Seismic evaluation of a nuclear installation is an important activity. With the changing seismic design and safety requirements, it is also important to re-evaluate existing nuclear plants. In this context, a seismic re-evaluation exercise of a fast breeder test reactor (FBTR) at Kalpakkam is carried out. The safety objectives identified for re-evaluation are

• Safe shutdown of the plant
• maintaining in safe shutdown condition,
• long term decay heat removal, and
• containment of radioactivity.

The best way to ensure that the software used in a safety critical system meets the required reliability is through formal verification, a process of proving certain properties in the designed algorithm viz-a-viz its requirement specification. Unfortunately, exhaustive formal verification is not always feasible due to difficulties involved such as state space explosion and in practical application of formal methods. Also, a major assumption in formal verification is that the requirements specification captures all the desired properties correctly. Further, software testing with large number of test cases without analyzing the quality/effectiveness of test cases, cannot give confidence on the reliability estimate. The widely used black box models (also called reliability growth models) have assumptions and hence, not suitable for safety critical systems.

An approach that combines results of software verification and testing to quantify the software reliability in safety critical systems is developed. In this approach, a method for generating efficient test cases, ensuring adequacy of software testing using appropriate software metrics such as ‘Modified Condition Decision Coverage’, ‘Linear Code Sequence and Jump coverage’ and mutation testing are adopted.

For ensuring adequate safety in nuclear power plants (NPPs), engineered safety systems are incorporated with redundancy, which is considered as a fundamental technique for fault tolerance. However, in redundant systems, common cause failures (CCFs) are considered to be the major contributor to risk and therefore, quantifying CCF is essential to demonstrate the reliability of a system. The approach as described in NUREG/CR-5500 is extended to derive plant specific coefficients for CCF analysis. In this approach, the impact vectors are modified to reflect the likelihood of the occurrence of the event in the specific system of interest. This method is also known as mapping. The mapped impact vectors are finally used to derive alpha factors for CCFA.

Safety Grade Decay Heat Removal system (SGDHRS) of Prototype Fast Breeder Reactor consists of four independent but 2 sets of identical loops each having 8 MWt heat removal capacity at a sodium temperature of 820 K.  The probability of failure on demand in opening of the dampers and its impact on SGDHRS reliability is assesed by modelling the relay logic, actuators, valves and power supply for both electrical and pneumatic actuated dampers in SGDHRS.

Most of the nuclear power producing sites in the world houses multiple units/plants. Such sites are faced with hazards generated from external events: earthquake, tsunami, flood etc.  Further, risk from a multiple unit site and its impact on the public and environment was evident during the Fukushima nuclear disaster that took place in March 2011. An in-house integrated approach for risk assessment of a multi-unit NPP site is developed by considering all categories of internal / external hazards.