Strugling with a siphon

[camillitech]

Hi Guys,

I'm looking for suggestions to stop air getting into my penstock or even help in getting it all out in the first place. I've done a few micro hydros here where I live. A Harris that runs from one of two lochs depending on water availability, either 93m of head and 600m of penstock or 120m of head and 800m all in 63mm MDPE pipe. I did that for a mate and it's been a great success with no issues other than a few bursts at the turbine end due to the pressure.

My own property runs off an ES&D Stream Engine from a small creek with 3 different penstocks of 50mm, 63mm and 90mm pipe 470m long with 40m head, each one supplying a different nozzle depending on water flow, again that works well with no issues.

However my next project has got me stumped as it's a little ambitious to say the least. The loch is over a mile away giving me 80m of head. The sensible thing would be to go straight down from the loch to the sea with 400m of 90mm pipe, stick my Powerspout on the end of it and run a mile of cable back to the inverter but at 900w that's expensive cable.

The silly option that I'm trying to go for is to run 800m of the 90mm MDPE pipe overland and use only 1000m of cable, mainly because the pipe is free but also because the turbine would be much more accessible for maintenance. The problem being that this route requires the pipe to siphon in a couple of places and I keep getting air locks.

The first siphon just over the lip of the loch I ran for two months just to see if it would work and all was peachy, the problem seems to be further along the pipe. I know that a normal bleed valve will not work on the negative pressure of the siphon but I've just ordered fittings and valves so I can pressurise the system to bleed it initially to see if that works. I was just wondering if you guys had any other suggestions before I revert back to plan A and the expensive cable. Seems a shame not to use all that free pipe, especially when its taken me weeks to lay it 

Cheers, Paul 

[TomW]

Paul;

I have read that over a certain drop with a siphon that it literally draws air bubbles out of the water making it very hard to avoid the air lock demon.

I am no expert but remember that from someplace in regards to something not related to hydro power installs.

Is your tunnel boring machine in the shop?

Sorry, couldn't resist.

Sounds like an awesome hydro resource

Tom

[taylorp035]

If I have my fluid dynamics right (I am doing these types of problems right now), the pressure should be higher than atmospheric at any point below the loch.  In order to get it started, you can have a valve just below the loch water level to stop the water initially.  Then fill that part up and then connect the intake to the loch.  Open the valve and it should start to suck the water down the drop.  If there are any "humps", ideally you would want the valve lower below the loch than the height of any of those humps.  But if there are large amounts of air in the pipe, it should compress and make starting it easier. 

Trying to suck the water/air to get it started from the bottom maybe be challenging, especially if their are any leaks.  Plus, you could only suck at an equivalent of atmospheric pressure, or 999*9.81*h = 101,000  --> height = 10.3 meters.

Keep us posted!

[wpowokal]

A couple of points.

The air will be liberated over time, therefore one solution if the pipe is reasonably accessible is to install a diaphragm hand pump at the high points with a good quality ball valve between pump and pipe, pump is typically available from boating places as hand bilge pumps.

The warmer the water the more readily the air is liberated, I assume (dangerous word that) your pipe is not buried.

The best solution is to dig the pipe in and avoid the high points, have you considered transferring the power at a higher voltage.

Allan

[damian]

   Tom W,

 You are right that the air is coming out of the water.  The reverse happens in a well pressure tank with no diaphragm, the air dissolves into the water under pressure.  At the high points in a siphon the negative pressure literally pulls dissolved air out of the water.  I had some friends many years ago who had a well up on a hill.  They were siphoning water from the well down to a point of use lower than the water level in the well and an air pocket would develop in the top of the siphon.  The air was coming out of the water itself, not only because of the pressure issue, but also because of the temperature issue (which someone mentioned here as well?).  The water was warmer at that point than it was in the well, at least in summer, which was not in our favor either.

