Effects of Selected Water Chemistry Variables on Copper Pitting Propagation: Experiments and Modeling

Effects of Selected Water Chemistry Variables on Copper Pitting Propagation: Experiments and Modeling

The use of copper tubes accounts for over 80% of all pipes used in potable water distribution systems in North America and Europe However, pitting corrosion occurs in some potable waters – accounting for over 60% of failures seen in copper piping.  Although the subject of much study, many of the precise details of the factors controlling pitting particularly its connection to water chemistry details remain uncertain. For instance, the manner in which a pit propagates in certain water chemistries after a pit site has stabilized has not been examined systematically. Some of the critical unknowns include the effects of water chemistry and corrosion product formation on the rate of pit propagation, whether pits are stifled at long times in certain water chemistries or continue to grow at slow rates as well as whether pitting propagation can be suppressed by altering the water chemistry once initiated.
The pitting propagation behavior of copper (UNS C11000) was investigated from an electrochemical perspective using the artificial pit method. Pit growth was studied systematically using a range of dilute HCO₃^-, SO₄^2- and Cl^- containing-waters at various applied potentials. Pit growth was strongly potential dependent at intermediate potentials in all solutions. Pit growth was consistent with Ohmic controlled propagation at intermediate potentials and mass transport control at high potentials. However, pit growth rate was independent of the bulk solution conductivity of various waters. Instead pit propagation was mediated by the nature of the corrosion products formed both inside and over the pit mouth (i.e., cap). Certain water chemistry concentrations such as those high in sulfate were found to promote fast pitting that could be sustained at low applied potentials. In contrast, Cl^- containing waters without sulfate ions resulted in slower pit growth. High HCO₃^- concentrations relative to Cl^- or SO₄^2- suppressed pitting. These observations were interpreted through understanding of the type and amount of corrosion products formed inside and over pits as well as their resistive nature.   Modeling work was performed at Swerea-Kimab using finite element software. A cylindrical pit geometry was assumed. A potential was applied between the mouth of the pit and the actively corroding pit bottom. The model included copper oxidation kinetics, transport by migration and diffusion, Cu(I) and Cu(II) solid corrosion product formation govern by equilibrium thermodynamics and a saturation index, as well as pit current and depth of penetration. Malachite, nantokite, cuprite, bronchantite and atacamite corrosion product formation were all considered as govern by the conditions inputted and ion accumulation.   Stifling was predicted based on a decrease in pit current assuming a corrosion product resistance proportional to film thickness and porosity. The findings of the modeling were in good agreement with artificial pit experiments. The ramifications of these finings towards pit propagation in potable waters will be discussed. Moreover, possible mitigation of on-going pitting by alterations in water chemistry during pit propagation will also be discussed.