Pragmatic application of the precautionary principle to deal with unknown safety challenges

Nuclear power technology has matured over a number of decades to the point where our understanding of the technology under a wide variety of circumstances is quite high. Despite this high degree of maturity, discoveries of new challenges occasionally surface. These may arise from either unusual or unexpected operational conditions or new experimental findings from ongoing research. With the early realization that such discoveries could occur, a conscious effort was made to take precautions against their negative impacts. Principles such as defence-in-depth, designing for high reliability, incorporation of robust safety margins and use of justified conservatisms are key examples of established practices that are embedded in national regulatory regimes of most, if not all countries with nuclear programs. Because of these provisions the safety cases of the current generation of reactors proved to be quite resilient to discoveries of earlier unrecognized challenges. A fundamentally important element in the management of "unknown unknowns" is a healthy research programme. Such a programme is especially necessary as a precondition for understanding potential impacts from changes in operating conditions or implementation of novel design features. A research programme helps minimizing chances of stumbling on "unknown unknowns", and allows resolution of emerging issues to by virtue of the accumulated understanding and capability to predict challenges to safety. In the few instances when discoveries occurred with recognized negative effects on safety, these spurred changes in operating conditions, maintenance or testing practices, design modifications, as well as required targeted research projects. This paper outlines several CANDU-specific "discoveries" in the field of thermalhydraulics, illustrating past "unknown unknowns" and the actions taken to address those. The main message, however, is to point out that both the industry and the regulator should maintain adequate provisions to deal with "unknown unknowns" and that a constant vigilance is necessary to avoid complacency. (author)