Brackish Water 5

[clay87]

We currently have vertical pumps circulating brackish raw water for cooling in the plant (50-95degF). Chlorides are about 1000 mg/L and usually two of the three pumps are running. The pump columns and castings are Ni-Resist, some other internals are 300 series stainless. We experience quite a bit of erosion, pitting, and cracking of the Ni-Resist parts. We are limping along by outsourcing braze repairs, metal stitching, etc. Pics attached.

1. Please provide insight as to what the degradation mechanisms in the photos are.

2. We are entertaining a proposal for new pumps in which the Ni-Resist castings are replaced with SS316 (cost, casting difficulties, weld-ability, etc.). I know this is a no-no in brackish water but it seems the stainless parts that are in the pump now are holding up well - alot better than the Ni-resist. Any thoughts?

ps. We are also being quoted for a duplex SS material but I predict that we cannot stomach the cost.

///Good, fast, cheap: pick any two...///

 

[rmw]

What else is in that brackish water besides chlorides?

And, we ought to be coming to you for answers. You are the first I've seen to be able to post multiple pictures. Slick how you did it. Congrats.

rmw

 

[unclesyd]

We have 5 large vertical pumps on a river near the head of a bay. The material is CI protected by extensive sacrificial anodes. This CP system also protects the sheet pile wall and the traveling screen. At our location we normally get a saltwater tongue on high tides and during particularly dry spells we get salt water for extended periods of time. Thee pumps are pulled once a year and have to replace the first stage bowel. About every 4 years we have to replace the first and second stage impellers. I can't recall any of these pumps having mechanical damage from operations.
I'll try to get some specific informato if I can't someone in yhr poweer house I know.

Addenda:
We have and an excellent location as at certain times of the year when we get a salt tongue one is able pick enough large shrimp to fill a 48 quart cooler in about 30 minutes.

 

[strider6]

At 1000 ppm Chlorides and 50 °C also SS should corrodes, i think that in your case it could be that the Ni resist is acting like a sacrificial for the SS. 1000 ppm is usually considered as the "safe" limit for SS like SS 316.
Stainless Steel is suitable for high chlorides conc if it's in service but in case of an idle pump there could be corrosion.

See this paper from NIDI:

http://www.stainless-steel-world.net/pdf/11004.pdf

and

http://www.corrosionist.com/Corrosion_Control_Seawater.htm

hope this help

S

Corrosion & Rust Prevention Control

 

[EdStainless]

Your 316 is holding up now because the Ni-resist is acting as a sacrificial anode. If you go all 316 it will be a disaster.

The usual approach is to make the most critical parts out of the highest alloy, and choose what is sacrificed.

So your shaft is a very good alloy.
Your impellers are slightly lower on the galvanic series.
And this is what you have.
The bowls are Ni-resist and allowed to corrode. The problem that I see is twofold. The cracking is probably due to improper stress relief anneal after casting.
Secondly you are allowing these to sit stagnant. This is causing much more sever localized corrosion than was designed for.
If you use a better alloy than Ni-resist or try coating you will only cause accelerated corrosion of your impellers.
One approach is to install plain steel sacrificial anodes. These will need to be replaced regularly, maybe every 6 months until you can figure out a correct size.
You would be better off keeping the pumps running all of the time, or when you shut one off back flush it with clean water.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[clay87]

rmw
other things in the water: crabs, clams, bryazoa, catfish, redfish, alligators...
more water sample stuff: 8.0 pH, 3600 microS/cm conductivity, 1000 mg/L Chlorides (Cl), 1500 mg/L Chloride (CaCO3)

strider
Thanks for the helpful links. Our max temp is 95 degF, avg is 75 degF. One out of three pumps is idle.

Ed
We are currently coating the castings with a Belzona or Devcon (as seen in picture)to prevent erosion and bio growth. Could this type of coating lead to corrosion problem on our SS parts?

 

[metengr]

Re; Cracks

The cracking observed in the diffuser vanes and discharge bowl webs looks to me to be caused by fatigue initiation and propagation from some type of flow induced stresses. You either need to increase the thickness or modify the geometry of the vanes and webs to avoid this problem. I also viewed cracking at previous weld or braze repairs along the diffuser vanes, which will be a problem with Ni-Resist castings, period. If the cracks are small blend them out and do not braze or weld repair.

The cracks in the outer column looks odd to me because there should be no signifcant dynamic loading on these pieces. These could be cracks that originated from casting defects that some how are propagating in service under dynamic loading conditions from flow.

 

[blacksmith37]

Which Ni- Resist and which type of CI (flake, nodular,)?
I have seen very good performance of Ni- Al Bronze in similar pumps. (keep in mind heat-treatment or casting "shake-out" temperature makes a difference in this alloys' corrosion resistance).

 

[clay87]

Which Ni- Resist

A436 Type 2.

which type of CI (flake, nodular,)?

The stuffing box is A395 60/40/18 or A536 65/45/12.
Was that the question?

I also viewed cracking at previous weld or braze repairs along the diffuser vanes, which will be a problem with Ni-Resist castings, period. If the cracks are small blend them out and do not braze or weld repair.

On that one I think it was a poor attempt at weld repair a long time ago. We currently send off to braze, braze weld, or metal stitch, which should be better?? We aren't too worried about minor cracks; it's the ones that are several inches long or go through entire vane/web.

If you...try coating you will only cause accelerated corrosion of your impellers.

Can someone clarify or corroborate this? We currently coat with Belzona/Devcon (ceramic polymer coating). Are we potentially causing another issue?

By the way, still waiting on quote for pump made of 2205 or A890 Grade 4A material.