Transformer overheating (external) - Circulating Current? 5

[KnicksJets]

We have a GSU transformer that has been running for over a year now. There are two support bars holding a conservator tank over the main tank. These bars are between phase A and C (two bars are right above phase B) low voltage bushing box.
We recently noticed overheating in the bolts in these support bars, temperature as high as 320 deg F. We were told it could be because of the circulating current.

Any input? If it's true, can anybody explain the phenomenon?

 

[itsmoked]

Unintentional circulating currents occur when a metallic structure near current carrying conductors can act like a single winding of a transformer. This can put voltage potentials on that metallic structure or induce currents large enough to heat them.

Perhaps something has changed to exacerbate the problem with your transformer. Unbalanced loading or higher loading? Or oil that may have conducted the heat away is no longer touching the bolts, etc?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[ScottyUK]

I assume the LV connection is via isolated phase bus. IPB has near-perfect magnetic field cancellation over the majority of its length, but at discontinuities such as Tee's and bushing enclosures the IPB usually has a significant residual magnetic field which is not cancelled by the image currents in the outer casing. The residual magnetic field will create eddy current heating. Check or mbad or missing connections to the IPB structure around the bushing housing.


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

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

 

[itsmoked]

Hey Scotty.

Question:
You could get induction heating which would be the shorted transformer situation.

And eddy current heating which would be a different mechanism.


Is this correct?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[ScottyUK]

Yeah, a shorted turn would certainly be possible but they're usually fairly easy to spot because anything large enough not to melt is going to be a reasonable size. The IPB casing is grounded at one, and only one, point with the rest of the duct being supported on pad insulators. Bridging the pad insulator can have 'interesting' effects on whatever forms the short.

Scaffold platforms erected near the IPB duct ends can get unexpected currents induced in them too. Smoking scaff tubes is the usual symptom. Obviously permanent structures would also be affected but design review should stop them even being built in that location.


BTW, the last sentence of my post above should read "Check for bad or missing connections to the IPB structure around the bushing housing."


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

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

 

[ScottyUK]

Meant to say that a broken or high resistance bonding conductor will cause poor image current cancellation, and these type of failures are the most common type of problem with IPB. Normally the main bonds are welded channel or box section spanning between the casings and these don't cause problems, so specifically look at the flexible conductors between the main duct and the bushing enclosure. If they are laminated aluminium foil, which is fairly typical, then corrosion can be troublesome.


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

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

 

[7anoter4]

I think that as the conservator and the support legs encircle only 2 of 3 [or 4] outgoing bars, the sum of currents between the legs is not 0 and there is a high magnetic field produced by this difference current. This close magnetic field produces, in turn, a circulating current which may heat the bolts. You have to deploy one of the legs in order to encircle all 3 or 4 bars.

 

[waross]

One or both of the supports may be changed to aluminum. An insulator may be installed in one of the supports.
As I visualize the installation, you have magnetic encirclement of only part of the magnetic circuit. In power wiring we expect heating problems to get severe when the vector sum of the encircled currents exceeds about 200 amps. Stated another way, put all your cables through the same hole in the steel.
When single conductor cables enter individual holes in a steel panel you will have issues. It just takes a hacksaw cut to join the holes to interrupt the magnetic encirclement and avoid the heating. I believe, from the description of the problem, that this may be the same effect.

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

 

[prc]

You have not mentioned about the GSU rating,type of bus connection on LV side ie isolated bus duct or open bus bars.Better to contact transformer manufacturer who can give solutions. In such cases, the solution is to give ample contact area for circulating currents so that heating will be reduced.In this case bolts are carrying current more than their cross sectional area will permit. One way is insulate bolts so that current wuil not go through them and by pass current through short copper wire connections.Experienced transformer experts will give solutions to you after studying the lay out etc.

 

[Kiribanda]

Pdshah,

Why don't you post a picture showing the support bars of your transformer for everbody's use?

 

[KnicksJets]

First of all thanks all for the input. I still have few questions, while I am attaching few pictures and answering your queires:

Attachment 1: Conceptual drawing of the heated bolts.


PRC: The GSU rating is 607 MVA 22KV on LV side and 345 KV on HV side. LV is connected to the isophase bus with rating 23 KV and 16,000 amps each phase.
Are you suggesting we connect a copper wire across the bolts (the lower two bolts in the X bar)?

Waross: Would appreciate if you clarify your suggestion.

7anoter4: The support legs are between Phase A and C. The legs are left and right to Phase B.

ScottyUK: Thanks for the explanations. Your response make sense and goes along with my recently developed theory. The LV bushing box is connected to the Isophase bus box by bolts and insulated by neoprene gasket. The tanks is leaking oil (from the top of the tank with neoprene gasket in the manway cover) suggesting the gasket might not be doings its job. Are we saying that the Isophase bus box should be completely isolated from the LV bushing box?