 What I came up with for that was a chamber located above the pipe at the high point of the siphon.  A valve below the chamber separated the chamber from the top of the pipe and another valve at the top of the chamber connected the chamber to the atmosphere.  To operate: shut the valve to the pipe, open the top valve and pour the chamber full of water, shut the top valve and open the valve to the pipe.  The air in the pipe escapes into the chamber and the siphon is maintained.  Come back in a week (in that case) and refill the chamber with water.  Of course the bigger the chamber the less often you have to refill.  I'm sure you could automate it or find another way to skin the cat.  For example, if the negative pressure is not too high or the rate of accumulation of air too great, make the top valve a check valve, connect a small vacuum pump to the outlet side of the check valve and turn it on and off with a float in the chamber.  A small solar panel and a battery could power this, or you could power it with muscle power.  As far as I can see, you are onto it, it's just that, whether you are carrying water or pumping air out under negative pressure, I think you will be supplying a bit of energy to the system at the high points of the siphons to keep it running.

Hope this helps,

Damian

[RP]

If I have my fluid dynamics right (I am doing these types of problems right now), the pressure should be higher than atmospheric at any point below the loch. 

In that case a simple check valve attached to the top of the humps should allow any air to escape.  By fitting a small petcock type restriction valve you could let it "leak" a little bit continually.

If it turns out that the pressure is below atmospheric when flowing then periodically closing off the pipe at the bottom would raise the pressure and "burp" out any air through the check valves.

[thirteen]

is the pipe buried or above ground?

[camillitech]

Cheers guys for all your replies which have given me some heart to carry on, I have to confess I was getting a little despondent.

RP,s 'burp' has encouraged me to try going down the check valve path that I'd dismissed due to the negative pressure.

Damian's solution I'd also considered theoretically so it's good to know that it works.

Digging the pipe in Allan is out of the question as it's mainly rock though the lip over the edge of the loch may be 'doable' and would remove at least one of the siphons, perhaps the most troublesome? As for high voltage I'm already up to 380v DC 

Taylor, I seem to be able to start it OK with a 2" waterpump 'teed' into the line from the loch end but I'm not sure if the pump is introducing air that I can't get rid of or the air is coming out as it warms up in the black pipe under negative pressure.

Tom, I'm sure I'd read that somewhere myself which made me wonder if I was 'flogging a dead horse'

There's a map here http://lifeattheendoftheroad.wordpress.com/2011/09/21/more-listers-2/ that will give you some idea of the terrain.

The red route I'm pretty sure will work as it's all downhill apart from the initial hump, which I may be able to remove anyway by either digging the pipe in, raising the loch level or a combination of both. Though I did actually have a siphon running out of it for a couple of months with no issues.
The green route only worked whilst the pump was running so it was obviously more than 10.3m (despite the GPS saying otherwise)
The yellow and purple routes are others that I've taken to try and lower the siphon, as yet to no avail.
I think I'll skip the black route as it puts the turbine in an awkward spot and does not really save that much cable in the grand scale of things. Each square on the map is 1Km and the cable has to go to where it says 'ford' on the map so you can see why I'm keen to go this way, far less cable and an easy to service turbine.

Still you can't bend the laws of physics 

Thanks, Paul 

[wpowokal]

WOW

Er could you have a small wind turbine driving a vacuum pump at the first hump out of the Loch, that would maintain the siphon. Keep the intake as deep as possible to be in the least aerated water.

Having seen it I believe the only siphon that will cause you any problem is the one at the loch wall, gravity will take care of any other high points, eliminate that and you have a winner.

Standard cable can be run up to 750V, if you can step it down OK.

Allan 

[camillitech]

Hi Allan,

if you think that the problem lies in the first hump then perhaps I'll have a go at burying it. One thing I haven't mentioned which may have some bearing on the matter is this. To cut down on the expense of valves the first few meters of pipe is only 63mm where the priming pump is tee'd in http://lifeattheendoftheroad.wordpress.com/2011/08/20/fair-worn-out/ (near the bottom of the page) Same at the terminal end 63mm valve and pipe. The 63mm ball valves are only 33GBP whereas the 90mm ones are over 100GBP, though since I posted that I've put a gate valve on the terminal end before the ball valve.

Anyway, I've ordered some valves and saddle clamps http://www.pipestock.com/saddles-clamps/reinforced-single-outlet/ that will allow me to break into the pipe without cutting it so I'll give that a go first.

