Problem with Double Wye Capacitor Unbalance Protection 4

[SMB1]

Dear All,

I am talking about Double Wye ungrounded Capacitor Unbalance protection using Neutral CT and neutral VT as explained in IEEE C37.99-2000 section 8.2.6.2.

i have 13.8KV capacitor bank double way ungrounded and both unbalance protection approches are used:

CT approach: CT will be connected to 51N relay in order to detect any fuse blown. the problem of this apprach in that no current will flow in the neutral in case of identical fuse blown in both legs of the same phase.

VT apprach: VT wil be connected to 59N relay to detect any unbalance. this apprach overcomes the 51N limitation.

the prblem i am facing is that 59N relay operated many time for other 13.8KV feeders faults while the 51N is stable.

any explanation please?

thank you,

 

[LionelHutz]

Info on how you installed this VT would be helpful. I'm assuming you went from the Y point to ground? If so, these faults likely unbalance the power network and cause the floating neutral of the capacitor banks to shift.

 

[SMB1]

thank you LionelHutz,

yes, the VT is connected from the neutral point to ground. the problem is what you sid. the netral point is shift due to any unbalance in the system which causes the relay 59N to operate.

So, what is the correct application of this type of protection?

 

[LionelHutz]

I have only ever used the CT.

I believe there are other methods shown in that standard which involving multiple PT's to account for the phase inbalance.

 

[SMB1]

But CT in the neutral can not detect identical fuse blown in both sides of the same phase.

 

[Marmite]

Unbalance protection is not meant to detect blown fuses. It is meant to protect the capacitors from being overstressed by series connected capacitor elements within the can shorting out and increasing the voltage experienced by the other elements in that leg.
Regards
Marmite

 

[SMB1]

Thank you Marmite,

I agree with you.

Usually we set the unbalance protection based on manufacturer unbalance calaculation which shows how much voltage will increase in other capacitor units.

the setting of the unbalance protection is set based on that calculation.

So, the objective is not to dtect the blown fuse but to protect the other capacitor units.

thanks

 

[busbar]

 
I think, per the formula in C37.99, that your scheme should detect open fuse conditions. Just make sure that the 59N relay is restrained for third-harmonic conditions.

 

[stevenal]

Perhaps the 59N should be delayed so that ground overcurrent line protection has a chance to clear the 13.8 faults prior to cap bank tripping.

 

[LionelHutz]

I missed your response, but I installed the CT to detect a shorted capacitor section just like Marmite has posted.

I never questioned it before, but is the 13.8kV system a grounded Y system? If so, you might be able to co-ordinate the 59N relay trip time with the other relays. However, if it's not a grounded Y system then that 59N relay connection will likely always be troublesome.

 

[SMB1]

thank you all,

busbar: yeas (blown fuse = open fuse)

Stevenal: i like the idea of delaying the 59N some time till the faulty feeder protection operate,

is it usual practice?

LionelHutz: shorted capacitor section will not cause voltage stress on other unit. I think we are taking about open fuse.

thank you

 

[stevenal]

Since I've never implemented this particular configuration of cap bank, I feel less than qualified to state what is usual. I would consider it unusual, though, to trip on out of zone faults. You may wish to look at Table 6 of IEEE 1036-1992, which lists permissible overvoltage versus duration points for capacitors.

 

[LionelHutz]

I'm not sure how you concluded that a shorted capacitor element won't over-stress the other elements. Maybe your capacitor bank has over-rated capacitors. Still, at some level of failure you will eventually over-voltage capacitor elements.

My primary concern when I built the double Y banks was to protect against shorted capacitor elements causing an over-voltage stress on the other capacitor elements.

I wasn't nearly as concerned with blown fuse detection, mostly because a blown fuse would not cause a capacitor can to explode. I considered the odds of either 2 capacitor elements or 2 fuses on the same phase failing at exactly the same time to be fairly slim enough I was not protecting against those cases.

I still believe one of those multi PT scheme's is the proper way to protect the bank against blown fuses. If you don't want to apply such a scheme then you will have to adjust the settings until you get the results you want, regardless of "usual practice".

Maybe you should consider a scheme such as Fig 11. Take a look at a Cooper Power iCP-440 relay.

 

[stevenal]

If the caps are fused, shorting will cause a fuse to blow. This condition should should be alarmed so that it is corrected before other fuses also blow. No need for the protection relay to operate for the short, it needs to detect the open fuse. Other fuses are likely to blow because the remaining elements in the series group will experience higher voltage and current. When enough fuses are open, the condition will quickly cascade. Now you have a condition where the faulted units cannot be easily found. Your protection should alarm on the open fuse condition, and trip when unit voltage exceeds 110%. See IEEE C7.99.

If using a fuseless capacitor, the protection philosophy shifts to detecting the short(s).

Both conditions cause overvoltage. Good equations for your spreadsheet can be found in C37.99.

 

[SMB1]

the attached paper explain in detail the unbalance protection in different cap. bank.

please see

 

[stevenal]

"See IEEE C7.99"

Meant C37.99.