Nowadays, a large focus is given to mass personalisation, and multiple path shop floors are suited to such production environments. Hence, this paper deals with the Job-shop scheduling problem that is used for modelling a manufacturing system. Meanwhile, a large attention is given to energy consumption of production systems, but few works consider power requirements of the production systems in order to process operations. In order to contribute in filling this gap, this paper considers the problem where the objective is to minimise both the total completion time of all operations and the instant available power required to process these operations. The problem results in the Bi-objective Job-shop Problem with Power Requirements (Bi-JSPPR). The goal of this paper is to provide a Pareto frontier of schedules minimising both criteria, considering that operations may consume a lot of power at the beginning of the process (consumption peak), more than its consumption after a while, which allows to model power profiles of manufacturing operations. To solve the problem two metaheuristic approaches are investigated: a hybrid Non-dominated Sorting Genetic Algorithm (NSGA-II) and an iterated Greedy Randomized Adaptive Search Procedure coupled with an Evolutionary Local Search (iGRASP×ELS). An efficient local search procedure is specifically designed to improve the quality of solutions in the Pareto frontier of the hybrid NSGA-II (hNSGA-II). Computational experiments and statistical tests are conducted to demonstrate the efficiency of the approaches. Results show that both approach are complementary, having the hNSGA-II showing better average performances, while the iGRASP×ELS is better when high peak power consumption are considered.