Joining the Renewable Heating Hub forums is completely free and only takes a minute. By registering you’ll be able to ask questions, join discussions, follow topics you’re interested in, bookmark useful threads and receive notifications when someone replies. When choosing your username, please note that it cannot be changed later, so we recommend avoiding brand or product names. Before registering, please take a moment to read the Forum Rules & Terms of Use so we can keep the community helpful, respectful and informative for everyone. Thanks for joining!
Safety update; RCBOs supplying inverters or storage batteries
And you've even chosen a C-curve, which means it's less susceptible to spurious tripping due to back-EMF from the heat-pump fan and compressor motors. 🙂
Most if not all type A/AC RCDs are bidirectional... Ironically, manufacturers seem to consider this too obvious to explicitly mention it in the datasheet 🤣 or even on the fine print on the device itself.
It's common to wire the input on the top or bottom, whatever allows most convenient wiring for the rest of the panel... unless the device has an arrow indicating direction.
I thought best to continue on this thread as @transparent explained the background and various members went into some detail into our understanding of what lead the IET to update the electrical regulations for the UK.
My initial view was similar to yours, but now my understanding is that:
in the UK this risk led to restrict the use of these devices
elsewhere and specifically in Germany the view was that requiring bidirectional would not be proportionate to the risk in Germany
We discussed this in a thread about DIY solar. Both approaches could be justified. I should say that the process for accepting solar DIY in Germany was very thorough, but unfortunately we do not have the detailed argumentation, even in German.
Perhaps you can help us to understand if the RCD from Noark that you shared is built in a way that removes the risk that led the IET to restrict their use?
I included below a description of the risk with some RCD (I mean it in the broad definition that includes RCBOs) and I am still sceptical about the risk assessment process.
I would encourage you to read the thread above as it has more info.
Thanks, that makes sense. The issue would occur just after the RCD trips (probably depending on the coil material, etc), in case the inverter is still providing power. We should bear in mind that all inverters approved for domestic use have anti-islanding, so when they detect a loss of grid, they should stop providing power within a certain time. I deduct it could also depend on how quickly the inverter stops providing power. And it would also be obvious that the RCD is faulty whenever the RCD test button is pressed.
As the real risk depends on the coil material and heat dissipation and the inverter timings involved, I can only hope that the IET made the decision after careful risk analysis of the equipment combinations (RCD and inverter) that had been historically deployed at the time. And how the risk would evolve.
So the IET officially named these as "bi-directional" RCDs which essentially requires that all installations use them and so are able to cope with an inverter where anti-islanding is not active. This also has the effect of allowing inverters with that type of fault to continue to work undetected (most would as they are probably on MCBs). I am glad I do not have any role in advising government in these matters!
This post was modified 2 weeks ago 3 times by Batpred
8kW Solis S6-EH1P8K-L-PLUS hybrid inverter; G99: 8kw export; 16kWh Seplos Fogstar battery; Ohme Home Pro EV charger; 100Amp head, HA lab on mini PC
I thought best to continue on this thread as @transparent explained the background and various members went into some detail into our understanding of what lead the IET to update the electrical regulations for the UK.
My initial view was similar to yours, but now my understanding is that:
in the UK this risk led to restrict the use of these devices
elsewhere and specifically in Germany the view was that requiring bidirectional would not be proportionate to the risk in Germany
Not quite... Correction required!
Those discussions here on the Forum related to bi-directional RCBOs.
@bobflux is writing about RCDs. That's a different subject area!
The quote you post from yourself to @robl also confuses the two issues.
We already have one topic flagged for deletion because it propagates incorrect information regarding electrical safety devices.
We now have this same problem leaching into other discussions. Not good 🙁
Perhaps you can help us to understand if the RCD from Noark that you shared is built in a way that removes the risk that led the IET to restrict their use?
There are bidirectional arrows in the manual.
However I think I understand the question a bit better... I think it's a matter of perspective...
In France (and the rest of the EU) we like to put the RCD at the left of the line of breakers, there's only one per line so it can be 2 modules wide, then the copper combs distribute power to the line of breakers from the top and circuits come out the bottom.
However the UK likes to put the copper comb on the bottom to distribute Live, but Neutral comes out of the breakers on pigtails to a screw terminal block, and circuits come out on top.
Then the Americans do it sideways.
Then I realized... this UK consumer unit has a ton of RCBO's, thus they have to be tiny 1 module wide units, what's the consequence of that?
How a RCD works:
RCDs use a current sensing transformer, whose primary winding consists of the wires carrying power (Live and Neutral, or more for three phase).
Without fault current, all current going through the transformer via one of the wires returns through other wire(s). For single phase, this means current in both wires is equal but opposite direction. Thus magnetic fields cancel, there is no flux in the transformer core, thus no voltage on the secondary to energize the trip coil.
With a fault, some of the current escapes through the fault and does not return through the other wires, thus the cancellation does not occur. The transformer core receives magnetic flux and there is voltage/current on the secondary (names "search coil" in the schematic).
If your RCD is specified as "Voltage independent" then the secondary provides enough power to energize the trip coil which trips the RCD. These devices are actually powered by the fault current. They do not require any voltage to operate (except the voltage required to create the fault current), and are bidirectional by nature unless specified otherwise, since the transformer is symmetrical. They don't even require both L and N wires to be connected, if you just connect L only (or N only) on both sides and more than the trip current flows in the wire, it'll trip.
