A 1600 mm ND butterfly valve was installed in the seawater outlet of a heat exchanger of a nuclear power station in 2001. During a routine inspection, after 34 months in service, inspectors noted that there was corrosion damage on the valve seating. We inspected the valve and also arranged for metal samples to be cut from the valve seat area for microscopic examination.

A general view of the valve is shown in Figure 1. For the most part it appears in sound condition. By contrast, the seating ring, which was originally nickel plated then overcoated with tungsten carbide, was very badly damaged.  Figure 2 shows the area where the coating has disappeared completely from the seating ring, leaving the carbon steel exposed to the corrosive effect of the seawater. It was estimated that 20-30 mm of carbon steel had been lost in some areas; suggesting a corrosion rate of the order of 10 mm/year.

Samples were cut from the seating ring where some coating remained in place. Figure 3 is a micrograph the substrate, the nickel alloy and the nickel/tungsten carbide layers are clearly visible. The thickness of each coating layer is in excess of 100 µm, as required by the specification.

On the other hand, the interface between the steel and the nickel alloy shows no evidence of a blast profile. The situation in Figure 4 is similar. Again, all three layers are clearly visible but there is no evidence of a blast profile. Moreover, in this figure we see a layer of scale between the nickel alloy and the steel substrate. This scale appears to have been present before the nickel was applied, rather than having formed subsequently.

This indicated that the cause of the premature corrosion damage was the coating contractor’s failure to prepare the steel surface by abrasive blast cleaning prior to hot spray application of the nickel alloy.

Click on an image for a larger view