There are 3x283 MVA 20kV / 420kV transformers employed as generator transformers to export generation to the switchyard at a power plant here. While 'back-charging' the transformers from HT (420 KV) side, sparks were observed at the instant of charging coming off the bolts that bond the base to the mid section body of the transformer. The sparks didn't appear to strike the earth. A feeble layer of soot was visible on the few select bolts that seemed to have sparked. This happened only in one phase of the transformer.
Has anyone else noticed something like this before? Can someone tell me why this occurs?

The bolts weren't tight enough.

Alan

Small air gaps can produce sparks at the instant as high charging current can cause induction.
Yes the above answer is correct.
Is that this particular phase the middle phase?

Thanks, Mr. Maltz and Mr. Indukuri.
The bolts were tight, they were verified again. There might have been air gaps in them, though.
Mr. Indukuri, yes, it was the middle phase. How did you guess that?