Operation and Management of Thermostatically Controlled Loads for Providing Regulation Services to Power Grid

The notion of demand-side participation in power systems operation and control is on the verge of realization because of the advancement in the required technologies an tools like communications, smart meters, sensor networks, large data management techniques, large scale optimization method, etc. Therefore, demand-response (DR) programs can be one of the prosperous solutions to accommodate part of the increasing demand for load balancing services which is brought about by the high penetration of intermittent renewable energies in power systems. This dissertation studies different aspects of the DR programs that utilized the thermostatically controlled loads (TCLs) to provide load balancing services. The importance of TCLs among the other loads lie on their flexibility in power consumption pattern while the customer/end-user comfort is not (or minimally) impacted.;Chapter 2 discussed a previously presented direct load control (DLC) to control the power consumption of aggregated TCLs. The DLC method performs a power tracking control and based on central approach where a central controller broadcasts the control command to the dispersed TCLs to toggle them on/off. The central controller receives measurement feedback from the TCLs once per couple of minutes to run a successful forecast process. The performance evaluation criteria to evaluate the load balancing service provided by the TCLs are presented. The results are discussed under different scenarios and situation. The numerical results show the proper performance of the DLC method. This DLC method is used as the control method in all the studies in this dissertation.;Chapter 3 presents performance improvements for the original method in Chapter 2 by communicating two more pieces of information called forecast parameters (FPs). Communicating improves the forecast process in the DLC and hence, both performance accuracy and the amount of tear-and-wear imposed on the TCLs.;Chapter 4 formulates a stochastic optimization model for a load aggregator (LA) to participate in the performance-based regulation markets (PBRM). PBRMs are the recently developed and practiced regulation market structure recommended by Federal Energy Regulatory Commission (FERC) in 2011. In PBRMs, regulation resources are paid based on both regulation capacity bids and the regulation performance including the provided mileage and the performance accuracy. In order to develop the income from the PBRM, the convention of California Independent System Operator (CAISO) is used. In the presented optimization model, the amount of tear-and-wear imposed on the TCLs are confined to prevent abrupt switching of TCLs.;In Chapter 5, a two-stage reward allocation mechanism is developed for a LA recruiting TCLs for regulation service provision. The mechanism helps the LA to distribute the total reward (earned from regulation service provision) among the TCLs according to their contribution in the whole provided service. In the first stage, TCLs are prioritized based on their service provision capability. In order to do so, an index called SPCI is presented to quantify TCLs capability/flexibility and therefore, prioritize them. After prioritization TCLs a priority list is constructed in the first stage. In the second stage, a reward curve is constructed representing the functionality of the possible total reward with respect to the number top TCLs in the priority list. Then, the allocated reward to individual TCLs is calculated by applying the incremental method on the constructed reward curve. This presented reward allocation mechanism is based on the definition of maximum service capacity (MSC) for a control group including TCLs. MSC is defined and its calculation method is presented before discussing the two stages of the reward allocation mechanism. The numerical results proves the suitability of the proposed prioritization method as it is observed the TCLs with higher rankings can contribute more to the total reward in comparison to the TCLs with lower rankings in the priority list.