Centrifugal pump thrust adjustment 2

[CVCS]

My plant has three multistage (eleven) centrifugal Pacific (Flowserve) pumps that are used to inject water at ~2800psi. The pump has a pressure reducing bushing, no balancing disk.

One of the pumps has always run hotter on the outboard kingsbury thrust shoes. We would like to change this.

If I adjust the rotor running position off the suction end, can I change the thrust enough to decrease temperature on the outboard shoes? If so, which way do I move the shaft, inboard or outboard?

 

[Artisi]

First up you need to establish if the thrust bearing is operating at an acceptable temperature, "hotter" is subjective and doesn't tell us anything. Is the bearing operating and giving normal life - if yes why worry?

 

[JJPellin]

I would not recommend trying to reduce the thrust bearing temperature by changing the running position. Instead, I would try to understand why one out of three identical pumps would have a different thrust bearing temperature than the other two. A few possibilities occur to me.

First, I assume that the pump has a tilt pad thrust bearing with embedded temperature probes (RTD's or thermocouples). A subtle difference in how the probe is embedded can result in a substantial difference in the temperature reading. Specifically, if the probe is closer to the surface or if there is slight bulge in the surface babbet because of the way the machining was done you can see a higher temperature reading. It can also make a noticeable difference if the instrumented pad is in the upper half or lower half. If the bearing is not functioning correctly because of a hang-up of a leveling pad or uneven shimming of the bearing halves you could see a higher temperature reading. I recently found a thrust irregularity in a high pressure coker jet water pump (3600 psi) that was a result of a distorted cage assembly resulting in uneven pressure among the bearing pads.

Then I would consider any differences in the lube oil conditions to the three machines. If the flow of oil to this bearing housing is different than the others, it could account for the temperature difference. An incorrect orifice plate or a plugged passage within the housing could explain the difference. Many of these bearings use floating brass rings as oil retainers on each side of the thrust bearing assembly. If the rings in this pump are tighter than the other pumps, it could be passing less oil flow. These bearings often run flooded but not pressurized. Is there supposed to be an overflow port at the top of the housing to allow the excess oil to spill over to the radial bearing section of the housing?

I would be very concerned about the condition of the balance line and pressure reducing bushing. Based on your description, I assume you have opposed impellers with a balance line. If the pressure on the two seals is not close to the same, there could be a restriction in the balance line or an opened up pressure reducing bushing. If the pressure in the two seals differs much, this will translate into a thrust imbalance. For that matter, any bushing or wear rings problems could affect the relative thrust contribution of one or more stages. A check of the overall pump performance relative to the test curve could reveal a problem.

I keep coming back to my original point. I would first try to understand the difference before I would try to correct it with any change to the pump. And as Artisi noted already, unless the temperature difference is resulting in failures, it might not require any action. If you could give us more details about the pump such as model, history or service conditions, we might be able to provide more specific advice.

Johnny Pellin

 

[CVCS]

Pump model: Pacific RLIJ 2 1/2
Service conditions: normal operation is 120gpm, 100F water. Motor is 900hp, 1800RPM with a 2.7368:1 speed increaser (shaft at 4900RPM). The pump is normally operated back on the curve. We have John Crane Mechanical Bellows 1-B seals on each end of the pump. The seals tend to leak even after replacement.

The outboard thrust shoes did fail in 2000 when mechanics discovered the Babbitt lining completely wiped off the shoes. I am assuming very high temperatures allthough exzct conditions remain unknown since we did not have RTD's at the time.

Recently we had another, much less significant wipe due to a non-prelubed test start by procedure. This time the temperature jumped 10-12 degrees. We have fixed the conditions for the dry start and have the opportunity and momentum to try to change the higher outboard bearing temperature. It has been a question here at the plant for years.

Our vendor manual allows a tolerance of +/- .03in from the running position off the suction end. We were considering changing the thickness of our thrust spacer to move the shaft out or inboard to effect a change.

We have conflicting theories as to which way to move the shaft. At first I was thinking that since the shaft is set in the correct position initially but riding outboard (higher temperature), we should move the shaft outboard to allow it to be balanced. Thrust collar would remain in the exact same position within the housing.

Then I saw a correspondence from the vendor from 2000 that said "Rotors that are set inboard will thrust inboard and rotors that are set outboard thrust outboard." I am trying to understand the hydraulics inside the pump that would make this so.

If we move the shaft...which direction if at all?

The pump is performing normally. We have prescribed limits for temperature, below 200F is considered normal, 225F warning and 245F shutdown. The outboard thrust bearing is operating at 175F in the summer months while the other two identical pumps are in the 130 - 140F range.

