Broken Conductor Relay

[deadzone]

Broken Conductor
Earth fault overcurrent element that detects current unbalanced condition can detect and alarms the broken conductor.
This is a new protection system which we are going to implemented. Most of our OHL are across the desert (High Resistive) so ordinary E/F relay is not responding.
Please what is the advantage and disadvantage of this new Protection System (new for us i.e.).
Is this system was successful on other countries.
Is there any other alternative protection that will solve our problem or Broken Conductor relay is best solution.

Thank you

 

[Bung]

A broken conductor hanging to the ground on the load side will only carry load current - nothing will see that if the load is very low downstream of the fault. On no load, all you will see is transformer magnetising of the d/stream trannies.

The fundamental problem is trying to pick up a mechanical fault by measuring the electrical signals. If the fault does not induce changes in the electrical parameters (amps and volts) that can be clearly and unambiguously separated from any possible load conditon, then no relay can ever "see" it.

We use sensitive earth fault (SEF) relays set at around 2% of nominal (400A CTs) in a residual connection. We have no phase to ground loads, all our loads are Phase -phase or 3-phase. We are also implementing a delayed reset on the timers of all SEF relays to avoid missing arcing or "pecking" when the conductor fizzes and pops on the ground.

We still miss some "conductor down" faults. There are some fancy arc fault detecting relays out there, but they are set by little more than "suck it and see", take minutes to make a trip decision, and have to be fine tuned every time they miss a fault. Not terribly practical in real life situations.


Bung
Life is non-linear...

 

[vic3fan]

There is an excellent reference about this type of fault on Areva (former Alstom) web site

http://www.areva-td.com.

They have developed this logic into MiCOM P141 relay, and it’s about calculating the I2/I1 ratio, (where I2=negative sequence phase current, I1=positive sequence phase current), because when this type of fault occurs, the phase currents will not increase and the usual methods used in detecting phase-to-ground, phase-to-phase-to-ground and three-phase faults do not apply. For details see their P14x Technical Guide, chapter 2 ‘Application Notes’ pages 73-75 of .pdf file – follow Products & Systems (left side menu) / Protection Relays / Overcurrent Protection / MiCOM P14x Series / MiCOM P140 technical guide / Application Notes-TG8612 Chapter 2 C English (unfortunately there is no direct path, I would give it if available).

I have seen a concrete application with this relay. Of course, there are other methods as well.

Bung is right, a negative sequence element does not operate at low load levels.

 

[busbar]

Literature seems to treat downed-line detection as a classical distribution-relaying quandary.

http://www.selinc.com/techpprs/6155.pdf

negative- and zero-sequence overcurrent principles

http://www.selinc.com/techpprs/6123.pdf

ground-detection limitations

http://scholar.lib.vt.edu/theses/av...35/unrestricted/ThesisMasterFabianUriarte.pdf

http://lib.hut.fi/Diss/2001/isbn9513859614/isbn9513859614.pdf

 

[Bung]

I once thought that the Areva I2/I1 ratio might afford some reasonable additional protection cover for downed or broken conductors. And it does, in some limited circumstances where the fault current at the point of fault is significantly above the load current at the relaying point. Unfortunately this just ain't so for most practical overhead distribution lines. And I defy anybody to truthfully state that they balance the single phase loads across all three phases - in most normal utilities it is not like that. Everything gets hung on the two outer phases 'cos they are easy for the linies to get to. I was going to set it (the I2/I1 protection) on some of our feeders that have Areva relays, and very soon gave it away as impractical.

But you have to do the analysis on your own system - it can work in some cases.


Bung
Life is non-linear...

 

[Bung]

Bear in mind too that the neg phase seq current (I2) is going to be half of diddly-squat when you have a downed conductor, so measuring I2/I1 is going to tell you a whole lot.

Bung
Life is non-linear...

 

[Bung]

Oops, sorry - correction, should read "is not going to tell you a whole lot."

Bung
Life is non-linear...

