20 replies to this topic

[go-fish]

Hi

If a pump is selected for 100 m3/hr at a given mass density of a fluid, will the pumping capacity of the pump change with variations in density due to temperature?

My understanding is that centrifugal pumps are volumetric machinery and hence, they work at a fixed volumetric flow even if there are density variations due to temperature. So if I am running hydraulic calcs for two densities, I should keep the volume flow same and vary the mass flow rather than vice versa.

Thanks

[ankur2061]

go-fish,

Density or more commonly the specific gravity of a liquid has no effect on either the volumetric flow rate or the head developed by a centrifugal pump. This essentially means that for a given volume flow rate a change in density will change the mass flow rate. Thus if the density increases the mass flow rate will increase and if the density decreases the mass flow rate will decrease.

The most immediate noticeable effect of density or specific gravity change if you are not measuring the mass flow will be in the absorbed power of the pump driver. This is evident from the absorbed power equation for any pump.

P = Q*H*ρ*g / 3600*η

where:

P = absorbed power, kW
Q = volumetric flow rate, m³/h
H = Differential head, m
ρ = density of liquid, kg/m³
g = gravitational constant , 9.81 m/s²
η = efficiency, dimensionless

Considering the volume flow rate, head and efficiency remain unchanged an increase in density increases the absorbed power and a decrease in density decreases the absorbed power for the pump.

This principle is important when a motor rating is to be provided for a pumped liquid which has a cold start-up. An example would be the motor rating of a Heat Transfer Fluid (HTF) pump. When sizing the motor for such a pump, you cannot use the density of the HTF at its normal operating temperature which would probably be 150-200 deg C more than the cold start-up temperature for motor sizing, since the motor would become undersized for the cold start-up temeprature. You wil have to arrive at the motor rating using the density corresponding to the cold start-up temperature.

Hope this helps.

Regards,
Ankur.

Edited by ankur2061, 19 March 2012 - 02:49 PM.

[S.AHMAD]

go-fish
1. You are absolutely right - the volume flowrate unchanged but the mass flowrate does. However keep in mind that the volume is at the operating conditions of the pump. Therefore, if you want to calculate the mass flowrate, need to know the density of fluid at the operating conditions.
2. Ankur's comment on motor power is a good one.
3. Another example is when using water for line flushing during commissioning where the density of the design fluid may be much lower than that of water. Electric motor may overload if necessary precautions are not considered during line flushing.

Edited by S.AHMAD, 19 March 2012 - 07:20 PM.

[go-fish]

Thanks for descriptive answers

[knapee]

go-fish
For a given mass flow rate of liquid its volumetric flow rate is also inversely proportional to its density. Hence the head-capacity curve for a given fixed-speed pump is the same for all liquids irrespective of the liquid density.

However the.discharge pressure will vary with the density of the liquid being pumped

[yurge]

hello everyone!

 

I have read a copple days ago at some brochure, that with change of density in air stream ( in centrifugal blower) it will change althought a static pressure system resistance.

 

And if a change of a density in a pump leaves u with the same head and flow rate, thats mean that your working point ( meeting between pump curve and system resistance) should be the same and without any changes. So how it possible that with change of density  not affects on system ressistance?

 

Now i know that a topic is about a centrifugal pump, but centrigugal blower is a volumetric machine too and basically works on the same principales.

 

Thank you.

[furfurylic]

I finally believe I have gotten an answer after a lot of searching, but just to be sure:

 

This is our setup:

One centrifugal pump for circulating water through 2 heat exchangers (+piping).

I have measured flow rate with water@ 25 deg.C/ 8.5 bar(a) with an ultrasonic device to 15.4 l/s.

So if I where to do the same measurement for water@ 175C / 10 bar(a) I would see the same flow rate?

 

I thought that maybe the reduced viscosity would result in a reduced head for the system, and since the characteristic head of the pump will not change, the result would be an increased flow rate?

 

I actually want to calculate the power transfer in one of the heat exchangers by monitoring the temperature drop, and thought I needed to compensate for both the density and the flow rate to get the mass flow?

 

Br,

Karl

[Raj Mehta]

Hi Karl, 

 

Head and capacity depends only on two things (refer Affinity laws article at: http://www.chemicalfiles.blogspot.com/

 

1. Motor rpm (N)

2. Impeller diameter (D) 

 

So reduced viscosity has no effect whatsoever on the head. It might effect your pump efficiency, but not head or capacity. 

 

As Temperature increases, density (specific gravity) of the fluid increases. Your voliumetric flow-rate depends on the upstream process and the control philosophy in the upstream of the plant, so unless your throttle the valve or the upstream process gets upset, you will not see any changes in the volumetric flow-rate.

