Analysis of investment decision making in power systems under environmental regulations and uncertainties

The dissertation focuses on the study of environmental policies and their impacts on the power systems’ planning. It consists of three parts, each of which addresses a single problem on environmental policies and generation expansion planning. In the first part of the dissertation, I compared the capand-trade policy and various carbon tax policies in a single period under the generation expansion framework. The problem was modeled as a bilevel problem where the lower level competing generation companies maximized their own profits under the regulations of the upper level. The policies were compared via their effectiveness and efficiency. Effectiveness referred to a policy’s capability to control the amount of carbon emissions, and efficiency was measured with respect to five criteria: emissions price, renewable energy portfolio, total generation, total profit of generation companies and grid owner, and government revenue. In the second part, the model was extended to multi-period planning to gain better views into market dynamics. Cap-and-trade and four variations of carbon tax policies were integrated in a game-theory based generation expansion planning model to assess their impacts on new investments in renewable energy generation capacity. The most efficient tax policy and variations were obtained using inverse equilibrium models. The third part complemented the previous parts by conducting a realistic case study on the generation expansion planning under uncertainty. It studied the formulation and solution of investment decisions in new generation under the explicit representation of environmental policies and their associated uncertainty. A three-layer framework was proposed to study the investment decisions. The operations layer was used to represent the transmission physical flows under economic dispatch in the network; the assessment layer completed comprehensive assessment of candidate investment plans under uncertainty; the optimization layer was designed to compare the optimal investment decisions for the decision makers based on the optimization criteria. Our framework was tested on a realistic 240-bus WECC network, taking into account representative scenarios and investment plans.