The Building Failing at Itself
Most property cover imagines an external peril acting on a passive building: fire arrives, water rises, wind strikes. Equipment breakdown cover imagines the opposite — the machinery that keeps the facility alive failing from within, with no external cause at all. A transformer's windings short. A switchgear lineup arcs into a fault. A chiller's compressor seizes. A generator that was meant to save the facility fails on start. These are not perils the building is exposed to; they are the building failing at itself, and in a data center they are the losses that recur.
The distinction is not academic, because standard property policies were historically built to exclude exactly this. The classic fire policy covers sudden and accidental damage from named perils, and internal mechanical or electrical failure was carved out — the reasoning being that machinery wearing out or breaking down is a maintenance matter, not an insurable fortuity. Equipment breakdown cover, descended from the old boiler-and-machinery lines that grew up around steam-age industrial plant, exists precisely to fill that carve-out. It responds to sudden and accidental physical damage originating inside covered equipment: electrical breakdown, mechanical breakdown, and in some forms pressure-system failure. The trigger is the internal failure itself, not an external event.
In a data center the covered equipment is the entire critical-power and cooling backbone, and the exposure is severe for a specific reason: the same failure that damages the machine often stops the facility, so an equipment-breakdown loss frequently arrives with a business-interruption loss attached. A failed transformer is not only a repair bill; it is potentially months of lost capacity while a replacement is procured, because the electromechanical heart of the facility runs on long-lead components. This is why equipment breakdown and business interruption are underwritten together in this sector — the physical loss and the income loss share a root cause and a recovery timeline.
The hazards worth naming are concrete. Transformers and switchgear carry the facility's power and fail in ways that are sudden, destructive, and slow to remedy. Generators and their fuel systems are a paradox — protective equipment whose own failure, often discovered only when they are called upon, converts a grid outage into a full loss. Chillers, pumps, and the cooling plant are mechanical systems under continuous load whose failure can force a shutdown even when power is intact, because dense electronics cannot run hot for long. And the lithium-ion batteries in the uninterruptible power supply are both an equipment-breakdown exposure and a fire exposure at once, since a cell failure can be the internal breakdown that then becomes an ignition. The categories blur, which is why the cover is written to work alongside the fire and property sections rather than in isolation.
What makes equipment breakdown distinctive as a discipline is that its losses are, to an unusual degree, preventable through inspection and testing. External perils are largely stochastic — you cannot inspect away a flood. Internal failures announce themselves. Insulation degrades measurably. Transformers can be monitored for dissolved gases that signal internal faults before they become catastrophic. Switchgear can be thermally scanned for hot connections. Rotating machinery reveals wear through vibration. This is why the cover has always been bundled with an inspection service rather than sold as pure indemnity, and why the insurers strongest in this line are engineering organisations. What they sell first is the survey and the recommendation; the indemnity is the backstop for the failure the inspection did not catch in time. For the operator, the practical consequence is that a rigorous, documented testing and maintenance regime is not merely good operations — it is the argument that secures the cover and prices it, and its absence is the fastest route to a denied claim on the grounds that the failure was a maintenance default, not a sudden accident.