Very high UPS Idle Power Consumption

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.

Layla

Game Engine Developer
Jun 21, 2016
173
130
43
39
I just setup a bunch of SMX2000RMLV2U and SMX3000RMHV2U, and just installed fresh batteries in them this week (they're each showing about 135.2V-135.6V on the batteries), and I'm finding them drawing 3 amps each (@ 125V), and 1.4A @ 218V (for the HV models) with no loads attached! That's like 300-375 watts!

These are set in "Green Mode" which is supposed to be up-to 98% efficient (it doesn't do any buck/boost/voltage output regulation when on Utility power), and the batteries are showing 100% charge on every unit (they are brand new). 98% efficiency on a 2000W load would imply < 40W of "waste" power consumption from the UPS (which would be inline with my expectations). I'm seeing roughly 10X that!

It seems that the only explanation is that the UPSes are charging the batteries (despite them being full and reading at 100% charge on the UPS), which is it's own kind of worrying, because charging them at 200W+ for an extended period of time when they're already full sounds like a fast way to boil off electrolyte.

When I turn off the output section, the power use drops to about 1-2A @ 120V. Which implies a couple things:
1. Charging is only about half of the current being used when they're on, and the other half must be the UPS itself.
2. Despite not doing any buck/boost/power regulation/etc. (reminder, they're in Green Mode), they're taking something like 1.5A to produce no power at 120V? That seems far higher than expected for an idle UPS not doing any power regulation. It also doesn't jive with the theoretical 98% efficiency figure; where the UPS would be expected to "waste" only about 40W on a 2000W load...

1632359991276.png

Has anyone ever tested their APC Smart UPS X idle power consumption? Or encountered this kind of waste from a UPS with fresh batteries?
 
Last edited:
  • Like
Reactions: paf and nickf1227

Layla

Game Engine Developer
Jun 21, 2016
173
130
43
39
Mystery solved, I have my answer. I hooked the 208/240V UPS up to an Eaton Managed ePDU with per-outlet metering. Got some immediately useful results which answer all of my questions about power consumption in the above post:

Current: 1.389A (@ 217V)
Active Power: 33W
Apparent Power: 302VA
Reactive Power: -300VAR
Crest Factor: 2.676
Power Factor: -0.107

If you don't know what Power Factor is, or why that explains why my UPSes are not actually consuming 300W though they are pulling as many amps as they would be if they were, here's the Wikipedia page: Power factor - Wikipedia

But, the TL;DR; answer is that because the power factor is not 1.0, many fewer Watts are actually being consumed than the Amps being provided by the power company (but thankfully, the power company bills in Watts, not Amps!). In this particular case, the negative power factor means I'm leaking extra power back to the utility, though, again, this will go away with a proper resistive load applied to the UPS.

I probably should have guessed this was the case; modern power-factor corrected switching computer power supplies in general tend to have power factors near 1.0 and low crest factors (e.g. 1.4), but that still only holds with an actual load against them.

I should also note that some utility companies will charge you a surcharge if you run large loads with a very low power factor!
 
Last edited: