Quantifying Aircraft Lead Emissions at Airports (2015)

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.

3. REVIEW OF EXISTING METHODS FOR QUANTIFYING AIRCRAFT-RELATED LEAD EMISSIONS In its simplest form, the methodology for quantifying lead emissions from an individual aircraft during a given airport operation (e.g., taxiing) requires only the following information: • The engine fuel flow rate during the operation in units of volume or mass of fuel consumed per unit time; • The lead content of the fuel in units of mass of lead per unit volume or mass of fuel consumed; • The amount of lead that is retained in the engine or exhaust system as a percentage of total lead consumed; and • The amount of time required to conduct the operation. However, the amount of information needed expands dramatically when quantifying the aircraft-related lead emissions at a given airport over any significant length of time. The expanded data needs typically include the following: • Relationships between engine fuel flow and engine load for all of the different types of engines used in aircraft operating at the airport; • Engine loads during each aircraft operating mode (idle/taxi, run-up, takeoff, climb-out, and approach) for all of the different types of aircraft in operation at the airport; • The gasoline lead content and lead retention rate for all of the different types of aircraft in operation at the airport; • The duration of each of the different operating modes for all of the different types of aircraft in operation at the airport; and • The number of each type of aircraft in operation at the airport and the operations in which they are engaged. -9-

Furthermore, to the extent that one seeks to use aircraft-related Pb emissions data in combination with air quality models to estimate ambient Pb concentrations at points in and around the airport, information is also needed regarding spatial and temporal patterns of aircraft activity as well as relevant meteorological data. Because of the scope and complexity of the information needed to quantify aircraft- related lead emissions for a given airport and the resources required to obtain that information, many simplifying assumptions are typically required. Obviously the types of information that are actually used and the nature of the assumptions that are made have the potential to substantially affect the quality of the estimates of Pb emissions occurring at an airport. Given this, the goal of this phase of the study was to critically review existing methods used to quantify aircraft-related Pb emissions in order to assess the information and assumptions upon which they are based and to develop approaches that can be used to improve quantification of airport Pb emissions. 3.1 Literature Search The review began with a search of the technical literature related to the quantification of aircraft-related lead emissions. Over 70 related documents were identified and are summarized in the annotated bibliography provided in Appendix A. As a review of the bibliography shows, many of the identified documents, while related to aircraft lead emissions, were not directly relevant to existing methodologies used for quantification of Pb emissions at airports. The methodologies described in the directly relevant studies are summarized below, along with key sources of existing information. 3.2 Summary of Existing Methodologies and Information Sources for Quantifying Aircraft-Related Lead Emissions 3.2.1 EPA Methodologies AP-42 – The general method by which air emissions are quantified from aircraft had its origination in the Compilation of Air Pollutant Emissions Factors, Third Edition (AP-42) published by the U.S. Environmental Protection Agency (EPA) (U.S. EPA 1977). Therein, times in mode of operation (TIM) for piston-powered general aviation (GA) and military aircraft within the Landing-Takeoff Cycle (LTO)—representing taxi-out, takeoff, climb-out, approach and taxi-in operations—are presented. Additionally, modal fuel flow rates and emissions factors for select criteria pollutants emitted from the Continental O-200 and Lycoming O-320 model piston engines are provided. The fuel flow rates reported for these engines are summarized in Table 3 below. Notably, a methodology for quantifying aircraft Pb emissions is not addressed in this document. -10-

