What should be a minimum tank level to operate pump? 6

[Abhi1712]

I am working with a small pump company. (Company is newly started). I am having centrifugal pump of maximum capacity of 132 GPM(crude oil transfer). Minimum suction pressure required at pump is 8 psi. I have no data available for NPSHR. So I have to consider crude oil tank level where the pump has to be install. Can you please tell me what should be minimum tank level to operate pump?

 

[vpl]

If you've already bought the pump, then the required net positive suction head is usually obtained from the pump manufacturer. My understanding is that, if you are the pump manufacturer, you test the pump to arrive at the minimum.

If you're designing the system, then your tank needs to be big enough for your system needs. Generally, for water systems (which I realize you don't have), a rule of thumb to prevent vortexing is 3 feet of head above the NPSHR.




Patricia Lougheed

******

Please see FAQ731-376: Eng-Tips.com Forum Policies for tips on how to make the best use of the Eng-Tips Forums.

 

[BigInch]

8 psiA and is that with water or crude?

if that is 8 psiAbsolute specified with water in the suction, it would mean that min NPSHR is 18.4 ft, with crude in the suction it would mean min NPSHR is 22.25 ft.

 

[Abhi1712]

To VPL:

Thank you for your reply. We are the Pump manufacturer. Can you please explain in detail how to calculate NPSHR? or any useful link to calculate NPSHR?

We are not designing the system. We are going to install pumps at various refineries.

 

[BigInch]

The pump would obviously have to be mounted outside the tank at a lower elevation in order to not waste all that tank shell height.

 

[Abhi1712]

To BigInch,

Thank you for your reply. Will you please explain in detail how to define tank level (suction side). I have to instruct operators for installing pump at various locations. Also if tank level is below required then we have to provide "Booster pump" to maintain suction pressure 8 psia.

 

[BigInch]

Its not really defining tank level unless your pump's suction is located in or very near the tank itself. To calculate NPSH Available, you normally would start with the lowest tank fluid level in the tank (considered as equal to the outlet nozzle height above pump centerline, subtract piping friction loss, head equivalent of the product's vapor pressure (subtract the elevation difference if the pump is higher than the nozzle).

To find a minimum tank lo-lo level pump trip setting, just work backwards from a known NPSH Required, add pipe friction loss, vapor pressure and subtract height to the nozzle, if the nozzle is above the tank.

 

[Artisi]

Suggest you do some reading on pump hydraulcs and pump applications, as a pump manufacturer ?? I would have thought you already understood basic pump technology.

 

[rmw]

Is the fluid you are trying to pump saturated?

To Artisi's point, if you are not part of the Engineering organization of your company, I suggest that you get in touch with them. Someone somewhere designed this pump and knows what its characteristics and requirements are.

Careful - NPSH curves are often developed for pumps using water data and have to be corrected to the actual fluid being pumped.

rmw

 

[Clvet]

NPSHr is the net positive suction head required and a functon of the pump, impeller size, and speed it is running at.

the required is not something that can be easilycalculated. It might be possible to calculate it analytically using computerized fluid flow software however even then, you still need to confirm it with real world testing.

Do you have any kind of performance curve available for the pump that would illustrate the pump's perfomrance(including NPSHr)?

 

[Abhi1712]

To rmw,

Thank you for your reply. This centrifugal pump will be use for crude oil transfer. We just developed pump and I think we have to prepare performance/characteristics curve and NPSHR curve for pump. But since it is going to be used for crude oil i am doubtful to do testing with WATER.

To Clvet,

Thank you for your reply. We don`t have performance curve available. We have to develop it. We are going to use (sell) it for crude oil transfer. Will you please guide me how to develop performance / NPSHr curve for pump?

 

[katmar]

Abhi, it is normal to do performance tests using water because it is usually the easiest and safest fluid available and (more importantly) because the results can be reliably interpreted for other liquids.

Both NPSHr and delivery head are given on pump curves in terms of feet (or metres) of pumped fluid because these are (almost) independent of the liquid SG. If your pump delivery head is 100 ft of water, it will also give 100 ft of hydrocarbon - but of course the delivery pressure (measured in psi or kPa) will be less for the hydrocarbon.

This rule can also be applied to the NPSHr, although strictly the NPSHr (measured in feet) for hydrocarbons is likely to be a bit less than that for water. It is customary to use the water figure and take the difference as a contingency, unless you are trying to wring every last drop of performance from the pump. This does not contradict what BigInch has written above because you need to know what fluid SG was implied in the specification of the NPSHr if it was given in pressure terms (psia) and you want to convert it to head.

It is a bit distressing that an employee of a pump supplier should be asking these questions, although it confirms my belief that purchasers of pumps need to understand in great detail the pump selection process. I suggest you spend some time reading the excellent web sites at

http://www.pump-zone.com

and

http://www.mcnallyinstitute.com

Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[Artisi]

katmar

Correction, that should read pump manufacturer not supplier if we assume that "we are the Pump manufacturer" and "We just developed pump" is in fact what was meant. This makes it even more distressing and once again it is a case of the unknowing trying to re-invent the wheel.

Give it some time a we will probably start seeing posts from unlucky end users who have purchased this equipment and have application / technical problems.

 

[d23]

Abhi1712,

If you were told the minimum intake pressure required is 8 PSI then someone in your company knows the NPSHR. You should get that information from them.

You need to ask if the 8 PSI is PSIA or PSIG. There is a difference between absolute and gauge. Not knowing the gravity of the liquid or if the pressure values are absolute or gauge makes BigInch calculations a little questionable.

You should also ask if the 8 PSI is for fresh water or the liquid (crude) you are pumping. There are NPSH reduction curves available for fresh water verses various other liquids.

Example reduction curve:

http://www.mcnallyinstitute.com/Charts/NPSH_reduction_chart.html

NPSHR can be calculated, but calculations are usually only used as a starting point for NPSH testing purposes. To calculate NPSHR you need to know the suction eye velocity. Most companies do not want to disclose this type of impeller design information. NPSH calculations are difficult and not very accurate.

D23