Voltage Drop - how much do splices cancel out upsizing?

Status
Not open for further replies.
J

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
I have a customer that I'm trying to set right about a voltage drop question, and I'm pondering the question below. Take this kind of a general theoretical question that I've put specifics on the for the sake of clarity.

Given: the equipment terminals can accommodate max 10awg.

The customer wants to use 2awg to avoid voltage drop. (Just take this as given, please. :lol:)

So I'd have to splice 2awg to 10awg to land in the equipment, probably using a polaris style connector or something. This probably has some resistance that at least partially cancels out the gain of upsizing the wire, right?

Thus the questions:
How much resistance does the additional splice have, and how would I find that out?
Is there a ballpark number that people use?
Is there a resource for the resistance of certain types of splices?
Would the manufacturer know this?
 
If you are going from 10 AWG to 2 AWG because of voltage drop over a distance, resistance in your connections should be fairly negligible.

Think about all the other connections you may have between this circuit and the source, they are resistance (presumably negligible as well) in the path between your load and the source. You have branch circuit device - breaker has the input and output terminal as well as the switch contact, fuse has input/ouput terminals as well as contact points on each end of fuse, then if you have all this on service feeder(s), metering equipment, connection at the source... maybe other points of connection. But all those points are supposed to be negligible when in good working condition, when those points begin to fail you get heat at those points that makes them get even worse.
 
ptonsparky said:
A good compression splice can have less VD than the wire itself. I've measured it.
Click to expand...

So let's say I've done the math and upsizing from #10 to #2 results in half-a-percent less VD. Would the splice claw back only a small fraction of that, or more than half? I guess 'less' is vague.

Also, would Polaris qualify as 'good compression splice'?
 
jaggedben said:
So let's say I've done the math and upsizing from #10 to #2 results in half-a-percent less VD. Would the splice claw back only a small fraction of that, or more than half? I guess 'less' is vague.

Also, would Polaris qualify as 'good compression splice'?
Click to expand...
My good compression splice was with a Blackburn tool that squeezes an H tap onto the wires. Irreversible.

I measured the VD on a distance of the wire, then the same distance with the splice included.
repeated it several times with similar results each time.

The only problem i see would be if you’re overloading the 10 to begin with. The splice will be negligible, unless you screw it up somehow.
 
junkhound said:
rule of thumb -50 millivolt drop per connection at rated current.
Click to expand...

So #10 at 24 amps would be, coincidentally, also 50 millivolt drop per foot. #2 is 9 mv per foot. So with only about 2.5 feet of #2 you equal the resistance lost in the two splices and after that you are gaining; Its not looking good JB - how about "accidentally" adding another zero to junkhound's figure and tell that to the client ;) Maybe compute the energy lost and show them the payback period?
 
To answer your question, if splice resistance was so significant that it negated the use of oversized wire, then I'd say someone was doing their splices wrong if that were the case.

On top of that, if every splice we made had significant resistance, imagine all the the heat generated in an average building full of spliced connections?!?!

It's simple ohms law, volts dropped multiplied by the given current across the connection equals watts of energy.
 
Cow said:
To answer your question, if splice resistance was so significant that it negated the use of oversized wire, then I'd say someone was doing their splices wrong if that were the case.

On top of that, if every splice we made had significant resistance, imagine all the the heat generated in an average building full of spliced connections?!?!

It's simple ohms law, volts dropped multiplied by the given current across the connection equals watts of energy.
Click to expand...

In this case the upsizing would be quite optional - it is about squeezing the most efficiency out of the application - and is not to deal with conductor length or basic function. The run is probably about 80ft.

If a splice equals 10 extra feet of wire, then at 500ft nobody cares and at 30ft upsizing is more or less pointless. None of this has to do with whether the heat energy associated with voltage drop is noticeable to someone standing around.

It's not simple ohms law if you don't know the resistance of the splice. ;)
 
jaggedben said:
So I'd have to splice 2awg to 10awg to land in the equipment, probably using a polaris style connector or something. This probably has some resistance that at least partially cancels out the gain of upsizing the wire, right?
Click to expand...

Up sizing the wire should make things more efficient. The real question is, how much more. Is it going to take 5 years or 50-500 years to pay for this upgrade in materials ( don't forget the extra labor).

