Multi-objective optimization of both pumping energy and maintenance costs in oil pipeline networks using genetic algorithms

This paper proposes an optimization model for the pipeline operation problem using a dual-objective non-dominated sorting genetic algorithm (NSGA-II). One and foremost objective is to minimize pumping energy costs. The second objective is to recognize the pipeline operators' concern on pumps maintenance costs by reducing the number of times pumps are turned on and off. This is commonly believed as a main source of wear and tear on the pumps. The formulation of the problem is presented in detail and the model is tested on a hypothetical case study (which is based on consultation with two industrial partners). The output results are promising since they would give operators a better understanding of different optimal scenarios on a "Pareto front". Operators can visually assess several alternatives, and analyse the cost-effectiveness of each scenario in terms of both objective functions.