19 replies to this topic

[newcastle]

Hello All,


I have been worked in engineering company and now, I have involved in one project which one of the job is designing siphon pipe.
The siphon pipe is used for transferring sea water from sea to intake basin in shoreline. Material of siphon pipe is FRP and its flow rate is 3,900 m3/hr.


I have searched some literatures and found that minimum velocity criteria is around 2.4-2.5 m/s.
At the other hand, we have been informed that one plant has operated siphon pipe with maximum velocity criteria of 1 m/s (it considers as gravity flow).


Since both above cases are very different which lead to different siphon pipe size, can anyone share regarding the siphon pipe design?


Your kind help is highly appreciated.


Regards

[katmar]

The key to siphon design is the Froude number. Set the search target in the menu bar above to "Forums" and search for "Froude". You will see that there have been many discussions on this subject here. A similar search on Google will find even more information.

Just one pointer based on the information you have been given - you will get different velocities in different diameter pipes working as siphons. The Froude number is the criterion, not the velocity.

[S.AHMAD]

Dear newcastle

You need to match the head available and the frictional pressure drop in order to obtain desired flowrate of 3,900 m3/h. Pls take note that, as the time progress and water level in the storage compartment (pond?)rises, flowrate reduces since head differential reduces.

Line frictional pressure drop is proportional to square of velocity. Higher velocity means higher pressure drop and you may not be able to get the desired flowrate if the differential head is limited.

Head differential is the difference between sea surface elevation and you destination pond elevation.

Edited by S.AHMAD, 15 August 2011 - 04:56 AM.

[newcastle]

Dear katmar and S.AHMAD,

Thank you for your quick reply. I have searched some forums and articles regarding Froude number
and it seems that Froude has related to gravity line sizing.

In my understanding, siphon system that we would like to design is not a gravity line, since the driving force of flow is level difference
(between sea and basin) and water flow is "against" gravity. Please CMIIW.

The system is also provided with vacuum pump to suck air out of the system during start-up.
Once the water flows, it will be continuous (depends on head difference as per S.AHMAD statement) unless there will be siphon break.

Herewith I attach the schematic diagram of siphon system.

Regards

Attached Files

•  Dwg-Siphon.pdf   23.36KB   480 downloads

[katmar]

These two aspects (i.e. the Froude number consideration and the friction analysis) are two sides of the same coin. Of course you must match the friction loss against the head available. In any pipe sizing application where the pipe runs full you have to consider this. The difference in head is due to the difference in levels between the sea and the basin, and this is what constitutes gravity flow - when liquid flows from a source at a higher level to a destination at a lower level without the action of an external driver such as a pump.

The way you have described removing the air with a vacuum pump is fine - it is often done that way. It will work as long as both ends of the line are always submerged, or you have valves at the end of the unsubmerged line that can prevent air being sucked in while the line is being evacuated. This is called "initiating the siphon".

The relevance of the Froude number is that it enables you to predict whether any air in the system will be flushed out. If the air was not being flushed out you could have two problematic scenarios. The first is that you would have to use your vacuum pump to remove every last bit of air in order to get the full capacity of the pipe, otherwise you will have a permanent bubble sitting in the horizontal section restricting the flow of the water. You have shown your "horizontal" section as slightly sloped - this is probably a good thing to do, but remember to connect your vacuum pump to the highest point or you will always leave a bubble in the pipe. The second scenario is that in operation any air that enters the system will be trapped. Air can be released from the water when it gets to the top point because the pressure there is below atmospheric. Also because the pressure is below atmopheric it is possible to leak air into the pipe at flanges and fittings. If you ensure the correct Froude number any air that enters the line will simply be flushed out. If not, the siphon will eventually break or you will have to keep your vacuum pump running.

The Froude number requirement is usually not difficult to meet. In your case with 3900 m³/h flowing in a 32" line you will have a velocity of 2 m/s and the Froude number will be about 0.7. This is certainly high enough to move any bubbles along the horizontal section, but it would be nice to have it a little bit higher in the final vertical drop into the basin. If you have a well constructed line with minimal leaks the 32" line will probably be OK and will get rid of small amounts of air even in the vertical section.

