Inductive heating

[GreyGoose]

Can anybody point me in the right direction or give a good reference to solve the following problem/question:

How much heating (i.e. max temp) will occur in a piece of steel conduit (sleeve) due to the magnetic field produced by a large power conductor run inside of it?

Although the NEC prohibits this (i believe), it has happened in the field anyway & i'd like to find out the impact.

 

[peebee]

While I can't answer your question, I can point out that you might not be asking the right question.

The right questions might include: how much current will try to flow through the conduit, how much voltage will be imposed on it, what problems can this cause with my communications equipment, are there any life-safety issues involved with the voltages that might be imposed on the grounding sytem, will my conduit couplings start sparking . . . . .

Yes, this is a code violation.

 

[busbar]

One reference is 99NEC300-20 Induced Currents in Metal Enclosures or Metal Raceways. An empirical followup is probably the only reliable course. You may want to anticipate significant cable-insulation deterioration.

 

[GreyGoose]

Well, there should be no current flowing through the conduit itself, as it is a single conductor insulated and jacketed cable. My pricipal worry is accelerated degredation of the cable insulation over time due to excessive temperature. The voltage level is about 15 kV (60 hz circuit) but the real worry should be the current flow, as that will determine the strength of the magnetic field produced. There are no communication issues as this conduit is only run alongside other concrete encased power ducts. I need to find out how bad the heating is going to get so I can get an idea of my options to mitigate the problem. Grounding is also not an issue at the moment (Unless the insulation fails, then look out). Just wanted to find out if there are any thermo or EMF wizards out there that can help.

 

[dpc]

If you are running a single phase conductor in a metal conduit such that there is a net current (and flux), there will be induced current in the conduit, even if the conductor is insulated. The flux creates eddy currents in the iron of the conduit, just like a transformer core.

Heavy current circuits - 1000 A and up - can create enough current in the conduit to cause it to heat significantly - to the point of glowing, so I've been told. We've had clients do this and they have experienced cable failures within a few days due to heat.

I can't quantify it for you, but it is certainly a very real concern. PVC conduit is one solution if you want individual phases in separate conduits.

 

[peebee]

Exactly. It's that current flowing through the conduit that causes the heating.

Or, if you open circuit the conduit at one end, instead of passing currents and heating up, it will develop voltages at the other end, possibly lethal voltages.

It's strongly recommended that the installation be corrected to meet code.

 

[GreyGoose]

yes, 3 phases in one duct is ok if you dont exceed the fill limits for the conduit size. check the NEC.
Also there is a recommendation in the NEC for running all 3 phases in a single ferrous duct (rather than a single phase or conductor) in order to negate this inductive heating effect.

BTW, peebee, i understand about EDDY current flowing in the duct due to induction, but it sounded to me like you were talking about the LINE current running through the conduit, which didnt make sense. I dont ever recall however voltages developing in one end of a plate or duct section due to eddy currents. if that was true, would you have to ground all the laminations in a transformer core (for example)? the section I'm talking about BTW is not a long run of steel conduit, but about a 2 foot sleeve near the cable termination. I would love to meet the code but that may not be possible this late in the game. i need to go to the field and take a good look as to what can be done...

 

[peebee]

If it's only two feet I wouldn't be too concerned about it (but keep in mind it's still a code violation. Just make sure they are well bonded together. Also, you should realize that if it's only two feet, then it would be relatively easy to correct. Replacing the 2 feet of conduit with a 2-foot pull box, for example, would be a cheap and easy fix.

Re: "I dont ever recall however voltages developing in one end of a plate or duct section due to eddy currents." -- this is about the same thing that will happen to the shield on a shielded MV conductor. Most cable manufacturers have information on this, as well as the IEEE Green Book.

Whenever you have two conductors (and a metal conduit falls into the category of conductor here) running physically parallel, the current in one induces a magnetic field which generally envelopes both, and the field induces a current in the other. It's basically a low-grade current transformer, and the conduit is the transformer secondary winding. The theoretical ratio is 1:1, indicating that the current in the conduit will try to match the current in the conductor, although the effective ratio will be less. If you open-circuit the conduit (the transformer secondary), this will interrupt the current flow. But the current will keep trying to flow, and will generate a voltage in doing so.

 

[busbar]

A lateral split in the ferrous sleeve will minimize inductive heating.

 

[peebee]

Re: "if that was true, would you have to ground all the laminations in a transformer core (for example)?"

The laminations are there to keep the currents from flowing in the transformer. It seems to me that the core would typically be grounded, but yes, I agree, that seems counter-intuitive to isolate the laminations and then bond them together again. Maybe you're just limiting the currents to the point where they are intentionally bonded.

I'm not that familiar with transformer construction, or with magnetic circuits in general. I'm much more of a practical electrical kind of guy. This is getting into the realm of speculation and conjecture for me, and I'd love to see some other comments on this.