Development of Mixed Integer Non-Linear Model for Hydrogen Distribution in them Refinery

Abstract

In the refining industry, growing amount of hydrogen is needed for petroleum conversion and clean fuel production. The design of hydrogen networks in industrial production sites aims at minimizing the consumption of fresh hydrogen by optimizing the operating parameters, more and more by trying to recycle the degraded hydrogen produced in the process units, and by integrating hydrogen purification units.
For this purpose Pinch Technology, linear model, nonlinear model is obtained in terms of flow rates and purities of streams handled by compressors and purifiers. The optimal design of hydrogen networks aims at minimising the consumption of fresh hydrogen by improving recycling and reuse of process hydrogen. In the literature this was achieved by LP and NLP models which computed the minimum hydrogen requirement, and made a preliminary selection of the compatible purification units. These models also integrated compressor units.
In the present work the distribution of hydrogen is carried out by MILP (Mixed integer linear programming) model which is solved by mathematical programming software GAMS. The detail modeling and solution techniques are discussed in the report. Along with this, some excel data sheets are provided which shows how the calculations were done in LP and NLP model . The results obtained from present study compare well with the published work. Further, some suggestions are given for future research.