Two things make 11kV switchgear different from the low-voltage boards downstream of it. The first is construction: everything that matters happens behind earthed metal, invisible to a walk-past inspection by design. The second is what failure means — an MV switchboard does not trip and inconvenience a floor; it feeds everything, and its failure mode of record is an arcing fault that can take the switchroom with it. Reading its condition early is therefore worth doing properly, and "properly" turns out to mean three instruments, not one.
Two ways a board dies.
Strip away the specifics and MV switchgear fails through one of two systems. The current path — busbar joints, cable terminations, breaker contacts — degrades by resistance: a connection loosens or corrodes, resistance climbs, and the joint cooks itself under load over months. The insulation system — bushings, spouts, cable boxes, support insulation — degrades by discharge: contamination, moisture, ageing or a manufacturing void lets tiny sparks form, and partial discharge erodes the insulation until, one switching surge later, it flashes over. Different physics, different timescales, different evidence — and no single instrument reads both.
What each instrument reveals.
Thermography — the current path, under load
The infrared survey reads heat where current meets resistance: terminations, joints, breaker stabs. On metal-clad gear much of this is indirect — reading accessible surfaces to infer what the enclosure hides — which is honest work with stated limits, not a weakness to hide. Findings are rated by temperature rise against a named reference, and the survey only means something under real load.
TEV — discharge inside, through the cladding
Partial discharge inside the insulation system induces transient earth voltages on the enclosure metal. A TEV sensor on the cladding reads them through the closed panel — no opening, no contact with anything live. Readings are compared panel-to-panel and against site background, because the pattern across a row of identical panels is what separates a discharging unit from a noisy site.
Ultrasound — the surface and the air
Discharge across surfaces — tracking over contaminated insulation, corona at HV terminations — radiates ultrasound that instruments hear at ventilation slots, door gaps and seams. It finds the failure mode TEV can miss: activity on the outside of the insulation rather than inside it, often the earliest sign of contamination or moisture doing damage.
What each misses alone.
This is the part a proposal rarely spells out. A camera-only survey of an MV room watches the current path and leaves the insulation system — the flashover family — essentially unmonitored, because discharge makes sound and voltage transients, not useful heat. A TEV-only screen can sit right past surface tracking that an ultrasound probe hears immediately, and neither hears a heating joint. And all three together still do not measure insulation resistance, verify protection relays, or check contact resistance — that is outage work, and the live screen is what makes that outage short and targeted rather than a substitute for it.
So "we scanned the switchroom" is an incomplete sentence. The question is: with what, and which failure family was left unwatched?
What a combined screen looks like in practice.
On the day, it is undramatic — which is rather the point. The plant runs normally. The surveyor works the switchroom panel by panel: TEV readings on each unit's cladding, ultrasound at the openings, thermal imaging of what line-of-sight and agreed covers allow, everything logged against the panel schedule. Because the room is a row of near-identical units doing similar work, the fleet is its own reference — the finding is rarely an absolute number and usually a panel that refuses to match its neighbours.
What comes out is a written, severity-rated picture per panel: which units are quiet in both failure families, which carry indications, and what each indication justifies next — monitor, investigate, or act. Where the evidence points deeper, the assessment deepens: that escalation — screen first, then oil work, earthing, shutdown testing where findings justify it — is how our Electrical Health Assessment is tiered, interest declared. And year on year, the same panels read by the same methods become a trend, which is where an MV room's condition stops being a guess and starts being a history.
Retrieved from https://amkatechnologies.com/insights/11kv-switchgear-condition-assessment
Published 30 July 2026. AMKA Technologies Sdn Bhd, SSM 202301041763 (1535682-T).