[en] A qualitative assessment of the fission product release capability of the ELOCA.Mk5 computer code was performed by simulating two transients from the sweep-gas experiment, FIO-133. Improved agreement between calculated and experimental trends in release was obtained by applying an interface pressure stress component to the pellet center. As well, results show that the current system for defining the reference temperature distribution for the thermal stress component is not always realistic. These results are being used in the development of a new, mechanistic pellet stress model. (author)

[en] Sweep gas experiments performed at CRL from 1979 to 1985 have been analysed to determine the fraction of the fission product gas inventory that is released on reactor shutdown and startup. Empirical equations were derived and applied to calculate the xenon release from companion fuel elements and from a well documented experimental fuel bundle irradiated in the NRU reactor. The measured gas release could be matched to within about a factor of two for an experimental irradiation with a burnup of 217 MWh/kgU. (author)

[en] The grain-boundary chemistry of used CANDU fuel exposed to dry air at 150 degrees C for a prolonged period has been investigated by X-ray photoelectron spectroscopy (XPS) and scanning electron microscopy (SEM). High degrees of surface oxidation have been determined using the chemical-shift effects for the uranium photoelectron emission, but these must be largely restricted to thin films. The observed distribution of segregated fission products implies an absence of major fuel restructuring and SEM examinations revealed mainly subtle changes in the UO₂ grain structure. These findings are consistent with metallographic evidence of pervasive grain-boundary attack, despite only slight bulk alteration of the fluorite-lattice structure. (author)

[en] An integrated approach to plant life management has been developed for CANDU reactors. Strategies, methods, and procedures have been developed for assessment of critical systems structures and components and for implementing a reliability centred maintenance program. A Technology Watch program is being implemented to eliminate 'surprises'. Specific work has been identified for 1998. AECL is working on the integrated program with CANDU owners and seeks participation from other CANDU owners

[en] The mechanical integrity of PHWR components based on fracture mechanics requires the determination of J-R curves obtained from laboratory specimens. These J-R curves are geometry dependent and the transferability of specimen J-R curves to component level is questionable. The influence of crack tip constraint or stress triaxiality has been emphasised recently in explaining the geometry dependent resistance of specimens and structures to ductile tearing. In this work 'Q' and 'h'are used as crack tip constraint indexing parameters. Two dimensional plane strain elastic plastic analyses have been carried out on Centred Cracked Panel (CCP), Three Point Bend Bar (TPBB) and Compact Tension (CT) specimen to study the characteristics of 'Q'and 'h' parameters. For each geometry, different a/w ratios (a/w = 0.1-0.7) are considered. For each a/w ratio, the effect of different strain hardening exponents (n = 3,5,10 and 20) is also investigated. The analysis has also been carried out for Indian PHWR PHT piping material also. Reference stresses have been obtained using standard plane strain small scale yielding solution. The behaviour of Q-parameter and triaxiality factor (h) over distances 1≤r/(J/σ_o)≤5 for several deformation levels is investigated. The variation of Q and h at different angular positions from the crack tip is also studied. It is observed that h is dependent on both distance and angular position. Hence the third parameter is required where h value has to be evaluated to consider it for local fracture criterion. Whereas, Q is independent of distance for low constraint geometries and is dependent on distance for high constraint geometries. It is identified that Q can be used as constraint indexing parameter along with the J-integral to characterise the crack tip stress state. A unique linear relationship can be obtained between h and Q that is irrespective of geometry for a given material for low constraint geometries. For high constraint geometries this relationship is not valid at high deformation levels. (author)

No abstract available

[en] All safety analysis codes require some representation of actual plant data as a part of their input. Such representations, referred to at Point Lepreau Generating Station (PLGS) as plant idealizations, may include piping layout, orifice, pump or valve opening characteristics, boundary conditions of various sorts, reactor physics parameters, etc. As computing power increases, the numerical capabilities of thermalhydraulic analysis tools become more sophisticated, requiring more detailed assessments, and consequently more complex and complicated idealizations of the system models. Thus, a need has emerged to create a precise plant model layout in electronic form which ensures a realistic representation of the plant systems, and form which analytical approximations of any chosen degree of accuracy may be created. The benefits of this process are twofold. Firstly, the job of developing a plant idealization is made simpler, and therefore is cheaper for the utility. More important however, are the improvements in documentation and reproducibility that this process imparts to the resultant idealization. Just as the software that performs the numerical operations on the input data must be subject to verification/validation, equally robust measures must be taken to ensure that these software operations are being applied to valid idealizations, that are formally documented. Since the CATHENA Code is one of the most important thermalhydraulic code used for safety analysis at PLGS the main effort was directed towards the systems plant models for this code. This paper reports the results of the work carried on at PLGS and ANSL to link the existing piping data base to the actual CATHENA plant idealization. An introduction to the concept is given first, followed by a description of the databases, and the supervisory tool which manages the data, and associated software. An intermediate code, which applied some thermalhydraulic rules to the data, and translated the resultant data to CATHENA structure is then described. Finally, the results of some validation work are shown. (author)

[en] The CATHENA thermalhydraulic computer code was used to simulate various scenarios following a CANDU 9 steam generator tube rupture (SGTR) event. The analysis included cases with class IV power and emergency core cooling system (ECCS) available and other cases with subsequent loss of class IV power (LCIVP) or impairment of ECCS injection. Two main approaches were followed in the analysis of each case. In the first approach, D₂O feed was credited to provide conservative data for input to radionuclide release and dose calculations. Also operator actions are credited. The other approach is designed to give conservative predictions with respect to the acceptance criteria of fuel and fuel channel integrity and to prove that in case of such event, the operator will have enough time to mitigate the consequences. This is done by not crediting makeup for the inventory loss and relying on the automatic operation of safety systems. The analysis of the cases of the first approach provided the required data for radionuclide release and dose calculations and gave a good insight into the required sequence of operator timely actions to mitigate the consequences of such event. On the other hand, the cases of the second approach confirmed compliance with regulatory requirements for pressure tube and fuel integrity. The runs with ECCS available, showed the ECCS injection is effective in filling and cooling the core and that regulatory requirement's for fuel and channel integrity are met. In the event of ECCS impairment, the earliest indication of late fuel heat-up is late enough to provide the operator with an adequate time to act in mitigating the consequences of this event. (author)

[en] If a break should occur in the inlet feeder or inlet header of a CANDU reactor, the rapid depressurization will cause the channel flow(s) to reverse. Depending on the gap between the upstream bundle and shield plug, the string of bundles will accelerate in the reverse direction and impact with the upstream shield plug. The reverse flow impact velocities have been calculated for various operating states for the Bruce NGS A reactors. The sensitivity to several analysis assumptions has been determined. (author)

[en] In 1994, NB Power installed engineered pressure relief panels in the Turbine Hall of the Point Lepreau station. An individual panel must open within a given set period of time to be considered available. In order to judge the effectiveness of the new panels and to define the operating criteria based on in-situ tests, a detailed behavioral mathematical model for the turbine hall pressure relief panel is developed. The mathematical model is converted to various program designs and algorithms. Based on the test performed using these algorithms a program design is selected for modelling these panels. (author)