Receding Horizon Optimization of Ethylene Cracking Operation and Scheduling under Supply Chain Fluctuations

ABSTRACT

The global supply chain is experiencing ongoing turmoil after COVID-19, which inevitably affects the supply and demand of the chemical industry. Ethylene plants are regarded as one of the most profitable chemical industry chains, and thus, there is a need to track fluctuations in the supply chain and responsively adjust their operations and scheduling decisions to achieve optimal economic benefits. However, previous studies about ethylene cracking optimization are mostly limited to fixed supply chain parameters and cannot meet the need for real-time updates according to supply chain fluctuations. To address this problem, we propose an ethylene cracking optimization framework which combines parameter prediction and receding horizon optimization (predictive RHO) for real-time optimal operation and scheduling under supply chain fluctuations. This optimization framework uses historical supply chain data (supply quantities, feedstock prices, and product prices) to predict future parameters and uses the predicted parameters to optimize the operation and scheduling of ethylene cracking furnace systems in a receding horizon way. Two industrial case studies were used to validate the efficacy of the proposed optimization framework. In case 1, the predictive-RHO framework achieved comparable economic profits of 147,370 and 146,483 $/day without violating daily inventory constraints (vs 151,204 $/day from the original model violating inventory constraints) and a faster calculation speed of 40 s (vs 32 min) under two fluctuating feed supply modes. In case 2, predictive RHO improves economic profits by 1.8% through integrating operations with scheduling and 1.1% by tracking the feed and product price fluctuations. These two cases show the strong capability of predictive RHO in tracking supply chain fluctuations and improving the inventory management level as well as the economic profits of ethylene cracking plants.