Transformer differential protection 1

[Gauss2k]

I am programming a digital relay that will be used to protect a transformer from overcurrents (50-51-50N-51N) and internal faults (87T).

The settings for the 50-51 were determined by a coordination study, but I am unsure of which values I should use for the differential protection.

I have looked in my local "electricity code" (Quebec, Canada), but I haven't found any details about that. The IEEE Buff Book talks about the differential protection but do not seem to mention any precise recommended settings.

The relay default settings are Ids=30%In and Id/It=15%
(where Id is the differential current: |Iprim-Isec| (normalized current) and It is the through current: MAX (|Iprim|,|Isec|))

Should I keep these settings? And why, or why not?

 

[dpc]

It's hard to give specific advice without knowing more about the transformer, the relay, the CTs and the service.

In the old days, there were few adjustments on transformer differential relays - we matched up the CT ratios as best we could and that was about that. The slope and pickup were limited in adjustment. With the digital relays, we have more options.

GE/Multilin SR-745 documentation has a little tech paper on setting the parameters - this should be available on their website. I suspect Schweitzer also has suggestions.

The default settings will probably work fine unless you have something very unusual in the xfmr. On the other hand, if your transformer does not have a load tap changer, you may be able to reduce the differential pickup setting. The default setting has to account for the possibility of an LTC causing the CT ratio mismatch to vary as the taps are changed (along with CT error and other factors).

 

[RalphChristie]

I agree with dpc.

Factors to take into account before applying /setting differential protection are:

Transformation ratio: compensation for the different CT-ratio's on the two sides of the transformer and also the possible mismatch between the CTs
Transformer connection: not really an issue with modern relays, (it is done automatically inside the relay) but it is needed for the possible phase shift between primary and secondary.
Inrush currents: also not really an issue with modern relays because they have harmonic restraint units or a time delay during an inrush condition (during energizing)
onload tap changer if the transformer has the benefit of changing the transformer ratio.

The determination of the differential settings will be such that there will be no differences in the currents in the restraint windings (to avoid mal-operations) due to the reasons stipulated above. Thus, your settings must be set on such a way that during the biggest normal unbalance (like for instance the tapchanging) there ain't a trip.

There are a lot of books written on this subject, (Blackburn, Elmore, Horowitz, etc) and it would also be most probably in the relay manual.

 

[Gauss2k]

The CT connection and ratio are already configured in the relay (it does allow to ajust the vector shift no matter how the CT are connected)

Since the diffential protection is 87"T", according to the IEEE buff book, it means it has an harmonic restraint unit, so the inrush current should not be a problem.

I'll have to check for the tap changer though, I'm not the person who went to see it and make the coordination study. If there is one, how can I know the imbalance it will create when it is operated? The transformer manufacturer? Basically, I understand that there is no "recommended" or standard slope and pickup value based on the type of transformer or other factors?

BTW, the relay is a "Merlin Gerin" and the software I use to configure it is "SFT2841". It's pretty good since it allow to customize about everything (all thresholds, messages, leds, etc..)

 

[RalphChristie]

In the ideal differential scheme there would be no difference in the secondary currents from the different sides of the transformer. For any unbalance your diff protection should then operate, because it indicates a fault. In the real world, however, there are always mismatches. In your instance there would be probably CT-errors, ratio-mismatches and a possible mismatch due to the OLTC. What you have to do is to determine the possible mismatch by calculating the secondary currents into the relay from the different sides of the transformer. For your OLTC you have to see what is the percentage change your tapchanger can give from the nominal tap to the lowest/highest tap. You will find it on the nameplate of the transformer. Your total value would be mismatch (%) + OLTC (%)
This is your maximum spill current (percentage) under normal conditions.


Elmore gives in Protective relaying theory and applications the formula to calculate mismatch (M) as:


M = [ ( (I low / I high) - (T low / T high) ) / S ] x 100%

where

I low, I high:
relay input currents, at same kVA base, for low and high voltage sides, respectively

T low, T high:
relay tap settings for low and high voltage sides, respectively (1 in your instance)]

S:
smaller of the two terms (I low / I high) or (T low / T high)​