Vibration in Vertical Pumps

I found the following article in EC&M Magazine to be of interest:

http://ecmweb.com/motors/solving-motor-vibration-problems-vertical-pumps

01-17-2014 05:46 PM

Top #2

Mike Trumbature

01-17-2014 05:46 PM

Assuming the pump is balanced here are some additional problems mentioned earlier in another group and here also.

VFDs, see last paragraph link SHAFT CRITICAL SPEEDS - https://docs.google.com/document/d/12f-bnSyaRxBtOxAvdLYZd-XYAo8FcAHxoCwmMpI5of0/edit
Still not sure how the block this on the variable speed engines?

Foundation problems - including the pump not being plumb especially with hollow shaft motors (rough on the thrust bearing) - https://www.linkedin.com/groups/VERTICAL-FIRE-PUMP-VIBRATION-PROBLEMS-144875.S.100161948?trk=group_search_item_list-0-b-ttl&goback=.gna_144875

Up thrust at high flows - especially hollow-shaft motors, when the force of the water pushing upwards on the impellers exceeds the weight of the shaft and impellers and down-thrust from the discharge pressure. Running the pump too far to the right of its curve in shallow pits with limited shaft weight can really make them jump up and down.. Here is one that vibrates so bad it kept breaking the tubing off of the pilot regulator on the main relief valve so the put SS tubing on it. See note #4
https://www.box.com/s/ak14tez5ie5a668lisgm

Here is more on cavitation vibration and diesel fire pumps like this one above
https://www.linkedin.com/groups/CAVITATION-VIBRATION-144875.S.101342720?trk=group_search_item_list-0-b-ttl&goback=.gna_144875

Vibration analysis is typically done with a seismic detector, just like earthquake detectors. Piezoelectric pick-ups are used for higher frequencies found on turbines and gear boxes (gear mesh frequencies) like right angle gear drives and speed increasers etc.
They connect to an analyzer like a radio with a tuning filter to monitor a specific frequency. These frequencies are multiples of RPMs and associated with various problems that cause them i.e 1 X RPM - out of balance, 2 - 4 X RPMs looseness, to misalignment, 7 X RPM, vane passing frequencies - flow related problems, 13 X RPMs - anti-friction bearings, you can even detect a cracked bearing race - "sideband chatter" You can pick up "standing waves" on the piping which can also resonate like a giant tuning fork with a long unsupported discharge header on multiple fire pumps. The "wave length" is usually vane passing frequency and the speed of sound through the liquid adjusted for pressure. You can see the high and low spots on the pipe with a vibration analyzer.

I occasionally worked with this unit in the 70's with Coastal States Pipeline
http://www.ebay.com/sch/sis.html?_nkw=IRD+350+VIBRATION+ANALYZER+DYNAMIC+BALANCER+
and quite frequently with its little brother
http://www.ebay.com/itm/IRD-Self-Powered-Portable-Vibration-Meter-Model-305-/160883340844

Many multiple fire pump wet pit installations will have more foundation and piping problems than a single well pump on solid ground with the hollow "drum lid" foundations and loose piping everywhere. That only aggravates the all the above problems. Long pipe runs of unsupported discharge piping on multiple pumps like 5 CAT 3412s are a major problem when on pump is .6 IPS and the one next to it is about half of that ant not even running. Typically about .3 makes people really nervous. The range of most seismic detectors is about 600 - 80,000 CPM
http://www.macanalysis.ca/vibration_training_manuals/velocity_vibration_severity_chart.php

Worn line shaft bearings, especially on a deep set pump, will "rock and roll."

01-17-2014 07:50 PM

Top #3

Mike Trumbature

01-17-2014 07:50 PM

Many process pumps, horizontal and vertical, have "machinery monitors" like this to monitor vibration on both axis, as well as motor winding and bearing temperatures, oil temperatures. It can even monitor shaft movement, like sleeve bearing to shaft or upthrust on hollow shaft motors down to .0001" This unit monitors sleeve bearing clearances on X and Y axis on an HSC pump. There is no vibration tuning filter on here.
https://www.box.com/s/lfs600ny1b4nrm9ve3je

It's a great article on how to solve the problems. First you need to determine what they are, hence the vibration analysis. You can either rent the equipment yourself or have a service some out and do it for you.

01-17-2014 10:17 PM

Top #4

Mike Trumbature

01-17-2014 10:17 PM

Jim,

It looks like the variable speed engine issue with shaft critical speed lockouts is falling through the cracks. I don't think anyone here knows, if they do, they have not said one way or the the other.

