Soft start for electric heaters

[Temp]

After some info/advice on breaker curves and/or soft start/in-rush limiting..

 

Our local village hall is having issues with its IR electric heaters tripping at switch on. There are 6 heaters each on their own C16 30mA RCBO. I don't currently know the wattage of each heater but it looks like they have 6 elements in each so I'm assuming they are quite powerful units. 

 

My background is in electronics rather than electrical engineering so I've just been asked to "translate" what the electrician is saying when they can get him to visit next time.

 

For the moment I'm assuming they are tripping due to inrush current rather than Earth leakage but will ask if he's checked that. The wiring doesn't look too old.

 

It looks like last time he visited he switched one of the RCBO for a B20. That one seems to be OK but three others on C16 are tripping.

 

As I understand it..

C curve trip between 5 and 10 times rated current within 1 second. So a C16 trips between 80 and 160 A.

B curve trip between 3 and 5 times rated current between 1 and 5 seconds. So a B20 trips between 60 and 100A.

 

Q1: So tentatively it sounds like the issue is the duration of the inrush current rather than the magnitude of the inrush current?

 

Q2: The whole lot are on one 100A RCD (I think) so I suspect he can't switch them all to B20 because 6 x 20A = 120A. They are normally all on at once so can't apply diversity? Or am I talking bollocks 🙂

 

Q3: Would it be reasonable to try 6 x B16 instead of C16 as they tolerate overload for a bit longer or am I misunderstanding how they work?

 

Have heard conflicting info from the Village Hall committee. One said the electrician was considering adding in-rush current limiting at £25 to £100 per heater, but another added "or rewiring".

 

I think each heater is on a 2.5mm^2 radial. Part clipped, part loose in the roof and part in metal conduit. I think a 2.5mm^2 radial is typically limited to 16-20A.

 

Any suggestions for how to diagnose the problem ?

 

Do electricians carry a fancy meter that can measure in-rush peak durations ? I suspect not. 

 

Edited Friday at 00:14 by Temp

[Nickfromwales]

1 hour ago, Temp said:

Do electricians carry a fancy meter that can measure in-rush peak durations ? I suspect not. 

A simple clamp meter wouldn’t suffice. This will be measured within a human ‘blink’.

 

1 hour ago, Temp said:

Or am I talking bollocks 🙂

Oi!!!! Get your own job  

A C curve breaker is usually the more tolerant of “surge current”, caused by “shunt resistance” in motors usually, or lots of fluo lights etc. I would have thought C curve would be a ‘pikey’ solution but @ProDave and @Onoff are the books on this vs my less qualified understanding.

 

1 hour ago, Temp said:

The whole lot are on one 100A RCD (I think) so I suspect he can't switch them all to B20 because 6 x 20A = 120A. They are normally all on at once so can't apply diversity?

Look at a domestic CU, where quite often or not (usually always) the total amps of each circuit massively outweighs the rating on the front of house isolator.   

 

I would suggest these were brought on by contactors which had programmable ‘delay on’ or even Shely’s perhaps, to bring them on one after the other. Depends on how they’re then used / turned on & off when room temps are achieved. 

[Temp]

3 minutes ago, Nickfromwales said:

I would suggest these were brought on by contactors which had programmable ‘delay on’ or even Shely’s perhaps, to bring them on one after the other. Depends on how they’re then used / turned on & off when room temps are achieved. 

 

As I recall you have to walk around the room switching on each heater in turn so I don't think the issue is multiple heaters being turned on at the same instant. 

 

[Onoff]

Is this a new issue?

 

Is the B20 RCBO he's fitted the same make as the C16 one it replaced?

 

The type of RCD section within RCBOs can vary:

 

https://www.electrical-installation.org/enwiki/Types_of_RCDs

 

The individual RCBOs should have an RCD test applied to gauge tripping and also their ramp times. 

 

The fact that it's been upped to a 20A and not tripping is perhaps telling. B curve is for low inrush currents, C for medium and D for high. If it's an inrush issue then it's usual to go from B to C curve or even D not the other way. 

