Pump Shaft Failure Possibility

[cokeguy]

I´he heard that when you restrict flow by partially closing a valve inmediately at the discharge of a centrifugal pump you run the risk of overheating the liquid and even breaking the pump shaft. In our plant we have several slurry pumps with HP ratings from 50 to 150 HP, no problems as of this moment with them, but we sometimes have to restrict their flowrate by closing the pump´s discharge butterfly-type valve, located a couple of feet form the discharge. The reason we do this is that flow regulating valves located much farther downstream from the pump sometimes fail or wear-out and therefore tend to increase flowrate and overload the motor, so we partially close the discharge to control flowrate until we can schedule a maintenance shutdown to replace flow-control valves (by the way we are in the process of replacing them with other type of valves more suitable for throttling slurry flow)

In the meantime, any risk on the pump by semi-closing discharge valve? How can we know if we are on the verge of breaking a pump shaft, or does that happens only with smaller or larger pumps? Our pump´s shafts range in diameter from 2.5" to 4", if my memory serves me right, and flowrates vary from about 1000 gpm to 3000 gpm.

 

[Artisi]

If you are closing a discharge valve to reproduce the head that is normally imposed by the system and regulating valves, I can see no reason for you to expect any problems.
However, if you are imposing more head on the pumps than normal and forcing the pumps to run at a higher discharge pressure, you could be operating the pumps in an area that might lead to problems.

Bear in mind, that "real" slurry pumps are usually very robust in their construction owing to the nature of their usual operating conditions - but, if you have a concern, my advice is to contact the manufacturer with all the operating information and ask for their appraisal of the pump duty.



Naresuan University
Phitsanulok
Thailand

 

[JJPellin]

As you describe it, I would think it very unlikely that you would have a problem with shaft fracture. This is no the normal failure mode I would expect for a pump running too far back on the curve. The pump may run hotter. At some point, it could increase the chance of cavitation. There may be more radial loads on the shaft (especially if the pump has a single volute). But, as noted, if you are only restoring the head requirement to compensate for a worn out control valve, you should see no affect at all on the pump. The valve, however, may be a different matter. Depending on the design of the butterfly valve, valve reliability could be compromised. Many block valves are not designed for throttling.

 

[bulkhandling]

For a good selected pump, normally, the operating point between 70% to 100% of the Best Efficiecy Point will provide a satisfactory operation with bearing life mostly longer than designed L10 life. To extend the range from 50% to 110%, pumps will run normally satisfactorily with reasonable bearing life. For each specific pump, the manufacturer has a recommended minimum flow. Under that flow, the bearing life will be considerably shortened.

When mention about slurry applications, we normally do not throttle valves to control the flow rate except that the slurry is not abrasive or have no choice .

For a abrasive slurry, variable speed control is always used for flow control. With throttling valve, the valve is always a major concern of maintenance, the pump impeller, wear ring and casing worn out faster, the bearing life is shorter, energy consumption is higher...

 

[whizzz]

There is no problem for you to control the flow rate at the discharge. The common cause of broken shaft could come from the pump equipment, not the process. For examples, commonly, the first question raised up after the shaft found broken is... do the pump vibrate? the unbalanced of pump impeller is the major problem which cause this, especially for those who always modify their impeller without accurate dimensions or balancing.
Also immediate impact inside the pump system also can cause this problem.

 

[cokeguy]

Thanks everyone for your input. Now, regarding this same issue, we will be making some process tests next week involving our 3000 gpm slurry pump (downstream piping changed a lot) and we were asked to prepare a recirculation path back to the slurry tank that feeds the pump because one of the two main downstream valves will remain closed for some time and the flow would be too small (perhaps 1000 gpm). Slowing down the pump is not an option, as we expect to open the other valve and thus go back to normal flowrate in perhaps one or two days at the most, without stopping the pump.

My question is, do we really require auxiliary recirculation paths in those cases? How far to the left from the BEP on a typical pump curve can you move without worrying, and for how long? Any document or link that covers specifically the perils of working too far left on a pump curve?

