There are some features of evolutionary plants that require new models and extension of the codes that have been developed for present day plants.

Some integral codes have special models that have been developed for particular plants. They therefore cannot be applied or it is difficult to apply them to new plants. It is also difficult to apply them to new experiments for the purposes of code validation. The MAAP code developed by EPRI is an integrated severe accident code, which has specific models for specific plants and phenomena. A special version AP600-MAAP has been developed for evaluation of AP600 safety (IAEA-TECDOC-752, 1994).

In general, in advanced LWR designs, there is a requirement for much stronger thermal-hydraulic coupling between the primary circuit and containment. This has led to the coupling of some system thermal-hydraulic codes, e. g. RELAP5 with the containment code CONTAIN.

Many evolutionary passive designs have large pools as heat sinks and condensers. To be effective, these need to be well mixed and the effectiveness of these pools needs to be established. The system thermal-hydraulic codes do not have the required mechanistic mixing models and therefore need to be benchmarked against CFD codes.

The system codes have limitations in their modelling of condensation, particularly in the presence of non-condensables or 3D effects.

Finally, it has been established that the performance of the system codes in buoyancy-driven situations is less robust, than in their application to the modelling of high-pressure forced convection flows, the regimes for which they were originally developed. Much effort has been expended in improving the performance of these codes in low-pressure applications in current generation reactors, e. g. in the modelling of shutdown accidents. Generally, later versions of the system codes, e. g. RELAP5 are much more robust (compared with earlier versions in this respect).