6 replies to this topic

[mechanical-engr]

Hi,

We are proposing to install a temporary water pipeline from a water source (point-A) at + 120 m MSL to the raw water reservoir in our plant located at an elevation of + 167m MSL (Point- A)

The pipeline is routed on ground (no pipe racks/pedestal) through a rough terrain where the elevation goes up to + 192 m MSL at one point (point-C)and from there the elevation gradually comes down to +167 m at the inlet of raw water reservoir (destination point- B ).

The total length of pipe between water source and reservoir is around 4000 m.

I have calculated the pump head as follows

H = static head + friction loss

Where,
Static head = 192 (highest point) - 120 (water level at source)

= 72 m

Regarding friction loss to be considered for selecting pump head, I am confused between the following two options.
1) Should I consider the friction loss for entire pipe work (from Point-A to Point- B ) to arrive at pump head?

2) Or shall I consider friction loss only up to Point-C (highest point in pipe route) from where the water can flow to its destination by gravity?

In case-2, there is significant reduction in pump head thus involving less cost than with case-1.

Kindly suggest.

Edited by mechanical-engr, 05 August 2011 - 12:00 AM.

[breizh]

Yes the entire pipe ( subject to friction).
For the static head , applying Bernouilli between point C and A , it will be ZC-ZA
Breizh

[Geordie87]

Is the line reporting to the resvovoir through an open end? If so the downhill section may be running only partialy full and this changes the nature of the flow through this section.

[katmar]

You are correct that the static head that you must design for is from point A to point C, ie the full 72m.

With the information you have provided it is not possible to say what friction you should take into account. It depends very much on the length of the pipe from point C to point B. As the water flows from C to B the static head will be recovered. If this recovered head (25m) is greater than the friction loss from C to B then this static head will provide the friction head required and your pump will not have to provide it again. If the friction head is more than the recovered static head then the pump will have to provide the excess friction head, ie the quantity exceeding 25m.

But this is a bit theoretical and you should not design that way. I agree with Geordie87 that the section of pipe from C to B might not run full, and I would even say that it should not be designed to run full. If it runs full it is possible to draw a very deep vacuum at point C though the syphon mechanism and if you are using plastic pipe it could be sucked flat. Even steel pipe can be collapsed under these circumstances.

It would be better to provide an air vent at the highest point - this can either be an open riser with a 180 degree bend on top, or a vacuum release valve. This will ensure that the pipe will not collapse. With an air vent in place you must ensure that the static head available from the highest point (the 25m) will definitely exceed the friction loss and this may require a slightly larger pipe from point C to point B to get it to run partially full. The diameter of this section of pipe will depend on its length. Under these circumstances you do not need to provide for the friction from C to B in the pump head.

You have spoken of only one high point, but if there are other undulations between point C and point B it will complicate your design. If possible you should route the pipe from C to B to give a steady downslope all the way.

[S.AHMAD]

[
Hi there,

During normal operation (pipeline filled with liquid) what matter is the parameters at the source and at the destination (reservoir). So you need to consider the whole piping systems (pipe + fittings)including the pump suction piping otherwise you may undersize the piping system (may not be able to deliver the flow rate required. Unless the length of pipe from the highest point to the destination is very short or frictionless. The highest point static head i.e 72 m is only required during start-up when the line is empty (the pump head must be able to overcome this). Once the line has been filled up with water, the static head concern will be 167m - 120m = 47m and add the reservoir liquid level. During normal operation, the pressure at the highest point (197m) will be vacuum (normal case). My suggestion is that you design based on the whole system and counter-check whether the pump head is high enough to overcome the highest point during start-up.

The pump head = elevation differential between A and B+friction for the whole system.

Elevation at B shall include the highest possible water level. Elevation at A shall consider the lowest possible water level

Edited by S.AHMAD, 09 August 2011 - 10:56 PM.

[katmar]

During normal operation, the pressure at the highest point (197m) will be vacuum (normal case).

This situation should be avoided. The best result you can expect under such a scenario is that the flow rate will be severely restricted because of the vapor generated by the boiling water at the highest point. The worst you can expect is a collapsed pipe and a catastrophic failure. A vent or vacuum relief valve must be installed at the highest point to prevent this situation from developing.

The only time you can take the static head recovery from the high point down to the discharge as a credit towards the head that the pump must develop is when there is a back pressure at the eventual destination that will keep the pressure in the line at the highest point high enough to prevent vaporization. In a situation as described by the OP where the water discharges freely into a reservoir it is best to design the section of pipe from the high point down to the discharge to run partially full so that there is no chance of generating a vacuum. This can be difficult if there are local high points between the highest point and the discharge because the same criteria will apply to each of these high spots and some may also require vents or vacuum relief valves.

[S.AHMAD]

[/quote]
This situation should be avoided. r
[/quote]
I fully agree with Katmar, vacuum should be avoided (if possible). In order to avoid formation of vacuum at the highest point, the line should be sized at the desired flowrate such that the friction pressure drop between point C (highest) to point B (reservoir) is greater than the static head between C and B. If the friction pressure drop is lower, then vacuum will be formed continuously during normal operation. However, vacuum will occur when the pump discharge valve is inadvertently closed or partially closed (flowrate is zero or lower than the design).

In order to see the relationship clearer, apply Bernoulli’s equation between C and B and include the frictional loss in the equation.

We can protect the line from crumpling by:

1. Operating procedure - when stopping the pump ensure that the reservoir inlet valve is closed first prior to stopping the pump. Then close the pump discharge valve tightly and only then stop the pump. Any leakage will eventually result in the formation of vacuum. After starting the pump, ensure that the pump discharge valve is opened first and followed by the reservoir inlet valve.

2. Install vacuum breaker or design the piping to stand vacuum that is equal to the static head differential between C and B (I.E 167 - 192 = -25m). I believe in this case, installing vacuum breaker is the cheaper option but ensure that periodic maintenance of the breaker is carried out appropriately.

The above two are not selective - both must be made available.

Lastly, do not forget to install thermal relief valve (3/4" valve is normally sufficient) to protect the piping from overpressure caused by sun radiation. Overpressure may occur when both end valves are closed and the water is trapped in between.

Edited by S.AHMAD, 10 August 2011 - 04:38 AM.