Table 3 Original Fuel Flow Rates from AP-42 Engine Model Fuel Flow Rate per Mode of Operation (lb/hr) Taxi/Idle Takeoff Climb-out Approach Continental O-200 7.68 48.4 48.4 21.3 Lycoming O-320 13.0 65.7 63.5 23.1 Source: U.S. EPA, “Compilation of Air Pollutant Emissions Factors, Third Edition” (1977) As part of the Emissions Inventory Improvement Program (EIIP) in the early 1990s, EPA updated its guidance for preparing emissions inventories in a ten-volume series. The updated guidance for aircraft emissions inventories was presented in Procedures for Emission Inventory Preparation Volume IV: Mobile Sources (known as Procedures Volume IV) (U.S. EPA 1992). With respect to GA aircraft emissions, the methodology was updated to provide an approach for quantifying emissions if the specific aircraft fleet mix and engines were known, as well as an alternate approach involving fleet average emissions factors if the only data available were the level of LTOs at a facility as reported in the FAA’s Air Traffic Activity publication. Procedures Volume IV also improved the methodology by providing guidance on the adjustment of approach and climb-out TIM calculations to account for local mixing height, generally defined as the atmospheric ceiling above which vertical mixing of air (and air pollutants) does not occur. Again, however, no methodology specific to estimating Pb emissions from piston-engine aircraft was provided, although the fuel consumption data provided for piston engines were expanded slightly to include the Continental TSIO-360C engine model. Table 4 summarizes the fuel consumption and TIM information for piston-engine aircraft provided in Procedures Volume IV. Values published for the O-200 and O-320 engine differ slightly than those previously provided in AP-42. National Emissions Inventory – Pursuant to the Consolidated Emissions Reporting Rule (CERR) promulgated in June 2002, EPA began preparing the National Emissions Inventory (NEI) on a three-year cycle (U.S. EPA 2002). The NEI catalogs emissions from point, non-point, area, mobile, and stationary sources by state and county. In estimating aircraft lead emissions for the 2002 and 2005 NEI, EPA relied on a methodology developed for use with on-road vehicles designed to operate on leaded gasoline (U.S. EPA 1998). The methodology accounted for piston-engine Pb emissions by taking the total gallons of aviation gasoline (avgas) produced in 2002 and 2005 and -11-

Table 4 Updated Fuel Flow Rates from AP-42 Engine Model Fuel Flow Rate per Mode of Operation (lb/hr) Taxi/Idle Takeoff Climb-out Approach Continental O-200 8.4 45.0 45.0 25.8 Continental TSIO-360C 11.4 133.2 99.6 61.2 Lycoming O-320 9.6 88.8 66.6 46.8 Source: U.S. EPA, Procedures for Emissions Inventory Preparation Volume IV: Mobile Sources (1992) factoring it by the American Society of Testing and Materials (ASTM) maximum allowable Pb concentration in 100 low-lead (100LL) aviation gasoline (2.12 grams Pb/gallon avgas) and assuming that 25% of the Pb consumed was retained in aircraft engines. The 25% Pb retention assumption was developed using data from measurements made in the exhaust from vehicles operating on leaded fuel. The resulting national Pb aircraft emissions estimates were then apportioned to 3,410 airports based on the level of piston-engine aircraft activity reported in the FAA’s Terminal Area Forecast (FAA TAF) for the year in question. In a technical support document and guidance document issued in 2008, EPA first described the calculation of Pb emissions from piston-powered aircraft based on the use of aircraft-specific emission factors and activity data (U.S. EPA 2008). This methodology focused on refining Pb estimates specific to the LTO cycle. These refinements are summarized below. 1. Computing single- and twin-piston-engine LTOs based on the FAA’s General Aviation and Air Taxi Activity Survey (GAATA) compiled in 2005. 2. Applying times in mode for piston-engine aircraft operations contained in Procedures Volume IV to fuel flow rates for piston-engine aircraft available in the FAA’s Emissions and Dispersion Modeling System (EDMS) to compute single- and twin-engine fuel consumption values per LTO in combination with an aviation gasoline density of 6 lbs/gallon. Table 5 presents the EDMS engine fuel flow data. 3. Computing a weighted-average fuel consumption value per LTO, according to the proportions that 90% of landings reported in the GAATA Survey were completed by single-engine aircraft and 10% were completed by twin engine aircraft. The resulting factor was 7.34 g Pb/LTO. 4. Reducing the assumed Pb retention factor from 25% to 5%. -12-

