric interactions in the pollutant plume, depends on the land use within 3 km of the project site. The land use typing for the dispersion analysis was based on the classification method defined by Auer (1978), using pertinent United States Geological Survey (USGS) 1:24,000 scale (7.5 minute) topographic maps of the airport areas. If the Auer land use types of heavy industrial, light-to-moderate industrial, commercial, and compact residential account for 50% or more of the total area, the U.S. Environmental Protection Agency (EPA) Guideline on Air Quality Models recommends using urban dispersion coefficients. Otherwise, the appropriate rural coefficients were used. Based on observation of the area surrounding the airport sites, rural dispersion coefficients were applied for each case study airport. Auxiliary Power Unit (APU) Operating Time For the base case, if APUs were contained in a particular aircraft, it was assumed that they operated for 26 minutes if pre-conditioned air/gate power units were not available at a particular gate. If pre-conditioned air/power units were available at a particular gate, then APUs were generally assumed to operate based on airport-specific data (about 7 minutes). These values are in line with EDMS defaults and FAA guidance (2010h) for APU use when the availability of pre-conditioned air/gate power is known. It should be noted that the AERMOD dispersion analysis was performed without incorporating particulate depletion due to gravitational settling, chemical transformation, and wet deposition. These principles are difficult to simulate. Receptors A receptor network was developed to capture and adequately define the area of maximum impact. Receptors used in the analysis include a discrete receptor grid and a polar receptor grid. The discrete receptors generally represent areas where high concentrations of pollutants are anticipated and areas where the general public has access. These receptors typically include terminal curbsides and access areas, public parking facilities near the ends of runways where aircraft are queuing and waiting to takeoff, and nearby parks, schools, and residential areas. The discrete receptor grid is described as:

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports D-7 • Boundary receptors – these receptors are located in areas along the airport boundary including runway ends at a spacing of 10° to 15°. • Terminal area receptors – these receptors are located within the main terminal curbside area. • Sensitive receptors – these receptors include schools, parks, residential areas, day-care centers and other public areas located in the vicinity of the airport. Polar grids were developed to contain 36 radii, spaced at 10°intervals. The inner polar grid extended from 500 to 2,500 meters from the center of the airports at intervals of 500 meters. The outer polar grid extended from 3,000 to 10,000 meters from the emission sources at intervals of 1,000 meters. Any polar grid receptor located on the restricted areas of airport property was discounted. This polar grid approach normalizes the spatial domain around airports of differing sizes and configurations, and it allows for a better comparison of the predicted concentrations at each airport (as the grid is exactly the same at each airport). Polar grid receptors located close to key sources of emissions (e.g., center of gates, taxiways, runways, and roadways) were removed as they would not be representative of public exposure. Terrain Data The AERMAP (Version 09040) processor was used to determine receptor elevations for all of the receptors. AERMAP uses digital elevation model (DEM) data to calculate terrain elevations and associated hill heights for use in AERMOD. DEM 1° format data within the vicinity of the airports were used. Receptors are placed at a height of 1.8 meters (typical breathing height) above ground level. Meteorological Data Meteorological conditions such as wind speed, wind direction, atmospheric stability, and air temperature were also specifically assessed for each case study airport. These meteorological data were acquired from the National Climatic Data Center (NCDC) and processed using AERMET (Version 06341), the meteorological processor contained within AERMOD. Figure 17 provides the annual wind roses for the meteorological data for the five case study airports and their analysis years. Each airport has a unique set of meteorological conditions related to wind direction, wind speed, ambient temperature, atmospheric stability, and turbulence indices. The meteorological analysis year was dictated by the emission analysis year for each case study airport. The analysis emission year was dictated by the availability of airport operational data in EDMS format. A summary of the meteorological data for each case study airport is shown in Figure 17. For example, the ATL analysis was based on operational data from a recently completed emissions inventory for 2008, while the MHT analysis was based on a recently completed emissions inventory for 2007. The meteorological data, ambient monitoring data and other year sensitive data were for the same year on a case study airport basis. The term “atmospheric mixing height” generally describes the height above ground level below which the atmospheric mixing of most air pollutants occurs (and above which little mixing occurs). Within the atmosphere, this height (expressed in meters or feet) is determined by an assortment of environmental factors including air temperature, humidity, solar radiation, wind speed, and topographic features on the ground (e.g., valleys, mountains, vegetative cover, reflective and impervious surfaces, water bodies). The atmospheric mixing height is dynamic and

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports D-8 varies both spatially and temporally throughout the day, season and year with corresponding changes in these above mentioned environmental factors. Figure 17 – Wind Ros es fo r Five Cas e Stud y Airports

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports D-9 Mixing Height In air quality assessments (i.e., emissions inventories and dispersion modeling), the atmospheric mixing height is used to define the vertical limits of a particular study area. In simple terms, this is the height of a figurative “box” within which airport-related emissions are assumed to occur and disperse, with the ground representing the bottom and the horizontal distances representing the sides of the box. As mentioned, the height of the mixing zone varies by time of day and by season. Typically, during summer daylight hours, the mixing height can be 6,500 meters. In winter, the mixing height may be as low as a hundred meters. Table 17 contains the annual average mixing heights for the five case study airports. The mixing height for Las Vegas McCarran International Airport (LAS) is generally higher due to its arid, high-altitude climate. Table 17 – Atmospheric Mixing Height (meters) ATL LAS MHT PHL SAN 811 2,207 661 796 853 To enable direct comparison of the PM2.5 emissions between the case study airports, the aircraft emissions reported in Chapter 6 were normalized to 914 meters (3,000 feet) (i.e., the mixing height was set to 914 meters for all case study airports). However, for the dispersion modeling, the original mixing heights were used. This approach was discussed with the FAA and, additionally, is in line with the EPA recommendations and the Airport Cooperative Research Program (ACRP) Report 11 on greenhouse gas (GHG) emissions inventories (ACRP, 2009). AMBIENT MONITORING DATA Local and state air protection agencies operate ambient monitoring networks to measure ambient concentrations as a means of assessing public health impacts and National Ambient Air Quality Standards (NAAQS) compliance. For PM2.5 concentrations, these ambient monitoring stations typically measure daily values at intervals of three to six days, from which the annual average concentrations are determined. Monitoring data were obtained from the EPA AirData database for stations within each case study airport’s air quality region. The representative background concentrations were then determined based on available data from the region. An ideal “background” concentration is designated by EPA as being located in a rural setting upwind of the airport. Background concentrations were added to the model estimated concentrations for the airport sources to obtain total pollutant concentrations. For the 24-hour period, the background is representative of the 98th percentile value of daily concentrations. Table 18 displays the monitoring sites near each case study airport and the value determined to represent background concentrations (shown in bold italics). Importantly, the direct use of background concentrations may cause an underestimation of the total concentrations, as contributions from non-airport sources in the same geographic area as the airport (such as power plants) may not be fully accounted for. Similarly, if a monitoring site situated very close to the airport is used as a background site, the resultant concentrations may be

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports D-10 an overestimation of the background values, as the monitoring station may already include a portion of the airport contribution. Nevertheless, given that the focus of the ACRP 02-23 project is the airport sources and the impact that alternative fuels will have on the emissions and ambient PM2.5, it is appropriate to select a representative value only so that an approximation of the total ambient PM2.5 concentrations can be obtained. Table 18 – Ambient Monitoring Data (µg/m3) Airport Monitoring Station ID Study Year Concentration (µg/m3) Location from Airport 24-hour Annual Distance (mi) Bearing ATL 13-063-0091 2008 28.2 13.6 2.8 SE 13-089-0002 19.1 11.0 8.7 ENE 13-121-0048 24.8 11.8 10.0 N 13-121-0032 24.0 13.4 12.8 N 13-089-2001 21.8 12.3 20.1 NNE LAS 32-003-0561 2008 22.5 9.07 6.2 NNE 32-003-2002 18.8 8.88 7.8 NNE 32-003-1019 12.9 4.93 23.3 SSW MHT 33-011-1015 2007 29.9 10.3 11.8 S 33-013-1006 26.6 9.67 13.8 N 33-015-0014 23.7 8.63 36.2 ENE PHL 34-015-5001 2004 29.0 12.4 3.7 SW 42-101-0136 29.5 12.7 3.9 NNE 42-101-0047 31.5 14.4 6.4 NE 42-101-0004 34.3 13.9 12.1 NE 42-101-0020 29.3 13.9 9.0 N 42-101-0024 33.4 12.8 18.6 NE 42-045-0002 30.5 15.0 7.4 WSW 34-007-0003 35.0 13.3 8.4 ENE SAN 06-073-1010 2008 24.8 13.2 3.2 SE 06-073-0006 21.5 11.8 7.9 NNE 06-073-1002 30.6 12.3 28.0 NNE 06-073-0001 22.7 12.0 10.4 SE 06-073-0003 26.0 13.4 15.0 ENE Bold italics represent data used for background estimates Source: U.S. EPA (2010b) Development of Impacts for Alternative Fuels For each scenario, the relative change in emissions between the base case and the alternative fuel scenario (scenario/base) were applied to the relevant source contribution’s dispersion modeled results for each receptor point (i.e., individual locations) as indicated in Table 19 and Table 20.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports D-11 Table 19 – Emission Changes and Application to Dispersion Source (Aircraft) Fuel and Source Type Scenario/Base Applied to Model Source FT (natural gas) jet aircraft main engines high thrust non-volatile emissions Takeoff Climb-out FT (natural gas) jet aircraft main engines all thrust sulfur emissions Approach Taxi-in Taxi-out Takeoff Climb-out FT (natural gas) jet aircraft main engines high thrust volatile emissions Takeoff Climb-out FT (natural gas) jet aircraft main engines low thrust non-volatile emissions Approach Taxi-in Taxi-out FT (natural gas) jet aircraft main engines low thrust volatile emissions Approach Taxi-in Taxi-out FT (coal) jet aircraft main engines high thrust non-volatile emissions Takeoff Climb-out FT (coal) jet aircraft main engines all thrust sulfur emissions Approach Taxi-in Taxi-out Takeoff Climb-out FT (coal) jet aircraft main engines high thrust volatile emissions Takeoff Climb-out FT (coal) jet aircraft main engines low thrust non-volatile emissions Approach Taxi-in Taxi-out FT (coal) jet aircraft main engines low thrust volatile emissions Approach Taxi-in Taxi-out 91/96UL AvGas for piston-engine aircraft approach None 91/96UL AvGas for piston-engine aircraft idle None 91/96UL AvGas for piston-engine aircraft takeoff None 91/96UL AvGas for piston-engine aircraft climb None FT (natural gas) APU Terminal/concourse FT (coal) APU Terminal/concourse FT (natural gas) turboprop and turboshaft aircraft high thrust non-volatile emissions None FT (natural gas) turboprop and turboshaft aircraft low thrust non-volatile emissions None Electricity to replace some APU use Terminal/concourse

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports D-12 Table 20 – Emission Changes and Application to Dispersion Source (Non-aircraft) Fuel and Source Type Scenario/Base Applied to Model Source Electric GSE Terminal/concourse LPG GSE replacing diesel GSE Terminal/concourse CNG GSE replacing gasoline GSE Terminal/concourse CNG GSE replacing diesel GSE Terminal/concourse E10 in gasoline-fueled GSE Terminal/concourse B20 in diesel-fueled GSE Terminal/concourse B100 in diesel-fueled GSE Terminal/concourse Natural gas road vehicles to replace diesel Internal roadways and parking (exhaust only) Electric road vehicles Internal roadways and parking (exhaust only) E10 in gasoline-fueled road vehicles Internal roadways and parking (exhaust only) B20 in diesel-fueled road vehicles Internal roadways and parking (exhaust only) B100 in diesel-fueled road vehicles Internal roadways and parking (exhaust only) SENSITIVITIES OF ANALYSIS As discussed in Chapter 2, there are four types of main aircraft engines: jet turbines, turbofans, turboprops (including turboshafts), and pistons. Within EDMS, aircraft PM2.5 emissions are generated using FOA3a based on the HC emission factor, fuel flow, and smoke number (SN) data from the International Civil Aviation Organization’s (ICAO) Engine Emissions Certification Databanks (2010). ICAO sets emission standards for jet turbines and turbofan engines greater than 26.7 kN of thrust, but not for turboprop, turboshaft or piston engines. Similarly, standards are not set for APUs. Jet Turbines and Turbofans There are some instances where EDMS does not calculate jet engine aircraft particulate matter emissions (i.e., where no ICAO related SN value exists or the value listed in ICAO is zero). Generally, EDMS uses the Calvert methodology (John, 2006), where a maximum SN exists and the relevant thrust SN is not available in ICAO. However, where the SN is listed as zero (as oppose to blank) in ICAO, EDMS assumes that the SN is zero, which results in no estimation of particulate matter emissions. The ICAO SN value is typically listed as zero because the SN is below the limit of detection for the equipment being used as oppose to actually being zero. A number of manufacturers have started to report “0.01” where the limit of detection has been reached, rather than zero. Therefore, in instances where EDMS assumes a value of zero for SN for the ICAO zeros, alternative calculations have been undertaken for the ACRP 02-23 project, using a value of 0.01 for SN. The FOA3a was then applied in its detailed form, as presented by Wayson et al., (2009), to these new SN values to calculate non-volatile and volatile particulate matter. Since Wayson does not provide an estimate of particulate matter related to lubrication oils, this was taken from the PARTNER15 Project Report (Wayson et al., 2009; Ratliff et al., 2009).

