Abstract: Abnormal events may cause supply chain disruptions or order delays, with the result that an abnormal, procurement lead time (PLT) is usually longer than a normal one. Enterprises need to balance the cost and service level to develop the optimal ordering strategy to deal with abnormal events. A higher level of order quantity can reduce the number of orders and form a robust ordering strategy; a lower level of order quantity can form a flexible ordering strategy. The correlation of PLTs is defined as the mean of the differential multiplier terms between abnormal and normal PLTs. Based on the two correlated PLTs, a cost service equilibrium model is established for retailer inventory decisions to study the decision-making of order batches. Results show that when the cycle service level is high, PLTs correlation does not affect the ordering strategy of an enterprise; when the cycle service level is lower than a certain value, enterprises tend to order higher quantity, as PLTs correlation increases, so as to form a robust ordering strategy. Simultaneously, there are multiple service level thresholds for the impact of PLT fluctuations on order batch quantity. In different service level ranges, enterprises can adopt different optimal ordering strategies. This finding is significantly different from that only one service level threshold for the impact of PLT fluctuations on order quantity in the independent PLT hypotheses. As PLT fluctuations increase, enterprises may adopt different strategies from those with independent PLT hypotheses.

Key words: correlated procurement lead time (PLTs), cost, service level, order quantity