Emissions of Volatile Particulate Components from Turboshaft Engines Operated with JP-8 and Fischer-Tropsch Fuels

Particulate emissions from two types of helicopter turboshaft engines operated with military JP-8 and paraffinic Fischer-Tropsch (FT) fuels were characterized as an objective of the field campaign held at the Hunter Army Airfield in Savannah, GA in June 2007. In general helicopter engines exhaust particles size distributions observed at the engine nozzle and 4.14 m downstream locations showing the geometric mean diameters smaller than 50 nm for all engine power settings investigated in this study. For both locations, the geometric mean diameter increased as the engine power setting increased; this trend also holds true for the emitted particle number concentration. The growth of particle geometric mean diameter was found significant, 7 nm, only in the case of the idle power setting.

is anticipated that the increased demand on air travel will increase output of undesirable emissions, such as carbon monoxide (CO), carbon dioxide (CO 2 ), nitrogen oxides (NO x ), sulfur dioxide (SO 2 ), hazardous air pollutants (HAPs) like formaldehyde, 1,3-butadinene, acetaldehyde, acrolein, benzene, ethylbenzene, naphthalene, unburned hydrocarbons and particulate matter. Emissions of gaseous species (e.g., HAPs, CO, and unburned hydrocarbons) are generally much higher during idling, taxiing, and low engine power operations than that at the cruise or max power setting. Particulate matter (PM) emitted by aircraft are in the ultrafine particle-size region with the peak mobility diameter typically less than 100 nm (see Rogers et al., 2005;Cheng et al., 2008) Petzold and Schröder, 1998;Wong et al., 2008). Species within these two classes play active roles in chemical reactions and aerosol microphysics.
For instance, soot particles are known to be hydrophobic and contain almost no watersoluble materials (Wyslouzil et al., 1994;Petzold and Schröder, 1998). Adsorption of sulfuric acid agglomerates and molecular clusters (formed through nucleation, for instance) could occur inside a jet exhaust plume in the atmosphere (Miake-Lye et al., 1994;Kärcher, 1996) leading to chemically modified soot particles that become hydroscopic with improved light-scattering efficiency. Frenzel and Arnold (1994) reported that the formation of sulfite in the plume is only restricted to plumes of age less than 10 ms. Gas-to-particle conversion and coagulation that lead to the buildup of sulfuric  (Corporan et al., 2004;Wade, 2004;Rogers et al., 2005;Chen et al., 2006;Jones et al., 2007) reported measurement of polynuclear aromatic A photograph taken during the T700/T701C en ng lines were used to tra gine emission test is shown in Fig. 1(a). The extractive sampling probes were mounted in a vertical rake with the probe tips approximately 40.6 cm downstream of the engine nozzle. A vertical plane at the distance of the probe tip is called the engine exhaust plane or EEP. Fig.   1(b) shows a close-up look of the sample probes mounted on the stainless steel rake.
Three tip-dilution probes were used for particle samples, three undiluted probes were used for gas sampling, and two thermocouples were used for measurement of exhaust plume temperature. Each probe was separated by 3.18 cm center-to-center. The particles were diluted with nitrogen gas immediately at the probe tip upon sampling Corporan et al., 2008). The diluted and undiluted probes had the same probe geometry and inlet diameter.
Bundled sampli nsfer the engine exhaust from the probes to the instruments located in a trailer (called trailer 1) approximately 23 m from the probe rake. The bundled line from the probe tips to the valve box (not shown in Fig. 1(a)) was maintained at 150°C and the line temperature from the valve box to the trailer 1 was maintained at 75°C to prevent condensation of water vapor and condensable organics within the sampling lines. Continuous particle measurements in trailer 1 included a TSI, Inc. Scanning Mobility Particle Sizer (SMPS ® ), a condensation particle counter, a smoke sampler, and a Taper-Element Oscillation Microbalance (TEOM). One of the sampling lines from the rake ( Fig. 1(b)) was connected to the box labeled as FPS in Fig. 1(a), where a prototype fastscan particle sizer (FPS) was located. The sampling distance between the probe tip and the fastscan inlet was approximately 0.9 m. The FPS data will be discussed in a separated paper (Mahurin et al., 2008) and not used in this paper.
A second rake containing two probes was

