Source-linked AI summary
Quantifying the Unintentional Islanding Risk: A Comparative Study on Active Distribution Network
Harshit Nayak, Edoardo Daccò, Jose Luis Rueda Torres, Davide Falabretti
TL;DR
Unintentional separation of part of a network makes the regulation mode of rotating machines remaining on the island especially consequential. The study uses electromagnetic-transient simulations of a real-network MV benchmark with synchronous generators and a grid-following inverter under different regulation modes. Fixed-setpoint islands clear in 12.5–21 seconds, while regulating synchronous generators sustain islanding indefinitely across investigated inertia values without reaching protection thresholds.
Problem
Unintentional separation of part of a network makes the regulation mode of rotating machines remaining on the island especially consequential.
Source: §I · p. 1Method
Results
Fixed-setpoint islands clear in 12.5–21 seconds, while regulating synchronous generators sustain islanding indefinitely across investigated inertia values without reaching protection thresholds.
Source: C. Results Summary and Protection… · p. 5Takeaways & Limitations
Active frequency and voltage regulation, rather than the specific rotating or power-electronic technology providing it, is the critical factor governing island persistence in the studied network.
Source: §V · p. 6
Abstract
from arXiv · showhide
The growing share of Inverter-Based Resources (IBRs) is changing the operation of modern Active Distribution Networks (ADNs), raising concerns for grid stability, protection, and reliability. In this paper, we investigate the dynamic behavior of a Medium-Voltage (MV) portion of an ADN during the transition from grid-connected to islanded operation. The study considers a Grid-Following (GFL) converter and Synchronous Generators (SGs) operated under two distinct regulation modes: a fixed-setpoint, non-regulating condition representative of present-day distribution networks, and an active frequency/voltage-regulating condition representative of SGs equipped with governor and AVR control. Detailed electromagnetic transient (EMT) models of both technologies have been developed in DIgSILENT PowerFactory that quantifies the sensitivity of island persistence to SG regulation mode and inertia. The results show that when DERs operate at a fixed power setpoint, the risk of forming a sustained, undetected electrical island is limited, since the isolated network drifts out of the protection thresholds within tens of seconds. Conversely, enabling frequency and voltage regulation on the SGs is sufficient to sustain an unintentional island indefinitely without triggering conventional protection, regardless of system inertia.
I. INTRODUCTION
The paper examines how synchronous-generator regulation affects the persistence of unintentional islands in active distribution networks, a question not systematically assessed in real-network studies against a grid-following inverter. It uses a real MV-network benchmark to quantify island survivability under different operating modes.
I. INTRODUCTION
- The paper compares IBRs, broadly divided into grid-forming and grid-following converters, with synchronous generators that can supply stronger short-circuit currents.The passage contrasts SG currents of five to seven times rated capacity with IBR contributions of little more than rated current.Source: §I · p. 1
- Unintentional separation is important because the regulation mode of rotating machines remaining on the island can affect its persistence.Persistent islands are discouraged because they can lead to asset deterioration, safety hazards, and out-of-synchronism automatic reclosing.Source: §I · p. 1
- Prior studies largely addressed anti-islanding schemes or intentional islands, and did not systematically quantify SG operating modes against a GFL inverter on a real distribution network.The paper identifies the transition to islanding with both regulating and non-regulating generators as a largely missing analysis.Source: §I · p. 1
- Fixed-setpoint operation leads to island collapse within 12.5–21 s, whereas active frequency and voltage regulation enables sustained islanded operation, independent of generation technology.The study uses a real network-validated MV benchmark.Source: §I · p. 1
- The study uses electromagnetic-transient simulations of a real Italian MV distribution network developed in DIgSILENT PowerFactory.The authors identify practical, network-validated tripping-time and persistence thresholds as a contribution.Source: §I · pp. 1–2
II. DYNAMIC MODELING OF GENERATION UNITS
The study develops calibrated dynamic models of synchronous generators and a grid-following inverter in DIgSILENT PowerFactory to compare their behavior under common operating scenarios. The SG model includes governor and AVR dynamics.
