Environ. Sci. Technol. 2008, 42, 1884–1891
Speciation and Chemical Evolution of Nitrogen Oxides in Aircraft Exhaust near Airports EZRA C. WOOD,* SCOTT C. HERNDON, MICHAEL T. TIMKO, PAUL E. YELVINGTON,† AND RICHARD C. MIAKE-LYE Aerodyne Research, Inc., 45 Manning Road, Billerica, Massachusetts 01821
Received August 16, 2007. Revised manuscript received December 20, 2007. Accepted January 14, 2008.
Measurements of nitrogen oxides from a variety of commercial aircraft engines as part of the JETS-APEX2 and APEX3 campaigns show that NOx (NOx ≡ NO + NO2) is emitted primarily in the form of NO2 at idle thrust and NO at high thrust. A chemical kinetics combustion model reproduces the observed NO2 and NOx trends with engine power and sheds light on the relevant chemical mechanisms. Experimental evidence is presented of rapid conversion of NO to NO2 in the exhaust plume from engines at low thrust. The rapid conversion and the high NO2/NOx emission ratios observed are unrelated to ozone chemistry. NO2 emissions from a CFM56-3B1 engine account for approximately 25% of the NOx emitted below 3000 feet (916 m) and 50% of NOx emitted below 500 feet (153 m) during a standard ICAO (International Civil Aviation Organization) landing-takeoff cycle. Nitrous acid (HONO) accounts for 0.5% to 7% of NOy emissions from aircraft exhaust depending on thrust and engine type. Implications for photochemistry near airports resulting from aircraft emissions are discussed.
1. Introduction 1.1. NOx Emissions and Air Quality. Although much recent research has focused on the effects of aviation on the chemistry of the stratosphere and upper troposphere (1, 2), much less has been devoted to its effects on the lower troposphere (3, 4). The chemical composition of aircraft engine exhaust emissions is markedly different across the wide range of thrust levels used at airports (e.g., idling and takeoff). The impact of these low-altitude aircraft emissions of volatile organic compounds (VOCs), particulate matter, and nitrogen oxides on local and regional air quality is not well-known. With global air traffic expected to increase greatly in the coming decades, the environmental impact of airport operations is becoming an increasingly salient issue (5, 6). The impact of NOx emissions and subsequent photochemistry (including the formation of ozone and secondary aerosol) on public health and the environment are well documented (7–10). Attaining a full understanding of the impact of airport emissions on local and regional air quality requires a thorough accounting of airport NOx emissions. Plans for a third runway at London Heathrow International Airport were postponed in 2003 because of concerns that it * Corresponding author e-mail:
[email protected]; phone: 978-663-9500 x219; fax: 978-663-4918. † Now at Mainstream Engineering Corporation, 200 Yellow Place, Rockledge, FL 32955. 1884
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 6, 2008
would lead to exceedances of the 2010 European Union directive limit for NO2 of 40 µg/m3 (∼21 ppb) in the surrounding area. Many air basins in the United States are in nonattainment of the AQS for ground-level ozone (an 8-h average of 80 ppbv), the chemical production of which is intimately connected to NO2. Moreover, the California Air Resources Board recently approved amendments to the California ambient air quality standard for NO2 that will lower the AQS to an annual average of 30 ppb (11). Emissions of CO, NOx, total hydrocarbons (VOCs), and exhaust opacity (“smoke number”) from new aircraft turbine engines must meet the certification standards of the International Civil Aviation Organization (ICAO). These ICAO standards make no distinction between NO and NO2 emissionssonly total NOx is specified. Although most combustion-produced NOx is emitted in the form of NO, it has been observed that NO2 can constitute a significant fraction of the NOx emissions from some sources of combustion (12, 13). This is of concern because NO2 is more toxic than NO, and primary NO2 emissions lead to enhancements in ozone production (14). In contrast to the advanced state of knowledge regarding NOx emissions from automotive sources (12, 13, 15), much less is known about aircraft emissions. The speciation of aircraft NOx emissions between NO and NO2 has not been well characterized and is usually assumed to resemble most combustion sources. 1.2. Impact of Airports on NOx and O3. Various studies have indicated that airports can have a significant effect on air quality in the surrounding area. Pison et al. (3) modeled the impact of airport aircraft emissions on ozone concentrations near Paris. They assumed an overall NO2/NOx emission ratio of 5% for aircraft engines, and their model results indicated that the greatest impact of aircraft emissions was a reduction of near-surface ozone concentrations near airports due to the nighttime titration of ozone by fresh NO emissions. A similar modeling study that focused on the impact of Hartsfield-Jackson International Airport on the Atlanta nonattainment area showed that consideration of the vertical distribution of the emissions was an important factor in determining the effect of aircraft emissions on local ozone photochemistry (4). Measurements of NO, NOx, wind speed, and wind direction at a number of sites near London Heathrow International Airport (16) indicated that 27% of both the ambient NO2 and NOx observed at the airport boundary were caused by airport NOx emissions. 