Improving synthetic inertia provision by Power Electronic Interfaced Power Sources to support future system stability

Increasing penetration of power-electronics-converter-interfaced generation and loads raises several challenges for the operation, control and protection of power systems. This paper investigates the impact of high penetration of Power Electronic Interfaced Power Sources (PEIPS) on frequency control, and aspects of provision of synthetic inertia (SI) by PEIPS. Contrary to directly connected conventional synchronous machines, which provide inertia to the system inherently, thus effectively counteracting large gradients in the system frequency (rate of change of frequency, RoCoF), PEIPS need to be operated under specially designed control schemes in order to provide synthetic inertia by varying their power output proportionally to the RoCoF. Implementing such control schemes for current-controlled PEIPS requires both accurate and fast frequency measurement and can rather approximately reproduce physical inertia added to the system by directly connected synchronous machines. While the transient behavior of conventional synchronous machines is defined by their well-known electromechanical properties, the dynamic behavior of PEIPS is mostly dictated and limited by their control strategy and pertinent measurements. Under this setup, it is worth contemplating strategies that go beyond mimicking the dynamic capabilities of synchronous machines. A synchronous machine provides inertia to the system inherently due to its rotating mass. However, current-controlled PEIPS implement synthetic inertia by measuring frequency, calculating the frequency gradient and providing power accordingly. The delays due to frequency measurement, RoCoF computation and current control reflect an inherent difference and disadvantage of synthetic inertia compared to real inertia and must be properly studied to avoid deterioration of system stability under a large penetration of PEIPS. While the behavior of synchronous machines is determined by their physical properties, the behavior of PEIPS may be altered by modifying their control regime in order to improve their transient behavior under higher RoCoF values. Furthermore, it is possible to implement non-linear strategies that do not depend only on the frequency gradient but also on the frequency deviation. One option to achieve this is to implement different activation areas for different signs of both frequency deviation and RoCoF. Such a zone-selective implementation of synthetic inertia, realized by an activation function, could avoid that synthetic inertia counteracts the effect of other control reserves during recovery from frequency deviations, i.e., after the frequency nadir has occurred. This paper investigates such options and their impact on the frequency stability of the power system when synthetic inertia is provided by PEIPS.