Dezincification of Brass in Potable Waters
This report was produced for the Urban Water Research Association of Australia, a now discontinued research program.
Report No UWRAA 84
Dezincification of copper/zinc alloys, particularly forged duplex brass, remains a serious economic problem to both the Water Industry and its customers. The present understanding of the mechanism(s) of dezinfication is reviewed, together with the influence of both the metallurgy of the brass and potable water composition. Previous attempts to use electrochemical methods to study corrosion rate are also detailed.
The development of a potentiostatic method to determine instantaneous corrosion rates of duplex brass in a variety of potable water types is described, and the methodology is shown to provide reliable estimates of dezincification rates by correlation with solution analysis techniques. The method has been used to compare the dezincification propensity of a wide range of Australian drinking waters ranging from very soft sources to high chloride supplies that are barely potable. The effect of a range of water chemistry parameters including alkalinity, chloride, temperature, pH and free chlorine residual have also been studied.
This work indicates that the traditional parameters of chloride and alkalinity have considerably less effect on corrosion rate of duplex brass than previously thought. Hard waters, with relatively high levels of chloride are in facts lightly less corrosive than soft, unbuffered supplies due to the lack of scale-forming abilities of the softer waters. However, the effects of pH and, particularly, free chlorine, are quite marked and if inadequately controlled can lead to unacceptably high dezincification rates.
Field testing of duplex brass components has shown the difficulties in correlating material composition and structure morphology with both field established corrosion rates and the currently accepted accelerated test method. The latter test showed, at best, a lack of repeatability in predicting actual field performance on even a trend basis. Correlation between dezincification penetration and structure morphology was likewise quantitatively poor, although a straight line correlation can be attributed to the relationship between volume fraction of beta and depth of penetration for duplex brasses showing discrete and finely distributed beta phase. No correlation existed for brasses with coarse, interconnected beta.
Both electrochemical testing and surface analysis (SIMS) on duplex, beta, alpha and arsenic-inhibited alpha brasses strongly indicates that at neutral pH the mechanism of dezincification is that of preferential dissolution of zinc. Alpha brasses, however, appear to be capable of a dissolution-redeposition mechanism at lower pH as suggested by Heidersbach and Verink. Beta brasses, by contrast, only dezincify through a preferential dissolution mechanism which suggest that arsenic inhibition will only work with alpha brasses when the dissolution-redeposition mechanism is operable.