Pilot valve instability and subsequent damage during design/development testing of a PWR plant reactor circuit relief valve, and development of a relief valve test facility

This paper discusses problems encountered during design/development testing of a PWR Plant Reactor Circuit Relief valve. It was found that test facility system and detail valve seating geometry factors led to pilot valve instability, resulting in material loss a Stellite 6 hard facing and subsequently leakage. Possible causes for the material loss were considered to be either cavitation, flow erosion-corrosion, or mechanical vibration/impact. Following metallurgical examinations, the most likely cause was considered to be flow erosion-corrosion, resulting in anodic dissolution of the matrix phase. The work concluded that the material 'scarring' present was inconsistent with cavitation damage. It was considered that a high flow rate was occurring when the poppet of the pilot valve was in close proximity to the valve seat, and that the flow rate was sufficient to strip away the protective oxide film. Theoretical simulation and actual measurements of pilot valve movement showed that, although this type of valve can appear to give a 'clean' lift, pilot oscillation can occur in the first 100 milliseconds of a lift. This paper presents the changes made to the test facility and a detailed design/manufacturing change - the introduction of a lapped profile to the pilot valve poppet and seat, in order to minimize pilot valve instability and ensure any instability could be tolerated. The results of the metallurgical examinations of the damaged pilot valve hard facings are presented, together with the results of theoretical modeling (using the RELAP5 analysis code) of system and valve interaction. (author)