Fuel Analysis
The JP-8 and FT fuels used in the engine tests were sampled and analyzed for JP-8 specification properties. The fuel-analysis results are shown in Table 1. The FT fuel is composed solely of iso-(82%) and normal (18%) paraffins with a distillation range similar to a typical JP-8. Details on the physical and chemical aspects of this fuel have been previously reported (Corporan et al., 2007a;2007b;DeWitt et al., 2007)

Particle Size Distributions
The shape of size distributions of particle in T700 and T701C engine emissions are similar to those displayed in Fig. 3. As an example, the averaged particle size distributions observed at the EEP for the T700 engine operated with JP-8 fuel for the three power settings are displayed in Fig. 3(a), which shows the geometric mean diameter (GMD) was generally less than 50nm for all three distributions. Also noted in Fig. 3(a) that the particle concentration at GMD and the GMD of a distribution both increased as the engine power increased. The GMD of the particles increased from 20nm at the idle power to 42nm at the max power. The variation in the number concentration of each distribution shown in Fig. 3(a) is approximately ± 10-15% of the average and independent of the engine power setting.
The averaged particle size distributions sampled by the environmental probe for the same three power settings are also displayed in Fig. 3(b) for comparison with those obtained at the EEP. The distributions in Fig. 3(b) were corrected for plume dilution using the ratio of CO 2 concentrations measured by the EEP probe to that measured by the environmental probe. The overall pattern of the two groups of particle size distribution shown in Figs. 3(a) and 3(b) are similar. In Fig. 3(b) from the environmental probe, the GMD and particle concentration increased again as the engine power increased as those found at the EEP. In contrast, the GMD value at the idle power condition as shown in Fig. 3(b) was increased by approximately 7nm from that shown in Fig.   3(a). The growth defined by GMD increase was found to be minimal, if any, for the two higher engine-power settings. There was 2nm growth in the GMD at the 75% max power condition and 1nm at the max power condition.
These growths were within the statistical fluctuations (± 5% of the GMD) embedded in the measurements.

Fuel Effect on the Formation of Nuclei
Particles engine running JP-8 fuel are displayed in Fig.   5. The sulfur and sulfate data were corrected for plume dilution using the same dilution ratio mentioned previously for particle size distribution. For testing with FT, the sulfur emissions were below the analytical detection limits and are thus not reported in this paper.
The three solid bars in Fig The left tail region of a measured particle size distribution displayed in Fig. 4 Fig. 4(c)), the curve did not show the upward pattern; instead, a small peak was found at 11nm. The source of this 11-nm peak was unclear because (1) there were no sulfur to form condensible sulfuric acid species and (2) most paraffinic hydrocarbons were burned completely at the max power. Without aromatic hydrocarbons in the FT fuel, much less soot was produced in the T701C-FT max power case (see Fig. 8(a)). So, it is plausible from the data displayed in Fig. 4(a), 4(b) and 4(c), the formation of the upward-tail section of an engine particle size distribution could be primarily attributed to the sulfur and aromatics in the JP-8 fuel.
Oxidation of SO 2 to the oxidized sulfur form like sulfate, bisulfate, or sulfite in the atmosphere is well-understood (e.g., Brasseur et al., 1999;Seinfeld and Pandis, 1998).
Current understanding indicates oxidation can take place in the gas or liquid phase. The gas phase takes place by addition of OH radical to SO 2 , with a rate coefficient of 9E-13 (cm 3 /molecule/s) at atmosphere. Sulfur trioxide (SO 3 ) then reacts with water molecules in the gas phase or gets uptake into droplets to form sulfuric acid. Sulfuric acid

Particulate Sulfur
The sulfur data for the emissions of T700  However, it is worth noting that sulfur conversion was indicated to be greater than that explained by our current knowledge of contrail physics and chemistry (DeWitt, 2003).
With the estimated residence time of 550 ms at the idle power to 180 ms at the max condition between the EEP and the environmental probe tip, the T700/T701C engine plumes cannot be considered as young (< 10 ms) plumes based on the previous findings (Frenzel and Arnold, 1994;Kärcher et al., 1995;Petzold and Schröder, 1998). On the other hand, the age of particles in all test cases in this study were not as old as those of the ambient particles (~hours to days), either.
There was little liquid water droplets present in the plume. Thus, one likely mechanism for fuel sulfur to be converted to sulfate in the plume was that the gas-phase sulfur dioxide oxidation. SO 2 was formed in the engine combustor, then reacts with •OH or singlet •O radicals to form SO 3 species in the plume which subsequently reacts with water molecules in the air or on the surface soot particles to form H 2 SO 4 . Alternatively, SO 3 could react with water molecules in the vapor phase to form sulfuric acid that collide and combine with soot particles in the turbulent plume flow.
Using the T700-JP8 data, two plots were made to Fig. 6 show (1)  The S(IV)/sulfur ratio at the three engine power settings is reasonably close to the value of one, which again indicates that the conversion of sulfur to sulfate in the T700-JP8 plume was independent of the engine power setting. The sulfate to EC ratio decreased from 0.8 to 0.2 as the engine power setting was increased from the idle to the max, supporting the reaction mechanism of sulfur conversion (Petzold and Schroder, 1998;Kärcher and Yu., 2009).

