Uma abordagem heurística para o problema de roteamento de sondas de intervenção bi-objetivo

Abstract

Throughout the productive life of oil wells, some interventions known as workovers are usually necessary. These repairs, are held by intervention rigs and are essential to maintain production or possibly improve it, these being scarce and expensive. Given a set of wells requiring maintenance and a heterogeneous fleet of rigs, each rig with a specific starting position, the Workover Rig Routing Problem (WRRP) consists in finding routes for the rigs so that the total loss of production of wells within a planning horizon is minimized. The route of a rig is considered feasible if it has the necessary equipment to carry out all the maintenance of associated wells and respect the stipulated planning horizon. In the literature, many authors have proposed models and solution methods for the WRRP mono-objective, i.e., trying to minimize just the total loss of production, not considering the rental of rigs. Therefore, this paper studies the WRRP bi-objective which minimizes both the total loss of production of the wells and the rental costs of rigs. We implemented an Adaptive Large Neighborhood Search metaheuristic for this problem, which was tested considering instances proposed in the literature as well as from new situations proposed in this paper, in order to approach the problem in real-life context. The results of this research show that although the minimization the total loss of production is an a very important objective, the variation of the rental cost of the probes and as oil prices affect in the definition process of routes.