Higher efficiency pelton wheel?

[MaxFlow]

If you have a tightly sealed pelton type wheel, could you increase its efficiency by either creating a vacuum in it or by pumping in a lighter than air gas ?

What gas might be a good candidate?

[Nando]

Pelton wheel with gas

Can you in detail try to describe what your thoughts are regarding your scheme ?.

Pelton Wheels are water specific.

Nando

[thunderhead]

Given that Pelton wheels are better than 90% efficient anyway, what are you hoping to achieve?

[MaxFlow]

lower drag should always be a thought to explore.

[Ungrounded Lightning Rod]

If you make it airtight (without appreciably increasing the friction on the seal) and arrange your tailstock correctly (so it fills with water yet doesn't appreciably retard the flow) you can get it to run at a slight vacuum.  This adds the head of the drop from the wheel to the wheel outlet to the water level near the tailstock outlet to the head from the headstock inlet to the wheel inlet.

It's as if your wheel were running submerged at the level of the outlet stream - but without having to actually submerge it.  convenient for keeping the genny dry.

You can only get a few feet this way.  But every foot helps.

Getting the air out of the tailstock, and keeping it out, may be a problem.  (For starters, even a slight leak will let the tailstock refill with air and you'll be back where you started.

[Ungrounded Lightning Rod]

Now that I think about it this may not work.  I don't thing peltons work well with the housing full of water.

There are other types that do, however.  The mitchell/banki for example.   It has a somewhat lower efficiency than a pelton.  But it runs fine full of water and over a wide range of heads.  (Peltons prefer high heads.)  A mitchell/banki is also very easy to fabricate out of sheet metal.

[Flux]

I don't think any impulse turbine will run submerged, you need a reaction type for this. It's ok for Thompson, Francis, Kaplan etc where all buckets run full but I can't see how it can work on a Banki, Pelton or Turgo.

Flux

[Ungrounded Lightning Rod]

I understand a mitchell/banki runs just fine full of water.  (They're used for fleapower generation to run at least one model of electrical plumbing fixture.)  At least one site that sells big commercial models recommends the tailstock-full-of-water-to-increase-head hack, too.

With a M/B the water enters through a narrow opening where the wheel is moving with it, and leaves through a wide one where the wheel is moving away.  With the difference in opening sizes you don't need to compress the incompressible to get a change in velocity, so being filled solid isn't a problem.  (Also:  Except for the openings the housing can by cylindrical, so there's little fluid friction there.)

While M/Bs are only about 80% efficient vs. Pelton's 90+ they're still no slouches, converting maybe a couple hundred KW in something significantly smaller than a bathtub.  Since they involve no compound curves you can weld 'em up out of sheet metal quite easily.  Also: they have no really narrow nozzle or other tight passages to clog (and anything that gets pasted up against the rotor vanes at the input side of the rotation gets blasted off at the outlet side).  So if you keep the tree limbs and stray animals out of 'em they're pretty low maintainence.

[maker of toys]

I agree, reduced drag is a worthwhile goal.  the question is, how much do you actually save?  on formula 1 race cars, (v10 engine, 16000+ rpm redline, 900+ hp, 2.5 litres displacement (IIRC), 2 hours between overhauls, budgets for engine development bigger than most state's lotteries, national and/or corporate pride on the line-- this is not mommies' minivan) some teams pull a vacuum on the crank-case to reduce air drag on all the oddly shaped ironmongery that flails around in there; and they pick up substanial horse-power over what the pump requires.

but they pay a price, too; the vacuum increases oil consumption, as the hot oil is that much closer to boiling, and the vacuum also draws dirt from the outside world into the case seals.  So reliablility suffers.

now to the question of pelton wheels:  I agree with other posters, that filling the wheelhouse with water would substantially increase drag.  But you could provide a calibrated leak, such that the wheelhouse never got to a low enough pressure to support the column of water represented by the tailstock, and avoid that pitfall.  And you'd probably gain a bit of efficiency from lower air drag.  but you pay for that gain in two ways:  1, you've got to pump out the air in the first place, and 2, you have effectively created a negative head on the tailstock, which makes it harder for the water to leave once it has given up its mechanical energy to the wheel.

If you've already got the wheel at the maximum head you can achieve with your site, then you're probably chasing a wild goose; you'll need some fall to get the tail water to overcome the vacuum.  If you haven't optimized your head, then that's the place to start, not more ironmongery.  

I think you'd have to have a really large setup to see any meaningful improvement from this.  the same amount of effort and money could reduce your electrical losses by the same absolute percentage and get MUCH more real advantage from something the size we here would build.

 

[Ungrounded Lightning Rod]

IMHO the point of pumping down the housing is not to reduce air drag (though it does that too, of course, as long as an air bubble remains).  The point is to increase the effective head.  Every PSI you lower the pressure in there is equivalent to adding about two feet to the head.  Doesn't matter if you're sucking the water in from inside or pushing it in from above - only the pressure difference matters.