Itsmoked: Thanks to your response. Loading has been same. We have more oil leak, sure, but this part of the structure (Support rods) were not being cooled anyway. It looks completely discolored.

Anybody sees any other potential problem than the bolts melting away and causing the support to collapse (we don’t see it happening as there are plenty other supports to the conservator tank)?

 

[ScottyUK]

Pdshah,

If you look at the IPB main bar casings you will see massive cross bars between the outer casings of the phases just before they reach the LV bushing box. They aren't just for mechanical bracing, they carry the full image currents. Once you are beyond these main bonding bars the image current cancellation is poor or non-existant and external fields are high. The bushing box is normally connected to the main structure of the IPB at a single point to maintain its connection with earth, but not to provide a path for image currents to flow. Some designs provide a supplementary bond in the form of a heavy copper or aluminium bar between the bushing boxes although I am unconvinced how much this contributes to external field reduction. If it is fitted you may wish to investigate the soundness of the joints. The external magnetic field in the region beyond the main bonds will be very high and any closed loop formed by supporting steelwork and the like will experience large circulating currents. From the sketch I expect that is what is happening.

Bill's earlier suggestion of an insulated pad may have some merit.


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

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

 

[KnicksJets]

I am attaching the picture of my LV Bushing box to the Isophase bushing box.
ScottyUK, I don't see the crossbar you refer to in our isophase box.
There was one suggestion that the isolation between the LV Bushing box and the isophase box (the neoprene gasket) might not be adequate.

 

[ScottyUK]

That's a big 'un.

I can see the flexible connections which extend the IPB casing onto the bushing box (just above the date), so I'm guessing that they are using one large LV bushing box and relying on the box itself to provide the interconnection between the three IPBs. If I'm wrong please describe how and where the three IPB casings are interconnected.

I can see a lot of red sealant - Hermetite? - which may be contributing to your problem if it is causing a high resistance path between IPB casings. I'd expect something like Biccon or Alcoa compound if that is supposed to be a current-carrying joint. With a 16kA image current you don't need much resistance before it becomes a problem.


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

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

 

[prc]

I think your problem is the heating of bolts marked in red.
As Scotty raised, normally common (for 3 phases) LV bushing box is used in such application.But it seems in your case they are separate.Then the problems of circulating current will be high.LV bushing turret is usually of aluminium,In case it is of non-magnetic stainless steel,leakage flux coming out of box ix more.This flux is impinging the two vertical supports of conservator.The induced voltage in these will be different.When they are connected by the X stiffeners,they will carry enormous current ( I have seen hundreds of amperes)This current will pass through bolts as painted surface will not allow current flow.Heating will be directly proportional to the load current in LV.

I have seen this type of heating in many large GSUs.Some times small air equaliser pipes may get melted.Then increase the pipe size to give more area for current flow.

Easiest way to solve the issue, is to stop this current flow.This you can do by insulating bolts on one side.( using frp tubes and washers)
Another way is weld small stainless steel pads on the X bars and vertical supports near the bolts and then use at least 150 mm2 copper cable with socket to connect between pads.This way you can by-pass current through the bolts.

Will you come back with the results for the benefit of all?

 

[ScottyUK]

Hi prc,

Am I mis-interpreting your post or are you suggesting interrupting the current flow at the flanges where the red sealant is evident? If that box forms the main interphase bond at the transformer end then the whole IPB will not behave as it is designed to because image current will not flow in the casing and the whole IPB will have a massive external field. If the joint is designed to carry current then the joint should be made using the techniques you would employ for a busbar joint, because that is essentially what you are jointing.


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

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

 

[prc]

Scotty, I was not referring to flanges,but just the contact area of the X cross bars used to connect the vertical supports of the conservator.I believe isophaase box is closing the currents of isolated buses or there is separate conducting jumpers between the outer case of the buses.Even if bus duct current is getting cancelled, I think external flux may be substantial to create potential difference on the vertical support.

 

[ScottyUK]

Agreed!


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

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

 

[waross]

Magnetic encirclement;
The primary cause of heating in the vicinity of large currents is stray flux. At lower current levels, cutting the magnetic path will introduce an air gap in the magnetic circuit that will dramatically reduce the flux density and the resulting heating. However with extremely high currents the magnetic fields may be strong enough to cause heating without magnetic encirclement.
I will defer to Scotty on this one. He has more expertise in this field than I.
However, i did some searching on Iso-Phase bus to brush up on the principles and found a case history of an almost identical problem and the resolution of the problem.

http://www.reliableplant.com/ArticleSendtoFriend.aspx?articleid=7947

I hope this helps.

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

 

[ScottyUK]

Bill,

That's a great real-world example of how important the bonding bars are and why the joints on them are so important. Thanks for posting it.

The guys involved probably thought it was mechanical bracing or similar and not a fundamental part of the IPB.


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

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