461970 A Game Theoretic Framework for Strategic Production Planning

Tuesday, November 15, 2016: 4:31 PM
Carmel II (Hotel Nikko San Francisco)
Philip Tominac and Vladimir Mahalec, Chemical Engineering, McMaster University, Hamilton, ON, Canada

Strategic planning allocates future production targets which link plant scheduling procedures to high level organizational decisions. In most literature strategic planning is driven by two common objectives: minimization of the costs associated with meeting fixed production objectives, or maximizing the profitability of a process according to a set of fixed product prices [1]. From a strategic perspective neither of these objectives addresses the potential for changes in market conditions nor the role of competition in organizational planning. The solutions obtained by these types of approaches represent cost effective plans and may yield reasonable production targets that maximize profitability, but such results may not represent competitive strategies, and the value reported by the model may not be realized, or be achievable, as a result.

All industries face competition and need to make decisions accordingly in order to avoid losses in profits or market shares. The study of competitive decision making processes falls into the field of game theory which has developed methods to quantify competitive behaviour. Our research examines methods of incorporating game theoretic descriptions of competition into strategic production planning, which connects organizational decision making to process models allowing competitive decisions to be made within the operating space of existing process equipment. The production targets emerging from such strategic planning methods represent competitive strategic planning results.

We present a framework for game theoretic strategic production planning based on the potential game formulation of a Cournot oligopoly model of market prices which incorporates the planning efforts of multiple organizations competing in different markets [2][3][4]. The framework is illustrated using strategic production planning models of crude oil refineries. In this framework competing organizations each produce multiple products which are sold in local and export markets. Product prices in those markets react according to the total supply of each product received from all suppliers. Refiners are forced to set production targets which take into account not only market conditions but also the behaviour of their competitors such that the resulting strategic production plans form Nash equilibria.

We demonstrate the use of the framework in strategic production planning and extend its use to market structure planning and process expansion planning. As Nash equilibria, the strategic results obtained are robust to changes in competitor behaviour and represent optimal responses to both market conditions and competitor behaviour taking into account the operating costs and limitations of the underlying process model. The resulting strategic plans have a property not shared by plans generated using cost minimization or profit maximization approaches; they are rational according to the description of competitive market forces, and thus avoid production plans which under fixed price analysis may be considered optimal but are unachievable in a game theoretic perspective.


[1] Papageorgiou, L., “Supply chain optimization for the process industries: Advances and opportunities,” Comput. Chem. Eng., vol. 33, 2009.

[2] Monderer, D., Shapley, L., “Potential Games,Game. Econ. Behav., vol. 14, 1996.

[3] Slade, M., "The Fictitious Payoff-Function: Two Applications to Dynamic Games," Ann. Econ. Stat., 1989.

[4] Slade, M., "What does an oligopoly maximize?" J. Ind. Econ., vol. 42, 1994.

Extended Abstract: File Not Uploaded
See more of this Session: Planning and Scheduling I
See more of this Group/Topical: Computing and Systems Technology Division