53,760 hours.

 53,670 hours is the number of hours in just over 6 years, and this week I decided after this length of time it was time for the bearings to be changed; this blog illustrates what everything looked like when I took the bearing-housing apart having never disturbed it in all that time.

I installed the bearings on 5th July 2017 and they were SKF E2 Energy Efficient bearings; they have run continuously ever since bar one period of 48 hours in October 2018 when there was insufficient water; the only other times when the turbine shaft has not been turning is when a nozzle is changed, or the stator is changed, and such stoppages are typically for only a matter of minutes.

Of particular note is that the bearings have not been greased at all; neither was grease preloaded into the bearing housing when they were first installed; they have operated only on the grease put in by SKF at manufacture.

I made the decision to change them only because 6 years seemed a long enough interval; there were no warning signs of impending failure that prompted the change; literature from SKF suggests that the bearings can be expected to be serviceable for up to 9 years; after this length of time the grease will have come to the end of its ability to lubricate and 10% of a batch of apparently identical bearings will fail; this metric for predicting the likelihood of a bearing failing is called the T10 life expectancy; it is very much dependent on the conditions under which the bearing is operating, especially the load it is carrying, the temperature it is operating at, and most especially whether water and other contaminants can get to the rolling parts.

6 years is thus within the T10 life expectancy for the bearings - except the big unknown is the conditions under which they are actually operating.

So here is a pictorial account with captions of what the seals and bearings looked like: -

on my turbine, I have a specially made cover to help prevent water ingress; its purpose is to give a metallic face for the V-lip seal to rub on; it fits over the plastic 'Top-hat' and is held in place only by being a tight fit. 

a puller was needed to get it off.

another modification I have made on my turbine is this deflector to discourage water from entering the drainage hole of the 'Top-hat'

the first glimpse of the condition of the shaft indicated that little moisture was getting to it

for comparison, this is a picture from the previous bearing replacement in 2017, which was done after just 14,448 hours of operation and before steps were taken to prevent water ingress; it shows limescale encrusting the shaft, indicating that quite a bit of water was getting in.

detail of, and explanation of, the marks on the shaft; the brown colouring was of silt-like consistency and rubbed off very easily with wire wool; the surface of the shaft was not scored where it had been polished by the seal.

I was surprised by how much the stainless steel of the cap had been worn by the rubbing of the V-lip seal; it was almost as bad as the wear on the plastic of the Top-hat in the next picture, except that the plastic wore to be like this in a matter of months.

when I had removed the dust shields of the two bearings, the grease around the balls still looked pretty good

a close-up of the balls shows they were still well lubricated

SKF's E2 bearings have a very different looking ball cage; this is what the reverse side looks like.

the radial shaft seal on the inner end of the Top hat was a bit mucky; it has two lips and the second picture is a close up of the space between the two lips showing it to be full of the silt like material present on the shaft.

In summary, the bearings looked good enough to do another few years, but the seals were in need of being changed. The feel of the shaft rotating in the housing was of very free rotation as if the bearings were well 'run-in'. By comparison, when I had put in new bearings, of the same sort, the rotation felt rather stiff and not so free.

Unsurprising then that when I powered up the turbine with its new seals and bearings, power output was 16 watts down on what it had been, - that's a 4% loss of efficiency, - down from 44% to 40%, taking efficiency here to mean whole system efficiency, ie water-to-wire.

Ah well, - better a small drop in generation now than having the bearings fail later, - possibly in winter and having to do the job of changing them on a cold, wet day, as an urgent rather than an elective undertaking !