II. DYNAMIC MODELING OF GENERATION UNITS
- Both generation-unit models were calibrated against actual DERs deployed in real active distribution networks for comparison under identical operating scenarios.The models were implemented in DIgSILENT PowerFactory.Source: §II · p. 2
A. Synchronous Generator Modeling
- The SG model represents a diesel-driven synchronous generator with limits on stator current, rotor excitation, turbine torque, and synchronism stability.Its governor supports isochronous and droop operation, while its AVR uses static excitation and PI voltage regulation.Source: A. Synchronous Generator Modeling · p. 2
B. Grid Following Inverter Modeling
The grid-following inverter model combines a phase-locked loop with cascaded current and power control. Its outer controller derives active- and reactive-current references from frequency and voltage measurements.
B. Grid Following Inverter Modeling
- The GFL model combines a PLL with cascaded inner-current and outer-power control implemented using built-in and user-defined PowerFactory blocks.The inverter is represented by the built-in static-generator module, whose inputs are set by the custom regulators.Source: B. Grid Following Inverter Modeling · p. 2
1) Outer Power Control:
- The outer power controller uses frequency and voltage measurements to set active- and reactive-current references through respective droop laws.The direct-axis current reference follows active-power droop; the quadrature-axis reference follows reactive-power droop, with voltage components used to decouple the current channels.Source: 1) Outer Power Control · pp. 2–3
2) Inner Current Control:
The inner current controller transforms measured converter currents into the PLL-aligned synchronous dq frame, regulates current errors with two PI controllers, and sends the resulting voltage references to PWM modulation.
2) Inner Current Control:
- Measured converter currents are transformed into the synchronous dq reference frame using the grid-voltage angle obtained from the PLL.The transformed quantities are id = ir cos θ + ii sin θ and iq = −ir sin θ + ii cos θ.Source: 2) Inner Current Control · p. 3
- Two independent PI regulators process the errors between reference and measured dq currents to produce dq-frame voltage references.The current-controller gain and integrator time constant are Ki,c and Ti,c, respectively.Source: 2) Inner Current Control · p. 3
- The dq-frame voltage references are transformed back to the stationary frame using the same PLL angle and used as PWM modulation references.Source: 2) Inner Current Control · p. 3
III. METHODOLOGY: CASE STUDY AND TEST MATRIX
The case study uses a real 20 kV radial MV network in Southern Italy with synchronous generators and a GFL converter. At t = 2 s, the MV network is disconnected from the main grid, and island persistence is assessed against DER protection thresholds.
III. METHODOLOGY: CASE STUDY AND TEST MATRIX
- The case study is a 20 kV radial MV network in Southern Italy with two 0.45 MVA SGs, one 5 MVA GFL, and a 25 MVA HV/MV transformer.Loads are modeled as constant impedances whose active and reactive absorption vary with the square of voltage magnitude.Source: §III · p. 3
- At t = 2 s, the MV feeder switch opens, disconnecting the network from the main grid and leaving it supplied solely by DERs.The network begins grid-connected with the upstream circuit breaker closed.Source: §III · p. 3
- The island is assessed against frequency limits of 0.95–1.03 p.u. and voltage limits of 0.85–1.10 p.u.; DERs initially supply exactly the load demand.This pre-event power balance minimizes exchange with the upstream grid to represent a worst case for island persistence.Source: §III · p. 4
A. Test Matrix
Both scenarios use the same islanding event and the same GFL baseline, isolating the SG regulation mode as the variable under test.
A. Test Matrix
- Both scenarios apply an identical unintentional-islanding event to the same portion of the grid.The GFL provides a common baseline and does not participate in autonomous frequency or voltage regulation.Source: A. Test Matrix · p. 4
- Scenario I operates the SGs at a fixed active-power setpoint with constant field excitation.The GFL regulates active and reactive power in both scenarios.Source: A. Test Matrix · p. 4
- Scenario II activates the SG governor and AVR for active frequency and voltage regulation.SG regulation mode is the variable tested across the scenarios.Source: A. Test Matrix · p. 4
IV. DYNAMIC PERFORMANCE ASSESSMENT
The assessment uses network voltage and frequency responses to check compliance with DER protection thresholds and identify possible undetected islanding.
IV. DYNAMIC PERFORMANCE ASSESSMENT
- Each numerical simulation calculates network voltage and frequency to check compliance with DER protection thresholds and identify possible undetected islanding.The supplied passage introduces this evaluation procedure.Source: §IV · p. 4
- The assessment compares DER regulation functions and rated power across scenarios while applying the same protection-threshold check.The table caption identifies the matrix scope, and the procedure passage specifies the voltage and frequency check.Source: Table IV · p. 4, §IV · p. 4
- Table IV summarizes the regulation functions and rated power assigned to each DER technology in each scenario.It presents the test matrix for the assessment.Source: §IV · p. 4, Table IV · p. 4
A. Scenario I: GFL + SGs(At Fixed P & Constant Excitation)
With fixed SG setpoints, island dynamics depend on inertia and frequency eventually reaches the protection threshold. With SG frequency and voltage control enabled, all tested inertia cases remain within protection thresholds and sustain the island.
