Coordination between Tie & Main Breakers: Spot Network 1

[Laplacian]

What is a general rule of thumb for coordination timing between the time-overcurrent clearing for main and tie breakers operating in a spot network configuration? If I add 0.3 seconds for every downstream device on our 13.2kV feeder, clearing time becomes significant at the mains.

This is our main substation utility interface at 13.2kV. The utility owns the transformers (69-13.2kV) and they do have transformer differential and directional relaying looking into our system.

Also, are there any good freely published documents regarding basic/advanced industrial relay coordination?

 

[RalphChristie]

Laplacian

You can download the old PRAG (now NPAG) book at the site of Areva for free in pdf format.
Use this link:

http://www.areva-td.com/servlet/Con...d=1017999014820&tab=Chapters&id=1056536208254

Chapter 6 deals with O/C and E/F Protection - you'll find some info there regarding coordination.

I am not familiar with the term "spot network configuration." Can you describe it?

A lot of feeders in a radial system can cause some coordination problems - the time become too long at the top feeders. Some things you can do to "win" time at the top are:
? Use instantanious units, especially at the last feeders, or at transformers.
? At feeders where you have breakers at both sides (cable feeders, line feeders etc.) use the same setting for both breakers.

Also, I assume you use electronic relays. The 0.3seconds time is a average value for time between breakers. It include a 0.1seconds for CT-error and 0.2seconds for breaker operating time. (add another 0.1seconds for relay overshoot if you use electromechanical relays) If you can ensure a smaller CT-error or faster breaker operating times, you can decrease your time between breakers accordingly.

Regards
Ralph

 

[dpc]

If you normally operate with the tie breaker open, closing it only for maintenance or emergency conditions, you can sacrifice coordination between the mains and the tie. This is very commonly done in industrial facilities with "double-ended" configurations.

If the tie breaker is normally closed, as it may be in this case, this is generally not acceptable. A partial differential scheme between main and tie breaker can improve overall coordination.

 

[Laplacian]

RalphChristie:
What we call spot network has two incomming mains and a tie breaker between both incommers. It is not secondary selective; no transfer scheme.

Instantaneous is out of the question due to the downstream unit substations that must clear their faults first.

These are the old Westinghouse CO-8 relays for phase and ground time-overcurrent protection.

Thanks for the link to the "Protective Relaying Application Guidelines"; I also found the "Art & Science of Protective Relaying" by GE on this site.

dpc:
We run our tie breaker normally closed; the equipment is rated for this service (high impedence transformers).

We do not have bus differential protection at this time; may propose a project to add a partial differential scheme as you suggest.


Thanks for your replies.

 

[RalphChristie]

So you actually run two trsf in parallel?

You can still use a instantanious unit - for instance on a downstream trsf. Just set the inst. unit at 125% of the through fault current. If the current is below the set-value, it will trip on the relay-curve, any current above the set-value means a fault on the primary side, thus inst. operation. With this aproach you can win a few milli-seconds.

The other way to win time is to have zone protection (differential schemes) like dpc suggested. If you have zone protection, you can decrease the different O/C levels. A diff scheme between the main and tie breakers will be a good option. But you can even use it on feeders, for instance on a cable feeder. (although distance can be a drawback) For an example, check the Argus-range of relays from Reyrolle.
www.

Another thing to note is the fact that downstream faults will divide (half of the fault-current in each leg) between the two breakers.

Failure seldom stops us, it is the fear for failure that stops us - Jack Lemmon

Regards
Ralph

 

[RRaghunath]

I would like to elaborate on partial diffrential scheme that dpc has suggested. I worked with the scheme while in Gulf and found it a very desirable one where ever 'bus coupler closed' operation of switchboards is involved.

I find the scheme specially beneficial from the point of providing grading between the incomer and bus coupler without the need to keep different settings.
This also doesn't require additional CTs, the existing normall protection class CTs in incomers and bus coupler are adequate.
No separate OC & EF protection needs to be provided once the partial differential protection is included. The relay used is normal IDMT OC / EF relay.
The protection receives summed up currents of the incomer & the bus coupler. There is no relay in the bus coupler, only one relay each in each of the incomers.

Whenever either of the OCEF relays acts it trips the corresponding incomer as well as the bus coupler. The relay in the other incomer doesn't see the fault considering that the current in buscoupler is in phase opposition to that in the incomer and no current comes to the relay.

With buscoupler closed operation,
* The current in the relay is same as that in the faulty outgoing feeder and hence the coordination can be worked out for the same currents in the outgoing as well as in the incomer feeders.
* The scheme doesn't use CTs in outgoing feeders.
* In case of bus fault, the corresponding relay in the subject incomer sees twice the current through the incomer and that in the other incomer zero current. Thus, only the subject bus-section is isolated.

The disadvantage in the scheme is that the relay cannot be set instantaneous and thus is not fast on bus faults.
Otherwise the scheme has all the benefits especially in providing discrimination at the incomer level as an IDMT OCEF protectio even in cases where full fledged bus differential protection is provided, in all cases requiring bus coupler closed operation.

 

[rconnett]

One other suggestion - using instantaneous elements for fault detection or sometimes called 'zone selective interlocking' is a cheap way to gain some speed & selectivity.

The principal uses an instantaneous element to detect a fault.

Logic is developed that compares the fault detectors (instantaneous element) on a main breaker & compares it with downstream devices or a tie breaker. If the main breaker senses a fault, and downstream devices do not, then a fault has occurred on the main bus & the main breaker trips instantaneously. This is especially appropriate if you can install slide in solid state relays.

This requires being able to get the inst element brought out to an available terminal & it has to be self resetting, as well.

A slide in solid state replacement might allow you to improve your margins a bit. I believe there are slide in replacements available for the CO-8 - see below:

http://www.basler.com/downloads/pd0bro.pdf

look on the bottom of page 20 - BE1-50/51B-214 or 219.

Good luck!

 

[RalphChristie]

Interresting rraghunath. I nominate you a star.

Laplacian: Sorry, I see I haven't given you the link to the differential relays from Reyrolle.
Check:

http://www.rmspl.com.au/handbook/Reyrolle/solkor_n_brochure_v2.pdf

http://www.rmspl.com.au/handbook/Reyrolle/solkor_rf_and_m_brochure_v1.pdf

This is just an example of feeder differential relays, there will be probably many other brands.

Regards
Ralph