Oxygen Tolerances in Carbon Steel

Don't know how much this relates to the oil and gas industry, but in the electric power industry, EPRI would have us believe that oxygen is our friend.

Carbon steel will form protective oxides - namingly magnetite and under higher oxygen concentrations will form more protective iron (III) oxides. These oxides are good at resisting general corrosion.

The predominant factors in determining whether the pipeline will be passive (protective) or non-passive are pH and potential (voltage) - the influence of these factors can be viewed on a pourbaix diagram for iron (steel).

However, if there are solid impurities in the pipeline, these may deposit out and can lead to the formation of concentration cells (and the onset of galvanic corrosion) in which case oxygen should be kept to an absolute minimum.

In the power industry, boiler oxgyen concentration limits of 10 ppb are not uncommon.
Hopefully this information gives you a bit of a start...

kclim,
Thanks for the post. It's been my experience that what you say is true - I used to work in steam plants and at high temperatures/high pressures, oxygen was a really bad actor (especially in stanless and inconel) as it was required for Stress Corrosion Cracking and Cl Stress Corrosion, but in carbon steel it was not a big deal.

What I was looking for is some support for the idea that oxygen is only a contributor to corrosion in mild steel under very limited environmental conditions. I brought up pH being a factor, but I missed potential. I wasn't familiar with a Pourbaix diagram (I guess continuing education is worthwhile after all). I've just gone through about 25 web pages on the subject and it does in fact help explain several phenomenon that I've observed in the field.

But, I'm still not clear on the role of oxygen in these reactions. The agressive corrosion cases (low pH and high postitive potential) are in the Fe++ and Fe+++ range and the surface-to-mild cases all have hydroxides. Seems to support the "oxygen is your friend" statement. Any idea what I'm missing here?

Thanks

David Simpson, PE
MuleShoe Engineering