Presuming it's a centre-tapped RLV arrangement, the secondary MCBs would need to be full double-pole (not single pole, nor switched N), and perhaps need RCD backup if the loop impedances weren't low enough to give 5s disconnection times (an RCD on the primary wouldn't see earth faults on the secondary side). C16 sounds about right.

A D25 on the primary is harder to judge - it all depends on the starting surge of the transformer (and any connected loads, if they might be present on power-up) - I'd say it's in the right ball park, but verify against the manufacturer's instructions and likely usage conditions. Probably a simple MCB will do rather than an RCBO unless there are extenuating circumstances (like it's in a bathroom or a TT system).

   - Andy.

More info needed.

Do you have the datasheet for the transformer, and if so what does it say about inrush - are inrush limiters built in, is it E-I core or toroidal ? (toroids use less wire and have  lower loss but much more of switch-on thump)

3.3KVA implies that with 25A inlet MCB there is scope to overload, so does it have thermal trips built in that give protection against that ? Also is that 3.3 KVA rating 24/7 or with some duty cycle ?

Generally the choice of load side MCBs will be more to do with expected loads, but for 'anything reasonable' 16A C type is a good start.

Mike

I am probably going to go with one of these:

First, a step back and a think about what's been posted so far ...

Most portable dual output 3.3 kVA site safety transformers have neither fuses nor RCDs at their output, but they do have:

• 13 A fuse at the input (in the BS 1363-1 plug); and
• a thermal overload trip

Does this mean they will not provide 5 s disconnection time under fault conditions? (Anyone wanting to think this through could assume an impedance voltage of, say, 5.6%.)

Does it mean they there is inadequate protection against overloading the socket-outlets?

Tom D said:

Can someone confirm if in place of FUSE 1 ca be used 25A RCBO Type D

If you are relying on the primary OCPD for ADS fault protection, you will need at least 100 A to operate a 25 A BS 88-2 fuse in 5 seconds ... a D25 needs 250 A (see Table in Fig 3A6 in BS 7671)

... but for general overcurrent protection  you'd need to compare adiabatic. If the D25 is operating in faults at less than 0.1 s, you would need to consider the I²t provided by the manufacturer.

There's not much benefit to this being RCBO as it won't pick up residual currents on the secondary (output socket-outlets).

Tom D said:

FUSE 2 16A MCB Type C  please?

Again, there are similar considerations of fuse vs mcb here ... although a C16 that could be reset by users (if in a DBO) might be advantageous, and provide selectivity.

I'm not clear what sort of overload if any needs protecting against assuming the thermal trip works and why that then needs to operate  within 5 seconds. Transformer heat up times at 100% overload are more likely to be closer to 5 minutes.

Note that the input current spec tells us that transformer is designed for more like 1500W steady state, and 3300 watts for short periods whatever that  'intermittent' spec really means. **

It may well be quite droopy at that full intermittent rated  load - could be more like 10%.  - so a dead short may not draw enough to prompt trip something on the primary either.
Mike.

** this one, is similar weight, and their the 3kVA rating is for  5 Mins on 15 Mins and half that indefinite.

Does this mean they will not provide 5 s disconnection time under fault conditions?

As it's portable, it depends on now long an extension lead it's plugged into on the 230V side 

The 5s disconnection time is a bit of a BS 7671 peculiarity - I think it's been said that some other RLV standards traditionally didn't insist on rapid disconnection on Earth faults on RLV systems at all - which makes a certain amount of sense since the voltage at the fault (presuming equal sized c.p.c.s) should drop to around half of 55V (or a bit more on a 3-phase system) so the risk of death directly from electric shock from such a system is inherently pretty low. There might be indirect risks of course, e.g. someone having an involuntary reaction to a perceived electric shock and falling from a height, so I can understand BS 7671 taking the precautionary approach.

So a portable transformer,  just plugged in and with things plugged into it, likely isn't covered  by BS 7671, so possibly no 5s disconnection time requirement to meet.

Hard wire it though, and you're responsible for meeting BS 7671 requirements (but then you do have a bit more control over loop impedances).

   - Andy.

mapj1 said:

I'm not clear what sort of overload if any needs protecting against assuming the thermal trip works and why that then needs to operate  within 5 seconds.

The thermal trip doesn't need to operate in 5 seconds ... nor does the OCPD providing overload protection.

The 5 seconds is for ADS.

For this particular size of transformer, only reason I can see for OCPD on the secondary side at all, is to provide selectivity in case of a fault L-L or L-PE.

The primary OCPD can protect primary and secondary against overload, and also theoretically the cables connecting to socket-outlets if we select things to meet adiabatic.

In terms of ADS I'd agree ... overcurrent protection perhaps not with the 13 A fuse in the plug, and consideration of non-adiabatic.

Considering ADS, let's have a go with the 13 A fuse. If we consider maximum voltage drop to the socket-outlet the portable transformer is connected to, plus the impedance voltage in the transformer, we have a total C of, say 0.85. Since we are looking at a fault to centre-tap, we need to consider that we need to half this again, so voltage factor C = 0.425. It takes about 60 A to operate a 13 A fuse in 5 seconds.

Using the formula of Regulation 411.4.4, Zs × Ia ≤ U0 × C or Zs ≤ U0 × C ÷ Ia, (NOTE at this point I'm reflecting the voltage factor back to the primary, and assuming that on the secondary R1+R2 is approximately zero as BS 7671 talks about Zs at the socket outlet of RLVS) giving a Zs at the 13 A socket-outlet into which the transformer is inserted of over 1.6 ohms ... a radial on a B20 can have Zs up to 2.19 ohms, and a B32 ring final circuit 1.37 ohms.