SAGD short stories:
Centrifugal Pump Noise
“SAGD short stories” is a blog series about SAGD and process operations.
Noise in a Centrifugal Pump: Cavitation . . . or Not?
When it comes to operating centrifugal pumps, unusual noise and vibration should always be a concern. As a process operator, troubleshooting and fixing noise and vibration issues correctly and quickly can get your process back in order and prevent downtime due to premature pump failure.
Let’s look at a few things that can cause a noisy centrifugal pump, as well as a few troubleshooting tips.
First, consider the root cause. Does the problem lie with:
- the pump itself?
- the liquid being pumped?
- how the pump is being operated?
Rule out a physically damaged pump.
- Check if the noise is coming from the bearing housing or impeller casing / volute to rule out a bad bearing.
- Consider if the noise is vibration from rotating parts that are damaged, out of balance, or operating in harmonic frequencies.
Rule out a contaminated stream.
- Check if the suction valves are lined up to the right supply.
- Check if strainers or filters are bypassed.
- Check for leaks where air could be drawn into the suction.
Knowledge of your process helps here, since the cause of the noise
could be something as simple as an upstream upset introducing some
solids into a normally 100% liquid stream.
For example, knowing freshly made filter beds introduce
a certain amount of filter material to a process stream
can lead you to check with upstream operators to see if that is the cause.
Troubleshoot pump operation.
Liquid pressure and temperature are linked in a relationship where the liquid’s vapour pressure and boiling point increase and decrease together. The pressure changes that occur in a centrifugal pump can result in a momentary lowering of the liquid’s boiling point in the impeller area. If the boiling point falls below the liquid temperature, some of the liquid will vaporize (boil) and then quickly condense back to liquid as the pressure and boiling point go back up.
In a centrifugal pump, Cavitation occurs when bubbles of vapour (think steam) spontaneously form in the suction liquid entering the low-pressure zone around the impeller eye. The spinning impeller forces the liquid and bubbles into a higher-pressure zone, which implodes the bubbles on the surface of the impeller.
A similar situation called suction liquid Air or Vapour Entrainment occurs when there are already bubbles of air or vapour in the suction liquid feeding the noisy pump. In this case, the noise is bubbles imploding as they are forced away from the impeller eye into the high-pressure zone.
Internal flow Suction Recirculation occurs when a low suction flow rate results in a recirculation of some of the liquid in the suction (you can consider it as a cavitation event that takes place entirely on the suction side of the pump). The recirculation flow increases and decreases the liquid velocity, and therefore its pressure, which causes some liquid to vaporize and then condense (following the cavitation pattern). This is most common in high suction energy pumps operating at the low flow end of its performance curve.
Regardless of the mechanism,
it is the collapse of a void in liquid that causes the noise and vibration;
much like condensate induced water hammer.
The implosions and shock waves can break off tiny pieces of the impeller . . .
. . . and there can be hundreds of these events per minute.
How do we tell the difference?
The first thing to consider is the job the pump is doing.
For example, if a foaming problem in a tank
is followed by a noisy pump drawing from the tank;
consider foam entrainment in the suction liquid as a possible cause.
Without a quickly obvious cause . . .
. . . begin to close the pump’s discharge valve and listen for changes.
If the noise doesn’t change much, the problem is not associated with the flow rate,
and points toward vapour entrainment in the liquid.
If reducing the discharge flow rate quiets or eliminates the noise, the problem is associated with a high flow rate.
Since reducing the flow rate reduced the problem, consider cavitation as the cause, as reducing the flow rate also reduces the pressure drop in the suction zone.
If throttling the discharge makes the pump noise worse, the problem is associated with an already too-low flow rate (the pump is operating below its performance curve).
Suspect internal suction recirculation, which most often occurs in high suction energy pumps operating below the optimum flow rate range.
All of these problems will damage a centrifugal pump, and operators should do everything possible to solve them.
Instances of pump noise need to be addressed a.s.a.p . . . . but you need to know what is really happening when asking for help or repair work.
As an added benefit, few things make an operator look better than being able to quickly and accurately assess pumping problems.
To find out more about how to operate in the Canadian Oil Sands,
you can view Contendo’s SAGD Oil Sands Online Courses.
Kevin Fox is a senior technical writer at Contendo.
He is a power engineer who has written process education programs for industrial clients since 2009.