Formal Calculation of RCD Size.

[Ray Pooley]

I have had a good search for this information but can't find it anywhere not even in my on site guide 17th edition.

So imagine you have a CU with 10 MCBs but no RCDs. You decide you want to install an RCD.

You do your max load and diversity calculations for each MCB circuit and you come out with 10 current values that will invariably be less than the rating of the MCBs.

How do you use that information to decide what size RCD to install assuming the MCBs are all in the same bus bar?

Do you add up the max demand and diversity currents? The current ratings of the MCBs? Or is there a formula similar to the max demand and diversity rules for MCBs that you deploy to the RCD size calculation? EG: All of highest current value, 80% of next highets and so on.

Is this calculation something that can lead to a decision to split the CU if a single RCD can't cover the entire unit?

Thanks

 

[telectrix]

as the load rating of a RCD is the max. it can safely handle, then you would just fit a 100A RCD, or, if splitting to 2 RCDs fit 2 x 63A. ssssssimplesss,

 

[Ray Pooley]

as the load rating of a RCD is the max. it can safely handle, then you would just fit a 100A RCD, or, if splitting to 2 RCDs fit 2 x 63A. ssssssimplesss,

I have read that on my various searches and it does sound simple I agree. Others have suggested rules of thumb. But I am interested in how you arrive at that decision. A formal diversity type formula or approach. Is there such a thing?

 

[SparkyChick]

One of the things about the 18th edition was the requirement to ensure the RCDs cannot be subjected to overload. So, if the upfront OCPD is 80A, the RCDs should be rated at a minimum of 80A a piece.

I'm sure if I'm wrong, someone will correct me, but I believe this is why Hager moved to supplying their boards with 100A RCDs, as it removes the guess work.

 

[Andy78]

As per SC's post the 18th asks for overload protection of RCDs unless the manufacturer's instructions agree that diversity can be used, and I don't yet know of a manufacturer that does.

 

[Timbo]

How do you decide what rating of main switch to install? Same principle.

 

[Ray Pooley]

One of the things about the 18th edition was the requirement to ensure the RCDs cannot be subjected to overload. So, if the upfront OCPD is 80A, the RCDs should be rated at a minimum of 80A a piece.

I'm sure if I'm wrong, someone will correct me, but I believe this is why Hager moved to supplying their boards with 100A RCDs, as it removes the guess work.

Is there anything on the subject of RCDs in the 18th edition that is similar to Appendix A Table A2 in the 17th but applying to RCDs instead of MCBs?

 

[Lucien Nunes]

That table is correlates loop impedance with OCPD rating & curve to achieve a sufficiently fast disconnection. But the current current rating of an RCD is unrelated to disconnection time, it's simply a maximum as for any switchgear or Iz for a cable, beyond which it might overheat.

Just as you would protect a 20A switch against overload by either 20A max OCPD upstream or a 20A max fixed load downstream, so it is with RCDs. E.g. 100A supply with 150A total MCB ratings: RCD rating must be at least 100A to be protected by the supply fuse. Or, 100A supply with 52A total MCB ratings, RCD rating must be at least 63A to be protected by downstream MCBs.

I've always made sure RCDs are protected against overload, I can't see any justification for not doing so.

 

[Ray Pooley]

That table is correlates loop impedance with OCPD rating & curve to achieve a sufficiently fast disconnection. But the current current rating of an RCD is unrelated to disconnection time, it's simply a maximum as for any switchgear or Iz for a cable, beyond which it might overheat.

Just as you would protect a 20A switch against overload by either 20A max OCPD upstream or a 20A max fixed load downstream, so it is with RCDs. E.g. 100A supply with 150A total MCB ratings: RCD rating must be at least 100A to be protected by the supply fuse. Or, 100A supply with 52A total MCB ratings, RCD rating must be at least 63A to be protected by downstream MCBs.

I've always made sure RCDs are protected against overload, I can't see any justification for not doing so.

So it's just a case of the bigger the better? No formula?

By formula I am talking about a rule for determining the minimum RCD spec for all MCBs on a single busbar. A formula that a computer program could apply to a set of values. For example in the case of MCBs it is possible to write a program that can process a list of device power ratings on a circuit and come up with a minimal MCB current. Because ther are rules like 100% of first, 80% and 60% of remaining sum etc.

In the case of RCDs a program that could process a list of MCB ratings or MDD values that would spit out a minimum RCD amps rating. That's what I am aiming at. Computerising the calculations. I just need to know the formula that can be used in the RCD calculation. If It exists of course and I am getting the impression that it doesn't.

 

[Wilko]

...
I just need to know the formula that can be used in the RCD calculation. If It exists of course and I am getting the impression that it doesn't.

Hi - sorry if I’m repeating earlier advice, but without a load analysis and design the best guide (in my humble opinion) would be “the RCD rating should equal or exceed the upstream Overload Protective Device rating”. So if the service fuse is 80A, then the RCD should be 80 or 100A, say.

 

[Lucien Nunes]

...come up with a minimal MCB current. Because there are rules like...

It's a completely different situation. When calculating maximum demand, you're working out what supply current will satisfy the user's likely requirements. If they try to exceed that current no switchgear gets overloaded, the supply OCPD operates. So the limit is one of functionality, not safety.

OTOH if you use a similar method to size the RCD and it's not protected by OCPD, then if the user takes more than you calculated, the RCD overheats.