An integrated order and demand management system design for a petroleum products distributor

The company supplies petroleum products both from domestic and foreign suppliers, stores them in its depots (terminals) all over the country, and sells them to its retail stations, aviation and marine customers, its distributors, and commercial-and-industrial customers. In this study we focus on the ‘commercial-and-industrial customers’ segment of the company, due to the relative convenience for new system implementations for this customer segment. These customers – including private and public sector customers such as factories or public institutions – expect the company itself to deliver the so-called black product (like fuel oil) within the time windows they require. The customers place their orders of varying sizes usually at random times. The order process begins with the agreement made between the customer and the company; then the company assigns one of its third-party carriers to the customer as the second step after which the order and distribution process goes on between the customer and the third-party carrier. Having observed the problems inherent in the current order and distribution management system of this customer segment, we propose a ‘Web-based Order Management System’ (WOMS), and a ‘Distribution Management System’ (DMS) together with an ‘Integration Module’ connecting the two systems. WOMS is designed to handle the order receiving and processing activities, whilst DMS is designed to be an operational decision support system for the logistics activities of the company related to the specified customer segment. The decision support is provided through an integer programming model, which is to be run by the company on a daily basis over a rolling horizon of five working days (a week). The model aims at minimizing the distribution logistics costs of the company, while deciding on the heterogeneous fleet size, the delivery quantities and dates, and the routes of the deliveries for the customer orders during the week ahead. We carry out a pilot study by considering the Aliaga depot, a representative of the 12 depots of the company, from where the customers in the Aegean region only are served. The model and its partitioned (easy-to-solve) version are both run using the software CPLEX for the six real scenarios, and the results are compared with the current system's actual results. The proposed DMS provides the company with solutions that result in a savings of 13.5% to 20% in distribution costs and a smaller number of third-party carriers, in about 2 hours of computation time with the gaps in the solution quality between 0.38% and 1.80% by means of the main model. By the help of the proposed system on the whole, it is expected that the company's control over the order and distribution processes can be improved, and that its working style can be changed from a reactive one to a proactive one.