Table 5 EDMS Fuel Flow Rates Engine Model Fuel Flow Rates by Mode of Operation (lb/hr) Taxi/Idle Takeoff Climb-out Approach Continental 6-285-B 72.1 153.0 166.0 83.3 Continental IO-360-B 8.1 103.0 71.7 36.6 Continental O-200 8.3 45.2 45.2 25.5 Continental TSIO-360C 11.5 133.3 99.2 61.0 Lycoming IO-320-D1AD 7.8 91.7 61.4 37.6 Lycoming O-320 9.4 88.9 66.7 46.5 Lycoming TIO-540-J2B2 25.0 259.7 204.5 99.2 Wright R-1820 88.9 1165.9 861.9 323.0 Source: Federal Aviation Administration Emissions and Dispersion Modeling System (version 5.0.2) This general methodology was also used with some revisions to generate the 2008 NEI estimates. (ERG 2011). For example, EPA expanded the list of data sources used to compute LTOs at GA facilities for which data are available (ERG 2011). These sources include the FAA Form 5010, FAA’s Operations Network (OPSNET) and ATADS and TAF databases, and the Bureau of Transportation Statistics (BTS) T-100 database. Furthermore, an approach was outlined to estimate LTO activity based on the number of GA aircraft based at a facility if LTO data are unavailable: LTOs = 1293 + 203 * (# based aircraft) + 0.0019 * (county population) – 473 * [Alaska – 144 * (AlaskaXaircraft)] Where: • LTOs = Landing Takeoff Operations (i.e., 1 landing + 1 takeoff) • [Alaska – 144 * (AlaskaXaircraft)] = correction factor to account for the effect that aircraft based in Alaska have on the suitability of the equation For facilities with neither LTO data nor aircraft-based data, EPA proposed that the bottom 10% of LTO values calculated according to the equation above was representative of the missing activity. Additionally, EPA considered the median number of LTOs reported at heliports as representative of helicopter activity at facilities where data are unavailable. Finally, EPA provided an estimate of Pb emissions occurring outside of the -13-

LTO cycle during the cruise mode of operation, apportioned to individual states based on their share of the national general aviation and air taxi LTOs used in the NEI. The NEI documentation acknowledges that its methodology would benefit from the following improvements (ERG 2011): 1. Use of airport-specific LTO and TIM data, and an improved process by which LTOs are computed from the number of aircraft based at the airport if LTO information is not available; 2. Use of gasoline Pb concentrations based on data specific to the fuel being supplied at an airport; and 3. Fuel consumption rates specific to the fleet mix operating at an airport. 3.2.2 Other Methodologies Harris and Davidson calculated Pb emissions from piston aircraft operating within the South Coast Air Basin (SCAB) of southern California by inputting 2001 LTO data from facilities within the basin into EDMS (Harris and Davidson 2005). For this assessment, aircraft activity was modeled in EDMS using the Cessna 172, Piper PA28 and Cessna 150 aircraft types, assuming a 64.9-minute total flight duration, and that 42.1% of this average flight activity occurs below the local atmospheric mixing height (27.3 minutes). Sulfur dioxide (SO2) emissions calculated by EDMS were converted to Pb emissions by applying a speciation factor of 0.739 and an uncertainty estimate of 17.5%, resulting in an overall emissions inventory of 267 kg of Pb per year in the SCAB. Detailed studies of Pb emissions have also been conducted at SMO, each of which included emissions inventories prepared using distinct methodologies. At the behest of the Santa Monica Airport Working Group, Piazza (1999) computed emissions of Pb at SMO using SMO-supplied aircraft fleet and activity data, engine emission factors from AP-42 and the FAA’s Aircraft Engine Emission Database (FAEED), and the calculations outlined in Procedures Volume IV. Notably, default LTO times in mode were applied in the study. On behalf of EPA, ICF International prepared a Pb emissions inventory based on 2008 piston-engine aircraft activity at SMO (ICF International and T&B Systems 2010, Carr et al. 2011). Activity data were obtained directly from the airport and used as inputs to the 2008 NEI methodology to estimate Pb emissions. Fuel consumption was calculated based on an 11.8-minute LTO cycle (compared to the 27.3 minutes used in previous studies). The ICF methodology applied at SMO included two modes of operation that were previously unaccounted for in the then-existent methodologies: aircraft run-up and landing. The inclusion of engine run-up was a significant improvement, as this mode commonly occurs in order to perform safety checks. Moreover, sensitivity analysis conducted by ICF showed engine run-up to be one of the most important factors related to total aircraft-related Pb emissions at SMO. The study also accounted for fuel -14-

