Impact of Wind Energy on the Operation of Power Systems

This dissertation identifies and proposes mitigating solutions for some of the adverse economical and technical impacts of integrating wind generation into existing power systems. The main focus is on the short-term operation looking seconds to hours ahead, and on day-ahead scheduling looking 12 to 36 hours ahead.;The main innovative parts of this thesis develop a hybrid deterministic/probabilistic Unit Commitment which models the day-ahead prediction of wind power and demand as random quantities. This formulation guarantees that the scheduled reserve and generation levels meet the demand with a probability above a minimum threshold, while satisfying all conventional generation and transmission constraints. The new hybrid UC schedules less reserve than the conventional deterministic method while keeping its solution less computationally intensive than a formulation based on stochastic programming.;The case studies illustrate the effects of integrating wind generation in power systems dominated by nuclear and fossil fuel generation. The evolution of long-term economical and operational quantities such as costs, energy incremental costs, CO2 emissions, reserve levels, and wind curtailment are estimated for different levels of installed wind power by simulating the operation of a realistic power system over one year.;First a thorough review illustrates how wind power uncertainty affects the operation of power systems composed of conventional fossil, nuclear and hydro power sources. Then, the impact of wind power variability is studied in the short-term operations (with a time frame of less than an hour) by estimating: (1) The probability distribution of power systems frequency deviations caused by the residual demand variability; (2) The additional operational cost of following the intra-hour wind power variability through existing centralized Automatic Generation Control mechanisms.