A. Scenario I: GFL + SGs(At Fixed P & Constant Excitation)
- In fixed-setpoint operation, the post-islanding transient settles within tens of milliseconds, and lower SG inertia produces larger, faster frequency excursions.The transient follows removal of the stiff grid reference; the SG rotor primarily absorbs the power imbalance because it operates without governor action.Source: A. Scenario I: GFL + SGs(At Fixed P &… · p. 4, Figure 4 · p. 5
- In fixed-setpoint operation, lower inertia causes faster frequency decline, while inertia above approximately 20 kg·m2 allows island persistence beyond 15 seconds, including the 4 second protection delay.Voltage remains above the 0.85 p.u. threshold across the investigated cases, so collapse is driven by frequency instability rather than voltage excursions.Source: A. Scenario I: GFL + SGs(At Fixed P &… · p. 4, Figure 4 · p. 5
B. Scenario II: GFL + SG (Performing Voltage and Frequency Control)
- All investigated inertia values with SG frequency and voltage control maintain both quantities within protection thresholds, resulting in sustained unintentional islanding.The SG control action balances the islanded network and prevents the excursions required for passive islanding detection.Source: B. Scenario II: GFL + SG (Performing… · p. 5, Figure 5 · p. 5
- With SG control enabled, the lowest-inertia case, J = 10.033 kg·m2, has frequency overshoot of approximately 1.033 p.u. and undershoot of 0.968 p.u.; higher inertia gives slower, more damped responses.The initial oscillations remain within applicable ride-through envelopes; the Fig. 5 caption identifies frequency and voltage responses for varying inertia.Source: B. Scenario II: GFL + SG (Performing… · p. 5, Figure 5 · p. 5
C. Results Summary and Protection Implications
The results distinguish finite island persistence under fixed-setpoint operation from indefinite persistence when synchronous generators regulate frequency and voltage. Active regulation also suppresses the deviations used by passive protection, motivating assessment of grid-forming converters.
C. Results Summary and Protection Implications
- Active frequency and voltage regulation keeps the island within protection thresholds, so conventional interface protection does not detect the sustained island.The reported maximum and minimum frequencies are 1.037 p.u. and 0.974 p.u., respectively.Source: C. Results Summary and Protection… · p. 5
- Scenario II persists indefinitely at every tested inertia, whereas Scenario I clears after 12.5–21 seconds as inertia rises.Table V summarizes the dynamic indicators for both scenarios.Source: C. Results Summary and Protection… · p. 5, Table V · p. 6
- SG governor and AVR regulation acts over several seconds yet sustains the island; faster grid-forming converter regulation is expected to have similar or stronger island-sustaining capability.The paper identifies grid-forming resources as a case for dedicated assessment.Source: C. Results Summary and Protection… · p. 6
V. CONCLUSION AND FUTURE WORK
The study finds that fixed-setpoint DERs do not sustain the island indefinitely, while active SG frequency and voltage regulation does so regardless of inertia. It identifies active regulation as the key persistence factor and outlines further grid-forming and network-sensitivity studies.
V. CONCLUSION AND FUTURE WORK
- Fixed-setpoint DER operation leads the island to drift beyond protection thresholds within tens of seconds, while active SG frequency and voltage regulation sustains it indefinitely regardless of inertia.The study analyzes a GFL converter and SGs in a real MV distribution network using detailed EMT models implemented in DIgSILENT PowerFactory.Source: §V · p. 6
- Active regulation, rather than whether the regulating resource is rotating or power-electronic, is identified as the critical factor governing island persistence.The paper expects grid-forming converters to pose equal or greater risk than the regulating-SG case, but leaves them for future assessment.Source: §V · p. 6
- Future work will assess grid-forming converters against SG and GFL baselines, test network-condition sensitivity, and pursue hardware-in-the-loop validation and techno-economic assessment.The planned sensitivity studies include load composition, DER penetration, and topology.Source: §V · p. 6