Just a thought chaps, you don't think my filter could be too small do you for 3lts/sec (45GPM) do ya http://lifeattheendoftheroad.wordpress.com/2011/09/03/filter-tipped/

Thanks for all your suggestions and input it's really made me determined to get the 'long route' working.

Cheers, Paul

[loadstone]

Hey camillitech, What a great view you have at your place....any land for sale nearby,

OK, first what is a penstock (sorry from Indiana living in Guatemala), since water can not be compressed (OK very very little) it transfers easily pressure applied to it or negative pressure as well. It cavitates or disassociates at 1 ATM (turns to gas) but your problem could be the accumulation of all the other humps also. Have you tried loosening the first joint over the hill to see if water is exiting. Then try free flowing the water into the down line to see if it exits at your house (make a call by radio to check) If it does then the problem is isolated. Equal the pressure on both ends (put the water into a 55 gal drum in the top of the barrel, just over the rise and connect the down line out the bottom) and it will flow. Discharge is 328 gpm (Hazen williams equation) from the measurements you gave, much more than the 45 that the pump can produce.

Jon

Doesn't look like the filter is the problem.

[wpowokal]

Hi Paul,
          If you can dig the top end of the pipe in such that it will always be below water level then no priming valve is necessary, gravity will do it for you, what would be handy is a rope tied to that strainer such that any initial fill of the pipe (with lower end valve shut) you can raise it enough to have a downhill slope on that top section. This will allow all air to be driven up to the top and out the suction end, then you will stop worrying about any up hill and down dale bits lower down.

If it remains necessary to retain the reduced pipes over a short distance then it is no big deal, the long run is where losses in pipe friction will occur.

You will already know but shut the lower end valve very slowly once the water is flowing.

That suction filter looks fine to me, no not too small, but the way you will tell is once in service if it gets clogged or was too small then it would get sucked in and horribly De formed. What is the bottom of that loch like for potential clogging material ie weed, silt  swimmers?

In your first post you said you ran a siphon for two months and you believed the subsequent loss of siphon was caused by lower areas, can you elaborate on this please. Provided there is a continual slope in the first section after the Loch wall at least until the elevation is several meters lower than Loch water level plus the number of meters it is lifting the water, then any slight rise should not stop the siphon. That said if the outlet was restricted and all air was not purged out of the line initially then said air will find it's way up the pipe and break your siphon.

From what I can see on your blog I believe you could establish and maintain a siphon, when priming with your trusty Honda how have you stopped all the priming water going out the filter end?

 
I don't know the elevation at the Loch but there is no way us mortals can siphon lift water 9M, the greater the lift the more critical all aspects become, 3M is more realistic. By your pics would I be right in saying you will have no effect on the Loch water level once you begin using the water?

If it was my project and I could not economically or practically dig that first section of pipe in I would......

1) Not tee in a priming valve at the high point, I would have it teed in near the filter such that it would be under water in service, say a meter or two away (reason for small distance will be evident when you finally place the filter in the water).
2) Pull the strainer out, connect trusty Honda and fill the line with the bottom closed off, have some wine and cheese on hand and wait for the line to fill.
3) Once full keep the strainer elevated, go home for the night have another bottle of red, return in the morning (in the intervening time any entrained air will have made its way out the top, this goes to my point in an above paragraph) Start trusty Honda and with it running lower the strainer into the water, go to the lower end and open the valve, return to the top and shut off the priming valve then stop trusty Honda and disconnect.
4) I would nevert fully close off the lower end, instead I would divert the water to dump, this will ensure that the siphon is not lost if there is any slight air ingress somewhere along the pipe.
5) If possible I would transmit the power as AC at full cable voltage utilizing transformers at each end. In my past life I had such a connection some 800M with 4mm sq cable which was quite happy transferring 3Kw with acceptable voltage drop.

Allan of the jungle

[taylorp035]

If photo #5 is your hill that you need to clear (initially), then i would use a windmill or two to pump the water up to the top of the hill (adding energy... good thing).  Then you can eliminate any negative pressure in the system, as long as nothing is higher than the top of that hill.

