A test is a moment in time.
I’ll relate a couple more anecdotes and you can decide for yourself if it’s considered 100% reliable by others.
Newer devices like the AFDD combine earth leakage protection with Arc fault detection in the same unit.
These recent designs have a self-test facility which is specified by regulation to self-test a minimum of once every 24hrs.
However, I note that the smarter manufacturers run the self-test every 15hrs to ensure a daily offset so that the test happens at different times every day, This ensures the device is tested during the range of conditions that it will endure.
Here is another little factoid that also demonstrates an earth leakage can be temporarily blinded:
Robin loop testers have a patented design that injects DC during certain tests so that can be done without disconnecting the standard earth leakage. (Which would otherwise trip during the test).
For more info google: Robin De-lock Loop Tester
Or:
I understand that this functionality still blinds a Type A earth leakage.
The only trouble now is that a type-B RCD costs can be prohibitively expensive. I did a quick check on RS Components, and a 63A type-B RCD is 13k! Even a 40A unit is over 5k.
Then to make matters worse, the “type” is such a confusing concept that the filter on that site is useless. Sometimes they confuse the overload characteristics of an RCBO (which can by type B, C or D) with the trip characteristics (which can be type AC, A, B, G or F).
So once I skipped all the Eaton breakers (because they seem to be miscategorised as type B), this was the cheapest type-B breaker I could find. And it is only 13A.
Looking for type-F instead, this was the best I could find.
So for the very cost conscious:
If you are going to ignore regulations anyway and do without a working E/L, it doesn’t get cheaper than not installing one at all. Going half-measure is a complete waste of money.
Or alternatively, if you are a Victron employee:
You could be pointing out that Victron LF inverters do not need an expensive Type B, and the competition does, and that should be factored into the customer’s price decision.
Edit: I have also noticed that the more functional Type B is often cheaper than a Type F.
Read here about a RCD per board which makes a lot of sense as we had that in Nov last year when the critical loads DB tripped “sommer net” the one day.
Stumped, worried, and panicky even, I accidentally found the cause.
Caught my 83yo mother in the garage busy filing a knife. Why I asked … she used the knife to get a plug out of a socket. Plastic handle … dankie tog!
So, with the multiple RCDs (5 DB’s) I also bought a LOT of new Hager, Schneider and ABB breakers for the DB’s. Everything Type C.
So I was reading this thread with dread re. non-Type C RCDs …
I think that’s been mentioned earlier, no need to beat that drum any louder
That’s the confusion I’m talking about. There is no “type C” rating for RCDs. That’s the overcurrent behaviour.
Let me explain that a bit more. Your type rating on a breaker tells you how quickly it trips, to dumb it down a lot. Your typical household breaker is a type-C, and that means it uses a combination of thermal- and magnetic measurement. With these breakers you can go over their rating for a short period of time, only when you exceed the breaker capacity by 5 times (5 x In) will it trip immediately (the magnetic portion of the protection). For current levels lower than that, it trips when the thermal component heats up enough to signal a problem. That means you can go 10% over, and the breaker might only trip after an hour.
This behaviour is acceptable, because the breaker will heat up before the wire it protects goes up in smoke. But it also means your max capacity is affected by ambient temperature, which can cause nuisance tripping below the max.
For overcurrent, a type-B breaker is faster-acting than a type-C breaker. Using Type-C in a residential setting is no problem at all.
Conversely, the tripping characteristics of an RCD has to do with all the stuff discussed earlier in this thread, and this is where some of the confusion comes in, because there is a type B residual current breaker!
When dealing with RCBOs (that’s an RCD with overcurrent protection in the same device), it becomes especially problematic, because now it is possible to have a type-B breaker (for the overcurrent part), which is at the same time a type-A RCD…
Now your job is to get the sales guy at the local electrical supplier to understand that…
RCDBO, is cheaper to get a 63a RCB with a 63A double pole breaker before it.
Also learned, the smaller the amp rating on the RCD, the bigger the price tag.
And that it is seriously advisable to have a larger RCD than the 2P breaker it connects to.
Then I hit the Type C 6kA and 3kA breakers. Oke at the electrical shops said get 6kA man, so much better, he said. So I did. What do I know?
Man, the price difference!
And why upgrade to 6kA if you already have 3kA breakers?