1.3. Previous Aircraft NOx Measurements. Few studies report both NO and NO2 emission indices from aircraft, and the results of these studies are conflicting at times. Diehl (17) and Miller et al. (18) measured NO and NO2 emissions from military turbojet engines and found that the NO2/NOx emission ratio varied from 25% at low power up to 100% at military (high) power. Such turbojet engines do not have a bypass around the combustor (i.e., there is no fan flow) and are used only in a small fraction of today’s global aircraft fleet, as turbofan engines are the predominant type of modern engine. The measurements of Spicer et al. (19) found that NO2 accounted for at most 41% of NOx emissions from two turbofan engines at idle, though the hot (150 °C) stainless steel sampling system used might have contributed to a conversion of NO2 to NO (20). Herndon et al. (21) measured NO and NO2 in advected exhaust plumes from in-use commercial aircraft at New York’s John F. Kennedy International Airport and concluded that the NO2/NOx ratio observed in takeoff plumes was consistent with the oxidation of primary NO emissions by ambient ozone to form NO2. 10.1021/es072050a CCC: $40.75
2008 American Chemical Society
Published on Web 02/20/2008
Most recently, Wormhoudt et al. observed that at low power NO2 accounted for up to 80% of the NOx emitted by a CFM56-2C1 engine during the APEX campaign (22). HONO was found to account for 4-7% of the total NOy (NOy ≡ NO + NO2 + HONO + HNO3 + organic nitrates + . . .) emissions at low thrust and 2-4% of NOy emissions at thrusts higher than 20%. HONO/NOy values between 1 and 2% were measured from a RB211-535-E4 engine measured during the EXCAVATE campaign (23) and from the TRACE engine measured during the NASA/QinetiQ campaign (24). This paper investigates the speciation and chemical evolution of nitrogen oxide emissions from commercial aircraft engines using measurements from the JETS-APEX2 and APEX3 campaigns as well as a chemical kinetics model of the combustion chemistry. Measurements from both dedicated engine tests (using stationary aircraft) and advected exhaust plumes of in-use aircraft are considered. We show that aircraft engines emit NOx primarily in the form of NO2 at idle thrust, which leads to formation of O3 in the exhaust plume on a time scale of minutes. Over an entire landing takeoff cycle NO2 accounts for 16-24% of total NOx emissions, depending on engine type and other parameters such as the thrust values used and time spent during each mode (idle, takeoff, climb-out, and landing). Measurements of nitrous acid (HONO), which is a precursor to hydroxyl radicals (OH) in the atmosphere, and NOy are presented. An overview of gas-phaseandparticulateemissionsfromalloftheaircraft-engine combinations studied during JETS-APEX2 and APEX3 is presented in Timko et al. (25, 26).
2. Experimental Section A full description of the instrumentation used for the Aerodyne contribution to the JETS-APEX2 and APEX3 campaigns is described in detail by Timko et al. (26). The aspects most relevant to NOy measurements are described here briefly. The JETS-APEX2 (Aircraft Particle Emissions eXperiment) campaign was carried out at the Oakland International Airport (OAK), August 22-29, 2005, and the APEX3 campaign took place at the Cleveland Hopkins International Airport (CLE), Oct 30-Nov 13, 2005. In both field campaigns, the gas-phase and particulate components of the exhaust from a number of aircraft-engine combinations were measured by a team of research groups in a manner similar to that of the original APEX campaign (27). For each engine test, an airplane was chocked in the testing facility and run through a test matrix of various thrust settings. The exhaust from one engine was intercepted by a number of stainless steel probes at 2 or more distances downfield from the engine (typically 1 and 30 m) and distributed to the research groups’ laboratories for analysis. Four different Boeing 737 airplanes with CFM56 engines were measured during JETS-APEX2, while nine airplanes with a greater variety of airframes and engines were characterized during APEX3. A second measurement approach used during JETS-APEX2 and to a smaller extent APEX3 was the measurement of advected exhaust plumes from in-use aircraft on the tarmac. On August 26, 2005, three of the research groups measured the exhaust from over 200 in-use aircraft during landing, taxiing (idling), and takeoff operations. Visually recorded tail numbers identified the engines, which after referencing to the ICAO database provided the relevant certification emission indices. The measurements presented here were taken aboard the Aerodyne Mobile Laboratory (28), a panel truck designed for the measurement of gas-phase and particulate matter species in ambient air and vehicle emissions. A brief description of the relevant instrumentation for the measurements described in this manuscript follows. NO2 was measured with an Aerodyne quantum cascade tunable infrared laser differential absorption spectrometer
