| OCR Text |
Show This approach has already been evaluated for tangentially fired furnaces by estimating the emissions for situations that were not pan of the database used to formulate the regressions. It can predict NOx emissions for coal ranks from subbituminous through Iv bituminous within 10 to 15 ppm of observed values. We also expect that the same basic approach would work as well in correlating coal quality impacts on plan area heat release rates, near-burner radiation loads, heat rates, furnace exhaust temperatures, and steam-side temperatures. SUMMARY The crucial distinction between FLASHCHAIN and all other models for coal devolatilization is that any test results that contain the coals' proximate and ultimate analyses and the operating conditions (time-temperature history and pressure) can be simulated. We have already simulated most of the devolatilization data reported in the open literature from the past 15 years and regularly scan journals to fmd additional cases. Other models cannot be validated this way because their highly specialized input data was never acquired when the tests were performed. Evaluations of the predicted ultimate yields have been reported for more than 70 coals, including almost 20 cases that illustrate the reaction dynamics. Coal samples spanning the entire .. rank spectrum and all geographic regions worldwide are included. Heating rates from 0.5 to 15,000 K/s.have been considered, along \\ith ranges of temperature that are broad enough to .illustrate complete primary devolatilization, from its onset through tar production to the end of second-stage release of CO and small amounts of Cf4. The cases on pressure effects represent pressures from vacuum to 7 MPa. Predicted coal-N release has been reported for almost 40 different coals, and resolutions of the oxygen species have been reported for almost 30 different coal samples. In addition to these cases that have been reported in the open literature, we have already conducted FLASHCHAIN simulations for our clients that take the total number of coal samples to well over 200. We now routinely perfonn blind predictions, in which our clients provide only the proximate and ultimate analyses of their coals and provide sufficient information to define the timetemperature history during their tests. Once the predictions are validated, the model is then used to assess coal quality impacts on boiler performance, as a vinual coal laboratory . This model is ready for various applications, including schemes that can define the devolatilization reactivities of individual coals for CPO simulations, as well as more engineeringoriented regressions of coal quality impacts. 11 |
