0.50 mol% benzene and 0.23 mol% water. Getting the simulation to converge to this steady state is quite difficult.
The calculated compositions of the reflux and recycle are compared to the guessed values. The reflux composition is quite close: 84.4/14.0/1.6 mol% B/E/W calculated versus 84.4/ 14.1/1.5 mol% assumed. The recycle composition is somewhat different: 4.7/41.6/ 53.7 mol% calculated versus 7.2/47.1/45.7 mol% assumed. Changing these compositions to the calculated values and rerunning the program give a new Bj composition of 0.926/ 98.92/0.157 mol% B/E/W. The calculated flowrate of the distillate D2 under these conditions is 0.0704 kmol/s versus the 0.06 kmol/s assumed.
Now that we have some reasonable guesses for the values of the recycle streams, the Design Spec/Vary capability can be used to drive the compositions of the two product streams to their desired values. The key feature in the first column is to keep enough benzene in the column to entrain out the water so that the bottoms is high-purity ethanol. On the other hand, if too much benzene reflux is fed to the column, it will go out the bottom and drive the bottoms off specification. A Design Spec/Vary is set up to maintain the benzene composition of the bottoms at 0.5 mol% by manipulating the REFLUX stream, which is consider a Feed rate on the list of choices given in the Vary, Specifications, Adjusted variable, and Type.
The initial guessed value of the reflux ratio in the second column was 2. The bottoms purity was very high. The reflux ratio was reduced to RR ~ 0.2 without affecting the bottoms purity significantly. A second Design Spec/Vary is set to maintain the ethanol composition of the bottoms of the second column at 0.1 mol% by varying the bottoms flowrate B2.
Several runs are made in which the guessed compositions of the reflux and the recycle are compared with those of the calculated organic stream from the decanter and the distillate D3 from the second column. When these variables are fairly close, the recycle/D2 loop is closed. The procedure for doing this involves three steps:
1. Delete the stream labeled "D2CALC" on the lower left in Figure 5.22.
2. Click the stream labeled "REYCLE" and reconnect it to the valve labeled "V2."
3. Go down near the bottom of the list of item on the Data Browser window and click Convergence and then Tear. This opens the window shown in Figure 5.24, where the dropdown menu is used to select RECYCLE.
When the program is rerun, it converges to the values shown on the flowsheet given in Figure 5.25. Figure 5.26 gives the ternary diagram with the locations of all the streams marked.
In theory, the next and final step is to close the organic reflux loop. The stream labeled "ORGREF" is deleted, the stream "REFLUX" is connected to the mixer "M1," and "REFLUX" is defined as a TEAR stream. Unfortunately, this loop does not converge even though the initial values of the guessed and calculated values are very close in both composition and flowrate. An alternative way to converge this system using dynamic simulation will be discussed in Chapter 8 after we have discussed the details of dynamic simulation.
Was this article helpful?