SEL-351S CT Saturation Protection can someone explain simply ?

[bdn2004]

On a recent project an old 15 kV fused switchgear was replaced by new switchgear that uses HV vacuum circuit breakers and SEL-351S relays on the feeders that land on step-down transformers. Looking at the curves for the new feeder relays vs old E-rated fusing … the instantaneous setting for the new relays is set multiples from where the old fusing curves show. This drives the arc-flash up to Danger levels. But when questioning the engineer who made the settings, he constantly repeats the following that is typed out below. It sounds like something from the manufacturer. And he won’t adjust the instantaneous settings. Not sure I’m understanding what he’s saying here…. Can someone put this in layman’s terms and is he right?

CT Saturation Protection
The SEL-351S phase instantaneous overcurrent elements normally operate using the output of a cosine filter algorithm. During heavy fault currents when the relay-detects severe CT saturation, the overcurrent elements can operate the adaptive current algorithm.
The adaptive current algorithm is only used for phase instantaneous overcurrent elements if and only if the corresponding pickup setting is greater than eight times the nominal phase current. For example, if 50PI, 50A1, 50B1 and 50C1 = 45A (in a 5A nominal phase current relay), then 50PI, 50A1, 50B1, and 50 C1 elements operate on an adaptive current algorithm. However, if 50PIP = 35 A, then the 50P1, 50A1, 50B1, and 50C1 elements operate on the output of a cosine filter algorithm. No other overcurrent elements use the adaptive current algorithm.
Based on the level of a “harmonic distortion index”, the adaptive current is either the output of the cosine filter or the output of the bipolar peak detector. When the harmonic distortion index exceeds the fixed threshold that indicates severe CT saturation, the adaptive current in the output of the bipolar peak detector. When the harmonic distortion index is below the fixed threshold, the adaptive current in the output of the cosine filter.
The cosine filter provides excellent performance in removing dc offset and harmonics. However, the bipolar peak detector has the best performance in situations of severe CT saturation when the cosine filter magnitude estimation is degraded. Combining the two filters provides an elegant solution for ensuring dependable phase instantaneous overcurrent element operations.

 

[System Protection]

He literally just copied that straight from the manual.

He is not answering your question.

 

[dpc]

The relay has two algorithms for determining the actual maximum current from the digitized sample data. One method is designed to deal with potential CT saturation for very high levels of fault current. This methodology should not impact the relay settings to any meaningful degree. It's quite possible that CT saturation isn't an issue at all in your system - this depends on the magnitude of the fault current and the quality of the CTs.

I think they are misinterpreting the quoted description in the manual and believes that the IT pickup needs to be high enough that the bipolar peak detector comes into play. That is incorrect, imo.

 

[bacon4life]

The setting engineer should be able to provide a rational for the settings they picked. Everything quoted is an irrelevant digression into the internal workings of the relay.

Regarding the arc flash danger increase, circuit breakers take at least a few cycles to operate. If the setting engineer is so far into the weeds, it might be good to verify there arc flash calcs used total clearing time rather the just the relay operating time.

 

[cranky108]

The 50 function in a microprocessor relay is not the same as what would be found in an electromechanical relay, and the difference in settings may be a higher setting, or a slight time delay, either way does not exactly follow the 50 function in the EM relay. When faced with a slight over current a 50 function in an EM relay will have a slight time delay (you can look up those curves) but the microprocessor relay will not.
This might be part of what you see. But if the old fuses were current limiting the breaker will not be fast enough to match it.

 

[9wearSuave]

I agree with other sentiments that the engineer should provide justification for the setting values. I think it’s worth noting that the instantaneous element of a relay usually isn’t used for mitigating arc flash, but more for protecting equipment from serious damage. Arcing current *usually* lands in the long/short time regions of the inverse curves

 

[davidbeach]

Low-set instantaneous elements activated when personnel are around the equipment are absolutely a means of arc flash hazard mitigation.

When one this sentence into the German to translate wanted, would one the fact exploit, that the word order and the punctuation already with the German conventions agree.

-- Douglas Hofstadter, Jan 1982

 

[bdn2004]

Thanks for the good answers. Sounds like the consensus is that he's not answering the question. Might be worth more investigation and contact Schweitzer to confirm.

Seems the brand new switchgear should have the quality of CTs installed for the application. The high arc flash occurs on the secondary side of the transformers at the 480V switchgears at the terminals of the mains. We had a Client once tell us ... they know it's dangerous - but they want to know if the calories are 50 cal per cm squared vs. 100 cal per cm squared. And that's the magnitude of this very high setting of the instantaneous is having on what it was with the fusing.

 

[9wearSuave]

Hi Davidbeach,

I understand maintenance mode settings drop the instantaneous setting low temporarily. I was referring to the af levels during normal operation.

 

[dpc]

Seems the brand new switchgear should have the quality of CTs installed for the application.

This is not always the case. The ANSI minimum for switchgear CTs is pretty low. If you want a higher accuracy class, you have to specify it.

For arc-flash reduction in this situation, a maintenance-mode setting for the instantaneous as David Beach referenced will give you a substantial reduction in arc-flash incident energy.

 

[SEL487]

Why contact Basler for a SEL-351S relay?

Does the engineer have access to software to model the TCCs? You're right to call them out on it.

 

[bacon4life]

CTs sized to be immune from saturation may not physically fit inside of medium voltage switchgear. It is perfectly reasonable to size CTs with some amount of saturation as long as the design aligns with the relay capability for detecting saturation.

 

[cranky108]

I don't know about that. We limit our fault current to about 10kA with a higher impedance transformer, and have 2000/5 CT's on 12kV so maybe a compromise needs to be included in the design to make this work.

 

[bdn2004]

oops dealing with Basler on another model. I edited it to Schweitzer. Maintenance mode is a good idea - if it comes with that.
That's a good idea... specify the transformers with higher impedance to limit the fault current.

This project was a part of a purchase of a large portion of an existing industrial plant and upgrades to be made into a separate company.
The engineering Contractor did model it on a power software like ETAP or SKM. But the entire Plants combined is modeled in EasyPower, which is why we are involved.
We just copied in what they did.

I sent the question to EasyPower also - they want the file, and let their experts have a say.