Industrial Gas Supply Chain Logistics: Global Liquid Helium Planning and SKU Rationalization Methodologies in Packaged Gas

Abstract

We present applications and developments of operations research methodologies in the area of industrial gas supply chain logistics. The first problem studied regards the liquid helium business. The global supply chain for liquid helium presents a complex structure due to increasing foreign demand, elaborate recovery techniques, and costly forms of distribution. Although the problem contains parallels to the liquid natural gas supply chain, supply requirements and problem-specific network constraints require a unique optimization model. We develop a large-scale, discrete time, path-based integer-programming model which solves optimally with CPLEX. Computational results implementing a rolling horizon structure and testing based on historical data are presented. A detailed sensitivity analysis demonstrates the effective use of our model, testing a variety of realistic parameter settings for the liquid helium supply chain.The second problem studied regards SKU rationalization in the packaged gas business. Effective SKU rationalization is advantageous when applied to businesses with a high variance and variety of product offerings. Advantages may include lower production costs, inventory simplifications, and system-wide reductions in transportation costs. We apply SKU rationalization in the form of a variant of product substitution, towards an industrial packaged gas supply chain problem which includes production, allocation, and distribution decisions. An effective mixed-integer programming formulation is developed, capable of handling additional line investment, varying degrees of substitution, economies-of-scale in production, as well as network-wide planning decisions in the supply chain. A case study based on historical data is used for testing, followed by computational results and policy implications in the form of customer incentivization.Lastly, we present extensions of the aforementioned SKU rationalization problem. A customer-selection heuristic is developed and shown to perform very well compared to a full MILP formulation, which includes customer selection for substitution as an explicit decision. Following this, incentivization strategies are studied in an effort to reduce lost demand via product substitution, while simultaneously ensuring adequate business profit. Multiple strategies utilizing stochastic customer behavior are proposed as valid alternatives for decision makers, dependent on risk-preferences and instance-specific customer behaviors imposed.