witness pump test - items to watch for

[someguy79]

I'm planning for a witnessed pump test at the factory. I haven't done a witnessed test before, so I would like to get some opinions on what to watch out for during this process. I don't expect the manufactuer or mfr's representative to do anything unscrupulous, but it's always possible, and so is the unexpected.

The application is a 35,000 gpm, centrifugal pump with 600 hp motor.
The motor used for testing will be the job motor (the one to be installed at my plant).
The test fluid will be water.
In service it will pump raw sewage.
The specification calls for testing to be per HI 14.6.

I would like to know if anyone has suggestions on things that must get my attention, or has had experiences where testing could've/should've been better.

Thanks in advance for any ideas and info.

 

[EdStainless]

Do you have a copy of his test plan?
Copies of his internal QA manual that cover testing? Sample documents (taken at test and final report)?
Plan on showing up the day before to review manufacturing documentation and familiarize yourself with their test area and equipment.
Check calibrations on pressure, flow, and temp devices.

You are not trying to catch him doing anything wrong. You need to know enough to explain the entire process to people back at your facility. And if, heaven forbid, something isn't right when you install it you will have insight into how it was tested and what may be wrong.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[LittleInch]

The only one to really watch is the NPSH test.

The temporary pipework they often use and the valves often allow air into the system when they start getting below atmospheric and hence the test goes a bit wonky.

Also you don't ever find the cavitation limit.

Other than that it's a fairly simple laid down procedure about testing at various points along you pump curve and when they get to a point often the tech will say - are you happy with that? and then record and move on.

So get a copy of the test procedure - read it a couple of times BEFORE the test and just tick it off as you go along.

Just watch for the water temperature going up as the volume they use is often quite low and as they throttle it all away back into a holding pond it can warm up over the test.

All pretty automated and signal driven nowadays, but on the soak test look out for small leaks, seal issues and strange noises....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Artisi]

First up, get a copy of the test standard , read it and understand it.
What are you testing, is there a guaranteed duty point in terms of flow, head, NSPH and efficiency or is it a general test to establish a general pump performance over a range of duties.
Normally a pump manufacturer will guarantee one (1) DUTY POINT ONLY, although they will or should produce a performance curve over a wider range of greater flow and a lesser flow.
Things to check, if you wish to be tough are:
Pressure gauge calibration certificates or in the case of a major pump company they will usually have gauge calibrating equipment and calibrate or test prior to and after the test.
How is flow being measured, orifice plate, Doppler or turbine meter - check calibration certificates.
How is power being measured, check voltage, check certificates.
As you are running job motor, speed needs to be checked and compared to pump speed selected for the guaranteed duty, necessary corrections will need to be made if speeds are different.
Check atmospheric pressure, water temp.

Run the test, if one point is guaranteed the test technician will concentrate around this point and should take a number of test points and average out the results.
If the duty point achieved in terms of flow/head is over it will be your decision to accept or reject provided efficiency and NSPH are correct. If the duty point and efficiency are not achieved, the standard will nominate what are acceptable tolerances. If you accepted the test code as the means of proving performance then you are locked into accepting the tolerances - there is no argument and the supplier has fulfilled the contract.

However, if under tolerance there may need to be some rework on the impeller - but this is another story that can be addressed here later.

One important aspect you need to be aware of, if the pump test is acceptable to you as tested in the test rig and there is a short-fall of performance on site, this is your problem not the pump and not the manufacturer.

Lots of luck with an always boring pump test.

PS. forgot to add - vibration and noise testing if guaranteed should be by mutual agreement.
Bearing temp. rise should be recorded, expect it to be a bit higher than what would be normal until such times as the pump has run for a few hours.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[EdwardDrg]

I once worked in a pump manufacture factory.
If you want to witness the pump test, pay attention to:
a. test pump capacity is several points, including the lowest and the highest.
b. when pump in these capacity condition, record the pump discharge pressure and power.
c. finnally you will have the test performance curve. this is the key for the test report.
e. test time should be long enough.

Hope these information is useful for you.


specialized in marine pump engineering

http://silipump.com