pump position relative to flow tank 1

[132413]

I have a 1.5 Hp centrifugal pump and I want to use it to move water through a tank to create a piece of test equipment for a flow-meter design. My question is, "Am I wrong to position the entire pump BELOW the tank?" In other words the suction port would take water downwards from one point in a 3'deep tank and force it directly upwards at another point, creating a boiling surface that was moving at a pretty good rate of speed. Would this work, or maybe be bad for the pump?

Peter

 

[Scipio]

There's a couple things you might run into problems with, depending on your flow rate and pressures. The first is the depth of the water - assuming the water stays cold, that 3' should be enough suction head to meet the pump's NPSHR, so I wouldn't expect a problem there at your power range. There is a chance you could form a vortex from the surface, however. I'm not sure how to calculate depth for vortex formation though, but it might be worth a putting a vortex breaker (basically just a steel "X" across the inside of the tank outlet) on the outlet to cut the chances. As far as the return being directed upwards at another point, the only thing I'd be worried about here is the jet being enough to extend above the surface in the tank (sort of like a little fountain). This could cause aeration of the water, which can result in a loss of performance in the pump, and flawed test results. If the tank is big enough, I'd return horizontally below the water level, or through some sort of diffuser to break the velocity. I can't say for sure, not knowing the tank geometry, if you'd have these problems, but I am sure the breaker and horizontal return would minimize the chance of problems.

 

[132413]

Thanks, Scipio, for your careful reply. Something is still unclear to me however, not really knowing the NPSHR and lacking the training to use the concept effectively. With a centrifugal pump, is it necessary to "suck" from ABOVE the water level, ie., to have some sort of resistance that way? With my intended arrangement, the suction intake is below the entire tank, so that the water flows downhill into the suction unit. The discharge is blowing upward into the bottom of the tank. With a 1.5 Hp pump, would there be problems with the resistance placed on the discharge end rather than on the intake? The tank is 2'long, 3'deep, and 8"wide. Total depth down to the pump, including pipes and hoses is 5'. I worry a little about cavitation and possible damage to the pump.

 

[Artisi]

It is a misnomer that pumps "suck" - pumps don't "suck", a pump unit lowers its internal pressure and atmospheric pressure forces the product to be pumped into the eye of the impeller.

But back to your question, basically there is no problem with the supply level being above the pump inlet - in fact this is a prefered method of supply to the pump inlet.
As for pumping back into the bottom of the tank - this does not represent any problem either.
The only problem I can see at this point without having all your data available is that the return flow into the tank does not directly impinged on the inlet flow - direct the return flow away from the inlet flow as unstable flow to the inlet could result in performance problems.

Naresuan University
Phitsanulok
Thailand

 

[quark]

This seems to be a weird setup. What is the reason for connecting pump discharge to tank bottom?

You can visit the following sites for a good understanding on pumping systems

http://www.mcnallyinstitute.com

http://www.fluidedesign.com

http://www.pumpworld.com

Scipio,

Minimum submergence required to avoid vortexing can be calculated by the formula(I think this holds good only for water)

S = D + 0.574Q/D^1.5

Where S is submergence in inches,
D is inner diameter of suction bell in inches and
Q is flowrate in USGPM

Reards,

 

[132413]

Artisi---
Thanks for your good response. I suppose it's a slight semantic question as to the meaning of "suck." When a baby "sucks" (from his/her mother), he/she lowers the pressure inside the mouth using tongue and mouth muscles, and by your corrective reasoning, atmospheric pressure forces the milk into the baby. Hey...but of course! Taking atmospheric pressure into consideration is a unique and novel way of viewing early childhood.

Quark---
The setup is "weird" in order to make for a smoothly flowing water surface in the tank. Roughly square in profile---a polyethylene plating tank---the pump is discharging into the bottom at the lower left corner and drawing out of the bottom at lower right corner. The 3'depth of the tank dampens the fountain effect of the discharge and a series of plastic plates in the upper left corner smooths out the turbulence. The biggest problem is how to vary the flow from the pump, since it's an induction motor and it's difficult to make a variable speed control for an induction motor (that one can afford). Many thanks for the websites.