 

If the cumulative load is too much for the incoming supply then you might need to think about spreading it across phases if you have 3P coming in.  Or timers. 

 

As for diversity on heating then in a domestic premises this would be 100% up to 10A + 50%. There are other ratings for other than domestic in the OSG. Find out the rating of each unit. 

 

[Temp]

@Onoff & @Nickfromwales I think it started getting worse as the weather got colder but then again they aren't used much in summer.

 

I was able to speak to the electrician briefly today. He confirmed..

 

Each heater is 3kW (approx 12.5A).

The insulation breakdown tests he did recently passed ok

He plans to change the RCBOs from C16 to C20s but might not be able to do them all. Think he said he needed to check the wire impedance to ensure they would trip if there was a short?

 

He also mentioned he's considering investigating NTC devices to limit inrush. Ive used these before as auto resettable fuses at much lower currents. But not for high power applications like this. I know they can be a bit tricky to get the right spec.

 

Dimplex who sold the heaters say they aren't made/supported now.  

[Alan Ambrose]

OK hold my beer....

 

Get your sparks to disconnect and measure the cold resistance of a heater and the voltage you actually have there on no-load (or if they're lazy, just assume 230V). Divide one by the other and that's your inrush current.

 

If you've ever seen an old fashioned resistance heater heat up, you'll remember it takes a minute or so to do that. That is, much longer than any RCBO curve, which are designed for machinery in-rush currents. While it's heating up, the resistance goes up and the current goes down ... until it reaches a steady state. Measure the current at this stage with a clamp meter and you'll know the steady-state current (and hot resistance come to that).

 

Now you've got inrush current and steady-state current. Just size your MCBs/RCBOs accordingly to fit the inrush current (actually cold-state current - there's not really an inrush currrent - that is usually associated with sub-1s timescale). Yeah, probably 20A, but worth checking to avoid further problems.

 

BTW, worth the sparks checking the nominal power of the heaters - on a 16A trip, each is probably 3kW.

 

Yeah, I appreciate they probably dont glow...

 

 

 

Edited Friday at 16:48 by Alan Ambrose

[Nickfromwales]

20 minutes ago, Temp said:

He plans to change the RCBOs from C16 to C20s but might not be able to do them all. Think he said he needed to check the wire impedance to ensure they would trip if there was a short?

I’d want to hear from a governing body and then have a method statement in place if doing something ‘bespoke’ and pushing these boundaries, given this is for a 3rd party public venue. I’d prob say go for B20’s not C20’s as there should be no need for the C capability.

[Onoff]

6 minutes ago, Nickfromwales said:

Think he said he needed to check the wire impedance to ensure they would trip if there was a short?

 

Did he say earth loop impedance? 

[Temp]

On 17/01/2025 at 16:49, Nickfromwales said:

I’d want to hear from a governing body and then have a method statement in place if doing something ‘bespoke’ and pushing these boundaries, given this is for a 3rd party public venue. I’d prob say go for B20’s not C20’s as there should be no need for the C capability.

 

Apparently he's experimenting with B and C 20. Jury still out. Got a mixture in at the moment. When I tried them those on C20 seemed reliable. The few on C16 not reliable (tripped 2 of 3 tries to turn on).

 

 

[Temp]

On 17/01/2025 at 16:55, Onoff said:

 

Did he say earth loop impedance? 

 

He said the Earth leakage was ok but I don't recall him mentioning the earth loop impedance. I thought he was concerned that a short might not draw enough current to trip the breaker but I might have misunderstood. 

 

I know he was going to test the wiring before deciding if 20A breakers were ok. Apparently that test went ok as he's fitted some C20 and I've heard he plans to do the rest.

 

Just to add they only seem to trip immediately at switch on not during use. 

 

[andyscotland]

1 hour ago, Temp said:

 

He said the Earth leakage was ok but I don't recall him mentioning the earth loop impedance. I thought he was concerned that a short might not draw enough current to trip the breaker but I might have misunderstood. 