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports D-13 Turboprop and Turboshaft Aircraft As previously noted, 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. As such, a number of turboprop and turboshaft aircraft particulate matter emissions data were used (as discussed in Chapter 2 and Appendix A) to compile a list of turboprop and turboshaft engine specific emission factors (particulate matter per fuel used in grams per kilogram, where particulate matter is assumed to be equivalent to PM2.5). This list of engine emission factors was compared with turboprop and turboshaft engines utilized within the five case study airports and appropriate substitutions were made. Where there was no matching engine, an average turboprop (including turboshaft) engine PM2.5 emission factor (g/kg) was used. Of note, the FOA3a methodology is not applicable to turboprop or turboshaft engines. Consequently, no estimate of the non-volatile and volatile PM2.5 emissions from turboprop or turboshaft engines can be made. Turboprop and turboshaft aircraft account for a small percentage of the total aircraft operations at larger airports such as ATL (only 3%), but up to 30% at airports such as MHT. Piston-engine Aircraft As discussed, EDMS does not estimate PM2.5 from piston-engine aircraft. Therefore, soot emission factors (see Table 8, Appendix A) were used for all piston-engine aircraft as described in Appendix A. It was assumed that all AvGas was 100LL for the base case. Again, piston- engine aircraft account for a very small percentage of the total aircraft operations at larger airports (1%) and only 4% at MHT.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-1 APPENDIX E: CASE STUDY AIRPORT RESULTS Chapter 6 presents the case study airport emissions inventories and the range of emissions and other air pollution impacts for the alternative fuel scenarios compared to the base case. This Appendix presents detailed information relating to the base case and alternative fuel scenario emissions inventories and air quality impacts. The base case information describes relationships between the emissions from each source category for each case study airport. The detailed alternative fuels scenario information describes the change in emissions and air quality impact as a result of each scenario at each of the five case study airports. BASE CASE To provide a means of evaluating aircraft, ground support equipment (GSE) and auxiliary power unit (APU) emissions more consistently between the case study airports, the emissions of fine particulate matter with a diameter of less than 2.5 micrometers (PM2.5) were normalized as a function of the number of landing and takeoff (LTO) cycles as shown in Table 21 and Figure 18. For GSE and APU emissions, this is based on the total LTOs. However, for aircraft, this was based on the number of aircraft for which the Federal Aviation Administration’s (FAA) Emissions and Dispersion Modeling System (EDMS) calculated PM2.5 emissions. As shown, the general trends are similar for each case study airport with the highest per LTO emissions from aircraft, followed by GSE and APU. The GSE emissions per LTO are higher at Philadelphia International Airport (PHL) and lower at San Diego International Airport (SAN). The PHL analysis year was 2004 and included a disproportionate amount of diesel GSE compared to other airports, since 2004 PHL has implemented a number of alternative-fueled GSE replacements, and, therefore, the GSE analysis is not a true reflection of PHL in recent years. In addition, SAN is subject to more stringent Californian Air Resources Board (CARB) requirements on emissions control, which are designed to reduce fuel consumption and/or reduce pollutant emissions. Table 21 – PM2.5 Emissions Inventory (kg per LTO) for Aircraft-Related Sources Source Category ATL LAS PHL SAN MHT Aircraft 0.087 0.073 0.092 0.097 0.097 Ground support equipment 0.020 0.014 0.063 (a) 0.005 0.021 Auxiliary power units 0.026 0.012 0.016 0.025 0.009 (a) The PHL analysis year was 2004 and included a disproportionate amount of diesel GSE compared to other airports, since 2004 PHL has implemented a number of alternative-fueled GSE replacements and therefore the GSE analysis is not a true reflection of PHL in recent years.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-2 Figure 18 – Aircraft-Related Air Emissions Inventory (kg per LTO) (a) The PHL analysis year was 2004 and included a disproportionate amount of diesel GSE compared to other airports, since 2004 PHL has implemented a number of alternative- fueled GSE replacements, and, therefore, the GSE analysis is not a true reflection of PHL in recent years. Aircraft PM2.5 Emissions Aircraft activity levels (i.e., aircraft arrival and departure operations) and aircraft and engine assignments were developed based on airport-specific information. As shown in Table 22 the aircraft emissions were further designated by operating mode for the assessment of non-volatile and volatile PM2.5 emissions, since the level of aircraft-related emissions is a function of the time that an aircraft operates in each of the operational modes (i.e., an LTO cycle). An LTO cycle consists of the following operational modes:  “Taxi/idle/delay” includes the time an aircraft taxis between the runway and a terminal, and all ground-based delay incurred through the aircraft route. The taxi and idle delay mode includes the landing roll, which is the movement of an aircraft from touchdown through deceleration to taxi speed to full stop.  “Approach” begins when an aircraft descends below the atmospheric mixing height and ends when an aircraft touches down on a runway.  “Takeoff” begins when full power is applied to an aircraft and ends when an aircraft reaches around 500 to 1,000 feet. At this altitude, pilots typically power back for a gradual ascent.  “Climb-out” begins when an aircraft powers back from the takeoff mode and ascends above the atmospheric mixing height. Although EDMS includes aircraft engine startup, it does not calculate PM2.5 emissions during this mode.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-3 Table 22 – Annual PM2.5 Emissions Inventory by Aircraft Mode (kg) Aircraft Mode ATL LAS PHL SAN MHT Non-volatile Emissions Taxi-in 1,275 576 712 167 83 Taxi-out 836 682 679 282 73 Approach 1,274 376 1,046 512 90 Takeoff 5,453 4,728 3,319 1,738 364 Climb-out 2,868 1,675 1,364 718 170 Total non-volatile 11,706 8,037 7,120 3,417 780 Volatile Emissions Taxi-in 3,994 1,064 1,423 238 138 Taxi-out 2,463 1,414 1,345 546 121 Approach 3,849 578 1,720 797 142 Takeoff 6,025 4,563 3,328 1,701 426 Climb-out 4,121 1,948 1,710 897 245 Total volatile 20,452 9,567 9,526 4,179 1,073 Grand total 32,157 17,604 16,647 7,596 1,853 As shown in Table 22, the approach mode represents between 5% and 17%, the taxi-in mode between 5% and 16%, and the taxi-out mode between 10% and 12% of the total aircraft PM2.5 emissions. By comparison, the takeoff mode represents between 36% and 53%, and the climb- out mode represents between 18% and 22% of the total aircraft PM2.5 emissions. For the entire LTO cycle, the non-volatile PM2.5 emissions represent between 36% and 46% and the volatile PM2.5 emissions represent between 54% and 64% of the total aircraft PM2.5 emissions. To conclude, the largest proportion of aircraft emissions occur during takeoff and, in terms of volatile and non-volatile split, the volatile emissions generally dominate in all modes of the LTO cycle. These findings reveal that a large proportion of the aircraft PM2.5 emissions are related to airborne emissions, which have a smaller impact on PM2.5 concentrations at ground level, even close to the airport than ground level emissions. The EDMS aircraft PM2.5 emissions were further segregated by aircraft type, as shown in Table 23. Note that Table 23 includes data from piston-engine aircraft for PHL only as these aircraft had user defined emissions included within the EDMS emissions inventory. Importantly, EDMS designations include business jet (e.g., Gulfstream V, Learjet 35), regional jet (e.g., CRJ-900, ERJ145, CL601), small jet (e.g., MD-83, 757-200, A320-200), medium jet (e.g., 767-300, 767-400, A330-300), and large jet (e.g., 777-300, A340-200). These results show that small jets account for the largest amount of aircraft PM2.5 emissions (between 48% and 86%, depending on the airport). This is followed by medium sized and regional jets. Business and large jets make up the remaining percentage and their contribution is mostly dependent of the type of service at the airport (e.g., commercial, cargo).

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-4 Table 23 – Annual PM2.5 Emissions Inventory by Aircraft Type (kg) Aircraft Type ATL LAS PHL SAN MHT Business jet 1,298 771 408 366 178 Regional jet 8,756 342 2,037 812 357 Small jet 15,382 15,225 10,391 5,606 1,101 Medium jet 4,108 632 2,171 728 161 Large jet 2,612 634 1,371 85 56 User defined piston 0 0 269 0 0 Total 32,157 17,604 16,647 7,596 1,853 Note: EDMS 5.1.2 designations include business jet (Gulfstream V, Learjet 35), regional jet (CRJ- 900, ERJ145, CL601), small jet (MD-83, 757-200, A320-200), medium jet (767-300, 767-400, A330-300), and large jet (777-300, A340-200) Again, to provide a means of comparing airports, the aircraft emissions normalized per jet LTO by aircraft type are displayed in Table 24 and Figure 19. 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 (see the Sensitivities of Analysis section at the end of this Appendix for an estimate of these aircraft emissions). Therefore, the emissions have been divided by the number of LTOs for aircraft that have PM2.5 emissions data in EDMS. The number of LTOs for some aircraft types at some of the case study airports is very low (such as medium jets at MHT). Thus, the emissions per LTO may be of limited value in these cases. In general, small, business and regional jets have lower PM2.5 emission per LTO compared with larger jets and PM2.5 emissions per LTO generally increase from small, to medium, to large jets. Table 24 – PM2.5 Emissions Inventory by Aircraft Type (kg per LTO) Aircraft Type ATL LAS PHL SAN MHT Business jet 0.109† 0.062† 0.074† 0.081† 0.082 Regional jet 0.047 0.073† 0.036 0.083† 0.167 Small jet 0.104 0.071 0.100 0.096 0.071 Medium jet 0.180† 0.092† 0.187† 0.218† 0.551† Large jet 0.437† 0.320† 0.715† 0.038† 0.120† Weighted Average 0.087 0.073 0.092 0.097 0.097 Note: EDMS5.1.2 designations include business jet (Gulfstream V, Learjet 35), regional jet (CRJ- 900, ERJ145, CL601), small jet (MD-83, 757-200, A320-200), medium jet (767-300, 767-400, A330-300), and large jet (777-300, A340-200). † Based on less than 10% of the total aircraft LTOs.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-5 Figure 19 – PM2.5 Emissions Inventory by Aircraft Type (kg per LTO) Ground Support Equipment The type of GSE analyzed included aircraft tugs, baggage tugs, belt loaders, fuel trucks, water trucks, lavatory trucks and cargo loaders. Air emissions resulting from the operation of GSE vary depending on the type of equipment, fuel type (e.g., gasoline, diesel, propane, electric), and the duration of equipment operation (engine run time). The type of GSE used depends on the aircraft type and the designated category of an aircraft operation (e.g., passenger, cargo). The results in Table 25 show that for GSE emissions, the majority (i.e., greater than 70%) of PM2.5 emissions are due to the operation of diesel-fueled equipment. Each of the five case study airports has its own airport-specific GSE fleet and fuel type, the latter partly a function of the geographic location, the airline’s preferences and the air quality regulations in the region. Secondly, the unavailability of airport-specific data required the use of EDMS default data, which can modify the results of an emissions inventory. Table 25 – Annual PM2.5 Emissions Inventory by GSE Fuel Type (kg) Fuel Type ATL LAS PHL SAN MHT Gasoline 3,230 713 555 19 102 Diesel 6,600 3,300 14,385 (a) 559 843 CNG - - - 3 - LPG - 101 - - - Total 9,829 4,114 14,940 582 945 Note: A dash (-) indicates that no data were available for these sources (a) The PHL analysis year was 2004 and included a disproportionate amount of diesel GSE compared to other airports, since 2004 PHL has implemented a number of alternative-fueled GSE replacements, and, therefore, the GSE analysis is not a true reflection of PHL in recent years.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-6 Table 26 contains the GSE emissions data by GSE type and fuel type. Electric equipment is not listed as it does not emit direct PM2.5 emissions. The majority of the PM2.5 emissions is attributable to aircraft tractors, baggage tractors, belt loaders, and support trucks (e.g., catering, cabin service and lavatory). Again, each of the five case study airports has its own specific GSE fleet mix and some airports have more or less gasoline/diesel equipment, some require deicers or air-conditioners and some have greater use of ground power units. For example, the PHL analysis year was 2004 and included a disproportionate amount of diesel GSE compared to other airports, since 2004 PHL has implemented a number of alternative-fueled GSE replacements and therefore the GSE analysis is not a true reflection of PHL in recent years. To some degree, the GSE emissions are also a function of the data (i.e., using default data or the collection of actual GSE vehicle types, fuel types and operating times) used for this analysis. For example, some airports have a greater amount of airport-specific data available, and this may yield estimates closer to actual values. In other words, the greater use of default data may result in underestimation or overestimation of the actual emissions. Nevertheless, the general conclusions on fuel and equipment trends and the resultant PM2.5 emissions still apply. Table 26 – Annual PM2.5 Emissions Inventory by GSE and Fuel Type (kg) GSE Type Fuel Type ATL LAS PHL (a) SAN MHT Aircraft tractor Diesel 2,535 556 1,002 219 110 Gasoline - 2 5 - - Air-conditioner Diesel - 8 - - - Air start Diesel 276 6 2,260 29 78 Gasoline - - - <0.5 - Baggage tractor Diesel - 421 5,174 55 93 Gasoline 1,479 150 154 6 37 LPG - 1 - - - Belt loader Diesel - 984 2,892 37 169 Gasoline 1,731 427 43 3 16 LPG - 1 - - - Cabin service truck Diesel 1,098 451 558 - 67 Gasoline - - 9 - - Cargo loader Diesel 982 206 489 32 57 Gasoline - - 4 - - Cargo tractor Diesel - 41 - - - Gasoline - 5 - - - Cart Gasoline - 1 - <0.5 - Catering truck Diesel 1,370 176 422 - 36 Gasoline - 27 6 5 - LPG - 98 - - - Deicer Gasoline - 3 - - 22 Fork lift Diesel 8 - - 3 - Gasoline - 1 - - -

