Determining existing conductor length help

Status
Not open for further replies.
M

mjmike

Senior Member
I am working on arc flash and short circuit calculations for a client. The facility is rather large with about 5 stories, 6 distribution boards, and about 50 branch panels. There is also an emergency system. The service main board is 480/277v 3-phase and about 3200a. The step-down voltage is 208/120 3-phase. Ther is no existing one-line.

I can easily create the one-line and get the breaker data, but the trouble I am having is determining the condutor length. The facility is older so most conduits are in the masonry structure and can't be visually traced or measured. I am looking for advise on how do figure out the conductor length. I can de-energize the facility so is there some type of electronics equipment, like a tracer, that will determine length? Thanks for any opinions.
 
tryinghard

tryinghard

Senior Member
Location
California
I believe there is an instrument to measure conductor length, warehouses use them but I don't know of a type. We?re labeling at our facilities as well and in most cases estimating the circuit lengths [tedious].
 
M

mjmike

Senior Member
I considered doing that. However, there are a couple issues. The first being the points where the measeurements are taken may be 300+ feet apart with floors in between. I could if I use 2 wires and connect at the end panel then read across the 2 wires at the switchboard, but then you have resistance of the connector. Secondly, if we are talking say 500mcm stranded, how would you possible insure you have all the strands at both ends? Thirdly, the number one rule of multi-meter testing, I would have to disconnect and isolate the conductor being measured. Good suggestion, I am just not sure how to accuratelly take this measurement. Maybe I am over-thinking it, if you could clarify, maybe it is a viable solution.
 
S

sgunsel

Senior Member
To measure a resistance on a long wire, it would probably be easiest to 1) remove all power, 2) short two wires together at one end, 3) measure resistance between the two wires at the other end. Diviide the measured resistance by 2 to get the resistance of the wire over its length. You don't have to worry about individual strands, they are already in intimate contact with each other. You will need a meter with adequate sensitivity for the resistance expected. The DC (which is what your meter will be) resistance of 500 kcmil Cu, for example, is 0.0258 ohms per 1000 ft. at 75C. Resistance varies with temperature. Your wires will most likely NOT be at 75C.
 
Z

ZinskI/E

Member
Location
New Orleans, LA
mjmike said:
I considered doing that. However, there are a couple issues. The first being the points where the measeurements are taken may be 300+ feet apart with floors in between. I could if I use 2 wires and connect at the end panel then read across the 2 wires at the switchboard, but then you have resistance of the connector. Secondly, if we are talking say 500mcm stranded, how would you possible insure you have all the strands at both ends? Thirdly, the number one rule of multi-meter testing, I would have to disconnect and isolate the conductor being measured. Good suggestion, I am just not sure how to accuratelly take this measurement. Maybe I am over-thinking it, if you could clarify, maybe it is a viable solution.
Click to expand...

One idea would be to:
- measure the resistance of a short piece of cable for a jumper, say #10
- install the jumper directly into the lugs on one end of two cables
- measure the resistance of the loop
- subtract the resistance of the jumper and divide by 2
- perform the calculations to determine conductor length

Some caveats:
- resistance will change with temperature of the conductors: overhead, underground, indoors, outdoors (though the difference becomes smaller as conductor size increases). I would survey the routing of conduit/cable and determine a weighted average of conductor temperature before I perform the calcs.
- all splices and terminations introduce additional resistance. If you're lucky the cables go directly from point A to point B without interruption.
- be sure to measure the resistance of your test leads, they will also introduce an error if not compensated in the calc
- lastly, the accuracy of your test equipment: my DMM reads to thousandths of ohms and is certified annually. Greater resolution would require more sensitive equipment.

IMO the measurements you gather would be relatively accurate, with accuracy improving with conductor size increasing. An unreasonable calculation could indicate a cable fault, splice or tap points, or any number of other unknowns. Short of actually measuring the cable or using footage markings on the insulation, I don't know how to do any better.

It would be a good idea to test this method against a similarly installed circuit of known length to determine an accuracy tolerance.

Good luck.
 
Last edited:
M

mjmike

Senior Member
That all makes sence and my concerns were the same as pointed out such as temp, splices, etc. I guess it would not be that bing of a difference to be concerned of. Will probably give it a try.
 
steve66

steve66

Senior Member
Location
Illinois
Occupation
Engineer
It seems like it would be much easier (and more accurate) to measure the operating current with the conductor under load. Then measure the voltage on each end. Use the voltage drop formula in reverse to find the distance.

Have you thought about just estimating the distances based on the distance between the source and load? With some good engineering judgement, that might be sufficient. (By that, I mean making sure your calculations aren't too sensitive to changes in changes in the length.)

Steve
 
Z

ZinskI/E

Member
Location
New Orleans, LA
steve66 said:
It seems like it would be much easier (and more accurate) to measure the operating current with the conductor under load. Then measure the voltage on each end. Use the voltage drop formula in reverse to find the distance.

Have you thought about just estimating the distances based on the distance between the source and load? With some good engineering judgement, that might be sufficient. (By that, I mean making sure your calculations aren't too sensitive to changes in changes in the length.)

Steve
Click to expand...

