Cavitation on suction side downstream from cooling towers

[Boiler1]

Dear all,

I am involved in extension of a chilled water installation where cavitation had occured in the past.
The cavitaion occured on suction side of the pump set installed some 5 meters ( 6 yards) downstream from the cooling towers ( this distance is hydraulic- includes for bends, etc).

The cooling towers are lifted up on a platform 0.5 m ( about 2 feet) above the pumps.

At some point in the past, in an effort to reduce cavitation a by-pass from the return ( to cooling towers) header with a three port valve have been insatlled immediately before the pumps.
This seem to have worked well and reduced the cavitation (allthough not sure how).

Now i have to replace the original set of pumps with larger set due to the building extension.

Is there anything that can be done to prevent cavitation occurring again?

Lifting the cooling towers up higher is not an option due to some planning costraints.

Please help

 

[mtngreen]

Sounds like the cavitation was occuring because of inadequete NPSHa.

1. Lower the pumps. If you can lower the pumps, install a sump.
2. Increase the pipe size from the towers and eliminate bends, etc.
3. Install pumps with very low NPSHr.

 

[imok2]

Sounds like pressure drop on suction side of pump is too big try to keep it no more then 5#

 

[ChrisConley]

The only thing that mtngreen doesn't mention, would be to lower the vapour pressure of the water (by lowering the temperature).

Not likely a solution, but the operation of the cooling towers could be arranged to produce cooler water.

But seriously, try the three things that mtngreen recommends.

 

[SAK9]

The three way valve you mention is meant for tower bypass in winter to keep the water temperature high enough so that the chiller may not trip on low pressure.Since it is passing water at high pressure( a bit less than pump discharge pressure after allowing for line losess)it is helping to raise the suction pressure thereby reducing cavitation.

I believe you need to find out from the pump supplier what is the required NPSH for the pump.At a glance you seem to have adequate head to keep the pump away from cavitation(0.5 m+water column height in the basin).I would suggest checking a few things like suction strainer within the tower sump for clogging,level of scaling inside the suction pipe,design check on pipe size,correct use of reducers on pump suction etc

 

[ordesign]

How's basin water level?

 

[Boiler1]

As far as i know the NPSHa = atomspheric head + static head - friction losses - water vapour pressure for given temperature.

With the above calculated i should be well in the safe zone so i think there must be some localised pressure drop causing the cavitation. Could not see anuthing obvious though.

Thanks for all the input it was really helpfull.

 

[marcoh]

I recently completed a project where I had a similar issue. The chillers were located about 50m from the cooling towers and a few meters lower and in theory there should not have been any NPSH issues but there were.

We had to relocate the cooling towers and used the opportunity to increased the pipe diameter and raised height of the cooling towers by 1500mm and no more issues.

I realise you are constrained and cannot lift the cooling towers but can you move the pumps to adjacent the cooling towers?

 

[quark]

Cavitation in cooling water system is rare. May be a clogged strainer or air vortexing in.

 

[DrRTU]

With only 2’ of NPSHa, chances are you are you are droppings below NPSHr. Have you checked the actual pump curves? The pumps that were installed maybe different than the ones specified. The impellers may have been trimmed. Make sure you have the correct impeller rpm curves. Suction strainers and undersized inlets cause a large loss in suction head. As other have pointed out, vortexing air will destroy an impeller fast. Have you looked into the basin during non bypass low load operation?

 

[SAK9]

I would tend to agree with Quark simply because a typical centrifugal pump can draw water from 6~8 m below(depending on water temperature) its impeller eye.So you need some abnormal conditions for caviatation to occur in a cooling tower installation

 

[sterl]

Air entrainment much more likely than saturation basis of NPSHR if this a ordinary cooling tower installation.

If you have something on suction side that represents a big restriction, air could be finding point of entry other than sump pipe stub..

Other than that: Lower and widen the pipe stub....You'll end up with a wide rectangle for your intake with a deliberate high point from which you run a vent to well above the cooling tower level.

 

[quark]

In the worst case scenario, if your water temperature is 50C, the vapor pressure will be about 0.123 bar (or 1.23 m) and to induce conditions of cavity inside the suction piping, you have to create resistance equal to 8 meters (I am considering 0.77 meters for entry losses and accessories).

If you have a suction pipe length of 10 meters then the the flow velocity that can induce cavitation will be as below(approximately).

1" - 4.4m/s
2" - 6.26m/s
3" - 7.92m/s
4" - 8.86m/s
5" - 9.9m/s
6" - 10.85m/s
8" - 12.53m/s
10" - 14.01m/s

These are tremendously higher velocities for a suction line (ofcourse, even for discharge line as well). Your problem is something else.

 

[rmw]

If you find no obstruction and your pipe size is adequate to prevent high velocities in the suction, I vote for air entrainment. How deep is your basin and does the basin exit have any vortex prevention means? Also, the air bubbles that are introduced into the basin water by the falling of the water through the cooling tower might be getting picked up and transported to the pump.

thread407-238094 has some very interesting reading on the subject by some very knowledgeable people. Read all of it as well as the thread linked in it and all of the threads linked there if you want to really get your mind around your problem. I just did because I am working on just such a problem.

rmw

 

[Gunner777]

Boiler1,

I may be getting the wrong end of the stick here....Assuming of course the correct pumps are installed and the pipe sizes are correct can the pump not be moved so that the cooling tower is on the supply side of the pumps and install a pressurisation unit on the suction side of the pumps. If air is a problem then install a pressure step de-airator on the suction side of the pumps before the above pressurisation unit. This will ensure that the suction side of the pumps has the correct pressure to operate properly and the de-airator will ensure that the water has very little air saturation. This will enable the de-airated water to naturally asorb any trapped air in the system on the way round..........

 

[imok2]

Here are some things to consider:
.All piping must pitch upward to the tower basin, if possible, to eliminate air pockets. Again, this is good advice, but in my experience, this alone will not ensure trouble-free operation. If there is a long run of relatively low-velocity condenser water piping between the cooling tower and the condenser water pump, entrained air has a tendency to come out of solution. It may not get back to the tower through the suction line. If this now free air is ingested into the pump suction, it seems to exacerbate, if not cause the noise problem. The elevation of the tower basin water level was approximately 6 ft above the pump suction.
The calculated pressure loss in the suction line was approximately 5.5 ft. This should have produced a slightly positive pressure at the pump suction. The pressure gauge at the pump flange read 0 psig as close as we could tell. The pump was noisy. Since the gauge pressure at the pump suction flange was less than the NPSHR, there was the real possibility that the cause of the noise was due to air cavitation.
I removed the gauge connected to the pump suction flange and opened the gauge cock. The noise in the pump went away. I concluded that we were venting enough entrained air through the gauge line to eliminate the problem. We raised the gauge line to an elevation above the water level in the tower basin to prevent drain-down on shutdown, left the gauge cock open and the gauge off, and considered the problem fixed.