Frequency Stability and Resilience Analysis of Grid-Forming and Grid-Following Inverters with a Partial Power Boost Converter Under Grid Strength Variation in PV-Integrated 22 kV Power Distribution Systems in Papua New Guinea

Abstract

Modern power distribution systems face increasing challenges in maintaining frequency stability and system resilience in power systems dominated by inverter-based renewable energy sources, which have higher shares of power electronics as they transition from conventional synchronous generators. This challenge is particularly relevant in power distribution networks with low inertia and variable grid strength, such as those seen in regions transitioning from conventional to renewable systems. For instance, Papua New Guinea (PNG) faces these operational challenges due to its increasing share of inverter-based resources (IBR) and weak grid infrastructure. This research addresses these challenges by comparing two inverter control strategies: enhanced PLL-based grid-following (GFL) control and droop-based grid-forming (GFM) control, and examining how these controls respond to changing grid conditions to maintain frequency stability and overall system resilience in a photovoltaic (PV)-integrated 22kV distribution network.

A manual grid strength variation approach was employed by adjusting the resistance (R) and inductance (L) of the distribution line to reflect different grid strength conditions. Unlike conventional short-circuit ratio (SCR) analysis, which assumes a fixed grid impedance, this study also varies the distribution feeder’s resistance-to-reactance (R/X) ratio by manually tuning resistance and reactance values during grid strength testing. This ensures the model captures realistic frequency behaviour and system resilience of the inverter controls, reflecting practical conditions in varying and mixed urban-rural networks, such as those in PNG. This method enables the system’s SCR to shift dynamically among strong grid (SCR = 15), weak grid (SCR = 5), very weak grid (SCR = 2), and islanded conditions, replicating real-world scenarios such as sudden grid faults, reconnections, and transitions between grid-connected and islanded modes. A 2.50 MW PV system was modelled in MATLAB/Simulink, incorporating a partial power converter (PPC) boost stage with maximum power point tracking (MPPT) and DC-link voltage regulation. The PV farm was driven by real irradiance and temperature data for the entire year 2024, collected for Port Moresby, PNG, ensuring that solar resources variability was faithfully represented in the system’s dynamic responses. The simulation was tested under full-load and half-load conditions, with a 10% voltage dip, to observe the behavior of the two inverter controls.

The simulation results confirm that the grid strength variation method and the tests employed in this study effectively reveal the difference in how each inverter control responds to real disturbances. While all inverter modes were under strong grid (SCR=15) conditions, the droop-based GFMI consistently maintained frequency deviations within ±0.03 Hz and voltage deviations within acceptable limits during the most severe transitions, including full islanding and resynchronization under very weak grid (SCR=2) conditions. In contrast, enhanced GFLI (EGFLI) models exhibited frequency oscillations, prolonged settling times, and phase imbalance. The PPC architecture further enhances system resilience by stabilizing the DC-link voltage and ensuring that the PV system maintains a consistent power output, even in varying local weather conditions.

The detailed analysis of dq-axis voltage and current waveforms, active and reactive power sharing, and load-side responses demonstrates that the GFMI control not only withstands abrupt breaker operations, islanded operations, and voltage sag, but also rapidly restores synchronism when the grid is reconnected. This confirms that the system can withstand faults, maintain waveform symmetry, suppress harmonics, and recover power quality under real solar input conditions.

These findings validate that the grid strength variation method offers a practical and rigorous way to test resilience, bridging the gap left by static SCR studies, and demonstrate that droop-based GFMI provides a robust pathway towards stable, low-inertia grids that do not rely on conventional generation. This research offers clear and practical insights for engineers, planners, and policymakers seeking to deploy advanced inverter technologies and integrate renewable energy reliably into fragile distribution networks, such as those in PNG.