Short Term Planning and Scheduling For Gasoline Blending in Oil Refineries

Autor: Lamyaa M. Dawood, Alla-Eldine Hassan, Faisal G. Zhqiar
Rok vydání: 2021
Předmět:
Zdroj: Journal of Petroleum Research and Studies. 2:169-214
ISSN: 2710-1096
2220-5381
Popis: Product blending is an important optimization task that is encountered in the operation and scheduling of important industrial plants like petroleum refineries. The key objective of blending is to mix various intermediate products to achieve desired properties and quantities of products with minimum cost. There are uncertain parameters which make it very difficult to attain the optimum allocation of available resources. Consequently, there is a need to develop computational optimization techniques to tackle the blending issues. In this research the main objective is to propose an approach to solve product blending issue in an optimum way. The blending problem can be formulated as an optimization where its objective is to maximize net profit while determining the optimal allocation of intermediate streams to produce optimum production mix of final products. The proposed approach is introduced for integrating short term planning and scheduling for product blending. Two mathematical models have been proposed. The first model deals with planning issue for product blending and the results are regarded as production guidelines. In the second scheduling model, scheduling will treat the production guidelines to verify optimum allocation for available resources. The approach was applied to different real time case studies form Midland Refineries Company, and the results show the efficiency and flexibility of this approach to solve the different case studies. Also minimize lead time from 72 to 24 hr in the second case according to reduction in re-blend process, in addition to minimize production cost depending on optimum allocation for available resources. The last case study which is a complicated one, were WIN QSB version 1.00 software is utilized. The results gained after 0.031 second CPU time for planning level and 3.375 second CPU time for scheduling level. This is considered as an advantage to the model.
Databáze: OpenAIRE
Externí odkaz: