Strange 3-Phase Readings? I need some 3-phase ninjas!

Status
Not open for further replies.

Darksabre

Member
Location
United States
Occupation
Electrical Engineer
Hey guys - long time browser, first time poster :bye:

I'm trying to help a customer troubleshoot a 3-phase power system (specifically a reverse overpower protection relay - Basler BE1-32R). They are getting nuisance tripping from the relay, so I had them take readings of the line-to-neutral voltage and the current from the CT's going to the protection relay sense inputs. Whoever took the readings had equipment that could measure phase angles of the voltage and current, so they sent this data to me to analyze. Unfortunately, my 3-phase knowledge is not the best :happysad:.

I'm attaching the phasor diagram for easy reference, but basically, I'm concerned about how far out of phase Ic is from Vc.

I have no knowledge of their load characteristics at this point, but it looks like it is all over the place. Ia is leading Va by 81 degrees. Ib is lagging Vb by 33 degrees. Ic is...well...leading by 224, or lagging by 136...? I'm not sure what to make of that. Does the cause of this stand out to any of you 3-phase guys?

BTW, the reverse power relay in question operates by measuring voltage and current, and then calculating the phase shift. It uses that information to tell which direction power is moving, and then trips out if it is flowing in the tripping direction (with an adjustable Watt setting and Time Delay setting). An engineer was onsite to commission the relay by injecting a test current and recording the trip points. It seemed to perform as expected during the tests and tripped at the appropriate current flow. Could this odd phase shifting possibly cause the relay to be tripping?

Thanks guys!
-Aaron
 

Attachments

  • BL at ATS Test 2.bmp
    106.7 KB · Views: 1

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
A leading current implies either a simple capacitive reactance (oversized VAR correction?) or a non-linear capacitive-like load such as a power supply.
It is uncommon in the wild, and is rarely seen in captivity.

My first thought would be that the measuring equipment is screwed up.
My second thought is that there is a really wild combination of single phase loads attached to the source.
My third thought is that there may be a large three-phase motor on site which is running single-phased and is acting like a phase convertor, back feeding one phase.

I am sure that there are more options to follow as others jump in.

PS: Never mind the original PS, I was having problems following the color code in the diagram....
 
Last edited:

Smart $

Esteemed Member
Location
Ohio
My initial impression is the B and C voltage and current relationships are reversed (Ib reported as Ic, Ic reported as Ib). Can't say for certain without knowing what the loads or reverse power sources were at the time the measurement was taken. Someone made the determination it was nuisance tripping... would help to know the criteria by which that decision was made.

PS: WELCOME :thumbsup:
 
Last edited:

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
A leading current implies either a simple capacitive reactance (oversized VAR correction?) or a non-linear capacitive-like load such as a power supply.
It is uncommon in the wild, and is rarely seen in captivity.

My first thought would be that the measuring equipment is screwed up.
My second thought is that there is a really wild combination of single phase loads attached to the source.
My third thought is that there may be a large three-phase motor on site which is running single-phased and is acting like a phase convertor, back feeding one phase.

I am sure that there are more options to follow as others jump in.

PS: Never mind the original PS, I was having problems following the color code in the diagram....
I agree, especially with thought #3 or the over sized PF correction system. I've seen that happen on systems where people use bulk PFC caps rather than having them come on with the offending motors. Even if it's an automatic PFC system, often times the available steps of VARs are too large to avoid over compensating, ESPECIALLY with a "wildly" variable single phase load distribution. It's very common for people to add on single phase systems willy-nilly after the initial installation, undoing all the hard work the original EC put into proper load distribution.
 

Darksabre

Member
Location
United States
Occupation
Electrical Engineer
Thanks for the replies and welcomes already. :)

I agree, knowing the load characteristics would be very helpful. I've sent out an email asking for more info on that. I've also asked for them to confirm the test lead connections (for what good that will do...) and take measurements on the load again.


I'm trying to nail it down, but from what they are telling me, they are seeing this "nuisance" trip when they transfer from Generator to Utility (two sources connected to a load via an Automatic Transfer Switch). The reverse power relay is supposed to protect the Utility from any potential backfeeding by the generator if the two sources are paralleled for too long during a closed-transition transfer. Basically, every time they run a generator run/transfer test, they come back and see that the relay has tripped.

I don't yet know if the relay is tripping before, during, or after a transfer has occurred. I definitely need to find this out! If it is tripping after the transfer from generator to utility, is it possible to see some backfeeding if the load has high capacitive characteristics? This is where 3-phase starts to hurt my brain. :? They also have the protective relay set very low (i think 1.5 Watts @ 1 second). This is the setting required by the utility company.
 

Darksabre

Member
Location
United States
Occupation
Electrical Engineer
Oh, I agree. I just wanted some quick looks at that 3 phase vector to see if anything was obviously wrong. I'll report what I hear back.
 

SG-1

Senior Member
I would like to know the service voltage & generator voltage during the re-transfer. If the generator voltage is higher it should be expected to send power into the utility. The relay is going to see that, especially if it is set near one watt. I do think the system should be re-checked for mis-wiring.

One easy solution I see would be to set the parallel time below one second, if the relay is set at one second.

Until then here is some interesting reading:
http://www.basler.com/downloads/solutionsforunconv.pdf

Welcome to the Forum !!
 
Last edited:

Darksabre

Member
Location
United States
Occupation
Electrical Engineer
I would like to know the service voltage & generator voltage during the re-transfer. If the generator voltage is higher it should be expected to send power into the utility. The relay is going to see that, especially if it is set near one watt. I do think the system should be re-checked for mis-wiring.

One easy solution I see would be to set the parallel time below one second, if the relay is set at one second.