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-7 GSE Type Fuel Type ATL LAS PHL (a) SAN MHT CNG - - - 3 - LPG - 2 - - - Fuel truck Diesel 117 90 381 33 86 Gasoline - - 25 <0.5 - Generator Diesel - 58 - - - Gasoline - - - <0.5 - Ground power unit Diesel 26 131 953 147 89 Gasoline 20 26 5 2 26 Hydrant truck Diesel - - 152 - - Lavatory truck Diesel 144 14 57 1 20 Gasoline - 23 49 <0.5 1 Lift Diesel - 35 - <0.5 - Gasoline - - - <0.5 - Other Diesel - 25 - 1 - Passenger stand Diesel - - - 1 - Gasoline - - - <0.5 - Service truck Diesel 43 97 46 - 37 Gasoline - 48 127 1 - Sweeper Diesel - 3 - - - Water truck Gasoline - <0.5 128 - - Total 9,829 4,114 14,940 582 945 Note: A dash (-) indicates that no data were available for these sources, either because the airport does not have the specified equipment/fuel type or the airport did not provide data for the specified equipment/fuel type (a) The PHL analysis year was 2004 and included a disproportionate amount of diesel GSE compared to other airports, since 2004 PHL has implemented a number of alternative-fueled GSE replacements, and, therefore, the GSE analysis is not a true reflection of PHL in recent years. Auxiliary Power Units APUs are onboard aircraft engines that provide power, heat and air-conditioning to aircraft while taxiing or at the terminal gate. APUs can also be used to start the engines before departing from the gate area. EDMS assigns default APUs based on aircraft assignments and also includes pollutant emission factors corresponding to the horsepower for each unit. Table 27 contains the PM2.5 emissions inventory results for APUs. The APU emissions tend to be higher for the larger airports and lower for the smaller airports. Table 27 – Annual PM2.5 Emissions Inventory (kg) for APU Source Category ATL LAS PHL SAN MHT Auxiliary power units 12,617 3,800 3,802 2,730 425

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-8 Road Vehicles For most airports, road vehicle emissions occur at parking lots and at on-airport and off-airport roadways, as shown in Table 28. The level of emissions that result from the operation of airport- related road vehicles depends on several factors, including:  Volume of road vehicles  Vehicle fleet mix  Road vehicle emission factors  Travel distance  Vehicle speed  Vehicle model year Road vehicles include privately owned vehicles (e.g. cars, vans, trucks, cabs, rental cars), mass transit vehicles (e.g., buses and vans), airport-owned vehicles (e.g., shuttles, buses), and delivery vehicles (e.g., trucks and vans). Emissions from road vehicles in parking facilities are a result of:  Total time vehicles spend idling in a parking facility  Distance vehicles travel in the facility  Speed of the vehicles  Type of vehicle Table 28 – Annual PM2.5 Emissions Inventory for Road Vehicles (kg) Roadways and Parking Lots ATL LAS PHL SAN MHT On-Airport Emissions Exhaust (non-volatile) 995 128 846 277 52 Exhaust (volatile) 474 18 416 70 20 Brake And Tire 634 243 283 60 25 Total Onsite 2,103 389 1,545 407 96 Off-Airport Emissions Exhaust (non-volatile) 11,802 1,349 6,749 1,109 22 Exhaust (volatile) 4,699 401 3,216 455 8 Brake and tire 5,265 1,325 2,256 462 10 Total Offsite 21,766 3,073 12,221 2,026 41 Roadway and parking lot road vehicle emissions were also computed on a per enplanement basis (i.e., passenger) for each airport to provide a better means of comparison (see Table 29). Airport enplanement values are readily available, and those values provide a general estimate of the roadway emissions. As the geographic roadway coverage varies from airport to airport, the emission rates per vehicle mile were also calculated, resulting in a range from 0.02 to 0.07 grams per vehicle mile. However, an airport’s vehicle miles traveled are not always readily available (where spatial data are not available) but, as it provides a more precise comparison of the roadway emissions between airports, it has been included.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-9 Table 29 – Road vehicle PM2.5 Emissions Inventory – Emission Factors Source Category ATL LAS PHL SAN MHT Emissions (kg) per 1,000 enplanement On-airport roadways 0.03 0.01 0.06 0.02 0.05 Off-airport roadways 0.28 0.07 0.5 0.11 0.02 Emissions (grams) per vehicle mile On-airport roadways 0.04 0.02 0.05 0.07 0.03 Off-airport roadways 0.03 0.02 0.07 0.03 0.03 Stationary Sources and Fire Training Stationary sources of PM2.5 emissions at airports include boilers, generators, snow melters, cooling towers, painting operations, aircraft engine test facilities, and fire training facilities. These sources, their size, fuel type, and use can vary greatly from airport to airport as a result of the airport’s size and climate. Of the five case study airports, LAS has a number of on-airport boilers and cooling towers and PHL includes a fire training facility (see Table 30). Table 30 – Annual PM2.5 Emissions Inventory (kg) for Stationary Sources and Fire Training Source Category ATL LAS PHL SAN MHT Stationary Sources Boilers 218 1,153 392 90 47 Generators 230 642 - 498 12 Cooling Towers - 3,664 - - - Miscellaneous - - - - 6 Total Stationary Sources 448 5,459 392 588 66 Fire Training Fire Training - - 2,819 - - Note: A dash (-) indicates that no data were available for these sources, either because the airport does not have the specified emission sources or the airport did not provide data for the specified emission sources Ambient Monitoring Data Background PM2.5 concentrations were determined for each case study airport. Table 31 contains those 24-hour concentrations (noted as the maximum of the 98th percentile) and the annual average PM2.5 concentrations. These data are used in support of the dispersion modeling discussed in the next section. Table 31 – Background Concentration (µg/m3) Source ATL LAS PHL SAN MHT 24-hour 19.1 12.9 29.0 21.5 23.7 Annual 11.0 4.9 12.4 11.8 8.6 Source: U.S. EPA (2010b)

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-10 Dispersion Modeling Analysis The AERMOD dispersion model incorporated in EDMS was used to estimate ambient (i.e., outdoor) concentrations of PM2.5 on, and in the vicinities 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. AERMOD is the EPA preferred dispersion model for general industrial sources. The dispersion modeling results are represented by the concentrations for the receptor at which the maximum concentration occurs, the radius of influence (ROI, defined as the distance that extends from the source (in this case, the airport reference point) to the farthest receptor distance at which the source has a concentration greater than a specific threshold for a given pollutant) and the influence area (i.e., the area within a threshold concentration level). Depending on the location of a receptor, the concentration (maximum or otherwise) may be dominated by aircraft, GSE and APU, road vehicle or stationary sources. Average Annual PM2.5 Concentrations The dispersion modeling analysis computed annual average PM2.5 concentrations in micrograms per cubic meter (µg/m3) for each receptor contained in the discrete and polar receptor grids set up for the case study airports. Table 32 displays, the maximum airport-related annual PM2.5 concentrations and the annual PM2.5 concentrations by source category, for the same overall maximum receptor, and for each airport. The background PM2.5 and the total PM2.5 concentrations are also shown in Table 32. The contributions of source categories to the maximum receptor concentrations are highly variable and depend on the exact location of the receptor and the receptors proximity to the various source categories. Table 32 – PM2.5 Dispersion Modeling Results – Maximum Annual (µg/m 3) Source ATL LAS PHL SAN MHT Aircraft 0.27 0.20 0.21 0.07 0.03 Gates 1.03 0.63 0.75 1.38 2.33 Roadways 0.12 0.46 0.28 0.75 0.01 Parking facilities 0.27 0.03 0.01 0.20 0.02 Stationary sources and fire training 0.01 1.46 0.01 0.07 0.00 Subtotal 1.70 2.77 1.26 2.47 2.40 Background 11.00 4.93 12.40 11.80 8.63 Total 12.70 7.70 13.67 14.27 11.03 NAAQS 15.0 15.0 15.0 15.0 15.0 As shown, the gate activities (i.e., GSE and APU) generally contribute the greatest percentage to the overall airport-related PM2.5 concentrations. Exceptions include LAS where stationary sources are a majority. These results are highly dependent on source emission strengths, source locations relative to receptors and meteorological data. Given MHT’s small size, the location of the public access receptors may be closer to the apron area and, thus, have a large contribution of

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-11 GSE and APU impact at the maximum receptor. Conversely, ATL may have a greater distance between the aprons and the public access receptors and, thus, have a smaller percentage contribution from GSE and APU. Gate-related PM2.5 emissions extend over a wide area. However, PM2.5 emissions from aircraft, roadways and stationary sources are generally confined to the runways and taxiways, along the roadways and in other isolated areas of the airport. The concentrations associated with these sources disperse more rapidly with distance. Table 33 and Figure 20 present the maximum annual concentration (airport contribution only) as a function of distance from the airport reference point (as provided within EDMS). Table 33 – PM2.5 Dispersion Modeling Results – Maximum Annual (µg/m 3) by Distance (meters) Distance (km) ATL LAS PHL SAN MHT 0.5 - - - 1.28 1.28 1.0 - 2.77 - 2.47 2.40 1.5 - 2.14 1.02 2.34 1.59 2.0 1.70 1.25 1.26 1.41 0.27 2.5 1.23 1.17 0.82 0.71 0.17 3.0 0.79 1.25 0.75 0.46 0.11 4.0 0.39 1.01 0.30 0.30 0.07 5.0 0.30 0.26 0.21 0.21 0.04 6.0 0.20 0.20 0.14 0.16 0.03 7.0 0.17 0.16 0.11 0.14 0.02 8.0 0.15 0.13 0.08 0.11 0.02 9.0 0.11 0.10 0.07 0.09 0.01 10.0 0.10 0.09 0.06 0.07 0.01 Note: A dash (-) indicates within airport property. Values do not include background concentrations. As shown in Figure 0, the annual average concentration at a distance of 5 km is much lower than the overall maximum concentration.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-12 Figure 20 – Air Dispersion Modeling Results – Annual Concentration (µg/m3) by Distance (meters) 24-hour PM2.5 Concentrations Table 34 presents the maximum 98th percentile 24-hour PM2.5 (represented as the seventh highest value at any given receptor) concentrations for each case study airport, the background concentration and the total concentration. These results are further broken down by each airport source category. The contributions of source categories to the maximum receptor concentrations are highly variable and depend on the exact location of the receptor and the receptors’ proximity to the various source categories. As shown, the terminal gate sources (i.e., GSE and APU) again generally contribute the greatest percentage to the overall concentration. The exception is LAS, where stationary sources are a majority. As with the annual PM2.5 concentration results, the gate source emissions are spread out over a wide area, while the airport runway and taxiway, roadway and stationary sources are comparatively more confined in their spatial distribution, especially at ground level. Table 34 – PM2.5 Dispersion Modeling Results – Maximum 24-hour (µg/m 3) Source ATL LAS PHL SAN MHT Aircraft 0.96 0.34 0.46 0.33 0.01 Gates 4.54 1.50 4.37 3.50 4.10 Roadways 0.58 <0.01 0.24 0.07 <0.01 Parking facilities 0.25 0.08 0.04 0.02 <0.01 Stationary sources 0.01 2.71 0.03 0.07 0.03 Subtotal 6.34 4.63 5.14 3.99 4.15 Background 19.1 12.9 29.0 21.5 23.7 Total 25.4 17.5 34.1 25.5 27.9 NAAQS 35.0 35.0 35.0 35.0 35.0

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-13 Table 35 and Figure 21 present the maximum 24-hour 98th percentile concentration (airport contribution only) as a function of distance from the airport center point. Table 35 – PM2.5 Dispersion Modeling Results – Maximum 24-hour (µg/m 3) by Distance (meters) Distance (km) ATL LAS PHL SAN MHT 0.5 - - - 2.24 4.15 1.0 - 4.63 - 3.99 3.90 1.5 - 3.85 4.53 3.62 3.62 2.0 6.34 2.54 5.14 2.32 0.95 2.5 5.52 2.95 4.46 1.68 0.64 3.0 3.64 2.67 3.23 1.11 0.49 4.0 2.11 2.01 1.95 0.82 0.31 5.0 1.53 0.69 1.30 0.57 0.23 6.0 1.04 0.51 1.09 0.56 0.18 7.0 0.92 0.43 0.99 0.46 0.15 8.0 0.88 0.35 0.70 0.40 0.12 9.0 0.73 0.29 0.56 0.32 0.09 10.0 0.65 0.25 0.52 0.30 0.08 Note: A dash (-) indicates within airport property. Values do not include background concentrations Figure 21 – Air Dispersion Modeling Results – Maximum 24-hour Concentration (µg/m3) by Distance (meters) Radius of Influence (ROI) For stationary source permitting, the EPA’s Prevention of Significant Deterioration (PSD) Program defines a pollutant concentration threshold to estimate the ROI. For PM2.5, these concentration thresholds are 1.2 µg/m3 for a 24-hour 98th percentile concentration and 0.3 µg/m3 for annual concentrations (U.S. EPA, 2010d). The ROI is the furthest distance from a source (in this case, the airport reference point within EDMS) where all receptor concentrations are below

