Coupling a turbine with a pump 5

[nilsadams]

Hello,

I would like to combine a turbine with a pump in order to lift water up to the pump position. I am considering two options:

Either (1) using a turbine pump
Or (2) using a turbine and then connect it to a pump

From financial and technical point of view, which one is a better option ?

I want to lift water up to the pump (turbine) position, which is 3 meters (bellow the pump position) using a discharge of 2 m3/s. Just beside to the pump position, there is a cannal where the pumped water needs to be discharged. The idea of using a turbine is because I have a water fall of about 6 meters to the pump (turbine) position with a flow rate of discharge of 0.9 m3/s.

Thanking in advance,
Nils

 

[nilsadams]

BRIS,

Thanks.

"Why do you not simply pipe from A to C to E and from B to D to E and place your turbines at E ?. "

Well, that could be a good option if I had no any constraint.

Do you mean two penstocks ?

The problem is that, just immediately after D, there is an electromecanical equipment designed to control and record the flow from both ACD and BD. The two discharges must combine together at the point of the equipment. The point where there is this equipment is on atm pressure. Afterwards, the pipe is under pressure upto the point E (River). Again, the area around E is not the property of my client. It means, it needs another legal procedure if we take the option to install the powerhouse at E. The other risk to the turbine is, point is closer to a big river with significant flooding in case of a 20 to 30 years return period. Thus we may need to invest a good structure for the power house.

There is a sort of siphon between A and C. As the rest, the way up to A as well as that of B is through Chanel (open).

That would create a lot of disorder if we remove the flow control equipment beside D. Again, on the surface between C and D there is a private road (for cars).

But let me assume that it may be possible to displace those equipment and can get permission for installation at E. Now my question : is it possible to use a sort of forked penstock ? I mean, just thinking to join a pipe AD and BD at point D and then use a single pipe EF as it is. As I said earlier, point A and Point B are not at the same level. How can you view the impact of using such penstock on the overall head at E.

Am working at a complicated cite :-( but it is also interesting to encounter such challenges

 

[waross]

As I understand your situation you want to move water from F to the river by the most economical means.
Siphons should work well. I would suggest multiple smaller siphons. You will need flow control to throttle or disable some siphons at periods of low flow. You will need a small pump and valves to start the siphons. An alternate is to use large pumps to start the siphon action and then let the pumps spin free in the siphon induced flow. This has the advantage that you can increase flow by starting the pumps during times of exceptionally high flow.
Basically design a conventional pumping scheme while keeping in mind that the layout should be conducive to siphon flow when the pumps are stopped.
respectfully

 

[nilsadams]

You mean to make hydraulic connection through FDE using siphon from F to D ?

 

[Warpspeed]

I have been thinking a bit more about siphons. The maximum theoretical suction lift of any system is limited to 10.3 metres, and the practical maximum is probably a bit less than that.

That is not likely be a problem here, but it is something to keep in mind.

If you do decide to use a siphon between F and E, no additional pump will be required to initially charge the siphon pipework. A high pressure water source is already available directly from either A or B.

Another advantage of a siphon between F and E is that there would be no machinery located at E except for a motorised gate valve, which could be remotely operated from F

 

[waross]

Yes. Is this feasible? The lower the discharge elevation, the better the siphons will work.
Of course there will be a lot of compromises before you decide on a scheme but compromise is often a big part of design.
You will need some controls but the power used in the controls and occasional pumping to start the siphons will be a small part of the power required to pump everything.
Are you trying to get run-off water to the river. Is the flow steady or dependant on rainfall or some other factor?
Tell us a little about the overall picture, please.
respectfully

 

[nilsadams]

I thank you guys !

The variation of the flow is not much. The flow what I described above are almost represents the 80 to 90 % flow time in my FDC. It means I have relatively constant flow sources.

As to using a siphon, the discharge elevation (D) is higher than the saction elevation (F). Most siphons I see go up from the suction elevation but the descharge elevation goes down below the discharge elevation. I would be glad if you make me clear a siphon with higher discharge elevation than the saction elevation can be operational. If I connect the discharge end to the pipe leading E, perhaps the link FDE can give me a discharge elevation lower than the saction elevation.

Regards,
Nils

 

[BRIS]

Sorry I have not read through all the replies so I may be stating something that has already been considered. I w agree that given your constraints a siphon could provide the best solution. Place your turbine at F and provide a closed pressure conduit from the outlet of the turbine from F to D to E. The siphon will prime. You may need to provide an air valve with a non return flap at D (a valve that lets air out but not in).

Clearly you will have to abandon/remodel the electo mecahnical arrangement at D. You can provide a flow meter and regulated valve at the turbine to control flows.

 

[nilsadams]

Warpspeed, waross , BRIS,

Thanks. How do you dimension the siphon in terms of height and performance. Do you know commercially available siphons of the kind BRIS mentioned ?

