Transient Voltage Spikes - What is normal or acceptable?

Status
Not open for further replies.
Some time ago I started a thread that called Three Phase Site - Multiple Electrical Issues. To recap and be brief this site is a three building wood frame complex for elderly care and built to residential standards, but has a 1600 amp three phase main panel. There have been and continues to be multiple electrical issues on both high and low voltage systems. We are one of the low voltage guys and experience device resets and lockups, but fortunately have not had to replace gear. Here is what has been replaced:
1) PTAC thermostats by the dozens
2) CF lights. Sometimes 4 go out at the same time.
3) Hallway Emg. Lights come on for no apparent reason.
4) 3 Fire panels have been replaced.
5) AFIC breakers trip randomly or so it seems.
6) UPS units log alerts when no power failure seems to occur.
7) Ballast have been replaced multiple times.
8) A maintenance worker reported being shocked when he touched the neutral or ground wire.

What we know. A power study was done and current harmonics was sometimes 15 to 20%. The installation company did not run a neutral for each branch circuit and all of the bedrooms with AFIC have shared neutrals. The first grounding system was inadequate and was three rods bonded was added to the 1600 amp service, but two separate buildings with their own panels were not grounded until recently. That EC left them un-interconnected. All three buildings are separately grounded now, but feed from the same 1600 amp main and transformer.

Over the last several months I have connected a power quality meter (Fluke 43b) and logged transient spikes of at least 50% RMS of 120VAC. In a three day period on several occasions it exceeded the 40 captures the meter stores and at other times it will range from 10 to 38 in a three day period.

My question for the forum is how many transients spikes of that degree (60VAC) is normal. I know if a perfect world zero is the answer, but this facility is not normal. Any other input is appreciated too.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
That is a transient increase to 180VAC for one or more cycles or just a single cycle spike on top of the AC?
Even one such spike can damage electronics built normally for 120V and not equipped with surge protection. Many per day guarantees failures.
Your job, Mr. Phelps, is to find the cause. Irregular grounding alone would not do it, but wrong bonding could make the effects of a compromised neutral much worse. If you have feeders from one building to another rather than service conductors to each, you really should have EGCs. But once again that is not the root cause.
Are you able to confirm whether the spikes are present at the service point? If they are, then you may have a series arcing POCO neutral (intermittent).
 
Over a year ago the utility was asked to check their equipment. I am rarely on site myself and certainly not for hours of waiting. The utility reported that their equipment was fine, but the staff on site never saw anyone show up. Not checking in or out makes one wonder if it was a drive by check.

A new engineer has been brought in and I believe is trying to get the utility to check (recheck?) their service/transformer as you mentioned. If nothing else, to eliminate that part of the puzzle. He specializes in troubleshooting and I hope can get them to be through.

I believe him and I share many of the same ideas in the problems. Somewhat like you have said, there maybe more than one issue at hand.

To answer your question on the spikes. They do come in waves and are consistently inconsistent as to time. Three or four in minutes and then 12 to 24 hours without any. One captured showed a rising spike on the downward waveform. Like the AC was going backwards?

I agree that finding the root cause(s) is the challenge, but knowing they are there is the first step to recovery IMO.

Thanks for the input. More as it unfolds.
 
Interesting comment. With the early sundown now, I was there late afternoon and one of the parking lot lights had a noticeable flicker. The maintenance supervisor noticed and said I need to call the company to replace that...again. They hire a company with a lift to change bulbs.
 

