Currently, pipeline is the most effective way to transport large-volume products over long distance. To effectively satisfy market demands for multiple refined products by delivery due dates, the multi-product pipeline network usually transports several refined products in sequence from refineries to certain destinations. The integrated scheduling of multi-product pipeline network, including inventory management, transport routes planning, batch sequence, batch volume, et al., is one of the most strategic problems due to its large-scale, complexity as well as economic significance. This subject has been widely studied during the last decade. However, most researches focus on large-size scheduling models whose computational efficiency greatly decreases for a complex pipeline network or a long time horizon. Aiming at this problem, the paper develops an efficient decomposition approach, which is composed of two mixed integer linear programming (MILP) models. The first model divides the entire time horizon into several intervals according to delivery due dates and optimizes the transport routes and total transport volume during each interval with considering market demand, production campaigns and inventory limits. Then the solved results are used by the second model which sets the objective function as the membership function based on fuzzy delivery due dates. Besides, a series of operational constraints are also considered in the second model to obtain the optimal batch sequence, batch volume, delivery volume and delivery time in each node. Finally, the proposed approach is applied to a Chinese real-world pipeline network that includes 5 complex multi-product pipelines associated with 6 refineries and 2 depots. The results demonstrate that the proposed approach can provide a guideline for long-term pipeline network scheduling with delivery due dates.

Considering market’s diversified demand and transport economy, large volumes of various refined products commonly move down the pipeline in batches, which are pumped at pump stations and delivered to respective delivery stations. The integrate detailed scheduling optimization is a sophisticated problem due to the characteristics of multi-product pipelines, such as market-oriented, fluctuated demand, various processing technique and complicated hydraulic calculation during batch migration. The integrate detailed scheduling optimization has been widely studied during the last decade, however, most of them studied pipeline scheduling and pump scheduling separately. Besides, the proposed methods are mathematical models, whose computational efficiency greatly decreases in large-scale pipeline scheduling, let alone in the problems coupling with pump scheduling. Aiming at this problem, this paper presents a novel depth-first searching approach based on flowrate ratio to deal with the detailed scheduling of operations in a multi-product pipeline with multiple pump stations. As for each single time interval, the proposed method decides an ideal flowrate ratio according to current status, then solves out the optimal flowrate that mostly conforms to the ideal ratio and satisfies all operational constraints, and finally updates information for next time interval. However, during the computational procedure, backtracking method would be adopted to modify the previous flowrate ratios and recalculate new flowrate when the actual delivered products are insufficient. Finally, a case tested on a Chinese real-world pipeline with 6 delivery stations is given to demonstrate the veracity and practicability of the proposed method. From the results, computing time of the case is within 1 minute, and the solved detailed scheduling plans can fulfill demand with stable pump operations. Besides, the proposed approach is scarcely influenced by the scale of pipeline structure and time horizon, so it is also applicable to the long-term scheduling of a pipeline with many delivery stations.