Novel rates of OH induced sulfur oxidation. Implications to the plume chemistry of jet aircraft

Holger Somnitz, Gö tz Georg Gleitsmann, Reinhard Zellner
2005 Meteorologische Zeitschrift  
Novel rate coefficients for the most important and rate controlling sulfur oxidation reaction, OH + SO 2 → HSO 3 , over an extended range of pressure and temperature have been derived from ab initio quantum chemical/RRKM dynamical calculations. From these calculations the rate of oxidation of S(IV) to S(VI) under typical conditions of a jet aircraft plume is predicted to be considerably slower than previously accepted on the basis of interpolations of experimental data (i.e. Tremmel and
more » ... , 1999). These kinetic results have been incorporated into a chemicaldynamical code of the jet regime of a B-747 airliner (BOAT code) to predict sulfur conversion efficiencies in this regime of less than 1%. It is concluded therefore that overall conversion efficiencies in the order of several % are caused by corresponding conversions in the combustor and/or turbine of the jet engine and not in the plume. concluded from our work, that the well-known conversion ratio of S(IV) to S(VI) of several % as confirmed by a number of airborn experiments, can only be reproduced assuming sufficient formation of SO 3 or H 2 SO 4 already inside the engine and/or the turbine. The plume effect on this ratio is less important. ACKNOWLEDGEMENT Financial support of this work by the EU-project PARTEMIS (GRD-1999-10819) is greatfully acknowledged. REFERENCES : In situ observations in aircraft exhaust plumes in the lower stratosphere at midlatitudes, J. Geophys. Res. 100, 3065-3074. Fulle, D., H.F. Hamann, and H. Hippler, 1999: The pressure and temperature dependence of the recombination reaction HO . + SO 2 + M → HOSO 2 . Franke, 1998: Observation and model calculations of jet aircraft exhaust products at cruise altitude and inferred initial OH emissions, J. Geophys. Res. 103, 10803-10816. Tremmel, H.G: and U. Schumann, 1999: Model simulations of fuel sulfur conversion efficiencies in an aircraft engine: Dependence on reaction rate constants and initial species mixing ratios. Aerosp. Sci. Technol. 3, 417-430. ABSTRACT: The DLR -Soot Generator was used as a variable and well defined soot source. The particle mean diameter of the log normal size distribution can easily be shifted between 6nm and 250nm. This soot loaded exhaust gas is sucked through a quartz fiber filter via a computer controlled gas sampler. The soot particles are trapped on a quartz fiber filter. This special quartz filter has a sampling efficiency better than 99,9% for particles between 6nm and 250nm. The carbon load on the quartz filter is burned in an oxygen atmosphere. The resulting carbon dioxide concentration is measured with a Fourier Transform IR spectrometer (FTIR). If the gas sampling volume, the gas cell volume and the carbon dioxide concentration is known, a soot mass fraction can easily be calculated. The corresponding size distribution, number concentration and volume concentration are measured by a Scanning Mobility Particle Sizer system (SMPS). Using the soot mass of the filter experiment and the soot volume of the SMPS measurements, a soot density was calculated. This soot density is based on the mobility diameter of the fractal soot particles. The quotient of graphite density and calculated soot density gives the fractal character of the soot. Due to the excellent trapping efficiency of the fiber filter, it is possible to get the absolute soot mass fraction of the exhaust gas. Before taking filter soot samples, the quartz fiber filters are heated to 800 °C, in order to burn all organic material on the filter. Then, with a computer controlled gas sampler, a known volume of aerosol was sucked through the filter. In a reaction tube, the volatile organic fraction of the soot loaded filter was burned at 480 K in an oxygen atmosphere. After that procedure the reaction tube was cooled down to room temperature and evacuated again. The remaining black carbon on the filter was burned in a oxygen atmosphere at 800 °C. After cooling down, the resulting CO 2 concentration was measured with a FT-IR spectrometer. The FT-IR system was calibrated with test gases of known CO 2 content. With known exhaust gas sampling volume and known FT-IR gas cell volume, the black carbon mass is calculated via the CO 2 Signal. The SMPS instrument classifies particles by the so