An Oxy-hydrocarbon Model of Fossil Fuels

This paper asserts a new method of analyzing fossil fuels, useful for sorting coals into well defined categories and for the identification of outlying ultimate analysis data. It describes a series of techniques starting with a new multi-variant approach for describing the lower Ranks of coal, progressing to a classical, but modified, single-variant approach for the volatile and high energy Ranks. In addition, for a few special cases, multiple low and high Ranks are also well described by the multi-variant approach. As useful as these techniques are for analyzing fuel chemistry in the laboratory arena, this work was initiated in support of Exergetic Systems’ Input/Loss Method. At commercial coal-fired power plants, Input/Loss allows the determination of fuel chemistry based on effluents. The methods presented allow equations to be developed independent of actual combustion stoichiometrics, for improved Input/Loss accuracy in determining “real fuel in real time”. PAPER72_Rev30.WPD Introduction Historically coal chemistry associated with low and non-volatile Ranks has defied descriptive analytics. For these fuels, there has been no technique allowing a predictive capability for fuel chemistry. For example, as seen in Figure 1, Irish Peat has a considerable variability (hydrogen versus carbon) even though the fuel was laid down at broadly the same time across the island. Similarly, a coal mined from the Decker region of the Powder River Basin (PRB) in North America, having an average of 70% dry fuel carbon by weight, cannot reasonably be compared to a coal mined from the Buckskin region of the same PRB having an average of 64% dry carbon. These two PRB coals differ in average dry calorific values by 5.6% (>1535 ÄkJ/kg or >660 ÄBtu/lbm). Figure 2 presents graphical results from several hundred ultimate analyses of coals from 17 PRB regions. As a further example, there has been little success in explaining the vagaries found in fuel chemistry associated with high energy coals, of the same Rank, but obtained from different continents, herein termed “High Seas” coals. High Seas coals are internationally traded high energy, low water, bituminous coals (similar to hvAb & hvBb), which are often purchased literally when cargo ships are underway. Although High Seas coals typically have uniformity of calorific values, Figure 3 illustrates a significant variability in fuel chemistry. Note that Figures 1 and 2 show poor Coefficients of Determination (R ), 2 62% for Irish Peat and 77% for PRB coals. The R value 2 for the High Seas coals of Figure 3 is better at 91%. However, experience with installing Input/Loss indicated fuels of moderate energy, down to the lignites, the highest R for these classical plots rarely exceeded 85% [1]. 2 Davidson [2], in a most comprehensive literature review of the chemistry of coal, notes that several authorities report there appears “to be no accurate method for the determination of molecular weights of coals”. Although the authors agree that seeking an absolute molecular weight for a particular coal Rank is a fool’s errand, we believe that a consistent normalized molecular weight, if consistently demonstrated, offers the touch-stone for substantiating descriptive analytics. It is true that the examination of molecular patterns of whole Ranks reveals an incredible complexity of interstitial layering, aromatic and aliphatic structures, which suggests that coal is a heterogeneous substance at both the macroscopic and microscopic levels. Nevertheless, we argue that at least at the macroscopic level, a consistently computed molecular