(371k) Stop-of-Supply Characterization as a Metric to Improve Operations in a Chemical Supply Chain

A random process as the demand of a chemical product can be characterized by a compound-Poisson distribution, when both the amount and number of orders are assumed uncertain. Then if the total demand, denoted S, in a certain period is estimated, the quantity (S-m)+ represents the amount in excess of the supplier capacity m. Here Stop-of-Supply, is defined as a characterization of the probability to stop the supply according to the capacity parameter m. At a certain time, a level of stock is planned under demand uncertainty. Probability of excess of demand is obtained for each period, to be turned to a parallel supplier. The problem is to fix a policy of the ratio for in-house to the outsource-supplier of the product to be delivered. In other words, we address the management of outsourcing a product in order to keep high levels of performance in the supply chain. Our main objective is to maximize profit considering several supply chain issues but also chemical process plants operative aspects, like plant capacity for processing the chemicals, multiple operation modes of processing units and optimal schedule. The framework we propose to solve this problem is based on two level planning: one focus on real time data processing to provide a density function. Stochastic techniques enable us to obtain our goal. The second high level planning is based on the surrogate model of the supply chain that involves processing plants to produce the chemicals. Here, the flow rate of raw materials, intermediate and final products become decision variables that define the inventories allocated in the process unit plants to operate under a certain operation mode to maximize a reward function subject to market prices, operation costs and holding costs of inventories and incurred penalizations for outsourcing to parallel suppliers. Thus, a wide set of optimal distribution of operations within the chemical plant is obtained for sensibility analysis of the stop-of-supply formulation. Operations management can be improved since smooth operation plans can be selected provided a level of risk of stop-of-supply is selected.