AC and DC in same trench?

[BeezFun]

Planning a solar installation using 600v, the conduit run between solar electronics in house and the panels happens to go right past a building where I need to pull 110v, 40A service. Question is can I drop two conduits in the same trench, pull wire for solar DC run through one, and wire for AC run through the other? Can the AC line put some kind of noise on the DC line that interferes with any of the electronics in the solar charging/inverter system? Or is there any other reason not to do this? If they need to be separated, how much distance between them?
tnx

 

[greasemonkeyok]

The answer to your question will depend on how far the lines will be parallel. The further they run parallel, the further away from each other they will need to be.

Hope someone here can give you more specific information on what the code requires.

 

[Gary Fowler]

Normally we are required to separate instrument cables (milliamp carrier ) by 12" from high voltage cables or put in a shielding of carbon steel plate. I have never heard of a requirement to separate high voltage cables but then on heavy construction, we dont have high voltage DC, only the low voltage control cables.
In any case, you should be ok with them if you keep them at least 12" apart.

 

[Gary Fowler]

The answer to your question will depend on how far the lines will be parallel. The further they run parallel, the further away from each other they will need to be.

The amount of EM radiation coming off a cable has nothing to do with length of the run. The higher the voltage, the more EM it produces. A coiled length of HV wire can actually overheat itself enough to melt the insulation due to EM but straight runs produce the same amount per given length regardless of overall length

 

[Steave]

Normally we are required to separate instrument cables (milliamp carrier ) by 12" from high voltage cables or put in a shielding of carbon steel plate. I have never heard of a requirement to separate high voltage cables but then on heavy construction, we dont have high voltage DC, only the low voltage control cables.
In any case, you should be ok with them if you keep them at least 12" apart.

I agree, we trenched and buried cable in pipe and placed DC, CCTV cables, and line voltage all in the same ditch but of course in different pipes. The manufacturer usually specified how far to keep communication cables away from line voltage and I've never seen one specify more than 12" apart. If we had several pipes sometimes we'd lay a layer cover with dirt then lay another layer.

There maybe something in the code about this but I've never had any problems just doing what the device manufacturer recommends.

We had camera cables and 110 v in the same ditch at a prison we wired and some on the runs were close to 1000'. The 110 v was for the heaters built in the cameras. The camera manufacturer also recommended 12" separation
"whenever possible" but at least 3".

I don't think you'd have any problems.

 

[joshuabardwell]

The amount of EM radiation coming off a cable has nothing to do with length of the run.

This is true, but the longer the distance over which the cables are parallel, the more they will "couple" and the more the one will induce EMI in the other.

To the OP: I am not an electrician, but I am a designer of computer networks and installer of computer networking equipment and cabling. We know that if we run data cable (Ethernet) parallel to high-voltage AC lines, the data cable must be protected from induced EMI or data corruption will occur. But we're talking about very low-level signals here. It's hard for me to imagine that the amount of EMI created in your situation would damage the components of the solar system. Surely anything that is handling 600 volt DC is beefy enough to take a little EMI from a parallel AC line.

For me, the main motivation for separating them would be the potential for an insulation failure allowing the lines to short out to each other. 600 volts DC is nothing to fool around with. I don't know what standard practice or code is in this circumstance, but if I were just making it up, I would probably put them in different conduit as an extra safeguard.

 

[BeezFun]

Thanks for suggestions, I talked to building department, they only know about requirement for phone/signal cable separation, showed no interest in trying to find an answer. I've put out the question to a few electrical engineers I've worked with, I'll see what they come back with.

 

[joshuabardwell]

Thanks for suggestions, I talked to building department, they only know about requirement for phone/signal cable separation, showed no interest in trying to find an answer. I've put out the question to a few electrical engineers I've worked with, I'll see what they come back with.

Can I suggest that you contact an electrician? An EE may have a theoretical understanding, but this doesn't exactly sound like the domain of most EEs I know.

 

[mikehaugen]

I would think that you would be ok at that voltage, as stated by others it is mainly a problem with low voltage wiring. Are you by chance using shielded cable? Probably not I would assume, but if so than I would definitely say it would be fine. If you are running shielded cable than be sure to only ground one end of the shield.

