Performance Characteristics of a Hot-Wall Furnace Fired With Coal Water Slurry (CWS) Using Gas/Air Atomization

Page 20

17 ~) would become progressively more important as the gas proportion increases and the accuracy of the required pressure correction for the metering gas rotameter likewise becomes more significant. The low data points would be corrected to the trend line if the thermal input was high by 50-150 MJ/hr. If the trend line is taken as the best estimate of the wall losses, then here also it would seem to imply that the closure is good and mostly lying in about the same range as for the oil data. 4.4 Firing Curves. These are illustrated in Figs. 7 and 8; Fig. 8 includes the data obtained with the CB-125 burner. The curves are described by the firing equation jj^ ^ (j n^ r H = H ° + H /a° (1 - H / H m) (D f f s s s where H ° is the idle heat at zero output (limit wall loss + idle-point POC loss); H m is the theoretical maximum output; and a is the dimensionless s intrinsic efficiency (13). The coefficients in the equation for drawing the curves were obtained by the usual procedure (13). Their values are listed in Table 4. Both curve sets (Figs. 7 and 8) show the clear effects of excess air. As before, the oil data (Fig. 8) show clear trends with relatively little scatter and with good agreement with the data obtained using the CB-125 burner. About two-thirds of the slurry/gas data (Fig. 7) are also reasonably coherent; again, the data points that are outside the majority trends are the earlier easurements, at both high gas fraction and high excess air. In this instance, however, the anomalous points will not be corrected to the firing curves by the gas flow correction suggested for the wall loss data. An increase of 50-150 MJ/hr is far too small to account for the difference. A more probable explanation may turn on the fact of high gas fraction as discussed later. m