Megawatt Transducer Calibration

[mgtrp]

Hi,

We are trying to check the calibration of a MW transducer and display gauge. The meter is configured for 2 CT inputs and 3 voltage inputs. Unfortunately, the technicians test gear isn't too happy creating the wye-connected B-phase-grounded system that the transducer is expecting to see.

Does anyone have any bright and simple ideas for testing this? I'm thinking of either insulating the meter case and connecting B-phase to the meter ground terminal (a nuisance, as some of the transducers are awkward to remove), or attempting to scrounge 3 small isolating transformers (we are in a remote area, so may not be able to find any).

Thanks.

 

[Skogsgurra]

Do you need an "NIST" calibration? Or just a sanity check? In the latter case, it is usually possible to measure (a precise recorder or data taker helps) the actual currents and voltages and do a manual calculation (if power math not supported by the recorder) so that transducer output can be compared to what you can assume being the correct value.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[mgtrp]

It's needed for a sanity check - we have several displays reading substantially different numbers, and we are trying very hard not to overload our poor generator that really should have been replaced so want some confidence in our readings.

I'll ask around, but I don't think that we have anything capable of measuring power factor to give us a reliable way of calculating MW based on the VT and CT signals while running.

 

[Skogsgurra]

A scope goes a long way. Even a single channel scope with external (or line) triggering will work. Two channels are, of course, better and four channels are excellent.

If you cannot measure, you can perhaps add known loads and see what display best reflects the added load. That one is probably correct.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[waross]

Do you have any capacitor banks in the system? You can get enough information from VARs and KVA to determine the power factor.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[racobb]

Measured Volts and Amps will tell you a lot, especially if you have some historical pf info, or even a close idea of what the pf might be.

Alan

Democracy is two wolves and a sheep deciding what to have for dinner. Liberty is a well armed sheep!
Ben Franklin

 

[pwrengrds]

Is this a delta connected PT?

If so the test set shouldn't have a problem with it. Use the voltage neutral as B phase and connect to B input, set A phase to 30 degrees and connect to A input, set B or C phase of the test set to 90 degrees and connect to C input. (that creates a delta PT secondary connection) Connect the currents as they are labeled and it should work. Then to change the power factor rotate the currents as needed.

 

[mgtrp]

This is a 3-phase, star/star-connected VT, 7.2 kV / 120 V nominal ratio giving 69V phase-neutral when running. The meters are connected to CT's from A and C phase, and the VT's for A, B, and C phase. B-phase is connected to earth, and the star point is floating. I don't know if this arrangement is used in industry much, but it is not entirely uncommon in generating plants of this vintage (1960's).

After a night's sleep, I have gone through the vector diagrams, and once the test set is free, will provide the following voltages from the test set:

Va-n: 120V angle 0
Ia: 5A angle -30
Vb-n: 0V
Ib: 5A angle -150 (note, not connected)
Vc-n: 120V angle 60
Ic: 5A angle 90

I believe that this will give me the same voltages and currents as the meter would normally see with unity power factor.

If this doesn't work, we will connect an oscilloscope and check the voltages and currents with the unit running, and calculate power manually to check against the meters.

Thanks.

 

[pwrengrds]

I have seen this, its a way of getting a delta type connection using YY connected PT's. The delta connection that I referred to above should work. Your way is the same with a 30 degree phase shift on voltage and current.

I find just shifting the voltage once by 30 degrees then moving the all the currents around to change PF makes things simpler and less error prone. That way at PF of 1 the currents are 1, -120, 120 degrees.

 

[mgtrp]

Thanks for that pwrengrds. Good idea with setting the currents as normal and only shifting the voltages.

 

[busbar]

Have you considered a single-phase test? Series the current inputs and parallel the potential inputs.

 

[ScottyUK]

That would work if the meter is three single phase units internally. If any of the meter's internal magnetic components are shared - a three-phase VT for example - then then strange things may well happen when fluxes sum together.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[mgtrp]

The meter is meant for unbalanced circuits, so I assume that it would work if we single-phased it.

Regardless, we have injected the currents and voltages into two sets of meters as suggested by pwrengrds, seemingly successfully - one set of meters seems to have inserted a 10 degree phase shift into it's readings, the other appears to be scaled incorrectly in SCADA.