A Decision support framework and an open source software for saving water in chemicals industry

Abstract

In this paper, we describe a general software framework that integrates optimization into daily production planning that is widely encountered in chemicals industry. We also describe how we have realized the framework as a decision support system which solves the product sequencing problem in chemicals processing industry. In chemical industry, washes at chemical processing mixers are typically required to prevent the contaminating effects of the previous products on the current product being mixed. The minimization of the number of required washes is possible through effective production sequencing, yielding a signicant saving in the consumption of clean water. As a case study, we implemented and tested our software at a chemical plant located in an industrialized region of Turkey. The plant produces industrial cleaning chemicals and is owned and operated by JohnsonDiversey Turkey, which produces a variety of products for local and global markets. The system reads critical production-related information from a database and generates optimal or at least "good" solutions sequencing the products to be produced at a mixer. The optimal solution is obtained for small problem instances through solving a Mixed Integer Programming (MIP) model, whereas a "good" solution is seeked for larger instances. Our study integrates a unique set of critical features for the rst time in literature such as defining product groups, handling urgent orders and effectively visualizing production schedules. The company reported to obtain substantial water savings since this new system was into practice by the production planners.