Heating of Ferromagnetic Material surrounding HV Cable 2

[WhiteyWhitey]

Hi,

I am aware that when a single phase cable is enclosed by a ring of ferromagnetic material that it can heat up and cause damage to insulation. This I have learnt from previous posts.

My problem is that I need to justify this in a report to my client.

We have a HV Cable 132kV with an earthed screen running through a steel support structure. The structure has four vertical posts (parallel to current direction) and cross members the whole structure is bolted together without an insulating plate or anything.

Can anyone point me in the direction of the physics behind the heating?

I would ideally like to run a calculation on paper to find the circulating current under fault conditions.

Thanks in advance for your help,

Andrew

 

[dpc]

It's due to the magnetic field created by the current in the cable inducing current in the conduit. For ac current the magnetic field is always changing. A changing magnetic field cutting a conductor induces a voltage and the voltage creates a current.

This is a load current problem, not a fault current problem, so I'm not clear what you are asking regarding "circulating" current.

 

[davidbeach]

You have a CT with a ratio of 1:1. From that point out the math is rather simple.

 

[WhiteyWhitey]

Hi DPC,

Thanks for the reply.

I am familiar with the situation of a steel conduit.

In this situation, I have steelwork that is surrounding the cable. I am told that under normal load, the steelwork may heat up.

Under fault conditions some sparks could be expected at the bolted connections.

I have drawn a quick sketch to try and make the situation a bit clearer. Hope it clarifies things.

The blue lines represent galvanized steel with a right angle cross section.

The red represents the cable.

Thanks,
Andrew

 

[ScottyUK]

132kV sounds like a British-influenced installation. If it is a 132kV system then presumably this will be a three-phase cable, otherwise it would be operating at the line-ground voltage of 76kV. A three-phase cable won't cause circulating current problems because the magnetic fields cancel, or at least they do when you move away from the immediate vicinity of the cable.

Can you confirm whether this is a single or three phase installation?


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[WhiteyWhitey]

Hi Scotty,

The Cable passing through the steel loop is single phase,
The system is 50Hz AC to clarify also.


David,

I agree that the steel structure surrounding the cable is like the Core of a CT. However, my understanding was that the core of a ct is then wrapped in copper coils. The flux linkage between the main conductor (primary) and the copper coil on the CT (Secondary cause a current to flow. The math for this is Ok. For this reason I see finding the current in the steel structure subtly different to the ct.

In this case, I only have the steel bar. I am concerned about eddy currents in the structure and also what happens under fault conditions.

Example 1, if I were to use a solid steel plate and pass a single phase conductor through a hole in the plate, you would expect the plat to heat up if the load current was high enough. What direction would the current be flowing in to cause the heating?

Example 2, I am told that during testing under fault conditions, the high currents in the steel work cause sparks to fly at the bolted connections between the cross members. This is apparently due to the high resistance galvanizing between the two steel members. What direction is current flowing? How do I work the magnitude?

I currently know that the magnetic field in the vicinity of the cable will be a circular field around the cable with magnitude.

B = uI / 2*Pi*r

The circular field will induce a current in the loops on the side view because the B field is cutting the vertical steel members in the same direction.

A changing magnetic field here will give a changing flux and therefore EMF in the loops on the side. This is classical mutual induction.

What I dont understand is where any current can come from a closed loop at 90 degrees to the direction of current flow.

Im trying as hard as I can to state the problem clearly.

Thanks for your help,
Andrew

 

[waross]

The Canadian code allows up to 200 amps before derating for the heat produced must be used.
The steel forms the core and the conductor of the 1:1 CT. An actual CT has a laminated core to prevent or greatly mitigate current flowing in the steel core.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[7anoter4]

I remember a similar problem with 750 mm^2 110 kV cables lying on racks in a tunnel.
The problem –as I suppose-was the phase-to-ground short-circuit heated the iron close loop mainly at connection point.
The posts were hanged between ceiling and floor by mean of bronze inserts and the racks were cast aluminum made.
By the way, using bronze bolts instead of steel it poses another problem since the bronze screw is not so resistant as steel and the thermal extension or retraction force in the iron structure may break the bronze screw down.

 

[WhiteyWhitey]

Ok, the picture I am getting is that the Loop of steel will act like the core of a CT. (Circular Magnetic Field)

Eddy currents will be formed around this circular field.

