A Matheuristic for the Drilling Rig Routing Problem

In this paper, we discuss the real-world Split Delivery Vehicle Routing Problem with Time Windows (SDVRPTW) for drilling rig routing in Siberia and the Far East. There is a set of objects (exploration sites) requiring well-drilling work. Each object includes a known number of planned wells and needs to be served within a given time interval. Several drilling rigs can operate at the same object simultaneously, but their number must not exceed the number of wells planned for this object. A rig that has started the work on a well completes it to the end. The objective is to determine such a set of rig routes (including the number of assigned wells for each object) to perform all well-drilling requests, respecting the time windows, that minimizes the total traveling distance. The main difference with traditional SDVRP is that it is the service time that is split, not the demand. We propose a mixed-integer linear programming (MILP) model for this problem. To find high-quality solutions, we design the Variable Neighborhood Search based matheuristic. Exact methods are incorporated into a local search to optimize the distribution of well work among the rigs. Time-window constraints are relaxed, allowing infeasible solutions during the search, and evaluation techniques are applied to treat them. Results of computational experiments for the algorithm and a state-of-the-art MILP solver are discussed.