Troubleshooting Amine Unit Simulations

Lili Lyddon

June 2, 2008

If problems occur during execution of an amine sweetening unit simulation, you should always first look at the Warnings list and Message Log for clues as to what went wrong. If there is too little or too much acid gas absorption, check the following:

If too much acid gas is absorbed, be sure the Column Type (Process Data tab) is set to TSWEET Kinetics and not Equilibrium. Also be sure a Residence Time is being calculated (Stage Data tab, Hardware Grouping) and that it is reasonable (approximately 1 - 4 seconds). If Residence Time is not set or calculated, the results will be the same as if the Column Type were set to Equilibrium.
If there is not enough acid gas absorption, be sure the Property Package is set to Amine Sweetening or Electrolytic ELR. If the Property Package is set to an EOS (SRK or Peng-Robinson), very little acid gas pickup will occur.
If there is not enough acid gas pickup and the rich loading is too high, check the circulation rate as it may be too low.
Also check the lean loading. If the lean loading seems too high, you should check the reboiler duty or steam rate, which should correspond to about 0.8 to 1.2 lb steam per gallon of amine solution circulated (0.096 to 0.144 kg steam per liter of amine solution circulated).
If a trayed absorber is modeled using ideal stages, be sure the Real/Ideal Stage Ratio (Stage Data tab, Hardware Grouping) is set correctly. For a packed absorber, be sure the total column liquid residence time is maintained if the number of stages is changed.

If the absorber fails to converge, the Convergence tab contains a number of parameters which can be manipulated to help the absorber column converge. Try changing the following parameters:

Try the "Boston-Sullivan Nonideal" Inner Loop model instead of the default "Boston-Sullivan" (Parameters Grouping).
If the outer loop error appears to be steadily decreasing but the Column does not converge before reaching the maximum number of loops, increase the Maximum Iterations parameter in the Convergence tab, Solver Grouping.
The absorber normally does not require the "Composition-Dependent" Enthalpy Model (Parameters Grouping) even for absorbers with high acid gas loadings, however, this model might be required in some cases. Try this model only after first trying the other parameters mentioned above.
Try enabling the Boston-Sullivan Kb method by selecting the checkbox (Parameters Grouping).
If the absorber fails to converge after making a very large change to the sour gas composition/flow rate or to the lean amine flow rate, try making the change in increments. You can also try the Delete Last Solution button on the Column Convergence tab, Parameters Grouping.

If the stripper fails to converge, the Column Convergence tab contains a number of parameters which can be manipulated to help the stripper column converge. Try changing the following parameters. Change only one parameter at a time initially and in the order indicated:

If reboiler temperature is set, DELETE THIS SPECIFICATION. The column will have difficulty converging since reboiler temperature is determined by the stripper operating pressure. Set reboiler duty or steam rate instead.
The "Composition Dependent" Enthalpy Model (Parameters Grouping) is usually required for amine stripper columns.
If specifications such as composition (e.g. lean loading) or reflux ratio are set, change to specifications which allow easier column convergence, such as reboiler steam rate and condenser temperature. After the column converges and a profile is established, you may change to the more difficult specifications.
If reboiler duty is specified, be sure the value is not too high or too low. The duty should usually fall within the range of 800 to 1200 Btu per gallon of amine solution circulated (206 to 309 kJ/liter or 49 to 74 kcal/liter). This duty corresponds to a steam rate of 0.8 to 1.2 lb steam per gallon of amine solution circulated or 0.096 to 0.144 kg steam per liter of amine solution circulated.
Check the rich amine feed temperature to the stripper. If it is too low, the column will have trouble converging. The minimum rich feed temperature should be about 180 F (82 C). Temperatures of 200-220 F (93 to 104 C) are typical.
If the Column Type (Process Data tab) is set to TSWEET Kinetics, try using TSWEET Alternate Stripper instead.
Strippers operating near minimum reflux can often be solved using the "Boston-Sullivan Nonideal" Inner Loop model. instead of the default "Boston-Sullivan" (Parameters Grouping).
If the outer loop error appears to be steadily decreasing but the Column does not converge before reaching the maximum number of loops, increase the Maximum Iterations parameter in the Convergence tab, Solver Grouping.
Try enabling the Boston-Sullivan Kb method by selecting the checkbox (Parameters Grouping).
If the stripper fails to converge after making a very large change to the sour gas composition/flow rate, reboiler duty/steam rate, or lean amine flow rate, try making the change in increments. You can also try the Delete Last Solution button on the Column Convergence tab, Parameters Grouping.

If the Recycle has trouble converging in the specified number of iterations, you can increase the Maximum Iterations parameter. Open the Recycle block dialog, click the Process Data tab, click the "Advanced..." button, then click the "Solver Options..." button. Increase Maximum Iterations to 30 or 40. Scroll up through the Message Log and observe the Recycle error for each iteration. If the error is not changing or is oscillating, it may be necessary to decrease the default Weight for some components. For example, if the Recycle is placed in the rich leg or the semi-lean leg of an amine sweetening process, you may have to decrease the Weight for the reactive components (acid gas, amine, water) to obtain flowsheet convergence. If Weights are decreased, be sure to check the sensitivity of the results to the weighting.

Authored by Lili Lyddon (BR&E Technical Support and Help Author)