Determining Pump Head

[rockman7892]

We have a pump here at our plant that pumps cement from a Mill up to the top of a silo. Lately we have been overheating the motor on this pump and I am trying to determine if for some reason the pump may be working too hard.

Now I know the basics of pumps and that the static head is related to the height of the discharge pipe above the pump suction resoivar. I also know that from a static head you can look at a pump curve and determine the CFM and HP of that pump for that given static head.

My question is regarding the static head distance in feet and how it relates to a pressure. In other words, lets say I have a static head of 100ft, how do I represent this as a head pressure since this is what pump curves have listed?

Also if using something to represent the head pressure do you need a pressure gauge down at the pump or up at the discharge of the pipe? If I wanted to watch the head pressure in this pipe to use it as a variable on the pump curve to look at flow rate and HP then where would be a good place to locate this pressure gauge?

 

[GPRnD]

I am not sure whose pump curves you are looking at, but all major manufacturers of pumps provide pump curves showing the pump head in ft or m.

If you are asking how to convert head to pressure, then it is simply a matter of multiplying the head by the fluid density and the gravitational constant.

So in your example:

100 ft of head = 30.5 m

Pressure = rho * g * h

Using metric units and assuming your fluid has the density of water (SG =1)

Pressure = 1000 * 9.81 * 30.5 = 299205 N/m2 = 2.9 bar = 43.4 psi

If you want to measure the pressure and convert it directly to the head produced by the pump, you should mount your pressure gauge at the shaft centerline of the pump.

If you mount it at a higher location, you need to take into account that vertical height difference when calculating the pump head.

So for example, say you mounted the pressure gauge 16 ft above the pump shaft centerline and your pressure gauge read 55 psi.

The pump head is then:

55 psig = 379232 N/m2

fluid head = 379232 / (1000 * 9.81) = 38.7 m = 126.8 ft

pump head = 126.8 + 16 = 142.8 ft

 

[PEorl]

Technically the pump head would also include line losses and any fitting losses on that 16 ft of pipe (they probably are small, but should be checked) above the pump shaft centerline.

 

[SNORGY]

This situation sounds to me like there may be other factors at play as well. If you are pumping cement with a centrifugal pump and are seeing variations in viscosity and density, your pumping efficiencies will also be varying, as will the amount of work demanded from the system. The downward driver on efficiency at a given volumetric flow will manifest itself in increased heat. You may also be seeing increased friction and out-of-tolerance shaft deflections that affect the driver-to-pump alignment and, consequently, the required work input and heat generated. It may even be something deteriorating in the motor itself.

If you know the volumetric flow rate, density, viscosity, pump differential pressure, and pump efficiency curve, then the current drawn by the motor (you might be able to get that from a panel or MCC) will be able to tell you if there is something amiss with work input to work output. You can probably get a good indication of the flow rate by using a Doppler meter (if you do not have a flow meter in the system) strapped to the outside of the bare pipe.

Regards,

SNORGY.

 

[cvg]

depending on the fluid pumped, flow velocity, fittings, valves and size of pipe - the line losses (friction and minor losses) may be small or quite large. These are added to your static head to get your total dynamic head which is the one that is shown on your pump curve. Given that the motor is now running hot, you may have to replace both the pump and the motor.

 

[Ron]

Keep in mind that cement has a specific gravity of 3.15 as compared to 1.0 for water. That makes a big difference. Further, the friction losses are much greater because of the abrasive nature of the material, even if considered a viscous liquid.

Most of your values for water need to be severely factored up when dealing with cement.

 

[SNORGY]

True enough.

I had assumed that if the motor has only been overheating lately, then presumably it was correctly sized and specified at some point in the past. I would be inclined to look for sources of downgraded efficiency - impeller wear, changed clearances, increased friction, deteriorating windings, change in fluid properties, rise in dynamic losses due to increased pipe roughness - that sort of thing.

Regards,

SNORGY.

 

[tkall]

You are pumping liquid cement with a centrifugal pump? I have seen that done with PD pumps. Or is it dry cement?

 

[Artisi]

Getting back to your question, if you have 100ft of head and you are pumping cement and assuming cement has an specific gravity of 3 then the pressure is:

100 / 2.3 x 3 = 130.5 PSI

 

[Ron]

tkall..."Liquid cement" would hydrate and harden rather quickly. The cement would have to be dry...very dry, and is being pumped into a silo for storage use.

 

[rockman7892]

Thanks for the responses.

We are indeed pumping dry cement, and the pump I am referencing is a Positive Displacement pump.

