How a Data Centre Solved Downtime with the Right Miniature Circuit Breaker (MCB) Nuisance tripping on a distribution board sounds like a minor inconvenience until it happens inside a live data hall. For colocation and enterprise data centre operators across the UK, an MCB that trips unnecessarily — or fails to trip when it should — can mean minutes of unplanned downtime, an SLA breach, and an awkward call with a customer. This is a composite, anonymised account of how a UK colocation data centre operator worked through a recurring nuisance-tripping problem by re-specifying the miniature circuit breakers protecting its PDU distribution boards, and what other facilities teams can learn from it. Need help specifying or sourcing miniature circuit breakers for your site? Talk to Our Expert Team The Challenge The facility, a Tier III colocation site in the south of England, had been experiencing intermittent trips on several PDU circuits feeding server racks with a high concentration of switch-mode power supplies. The breakers involved were general-purpose Type B MCBs, originally specified for what the design brief assumed would be a fairly linear, resistive-leaning load profile. As rack density increased and more IT hardware with capacitive inrush characteristics was deployed, the site began recording several nuisance trips a month — each one triggering an incident ticket, a site visit, and in two cases a customer-facing SLA credit. Root cause analysis on the affected circuits pointed to inrush current at power-on, and to the cumulative effect of several devices switching on in close succession — both of which a Type B curve is not well suited to tolerate without tripping, even though the steady-state load was comfortably within the breaker's rated current. Diagnosing the issue took longer than the fix itself. The facilities team initially suspected a faulty batch of breakers, since the trips were intermittent rather than constant — a classic symptom that gets misread as a hardware defect. Power quality logging on the affected circuits over a two-week window told a different story: current spikes at roughly 8-12 times the steady-state load lasting only a few milliseconds, coinciding almost exactly with server power-on events and UPS-to-mains transfer tests. A Type B curve's magnetic trip element operates at 3-5 times rated current, which meant these entirely normal inrush events were being read by the breaker as a fault. What made the problem harder to spot initially was that the trips weren't happening on every power-on — only when several devices in the same rack cycled within a short window of each other, which pushed the cumulative inrush high enough to cross the Type B threshold. This intermittency is exactly why trip-curve mismatches are so often misdiagnosed as random hardware failures rather than a specification issue. The Solution Working with their supply partner, the facilities team re-specified the affected circuits with ABB Type C miniature circuit breakers, which tolerate a higher instantaneous inrush multiple before the magnetic trip element operates, while keeping the same thermal (overload) characteristic appropriate for the cable and load. A small number of circuits feeding UPS-backed inductive loads were moved to Type K breakers for an even higher inrush tolerance. Pole count and breaking capacity were re-verified against an updated prospective fault current calculation for the board, since the original design figures pre-dated a transformer upgrade elsewhere on site. Products Used The re-specification used ABB System Pro M Compact S200 Type C breakers (4-pole, 16A, 6 kA breaking capacity) for the majority of PDU circuits, ABB ComPact Home SH200 Type B breakers (1-pole, 16A) retained where the load genuinely was resistive/linear, and ABB System Pro M Compact S200M UC Type K breakers (2-pole, 10A, 10 kA) for the highest-inrush circuits. All units were sourced with same-week UK delivery to minimise the change-window duration. Selecting between Type B, C and K wasn't a blanket decision — each PDU circuit was assessed individually against its actual connected load, not the load the original design brief had assumed five years earlier. Circuits feeding lighting, general power outlets and other genuinely resistive loads were left on Type B, since there was no inrush problem to solve there and no reason to over-specify. Circuits feeding blade servers, storage arrays and network switches — where switch-mode power supplies produce a brief but sharp inrush at power-on — moved to Type C. The two circuits feeding UPS output boards, where transfer events could stack inrush from multiple downstream devices simultaneously, moved to Type K for the extra margin. Trip Curve Magnetic Trip Threshold Typical Use Case Type B 3-5x rated current Resistive/lighting circuits, general power Type C 5-10x rated current Server PSUs, general IT/industrial inrush loads Type K 8-12x rated current Inductive loads, UPS transfe…