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
You have detailed enough trouble here to warrant additional spending on finding the cause and solution. At the risk of promotion (because my company just bought this company), there is a product called an I-Sense made by Soft Switching Technologies that you should look into. SST makes power quality systems, mainly a type of surge and sag protection system that does not use batteries or capacitors to prevent shutdowns. The I-Sense is a tool they developed for helping customers understand how bad their power quality is. It connects to the utility connections and records, then reports, any disturbances. The cool thing about it is, users can elect to share that data to a server called I-Grid, so if there is an event, you can ask if that event was seen by others nearby who have an I-Sense unit installed and reporting. It can be a real eye opener, especially when you discover that your neighbors have been chasing the same phantoms you have been chasing, when in fact you are all experiencing the same things at the same times, which means it was not YOU!
https://www.igrid.com/igrid/whatisigrid.jsp
 

mike_kilroy

Senior Member
Location
United States
May be very low cost help:

http://www.ditekcorp.com/Docs/ProdGuides/120-240CM+ Datasheet.pdf

Stick one on side of each power panel for each building... If your issue is very low overvolts like you suggest (180vac on 120v circuit) then it will do nothing for you since it doesn't start clipping until 500v (350vac), but if your issue is really normal higher short duration spikes, then it might. There may be other brands/models that clip at lower voltages.

In any case, with your issues, it may be cheap insurance (I have used this brand surge protector and have no other relation to them).
 

ATSman

ATSman
Location
San Francisco Bay Area
Occupation
Electrical Engineer/ Electrical Testing & Controls
Spec Resolution

Spec Resolution

You have detailed enough trouble here to warrant additional spending on finding the cause and solution. At the risk of promotion (because my company just bought this company), there is a product called an I-Sense made by Soft Switching Technologies that you should look into. SST makes power quality systems, mainly a type of surge and sag protection system that does not use batteries or capacitors to prevent shutdowns. The I-Sense is a tool they developed for helping customers understand how bad their power quality is. It connects to the utility connections and records, then reports, any disturbances. The cool thing about it is, users can elect to share that data to a server called I-Grid, so if there is an event, you can ask if that event was seen by others nearby who have an I-Sense unit installed and reporting. It can be a real eye opener, especially when you discover that your neighbors have been chasing the same phantoms you have been chasing, when in fact you are all experiencing the same things at the same times, which means it was not YOU!
https://www.igrid.com/igrid/whatisigrid.jsp

If I am reading the spec sheet correctly:
Voltage Deviation Event detection trigger ?-cycle rms voltage ? 87% or ? 115% of set nominal
Adaptive waveform deviation detection of transient events.

The quickest wave capture is 1/2 cycle which converts to 1/2 x 16.6ms = 8.3ms.
Normally transients we see are in the microsecond range. Can you explain?
 
So is this a typical week?

Yes, this seems to be typical. The first few times I started recording and came back 3 days later and the unit had switched screens. I talked to Fluke and they gave me a simple answer that the Fluke 43B stores a max of 40 transient events. The tech added that 40 in three days was "unacceptable." He summarized it by saying each spike is like a drop of water in a water torture. At some point everything will give, it is just a matter of time. Since the site had a power quality study (Fluke 1750 monitoring for 3 weeks) we know the Current Harmonics runs 15% and occasionally peaks to 20% THD. I wonder what happens if a spike occurs when the current is at that 20% mark?

I will have to look up the make and model, but each building does have surge suppression on the three phase system next to each buildings main box.

Thanks for the input and assistance.



Here is the Transient log/times for last week
Monday
1) 11/10/14 18:47:37

Not one on Tuesday

Wednesday
2) 11/12/14 07:10:26
3) 11/12/14 09:00:01
4) 11/12/14 12:56:31
5) 11/12/14 12:57:25
6) 11/12/14 14:12:16
7) 11/12/14 14:12:42
8) 11/12/14 14:14:24
9) 11/12/14 14:18:13
10) 11/12/14 14:25:10
11) 11/12/14 14:31:30
12) 11/12/14 14:32:37
13) 11/12/14 15:21:31
14) 11/12/14 15:48:22

