In this paper the problem of the Job-shop is extended to support energy constraints. The objective is to propose scheduling tools for manufacturing systems considering consumption threshold that must not be exceeded. The operations are supposed to consume more energy at beginning and thus representing a consumption peak that is often present in machine tools. This assumption results in considering that an operation is divided into two sub-operations. The goal is then to propose the best schedule considering the energy threshold, the consumptions of operations and duration of consumption peaks as given data. A Mixed Integer Linear Model (MILP) for the problem solving is proposed; it is based on flow approach to take into account the energy threshold. Since it is difficult to find exact solutions for medium and large size problems, a metaheuristic based on a GRASPxELS is proposed. Small scale instances for the problem have been generated, and results expose the relevance of the metaheuristic approach.