That is a good idea, but I'm not sure how much easier it would be. I think the only steps saved would be the jumper and a few resistance measurements. All the same caveats still apply - anything that changes the resistance will also change the voltage dropped (temperature, splices, et al). Also, I think the resolution down to thousandths of ohms would be more accurate than tenths of volts and amps, which is how my DMM would handle it. The other advantage to using the resistance is that all testing is done de-energized, so no additional PPE is required.

Since I've never tried either method, it would be interesting to perform both tests on circuit of known length and see how closely they agree. If the accuracy of both was reasonable I'd pick the easier of the two.
 
Electric-Light

Electric-Light

Senior Member
A tiny section that isn't connected tightly isn't a huge deal, just as a foot of 16AWG wire in a section of 100' 10AWG extension cord won't affect it much.

I would use low voltage and measure on grounded side, so the floating voltage won't affect the instrument.

You'll need two DMMs. Use the 10A range on a DMM and place the DMM and the looped conductor in series. Push a few amps through it using a 12v battery and MR16 12v bulb or the like as a load.

On DC mV range, measure the voltage drop across the conductors under test as well as the A reading on the ammeter. Try to get the readings as close to the same time as possible.

You now have xx mV drop / xx A.

The reason for using a higher amperage is to provide higher resolution.

Suppose you've got 25mV drop at 4.8A. You can calculate the resistance using the formula V/I

0.025/4.8 =5.208 mOhm

If you were to push 48amps, you can get ten times the resolution using the same volt meter. You will get 250mV drop.

48A is not easy to push through, but 5A or so is easily done with a battery pack and some light bulbs. 5A is FAR more current than any ordinary DMMs provide. You can either use an expensive high-current four wire ohmmeter or get a pretty decent, high resolution low resistance reading using this method.
 
Last edited:
G

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
110402-1650 EDT

mjmike:

You have had several good suggestions using a voltage and current measurement method. You do not need an expensive 4 terminal instrument to make a good measurement because you can create your own 4 terminal measurement.

A 4 terminal resistance measurement is nothing more than a resistive element in which the test current is injected into the resistance thru terminals separate from the voltage measuring terminals. The voltage measuring terminals are located inboard of the current terminals. This arrangement eliminates any error from the voltage drop thru the termination of the current to the resistance. Since the voltmeter draws very little current there is little error from its connection to the resistance.

See my photos P19 - P21 at http://beta-a2.com/EE-photos.html for illustrations of 4 terminal connections. For your purposes you could simply poke the voltmeter probe tips into the wires being tested.

I believe you are talking about 500 MCM copper and if I calculated correctly this is about 0.021 ohms per 1000 ft at 20 deg C. You have to assume the wire size and material are the same throughout the run to be measured, and there are no bad connections. At the far end you could even use a #12 jumper, but #2 would be better. It only needs to be very short.

For a 300 ft run there would be 600 ft of wire, and the resistance would be about 0.012 ohms. I would want at least 10 A for a test current. I would suggest AC rather than DC as the current source. You can compensate out the voltage drop at the shorted end by measuring the voltage drop from one 500 MCM wire to the other at the jumper point. Then this voltage drop is subtracted from the voltage measured at the source end. 10 A at 0.012 ohms is going to produce 0.120 V drop. Use something like a Fluke 27 or 87 that can measure millivolts. Shorter conductor lengths will produce lower voltage drops.

If for some reason there is substantial induced voltage in your loop under test, then you would need a DC current source.

In all likelihood your current source will be a voltage source with a series current limiting resistance. From a 120 V source you can about 12 A using a 1500 W portable heater in series.

.
 
steve66

steve66

Senior Member
Location
Illinois
Occupation
Engineer
ZinskI/E said:
That is a good idea, but I'm not sure how much easier it would be. I think the only steps saved would be the jumper and a few resistance measurements. All the same caveats still apply - anything that changes the resistance will also change the voltage dropped (temperature, splices, et al). Also, I think the resolution down to thousandths of ohms would be more accurate than tenths of volts and amps, which is how my DMM would handle it. The other advantage to using the resistance is that all testing is done de-energized, so no additional PPE is required.

Since I've never tried either method, it would be interesting to perform both tests on circuit of known length and see how closely they agree. If the accuracy of both was reasonable I'd pick the easier of the two.
Click to expand...

PPE would be required for the method I suggested.

As far as being easier, I was thinking this method doesn't required disconneting the 500KCM cables. That alone seems like a big advantage. It can be an even bigger advantage if places where shutting off power to anything requires an act of Congress.

Steve
 
Jraef

Jraef

Moderator, OTD
Staff member
Location
San Francisco Bay Area, CA, USA
Occupation
Electrical Engineer
The TDR is the way to go. I use them to find faults in a cable, it tells you where it is within inches. Resistance is irrelevant, so it ignores properly terminated junction box connections (unless the short or break is there).

I can't speak for that particular TDR because this is the first time I have seen one from Megger. But so far nothing of theirs that I have used has been anything but top shelf, so I would imagine it's a good one.
 
G

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
110404-2247 EDT

Jraef:

How do you determine the propagation velocity for a particular cable installation to be able to determine within inches the break location on a long cable?

On one sample of CAT-5 I got about 76% the speed of light, and a different manufacturer about 67%.

.
 
Status
Not open for further replies.
Top