 

[joshuabardwell]

I would think that you would be ok at that voltage, as stated by others it is mainly a problem with low voltage wiring. Are you by chance using shielded cable? Probably not I would assume, but if so than I would definitely say it would be fine. If you are running shielded cable than be sure to only ground one end of the shield.

With Ethernet cable, you can buy it with foil shield in the insulation. If you can't get a hold of that, you can accomplish the same thing by putting the cable in metal conduit and grounding the conduit. Which leads me to the question... if you're putting metal conduit in the ground, do you really need a separate earth ground? What are you going to do? Put a ground rod in and attach a wire to the conduit? The whole conduit is a ground rod!

Maybe the "metal conduit" trick only works above ground.

 

[greasemonkeyok]

The amount of EM radiation coming off a cable has nothing to do with length of the run.

Uhmm, I'm not an EE, but that is just flat wrong relative to the question of how much EM is being transferred from the primary to the secondary. Think of it as a transformer laid flat instead of coiled.

The OP makes it sound like the solar run is DC, so that high voltage isn't going to affect the 110v.

And if you use metal conduit, there is the issue of capacitance which is way above my pay grade. So Joshuabardwell has the best suggestion of all.

 

[mikehaugen]

I thought about that too, though I don't think it is common to bury metal conduit. Most people that I know use some sort of plastic.

 

[joshuabardwell]

Think of it as a transformer laid flat instead of coiled.

That's a perfect analogy.

So Joshuabardwell has the best suggestion of all.

I hope you mean my suggestion about asking a professional electrician!

 

[Redneck in training]

There are two things to consider.
1.) The EM radiation decreases with cube of distance. In other words even few inches of separation make big difference.
2.) The common direct burial cable is twisted. The twist cancels out crosstalk to large degree to parallel conductors.

Therefore I think that to put 600 V DC and 110 AC in the same trench but in separate conduits is perfectly OK.
I am guessing that you will be using plastic conduit.
FYI As long as metal conduit contacts ground it doesn't work as signal shield. Shield works only if it is impossible to pass current through it. In other words it has to be isolated and then connected only on one end to common potential. Think about (in example) lightning strike causing current flow through the conduit. The EM will effect whatever is in the conduit. In example it could cause voltage spike in 110 VAC.

 

[joshuabardwell]

There are two things to consider.
1.) The EM radiation decreases with cube of distance. In other words even few inches of separation make big difference.
2.) The common direct burial cable is twisted. The twist cancels out crosstalk to large degree to parallel conductors.

1) is especially true given near-field vs. far-field effects. With regard to RF emission, when a non-emitting conductor is in the near-field of an emitting conductor, the non-emitting conductor has current induced in it and re-radiates as if it was a conductor itself. When the non-emitting conductor is in the far-field of an emitting conductor, the non-emitting conductor will still have current induced in it (this is how antennas are able to receive RF transmissions) but it will not re-radiate to a significant degree.

The exact size of the near-field and far-field depends on a number of factors. At 60 Hz, the wavelength is about 16 feet. Assuming your run of cable is at least 8 feet (1/2 wavelength), then the near field is equal to D^2 / 8, where D is the length of the cable. As you can see, the near field gets longer as the cable gets longer, which is why the potential for re-radiation (and induction of EMF) is so high with longer cables. On the other hand, once the cable is past a reasonably short length, you are basically guaranteed that any other cable in the same trench is going to be in the near field, so in a way, this makes the problem simpler, since moving the cable out of the near field is out of the question.

Basically, shielding the cable becomes your only option. But since we are talking about raw DC power lines and not sensitive, high-frequency/low-voltage data lines, the real answer, in my humble opinion, is to have the equipment on the line be sufficiently protected that the EMF doesn't affect it. It won't take much to filter out the induced current from an adjacent AC line, and I think it's likely that your DC power equipment is already doing this. But that's not my area of expertise, and it's just speculation.

Regarding point 2) yes, the twisted cable gives the ability for the equipment at the end of the cable to cancel out interference and cross-talk, but as far as I know, that requires that circuitry in the equipment at either end perform certain signaling techniques, such as common-mode rejection or differential signaling. I'm not aware of any standard AC equipment that does this. As far as I know, these techniques are exclusive to data transmission.

EDIT TO ADD: BTW, the near field of long cables is very large, but the inverse square law still applies. In short, the RF signal gets weaker proportional to the square of the distance. So at a sufficient distance, the re-radiated power is negligibly weak even though the receiving conductor is technically within the near-field of the transmitting conductor.

