⚡ electrical engineering

Performance of Differential Protection Applied to Collector Cables of Offshore Wind Farms with MMC-HVDC Transmission

This paper investigates the limitations of conventional differential protection for offshore wind farm collector cables connected to MMC-HVDC systems and proposes enhanced strategies based on sequence components, validated through PSCAD/EMTDC simulations to address the unique fault current characteristics of inverter-based resources.

Original authors: Moisés J. B. B. Davi, Felipe V. Lopes, Vinícius A. Lacerda, Mário Oleskovicz, Oriol Gomis-Bellmunt

Published 2026-01-26

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Moisés J. B. B. Davi, Felipe V. Lopes, Vinícius A. Lacerda, Mário Oleskovicz, Oriol Gomis-Bellmunt

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: A New Kind of Power Plant

Imagine a massive offshore wind farm. Instead of sending electricity through a long, traditional wire to the shore, it uses a high-tech "super-cable" (called MMC-HVDC) to send power over long distances.

In the past, these wind farms were connected to the main grid like a standard power plant. But now, they are connected to a "smart" grid where everything is controlled by computers and inverters (devices that change electricity types). The paper asks a simple question: If a short circuit (a "fault") happens inside the wind farm's internal cables, will the safety alarms (protection systems) still work correctly?

The Safety Alarm: The "Differential" Guard

To understand the problem, imagine a security guard standing at the entrance and exit of a hallway (the cable).

Normal Day: The guard counts people entering and leaving. If 10 people enter and 10 leave, the hallway is safe. The guard does nothing.
The Alarm (Differential Protection): If someone trips and falls inside the hallway, the number of people leaving won't match the number entering. The guard sees a "difference" and sounds the alarm to clear the hallway.

This paper tests how well this "guard" works when the people (electricity) are being controlled by very strict, fast-acting computers (Inverter-Based Resources).

The Problem: The "Silent" Fault

The researchers found a major glitch. In these modern wind farms, the computers controlling the wind turbines and the main converter are programmed to be very careful. If they sense a problem, they actively try to stop the flow of electricity to protect the equipment.

The Analogy:
Imagine the security guard is trying to count people, but the people inside the hallway are told, "If you see a fire, freeze and don't move."

If a fire starts (a fault), the people freeze.
The guard at the entrance sees 0 people entering.
The guard at the exit sees 0 people leaving.
The guard thinks, "Everything is fine! 0 minus 0 is 0."
Result: The guard doesn't sound the alarm, even though there is a fire. The safety system fails because the "difference" it looks for disappears.

What the Researchers Tested

The team built a detailed computer simulation of a real wind farm in Brazil and Spain. They tested two different "personality types" for the computer controllers:

Controller Type 1 (The "Strict" One): This controller tries to completely stop any weird, unbalanced electricity flows (called "negative-sequence" currents) during a fault.
- Result: The safety guard went blind. For many types of faults (like a wire touching the ground or two wires touching each other), the guard saw no difference and failed to trip. The cable remained unprotected.
Controller Type 2 (The "Flexible" One): This controller allows some of that weird, unbalanced electricity to flow during a fault.
- Result: The safety guard could see the difference again. The alarm worked correctly, and the fault was cleared.

The "Magic" of Zero-Sequence

The paper also looked at a special type of electricity flow called "zero-sequence" (which happens when a wire touches the ground).

They found that if the transformers (the devices stepping up/down voltage) are wired a certain way, this "zero-sequence" flow can act as a backup signal.
Even if the main guard is confused, this backup signal can sometimes wake the alarm up. However, this only works if the "Strict" controller isn't too aggressive and if the transformers are wired correctly.

The Good News: No False Alarms

The researchers also tested what happens if a problem occurs outside the hallway (in the rest of the grid).

Result: The guard was very smart. Even with the confusing computer controls, the guard correctly realized, "That problem is not in my hallway," and did not sound a false alarm. The system remained "selective" (it only trips when it should).

The Bottom Line

The paper concludes that as we move to more computer-controlled power grids, the old safety rules might not work.

The Risk: If the computers are too good at suppressing electricity flows during a crash, the safety alarms might not see the crash happening.
The Lesson: Engineers cannot just rely on the standard "difference" alarm. They need to look at different types of electrical signals (like the "negative" and "zero" flows) and ensure the computer controllers don't suppress the very signals the alarms need to see.

In short: You can't have a safety system that relies on seeing a difference if the computers are programmed to make everything look exactly the same during an emergency.

