discharge pressure centrifugal pump 5

[Muud]

I am trying to determine the discharge pressure of a centrifugal pump in a closed water system.

The centrifugal pump sucks the water from the bottom of an open tank (open top and under atmospheric pressure) and circulates it through different small vessels before finally driving the water back into the same open tank (but through a different inlet).

From the pump performance data I can see head at discharge at specific flow rate. Since in this case there is no elevation difference as water is taken and recirculated back to same tank I guess the static head will be zero. I believe if I just take the discharge head and by dividing it by 2.31 the result will be the discharge pressure of pump. Is this way of calculating the discharge pressure in this closed loop system correct?

 

[Artisi]

It's not a closed loop.
The discharge head (pressure) will be the static head which is inlet level in the open tank to the highest elevation you pump to, plus friction losses in the pipework plus head losses thru' the "small vessels".
Here could be some head recovery from any syphon effect but without a layout sketch of the pipework showing elevations etc - it can't be answered on the current info.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[bimr]

As Artisi states, it is not a "closed pumping system" because it is open to the atmosphere.

 

[georgeverghese]

Performance curve data shows differential head dh vs flow, where dh = discharge head - suction head. In this case, suction head = static head at tank - friction loss in suction piping, expressed as head loss.

 

[LittleInch]

Artisi has it correct as usual.

1) It's not a closed system , but is a circulating system where the static arrival head into the tank at the end of your piping system equals the inlet head into the pump so they will cancel out.

2) If the water circulating pipework is actually at the same elevation or only a few feet max higher than the pump then the pump differential head will equal the friction losses from discharge to return into the tank. If there is a significant high point in your pipework, this could impact on the head required to make the system flow.

3) The actual discharge pressure will be the differential head PLUS the inlet head of water coming into the pump. The end pressure of your system would also be a positive pressure due to the head of water above the tank inlet. This may be a negligible amount or it might be significant.

4) You will need to post pump curve, system elevation data, height of water above tank exit and inlet and then we can see if your supposition is correct.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Muud]

Thanks everyone for your input.

I have made a diagram showing the test setup with elevation data. Also I am posting the pump performance data.

I hope it makes the description of the system clear.

 

[Muud]

Here is the pump performance data

 

[LittleInch]

Muud,

you've posted the pump data twice, but not the system description.

Also assume H is the differential head?

Is this a test report?

Is there a graphical curve available rather than the table?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Muud]

Sorry my bad. I have fixed the reuploaded the file in the same post.

Yes H is the differential head (Hd - Hs).

The is the only pump performance data that I could dig out from the old pump documentation. There was no other pump curve data. I plotted the Q vs H to get a curve myself but I got a rather strange looking curve as you could see in the values as well. The H goes up and down for increasing flow. No idea why is that so.

 

[Muud]

I uploaded the correct attachment but for some reason I still see the old one. Anyways please see the system description schematic attached with this post.

 

[LittleInch]

[Edit] Ok.finally worked.

Those data points look a bit odd especially around the 15.7 to 15.9 flow range...

so what do you think the flow is or are you measuring it?

The pump curve looks very flat. What sort of flow are you getting?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[bimr]

You may use the formula in McNally's to calculate the system head:

McNally Website

 

[georgeverghese]

Assuming velocity in the discharge line to the first tank (vessel 1) is low, as a first approximation, discharge pressure will be approx 72cm of water column, let say it is 75cmH2O. This converts to a discharge press of (0.75/10.2)*14.51 = 1.07psig, since 10.2m H20 = 14.51psi.

 

[LittleInch]

OK,

Let's review what we have here.

You have an re-circulating system via an open atmospheric tank.

You have a pump which has a remarkably flat pump curve which varies from basically 41.5m to 37.5m.

Your inlet head into the pump, whilst not stated, is a max of 1.5m so basically lets say 1m. Essentially negligible.

Your maximum head into your presumably closed vessels full of liquid is a maximum of 75cm from pump centerline so again virtually negligible compared to the pump discharge.

Therefore, although you've given us no flow data to use, we can only assume your pump operates somewhere between lets say 5 to 20 m3/hr.

At that flow your discharge head will be probably 41m head +/- 1m. That's equivalent to 4.02 bar or 58.3 psi.

Makes sense?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Muud]

@ LittleInch. Sorry for my late reply.