 

"earth fault loop impedance" is the technical term for "wire impedance" - e.g. (in very simplified terms) the resistance of the loop of cable from the live on the generator at the power station to the end of the circuit and back through the earth to the generator.

 

This resistance/impedance determines the amount of current that can flow through a dead short from live to earth. The Earth Fault Loop Impedance is therefore one of the factors governing the size of the protective device : as you say it has to be low enough that a dead short will allow enough current to flow that the breaker will trip within the required disconnection time for the type of circuit.

 

Hence as you have correctly understood, the electrician is saying he needs to measure the earth fault loop impedance at the far end of each circuit before he can confirm that it's appropriate to change the overcurrent protective device (as the limit for a 20A breaker is lower than that for a 16A breaker, and indeed the limit for a type C is lower than for a type B).

 

1 hour ago, Temp said:

Just to add they only seem to trip immediately at switch on not during use. 

 

This does sound like it is due to the cold resistance (and therefore inrush current) of the circuit. If there is a single heater per circuit then there's not a lot you can do other than change the type/rating of the protective device or put an inrush limiter in front of the heater.

 

On 17/01/2025 at 00:11, Temp said:

Q2: The whole lot are on one 100A RCD (I think) so I suspect he can't switch them all to B20 because 6 x 20A = 120A. They are normally all on at once so can't apply diversity? Or am I talking bollocks 🙂

 

The ratings of the protective devices do not of themselves necessarily equal the maximum demand of the installation.

 

BS7671 does not provide any mandatory method for calculating maximum demand : it is up to the designer to use their expertise to make a reasonable assessment.

 

If you have 6 x 3kW heaters on fixed wiring with no other appliances connected (and no sockets in the circuits) then your maximum demand is 18kW = 78 Amps at UK nominal voltage (between 71-83 amps on a supply within the voltage tolerance range). This is without applying any diversity, which probably wouldn't be appropriate in this situation.

 

Changing the protective devices doesn't in any way change the maximum demand of the heaters.

 

Therefore, it's entirely reasonable to size the protective device for the demand, current carrying capacity and loop impedance of each individual circuit. It is not automatically a problem to have breakers adding up to 120A downstream of an overall 100A breaker.

 

Of course, you may find that if you make each individual breaker big enough to carry the inrush current, you move the problem upstream and end up with the combined inrush now tripping the 100A protective device. This however could be managed by adding control gear to switch one or two heaters slightly on after the others.

 

On 17/01/2025 at 00:11, Temp said:

I think each heater is on a 2.5mm^2 radial. Part clipped, part loose in the roof and part in metal conduit. I think a 2.5mm^2 radial is typically limited to 16-20A.

 

2.5mm² T&E is rated for 23A in conduit/trunking on a wall surface, 27A clipped direct to the surface. If the cable is buried / passes through or under insulation at any point, or is bunched with others / at a nonstandard ambient temperature this will reduce those ratings. So from that perspective a 20A RCBO would likely be fine for the current carrying capacity of the cable.

Edited 7 hours ago by andyscotland

[Onoff]

Meanwell for example do relatively cheap, 6kW inrush limiters, 2 module wide and DIN rail mounting but only 16A. Mix and matching these with whatever RCBOs are in the cu could cause you incompatibility issues and indeed invalidate insurances as going against MIs. That is even if there is space in the cu.

 

They don't do a 20A unit but a 28A but I think that is 3 modules wide. 

 

A bfo, separately enclosed, upfront current limiter serving all the RCBOs is an option but if that fails ever you potentially lose power to all RCBOs and thus heaters. 

 

I'm thinking if this is the route taken, then an enclosure with a current limiter per heater feeding an enclosure with just the heater RCBOs. That gets over the different makes in one cu issue. Unless you can get inrush limiters the same make as the RCBOs. 

 

A conundrum for sure which is why the sparks is perhaps taking his time. Put it to him maybe. Should be cheaper than a whole new board.