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-14 the threshold. Thus, concentrations at all receptors beyond this distance would be below the threshold. However, not all receptors within this distance would be above the thresholds as source location and meteorological conditions affect the direction and extent of the concentrations. These ROI concentration thresholds are a small fraction (about 5%) of the NAAQS and could be considered analogous to an audible noise level. The ROI concentration thresholds are not to be compared or related to any significance threshold for the National Environmental Policy Act (NEPA) and project-related effects. Table 36 provides a summary of the ROI for each of the five case study airports. Table 36 – PM2.5 Dispersion Modeling Results – Radius of Influence (meters) Averaging Period ATL LAS PHL SAN MHT 24-hour 5,674 4,614 5,487 2,920 1,953 Annual 4,968 4,952 4,012 4,018 1,988 Influence Area Based on the concentrations and the ROI concentration thresholds, an influence area (in acres) can be determined. The influence area is designated as the area represented by the receptors above the ROI concentration thresholds. Table 37 provides a summary of the influence area for each of the five case study airports. This influence area is a subset of the ROI. Thus, the concentration isopleth fits within the ROI with the furthest extent of the isopleth equal to the ROI, but the area represents only those receptors greater than the ROI concentration threshold. Of note, a portion of the influence area would be on-airport and off-airport and, in part, is a function of the size of the particular airport. Table 37 – PM2.5 Dispersion Modeling Results – Influence Area (acres) Averaging Period ATL LAS PHL SAN MHT 24-hour 14,322 3,647 11,515 1,390 364 Annual 9,129 5,079 6,445 2,511 454 SENSITIVITIES OF ANALYSIS As noted in Chapter 4, 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. The following section presents a sensitivity analysis of the estimated PM2.5 from those aircraft for which EDMS does not include emission calculations. This comparison demonstrates the sensitivity of the local PM2.5 impacts from an aircraft fleet mix at airports with high numbers of turboprop, turboshaft and piston-engine aircraft, and other aircraft not estimated by EDMS (see Table 38). The second portion of Table 38 shows the number of aircraft LTO cycles for which EDMS does and does not provide PM2.5 emission calculations. This potential issue has implications for the calculated ambient (dispersion modeled) concentrations.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-15 Table 38 – Annual PM2.5 Emissions Inventory (kg) for Aircraft Condition ATL LAS PHL SAN MHT Aircraft Emissions PM2.5 emissions within EDMS 32,157 17,604 16,647 7,596 1,853 PM2.5 emissions not within EDMS (aircraft specific) 8,920 2,597 4,906 2,272 2,432 PM2.5 emissions not within EDMS (scaled) 183 421 1,033 151 539 Total 41,260 20,622 22,586 10,019 4,823 Aircraft LTOs PM2.5 emissions within EDMS 340,180 191,690 182,560 59,892 19,117 PM2.5 emissions not within EDMS 148,920 112,696 54,678 50,055 26,719 Total 489,100 304,386 237,238 109,947 45,836 Table 39 depicts the percentage of increased emissions relative to the EDMS-generated emissions. Results from the sensitivity analysis indicate that, at the case study airports, aircraft emissions could more than 17% higher than reported by EDMS. At smaller airports, where the proportion of piston-engine, turboprop and turboshaft aircraft is much higher, the increase in aircraft emissions could be much more substantial. Table 39 – Annual PM2.5 Emissions Inventory (percent emissions increase) for Aircraft Condition ATL LAS PHL SAN MHT PM2.5 emissions not within EDMS (aircraft specific) 28 15 29 30 131 PM2.5 emissions not within EDMS (scaled) 1 2 6 2 29 Total 29 17 35 32 160 Of note, the dispersion modeling analysis shows that aircraft activities do not provide the largest contribution to the maximum concentrations. This remains the case even if the results are scaled in line with the above uncertainties. This is due to the distance the emissions have to travel to the receptor after the aircraft leaves the ground. Secondly, the largest aircraft emissions contribution is during the takeoff mode and other above-ground operating conditions. Thus, although important to the magnitude of the emissions inventory, the discounting of turboprop, turboshaft, and piston-engine aircraft PM2.5 emissions inherent in the EDMS may not change the dispersion modeling results or conclusions materially. As described in Chapter 5, the aircraft emissions inventory was conducted using an atmospheric mixing height of 3,000 feet (the default value within EDMS) for all airports. However, actual mixing heights vary as a function of climate, geography, altitude, ground cover, and the proximity to urban areas and bodies of water. Table 40 provides a comparison between the aircraft emissions with the airport-specific mixing height and the 3,000 foot default value. The mixing heights for LAS are generally higher due to the arid, high-altitude climate. As a result, the actual emissions are much greater for LAS than those assuming a mixing height of 3,000 feet.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-16 Table 40 – Annual PM2.5 Emissions Inventory (kg) for Aircraft by Mixing Height Condition ATL LAS PHL SAN MHT Actual Mixing Height 30,787 24,074 16,250 6,783 1,768 Mixing Height of 3,000 feet 32,157 17,604 16,647 7,596 1,853 ALTERNATIVE FUEL SCENARIOS The following tables present the detailed data and results of the base case and alternative fuel scenarios for each of the five case study airports. These results provide the data from which the summaries were presented in Chapter 6. Table 41 presents the airport wide annual emissions inventory for the base case and alternative fuel scenarios, while Table 42 presents the magnitude change and percentage change in airport wide annual emissions for the alternative fuel scenarios compared with the base case. A similar set of tables are presented for the following:  Table 43 and Table 44 present the maximum 24-hour 98th percentile concentration results.Note: The implied increase in emissions for the “100% CNG GSE replacing gasoline GSE, where model available” scenario is a theoretical modeling output related to the emission factor source data used, and is not likely to be observed in actual practice.  Table 45 and Table 46 present the annual maximum concentration results.  Table 47 and Table 48 present the 24-hour 98th percentile ROI.  Table 49 and Table 50 present the annual ROI.  Table 51 and Table 52 present the 24-hour 98th percentile influence area.  Table 53 and Table 54 present the annual influence area. Of note, the results represent analysis of the five case study airports; other airports may fall into a similar range of results or outside the range found for the case study airports, depending on their specific operating conditions. Table 55 presents the emissions inventory for the alternative fuel scenarios related to turboprop (including turboshaft) and piston-engine aircraft.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-17 Table 41 – Annual PM2.5 Emissions Inventory for Base Case and Alternative Fuel Scenarios (kg) Condition ATL LAS PHL SAN MHT Base case 78,920 34,440 52,366 13,928 3,426 100% FT (natural gas) aircraft and APU 56,748 23,504 42,235 8,773 2,288 100% FT (coal) aircraft and APU 58,655 24,849 43,339 9,331 2,414 100% Gate power and pre-conditioned air (PCA) to replace APU use 72,830 33,267 50,822 12,346 3,332 100% Electric GSE, where model available 69,535 30,740 40,416 13,421 2,704 100% LPG GSE replacing diesel GSE, where model available 74,528 32,658 41,615 13,551 2,801 100% CNG GSE replacing gasoline GSE, where model available 78,594 34,386 52,212 14,249 3,373 100% CNG GSE replacing diesel GSE, where model available 74,528 32,658 41,615 13,551 2,801 E10 in 100% gasoline-fueled GSE 77,628 34,421 52,144 13,927 3,385 B20 in 100% diesel-fueled GSE 78,129 34,415 50,641 13,895 3,325 B100 in 100% diesel-fueled GSE 75,806 34,342 45,578 13,795 3,028 B100 in 50% diesel-fueled GSE 77,363 34,391 48,972 13,862 3,227 Natural gas road vehicles to replace diesel (9% of market) 78,920 34,440 52,365 13,928 3,426 Natural gas road vehicles to replace diesel (32% of market 78,918 34,440 52,365 13,928 3,426 Natural gas road vehicles to replace diesel (100% of market) 78,914 34,440 52,363 13,928 3,326 Electric road vehicles (5% of market) 78,847 34,433 52,302 13,911 3,423 Electric road vehicles (10% of market) 78,773 34,425 52,239 13,894 3,419 Electric road vehicles (100% of market) 77,452 34,294 51,104 13,581 3,355 E10 in 100% gasoline-fueled road vehicles 78,793 34,383 52,251 13,793 3,419 B20 in diesel-fueled road vehicles (12.8% of market) 78,911 34,440 52,351 13,928 3,425 B20 in diesel-fueled road vehicles (100% of market) 78,851 34,439 52,249 13,927 3,420 B100 in diesel-fueled road vehicles (2.6% of market) 78,913 34,440 52,354 13,928 3,425 B100 in diesel-fueled road vehicles (6% of market) 78,904 34,440 52,338 13,928 3,425 B100 in diesel-fueled road vehicles (100% of market) 78,647 34,438 51,905 13,924 3,401 Note: The implied increase in emissions for the “100% CNG GSE replacing gasoline GSE, where model available” scenario is a theoretical modeling output related to the emission factor source data used, and is not likely to be observed in actual practice.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-18 Table 42 – Alternative Fuel Scenario versus Base Case Annual PM2.5 Emissions Inventory Condition Change in Annual Emissions (kg) from Base Case Percent Change from Base Case ATL LAS PHL SAN MHT ATL LAS PHL SAN MHT 100% FT (natural gas) aircraft and APU -22,172 -10,936 -10,131 -5,156 -1,138 -28 -32 -19 -37 -33 100% FT (coal) aircraft and APU -20,265 -9,591 -9,027 -4,597 -1,012 -26 -28 -17 -33 -30 100% Gate power and PCA to replace APU use -6,090 -1,173 -1,544 -1,582 -94 -8 -3 -3 -11 -3 100% Electric GSE, where model available -9,385 -3,700 -11,950 -508 -722 -12 -11 -23 -4 -21 100% LPG GSE replacing diesel GSE, where model available -4,392 -1,781 -10,751 -378 -626 -6 -5 -21 -3 -18 100% CNG GSE replacing gasoline GSE, where model available -326 -54 -154 321 -53 0 0 0 2 -2 100% CNG GSE replacing diesel GSE, where model available -4,392 -1,781 -10,751 -378 -626 -6 -5 -21 -3 -18 E10 in 100% gasoline-fueled GSE -1,292 -19 -222 -2 -41 -2 0 0 0 -1 B20 in 100% diesel-fueled GSE -791 -25 -1,725 -34 -101 -1 0 -3 0 -3 B100 in 100% diesel-fueled GSE -3,114 -98 -6,788 -133 -398 -4 0 -13 -1 -12 B100 in 50% diesel-fueled GSE -1,557 -49 -3,394 -66 -199 -2 0 -6 0 -6 Natural gas road vehicles to replace diesel (9% of market) -1 0 -1 0 0 0 0 0 0 0 Natural gas road vehicles to replace diesel (32% of market -2 0 -1 0 0 0 0 0 0 0 Natural gas road vehicles to replace diesel (100% of market) -6 0 -3 0 0 0 0 0 0 0 Electric road vehicles (5% of market) -73 -7 -64 -17 -4 0 0 0 0 0 Electric road vehicles (10% of market) -147 -15 -127 -35 -7 0 0 0 0 0 Electric road vehicles (100% of market) -1,469 -146 -1,262 -347 -72 -2 0 -2 -2 -2 E10 in 100% gasoline-fueled road vehicles -127 -57 -115 -135 -7 0 0 0 -1 0 B20 in diesel-fueled road vehicles (12.8% of market) -9 0 -15 0 -1 0 0 0 0 0 B20 in diesel-fueled road vehicles (100% of market) -70 0 -117 -1 -6 0 0 0 0 0 B100 in diesel-fueled road vehicles (2.6% of market) -7 0 -12 0 -1 0 0 0 0 0 B100 in diesel-fueled road vehicles (6% of market) -16 0 -28 0 -2 0 0 0 0 0 B100 in diesel-fueled road vehicles (100% of market) -274 -2 -461 -4 -25 0 0 -1 0 -1 Note: The implied increase in emissions for the “100% CNG GSE replacing gasoline GSE, where model available” scenario is a theoretical modeling output related to the emission factor source data used, and is not likely to be observed in actual practice.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-19 Table 43 – Maximum 24-hour PM2.5 Dispersion Modeling Results for Base Case and Alternative Fuel Scenarios (µg/m 3) Condition ATL LAS PHL SAN MHT Base case 6.34 4.63 5.14 3.99 4.15 100% FT (natural gas) aircraft and APU 4.92 4.14 4.52 2.52 3.56 100% FT (coal) aircraft and APU 4.96 4.16 4.55 2.55 3.56 100% Gate power and pre-conditioned air (PCA) to replace APU use 5.76 4.55 4.77 2.41 3.95 100% Electric GSE, where model available 3.96 3.87 2.36 3.50 2.01 100% LPG GSE replacing diesel GSE, where model available 5.25 4.40 2.66 3.63 2.26 100% CNG GSE replacing gasoline GSE, where model available 6.27 4.61 5.10 4.34 4.09 100% CNG GSE replacing diesel GSE, where model available 5.25 4.40 2.66 3.63 2.26 E10 in 100% gasoline-fueled GSE 6.03 4.63 5.08 3.99 4.07 B20 in 100% diesel-fueled GSE 6.14 4.63 4.74 3.96 3.84 B100 in 100% diesel-fueled GSE 5.56 4.63 3.56 3.86 2.94 B100 in 50% diesel-fueled GSE 5.95 4.63 4.35 3.93 3.54 Natural gas road vehicles to replace diesel (9% of market) 6.34 4.63 5.14 3.99 4.15 Natural gas road vehicles to replace diesel (32% of market 6.34 4.63 5.14 3.99 4.15 Natural gas road vehicles to replace diesel (100% of market) 6.34 4.63 5.14 3.99 4.15 Electric road vehicles (5% of market) 6.33 4.63 5.13 3.99 4.15 Electric road vehicles (10% of market) 6.32 4.63 5.13 3.99 4.15 Electric road vehicles (100% of market) 6.16 4.60 5.07 3.97 4.15 E10 in 100% gasoline-fueled road vehicles 6.28 4.62 5.13 3.98 4.15 B20 in diesel-fueled road vehicles (12.8% of market) 6.34 4.63 5.14 3.99 4.15 B20 in diesel-fueled road vehicles (100% of market) 6.34 4.63 5.13 3.99 4.15 B100 in diesel-fueled road vehicles (2.6% of market) 6.34 4.63 5.14 3.99 4.15 B100 in diesel-fueled road vehicles (6% of market) 6.34 4.63 5.14 3.99 4.15 B100 in diesel-fueled road vehicles (100% of market) 6.33 4.63 5.11 3.99 4.15 Note 1: Concentration represents airport contribution (does not include background) at the maximum receptor. Note 2: The implied increase in emissions for the “100% CNG GSE replacing gasoline GSE, where model available” scenario is a theoretical modeling output related to the emission factor source data used, and is not likely to be observed in actual practice.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-20 Table 44 – Alternative Fuel Scenario versus Base Case Maximum 24-hour PM2.5 Dispersion Modeling Results Condition Change in Maximum 24-hour (µg/m3) from Base Case Percent Change from Base Case ATL LAS PHL SAN MHT ATL LAS PHL SAN MHT 100% FT (natural gas) aircraft and APU -1.43 -0.49 -0.62 -1.47 -0.60 -22 -11 -12 -37 -14 100% FT (coal) aircraft and APU -1.38 -0.47 -0.59 -1.45 -0.60 -22 -10 -11 -36 -14 100% Gate power and PCA to replace APU use -0.58 -0.08 -0.36 -1.58 -0.20 -9 -2 -7 -40 -5 100% Electric GSE, where model available -2.38 -0.76 -2.78 -0.49 -2.14 -37 -16 -54 -12 -51 100% LPG GSE replacing diesel GSE, where model available -1.09 -0.23 -2.48 -0.37 -1.90 -17 -5 -48 -9 -46 100% CNG GSE replacing gasoline GSE, where model available -0.07 -0.02 -0.04 0.35 -0.06 -1 0 -1 9 -1 100% CNG GSE replacing diesel GSE, where model available -1.09 -0.23 -2.48 -0.37 -1.90 -17 -5 -48 -9 -46 E10 in 100% gasoline-fueled GSE -0.31 0.00 -0.05 0.00 -0.08 -5 0 -1 0 -2 B20 in 100% diesel-fueled GSE -0.20 0.00 -0.40 -0.03 -0.31 -3 0 -8 -1 -7 B100 in 100% diesel-fueled GSE -0.78 0.00 -1.58 -0.13 -1.22 -12 0 -31 -3 -29 B100 in 50% diesel-fueled GSE -0.39 0.00 -0.79 -0.07 -0.61 -6 0 -15 -2 -15 Natural gas road vehicles to replace diesel (9% of market) 0.00 0.00 0.00 0.00 0.00 0 0 0 0 0 Natural gas road vehicles to replace diesel (32% of market 0.00 0.00 0.00 0.00 0.00 0 0 0 0 0 Natural gas road vehicles to replace diesel (100% of market) 0.00 0.00 0.00 0.00 0.00 0 0 0 0 0 Electric road vehicles (5% of market) -0.01 0.00 0.00 0.00 0.00 0 0 0 0 0 Electric road vehicles (10% of market) -0.02 0.00 -0.01 0.00 0.00 0 0 0 0 0 Electric road vehicles (100% of market) -0.18 -0.03 -0.07 -0.02 -0.01 -3 -1 -1 -1 0 E10 in 100% gasoline-fueled road vehicles -0.06 -0.01 -0.01 -0.01 0.00 -1 0 0 0 0 B20 in diesel-fueled road vehicles (12.8% of market) 0.00 0.00 0.00 0.00 0.00 0 0 0 0 0 B20 in diesel-fueled road vehicles (100% of market) 0.00 0.00 -0.01 0.00 0.00 0 0 0 0 0 B100 in diesel-fueled road vehicles (2.6% of market) 0.00 0.00 0.00 0.00 0.00 0 0 0 0 0 B100 in diesel-fueled road vehicles (6% of market) 0.00 0.00 0.00 0.00 0.00 0 0 0 0 0 B100 in diesel-fueled road vehicles (100% of market) -0.01 0.00 -0.03 0.00 0.00 0 0 0 0 0 Note: The implied increase in emissions for the “100% CNG GSE replacing gasoline GSE, where model available” scenario is a theoretical modeling output related to the emission factor source data used, and is not likely to be observed in actual practice.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-21 Table 45 – Annual PM2.5 Dispersion Modeling Results for Base Case and Alternative Fuel Scenarios (µg/m 3) Condition ATL LAS PHL SAN MHT Base case 1.70 2.77 1.26 2.47 2.40 100% FT (natural gas) aircraft and APU 1.34 2.53 1.09 2.02 2.06 100% FT (coal) aircraft and APU 1.35 2.54 1.10 2.03 2.06 100% Gate power and pre-conditioned air (PCA) to replace APU use 1.52 2.72 1.20 2.07 2.34 100% Electric GSE, where model available 1.19 2.46 0.78 2.33 1.17 100% LPG GSE replacing diesel GSE, where model available 1.46 2.65 0.83 2.36 1.29 100% CNG GSE replacing gasoline GSE, where model available 1.68 2.76 1.25 2.61 2.37 100% CNG GSE replacing diesel GSE, where model available 1.46 2.65 0.83 2.36 1.29 E10 in 100% gasoline-fueled GSE 1.63 2.77 1.25 2.47 2.36 B20 in 100% diesel-fueled GSE 1.65 2.77 1.19 2.46 2.22 B100 in 100% diesel-fueled GSE 1.53 2.77 0.99 2.42 1.70 B100 in 50% diesel-fueled GSE 1.61 2.77 1.12 2.45 2.05 Natural gas road vehicles to replace diesel (9% of market) 1.70 2.77 1.26 2.47 2.40 Natural gas road vehicles to replace diesel (32% of market 1.70 2.77 1.26 2.47 2.40 Natural gas road vehicles to replace diesel (100% of market) 1.70 2.77 1.26 2.47 2.40 Electric road vehicles (5% of market) 1.69 2.77 1.26 2.45 2.40 Electric road vehicles (10% of market) 1.68 2.76 1.26 2.44 2.40 Electric road vehicles (100% of market) 1.50 2.72 1.24 2.38 2.38 E10 in 100% gasoline-fueled road vehicles 1.68 2.75 1.26 2.42 2.40 B20 in diesel-fueled road vehicles (12.8% of market) 1.70 2.77 1.26 2.47 2.40 B20 in diesel-fueled road vehicles (100% of market) 1.69 2.77 1.26 2.47 2.40 B100 in diesel-fueled road vehicles (2.6% of market) 1.70 2.77 1.26 2.47 2.40 B100 in diesel-fueled road vehicles (6% of market) 1.70 2.77 1.26 2.47 2.40 B100 in diesel-fueled road vehicles (100% of market) 1.68 2.77 1.25 2.47 2.39 Note 1: Concentration represents airport contribution (does not include background) at the maximum receptor. Note 2: The implied increase in emissions for the “100% CNG GSE replacing gasoline GSE, where model available” scenario is a theoretical modeling output related to the emission factor source data used, and is not likely to be observed in actual practice.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-22 Table 46 – Alternative Fuel Scenario versus Base Case Annual PM2.5 Dispersion Modeling Results Condition Change in Maximum Annual (µg/m3) from Base Case Percent Change from Base Case ATL LAS PHL SAN MHT ATL LAS PHL SAN MHT 100% FT (natural gas) aircraft and APU -0.36 -0.24 -0.17 -0.46 -0.34 -21 -9 -13 -18 -14 100% FT (coal) aircraft and APU -0.35 -0.23 -0.16 -0.44 -0.34 -20 -8 -12 -18 -14 100% Gate power and PCA to replace APU use -0.17 -0.05 -0.06 -0.41 -0.06 -10 -2 -5 -16 -3 100% Electric GSE, where model available -0.51 -0.30 -0.48 -0.15 -1.23 -30 -11 -38 -6 -51 100% LPG GSE replacing diesel GSE, where model available -0.24 -0.12 -0.43 -0.12 -1.11 -14 -4 -34 -5 -46 100% CNG GSE replacing gasoline GSE, where model available -0.02 -0.01 -0.01 0.14 -0.03 -1 0 0 6 -1 100% CNG GSE replacing diesel GSE, where model available -0.24 -0.12 -0.43 -0.12 -1.11 -14 -4 -34 -5 -46 E10 in 100% gasoline-fueled GSE -0.07 0.00 -0.01 0.00 -0.04 -4 0 -1 0 -2 B20 in 100% diesel-fueled GSE -0.04 0.00 -0.07 -0.01 -0.18 -3 0 -5 -1 -7 B100 in 100% diesel-fueled GSE -0.17 0.00 -0.27 -0.05 -0.70 -10 0 -22 -2 -29 B100 in 50% diesel-fueled GSE -0.08 0.00 -0.14 -0.03 -0.35 -5 0 -11 -1 -15 Natural gas road vehicles to replace diesel (9% of market) 0.00 0.00 0.00 0.00 0.00 0 0 0 0 0 Natural gas road vehicles to replace diesel (32% of market 0.00 0.00 0.00 0.00 0.00 0 0 0 0 0 Natural gas road vehicles to replace diesel (100% of market) 0.00 0.00 0.00 0.00 0.00 0 0 0 0 0 Electric road vehicles (5% of market) -0.01 0.00 0.00 -0.02 0.00 -1 0 0 -1 0 Electric road vehicles (10% of market) -0.02 0.00 0.00 -0.03 0.00 -1 0 0 -1 0 Electric road vehicles (100% of market) -0.19 -0.05 -0.02 -0.09 -0.02 -11 -2 -2 -4 -1 E10 in 100% gasoline-fueled road vehicles -0.02 -0.02 0.00 -0.05 0.00 -1 -1 0 -2 0 B20 in diesel-fueled road vehicles (12.8% of market) 0.00 0.00 0.00 0.00 0.00 0 0 0 0 0 B20 in diesel-fueled road vehicles (100% of market) 0.00 0.00 0.00 0.00 0.00 0 0 0 0 0 B100 in diesel-fueled road vehicles (2.6% of market) 0.00 0.00 0.00 0.00 0.00 0 0 0 0 0 B100 in diesel-fueled road vehicles (6% of market) 0.00 0.00 0.00 0.00 0.00 0 0 0 0 0 B100 in diesel-fueled road vehicles (100% of market) -0.01 0.00 -0.01 -0.01 -0.01 -1 0 -1 0 0 Note: The implied increase in emissions for the “100% CNG GSE replacing gasoline GSE, where model available” scenario is a theoretical modeling output related to the emission factor source data used, and is not likely to be observed in actual practice.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-23 Table 47 – Maximum 24-hour ROI PM2.5 Dispersion Modeling Results for Base Case and Alternative Fuel Scenarios (m) Condition ATL LAS PHL SAN MHT Base case 5,674 4,614 5,487 2,920 1,953 100% FT (natural gas) aircraft and APU 4,730 4,339 4,861 2,456 1,907 100% FT (coal) aircraft and APU 4,765 4,369 4,875 2,483 1,908 100% Gate power and pre-conditioned air (PCA) to replace APU use 5,437 4,591 5,048 2,680 1,923 100% Electric GSE, where model available 4,597 4,532 4,575 2,871 1,755 100% LPG GSE replacing diesel GSE, where model available 5,232 4,565 4,629 2,885 1,811 100% CNG GSE replacing gasoline GSE, where model available 5,650 4,613 5,457 2,945 1,949 100% CNG GSE replacing diesel GSE, where model available 5,232 4,565 4,629 2,885 1,811 E10 in 100% gasoline-fueled GSE 5,584 4,614 5,443 2,920 1,948 B20 in 100% diesel-fueled GSE 5,604 4,614 5,144 2,918 1,937 B100 in 100% diesel-fueled GSE 5,374 4,614 4,801 2,910 1,874 B100 in 50% diesel-fueled GSE 5,531 4,614 4,937 2,916 1,920 Natural gas road vehicles to replace diesel (9% of market) 5,674 4,614 5,487 2,920 1,953 Natural gas road vehicles to replace diesel (32% of market 5,674 4,614 5,487 2,920 1,953 Natural gas road vehicles to replace diesel (100% of market) 5,674 4,614 5,487 2,920 1,953 Electric road vehicles (5% of market) 5,673 4,614 5,484 2,919 1,953 Electric road vehicles (10% of market) 5,672 4,614 5,480 2,918 1,953 Electric road vehicles (100% of market) 5,649 4,614 5,423 2,899 1,952 E10 in 100% gasoline-fueled road vehicles 5,667 4,614 5,481 2,912 1,953 B20 in diesel-fueled road vehicles (12.8% of market) 5,674 4,614 5,486 2,920 1,953 B20 in diesel-fueled road vehicles (100% of market) 5,674 4,614 5,481 2,920 1,953 B100 in diesel-fueled road vehicles (2.6% of market) 5,674 4,614 5,486 2,920 1,953 B100 in diesel-fueled road vehicles (6% of market) 5,674 4,614 5,486 2,920 1,953 B100 in diesel-fueled road vehicles (100% of market) 5,671 4,614 5,462 2,920 1,953 Note: The implied increase in emissions for the “100% CNG GSE replacing gasoline GSE, where model available” scenario is a theoretical modeling output related to the emission factor source data used, and is not likely to be observed in actual practice.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-24 Table 48 – Alternative Fuel Scenario versus Base Case Maximum 24-hour ROI PM2.5 Dispersion Modeling Results Condition Change in 24-hour Radius of Influence (m) from Base Case Percent Change from Base Case ATL LAS PHL SAN MHT ATL LAS PHL SAN MHT 100% FT (natural gas) aircraft and APU -944 -275 -626 -464 -46 -17 -6 -11 -16 -2 100% FT (coal) aircraft and APU -909 -245 -612 -437 -45 -16 -5 -11 -15 -2 100% Gate power and PCA to replace APU use -238 -23 -439 -241 -30 -4 -1 -8 -8 -2 100% Electric GSE, where model available -1,077 -82 -912 -50 -198 -19 -2 -17 -2 -10 100% LPG GSE replacing diesel GSE, where model available -442 -49 -858 -36 -142 -8 -1 -16 -1 -7 100% CNG GSE replacing gasoline GSE, where model available -25 -1 -30 24 -4 0 0 -1 1 0 100% CNG GSE replacing diesel GSE, where model available -442 -49 -858 -36 -142 -8 -1 -16 -1 -7 E10 in 100% gasoline-fueled GSE -90 0 -44 0 -5 -2 0 -1 0 0 B20 in 100% diesel-fueled GSE -71 0 -343 -3 -16 -1 0 -6 0 -1 B100 in 100% diesel-fueled GSE -300 0 -686 -10 -79 -5 0 -12 0 -4 B100 in 50% diesel-fueled GSE -144 0 -550 -5 -33 -3 0 -10 0 -2 Natural gas road vehicles to replace diesel (9% of market) 0 0 0 0 0 0 0 0 0 0 Natural gas road vehicles to replace diesel (32% of market 0 0 0 0 0 0 0 0 0 0 Natural gas road vehicles to replace diesel (100% of market) 0 0 0 0 0 0 0 0 0 0 Electric road vehicles (5% of market) -1 0 -4 -1 0 0 0 0 0 0 Electric road vehicles (10% of market) -3 0 -7 -2 0 0 0 0 0 0 Electric road vehicles (100% of market) -25 0 -64 -21 -1 0 0 -1 -1 0 E10 in 100% gasoline-fueled road vehicles -7 0 -6 -8 0 0 0 0 0 0 B20 in diesel-fueled road vehicles (12.8% of market) 0 0 -1 0 0 0 0 0 0 0 B20 in diesel-fueled road vehicles (100% of market) -1 0 -6 0 0 0 0 0 0 0 B100 in diesel-fueled road vehicles (2.6% of market) 0 0 -1 0 0 0 0 0 0 0 B100 in diesel-fueled road vehicles (6% of market) 0 0 -2 0 0 0 0 0 0 0 B100 in diesel-fueled road vehicles (100% of market) -3 0 -25 0 0 0 0 0 0 0 Note: The implied increase in emissions for the “100% CNG GSE replacing gasoline GSE, where model available” scenario is a theoretical modeling output related to the emission factor source data used, and is not likely to be observed in actual practice.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-25 Table 49 – Annual ROI PM2.5 Dispersion Modeling Results for Base Case and Alternative Fuel Scenarios (m) Condition ATL LAS PHL SAN MHT Base case 4,968 4,952 4,012 4,018 1,988 100% FT (natural gas) aircraft and APU 3,893 4,836 3,926 3,235 1,963 100% FT (coal) aircraft and APU 3,913 4,853 3,933 3,319 1,964 100% Gate power and pre-conditioned air (PCA) to replace APU use 4,187 4,938 3,979 3,668 1,971 100% Electric GSE, where model available 3,993 4,898 3,808 3,907 1,894 100% LPG GSE replacing diesel GSE, where model available 4,544 4,926 3,829 3,934 1,921 100% CNG GSE replacing gasoline GSE, where model available 4,945 4,951 4,001 4,127 1,985 100% CNG GSE replacing diesel GSE, where model available 4,544 4,926 3,829 3,934 1,921 E10 in 100% gasoline-fueled GSE 4,876 4,952 3,999 4,018 1,985 B20 in 100% diesel-fueled GSE 4,903 4,952 3,976 4,007 1,980 B100 in 100% diesel-fueled GSE 4,694 4,952 3,894 3,983 1,951 B100 in 50% diesel-fueled GSE 4,838 4,952 3,949 3,997 1,972 Natural gas road vehicles to replace diesel (9% of market) 4,968 4,952 4,012 4,018 1,988 Natural gas road vehicles to replace diesel (32% of market 4,968 4,952 4,012 4,018 1,988 Natural gas road vehicles to replace diesel (100% of market) 4,967 4,952 4,012 4,018 1,988 Electric road vehicles (5% of market) 4,963 4,952 4,008 4,012 1,988 Electric road vehicles (10% of market) 4,958 4,952 4,004 4,006 1,988 Electric road vehicles (100% of market) 4,871 4,950 3,985 3,937 1,987 E10 in 100% gasoline-fueled road vehicles 4,960 4,951 4,005 3,982 1,988 B20 in diesel-fueled road vehicles (12.8% of market) 4,967 4,952 4,011 4,018 1,988 B20 in diesel-fueled road vehicles (100% of market) 4,963 4,952 4,004 4,018 1,988 B100 in diesel-fueled road vehicles (2.6% of market) 4,967 4,952 4,011 4,018 1,988 B100 in diesel-fueled road vehicles (6% of market) 4,967 4,952 4,010 4,018 1,988 B100 in diesel-fueled road vehicles (100% of market) 4,950 4,952 3,996 4,017 1,988 Note: The implied increase in emissions for the “100% CNG GSE replacing gasoline GSE, where model available” scenario is a theoretical modeling output related to the emission factor source data used, and is not likely to be observed in actual practice.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-26 Table 50 – Alternative Fuel Scenario versus Base Case Annual ROI PM2.5 Dispersion Modeling Results Condition Change in Annual Radius of Influence (m) from Base Case Percent Change from Base Case ATL LAS PHL SAN MHT ATL LAS PHL SAN MHT 100% FT (natural gas) aircraft and APU -1,075 -116 -86 -783 -25 -22 -2 -2 -19 -1 100% FT (coal) aircraft and APU -1,055 -99 -80 -699 -24 -21 -2 -2 -17 -1 100% Gate power and PCA to replace APU use -781 -14 -33 -350 -17 -16 0 -1 -9 -1 100% Electric GSE, where model available -975 -54 -204 -111 -94 -20 -1 -5 -3 -5 100% LPG GSE replacing diesel GSE, where model available -424 -26 -183 -84 -67 -9 -1 -5 -2 -3 100% CNG GSE replacing gasoline GSE, where model available -23 -1 -11 108 -3 0 0 0 3 0 100% CNG GSE replacing diesel GSE, where model available -424 -26 -183 -84 -67 -9 -1 -5 -2 -3 E10 in 100% gasoline-fueled GSE -92 0 -13 -1 -3 -2 0 0 0 0 B20 in 100% diesel-fueled GSE -64 0 -36 -12 -7 -1 0 -1 0 0 B100 in 100% diesel-fueled GSE -274 0 -118 -35 -37 -6 0 -3 -1 -2 B100 in 50% diesel-fueled GSE -130 0 -63 -21 -16 -3 0 -2 -1 -1 Natural gas road vehicles to replace diesel (9% of market) 0 0 0 0 0 0 0 0 0 0 Natural gas road vehicles to replace diesel (32% of market 0 0 0 0 0 0 0 0 0 0 Natural gas road vehicles to replace diesel (100% of market) 0 0 0 0 0 0 0 0 0 0 Electric road vehicles (5% of market) -5 0 -4 -6 0 0 0 0 0 0 Electric road vehicles (10% of market) -9 0 -8 -12 0 0 0 0 0 0 Electric road vehicles (100% of market) -97 -2 -27 -82 -1 -2 0 -1 -2 0 E10 in 100% gasoline-fueled road vehicles -8 -1 -8 -36 0 0 0 0 -1 0 B20 in diesel-fueled road vehicles (12.8% of market) -1 0 -1 0 0 0 0 0 0 0 B20 in diesel-fueled road vehicles (100% of market) -5 0 -8 0 0 0 0 0 0 0 B100 in diesel-fueled road vehicles (2.6% of market) 0 0 -1 0 0 0 0 0 0 0 B100 in diesel-fueled road vehicles (6% of market) -1 0 -2 0 0 0 0 0 0 0 B100 in diesel-fueled road vehicles (100% of market) -18 0 -16 -1 0 0 0 0 0 0 Note: The implied increase in emissions for the “100% CNG GSE replacing gasoline GSE, where model available” scenario is a theoretical modeling output related to the emission factor source data used, and is not likely to be observed in actual practice.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-27 Table 51 – Maximum 24-hour Influence Area PM2.5 Dispersion Modeling Results for Base Case and Alternative Fuel Scenarios (acres) Condition ATL LAS PHL SAN MHT Base case 14,322 3,647 11,515 1,390 364 100% FT (natural gas) aircraft and APU 8,431 2,773 9,407 880 278 100% FT (coal) aircraft and APU 8,686 2,854 9,554 903 278 100% Gate power and pre-conditioned air (PCA) to replace APU use 11,331 3,474 10,852 946 306 100% Electric GSE, where model available 9,038 2,912 6,645 1,251 113 100% LPG GSE replacing diesel GSE, where model available 11,722 3,307 7,119 1,287 149 100% CNG GSE replacing gasoline GSE, where model available 14,158 3,637 11,453 1,492 348 100% CNG GSE replacing diesel GSE, where model available 11,722 3,307 7,119 1,287 149 E10 in 100% gasoline-fueled GSE 13,650 3,647 11,425 1,389 348 B20 in 100% diesel-fueled GSE 13,874 3,647 10,793 1,381 324 B100 in 100% diesel-fueled GSE 12,519 3,647 8,667 1,354 221 B100 in 50% diesel-fueled GSE 13,429 3,647 10,076 1,372 289 Natural gas road vehicles to replace diesel (9% of market) 14,322 3,647 11,515 1,390 364 Natural gas road vehicles to replace diesel (32% of market 14,322 3,647 11,515 1,390 364 Natural gas road vehicles to replace diesel (100% of market) 14,322 3,646 11,515 1,390 364 Electric road vehicles (5% of market) 14,311 3,646 11,496 1,385 363 Electric road vehicles (10% of market) 14,299 3,646 11,476 1,380 363 Electric road vehicles (100% of market) 14,088 3,637 11,109 1,283 361 E10 in 100% gasoline-fueled road vehicles 14,259 3,643 11,480 1,350 363 B20 in diesel-fueled road vehicles (12.8% of market) 14,321 3,647 11,511 1,390 364 B20 in diesel-fueled road vehicles (100% of market) 14,313 3,647 11,479 1,390 363 B100 in diesel-fueled road vehicles (2.6% of market) 14,321 3,647 11,512 1,390 364 B100 in diesel-fueled road vehicles (6% of market) 14,320 3,647 11,507 1,390 364 B100 in diesel-fueled road vehicles (100% of market) 14,285 3,647 11,371 1,389 363 Note: The implied increase in emissions for the “100% CNG GSE replacing gasoline GSE, where model available” scenario is a theoretical modeling output related to the emission factor source data used, and is not likely to be observed in actual practice.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-28 Table 52 – Alternative Fuel Scenario versus Base Case Maximum 24-hour Influence Area PM2.5 Dispersion Modeling Results Condition Change in 24-hour Influence Area (acres) from Base Case Percent Change from Base Case ATL LAS PHL SAN MHT ATL LAS PHL SAN MHT 100% FT (natural gas) aircraft and APU -5,891 -873 -2,109 -510 -86 -41 -24 -18 -37 -24 100% FT (coal) aircraft and APU -5,635 -793 -1,961 -487 -85 -39 -22 -17 -35 -23 100% Gate power and PCA to replace APU use -2,991 -172 -664 -444 -58 -21 -5 -6 -32 -16 100% Electric GSE, where model available -5,283 -735 -4,871 -139 -251 -37 -20 -42 -10 -69 100% LPG GSE replacing diesel GSE, where model available -2,600 -340 -4,396 -103 -215 -18 -9 -38 -7 -59 100% CNG GSE replacing gasoline GSE, where model available -164 -9 -62 102 -16 -1 0 -1 7 -4 100% CNG GSE replacing diesel GSE, where model available -2,600 -340 -4,396 -103 -215 -18 -9 -38 -7 -59 E10 in 100% gasoline-fueled GSE -672 0 -91 0 -16 -5 0 -1 0 -4 B20 in 100% diesel-fueled GSE -448 0 -722 -9 -39 -3 0 -6 -1 -11 B100 in 100% diesel-fueled GSE -1,802 0 -2,848 -36 -142 -13 0 -25 -3 -39 B100 in 50% diesel-fueled GSE -893 0 -1,439 -18 -74 -6 0 -12 -1 -20 Natural gas road vehicles to replace diesel (9% of market) 0 0 0 0 0 0 0 0 0 0 Natural gas road vehicles to replace diesel (32% of market 0 0 0 0 0 0 0 0 0 0 Natural gas road vehicles to replace diesel (100% of market) 0 -1 -1 0 0 0 0 0 0 0 Electric road vehicles (5% of market) -11 0 -19 -5 0 0 0 0 0 0 Electric road vehicles (10% of market) -23 -1 -39 -10 0 0 0 0 -1 0 Electric road vehicles (100% of market) -234 -10 -406 -107 -3 -2 0 -4 -8 -1 E10 in 100% gasoline-fueled road vehicles -63 -4 -35 -40 0 0 0 0 -3 0 B20 in diesel-fueled road vehicles (12.8% of market) -1 0 -5 0 0 0 0 0 0 0 B20 in diesel-fueled road vehicles (100% of market) -9 0 -36 0 0 0 0 0 0 0 B100 in diesel-fueled road vehicles (2.6% of market) -1 0 -4 0 0 0 0 0 0 0 B100 in diesel-fueled road vehicles (6% of market) -2 0 -9 0 0 0 0 0 0 0 B100 in diesel-fueled road vehicles (100% of market) -37 0 -144 -1 -1 0 0 -1 0 0 Note: The implied increase in emissions for the “100% CNG GSE replacing gasoline GSE, where model available” scenario is a theoretical modeling output related to the emission factor source data used, and is not likely to be observed in actual practice.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-29 Table 53 – Annual Influence Area PM2.5 Dispersion Modeling Results for Base Case and Alternative Fuel Scenarios (acres) Condition ATL LAS PHL SAN MHT Base case 9,129 5,079 6,445 2,511 454 100% FT (natural gas) aircraft and APU 5,977 4,271 5,668 1,742 351 100% FT (coal) aircraft and APU 6,126 4,345 5,733 1,784 355 100% Gate power and pre-conditioned air (PCA) to replace APU use 7,366 4,917 6,196 1,945 407 100% Electric GSE, where model available 6,729 4,618 4,835 2,310 200 100% LPG GSE replacing diesel GSE, where model available 7,912 4,859 4,993 2,361 234 100% CNG GSE replacing gasoline GSE, where model available 9,045 5,072 6,424 2,626 428 100% CNG GSE replacing diesel GSE, where model available 7,912 4,859 4,993 2,361 234 E10 in 100% gasoline-fueled GSE 8,817 5,079 6,415 2,511 436 B20 in 100% diesel-fueled GSE 8,915 5,079 6,212 2,498 419 B100 in 100% diesel-fueled GSE 8,292 5,079 5,535 2,458 313 B100 in 50% diesel-fueled GSE 8,709 5,079 5,988 2,485 385 Natural gas road vehicles to replace diesel (9% of market) 9,128 5,079 6,445 2,511 454 Natural gas road vehicles to replace diesel (32% of market 9,128 5,079 6,445 2,511 454 Natural gas road vehicles to replace diesel (100% of market) 9,126 5,079 6,445 2,511 454 Electric road vehicles (5% of market) 9,102 5,078 6,437 2,503 454 Electric road vehicles (10% of market) 9,076 5,077 6,429 2,494 454 Electric road vehicles (100% of market) 8,600 5,056 6,285 2,334 450 E10 in 100% gasoline-fueled road vehicles 9,004 5,070 6,431 2,443 454 B20 in diesel-fueled road vehicles (12.8% of market) 9,125 5,079 6,443 2,511 454 B20 in diesel-fueled road vehicles (100% of market) 9,103 5,079 6,430 2,511 454 B100 in diesel-fueled road vehicles (2.6% of market) 9,126 5,079 6,444 2,511 454 B100 in diesel-fueled road vehicles (6% of market) 9,122 5,079 6,442 2,511 454 B100 in diesel-fueled road vehicles (100% of market) 9,027 5,079 6,387 2,509 453 Note: The implied increase in emissions for the “100% CNG GSE replacing gasoline GSE, where model available” scenario is a theoretical modeling output related to the emission factor source data used, and is not likely to be observed in actual practice.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-30 Table 54 – Alternative Fuel Scenario versus Base Case Annual Influence Area PM2.5 Dispersion Modeling Results Condition Change in Annual Influence Area (acres) from Base Case Percent Change from Base Case ATL LAS PHL SAN MHT ATL LAS PHL SAN MHT 100% FT (natural gas) aircraft and APU -3,151 -808 -777 -769 -104 -35 -16 -12 -31 -23 100% FT (coal) aircraft and APU -3,003 -734 -713 -728 -99 -33 -14 -11 -29 -22 100% Gate power and PCA to replace APU use -1,762 -163 -249 -566 -47 -19 -3 -4 -23 -10 100% Electric GSE, where model available -2,400 -462 -1,610 -202 -254 -26 -9 -25 -8 -56 100% LPG GSE replacing diesel GSE, where model available -1,216 -220 -1,453 -151 -220 -13 -4 -23 -6 -48 100% CNG GSE replacing gasoline GSE, where model available -84 -7 -21 114 -27 -1 0 0 5 -6 100% CNG GSE replacing diesel GSE, where model available -1,216 -220 -1,453 -151 -220 -13 -4 -23 -6 -48 E10 in 100% gasoline-fueled GSE -312 0 -30 -1 -18 -3 0 0 0 -4 B20 in 100% diesel-fueled GSE -214 0 -233 -13 -36 -2 0 -4 -1 -8 B100 in 100% diesel-fueled GSE -836 0 -910 -54 -141 -9 0 -14 -2 -31 B100 in 50% diesel-fueled GSE -420 0 -458 -26 -70 -5 0 -7 -1 -15 Natural gas road vehicles to replace diesel (9% of market) -1 0 0 0 0 0 0 0 0 0 Natural gas road vehicles to replace diesel (32% of market -1 0 0 0 0 0 0 0 0 0 Natural gas road vehicles to replace diesel (100% of market) -3 0 0 0 0 0 0 0 0 0 Electric road vehicles (5% of market) -27 -1 -8 -8 0 0 0 0 0 0 Electric road vehicles (10% of market) -53 -2 -16 -17 0 -1 0 0 -1 0 Electric road vehicles (100% of market) -529 -24 -160 -177 -5 -6 0 -2 -7 -1 E10 in 100% gasoline-fueled road vehicles -125 -9 -15 -68 0 -1 0 0 -3 0 B20 in diesel-fueled road vehicles (12.8% of market) -4 0 -2 0 0 0 0 0 0 0 B20 in diesel-fueled road vehicles (100% of market) -26 0 -15 0 0 0 0 0 0 0 B100 in diesel-fueled road vehicles (2.6% of market) -3 0 -2 0 0 0 0 0 0 0 B100 in diesel-fueled road vehicles (6% of market) -7 0 -4 0 0 0 0 0 0 0 B100 in diesel-fueled road vehicles (100% of market) -102 0 -59 -2 -2 -1 0 -1 0 0 Note: The implied increase in emissions for the “100% CNG GSE replacing gasoline GSE, where model available” scenario is a theoretical modeling output related to the emission factor source data used, and is not likely to be observed in actual practice.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports E-31 Table 55 – Annual PM2.5 Emissions Inventory (kg) for Turboprop and Piston-engine Aircraft Condition Scenario ATL LAS PHL SAN MHT Turboprop Base case 3,337 985 4,119 1,266 2,189 FTG1 1,684 499 2,094 642 1,107 Piston-engine Base case 2.3 0.6 289 1.3 2.68 AGUL1 0.06 0.02 7.1 0.03 0.07

