Ternary DRD table lookup procedure:
1. Classify a system by writing down each position number in ascending order of boiling points.
• A position number is not written down if there is no azeotrope at that position.
• The resulting sequence of numbers is known as the temperature profile.
• Each temperature profile will have a minimum of three numbers and a maximum of seven numbers. 44
• List multiple temperature profiles when you have incomplete azeotropic data.
• All seven position numbers are shown on the diagram.
2. Using the table, look up the temperature profile(s) to find the corresponding DRD #. * Table 13-16 and Fig. 13-59 developed by Eric T. Peterson, Eastman Chemical Co.
4) and tray 18. The temperature remains constant at about 93°C throughout the pinch. Product specifications are met.
When the feed composition becomes enriched in water, as with Case B, the column profile changes drastically (Fig. 13-62b). At the same reflux and boil-up, the column no longer meets specifications. The MIPK product is lean in MIPK and too rich in water. The profile now tracks generally up the left side of Region II. Note also the dramatic change in the temperature profile. A pinched zone still exists
FIG. 13-61 MEK-MIPK-water system. (a) Approximate bow-tie reachable compositions by simple distillation.
between trays 4 and 18, but the tray temperature in the zone has dropped to 80°C (from 93°C). Most of the trays are required to move through the vicinity of the saddle. Typically, pinches (if they exist) occur close to saddles and nodes.
In Case C (Fig. 13-62c), increasing the boil-up ratio to 6 brings the MIPK product back within specifications, but the production rate and recovery have dropped off. In addition, the profile has switched back to the right side of the region; the temperatures on trays in the pinched zone (trays 4-18) are back to 93°C. Such a drastic fluctuation in tray temperature with a relatively minor adjustment of the manipulated variable (boil-up in this case), can make control difficult. This is especially true if the control strategy involves maintaining a constant temperature on one of the trays between tray 4 and 18. If a tray is selected that exhibits wide temperature swings, the control system may have a difficult time compensating for disturbances. Such columns are also often difficult to model with a process simulator. Design algorithms often rely on perturbation of a variable (such as reflux or reboil) while checking for convergence of column heat and material balances. In situations where the column profile is altered drastically by minor changes in the perturbed variable, the simulator may be close to a feasible solution, but successive iterations may appear to be very far apart. The convergence routine may continue to oscillate between column profiles and never reach a solution. Likewise, when an attempt is made to design a column to obtain product compositions in different distillation regions, the simulation will never converge.
Was this article helpful?