Breaker Monitor Device vs Relay for HV circuit breaker diagnostics

[rockman7892]

I'm looking into several options for monitoring breaker diagnostics as part of increased predictive/condition based maintenance for Substation HV breakers. Typical application is 138/230kV Substation with single HV breaker with radial feed to transformer.

In doing some research I came across dedicated breaker monitoring devices that can provide an array of breaker diagnostics from a single stand along monitoring device (INCON Optimizer 3 Breaker Monitor for example).

Looking at several of these monitors they all appear monitor key breaker diagnostics including the following:

- Breaker Timing
-Operations Count
-Trip & Close coil circuit integrity
-Charging motor current draw & run time
-Contact wear & restrike detection
-Sf6 gas monitoring

Does anyone have any experience with using this type of breaker monitor or similar device that have found it to be useful for breaker predictive based maintenance? Any recommended manufactures or models?

I also have seen applications where several of these monitoring functions can be programmed into a protective relay associated with the breaker (Breaker timing, TC monitoring, etc..)

Has anyone had similar sucess with using relay to perform monitoring for predictive maintenance vs using an external monitoring device?

 

[thermionic1]

- Breaker Timing - We use SEL 400 series relays for this via a 52a contact. The alarm for this goes to the equipment specialist rather than dispatch. It's important to capture the difference between main contact and aux contact timing during commissioning.
-Operations Count - These generally get screwed up over time during maintenance and commissioning. SEL relays have a breaker wear that's probably more accurate. Upgrading the protection on a two trip coil breaker will likely result in over 100 operations just doing commissioning.
-Trip & Close coil circuit integrity. We use SEL 400 series relays for this. The alarm goes to SCADA.
-Charging motor current draw & run time. Our breakers have an alarm timer for running too long, which is monitored by SCADA. I suppose the current monitor is handled by the Motor Loss of AC or DC supply UV relay.
-Contact wear & restrike detection. We use SEL 400 series relays for this. The alarm for this goes to the equipment specialist rather than dispatch.
-Sf6 gas monitoring. Our breakers have a low gas alarm and a low low lock out alarm. These are monitored by SCADA I/O.

Our preferred relay to implement this on is the SEL-451, which also acts as the BF, reclose and sync check device. In other special cases where a 451 is not installed, the same logic will go on a 487E, 411L, etc. The same proven logic can be copy / pasted between relays with ease.

I was involved with quite a bit of apparatus monitoring years ago. Maintenance crews sometimes view these devices as a threat and treat them as such. Too many nuisance alarms and people loose faith in the devices. Sunlight entering the windows in the breakers can cause alarms on FO cables from these devices. Sunlight entering the windows in the breakers can cause front screens and displays on these devices to fail.

I personally think using the SEL (or your choice) of relays is probably better long term. The relay, SCADA and commissioning people are already familiar with programming and operation of these devices. We have on line oil analyzers or various vendors, bushing PD, etc on some larger transformers. Many of these devices are problematic for a variety of reasons, including difficult programming, poor access to SW, exposure to elements (LCD screens failing).

 

[bacon4life]

For a radial sub feeding a transformer, there probably are going to be very few operations. My experience has been that contact wear and operations count fail to correlate well with end-of-life for most breakers. The exception might be for things that regularly operate like like capacitor banks.

One challenge with using IEDs for things like operation counts and breaker wear is that you have to transfer the maintenance records into new IEDs whenever IEDs are replaced. Adding that task into the workflow for relay technicians can be a significant change to how they archive & track relay settings. Even though SEL relays have good longevity, I doubt many of them will last as long the actual breaker.

For monitoring breaker wear with the SEL relay, you need to adjust how you test the relay such that currents injected during your relay testing do not count towards breaker wear.

If you use the SEL as the operation counter, the SEL must be online during all breaker maintenance activities that could change the state of the breaker. Depending how your crews are organized, using the SEL as operations counter might require additional coordination. Often our breaker maintenance crew is working on timing tests out in the yard while a totally separate crew might have the relays inside the control house isolated doing relay testing or relay firmware updates.

I do use SEL relay for trip coil monitoring, as well as collecting SCADA status from the SF6 pressure/density alarm. I have not tried using it for breaker timing monitoring, but that sounds promising.

 

[cranky108]

Strange you say that. I do have SEL relays that are more than 25 years old. Exactly how long are you expecting modern breakers to last? Even if they have few operations?

That said, the company I was with made part of replacing a battery set, to also replace the battery charger. So why not replace a breaker and a breaker monitor relay at the same time?

 

[bacon4life]

We have lots of transmission circuits breakers more that 40 years old, with some more than 50 years old. Hopefully the breakers we are currently procuring last at least that long.

We have ended up replacing a fair number of 100, 200, and 300 series SEL relays. Of the replacements, many were due to wanting features on 400/700 series relays rather than due to hardware failure concerns. We are also considering replacing SEL-451 relays just a few years old with newer SEL-451 relays that have harmonic restrain blocking. If the high impedance fault detection algorithms are ever perfected, there might be desire to upgrade again in a few years.

Jumping to medium voltage switchgear, we are ending up prioritized replacing non-SEL microprocessor relays as more urgent than replacing much older electromechanical relays. We have had several cased of this particular model of ~20 year old microprocessor relay failing in service.

 

[cranky108]

Have you looked into firmware changes in place of replacing the relays?

 

[davidbeach]

Firmware changes can’t add second generation 4xx features to first generation 4xx relays.

So would those ~20 year old relays be a “protection unit” that doesn’t monitor the function of the A/D? So they quietly fail with no notice? Any place we had them they were paired with something else set slightly slower with no reclosing. So any unexpected one-and-done on a feeder with those starts out as a suspected relay failure. Most are gone now, fortunately, and the rest are on their way out.

But, yeah, the EM relays will just go on and on. I won’t get to retire the last EM line relays on our system, don’t know if my eventual successor will get to either. Other drivers may result in the retirement of the last EM feeder relays before the last EM line relay. But we’re within a project or so of the 411L being our most common line relay, overtaking the 321.

I’ll see your silver lining and raise you two black clouds. - Protection Operations

 

[cranky108]

Such a shame the 321 relays are being replaced, and they were such a fast relay, but lacking in features. Good to pair them up as secondary relays, and there are very few actual relay failures in service. Maybe 1 out of 132 relays in service per year.
I once had a manager who thought why put in a second relay, as there are very few relay failures, and the second relay is doubling the cost of projects. He was very concerned about customer costs, but not about reliability.
I suggested we start wiring the alarm outputs to the trip, so we would not have any unprotected equipment.
I changed jobs shortly after that. I don't know what they did.

 

[thermionic1]

I hope you didn't! On SEL relays when Access level 2 is made from logging in, the Alarm output pulses for a second. Imagine someone from protection logging in remotely to change a setting, etc. and tripping a breaker. We open the alarm cut out in the field if we need to get into the relay, so we don't drive OPs nuts.

 

[bacon4life]

As David indicated, there are hardware improvements on newer 4xx relays that add additional features.

I am not sure the exact failure mode of the protection units.

One of the drivers for swapping out 300 series to 400 series relays is the avoidance of tying 2 CTs together for ring bus applications. With a single set of CT inputs, it can be easy to cause a misoperation during breaker testing/maintenance.

We had some worries that putting a test switch on the alarm output might result in the test being left open. Instead the dispatcher can inhibit alarm notifications for the expected duration of the relay work.