 

Since density = mass / volume & volume flow-rate is constant, so your mass flow-rate has to chnage with density. 

 

Hope things are clear. 

 

Regards,

Edited by Raj Mehta, 11 December 2013 - 03:35 AM.

[ankur2061]

Raj,

 

 

As Temperature increases, density (specific gravity) of the fluid increases.

. 

 

 

Raj,

 

Density or specific gravity decreases with increase in temperature.

 

Regards,

Ankur

[Raj Mehta]

My apologies to all. 

 

Ankur sir is right, Density decreases with increase in temperature. 

[furfurylic]

Thank you very much for your valuable feedback.

 

I don't have any trouble with accepting that the pump characteristic doesn't change with a decrease in density and viscosity, but I don't quite get that the decrease in viscosity doesn't affect the flow resistance (head) of the rest of the system. In my head a lower viscous media will have less pressure drop flowing through a defined system than a higher viscous media, when both have the same volume metric flow rate. If then the pump characteristic is the same (same P vs. V curve) then the whole system should settle at lower pressure drop but higher flow. (The media should have lower static pressure but higher dynamic pressure).

 

I know that in my system the circulating media (water), changes properties from when I'm measuring till the operating point:

Dynamic viscosity: from 0.89 to 0.15 cP (83 % reduction)

Density: from 997 to 890 kg/m3 (11 % reduction)

Heat capacity, cv: from 4.1 to 3.5 kJ/kg/K (15 % reduction)

Heat capacity, cp: from 4.2 to 4.4 kJ/kg/K

 

Working on this now, as I write this I'm thinking...

Because when the density goes down, so does the dynamic pressure (q=1/2*density*v^2) of a constant volumetric (speed) flowing media. So when a centrifugal pump is working of such a media, at lower densities the maximum possible dynamic pressure of the media flowing through the pump, (defined by impeller area and n (=Volume/time)), is lower. So the "transfer" from static to dynamic pressure (P vs Q curve can be seen as a Pstatic vs Pdynamic curve), "requires sacrifice of" more static pressure for a given amount of dynamic pressure.

Using this on my system: When my water circulates at 175 deg C, the static pressure is the same, but the dynamic pressure is lower by the same amount that the speed (=volumetric flow rate) of the water is the same. Hmm.. just don't see how the decrease in viscosity effects this though...

 

(Sorry for my quite chaotic mind ramblings.)

 

Again, thank you very much for your responses.

 

Br,

Karl

[curious_cat]

 

 

I don't have any trouble with accepting that the pump characteristic doesn't change with a decrease in density and viscosity, 

 

Even that is only true within bounds, I think. If you used a water derived head curve to extrapolate to say a viscous crude your head curve is going to be off.

Edited by curious_cat, 14 December 2013 - 02:30 AM.

[karthik.kattupalyam]

Dear sir,

 

Can anyone pls tell me how to find flow rate of pump by using pump curve and I got some curves but there is so many graphs i cant understand.

Actually I need to find flow rate of pump  because here that pump have not any kind of details, So I kept a press gauge at discharge side I found 2kg/cm2 pressure and motor having 5Hp and 2900Rpm and I calculate head is 20m and by using power formula  

Power(Kw) = q(m3/hr) ρ(kg/m3) g(m/sec2) h(m) / (n*3.6*10^6),

I found 50m3/hr with 0.75 efficiency ,So please  tell that is correct or not and I need pump curve also.

[breizh]

Karthik ,

 

Consider this resource , it should help you to find answer to your query .

For your specific question you can read chapter 5 , I'm encouraging you and others to use this excellent "free" book .

 

Breizh

 

[fallah]

Dear sir,

 

Can anyone pls tell me how to find flow rate of pump by using pump curve and I got some curves but there is so many graphs i cant understand.

Actually I need to find flow rate of pump  because here that pump have not any kind of details, So I kept a press gauge at discharge side I found 2kg/cm2 pressure and motor having 5Hp and 2900Rpm and I calculate head is 20m and by using power formula  

Power(Kw) = q(m3/hr) ρ(kg/m3) g(m/sec2) h(m) / (n*3.6*10^6),

I found 50m3/hr with 0.75 efficiency ,So please  tell that is correct or not and I need pump curve also.

 

Hi,

 

You should apply pump differential head, then you should use discharge head minus suction head instead of the discharge head alone. Indeed, if you have the pump curve in hand, you can check the calculation result on such curve...

[S.AHMAD]

Karl

Just to answer your question about the effect of temperature. I have mentioned it on post#3 that the volume of the pump curve is at the operating conditions. This means that you need to convert the volume measured at 25^oC to the volume at 175^oC if the pump operating conditions is at 175^oC. Pressure is normally has insignificant effect on volume of incompressible fluids.

