Pressure Drop Through a Water Pipe with Big Multiple Elevation Changes

pictures are worth 1000 words, and a few more words would not hurt also in this case.

I suspect some measurements are more than a little bit off, or the conversion from one pressure unit to another, perhaps?

Also, did you account for the depth of sumps at the point where water is being removed? If pumping and not free drainage, did you have accurate, correct pump curves, corrected for NPSH?

Where are you taking the pressure readings?

Can you draw a diagram?

It would also help to identify if this is a static (zero flow) pressure measurement or when fluid is flowing.

I suspect that somehow you may be getting sizable vertical air bubbles in the piping, each of which adds or subtracts from overall pressure drop.

One of the things to look for is water coming out of the pipe where it isn't expected.

Another is to look for tree roots in the pipes where they would not be expected.

There aren't any equations for those things.

There's no equation for a bullshine on an as-built drawing, nor a partly-closed valve.

Keep digging. You'll find it!

Here is the skech of the system, we had vented the high point a few of times, and the pressure drop is based on the pump discharge pressure gauge which had been calibrated and it is aligned with pump curve. The flow meter had been calibrated.

When I did the calculation, I considered all the elbows, also the pressure drop calculation is baed on the pipe ID, Hazen-Williams equation was used for the calculation, and C is 100 in the calculation. The calculated pressure drop is 1.25 bar, while the real pressure drop is 3.2 bar (The pressure drop from flow meter had been excluded)

Something is causing the pressure loss and it is not air.

Is this a new installation?

Are your calculations wrong?

Does it really matter?

Are your pumps able to cope with the demand?

Does it really matter?

Why?

As high point is 14.4m above grade, while atmospheric pressure ~ 10m, there is a free fall after the high point, due to the air in/out vent. So the static lift is 1.4 bar ignoring level in sump 1 below grade (which you don't give). So it must be greater than 1.4 bar. You don't give horizontal distances, but perhaps if you work out friction loss to that point and add 1.4 bar you might get nearer to the 3.2 bar.

BTW I much prefer Darcy-Weisbach/Moody formula for headloss calcs.

Thanks, this is the reply I expected, and I expect some studies/back ground more in details on the topics

I am assuming your starting pressure is at the output of the pump, 3 meters above grade. Where and how are you measuring the outlet pressure? I would assume it is the head pressure calculated by the depth in sump B of the outlet.

How do you measure the elevations of the start and end elevations? (Is "grade" elevation the same at both ends?)

Pressure difference has two components, elevation and fluid drag. You can separate these components by measuring with the pump on and off, and possibly isolate where your discrepancy lies.

Thanks, I had considered the elevation difference and flow meter. Actually the pressure 1.25 bar I menioned is only for friction loss. And the 3.2 bar is the total pressure drop minus elevation difference and pressure frop from flow meter

This becomes more ridiculous and laughable with each and every one of your responses.

<Unsubscribe>

Well then, what actually is the problem, seems it is about right.

If the high point is 14+ metres above grade, then the pump is effectively lifting to a virtual waterfall at that point. What is happening downstream of that point is that the fall of water to the destination is pulling a vacuum in the pipe behind it. The pump cannot "see" what is happening downstream of the high point as a result of the vacuum, and cannot be assisted by it.

Considering the pipe as one installation is therefore inapplicable. The calculations need to be revisited with the above in mind. What is being experienced is driving the calculation, and not the other way round.

Thanks, could you please explain a little bit more in detail on the reply?

The calculations need to be revisited with the above in mind. What is being experienced is driving the calculation, and not the other way round.

In my #17 I was assuming your vent valve at the high point admits air when pipe pressure is below atmospheric, as well as venting. If that's not the case, or if the vent is a manual valve, as implied by your #6, there could be a partial vacuum downstream of the high point, which would reduce the head seen by the pump.

But if the pressure at pump discharge is 3.2 barg, unless there is high headloss between pump and high point, pressure at the high point is well above atmosphere, and it can just be treated as a pipe length, ignoring the ups and downs.

Need to ensure the discharge into sump B is drowned, to give syphonic assistance.

Still happy to estimate losses, if you give data.

Er, the calculation is based upon a model that doesn't match reality?