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports F-1 APPENDIX F: GUIDANCE DOCUMENT Note: The spreadsheet tool referenced herein is available for download on the publication summary page on the TRB website. Access to this page is provided via the “ACRP Web-Only Document 13 Web Page” link on this document’s bookmark menu. This Guidance Document has been written as a key output from the ACRP 02-23 project Alternative Fuels as a means to reduce PM2.5 Emissions at Airports, Principal Investigators: Dr. Hazel Peace and Damon Fordham of PPC/AEA with contributing authors Jamie Beevor, Dr. Mark Broomfield, Dr. John Norris and Dr. Brian Underwood; Mike Kenney, Mike Ratte and Paul Sanford of KB Environmental Sciences Inc.; Dr. Mary E. Johnson and David L. Stanley of Purdue University; Mary Vigilante of Synergy Consultants Inc.; and Richard Altman. BACKGROUND The U.S. National Ambient Air Quality Standards (NAAQS) for fine particulate matter with a diameter of less than 2.5 micrometers (PM2.5) are set primarily for the protection of public health. The current 24-hour NAAQS for PM2.5 is 35 μg/m3 and the annual standard is 15 μg/m3. Over 50 commercial airports in the U.S. are in areas that are classed as PM2.5 “non-attainment” (i.e., in breach of the NAAQS). Proposed improvement projects at Los Angeles International Airport and Philadelphia International Airport are facing agency review because of the potential impacts to local and regional PM2.5 air quality. Other airports around the country (e.g., Chicago O’Hare International, Seattle-Tacoma International, and George Bush Intercontinental/Houston) have all experienced similar public concerns about the potential health effects associated with the combustion of jet fuel, principally due to emissions-related to particulate matter. It is anticipated that expansion of other airports to address capacity needs will face increased pressure to consider particulate matter impacts and emissions of related local pollutants. One of the ways in which airports can assist in reducing PM2.5 impacts is by increasing the availability and use of alternative fuels. This guidance has been produced as an outcome of the ACRP 02-23 research project and is based on the project’s findings. The ARCP 02-23 project was undertaken from July 2010 to December 2011. The ACRP 02-23 project’s aim was to investigate the impact that alternative fuel use could have on emissions and ambient air pollution concentrations of fine particulate matter (PM2.5) at airports. The results were based on modeling of emissions and ambient air pollution concentrations at five case study airports for those sources that contribute most to PM2.5 emissions. Alternative fuels were selected for analysis primarily based on their potential to reduce PM2.5, and were limited to those with short-term (i.e., fewer than 10 years) commercial availability and available emissions data. This Guidance Document provides airport operators, and others, with an understanding of the relative potential benefits of alternative fuels as a means of reducing the impacts of PM2.5 emissions and aims to support decision-making by providing technical supporting material. While airports are the primary audience for this document, other non-airport stakeholders, particularly airlines, fuel providers, equipment manufacturers, and ground support providers (e.g., airside operations, passenger transportation operators, and construction equipment operators) may benefit from using this guidance.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports F-2 LEVEL OF ANALYSIS – A TIERED APPROACH Different users of this guidance will wish to approach the selection of alternative fuels at different levels of detail. Therefore, this guidance is split into three tiers:  Tier 1 is high level guidance based on a number of key criteria, which is suitable for airport executives, senior managers, and their clients/service providers, and will help establish key messages as part of stakeholder engagement.  Tier 2 presents a spreadsheet tool based on the emission results from the five case study airports analyzed in the ACRP 02-23 project. The tool allows users to combine the impacts of different alternative fuel scenarios at those airports and to alter penetration factors to enable them to understand the different source and alternative fuel impacts. This tool is aimed at airport environmental managers.  Tier 3 refers the reader to the ACRP 02-23 Final Report and is intended for those who wish to undertake a detailed study of their own airport following the methodology used in the ACRP 02-23 project. The ACRP 02-23 Final Report provides more detail on the information presented in each tier of this guidance. TIER 1 The primary purpose of the Tier 1 guidance is to help airports to undertaken a high level assessment of the suitability of various alternative fuels as substitutes for conventional fuels for a particular emission source. Each combination of alternative fuel and emission source was rated in the ACRP 02-23 project in terms of key criteria. The definitions for each of the criterion and their ratings are shown in Table 56. After Table 56 each of the major airport emission sources—jet-fueled aircraft, AvGas- fueled aircraft (i.e., piston-engine aircraft), auxiliary power units (APUs), ground support equipment (GSE), and road vehicles—are briefly discussed, followed in turn by a table highlighting each alternative fuel and the ratings that were assigned for each of the key criteria in the ACRP 02-23 project. The information in these tables will allow readers to assess which alternative fuels may be appropriate for consideration at their airports. The ACRP 02-23 project determined that other airport sources of PM2.5 were generally small in comparison to those sources listed above, at least for the case study airports analyzed. Consequently, other sources are not included in this guidance. However, some airports may have other equipment or machinery that does represent a significant emissions source, such as oil or solid-fueled power and/or heat generation. These sources will need further consideration on a case-by-case basis. Further detail is provided in the ACRP 02-23 Final Report.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports F-3 Table 56 – Alternative Fuels – Criteria and Definitions Criterion Definition Rating Change in PM2.5 emissions (H, M, L) The relative decrease in emissions compared with the dominant existing fuel/engine (or vehicle) H = >75% reduction M = Between 25% and 75% reduction L = <25% reduction Availability of fuel (H, M, L) Is the fuel currently available? H = Widespread availability of fuel/blend in many states, though some regional variability M = Frequently available, but not at all sites/locations and would often require additional infrastructure (e.g., tanks) L = Limited/not readily available Availability of new vehicles (H, M, L) Are vehicles that can use this fuel currently available or are they likely to be available in the short-term? It should be noted that model availability depends on purpose H = Many model types readily available for this fuel type and many being used M = Many model types available that can use this fuel, though not universal L = Not many models available (if any) that can use this fuel Cost to convert existing vehicles (H,M,L) How much is it likely to cost to convert a typical vehicle? H = >$20,000 M = Between $200 and $20,000 L = <$200 N/A = no cost associated (i.e., for drop-in fuels). Drop-in fuel for existing vehicle? (Y/N) Can the fuel be used in existing vehicles with no modification? Y/N or N/A GHG life-cycle emissions (H, M, L) Greenhouse gas (GHG) emissions of the alternative fuel relative to the primary conventional fuel. This figure includes the fuel processing (i.e., “well to wheel”) emissions H = >90% of conventional fuel M = Between 40% and 90% of conventional fuel L = <40% of conventional fuel Emission data source reliability (H, M, L) Is the source of the proposed emission factors based on reliable data? H = Widely tested, many high-quality (government or referred journal) published studies with similar results for a range of vehicles M = Published studies, but limited to one or two vehicles L = No specific data, assumptions based on similar source (e.g., road vehicle for GSE) or based on calculations Cost of fuel compared with conventional (H, E, L) This is the marginal increase in fuel cost compared with the dominant existing fuel H = >125% of conventional fuel E = Equivalent price to conventional fuel – between 75% and 125% of conventional fuel L = <75% of conventional fuel (N/A where no data on cost are available. Variable and N/A assume worst case, high cost) Cost of vehicles compared with conventional (H, M, L) This is the marginal increase in vehicle cost compared with the dominant existing vehicle type H = > 200% M = between 110% and 200% L = <110% N/A = no additional cost (i.e., for drop-in fuels).

