Sizing the string breaker for a lithium retrofit, not reusing the VRLA one
The breaker bolted to the old VRLA string was sized for a battery that is about to leave the cabinet. Reusing it without checking is a guess wearing a rating label.
Key figures at a glance
| Parameter | Value | Where it comes from |
|---|---|---|
| What the old breaker was sized for | The VRLA string's fault current and the plant's own distribution limits | DC protection design |
| What doesn't change with the battery | The site's cable run and the ampacity it was sized for | Site design |
| What does change | The pack's own overcurrent / short-circuit response, set inside the BMS | Manufacturer datasheet |
| What to request before ordering | The BMS's max continuous discharge current and its internal protection trip curve | Manufacturer datasheet |
| Why coordination matters | An upstream breaker and the BMS's internal cutoff both react to the same fault — one has to lead | DC protection design |
What a breaker rating actually protects against
A DC string breaker or fuse is sized against the worst case the wiring and the connected battery can produce: the pack's maximum fault current, the cable's ampacity for its length and gauge, and the plant's own distribution limits. VRLA and lithium fail very differently under a short. A VRLA string's fault current is set almost entirely by its internal resistance and can be large and fast; a lithium pack's is capped and shaped by its BMS, which usually limits and interrupts current on its own terms — numbers a breaker chosen for a VRLA string was never asked to coordinate with. Same amp-hours on the label, different behaviour at fault.
What genuinely carries across from the old installation
The cable itself, if it was sized correctly for the run in the first place. Ampacity and voltage drop are functions of length, gauge and the site's actual load current — none of which change because the chemistry hanging off the far end of the cable changed. It is worth re-measuring the real load rather than assuming the old cable carried headroom that no longer applies, but the cable is not automatically wrong just because the battery is new.
The document to request, and why a single amp number isn't enough
Ask for the BMS's maximum continuous discharge current and its internal short-circuit response — both the trip threshold and how fast it acts. A breaker sized only against the pack's continuous current rating, with no view of what the BMS itself does during a fault, can end up racing the BMS to open first, or worse, never seeing a fault that the BMS already cleared internally before the breaker's slower curve engaged. Either outcome means the breaker was chosen without the one piece of information that actually decides whether it does its job.
Protection coordination is a system decision, not a component swap
The upstream breaker and the BMS's internal protection both watch the same physical fault, and one of them has to be the device that actually clears it. Normally that is the BMS's own disconnect, with the breaker acting as backup for a fault the BMS cannot clear itself — a failure inside the BMS, for instance. Deciding which one leads is a coordination exercise done with both curves in hand; it is not a decision that gets made correctly by reusing whatever rating happened to already be bolted to the shelf.
Equipment this guide covers
Common questions about this family
Is it safe to just keep the same breaker rating as the VRLA install?
Often it will sit there working fine without ever being tested by a real fault — which is exactly the risk. The rating that actually matters is the BMS's protection curve, not the number stamped on the old breaker, and the only way to know the two coordinate is to compare them, not to assume it from the amp-hours being similar.
The BMS already has its own overcurrent cutoff. Do I still need an external breaker?
Yes. The BMS protects the pack against faults within its own reach, but it cannot protect against a fault in the cable or busbar between itself and the breaker, and it cannot cover for its own internal failure. The two are complementary devices, not redundant ones.
Other guides
- Is the Emerson R48-2900U the same as the Vertiv R48-2900U?
- Can I mix R4850G2 and R4875G1 modules in the same shelf?
- What float voltage do I set for lithium on a Flatpack2?
- Can I put a 51.2 V rack battery in my −48 V telecom plant?
- How do I disable temperature compensation on a NetSure controller?
- Is the Eltek Flatpack2 discontinued?
- Where do I set the low-voltage disconnect on a lithium retrofit?
- What paperwork does a telecom lithium battery need to ship?
- How do I identify my 48 V DC plant?
- Does my controller need a firmware update before a lithium retrofit?
- How does a lithium BMS talk to a telecom DC plant?
- How do I size a lithium bank to replace my VRLA string?
- Does the grounding scheme change when I retrofit a −48V plant to lithium?
- Can a VRLA string and a lithium string run in parallel during a migration?