Corona Discharge on HV Transformer Terminals

I don't think at 22 kV it could be corona effect damage, but it seems to me like a surface discharge due to a small creepage path. There is a cable termination provided there?

Sharp edges and points intensify the electric field and result in corona discharge.

What...no stress cones???

The transformer passed PD test about a year ago and the next scheduled test would be around this period. (Murphy's Law)

The transformer was tripped by instantaneous O/C earth fault protection.

Can the fault be initiated by this "corona" discharge at the termination which is some distance away from the top of the coil (Phase-A) as shown? It is not known when the discharge had started and for how long.

Rixter has already given a good hint that sharp edges give rise to corona.

In your case from Photograph it does not look like corona (as Corona is ionization of air around points of electric stress concentration. This does not causes erosion of live parts).

From photograph it appears as if after initial corona, arc has struck. To confirm if this was arcing, did you find similar spot on box around the terminal (may be on any projecting out bolt head or similar sharp corner at shortest distance to the live terminal).

As said by Rixter, only covering with insulation will not help. To avoid Corona or arc, first we shall smoothen off sharp edges or corners with file to round these off.

Reason being that dv/dx (Voltage gradient) is maximum at sharp points (theoretically infinity) and minimum at spherical surface.

Hence sharp edges will result in breakdown of air around it and thus effectively reducing distance of live terminal to earth and ultimately resulting in flash over specially during high humidity (rainy seasons).

After smoothening all sharp points on live terminals, fill up uneven surfaces/gaps with insulation putty filler. We get commercial fillers for such application (very common to use these in jointing kits of HV Cables). These does not harden and can be easily removed for any maintenance activity.

Next cover it with one or two layers of Half Overlapped splicing tape.

If this is a 22kV termination, and a Raychem heatshrink term then you obviously have NOT completed this termination as per manufacturers specification. Clearly the person trying to offer you his services has no idea as to how to produce a termination on any MV system. Why are the bushings exposed at this voltage?

Good advice: remove this connection, strip it back and check for carbonising, clean, check and remake the entire termination and encapsulate the bushing with Scotchfill putty, 23 tape and 33 tape. Or alternate method of sealing and rounding/smoothing the bushing connection.

Is is just me or has this field of work suddenly filled with chancers?

thanks for the advice.

I've looked at the specs. on Schotchfill putty, appears good for voltage up to 600V but not too sure for 22kV applications.

I agree that there shouldn't be exposed metal part on the termination, however, other transformer manufacturers' catalog also depict similar connection methods. Perhaps the clearance distance is adequate in the design such that it does not require full encapsulation.

I am keen to find out the probable cause/s to prevent future occurance. I coin the word "corona" discharge because there was no obvious sign of arcing seen on other parts inside the transformer enclosure (apart from the flash-over between the coils on the top portion, but I think it was the result of the discharge). Indeed the connectors making the delta connection on the HV windings has clear signs of erosion and pitting.

Dave, having helped design the product, I can assure all, Scotchfill is ideal and it is per layer rated at a highly stretched application. The alternate is the green BICC putty which leeches oils after a few years use. If you don't want work, then install bushing boots from 3M or Raychem.

This is clearly discharge from the bushing and carbonizing will have started in the cores of the termination. The next step, it will track to the stress cone and the term will fail, rather quickly. There is no penciling on the term core at the lug, it is clear that there is a square cut edge where the lug meets the insulation.

I will disagree with you on clearance. It clearly requires insulation over the bolted connection, whether it is 11, 22, 33kV, the bushing requires the exposed metal and lug to be insulated.

Your closest and weakest path to earth is this termination and this discharge is happening all the time the cable is energized, and it will grow worse each day until it fails. Once it starts it proceeds quite quickly from the lug to the top of the stress cone and here it will blow out.

My specialty is cables, terminations and jointing up to 500kV. This is not a good termination connection and it really needs urgent attention before a failure occurs costing more money than is necessary. Just good free advice, to be accepted or rejected. I shall leave this for you to decide.