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports F-4 Criterion Definition Rating Additional infrastructure needed (H, M, L) What additional infrastructure is needed for the fuel to be used? H = Additional equipment such as compressors, high pressure buffers and tanks needed M = Additional tanks, similar to those already in existence, would be needed (e.g., for different blends) L = Assumes that diesel, electricity and gasoline are readily available on, or near, the site N/A = no additional cost associated. Warranty validity issue (Y/N) Could the use of this fuel result in vehicle/engine warranty being invalidated? Y/N Note that “vehicle” is used here to refer to aircraft, APU, GSE and road vehicles Jet-fueled Aircraft ASTM International has approved an alternative jet fuel specification in annexes to ASTM D7566 for Fischer-Tropsch (FT) and hydroprocessed renewable jet (HRJ) fuels blended with at least 50% conventional jet fuel. This means that, in theory, these fuels can now be produced and sold as a “drop-in” fuel for aircraft (i.e., no modifications are required to the aircraft to use this fuel). However, current availability is limited. Particulate matter emission reduction data for HRJ fuels were not finalized at the time of writing, so they are not included here. However, it is likely that emission reductions will be similar to those for FT fuels. With 50/50 FT blended fuels, total particulate matter emission reductions are in the region of 50% for aircraft engines and APUs. The cost of FT and HRJ fuels are currently high compared to costs of conventional jet fuel. However, as commercial productivity and demand increase, the cost is likely to reduce. Other considerations are shown in Table 57. Table 57 – Jet-fueled Aircraft Alternative Fuels C ri te ri on C ha ng e PM 2. 5 e m is si on s (H , M , L ) A va ila bi lit y of fu el (H , M , L ) A va ila bi lit y of n ew v eh ic le s (H , M , L ) C os t t o co nv er t e xi st in g ve hi cl es (H ,M ,L ) D ro p- in fu el fo r ex ist in g ve hi cl e? (Y /N o r N /A ) G H G li fe -c yc le e m iss io ns (H , M , L ) Em iss io n da ta so ur ce r el ia bi lit y (H , M , L ) C os t o f f ue l c om pa re d w ith co nv en tio na l ( H , E , L ) C os t o f v eh ic le s c om pa re d w ith co nv en tio na l ( H , M , L ) A dd iti on al in fr as tr uc tu re n ee de d (H , M , L ) W ar ra nt y va lid ity is su e (Y /N o r N /A ) Low-sulfur Jet-A for aircraft L L H N/A Y H L E N/A L N FT (natural gas) aircraft M L H N/A Y H M H N/A L N FT (coal) aircraft M L H N/A Y H M H N/A L N FT (biomass) aircraft M L H N/A Y H L H N/A L N HRJ (biomass) aircraft M L H N/A Y H L H N/A L N

