(186q) Dynamic Production Planning and Scheduling for an Chemical Plant

Chemical/petrochemical industries are facing increasingly challenges to sustain their profitability nowadays due to volatile raw material and product markets, increasingly rigorous environmental and safety regulations, and global competitions. As the key player of chemical industry, the olefin plant is the most important sector for the entire value-added chemical product chain. An olefin plant employs multiple cracking furnaces in parallel to convert various hydrocarbon feed stocks to smaller hydrocarbon molecules, mostly ethylene and propylene. The continuous operational performance of cracking furnaces gradually decays because of coke formation in the reaction coils, which requires each furnace to be periodically shut down for decoking. Thus, production and decoking scheduling of the cracking furnace system is an importance and challenging task for olefin plants. In the previous studies, olefin plant scheduling works were almost exclusively aiming at maximizing the total profit of the furnace system based on the available raw material types and quantities, performance of each furnace, as well as the market prices of raw materials and products. The front-end feedstock procurement planning and the back-end production response to product market have been simplified or just neglected.

In this study, a systematic methodology for the optimization of the entire supply chain of olefin plants have been developed, which covers feedstock purchasing and storage management, olefin production, and productivity control and response to short and long-term product markets. It couples the production planning for both front-end feedstock procurement and back-end production response to product market, as well as the olefin plant scheduling for the optimal olefin product manufacturing. The planning solution optimally determines how much of each major product to produce with respect to time; when and how much of each of feedstock to buy; and solutions of feedstock inventory management. The scheduling solution will determines the detailed production information such as batch processing numbers of each furnace during the given scheduling time, feed allocation, starting and ending time of each batch of each furnace, and the decoking sequence. Such dynamic planning and scheduling results will be simultaneously obtained through a novel MILP (mixed-integer linear programming) model. A case study with real olefin plant data is employed to demonstrate the efficacy of our development,