Industrial site electricity supply expected infrastructure losses with 200KW load

MPC said:

The estate has several units each one sub metered from a resident 500KVA transformer.

One transformer for the whole estate, or one transformer per tenant? If the latter, I can imagine iron losses being significant, especially if the transformers predate the Ecodesign Directive and there are quite a few of them. This would also be quite oversized for what is now a 200kVA load so might be thinking the load’s dropped well below what the system was designed for, so measurement apparatus may be over-rated.

(Yes I know this doesn’t suggest an ideal design; I’m picturing the likes of ex-airfields using the infrastructure that’s been there since forever.)

If one transformer then LV losses as Mike says are probably a big part of the answer, depending on layout

Thanks for the thoughts. One transformer onsite serving thirteen tenants. The submeters are in the same room as the transformer so I'm thinking the LV losses should be low? Leaving only the 60Metre HV cable from the main site meter to the transformer loss and the transformer losses.

Thanks for the thoughts. One transformer onsite serving thirteen tenants. The submeters are in the same room as the transformer so I'm thinking the LV losses should be low? Leaving only the 60Metre HV cable from the main site meter to the transformer loss and the transformer losses.

Hmm. The whole point about an HV line is that losses are lower and line length is not important- I assume its 11/6.6 kV, and not something exotic like 3kV ?

As a very poor rule of thumb, a cost effective line voltage (i.e one where cable costs and losses and the cost of extra transformers are more or less optimized)  usually ends up within a factor of two of the 1v/m  or so.
So LV at 230V going more than a few hundred m gets tricky and it is worth considering more transformers and 11kV lines are not worth stepping up to 33kV unless  distances between sites are tens of km (so you will see 33kV lines to remote villages in Scotland and Wales, and a wider spacing of the transmission step down sites, where in the south of England, where nowhere is more than a few km from somewhere else, 11kV links are the rule but there are far more EHV feeds into it..)
I digress but the point is  50m more or less of  HV line is like a few feet of LV as a percentage loss.

Mike.

Again thanks for the thoughts. Its 11KV input. Transformer is a WODEN weighing some 1780KG.  Trying to get some idea of efficiency without measuring equipment on the transformer I did the following calc. Assuming a 5m^2 cooling area as its a big transformer. Assume a Convection Coefficient of 10W/m^2 degC (pretty poor). The power needed to raise the transformer 20degC is 10 x 5 x 20 = 1KW.

So the 20degC rise in surface temperature suggests the transformer is not the where the 20KW loss is. Hmm.

I suppose there's also the question of meter accuracy. If you have "many" LV meters I guess they should tend to even out, but if the one HV meter is reading high, that might account for some discrepancy, especially if it's old or hasn't been checked in a while.

   - Andy

1kW at middle loading is entirely credible. Lightly loaded a transformer of that size might dissipate a few hundred watts in the magnetic losses in the core, and at full load maybe a couple of kW in the windings. As Jam notes it is not a modern high efficiency model, but the old ones were not that much worse, usually heavier and some nasty oils in the very old ones, but at most twice the dissipation of a modern one.  Consider that if it dissipated a lot more say 5-10 kW you'd have the oil at chip pan temperatures and a cooling problem . (which you can estimate from how much hotter the room would get with a 3 bar heater in it.)

So now we are not sure what is happening.
Mike.