Scipio---
Thanks much for the formula. I'll check it out.

 

[checman]

132413, The pump below the supply is one of the preferred placements of a pump. The suction line should be of a design that does not allow air pockets to be trapped. You most likely need to control the discharge of the pump. If you don't it is possible that it will run out on its curve and this could damage your pump. If your motor does not over load or run hot you may have enough back pressure through your piping to keep it from running out on or off its curve. A pressure gauge on the discharge will allow you to plot the pump on its cure. If the tank is vented to atmosphere you will have atmospheric pressure plus the distance from the top of the liquid to the centerline of the pump minus piping loses and vapor pressure for water at the temperature pumped for a rough NPSHA number. I would guess and say you have plenty. Also minimum submergence is based on inlet piping velocity. But if you are running the pump on its curve and piped without reducing the suction piping you should be OK with three feet. If you supply us with more details we could be more precise. But what kind of flow meter would you test in this system?

Regards checman

 

[Artisi]

Having just rechecked the info., have I read the tank width correctly at 8" (inches)?
What is the estimated thru-put of the pump under your run conditions?

As for flow control, you could throttle the pump discharge so long as you stay within the operating band of the pumps performance, although this will increase the water temperature at a greater rate over time.


Naresuan University
Phitsanulok
Thailand

 

[132413]

Checman---
Thanks for your thoughts. I'm somewhat relieved because of them. It's a 2.5"dia. propeller-driven flow meter that uses tiny rare earth magnets placed at 180 degrees, N and S poles, to make a near sine wave signal against a coil with a soft iron core. Your basic "magnetic reluctance" pickup.

Artisi---
I also appreciate your concern coming from so far away. The pump I have, discharges about 38 GPM at 100 feet of head. I can't quite tell what it can be expected to do with only around 5 feet of head. Yes, the tank is 8" wide, an odd shape but it's all I need to test out a propeller flow meter, run some data on it, and so forth. I may have purchased a pump that is too powerful for this particular system, I don't know.

 

[Scipio]

132413,

That's a huge flowrate compared to what you need, you'll need a choke valve on the pump discharge to throttle the flow and make up the difference in head, otherwise the pump will run off the end of it's curve, which introduces all sorts of problems, assuming the motor doesn't overamp and trip out. For a centrifugal pump, you should start against a closed discharge valve, then open the valve to increase flow from zero. A globe valve will probably work the best for your application. You may also have temperature problems with the water, every time it passes through the pump it'll get hotter, and it looks like you'll completely exchange the water in that tank in a little under 6 minutes.

 

[Artisi]

such a huge pump (relative) for the application - can you change the pump unit for a smaller unit -- alternately - reduce the pump impeller diameter to reduce flow and head, this will also reduce power consumption and lead to less throttling of the discharge and less heat build up in the water supply.
Is it a motor-pump? can you change motor, if 2 pole motor changing to 4 pole will reduce head by factor of 4 and flow by half.
It is unlikely you will have any overload problems - pumps of this size tend to be non-overloading.

Naresuan University
Phitsanulok
Thailand

 

[quark]

Apart from the problems of overloading of pump and heating up of circulating fluid, already mentioned above, you will have problems of inaccuracy of flowmeter. Though you provide straight lengths either side, flow will not be straightened with an almost closed valve(to maintain 5ft. head from 100ft.) with in short piping distance. It will also becomes difficult to test the meter at various flowrates. I concur with Artisi for the solution.

Regards,

 

[JStephen]

If you don't want to use a valve to cut the flow rate way down, you could also put a bypass and a valve around the pump itself, so that a variable percentange of the flow just went around the loop, and not through the tank. It might work out better for the pump that way.

 

[25362]

If I understood the query rightly, I was thinking more in the lines of putting the meter under a constant hydrostatic head (water level) maintained, for example, by using overflows. The downward flow going to a receiver on which the collected water rate can be measured for meter calibration, and returned to the existing "prismatic" tank by a pump, if so desired. How far am I from the original "correct" idea ?