Sizing of Energy Storage System for Virtual Inertia Emulation

Abstract

The infusion of renewable energy sources into the conventional synchronous generation system decreases the overall system inertia and negatively impacts the stability of its primary frequency response. The lowered inertia is due to the absence of inertia in some of the renewable energy-based systems. To maintain the stability of the system, we need to keep the frequency in the permissible limits and maintain low rotational inertia. Some authors in the literature have used the virtual synchronous generators (VSG) as a solution to this problem. Although the VSG based distributed recourses (DER) exhibits the characteristics and behavior of synchronous generators (SG) such as inertia, frequency droop functions and damping but it does not optimally solve the question of frequency stability. This paper presents a solution for these problems via an empirical model that sizes the Battery Energy Storage System (BESS) required for the inertia emulation and damping control. The tested system consists of a Photovoltaic (PV) based VSG that is connected to a 9-Bus grid and the simulation experiments are carried out using EMTP software. The VSG transient response is initiated by a symmetric fault on the grid side. Our simulations show the battery energy sizing required to emulate the virtual inertia corresponding to several design parameters, i.e., the droop gain, K <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">w</inf> , the droop coefficient, K <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">d</inf> , and the VSG time constant Ta. The simulation results show that to limit the rate of change of frequency (ROCOF) the battery needs to absorb a peak power of 0.57 pu and supply a peak power of 0.63 pu when the time constant is 4s, and the battery needs to absorb a peak power of 0.62 pu and supply a peak power of 0.69 pu when the time constant is 10s. Moreover, the findings suggest that to achieve a better ROCOF, K <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">d</inf> and T <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">a</inf> can be increased, however, this requires a larger battery rating. A better ROCOF can also be achieved by increasing the droop gain, K <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">w</inf> , without increasing the battery size.