ronnie497
Electrical
- Apr 4, 2004
- 4
Hi Everyone,
I have been involved in the post mortem of a 60MVA H2 cooled turbo generator failure.
The machine was rewound. Originally it had "straight" bars, overhang clip transpositions. This was to be replaced with 360 deg roebel. Due to space constraints, 180 deg roebel was employed together with overhang clip transpositions. (96 slots, single circuit)
A top coil limb leading to the Blue interphase vapourised for over 30cm. (Very short time on load). Dissection has shown serious inter-adjacent strand arcing in the limbs leading to the other interphase connections.
After tracing the transpositions through the winding I have developed a theory that the 180 deg roebel can cause a voltage difference to appear between adjacent strands that start about the centre of the bar as they will depart from each other and only recombine at the bar end (Left Up, Right down). All the bars are of the same twist (Clockwise). As the stacks are reversed at the series (interphase) connection, this voltage increases from the star point up to the series connection and then reduces towards the line connection.
Not having any design experience I am unable to qualify this theory. Does the main flux density vary with depth or is it only the quadrature flux? (How much, can I guesstimate it?).
All suggestions and help are welcome.
Thanks
I have been involved in the post mortem of a 60MVA H2 cooled turbo generator failure.
The machine was rewound. Originally it had "straight" bars, overhang clip transpositions. This was to be replaced with 360 deg roebel. Due to space constraints, 180 deg roebel was employed together with overhang clip transpositions. (96 slots, single circuit)
A top coil limb leading to the Blue interphase vapourised for over 30cm. (Very short time on load). Dissection has shown serious inter-adjacent strand arcing in the limbs leading to the other interphase connections.
After tracing the transpositions through the winding I have developed a theory that the 180 deg roebel can cause a voltage difference to appear between adjacent strands that start about the centre of the bar as they will depart from each other and only recombine at the bar end (Left Up, Right down). All the bars are of the same twist (Clockwise). As the stacks are reversed at the series (interphase) connection, this voltage increases from the star point up to the series connection and then reduces towards the line connection.
Not having any design experience I am unable to qualify this theory. Does the main flux density vary with depth or is it only the quadrature flux? (How much, can I guesstimate it?).
All suggestions and help are welcome.
Thanks