Until then here is some interesting reading:
http://www.basler.com/downloads/solutionsforunconv.pdf


Thanks for the link - i will check it out.

The transfer switch has a max voltage difference setpoint before it will let the two sources sync for a closed-transition. I will have to check what it is, but it is usually pretty low, like 2-4% (then again, the relay setpoint is pretty low also, haha...). As far as parallel time goes, the max allowed parallel time is 100ms via a parallel limit timer, separate from the relay. This particular utility company requires a reverse power relay in case the automatic transfer switch faults somehow and is stuck in a parallel situation for more than a second. An actual CT transfer takes maybe 20-30ms. That's why I was thinking it's something with how the load is acting after a transfer back to utility
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
Thanks for the link - i will check it out.

The transfer switch has a max voltage difference setpoint before it will let the two sources sync for a closed-transition. I will have to check what it is, but it is usually pretty low, like 2-4% (then again, the relay setpoint is pretty low also, haha...). As far as parallel time goes, the max allowed parallel time is 100ms via a parallel limit timer, separate from the relay. This particular utility company requires a reverse power relay in case the automatic transfer switch faults somehow and is stuck in a parallel situation for more than a second. An actual CT transfer takes maybe 20-30ms. That's why I was thinking it's something with how the load is acting after a transfer back to utility
There was a thread earlier in which a reverse power relay was being tripped because of the presence of very large UPS systems which backfed the AC supply when the mains voltage was lower than the generator voltage or the UPS output voltage. I do not remember the details, but the symptoms were similar.
Part of the cause was that the additional sources took longer to stabilize after the transfer than the reverse power relay would allow.
That thread may have included multiple feeds to the complex or interesting control of UPS function during the moments of transition from mains to generator. I will try to find it.
The ultimate diagnostic was that a power monitor did show substantial reverse current just after the transition.
 
Last edited:

mivey

Senior Member
My initial impression is the B and C voltage and current relationships are reversed (Ib reported as Ic, Ic reported as Ib).
My impression was that the currents were mis-labled where a,b,c were labeled as c,b,a PLUS the clamps were put on backwards. Changing this gives:
Va = 295.333@0.00d
Vb = 296.423@239.00d
Vc = 296.579@119.00d

Ia = 40.126@345.00d
Ib = 39.631@205.00d
Ic = 29.228@79.00d

and power factors of
a: 0.966 (15d lag)
b: 0.829 (34d lag)
c: 0.766 (40d lag)
 

Smart $

Esteemed Member
Location
Ohio
My impression was that the currents were mis-labled where a,b,c were labeled as c,b,a PLUS the clamps were put on backwards. Changing this gives:
Va = 295.333@0.00d
Vb = 296.423@239.00d
Vc = 296.579@119.00d

Ia = 40.126@345.00d
Ib = 39.631@205.00d
Ic = 29.228@79.00d

and power factors of
a: 0.966 (15d lag)
b: 0.829 (34d lag)
c: 0.766 (40d lag)
While I can picture that if circumstances where normal, but not for a reverse-power trip condition...??? The phasor diagram source, timing, and likely several other pertinent factors are missing. Trying to nail it down from here is like swinging your arm as if to strike the nail, but without the hammer :happyyes:
 

mivey

Senior Member
While I can picture that if circumstances where normal, but not for a reverse-power trip condition...???
So you are thinking the data was captured during a reverse power condition?

The phasor diagram source, timing, and likely several other pertinent factors are missing.
As well as any transient data and wave shapes. That's why I thought we were seeing more of a steady-state set of values.
 

Smart $

Esteemed Member
Location
Ohio
So you are thinking the data was captured during a reverse power condition?

As well as any transient data and wave shapes. That's why I thought we were seeing more of a steady-state set of values.
Well, yes. Some condition is tripping the relay. If not tripping during steady state condition, I can't imagine why they would send a capture of such phasor diagram.
 

Phil Corso

Senior Member
Gentlemen...

Deriving power, P, or reactive volt-amperes, VAr, by calculation will only work if the wave shapes are sinusoidal!

On the other hand using a Watt-meter will always provide a true power measurement, regardless of wave-shape! So, what are the associated power measurements?

Darksabre...

If the measurements shown are indicative of reverse-power operation of the generator, w/excitation, then they represent V and A of a synchronous-motor! Hence, the currents shouldn't be different!

Regards, Phil Corso
 

ATSman

ATSman
Location
San Francisco Bay Area
Occupation
Electrical Engineer/ Electrical Testing & Controls
Darksabre
We need to know at what instant in time were those phasor values recorded and at what point in the system:
>Steady state before transfer
>During transfer of load
>Steady state after transfer
also would like to know the nameplate info of the ATS, manufacturer? model#?, etc.
 

ATSman

ATSman
Location
San Francisco Bay Area
Occupation
Electrical Engineer/ Electrical Testing & Controls
I was too late in the editing so here it is:

We need to know at what instant in time were those phasor values recorded and at what point in the system:
>Steady state before transfer
>During transfer of load
>Steady state after transfer
If it was recorded during transfer then the readings could be normal and the 1 Watt trip setting could be set too low.
Are any breakers tripping other than the one controlled by the 32 relay?
also would like to know the nameplate info of the ATS; manufacturer? model#?, etc.
100ms TD for paralleled sources is pretty much standard.
 

mivey

Senior Member
Well, yes. Some condition is tripping the relay. If not tripping during steady state condition, I can't imagine why they would send a capture of such phasor diagram.
Could be either way. some relays have maximum torque with positive values (I think for those that read single direction) and others have maximum torque for negative values (bi-direction readings).
 
Status
Not open for further replies.
Top