I am impressed by the scale of your hydro set up that you are trying to conquer.  Is there an estimated height of the top of the hill?

[scoraigwind]

Based on the good suggestions above I have concocted this scheme to operate the syphon.  Ignore the dotted line in the above sketch. 

Open valves A and C to prime with pump.  Other valves closed.  After priming, close A and open B. Use valve C to control flow.  Never close valve B except during priming process.  Close valve C during bleeding process for best results.

  The dead leg of pipe between valves D and E should slope up gently and be as long as possible to store water/air.  Tee and elbow arranged so that air rises easily into this leg. Valve E should not be more than 10 m above loch level.  Bleed air out as follows. Open E and fill dead leg of pipe with water.  Close E and open D.  After a while, Close D and open E and refill pipe with water.  Continue until satisfied there is no more air in pipe.   Leave valve D open.  E is normally closed except whilst bleeding air.

Any high spots on the way down the hill can be bled using a valve since they will be under positive pressure.  I sometimes drill a very small hole instead and seal it with a stainless steel screw.

[rossw]

Open valves A and C to prime with pump.  Other valves closed.  After priming, close A and open B. Use valve C to control flow.  Never close valve B except during priming process.  Close valve C during bleeding process for best results.

Strikes me that to prime it'd be quicker and easier to close B and C, open A, D and E.
Start the pump, fill until water comes out E. Water *should* have fully filled the pipe from C-D and A-C, any air should have (or should quite quickly) rise from the downhill pipe segments and escape out D-E.

When primed, Close A. Close D. Open B and C.

[camillitech]

Hey camillitech, What a great view you have at your place....any land for sale nearby,

OK, first what is a penstock (sorry from Indiana living in Guatemala), since water can not be compressed (OK very very little) it transfers easily pressure applied to it or negative pressure as well. It cavitates or disassociates at 1 ATM (turns to gas) but your problem could be the accumulation of all the other humps also. Have you tried loosening the first joint over the hill to see if water is exiting. Then try free flowing the water into the down line to see if it exits at your house (make a call by radio to check) If it does then the problem is isolated. Equal the pressure on both ends (put the water into a 55 gal drum in the top of the barrel, just over the rise and connect the down line out the bottom) and it will flow. Discharge is 328 gpm (Hazen williams equation) from the measurements you gave, much more than the 45 that the pump can produce.

Jon

Doesn't look like the filter is the problem.

Hi Jon,

The house and a bit of land will be going up for sale as soon as I've got the hydro sorted and built another 

The penstock is the supply pipe and whilst I have done what you suggest by splitting the pipe I've never left it open long enough to see if air builds up. The flow as you rightly point out is much more but I use 45GPM as that's what will be going through the turbine.

Thanks, Paul

[scoraigwind]

Strikes me that to prime it'd be quicker and easier to close B and C, open A, D and E.

Yeah maybe except it's a long way to go to close C (at the turbine) and I am not convinced that it would go, whereas if C is open it will go for sure.

[camillitech]

Hi Allan,

"In your first post you said you ran a siphon for two months and you believed the subsequent loss of siphon was caused by lower areas, can you elaborate on this please. Provided there is a continual slope in the first section after the Loch wall at least until the elevation is several meters lower than Loch water level plus the number of meters it is lifting the water, then any slight rise should not stop the siphon. That said if the outlet was restricted and all air was not purged out of the line initially then said air will find it's way up the pipe and break your siphon."

To asses the viability Allan I ran a siphon just over the lip of the loch for a couple of months to see how much I could extract without lowering the level. As it ran perfectly the whole time I thought that I was 'onto a winner'. However the pipe I used was 6m long lengths of PVC ducting cemented together and the only valve in the system was at the outlet end. I simply got the siphon going by opening the valve, submerging the whole pipe in the loch, then closing the valve whilst I dragged the pipe over the hill. I use the word simply as in simpler than trying to fill the pipe up with mouthfuls of water 

Obviously now I have all the valves, joints, different size pipe and extra hills to climb it's a little more complex than I'd envisaged 

"From what I can see on your blog I believe you could establish and maintain a siphon, when priming with your trusty Honda how have you stopped all the priming water going out the filter end?"