consumption during landing based on the assumption that engines would operate at full load. Run-up fuel consumption rates were obtained from operational manuals for the IO-360, IO-320, GSO-480, IO-550, TIO-540-J2B2, and TSIO-550 engines. When averaged, these yielded a rate of 13 gallons per hour for twin-engine aircraft and 7 gallons per hour for single-engine aircraft. Based on the two most common piston engines in the aircraft fleet at SMO (i.e., the IO-360 and the IO-320), a fuel consumption rate of 5 gallons per hour was ultimately selected. Piston-powered helicopter activity was accounted in the ICF study based on a 20-minute LTO cycle and an average Pb emission rate of 6.6 grams per LTO. Notably, 25% of the LTO activity reported for helicopters was assumed to be conducted by piston-engine powered machines, based on estimates from the airport operator. 3.2.3 Swiss Federal Office of Civil Aviation Methodologies and Data The Switzerland Federal Office of Civil Aviation (FOCA) has collected emissions and fuel consumption data on many types of piston-powered aircraft and helicopter engines, and has published guidance on the calculation of emissions and fuel consumption resulting from their operation. Table 6 summarizes piston-engine fuel consumption rates per mode of operation collected by FOCA (Switzerland Federal Office of Civil Aviation 2007). FOCA has also developed a relationship between fuel flow and shaft horsepower (HP) for piston helicopter engines (Switzerland Federal Office of Civil Aviation 2009). Times in mode associated with the FOCA helicopter emissions estimation methodology comprise 5 minutes total ground idle time, 3 minutes takeoff time, and 5.5 minutes approach time. Notably, fuel flow rates derived for 172 distinct helicopter airframe/ piston-engine combinations, as computed using the equation below, are also provided. Fuel Flow (kg/s) = 1.9*10-12* SHP4 - 10-9*SHP3 + 2.6*10-7*SHP2 + 4*10-5*SHP + 0.006 Where: • SHP = Shaft Horsepower; assuming operation at 20% maximum SHP during idle; • 95% maximum SHP during takeoff, 60% maximum SHP during approach; and • 90% maximum SHP during cruise. -15-

Table 6 Swiss FOCA Fuel Flow Rates Engine Model Horsepower (HP) Fuel Flow Rate per Mode of Operation (lb/hr) Taxi/Idle Takeoff Climb-out Approach IO-320-DIAD 160 7.9 92.1 61.1 37.3 IO-360-A1B6 200 11.1 107.9 84.1 49.2 IO-540-T4A5D 260 19.8 132.5 117.5 58.7 IO-550-B 300 30.2 144.4 142.9 77.8 O-200 100 7.9 45.2 45.2 25.4 O-320-E2A 150 10.3 79.4 63.5 38.1 O-360-A3A 180 12.7 95.2 81.0 42.9 O-540-J3C5D 235 12.7 131.7 111.1 52.4 Rotax 582 DCDI 64 4.8 31.7 28.6 12.7 Rotax 912 80 9.5 30.2 24.6 14.3 Rotax 912S 100 4.0 42.4 32.7 18.3 Rotax 914 114 14.3 57.1 44.4 23.0 TAE-125-01 135 2.4 50.8 40.5 19.8 TIO-540-J2B2 350 25.4 259.5 204.8 99.2 TSIO-360C 225 11.1 133.3 99.2 61.1 TSIO-520-WB 325 48.4 214.3 182.5 111.1 Unspecified < 200 HP 150 9.5 88.9 66.7 46.8 Unspecified > 500 HP 1200 7.9 1780.2 356.4 174.6 Unspecified 201 to 300 HP 225 11.1 133.3 99.2 61.1 Unspecified 301 to 500 HP 350 25.4 259.5 204.8 99.2 Source: Switzerland Federal Office of Civil Aviation, 2007. 3.2.4 Additional Piston-Engine Fuel Consumption Data Fuel consumption rates from select piston aircraft engines have also been collected through a series of technical reports prepared by both the FAA and the Coordinating Research Council (CRC) as part of a series of tests conducted on the viability of unleaded fuel alternatives to avgas (Atwood 2007, Atwood 2009, Atwood and Camirales 2004, Atwood and Knopp 1999, Coordinating Research Council 2010). These rates were -16-