You will need a level difference of about 900 mm between the sea and the basin. If it is more than this your flow will be higher than 3900 m³/h and then you will definitely flush out any air.

[S.AHMAD]

Dear newcastle

You can use following equation:

Delta-H = 2f(L/D).V^2/g

Where
f = fanning friction factor (dimensionless)
Delta-H = elevation differential (m)
L = equivalent length - pipe + fittings (m)
D = diameter (m)
V= velocity (m/s)

Alternatively you can express V in terms of total flow and diameter since V= Q/a.

In the equation, there are 2 unknowns f and D. Assume D and iterate.
You can accomplish the iteration easily using EXCEL spreadsheet GOALSEEK function or SOLVER.

Edited by S.AHMAD, 16 August 2011 - 02:01 AM.

[newcastle]

katmar, thank you for the explanation.

You have mentioned that Froude number will be 0.7 but you also said that it seems fine.


I have read articles which said that Froude number should be equal or less than 0.3 to ensure self-venting flow.
If it is the case, then siphon pipe size will 48" (huge impact in terms of either cost of material or construction).

Actually, what will be happen if Froude is bigger (say 0.7)?
Since the inlet and outlet pipe is submerged, can we assume that air will not entering the pipe?
Or pipe leaks should also be considered?

My client has referenced plant operating siphon pipe at less flow (1,700 m3/h) but unfortunately the siphon pipe is 36".
That's why they have hesitation since it is a common sense that bigger flow should lead to bigger pipe.

[S.AHMAD]

My client has referenced plant operating siphon pipe at less flow (1,700 m3/h) but unfortunately the siphon pipe is 36".
That's why they have hesitation since it is a common sense that bigger flow should lead to bigger pipe.

Dear newcastle
I have done some calculation using spreadsheet. For a chosen pipe diameter, the flowrate accomplish depends on the head differential availability(between sea level and basin level).
Expected Flowrate, m3/h
Head,m 32" 36" 48", pipe diameter respectively
0.05 1,226 1,645 3,377
0.10 1,733 2,327 4,776
0.15 2,123 2,849 5,849
0.20 2,451 3,290 6,754
0.25 2,740 3,679 7,552
0.50 3,875 5,202 10,680

As you can see from the above table, for 32" pipe you need at least 0.5 m head differential in order to obtain flowrate of 3900 m3/h. Your client reference of 36" for flowrate of 1,700 m3/h, the head differential is 0.05m. Therefore, if you are using the same basis head as the reference, then you need at least 48" diameter.

I hope the above helps.

Edited by S.AHMAD, 17 August 2011 - 10:57 PM.

[breizh]

Hi ,

An additional resource :

http://www.fao.org/d...44e09.htm#3.7.1 types of siphons

Hope this helps

Breizh

[katmar]

You definitely do NOT want self venting flow. That would guarantee that any air in the pipe could never be flushed down the final vertical pipe. In a vertical pipe, as the Froude number increases above 0.3 air starts to be drawn down with the liquid. It is difficult to pinpoint a value where all the air is drawn down the pipe because it is a time and quantity trade off. The more air you have the longer it will take to flush it out. A typical design range for horizontal pipe would be 0.5 to 0.7 so you have no problem there. For vertical design I have seen recommended values in the range 0.7 to 1.0. That is why I said that if you have a well constructed line it should be OK, but maybe the final vertical downleg could be smaller if you have spare head available. In both the horizontal and vertical cases if you have a flow above the recommended range it is not a bad thing in terms of air removal - it just means higher pressure drop.

The table of flowrates vs head presented by S.AHMAD seems to be based on 150 m of straight pipe. You should include allowances for resistances like an inlet screen (you dont want fish or plastic bags coming through), bends and valves, and the exit loss. Another very important factor in sea water piping is the possibility of encrustation with barnacles etc. This reduces the pipe diameter and severely increases the pipe roughness. That may be why your reference pipeline seems overly big.

