Application of energy storage for frequency response

Abstract

With more integration of renewable energy sources such as wind turbines and photovoltaic cells, current power system is experiencing a significant change both in control, operation and planning. The power output of renewable energy sources is susceptible to the environment, and thus, it has the characteristics of variability and uncertainty. The intermittent and uncertain nature of renewable energy sources causes frequent power change and it will make the synchronous generators difficult to catch up. This means that keeping power balance between the generation and load demand is more challenging than before. On the other hand, the integration level of renewable energy sources is still growing and is expected to even replace the power plants based on fossil fuels. This makes that inertia and damping will decrease continuously. When the power imbalance happens, frequency maybe undergoes large oscillations in the transient process. If the traditional control strategies can not stabilize the frequency timely, relay protection devices will be activated undesirably to prevent most equipments from damages.

Energy storage is kind of devices that can store energy at one time and output it at another time. This characteristic makes that energy storage can help synchronous generators to catch up the load by absorbing energy when there is power surplus and output energy when the load level is high. Moreover, the response speed of energy storage is extremely high and energy storage can output power almost immediately to its power capacity. This characteristic makes that energy storage can participate in frequency control. However, energy storage cost is still not cheap, how to place and size the limited energy storage to participate in power balance as well as frequency control becomes an important problem. A framework for the planning of energy storage considering the frequency constraints is proposed in this thesis. RoCoF and frequency nadir are constrained in this model, and the optimal locations and the corresponding capacities of energy storage can be obtained. On the other hand, energy storage can mimic the control strategy of synchronous generators to participate in the frequency control through emulating virtual inertia and damping. Considering that energy storage will be widely deployed in future power system, another framework is built in this thesis to give a solution to the optimal allocation of virtual inertia and damping for energy storage based on their locations, power input or output and state of charge. Through this framework, the appropriate energy storage can be selected to provide suitable virtual inertia and damping.