Pump Thrusts

[sow]

We are having problems with some of our wet well submersible mixed flow (shrouded) vertical sewage pumps.

Pump end bearings are failing and the seals are failing. Motor end bearings are also failing.

The pumps are running out on the curve when they start for about ten seconds or so - a short period but they start quite often and it soon adds up.

Also the pumps are "turbining" when they are switched off -this is required as the pumps are in a 20m sump, therefore vacuum breakers are needed in the valve house.

I am convinced that the increased radial thrusts known to happen to the right of BEP are causing a problem during start up.

My intuition (not been calibrated for a while admittedly) tells me that during the turbining phase the axial thrust may reverse. Obviously it would need to overcome the pump dead weight. My main concern here is that the bearing arrangement in this particular pump only takes thrust in the downward direction.

However I cant find any calcuation for thrust during this operation (Ive looked at Sulzer, Karrasik) or even refernce to the possibility of axial thrust changing direction.

Any comments or suggestions greatly appreciated

 

[Artisi]

Are you are talking about line-shaft driven vertical pumps or close-coupled motor/pump units.

 

[MikeHalloran]

How big are the vacuum breakers relative to the pump discharge size?

Have you measured how fast the pumps spin when 'turbining'?



Mike Halloran
Pembroke Pines, FL, USA

 

[JJPellin]

I am not certain that I understand the configuration of your pumps. But I think it would be much more likely that you get reverse thrust when running off the end of the curve. This is common in vertical turbine pumps, especially pumps with very many stages. In order to reduce the net thrust running at normal flow, the manufacturer adjusts the impeller geometry. Then, at very high flow, without the benefit of discharge pressure, the thrust reverses and can be great enough to buckle the shaft. This was well documented in a paper presented at the Pump Users Symposium in 2005.

I understand that you do not want to pull a hard vacuum when the pump shuts down. But can you install a check valve at the bottom of the line, down by the pump to keep the line flooded and avoid both the reverse flow and the high flow at start-up. If that will not work, then you could adapt an anti-reverse, ratcheting mechanism so the pump cannot spin backwards. This is commonly used on deep water well pumps.

 

[sow]

All - thanks for your comments.

The pumps are wet well pumps as in the motor is submerged by the sewage. The motors are close coupled and are vertical.The pumps are single stage.

Configuration similair to that shown on this web site ( note this company has nothing to do with this particular pumping station)

http://www.cwsalescorp.com/station/

The total head is around 80m. There is a 20m lift to the valve chamber, the rest of the head is in losses.

I havent been able to measure the speed of the turbining as the pumps are submersible close coupled pumps.

The vacuum breakers consist of about 4No air valves the size

 

[MikeHalloran]

I'm sorry, I don't understand why the vacuum breakers are necessary.



Mike Halloran
Pembroke Pines, FL, USA

 

[sow]

No problem Mike - the vacuum breaker admits air when the pumps are stopped and releases air when they start up. This is beacuse there is about 20m between the NRV and the water level in the sump (at atmospheric) therefore to my mind there would be a full vacuum in the pipe - this could cause problems sucking air in and pose surge and cavitation problems with starting the pumps

 

[MikeHalloran]

So, without the vacuum breakers, after stopping the pump and waiting for equilibrium, there would be a column of black water maybe 10m high at the pump discharge, and a vacuum above that. But the pressure at the pump outlet would be nearly atmospheric, because of the weight of the water column. Would the pump have a problem starting against that?





Mike Halloran
Pembroke Pines, FL, USA

 

[sow]

Its not so much the starting against low head im worried about, this can be catered for by properly priming the main by use of the VSD's. We are hoping to extend the ramp up time to reduce any off the curve problems.

The designers had a surge analysis carried out that shows surge would be a problem if we did not include the vacuum breakers . Surge may not be an issue for shorter mains however this main is 10km long. Pumping against vapourized water / air mix could cause a column collapse that can even exceed Joukowski heads. In fact the difference in level is more like 30m plus the horizontal leg therfore you could have a 40 -50m length of vapour up against your NRV. I suppose theoretically you could try and re-pressurise the vapour with the pumps ,it is very difficult if at all possible to predict what will go on during this phase. I think even if you got away with this you would cavitate your pipework and valves in no time as vapour bubbles implode.

Also it would be likley that air would be drawn through the gaskets (along with the gaskets possibly). Again a problem with surge.

We have considered adding NRV's in the wet well to remove this potential problem all together, however the sump is 30m deep pumping raw sewage there is no standby well- the operators quite rightly dont like the idea of nrv's down there.

Therefore i was interested in the direction of the axial thrusts, as the damage shows all the signs of reverse thrust (i.e. upwards) and radial thrust for that matter. The pumps installed are fitted with bearings that are not designed to take thrust upwards - is this normal practice and could it be a problem? A quick fix may be to fit bi-directional bearings. Im not too sure about this argument as the impellor back plate would stop up against the casing.

Cheers

 

[Artisi]

The fitting of bi-directional bearings has merit, however if you haven't contacted the supplier/manufacturer suggest this might be a good place to start.

"Im not too sure about this argument as the impellor back plate would stop up against the casing." --- What do you mean by this?

You have said that the pump/motor is close coupled - but talk about pump bearings and motor bearings - is the pump fitted with bearings - how is the pump connected to the motor?

 

[d23]

sow

I have never been involved with sewage pumps, so this post is only general information.

There is some missing information that could help the replies like the actual rotating speed and the pump specific speed.

A mixed flow stage provides lift by both centrifugal force and diffusion. A mixed flow stage will usually have a very flat power curve, which would indicate that the lift is reasonably balanced between the two forces. If the power curve increases as flow increases there will be more centrifugal force than axial or diffusion indicating a low specific speed. If the power curve decreases as the flow increases there will be more axial or diffusion than centrifugal force indicating a high specific speed. Frankly if I were pumping solids or viscous fluids I would prefer a higher specific speed pump.

A submergible mixed flow pump will have a bowl or diffuser verses a volute. The centrifugal force generated by the impeller will completely surround the impeller (360 degrees) creating a balanced hydraulic environment that dramatically reduces any radial thrust. Being a sewage pump you may have some vane plugging causing adverse radial loading. Cleaning the pump would be the only fix for this.

Without seeing the failed equipment I have to believe axial thrust should represent most of your problems. We have performed extensive load cell test on our mixed flow stages and found that for low specific speed impellers the maximum thrust occurs approximately half way between BEP and shut-in. As the specific speed increases the maximum thrust moves closer to shut-in indicating an increase of axial force. A low specific speed impeller will have no down thrust in fact transitions to up-thrust approximately ¾ between BEP and open flow.

The static axial thrust or the weight of the impeller and shaft will affect where maximum down-thrust and the transition to up-thrust occurs. The approximations above only represent pumps that we manufacture. The point being that axial thrust is a major part of determining a recommended operating range for a pump. You need to consult with your pump supplier for min-max range and where the pump could transition to up thrust. Up thrust can affect the motor seal and bearing.

The industrial motor you use will have a tilting pad type thrust bearing. Tilting pad thrust bearings can withstand very high thrust loads at there rated speed, but require a lubricating film to maintain a clearance between the pad and the collar. This film depends on a rotating force. At low speeds there will not be any film or very little film available. During any extended or slow starting and stopping process or anytime the pump is not operating and allowed to spin backwards at a low RPM a tilting pad type bearing could fail. Numerous starts and stops contribute a large part of your bearing failures. I don’t know the type of motor you use, but oil filled motors verses water filled would help the bearing.

D23

 

[Artisi]

For interest, what size of unit are you talking about.