called mobility diameter. Using a sphere model gives the aerosol volume. Our system can work in two measuring ranges, the low flow mode and the high flow mode. If the mean diameter of the measured size distribution is in the mid range of the measuring range, high flow and low flow measurements should give the same volume. SMPS measurements in low flow mode as well as in high flow mode have been done before and after filter sampling. The mean SMPS soot volume in high flow mode was 3.03E+11 [ nm 3 / cm 3 ]. The mean SMPS soot volume in low flow mode was 3.78E+11 [ nm 3 / cm 3 ]. The gives a soot density of 0.74 g/cm 3 (high flow) and 0.59 g/cm 3 (low flow). The mean soot density of a ethylene flame at 1 bar is d MD ~ 0.66 g/cm 3 . (We name the density d MD ,because it is based on the mobility diameter.) Assuming the primary soot particles consist of graphite similar structures, the density of the primary particles is ~2.2 g/cm 3 like graphite. Therefore a "soot fractal character" can be calculated by dividing the graphite density through the soot density. For our ethylene soot the determined "fractal character" is 2.2 g/cm 3 / 0.66 g/cm 3 = 3.33 =fc MD (again based on mobility diameter). CONCLUSION A method to determine soot mass, soot density and soot fractal character is shown. Flames under different conditions, or flames of different fuels will have different soot density and soot fractal character. Whenever soot densities are published, it is absolutely necessary to give the method of determine the particle diameter or volume. Because of the fractal character of the soot agglomerates, only soot densities determined with the same method can be compared. ABSTRACT: EXCAVATE was conducted at NASA Langley Research Center in Hampton Virginia during January, 2002, and obtained detailed measurements of emissions from Langley's T-38 and B-757 aircraft to determine ion densities; the fraction of fuel S converted from S(IV) to S(VI); the concentration and speciation of volatile aerosols and black carbon; and gas-phase concentrations of hydrocarbons in the exhaust of a typical commercial airliner, all as a functions of engine power, fuel composition, and plume age. Our observations indicate that chemiion densities were very high in the exhaust plumes, consistent with values that are presently being used in microphysical models of aerosol formation in exhaust plumes. Both aircraft were found to emit high concentrations of organic aerosols, particularly at low power settings and to produce black carbon concentrations that increased with engine power. Total particle emission indices were typically a factor of 10 higher at 35 meters than at 1 meter behind the engines due to formation of new particles. The concentration of sulfate aerosol was directly dependent upon the fuel sulfur level and increased considerably as sampling took place progressively further downstream from the exhaust plane. For the B757, organic aerosol emissions were very high at engine start and took several minutes to reach much lower, equilibrium values after changes in engine power. This was particularly notable when the engines were reduced from high to low power as might occur during taxi and landing. ABSTRACT: A microphysical model has been used in order to study volatile particles formation in the sampling system of the PartEmis European experiment. The fraction ε of the fuel sulphur S(IV) converted into S(VI) has been indirectly deduced from comparisons between model results and measurements. ε has been found to be in the range 2.5 % to 6 %, depending on the combustor settings and on the microphysical approach used. Different processes have been investigated, comprising soot particles activation and possible growth. Growth factors of monodisperse particles transported in the line and then exposed to high relative humidity (95 %) have been calculated and compared with experimental results. Results show interesting trends of increasing growth factors with decreasing size. ABSTRACT: In the present study, we investigate the chemistry of expanding aircraft exhaust plumes for a range of conditions (latitude, altitude, season, plume expansion rate, time of day of emissions, aircraft type, composition of emissions and background air mass). The effect of plume chemistry on 2-D global model calculations of the impact of subsonic aircraft emissions on ozone is discussed.
doi:10.1127/0941-2948/2005/0060 fatcat:6p2j4dyamfc5tm7ehe2tn66y2u