16.2 - Severe Accident Analysis (RRC-B) - EDF Hinkley Point

SUB-CHAPTER : 16.2PRE-CONSTRUCTION SAFETY REPORTCHAPTER 16: RISK REDUCTION AND SEVEREACCIDENT ANALYSESPAGE : 141 / 295Document ID.No.UKEPR-0002-162 Issue 04Selection of sacrificial concreteThe composition of the sacrificial concrete in the reactor pit is specifically selected to meet therequirements of temporary melt retention. The concrete aggregate consists mainly of Fe 2 O 3 andSiO 2 in approximately equal proportions, with 15% by weight common Portland cement in thedry concrete mixture as a binder.Fe 2 O 3 helps to oxidise the chemically aggressive metallic components Zr and U in the substoichiometricmelt. The reaction by-product, Fe, does not alter the thermo-chemicalcharacteristics of the melt. In addition, after the dissolved metal inventory in the oxidedecreases, surplus Fe 2 O 3 accumulates as FeO and Fe 3 O 4 in the oxide melt, which reduces theliquidus temperature and, correspondingly, the temperature level at which the MCCI takes place.This effect is beneficial as, in combination with the formation of silicates, it contributes toreducing the fission product release from the MCCI pool.Fe 2 O 3 and SiO 2 are provided in their natural form as iron ore (hematite) and siliceous pebbles.This also introduces, MgCO 3 (dolomite) and CaCO 3 (limestone) into the concrete system, butonly in small quantities. Ordinary Portland cement is used to bind the concrete aggregates.Hardening tests with the described composition have shown that a steady state water content ofless than 5% can be achieved. Sub-section 16.2.2.4 - Table 1 lists the composition, density anddecomposition properties of that concrete called FeSi/PZ15/8 for a water-content of 5%.2.4.1.2.1. Validation strategyA precondition for the corresponding analysis is the characterisation of the melt release from theRPV. This information includes the masses of oxidic and metallic components, theircorresponding thermo-dynamic and thermo-chemical states, and their order of discharge intothe reactor pit. These parameters can be obtained from integral codes such as MAAP-4(Appendix 16A).However, because late in-vessel melt progression is an ongoing R&D issue, the modelsincorporated in these codes and the resulting predictions involve a high level of uncertainty.Moreover, the feedback of thermal radiation emitted from the MCCI pool in the pit to the RPVlower head, which, under EPR conditions, contributes substantially to lower head failure, is notadequately reflected.Due to these deficiencies, the validation of the accumulation function of the pit for the EPR isperformed based on generalised melt release sequences. Though still being supported byMAAP-4 predictions for the state and amount of the released melt, they involve a wider variationof key assumptions and thus envelop the release sequences and conditions obtained withMAAP-4. This strategy drastically reduces the spectrum of melt release scenarios and releasemodes to be considered.The principal tool to perform the assessment is the MCCI code COSACO (Appendix 16A), whichhas been extended by models that simulate the lower RPV and its coupling with the MCCI.2.4.1.2.2. Modelling approachGeneralised release casesThe applied generalised melt release sequences are based on the modelling scheme outlined inSub-section 16.2.2.4 - Figure 2.