Summarizing and Interpreting Aircraft Gaseous and Particulate Emissions Data (2008)

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.

S U M M A R Y The commercial aviation community is faced with the need to assess the impacts of avia- tion emissions on air quality. Until recent government-sponsored tests were undertaken, emissions of particulate matter (PM) from commercial jet engines were not well under- stood. Prior to the mid-1990s, jet engine PM emissions were identified as smoke and were exclusively quantified using the Society of Automotive Engineers (SAE) Recommended Practice 1179—Smoke Number (SAE 1991). The smoke number does not identify the key characteristics of the PM (morphology, chemical composition, distributional accounts of size and volume, or number and mass concentration) and is, therefore, of limited value to those parties analyzing environmental and health impacts of aviation emissions. As a result, alternative methods for characterizing PM emissions based on these key characteristics were developed. These methods were first applied to quantify PM emissions for military engines (Spicer et al. 1992, 1994; Howard et al. 1996; Whitefield et al. 2002). Although these meth- ods and data served to improve the scientific community’s understanding of aircraft engine PM emissions, these studies focused on engine technologies that are different from those currently used in the commercial fleet. A series of tests was devised and conducted by NASA and the FAA’s Partnership for AiR Transportation Noise and Emissions Reduction (PARTNER) Center of Excellence and other parties to address the need for data representative of engines in the commercial fleet. The data from these tests have recently been made public, however, until this report, they have not been distilled into a form directly useable and, in some cases, understandable by the airport community. Such a synthesis is the primary goal of this report. To facilitate under- standing of test results, this report begins with four primer sections on • PM characteristics, sources, air quality effects, and health consequences; • Hazardous air pollutants; • Field test methodologies; and • Models for the estimation of emissions, air quality effects, and health consequences. These primer sections are followed by a summary of test results and a review of relevant published material. The report is supported by four appendices that provide additional test details. Test data is available from the FAA and plans are being discussed on how to provide this information on a publicly accessible website. The cumulative dataset from these tests is extensive. It includes studies that assist with understanding how emissions evolve with distance from the engine. For several stationary commercial aircraft, emissions were measured in the near-field plume, referring to the exhaust stream within 1 to 50 m (3 to 164 ft) from the engine exit. Measurements were also taken of emissions downwind (>100 m [>330 ft]) from moving aircraft during routine operation at Summarizing and Interpreting Aircraft Gaseous and Particulate Emissions Data 1

2two large commercial airports. Each set of tests measured a range of particle characteristics, including number, size distributions, mass, and composition, as well as gaseous emissions concentrations, including nitrogen oxides (NOx including nitric oxide, NO, and nitrogen dioxide, NO2), carbon monoxide (CO), hydrocarbons (HC), and sulfur dioxide (SO2). Gas-phase measurements complement particle measurements providing a more specific description of aircraft engine emissions than has been available to date. This more complete dataset will improve estimates of airport contributions computed by air quality models. The primary observations and conclusions from these studies that are of interest to the airport community are • Emissions data at the exhaust nozzle from a subset of engines operating in the commercial fleet have been collected. These include – CFM56-2C1 on a NASA DC8, – CFM56-3B1, -7B22 on B737s, – JT8D on MD-88s, – CF6-80 on B767s, – PW2037 on B757s, – PW4158 on A300s, – RB211 on B757s, and – AE3007 on ERJ 135/145s. Combined, these engine types are present on more than 70% of current aircraft operations in the U.S. domestic fleet. • Particulate matter number and mass concentrations have been normalized by the amount of fuel burned to produce emission indices that allow the quantification of emissions per kilo- gram of fuel burned. The mass-based emission indices can be used to develop emissions inventories for the aircraft and engines studied. The PM First Order Approximation (FOA), which is implemented in the FAA’s Emissions and Dispersion Modeling System (EDMS), is an application of this technique. Prior to the Aircraft Particle Emissions eXperiment (APEX) studies, it was not possible to compute an emissions inventory of aircraft PM that was repre- sentative of current and future aircraft fleets. • In all cases, gaseous emissions and engine operating parameters revealed that the engines were operating in a representative manner (i.e., their combustion gas emissions were within the un- certainties of the emissions measurements conducted for certification). This being the case, it is reasonable to assume that the measured PM emissions are also representative and that the results reported should be used with confidence to develop emission inventories. • Unburned hydrocarbons are emitted as a variety of compounds including ethylene, formalde- hyde, acetaldehyde, and benzene. Most of these compounds are emitted at thrust levels <30%. Emissions of the various hydrocarbon (HC) species rise and fall with one another, regardless of engine type or thrust setting. Even when the absolute magnitudes increase by a factor of 10 or more (as is the case for older engine technology or for operation at low thrust condition or low ambient temperature), the ratio of one HC species to the next remains constant within the uncertainty of the measurement. • Measurements were made both at the exhaust nozzle and at locations in the near-field plume (downstream). The non-volatile PM (i.e., particles that exist at engine exit plane temperature and pressure conditions) mass and size did not change appreciably between the exhaust noz- zle and the downstream sampling points. Volatile PM (i.e., particles formed as the exhaust cools, from condensable gases such as sulfur oxides, HC, and engine oil) was observed at the exhaust nozzle in small quantities and increased by about a factor of 10 at the downstream sampling locations. Volatile materials entered the particle phase as new particles (<20 nm) and as coatings to preexisting particles.

