Operation Off the Pump Curve

That question is difficult, but as you already know...... the pump curve is at a base reading of water such as 68 or ?? degrees...... and condensate return the temps are higher........ I think your pump supplier would have more input on this.

Most of the time this water is fairly cold, as it sources from an outside CST, and is make-up water that originally sourced from city water, through a purification system that heats the water to a maximum of approximately 75 °F. This will affect pumping performance slightly as to the curve, but it makes low dynamic head pumping more likely to cavitate due to higher vapor pressure of water.

I know very little about pumps but in general when someone is demanding that something be used in a manner that doesn't comply with the data sheet then they must understand that they own whatever happens. It is very common for "they" to decide that we can go off the data sheet but that it is "my" problem when the smoke leaks out of the equipment.

To be honest, I thought something was wrong when an operator started asking me about this pump (that I have absolutely no responsibility for selecting or setting up), stating that "it makes a lot of noise", but every time he mentions this to higher ups, he is told to "leave it alone". I respecfully submit we should not operate the pump with the destination valve wide open, or the recirc valve more than just cracked.

Sounds like you are as far away from the "knee" as you can get. Never a good thing.

I'd say you are "probably" right, but that's based on a guess.

"Do at risk of damage to pump, no warranty" would be my response.

It is a Goulds pump model 3656 3x4-7, and from the pump curve in the brochure, should never be operated below approximately 20 ft TDH. Here we are considerably below the TDH minimum, and cavitation is definitely taking place (1) postion well below the operating curve, and (2) the terrible noise it makes while running. From my earlier calculations, we are operating with no more than 10 ft TDH when the pump is being run with full open outlet.

As I have already stated: Back-pressure the pump either at the discharge, or at the point of release to the hot-well to a pressure of 90 TDH (add 15-20 for CST level above pump, so 105-110 ft pressure (45-48 psig), in order to (1) provide maximum pump output, and (2) eliminate cavitation.

This information was sent to the operator initially posing the question.

That's about all you can do.

Maybe management doesn't care about periodic replacement of the impellers/pumps.

I did a cooling tower with that required 200 HP, 17 ft of suction head, 78 PSI and 3000 GPM for each pump hooked in parallel. Goulds were in the running and actually selected, at the last moment, I was reviewing the numbers for the 4th time and I panicked, (due to being over tired) and I called our suppler that I required a lower head...... they supplied a new pump line they just acquired and had just complete their first install called Rohlpumpen where it only required 6 ft of suction head. Very happy with it.

So wind down the outlet isolation valve a bit, get the pump up its curve a bit, and the cavitation will stop.

Are there any Engineers at the facility?

No, just MBA's and accountants.

There are people with engineering degrees who apparently specified this pump, but apparently are too broke to pay attention to the correct operation of the pump.

Well, that's nothing to do with CR4.

I'm not directly familiar with how to read pump curves, but looking at your situation, (Question for all readers) with a static lift of 18 feet is it possible that there is cavitation potential at the inlet of the pump?

I suppose I'm asking whether you need to consider not just the TDH, but also the situation at each of the pressure interface zones.

Possible, if the water supply has air entrained in it.... And not enough time or level for the air to settle out.

We are not "lifting" the supply, it is flooded, but flooded suction counts as negative suction lift in the TDH calculation. As to air, yes, this is water that is fairly cool with no real options for warming it up and degassing it. Therefore the soluble gases of oxygen and nitrogen account for approximately 20-30 ppm saturated level, and will be a serious problem if the suction is not throated properly, i.e. no real increase in velocity that results a substantial reduction of overall pressure at the suction, if so, then there will cavitation no matter what we do.

Notwithstanding 991's comment (which is probably not applicable) about entrapped air, cavitation is possible for any number of reasons including static lift, water temperature and SG, altitude of the plant, inlet plumbing configuration and the turbulence it causes and likely some other reason that don't come to mind.

there is no static lift applicable in this situation. The pump has a flooded suction with 15-20 average ft water head, the overall TDH = static head of discharge + suction lift + frictional loss head and comes out to about 10-15 ft best case. We are definitely below the lower limits of TDH of the pump curve for this pump. Closing the valve at the discharge point will increase TDH possibly up into the optimal area of the pump curve, and actually more output flow will take place. Sometimes people just aren't the brightest light bulbs around here.

A more comprehensive pump curve, including NPSH_R curves, could help. The NPSH_R curves are typically a skewed U shape. The right side curves up moderately to the right; the left side curves up more sharply as flow is throttled to near shutoff. Cornell pump curves are among the best in showing these details (disclaimer). You have the situation pegged about right, including throttling the discharge, and gradually opening it so long as stability is maintained.