Repeated Feed Pump Bearing Failures 2

[jsalas1]

Hi all, longtime lurker, first time poster.

I have a situation which is made more difficult in that I am getting only word of mouth from two middle-men, but I will try to be as detailed (or concise) as possible.

We have a customer who has two feed pumps, both KSB HGC 3/10 (size 3 pump, 10 stages). Both pumps meet rate but one of the pumps (lets call it the south pump) continually wipes bearings and has had multiple occasions where the balance disk has failed. The other pump (call it north pump) runs without issues. My shop repaired both pumps so I'm inclined to believe that we can get to the bottom of this since one of the repairs was successful.

Both pumps are motor driven and coupled via Voith variable speed fluid coupling. The pump cases are center-line mounted and the fluid coupling is floor mounted, which I think is one issue. The techs are aligning the pump and motor under a cold condition. My gut tells me they should align it, run the pump to warm up both the case and coupling, and then re-align it. I don't know much about these fluid couplings so they may be more forgiving than that but it seems unlikely that a boiler feed pump could run hot when it is aligned cold. In addition, the techs are starting the pump with the fluid coupling 85% engaged, which means it spins right up and I wish they would just get rid of the dang thing (the fluid coupling not the tech).

Furthermore, the techs recently found that the north pump has dowels which encourages the case to grow away from the driver. These dowels are on the drive-end of the pump at the centerline. The south pump does not have these dowels, which I think is obviously an issue. By not doweling the south pump, I would think the case is allowed to grow axially in both directions instead of favoring one, which I think could work in concert with the fact that the train is cold aligned. I don't like either of those things and yet somehow that is procedure.

The last thing and the reason I posted this thread is that the bearing lube oil is discolored and eventually both bearings will get wiped (drive-end goes first). These pumps have a common bearing design that can be run in multiple configurations (pressure fed, oil rings, flooded housing), but these pumps both use oil rings. On both pumps, the bearing temperatures (drive-end and non drive-end) don't exceed 100F during operation. However, the south pump bearing oil is discolored and eventually the bearings rub. I believe the techs are over-filling the reservoirs which is causing too much drag for the oil rings to be effective, but I'm not sure that would discolor the oil or cause a bearing rub. They have done contamination checks and they were clear. They also found that the bearing housings we're not square with the case so they squared them up but it changed nothing. A vibe tech also went out and measured vibration, I'm not sure where but his results were that the dominating response was synchronous and there was a slight broadband response around 2X, but maybe not enough to be conclusive. I can share photos to provide a visual aid but it seems I'm limited to one per post.

Thoughts and suggestions are welcome and I apologize for the novel. If you have made it to this point kudos to you and thanks for your attention.

 

[MFJewell]

I thought you were talking about the pumps in my plant for a second, then I got to the "electric motor" part (ours are turbine driven).

 

[georgeverghese]

These high temperature KSB pumps have low permissible suction / discharge nozzle mechanical force limits that may be thermally induced by piping. Is that an expansion bellows on the discharge, encased in insulation ? What about the suction side ? Thee can be tricky to select and install, and I've heard of rotating machines ( low temp gas expanders / HT pumps) with mechanical problems (high vibration/ bearing failures), the root cause of which were these bellows..

 

[romke]

the fact that the bearing oil in one pump gets discoloured (darkened I assume) indicates that it is exposed to a higher temperature. that may well be locally so that simple temperature measurements may not show it - the 100 F measured is a temperature that should not lead to discoloration. the fact that the bearings get wiped and the oil changing in color both indicate locally higher temperatures. what caused it however, is hard to say. it's like the chicken and egg problem - got the bearings wiped because the oiltemperature was raised with a lower viscosity as a result that was no longer capable of sufficient lubrication with subsequent failure of the bearing? or was the bearing running hot due to misalignment causing wiping and oil oxidation? whatever the cause - there is a reason for the difference in temperature between both pumps and finding out the cause of the difference in thermal regime may well point you in the right direction.

 

[Artisi]

If it was my problem, and it is insisted that the train is only cold aligned, I would set up a number of dial indicators in the cold condition and monitor the installation as it came up to operating temp. this should remove the guess work in thermal expansion concerns and could give you something meaningful to work with.
Further, have you had a failed bearing analysed by the bearing manufacturer?