collected at varying engine power settings, many of which do not directly correspond to those assumed in the existing emissions inventory methodologies. Figure 4 shows the covariation of mass fuel flow (in lbs/hr) with brake specific fuel consumption (BSFC, in lbs/hp-hr) for a sample of engine test data taken from these studies, representing a variety of engines (e.g., IO-540-K), power settings (e.g., 100%, 85%, 75%), and fuel blends (e.g., 100LL, UL). Relationships shown by the data in Figure 4 include the following: • BSFC and mass fuel flow values typically decrease as engine power setting decreases; • TSIO and TIO series engines typically have the highest fuel consumption rates in the data set in terms of both BSFC as well as mass fuel flow, indicating that these engines operate at higher horsepower ratings; • The IO-320 series engine consumes the least amount of fuel compared to other engines for which data are available; and • With a few exceptions, each engine data series shows good linear correlation between BSFC and mass fuel flow (r2 > 0.9). Most significantly, Figure 4 demonstrates the utility of using BSFC to estimate fuel consumption for estimating Pb emissions from piston-engine aircraft because it allows fuel consumption to be varied as a function of engine load instead of having to obtain measured or estimated mass fuel flow rates for each power rating. 3.3 Issues Identified During the Review The major issues identified with the information and assumptions used in current methodologies for estimating aircraft-related lead emissions are discussed below. 3.3.1 Information Regarding Airframes and Engines Fuel consumption rates can vary between fixed-wing, experimental, light-sport, and rotorcraft types of airframes defined by FAA. However, existing emissions inventory methodologies do not segregate aircraft operations by airframe type for the purposes of calculating lead emissions from those aircraft equipped with piston engines. This is due in part to the lack of readily available data sources that adequately characterize operations by airframe type and also indicate how engine technology and usage (i.e., type and number of engines equipped to the airframe) can vary by airframe type. -17-

Figure 4 Covariation of BSFC with Mass Fuel Flow (Legend: Engine, Fuel, % Throttle) -18-

To address this issue, data regarding observed aircraft operations or data from provided airport records could be used in combination with the FAA’s Tail Number Registry and Type Certification database to determine specific airframe types and engines in operation at an airport, to the extent tail number information is available for the facility. Alternatively, the FAA’s TFMSC (previously the Enhanced Traffic Management System Counts, ETMSC) database could be used to obtain type data for aircraft using the facility. However, this approach is not likely to be as robust as the approach described above because, although TFMSC captures the vast majority of operations, it may not provide full coverage of certain localized types of flights. TFMSC data report predominantly instrument flight rules (IFR) flights, which are a minority of the types of flights conducted at many general aviation airports where visual flight rules (VFR) flights (e.g., training flights, recreational flying) are a significant fraction of total piston-engine powered aircraft operations. With that understood, cross-referencing of the TFMSC data with state/local level records from the Tail Number Registry could aid in engine selection by highlighting which engines are the “most flown” in the area for a given TFMSC aircraft type. 3.3.2 Engine Fuel Consumption Rates and Modal Load Assumptions Engine fuel consumption rates in existing methodologies are cast in terms of the mass of fuel used per unit time in a given mode and are generally based on data from EDMS that represent only a limited number of piston engines. Furthermore, these limited data are then averaged to yield one single-engine and one twin-engine fuel consumption rate, which are then combined into a single fuel consumption rate based on the assumed proportions of single- and twin-engine aircraft. It is further assumed that the fuel consumption rates are transferable across aircraft of varying sizes and with different engine technologies. Similarly, as noted previously, during recent assessments of Pb emissions at SMO, the methodology used was improved by accounting for Pb emissions during engine run-up mode. However, the fuel consumption rates for run-up were averaged based on single- and twin-engine aircraft most frequently in use based on the available operational data. Additionally, landing mode fuel consumption rates used in the most recent SMO study incorrectly assumed that engines would be operating at full load when, in fact, they operate at idle. The issues identified above could be addressed by developing BSFC rates for piston engines rather than using a generalized mass fuel flow rate for a limited set of single- and twin-engine aircraft. Using this approach, modal variations in horsepower and load factor can be incorporated into the fuel consumption and emissions calculations for the taxi/idle, takeoff, climb-out, approach, landing and run-up modes of operation, accounting for variations in engine performance within these modes. BSFC values can be extracted from either manufacturer specifications or from existing data published by the Switzerland FOCA, the FAA, and the CRC. The lack of available BSFC information for every type of engine may, however, preclude a complete characterization of total fuel consumption across the entire piston-powered aircraft fleet at a facility. Where data are unavailable for a specific engine(s), it is possible to estimate BSFC based on data available for other similar engines. -19-

