Research report summaries 2013–2014

EVS 2010 is a statistical tolerance limit solution to the NOP trip setpoint problem. This report presents my evaluation of the theoretical correctness and practical fitness for purpose of EVS 2010 for that problem. On the basis of a technical review of the EVS 2010 theory and three separate benchmarking exercises to assess its fitness for purpose, this report reaches a number of conclusions supported by nine principal findings.

My technical review found that the EVS 2010 theory is mathematically and statistically correct, and the benchmarking found that in the great majority of tests EVS 2010 provided adequate protection in terms of the tolerance limit coverage.

In the context of the NOP trip setpoint problem, ‘adequate protection’ means that the method provides at least 95% assurance of a trip operating early enough during a slow loss of regulation event to prevent the risk of dry-out in 95% of those events. These positive findings suggest that EVS 2010 is basically sound, at least in theory, and has the potential to provide a practical solution to the NOP trip setpoint problem.

However, a number of other findings make it clear that at present there are several outstanding concerns regarding its use for determining NOP trip setpoints in practice.

In particular, benchmarking indicates that there are three specific situations where EVS may not provide adequate protection against the risk of dry-out. Furthermore, the inherent limitations of the generic benchmarking employed in this work mean that the encouraging performance in benchmarking tests may not hold for a particular real application.

In any proposed practical application it will be essential to provide assurance that EVS 2010 will deliver good tolerance limit behaviour in that application.

My "verification and benchmarking" task formally yields only these conclusions, and in particular is not prescriptive about how the assurance demanded in this third conclusion should be provided. However, the section in my report entitled "Summary, Conclusions and Recommendations" ends with some recommendations for future progress in the NOP trip setpoint problem, including three ways to achieve assurance of good tolerance limit behaviour in proposed applications of EVS 2010.

Read the RSP-0293-Final Report (PDF)

The main goal of the project was to develop a comprehensive database and models on the steam generator tube degradation process and resulting leak rates that permit the CNSC to independently evaluate the integrity of steam generator tubes as plants age and degradation proceeds, new forms of degradation appear, and thus new defect-specific management schemes should be implemented. The scope of the present research includes consolidation of data on CANDU and PWR steam generator tube degradation mechanisms (Primary Water Stress Corrosion Cracking (PWSCC), pitting and fretting) and the development of flaw and their growth based on fracture mechanics models. An experiment and modeling program was initiated. Experimental data on leakage rates for CANDU SG relevant tube cracks was collected and a Homogeneous Equilibrium Model (HEM) and Homogeneous Non-Equilibrium Model (HNEM) were derived from the conservation equations, providing a physical basis for prediction. Several tasks were carried out under this project:

• An experimental program was created which includes: the design and building of a experimental test facility; development of a test matrix; the design and manufacture of simulated crack test specimens; conduction of experiments for various operating conditions.
• A consolidated database on choked flow for small length to diameter cracks was compiled by running tests on the Facility for Leak Rate Testing (FLRT) at the Multiphase Flows and Fuel Cell Research Labs (MFRL) at Purdue University by simulating the pressure difference of a steam generator tube across crack samples conducting a constant area leak rate tests.
• Development of a new choking flow model and validation against experimental data.
• Application of the RELAP5 code model for predicting choking flow and assessing its predictive capabilities.
• Recommendation of an optimal choked flow modeling approach for application for SG tube integrity assessments.

Read the RSP-0294-Final Report (PDF)

RSP-0295 – Investigation of the fatigue cracking and leakage rate potential of U-bend tube bundles subjected to flow-induced vibrations

The steam generator (SG) tubes in a CANDU reactor comprise most of its primary coolant pressure boundaries. Maintaining the integrity of SG tubes is a major safety issue, since they ensure the separation of the two fluids.

These devices house several tubes, around which pass a large amount of energy in the form of high-speed fluid flow. Serious problems can arise if even a small portion of this energy is converted into mechanical energy, since this can cause violent tube vibrations. In turn, this can cause SG tube failure because of fatigue and/or wear at the support locations. Measures are therefore taken to reduce the potential of fretting wear, by stiffening the structure and reducing clearances at the support. Fretting wear from normal operation should be accounted for in the design stage. However, in some situations, supports located on the straight part of the tube may deteriorate to the point where extremely large clearances or even total wastage of the supports may result. One example of this is the problem experienced in Bruce Unit 8, where severe degradation of the tube support plate (TSP) occurred. This degradation was revealed by eddy current testing and was later confirmed by visual inspection. The finding was described as metal loss caused by flow-accelerated corrosion of the carbon steel trefoil support plate, and varied from minor to complete loss of the TSP ligaments. This loss of TSP ligaments resulted in lack of support for the adjacent tubes, making them more susceptible to fretting-wear damage and fatigue cracking at these locations. In addition, this had the potential to affect the rate of wear in the U-bend portion of the tube due to the evolution of unstable modes. The integrity could have been seriously breached as result of a potential support loss. Therefore, remedies were proposed and installed. Such remedies included adding flat bars at U-bend locations, as well as inserting so-called comb support at the place of the corroded broached-hole support. Previous investigation by the University of Guelph’s Fluid-Structure Interaction Laboratory showed that these remedies were effective in reducing the vibration amplitude and the resulting fretting wear damage when all U-bend support clearances were kept under 0.2 mm. 