I would say each one of the high points needs a bleed off valve.....sounds to me like there is air trapped in the line.....either that or your elevation readings are off....What are you measuring altitude with?

http://www.lmnoeng.com/hazenwilliams.php

http://www.pipeflow.com/pipe-pressure-drop-calculations/pipe-elevation-changes

Thanks, actually we installed high point vent valves for all the high point, and we vented all the high points before we tested the line, still the flow is not changed, I do mean the pressure drop is still the same no matter vent or not.

Maybe you got the wrong pump...

Thanks, the presure drop was measured by pressure meter, it seems it is not related to the pump. Any way, it seems the pump curve is OK, and the running pump is aligned with pump curve.

If the pump is running in it's recommended efficiency zone, then it's ok....

http://jmpcoblog.com/how-to-read-a-pump-curve-series/

Dumb Question, but I'll ask it anyway:

Are the Units of Measure consistent for all of the calculations?

If not, don't feel too bad - NASA blew a big chunk of change a while back because some genius forgot to convert metric to English units.

Thanks. Yes, I did the hydraulic calculation in a few of methods actually to confirm if something wrong with original Hazen-Williams method, and all the results are similar (Within 20%)

Your "problem" may not be a problem at all.

If you are measuring your outlet pressure anywhere near the opening of the pipe, then you might only read the head in the finished tank plus a little of the pipe friction representing the final few meters/yards of pipe.

The unconstrained pipe end means that flow measurements are meaningful, pressure at pump discharge is meaningful, but there will be a section near the pipe end where those calculated pressures will not be matched. (Unless the flow is actually restricted by the outlet.)

Imagine if you tried to measure the pressure 25mm (1") from the pipe end.

Thanks, we measured pump discharge pressure only, and I assume the end pressure of the pipe is atm as it is into open sump. The calculated pump discharge pressure should be 1.25 barg, while we need 3.2 barg to deliver the flow calculated (The DP across flow meter had been excluded,).

What is the equivalency between bargs and Bar?

Barg = bar gauge; bara = bar absolute; bard = bar differential.

Thks! we use kPa for such calculations

And I, being from a different time, still use PSIg, or a, or whatever.

All of the units are easily convertible. I find the need to use many of the various pressure units, depending on the situation. MPa, KPa, Bar, Atm, Ft Water, Inches Mercury, mm Hg, etc. ad nauseum.

If you give us the flow, pipe ID, pipe length and no. of bends from pump discharge to the 14m high point, and the pipe length and no. of bends from the high point to Sump B, we can estimate losses and see if we agree with your figures.

Thanks, do you have any different calculation method other than some typical method such as Hazen-Williams, Darcy-Weisbach/Moody formula ? The calculation we had ever been checked by a few of engineers in different method, and different results are close. But the real pressure drop is much too high. Let's assume there is nothing wrong with the piping system, can we say the calculation method we utilised is not approriate in our case (Big up and down of piping lay out.

Have you read my #32?

I prefer Darcy-Weisbach/Moody formula. In Hazen-Williams the constants look pretty arbitrary and plenty of scope for confusion, specially if converting from imperial to metric. Also I believe (unless I'm confusing it with a different formula) you need to adjust the constant for very low or high Reynolds No. Darcy-Weisbach gives a friction factor f nearly always in range 0.004 - 0.007, and no. of velocity heads V²/(2*g) which you should have a feel for so alert you to a mistake. If you've tried other formulas and got reasonable agreement with Hazen-Williams it should be OK, and we're back to explaining the discrepancy with observed pressure loss.

If the vent valves are manual, does water flow out when opened, or suck air in?

You've now given the pipe ID but seem reluctant to give other system details

18 psi discharge...? I don't think so....

_{Just passing by.}

_{The OP is clueless.}

You have got a blockage somewhere along your pipe (tree roots like #3) if not, then some other foreign body trapped inside the pipe somewhere. Any tools missing ?? Has it been quashed or dented somewhere ??

I'll ask the "dumb" question.

Have you calculated using pipe external diameter? (Instead of internal.)

Your pipe supplier will be able to give you the relevant information of the pipes you are using. Otherwise you have some unidentified obstruction, like some cheapskate using an 80mm valve in a 100mm line.

Thanks for the remiding, we used ID instead of OD (14" line, sch 20, ID 339.6 mm)