Pump NPSHA 1

I am a little bit confused also but mostly about your description and the respective elevations.
A diagram showing an elevation view will help. Also, is the pump to withdraw propane or to fill the tank?

Thank you.

Let me clarify,
This is a vertical turbine pump withdrawing propane from an elevated horizontal closed tank.

- Height of the liquid inside the horizontal tank (worst case) = 1'
- Height from the bottom of the horizontal tank to the suction of the pump = 3'10"
- Height from the suction to the bottom impeller = 6'2"

So for calculation NPSHA, the height of the liquid from surface to the center of the impeller = 11 feet.

I am just confused about my NPSHA calculation since it resulted in (-) negative number. The NPSHR from pump manufacturer is 8.4feet in water (old 1966 document that I found). I know that for hydrocarbon, because of the vapor pressure compared to water during pressure drop, the actual NPSH required is less than of water.

If I do the calculation like the above using TVP @ 100F which is 176.1 psig, I would get negative NPSHA number using 140 psig of liquid propane, yet I know for sure that the vertical turbine pump works. The site has average ambient temperature of 65F, but the maximum during summer can reach to 100F.

Calc: 154.7*2.31/0.50 + 11 - 3.4 + 190.8*2.31/0.5 = -159 ft
Did I do my calculation incorrectly?

I think i found the answer to my questions.


Thank you all! Just use the search button *doh*

For propane from a tank as you've described, vapor pressure = system pressure so they cancel each other out when calculating NPSHA. Negative NPSHA isn't possible because it would mean the system pressure at the pump impeller would be less than bubble point pressure and you'd actually have a boiling mixture.

NPSHA is simply the suction pressure ABOVE the fluid's vapor pressure at the pump suction in feet of fluid. Re-arranging the 'classic' NPSHA equation shows this more clearly.

Assuming the 8.4 feet NPSHR is at the first impeller elevation and your dimensions, I agree with your 11 feet elevation. Given your line loss value of 3.4 feet, you have NPSHA of 8.6 feet, marginally above what is required. Since the level in the propane drum is not usually at the minimum level, you will have more NPSHA than this though the margin between NPSHA and NPSHR is marginal at minimum level.

For vertical pumps, I prefer to specify the NPSHA at the inlet suction flange and then let the manufacturer determine how much elevation below the suction nozzle they need to provide given the NPSHR and the line losses through their inlet piping. Are those losses included in your 3.4 ft estimate of the suction line losses?

From your first post:
* Surface pressure = 140 psig
* Vapor pressure = 176.1 psig

How is it that your surface pressure is less than the vapor pressure?

Also, the NPSHR in the file should not be in terms of "feet of water." A statement of NPSH is in terms of the height of a column of the fluid being pumped.

The specific gravity is used with your 2.31 constant when converting psi to feet of fluid, but NPSH values are not necessarily in terms of feet of water.

TD2K,
No, I did not account for the line losses through their inlet piping. The 3.4' is the line losses from piping length and the elbows only.

Wilbur55,
According to operator the pressure of the vessel varies between 120-160 psig. I am using average of 140 psig.

I think my error was on the TVP, the values of the surface fluid pressure and TVP should cancel each other (saturated liquid) in closed tank. I wasn't understanding this correctly.

But, I read that typically pump manufacturer got the NPSH# using cold water as the pumped liquid. So the NPSHR# for water and propane in this case using the same pump would be different because of their SG no?

The information from the operator doesn't square with your evaluation of the vapor pressure at 100°F for "(saturated liquid) in closed tank."

You may be using a higher temperature than exists, or something is amiss in the information from the operator: the gauge is wrong, at some other location, in different units, etc.

Again: How is it that your surface pressure is less than the vapor pressure?

How do you reconcile "the operator says so" with the physical properties of the material involved?

The 100F TVP value is used from what I found in the lab analysis of the propane that is available at the plant.

Again: How is it that your surface pressure is less than the vapor pressure? The reason is that I did not understand TVP definition clearly. After searching the forum and looking back at my book, I think I get it now. For saturated liquid in closed tank, the surface pressure and the TVP is matching one another.

How do I reconcile "the operator says so" with the physical properties of the material involves? I was wrongly using arbitrary number of average surface pressure. If I am using 140 psig surface pressure, my corresponding TVP should be at 140 psig as well, because TVP is based on temperature only, and I believe the tank is as well.

Manufacturers do test their pumps using cold water because it is convenient to do so.

Centrifugal pumps are "constant head" machines.

Neglecting the effects of viscosity and variations in pipe friction, the pump doesn't care what fluid is being conveyed. A pump connected to 30 feet of vertical pipe may send 100 gpm of any fluid (water, gasoline, mercury) out the top of the pipe. The motor horsepower will change with fluid density, but not TDH or NPSHR.

See:

No, NPSHA is feet of liquid, not feet of cold water. If your NPSHA is 8.4 feet, you don't need to supply 2*8.4 feet to compensate for the density difference between water and propane (which is approximately 2:1).

