help writing pump alarm

[TheWaterboy]

I know this is an academic excercise that only uses pressure but I find it interesting.
Now I've not tried this solutionmyself because we use devices that monitor the current and voltage to detect this very thing, but I would think that with enough resolution, there would be enough change in current from normal to detect the motor responding to the potential causes of cavitation. Or for that matter anything outside the pump curve.
I don't understand how suction pressure by itself could detect cavitation if it stayed above atmo. Low suction pressure would tell you that cavitation was possible, but not that it was happening unless the pump curve and service curves were somehow factored in.

 

[OkiePC]

I think what Tom is inferring is the pressure differential compared with the chart, so yes a look-up table, but he (and you) are the pump experts.

I also think that he is talking about preventing cavitation, and what I have done with the available flow detector or discharge pressure sensor with a couple of timers is to detect long after cavitation has occurred to the point the pump is gulping air.

Please carry on, though, this is very educational. And do correct me if I misunderstood.

Paul

 

[Tom Jenkins]

Low suction pressure doesn't detect cavitation, it causes it!

Remember, cavitation is water boiling at low pressure areas inside the pump, usually near the impeller eye. Just like water boils at a lower temperature in Denver because of the lower atmospheric pressure, it will boil at 68F if the pressure gets low enough. then, when pressure builds up at the impeller OD the water "unboils", that is the vapor bubbles collapse. This causes erosion of the impeller.

Look at the attached example. If the pump has a 6 ft. suction lift there will be cavitaiton because the Net Positive Suction Head Available (14 ft) is less the the Net Positive Suction Head Required (15 ft.), even though the flow is 2300 gpm and the power draw is 70 hp. There will be flow until the impeller erodes so much the pump fails, that is.

Now, if the pump suction was flooded by 6 ft. instead of having to lift the water 6 ft the NPSHA would be 26 ft and the pump would be fine.

If the water were hotter then the pump might be in trouble again because the water would boil at a higher pressure.

If rags plugged the line and caused another 12 ft of pressure drop then the pump would cavitate at 68 F water temp, even with the flooded suction.

That's why you need to select the pressure transmitter on the pump suction carefully. You need to know what the inlet conditions will be. An absolute transmitter or a compound transmitter will always work. If the pump suction is flooded then a guage pressure transmitter would work. If the water level in the sump is below the pump inlet then a suction (negative pressure) transmitter will work.

 

[TheWaterboy]

I get what you say, but the question was for a "pump cavitation alarm". I interpreted that to mean "detect the occurrence" but I suppose it could mean to "warn about the possibility".
Preventing it is preferred of course, but detecting it in a factory setting would be in interesting exercise.

 

[Tom Jenkins]

Well, it is a matter for some interpretation, I guess. I took it to mean alarm on impending cavitation, since actual cavitation occurrence is a destructive phenomenon. I'm not sure how you could detect cavitation unless you monitored for sound or vibration. In testing a pump you can hear cavitation, and it isn't a pretty sound!

 

[ganutenator]

thanks all