As shown in Figure 8.19b, the distillate and bottoms product purities return to values close to their specification levels. However, there are very large dynamic drops in product purities. An increase in feed flowrate causes large drops in distillate purities of the distillate in column C1 and the bottoms in column C2. The reverse effect occurs for a decrease in feed flowrate.
We would expect that the deviations in the C2 bottoms purity could be improved by using a steam to feed ratio in this column. However, this might cause a more rapid increase in heat input in column C1, which could make the deviations in the distillate purity of column C1 even worse. This control structure is shown in Figure 8.20.
Setting up the ratio requires converting the two signals to metric units. The steady-state value of F2 is 0.49137 kmol/s (1768.9 kmol/h), and this is Inputl to the multiplier. The steady-state value of reboiler heat input is 21.80 MW (78.48 GJ/h), so this should be the output of the multiplier. The value of Input2 to the multiplier is calculated to be 0.04437. Change the initial value of the temperature controller output to 0.04437, and specify the output range of the temperature controller to be 0-0.1.
The TC2 controller is retuned with TC1 on manual, giving KC = 1.06 and ti = 9.2 min. Then TC1 is retuned with TC2 on automatic, giving KC = 1.16 and ti = 14.5 min, which is quite different from the previous values of KC = 0.535 and ti = 33 min. These results indicate that much tighter control is possible with the steam to feed ratio.
The performance of this control structure for 20% step changes in feed flowrate are given in Figures 8.21a and 8.21b. The improvement is very striking. The peak transient deviations in product purities are reduced by an order of magnitude. However, the performance of the system for feed composition disturbances is not very good.
Was this article helpful?