Evil Math & NPSH margin

Evil Math & NPSH margin

Ok maybe not so evil, but I was struggling for a catchy title....

In my previous post I talked about Hydraulic Institute (HI) 9.6.1 (2017) and how it can help guide to selecting the appropriate NPSH margin for your pump application.

However some applications are not covered by the standard (such as water injection). Additionally I know some people might want a more math based approach to margin determination. So fear not. If you are one of those people who enjoy some non-evil math, read on.

The formula I'm going to cover is a simplification of the one we use internally. I've tried to spare you the reader the need to wade through several different charts and extrapolate between them. If you are concerned that your application falls outside the limits of the formula, feel free to contact me and I can give you more specific guidance.

So onto the formula, which is applicable for pumps pumping water and operating in the POR (Preferred Operating Range of 70% to 120% of BEP) with a Nss of 11,000 US (231 metric) or less. It requires that you know 3 pieces of information:

  1. The eye diameter of the impeller (D1 in inches, see the diagram below)
  2. The rotational speed of the pump N in RPM
  3. The NPSH3 (NPSHr) in ft. of the pump at the operating point of interest


No alt text provided for this image

Note that MNPSH is expressed as a non dimensional number.

No alt text provided for this image


The formula reflects the general sensitivity of pumps to the velocity at the impeller eye. It is well established that pumps with very high eye velocities ( > 80 ft/s, 24 m/s) require significant NPSH margins. ( I'd encourage you to try the different eye velocity values and see the effect on recommended margin).

Ok so now for a worked example to show you the formula in action:

The example is a pump @ 3560 RPM with an eye diameter (D1) of 7 inches (178mm) and a NPSHr of 26 ft. (7.9m)

u1 = π * 3560 * 7 / 720 = 109 ft./s

Putting this into the formula gives:

No alt text provided for this image


So in this application the required NPSH margin is 1.91 and the hence the recommended NPSHa is 1.91 * 26 = 49.7 ft.

I'd encourage you to take a look at any problem pumps you have in your installation and evaluate them where appropriate using this formula or HI 9.6.1. You may find it a useful tool in judging the quality of the installation and help guide as to whether insufficient NPSH margin is a contributor to the pump's reliability.

Any feedback is as always gratefully received.

Beatus Centrifuga

Thanks!, Simon Bradshaw what needs to be change to apply this formula for refined hidrocarbons such as Diesel and for crude oil?

Like
Reply

Hi simon, just saw your comment with a link to this article. Thank you for that. Just a question -- what if 2 pumps are connected in series. How could we calculate if cavitation will occur on the second pump? Would a turbulent flow on the impeller eye cause cavitation?

Like
Reply

Simon, two more questions: 1.What do you think about defining operation zones for Crude Oil pipeline pumps (BB3), according to Process Industry Practice PIP RECP001 (Nss vs. %BEP)? 2. Is it possible to use your formula for viscous products (crude oils)? What is it needed to use with crude oil? Thanks!

Like
Reply

Simon, How I can to apply the NPSH margin formula for a double-entry impeller?

Like
Reply

Interesting many thanks

Like
Reply

To view or add a comment, sign in

More articles by Simon Bradshaw

  • Testing Centrifugal Pumps Part 4

    So I'll admit I didn't expect to be writing a Part 4 of this series. However here we are.

    15 Comments
  • Testing Centrifugal Pumps Part 3

    Welcome to 2026. If 2025 was the year when AI exploded in media mentions, hyperbole and usage for generating animated…

    10 Comments
  • Lowering Vane Pass Vibrations (Part 2)

    Welcome to Part 2 of the article series on how to effectively manage vane pass vibrations. If you've arrived here…

    13 Comments
  • Lowering Vane Pass Vibrations (Part 1)

    One of the ongoing debates when working on any pump design standard (such as API 610), is how prescriptive the standard…

    29 Comments
  • Edward “Ed” Smylie: The Quiet Engineer Who Helped Save Apollo 13

    In a sharp deviation from your normal and expected programming regarding pumps, I wanted to remember an Engineer who…

    3 Comments
  • Standards & Margins - a key to reliable pumps

    Welcome to the first Pumping Corner of 2025. I want to take a moment to thank each of you for subscribing and being…

    30 Comments
  • Pump piping Forces & Moments - how much is too much ?

    Truth be told I'm highly guilty of spending excessive late night hours randomly browsing the interweb and then zooming…

    20 Comments
  • Testing Centrifugal Pumps Part 2

    Welcome to Part 2 of this indeterminate length series into the cornucopia of curiosities that relate to the testing of…

    12 Comments
  • Testing Centrifugal Pumps Part 1

    So it's been a while since I last visited the topic of testing pumps. While I've covered specific aspects of testing in…

    24 Comments
  • Things API 610 got wrong Part 9

    After the explosion of commercialism and over indulgence wrapped in crinkly paper that marks the end of the year in…

    22 Comments

Others also viewed

Explore content categories