Valve B in Hugh's diagram

"If it was my project and I could not economically or practically dig that first section of pipe in I would......

1) Not tee in a priming valve at the high point, I would have it teed in near the filter such that it would be under water in service, say a meter or two away (reason for small distance will be evident when you finally place the filter in the water).
2) Pull the strainer out, connect trusty Honda and fill the line with the bottom closed off, have some wine and cheese on hand and wait for the line to fill.
3) Once full keep the strainer elevated, go home for the night have another bottle of red, return in the morning (in the intervening time any entrained air will have made its way out the top, this goes to my point in an above paragraph) Start trusty Honda and with it running lower the strainer into the water, go to the lower end and open the valve, return to the top and shut off the priming valve then stop trusty Honda and disconnect.
4) I would nevert fully close off the lower end, instead I would divert the water to dump, this will ensure that the siphon is not lost if there is any slight air ingress somewhere along the pipe.
5) If possible I would transmit the power as AC at full cable voltage utilizing transformers at each end. In my past life I had such a connection some 800M with 4mm sq cable which was quite happy transferring 3Kw with acceptable voltage drop."

All points noted, especially the digging in if possible and underwater valves/connections plus I'll try and increase the first few meters of 63mm  pipe to 90mm as I feel that this is creating more negative pressure.

However AC is out as the turbine is on its way from New Zealand as we speak 

Thanks, Paul

[wpowokal]

Paul, your in transit turbo/alternator produces AC that is rectified to DC, I do not fully understand your electrical requirements but suffice to say it can be done, I suspect the alternator (EX motor) is being used stock standard so it is 3 phase albeit most probably with a control board of some kind . Your existing cable if already in place may limit you but..................

I detailed how I would do it for the hydraulics and rest my case, you do realize I am at a disadvantage here, as I am down under and every time I look up, well you know what a Scotsman wears under his kilt.

Ross don't forget we are down under and water swirls the other way going out a hole.

Allan

[camillitech]

Paul, your in transit turbo/alternator produces AC that is rectified to DC, I do not fully understand your electrical requirements but suffice to say it can be done, I suspect the alternator (EX motor) is being used stock standard so it is 3 phase albeit most probably with a control board of some kind . Your existing cable if already in place may limit you but..................

I detailed how I would do it for the hydraulics and rest my case, you do realize I am at a disadvantage here, as I am down under and every time I look up, well you know what a Scotsman wears under his kilt.


Allan

To be honest Allan my original plan was to go for a Powerspout HE, the high voltage AC version for the reasons you suggest but they're reluctant to supply them overseas for some reason. I'm sure that I could have overcome this hurdle but my plan is to use the GE, grid enabled model and feed into my own 'off grid' set up via a Sunny boy. This will give me much more flexibility whilst I'm building my new barn and house as I can put the Sunny boy into the barn and run armoured mains cable the 200m to the current dwelling rather than take HV down to the battery bank then 230v  back up to the new site.

The new house being 200m closer to the new hydro than my present location, going HV may save me a few $$ in cable but it means that to get the benefit I'd have to shift all my batteries, inverter and generator now.

Cheers, Paul of the peat bog   

[scoraigwind]

The Powerspout GE400 is quite a nice package with a built-in voltage-limiting circuit and dump load, so that you are protected from the potentially rather extreme voltages and overspeeds you would get if something went wrong.  It also protects your inverter, and the Grid-tie inverter handles the MPPT side of things, automatically adjusting the turbine to its best operating point.

[thirteen]

Could you have a small crack somewhere piping system or even a possible bad joint? You said it worked for a couple of months then problems started. Just wondering but I have not read every ones replies so I might have missed things. Just an idea to toss or look at.

