Supplying the holy city of Medina with water from the desalination plant at Yanbu in Saudi Arabia was a substantial undertaking. The project required installing 176 km of pipeline and around 600 valves, of which 38 were main line ball valves of up to 40 inches diameter.

Water delivery commenced in July 1998; but within four months it was clear that most, if not all, of the main line valves were suffering from corrosion of the chromium plated cast steel balls (see figure). This sparked a dispute between the Saudi client, the Argentinean contractor and the Italian valve supplier; with the contractor and valve supplier averring that the client’s desalination plant had fed corrosive water into the pipeline.

We were called in to determine the cause of the problem. An examination of the water analysis records revealed that out-of-specification water had, as alleged, occasionally been supplied by the client. However, a detailed chemistry and corrosion assessment demonstrated that this alone could not have accounted for the observed corrosion.

Attention then moved to the valves themselves, and in particular the quality of the chromium plating. Microscopic investigation of a plating sample from one of the valves, together with a detailed assessment of the valve plating specification and quality control documents, confirmed that the cause of the corrosion was the quality of the electro-plating.

When applied correctly chromium plating provides a hard, wear-resisting surface that is, in itself, highly resistant to corrosion. However, chromium electo-plate has a tendency to form microscopic cracks. Accordingly, if chromium plating is applied directly to steel intended for corrosive service then there is a risk that fluid could penetrate the micro-cracks and attack the substrate. In these situations it is standard practice to apply a primary protective layer of nickel plating which does not crack. It is also normal to minimize the depth of micro-cracking in the chromium. This is achieved by maintaining the plating bath high temperature and by plating slowly.

In the case of the valves supplied for this project, unfortunately, the manufacturer’s specification omitted pre-plating with nickel. Moreover, it also failed to specify any control of the plating bath conditions. As a result, when the plated ball valves entered service, water penetrated the micro-cracked chromium to reach the steel. Rusting initiated immediately. Furthermore, the rate of corrosion was accelerated due to electrolytic interaction between the chromium and steel (galvanic corrosion). This explained the speed with which the problem became apparent.