As you rightly mentioned the performance data seems odd to me as well.

We measure the flow and depending on the test, we commonly set flow at either 1 m3/hr, 4 m3/hr or 10 m3/hr. (tests at 4 m3/hr and 10 m3/hr are more frequent though). Thus typical flow range for our system 1 to 25 m3.

The inlet head typically around 1m as you have supposed.

You approximations were spot on

It does makes sense and I had calculated a similar discharge pressure around 4 bars @ 23 m3/hr and an suction head of 1m. I have learnt more from this now then just assuming things

I believe I do need to factor in the friction losses through the piping and small vessels on top of that? Correct?

 

[LittleInch]

Nope.

I don't know how you're controlling flow, but I can only guess that you have some sort of variable resistance device ( aka a control valve). The total pressure resistance of the system will equal the output of the pump. some will be taken by the valve, the rest by the friction losses in the tubing and vessels.

BTW, running a pump designed for up to 20 - 25m3/hr at 1 m3/hr is not a good idea. I know it's fairly small, but it's still not the way to do it.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[georgeverghese]

Sorry, forgot to mention that previous calc was based on the assumption that it is atmospheric press at vessel 1 and that there are no restricting/ control devices in the line. With these later posts, it is now clear vessel 1 pressure is much higher and that there is an FCV or similar on this line.

 

[Muud]

@ LittleInch, There is a butterfly valve after the pump to regulate the flow manually.

We do have a bypass line that goes from the pump to the recirculation line. This bypass line is normally kept open when the pump is running. I have attached an updated drawing showing that bypass line. However, I have noticed that there is a lot of noise, bubble formation and turbulent flow in the system with bypass line open. When carrying out the test at 10 m3/hr I sometime do close the bypass line and there is much less turbulence and noise from the system. However, in this case the temperature of water does rises. There is no temperature sensor or probe installed so I can't measure how much is the change.

I am doubtful if the recirculation loop is appropriate (the point at which it presently enters the suction tank or directly into the suction line or other possibility). Besides I also wonder if mixing the minimum flow line and recirculation as shown in the drawing is a good idea because it might also be generating turbulence and air bubbles in the suction tank which then may enter the pump.

What do you think about the bypass line and recirculation loop?

 

[LittleInch]

Ah, the plot thickens.

Please also add to your diagram or state where you measure flow and how, if at all, you prevent the pump from going to beyond its end of curve now you're given the water a very short ride back to the tank from which it came???

My guess is that you are measuring flow downstream of your butterfly valve ( shown as a ball valve on your sketch) and in fact the flow through your pump is actually very high, causing possible vortexing of air into the pump inlet, cavitation of your pump and damage to your pump??

Bypass lines need to provide for a fixed flow and provide a pressure drop so that the pump doesn't exceed its max flow. It might be that there is an orifice plate or similar restriction in the bypass line which is actually causing pressure to fall below vapour pressure in the orifice plate which is causing the bubbles and vibration

If you get it right then adding it into the line going back to the tank is Ok, but as said I think you have a high flow in the bypass line.

The temperature rise doesn't make a lot of sense, but over time the friction losses will increase the temperature of the water if the open water tank is quite small relative to the flow through your system.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Muud]

I have updated the drawing now with the flow meter, pipe lengths as well as the vessel heights.

Yes the flow in our system is measured downstream the control valve, after the last vessel as shown in the drawing. Indeed we are only measuring the flow through the system downstream and not the bypass flow. The bypass line is smaller in size than (~ 1/2 of main flow lines) and has a number of resistances (valves) some of which can be taken out I suppose.

Regarding exceeding the max pump flow, I suppose that would be the last point on the performance curve is around 23 m3/hr and indeed now considering the bypass flow, we may be exceeding the max flow of the pump. The bypass flow is not measured so I don't know how much is it. Although that may not be happening frequently.

In the bypass line, there is no restriction orifice plate, it's just a smaller line with an elbow, a check valve and a control valve. I think the bypass line is not optimally designed with regards to position, and causing unnecessary pressure drops. I would like to redesign and reposition it perhaps from top of the tank so that the recirculation and bypass flow are not mixed to avoid creation of vortices and air bubbles. Could you please give me some pointers regarding the redesign of bypass line?

Should we also add a flow meter to the bypass line? or can it be calculated somehow?
I