We have had outboard bearing issues for other reasons so this topic is important for us to understand and try to mitigate. Translation: we want similar equipment to operate the same way.

Temperature probes: you are correct to deduce that we have imbedded temperature RTD's in the lower pads of the bearing. We are replacing the thrust shoes in this repair outage so we may see a difference.

JJPELIN: Question, how did you determine that the cage assembly was distorted?

Oil control ring: we do have identical oil contorl rings on eash of the three pumps. They have an orifice at the top of the ring that allows oil to exit the thrust bearing area and return to the sump. Other than that I believe the oil is moving properly through the system. Again, all three pumps are identical in this regard. Oil is brought up from the supm via a shaft driven positive displacement pump and delivered to each bearing area (in/out journal, in/out thrust) via ports on the bearing housings. We inspected these and they are normal.

We do not have imposing impellers. All impellers are facing the same direction similar to Figure 8 in Marks' Handbook. The vendor drawing of the balance drum looks exactly like figure 11 (not like the one in figure 8).

Our overall pump performance is tested every three months and full flow testing is done every eighteen months. All tests have been SAT as compared to the test curve.

 

[JJPellin]

Thank you for the detailed information. The distorted cage I referred to in my first reply was found during an inspection of the bearing on a granite table. They laid it out to measure the overall thickness of the bearing, complete with shims and pads. They found it would rock when placed pads up on the table indicating that the cage assembly was warped. If you have trapped shims in each half of a cage assembly, could they have been shimmed differently by mistake?

As far as your original question, I would not expect a small change in axial running position to have any affect on thrust direction or load. I am assuming that the wear rings are all straight radial rings with no steps. The balance drum you describe has a straight radial fit that would be unaffected by a small axial move. I really can't think of any reason why a small axial move could reduce the thrust load and solve your problem.

When you say to the outboard thrust shoes were wiped, I assume you mean the thrust shoes furthest away from the coupling. This would normally be the inactive bearing. If I am correct in my assumption, this tends to rule out my concern about a balance line restriction. With a typical configuration, if the balance line was restricted, the higher pressure on the outboard end would cause the shaft to thrust hard toward the coupling.

If the thrust failure was toward the coupling, then a balance line restriction or a pressure difference between the two seal chambers would be my candidate for most likely source of your problem.

The last thing that I would check carefully is the influence of the coupling and driver. Is there any chance that the coupling or driver are pushing this pump rotor toward the outboard? Incorrect coupling spacing, a motor not set on magnetic center, a locked up gear coupling could all push back on the pump rotor and cause it to run hot to the inactive.

If none of this pans out, I would worry that something within the pump is assembled wrong or is damaged. Most errors of this sort would also cause a loss of pump performance, so this might be a long shot.


Johnny Pellin

 

[CVCS]

We ran across a note from a report in 2000 when this same problem occurred only an order of magnitude more severe. It says "Depending how many impellers are set toward suction or discharge may effect the direction that the rotor will want to thust. One of the things that can influence the direction of thrust is the axial position of the rotor. Rotors that are set inboard thrust inboard and rotors that are set outboard thrust outboard. This is particularly true due to larger sectional areas behind the impeller shroud wall in the direction of rotor offset."

We are confirming this with the vendor but we (engineering) cannot validate or dismiss this statement especially the last sentence!

Since I believe we have a straight through pressure reducing sleeve, i.e. no variable inner chanber, I agree with you that positioning the balance drum will have little or no effect.

For this pump we actually set the running position off suction stop. That is done by installing the inboard thrust shoe then setting an indicator to zero, removing the inboard bearing, pushing the shaft to the inboard until it stops and measuring this distance. This is the running position.

I'll have the mechanics check the cage next week. Also the vendor is coming in next week thursday to help us out. Even if in the end we putting everything back together and we have a higher temperature on the outboard thrust bearing, we will still be with in the normal range. Just at a lower margin to the limits than our other two pumps.

Thanks for the great advice!

 

[JJPellin]

I believe you may have misinterpreted the note from 2000 that you quoted. When I read this, I believe it is referring to a pump with opposed impellers. What I think it means to say is that impellers (not rotors) that are facing toward the outboard will produce thrust toward the outboard and impeller that are facing toward the inboard will produce thrust toward the inboard. When it says this is true because of difference "sectional" areas on the backside of the impeller shroud, I interpret this to mean that the wear-rings on the front sides of the impeller are a different area than the bushings on the backsides of the impellers. You indicated that your pump does not have opposed impellers, so I think this statement is not relevant to your case. In a pump with all impellers facing the same direction, the thrust can still be in either direction. But it is only dependent on the design of and the condition of the balance piston. The impellers always thrust toward their suction eyes. And the balance piston always thrusts in the opposite direction, toward the outboard. Nothing you do with rotor running position can change that.

