Optimal Planning of the Supply Chain at Perstorp Oxo AB
The purpose of this thesis is to develop a mathematical optimisation model that can be used as a decision support tool for the supply chain planning at Perstorp Oxo AB. Perstorp Oxo AB is a subsidiary to Perstorp AB which is global company in the process industry. Perstorp Oxo AB produces chemicals at the site in Stenungsund to customers in a variety of branches and for further refinement at other Perstorp plants. In total, 11 oxo products (2-EH, 2-EHA, 2-EHAL, DOP, DPHP, IBAL, IBUT, NBAL, NBUT, PRA, PRAL) are produced in Stenungsund. Further refinement is carried out at the sites in Gent (NX795 and Bis-MPA), Castellanza (propionates) and Perstorp (Neo and TMP). The customers are mainly in branches such as food and feed, leather and textile, plastic and safety glass production. Since Perstorp AB acts on a global market, Perstorp Oxo sells products to customers worldwide. Therefore, Perstorp Oxo has two large inventory facilities in Antwerp, Belgium and Tees, United Kingdom for five product types each and two smaller in Philadelphia, USA and Aveiro, Portugal for one type respectively. The inventories in Antwerp and Tees are replenished from Stenungsund, whereas the inventories in Philadelphia and Aveiro are replenished from Antwerp, i.e. indirectly from Stenungsund.
Today, Perstorp Oxo AB does not use any advanced planning tools or optimisation algorithms when planning the supply chain at the site in Stenungsund. It is believed that by having an optimisation tool for the planning of the supply chain, the supply chain at the site could be improved and be more profitable. Also, there is a need to involve profit margins into the planning process in order to make plans that improve the profitability.
We have, in order to fulfil the purpose of the thesis, carried out different activities. To begin with, we started to map the supply chain and to identify the theoretical frame of reference. We chose three scenarios that were to be investigated during the experiments. This called for a working optimisation model, which was developed dynamically, with conceptual modelling, data gathering, model formulation and programming and validation carried out in an iterative fashion. The scenarios that were investigated and compared to a baseline scenario during the experimentation phase treated: an optimised baseline scenario, an unplanned stop at the production of 2-EH and 2-EHAL and an unplanned stop at the Neo plant in Perstorp.
The developed model is of a mixed-integer programming type. The objective function maximises the profit margin, i.e. the difference between the selling price and the cost of production, transports, inventory carrying and possible purchases from external suppliers. The constraints treat maximum and minimum capacities in production, inventories, transports and sales. Further, the model also treats setups between DOP and DPHP, production rate changes and relationships in the production. The optimal solution shows the quantities to be transported between the different sites, production rates, inventory levels, setups and purchases from external suppliers, each with its respective cost.
The results of the experiments show that there is a potential to increase profit margin by using a decision support tool based on an optimisation model as the results of scenario 1 are studied. The results from scenario 2 and 3 are ambiguous and they give no univocal conclusion to however an optimisation model can be used as a decision support tool when severe production decreases occur in Perstorp Oxo AB’s supply chain.
However, the model shows signs of instability, thus providing a non-robust optimal solution. This is considered to be a result of partially invalid parameters and the solution method. Therefore, the results provided in this thesis are not extensive enough to draw any conclusions upon however a decision support system based on optimisation can be used to improve the planning at Perstorp Oxo AB.
AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)