The current density in the object will be the main variable dictating heat dissipation.

Current Density seems to be higher at the corners and connecting pieces.

I suspect to get a good quantitative answer I'll need some 3D simulation going on.

Thanks for the help!

Andrew

 

[ghostbuster7]

On another forum I just posted a situation where the steel support framework was heating up underneath the power cables.Infra-red pictures have been included.At the end of the thread I also posted the answer/explanation.I hope this helps,here is the link.

http://forums.mikeholt.com/showthread.php?t=119373

Ghost

 

[7anoter4]

It is not entirely correct if we think the close loop iron angle as a transformer. May be it is a transformer without secondary. All the losses are in the "laminate".
If we neglect the hysteresis losses and we'll take only eddy current losses the usual formula is:
[see -for instance:

http://www.utexas.edu/research/cem/images/CharacterizationPowerLossesSoftMagneticMaterials.pdf

Pe=(pi()*B*t*f)^2/(6*10^7*d*ro) w/kg
where flux density (B) is in Gauss, frequency (f) in Hz, laminate thickness (t) in cm, density (d) in g/cm3, and electrical resistivity (?) in ??-cm.
If the "laminate" thickness will be 5 mm and the iron properties:
d=7.8 g/cm^3; ?=0.1 ohm.mm^2/m=10 ??-cm; miufe=2000 relative permiability what we need is B.
Let's say cross members will be arranged in a square 0.5*0.5 m. The periphery length is then 2 m.
If will be a non-magnetic gap in this periphery of lengap [m] length, the "magnetic circulation" will be:
Sum(H*L)=Sum(I) that means:
Hfe*(lengfe-lengap)+Hgap*lengap=I[A] where Hfe and Hgap are the magnetic field intensity.
Hfe=B/miuo/miufe ; Hgap=B/miuo.
B/miuo/miufe*(lengfe-lengap)+B/miuo*lengap=I[A]
B=I[A]*miuo/[(lengfe-lengap)/miufe+lengap] Wb/m^2
miuo=4*pi()/10^7 H/m vacuum permiability
If lengap=10 mm[0.01m] then B=0.114 Wb/m^2 =1143 Gs and then Pe=((pi()*1143*0.5**50)^2/(6*10^7*7.8*10)=1.7w/kg
the weight of a iron square could be 0.5*5*7.8*200=3900 g=3.9kg
The total losses with gap=3.9*1.7=6.63 w
If lengap=0 then B=1.25 Wb/m^2 and Pe=208 w/kg
The total losses without gap=3.9*208=811 w and in free-air of 30 oC the iron temperature is 100-120 oC.
If the current will be 1700 A B=2.1 [saturated state] total Pe=2340 w and the iron temperature may be
300 oC.

 

[wareagle]

WhiteyWhitey
You said you have a single phase cable with a grounded shield. If so, I would think this shield would not allow the field to cause a problem. Maybe some one else could confirm this concept.

 

[itsmoked]

Hi wareagle.
Shields are really all about capacitive shielding not inductive shielding. These heating effects are caused by inductive coupling.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[WhiteyWhitey]

GhostBuster,

Thanks for the post! Thats a very similar situation to what I have.

7anoter,

Whoa! Thats the sort of detail I needed. I havent fully tried the application myself but I'll spend some time trying it out. Thansks. (Star for you!)

Wareagle,
itsmoked is quite right. The effects I am concerned about are from inductive coupling, I.e Electromagnetic induction. The magnetic field will permeate the shield.

Thanks again.
I'll post up the solution once I have applied it to my particular situation.

 

[xxjohnh]

Need a good mechanical heat transfer man. Ran into a similar problem in China. The contractor had installed a duct bank of several kilometers with rebar between ducts that contained single phase 110 Kv cable, a design defect in the engineering world. The contractor produced a calculation by a well known college professor who calculated the heating effect. It was neglible. This was accepted by the government body.
It is only the steel or iron between conductors that is the problem. not around all three. Maybe easier to change the steel to stainless or cut it out and design a structural retrofit.

 

[7anoter4]

I had some problems with the temperature appreciation [the Nusselt, Reynolds and Grashold coefficients was partially in SI units and partially Imperial also].So actually the temperatures are more elevated.
Then for "no-gap" and 1000 A the iron temperature [cooled in natural ventilation of 30oC air] will be 300 oC. and for 1700 A [through 132 cable] will be 700oC.