Integrated energy systems modeling for policy analysis and operational planning

Abstract

This research focuses on the application of mathematical modeling for analyses at the levels of long-term national energy policy and medium-term operational planning for electric utility in India. bong-term policy concerns addressed are: i) optimal technology/fuel mix and emissions projections; ii) sensitivity to macroeconomic and policy parameters like pectoral demand growth, energy prices, discount rate, demand side management (DSM) policies, and carbon emissions mitigation targets; m) hedging strategy of investments under long-term uncertainties. Medium-term utility level concerns addressed are: i) sensitivity to capacity mix; ii) hedging strategy under uncertainty in water inflows. Long-term policy involves consideration of complex inter linkages among fuels, Supply demand technologies, and demands. Energy demand in a sector is driven by its growth and technology mix. Hence, an exogenous integration of energy system model, end user sector models, and demand projection is attempted. Planning horizon spans 40 years. 22 electricity generation technologies and 86 demand technologies are considered. End-use sector modeling is done for 15 sectors, including 10 industries. Technology and fuel costs exhibit regional variation in a large country like India due to its diverse resource endowments. Traditional models with single costs are inappropriate in this context. Multiple cost structure, capturing the non-linear distribution of reality, is adopted. Stochastic programming is used to incorporate uncertainties. Analysis with long-term modeling suggests the following: i) Under business as usual case, Indian energy system is expected to remain coal dependent over long run. Some switch away from coal is expected to occur, mainly due to penetration of gas based thermal technologies in power sector, and coal to electricity substitution in steel and cement industries. Peak :off-peak ratio of long run marginal cost of electricity stabilizes at 2.6. ii) Strict carbon mitigation targets can significantly increase the cost of production of industrial products like cement, steel, brick and aluminum iii) In earlier periods. peak shifting DSM policies are more effective than energy conservation policies, with a potential of i5% savings in power sector investments. iv) Uncertainty in carbon mitigation target has high expected value of information (Rs. 80 billion over 40 years). Consideration of joint uncertainties increases the expected value of information. While uncertainties in gas price and carbon mitigation target require no specific hedging strategy over the business-as-usual case, hedging against uncertainty in growth requires large investments in energy technologies. Medium-term concerns at electric utility are modeled separately using the example of Tata Electric Company (TEC). Stochastic programming is used to incorporate uncertainty in water inflows. The model spans one year horizon. It captures major concerns of a utility like priority for irrigation, trade-off between storing water for hydro generation during high peak demand months and loss due to evaporation. Results show that about l0% improvement in annual utilization is possible by considering future water inflows in the plan. Additional improvements are possible over long run by changing the peak: base capacity mix. Hedging strategy for TEC is to store more water in pre-monsoon periods. Thus high inflow scenario has the highest value of information of Rs. 1.2 million over a year. This will require TEC to prepare for the export of electricity and prevention of spillage during later periods in case of occurrence of high inflows.