Additionally, although modal load factors are available for the more traditional modes of operation within the LTO cycle, factors for new modes (i.e., run-up/maintenance), as well as refinements to existing modes (i.e., continuous lowering of load during approach and landing), can be developed using assumptions regarding average engine load either assigned directly from available engine test data or interpolated using fuel consumption relationships obtained from the Swiss FOCA data or other similar sources. 3.3.3 Aviation Gasoline Lead Concentrations The general practice at present is to use a fuel-based Pb emissions factor developed using the ASTM maximum allowable concentration of Pb in 100LL aviation gasoline, which is 2.12 g/gallon. However, a recent survey of 89 aviation gasoline samples from FAA fixed base operators (representing nine refineries) indicated Pb concentrations ranging between about 1.3 and 2.1 g/gallon. Additionally, 23 aviation gasoline samples obtained from engine manufacturers for use in certification testing exhibited Pb concentrations ranging between 0.3 and 2.2 g/gallon (Coordinating Research Council 2011). Material Safety Data Sheets (MSDS) available from aviation gasoline producers (British Petroleum 2011, Chevron Global Aviation 2003, ConocoPhllips 2010, Shell Energy North America 2003, Petro-Canada 2009, Phillips Petroleum 1998) corroborate this variance in the concentration of Pb actually contained in 100LL aviation gasoline blends compared to the ASTM maximum, demonstrating that the local sources of aviation gasoline supply may be influential in refining airport emissions inventories. Given that aircraft-related Pb emissions are directly proportional to the amount of Pb in the fuel, under ideal circumstances the actual concentration of lead in the fuel being used by each aircraft would be available for use in preparing an airport Pb emission inventory. Given that these data do not exist, the next best alternative is to use data regarding the Pb concentrations present in the fuel being dispensed at the airport. In general, Pb concentration data should be available for each load of aviation gasoline delivered to an airport or each batch of fuel to the extent that aviation gasoline is delivered to an airport by pipeline. However, these data may be unavailable from airport operators or fuel suppliers, and manufacturers may consider the information proprietary. Alternatives include historic data and forecast Pb usage levels from airport fuel suppliers or the best available existing data regarding Pb concentrations as a function of fuel grade, geographic region, and season from published sources. 3.3.4 Engine Lead Retention The fraction of total lead in the fuel consumed by an aircraft engine that remains in the engine and exhaust system is another key parameter that has to be accounted for during inventory preparation. Ideally, lead retention data, which would be expected to vary somewhat with engine technology and exhaust system design, would be available on an aircraft-specific basis; however, data at this level of detail are not available. -20-

As noted above, EPA’s original methodology for calculating Pb emissions from piston aircraft within the NEI assumed that 25% of the Pb consumed by the engine was retained in the engine oil based on data from automobiles designed to operate on leaded fuel. More recently, Petersen (2008) performed a study that estimated lead retained in aircraft engines to be 5%. This study included quantification of the concentrations of lead in used and new fuel lubricants for a sampling of in-use piston engines, the results of which are summarized in Table 7 to estimate the amount of lead retained in engine oil. A representative value for lead retained in engine oil and an estimate of lead retained in the engine itself from the Swiss FOCA were then used along with an estimate of the lead consumed in fuel to arrive at the 5% retention value. In the absence of additional data, this 5% retention rate currently represents the best available information for inventory preparation. However, additional research in the area would be useful to confirm the value. Table 7 Piston-Engine Oil Data Sample Lead (ppm) Test Hours Sump Capacity (quarts) IO-360 2,453 50 8 O-300 3,605 25 8 O-320 1,726 20 8 O-320 2,911 20 8 C-85 3,747 21 4.5 IO-360 2,017 40 8 O-235 L2C 5,797 100 6 O-300 4,456 40 8 IO-550 5,536 50 12 O-320 D2J 10,286 100 8 New Unused Oil 226 0 0 Source: Petersen 2008 3.3.5 Time in Mode The piston-engine aircraft TIM data currently utilized in quantifying aircraft-related lead emissions are summarized in Table 8. EPA’s NEI methodology has historically utilized the EPA/ICAO (International Civil Aviation Organization) standard TIM reported in Table 8, while the Switzerland FOCA methodology has its own set of standard times. For comparison purposes, times in mode for select GA airports are also provided in Table 8, demonstrating that facility-specific data, including fleet mix and taxi patterns, are important factors to consider in estimating TIM values. Specifically, taxi times -21-