Thursday
15) 11/13/14 13:44:11
16) 11/13/14 13:46:23
17) 11/13/14 13:46:33
18) 11/13/14 13:46:37
19) 11/13/14 13:49:46
20) 11/13/14 13:55:28
21) 11/13/14 13:56:44
22) 11/13/14 13:56:52
23) 11/13/14 13:57:03
24) 11/13/14 13:58:21
25) 11/13/14 14:02:56
26) 11/13/14 15:12:52
27) 11/13/14 15:51:50
28) 11/13/14 15:54:05
29) 11/13/14 17:43:39
30) 11/13/14 17:52:24

Friday
31) 11/14/14 07:05:30
32) 11/14/14 07:14:38
33) 11/14/14 07:19:52
34) 11/14/14 07:32:48
35) 11/14/14 07:50:29
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
The fact that both spikes are near the peak is either telling us something or is just an indication that trigger for logging is based on absolute voltage rather than dV/dt or other method that would show spikes anywhere on the waveform.
 

robbietan

Senior Member
Location
Antipolo City
our experience in the utility often points to capacitors on the line as the main cause of voltage transients

it would be helpful if the presence of capacitors near the building or inside could be seen

I had a building here whose power factor capacitors switching matched the transients recorded on the 5th to 19th floor of the building
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
141117-0814 EST

The first scope trace appears to show a single full cycle moderately high frequency, relative to the power line frequency, oscillation that is triggered somewhere near the line voltage peak. Likely related to an LC circuit. The L could be partly from the distribution transformer and/or cable inductance.

Power line communication could be another source, but I do not think it is.

The second scope trace also looks like a resonant circuit producing the oscillating voltage. But, note, the oscillation grows with time, then quits. It is not just a single full cycle as in the first scope picture.

If you have have an LC resonant circuit with some resistive loss and start with some energy in the circuit, and add no more energy after a certain time point, then the resulting oscillation after that time point is called a damped oscillation. Energy is constantly being lost with time and thus the amplitude has to diminish..

How do you get a growing oscillation? Energy has to be pumped into the circuit in phase with the oscillation. View this as a child's swing where someone adds a push each cycle to increase the height of the motion.

Look for something synchronized to turn on near the power line voltage peak. Is it only on positive peaks? What would be something to cause an increasing waveform?

.
 
Here are two more pictures that show other transients, not at the upper peak. I would categorize a larger percentage of spikes at the peaks now that it is pointed out. I find it interesting that all three buildings have TVSS and yet we see them as frequent as we do. Does anyone have any experience with Liebert TVSS? Do they truly help?
 

Attachments

  • 20141103_172309[1].jpg
    20141103_172309[1].jpg
    142.9 KB · Views: 3
  • 20141103_171711[1].jpg
    20141103_171711[1].jpg
    137.6 KB · Views: 3
  • 20141117_110231[1].jpg
    20141117_110231[1].jpg
    137.1 KB · Views: 3

junkhound

Senior Member
Location
Renton, WA
Occupation
EE, power electronics specialty
The waveforms look very much like an SCR rectification or inverter circuit reverse recovery commutation.

Also, the recorded transients are suspiciously grouped, implying a turnon or activity of a few specific pieces of equipment.

Are there items such as treadmills, etc on the site? Esp. if no one uses them on Tuesdays?

Have also seen golf cart chargers produce such transients on-line.

Does a physical therapist show up mostly on Wed and Thurs. but never on Tuesday? Possibly a portable piece of equipment is the culprit.


I'd get the schedule of events at the complex and see if there is any correlation with anything, however seemingly unrelated on the surface.
 
Last edited:

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
141117-1421 EST

well grounded:

I don't see any transients in your latest waveforms. All I see is the small square ripple on the sine wave which appears to be your instruments sampling rate. If this is a modulation of/or on the sinewave, then it is about 30 to 40 times 60 Hz.

A transient voltage limiter is a clamping device that is moderately soft and starts to clip well above the peak of your sine wave. Do a Google search on the operation of MOVs.

.
 
Status
Not open for further replies.
Top