This thesis explores the potential risk implications that a large penetration of intermittent renewable electricity generation -- such as wind and solar power -- may have on the future electricity generation technology mix, focusing on the anticipated new operating conditions of different thermal generating technologies and their remuneration in a competitive market environment. In addition, this thesis illustrates with an example how risk should be valued at the power plant level in order to internalize the potential risks to which the generators are exposed.

This thesis first compares the impacts of three different bidding rules on wholesale prices and on the remuneration of units in power systems with a significant share of renewable generation. The effects of bidding rules are distinguished from the effects of regulatory uncertainty that can unexpectedly increase renewable generation by considering two distinct situations: 1) an 'adapted' capacity mix, which is optimized for any given amount of renewable penetration, and 2) a 'non-adapted' capacity mix, which is optimized for zero renewable penetration, but that operates with a certain non-zero renewable capacity, added on top of an already adequate system. The analysis performed stresses the importance of sound mechanisms that allow the full-cost recovery of plants in a system where the intermittency of renewables accentuates nonconvex costs, without over-increasing the cost paid by consumers for electricity. Additionally, the analysis quantifies the potential losses incurred by different thermal technologies if renewable deployment occurs without allowing for adaptation.

Methodologically, this thesis uses a novel long-term generation investment model, the Investment Model for Renewable Electricity Systems (IMRES), to determine the minimum cost thermal capacity mix necessary to complement renewable generation to meet electricity demand, and to extract hourly wholesale prices. IMRES is a capacity expansion model with unit commitment constraints whose main characteristics are: 1) reflecting the impact of hourly resolution operation constraints on investment decisions and on total generation cost; 2) accounting for the chronological variability of demand and renewable output, and the correlation between the two; and 3) deciding on power plant investments at the individual 5 plant level. These characteristics allow for a detailed analysis of the profits obtained by individual plants in systems with large renewable penetration levels. In addition, this thesis tests the performance of a heuristic method that selects four weeks from a full year series to optimally represent the net load duration curve (i.e., the difference between demand and renewable output, decreasingly ordered). For each application of this heuristic method, three metrics are proposed to reflect that the approximation also represents the chronological variability of the net load.

Lastly, this thesis explores the role of risk in the valuation of electricity generating technologies and shows how to incorporate standard risk pricing principles into the popular Monte Carlo simulation analysis. The exposition is structured using the standard framework for a typical Monte Carlo cash flow simulation so that the implementation can be readily generalized. This framework stresses the necessity of an asset pricing approach to assess the relationship between the risk in the assets cash flows and the macroeconomic risk with which the financial investors are concerned. The framework provided is flexible and can accommodate many different structures for the interaction between the macroeconomic risk and the risks in the asset's cash flows (such as those from shocks in renewable deployment).