Self-Powered Relay Limitations

[rba10]

What are the limitations of self-powered relays?

We are pushing a vendor of our 34.5kV switchgears to provide 3A pick-up on ground faults with a self-powered relay. The vendor says that he doesn't believe one exists. If that is true, what is the reason?

I'm new to the company and I'm not directly involved (at least not yet) with the issue but I want to understand it better. I have some basic knowledge of relays but obviously not to this degree (at least not yet).

 

[cranky108]

Electromechinical relays are self powered and have been around for over 50 years. And several we have have taps lower than 3 amps.

The limits are you don't have flexable curves.

 

[davidbeach]

There are also self-powered (actually CT-powered) solid state relays also. The DIAC family from GE and the BE1-50/51B family from Basler immediately come to mind. I know the Basler allow selection of curves.

 

[racobb]

I believe ABB may have them as well with their Micro AT line.

Alan

 

[cranky108]

I believe most of the relays will react like the Basler self powered relays, in that there is a slight delay caused by the power up sequence of these relays. If that dosen't bother you then these are a good choice.

However, if this delay could skuew your curve enough to cause a coordination concern, look at the electromechinical relays.

Another concern is the elecrromechinical and solid-state relays operate a little different on the IOC units. And to think or it there maybe a difference in how the relays respond to harmonics.

 

[davidbeach]

The start up times are very short, only a very small percentage of the tripping time. CT errors will generally cuase more trip time error.

 

[stevenal]

Don't the published curves include start up time?

 

[davidbeach]

There are published start-up "time add" curves. The published trip curves assume some level of load current so that the relay is turned on before the fault. Start-up time will be the least of the timing errors in a real system and if start-up causes miscoordination there was not enough margin to begin with.

 

[davidbeach]

All that said about self (CT) powered relays, the truth of the matter is that they should only be used as replacements for other older self powered relays. No new installation should use them, anything new should use a modern numeric relay that can provide more information after an event than just a target.

 

[cranky108]

There are dual powered relays available, where they can be powered by a PT and CT (so they should be powered most of the time).
But the questions is how do you intend to trip your breaker?

The capacitor trip is hard to find, other than the Basler unit. But the breaker needs to be sized for it (and the relay contacts).

Or is this a replacment application?

 

[rba10]

First, thanks for all the input.

Basically it was realized that the relay protection did not fully meet our desired pickup amp specification. We need a self-powered (or dual-powered) unit for safety reasons relating to our setup.

I understand the data collection advantages or maybe even a slight time advantage, what I'm having trouble understanding is why there is a difference in the pick-up amps between a self-powered relay or one with an external power supply. Is it an issue with CT powering of the relay at lower amps?

 

[davidbeach]

If the relay is powered from the CT circuit and there is no current in the circuit, how can the relay be powered? You don't want a lot of burden from the relay, so there is some minimum amount of current necessary to get a usable voltage and power level for the relay.

 

[rba10]

Ok that makes sense there is a minimum number of amps required for any CT to power a relay. In this case then I guess our vendor is saying that 3 amps is below that minimum value.

This may be a dumb question, but isn't it possible to power the relay using a CT or PT on one of the phase lines?

 

[cranky108]

There is a very limited area where a relay can be powered by either a CT or PT. The PT voltage can approch zero volts during a fault, so alone a PT powered relay won't work.
However a PT powered relay can have a pickup setting much lower than a CT powered relay.

So a dual powered relay is one option. But if you are using a battery to trip your breaker, then consiter a DC powered relay. If you don't have a battery, then a CT or dual powered relay is required.

If you are using a capacitive trip device, then higher rated contacts may be required.

 

[racobb]

What did your spec require or in other words, what did you buy? If you have station batteries maybe it is time now to do a change and get what you want up front.

Alan

 

[rba10]

The current setup uses power from the CT's to charge a trip capacitor to trip the breaker.

Our spec. requires a pickup between 3 & 10 amps for a ground fault. Where we were even getting ground overcurrent protection I believe it was like 50 amp pickup which is greater than the rated current let alone anywhere near the 20% mark.

 

[racobb]

Have never seen or heard (I guess until now) of the CT being used to charge a cap before it could trip a breaker. How does that work? Is there no AC available to keep the cap trip unit charged?

How do you manually trip it....electrically or with the manual spring release button?


Alan

 

[marks1080]

Agreed, this is strange.

Is there no DC system? Is there no AC station service? Is there lighting in this building? If so, there's got to be a better way of powering that relay. I would not trust a charged cap to supply enough trip current for a breaker on intuition.

 

[racobb]

rba10,

Are you sure you don't have a standard cap trip device that is powered by AC to keep the unit charged and trip the breaker?

Alan

 

[cranky108]

Some of the older capacitor trip devices did have a CT charging in addition to charging by a station power source.

Some relays (not of american design) can charge an internal capacitor from a CT alone. But given a large fault will charge the capacitor very quickly, and a smaller fault will charge the capacitor slowly, which I supose fits in with some relay curves.