Loose wedge testing for Brush generators carried out on site by manual tapping
Loose stator winding wedges allow the stator winding in the core slots to move because of the electromagnetic bar bounce forces and vibration. This can in turn cause the ground-wall insulation of the winding to wear and fret against the stator core iron and eventually fail by ground fault. Therefore it is necessary to maintain the stator winding wedges tight to avoid such problems.
I will try and describe and explain how a loose wedge test is carried out on site and why it is important, especially on Brush machines that only skilled and experienced engineers attempt to replace wedges. First let's explain how Brush wedged machines differ from other major manufacturers
Brush use an epoxy felt system where the slot has felt at the bottom of the slot, in the middle and at the top underneath the wedges. This system is designed to take out any inaccuracies in the coils and wedging materials and completely fill the under-the-wedge areas. This felt is compressed using an impounder type device after which the coils are tapped in over the felt, the epoxy loaded felt will rise to fill small deviations in the coils and wedging material and hence a fully bonded wedge.
After all the wedges are in position and the rest of the generator winding has been completed the whole generator is put in a large oven and heated up to a Brush designed temperature and duration. From this operation the epoxy filled felts are now all completely cured and a ‘tight wedge’ situation is achieved from which in an ideal world there should never be a case for re-wedging the machine. In reality though badly supplied materials in the past, not Brush’s fault, over a period of time gave way to shrinkage of the wedges and non-fully cured epoxy felts which in turn left some loose wedges.
The first case of loose wedges that I came across was some 30 years ago after a failure in a Middle Eastern generation plant. These loose wedges contributed to mechanical damage to the sides of the coils and eventual failure and loss of generation. From this experience myself, the chief electrical design engineer and senior insulation engineer at the time came up with a test to identify loose wedges in a generator and a repair procedure to overcome the problem.
The procedure for finding loose wedges was to tap the length on the wedge with a small blunt metal instrument (e.g. an 8 oz ball peen hammer) and listen and feel for the vibration of the wedges recording the degree of looseness on a chart (see below), this was a simple test but very effective and for many years this process has been used. In today's world, electronic testing devices, like ElCid developed by IRIS Power/Adwel are used, but, on these Brush machines there is no replacement for experience when dealing with a felt system under the wedge.
A criteria was established to classify what is considered to be a loose wedge and one that is not and what percentage of a complete slot of wedges would have to be loose before they had to be replaced.
This criteria states that if the first 250 millimetres of wedge at each end of a slot was tight and any other 25% of that same slot wedge was tight it is deemed as acceptable with no re-wedging required. However in a lot of cases some very severe loose wedges have been found towards the centre of the slot where they had vibrated so badly from side to side that fretting and material loss has be noticed, in this instance the wedges should be replaced. This brings up another problem in the fact that if most of the wedges are classed as acceptable some may still have to be removed to access the badly damaged ones and this is where a great deal of care must be taken.
To remove a tight wedge from a slot could cause some severe damage to the top of the coil if great care is not taken as the wedge and coil can be bonded by the epoxies. The operation of removing the good and tight wedge can result in the top layers of insulation being badly damaged with a prospect more damage being caused than repaired.
This is where you have to have experience personnel to carry out the type of work and the correct tooling to achieve a permanent and safe re-wedge.
I have personally been to two sites where a third party company has re-wedged a Brush machine where during the process of removing the wedges they have caused so much damage that eventually the stators had to be rewound due to failures to the coil insulation.
When replacing wedges in Brush generator a total understanding of the system is required and procedures must be followed very carefully. This will involve some slight re-designing of the new wedges, new epoxy filled felts cut to specialised sizes, an epoxy filler tape, graded varnish’s, epoxies and the special tooling required to apply pressure to the newly added felts.
Under no circumstances should the old epoxy felts be removed from the top of the coils unless, they are so badly damaged by corona activity, it deems they have to be replaced.
Richard Pipes - Senior Winding Engineer GustoGen