EDITED TO ALSO ADD: This conversation is what you get when you ask an RF engineer to talk about electrical wiring. There is some overlap there, but ultimately, if an electrician disagrees with me, you should do what he says. But if an electrician disagrees with me as to how to wire up a radio antenna, then listen to me.

 

[Redneck in training]

1) is especially true given near-field vs. far-field effects. With regard to RF emission, when a non-emitting conductor is in the near-field of an emitting conductor, the non-emitting conductor has current induced in it and re-radiates as if it was a conductor itself. When the non-emitting conductor is in the far-field of an emitting conductor, the non-emitting conductor will still have current induced in it (this is how antennas are able to receive RF transmissions) but it will not re-radiate to a significant degree.

The exact size of the near-field and far-field depends on a number of factors. At 60 Hz, the wavelength is about 16 feet. Assuming your run of cable is at least 8 feet (1/2 wavelength), then the near field is equal to D^2 / 8, where D is the length of the cable. As you can see, the near field gets longer as the cable gets longer, which is why the potential for re-radiation (and induction of EMF) is so high with longer cables. On the other hand, once the cable is past a reasonably short length, you are basically guaranteed that any other cable in the same trench is going to be in the near field, so in a way, this makes the problem simpler, since moving the cable out of the near field is out of the question.

Basically, shielding the cable becomes your only option. But since we are talking about raw DC power lines and not sensitive, high-frequency/low-voltage data lines, the real answer, in my humble opinion, is to have the equipment on the line be sufficiently protected that the EMF doesn't affect it. It won't take much to filter out the induced current from an adjacent AC line, and I think it's likely that your DC power equipment is already doing this. But that's not my area of expertise, and it's just speculation.

Regarding point 2) yes, the twisted cable gives the ability for the equipment at the end of the cable to cancel out interference and cross-talk, but as far as I know, that requires that circuitry in the equipment at either end perform certain signaling techniques, such as common-mode rejection or differential signaling. I'm not aware of any standard AC equipment that does this. As far as I know, these techniques are exclusive to data transmission.

EDIT TO ADD: BTW, the near field of long cables is very large, but the inverse square law still applies. In short, the RF signal gets weaker proportional to the square of the distance. So at a sufficient distance, the re-radiated power is negligibly weak even though the receiving conductor is technically within the near-field of the transmitting conductor.

EDITED TO ALSO ADD: This conversation is what you get when you ask an RF engineer to talk about electrical wiring. There is some overlap there, but ultimately, if an electrician disagrees with me, you should do what he says. But if an electrician disagrees with me as to how to wire up a radio antenna, then listen to me.

Thanks for depth of your explanation. I am working in process control/instrumentation field. I know how to troubleshoot and deal with interference but not necessarily why.
I think that you missed 3 zeros at the wave length at 60 Hz. It should be 16000 ft.

 

[k0ua]

I was going to say that the wavelength = constant/frequency in Hz. so :
Lambda = 3x10^8 meters /60 so Lambda the wavelength is 5000000 meters or about 16404199.48 feet.. approximately as I used an approximation for the speed of light as 3x10^8 meters per second.. soooo the old wavelength of 60 hz is pretty long... 60 MEGAherts would be about 5 meters or about 16.4 feet.. but now we are talking about vhf TV channel 2 lower end here. not 60 hz.. which is darn close to DC in the big scheme of things.

James K0UA

 

[darg]

The insulation on {all of the wire} should be rated 600 volts to be in the same ditch.

darg

 

[greasemonkeyok]

Yes, I meant the recommendation to consult an electrician because they have the code books to look stuff (a technical term for us laymen) up.

I asked the same thing once, except it was copper tubing for propane parallel to 220v AC and they said "#@$$ no." An explanation followed, much like the above discussion. All I got out of it was, "don't put nutin in the same trench with electricity" especially copper pipe to the propane tank. Propane delivery man won't like it.

 

[tcartwri]

Irregardless of all the fine answers here, code requires you to have a disconnect at the solar array for the DC run. Because it's DC, these disconnects run around $1000 for a 5000 W array.... Does that answer your question?

You should have your inverter(s) at the array, and make your home run to the meter in 240V AC. The inverter(s) will have the DC switch built in, for your required cut off at the array.