this week i had to go check a problem for a long term customer that she was having with her furnace. she had had some company come in and do a furnace service, and a few days after they left, she was having water leaking from near the furnace. i just assumed they broke something or knocked it loose.

i went and had a look and was shocked to find that a brass nipple (1/2" MIP to 1/2" pex barb) had partially dissolved. yes, actually dissolved. it had actually lost about half of its total mass (compared to another identical fitting) and the dissolved portions matched identically to the probable flowpath of the liquid condensate. the fitting was acting as a connector between the furnaces condensate collection port and a 1/2" drain line to a floor drain. it was threaded into a nylon collection tray and connected to a plastic drain line.

i have never in my life seen anything (outside of process equipment) like this. whats even crazier is that the furnace and all its plumbing was replaced 3 years ago. my thoughts are:

a) dielectric dissolution (galvanic corrosion) - if there is a small current leak through the furnace condensate to ground in the drain (current flowpath is the water), its theoretically possible there could be some low level ionization in the water, and thus some corrosion. this seems so unlikely as to be almost unmentionable.

b) the furnace cleaners used some sort of acid to clean out the furnace and it ate through the fitting. i know of lots of acids capable of doing this but why would they use it considering how much other damage it would do inside the furnace. i just cant see this either.

c) could the condensate water be acting like distilled water (i guess the interior of a furnace is really is like a distillation tower if you think about it). over time distilled water can dissolve just about anything.

d) could carbon dioxide from the combustion chamber be dissolving into the water, making dilute carbonic acid - causing the corrosion?

if anyone has ever seen or heard of this or knows whats going on, i would appreciate the input



  • Brass may contain some lead (2% is one typical mixture). Given the news from Flint reminding us of the hazard of improperly treated water,, it might be a good idea to check chemistry/pH of both the water in the system and water coming into the house...
    – keshlam
    Mar 6, 2016 at 3:08
  • 1
    Cleaning the evaporator coils a cleaning solution is normally used but in 3 years that is fast. I have seen DI water eat everything but Teflon and stainless. Your idea about galvanic corrosion sounds the most plausible to me as I have seen galvanized to copper connections fail in a similar time frame metal to metal.
    – Ed Beal
    Mar 6, 2016 at 3:34
  • Not likely to be an acid used by furnace service people. It takes a fair amount of concentrated nitric or sulfuric to dissolve grams of metal as you describe. If they were dribbling that much around, you'd see evidence elsewhere. May 6, 2016 at 15:47
  • 1
    sorry - should have updated this. turns out it is a mild acidic solution of carbonic acid that comes from the combustion gases and water vapour. i talked to an hvac contractor i use and he said that some types of high efficiency furnaces are known for this. even to the point that some of them have stainless steel heat exchangers. the secret is to plumb any drain lines with only plastic fittings. its also been known to dissolve the internal copper pipes in older floor drains, so it should be something anybody should be wary of. May 6, 2016 at 15:59
  • @wayfaring stranger - you are incorrect. it takes very little concentrated nitric or sulfuric acid to dissolve almost any metal. it only takes about 400 millilitres (about a cup and a half) of 70% nitric acid (common grade) to dissolve 5g of brass (66%zinc, 34%copper). Zn + 2HNO3 ----> Zn(NO3)2 + 2[H] Cu + 4HNO3 ----> Cu(NO3)2 + 2NO2 + 2H2O May 6, 2016 at 16:12

3 Answers 3


I would think electrolytic corrosion to be unlikely due to the intermittent nature of condensate flow and the fact that the brass fitting is isolated between nylon and PEX.

Your thoughts around acidic condensate are intriguing. Boiler feed water for industrial steam systems is treated with inhibitors to prevent yellow metal corrosion in the condensate collection systems, and industrial closed-loop cooling systems are treated for the same reason, so it is a known fact that yellow metals are susceptible. Pure water easily absorbs carbon dioxide just from being exposed to air, so the gas-rich environment of the combustion chamber may be accelerating the process.

I would replace the nipple with a plastic component, that should take care of the issue either way.

  • i agree that the electrolysis idea seems unlikely, but there has to be a cause thats reasonable. "yellow" metals (copper alloys) are prone to accelerated corrosion due the electroreactivity of copper as an alloying member, and is why brasses and bronzes are used instead of pure copper - but this is a residential furnace. the offending connector was replaced with glassed nylon. i just want to know WTF happened Mar 6, 2016 at 22:39

As I understand the fitting was between a nylon tray and plastic hose; If so, galvanic, stray current, dissimilar metals had nothing to do with the corrosion . Just corrosion by soft acidic ( CO 2 ) water and nothing else. It would have had a good chance of dezincification but an examination would be needed. Just replace it with plastic. Brasses are usually good in water EXCEPT soft acidic waters. I have assumed this is a modern very high efficiency furnace.

  • Copper alloys do not work in soft acid water. That is why home AC units are aluminum and not copper like old auto radiators. ( Although the aluminum corrodes in several years. ) The AC condensate is bad, but your combustion condensate causes even more corrosion. Feb 17, 2018 at 14:15

On a high efficiency gas combustion appliance nitrogen dioxide and nitrous oxide dissolve into the water vapor to form nitric acid. Brass reacts with nitric acid.

  • It's only the NO2 that does this, not the N2O, but yeah -- NO2(g) + H2O(l) <-> HNO3 (aq) Feb 17, 2018 at 14:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.