Of course that means the entrance to your tailstock has to be that much below atmospheric pressure to get the water back OUT without losing your vacuum.  So you need a tight seal, a completely below-surface tailstock outlet, as well as something to entrap air when there's more in the turbine housing than you want.  (When you're pulling a vacuum you can't bleed the air out through a hole - any leak will bleed air IN.)

If your turbine works better with air, rather than water, around it, you need some way to insure the necessary air bubble is maintained.  IMHO that should be doable by arranging the mechanism that entraps the air in the outlet water so it stops working when the bubble is reduced to the desired size.  You probably don't need to have an explicit bleed hole to keep the bubble from getting too small:  The lower pressure will encourage disolved air to come out of solution and bubbles coming in to stick around.  The trouble will be to keep from sucking air into the tailstock outlet and completely break your vacuum, losing the advantage of converting tailstock height to head.

[maker of toys]

. . . the bleed hole was actually for maintaining the bubble in the case.  but, as you say, there are plenty of other sources of gas for maintaining the bubble. . .  i was in thought-experiment mode, and cases there are perfectly sealed,  and fluids are homogeneous. <G>

a venturi pump (familiar to chemistry students at impoverished community colleges) can use a small mass flow of water at moderate velocity and head to pull quite a suprising vacuum on a moderate sized volume. If you had sufficient water that you could give up 2 liters a minute, you could go that route.  You'd have to balance the benefits of the vacuum over what 2 kg/min of extra flow would gain you.

for a pelton wheel, I don't know as .25-.5 bar of reduced case pressure is going to affect the jet velocity all that much;  I suspect that the nozzle is a choked flow at maximum efficiency, and so increasing the pressure differential by 5%  won't add up to much.  But; with a high-rpm wheel spinning in an air- and water-filled housing, I think there's quite a lot of potential pumping loss that could be attacked. . . 30% reduction in that could be worth doing if the complexity didn't get out of hand.  (which it inevitably will, but, hey, this is supposed to be fun, right?)

(mmmm.  I think me and my old fluids book are going to be reacquainted soon. . .)

[Ungrounded Lightning Rod]

A venturi pump (familiar to chemistry students at impoverished community colleges) can use a small mass flow of water at moderate velocity and head to pull quite a suprising vacuum on a moderate sized volume. If you had sufficient water that you could give up 2 liters a minute, you could go that route.  You'd have to balance the benefits of the vacuum over what 2 kg/min of extra flow would gain you.

Unfortunately, you can't use the venturi pump to depressurize the case - since it's being fed by a high-pressure water supply (and must be, in order to work).  The primary depressurization must be through the tailstock.  (Otherwise you'd be able to make much better use of the water power running the venturi to run the turbine.)

But you should be able to use it to DEBUBBLE the case:  Put a vent hole where you want the bottom of your bubble, hook it to the vacuum port of a small venturi, and bypass a tiny stream of your feedwater to power the venturi, which sucks even harder than the tailstock but only on the tiny hole.  Air bubble gets sucked out until the water level rises to the hole, after which you only suck out "used" water, in parallel with the tailstock.

[maker of toys]

yeah, 'Bubble control' was what I had in mind with the 'calibrated leak' and venturi pump. Sorry if that wasn't clear to everyone.  I thought the futility of trying to pump away the xx kg/sec waterflow with 2 kg a minute through a venturi was obvious. My thinking was, if you could have the tailstock sealed and drawing away the water mass, you could then reduce the air pressure in the case to reduce 'pumping losses' from the centrifugal effects of the spoons. put the 'leak' at the bottom of the case, the line to the venturi to the top, and voila- partial vacuum in the case.  makes tailstock design a little simpler; you'd just need a 'trap' type affair somewhere to keep from overwhelming the venturi pump.

Works pretty good in my head, but that may be solely due to the prevailing vacuum in there. . . .<G>

(I wonder how much angular momentum is stolen from the wheel by water droplets being flung off?  Not that there's any way I can see to cure that.)

<G>  In retrospect, i'll have to remember that we're not all engineering types here.

I'm still of the opinion that, as fun as all this thinking is, if you've got the tailstock fall to overcome a reduced case pressure, then moving the pelton to take optimise the head will probably return a better result for the effort. . . .  but there  might be complicating factors for positioning the wheel optimally, so what the heck.

-Dan

[Gog]

This sounds like an interesting idea to me for non optimal turbine placement or to keep the turbine out or a flood prone area

My late thoughts on maintaining an air bubble.

Assuming the tail pipe is supplying a good vacuume and is sized to deal with a greater flow than the head pipe.

A toilet ballcock could be installed "inverted" in the bottom of the sealed turbine box with its inlet upen to atmospheric  air pressure

If the water rises , at a predetermend point the ballcock will open and introduce air into the box and lower the effective head  for the head pipe untill the water drains  through the tail pipe and then close again when the level drops  , reintroducing full head to system.

It shouldnt break the tail pipes vacuume.

As this would cause surging in the turbine it would be for a battery charging set up only

Possible ?

Silly ?