Pumps with sliding base plate

[lcmechengr]

To help eliminate the lateral forces placed on a pump suction nozzle, I was thinking I could place the pump on a sliding plate. Depending on the amount of displacement required, I would either have slotted bolt holes or devise some type of clips to restrain the pump during start-up. I would check the frictional forces and make sure that these would not exceed allowable nozzle loads. I was curious if anyone has had any experience with this or has done this in the past. Ideally, I would like to decrease nozzle loads, but I am limited by real estate, NPSHr of the pump, and not being able to use expansion joints. Besides retraining the pump against start-up forces, are there any other force or vibrational issues that need to be condidered?

 

[JJPellin]

We have 1400 centrifugal pumps in our refinery. Of those, 4 are on floating base plates. Two of them are in our cokers. These bases are supported on 6 large springs. This was done for the reason you are discussing. Plot space was valuable and the piping needed to be short with no chance for expansion joints. These pumps have been bad actors since they were installed in 1963. The other two are on sliding bases very much like you described. This was done by the OEM at orignal purchase and was not a home-made fix by us. the bases are welded to plates that are bolted to the foundation loosely. The holes are slotted and the hold-down studs are double nutted with a certain amount of room for movement. These pump are also bad-actors with a very poor reliability history since they were installed in about 1984. We have had vibration problems, bearing problems, wear ring rubs, coupling and alignment problems. The pump design is poor, but I blame a lot of the trouble on these sliding bases. I am old fashioned. I think a pump should be bolted and grouted down to a large foundation block and the piping should be made flexible enough to limit the nozzle loading. In a similar pair of pumps in another coker unit, they took the bottoms line from the fractionator and made a big loop all the way around the skirt and then back to the pumps. This loop is supported off the skirt and makes the piping flexible enought to meet nozzle load requirements. These pumps are much more reliable than the ones on springs or sliding plates.

 

[electricpete]

Needless to say I assume the motors are mounted on the same plate as the pump.

It sounds good in theory to a non-mechanical guy, but that's not saying much. Maybe a related question: why is a rigid attachment to a massive foundation important to begin with.

jjpelin - what kind of problems do you see?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[JJPellin]

We have a pump with fairly tight internal clearances between rotating and stationary parts. The wear ring clearances might be about 0.010" per side. We have a coupling between the driver and pump that we have aligned with an allowable misalignment tolerance of +/- 0.002". With these machines bolted to a sliding plate, there is no good way to maintain the internal and external alignments within acceptable tolerances. Radiant heat on the top of the sliding plate causes it to curle and distort. This changes the coupling alignment. The support pedestals under the pump distort and toe out from the heat of the pump case, this puts stresses on the pump case that affects the concentricity of the rotating rings within the stationary rings. We have seen bearing failure, mechanical seal failures, wear ring rubs, coupling failures and high vibration that was indicative of misalignment. Bolting everything down to a massive foundation does not prevent all of this, but it greatly reduces the affect. In our case, one pump has a motor driver and the other has a steam turbine driver. Yes, the drivers are bolted to that same floating plate with the pumps.