304 SS Cooling Coil Failure

[goomba]

A 304 SS finned cooling coil is installed in an oven and is connected to a chilled water loop that operates at 50°F and 40-60 psig. The ovens are operated up to 165°C for approx 4 hours. Then the water is turned on for the cooling cycle which lasts approx 1/2 hour. The ovens have operated continuously for 1 year. Recently, 4 coils have developed leaks. One coil has been sent to a materials testing lab for investigation. There is very little presence of chloride in the grain boundary (0.2%). I'm having difficulty getting a water sample due to a customs/shipment issue. Haven't ruled out thermal shock (or anything else) as the culprit. The ovens are used for epoxy curing. If anyone has an idea for a solution it would be much appreciated.

 

[kenvlach]

Hi,
My diagnosis is stress corrosion cracking. It would be good to have more details regarding the water and the HXer materials and fabrication, but 304 SS has well-known susceptibility to SCC (stress corrosion cracking) in the presence of chloride. I don’t know what kind of laboratory you used for testing or the method of analysis for chloride, but I think 0.2% (2000 ppm) inside an SS grain boundary is enormous and obviously the major contributor. You probably have severe corrosion during the 165 C heating cycle due to salt and other impurities in the residual water trapped inside, plus the inrush of cold water flashing to steam having an erosive effect on any protective film. Are your coils failing in areas of maximum deformation (bends)? If the coil has sharp bends, the thermal expansion and contraction may have contributed stress, plus if the valves are closed tightly, there will be about 100 psi (0.70 MPa) steam pressure inside during the heating cycle.

As to what you should do, you are fortunate to have a closed loop system since it is cheaper to treat the water with corrosion inhibitor than to buy HXers made from more expensive grades of SS or titanium. I suggest that you find a local supplier of boiler water treatment chemicals and explain the situation. Flush out the system* with the best purity water you have available, then add your treatment chemical(s).

*If your HXers are welded entirely of 304 SS, it would also be a good idea to first have them annealed (unless you can find out that the manufacturer did this) and passivated (including the insides). The lab should be able to tell you whether the HXers were annealed if they did a proper metallurgical examination.

 

[kenvlach]

Hi again,
I think it’s maybe a good idea to give some references (although you knew to look for chloride, so you already have some).
***********************************************************
‘CORROSION RESISTANCE’ from Atlas Steel gives a concise description of SCC.
(

http://www.atlas-steels.com.au/techinfo/index.html)

“Stress corrosion cracking (SCC) Under the combined effects of stress and certain corrosive environments stainless steels can be subject to this very rapid and severe form of corrosion. The stresses must be tensile and can result from loads applied in service, or stresses set up by the type of assembly e.g. interference fits of pins in holes, or from residual stresses resulting from the method of fabrication such as cold working. The most damaging environment is a solution of chlorides in water such as sea water, particularly at elevated temperatures. As a consequence stainless steels are limited in their application for holding hot waters (above about 50°C) containing even trace amounts of chlorides (more than a few parts per million). This form of corrosion is only applicable to the austenitic group of steels and is related to the nickel content. Grade 316 is not significantly more resistant to SCC than is 304. The duplex stainless steels are much more resistant to SCC than are the austenitic grades, with grade 2205 being virtually immune at temperatures up to about 150°C, and the super duplex grades are more resistant again. The ferritic grades do not generally suffer from this problem at all.

In some instances it has been found possible to improve resistance to SCC by applying a compressive stress to the component at risk; this can be done by shot peening the surface for instance. Another alternative is to ensure the product is free of tensile stresses by annealing as a final operation. These solutions to the problem have been successful in some cases, but need to be very carefully evaluated, as it may be very difficult to guarantee the absence of residual or applied tensile stresses.

From a practical standpoint, Grade 304 may be adequate under certain conditions. For instance, Grade 304 is being used in water containing 100 - 300 parts per million (ppm) chlorides at moderate temperatures. Trying to establish limits can be risky because wet/dry conditions can concentrate chlorides and increase the probability of stress corrosion cracking. The chloride content of seawater is about 2% (20,000 ppm). Seawater above 50oC is encountered in applications such as heat exchangers for coastal power stations.”

************************************************************
The Nickel Development Institute (

http://www.nidi.org)

has a lot of good, free publications. If your HXers are entirely SS (welded fins, not brazed), ‘Specifying Stainless Steel Surface Finishes’ explains corrosion of welds, and how to clean and passivate them (and for your case, include the insides):

http://www.nidi.org/index.cfm?fuseaction=tech_lit&ci_id=3092&la_id=1&type=pdf

Another relevant document is ‘PREVENTING STRESS-CORROSION CRACKING OF AUSTENITIC STAINLESS STEELS IN CHEMICAL PLANTS’, 1992. This report discusses stress corrosion cracking, SCC, that occurs with stainless steels in chemical plants. It focuses on SCC caused by chlorides and how to prevent it. It discusses HXers and suggest keeping the water pH about 9 to 12, using pure water, changing types of metal ions in solution [doesn’t say which, though].

http://www.nidi.org/index.cfm?fuseaction=tech_lit&ci_id=3170&la_id=1&type=pdf

STAINLESS STEELS FOR EVAPORATORS/ CONCENTRATORS (1981)

http://www.nidi.org/index.cfm?fuseaction=tech_lit&ci_id=3139&la_id=1&type=pdf

”Concentration of liquids, or the removal of moisture from products by evaporation, are process operations used in many industries” -- This is the situation inside your HXer during the heating cycle.
If you have a choice of materials for new HXers, this report gives recommendations on alloy selection for resistance to various types of corrosion such as SCC.