Unfortunately we have no pressure or flow or temperature indications on the discharge side that could help us identify actual operating point or temperature rise, only motor amps and downstream process behaviour. We can live with the pressure rise downstream, no problems expected there, my only worry is the pump in itself. Any comments greatly welcomed.

 

[Artisi]

my suggestionn is to ask the manufacturer, they are the only ones qualified to make this sort of recommendation. We, the forum guys here have not even be made aware of the pump pump manufacturer or the full duty/ installation details, so how can we realy give you good advice.

Naresuan University
Phitsanulok
Thailand

 

[cokeguy]

You are correct, I´m indeed relying on our pump supplier to help us with this specific test next week, but I´m just trying to get more general info/literature on that particular subject because I´ve yet to find a good reference that covers it. Bulkhandling´s post about bearing life and variable speed for slurry pumps was very helpful and I´m looking into it, but I´m always trying to find and learn more since I´m relatively new at this job. For the past six months I´ve detected and corrected several problems, mispractices and engineering flaws here at the plant thanks in no small part to the cases and references that have been discussed and suggested in this forum. Thanks all for your help.

P.S. by the way, this particular 3000 gpm pump is not (I believe) a "real" slurry pump, it was not specified correctly and therefore we have had to change impeller and liner once already , as well as a damaged mechanical seal, in less than 1 year of operation. The other pumps are real slurry pumps with centrifugal seals and natural rubber (?) impellers and liners, absolutely no trouble with them for almost 3 years now.

 

[bulkhandling]

It's not my intention to replace your vendor's job since we do not know the details of your curve and piping configuration.

If you only run the smaller flow for one or twice and each time one or two days, you may not have to recirculate it since 1/3 of the design flow may still be more than recommended minimum flow - your job to check. If the low flow is kind of regular case, a recirculation could be a good choice.

The recirculation line does not have to be very big - big enough to make the pump run at reasonable operating point on the curve can be good enough. A manual throttling valve (a pinch valve is good for your application if no chamical concern about the pinch valve rubber material) on the recir-line is suggested to get the back pressure to deliver the designed flow rate in the main pipe.

 

[bulkhandling]

One thing nobody has mentioned yet is the critical velocity for the slurry. When the flow reduced in the main pipe, the velocity is also reduced to which can be considerably lower than the critical velocity. You may have to have a check on this. Suggest to get a manual from GIW or Warman to have a check.

 

[Artisi]

bulkhandling (Mechanical)- you took the words right out of my mouth, today I planned to comment on both points you raised;

1. slurry control valve, these are available and should be used if you intend to vary the flow in this manner.
2. the critical velocity is a very important factor and must be looked at if you are going to use speed control.

One other point you should consider is replacing the incorrectly spec'd pump with a slurry pump (that's if it not already a slurry pump), as it will probably be more economical in the long run.

I sure most of the guys / girls in here are more than willing to help with answers and advice - but it helps if you give us specific questions so that we can give meaningful answers and not be in the position where we need to answer in general terms.



Naresuan University
Phitsanulok
Thailand

 

[vanstoja]

For an extensive discussion of the estimation of minimun operating flows for centrifugal pumps see vanstoja and others' replies in Thread407-104691.

 

[bulkhandling]

If you have not replaced your regular centrifugal pump with a slurry one, here are a few tips to extend your pump life and reduce maintenance cost.

a. make an accurate hydraulic calc on the piping system. To trim the impeller to which when pumping the maximum flow, the control valve is NEAR the fully open position (you may need some contingency to make you safe and some room for wear). This way you get a smaller impeller tip speed and less power consumption.
b. Add a seal water to the mechanical seal. This will make the mech seal last longer.
c. Buy duplex stainless steel impeller instead of use corbon steel impeller. Flowserve and Sulzer have duplex ss as their standard impeller material and the cost is normally not much higher. Duplex SS is much harder than carbon steel and for sure lasts much longer.