Probabilistic aspects of surveillance test interval extension

By increased number of operating nuclear reactor years and experience it can be found out that reactor shutdowns may be required in situations which are not safety significant or even the shutdown is not the safest action at the moment. A method is proposed for establishing probabilistic aspects of surveillance test interval determination. The technical basis for test interval extension justification from one to three months for scheduled online tests for the Engineered Safety Features Actuation System (ESFAS) which served as a case study are presented. For ESFAS, possible increased test interval first by the use of generalized system model has been determined. The results show that if test interval is extended from one month to three months where unavailability is the lowest, unavailability decreased for more than 20%. So the test interval extension seems to be reasonable. (author). 10 refs., 3 figs., 1 tab