Transformer standby losses

[moltenmetal]

Hi All: thanks in advance for your help.

We have a light manufacturing facility which seems to draw quite a bit of power when nobody is in it...and yes, we've looked for hidden grow-ops and haven't found any!

Peak draw during peak production is on the order of around 220 kW, plus more in summer when the air conditioning is running. However, evenings and weekends when nobody is in the place, the draw is still on the order of 30 kW- to me that's a staggeringly high fraction of our peak demand. That's after we found and killed some of the phantom night-time loads. Some of the remaining nighttime draw adds up- emergency and landscape lighting, servers, similar stuff we can account for- but some of it is still unknown. One suspicion is standby losses from our internal distribution transformers.

We're planning an expansion facility and have the chance to do things right this time, or at least to try. So here are my questions:

1) What percentage of the design draw on a transformer is typically lost due to standby losses when there is no load or nearly no load on the secondary? Is it 1%? 5%?

2) Given that our large users, which are lighting (on 347 V, which is phase to neutral from our nominal 575 V 3 phase supply) and welding machines (575 V 3 phase) are usually off 15 hours per day and all weekend, is there any energy efficiency/power cost savings benefit to be had by switching off the primaries of the transformers for these loads during unused periods?

There would be costs to do this, including having some smaller transformers to supply the various loads that are required 24/7, as well as the contactors and timer to switch off the primaries.

We will of course request data on the transformers before we let the builder buy them, which we didn't last time around. But some idea of the magnitude of the standby losses of typical transformers would be helpful to let me know if this is sensible or just a total waste of time.

 

[dpc]

Transformer no-load losses are probably less than 1%. Very small in relation to any other losses in your system. I would not de-energize your transformers until you do a systematic search to find where this 30 kW is going. Should be relatively simple to take a clamp-on ammeter and start opening breakers during your off-peak hours. You can then see which circuits are using the power. You need to narrow down your search.

Power transformers are incredibly efficient.

 

[waross]

Are you sure that the draw is 30 KW? 30 KVA may be more reasonable. I have seen unloaded transformers running at 10% PF. Very little KW demand but a 90% PF penalty on what load that there was.
If you are calculating the draw based on Amp readings then you are not calculating KW, you are calculating KVA which may be many times higher than KW for unloaded transformers.
If you give us a description of the revenue meter in place we may be able to provide a tutorial on determining true power usage base on information on the meter face and a stop watch.
Some electronic meters are programed to display the instantaneous demand.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[prc]

Waross,in modern transformers the no-load pf is quite high 0.5 to 0.9. This happened due to the drastic reduction in exciting current-from better core steels and improved construction.

 

[waross]

Thanks for the heads up prc.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[ScottyUK]

30kW doesn't sound excessive - do you have a large floodlit car park? A server room with HVAC?

 

[moltenmetal]

Landscape lighting on this property is pretty modest and so is the demand of our servers. 30 kW is a very significant fraction of our peak load. The figures given are from our utility bill and so they're kW, not kVA. We did a load survey but regrettably I don't have a circuit by circuit breakdown which sums up to the losses showing up on the utility meter- just some vague impressions of what the loads are.

1% losses don't sound like much- unless they're 1% of the full design load of the transformers, which in our case would be something like 2-3kW given that we're not challenging the design loadings of these transformers even during our peak periods. If the losses are really 3%, which would not surprise me at all given the lack of attention to detail given to the procurement of this gear when we had the building built for us, that could be as much as 1/3 of the load we're seeing, which would certainly add up during the ~ 14 hours per day that the transformers are sitting idle. Regrettably, it's a load which we can only test at the meter by shutting down everything else, some of which you never want to turn off, so you need to do it by difference. Not easy...