Research On Multi-network Integration Optimization Methods For A Refinery

With the economic and social development, energy saving of the oil refining enterprises has aroused public concern. This will make higher demands for energy saving of refinery and effective methods is urgent needed to optimize the entire process of refinery production to achieve the refinery business profits maximized, to reduce resource consumption, to reduce costs, to improve production efficiency and to reduce environmental staining. As the refinery production process is a continuous, complex, multi-process production process, if the entire refinery production process is integrated into a model, the model will be very large, which brings difficulties to the optimization calculation. In order to overcome computational difficulties, this paper used refinery multi-network coordination optimization model to optimize the whole production process.First of all, this paper briefly reviewed the research development status on the oil refinery optimization and the technology of approximation. On the basis of characteristics of the whole production process for the refinery, the production process was decomposed into the network of production, the network of hydrogen, the network of stream and the network of water.Secondly, the paper analyzed the multi-network coordination optimization framework and explained the reason of selecting collaborative optimization framework as the framework of refinery multi-network coordination optimization. To address the difficult which caused by the system-level consistency constraints of collaborative optimization, the paper used dynamic relaxation factor method to relax the system-level consistency constraints and one example indicates that the method of system-level consistency constraints relaxation can reduce the computational. In order to further reduce the optimization calculation, the paper used the Kriging approximation method to optimize the subsystem level and this method can reduce the computational significantly which verified by one example.Again, according to multi-network collaborative optimization framework, a multi-network integration optimization model which is based on Kriging approximation for a refinery is established. Using the iSIGHT software platform and validating by experimental data, the results show that the optimization model for refinery multi-network integration can increase the overall refinery benefits.Finally, I made a summary and outlook for this paper, and pointed out the further research direction.