OCR Text |
Show summed for all the tubes this gave the useful output, H . 4.3 Wall Loss and Closure. Wall losses, H , have not been independently - w calculated as yet beyond a rough estimate based only on the roof temperatures (which are higher than average). A direct determination of the closure is therefore not yet possible. Instead, the wall loss (plus errors) was calculated by the differences: H = (Hr - H - H ). Figures 5 and 6 w f g 8 illustrate the variation of H (wall loss + errors) with output, Hc, for w b the slurry and oil experiments respectively. As can be seen, the oil data (Fig. 6) show a roughly linear rise with output as expected (13); in addition, the independent rough estimate of the wall loss, which is marked by an X on Fig. 6, (at 390 MJ/h) was a little lower than the mean line. This suggests that the closure on the heat balance is reasonably good, possibly in the range 95 to 105%. Additional data pointsj (Fig. 6) fired by the CB-125 Burner are on the trend line of the double-swirl burner (CB-125 nozzle). The scatter on the slurry plot (Fig. 5) is rather greater, but there is reason to believe that the trend line on the curve is reasonably representative of the best data. There are 5 or 6 points significantly below the trend line that are clearly too low by independent estimation on any basi The value obtained by independent estimation is about on the trend Lint . which is marked by an X on Fig. 5, and the trend line data are about 20 higher than the oil data in Fig. 6, roughly in accordance with the differences in roof temperature (used as a guide) for the two I ii Lng systems. The lowest points on the graph were also some of the firsl measurements made so there may be a factor of improved measurement technique for the other points; they are also high excess air and high gas proportion. 01 these last two factors, the second may be the more significant as the accuracy of the gas rate |