One board, the main one, is going 6kA, the rest, bugger that, are all 3kA.
The “impulse withstand” rating. Imagine there is a short circuit in your installation. Imagine that for 500ms, a current of 4000A flows. If you have a 3kA rate on the breaker, your breaker explodes in a big firey ball of electrical arcing. If you have a 6kA breaker, it trips and nothing bad happens.
Whether you need a 6kA breaker depends on the maximum short circuit current at your premises. In your typical single-phase setup in our parts, having more than 1000A short circuit is somewhat unusual…
A sparky can measure the maximum short circuit current with the correct equipment (they don’t actually short circuit it, they do an impedance test).
The only time you must upgrade to the larger breaker, is if you find that your maximum short circuit exceeds what the breakers can handle.
Yes, ties back to the transformer in the street, I read briefly. So I checked with the sparky first. Cause I already got the 6kA, thought I would protect the main DB with them, everything hangs off the main DB in any case.
EDIT: So effectively I have upgraded the main DB from 3kA to 6kA, the rest hanging off it, being 3kA. I think I’ve got both bases covered now.
Yes, but also the cable coming to your house. That has a certain inherent resistance which limits the max current that can flow. Your breaker must be able to safely handle that surge and trip without setting anything on fire.
A larger post, with pictures, about all this type rating nonsense. I’m going to use a common CBI breaker we use in SA.
This is a CBI QA17C. The C at the end means something, it means this device switches only, it doesn’t have overcurrent protection. You can also see that because it has two green levers. The CBI overcurrent breakers have white levers. Note that this is a type-A RCD, by the part circled in red.
Below is a QA17A. This is an RCBO, it includes overcurrent protection. It is also double the price. Note that this is a type-C breaker (red arrow), but a type-A RCD.
Note also the white lever in the Live-side. It has overcurrent protection.
This indicates the rupture capacity of an MCB. It is the maximum amount of current that MCB can interrupt during a fault.
Naturally, it isn’t a technical issue to vastly exceed what is necessary, but why pay more?
A general rule of thumb is if you live close to the substation (source) that supplies you, you will probably need higher-rated rupture capacity MCBs.
However, determining the fault level at your DB is a standard COC test that should be listed on the COC documentation.
Where I would spend the money would be on the rupturing capacity of fuses or MCBs that are involved in your batteries’ circuit. That’s the place it’ll be overlooked.
That is what I learned in Dec, the price tag and secondly, keep the functions apart if you want to save some monies, the overcurrent ideally being smaller than the RCD so that it “protects” the RCD against overcurrent.
Hence me installing 63a RCD’s with like 40a overcurrent 2p breakers on some of the DB’s.
Combining them saves space in the DB though… which can under some circumstances be an overriding factor
It obscures faults though (in my opinion anyway). When the devices are combined, and it trips, you don’t know if it is an overcurrent event, or a residual current event.
Oh good, and because I only have MPPTs at this point, I also don’t need to worry about it?
Say I get a Fronius at some point, presumably I would then need to fork out some extra monies on protection equipment for HF designs?
This raises the question: What % of installations is simply not adequately protecting humans purely due to the regulations not being applied (due to ignorance or incompetence)…
My interpretation is that unless the manufacturer comes right out and states their HF inverter still has a transformer that galvanically isolates the PV (DC) side from the AC side and that a Type A RCD is approved for usage, Type B is the default.
SMA does that.
Now, Fronius and ABB also state that a Type A is suitable but then go on to recommend a Type A with a sensitivity higher than what most countries’ regulations allow.
I have not seen a mention of the DC to AC side isolation from them.
So I’d be inclined to contact Fronius and enquire and if there is no isolation I’d revert to a Type B.
LF designs have this isolation as standard.
I have no idea, I gather it is common knowledge in the UK and US.
My opinion is that given the price of Type Bs, it would’ve elicited many a forum discussion amongst the Sunsynk and Solis fanbois. Discussions that I haven’t seen.
Nevermind just the price. The general availability of anything past basic RCDs in this country is severely lacking. Try to buy a “high immunity” type-A RCD at your local supplier and watch the blank stare that follows…
I’m convinced a large part of this is simply supply and demand. The prices will come down if the volumes go up. It’s really not rocket science.