Technical Summary: Performance of Differential Protection Applied to Collector Cables of Offshore Wind Farms with MMC-HVDC Transmission

Problem Statement
The global transition toward low-carbon energy has accelerated the integration of offshore wind farms connected via Modular Multilevel Converter-based High-Voltage Direct Current (MMC-HVDC) transmission. A critical protection challenge arises in the collector cables connecting wind turbines to the Offshore MMC (OMMC). Unlike traditional grids, both terminals of these cables are supplied by Inverter-Based Resources (IBRs). The control strategies of these IBRs actively limit fault current magnitudes and alter their characteristics, deviating significantly from the behavior assumed in conventional protection designs. Specifically, many control strategies suppress negative-sequence currents during faults. This suppression can result in near-zero fault currents for certain fault types, potentially rendering conventional differential protection schemes insensitive and unable to reliably discriminate between internal and external faults. While existing literature acknowledges these limitations, it often explores restricted fault scenarios that do not fully represent the diversity of fault types and resistance levels encountered in practice.

Methodology
To address these gaps, the authors conducted a comprehensive performance evaluation using electromagnetic transient simulations in PSCAD/EMTDC. The study utilized a realistic model of an offshore wind farm consisting of two 450 MVA clusters of full-converter (Type 4) wind turbines connected to an OMMC via 230 kV collector cables.

The methodology involved:

Protection Schemes: A comparative analysis of three differential protection strategies:
1. Conventional Phase Differential (87L): Based on phase currents.
2. Negative-Sequence Differential (87Q): Based on negative-sequence components.
3. Zero-Sequence Differential (87G): Based on zero-sequence components.
  All schemes employed percentage restraint logic with a 50% slope and a 0.3 p.u. minimum pickup.
Control Scenarios: Two distinct OMMC control strategies were evaluated:
- Control C1: Active suppression of negative-sequence current during faults.
- Control C2: Prioritization of negative-sequence current injection (no forced suppression).
Fault Scenarios: The study simulated internal faults at various locations (0% to 100% of the cable length) and external faults. Fault types included single-phase-to-ground (AG), phase-to-phase (AB), phase-to-phase-to-ground (ABG), and three-phase (ABC) faults. A wide range of fault resistances was applied (0 to 50 Ω depending on the fault type).
Analysis Metric: Performance was assessed by plotting the operating current ( $I_{OP}$ ) versus the restraint current ( $I_{REST}$ ) 50 ms after fault inception to determine if the relay operated within the defined security region.

Key Results
The simulation results highlighted significant vulnerabilities in conventional protection when specific control strategies are employed:

Impact of Negative-Sequence Suppression (Control C1):
- Phase-to-Ground (AG) and Phase-to-Phase (AB) Faults: When the OMMC and wind turbines both suppress negative-sequence currents (Control C1), the negative-sequence circuit effectively becomes an open circuit. Consequently, the fault current is practically null. Under these conditions, neither conventional phase differential (87L) nor sequence-based differential (87Q, 87G) protections demonstrated sufficient sensitivity to operate for internal faults, even at low fault resistances.
- Phase-to-Phase-to-Ground (ABG) Faults: The presence of a zero-sequence path (dependent on transformer winding configurations) allowed for non-zero fault currents. However, even with this path, the 87Q element performed unsatisfactorily, and phase elements (87L) only operated for low-resistance faults. Only the 87G element showed sensitivity across all resistance levels, though its performance was still heavily dependent on the availability of a zero-sequence path.
- Three-Phase (ABC) Faults: Since these faults involve only the positive-sequence circuit, the conventional phase differential protection operated satisfactorily regardless of the control strategy or fault resistance.
Impact of Non-Suppression (Control C2):
- When the OMMC did not suppress negative-sequence currents (Control C2), all evaluated differential elements (87L, 87Q, and 87G) operated with proper sensitivity and selectivity for all internal fault types and resistance levels.
Selectivity:
- For external faults, the operating quantities for all differential schemes remained practically null across all control strategies and fault types. This indicates that the schemes maintained high selectivity and did not exhibit spurious tripping for external faults, assuming ideal synchronization and measurement conditions.

Significance and Claims
The paper claims to provide a deeper conceptual understanding of protection challenges in converter-dominated grids. Its primary contributions are:

Identification of Critical Vulnerability: The study demonstrates that conventional differential protection, historically considered highly sensitive and reliable, may fail to detect internal faults in offshore collector cables if the connected IBRs (both wind turbines and OMMC) employ negative-sequence current suppression strategies.
Limitations of Sequence Components: While negative- and zero-sequence differential elements offer enhanced sensitivity in some contexts, they are not a universal solution. Their effectiveness is strictly contingent on the control strategies of the converters and the transformer winding configurations (specifically the existence of a zero-sequence path).
Need for Innovation: The findings suggest that as power systems transition to converter-dominated topologies, relying solely on conventional relay functions—even differential ones—may be insufficient. The authors argue for the development of innovative protection approaches and future-proofed standards that account for the specific fault current characteristics imposed by IBR control strategies.

The paper concludes that ensuring sensitive and selective protection for these critical transmission paths requires moving beyond traditional assumptions about fault current behavior, particularly in scenarios where negative-sequence currents are actively suppressed.