 

[waross]

The discharge end of a siphon is below the inlet end. The water flows from the end with the higher elevation to the lower elevation.
You must arrange your discharge so that the discharge is lower.
A basic siphon is just a pipe from point "A" to point "B" at a lower elevation. You can use the difference in head between the inlet and the discharge as the pressure drop and work backwards from the pressure drop tables to determine the flow rate with different sizes of pipe.
Starting the siphon. The pipe must be completely full of fluid.
Method #1 is to close the discharge end with a valve and fill the pipe with a small pump. The inlet must be closed. A check valve may work. The air has to be vented at the high point. When the pipe is full of fluid and the air has been removed, the valve at the discharge is opened and the flow starts.
Method #2 Is to use a pump with a great enough volume to flush the air out of the pipe ahead of the fluid. Once the flow of fluid has purged the air the pump may be stopped and the flow will continue.
yours

 

[Warpspeed]

My understanding of sizing a siphon is that the difference in elevation from inlet to outlet is the driving pressure. Flow stabilises at a point where total pressure drop down the length of pipe equals the driving pressure.

Possibly the most trouble free way to do this would be to dig a pit or a well at F down to at least the same depth as the river bed at E.

Build your siphon so that both ends are completely submerged down to the same level. In other words have no deliberate built in fall. Prime your siphon, and open both ends. The water in the well will then exactly equal the level in the river. Flow could be in either direction, but the water levels will always end up exactly equal.

As turbine discharge water flows into the well, the level will rise, and the siphon will begin to flow. Flow rate will be proportional to the difference in water level at each end.

That difference in level could be kept fairly small by having a large enough siphon pipe flow area. In fact the pipe could never be made too large.

Just design the siphon pipe so that it might flow for example 3 cubic metres per second at a one metre pressure head, for whatever total pipe length is required. Or choose any figures you like.

Doing it that way requires no flow control on the siphon. You can shut down your turbine and the siphon will just stop flowing too. Or if turbine flow increases beyond the design flow, the well just runs a little higher.

I think I would run more than one siphon pipe for reliability and ease of starting. I might also build a well at both E and F to ensure neither end of the siphon could ever become exposed and admit air, even if the river runs very low. There should easily be enough flow through both wells to constantly sweep them clean, and keep them completely free of mud and silt.

These wells do not need to be large, just deep, with the pipe ends running well below the minimum possible river water level. The only way the siphon can ever fail is if air enters the system somehow. A couple of suitably deep wells would certainly make that much more unlikely.

 

[waross]

I like that Warpspeed. You can extract the air at the high point with a small water sealed rotary vacuum pump and the siphon will start by itself. That way you do not need any valves to start the siphon. Simple cheap and reliable.
I hope nilsadams can use the concept.
Respectfully

 

[Warpspeed]

I had not thought of that, but you are quite right. The siphon pipe could remain completely open at both ends, and all the air extracted from the high point. It might be a fairly lengthy process to fill the entire siphon starting from a completely dry pipe. But it should only need doing once. If there were multiple siphon pipes it may still be a fairly economical way to do it, as only one vacuum pump would be required.

If both the siphon pipe ends can be sealed, the whole thing could then just be filled with water, and fairly quickly, which may be faster and cheaper.

I suppose it depends on the relative cost of a suitable vacuum pump, versus multiple large gate valves.

 

[BRIS]

You don't need to consider it as a siphon if you connect a pressure pipe to the delivery side of your turbine and pipe it to point E. You then have a closed system from pint A to E with your turbine somewhere along it. The head across the system is elevation A-E and the head across your turbine is elevation A-E - friction and other losses. You will need air relief valves on the crest(s) and these need to let air out not in. (not vacuum breakers). The size of the siphon pipe is determined from economics, smaller the diameter greater the head loss and lower the head across the turbine. I would suggest size for a velocity of o.75 to 1.0 m/sec.

You need to consider NPSH and lower the turbine to achieve adequate positive pressure. (this may be higher or lower than point F depending on losses)

If you adopt the open well solution at F as suggested by others a simple method of priming your siphon would be to provide a bypass around your turbine which tees into your siphon pipe. You will need a foot valve on the bottom of your siphon pipe in the well at point F. To start run the flow through the by pass to flush out the air. Close the by pass and the siphon will start to suck from your wet well at F. Open your turbine and the system is operating. (you will need the by pass in any case for maintenance).

 

[nilsadams]

I thank and appreciate your devotion here to figure out and suggest options to this problem. Using siphon is taking the momentum. Just to give you one more figure, the aerial distance between A to E is about 100 m. Using a siphon from A to E (or F to E) will not be convenient due to some private property structures that will not allow to pass the siphon up to the river.

If am going to use s siphon (s), it would rather be from F to D. The problem that I couldn't figure out yet is the elevation of the siphon outlet at D is higher than the Elevation at the turbine (F). From your comments, the siphon requires driving pressure to be operational. But from F to D, the pressure head would be negative. If I were to assume to connect the siphon from F to E, it would be possible to create pressure head or maintain similar water level in two wells as warpspeed suggested. But from F to E, I cannot go on the surface or make some earth works to pass the siphon. That option may then take me back to the use of a tunnel.

