Allowable Chlorides in Cooling Tower Circulating Water 3

[Mark31]

Cooling Tower suppliers nominate an allowable chloride level of approx. 750ppm (50C, pH=6-8) for circulating water. Corrosion resistance tables nominate a chloride level of approx. 150 to 250ppm for aqueous solutions in contact with 316 SS.

Can anyone tell me why there is such a discrepancy given that the cooling tower internals are of stainless steel construction.

Thanks

 

[Ashwini]

Chloride levels are varying. SS 316 can tolerate even around 3000 ppm chloride provided there is no fouling. However, in most of the cooling water applications there is likely hood of some fouling that is manufactures recommend a safe limit of around 500 -750 ppm of chlorides. Under fouled conditions around 100 to 150 ppm are known to cause pitting, crevice corrosion. Regards
Ashwini Sinha

 

[mcguire]

Also bear in mind that there is a great deal of variability in the corrosion resistance of 316 depending on the surface finish and composition of a given heat of steel. A sulfur level of 0.008% which is common for TIG welded material renders 316L no better than a 304 with 0.001% sulfur.

 

[rustbuster1]

I'm curious about McGuire's comment: surely if S was such a strong lowerer of pitting resistance it would be included in the PREN or at least flagged as such by the SS makers. I have never heard of this effect and would like more info.
A reason we ascribe for lower corrosion resistance of 316 welds - including TIG welds - is that they are partly ferritic and the Mo segregates out of the austenite into the ferrite. This results on parts of the cast structure having Mo content as low as 0.5%. The common remedy is to upgrade the filler metal to one that has more Mo, to mitigate the segregation effect.

I'll be interested to see where the S effect is referenced.

 

[mcguire]

Rustbuster1
My best information from some unpublished research done by Ugine when I was in charge of R&D for one of their units. 5t showed that the sulfur levels requested by US stainless pipe producers for easy TIG welding reduced the pitting potential of 316 to that expected of 304. Similar observations were made by Osozawa et al in "Stainless Steels '87" and Haynes in " Applications of Stainless Steel 1992". I am about to try to do the definitive experiments to refine the PREN formula to account for sulfur. I expect it to approximate:
PREN=%cr+3.3x%Mo +16x%N-1000x%S.
The sulfur in the form of MnS acts as the nucleus for pitting sites. Passivating with acid removes surface sulfides, removing the effect until they are exposed by abrasion, welding, or general corrosion. So the formula above would hold for surfaces whose composition represented the bulk composition, which is seldom the case for prepared surfaces, but always potentially there for steels with unnecessarily high sulfur.

 

[mcguire]

Rustbuster1
An addendum. The stainless steel producers do know of this effect, but the pipe and tube producers are their customers, so they stay quiet. The tubing producers could get rid of the need for sulfur by switching to plasma or laser welding which doesn't benefit from sulfur. The steel producers would be happy also since sulfur also causes defects for them. As far as I know only Copperweld is converted to laser for stainless.
This isn't very relevant for ferritic stainless which generally has titanium stailization. The titanium also prevents the formation of MnS, forming an insoluble TiS which doesn't nucleate pits.
Likewise duplex steels are so deoxidaized and desulfurized for reasons of steel production that they are never found with deleterious levels of sulfur. So, essentially it's a sickness confined to 304, 304L, 316, and 316L, but this is the great majority of stainless pipe and tube.

 

[Mark31]

I think my question has been lost to a discussion on TIG welding?
Cooling tower suppliers nominate an allowable chloride level of 750ppm. I assume this is due to experiance with the material and using TIG welds. SS suppliers quote a figure which is much lower. The question is...Why?

 

[mcguire]

Sorry Mark31
We can get carried away. The point is that the fabricators don't know! So, they play it safe. If you're looking for expertise, you will NOT find it with manufacturers of cooling towers. Even the stainless steel producers themselves have eliminated their experts in their severe financial losses of the last few years.

Bottom line, variability in recommendations reperesents lack of knowledgte. Scary, huh?

 

[Mark31]

mcguire....I am still confused.

The cooling tower number is HIGHER (allow.=750ppm) than the SS suppliers number (allow.=250ppm). This would suggest that the cooling tower supplier knows more about SS in chloride rich environments than the material supplier.

And yes, this is very scary.

 

[mcguire]

Let's put the difference in perspective. 316 has a possible variation within the composition specification of 5.9 PREN units. This alone implies a difference in tolerance to chlorides of a factor of 100. This is even discounting the effect of sulfur, which could easily make that difference a factor of 1000.
Based on that the difference between the two maxima isn't very great. It should caution you the specifications are generally too broad to be useful and you need to require the vendors to support their numbers with data.
If you want to be safe, you can write your own materials spec. ASTM allows you to tighten up specs. Then you can really design with some degree of certainty.

 

[cynwrig]

Long time ago switched from Hypo dosing to Chlorine dosing, finally to an organic chloride as biocide. The latter worked as well as the others and was easier to control. I remember that Betz were the suppliers - not an advert!!.
What cycles of conc are you operating at and what is the corrosion treatment if any?
Is this a new installation ? What is the source of the make-up water?
My instinct says that you are probably safe - but try to get lower Chlorides as very localized hideouts, eg fouling or deposits, will wreak their revenge. Most operators defy the text book by good mtce practices , includig system cleans either in total or as individual units.
Wished I could be more specific - more of an operator than a materials man!!