Multi-aspect efficiency measurement of multi-objective energy planning model dealing with uncertainties

Abstract

In this study on energy planning, a combination of multi-objective optimization and efficiency measurement is proposed as a means for determining an efficient energy mix that considers the multi-dimensional nature of energy planning and its associated uncertainties. Various multi-objective functions are appended to the proposed optimization model to meet requirements related to energy need, cost, environmental impact, security, social impact, and social benefit. A slacks-based measure methodology is applied to determine the best energy mix from the alternatives produced by the appended model. The energy efficiency of each energy mix is measured from the linear combination of its defined inputs and outputs. The outputs to be maximized include total generated electricity, direct employment, and percentage of generated electricity from renewable energy, while the inputs to be minimized consist of total economic cost, carbon dioxide emission, total social cost, and power-plant-type dependence score. To demonstrate the applicability of the proposed model, a case study of Thailand’s power development plan is featured. Various types of power plants, both fossil fuel-fired and renewable energy-driven are considered in the empirical analysis. The results show that the proposed method can contribute significant improvements, including a reduction in total emissions and in the power-plant-type dependence score (by 31.41% and 25.59%, respectively). It also increases total employment and the proportion of generated electricity from renewable energy plants (by 25.73% and 47.39%, respectively), with marginal tradeoffs of total costs and total social costs (which increase by 8.94% and 13.89%, respectively). Quantitative results from the model could help policy makers efficiently determine an appropriate energy policy—one that optimizes all the various aspects, under a given set of constraints and scenarios of uncertainty.