Essays on investment planning in electricity generating capacity
Hartley, Peter R.
Doctor of Philosophy
In the first part of this study we develop and analyze two mathematical models that incorporate a time changing demand for electricity and uncertainty of input prices. The first model highlights the shortcomings in assuming a constant plant utilization under uncertainty of input prices and the effects of such assumption on the optimal investment in electricity generating capacity in a simple two period model. The second model presents sufficient restrictions to the optimal investment in electricity generating capacity problem to allow for a recursive solution. The necessary restrictions are extremely limiting to the extend that we found a solution for very simple scenarios. In our opinion, the problem is better handled in a case by case basis rather than under a general dynamic framework. Following the spirit of our conclusions of the first part of our study, in the second part we provide a methodology to simulate long-term natural gas prices, we analyze the investment prospects of nuclear and natural gas generating capacity in Mexico and provide a constraint approach for the optimal generation of hydroelectric plants in the Mexican hydroelectric system. These three problems belong to the solution of the optimal investment in electricity generating capacity in Mexico. To simulate the uncertainty of natural gas prices, we assume that natural gas prices are the sum of two stochastic processes: short-term and long-term variability. We characterize the short-term variability of natural gas prices using an Exponential General Autoregressive Conditional Heteroskedastic (EGARCH) model. The uncertainty of the long-term variability of natural gas prices is based on the long-term natural gas prices scenarios of the National Energy Modeling System of the Energy Information Administration. Equipped with a methodology to simulate long-term natural gas prices, we investigate the investment prospects of nuclear and natural gas generating capacity in Mexico using the levelized cost methodology. Finally, we derive an algorithmic solution for a constraint version of the optimal generation of hydroelectric plants, then we provide a guide for its application to the Mexican hydroelectric system.
Energy; Economics; Applied sciences