This thread is a technical continuation of the discussion on failures of copper heat exchangers in gas heaters in pools/spas discussed in this thread starting at this post.
This link and this link give nice summaries of copper pipe corrosion. This link provides more technical detail when chlorine is present, as is often found in household plumbing (up to 1 ppm FC with no CYA) and is of course present in pools (though with CYA in the water the equivalent amount is about 0.1 ppm FC). The description of the possible corrosion mechanism sounds suspiciously similar to that of stainless steel in terms of disruption of a passivity layer (in this case, of copper oxide). It's interesting to note that Dissolved Inorganic Carbon (DIC) plays a significant role by enhancing the solubility of copper(II). In pools, DIC is carbonates and the 0 to 20 mg C/L where the effect is strongest is equivalent to a TA of 0 to 153 ppm (ratio of roughly 100/12 calcium carbonate to carbon molecular weights and adjustment at pH 7.5 of 0.915 bicarbonate to total carbonate) and that this effect is stronger at higher pH above 7.5.
This research at the EPA showed that pitting corrosion occurred at low DIC levels of 5 and 10 mg C/L which is a TA of 38 and 76 ppm and high pH water of 9 in the presence of chloride at 20 mg/L (that's VERY low) and was not observed at pH of 7 or 8. This link seems to confirm chloride's role in the mechanism of pitting attack on copper. This link from Health Canada implies that low levels of chloride may reduce copper corrosion, but at high chloride levels it may increase pitting corrosion -- note that "high" in this context won't be very high with respect to pools since this is about drinking water (too bad they don't give more specific levels).
It looks like the general consensus from the above and other articles is that sulfates can play a significant role in pitting corrosion and that chlorides also increase pitting corrosion. This is very similar to what is said about stainless steel and appears to be consistent with the inhibition of forming a passivity layer. When one combines this with chlorine oxidation and, most importantly, with water flow rates through the tubing, corrosion seems inevitable. This link gives a lot of detail of the various factors and types of corrosion, but unfortunately nothing definitive on specific chloride levels.
Unfortunately, none of this would specifically determine how much faster corrosion would be at higher salt levels such as found in SWG pools. It only indicates that such corrosion is more likely to occur at faster rates at higher levels, but not specifically by how much. This link from the EPA on stainless steel corrosion of various alloys says the following on page 28:
Non-halide salts have little effect on stainless steels, but chlorides particularly tend to promote pitting, crevice corrosion, and stress-corrosion cracking. In some cases sulfates seem to aggravate the effects of chlorides. Chlorides present in amounts as little of 0.3% with sulfates present can produce severe corrosion. Even quite low concentrations of chlorides can cause corrosion when concentrated by occlusion in surface films.
So it may just be that the thinner less expensive heat exchangers combined with high flow rates, U-tube bends, and higher salt levels push corrosion rates higher so that they become more visible. Fixing any of these factors would likely help. The easiest approach is to use more corrosion-resistant materials such as cupro-nickel or titanium. So if you are in the market for a gas heater, it would probably be worth getting one with such materials, independent of your pool type but most especially if you have the higher salt levels in an SWG pool.
Richard
This link and this link give nice summaries of copper pipe corrosion. This link provides more technical detail when chlorine is present, as is often found in household plumbing (up to 1 ppm FC with no CYA) and is of course present in pools (though with CYA in the water the equivalent amount is about 0.1 ppm FC). The description of the possible corrosion mechanism sounds suspiciously similar to that of stainless steel in terms of disruption of a passivity layer (in this case, of copper oxide). It's interesting to note that Dissolved Inorganic Carbon (DIC) plays a significant role by enhancing the solubility of copper(II). In pools, DIC is carbonates and the 0 to 20 mg C/L where the effect is strongest is equivalent to a TA of 0 to 153 ppm (ratio of roughly 100/12 calcium carbonate to carbon molecular weights and adjustment at pH 7.5 of 0.915 bicarbonate to total carbonate) and that this effect is stronger at higher pH above 7.5.
This research at the EPA showed that pitting corrosion occurred at low DIC levels of 5 and 10 mg C/L which is a TA of 38 and 76 ppm and high pH water of 9 in the presence of chloride at 20 mg/L (that's VERY low) and was not observed at pH of 7 or 8. This link seems to confirm chloride's role in the mechanism of pitting attack on copper. This link from Health Canada implies that low levels of chloride may reduce copper corrosion, but at high chloride levels it may increase pitting corrosion -- note that "high" in this context won't be very high with respect to pools since this is about drinking water (too bad they don't give more specific levels).
It looks like the general consensus from the above and other articles is that sulfates can play a significant role in pitting corrosion and that chlorides also increase pitting corrosion. This is very similar to what is said about stainless steel and appears to be consistent with the inhibition of forming a passivity layer. When one combines this with chlorine oxidation and, most importantly, with water flow rates through the tubing, corrosion seems inevitable. This link gives a lot of detail of the various factors and types of corrosion, but unfortunately nothing definitive on specific chloride levels.
Unfortunately, none of this would specifically determine how much faster corrosion would be at higher salt levels such as found in SWG pools. It only indicates that such corrosion is more likely to occur at faster rates at higher levels, but not specifically by how much. This link from the EPA on stainless steel corrosion of various alloys says the following on page 28:
Non-halide salts have little effect on stainless steels, but chlorides particularly tend to promote pitting, crevice corrosion, and stress-corrosion cracking. In some cases sulfates seem to aggravate the effects of chlorides. Chlorides present in amounts as little of 0.3% with sulfates present can produce severe corrosion. Even quite low concentrations of chlorides can cause corrosion when concentrated by occlusion in surface films.
So it may just be that the thinner less expensive heat exchangers combined with high flow rates, U-tube bends, and higher salt levels push corrosion rates higher so that they become more visible. Fixing any of these factors would likely help. The easiest approach is to use more corrosion-resistant materials such as cupro-nickel or titanium. So if you are in the market for a gas heater, it would probably be worth getting one with such materials, independent of your pool type but most especially if you have the higher salt levels in an SWG pool.
Richard