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports F-5 AvGas-Fueled Aircraft (Piston-engine Aircraft) While alternative AvGas fuel not is yet commercially available in the U.S., it could be in the future. Compared with 100LL (the AvGas used in the U.S.), 91/96UL produces emission reductions in the region of 90% for piston-engine aircraft. Other considerations are shown in Table 58. Table 58 – Piston-engine Aircraft Alternative Fuels C ri te ri on C ha ng e PM 2. 5 e m is si on s (H , M , L ) A va ila bi lit y of fu el (H , M , L ) A va ila bi lit y of n ew v eh ic le s (H , M , L ) C os t t o co nv er t e xi st in g ve hi cl es (H ,M ,L ) D ro p- in fu el fo r ex ist in g ve hi cl e? (Y /N o r N /A ) G H G li fe -c yc le e m iss io ns (H , M , L ) Em iss io n da ta so ur ce r el ia bi lit y (H , M , L ) C os t o f f ue l c om pa re d w ith co nv en tio na l ( H , E , L ) C os t o f v eh ic le s c om pa re d w ith co nv en tio na l ( H , M , L ) A dd iti on al in fr as tr uc tu re n ee de d (H , M , L ) W ar ra nt y va lid ity is su e (Y /N o r N /A ) 91/96UL AvGas for piston-engine aircraft H L L L Y H M E N/A M Y APU APU emissions can be reduced in two ways. First, by reducing APU use (e.g., by providing and encouraging use of alternatives, such as fixed electric ground power (FEGP) and pre-conditioned air (PCA)). Second, by using alternative jet fuel instead of conventional fuels (refer to Jet-fueled Aircraft, above). FEGP and PCA are likely to require high up-front investment, although grants are available through the Federal Aviation Administration’s (FAA) Voluntary Airport Low Emissions (VALE) program. Other considerations are shown in Table 59.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports F-6 Table 59 – APU Alternative Fuels C ri te ri on C ha ng e PM 2. 5 e m is si on s (H , M , L ) A va ila bi lit y of fu el (H , M , L ) A va ila bi lit y of n ew v eh ic le s (H , M , L ) C os t t o co nv er t e xi st in g ve hi cl es (H ,M ,L ) D ro p- in fu el fo r ex ist in g ve hi cl e? (Y /N o r N /A ) G H G li fe -c yc le e m iss io ns (H , M , L ) Em iss io n da ta so ur ce r el ia bi lit y (H , M , L ) C os t o f f ue l c om pa re d w ith co nv en tio na l ( H , E , L ) C os t o f v eh ic le s c om pa re d w ith co nv en tio na l ( H , M , L ) A dd iti on al in fr as tr uc tu re n ee de d (H , M , L ) W ar ra nt y va lid ity is su e (Y /N o r N /A ) FT (natural gas) APU M L H N/A Y H L H N/A L N FT (coal) APU M L H N/A Y H M H N/A L N FT (biomass) APU M L H N/A Y H L H N/A L N HRJ (biomass) APU M L H N/A Y H L H N/A L Y Low-sulfur Jet-A for APU L L H L Y H L E N/A L N Electricity to replace some APU use M H H N/A Y V H V N/A H N GSE Grants for purchasing alternative-fueled GSE are available through the FAA’s VALE program. In addition, drop-in fuels such as biodiesel (B20) and ethanol (E10) can be used to replace diesel and gasoline, respectively, as a drop-in fuel. However, the greatest PM2.5 emission reductions are gained when diesel GSE are replaced by (in increasing order) gasoline, CNG, LPG or electric GSE, where an appropriate alternatively fueled GSE exists. The costs of alternative fuels for GSE are typically either equivalent or slightly higher when compared to the conventional fuel, with the exception of the high-biofuel blends (e.g., E85 and B100). The high-biofuel blends may also create problems in terms of warranty invalidation for GSE. Other considerations are shown in Table 60. While retrofit technology is not the subject of this report, it could be advantageous to fit equipment (e.g., particulate matter traps) to existing GSE diesel engines given the uncertainties of particulate matter emissions and to be cost-effective. Where vehicle replacement is an option, electric GSE is better when compared with other alternative fuels in terms of reducing directly emitted particulate matter (U.S. FAA, 2010a). Around the world, electric vehicles are available as replacements for baggage tugs and belt loaders. A few other specialist airside electric vehicles have been trialed, and there are a few makes of electric aircraft push-back tugs. However, their relatively modest capacity suggests they would not be very flexible and unable to deal with larger aircraft.

