UPS efficiency

[Anonymous user]

Does anybody know the real efficiencys for the 160 kVA & 250 kVA PX models with real computerload? - not linear load that is stated in the datasheet.
Can´t find the figures anywhere.

[upsguy_apc]

UPS efficiency is just as published. PF is a different story and that doesn't play into efficiency as its not "real" power being consumed by the load. Even with a .9pf load on a UPS its total KW in vs KW out should remain extremely close to the published specs.

[BillP]

UPS efficiency is affected by Power Factor, whether leading, lagging or distortion. The losses (or inefficiency) in the UPS come from carrying current through the UPS. As Power Factor drops, current rises for a given kW. The UPS simply has to do more work to provide the same kW at a lower Power Factor. At Power Factor .90 the UPS with carry 10% more current and have 10% more loss. A modern UPS, such as SYPX 250 may have an efficiency of .96 with a linear load. At Power Factor .9 the efficiency will be closer to .956. calculation for losses 100 - (.04*100+(.04*10)) (A 100 kW load at PF=1.0 has 4kW loss, a 100 kW load at PF=.9 has 4.4 kW loss)

That being said, modern "real computer" loads are now Power factor corrected with listed Power Factors approaching 1.0.
So, a modern UPS with a modern load with a Power factor of .99 will have an efficiency of .959 calculation for losses 100 - (.04*100+(.04*1))

What this means is that the efficiency rating of the UPS with a linear load is valid for "real computer" loads.

[Anonymous user · Accepted solution]

I'm sorta scratching my head over the apparent efficiency of our local UPS bank (5x APC MGE EPS 8000 625kva, 480v output).

All 5 are operational & feed a common buss.  The average loading of the individual UPSs is 40-45%.  Based on some APC documents I found online, these UPSs should be 90% efficient, or more, starting at 20% load...with a peak efficiency of ~95% from 40-70+% load.

However, what I am seeing is Amps out/Amps in = only ~80%.  I see in the above calculations where KW out/in is the normal means of calculating efficience (or loss), but is there some reason that using Amps would result in such an apparently horrible efficiency number?