Note that it takes about 3-4 h for the 95% boiling point loop to settle out. Note also that the temperature on stage 2, which is given in the bottom right graph in these figures, does not change much at the new steady-state conditions for these disturbances. This suggests that the use of temperature control instead of boiling point control might work pretty well.
To test this theory, the control structure is modified to manipulate reflux to hold stage 2 at 171.6°C, as shown in Figure 11.30. The TC2 temperature controller has a deadtime of 1 min and is tuned in the usual way, yielding tuning constants KC = 0.90 and tj = 5.3 min (with a temperature transmitter range of 150-250°C and a maximum reflux flowrate of 70,000 kg/h). Note that this reset time is much small than that of the 95% boiling point controller, so faster closedloop dynamics can be expected.
Figure 11.31 shows that this is indeed true. The responses with temperature control are the solid lines. The response with 95% boiling point control are the dashed lines. The disturbance is a swing in crude oils to less Oil-1 and more Oil-2. The process steadies out in about 90 min with temperature control. The naphtha boiling point is not held exactly at 191 °C, but ends up at about 189.3°C. The naphtha yield is somewhat smaller, with more bottoms.
Was this article helpful?