Reliability-based inspection of prestressed concrete containment structures

A study was undertaken to develop a reliability-based approach to the planning of inspection programs for prestressed concrete containment structures. The main function of the prestressing system is to ensure the leak integrity of the containment by maintaining a compressive state of stress under the tensile forces which arise in a hypothesized loss of coolant accident. Prestressing force losses (due to creep and shrinkage, stress relaxation or tendon corrosion) can lead to tensile stresses under accident pressure, resulting in loss of containment leak integrity due to concrete cracking and tensile yielding of the non-prestressed reinforcement. Therefore, the evaluation of prestressing inspection programs was based on their effectiveness in maintaining an acceptable reliability level with respect to a limit state representing yeilding of non-prestressed reinforcement. An annual target reliability of 10-4 was used for this limit state. As specified in CSA-N287.7, the evaluation of prestressing systems for containment structures is based on the results of lift-off tests to determine the prestressing force. For unbonded systems the tests are carried out on a randomly selected sample from each tendon group in the structure. For bonded systems, the test is carried out on an unbonded test beam that matches the section geometry and material properties of the containment structure. It was found that flexural testing is useful in updating the probability of concrete cracking under accident pressure. For unbonded systems, the analysis indicated that the sample size recommended by the CSA Standard (4% of the tendon population) is adequate. The CSA recommendation for a five year inspection interval is conservative unless severe degradation of the prestressing system, characterized by a high prestressing loss rate (>3%) and a large coefficient of variation of the measured prestressing force (>15%), is observed