Optimization On Transfer Hub Scheduling Of A Multi-product Pipeline Network And Fuel Replenishment

Multi-product transfer hub receives multiple oil products from the upstream pipelines,stores them and delivers them into downstream pipelines.The schedulers draft the operational plan of the tank hubs based on a trial-and-error method at present.In this method,the schedulers cannot always determine a feasible operational plan even if it exists because of a wide variety of operation rules,limited storage space,and complex topological structure,particularly when the total inventory of a type of product is close to the upper and lower capacity limitations.Therefore,it is necessary to explore the operational scheduling problem for the transfer hubs of multi-product pipeline networks.In this paper,An MILP mathematical formulation of multi-product transfer hub with multiple tank farms(M-TH)is proposed.This MILP model is formulated based on a new time representation with dynamic and static time slots.The scheduling horizon is partitioned into several static time slots by the time points of batches arriving at and leaving from the M-TH or the flowrate changed.Each static time slot is divided into several dynamic time slots,whose duration and starting/ending points are determined by the optimization process.The proposed formulation takes into account the capacities of the tanks,operational rules of the tanks and tank farms,structural constraints,and settling time after the loading operation while minimizing the switch operations between the tanks and switchovers between the tank farms.Moreover,the MILP formulation can be solved for one separate product at one time as there is no interference among the single-product scheduling problems within the transfer hub.The proposed formulation is validated using the optimal results of a realworld case.A time horizon partitioning strategy is employed to deal with the long-term horizon scenario.The results help verify that we can substantially save computational time and obtain a satisfactory solution;however,the optimality is compromised.Although the optimization problem of the M-TH can be solved by mathematical programming which can obtain the globally optimal solution,the computation times would be exponentially increased along with the increasing size of the instance.A feasible solution cannot even be attained by CPLEX sometimes.Therefore,in this paper,a new model based on event tree is proposed for the scheduling optimization of the M-TH.In this model,the concept of the event tree,the nodes,the elements of the event tree and the treatment of each event are defined.The model is formulated by the natural language so that it is intuitionistic and understandable.A depth first search strategy is used to solve the event tree model and an effective pruning method is used to save computation times.As the final link of the refined oil distribution,the fuel replenishment problem is to collect multiple products from the depot and deliver them to the gas stations with a fleet of vehicles.This paper deals with a multi-compartment,multi-trip and split-delivery routing problem that occurs in the context of petrol station replenishment.We propose a new variant of the split-delivery vehicle routing problem(SD-VRP)with multiple compartments and multiple trips.A Mixed Integer Programing(MIP)formulation is developed for the described problem.To benchmark,the exact solutions for some small instances are obtained using CPLEX.A column generation approach is additionally developed,and a lower bound is obtained as an additional benchmark.To accelerate solving the column generation,a heuristic is proposed to generate columns with negative cost based on Tabu Search.Besides,the round capacity inequalities are used in the master problem of Column Generation to get the tighter lower bounds.Two heuristic separation methods are used to separate the valid round capacity inequalities,named Robust Route and Partial Enumeration,respectively.This work also explores how the parameters of the instance impact on the computational difficulty of the instance,namely,the number of vehicles,the number of compartments in each vehicle,the capacity of the compartment and the number of the product types.The well-known Adaptive Large Neighborhood Search(ALNS)heuristic is adapted to solve the fuel replenishment problem for larger instances.To adapt to the features of the fuel replenishment problem,namely,split-delivery,multi-compartment,and multitrip,a Best Visit Insertion Strategy is proposed.Also,Trip Removal Operator,Longest Route Removal Operator,Visits in Longer Routes Removal Operator and Greedy with Split Preference Insertion Operator are proposed to accelerate the convergence.The solutions obtained by the ALNS approach are compared to the solutions of both the MIP model and the lower bound obtained by the column generation.