A Case Study: Balanced Globe Valves Failure, Root Cause, and Recovery

During the pre-operational blowdown testing of the AP1000® Automatic Depressurization System (ADS), it was observed that several ADS Stage 1-3 balanced globe valves did not fully open; however, the valves did sufficiently open to provide adequate ADS flow to meet the ADS System depressurization safety function. Since the Main Control Room did not receive a full-open signal indicating that the ADS valves fully completed the valve stroke, the valves were disassembled to determine the potential cause and scoring was observed on the valve disk and valve body guide surfaces. A technical design review team was formed and determined that the primary cause was unequal thermal expansion between the thin walled valve disk and the thicker valve body resulting in interference during valve stroking. Classical thermal and CFD analyses were used to confirm that, once a normally closed globe valve received its signal to open, there was not sufficient “soak time” for the thicker valve body to reach thermal equilibrium with the thinner valve disk; thus, as the valve disk stroked open as it moved along the guide surfaces, there was not sufficient clearance resulting in scoring between the disk and the guides preventing the disk from stroking fully open. The solution was to maximize the clearance between the disk and body guides to allow for thermal growth during the ADS blowdown event. However, changing the clearances between the body guides and disk may compensate for thermal growth but may also impact the valve’s “balancing” function; therefore. disk piston rings were added to restore the tight clearance needed for “balancing” yet sufficiently provide relief from the effects of thermal growth. A prototype valve was retrofitted, and small scale validation tests were performed using air to challenge the changes. The preliminary performance test results were utilized by system designers to better understand the valve function in the event of an ADS blowdown. Although no full-scale retrofit testing was practical using high temperature saturated steam, as no suitable test facility was available within the plant operational test schedule time frame, the quality of the recovery effort yielded a high degree of confidence that the system retest would be successful. The project came to a successful conclusion upon two successful operational ADS blowdown tests with the retrofits installed.