How big a shaft for driving a 20kW alternator

[longstroke]

It sounds like someone has previously tried to repair the coupling.

The picture makes the top of the drive pin look like a part above it was fractured off previously - can you post a close up?

Then someone has tried to make up adapter pieces to be able to reconnect to the vertical shaft above.

It would appear that you could try running the wheel as is without damaging the top of the shaft - after trying to see how good the bearing under the wheel is, which presumably incorporates a thrust bearing to support the weight of the wheel and shaft, and the reaction force due to the water.
Have you been able to turn this by hand?

This is a really exciting project - I'd personally love a place with a water turbine, or potential for!

With regards to frequency stability - what was previously done with the generated power as this will influence the possibilities?
I'm guessing this is in France, so I've no idea on the regs for connecting to the grid?

[jonesy]

I imagined the webs on the slotted piece being on the underside. ie the flat part with the slot being the top. The circular tube being to locate the part on the lower  shaft.

Hadn't thought of that. It makes sense.

There are quite a few broken bolts, but no worn or broken flanges etc. My gut feeling is that the drive pin has fractured, but the current collection of bits will work with a pin shorter than it currently is.

As you say, the next job is to open the sluice and spin the wheel. From there it should be quick enough to fit the 'new' coupling, the tyre and bolt it up.
It is an exciting project, however, it takes 2nd place to getting a house comfortable. It was it a real state, having been abandoned for many years. It's in good shape now, but plenty still to do.
The mill is in France, not far from Lourdes. Despite being close to a main road, there are no services near by. IIRC, the cost for a small supply was around €60k (after a substantial grant) and more if you want to  inject. That gets the supply to the boundary, then there would be the costs for 100m or so to the mill.
There is what looks like a sluice control system on the wall. However, there are no actuators, just limit/open/close switches. If restored, this could control the frequency, but manual tweaking will be required to see how the system copes under load. The current electrical installation is basic, so it's not obvious what was previously connected. There are a lot of halogen lights around.
If I can drag my friend away from building, I'll go back over and we'll spend some time trying to get it spinning.

[jonesy]

A quick update. Opened the sluice and the water wheel turned.
https://1drv.ms/u/s!AjwyWSZbtvVcugz2aic36flbT_2P
Not the most exciting video, but you can see how out of true it all is. Internally, the spigot was running oval - about 5mm per revolution.
The speed is hugely down on the required 3.6 rev per second. Getting about 0.8 rev per second, which works out to ~11Hz. The race could have been more full - maybe another 30cm.

[knighty]

I get a "sorry there was a problem playing the video" message

[Westie]

I can't get it to play online but it does download and play with VLC, it's an mp4 video.

The whole thing looks under bodgeengineered to me.  The main drive shaft looks undersized, if it had a bearing support just above the wheel it would be a lot more stable. Have you looked at the bearing arrangement under the wheel as it looks knackered to me or the shaft has bent.

 

[jonesy]

I despair of onedrive. Even as the owner I can't get the darn thing to play. As Westie says, it can be downloaded (28M) and played as mp4, but to cut to the chase, you'll see a 1m diameter horizontal wheel with ~100mm run out, on a long vertical shaft moving side to side. The noise and vibration was impressive inside, but probably normal for a mill!
The wheel and shaft weighs a good amount and could be hoisted up, down and out. The thrust bearing is on a concrete pad and does seem to be knackered. I hadn't thought about a bent shaft.

The wheel is pretty worn too from use. It's not known whether this was from leccy generation or milling.
Anyone got any good suggestions on new suitable products. We get a 1.4m head and a flow of ~300 l/s. Can improve the head by maybe 100mm, and there is scope to improve the flow by widening the current overflow channel from ~200mm wide. The race is ~3m wide.

[knighty]

downloaded it (can't believe I didn't see the download link!)

watched it on repeat, it's hard to tell, but I think your bottom bearing could be causing a lot of your problems, if the problem was at the top of the shaft you'd expect the shaf to be constantly off to the same side/position - also if the problem was inside at the top it would (probably) be pretty obvious when you were standing next to it

does it have a bearing or a bush ?  a bush would be more fitting with the age

and I think water could be used for lubrication in some old bushs ?

[jonesy]

I see what you mean - it could well be a bush. It was lightly seized. It's not possible to see the bush/bearing with the wheel in place.

There was plenty of wobbling inside, so I did even wonder if there was a top bearing.

[guydewdney]

I think it’s time for a low head turbine from our sponsors....