Double Suction Split Case Centrifugal Pump

[jdogg05]

Hi,

At the plant that I work at there are two double suction centrifugal pumps that I have identified (there are probably lots of others). They are the HP boiler feed-water and fire-water pumps. I assume double suction was specified for these because of their critical application.

I have included a sectioned image of the fire-water pump. I am just confused about how these double suction pumps are internally routed... Does the suction "T" and split the incoming fluid to the outboard faces of the impeller (green and turquoise highlighted); and then through the closed impeller out to the discharge (yellow highlighted)? Looking at the sectioned view it is kind of difficult to establish the actual flow path of the fluid.

Also, my understanding of double suction centrifugal pumps is that they are designed to balance the axial forces imparted on the impeller. Why is the right side of the casing (turquoise highlighted) larger than the left side of the casing (green highlighted).

The driver for this fire-water pump is a turbo'd diesel. It uses the process fluid (high pressure side of pump) in the intercooler and engine coolant exchanger as the exchanging fluid.

Thanks for any help.

 

[DubMac]

The pump you are showing is not a double-suction splitcase; but a somewhat different animal. It is a two-stage SINGLE SUCTION (2nd impeller doubles the head pressure)splitcase, which were originally developed for boiler-feed applications by Worthington years ago (the LLR model).

In a typical single stage DOUBLE-SUCTION splitcase you would be correct that the water path is split and diverted equally to each side of the impeller suction, the impeller would have suction entry on each side and discharge vanes in between the shrouds. Not in this pump though; all of the water goes into the first stage impeller, and then after discharging, it goes through what is known as a crossover to take the water from 1rst stage discharge to 2nd stage suction; that is the reason for the expanded casting size on the right. The larger casting size is to accomodate the crossover channel to; it can be seen as the white void looking areas in that larger suction chamber.

One important thing to remember about these handy pumps is that the seal chamber on the 2nd stage impeller sees discharge pressure from the 1st stage.

 

[jdogg05]

Thanks for this information. I thought I had this pump understood but apparently it is more intricate than I thought!

Looking at the sectioned drawing that image was cropped from, it labels the right side impeller the 1st stage impeller and the left side the second. How I undertand Bernoulis volume flow rate law (relating velocity and pressure) is that when the velocity decreases the pressure increases. It seems kind of counter-intuitive that the right side would have a larger casting than the left if fluid is being pumped from right to left (bigger volume to smaller volume). This would mean the velocity was increasing from first to second stage, implying the pressure would be decreasing...

Sorry, I am now even more confused!

Also, in regards to your last point, the second stage packing seal (on the left side I assume) sees the discharge pressure from the 1st stage impeller. Is that why there is a line from the discharge of the 2nd stage to the lantern ring of this second stage packing seal? Is it sort of acting like a barrier fluid?

Thanks so much for your help.

 

[DubMac]

Forget about Bernoulli jdogg, the casting design is not trying to build pressure through enlarging/constricting casing volumes. The casing flow paths are designed to move the fluid as smoothly as possible into and out of the impellers, and at the same time keep out of the way of the mechanical elements of the pump. The reason the one side looks larger than the other is because it has to make room for the crossover (white area), thats all.

The line from discharge back to seal is just to provide a flush of the 2nd stage seal chamber using the higher pressure from the 2nd stage discharge.

 

[jdogg05]

Thank you for the clarification DubMac. I think I understand the pump a lot better now.

One last point: although this is not a double suction pump, it still has an axial balancing effect due to the orientation of the impellers (back-to-back). With large multi-stage BFW pumps are the succession of impellers orientated so as to balance the axial forces on the shaft? i.e. are they oriented in back-to-back pairs for this reason?

 

[JJPellin]

Thrust balancing in multi-stage pumps can be accomplished in a number of ways. Opposed impellers is one method. Other pumps use balance drums or disks. There is an FAQ on this subject.

http://www.eng-tips.com/faqs.cfm?fid=1306

Johnny Pellin

 

[DubMac]

Yes, impellers are back to back on large multistage volute style pumps to provide a measure of axial balance. Even your pump is not completely axially balanced though; you don't have the same suction pressure on both sides, but it is a lot better than in series.
Not all multistages have opposed impellers. A diffuser style pump (such as many super high pressure barrel pumps) has impellers in series and each impeller will have its own diffuser to channel the fluid from impeller to impeller. These pumps don't require the crossover (which becomes fairly significant in a 10 or 12 stage and castings can be tricky to pour) and diffusers can be fine tuned by polishing and playing with vane angles to higher efficiencies than volute styles (for the most part).

Diffuser style drawback is they have a bunch of parts and pieces and really need a high level machine shop and expertise to work on.; they of course need good axial thrust balancing, usually by a combo of bigass bearings and a hydraulic balancing device. Really ends up being customer preference/familiarity in deciding whether to go volute or diffuser style. Me, I'm a diffuser type guy; thats just the way I roll.

Go look at Flowserve, Sulzer, or Goulds websites, I would imagine they have pretty good info explaining all this stuff much mo betta.