 

Volume @175^oC = Volume @ 25^oC x density @25^oC / density @175^oC.

 

However, in industrial practices we normally ignore the effect of temperature, since the error in measurement of the flowrate is also considerable. Therefore to simplified our calculation we normally use whatever volume measured at 25^oC.

Edited by S.AHMAD, 31 January 2014 - 02:35 AM.

[S.AHMAD]

Dear sir,

 

Can anyone pls tell me how to find flow rate of pump by using pump curve and I got some curves but there is so many graphs i cant understand.

Actually I need to find flow rate of pump  because here that pump have not any kind of details, So I kept a press gauge at discharge side I found 2kg/cm2 pressure and motor having 5Hp and 2900Rpm and I calculate head is 20m and by using power formula  

Power(Kw) = q(m3/hr) ρ(kg/m3) g(m/sec2) h(m) / (n*3.6*10^6),

I found 50m3/hr with 0.75 efficiency ,So please  tell that is correct or not and I need pump curve also.

 

Hi

1. The 5HP is the motor design but the actual motor power is normally less than 5 HP. So your calculation is incorrect.

2. As we all know that the pump performance curve is the relationship (graph) of pump head (in m or ft) versus flowrate If we know the pump head then we can read the flowrate from the curve.. This is the case if the pump is performing as per design. Calculate the hydraulic power (without the efficiency)

3. If you are doing pump performance analysis, you should also measured the motor voltage, current and power factor and convert this data to actual motor power consumption.

4. The ratio of hydraulic power to motor power shall gives you the actual pump efficiency. Compare the actual with design efficiency. If about the same then the pump is performing as per design. If the actual efficiency is too much lowerr than the pump design efficiency, this means the pump is not performing as per design - requires some repair.

5. Alternatively, you can determine the actual flowrate by using the method that you used earlier. Instead of using the 5 HP use the  actual motor power. If the calculated flowrate is very much lower than shown by the pimp curve, that indicates pump problem.

Edited by S.AHMAD, 31 January 2014 - 03:24 AM.

[karthik.kattupalyam]

 

Dear sir,

 

Can anyone pls tell me how to find flow rate of pump by using pump curve and I got some curves but there is so many graphs i cant understand.

Actually I need to find flow rate of pump  because here that pump have not any kind of details, So I kept a press gauge at discharge side I found 2kg/cm2 pressure and motor having 5Hp and 2900Rpm and I calculate head is 20m and by using power formula  

Power(Kw) = q(m3/hr) ρ(kg/m3) g(m/sec2) h(m) / (n*3.6*10^6),

I found 50m3/hr with 0.75 efficiency ,So please  tell that is correct or not and I need pump curve also.

 

Hi

1. The 5HP is the motor design but the actual motor power is normally less than 5 HP. So your calculation is incorrect.

2. As we all know that the pump performance curve is the relationship (graph) of pump head (in m or ft) versus flowrate If we know the pump head then we can read the flowrate from the curve.. This is the case if the pump is performing as per design. Calculate the hydraulic power (without the efficiency)

3. If you are doing pump performance analysis, you should also measured the motor voltage, current and power factor and convert this data to actual motor power consumption.

4. The ratio of hydraulic power to motor power shall gives you the actual pump efficiency. Compare the actual with design efficiency. If about the same then the pump is performing as per design. If the actual efficiency is too much lowerr than the pump design efficiency, this means the pump is not performing as per design - requires some repair.

5. Alternatively, you can determine the actual flowrate by using the method that you used earlier. Instead of using the 5 HP use the  actual motor power. If the calculated flowrate is very much lower than shown by the pimp curve, that indicates pump problem.

 

Dear sir,

 

I checked amp of that time it shows 5amps so i calculated 1.73*V*A*cosphi so 1.73*430*5*0.997=3.72Kw appx 5hp it runs on full load.

[breizh]

Hi ,

The power factor seems to me  high ( 0.997 ) .

Take a look at the appendix E .

 

Additional resource :http://www.pumped101.com/

 

Breizh

Edited by breizh, 03 February 2014 - 08:57 AM.

[karthik.kattupalyam]

Hi ,

The power factor seems to me  high ( 0.997 ) .

Take a look at the appendix E .

 

Breizh

 

dear Mr.breizh,

 

I checked that in my plant electrical records, Electrical engg also said they are maintaing with high power factor in my divison plant

[S.AHMAD]

Dear sir,

 

 I got some curves but there is so many graphs i cant understand.

Hi

Can you share with us the kind of graphs that you have?