Short answer: a battery breaker is not a PV breaker with a different label. Battery current flows in both directions — charging and discharging — while a conventional PV string breaker is built for one direction, so the wrong device can fail to interrupt a reverse fault. On top of that, a DC rating is not a single number: on the CHINT NB1-63DC, one pole gives DC 250 V, two poles give DC 500 V and four poles give DC 1000 V, with Icu 6 kA, because the poles are wired in series to split the arc voltage. Order by the pole count that matches your bank voltage, and confirm polarity behaviour on the exact code.
We are an independent trading company — not a manufacturer, and not an appointed distributor of any brand — so the ratings here are quoted per exact ordering code, and the datasheet for the code you buy is what governs.
Why DC from a battery is harder than DC from a string
A PV string’s short-circuit current is limited by the modules themselves, roughly 1.25 times the operating current, and it collapses when the array is shaded. A battery bank is the opposite: a low-impedance source that delivers a very large current into a fault and keeps delivering it. The breaker is therefore asked to interrupt a far more energetic arc, on a circuit whose inductance depends on the cabling and the inverter’s DC link.
Arc extinction in DC is done by stretching and cooling the arc with a magnetic blow-out field plus splitter plates. That is why DC devices are deeper than AC devices of the same rating, why the voltage rating is quoted per pole, and why the number of poles in series is a design decision.
| Property | PV string side | Battery bank side |
|---|---|---|
| Fault current source | Limited by the module Isc, self-limiting under shade | Low-impedance bank, high prospective fault current held for as long as the bank is charged |
| Current direction | One direction only, array to inverter | Both directions, charge and discharge, plus reverse current between parallel racks |
| Voltage during the day | Rises with irradiance and falls with temperature | Sags under load, recovers at rest, and can stay high after the inverter is off |
| Device requirement | DC-rated breaker matched to the string voltage, often a polarised design | DC-rated, non-polarised / bidirectional device rated for the bank voltage and the prospective fault current |
| Isolation for service | String isolator, plus the combiner arrangement | Rack-level isolation with a lockable, visible break, because the bank cannot be switched off at the source |
Bidirectional, not just “DC rated”
This is the single point where BESS projects get caught. A polarised DC breaker relies on current flowing in the marked direction so that the magnetic blow-out field pushes the arc into the splitter plates. Wire it so that fault current arrives from the other end, and the field pushes the arc the wrong way: the contacts may not clear, and the device that was supposed to protect the cabinet becomes the fault.
In a PV system the current direction is fixed by physics, so a polarised device is acceptable when installed the right way round. In a battery system the inverter both charges and discharges, and a rack can be back-fed from the DC bus, so the device has to interrupt in both directions.
The NB1-63DC pole and voltage matrix
| Poles in series | DC rated voltage | Typical bank configuration | Notes from the range data |
|---|---|---|---|
| 1P | DC 250 V | 48 V telecom or small UPS strings, 110 V DC control supplies | Icu 6 kA at the stated voltage; electrical life above 1500 operations; IP20; pollution degree 3 |
| 2P | DC 500 V | Low-voltage rack systems and 400 V class DC links | Same Icu basis, voltage rating doubles with the second pole in series |
| 4P | DC 1000 V | High-voltage rack BESS and 800 V DC architectures | The usual choice where the bank or the DC bus sits above 500 V |
Two consequences follow. First, a 4-pole device used at 1000 V is four arc chambers in series, and all four poles must be wired in series — using them as two separate circuits destroys the rating. Second, the current rating is unchanged by pole count: the poles are for voltage, not current.
The NB1-63DC family is specified to IEC/EN 60947-2, the same moulded-case standard used for the AC side, with In up to 63 A. Above that, the AC-side logic of moving to a moulded-case frame applies, and the NM8N range is documented with DC versions alongside its AC frames.
Where the breakers go in a BESS cabinet
A rack-based system usually has four protection positions.