[camillitech]

It can't be far away now Hugh

Hi Thirteen,

the siphon that I had running for a couple of months was this one http://lifeattheendoftheroad.wordpress.com/2011/03/20/early-even-for-me/ (about half way down the page) it was only on a 60m long pipe just over the first lip of the loch and all was peachy. It started to go 'pear shaped' when I ran the full 800m 

Air leaks are a possibility and I'm going to do further tests shortly but I fear it may be dissolved air coming out solution under negative pressure 

Cheers, Paul

[XeonPony]

and all this time I just gave beer to the hydro guys to lose 2 pole transformers! transmiting at 14.4Kv allowes you to use ridiculuously small cable! mind you ya got to use poles now but in your case seems like good option.  Syphones will be a never ending head ach!

[camillitech]

and all this time I just gave beer to the hydro guys to lose 2 pole transformers! transmiting at 14.4Kv allowes you to use ridiculuously small cable! mind you ya got to use poles now but in your case seems like good option.  Syphones will be a never ending head ach!

Hi XP,

     "Syphones will be a never ending head ach!" you're not kidding, and the poles would have been an option 20 years ago when I was given 100 
Not only that but they were still in the ground, with cable on them and passed within 1Km of the none siphon site and virtually past the siphon site 

The telecom company laid a new 10 mile underground phone line to my house and in doing so got one of their tractors well and truly bogged. I removed it for them and took all the last 100 telegraph poles as payment.

In all honesty they probably got a good deal because it would have cost them more to remove them and take them away but I was happy as it gave me lots of strainer posts, uprights for buildings and some rather smelly firewood 

Sadly they're all gone, as is most of the wire which would have been perfect, using poles that were already in the ground would have been a good option but I think I'd have 'planning issues' putting some back 

I'm off work for a week now and made a little progress last week http://lifeattheendoftheroad.wordpress.com/2011/10/01/how-did-calum-do-it/ gonna move the pipe into the trench, re route it slightly and put a vacuum/pressure gauge on some of the high spots to see whats happening as well as working my way through some of the great suggestions you guys made.

Cheers, Paul

[roosaw]

Not to be johnny raincloud but is the height between the lock surface and the high spot in the pipe (transfer pipe not your "air catching" one) more than 32.2 ft?
Lots of folks are talking about "pulling air out of the water" when in fact what you are doing is lowering the pressure so much that the water boils at "room temp"  No amount of engineering is going to fix that.  You either have to pump the water over the hump or remove the hump.
Doing some research on water manometers may provide some insights

[whythehecknot]

Looks like a great sight for wind power as well.

[camillitech]

Not to be johnny raincloud but is the height between the lock surface and the high spot in the pipe (transfer pipe not your "air catching" one) more than 32.2 ft?
Lots of folks are talking about "pulling air out of the water" when in fact what you are doing is lowering the pressure so much that the water boils at "room temp"  No amount of engineering is going to fix that.  You either have to pump the water over the hump or remove the hump.
Doing some research on water manometers may provide some insights

Hi Roosaw,

sorted http://lifeattheendoftheroad.wordpress.com/2011/11/09/siphons-are-to-be-avoided/ I managed to get the height of the siphon down to around 8' and used a variation of Damians idea.



Been running now for six weeks or so with no issues other than a few burst hoses.

Cheers, Paul

[camillitech]

Looks like a great sight for wind power as well.

A great place indeed Whythehecknot,

I've a 2.5Kw Proven that's been up and running now for six years with no real problems other than a couple of broken springs and worn yaw rollers. Just had sustained storm force winds over Christmas between 60 and 75mph and she pumped out 2.4Kw for days on end 



Cheers, paul

[whythehecknot]

Im just amazed at all of the power resources you have at your disposal. Of course it took you some work and forsight but man!!! Im glad to see that you have discovered a solution to your dilema, and I wish you the best of fortunes. I just completed a water turbine project myself ( nowere nere the scale of yours ) and had some issues with air in the line. Didnt have to syphon but had some highs and lows that prevented water from flowing like it should (trapped air). But all it took was running the water at full flow and water velocity took care of it nicely.

  Take care!!!

[camillitech]

Cheers Whythehecknot, was pretty impressed with your turbine and pictures too http://fieldlines.com/board/index.php/topic,146182.msg998672.html#msg998672 specially the old one and the straw bale house.

Keep up the good work and have a great New Year, Paul