How many times have we heard sounds emanating from pump casings like ball bearings rattling around inside a tin can? The first thing that comes to mind is cavitation. We know what it sounds like and we see the damage it can do to impellers, but do we know what to do about it? In many cases - no! It has been classified as "the nature of the beast" and other non-technical terms but it is allowed to continue until the inevitable happens when the impeller disintegrates. That is when maintenance can be expensive.

On one site that I visited, there were four impellers in the warehouse just for one pump because they could not understand why the working impeller failed through severe pitting. This had been going on for nearly three years since the pump went into operation. They thought the rattling noise was due to the fluid contents being pumped. As the suction pressure head could not be increased, the fluid temperature was lowered out of its saturation temperature range and the noises stopped. This was a simple solution to a complex problem. But, in some cases where cavitational noises are heard and the above solution is not effective, impeller damage is still taking place. What is the problem?

A cavitation sound-alike problem was causing havoc in one pump on an oil platform about 120 miles offshore in the Gulf of Mexico. Because of higher production outputs, some of the existing pumps were "redesigned" to produce more. On this particular pump, a new motor was installed which increased the power to10 horsepower above the original design specification requirements. The original thought being that more power increased output. Unfortunately, the design change did not consider any increase in output piping dimensions to cope with the greater increase in fluid flow. Due to this oversight, severe backpressure in the discharge area of the pump's impeller and volute resulted and the pressure energy that was created, converted to heat energy. The subsequent flash-off created, what the operators considered to be, cavitation. From that point on for the next two years, they had unsuccessfully tried to eliminate cavitation. What they were actually experiencing was recirculation and not cavitation.

Simply put, cavitation is usually confined to the low pressure or suction side of the impeller whereas; recirculation is found in the impeller's exit area. The visual results of damage are the same in each condition but they are confined to the areas discussed. The implosions of the vapor bubbles have a devastating effect on the surface of surrounding metals, one expert estimated that the force of the implosion was in the area of (14.7 lbs/sq. ins)10 . This is a highly destructive condition in which pump impellers will fail. If vibration analysis, or ultrasonic programs are being used to support a predictive maintenance program, both cavitation and recirculation problems can be detected early enough to avoid problems. The following spectrum visually shows the difference between these two phenomenons.

Cavitation

The area of cavitation is within the eye of the impeller, therefore, there are no vane pass frequencies. Cavitation can cause axial surging, which, if excessive, will cause damage to both bearings and seals. The spectra produced have a very broad frequency range with low amplitude. The highest amplitudes are most often found in the suction area of the pump and are usually highest in the axial direction. It is also significant to note that in cavitation the peaks are non-synchronous, and there is also a lot of broadband noise.

Recirculation

Note the difference between cavitation and recirculation. It is very noticeable that there is now a clear peak at the impeller vane pass frequency. The cure for recirculation is to operate the pump at or close to its rated capacity for the operating conditions and to avoid excessive throttling on the discharge side. Recirculation can also be identified by washout of the impeller exit faces. This usually shows up as a polished, sharp edge on the impeller exit face.

From experience, it is essential that those people, who are assigned to troubleshooting activities on pumps and other machinery, fully understand the basic theory of how their charges work. An experienced operator should be able to recognize and control such conditions as cavitation and recirculation without having to bring in an engineer to determine the problem. If this basic knowledge were introduced during the initial training process, the amount of money spent on maintenance would be minimal. The personnel operating and maintaining the equipment would be much happier too.