A decision support tool for process optimization of sulphur free diesel production

This work is a contribution to the Sulphur Free Diesel (SFD) production. The research presented in this paper is performed in one of the Dutch refineries, which is going to be one of the front producers of SFD from high sulphur crude. The challenge is to adapt the process operation to changes in the specification requirements without capital investments, i.e. to find optimal process settings which maximize the margin taking into account blend values, deactivation costs of the catalyst, shutdown margin losses, product specifications and quality margins and considering the organizational complexity, too. The decision support tool, called SFD optimizer, determines production settings on a weekly basis to decide on the weekly schedule by Economics and Planning Department and in case of disturbances. The NLP-optimization problem with two types of decision variables (desulphurization depths for three HDS and five quantities of components sent to product pools) and with 22 (non)-linear (in)-equality constraints is solved with generalized reduced gradient method. The optimal solution results in maximal SFD margin, i.e. the total blend value minus the deactivation costs. The SFD decision support system presented in this paper was proven to effectively assist decision makers at the Economics and Planning Department.