Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations KootK on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

UPS, Waveform, THC, and Phases 2

Status
Not open for further replies.

Townshend

Student
Oct 8, 2022
5
At my house I have an 1100VA UPS (WB-OVRC-UPS-1100-1) where I have been encountering issues since Day 1 back in March. Periodically for periods of time the UPS will alternate rapidly between line and battery, about a second at a time (attached is a video of the occurrence). I had an electrician run a 20A dedicated circuit back to the panel to see if that resolved the issue, which it did not. I tested on other outlets/circuits throughout the house with no luck. At the time, the only possible connection I could make was the morning after it rained the issue seemed to occur. Not 100% of the time but that seemed to correlate and wondered if there was interference from the sump pump. I then tested by turning off every circuit in the house except for the circuit to the UPS, again, no luck.

I then replaced the UPS with a brand new one in case it was an issue with the UPS itself. It seemed to be pretty solid for nearly a month and then the issue would periodically happen again. Coincidentally during this time it was also nice and dryer here although was not positive that had any relevance. Eventually it began occurring with some regularity and to the point of near constant and replaced the UPS with a new one for a third time.

The third one continued the trend of near constant alternation between line and battery. At this point, I got a multimeter and monitored the frequency and voltage. Frequency ranged from 59.89-60.00 and voltage from 122-125, even during episodes where the UPS would alternate. At this point I began to look at a greater scale. I had my utility have a monitoring device at the meter for about a month and they noted no anomalies. I then took the UPS to another house a few miles down the road (as with most residential, single phase), which had the same issue. I then took to a local Starbucks which has a 3 phase pad mount. UPS never went away from line power during a 15 minute test, at my and another residence the alternation would occur almost immediately. I then went to a local school and tested there, also a 3-phase padmount feeding to the building. Again, 15-minute test, no issues. Tested at a 3rd residence, 1-phase, issues occurred. Local gas station, 3-phase open delta bank, had issues. Last, McDonalds 3-phase padmount, ran without issue for 15-minutes here as well.

I am not an electrician and learning and testing as I look to determine the root cause. A few questions:

1. The only consistent success has been the 3-phase padmount electrical feed. Obviously this is not feasible to my home, but would there be a specific reason the UPS would function without issue in this scenario versus 1-phase or even 3-phase open delta?

2. Since voltage and frequency appear within spec, the cause seems to be narrowing down to a waveform issue as it relates to total harmonic distortion (THD). Could I be on the correct track here?

3. If so, any recommendation on a Harmonic Analyzer or device to view the waveform and determine what might be out of spec to trigger the alternation? In looking around these are quite expensive. Are there professionals that I can look for that may have one I can hire or a recommendation of where to rent one?

4. Finally, if anyone has seen a similar scenario as mine and has any suggestions on ways to mitigate? Better grounding or an alternate method? The simple solution might simply be to switch to a less sensitive UPS, however, at this point it's also worth digging into for the educational aspect.

Thank You!

 
 https://files.engineering.com/getfile.aspx?folder=2ac01901-5a12-428a-be96-074f301ded81&file=Wattbox_UPS_Issue_10-5-22.mp4
Replies continue below

Recommended for you

All three of the exact same model?
Have you contacted the maker of the UPS?

See also:
Your main problem is, unless yours is the only house on the transformer and substation, the source of electrical problems could be anywhere, perhaps somewhere far away if a substation fault protection system is involved. In the case of pad mounted transformers when the customer is a business, that transformer is not shared with anyone else.

When you contact the UPS maker ask what the criteria are for switching so you know what to look for.
 
Townshend What is most likely happening is that your UPS is tripping on loss-of-frequency on the incoming AC line. Anything that is connected to that line (even if it's outside your home) and is operating as a non-linear load (e.g., computer, variable frequency drive, solar generation, wind tower, etc. is putting out distortion to the "pure" AC waveform. If it gets bad enough, the UPS sees it as multiple zero crossings - which makes it think it's lost line frequency even though that's not the actual case. As a result, the UPS goes to battery power for some predetermined interval. Then it checks the incoming waveform - if it's good, it reconnects to the line. If not, it stays on battery until the next scheduled "check". The battery will recharge once you're connected to a "good" AC line for a while - exactly how long is dependent on the UPS model/manufacturer. Most of the time, the recharge process is a "trickle" charge - which means if the disturbances are frequent enough or of long enough duration, the UPS can't ever get back to a full charge. Hence the appearance of constant switching between line and battery mode.

Just so you know - this is a very common occurrence on smaller islanded systems (like ships and oil platforms). The way to get around it is to either eliminate the source of the distortion entirely (usually not all that practical), or choose a different UPS design.

Converting energy to motion for more than half a century
 
3DDave, yes, same exact model. I may actually be able to find out how the grid is aligned to determine where a fault might be generated from. The one residence I tested at that exhibited the same behavior is 3.5 miles away in another city. In reading the thread you posted, it looked like he was able to determine a 14% harmonic distortion of the voltage waveform. Do you know how I could go about detecting if this is occurring?

I have reached out to the UPS maker, I am waiting to be put in touch with an engineer who has a better knowledge of it.

Gr8blue, very interesting, thank you. Would a Fluke 179 be able to detect the loss-of-frequency on the incoming AC line? Quite interesting to be aware this is not uncommon on smaller islanded systems. In my case, I am in a suburb of Chicago, of course if the source of distortion is somewhere serving 100k+ people it may be impossible to determine what that is and beyond my control. To your point, ultimately a different UPS design may be needed. Is there a piece of equipment that can verify the thesis otherwise?

Thank you both!
 
Follow up question. I purchased an Amprobe ACD-51NAV and on A~ mode under THD when the issue is occurring while the UPS is on battery power it states rd.y but then when it flips to line power momentarily the meter reads between 68.8 and 70.9. This is on a Klein x10 line splitter. Am I reading it correctly that there is total harmonic distortion between 6.8-7.1% (dividing by 10 for the x10 if I need to?)? Thank you!
 
Follow up question. This morning I got a Y cable and connected the UPS to one end and the Amprobe to the other. What's interesting is I noted the following:

THD: 9.5-9.6%
h01: 100%
h03: 9.0%

From what I understand, h01 is referencing itself and therefore will always be 100%?

However, h03 would be 3x frequency which in my case is 60Hz. Therefore at 180Hz there is also harmonic distortion of 9.0%. Would that typically be indicative of anything. All of the other harmonics were either 0% or >1%.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor