Vertical pump impellor damage, Water service 1

[sbeville]

After less than two years of service a five stage, submerged suction, bronze impeller pump stop pumping. The first three impeller are destroyed with the vanes completely gone. The last two impellors are heavily damaged. This pump is in water service, approximately 800 gpm @ 360ft. There were no reports of high noise levels.

I'm assuming cavitation but it seems unlikely that this much cavitation damage could be done in such a short time.

Has anyone experienced this before? Could galvanic corrosion be the culprit?

Thanks

 

[hydrae]

Cavitation is very fast yes two years is not unreasonable
have you checked the actual operating points?
I am thinking you may not be pushing to 360 ft since the pump continued to pump with 3 of 5 impellors gone.
what are the velocities near the 1st stage eye?

I would doubt corrosion, since bronze is more noble than most metals in a pumping system, I assume the other metals in the area are iron based. Also if it was galvanic then I suspect all the impellors would be equally gone.

Hydrae

 

[JJPellin]

What does the damage look like? Cavitation damage should be localized, deep, jagged pitting. If there are any solids in the water, you could have erosion damage. That would be localized (in different locations) and have a smoother appearance. Corrosion could be somewhat localized if the corrosion mechanism depends on velocity (corrosion/erosion). I agree with hydrae that galvanic corrosion is very unlikely and any corrosion mechanism would have probably attacked all stages. I have seen extreme damage to a pump like this from cavitation (cooling tower service), erosion (water wells with sand) and corrosion (acid carry-over). A more detailed description of the damage or a picture would be very helpful.

In order to determine if cavitation is likely, you need to verify that you have adequate suction head margin and adequate submergence. Even if both of these are good, any pre-rotation of the incoming fluid could cause cavitation. Operation at very low flow could result in suction recirculation cavitation. But for the damage to progress through three stages, I suspect abrasive erosion or low NPSH cavitation.


Johnny Pellin

 

[TenPenny]

What sort of water? What's the water analysis?

When you say the impellers were destroyed, what exactly do you mean? Any pictures?

 

[rmw]

How deep is it submerged and how well designed is the well (sorry no pun intended)? Have you ever seen vortex swirls on the surface of the water near the pump if it is visible?

rmw

 

[Artisi]

Photo's tell a much better story than trying to describe it.

 

[dabluffrat]

Cavitation damage would be seen heavy on the first stage impeller, not so much on series stages. Is this salt water? How do the bearings look? Open lineshaft or enclosed (tube) lineshaft? If the lineshaft and bowl bearings are bronze as well, I would expect similar wear and it would be a material application problem. I assume the bowls are cast iron and coated? how did they look? If the bowls are OK, i would try going to stainless steel impellers (316 or duplex) with peak wear rings.

Did you know that 76.4% of all statistics are made up...

 

[wimple]

sbeville

A possible explanation is to look for ppm content of NH4 in the water. Particular combination of pump material (impeller, casing etc) could activate galvanic corrosion on pump impeller as it bronze which has the lowest electrolytic potential. When it happens It does the job very fast.

You should also look to the high temperature of the water. Normally it should be limited to 45/50°C, above this temperature a special design is usually quoted (duplex SS etc).

I would be very interested to know the results of the investigation of your case

Thanks in advance
Rgds
Wimple

 

[JJPellin]

The suggestion about ammonia is excellent. We have been converting lube oil coolers from bronze tubes to stainless steel because of the potential of damage from ammonia in the boiler feed water used to heat up the oil for unit start-up. We also holed through some bronze tubes in a surface condenser when ammonia built up in the system. i believe that the ammonia was found to be originating in a recovered process condensate stream from a hydrogen plant.

Johnny Pellin