Severe accident progression in a PHWR is strongly influenced by the large variability that exists amongst various fuel channels and fuel entities within channels. A best effort prediction necessitates consideration of all risk sensitive phenomena and processes in a detail that reduces uncertainties in rates as well as magnitudes of source terms. A review of available computational codes points to a need of a much greater than hitherto undertaken, detailed modelling of the horizontal PHWR reactor channels, fuel bundles, end fittings and feeders with an advanced, more reactor geometry specific consideration of solid debris behaviour in the Calandria vessel. A new computer code ROSHNI improves upon existing computational aides firstly in detail in which the reactor is modelled and further in its fuller consideration of important PHWR related severe accident progression pathways and phenomena. The paper details the code modelling approach used in capturing the important system response parameters. It is expected that the source terms of Deuterium gas and fission products can now be evaluated with lesser uncertainty and with greater degree of control available to the analysts. Users are able to perform parametric and uncertainty analyses with greater ease. Initial modelling of progression of a severe core damage accident is being targeted for a CANDU 6 reactor. However, the code structure is generalized to extend the ability to other PHWR designs including multi-unit CANDU stations and Indian PHWRs. After Fukushima, serious concerns have been raised about severe accident mitigation capabilities of all operating nuclear power plants including CANDU reactors whose design concept presents specific vulnerabilities not common with other reactor types. While operating organizations and their supportive regulating bodies are cautious, to the point of lethargy, about design upgrades to operating reactors, only a best effort modelling of realistic accident progression pathways can empower them with the information required to justify the required risk reduction upgrades to reactors that were never designed with severe accidents within their design basis. Reduction of uncertainties by better modelling can also assist in the development of more effective severe accident management options and better operator training as the lessons learnt from Fukushima must be taken seriously. Otherwise decisions on hardware upgrades will remain handicapped by larger than necessary uncertainties inherent in the available but obsolete PHWR reactor severe accident modelling techniques. This paper concentrates on the rationale for development of a new code; its overall modelling approach and details of reactor core modelling. In summary, the code models transient behaviour of each and all fuel channels; models all individual fuel bundles with each represented by 16 fuel and Zircaloy rings; computes the response of the associated end fittings and carbon steel feeders; considers fuel heatup during staggered boiloff in channels followed by differentiated dry heatup of each fuel channel in steam and deuterium with consideration of moderator depletion. In all, fuel temperatures are evaluated at about 80,000 locations with separate consideration of thermal behaviour of 380 pairs of end fittings and the same number of feeders whose thermo-chemical response to severe accident conditions has a potential of generating additional combustible gas, not previously evaluated. It considers disassembly and thermo-chemical behaviour of individual fuel bundles to multiple fields of debris first suspended over underlying intact channels and gradually relocated to the Calandria vessel with consideration of impact of about 100 in-core devices. It also considers effect of failure of the Calandria vessel under thermomechanical loads and evaluates energy and fission product source terms for containment response. Deuterium gas source
(View PDF for the rest of the abstract.)
