Appendix A. Aspen Plus Troubleshooting Guide for Separations

Although specific for Aspen Plus, many of these suggestions are helpful for other simulators. Although Aspen Plus is not perfect, probably 99% of problems with Aspen Plus are due to operator error. If Aspen Plus is too slow, close other programs, and close extra windows within Aspen Plus. If none of the correlations fit the equilibrium data, see Appendix B.

If something is not working, check the following:

1. Probably the most common error occurs when specifying streams. You must specify total flow and the composition of each component. The default for composition is molar flows. If the sum of the component molar flows is not equal to the total flow, Aspen Plus still runs the simulation but normalizes the component molar flows into mole fractions. Best practice: Always use mole or mass fractions to specify composition. Another common error is to input the wrong numbers or with the wrong units (e.g., using a feed temperature of 30.0 K when you mean 30.0°C. Separation modules do not work with a feed this cold.)

2. If you expect two liquid phases and one vapor phase, enable this condition. In Setup, for Valid phases choose Vapor-Liquid-Liquid from the menu. Do the same for Analysis. For Flash, use Flash3, not Flash2.

3. A common error in distillation is to connect a product to the vapor stream from the condenser and then in the configuration page for the column in the menu for Condenser select Total. Aspen Plus gives an error message for this one. You need to disconnect the product and connect it to the liquid distillate product of the distillation column. This and similar errors can be avoided by paying attention to the balloon that identifies the connection when you hover the cursor over the connection point before clicking.

4. Aspen may ask you for conditions in an intermediate or product stream, and you do not know or want to specify these. If this persists and Aspen does not let you run when you click the Next button, delete the block, add a new block, and reconnect it to the streams. An alternative is to delete the stream and then add a new stream. One of these procedures usually solves this problem.

5. Convergence problems:

a. Open up the block in the data browser. (You can always get to the data browser from the menu—go to data, and look at the bottom of the menu.) After opening up blocks and then the desired block, click Convergence. Increase the number of iterations from 25 to 75. This often helps but does not always solve the problem. Running the simulation again with no changes sometimes works. Since Aspen Plus requires a change to run the simulation again, increase the number of iterations by one and then run again. If close to convergence, this may work. An alternative is to reinitialize and run again.

b. Make sure that your numbers are compatible. For example, if the feed to a distillation column is 100.0, D cannot = 110.0. If F and D are both 100.0, then B = 0, which is also probably not what you want to do. Selection of Reflux rate and Boilup rate (not ratios) are compatible and should result in a successful run. Conditions can then be adjusted to obtain the desired purities.

c. If you have tried other solutions and they have not converged, change the convergence procedure. For absorbers and strippers go to Block→Setup→Configuration→ Convergence. Select Petroleum/Wide boiling from the menu. This is also necessary for distillation if the boiling range is very large (e.g., a feed containing both methane and n-octane).

d. For very nonlinear systems such as those forming two liquid phases (e.g., butanol and water) select Convergence as Strongly non-ideal liquid.

e. For recycle problems, convergence often requires a trick:

Recycle Example 1, Extractive distillation. First, solve without solvent recycle by using all fresh solvent. Then connect the recycle and reduce solvent in steps down to steady-state value. There must be a place for the excess solvent to leave the system.

Recycle Example 2, Two-pressure distillation system to break azeotrope: Initially set B in one column and a low D value in the other. Then, without reinitializing, increase D until it approaches the desired recycle value.

Recycle Example 3, Two-column binary separation with heterogeneous azeotrope. Set up the system with total condensers and modest reflux ratios (say 0.5 or 1.0). Run the system, and reduce L/D in both columns in steps without reinitializing to very low values that approximate all of the reflux coming from the decanter. Specify the value of B for one column but not the other (try boilup ratio in the stripping column).

6. In absorbers and strippers you need to draw RadFrac with a vapor distillate product. In the configuration page for the block pick condenser = none and reboiler = none in the menus. The operating specifications section (where you would normally pick L/D and D) should turn gray—no menu items will be available. If you try to change to no condenser and no reboiler after picking operating specifications, Aspen Plus may become confused and does not adjust (the menu items for L/D and D remain in place). To correct this problem, delete the block for the absorber/stripper, install a new block, and reconnect all lines. Now pick condenser = none and reboiler = none, and everything should work.

7. When you cannot obtain the desired purity:

a. In distillation, this is often due to inappropriate choice of D (or B). Do not select both! For high purities redo your external mass balances (accurately!), assuming complete separation, and use these values of D and B.

b. If you have columns in series a low-purity requirement in the first column can prevent the second column from achieving the desired purities. For example, suppose there is a LNK in the second column that is the distillate product in the first column. All of the LNK in the bottoms of the first column will automatically be in the distillate of the second column. If the purity requirement for the LK in the second column are tight, the purity requirement for LNK in the bottoms of the first column must also be tight.

c. In absorption or stripping you may need to make the equilibrium more favorable by changing the column pressure or by changing the temperature of input streams. An alternative is to use a purer solvent stream.

d. If mole fractions of impurity you are trying to obtain are too low, Aspen Plus may never reach these values if the convergence parameters are not set tight enough. A more stringent value for mole fraction (say E-06) will probably work.

8. Try Help in Aspen Plus.

Although often not too helpful, Help sometimes explains what the difficulty is.

9. Miscellaneous.

There are a number of other things that can happen. Although hard to decipher, the Aspen Plus error messages often give a hint of where to look for items to change. These errors may occur when you try to do something you have not done before (e.g., set a pressure drop on every stage). If you are trying to do something different and get an error message, stop doing what is different (e.g., return to constant pressure in the column), and see if Aspen Plus will run. If it will, figure out a way to approach what you want to do in small steps (e.g., use very small pressure drop on each stage). This approach may show you what the problem is.

10. As a last resort, log out of Aspen Plus and then log back in using a blank simulation. Rebuild your entire system. Do not use a saved file as it is probably corrupted. Do not do this step on automatic pilot. If you do, you are likely to repeat the error without thinking about it.