The purpose of this study is to develop a strategy for investment in power generation technologies in the future given the uncertainties in climate policy and fuel prices. First, such studies are commonly conducted using deterministic methods which assume a given likelihood of the carbon and gas price levels. In this study a probabilistic approach is used to address these uncertainties. Secondly, capacity expansion models conventionally apply average estimates to predict the amount of power that each generator will produce based on the technology chosen. I propose an alternate method which determines the actual generation hour-by-hour of a generator. Using this method, I also capture the variability of wind generation across the year.

To accomplish this goal, I used the Electric Reliability Council of Texas (ERCOT) as a case study. I investigated the effect of different scenarios of generation technology investments projected over a period of twenty years. I conducted two sets of analyses; first assuming that Carbon Capture and Storage (CCS) technologies will be available after 2020, then assuming that they will not. Using a dispatch model, I simulated the hours of a load duration curve for 2020 and 2030. In the first period 2010-2020, I assumed the price of carbon to either be $0 or $50/ton CO2. In the second period, I take the carbon price to be at either a low of $25/ton of CO2 or a high of $100/ton of CO2. The price of natural gas used was either a high of $15/MMBtu or a low of $3MMBtu in both periods. Using a Monte Carlo, I sample the wind generation based on the season and the time of dat. The system costs with the new investment scenarios were then evaluated in a decision tree to establish the socially optimal solution.

I find that the optimal strategy to be taken today depends on the availability of CCS technologies in 2030. Assuming that there is CCS in 2030, the more dominant strategy would be to build natural gas generators today. If we assume that there is no CCS in 2030, the strategy would depend on the probabilities of the levels of gas and carbon prices in 2020.