6.6kV Terminal Box explosion 1

[neps3]

Hi

This is an intersting piece of puzzle and I would like to welcome those interested to analyse and contribute.

Transformer size -20MVA, 15.75/6.6kV
15.75kV connection is via the PIB.
Terminal Box 6.6kV- 3 cables per phase. 400sqmm Aluminium.
6.6kV bushing is vertical and the palm is Copper.
Aluminium Cable with Aluminium lugs 400sqmm.
Connection - with two bolts 12mm dia. per lug. 70mm long x 50mm width. Two holes per lug but very close. Since the hole on the bushing palm did not align the hole was drilled on the lug. This results in one hole to overlap 50% on the other.

SOE Sequence of Events show that a 900HP Pulveriser motor was started before the explosion of the terminal box which ripped off the bleed pipe to the Conservator and caught fire.

Although the transformer did not fail, the rewinders advised that the coils have moved. The business warranted rewind of the transformer.

On receipt I did some modification.

I would welcome your analysis please.

Regards

 

[DanDel]

Agreed, dpc...

 

[neps3]

dpc

I have read most of the questions and answers. I find that some go around in circles. Some are very explicit and technical/engineering, but some?.

Anyway, I thought this is a useful problem of a root cause analysis which I wanted others to know and have an approach to engineering solutions.

I find that some non technical ask fundamental questions and some take their time to answer. Very basics. Things like someone working in a power station saw smoke coming from the Alternator and what the issue is?

Anyway thanks for the talk.

Regards

 

[Shortstub]

Hopefully, you will address the terminations at the switchgear as well as the JB!

 

[neps3]

Shortstub

The 3 x 3 -400sqmm cable is 160 metres long. The Cable were spliced around the 15 metre mark from the transformer TB with new straight through joints.

No work was done at the switchgear end. Inspection at the switchgear end proved no issues. The cable cores were checked for IR/PI and was very good in the order of GOhms.

Thanks

 

[Andiri]

Kantor, do you do infra red monitoring/imaging on your system as part of a condition monitoring program?
Loose connections can be spotted and repairs effected before they become a problem by comparing the temperatures of adjacent connections in the same box. We have an extensive infra red monitoring program implemented at our mine in Namibia and have purchased our own camera for the job, enabling us to repair connections and follow up on their post repair thermal condition soon after the repair. It has proven itself worth the investment since thermal imaging contractors are pricey and the long wait for post repair follow-ups.

 

[neps3]

Adar

Thank you for the info. Yes we have an $80,000 Thermal imaging camera where we do online monitoring of thermal images. These extend to 300MVA 275kV transformer connections.

For the above the terminal box is on top of the transformer and the problem was different.

During our thermal imaging program we have picked up hot joints and have rectified. The work is on a routine cycle.

Thanks

 

[neps3]

sengn

Thanks for the info. Yes it can be fitted and good. But justifying the installation with detection system in transf and motor terminal boxes is quite difficult in the modern era. I will stick to elekies doing a good job in the termination with the engineer providing a good procedure.

 

[sengn]

kantor

do you know the costs??

 

[neps3]

sengn

We have installed in a switchboard and the cost ran to tens of thous.

We have 84 HV 6.6kV motors, 6x300MVA,6x20MVA and 100s of other transformers. I am now modifying the 6.6kV motor terminals with self amalgamating tape and Raychem "COLD BOOTS". This means the terminal is completed covered and any arcing is contained. It only costs $200 for a set of three. Why would I want to go for back up detection systems when if I have not eliminated the root cause. Thats my philosophy. "Keep the house clean", "No bugs then".

Even otherwise, I strongly would recommend to contact ABB or GE on this issue.

I hope I am making the message clear.

 

[jbartos]

Suggestion: Generally, the nature of the malfunction as described and somewhat discussed above would indicate an abrupt failure in insulation and or isolation impedance between line conductors and potentially the ground. There, cannot be this kind of failure with the good insulation and the good isolation between the line conductors. It might have been a three phase grounded fault at very high fault level, perhaps at the higher fault level than the installation was engineered and designed.

 

[Shortstub]

But the root-cause was a poor terminal connection not ph-ph or ph-ground insulation. Arcing resulted by the additional current caused when the 900Hp motor was started. Arc products then caused the ground! Once initiated, arcing product dispersion also enveloped the other phases resulting in the 3-ph fault. The e/f relay failed to quickly clear the fault. Another point is that an arcing ground-fault will produce an effective or rms value far below the theoretical value. Of course, this will affect the response of the e/f relay. (Kantor, can you determine actual timing?)

Arc-detection as a cure seems to me to be another case of "catching the mice, and letting the elephants go free!" But, if one has the money anything is possible.

 

[Shortstub]

Kantor,

One last thought about the connection hardare. Unless the bolt is steel, and the reinforcing plates as well, the problem will not go away!

 

[neps3]

Jbartos

The palm and the connection were exposed inside the outdoor closed terminal box. The design then must have taken the clearances into consideration and I checked the clearances and it complies with the requirements. But the root cause is not insulation issues, but the bad joint as shortstub points out to us.

shortstub

yes the hardware is as you have stated.

Regards