Load Encroachment in Distribution System

[PTCruiser85]

Hi,

I had asked a similar question in the past regarding a coordination study where the distance between fault location and feeder was only 700 m. It was very difficult to coordinate.

this forum had some interesting contribution in terms of views from you guys. After doing some more research, I found a name to this phenomena and it is called LOAD ENCROACHMENT. This term is more commanly used in transmission system protection.

Is it available in Distribution system as well, if yes how often it is. Most of the literature I have seen talks
about transmission aspect. Could you advise.

 

[dpc]

Load encroachment relates to the influence of load current on distance (mho) relay sensitivity and settings. It would be unusual to use a distance relay on a distribution feeder.

The issue of coordinating close-in faults is not directly related to load encroachment, at least as I understand the term.




David Castor

http://www.cvoes.com

 

[m3ntosan]

dpc is right, load encroachment is about magnitude and angle of the load that sometimes can enter in the tripping diagram for zones 2 and 3 (if forward). It's an issue for mho and quad type also.

Maybe you're looking for close-up faults and dificulty for protection to see the fault forward/backward, I guess is it call polarisation problem, that's why digital protections use voltage memory.

May you grow up to be righteous, may you grow up to be true...

 

[davidbeach]

Load Encroachment can be used with overcurrent relays as well as distance relays. The SEL-351 and -451 include load encroachment but have no distance elements. The results of the load encroachment logic can be used to torque control overcurrent elements.

An interesting use could be for cold load pickup, where the current is load but higher than the lowest instantaneous setting.

What load encroachment can't help with is distinguishing fault location from one side of the next relay to the other. If you want one relay to clear all faults between it and the next relay instantaneously you will have to have communications of some sort between the relays.

You could set an element in the upstream relay to trip after 3-4 cycles if not blocked and then set an element in the downstream relay to send a blocking signal if it also sees fault current.

 

[SMB1]

Hi,

there is no restrection to use distance in Distribution netwok, the Load Encroachment function can be set On and you need to enter your max. load impedance with your load angle.

i do recommend to use it.

thanx
SMB

 

[m3ntosan]

SMB,

What about price,weak infeed and low impedance?

May you grow up to be righteous, may you grow up to be true...

 

[PTCruiser85]

Gentleman,
Thanks for your comments ! I guess my question steered the discussion in a direction which I am not familiar with actually.

I was trying to find out that in situation as I discussed earlier in my pervious post where fault on the customer that is close to the substation is approx. same as on the feeder substation itself. It is difficult to coordinate instantenous elements function but the only way it can be resolved or coordinated is by using a load encroachment function on a SEL 351 relay. So the fault that is close to the substation on the feeder can distinguished and the instantenous element of the substation breaker can be inhibited or blocked till the customer clears fault or in another way of distinguishing or cheating substation breaker not to operate even though the fault current value is more that inst. over current element setting and it is supposed to trip and it will not trip.

I think now I made my self clear that if I used load encorachment function. Will this eleviate my problem.
I would appreciate an input on this issue.

Thanks

 

[jghrist]

If the concern is with a fault on the customer's system, load encroachment does not seem to be the solution. Normally, if instantaneous is used on a distribution feeder, reclosing is used and the instantaneous is disabled after the first trip. Inverse time overcurrent relays are used for subsequent trips with enough time delay to coordinate with tap and customer overcurrent protection.

 

[PTCruiser85]

Thanks jghrist. This is what I was looking for.

I was advised by one of the leading relay vendor application engineers probably wanted to market their relay.

I was told that I can use the load encroachment function and differentiate between fault and load and now if I think it does not make sense coz in my case there is no issue of load. it is a fault problem.

That means Jghrist utilities have to use an extremeley inverse function but what about the cable and equipment on the way to the fault that have to go through the trauma of fault current for .5 or probably 1 sec.

how can we quantify that damage in case of a 500 mcm xlpe cable going through 10000 a fault current for 1 sec.

 

[cranky108]

There are only two ways make coordination work. Time and communication.
If you are using a SEL-351 relay, then you can add a time delay with the 67P1 element.

The problem of having a tap just outside a substation always causes you to use a delay on the 50 elements, and to remove them completly. I've tried to explain this to people, but I've been told this is a protection problem. It isen't just a protection problem, the damage on equipment, mostly transformers(which is cumulitive), is a problem for everyone.

Has anyone looked at setting a minumim first tap rule (like a quarter mile)? Probally won't work on distribution, but maybe on subtransmission.

 

[PTCruiser85]

Bravo CraNKY108 !

You said what the reality of the problem is.
I have been trying to find a real solution to this problem for a month now and unfortunately people in utilities dont tend to change anything and want to come up with a solution without understanding real issue.

When I came across this problem and explained this to my fellow colleagues their first reaction was that How can you put a time delay on instantenous or disable instantenous elements.

Now what do I say

I sound like I am saying some thing that is beyong norms coz then I am asked questions like what will happen to cables,splices etc.

how can the directional element help.

 

[cranky108]

You can turn off the directional, which leaves you with a time delayed 50. Yes I have had questions about placing a time delay on a 50 function, but the people asking those questions have never done this job.

And in fact if you look at the SEL-50 function, it operates on the offset part of the wave form, which most of us don't calculate in studies. We all calculate the fault levels at the about 3 to 15 cycle area. So the SEL-50 function will sometimes appear to misoperate. We would have to rise or delay the 50 anyway.

Or said another way, look at the IAC 50 curve, the SEL relays don't have a curve.

So don't tell anyone you added a delay.

 

[PTCruiser85]

Thanks Cranky 108. Your comments are very useful.

1)Could you kindlly advise, how to quantify damage on cables and equipment for the fault level.

2) SEL 50 means SEL 501. I had an impressions that all digital relays has a algorithm that filters
out that DC component and activates only when the symmetrical component passes the threshold of overcurrent.Please correct me, if I am wrong.

Thanks

 

[cranky108]

1) I've been trying to qulify the damages for a while now but it's hard to calculate cumulated damages, then there are people who will debate you because it is in there interest.

2)By SEL 50 I mean the 50 element of any SEL relay. But even with the filter, the asymetric factor must still be consitered. A fault study program normally won't calculate that. It only lasts a few cycles.
In the old electromechinical relays there was a delay for currents just above the pickup value. That delay dosen't exist in modern relays (you could use a two step scheme).