Now, from your comments,am trying to figure out how to use a siphon from F to D using a valve to remove air at the siphon crest Or if I can use a vacum pump ? I would be glad if you put your further comments on this. All your comments have been very helpful to figure out a solution to such a real challenge.

Regards,
Nils

 

[waross]

A siphon transfers fluid from the high point to the low point. If "D" is higher than "F" the siphon will work but unfortunately in the wrong direction.
I understand that you have a tunnel now between "D" and "E".
Is it of sufficient size that you can install a pipe inside it to provide the needed fall for "F" "D" "E"?
What is the source of the water? If "F" is process waste water, is there anything that can be done to increase the elevation at which it is collected?
If you do go with a turbine at "D" there are some types that do not have to be installed at the absolute lowest part of the flow. If the discharge from a suitable turbine at "D" was closed and able to withstand negative pressure the efficiency would benefit from the additional 4 meters head from the drop to the river.
It looks like you can get a combined drop of 7 meters.
You mention measurement and control at point "D". Measurement is cheap and easy. Control is difficult to understand. It would seem to be very difficult to control the flow of a waterfall from the bottom.
Suggestion #1 Use a siphon with a siphon pipe installed inside the existing tunnel from "D" to "E".
Suggestion #2 Install a turbine at "D" and close the discharge from "D" to "E" so that it will withstand negative pressure. Instal a pump to pump from "F" to "C". Add this flow to the total flow. The pump can be direct coupled to the turbine or belt driven from the turbine.

Suggestion #3 Close the piping from "C" to "E". Install a venturi at "D" to suck water from "F".
If this is physically feasible it will be passive, relatively cheap, and automatic. As long as there was flow from "C" to "E" the venturi will be drawing from "F". It should handle a mixed air and water flow as well as a solid water flow.
respectfully

 

[ozmosis]

Not sure if this would be of benefit to your needs but the "Blake Hyrdam" has been in operation for over 150yrs now. Simple, effective and no power required:

http://www.allspeeds.co.uk/blake_hydram_performance.asp

 

[nilsadams]

waross:

Yes, I am using process water with fairly constant discharge. The current discharge line is through ACDE. There is no flow down to F. I just wanted to install a turbine at F to increase pressure head from A to F if I find economical option to lift water back to D. The reason to go back to D is simply to use the existing discharge pipe from D to E. In your 3rd suggestion, the venturi stuff can also be interesting. I have to look at the capacity and pricing.


sed2developer:

"Hyrdam", is it like a ram pump ? Do you have experiences in it ? I will look at in detail. For me it could be interesting if the output discharge is almost the same as the input discharge in order to either (1) generate higher power by directly connecting the penstock from the lifted water tank or (2) lift equivalent dicharge from F to D so that there will not be storage at F do discharge differences by using "hyrdam".

With regards,
Nils

 

[nilsadams]

sed2developer:

I checked the hydram site. looking at the performance chart, I don't think it has a capacity that match my flow-in and flow-out requirement. The daily discharge I am considering in this problem is much higher than the max values listed in the chart.

 

[waross]

Hello nilsadams;
I think that I may have misunderstood something. Is there a source of water at "F"?
If the only water sources are "A" and "B" and you have an existing tunnel or pipe from "D" to "E", things change a little. Then run the flow "A" "C" "D" through a turbine at "D" and enclose the discharge so that the the drop from "D" to "E" adds to the total head. You may have to add a small water sealed vacuum pump to keep the discharge free of air and the final discharge should be below water level if possible. If you can divert the water from "B" to the inlet at "C" you will have a total of 2.9 cubic meters per second.
The head from "C" to "D" to "E" will be 3M + 4M = 7 M.
I make that to be about 200 kw at 100% efficiency.
Dropping the water to "F" to get more head and then returning it will take more energy than you will gain.
The best you will be able to do is use a turbine that can use the head of the discharge as well as the head of the inlet and close the pipe from the turbine to the river.
at the end you will get 200 kw less efficiency losses. In the real world, I would expect anything ovewr about 125 kw to be a bonus and wouldn't be surprised to see 100 kw or less.
respectfully

 

[nilsadams]

waross,

thanks a lot. Yes you got the point. I don't have water source at F. The use of vacuum pump at D and pressure pipe from DE was not in my mind.

If I understood well your suggestion, Everything from A to E (if I pass through ACDE) should be a closed system. Thus, I will have my penstock up to D (turbine + vacuum pump ?) + pressure pipe upto E, right ? How do you couple the vacuum pump with the turbine at D or where do you put it ? Or it can be anywhere between D and E ? If you can elaborate me what you said "You may have to add a small water sealed vacuum pump to keep the discharge free of air and the final discharge should be below water level if possible. "
It sounds good to me to go ahead with this option as the others would be complicated and costly.

Regards,
Nils