(a) Increasing needs for accuracy and detail of codes result in corresponding needs regarding the details to be provided by experiments. For the development and validation of CFD codes one needs primarily single effect experiments with very detailed instrumentation. Instrumentation should be developed to provide improved local and field data for CFD codes. For design purposes and for system code improvement and validation, relatively realistic large-scale experiments are of interest. Thus some large scale more complex experiments with detailed instrumentation are likely to be needed to validate the correct interrelation between the different models used in the CFD codes;

(b) In some cases, uncertainties exist because of a lack of data for friction loss modelling and condensation (more general: phase interaction) particularly at low pressure. Perhaps dynamic flow charts can be useful here;

(c) Further experiments are likely to be needed for code validation related to the effects of non-condensable gases;

(d) As natural circulation systems used in different advanced designs are widely different in operating conditions and geometry, work on the scaling of natural circulation phenomena demands attention. Effective scaling laws will enable inter-comparison of data generated by various facilities.