Particulate Carbon
Using a PM-1 sampler similar to that used to sample particle sulfur and sulfate, particulate carbon was measured on a separate set of filter samples from those for sulfur taken at the environmental probe downstream from the EEP. The particulate carbon data were corrected for plume dilution using the same dilution ratio mentioned previously for particle size distribution and sulfur/sulfate.
The total particulate carbon emission increased as the engine power setting increased as shown in Fig. 7 Table 1). The absence of aromatics results in significant reduction of the fraction of EC in the soot (Corporan et al., 2007).
Operational breakdown of the total particulate carbon into elemental and organic carbon (i.e., EC and OC) leads to investigate the formation of soot and non-soot fractions in the carbon emission. In general the higher engine power leads to higher EC emission that is consistent across different engine and fuel.
The relationship between EC and the engine power setting is similar to that found for total particulate carbon shown in Fig. 7. The EC was negligible when FT fuel was used in the T701C engine operated at the idle and 75% max power settings. Even comparing the three bars of the T701-FT with those of the T700-JP8 and T701C-JP8 the reduction of elemental carbon (EC) mass concentration shown in Fig.   8(a) was substantial, some 50%-85% by average. Also noticed in Fig. 8(a) is the dramatic increase in the EC emission from the 75% max power to the max power setting.
The emission of organic carbon as shown in Fig. 8(b) was more complicated than those of the EC in Fig. 8(a). In general the OC emission consistently increases as the engine power increases for all three engine-fuel cases The reduction in the OC emission at the idle power setting by the FT fuel is observed in Fig.   8(b), but not for the other two higher power settings. In the 75% max power setting, the emissions of T701C engines (with JP-8 and FT) are lower than that of the T700-JP8, but both OC emissions of T701C engines are Ratio of OC to EC statistically identical. In the max power setting, the emissions of two T701C engines are however higher than that of the T700 engine.
Again, two OC emissions by the two T701C engines at the max power conditions are statistically identical.
The ratio of OC to EC has been used to interpret data on secondary aerosol formation and to trace/distinguish emission sources (e.g., He et al., 2004;Chu, 2005;Jaffrezo et al., 2005;Aymoz et al., 2006). For instance, He et al. (2005) found that average OC/EC ratio was 2.4, 4.2, 5.0 and 6.6 for Beijing, Gwangju, Kyoto and Ulaan-Battor, respectively. Jaffrezo reported the size-resolved OC/EC ratio and found most particles smaller than 50 nm have OC/EC ratios greater than 10. In these works, the ratio was considered a useful indicator for distinguishing the primary organic carbon from the secondary one. To our knowledge, there is no such OC/EC ratio for aircraft emission as a function of engine power. Fig. 9 shows the relationship between OC/EC and the engine power setting for the engine and fuel. The trend shows that the ratio decreased as the engine power increased for a given engine-fuel combination. Since OC measured in our study was mostly in the near field of aircraft emission, we have doubt about significant formation of secondary organic compounds as compared to the primary OC emitted by the engine. Thus, primary OC tends to increase as the engine power increased as shown in Fig. 8(b).
The increase of primary OC slows down as the power increased from the idle to the 75% max in Fig. 8(b), as compared to those of EC shown in Fig. 8(a). The OC/EC ratio is therefore expected to decrease as the engine power increased. This is exactly shown in Fig.   9 for T700-JP8 and T701C-JP8. It is unknown at the present on the cause of the large bar associated with the T701C-FT at 75% max power. Without this outlier, the OC/EC ratio for T701C-FT at the idle power is larger than that at the max power, which trend is consistent with the rest of the results of T700-JP8 and T701C-JP8. Also since the EC value for T701C-FT at 75% was small (see Fig. 8 (a)); it is possible that dividing a small number would result in a large outlying OC/EC ratio as shown in Fig. 9.