Impacts of smart grid functions on load-side frequency
University of New Brunswick
Several power systems have made notable progress in implementing smart grid functions as means to achieve load-side control activities. These new control approaches have shown promising potentials to reshape and optimize patterns of energy supply and demand in industrial, commercial, and residential loads. Demand response, peakload management, and direct load control are recognized as smart grid functions that aim to improve the power system efficiency, system reliability, and increase levels of sustainable energy generation and distribution. Several research directions and new business models have been introduced to facilitate and improve the implementation and deployment of smart grid functions in power systems. In general, the fundamental objective of smart grid functions is to regulate the electric power delivered to a set of loads or a set of load centers, where the principal focus is on the active power. Such an objective can introduce new challenges for the operation, control, dynamic conditions, and stability of the host power system. These challenges are centered around the frequency dynamics and margins of frequency stability. As a result, new methods and procedures are sought to carry out accurate assessment of possible impacts from smart grid functions on frequency dynamics and frequency stability margins. The successful development of such methods and procedures will allow setting stable ranges for the load-side control activities, which are associated with smart grid functions. This thesis aims to develop, implement, and test a new method for estimating the frequency changes at load buses that are subject to smart grid functions. The estimation of frequency changes is sought to provide stable ranges for the regulation of load power demands, as carried out by smart grid functions. The developed method is based on utilizing the ZIP model that is capable of providing a relationship between load power demands and frequency changes. Such a relationship can be formulated as a set of nonlinear equations, which can be numerically solved to estimate the frequency changes. The employment of the ZIP model is justified by the implementation of smart grid functions, which require the use of load models. The developed loadmodel based method is implemented for performance evaluation using the Barbados power system, which has 118 buses. The developed method is demonstrated to be accurate, simple, and insensitive to durations and/or magnitudes of power demand changes.