Fleet and backup tank sizing under regular and impulse stochastic demand patterns

The focus of this study involves logistics issues surrounding the industrial gas business. The industrial gas business is a major industry (approximately {dollar}30 billion/year) and has been experiencing substantial growth in recent years.; The industrial gas business supports two types of customers. The first type is the traditional merchant customers who rely entirely on the logistics system for product. Their demand is of a stochastic nature. The second type of customers is referred to as customers. These customers have air separation plants on their own sites and they must rely on the logistics system when their separation plant fails. To assist in providing a continuous flow of product, all onsite plants also have a backup tank which can provide sufficient product for relatively short durations of time, 24 to 96 hours of supply. The onsite separation plants are approximately 98% reliable, but when they fail, the customer generally requires a significant amount of product. This type of demand pattern will be defined as impulse stochastic.; This study focuses on two areas of the logistics support system. The first part deals with the economic sizing of onsite backup tanks. For a given onsite, we determine the most economical choice for a backup system considering numerous characteristics of the onsite customer as well as the logistics system. In the second part, we develop methods for determining the optimal fleet size under both regular and impulse stochastic demand patterns. The methods provide support for long range forecasting, short-term vehicle reallocation and private verses contract carrier vehicle mix determination. The methods further provide a means to describe the effect future onsite growth will have on distribution as it is related to vehicle fleet size.