DescriptionThe current types of rate structures for the electricity retailing create a disconnection between retail and wholesale markets. The rate structures expose utility providers to the full risks of wholesale markets, while prohibiting the response of end-use customers to market dynamics. This disconnection is considered to be the primary cause of unusual volatility in electricity markets. In this dissertation, the author proposes a Cost-for-Deviation (CfD) retail pricing scheme, which is designed to minimize the demand uncertainty of individual customers. A series of experiments demonstrate that CfD pricing is able to reduce the demand uncertainty by 10%, as measured by the root mean squared deviation of the demand. Consequently, the community’s cost of hedging the quantity risk in the real-time market is reduced by 38%. This dissertation also demonstrates the formulation and solution techniques for the day-ahead planning and real-time tracking optimizations that each customer faces under CfD. An efficient and robust approach, called Optimal Strategy Pool (OSP), is introduced to solve simulation-based on-line planning problems; and dynamic programming is adopted in neural network model-based predictive control. Both centralized and distributed mechanisms are studied for customers to reduce CfD charges via collaborative demand management. Overall, CfD pricing effectively reduces the demand uncertainty by promoting a predictable consumption behavior.