CR375 ASPEN+ Assignment #3

For your next assignment you are to use the ASPEN+ simulator for design of distillation columns. You are to again use the Soave-Redlich-Kwong equation of state for property estimation. As stated previously, the rigorous distillation program, RADFRAC, is a simulation program, which means that you have to specify the total number of stages (including condenser and reboiler) and the feed stage location. For a given distillate flow rate, L/D ratio, and feed properties, the program calculates the overhead and bottoms compositions and component flow rates. Thus, if you are doing design and you want to determine the number of stages required for a given specification you must run multiple cases. As already mentioned, shortcut methods can be used to estimate the number of stages and feed stage location for a given separation. Then the results from these methods can be used as a starting guess in the rigorous distillation program. This is the approach we will use.

You are to initially design a benzene-toluene column for the base case in the MCCABE project. The base case was x= 0.5, F = 100, q = 0.4, L/D= 3.0, for a fractional recovery of 0.96 of benzene in the distillate, and 0.96 of toluene in the bottoms. You are to start off by running the shortcut program in ASPEN+, DSTWU, which uses the Winn-Underwood method. You are to take the results obtained for the total number of stages and feed stage location and use them as input to RADFRAC. We will run RADFRAC with a 0.96 fractional recovery of benzene in the overhead. The optimum feed location as determined by DSTWU may not in fact be the optimum for RADFRAC. In order to check for the optimum, we will use the SENSITIVITY command and obtain results for the reflux ratio. The feed stage, which results in the lowest reflux ratio, will be the optimum. This can best be noted from a generated plot. Again since DSTWU is not exact, the resulting recovery of toluene in the bottoms may not be close enough to the spec (L/D=3.0). We will need to increment the number of stages by one and rerun RADFRAC. This operation will be repeated until the spec. is met as close as possible. This will be your final result.

In addition to this base case you are to also design a column for the following three component separation: F= 100, q=O.4, 0.35 mole fraction benzene in the feed, 0.35 mole fraction toluene and 0.30 mole fraction m-xylene. You are to again recover 0.96 of the benzene in the overhead and 0.96 toluene in the bottoms, with L/D=3.0, using the optimum feed stage location. This will require two RADFRAC design specs. In addition, you are to determine the column diameter required for both the upper and lower portions of the two distillation columns designed above. In both cases you are to use a flooding factor of 0.75 and a downcomer area of 0.10, and default settings for the remaining variables. The binary column is to use sieve trays while the ternary column is to be designed using Glitsch Ballast trays.

Your report should include the final design results as well as the results from short cut distillation for each design. Also include plots of composition versus stage number for your ternary column. A very brief discussion should be included, including explanation of your plots, especially for the ternary composition plot.