What are the Adjust and recycle logical operations?
Adjust and recycle logical operations are examples of iterative calculation routines. That is, these operations take an initial guess and run the calculation with that guess. The result is compared against a target value, and if it matches within tolerance limits, the calculation is finished. If not, the initial guess is adjusted by a program algorithm, and then rerun. These routines requires experience in their use before they can be used effectively.
Is there any guideline to use them?
Most of the current advanced simulation software such as HYSYS, ASPEN Plus, ProII, CHEMCAD, BRE Promax, etc. has logical operators which includes the adjust and recycle operations.
Following are some of the guidelines to remember when setting them up:
- Adjust operations need to be carefully placed and used. Improper placement can cause these operations to come into conflict with one another. A typical example is when two adjust operations are used to control the hot and cold side flows in a heat exchanger. In addition, unnecessary use of adjust operations will increase calculation complexity and time requirements.
- Adjust operation boundary conditions should be properly set. Users should be careful to establish reasonable lower and upper ranges for the adjusted variable. For example, no pressure variable should be allowed to drop below 0 psia (or 0 absolute), though the computer program may allow it. Reasonable boundary conditions may be based on the normal operating ranges for a particular piece of equipment.
- Recycle operation use should be minimized. These should be used only if there are undefined or unknown stream variables (observe the rules regarding degrees of freedom in a calculation). Too many recycles slow calculations and may force one recycle operation into conflict with another.
- The initial guess for the adjusted variable or recycle stream should be as accurate as possible to minimize the number of iterations needed to converge on a solution.
- Step sizes affect the number of iterations required for convergence of solutions. Too large a step size may cause these operations to “miss” the convergence or solution by making them jump around a mathematical “saddle point." Step sizes that are too small may require an excessive number of iterations to converge.
- The iteration count limit should be sufficient to allow a solution to converge. Too small a number requires frequent user intervention during calculation runs.
- Calculation priority should be used when adjusts or recycles are set up in series to force the program to execute and converge the operations in the proper sequence. Failure to do so can extend the number of iterations required, cause a solution convergence failure, or in the worst case, cause both to happen.
- Users should check to make sure all operations have properly converged before reporting results. Simulation programs will occasionally stop calculations without generating an error report, and then fail to converge all logical operations (recycles and adjusts).