Flowsheet Tolerances (Input menu)

Many of the calculations in PRO/II are iterative, and require certain relationships to be satisfied within specified tolerances to reach a solution. Setting the correct tolerance levels is important because tightening or loosening them will have significant impact on the simulation results and execution time. A rigorous application might need tight tolerances, whereas substantial computer time may be saved by loosening the tolerances for a less rigorous application.

Some of the tolerances, such as those for flash calculations with the "Version 5" and the "Alternate" Flash Algorithms, are built into PRO/II and cannot be changed. For other product and unit performance specifications the default tolerances are set using this window. Tolerances for individual specifications can be changed using Specification Tolerance Window which is usually accessible from that specification.

The global tolerance for the "Default" Flash Algorithm may be set by the user in this window. A tighter tolerance tends to increase CPU time. If tightened too much, it will result in flash failures. On the other hand, a looser tolerance will solve the flash more quickly, but typically adds "noise" to the derivatives. When encountered by controllers, optimizers, and recycle algorithms, this noise may hinder or even prevent convergence to a satisfactory solution. If you choose to adjust the flash tolerance, make small adjustments. Make no other changes to the flowsheet, and note the effects upon the results in successive runs.

Tolerances for recycle loops also default to the values set in this window, but can be changed later either globally for all loops, or for individual loops. See Recycle Loops for more details.

Usage

From the Input menu, select "Flowsheet Tolerances" command. The Default Unit Specification Tolerances window displays all the tolerance levels that will be applied for all iterative calculations in the simulation. You can override the default values provided by PRO/II if desired.

The tolerance can be expressed in either Absolute or Relative basis. Tolerance values can be specified as follows:

Temperature

This sets the error tolerance for unit operation temperature specifications. The default is an absolute tolerance of 0.001 F, or equivalent.

Pressure

This sets the error tolerance for unit operation pressure specifications. The default is a relative tolerance of 0.001.

Duty

This sets the error tolerance for heater or cooler duty specifications. The default is a relative tolerance of 0.0001.

Flash

This sets the global tolerance for the "Default" Flash Algorithm (see Miscellaneous Data Main Window). The default tolerance is 3.00e-006. It does not apply to the "Version 5" or the "Alternate" Flash Algorithm.

Miscellaneous

This sets the error tolerance on all other types of specifications, including unit operation parameter specifications. The default is a relative tolerance of 0.001.

Mass Balance

This sets the over-all mass balance tolerance for flow sheet. If over-all mass balance error exceeds this value, then flowsheet is shown as un-converged (Problem solution not reached) even if all unit operations and streams in the flowsheet are solved. The default value is 1%.

The tolerance default changes between PRO/II v9.4 and PRO/II v9.3 are summarized as below:

If you have selected absolute tolerance, you can change the dimensional units of temperature, pressure and duty using the Unit Of Measure Selection Window, which is displayed when the UOM button on the Toolbar is pushed. Note that this tolerance value is converted to the unit selected as the preference for the current problem when this window is displayed again.