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports F-7 Table 60 – GSE Alternative Fuels C ri te ri on C ha ng e PM 2. 5 e m is si on s (H , M , L ) A va ila bi lit y of fu el (H , M , L ) A va ila bi lit y of n ew v eh ic le s (H , M , L ) C os t t o co nv er t e xi st in g ve hi cl es (H ,M ,L ) D ro p- in fu el fo r ex ist in g ve hi cl e? (Y /N o r N /A ) G H G li fe -c yc le e m iss io ns (H , M , L ) Em iss io n da ta so ur ce r el ia bi lit y (H , M , L ) C os t o f f ue l c om pa re d w ith co nv en tio na l ( H , E , L ) C os t o f v eh ic le s c om pa re d w ith co nv en tio na l ( H , M , L ) A dd iti on al in fr as tr uc tu re n ee de d (H , M , L ) W ar ra nt y va lid ity is su e (Y /N o r N /A ) Electric GSE H H H N/A N V H V H L N LPG GSE replacing gasoline GSE L H L L N H H H M M Y LPG GSE replacing diesel GSE H H L H N H H H M M Y CNG GSE replacing gasoline GSE M H L M N H H L M H Y CNG GSE replacing diesel GSE H H L H N H H L M H Y Low-sulfur diesel GSE L H H N/A Y H L E N/A L N E5 in gasoline-fueled GSE L H M N/A Y H L E N/A N/A N E10 in gasoline-fueled GSE M H M N/A Y H L E N/A N/A N E15 in gasoline-fueled GSE M L L L N H L N/A L L Y E85 in gasoline-fueled GSE L M L L N M L H L L Y B5 in diesel-fueled GSE L H H N/A Y H L E N/A N/A N B10 in diesel-fueled GSE L L M L N H L N/A L M Y B15 in diesel-fueled GSE L L M L N H L N/A L M Y B20 in diesel-fueled GSE L M M L N H L E L L Y B100 in diesel-fueled GSE M M L M N M L H L L Y Road Vehicles Drop-in fuels, such as B20 and E10, can be used to replace diesel and gasoline respectively. However, the greatest PM2.5 emission reductions are gained when diesel vehicles are replaced by (in increasing order) gasoline, CNG, LPG or electric vehicles. The cost of alternative fuels for road vehicles are typically either equivalent or slightly higher than the convention fuel, with the exception of the high-biofuel blends (e.g., E85 and B100) which are more costly. The high- biofuel blends may also create problems in terms of warranty invalidation for road vehicles. Although there is a limit to the number of road vehicles that airports can influence beyond their

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports F-8 own fleet, mechanisms such as structured parking lot charges and taxi licensing can help to encourage alternative fuel use in road vehicles. Other considerations are shown in Table 61. Table 61 – Road Vehicle Alternative Fuels C ri te ri on C ha ng e PM 2. 5 e m is si on s (H , M , L ) A va ila bi lit y of fu el (H , M , L ) A va ila bi lit y of n ew v eh ic le s (H , M , L ) C os t t o co nv er t e xi st in g ve hi cl es (H ,M ,L ) D ro p- in fu el fo r ex ist in g ve hi cl e? (Y /N o r N /A ) G H G li fe -c yc le e m iss io ns (H , M , L ) Em iss io n da ta so ur ce r el ia bi lit y (H , M , L ) C os t o f f ue l c om pa re d w ith co nv en tio na l ( H , E , L ) C os t o f v eh ic le s c om pa re d w ith co nv en tio na l ( H , M , L ) A dd iti on al in fr as tr uc tu re n ee de d (H , M , L ) W ar ra nt y va lid ity is su e (Y /N o r N /A ) Low-sulfur diesel road vehicles N/A N/A N/A N/A Y N/A N/A E N/A N/A N Natural gas road vehicles to replace diesel H H M M N H L L M H N Electric road vehicles H H L N/A N V H V H L N E5 in gasoline-fueled road vehicles L H H N/A Y H L E N/A N/A N E10 in gasoline-fueled road vehicles M H H N/A Y H M E N/A N/A N E15 in gasoline-fueled road vehicles M L H L N H L N/A L L Y E85 in gasoline-fueled road vehicles L M H L N M L H L L Y B5 in diesel-fueled road vehicles L H H N/A Y H H E N/A N/A N B10 in diesel-fueled road vehicles L L H L N H H N/A N/A M Y B15 in diesel-fueled road vehicles L L H L N H H N/A N/A M Y B20 in diesel-fueled road vehicles L M H L N H H E N/A L Y B100 in diesel-fueled road vehicles M M L M N M H H M L Y

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports F-9 TIER 2 Tier 2 is a spreadsheet-based tool “AIRPORT PM2.5 EMISSIONS ALTERNATIVE FUELS IMPACT TOOL.xlsx.” It combines the results from the five case study airports analyzed in the ACRP 02-23 project in a format that allows the user to combine the emission impacts of different alternative fuel scenarios at those airports. The tool is limited to providing a range of results based on the five case study airports only. The results are displayed in a similar format to the results in the ACRP 02-23 Final Report, with the exception of one GSE emissions scenario, for which the lower bound has been set to zero rather than displaying the theoretical increase in emissions that resulted from one case study airport’s modeling output, due to the emission factor source data used for that case study. Instructions for use are below, and are also included in the spreadsheet tool itself. Instructions Open the file “AIRPORT PM2.5 EMISSIONS ALTERNATIVE FUELS IMPACT TOOL.xlsx.” 1. The tool should open with the “Instructions” sheet (Figure 22) for the user’s review. These are reproduced below. Figure 22 – Tool Instruction Sheet

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports F-10 2. Click “Go to INPUT” (Figure 23). 3. There are three columns on the “Input” sheet: a. The first shows the “Source Group” to which the alternative fuel scenario is applied (i.e., an alternative jet fuel is only applicable to jet-fueled aircraft and APU). b. The second shows the “alternative fuel scenario.” c. The third column shows the “Percent of Total Fuel Usage,” which can be changed by the user. The Percent of Total Fuel Usage are originally set to “0” for all alternative fuel scenarios and “100” for all base case scenarios. Note that the Percent of Total Fuel Usage is the equivalent of the penetration factor discussed in the ACRP 02-23 Final Report, multiplied by 100. Figure 23 – Tool Input Sheet 4. In the third column of the “Input” sheet enter, in the yellow cells, the required Percent of Total Fuel Usage (between “0” and “100”) where: a. A value of “0” indicates no use of the alternative fuel listed in the second column. b. A value of “100” indicates that 100% of this source group (listed in the first column) use the alternative fuel listed in the second column. However, in some instances (e.g. GSE electric push-back tugs for large aircraft), an alternatively fueled model does not exist and therefore the 100% refers to 100% where a replacement option is available. c. The total of all the penetration factors (third column) for one source group should equal 100 (otherwise a red error message will appear in the fourth column).

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports F-11 5. Once all the Percent of Total Fuel Usage have been entered correctly and there are no red error messages, click “Go to RESULTS” (Figure 24). 6. The “Results” sheet shows the percentage reduction for each source group and the total airport annual emissions compared with the base case. This is presented as a range of values based on the five case study airports. The results are shown in tabular and graphical formats. 7. Results can be saved by using the “File > Save As” function. Figure 24 – Tool Results Sheet

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports F-12 TIER 3 In some instances, an individual airport may wish to undertake its own, more-detailed study. The types of data that airports would need to collate to generate a base case emissions inventory and undertake dispersion modeling are summarized in Table 62. In addition, Appendix A of the Emissions and Dispersion Modeling System (EDMS) user manual (U.S. FAA, 2009a) provides an overview and screen shots of the data needed and how to enter them into EDMS. Meteorological data would need to be obtained in AERMOD format, and the EDMS manual lists where those data can be obtained. A flow chart summarizing the methodology is shown in Figure 25. To generate the alternative fuel scenarios, analysts should refer to the ACRP 02-23 Final Report, and in particular, to the methodology described in Chapter 5 and Appendix D. For most alternative-fuel scenarios, alternative fuel emission factors can be applied. However, some scenarios for GSE and road vehicles cannot be directly scaled. For those GSE scenarios that cannot be scaled, alternative fuel scenarios can be investigated via EDMS (refer to the user manual (U.S. FAA, 2009 and 2009a)). Stationary source alternative fuel scenarios can also be investigated via the EDMS interface. For road vehicle scenarios that cannot be scaled the alternative fuel scenarios can be investigated using the MOBILE model (or potentially the MOVES model), and, again, the user manuals for the relevant models should be consulted. Table 62 – Typical Data Used in PM2.5 Emission Inventories Source Category Data Aircraft LTO by aircraft type and engine type Taxi-in, taxi-out, delay times (aircraft time in mode) Profiles of quarter hour, daily and monthly activity levels Runway and taxiway assignments and coordinates Terminal/gate assignments and locations Ground support equipment (GSE) Number and type by aircraft type Fuel type Size and load Operating times Auxiliary power units (APU) Percent of gates with fixed power units Percent of gates with fixed pre-conditioned air Road vehicles Location, by segment Vehicle fleet mix by segment Roadway traffic volume by segment Average speed Emission factors (generated using either MOBILE6.2 or EMFAC2007 models) Parking facility Location by parking lot Vehicle fleet mix by parking lot Traffic volume by parking lot Travel distance Idle time Stationary sources Type and location Fuel type and quantity Stack height and diameter Exhaust temperature and velocity

Airport Cooperative Research Program Project ACRP 02-23: Alternative Fuels as a Means to Reduce PM2.5 Emissions at Airports F-13 Figure 25 – Methodology Flow Chart