Depending on how your basin is constructed, the pipeline can be a very elegant self controlling design. You are wanting a flow of 3900 m³/h, so I guess this is the flow that will be pumped into your plant from the basin. If the flow into the basin is higher than the outflow of 3900 m³/h then the level in the basin will rise and this will lower the differential head available to the siphon, thus automatically decreasing the flow in the siphon and the level will stop rising. The opposite occurs if you pump more than 3900 m³/h out of the basin. So if you design it with some flexibility in the operating level the basin can be nicely self controlling.

If you use a larger diameter siphon pipe it will be more expensive, but it will not flood the basin - the basin level will simply rise until the in and out flows balance (provided of course that it has been designed for this range of levels). The Froude number will be lower, but if your reference plant is working well then it means that my concerns over air ingress are probably too conservative.

[newcastle]

breizh:

Thanks for the reference. Actually, I have done my previous calculation based on this reference,
where I took the recommended minimum velocity around 2.4 - 2.6 m/s

S.AHMAD:

I have reference using formula from Bos (1989), based on article "Discharge measurement structures":
http://www.vl-irriga...419&type=5#2140

Q = PI/4*D^2 * ((2*g*DELTA_H)/(1.9 + f(L/D)))^0.5

It seems similar, but later on I find that this formula can be applied for large diameter pipe with short length.
Do your formula can be applied for all cases?

katmar:

Thanks again for your explanation which brings me much better understanding.
After all, what we concern is air entrainment in the siphon pipe which can lead to pulsation or vibration.

I have read your post in other thread: http://www.cheresour...drop-line-vent/
which said that:

"If the vent is simply closed off and no other changes are made there will still be vibration because the water will draw a vacuum and there will be boiling (vaporization) and cavitation (re-condensation) as the water travels down the leg."

Is it also possible in my case? Or if it is not, can I conclude that:
1. Air entrainment can be avoided by using Froude criteria of 0.5 - 0.7 in horizontal section and 0.7 - 1.0 in vertical downleg section
2. There is no criteria of maximum velocity (which seems < 1 m/s for gravity line)
3. My client's referenced siphon pipe is bigger because they might be consider encrustation but not velocity

Herewith I attach my calculation for reference.

Many thanks and Regards

Attached Files

•  Siphon Design.xls   183KB   438 downloads

[katmar]

The water in the siphon will boil if the top of the pipe at the highest section is more than about 8 m above the sea level. You have not given this dimension in your sketch. 8 m gives you some safety margin over the theoretical static head differential so that you can accommodate the friction head loss as well.

Note that you do not want to avoid air entrainment. On the contrary, you want to ensure air entrainment so that if any air finds its way into the siphon it will be entrained into the water and flushed out.

[S.AHMAD]

Dear Newcastle

The equation that you have is the same formula except that yours is expressed in Q instead of V. The factor 1.9 is for the entrance and exist loss and probably it includes pipe fittings as well.

However, please take note that the friction factor in your equation is Darcy's which is 4 times fanning friction factor.

In your case vacuum will be formed at the top section (whether you like it or not or whether you believe it or not). No vacuum means no flow of sea water. It is this vacuum that sucks the sea water up to the top section then down into basin under gravity. This is common for siphoning system and you do not have to worry about it as long as the absolute pressure (vacuum + 1 atm) is greater than the vapor pressure of sea water. Under this situation, there is no boiling and hence no vapor will be formed. The vapor pressure depends on the sea water temperature and the vacuum depends on the head differential top section and the basin level. The higher the differential head, the larger will be the vacuum. If you know the elevation of the top section, you can calculate the vacuum (pressure) using Bernoulli equation that includes frictional pressure drop terms.

P1 + Z1 = P2 + Z2 + Delta-P (friction)

Where 1 refers to top section and 2 refers to basin level. P2 is atmospheric pressure (1.013 bar a)
I have not looked at your calculation since the equation that you have is correct and similar to mine. As long as your input data is correct, it will give you the correct answer.

