DEAR ALL,

THERE HAS BEEN A FREQUENT PROBLEM OCCURRING IN 3 OUT OF 4 OF THE VERTICAL SEA WATER PUMPS WHEREIN THE VIBRATION OF THE MOTOR INCREASES WITH RESPECT TO TIDE.
THE PUMP IS WET VERTICAL DOUBLE SUCTION PUMP WITH DISCHARGE THROUGH THE COULMN PIPE. DRIVEN BY INDUCITONMOTOR CONSTANT SPEED - 1495 RPM.
THE VIBRATION INCREASES TWO TIMES IN A DAY DURING THE MORNING TIDE AND THE NIGHT TIDE.
THE VIBRATION IS MONITORED THROUGH THE BENTLY SEISMIC VELOCITY PROBE IN THE MOTOR NDE.
IT VARIES FROM 0.2 IN/S AT LOWEST TIDE TO 0.65 IN/S AT THE HIGHEST TIDE.

SEEMS LIKE IT MAY BE RESONANCE DUE CHANGE IN THE SYSTEM STIFFNESS DUE TO VARYING IMMERSION. YOUR VIEWS ON THIS PLEASE.

THREE PUMPS ARE IN PARALLEL - THERE IS NOT MUCH OF CROSS TALK BETWEEN THE PUMPS.
BEATING IS NEAR IMPOSSIBLE AS THE VARIATION DURATION IS AS LONG AS 8 TO 12 HOURS.

ONE OBSERVATION MADE IN TWO OF THE PUMPS IS THAT THIS PHENOMENON WAS OBSERVED AFTER THE PUMP OVERHAUL. THOUGH ONE PUMP IS AN EXEPTION - AS IT SHOWED NO SUCH CHANGE AFTER OVERHAUL.

DURING OVERHAUL ACTIVITIES THAT ARE ALWAYS DONE ARE
IMPELLER INSPECTION /REPLACEMENT
WEARINGS RENEWAL.
LINE SLEEVES AND BUSHES REPLACEMENT -(BUSHES ARE RUBBER LINED)
STUFFING BOX PACKING RENEWAL
AS THE SERVICE IS SEA WATER THE WETTED PARTS ARE OF AL BRONZE CONSTRUCTION, SHAFT OF MONEL.THE DISSIMILAR WELD JOINT BETWEEN THE PUMP BASE AND DISCHARGE ELBOW USUALLY GETS CRACKED AND WELD REPAIR IS DONE. THE DISTORTION AS A RESULT OF THE REPAIR IS CORRECTED BY MACHINING THE BASE (WHERE THE COLUMN PIPE IS LOCATED) AND THE MOTOR STOOL SEATING AREA ENSURING THEIR PARALLELISM AND CONCENTRICITY OF THEIR FACES AND BORES RESPECTIVELY.

NOW THE QUESTION IS SINCE IT IS LIKELY THAT OVERHAUL IS CAUSING THIS VIBRATION PROBLEM WHAT COULD HAVE GONE WRONG DURING OVERHUAL.

THE I X COMPONENT IS DOMINANT AND IT IS THIS 1 X THAT CHANGES (INCREASES) WITH THE TIDE LEVEL.

INSTALLATION OF A DYNAMIC ABSORBER TUNED AT 1 x DID NOTHELP MUCH.

WHAT CAN HAPPEN IF THE VERTICALITY OF THE PUMP IS SLIGHTLY OFF VERTICAL BY A FEW DEGREES.WILL THIS MAKE THE PUMP MOTOR SYSTEM SENSITIVE TO TIDAL VARIATIONS.

REGARDS

RAO

Can you check the foundation bolts. Check for any loose bolts.

If possible, send us the picture of your pumps.

I have a similar issue with our Raw Water Supply Pumps.

Zee,
The foundation bolts have been checked. Found intact.
Currently I dont have the picture. Shall attach one when I get it.

Thanks

Rao

I found one .

Note that the Discharge head is not of bolted design but welded construction. Which means that there are dissimilar joints as shown in the attachment.
Other difference is the pump shaft is not enclosed lubricated design and pump has a thrust bearing 72 series and driven through Metastream coupling MOD series.

Pump_CS.doc (124 Kb, 81 downloads)

Did u suspect imbalalance of the shaft at any point?

Please share spectrums and waterfalls if you can.

Maybe high tides is unbalancing the shaft. Its only a guess at the moment. Please send some datt.

One thing I missed out inmy first note
-
During overhauls the impellers are dynamically balanced.

Motor solo runs have been taken and the vibration levels are as low as 0.03 in/s.

Your point that the impeller goes into imbalance whent the tide level increases is not convincing - please elaborate.

Rao

As I mentioned that it was a wild guess.

