A solid-state circuit breaker for DC systems must do more than detect an overcurrent and open a switch. Because DC current does not naturally pass through zero as AC current does, fault current can continue to rise until the source, circuit impedance or protection action limits it. The breaker must interrupt that current quickly while also managing losses during normal operation and the energy stored in the circuit’s inductance.
A prototype study from Aalborg University examined a coordinated semiconductor topology intended to balance these demands. Its central idea is to use different device types for different roles rather than relying on one semiconductor technology to provide both low-loss conduction and fast interruption. The work built and experimentally tested a low-voltage proof of concept, comparing it with an IGBT-based breaker in a similar power range.
Why DC microgrid protection is challenging
DC microgrids can connect distributed sources, energy storage and loads through a common bus. They may be used in systems that include photovoltaic generation, wind power conversion and DC loads. A fault on one branch can affect equipment beyond the fault location, so protection must detect abnormal current and isolate the affected circuit.
In an AC circuit, current crosses zero periodically, which can help an interrupting device extinguish an arc. DC has no equivalent natural zero crossing. Under a short circuit, current is instead limited by the source voltage and the resistance and inductance between the source and the fault. A low-impedance path can therefore produce a rapid current rise. The fault’s location and the connected sources also influence the current available to the protection device.
Protection choices involve trade-offs. Fuses are simple and have low conduction losses, but they are one-time devices and may not provide the required response across the full range of conditions. Mechanical breakers can be reused, but interruption can be slower and arcing is a concern. Solid-state breakers can act quickly without mechanical contacts, but semiconductor conduction losses and cost are important design considerations. Hybrid approaches combine mechanical and semiconductor devices, with their own complexity and response-time trade-offs.
How the solid-state circuit breaker topology works
The studied topology coordinates semiconductor devices with different operating characteristics. Thyristors provide a main current path with comparatively low on-state losses. A semiconductor transfer switch provides a means to change the current path when a fault is detected. The broader design also includes a secondary branch and an overvoltage protection element to handle the interruption transient.
During normal operation, the breaker conducts through its designated main path. When the control system detects a fault, it changes the state of the transfer switch so current moves away from the normal path and into the secondary branch. The sequence matters: the secondary path must be established as the main-path current falls. The thyristor cannot be turned off simply by removing its gate signal; it enters its blocking state only after its anode current reaches zero.
Once current has been interrupted, the energy stored in the line inductance still needs a path. A sudden interruption can produce an overvoltage as the inductor opposes the change in current. The design therefore uses a discharge or clamping path, such as a metal-oxide varistor (MOV), to absorb the resulting energy and limit voltage stress. The timing of the control signals and the capability of the energy-management components are both important to a successful interruption.
Managing losses and thermal limits
Reducing conduction losses is a key motivation for combining device types. Semiconductor losses depend on the current path and the characteristics of the conducting devices. For a MOSFET, conduction loss is related to its on-state resistance and the square of current. Thyristor and IGBT losses are commonly evaluated using their on-state voltage and current, with resistance terms also relevant to the device model.
These losses become heat, so electrical efficiency cannot be considered separately from thermal design. Junction temperature must remain within the device manufacturer’s limits in both steady-state operation and switching transients. Heatsinks and the thermal resistance between the semiconductor junction and ambient conditions help determine the temperature rise. Because device electrical characteristics vary with temperature, thermal assumptions also affect loss estimates.
For a breaker expected to conduct continuously, the steady-state current path deserves particular attention. Switching losses may be less significant when switching is infrequent, but conduction losses persist for as long as current flows. A design that lowers those losses can reduce the thermal burden; however, the overall result depends on the selected devices, current, operating temperature and circuit configuration. The study’s reported comparison supports the topology’s potential for lower losses than an IGBT breaker of similar power range, without establishing a universal result for other ratings or applications.
Control and validation considerations
Device coordination increases control complexity. MOSFETs and IGBTs are voltage-controlled and require suitable gate drive, while a thyristor requires gate current to turn on and cannot be switched off through its gate in the same way. High-side devices may also require electrical isolation between control and power circuits. Gate resistors influence switching behavior, and driver current limits must be respected.
A practical control strategy must therefore manage detection, gate-drive signals and the order in which current paths change state. It should also account for the fault current’s decay, the thyristor’s transition to blocking, and the voltage transient handled by the discharge branch. The study proposed a simple control strategy for its prototype, but the need to coordinate unlike devices remains a defining feature of this breaker approach.
The experimental work was framed as a low-voltage proof of concept, using a 16 V DC test system with a stated conduction power rating of 640 W and a maximum non-repetitive fault current of approximately 100 A. Those conditions define the scope of the prototype; they should not be read as ratings for a commercial or medium-voltage breaker. Validation at a different voltage, current or system impedance would require design-specific assessment.
Practical implications for system designers
For engineers and integrators evaluating DC protection, the study highlights several linked design questions:
- Fault behavior: Identify likely fault locations, source contributions and the impedance that limits current.
- Normal-operation losses: Assess the devices carrying continuous current, not only the interruption event.
- Thermal design: Check junction-temperature limits and size the heat-removal path using expected losses.
- Interruption sequence: Confirm that current is transferred and reduced in the correct order for each device type.
- Energy management: Provide a suitable path for inductive energy and control the resulting overvoltage.
Coordinating thyristors with semiconductor transfer and protection components offers a way to balance low-loss conduction with rapid, contactless interruption. The approach also places greater demands on gate-drive design, control sequencing, thermal management and transient protection. For plant and system designers, the practical lesson is to evaluate the entire protection path—from fault detection through current commutation and energy discharge—rather than selecting a breaker based on switching speed or device type alone.