This report presents the finding of a work aimed at investigating the crack fatigue potential caused by such an accident. Numerical simulations were employed for the full U-bend tube subjected to a variable flow field typical of a CANDU steam generator configuration. Both deterministic and probabilistic evaluations have been utilized. It has been shown that the suggested remedies re effective in reducing the damage potential if the radial clearances were kept within 0.2 mm. In addition, the scallop bar supports at the U-bend apex were proven to be critical. Crack propagation for an assumed surface circumferential and through-wall cracks was simulated. The leakage rate in the through-wall crack was also calculated. Charts providing the probability of life and leakage rates exceeding certain thresholds were also presented.

Read the RSP-0295-Final Report (PDF)

RSP-0296 – Statistical analysis of common-cause failure data to support safety and reliability analysis of nuclear plant systems

This report describes the findings of a project entitled “Statistical Analysis of Common-Cause Failure Data to Support Safety and Reliability Analysis of Nuclear Plant Systems for the CNSC” under the contract No. 87055-12-0221.

Analysis of common-cause failures (CCFs) is an important element of the probabilistic safety assessment (PSA) of systems important to safety in a nuclear power plant. Based on the conceptualization of the CCF event, many probabilistic models have been developed in the literature. This report provides a comprehensive review of CCF modeling techniques, and shows that a modern method called the General Multiple Failure Rate Model (GMFR) is the most suitable for probabilistic modeling of CCF events. Therefore, the GFMR is described in detail in the report and adopted for the case studies. To estimate the parameters of the GMFR model, the Empirical Bayes (EB) method is adopted. The report describes the data mapping methods and the EB method for combining data from different component groups and plants in the statistical estimation.

This project presents detailed case studies to illustrate the data mapping and EB method. The case studies are based on CCF data for motor-operated valves. These case studies serve as templates to analyze CCF data from other safety systems. The report provides analysis methods to the CNSC staff to analyze CCF rates and evaluate the adequacy of input data used in the PSA of Canadian plants.

This project demonstrates the development of the capacity to analyze CCF data in line with international best practices.

Read the RSP-0296-Final Report (PDF)

RSP-0297 – Evaluation of default annual limit on intake (ALI) for yellowcake and uranium ore

In October 1994, the Atomic Energy Control Board (AECB) recommended default values for the annual limit on intake (ALI) for uranium ore and two types of yellowcake to the Saskatchewan Labour Working Group on Long-Lived Radioactive Dust. Since then, new solubility data has become available and the International Commission on Radiological Protection (ICRP) has recently evaluated its biokinetic models. Thus, it is appropriate to review the ALIs using the updated solubility data and the current ICRP biokinetic models to determine if changes to the default ALI values are required.

The primary objective of this study was to develop appropriate default ALIs for workers at uranium mines and mills in Canada using the most recent available data on solubility and dosimetric methods for uranium ore, calcined yellowcake and non-calcined yellowcake. It is acknowledged that future uranium mines and mills in Canada could produce ore and yellowcake with somewhat different characteristics than at present. Therefore, this study also provides high-level guidance on how to develop appropriate ALIs for new ores and for yellowcake of different origins and/or processes, or for future North American technology.

The first step in the study was to complete a literature review. Upon completion of the literature review, efforts were made to collect the most recent solubility data available for Canadian uranium mining and milling operations and to select solubility data that was considered representative of current uranium mines and mills in Canada. Using this solubility data, absorption parameters were derived and provided as inputs to the dosimetric model. The dosimetric model used for this study is the same as that used by the ICRP to derive dose coefficients for intakes of radionuclides. Finally, the results of the modelling were used in the process for calculating the revised ALI values. It should be noted the proposed ALIs in this report are based on radiological dose; however, in some instances, chemical toxicity the limiting factor for ALIs is.

Read the RSP-0297-Final Report (PDF)