Pump Killers: How to fight the 13 most common centrifugal pump failures? Number 8.

In the series “Pump killers” we investigate the 13 most common causes of centrifugal-pump failure and take a look at the measures that can be taken to fight these ‘killers’. Every week we will share a new, even bigger, killer. Our goal: to reduce the number of victims by sharing our knowledge.

In the pump industry it is often said that a pump "is running off its curve". You almost assume that people know what is meant by this. Nevertheless, in practice this does not always appear to be the case.

The performance of a pump is shown by the pump manufacturer in a pump graph. This graph often shows the relationship between capacity and head, capacity and efficiency, capacity / absorbed power and capacity / NPSHr. In some cases a part of the pump curve is shown in dotted lines, with which the pump manufacturer wants to indicate that the pump should not be selected in that part of the pump curve.

The line that we call the pump curve, shows how the pump produces a certain capacity at a certain head. You can see from the line that whenever the capacity increases, the head will go down. This is because the pump can only give a certain amount of velocity energy to the medium. The more medium it needs to pump, the more the velocity energy will be divided among the medium.

How does this phenomenon arise? In addition to this “pump curve”, the presence of pipe losses in the suction pipe and discharge pipe will also create a relationship graph with regard to capacity and pipe losses. We call this the line characteristic.

This line characteristic (green dotted line) can be shown in the pump curve and will intersect the pump curve at some point. This is also the intention and in this case the pump runs "on its curve".

If the pump curve and the pipe characteristic do not touch each other, we speak of a pump that is "off its curve".

What are the consequences? When a pump "runs off the curve", the pump will try to deliver enough capacity untill the pump curve line and the pipe characteristic line intersect again. The pump is not designed to deliver this excessively large capacity that is out of reach and, in consequence, will fail very quickly.

I often use the following example to explain it:

"Put someone on a track bicycle, so without brakes and without the possibility to pedal back, on a mountain and let them go. How does that feel? You are spinning, and that happens with that pump too."

The consequences of this killer are enormous. It is literally a whole list of with damage: shafts that break, stainless steel impellers where the welds come loose, cast iron impellers that develop a crater pattern at the end of the blades, stainless steel impellers that break loose from the core, excessive vibration and noise production, just to mention a few.

We have sometimes come across cases in which 100 minutes was enough to break an axle. Of course, this always happens at times when the pump cannot be missed.

How can this be prevented? Always make sure that the pump and the pipe system are properly aligned. If the pipe characteristic intersects the pump curve near the highest efficiency point, the calculation has been successful.

In the event that this deviates slightly, read this very literally, and the pipe characteristic does not touch the pump curve, you want to realize the following:

This can be done by putting a:

• Control valve: by squeezing this you can ensure that the resistance and therefore the pipe losses increase and that it still cuts the pump curve. The disadvantage of this, however, is that in practice, out of ignorance, people sometimes want to open this valve, causing the pump to run outside its curve again, with all the associated consequences.

Picture: Orifice plate

• Orifice plate: you place this between two flanges and it also increases the resistance and thus the pipe losses. This is a better option compared to a control valve, because it cannot be turned on by an unknowing user;

In both cases, it must be said: it is again, as we have already mentioned in an earlier article: accelerate with your foot on the brake.

• Frequency converter: actually the best option, it revs the speed back and is therefore the most economical of the three.

In conclusion, we can say that all of the above solutions are basically just fixes, because a pump running "outside its curve" is actually wrongly selected. If the choice of pump is really wrong, then it is better to use another pump, with the lines preferably intersecting each other as close as possible to the highest efficiency point.

We are curious whether we have succeeded in making this, in general, highly theoretical story somewhat manageable. Please let us know if anything is unclear or if you have any additions. Experiences with this killer are also more than welcome!

Next week: number 7.