OCR Text |
Show data from 13C NMR analyses is required for every sample with CPo. Whereas FLASHCHAIN entails no specialized laboratory tests whatsoever, FG-DVC has grown into a framework for extensive laboratory work involving state-of-the-art experimental facilities (Solomon, 1993). Although these models' underlying mechanisms share much in common, there are also tangible performance aspects to consider. To date, FLASHCHAIN has been used to predict the devolatilization behavior of more than 200 different coal from every geographical region worldwide. No other model comes remotely close to this level of performance, simply because only FLASHCHAIN simulations can be performed without specialized and expensive laboratory tests. Since they require specialized input, neither CPO nor FG-DVC can be used to make predictions for the substantial backlog of experimental studies in the literatw'e, because the information needed to defme their input parameters was never acquired. Following several years of development and validation, FLASHCHAIN is now ready to provide important coal-specific information on rapid-heating volatiles yields, the partitioning of coal-nitrogen among volatiles and char, detailed product distributions, and char characteristics. Yet the only sample-specific information needed to simulate these coal quality impacts is the proximate and ultimate analyses of the coal. And full simulations require only a few seconds on modem personal microcomputers. REPRESENTATIVE PERFORMANCE EVALUATIONS As illustrated in Figure 1 and other publications (Niksa 1994; Niksa 1995), FLASHCHAIN's high degree of evaluation is the basis for an unparalleled ability to capture the distinctive devolatilization behavior of individual coal samples. Figure 1 presents comparisons among the predicted and measured ultimate values of weight loss and tar yield for atmospheric devolatilization at 3000 K/s to 1037 K, followed by a 4 s hold period. The predicted weight loss is within 4 wt. % of the measured values in 14 of the 17 cases. The predictions also track the perturbations from a smooth, monotonic trend that is evident in the data. Similarly, predicted tar yields are within 4 wl % of the observed values in 14 of the 17 cases, and also depict the rather erratic relation with carbon content that is observed, as well as the broad maximum for hv bit coals. The only sample-specific inputs for these simulations are the reported ultimate analyses. In Figure 2, the FLASHCHAIN predictions display a finn basis for extrapolations to the very rapid heating conditions in combustors and gasifiers. The predicted weight loss and tar yields from both coals are within experimental uncertainties throughout. For the TIL #6, tar yields double over this range of heating rates, incrementally increasing ultimate weight loss by almost the same amount. But there is virtually no enhancement with the low volatility coal. FLASHCHAIN captures this important coal-type effect. 3 |