The classes, and why Class 0 is the one
EN 60903 runs from Class 00 up to Class 4, by maximum working voltage:
- Class 00: 500V AC
- Class 0: 1000V AC
- Class 1: 7,500V AC
- Classes 2 to 4: higher still, for utility and substation work
Vehicle traction systems sit comfortably under 1000V, which is why Class 0 is the automotive standard. Class 00 is not enough for most EV work, and the higher classes are thicker, stiffer and make delicate work harder without adding useful protection.
Overgloves are part of the system
Insulating gloves are rubber. Rubber does not get on well with sharp edges, swarf, bonnet catches or the underside of a bracket, and mechanical damage is what ends most gloves' working lives rather than electrical failure.
Leather overgloves go over the insulating glove and take that abuse. A workshop that buys insulating gloves and skips the overgloves will be buying insulating gloves again shortly, and will more likely be using a damaged pair in the meantime without knowing.
The test before every use
Two steps, and they take under a minute:
1. Visual. Inside and out, looking for cuts, tears, punctures, embedded objects, chemical attack, swelling, stickiness and any change in texture.
2. Air test. Roll or inflate the glove to trap air inside, then listen and feel for escaping air and watch for the glove failing to hold pressure. A hole you cannot see will still fail this.
Do it every time. Gloves are damaged between uses, not during them, and the gap between putting them away intact and taking them out damaged is exactly where this test lives.
Retesting and record keeping
For Class 00 and Class 0 gloves, EN 60903 treats the visual check and air test before every use as enough. The six monthly electrical retest in the standard applies to Class 1 and above, and no UK regulation sets a retest interval for Class 0. If your insurer, the manufacturer or your own policy gives a retest or replacement date, follow it.
Keep the record. A glove test log is one of the things an insurer or an auditor asks for after an incident, and reconstructing it afterwards is not possible. Date of first issue, date of each test, result, and who did it.