Optimization of India's electricity generation portfolio using intelligent Pareto-search genetic algorithm

Optimization of power generation mix is a significant strategy of climate change mitigation for countries like India. This involves multi-objective optimization of cost reduction, emissions reduction and risk mitigation taking into account relevant constraints. We use a variant of portfolio optimization technique to generate India's 12th five year plan electricity generation portfolio taking into account the carbon costs. For fitness evaluation of a generation portfolio, we use levelized generation costs and a Comprehensive Risk Barrier Index (CRBI), the latter capturing the cost risks modulated by project implementation barrier indices. For constrained optimization, we develop a fast hybrid algorithm, namely, Intelligent Pareto-search Genetic Algorithm (IPGA), which systematically evolves successively efficient frontiers and finally converges to the global Pareto-optimal front. This algorithm combines non-dominated sorting and separate elite population, while utilizing dual mode search for faster convergence and cluster reduction strategy for enhancing diversity. Halting mechanisms have been proposed for local and global Pareto convergence. We apply this generalized algorithm to simulate the impact of carbon costs, risks and barriers on India's optimal generation portfolio.