I assume that your pump is a diffuser style rather than volute style. It is probably a barrel configuration given the very high pressure. In most pumps of this type, the proper running position is very close to the center of the mechanical float. As you described the setting procedure, I assume that the manufacturer gives you a setting dimension from hard inboard. This is common with many of our larger barrel pumps. A typical pump might have a total float of about 0.650" and have a proper running position of 0.375" from hard inboard. This ideal running position is provided by the manufacturer as the position that will result in the ideal alignment of all impellers with their diffuser sections. But in the absence of this setting number, a rotor placed at mechanical center will generally work quite well.

If you are going to have the thrust bearings out of the pump anyway, I would suggest doing a full bump with no bearings installed. Following the measurement of running position that you described, zero the indicator with the rotor hard to the inboard and then thrust it hard to the outboard; hard stop to hard stop. This should be about twice the running position measurement that you set to. It does not have to be 50% / 50%, but if it was much farther off center than about 60% / 40%, I would question the running position setting that the manufacturer had provided.

Please come back and post any results from the work you have planned. Let us know if you find the problem. Good luck.

Johnny Pellin

 

[JJPellin]

The more I read this and think about it the more I believe that your problem is because of some difference in the impeller wear rings or balance drum in this pump. It has been common in some pumps to repair a damaged balance piston by making a clean up cut on the drum head (the stationary bushing around the drum) and then manufacturing a new balance drum to have the proper clearance to the re-cut drum head. Because of your very high pressures, this practice could result in a large change in thrust characteristics. For example:

If you had a balance drum that was 5.000" inches in diameter but increased it by 0.030" to correct damage from a failure, the net thrust would change drastically. With a differential pressure of 2800 psi, this would result in a thrust change of over 660 lbs toward the outboard. Similarly, any dimensional change to the impeller wear rings (because there are 11 sets of them) can produce a large cumulative thrust change. Do you have any repair reports from the 2000 failure that would provide wear ring or balance piston dimensions that can be compared to the original dimensions in the OEM manual?


Johnny Pellin

 

[CVCS]

Our pump is a barrel type and has all the impellers facing one direction using diffusers as you say. I see exactly your point as there can be no geometrical change in the areas inside the pump that are used for balancing by moving the axial position. This is why we were struggling with the statement from the vendor.

My research has found that we did in fact replace this pumping element after the events of 2000...with a rebuilt element. I don't see the balance drum or sleeve listed on the inspection report which leads me to believe that we did not send it to the vendor. It is, of course, possible that the balance drum clearance is just slightly smaller on the one pump than the other two. Vendor manual dimension is .012 to .014in.

A page in the inspection report lists non-conformances. It says that wear rings should be .010 to .013. As found dimensions are: 1st stage - .00098", 2nd - .0148 to .0158", 4th .014 and 11th .0136". First stage was repaired and the others left as-is. Isn't it true that as the wear rings wear or increase clearance, the pump will thrust more inboard? I believe this is why the balance drum is set up to thrust outboard slightly.

Our maintenance procedures and the vendor manual give the mechanics guidance into measuring the axial position of the shaft. We measured the full axial float at .362 and the running position off inboard stop .195. The procedure says, just as you stated, to set the position at one half of the total axial float if the best axial position is not known. It also has a section for the mechanic to measure the best axial position but this requires taking the head off the pump which I am not willing to do right now. Like I said we did find the repair report from the vendor for this element and in it, they calculated the best running position at .203. Again, the vendor manual allows +/- .030 from this position which is where we thought we could change thrust. With your help we are seeing that changing the running position will probably not change thrust.

 

[JJPellin]

The primary dimensions that I would be concerned about for the balance piston and wear rings is not so much the clearance as it is the diameter of the rotating ring. For an impeller wear ring, similar repairs can be made. If the wear rings had rubbed hard and damaged both the stationary and rotating rings, I might not want to replace all of those parts. Since there are eleven stages, that would be an expensive and long repair. But, I could take a slight clean-up cut on the impeller rings to get them smooth and round and then replace only the stationary rings. If I did this on eleven stages, the change in cross-sectional area, times the pressure drop across each ring, would produce an increase in thrust toward the outboard. Changing all of the wear ring diameters by only 0.030" would result in the same 660 lb.net thrust increase toward the outboard.

I am not sure now the thrust would change as the wear rings wear. The thrust contribution from each stage would tend to increase toward the outboard. But the counter-thrust contribution from the balance piston would tend to decrease toward the outboard. The affects might cancel each other out. But any significant change in clearances should affect the overall performance of the pump. You indicated that the pump capacity and pressure are good. So, I don’t think clearances are probably the problem.


Johnny Pellin