1. Broadened Specification") in the late 70s and early 80s in various operational gas turbine engines found that physical (rather than chemical) properties of the fuel were the dominant factors in most aspects of gas turbine engine performance. Only for smoke (soot) emissions and combustor liner heating (through radiation from soot) were chemical properties important, and they seemed to be best correlated against overall fuel H/C ratio or hydrogen content13-15as shown in Fig. 2. These results were relatively independent of aromatic concentration or nature (e.g., single or fused rings). Others disagreed, typically citing naphthalene content as a secondary factor that could not be
2. neglected16-18. Combustors with very fuel rich primary zones seem to be most sensitive to fuel composition effects on soot, while leaner-operating engines show less effect15. None of these studies, however, considered the impact on particle size, which has received increased attentionrecently. direct liquefaction process20. Itisnotedthatanyhydrocarbonsource(coal,shaleoil,biomass,etc.)canbeused as the feedstock for the F-T process. An example F-T fuel is "S-8", a synthetic jet fuel blended with JP-8 and flight tested in a B-52 in September and December, 2006. A surrogate for the F-T fuel would likely have an isoparaffin as a major component, while a surrogate for the naphthenic fuel would be rich in decalin and alkyl cyclohexanes. A broad chemical characterization of these fuels is provided in Table 3 (see Ref. 9). A detailed analysis of the F-T Jet A-1 (or S-8) has been performed21using GC-MS that demonstrates that a large number oftheparaffinsare alkaneswith oneor two attached methyl groups. Table3. Alternativefuel composition9.
3. Experimental and modeling results for a wide variety ofjet fuel/kerosene surrogates have been described in the literature (earlier reviews are available23,24). One reason for the large variation in surrogate composition is the wide variety of jet fuel applications, and the wide variation in composition sensitivity of these applications. In contrast to gasoline engines, where heptane/iso-octane surrogates have been employed since the 1920s for knocking/octane number estimation, surrogates have only recently been employed for gas turbines. Wood et al.24published work in the late 1980s describing the performance of a number of JP-4 and JP-5 surrogates. These surrogates were burned in a swirl-stabilized laboratory combustor, where the intent was to match fuel boiling range and composition. This required surrogates with more than 10 components (12 for the JP-5 surrogate), many of which were quite expensive. The surrogates matched the burning behavior of the fullboiling-range fuels, aside from soot formation. A similarly complex 12-component surrogate was developed by Schulz during the same period for liquid phase oxidation studies25. Schulz'conclusionwasthatsurrogateswere not useful for these liquid-phase studies, where rate of oxidation and deposition are controlled by trace species rather than bulk fuel composition. For liquid phase properties across a wide temperature range, it has been found that using dodecane in a physical code such as Supertrapp to calculate fuel properties produces results comparable to the sparse experimental data26. Modeling of multi-phase behavior such as vaporization would require amorecomplex surrogate.
4. Dagaut et al.11modeled kerosene combustion in low-temperature jet-stirred reactors using n-decane as a reference hydrocarbon while neglecting the aromatic components and captured major species profiles adequately. Guerèt et al.28modeled kerosene oxidation via quasi-global models for n-decane, and an aromatic component (n-propylcyclohexane, trimethylbenzene, xylene, toluene, or benzene). Concentration profiles of molecular species in the flow reactor were similar for the surrogate and kerosene; however, the need for further refinement of the aromatic models was recognized. A number of studies compared various aromatic compounds in surrogates, generally concluding that alkyl-substituted aromatics were the best aromatic components47-52. 5BA number of recent investigations5,30of surrogates have been based on the sixcomponent "Violi" surrogates. As shown in Table 4, these surrogates are intermediate in complexity between the 12- component Wood et al surrogate (labeled "UCI" in Table 4) and the simpler two-component aliphatic/aromatic surrogates of (for example) Guerèt et al.28. Similar in complexity are the extensive investigations at Drexel involving a large number of variations of a 6-component surrogate53. The Drexel surrogates are notable for including a component from all of the major compound classes - n-paraffins, iso-paraffins, oneand two-ring cycloparaffins, and one- and two-ring aromatics. Several other recent surrogate investigations, such as the ones labeled CSE54and REI32in Table 4, have taken the "smaller is better" approach tosurrogates.