Distance protection testing - Accuracy lines 1

[521AB]

All the manufacturer (to my knowledge so far) give the static accuracy of the distance protection relay at the reach point, at line angle.
In practice the tolerance is not defined for all the characteristcs (unless you take the accuracy data for voltage and current measurement, and get crazy to translate it in an impedance error...).
More: test sets define the accuracy lines in their own way. 5% here is not the same as 5% there..
Are you also in this black hole? Any idea on how to get out of it? Or.. do we need to get out of it?

 

[riteshbharat]

Distance relays are fast being replaced by line differential relays in most parts of the world, and hence the inaccuracies and under/over reach shall no longer be an issue. Until then, a distance relay accurate at reach point at line angle, is good enough.

My Personal View: I have seen that many users make unnecessary issues by comparing a relay with 1% accuracy to another with 2%. I think that relays are not meant to be 'super' accurate devices; the only thing they should do is operate for a fault and not operate in healthy conditions. As long as they do this, I do not care for their preciseness.

 

[DTR2011]

"Distance relays are fast being replaced by line differential relays in most parts of the world, and hence the inaccuracies and under/over reach shall no longer be an issue. Until then, a distance relay accurate at reach point at line angle, is good enough."

Hold on there cowboy.....

Please tell us how a 400 mile line is protected with line diff relays?

Please also explain where the budget for all of the comm equipment comes from?

There are still many many lines protected by simple step distance schemes out there. There are also many that use DCB/POTT/PUTT and line carrier equipment (economical comm via power line carrier).

I travel the USA and see this in all kinds of substations. Of course for new stations, this may be, but there is the old saying "if it ain't broke, don't fix it".

Perhaps locally, you may see line diff used in shorter line applicaitons, but there are physical limitations as to the fiber length (even with repeaters).

Even if it is possible to hop onto a larger network, there are always the same issues of dependability vs. security.

Sorry, if I did not address the OP question.

 

[521AB]

I also do not agree with the fact that distance relays will be replaced by line differential protection. There are too many advantages in distance relays.
Anyway I agree on the accuracy, but when you have to set the tolerance lines on the test set, you have to do it otherwize automatic tests will give you a report of 20 pages with "FAIL" !!!

 

[riteshbharat]

Smallgreek, I agree with the point that in some cases budget might be a constraint and in some cases it might be impractical (for various reasons), but good things always come at a price.

I have no doubt that current differential is more suitable for line protection. OPGW would be a good option for communicating as it runs along the transmission line.

521AB, while using automatic testing, one must increase the tolerance to 10% otherwise no relay would pass the test in all conditions. Also if you could enlighten me by listing the advantages of distance relays over current differential, it would be great.Thanks.

 

[521AB]

........if you could enlighten me by listing the advantages of distance relays over current differential.....

distance protection measures local currents and local voltages. Line differential is dependent on the communication link. If it is out of service, protection is out of service. THIS means you cannot just use it! You need another protection as back-up!

Also distance protection uses communication schemes, put if it is not availabe, distance protection is "back-up of itself" and still protects the line, maybe with an additional delay on the trip time, in some parts of the lines.

Line distance protection can only understand "fault inside" or "no fault inside". It doesn't understand "fault outside". I.e. it cannot be used as remote-back-up protection (this means: if one CB does not trip, for any reason, a remote circuit breaker will trip. This is usually done with some other protection zones (back-up zones are often called).

Line differential protection needs today a digital communication link, often this is NOT under the control of relay personnel, but belongs other departments if not other companies. If there are trouble (and there are) on the protection, you never know if they are on the communication media or in the relay (actually relays), because "people do not talk to each others" (I have simplified, hope you understand what I mean).

Testing (commissioning and maintenance) line distance is also difficult, if one "box" belongs one company, and the other box belongs to another company. this because you need to coordinate different people.
Line distance is not affected by that, I just need to verify that if the carrier is received, and local fault detection appears, MY relay trips instantaneously.
If the carrier is not received, it is clearly NOT my problem (communication problem or other relay's problem). I can prove in few minutes that the probem is not in "my station" (by using digital disturbance recorder built-in in the todays relays).

Distance protection relays are more complicated in settings, but: line distance protection has also a lot of settings: communication card in the relay, multiplexer (clock master or not?), alarms for short interructions etc... At the end... I prefer distance protection relays.

 

[riteshbharat]

Thank you very much 521AB for your valuable inputs. I agree with your points regarding backup protection in non carrier aided distance relays.

Now a days, most of the line differential relays have distance back up as well, which get enabled when the communication link fails - such as MiCOM P546, 7SD522, SEL 311L, L90, REL551 etc. Hence these relays provide best features of both differential and distance. In Australia these relays are very widely used. In places where there is no communication, they use it as pure distance relays for the time being...and in future when the link is established they will only have to enable the differential element.