If the impellers are imbalance and during the high tides, the imbalance force increases, which causes the high vibration as you mentioned.

You can ask the overhaul shop about the balancing results.

If the vibrations are directly proportional to immersion level, change in flexibility and distortion of the column pipe plays the role.

Rao,

Different stiffness at highest vs. lowest tide is a good working hypothesis. Why not do a bump test for both cases? Also you may possibly see damping change instead of stiffness change.

David

This message has been edited. Last edited by: David_G,

Just jumpin' in!!!

Have pump curve and all values. Design and check load on bearings. NPSH may have an influence operating outside design parameters affecting bearing loading.

Ensure setup is correct.

Mr David,

It is 450 kw motor. and there is lot of back ground vibration/ noise. We once attempted to do an impact test to no avail.

One question - In constant speed machines where in the system stiffness changes and pushes it close to resonance do Dynamic absorbers work?

When the dynamic absorber was installed for the current problem we could notice the absorber vibrating violently indicating it was well tuned at 1490 rpm(the motor rpm) . But i could not conclude whether it was doing the job - of creating a node .

We abandoned the whole exercise for the risk of pushing the machine futher into resonance.

Your views and experience could shed some light on this.

I am attaching a trend to give you an idea of the problem


Regards

Rao

This message has been edited. Last edited by: Ramesh Rao,

Trend.doc (58 Kb, 43 downloads)

what kind of line shaft bushings are used in the rebuilt pumps?

Bushings are rubber lined

Rao,

If you do a dual channel impact test with an instrumented hammer and right tip you'll likely succeed even in presence of noise; the correlation factor will tell you if it is a good test.

David

I agree with Sam. First check discharge and pr during tide and confirm that it lies between 70 % to 120% of the BEP discharge. If not, vibration may increase tremendously. API 610 permits operation of pump between these limits only.
Regards
Irshad Akhtar

Akthar,

The BEPflow is about 2000 cum/hr and the pumps are operated at around 1650 cum/hr during 3 pumpparallel operation and at 2100 during two pump parallel operation.
The figures meet the API requirement.
But the question is for tidal variation of 1.5 m there is such a big change in vibration -
The improvement of 1.5 m in the suction condition changes the operating point by only about 100 - `120ncum/hr as per the curve.

Regards

Rao

If the vibration increases at high tide, then the motor-pump structure has a natural frequency just above service speed (1xSS) when the Tide is low. High tide increases the water level around the pump column and thereby increases mass. The high mass lowers the Natural frequency. A tuned absorber (dynamic mass) is not effective because the natural frequency of the structure is changing. I have used a rigid mass attached to the top of the motor to reduce natural frequency well below 1xSS.

Why was the natural frequency test ineffective? Vertical pumps typically have two closely spaced frequencies for each Vibration mode due to the asymmetric structure.
Walt

Thanks Mr. Walt,

I have not tried with an instrumented hammer.

Shall try it soon.

Rao

Rao,
The trend data is over 2 years old - I am not sure exactly what you consider to be a problem. Are you experiencing failures as a result of the high vibration, or are you concerned at the high levels that occur twice each day?

These pumps are typically 2.3/3.5m high and 0.65 in/sec is nothing to worry about when measured at the top on the motor.

If you are concerned about vibration levels, then maybe you could consider a simple single plane balance on the motor fan - drill a hole at the end of each blade and take a once-per-turn reference pulse from the exposed coupling area. You would need to check that phase angle is repeatable at the times of each period of high vibration. Small nuts bolts and washers fitted to the predrilled holes should effect a good result.

Such a balance will not stop the variation in vibration, but it will reduce the maximum levels

Something else to consider: is the motor load different at the times of high and low vibration? If you are running close to a resonance, maybe load induced vibration (thermal unbalance)is changing the excitation force causing vibration to change in the way that you have seem. You would need to check current draw and exact speed at points of high and low vibration to see if this is the case.
If this is the case, then a single plane balance of the fan would reduce vibration.

And finally, the cynic in me wonders who is doing the weld repair and machining - is it the OEM or a local fab shop? What measures are in place to ensure that that work is properly carried out - who writes the repair spec and checks dimensions afterwards?

Thanks Buzz,

The trend as i had mentioned was only to explain the nature of the problem. I was posting from my residence wherein I found the trend in one of the files and posted the same.

The problem is current and have been observed in majority of the pumps. (though initially it was first observed about TWO years back in one of the pumps).

Your views on balancing to bring things under control is noted.

The load on the motor is steady. The cooling water demand is steady and seldom changes.

Repairs are done in the local fabrication shop with OEMs supervision.
Joint inspection with OEM and us is done for final dimensional checks.

Rao