Things API 610 got wrong (Part 6)

As I started to write this post I checked back on the previous entries in my imaginatively titled "Things API 610 got wrong" series and was shocked to note that more than a whole year had passed since I started on the first one. I'm painfully aware and grateful for being one of the lucky group that could work remotely and as such was largely insulated from the negative effects of the 2020 pandemic. So many were put into a bad situation: loss of employment, illness and death.

If we are to take away a lesson from 2020 I would hope that it would be an appreciation of how much our fates are connected to the well being of each other and that we each strive to be a little kinder and less quick to anger.

In other good news API 610 12th edition has finally been released. Kirit Domadiya has published a nice summary of the changes from 11th to 12th edition and I'd recommend taking the time to review it.

My goal with this series of posts on API 610 is to inform , spark debate and occasionally be accused of entertainment. I will be assured that I've accomplished this if you the reader find this installment more fulfilling and significantly less hazardous than Wombat herding. (Observant readers may note in my blogs that Wombats have a habit of contributing to many hitherto unexplained pump failures. In fact I fully expect that they will eventually have their own chapter in my favourite pump book (one that I highly recommend you own).

I'll warn you in advance this isn't going to be an easy read (even by my optimistic standards of technical density), but hopefully worthy of your time.

Anyway with that onto something less infectious, let's talk about API 610:

Caveat Lector

Before I continue with part 6 of the topic, I'd just like to recognize what API 610 got Right and the tremendous value of API 610 in guiding pump manufacturers and purchasers alike into specifying and building safe, reliable products. What I'm writing about in this series are quibbles with what is 99.9% otherwise an excellent standard.

I'd also like to mention the largely unrecognized Engineers who collectively put in many thousands of hours of time and care into crafting the standard. They are the ones very much deserving of your consideration and praise. A few of them such as Ron Adams, Morg Bruck and Bill Goodman I've been very lucky to work with and have benefited greatly from their wisdom.

What API 610 got Wrong #6

7.4.2 - Outsourcing recommendations on appropriate pump instrumentation to API 670 while failing to ensure API 670 adequately covers the monitoring needs that are specific to pumps.

This clause of API 610 states:

7.4.2.1 If specified, accelerometers shall be supplied, installed and tested in accordance with ANSI/API 670

7.4.2.2 If specified for equipment with hydrodynamic bearings, provision shall be made for mounting two radial-vibration probes in each bearing housing, two axial- position probes at the thrust end of each machine, and a one-event-per-revolution probe in each machine. The purchaser shall specify whether detectors shall be supplied. The detectors and their mounting and calibration shall be supplied, installed and tested in accordance with ANSI/API Std 670.

7.4.2.3 If specified, hydrodynamic thrust and radial bearings shall be fitted with bearing metal detectors. If pressure-lubricated hydrodynamic thrust and radial bearings are supplied with temperature detectors, the detectors and their mounting and calibration shall be supplied, installed and tested in accordance with ANSI/API Std 670.

7.4.2.4 If specified, monitors with cables connecting to vibration, axial-position or temperature detectors shall be supplied, installed and tested in accordance with ANSI/API Std 670.

So in summary this clause of API 610, while it has some specific requirements around monitoring of bearings, directs the reader to API 670 for everything else.This leads us to the main problem, specifically how well does API 670 guide the reader on how to successfully monitor pumps....

Measuring the Wrong Parameters

It is a subject of some concern to me that many of the monitoring schemes for centrifugal pumps critical to plant operation seems to be patterned on a similar scheme to that used for centrifugal compressors. API 670 demonstrates this particular thought process writ large. API 670 views them as being so interchangeable that it uses the same arrangement diagram (Figure H.3 shown below) for both compressors and pumps. This way of thinking can be paraphrased (by me) as:   

Because of course centrifugal compressors are typically much larger, more expensive and more critical pieces of machinery running at higher speeds - so by definition any monitoring scheme that works with centrifugal compressors must easily be good enough for the simpler centrifugal pump - Right ?

Not quite...

The (big) problem with this line of logic is that compressors suffer from issues that pumps generally don't (and vice versa): Rotordynamic issues from high speed operation, surging and fouling are all issues (almost) uniquely seen on centrifugal compressors. However centrifugal pumps have their own unique issues related to cavitation, NPSH and the potential for rapid pumped liquid phase changes in some processes such as light HC and boiler feed (so called Loss Of Suction events). These require a monitoring approach specifically tailored to centrifugal pumps.

(I'd further note that API 670 devotes a whole section of the standard (Section 9), to the compressor specific issue of surge but discusses pump issues such as cavitation only in passing in relation to the expected vibration signature in the Appendix N. (Using vibration monitoring to detect Loss Of Suction events is not appropriate since by then the event has already happened as discussed below).

Specification Gaps = Broken Pumps

I made a post some time back about two failures of pumps during commissioning. The processes, plant, conditions and operators had nothing in common between them. The things they did have in common were:

1. The services were ones where a Loss Of Suction (LOS) event was possible (One service was light hydrocarbons, the other was boiler feedwater)
2. The pumps were a multistage design (BB5 radially split barrel casing)
3. In both cases the pump was critical to the plant operation

In each case, the pumps tripped on sudden high shaft displacement (vibration) which in both cases was the subject of continuous monitoring. Subsequent inspection revealed that the pumps had been starved of liquid and partially run dry with heavy contact at the wear rings. This contact resulted in significant damage to the pump.

On both of the installations we are discussing here, the condition monitoring did not include direct measurement of the fluid conditions at the suction inlet of the pump. This was a significant engineering oversight in the guidance given by API 670 to the plant designers.

At a minimum for this to be a viable monitoring scheme both the pressure and temperature of the fluid at the pump suction would need to be monitored. This is because we need to know the NPSHa of the fluid and how close it is to boiling/flashing if a sudden pressure transient occurred.

Additionally the balance line pressure and temperature should be monitored as this can give good indication of the health of a multistage pump. If the pump is on a service prone to LOS this monitoring becomes critical to ensure that no fluid flashing occurs in the balance line. If it does it can damage the mechanical seal and even result in rotor damage if the Lomakin effect in the balance drum is severely reduced.

I've summarized my recommended scheme in the diagram below for a pump with pressure fed hydrodynamic bearings. The significant differences from API 670 H.3 are the incorporation of elements to monitor the fluid pressure and temperature at various points in the pump and reflects the inherently 2 phase nature of the fluid being handled and the wide range of fluid temperatures that can be present.

That's it for this posting. As always I am very grateful to any of my readers who have made it to this point. If there are any comments, criticisms or questions those too are gratefully received.

Obsesio positivum

#pump #NPSH #API610 #instrumentation #centrifugalpump