If I were looking at this I try to figure out voltage drop and power loss under the present conditions.
 
jaggedben said:
...
Thus the questions:
How much resistance does the additional splice have, and how would I find that out?
Is there a ballpark number that people use?
Is there a resource for the resistance of certain types of splices?
Would the manufacturer know this?
Click to expand...

Not answered in order:
I have not seen any mfg data.
I don't know of any resource except personal tests.

I have only one set of test data.
Application was 24V battery starting a turbine generator, 2000A peak, tapering off to 600A within 70 seconds.
Conductors were parallel 262.5 DLO, 25'
Connections were crimped lugs, bolted to copper bus.
​
1 milliohm at 1000A is 1 Volt - which is really a lot at 24V.
We measured the connection resistance with a 4 wire DLRO - one clamp on the bus, one clamp on lug. The setup did not measure the resistance of the crimped connection - just the lug to bus connection. So our data may not apply to your application. As I recall*, test sample of 4 connections, resistance
was 0.5 milliohms.

How to measure:
Say your application consists of 1' - #10 >> Splice 80' - #2 >> Splice >> 1' - #10
Ignoring the splices for the moment,
One-way conductor resistance Rc = 1.24 X 1 (foot) / 1000 + .194 X 80(feet)/1000 + 1.24/1000 = .0180 ohms
If the two splices added 1 milliohm, that would change the conductor overall resistance from 0.0180 ohms to 0.0190 ohms. Probably really hard to measure. And without a precision current source, really hard to measure the difference in voltage drop.

If my client had a burning desire to know, I would set up a test using 1'-#10 >> splice 1"-#2. Use a 4W ohmmeter to measure the combined resistance. Subtract off the 1.2 mohms for the conductors.

The instrument accuracy needs to be at least .1 mohm. You can make up a 4W ohmmeter using a D cell and two decent DMV (one set to measure millivolts, the other set to measure <10A) A short circuit D cell will put out 4A - 7A. Or rent a fluke DLRO 10X (retail ~$5K)

OR,

You could say, "I have access to some test data, connections are on the order of .5milliohms. Two splices will increase the resistance of the 80' run
1 - .019/.018, ~5%.. For a 30A, Vd changes from (30 X (2 X .018)) = 1.08V to (20 X (2 X .019)) = 1.14V"



the worm
* "as I recall" translates to "that was 15 years ago"
 
I would call losses in properly applied splices as down in the noise, but figure your total resistance with all small wire and then again with the larger wire ignoring the splices. Calculate the difference and divide by two. That's the resistance each splice would have to have for it to be a wash; does it seem reasonable?

Just for grins, figure the power dissipation the splices would produce if they matched the resistance of the difference. How hot do you think they would they get?
 
growler said:
Up sizing the wire should make things more efficient. The real question is, how much more. Is it going to take 5 years or 50-500 years to pay for this upgrade in materials ( don't forget the extra labor).
Click to expand...
The answer might be 'never.' It depends on the application. For example, with incandescent lighting, unless you're going to install a higher-wattage bulb to compensate for reduced light from voltage drop, the line losses will actually reduce operating cost slightly (and increase bulb life.)

In the case of HVAC, a heater would need to run longer to compensate for less instantaneous heat output. If the line-loss heat adds to the building heat, the net loss is reduced, but worsens during cooling season, and the compressor motor uses more current as a constant-power load.
 
Thanks iceworm and ggunn for the food for thought.

For others who might not have looked at my profile or other posts, this is a grid-tied PV application so the customer will actually get paid for the energy production that's gained by not losing it to voltage drop.

Final question...
Any guesses at the resistance for a wire-nut joining a single 6awg to a 10awg copper?
 
jaggedben said:
Final question...
Any guesses at the resistance for a wire-nut joining a single 6awg to a 10awg copper?
Click to expand...
My guess: Low enough (well made and tight) that it's worth introducing where it's definite that the larger conductor is beneficial.
 
LarryFine said:
My guess: Low enough (well made and tight) that it's worth introducing where it's definite that the larger conductor is beneficial.
Click to expand...

I agree, and if made up correctly, I would not be surprised if the resistance on that connection is less than the actual wire itself because there will be more surface contact between the 2 conductors than the cross sectional area of the conductor.
 
oldsparky52 said:
I agree, and if made up correctly, I would not be surprised if the resistance on that connection is less than the actual wire itself because there will be more surface contact between the 2 conductors than the cross sectional area of the conductor.
Click to expand...

Good point. Thanks.
 
Status
Not open for further replies.
Top