• The following conclusions were drawn when emissions were sampled at the exhaust nozzle: – The measured PM parameters for each engine type (i.e., JT8D, CFM56, CF6, RB211, etc.) are unique. For example, in the case of the RB211, JT8D, and PW4158, the mass-based emission indices measured as a function of fuel flow ranged from 0.04 to 0.70, <0.01 to 0.32, and <0.01 to 0.18 g/kg-fuel respectively. – The measured PM parameters for engine subtypes are also unique. For example, for the CFM56-3B versus -7B engines, the ratio of their mass-based emission indices at takeoff was found to be 4:1 (-3B:-7B). – Credible inventories based on nozzle emission rates will require engine-specific data like that measured in these studies. – Black carbon PM (i.e., non-volatile particles) constitutes more than 80% of the mass of PM emissions at all thrust conditions. At takeoff thrusts, more than 95% of the total PM mass is black carbon PM. • The following trends were observed when emissions were sampled downstream in the plume (greater than 10 m [33 ft] from the exhaust nozzle): – As the plume cools, condensable exhaust gases convert to the particle phase by nucleating new particles and by condensing onto black carbon PM. Collectively, the new particles and black carbon PM coatings are referred to as “volatile PM.” – Newly formed volatile particles outnumber black carbon PM by a factor of 10-100 in the cooling exhaust gases. (The number of particles formed in the cooling plume is determined by sulfur from the fuel, the amount of black carbon PM surface area available for conden- sation, and ambient conditions.) – Besides sulfate and organic substances, no other volatile materials are present at concen- trations greater than in the ambient background. – For most engines, HC sourced to incomplete combustion and lubrication oil constitute >95% of the volatile organic material that can be accurately characterized. The ratio of par- tially burned HC to lubrication oil depends on engine technology and thrust setting. For certain engines, lubrication oil constitutes up to 90% of the organic PM emitted at high thrust where combustor efficiency is maximized and unburned fuel is at a minimum. – The mass of particles in the plume does not change within experimental uncertainty as the plume travels downwind, but the number of particles increases by at least an order of magnitude. The large increase in PM number coupled with constant PM mass indicates that these newly formed particles do not contribute significantly to the total PM mass in the plume. This result indicates that a mass-based inventory alone will not capture this signif- icant volatile PM production. • Atmospheric conditions impact the measured parameters and need to be taken into account when measuring emission rates for inventory development. For example, a decrease in ambi- ent temperature from 35°C to 26°C (95°F to about 79°F) at one test site (NASA Dryden, APEX1) resulted in the formation of new additional volatile particles at 30 m (nearly 100 ft) downstream of the exhaust nozzle, not observed at the higher temperature. These new parti- cles had a mean diameter of <15 nm and increased the total number-based emission index by an order of magnitude at this distance. • The length of time that the engine has been running also impacts the measured PM parameters. For example at JETS APEX2, the number-based emission index measured at idle for a CFM56-7B engine was 50% lower after the engine was fully warmed up, compared to the ini- tially measured value. • It was not possible to statistically determine if there was any variation among subsets of various engine types due to the limited numbers of engine variations tested. • The engine-specific nature of PM emissions as described above indicates that additional studies will be needed to understand variability in PM emissions among engine types. Specifically, 3

4tests that address the 50,000- to 100,000-lb thrust class employed by such aircraft as the B747, B757, B767, B777, B787, A300, A310, A330, A340, A350, and A380 have not been investigated. These studies have improved our understanding of aircraft emissions. They have yielded data from more relevant engine technologies than had been available previously. The mea- surement methods developed in these studies provide an excellent foundation for future studies. The data can be confidently applied to assess and improve current predictive tools such as the FAA Aviation Environmental Design Tool/Emissions and Dispersion Model- ing System (AEDT/EDMS) and their predictive subcomponents such as the PM FOA methodology.