Table 8 Time-in-Mode Comparison Facility Piston Aircraft Times in Mode (minutes) Run-up Taxi (Idle) Takeoff Climb- out Approach Total EPA/ICAO Default --a 16.0 0.3 5.0 6.0 27.3 FOCA Default -- 12.0 0.3 2.5 3.0 17.8 DMW (Carroll County Regional Airport, Maryland) – 5 1.2 1.7 7.4 15.4 DPA (DuPage Airport, Illinois) – 13.9 1.6 2.2 8 25.8 GED (Sussex County Airport, Delaware) – 9.9 1.6 3.1 7.7 22.3 LSZB – circuit (Bern Airport, Switzerland) – 11.1 0.3 1.3 3.6 16.3 LSZB - LTO – 11 1.0 3.5 7.5 23.0 MWC (Timmerman Airport, Wisconsin) – 8.6 1.8 1.9 7.2 19.4 SGJ (Northeast Florida Regional Airport) – 7.3 2.1 2.4 7.9 19.7 SMO 1.5 7.4 0.3 1.3 1.3b 11.8 TPF (Peter O Knight Airport, Florida) – 3.6 1.6 2.3 7.2 14.5 VDF (Tampa Executive Airport, Florida) – 5.2 1.5 2.3 7.2 16.1 a. Dashes indicate that data are not available. b. Includes landing Sources: Switzerland Federal Office of Civil Aviation, U.S. Environmental Protection Agency, and KB Environmental Sciences, Inc., Carr et al. 2011. utilized in emissions inventories at the reported facilities are significantly lower than both EPA/ICAO and FOCA defaults. Conversely, takeoff times utilized in the emissions inventories are somewhat higher than both the EPA/ICAO and FOCA default values. For climb-out, the FOCA default TIM approximates the airport-specific times, while the EPA/ICAO default may be more representative for the approach mode of operation based on the presented data. Another important component to TIM calculation for takeoffs and landings is the local atmospheric mixing height above which Pb emissions are not allocated to the airport emission inventory. -22-

Overall, the data show considerable variation in TIM values, and TIM data collected at specific airports show that fleet mix, the local mixing height (which affects TIM for takeoffs and landings), and taxi patterns can have a significant bearing on actual TIM values. Given the above and the importance of accurate TIM data to the accurate quantification of Pb emissions, the best approach is to develop airport-specific TIM values. This could be done through surveys of airport personnel designed to obtain as much airport-specific information regarding TIM as possible or through a dedicated on- site data collection effort. The latter approach should include observations on aircraft speeds, taxi-paths, and taxi-path distances, and applying them to available runway usage data for various types and/or categories of aircraft. 3.3.6 Total Aircraft Operations, Aircraft Fleet Operations, and Temporal Variations Total Operations – Ideally, the data used to estimate total aircraft operations at a facility should be segregated by engine category (e.g., turboprop) and operational category (e.g., military); however, review of existing methodologies and data sources indicates that robust data are readily available only by operational category. In light of this, the primary approach recommended for this parameter is to acquire data from airport flight strips, counter systems, fixed base operator (FBO) logs, and/or other sources such as direct observations that may be available at a given airport. Absent the types of data described above, the FAA’s ATADS provides another source of total aircraft operations data, with those operations categorized as (1) air carrier, (2) air taxi, (3) general aviation, or (4) military. Furthermore, ATADS data are available for approximately 540 airports within the FAA’s National Plan for Integrated Airport Systems (NPIAS). The remainder of the operational data for the over 3,400 hub, non- hub, and general aviation facilities in the NPIAS are based on the FAA’s Terminal Area Forecast (TAF) database. Aircraft Fleet Operations – Ideally, once total operations in each category are computed, they should, as discussed below, be allocated to each airframe in the fleet, with emphasis on being able to adequately represent the proportions of operations within each category at the facility. As above, airport-provided data sources such as flight strips, counters, and data logs would be primarily consulted. An alternative approach, to be used in instances where the above approach is not feasible, could be to use the FAA’s TFMSC database to obtain operational data for each type of aircraft using the facility. TFMSC provides information on operations by category (e.g., air taxi), airframe (e.g., Learjet 35) and engine type (e.g., jet) for the subset of operations conducted under filed flight plans, or recorded under instrument flight rules (IFR). This approach is considered less desirable because, although TFMSC captures the vast majority of operations, it may not provide full coverage of certain localized flight types common to general aviation airports. Temporal Variations in Aircraft Fleet Operations – In addition to characterizing aircraft fleet operations in general, it may be important—particularly if modeling of ambient Pb -23-

