Motor Differential Protection

[EP007]

Our team is discussing the need for motor differential protection on a 3000 HP motor, 4.16 kV. There is high resistance grounding, 10 A, which is important since a stator ground fault, is low current, and can be cleared quickly by the MPR (motor protection relay), without any time advantage of differential protection. Likewise, high level or solid 3L, LL faults and unbalance can be cleared quickly.

What we are left with appears to be low current LL stator winding fault, with a high impedance in the fault circuit, if that is possible. Does this sound correct?

 

[ScottyUK]

Most heavy faults start out as low-current faults, if only for a few cycles before they become heavy faults. Detecting and clearing the fault in those few cycles makes the difference between a rewind and a new stator core. A 3000HP motor isn't going to be off the shelf - do you have redundancy in your process or is this a critical motor? A rewind could be turned around in a week or so, maybe a little less. A new stator will be a month or more away.

Don't look at the cost of protection compared to the cost of the motor, look at it as an insurance policy for your process.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[EP007]

Good points ScottyUK. There will be redundancy in motors. I was thinking the high resistance neutral grounding, limiting the ground fault to 10 A would allow us to trip instantaneously on 1 or 2 A ground fault. However, if it is lower, that would be something else, assuming we would even set a differential that low. After your reply, I think I am now on the side of paying the small extra money, even if the benefit is not precise.

 

[ScottyUK]

Plain E/F protection won't see a fault close to the neutral point, or not before it has gone terminally bad. Stator diff protection will. That's the big benefit.

I'd definitely want a stator diff relay on a motor that size, and with the (relatively) low cost of a multifunction relay these days it's an even easier decision.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[oldfieldguy]

Most of my former clients used motor differential protection starting at the 2000-2500 horsepower level.

Using the core balance scheme on six-lead motors, though, may present some problems with long leads and low-saturation CT's.

old field guy

 

[EP007]

It looks like they want to have the core balance CT at the motor, in which case we will need to run CT wires back, sized accordingly. In this case they are only 50-80 m from the relay.

 

[electricpete]

There is no doubt the conventional wisdom that large and expensive motors are protected by differential and the inherent assumption seems to be that somehow differential provides better protection.

The analysis of the original post makes sense to me. In theory a ground relay should be able to detect anything that a differential relay detects except a phase to phase fault ( at what sensitivity it can be detected may bear further discussion... but seems to me a core balance CT might be more sensitive to ground fault than differential CT).. Phase to phase fault without phase to ground seems very unlikely....and if it should occur it would surely be in the endwinding... no core to worry about there.

So I agree it's an intriguing question. Here's what IEEE 37.96-2000 says on the subject:

Motor circuits complying with the NEC are required to have one overcurrent unit (series tripping device, protective relay, or fuse) in each phase conductor, or other approved means. When fuses are used, the consequences of unbalanced operation and backfeeding of faults following a blown fuse must be considered. IOC relays (50) are used to detect motor supply cable faults as well as severe stator faults. They may be connected to trip directly, or through a short time delay to coordinate with the asymmetrical starting current when set just above the locked-rotor current (see Figure 37 and Figure 38). When the motor kilovolt ampere (kVA) rating is less than half that of the transformer (as a rule of thumb), instantaneous relays can be used for phase protection. Where the starting current value approaches the fault current (motor kVA) greater than half of the transformer rating), differential relays should be used.

So... the decision on whether or not differential protection is required seems to depend on rating of motor compared to transformer!?! Is it related to short circuit characteristics of the supply? I doubt it. Related to some kind of selecting coordination of motor tripping compared to transformer tripping? Doens't quite seem right. I have to admit I can't quite make sense of it at the moment.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

I re-checked the quote and it is quoted above exactly as in the standard, including unmatched end-parantheses in last sentence. Presumably is supposed to read:
"Where the starting current value approaches the fault current (motor kVA greater than half of the transformer rating), differential relays should be used."

So, maybe they are worried that the instantaneous overcurrent cannot be set tight enough and a phase to phase fault might occur at some level below the instantaneous overcurrent trip?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[edison123]

I'm with Ccotty. At this power range, a stator core fault will be ugly.

Scotty - They turn around motors after full rewind in one week in UK ?

Muthu

http://www.edison.co.in

 

[edison123]

Sorry Scotty. Grrr...

Muthu

http://www.edison.co.in

 

[ScottyUK]

You think that is fast... or slow?


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[edison123]

That's too damn fast. My shop takes about 3-4 weeks for that kinda rewind. Do they keep the pre-made stator coils in stock ?

Muthu

http://www.edison.co.in

 

[ScottyUK]

Yeah, we have a stock of pre-formed coils for critical machines and will pay for 24/7 working to get it back. Expensive, but so is downtime.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[edison123]

A ha, that explains it. Thx.

Wish I had clients like you.

Muthu

http://www.edison.co.in

 

[electricpete]

I don't think the original poster or myself would disagree that a core fault takes much longer to repair. But the question is why does differential protection provide better protection? Does it offer more sensitive or faster tripping than ground protection and instantaneous overcurrent protection for some scenario? What is the scenario?

(I dont' doubt there is a scenario since as I stated above there is conventional wisdom... but the scenario is not clear to me).

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

Stepping back, I think Scotty's response was a good one.

To make it a little more satisfying for myself I add the following: Perhaps adding differential provides redundancy and diversity with other protection which increases the reliability and quickness of response. I could buy that, since response of relays to faults may not be 100% predictable.



=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[ScottyUK]

Thank you, ePete.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[EP007]

With high resistance grounding, and sensitive ground fault protection, we try for 5 A NGR, but this may be 10 A. Add to that the modern microprocessor based relay, and the situation has changed. We can limit the ground fault to low values, and we can sense low level ground fault currents and also clear quickly.

I think the philosophy of differential is from days prior to the above, where it was also either solid ground or low resistance grounded system, and old style relays. I think the damaged stators some repair shops report are from these types of systems.

All I can see, concurring with electricpete, is redundancy. Relays do fail. I saw in a generation plant, a failure, but they had redundant relays from different manufacturers.

But if the relay fails in my case, maybe even that core balance diff out at the motor will go to a failed relay.

A better alternative may be to use a redundant relay at the switchgear.