Understanding Variability and Uncertainty in Energy Generation Portfolios Using SMART-Invest: A Stochastic Dynamic Planning Approach

Abstract

The objective of this thesis is to develop a precise model of variability and uncertainty that accurately captures the barriers to higher penetration of solar and wind power in the energy mix. The thesis will build upon the framework of SMART-Invest, the stochastic dynamic planning model introduced by Javad Khazaei and Warren Powell in 2015, with a particular focus on quantifying the e ects of solar variability and forecast uncertainty on the optimal energy generation portfolio. Using data from the PJM Interconnection, several classes of policy studies will be conducted under different scenarios of variability and uncertainty. Ultimately, the results of this analysis will be used to evaluate the reliability of current renewable energy forecasts and the plausibility of some of the country's energy policy objectives for the next 15 years.