concentrations at an airport is going to be performed—to characterize temporal variations in those operations. For example, given that aircraft usage can vary by season, month, or day of the week, resolution of that variation will improve airport Pb emission inventories that are prepared for similar time frames. Again, the primary recommendation for addressing temporal variations in the aircraft fleet is to use data based on direct observations at airports. Alternatively, the TFMSC database could be used, again with the caveat that it may not provide full coverage of certain localized flight types. 3.3.7 Non-Combustion Sources of Lead While current methodologies for quantifying aircraft-related lead emissions focus only on emissions from the combustion of leaded aviation gasoline, monitoring studies have identified elevated concentrations of Pb in soils at and surrounding airports, ranging between 21.7 and 232.5 µg of Pb per kg of soil. However, most of the research in this area has failed to demonstrate a clear spatial relationship linking soil concentrations with airport activity (Conor Pacific Environmental Technologies 2000, ICF International and T&B Systems 2010, Lejano and Ericson 2005, Young et al. 2002). Regardless of whether lead in soils at airports is due to aircraft operation, assessments of total Pb emissions at airports and, in particular, studies focused on the contribution of airports to Pb concentrations in TSP or PM10 should consider contributions from dust resuspension. Ideally, in order to estimate total airport lead emissions and airport contributions to ambient lead concentrations, data would be available regarding lead concentrations in soils and dust at the airport, as would data regarding soil entrainment and dust re-suspension rates. Unfortunately, these data are not generally available and are difficult to estimate. In the absence of data addressing non-combustion sources of lead from special airport-specific studies, the alternative is to use some combination of default local airport soil lead concentrations in conjunction with current EPA methodologies for estimating dust re-entrainment. 3.3.8 Validation of Emission Estimates As discussed in detail above, quantification of aircraft-related Pb emissions is complex and requires many assumptions that are likely to have varying degrees of accuracy. Given this, some studies use air quality modeling and gather ambient Pb monitoring data in order to validate the accuracy of emissions inventories. This type of study requires that the emission inventory include highly detailed information regarding the temporal and spatial distributions of aircraft operation and Pb emissions. Results are usually presented in terms of a comparison of modeled to monitored emissions for specific time periods. Figure 5 provides an example of data from such a study performed by Carr et al. (2011) at SMO. The model tends to be biased high (especially in the winter) but, overall, there is good agreement between measured and modeled PM-Pb concentration values, with an absolute fractional bias of 0.29 for the winter data and 0.07 for the summer data. -24-

Figure 5 Model-to-Monitor Comparison at SMO 3.3.9 Proper Documentation A general finding throughout the literature search of existing emissions inventory methodologies, as well as airport-specific emissions inventories, was a lack of sufficient detail to allow emissions inventory results to be recreated. Inadequate documentation included, but was not limited to, the following: • Specification of operational data sources with inadequate detail on how they were accessed, used, and manipulated for the purposes of preparing an emissions inventory; • Specification of TIM with inadequate detail on how it was either empirically observed or calculated from obtained data; • Lack of detail on averaging methods and operational assumptions (i.e., load points, horsepower) in developing fuel consumption rates used for piston-engine aircraft in existing emissions inventories; and • Incomplete documentation of supporting data or communications that guided underlying assumptions. 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 M od el ed P b, n g/ m 3 Measured Pb, ng/m3 Summer Winter -25-

Clearly, proper documentation of aircraft Pb inventories is important because it ensures reproducibility, which aids in validation, and helps identify potential sources of erroneous results that could be avoided by refining assumptions or adjusting the fidelity at which each inventory parameter is treated. In order to ensure that proper documentation is being provided, the technical issues outlined below need to be addressed in detail. Fleet Identification • For each aircraft category (i.e., GA, air taxi, helicopter), specify which airframe/engine combination was assumed. • Specify how many engines are equipped to the aircraft and, to the extent possible, how the engine is configured (i.e., HP rating). • Detail sources of fleet data, when they were accessed, and which time period(s) they cover, and summarize how they were processed or manipulated. Provide end results (i.e., fleet mix). Operational Specifications • Specify the level of operations for each aircraft/engine combination in the fleet, which sources of data were consulted to develop the operational levels (and when), and how the operational levels were derived. • Identify data sources and assumptions used to compute aircraft TIM and how TIM was computed, and summarize results for each fleet member and each operational mode. • Per mode of operation, indicate assumptions or empirical data used to assign engine load points to aircraft fleet. Emissions Factor/Fuel Consumption Derivation • For each member of the fleet, identify sources of modal fuel consumption data and present calculation steps based on observed/assumed operational parameters. Provide all rates utilized in the emissions inventory, per fleet member and operational mode. • Indicate the fuel Pb concentration(s) used in emissions factor development and present data and the rationale for these values used. ### -26-