Interesting question and responses. I tend to agree that the RCD appears to have gained a sort of false cult status protecting against fatal electric shock no matter what else is wrong with an electrical installation. A catch all.
I suspect that all those with experience recognise that it only forms part of a protective system (an automatic disconnection system) along with main protective conductors, equipotential bonding (where required) and fully compliant wiring. We have the infamous 50V/1667 ohms =30mA offering hollow assurance that our pipes and other extraneous metals won't kill us.
The 17th edition seems to essentially promote RCD protection on all circuits in the domestic world (you don't see much MIC or 50mm deep cables in an average dwelling these days) which does nothing to undermine the dubious status of RCDs.
My understanding
- A 30mA RCD will do little to prevent electric shock (fatal or otherwise) when a person comes into direct contact with a live conductor. True that the RCD will see the differential current but a person may perceive an electric shock at currents under 1mA and automatically pull away often smashing their hand into something else, falling off a ladder or throwing a screwdriver across the room. Occasionally their heart may go into VF as a result of the current flow across it or simply being startled by the incident. An electric shock drawing currents above or below 30mA may alter the heart rhythm causing problems now or later. Peripheral nerve damage can also be caused by relatively small currents though not common.
- A fatal electric shock may occur well below 30mA. There are many variables and no guarantees.
I have been told about a lab medical study which shows that a current exceeding 50mA as measured across the longitudinal axis of the heart can induce VF but I suspect that theory bares little resemblance to the real world.
- An RCD is intended to automatically disconnect the supply during an earth fault. It is to prevent extraneous conductive parts (pipes, washing machine cabinets, the handle of your electric kettle) from sitting at a potentially hazardous voltage for a protracted period of time.
- A properly installed electrical installation should offer the lowest possible earth fault impedance offering an RCD the optimum opportunity of disconnecting rapidly (not to mention the MCB but obviously at higher currents)
- A servicable RCD will disconnect well within 200ms (often within 20ms) when an earth fault develops in an installation that has been designed with appropriate protective conductors in place and has a reasonably low ELI.
- 20ms is still one whole cycle so if you are hanging on two pipes that are not equipotential then you might see the light sooner rather than later.
- Sometimes an RCD fails to disconnect! have seen a number of failed RCDs. It is not the first line of defence.
The only way I test an RCD is with my multifunctional though I do remember doing it with a resister and scope - those were the days. Having said that, we carry an electric kettle in the van that is so leaky it could be used to test RCDs