******************************************************
goomba,
As I considered various solutions to your problem, a very simple solution (better than using boiler water treatment) occurred to me, but you may need to slightly modify your system. I have presumed that you have an inlet valve, that fluid enters under 40-60 psi pump pressure, flows through the HXer and then drains freely to the return tank. But, if the return line is under some pressure, even when the water flow is turned off, then maybe you can use an automotive-type ethylene glycol-type radiator fluid (ethylene glycol has a b.p. of 197 C). If necessary, install a spring-loaded check valve in the return line, the idea is to keep the HXer filled with a corrosion-inhibited liquid during the heating cycle. If you cannot retain much residual pressure, you may have to add pure ethylene glycol to boost the boiling point.
In the USA, Prestone sells a product for automobiles which has a boiling point of 167.8 C (at 1 atm) before it is diluted to 50% with water. It consists of about 88% ethylene glycol, 4% diethylene glycol, 1% corrosion inhibitors, balance water. I suggest you find a similar product in your comppany, preferably a heavy-duty radiator coolant used for large diesel engines.

 

[mcguire]

SCC could have occurred with these tubes. The temperature and environment are right. But, it is easy enough to see if there are cracks. The cause could be SCC or simply pitting perforation. But since pitting is almost always the initiation site for SCC, it's better to simply address the pitting. At this temperature and chloride level, 304 pits. It does it even faster in tubing because tubing usually has higher sulfur ( 0.005 to 0.015%). MnS particles act as initiation loci.
It sounds like you need to move up to 316, which can also pit and SCC but is more resistant. Or you can use 2205, which is almost immune to SCC and won't pit in that environment.

 

[goomba]

Kenvlach and Mcuire,
Thank you very much for your responses. This is very useful information. This application is asking a lot from this cooling coil! Now the customer wants to investigate going to 300°C. I believe a secondary fluid pumped through the coil and cooled with a hx by the chilled water loop would be the best approach. That would eleminate the thermal shock and would allow for tighter control of the fluid quality. There would be a loss of thermal performance. Your thoughts on this would be appreciated.

 

[mcguire]

Whoa! That's getting nasty. Best economics for the low temp coil dictates 2205 in low chloride water at 70 to 80 C.

 

[kenvlach]

goomba,
For the 300 C, conditions, you are correct to switch to a non-aqueous fluid due to pressure problems with water-based.
I suggest that you contact a HXer mfr. such as Tranter

http://www.tranterphe.com/Default.htm

to get heat transfer coefficents, HXer sizing, etc., for a silicone oil (expensive) or other high temperature fluid medium. They have some on-line info, and a very good printed manual (which you can request from their rep. for your region, check their website representative listing).
With a non-corrosive fluid, you can use a less expensive HXer material.
For the intermediary HXer, if you use high quality water with treatment as previously suggested and maintain a continuous flow through the low temperature side, the conditions will be much less severe than previously, and 304 should work, but I would recommend 316 because you maybe cannot guarantee good future maintenance of the water system.

 

[goomba]

Some units have an all-copper finned cooling coil. Are there any potential problems with this given this history of failure with the 304 SS?

 

[kenvlach]

Copper is a very good HXer material if you use properly treated water (such as heavy-duty automotive type coolant) and observe the mfr.’s limits for pressure & temperature.

Under the previous conditions which corroded 304 SS, I think you would have serious problems. Regardless of material, you should maintain some back pressure in cooling line to avoid evaporation of liquid. This would also reduce stresses in comparison to cold water entering hot, empty line. Depending upon pressure limit of coil & the max. possible temp. of furnace, probably include a pressure relief valve & have it drain back to cooling tank. Don’t want to burst the coil if furnace accidentally overheats.

Please explain how the system was operated. Was I correct earlier?
“I have presumed that you have an inlet valve, that fluid enters under 40-60 psi pump pressure, flows through the HXer and then drains freely to the return tank. “
If so, the corrosion situation is very severe during the heating cycle, when residue from the cooling water is overheated under moist atmosphere. Even some anti-corrosive additives for liquid phase could cause problems.

 

[goomba]

The entering pressure is 40-60 depending on the actual chilled water system - there are 5 at various locations around the world. The outlet pressure is typically 20-30 psi lower than the inlet pressure. So there is always a back pressure on the coil. But not enough pressure to prevent the water from flashing to steam when the unit is at 165°C. The chilled water loop is fed originally from a reverse osmosis system and then treated with chemicals for anti corrosion, ph and bacteria etc. The treatment is supposed to be monitored by a 'professional' third party but not sure if they are watching for chloride. I'm trying to get water analysis done on all locations.

 

[kenvlach]

I suggest a water-50% diethylene glycol solution to avoid boiling (of course, some anti-corrosive package). Diethylene glycol has a b.p. of 245.5 C, so with some back pressure, the solution should handle 165 C.

Like I suggested previously, major causes of corrosion were the heating of wet residues inside the coil up to furnace temperature, erosion when water re-entered, plus shock. No way to maintain a protective film on inside of coil.