(QC-TILDAS) operating at 1606 cm-1 (e.g., ref (29),). At JETSAPEX2, NO was measured using a ThermoElectron model 42C chemiluminescence analyzer. Calibrations were performed by successive dilutions of a standard NO cylinder (Scott Specialty, accuracy +/- 2%) with NO-free compressed air. At APEX3, NO was measured using a lead-salt TILDAS at 1900 cm-1 and NOy was measured using a chemiluminescence analyzer operated in conjunction with an external molybdenum converter. Transmission of HNO3 and other “sticky” compounds through the inlet system was probably negligible and all measurements are interpreted as the sum of NO, NO2, and an unknown fraction of total peroxy and alkyl nitrates (30), though they are referred to as “NOy” in this document to distinguish them from the quantity “NOx” which is defined as the sum of [NO] and [NO2]. During APEX3, ozone (O3) was measured by a dual beam UV photometer (2BTech, model 205) at 254 nm. Carbon dioxide (CO2) was measured with two nondispersive IR absorption photometers (LICOR model 6262 and 820). Wind speed and direction were measured with a Vaisala anemometer. The uncertainty of all measurements described above was less than 5%. Nitrous acid (HONO) was measured during APEX3 using a lead-salt TILDAS operating at 1660 cm-1, similar to the measurements described in Wormhoudt et al. (22). The absorption linestrengths used for the 1660 cm-1 measurement were based on the spectral assignments of M. Herman (31–34). These relative intensities have been scaled to the absolute values in the ATMOS database (35) and checked with more recent laboratory measurements (36, 37). The linestrengths used are those of the cis isomer of HONO and are adjusted to account for an assumed equilibrium distribution of isomers. The accuracy of the HONO measurements is estimated as 50%, though this figure does not account for potential creation or destruction of HONO in the probe sampling lines. All gas-phase species were measured after passing through a 0.45-µm PTFE Teflon particulate filter.
3. Results and Discussion 3.1. Dedicated Engine Tests. Emission indices (EI, in grams of pollutant emitted per kg of fuel combusted) are calculated as described in Herndon et al. (21). Emission indices of NO, NO2, HONO, and NOy are all expressed in NO2 mass equivalents, which enables them to be compared quantitatively. Figure 1 depicts the observed NO, NO2, and HONO emission indices as a function of engine thrust setting for a CFM56-3B1 engine mounted on a Boeing 737-300 airframe (tail number N14324). EI NO and EI NOx increase with increasing power, in accordance with the ICAO certification data (displayed as black squares). The NO and NO2 data presented in Figure 1 were collected from the 30-m probe, and the HONO data were collected from the 1-m probe. The NO2 fraction of NOx decreases with power, from over 98% at the lowest power setting (4% rated thrust) to under 10% at higher powers (65-100% rated thrust), similar to the measurements presented in Wormhoudt et al. for a CFM562C1 engine during the APEX campaign. The profile of EI NO2 is dominated by a decrease with power, though there is a small increase in EI NO2 from 4% to 7% and again from 65% to 100% power. Similar trends were observed in the other 12 APEX2/3 engines (26), though the CFM56-3B1 was the only engine in which EI NO2 was higher at 7% thrust than at 4% thrust. HONO emission indices increase slightly with thrust, though HONO/NOy decreases. The observed EI HONO values for the CFM56-3B1 vary greatly depending on sampling probe. At the 1-m probe, HONO/NOy varied from 2% at low thrust to 1% at high thrust, whereas at the 30-m probe HONO/NOy varied from 4% to 0.5%. EI HONO also varied greatly between VOL. 42, NO. 6, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1885
FIGURE 1. Emission indices of NO, NO2, and HONO from a CFM56-3B1 engine measured during APEX3. The emission indices are displayed such that the contribution of each compound is added to those below it, such that the sum of HONO, NO, and NO2 is represented by the uppermost trace. ICAO certification values (uppermost squares) are added for comparison. The error bars for the HONO points are the standard deviation of the measurements, which for some points is higher than the 50% instrumental uncertainty for HONO. enginessfor example, HONO/NOy for the PW4158 engine ranged from as high as 7% at low thrust to 0.5-1% at high thrust. Total NOy emission indices (not shown) agree with the NOx emission indices within the experimental uncertainty (at most 6%). The NOx emission indices are approximately 10% lower than the ICAO certification values. The “correction” for the ambient temperature and humidity has only a small (
