Optimization of Inertia-less Variable Renewable Energy Sources for Fast Frequency Response in Low Inertia Power Grids

Abstract

Globally, modern power grids have experienced an unprecedented increase in the integration of variable renewable energy sources (VRES) in recent years. In 2023, global VRES installations reached a record 462 GW (346 GW solar, 116 GW wind), while Africa added 1.686 GW (0.909 GW wind, 0.777 GW solar). Kenya notably contributed 0.102 GW of solar in 2022. Increased uptake of VRES replacing legacy synchronous generators (SGs) presents new challenges to grid operators in handling frequency stability and control. The displacement of SGs by VRES reduces grid inertia, increasing vulnerability to frequency instability, as even minor disturbances can cause rapid frequency deviations without adequate damping. Since SGs provide ancillary services to the grid such as frequency support, it is imperative that researchers explore the viability of VRES to provide these services. This research developed a climate-friendly, techno-economic approach to optimize the deloading requirements of inertia-less VRES for Fast Frequency Response (FFR), specifically targeting underfrequency events in power grids with declining inertia. To achieve the goal of this research, first, a dynamic model of a deloaded utility scale solar photovoltaic power plant (SPVPP) representing inertia-less VRES with FFR capability was developed using DIgSILENT Simulation Language (DSL). The SPVPP dynamic model was incorporated in a modified IEEE 39 bus system and extensive dynamic analysis performed to evaluate its frequency response performance following a generator outage. The loss of a major load typically leads to an over-frequency event, which was not considered in this study, as the focus was specifically on underfrequency scenarios caused by generation deficits—conditions that are more critical in low-inertia grids. The study specifically targeted the worst-case scenario for frequency stability by analyzing VRES that contribute no rotational inertia—a characteristic exhibited by SPVPP. Wind power plants, which possess extractable kinetic energy through rotating masses, were therefore excluded to maintain this strict focus on the most critical and challenging case. Next, a techno-economic and environmental multi-objective optimization problem was formulated to determine the minimum deloading level of SPVPPs. A python script was developed based on Particle Swarm Algorithm to solve the optimization problem formulated. To illustrate the efficacy of the proposed frequency regulation strategy, it was applied to a modified IEEE 9 bus and IEEE 39 bus test systems. The results indicated that FFR action from optimally deloaded SPVPPs improved post contingency frequency response dynamics. For instance, considering the modified IEEE 39 bus system, when the proposed strategy was compared to BESS performance, the rate of change of frequency results shows an improvement of 3.1% at 10% penetration and 8.2% at 30% penetration with optimal deloading levels of 1.00% and 4.79% respectively. In addition, optimization of the deloading levels of SPVPPs achieved more significant results with the modified IEEE 39 bus system than the IEEE 9 bus system due to its higher inherent inertia enabled by a greater number of SGs. The primary contribution of this work is the development of a dynamic SPVPP model with FFR capability using DSL. The model uniquely considers both the RoCoF and frequency nadir following a disturbance, enabling a more comprehensive and effective frequency regulation strategy. This advancement offers a pathway to integrate VRES more effectively without significant concerns over inertia and frequency stability. Among its key recommendations, this study highlights the need to enhance the role of VRES, such as solar PV and wind, in providing voltage regulation and reactive power support to bolster grid stability