8000 12000 16000 f (min)

Figure 16.14 Relay-feedback test results for CS7 pairings.

of the purities settles down to a new steady state above its setpoint, the other one goes to a new steady state that is too far below its setpoint.

This problem occurs because the two temperature controllers in the CS7 structure give different initial responses as a result of their opposite actions. An increase in vapor boilup produces an increase in the temperature on both control trays (tray 16 and tray 26). The tray 16 temperature controller has direct action, so it increases the F0A feed flowrate. However, the tray 26 temperature controller has reverse action, so it decreases the F0B feed flowrate. Not enough reactant B is added in a timely enough manner to prevent light reactant A from going to the upper section of the column. Once this breakthrough occurs, the column moves to a different operating condition where the impurity of A in the distillate stream increases and product purities are lower. Thus, we can conclude that control structure CS7 is not a good choice for controlling this column/side reactor process.

Control Structure CS5. As mentioned in the previous section, it is desirable to use inferential temperature measurements instead of direct composition measurements whenever possible. However, the results show that control structure CS7 is not an effective control structure, at least with the optimum steady-state design studied here. It appears that some direct composition information about the reactant inventory inside the system is required for a more effective control system because the column is designed for neat

TABLE 16.3 Tuning Parameters for Control Structure CS7



Was this article helpful?

0 0

Post a comment