UC Berkeley

Electricity is the most common way to transport, transform and consume energy. However, the total electricity generation accounts for 40% of worldwide CO2 emissions. With the threat of climate change, it is crucial to lowering emissions related to electricity generation and grid operation.

While renewable energy sources provide a solution to decarbonize the electric grid, they bring new challenges due to their inherent volatility. The grid requires balancing capacity to maintain a stable frequency and storage to adapt generation hours to consumption hours. Meanwhile, 54% of the world's electricity consumption happens in countries that rely on competitive markets for efficient dispatch [4]. Therefore, it is essential to provide correct incentives and dispatch mechanisms on the electricity markets to integrate renewable energies and develop new market mechanisms for balancing renewable energy variability.

This dissertation explores methods to integrate flexible resources into the electricity market. First, we develop an optimal bidding strategy for grid-scale storage on a wholesale electricity market. The specific role of the storage system creates market power which challenges the design of the bidding algorithm.

Secondly, the dissertation focuses on flexible resources located on the distribution grid. Indeed, the simplest way to lower overall CO2 emissions with a decarbonized grid is to electrify all energy use. This electrification entails the deployment of electric vehicles, electric heaters, or electric stoves. In addition, rooftop solar panels, household batteries, and other local energy resources are installed on the distribution grid, known as distributed energy resources (DERs). As a result, the distribution grid faces dramatic changes and requires precise monitoring and maintenance. Controlling DERs and integrating them into the electricity market could provide new flexibility to the grid. This dissertation introduces a new market dispatch model for the day-ahead wholesale electricity market where DERs are integrated as stochastic sources of flexibility.

However, the control and integration of DERs on the distribution grid rely on having a complete and accurate distribution grid model. Because the distribution grid lacks sensors and monitoring equipment, developing a detailed grid model is challenging for the system operator. Research on this subject is recent but provides diverse solutions dependent on modeling assumptions and data acquisition. In order to give future researchers a good understanding of existing methods and challenges left to tackle, this dissertation provides a detailed review and analysis of methods to estimate distribution grid topology.