Getting information from the operator and the lab are necessary. I suggest continued dialog with them. However, implicit trust in their answer to the question they thought you asked them could be problematic.

Getting the "lab analysis of the propane that is available at the plant" may be essential if they are using some mixture of gases. For the pure substance, the lab numbers should be comparable to values in published tables.

The vapor pressure stated for propane at 100°F is comparable to a value from the graph here:

I trust the lab's number.

I don't trust the operating pressure of 140 psi at 100°F for a closed tank of pure propane.

Propane at 100 F on that chart shows to be 180 psia - 14.5 = 165.5 psig



Since you basically start out at vapor pressure at the tank, at minimum tank levels and maximum temperature it is common to find 2-phase flow in the suction line to the pump and no NPSH at the entrance to the canned pump. You are using a canned pump to try to build up the suction pressure (and NPSHa) by providing a level of head of fluid contained in the can itself.

From "BigInch's Extremely simple theory of everything."

Thank you. In the future, I should be careful to implicit trust to the answer from the operator without matching their answer with given data.


Thank you everyone!

Slightly off topic,

In the old document I found this under performance:

NPSH REQ'D (WATER), FT: 8.4' @ IMPELER EYE

and when I contacted this old pump mfq, they generated a new computer graph with NPSHR Required which is graphed at 8.4ft as well approximately.

I read the McNally Institute and came to this Chart Reduction

"NPSHR (net positive suction head required) is defined as the NPSH at which the pump total head (first stage head in multi stage pumps) has decreased by three percent (3%) due to low suction head and resultant cavitation within the pump. This number is shown on your pump curve, but it is going to be too low if you are pumping hydrocarbon liquids or hot water."

How low??

Per the chart, I have maximum NPSHR reduction > 10ft at 100F???

Probably too low more so for hot water than for hydrocarbon liquids, where you can typically get away with a little less, but that is not the case with the high vapor pressure hydrocarbons such as LPGs where you would like to be conservative, but due to the high vp, you just simply can't get any more anyway. To cover the outside edge of the envelope of where cavitation can begin, values of 20% more NPSHr than shown by the cold water curve may be needed, especially in cases of hot water.

From "BigInch's Extremely simple theory of everything."

Since you have difference sets of elevation values in your OP and your first reply, I am attaching of a JPEG of a general analysis showing tank and pump setup and the application of the first law of thermo. I don't think that you fully understand where the NPSH comes about so, look at the elevation sketch of the energies invovlved.

Maybe I've read something wrong, but just to be sure:

If the pump curve says it has an NPSHR of 8.4 feet of cold water at your flow, then you MUST divide that 8.4 by the specific gravity of fluid to get NPSHR value for pumped fluid.

In your case, 8.4/.5 = 16.8 feet of propane required; not the 8.4 as I think I saw above. SOme people find it easier to convert everything to PSI to remove possible confusion.

A couple of other thoughts:
you say this pump has been running 40 years; then surely it has been rebuilt several times. Make certain the documents you are working with reflect the pump in the hole exactly. (There is a chance the first stage impeller could have been switched out to a lower NPSHR model, maybe the pump length has been changed?)

Also, has the flow changed over the years? If you are actually running at a different flow, the NPSHR value would be different than the originally spec'd.

This has been the first interesting post in quite a while, thanks Techdrone.

The NPSHR was determined to be ~8 feet of cold water when cold water was the pumped fluid. The NPSHR will be ~8 feet of liquid propane when the pumped fluid is liquid propane. This is before the correction for the caveat that the NPSHR was measured when the pump head was diminished by cavitation during the test, as mentioned by in the later cite by the OP.

NPSHr is not a measure of distance, height, or pressure. NPSH is a measure of total energy required to smoothly enter the pump. 8.4 ft of total energy, potential energy + kinetic energy (the kinetic component is small and usually ignored) is 8.4 feet for gasoline, 8.4 feet for diesel, 8.4 feet for cold water, 8.4 feet for hot water, 8.4 feet for propane and 8.4 feet for a ROCK. Although you might need more than 8.4 feet of hot water head to prevent cavitation of that fluid.

8.4 feet of head (energy) for water = 8.4 * 62.4/144 = 3.61 psi
8.4 feet of head (energy) for diesel with SG of 0.835 = 8.4 * 62.4/144 * 0.835 = 3.02 psi The pressure for water and diesel are different, that's why the NPSHr curve is not given in pressure units. If that were true, the pump manufacturers would confuse all of us engineers out here by making us do those conversions you want to do every time we changed products going through the pump. It is much easier to specify energy required and let that apply to each and every liquid and simply let pressure be whatever pressure has to be.

You must convert a fluid's vapor pressure to head to calculate NPSHa, remember the Pressure/Density term from Bernoulli, but that's where it stops.

From "BigInch's Extremely simple theory of everything."