Abstract: The optimization of pumping scheme at each pump station along the products pipelines during the operation plays avital role in cutting down the energy consumption of products pipelines. The previous studies on this issue mostly make use ofdynamic programming and artificial intelligence algorithms to get the solution, which have certain limitations, but rarely establisha mixed integer linear programming model (MILP) and take the time limit of pump start-stop as the constraints of MILP model.In this paper, the MILP mathematical model was established on the basis of the known batch scheduling plan. In this model, theminimum cost of normal operation and start-stop of pumps is taken as the objective function, with the consideration of constraintconditions, e.g. the time limit of pump start-stop, the node pressure along pipelines and the flow rate being pumped. Then, themodel was solved by means of branch and bound algorithm. Finally, a case study was carried out on the transportation plan ofone in-service products pipeline. It was introduced into the MILP model for optimizing its pumping scheme. And the calculationresults indicate that the model can greatly reduce the energy consumption of pipeline operation and improve the stability ofpipeline operation to satisfy the requirements of on-site operation process.