01-18-2014 12:53 AM

Top #5

Debra Ingram

01-18-2014 12:53 AM

On a vertical pump, you touched on something earlier of shaft criticials but you didn't quite take it far enough. The majority of the problems that I find on vertical pumps being driven by variable speed is the bearing spacing. Most fire pumps have 10 foot bearing centers. You will almost 98% of the time hit a critical with 10 foot bearing centers. If you are running a vertical turbine with a variable speed device, always, always, always specify that the bearing centers do not exceed 60". Assuming that the foundation is solid and there are no natural frequencies that haven't been "dialed" out, you should be in good shape.

01-18-2014 03:28 AM

Top #6

Mike Trumbature

01-18-2014 03:28 AM

Obviously, the more supports you have, the better you can hold down a "whipping rope" and eliminate the "bend" in the middle.

The shaft critical speeds are calculations based on several factors including bearing bearing supports. http://en.wikipedia.org/wiki/Critical_speed

That info is required to be supplied by the pump people to the VFD controller people per NFPA 20, 10.10 .12.1 to pre-program these "skip" speed settings. If additional bearing supports are required for VFD operation, then the pump people need to put them in and not expect the customer to have enough smarts to request them. I presume the largest majority of these units are still being sold as a complete package by the pump OEMs so they have the knowledge and control (unit responsibility) of the situation to make sure everything works correctly?

You can lock those frequencies out with a VFD. What can you do about a variable speed diesel????

01-18-2014 06:08 AM

Top #7

Debra Ingram

01-18-2014 06:08 AM

I agree that the pump OEM still controls what goes into the design and selection and they should be held accountable for not taking into consideration of the bearing spacing. You can lock out the frequency that is bad on a standard VFD. On the variable speed engines, I haven't worked with them enough to know so this may be better answered by our friends at Clark; Justin Strouss and team are an excellent resource on this. Even if there isn't a way to lock out the critical frequency on the engine, it can still be avoided by going with 5' max bearing centers on the pump. If you have harmonics, this will need to be worked out in the field. I've seen isolators used under the feet and with the soleplate. I strongly recommend a soleplate be sold with the vertical turbine pump as standard even if it isn't specified. Again, the pump vendor is the expert in what their equipment requires to operate and also be per HI.

01-18-2014 08:31 AM

Top #8

Mike Trumbature

01-18-2014 08:31 AM

Looking at the "controls" on the governor of one engine unit, it looks like a pilot regulator controlling a cylinder like the controls for diaphragm in a PRV. Instead of connected to a valve, the positioner is connected to the governor.

See previous discussion referenced above.
https://www.linkedin.com/groups/VARIABLE-SPEED-ENGINES-SHAFT-CRITICAL-144875.S.162327272?trk=group_search_item_list-0-b-ttl&goback=.gna_144875

I see no one took into account HSC pump shaft critical speeds like this one
http://www.peerlessxnet.com/documents/B-1300.pdf

NFPA 20 recently addressed verticals only and VFDs, not the engines. Having been in the vibration analysis business, I've seen rotating assys. just like it "wobble" all over the place through three speeds from 1800 RPMs down on a Schenk "soft-pedestal" (leaf-spring supports) balancing machine" Most machines are "rigid pedestal" balancers and can only be seen on the analyzer meter.

01-18-2014 10:45 AM

Top #9

Debra Ingram

01-18-2014 10:45 AM

Typically the critical speeds are much higher in a horizontal pump and not as big of an issue as they are with a vertical turbine, but yes they do need to be examined during the pump selection process. Peerless and Patterson publish their critical speeds so it takes the question out of the process when selecting one of their pumps.

You also don't want to use an insert type coupling since they can back out when the speeds are going up and down. My preference would be a Falk type or something similar so that the inset can't be thrown out of the coupling during operation.

01-18-2014 01:29 PM

Top #10

Mike Trumbature

01-18-2014 01:29 PM

VFDs have been out for some time. It appears shaft critical speed lockout was added in the 2010 NFPA 20.

I wonder what happened to all the ones before that? Anyone have any ideas?

I was aware of shaft critical speeds back in the 70's on pipelines from vibration analysis school and variable speed controls on diesel engines with PID controllers and electronic governors.

01-18-2014 03:30 PM

Top #11

Mike Trumbature

01-18-2014 03:30 PM

Speaking of coupling problems -
https://www.box.com/s/84bcc95b227c778ddeb8

Other than the greasy mess, I'm a Falk SteelFlex fan for reliability. They can't "separate" like jaw and flange couplings

#2 already threw out one insert and this one is not far behind it. I pulled the front of a CAT 3408 over 3/8" to straighten it out. One advantage of the jaw type on old units like #1 - you loose a a spider, you still have a "coupling" On flange couplings, you loose an insert in the middle of a fire, you are screwed, blued and tattooed.