 

[veritas]

As mentioned the problem of detecting a downed conductor is not an easy one. High resistance to earth at the point of contact only exacerbates the matter. In the extreme case a downed conductor is really an open circuit condition. Negative phase sequence (nps)elements are only useful under the following conditions :

1)Radial lines
2) Reasonable load
3) No remote zero phase sequence (zps)source(s)downstream of the source.

1) A downed conductor/open circuit on an interconnected network will yield very little nps whilst substantial zps current will exist. This is because I2 is determined by the system as well as load impedances whereas I0 is determined by the system impedances only (and only those system impedances which are part of the parallel or interconnected circuit).

2) As mentioned more load, the better the chance of getting higher I2 currents.

3)The presence of remote zps source (most commonly star points of trfr neutrals) also lowers the zps impedance of the network substantialy compared to the nps impedance (Z2 load included).

I have actually done quite a detailed study quite a while ago to try and use NPS elements in the SEL-351 relay to detect open circuit/downed conductor conditions since conductor theft is a serious problem.

Hope this helps.

Regards.

 

[busbar]

Another marketing ploy is some of the “deluxe” downed-line relaying is deemed proprietary, and they’ll let you in the on the details after you buy a palletload of their boxes… and sign a nondisclosure agreement.

[One guy 'hinted' at second harmonics, but would say no more without a purchase order.]

 

[deadzone]

Thank you a lot for your participation

 

[AppEng1]

Another relay to consider for downed conductor sensing is the GE F60 with the HiZ capability.
The HiZ algorythm consists of a couple of different elements that is used as a group to establish if a downed conductor is detected. It uses loss of load, and arc detecting algorythms, that was established by detailed analysis of actual downed conductor testing on various surfaces.
For more info refer to:

http://www.geindustrial.com/products/manuals/f60/f60-j1_5.pdf

http://www.geindustrial.com/industrialsystems/pm/pr/hiz_fault_dist.pdf

http://www.geindustrial.com/cwc/products?pnlid=6&famid=31&catid=216&id=f60&type=product+information

http://pm.geindustrial.com/FAQ/Documents/F60/GER-3796.pdf

 

[Bung]

I bought one of the F60's predecssors once (the DFM) with the arc fault algorithn to try it out. When I tried to set it, I realised I was on a hiding to nothing, and it is still in its box on my desk 5+ years later. As a trial it wasn't going to work, as it wouldn't be exposed to enough faults to tune the settings, never mind prove anything about its effectiveness. I cannot justify buying a hundred or so to set up a full trial that would give significant results, and in any case, I still don't know how to set them in a way that allows me to see the effects of changing settings and still have the things do some useful work on the system.

Maybe they are smarter now, but it is still going to take a fair amount of trial and error to set up. Until a manufacturer tells me how I'm supposed to set it to trip in a reasonable time (say <15sec) based on my feeder arrangements, I'm going to remain very sceptical of them.


Bung
Life is non-linear...

 

[mc5w]

Arc Fault Circuit Interrupters for 120 volts are also a technology that is NOT ready for prime time.

One problem with AFCIs is that houses around here have the stairway lights wired by tapping any hot wire on the first floor and tapping a neutral on the second floor. I like to call this a haywire circuit.

Both AFCIs and GFCIs can also nuisance trip when a load turns off or a split phase motor tries to start. The mechanism is that circuit inductance forces a phantom ground fault to flow through the capacitance to ground. If only 0.2% of the starting current of a 1/2 hp 120 volts split phase motor diverts into the iron an AFCI will read that as a ground fault and trip. I keep telling SquareD and GE that the motor sorks fine on a 15 amp circuit breaker and we even swapped hot amd neutral to check for neutral ground faults.


The compatibility testing for AFCIs and GFCIs was done with NEW electrical devices. When equipment gets to say 10 years old the insulation resistance and capacitance gets to be a problem. Engine block heaters for trucks are the worst because they get salt water in them. New 120 volt motors are also bad because they HAVE to be cheap.

The only utility distribution that could detect fallen wires would be an ungrounded system that is periodically tested by "grounding" it to a 400 Hertz power source.

Mike Cole, mc5w@earthlink.net