Yield Based Planning Example of a Single Refinery

Yield Based Planning Example of a Single Refinery 

In this section, we use a simplified refinery model in order to demonstrate the concept of using mathematical programming for the planning of process-yield-based 

nc4 measured nc4 predicted (simulator) nc4 predicted (ANN) c4- measured C4= predicted (simulator) C4= predicted (ANN) 

This example illustrates the planning of petroleum refinery operations as an ordinary single objective linear programming (LP) problem of total daily profit maximization. 

The minimum number needed to define a system fully should be identified. In this example, it is three since, for example, fixing x3 determines x7, %4, Xg, %9, and xw fixing %2 then determines xu, x16, x3, x14, x17, x2o, and xis and finally fixing xs then determines xu, xlg, x13, x19, and x6. An LP model could be formulated using only these three structural variables or any other suitable three variables. In this case, the solution would only give values of the three variables and the remainder (if needed) would have to be calculated from them afterwards. It is usually more convenient to include additional variables in LP programs. 

---