As I mentioned, we are pumping cement into a storage silo via a 12" transport line. There are multiple silos and when pumping into other silos the pump seems to be fine, but when pumping into one particular silo the pump motor seems to overheat and burn up. We have smoked several motors.

It sounds like you guys are saying that I can put a pressure transmitter at the centerline of the pump shaft and use it to watch the head pressure. Since this head pressure is a combination of static head and friction head with static head staying the same, then an increase in pressure at this transmitter will tell me if there is additional friction pressure in the transport pipe?

Can I then use this pressure reading and convert it to a head in ft to comare against the pump curve to see what HP it equates to? I'm hoping that I can do this to prove to others that this head pressure is causing the pump to run harder and thus burn up the motor. Others seem to think there is a problem with the pump itself.

 

[GPRnD]

rockman, I think your plan sounds reasonable.

If the pressure gauge shows that the head at the pump is much higher when pumping to a particular silo, you have found the problem in that the piping to that silo must have something in it causing more frictional loss.

Since you are using a PD pump, the pump will simply respond to that increased head by drawing proportionally more power from the motor. If you know the efficiency claimed by the manufacturer, it should be a simple task to calculate the motor power.

 

[rmw]

I suspected from reading your initial post that you were pumping powder. I have not seen it pumped as a fluid per se, but I have seen it when a line would burst and a silo or hopper would empty itself out with the stuff flowing just like water, flowing all over the place. Nasty.

I also suspect that you have some place in the line to the silo in question where you have got a blockage of some type. Either moisture has gotten in and caused some of the material to set up, or some of the material when the pumps were shut off settled in the lines and now that it can't be re-aerated, it can't be lifted up and transported on to the silo and it has reduced your flow area. The pump would have to overcome that reduction in flow area in order to try to pump its design output.

Check your lines for blockage.

rmw

 

[rockman7892]

I went and looked at the pump closer and found that there are (4) pressure gauges and tranmitters already installed on the pump. There is one pressure gauge and transmitter located on the outlet transport line, one located on the inlet air line from the compressor and the other two are located on each side of the pump.

I noticed when pumping cement last week the outlet gauge on the transport line was reading about 19psi. The inlet gauge from the compressor line was reading about 24psi. Pump seemed to be running o.k. with these readings.

Obviously we can watch the outlet pressure transmitter when switching between silos to see if the pressure increases. If this PSI head increases from 19psi to some greater pressure then this would indicate friction in the pipe and could indicate the problem.

Could the other pressure gauges mentioned be used to indicate if there is a problem with the pump iteself as opposed to the transport line?

 

[rockman7892]

In looking into all of this I'm also trying to lean the main differneces that this positive displacement pump has from a centrifugal pump.

I know that with a centrifugal pump the volumetric flow will be dependent on the pressure head, and as the pressure head decreases the volumetric flow will increase. The HP of this type of pump will increase in proportion with the flow.

With my limited knowledge about a PD pump I know that the volumetric flow will be constant independent of the pressure head. So if the flow rate stays the same now matter what the pressure head is, then how does the hp of the pump react to changes in pressure or flow? Is the constant flow determined by the size of the pump only?

 

[GPRnD]

For a PD pump the volumetric flow is based solely on the speed of the pump (for a given pump size).

If you want more flow with a PD pump you run the pump faster.

With a PD pump, the absorbed power is function of the head x flow.

Since the flow is fixed for a given speed, the power varies in direct proportion to the pressure.

 

[rockman7892]

O.K. so that I understand.

A PD pump will put out a constant volumetric flow rate that is determined by the speed of the pump and will be the same regardless of any pressure or friction head. So even with friction in the line the pump will output the same flow rate and the hp of the pump will increase in proportion to this head.

A centrifugal pump in contrast will have a flow rate that is determined by the amount of head. So as the head increases the flow rate will decrease. The hp in a centrifugal pump will also be a function of head x flow however since flow will not be held constant the hp will vary as a function of both. As the head increases the flow will decrease however the hp of the pump will increase?

 

[Artisi]

Correct for a centrifugal pump with a continually rising H/Q curve - for an axial flow pump (can be classed as a centrifugal pump) power will normally increase with rising head and reducing flow.

 

[rockman7892]

If the pump is being overloaded from the input side will any of the pressure readings I mentioned indicate this overloading condition? Maybe a differential pressure across the pump?

 

[GPRnD]

Yes

If you measure the differential pressure across the pump you should be able to compare it with the pressure you get pumping to the other silos.

Then you can simply use a ratio of the two pressures to calculate the power increase you are seeing with this one particular silo. (This is assuing you have some data on the pump efficiency or a readout of current and voltage from the motor).