Rack level. A device per rack, or a fuse plus a disconnect, to isolate one rack for service and clear a rack-internal fault without dropping the whole DC bus. This is where the bidirectional question bites hardest: a faulted rack can be fed from the bus through the other racks.
DC bus to inverter. The main DC breaker, sized on the inverter’s maximum DC input current and rated for the full bus voltage at the matching pole count. It must interrupt the bank’s prospective fault current.
DC bus surge protection. A DC-rated SPD at the bus, coordinated with the breakers. AC-side SPD rules do not carry over — Uc is chosen against the DC bus voltage.
Service isolation. A lockable, visible break that a technician can prove dead — a tripped mechanism is not proof of isolation.
Choosing the current rating
Size on the maximum continuous current the circuit will carry — the larger of the inverter’s maximum charge and discharge current at the lowest bus voltage, not the battery’s kWh and not kW divided by nominal volts. Then derate for the enclosure: a BESS cabinet is frequently the hottest place on site. Finally, check the prospective fault current against the Icu at the actual bus voltage, remembering that a 4-pole device at 1000 V and a 2-pole device at 500 V do not have the same margin.
Where the available fault current exceeds the breaker’s rating, the answer is a fuse with a higher interrupting rating upstream or a change of device class — never a bigger breaker in the same frame. Mixing a fast fuse and a breaker needs a discrimination check.
Mistakes that show up in commissioning reports
An AC MCB on a DC circuit. It will appear to work at low current and fail at the first real interruption. The arc does not extinguish at a zero crossing that never comes.
A polarised PV breaker on a battery. Installed the “correct” way round for discharge, it is the wrong way round for charge or for a reverse rack fault.
Two poles used as two circuits. Each pole was counted on for its share of the series voltage; splitting them halves the rating and voids the assumption the design was built on.
Ignoring circuit inductance. DC breaking capacity is quoted for a defined circuit time constant, and a long inductive DC run is not the test bench circuit the Icu figure assumes.
Paralleled racks without individual protection. One faulted rack then draws current from all the others through cable that was never sized for it.
What to send us for a quotation
Bank chemistry and configuration · nominal and maximum DC bus voltage · maximum charge and discharge current · prospective fault current at the breaker position · number of racks and whether they are paralleled · whether the device must interrupt in both directions · enclosure ambient · required certificate and destination country · quantity. You get unit price, packing data and the certificate document for the exact ordering code. Request a quotation.
FAQ
Q: Can I use a solar DC breaker on a battery bank?
A: Only if the device is rated for current in both directions and for the bank’s prospective fault current. Many PV string breakers are polarised for one direction, which makes them unsafe where the battery charges and discharges through the same device.
Q: Why does the DC voltage rating change with the number of poles?
A: Each pole contains its own arc chamber. Wiring poles in series splits the arc voltage across them, so a 1-pole device is rated DC 250 V, a 2-pole DC 500 V and a 4-pole DC 1000 V on the NB1-63DC. All poles must be in series to get that rating.
Q: Does adding poles increase the current rating?
A: No. Poles are for voltage. The current rating stays at the device’s In regardless of how many poles are stacked, which is why a 63 A four-pole device is still a 63 A device.
Q: Is a breaker enough to isolate a battery rack for maintenance?
A: Not on its own. A tripped breaker is not proof of isolation. Use a lockable disconnect with a visible break at rack level, and treat the DC bus as live until it has been proved dead.
Q: Do I need a DC SPD as well as a breaker?
A: Usually yes on the DC bus. The breaker protects against overload and short circuit; it does not clamp a surge. The SPD must be DC-rated and selected against the bus voltage, with its backup protection coordinated with the breaker.
Related: MCB range · MCCB range · DC MCB for solar, 250–1000 V · Breaker between inverter and grid · DC SPD for solar PV · PV combiner box components · IEC 60898 vs 60947 · How to verify genuine product