Good luck

Edited by S.AHMAD, 21 August 2011 - 11:30 PM.

[chaubey]

Sir
I wish to share the case of one of the sea water intake installation in India. Apparently the water requirement for one of the power plant is being met through the sea water intake system by putting caisson in shore area which is 2Km from Intake structure to landfall point and further 3 km from landfall point to plant . I have been given to understand that the sea water from intake point (at offshore) to basin at plant area is by gravity . As I understand from this thread , the sea water intake pipe generally works on siphon basis , now my quiry pertaining to subject matter is below

1) Flow parameter :- 10000M3/hr , pipe dia :- 1.4mtr ,length 6km . Now I don’t see any vacuum pump in this installation (please refer the attached drawing also) , then in such case how does the plant is ensuring the siphon flow of water ?.
2) Secondly for a given process parameter , the head requirement is apprx 7 MWC , then in such case do you think it will be possible to match the dirrerential head by maintaining the difference in levels of sea and basin. ?

[katmar]

1) Using a vacuum pump is not the only way to prime a siphon. You did not attach the drawing so it is impossible to comment.
2) 7 MWC is not enough if the pipe is 6 km long. The siphon probably works over the 2 km from the intake to the caisson, and then there will be pumps to drive the water from the caisson to the plant. In sea water lines it is usual to allow a considerable pipe roughness factor because of encrustation inside the pipe. Although the roughness of steel pipe is usually taken as 0,05 mm you will probably find that in this case a roughness of 2 to 5 mm will have been used, depending on the company's maintenance and cleaning policy.
3) Rather than tacking a new question onto an old thread like this, you should rather start a new thread. You can still quote or refer to the old thread, but it is better to keep each thread intact.

[chaubey]

PLEASE IGNORE PREVIOS MAIL WITHOUT ATTACHMENT

Dear sir

I esteem with the great kindness for your response
I wish to re-iterate on what has been talk

1) Firstly , iwas wrong in exuding the distance . The distance from sea water intake to pump landfall point is 1.7 Km and further 400 mtr from land fall point to pump house.
2) I have attached the drawing showing the layout from intake to pump house . Where intake is 7 mtr below chart datum . Now having seen the drawing it is concluded that there is no vacuum equipment to create siphon . In such case , how the plant is ensuring the siphon ?

3) Sir what is difference between caisson intake structure and velocity cap intake

thanks in anticipation

Attached Files

•  Sea water intake pump house 1-Model (1).pdf   103.93KB   196 downloads

[breizh]

You may consider an ejector to prime the line !
waiting for feedback from others.
breizh

[S.AHMAD]

Chaubey
1. You are "piggybacking".. As mentioned by Katmar, it is a good idea to open a new thread.
2. It seems that your syatem is not using syphon for the seawater intake., therefore vacuum pump is not required.
3. With or without syphon, the flowrate of seawater to the pond is based on the same princinple namely the head differential between sea level and pond level and the line lenght, size, fittings and configuration.
4. For syphom to work, it requires vacuum pump initially (when the line is empty). Once the line filled up with seawater, vacum pmp is no longer required.

[katmar]

1. In my earlier post I took the length of the siphon line to be 2 km and now it is 2.1 km. Given the accuracy of the rest of the data those two lengths are the same.

2. Provided there is a non return valve (foot valve) at the sea inlet you can fill the line by pumping water into the highest point. Vacuum is not the only way to initiate a siphon.

3. I don't know.

[chaubey]

Dear sirs

I happen to review the process design data of one of the installed river water intake project and below is the parameter . The river water intake system designed to pump the raw water to thermal power plant

1) Flow of river water :- 300m3/hr
2) Pipeline diameter:- 400mm
3) Pipeline length:- 19km
4) Velocity inside the pipe:- 0.6m/s

1) No my observation to the parameter is , velocity inside the pipe seems to be very less as there can be potentially concern of algae deposition . Please correct me with my understanding

2) Secondly in case of sea water / river water intake pipe line , where distance of transport is high , what shall be the minimum and maximum velocity recommended