Back to the single module width RCDs, or rather RCBOs.
Is the transformer in these too small to power the trip coil on its own? I don't know. The documentation on some that were linked in the other topic says they use an amplifier or some kind of electronic detector to trip the coil. This requires power, which raises the question of which side of the RCD the power will come from. In this case, not bidirectional. Also... if it needs power (L and N) to trip, then if Neutral is cut upstream but Live is still energized, then the RCD safety feature is inoperative.
Then I noticed on the schematic printed on the unit: it doesn't break the N, only the L.
...which is the final clue to solve this problem I guess.
If it doesn't break both wires, besides being unfit for purpose as safety device, it means fault current can still flow in the N wire for too long and damage the device as was said before.
These devices feel really weird. Never seen one before.
Is the transformer in these too small to power the trip coil on its own? I don't know. The documentation on some that were linked in the other topic says they use an amplifier or some kind of electronic detector to trip the coil. This requires power, which raises the question of which side of the RCD the power will come from. In this case, not bidirectional. Also... if it needs power (L and N) to trip, then if Neutral is cut upstream but Live is still energized, then the RCD safety feature is inoperative.
You're on the right track.
An RCBO uses an electronic circuit to fire the solonoid coil, rather than the current created from a toroid being 'out of balance'.
The diagram shown on that Proteus RCBO is simplified in order to fit in the 18mm-wide space. But it does show the Neutral being broken as well as the Live. Both poles get isolated.
The full diagram for a bi-directional RCBO has three sets of contacts, not just two:
There are usually capacitors on the electronics circuit board which hold sufficient charge to cause the trip to operate. Each manufacturer has their own way of doing this.
The 'problem' which gets discussed here is the possibility of the electronics continuing to be powered after the solonoid has fired.
In the UK we can have Consumer Units with the Main Switch on the left, or on the right. That too can create confusion for non-Brits.
But it's not (yet) common to have a Consumer Unit filled with a complete line of RCBOs. 🤨
We're more likely to come across such an arrangement from members of this forum because those posting here are likely to be at the forefront of Low Carbon Technology, with multiple generating units, storage batteries, a heat pump and a couple of EV chargers thrown in for good measure!
That's good. It allows us to discuss an issue which isn't yet commonly found amongst the wider UK population. We are 'ahead of the curve' which is where a good Forum Topic should be!
The datasheet says the driving circuit is latching: once a fault current is sensed it will keep driving the trip coil until power is removed. So if power comes in from the GRID side, it will trip and remove its own power, then all is fine. But if power comes from the LOAD side it'll just stay powered, keep driving the coil, and it will burn exactly as you said.
Overcurrent is handled by a a magneto-thermal actuator as usual. It interrupts Live, but not Neutral, which makes its safety dubious IMO.
I also tested a french Schneider RCD:
I only ran one wire through, so the only power it receives comes from the fault current being simulated by the lightbulb. Since the trip coil is powered by the current transformer, when plugged in the lightbulb flashes briefly, then the RCD trips immediately and cuts off both L and N. So if you need a "bidirectional" RCD, here's one 🤣 but it is a RCD not a circuit breaker, so you need a separate MCB for overcurrent protection, that will take much more space in the consumer unit, 3 modules instead of 1, plus some jumper wires.
So it is less convenient but safer as 1) it cuts both L and N, and 2) it requires no power to trip. So it will work even if Neutral is disconnected, or during a brownout causing a drop of Live-Neutral voltage (that could turn off the amplifier in the other RCBO), or other abnormal conditions. Also it is very unlikely to die silently in case of transient spikes, lightning, etc, since there's no electronics inside to get zapped.
I only ran one wire through, so the only power it receives comes from the fault current being simulated by the lightbulb. Since the trip coil is powered by the current transformer, when plugged in the lightbulb flashes briefly, then the RCD trips immediately and cuts off both L and N. So if you need a "bidirectional" RCD, here's one 🤣 but it is a RCD not a circuit breaker, so you need a separate MCB for overcurrent protection, that will take much more space in the consumer unit, 3 modules instead of 1, plus some jumper wires.
So it is less convenient but safer as 1) it cuts both L and N, and 2) it requires no power to trip. So it will work even if Neutral is disconnected, or during a brownout causing a drop of Live-Neutral voltage (that could turn off the amplifier in the other RCBO), or other abnormal conditions. Also it is very unlikely to die silently in case of transient spikes, lightning, etc, since there's no electronics inside to get zapped.
Thank you for sharing.
So it seems that your Schneider RCD is functionally bi-directional 🙂. The IET rule is very precise about the device not having ports labelled in and out and perhaps other conditions. 😉
I agree the RCBOs that we are discussing are compact and so ideal. But we can only use it if and only if that RCBO of the correct type, curve, amps, sensitivity, etc. Otherwise, as you stated, you may need to combine a bi-directional RCD with an MCB.
8kW Solis S6-EH1P8K-L-PLUS hybrid inverter; G99: 8kw export; 16kWh Seplos Fogstar battery; Ohme Home Pro EV charger; 100Amp head, HA lab on mini PC