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.)

 

[JJPellin]

Based on my experience, the missing dowel pins seems to be the key point. This is a very long pump. I have measured axial thermal growth on a similar sized pump at 1/8 inch. It is not just a matter of directing the thermal growth away from the coupling. If the casing is not allowed to grow as designed, it will distort. As it distorts, the bearing will no longer be aligned to the shaft, resulting in bearing failure.

I am not familiar with this manufacturer or model. But I would expect a very specific arrangement to direct this thermal growth. The inboard feet should be fixed axially. There may be a pin block on the base under the inboard end. The inboard feet may be doweled. But, the arrangement of the outboard is more important. Those feet must be free to slide. There may be a key block on the base under the outboard end to direct the growth. There may be blocks along the sides of the outboard feet. The hold-down bolts on the outboard must be loose. They may be sleeved to leave a gap under the washers. They may be torqued to a low value to allow movement.

The alignment concern is real, but seems less serious. I would measure the support temperatures on the running pump and calculate the growth. I would use those results to cold-align the pump with offsets to accommodate the growth.



Johnny Pellin

 

[CouplingGuru]

jsalas1,

Fluid couplings are great, but they aern't really "couplings" per say. A power transmission coupling's primary responsibility is two fold; Transmit torque and Take up misalignment. The fluid coupling is more like a clutch it does great attenuating torque spikes and allowing for line starts, but it is terrible in taking up misalignment. In fact almost 90% of the time I see them installed, there is another coupling present to take up misalignment. For thermal growth applications such as yours, center line mounting a fluid coupling is paramount, however, depending on coupling model that may not be possible. The dowel pins are a nice way to push misalignment in the opposite direction, but that will not solve the misalignment perpendicular to the axis of rotation. Thermal growth causing misalignment perpendicular to the axis of rotation may not be of terrible concern since the pumps are centerline mounted, but that would be assuming uniform growth and no other connections functioning like the dowel pins.

Doing an alignment check in the hot condition of both pumps should give you insight in to how well each are holding alignment, then you can draw conclusions on to how much of an impact the differences in set-ups is effecting alignment in the running condition.

When it comes to couplings we are always here to help.

http://WWW.PSCCOUPLINGS.COM

 

[spciesla]

Consider consulting the Operations and Maintenance Manual for the pump and/or the OEM regarding alignment offsets for thermal growth.

 

[JJPellin]

We have only one pair of pumps that are similar to yours. Ours are a 13 stage stack-up. For our manufacturer, once the pump gets at or above 10 stages, they add a tapered case section in the center to accommodate the sag in the rotor. It is very important that that special case section is in the correct location and correct orientation. You did not mention the temperature of your service. The hotter the service, the more important to allow for the axial thermal growth using keys and pins. The brochure I found on line for this pump shows transverse keys under the inboard pump feet and an axial key-block on the outboard end. All hold down bolts are shown as double-nut to allow for the feet to slide. If your pumps have those features, they must be properly installed and adjusted. If they are not, I would expect the exact sort of bearing failures that you describe. Wiped bearings and dark oil with no increase in temperature suggests internal misalignment caused by case distortion.

Johnny Pellin

 

[jsalas1]

Holy cow thanks for all the input you guys!

The current climate surrounding this project is not great because its become a warranty issue but we hold that if the customer is performing their installation incorrectly, we can't be held liable. ANYWAY, I had the techs check the bearing constant levelers and they state that they are properly filled and that their vents are all sealing properly so they believe the discoloration is coming from oxidized oil and/or babbitt (as opposed to some foreign contaminant).

I got the Operation and Maintenance Manual from the customer this morning and on pg. 13 is the procedure for DOWELING THE INBOARD CASE FEET.

I am still going to chase down the hot alignment vs cold alignment (if they let me) to show that these pumps distort on their way to 330F (the techs think physics is hooplah).

I hope this licks the issue and will update when I hear back. Thanks again!

 

[JS_]

Use a laser alignment system such as PRUFTECHNIK with live trend to monitor the position / thermal growth from a cold to hot running condition.
[sup][/sup]