 

[521AB]

Which means that you are always using a back-up protection for the line diff. Line diff cannot be used as "independent" relay.
The fact that you are using integrated functions in one device, it is just a technological possibility. Still you are using line diff and something else..

Anyway, my topics was not about that... :-(.
Anybody with some other inputs on static accuracy of distance protection?

 

[PhilGavin]

Typically, I use 5% for static and some digital, 3% on certain digital relays, 7 % on some electromechanicals.
I have had to stretch some starting reaches to 10%, but I use 10% tolerance rarely.

As a rule though I would use 5% as a starting point and let empirical experiance dictate a tightening or slackening of tolerance.

I use the omicron description of Z tolerance: That is % of reach at the intersection of the line angle, transfered to the whole characteristic. This has limitations as changeing the line angle changes the tolerance.

Time tolerance would be typically, setting plus a typical operate time with +/- 5%

The power utility's setting policy itself also dictates the acceptable tolerances to a degree.

For example: A utility setting Zone 1 at 90%, is reliant on smaller errors than a utility that typically sets Zone 1 at 80%. Last thing you want is instantaneous tripping for a fault on an adjacent line.

I prefer to see a 'Fail' every now and again, and have to explain it due to testing to tight tolerances, than to create a false security with very slack tolerances.

 

[521AB]

Thanks PhilGavin,
the only think is that my measuring the border accuracy you measure nothing about the relay (almost nothing). If you slowly ramp, than you are just measuring the accuracy of the measuring element. The fault does not "slowly ramp".
Some numerical relays do not even tolerate a slow ramping, as they are based on jumps in the measured quantities.
If you "jump" from a steay state pre-fault situation into the steady-state fault situation, you also know nothing about overreaching because the dc transient is not injected, CTs are not saturated, load is not exported and neither imported.... so you cannot extrapolate anything.
Testing with constant fault current does not even allow you to control the source impedance behind the relay....
The only think I believe we can just say about the relay characteristic test, is that "it is healthy", so who cares about 5%, or 10% or whatever?
Sequences of steady state are helpful, but to my opinion not to "judge" the relay by testing the accuracy of the characteristic. Of course if you see a big error, like 30%, maybe the relay has problems, but not if the error should be 2% or 5% or 3.5%. For instance we can simulate shots to test autorecloser, weak infeed situations etc. From this point of view I see no need to inject the fault with the dc transient, but for overreaching of zone1.... no. The performances are in the technical documents of the manufacturer, not in the test set.

Another think I believe is good for testing characteristic is that "in one shot" you verify your settings. It is difficult to detect that one setting is 51 ohms instead of 5.1 or viceversa, among maybe 20 different settings..

Where I maybe can see need for accuracy test is close to the load area. There the load can "slowly linearly ramp" into the characteristic, but then things change again when the circuit breaker opens, healthy phases change etc. So, even there, mayb, it is not worth to investigate on the accuracy.

Just some thoughts...

 

[DiscoP]

We test the reaches on our distance relays using rapdly changing magnitudes of current that hunt inwards to find an operate value - it is done using a Doble Pro-test macro.
We use a single shot fault to test the timers are set correctly - usually at 85% of the reach.
Tests like transient overreach, we would class as type testing which we would do on one relay during evaluation if necessary.
Maybe we are missing something with this approach so I am reading this topic with interest.

 

[PhilGavin]

Hi 521AB,

The flaw in your logic is the implication that a relay that is innacurate during steady-state tests, will suddenly be accurate during dynamic conditions.

In essence, this is the worth of steady state testing.
If a a relay is innaccurate steady-state, it hasn't got a hope of being accurate during dynamic conditions.

 

[521AB]

Phil,
I did not mean that.
What I am trying to say is that accuracy in steady-state testing is not related to performances in dynamic conditions (performances is different than accuracy).
So I cannot extrapolate any information for the real behavior of the relay (I am talking about distance protection right now) by testing accuracy of the caracteristic within 2% for instance.
On the other side, steady-state inaccuracy might tell you some information on how the relay is healthy or if the settings are wrong, so the test is not useless (for vommissioning purposes).
The manufacturer should perform the dynamic tests and publish the results for the dynamic performances, and these results, to my understanding, cannot be extrapolated from the results od steady-state tests.

N.B. Steady state test is, from the electric point of view: injection of steady-test prefault and steady-state fault quantities, NOT from the relay point of view, where a steady state test is a test that does not have any "jump" in the measured quantities (this would be dynamic test for the relay).
For dynamic test I also mean from electric point of view, the fault contains the dc component transient.