Generally, 0.3 and 0.6 ips (8 and 16mm/sec) are considered to be "worry" and "do something" vibration limits for machinery including centrifugal pumps and their drivemotors. I believe the Hydraulics Institute Standard now recognizes the need for higher allowable vibrations for topside measurements on vertically mounted pump-motor sets that vibrate as cantilever beams fixed at the bottom. Nonetheless, your indicated ratio of 4 to 1 between base and motor top vibration levels suggests a cantilever beam that is too flexible over its length of 3.2 meters for comfort of the bearings and maintenance of the motor airgap clearance. You should probably brace the top of the motor or stiffen up the base support structure.

Single vane pump impellers are notoriously hard to balance.  They become out of balance easily due to wear.  At the next maintenance interval, inspect the impeller for wear, and rebalance if needed.

Arkie Boy,

The pump in question is new & has just been commissioned.

Clearances were checked before installation. Pump came with a balancing certificate which checked out OK. Wear is not an issue.

VANSTOJA

Due you have any rules of thumb for acceptable vibration ratios you mentioned (for a pump of this height).

Thanks Guys,

Berko

I don't have the latest HI standard on vertical pump vibration limits but I would guess that the limit isn't increased by a factor more than 1.5, so 0.4 to 0.45 ips might be the worry value with twice that value for do-something.
   Vertical pumps mounted like cantilever beams often have a 1st beam bending mode in the range of synchronous speeds of 30Hz and 60 Hz for 4 and 2 pole motors. You might do well to estimate the 1st cantilever mode frequency using the total weight of the combined pump and motor with weight both concentrated at the motor's center of gravity and distributed along the total beam length and to compare this with actual running speeds including motor slip. If the pump is mounted on a shell-like structure, the shell ID and OD control the beam inertia used to calculate beam frequency. If the pump has feet mounting, try using casing/volute diameters in the inertia equation. For more reliable verification of the cantilever beam bending frequencies, whack the top of the motor in several radial azimuths and record accelerometer responses on a portable noise analyzer using peak-hold mode to capture spectral "bursts". I had a 17,000 lb., 36in, diameter, 120 in. high vertical pump-motor unit mounted on a shell of 1-2in. wall thickness that calculated to be 30-55 Hz for distributed to point loading in the cantilever mode. It ran at both 30 and 60Hz synchronous speeds so the exact location of the beam frequency became very important.

You, as the end user, cannot reliably balance a single vane impeller as you would any other impeller.  Single vane impellers must be balanced, essentially, by an analysis of the forces acting on the impeller at the intended duty point and is done by the manufacturer.

Tim

Berko,
There is an IRD vibration severity chart that has been around for a long time. I think if you do a web search you may be able to get it.
Anyway I have several questions.
Questions:
1) Is the vibration directional? Is it vibrating more in the direction of the discharge piping?
2) What is the primary vibration spike? It at 1X running speed? Is it sub-synchronous? Is it super-synchronous?
3) What balanace grade was it balanced to? Was it balanced to ISO grade 2.5 or 1.0 or API 4W/N?

We have pumps here at our power plant that would occasional get to 27 mm/s . This is not normal but the vibes we noted were related to the cleaning of the travelling screens. As the screens became plugged the vibes went up. The majority of the spike was subsynchronous and not related to the balance of the pump.

A thorough vibration analysis is required and possibily determination of any resonace as per Vanstoga's recommendation.
Regards,
James Chin Yut

James Chin Yut

I recommend you to use ISO 10816 Part 3. There is no reason to allow more level on the motor top. The motor does not know where it sits and cannot adjust its wear and tear to allow higher levels. You should demand as delivery status below 2-3 mm/s rms and for an action to be motivated you should react when above 4-5 after a fairly long period of operation, like the warranty period. We use a standard compiled from ISO which is called SSG3030 in swedish and english language coering balancing quality, installed vibration and resonance rules for all sorts of machinery when commissioned and/or revised. (See also www.ssg.se, a company owned by most swedish paper mills, compiling all sorts of standards). Mail me is you need more info. arnelind@bluebottle.com

Is the pump operating at it s duty point? I have experienced such pumps vibrating excessively due to process conditions not being as per specification. Is the unit fitted with a VSD and perhaps the speed has been changed or it vibrates at a particular point on its curve?

I would also invrestigate the suction conditions as there may well be imbalance on the impeller caused by submersible eddies, swirling etc.

Your problem may not be the pump but the manner in which it is installed or the duty performed.

You need to work with the pump manufacturer re the design of the installation.

I would refer to Pumping Station Design by Sanks for information on station design. The book contains some godd and bad examples. You can download the book by subscribing to www.knovel.com