Improvement of JASMINE code for ex-vessel molten core coolability in BWR

• Published in: Annals of nuclear energy 2022-12, Vol.178, p.109348, Article 109348
• Main Authors: Matsumoto, Toshinori, Kawabe, Ryuhei, Iwasawa, Yuzuru, Sugiyama, Tomoyuki, Maruyama, Yu
• Format: Article
• Language: English
• Subjects: Agglomeration; JASMINE; Melt spreading; Molten core coolability; Severe accident; Wet cavity
• ISSN: 0306-4549; 1873-2100
• DOI: 10.1016/j.anucene.2022.109348

•JASMINE code was improved for evaluation of the molten core coolability under-water in BWR.•Overlapping and enthalpy of the particles were considered in agglomeration model.•Shooting flow and termination condition for melt spreading model were implemented.•DEFOR-A and PULiMS experiments were analyzed with improved JASMINE code. Evaluation of the molten core coolability in the wet cavity needs techniques to treat under-water melt behaviors including particle formation by break-up, horizontal spreading at the cavity bottom, and so on. The Japan Atomic Energy Agency extended the applicability of their fuel-coolant interaction analysis code JASMINE to simulate the relevant phenomena of molten core in a severe accident. It is assumed that three types of fuel debris with different coolable geometries would be produced in the wet cavity as particulate debris, agglomerated debris and cake debris. An agglomeration model that considers the fusion of hot particles on the cavity floor was implemented in the JASMINE code. Another significant improvement is the introduction of the melt spreading model based on the shallow water equation that takes crust development at the melt surface into account. These models have adjusting parameters, such as frequency of agglomeration occurrence between contacting particles as a function of melt enthalpy, mechanical strength of the surface crust governing the termination of melt spreading. For optimization of these parameters, we referred to data from the agglomeration experiment DEFOR-A and the under-water spreading experiment PULiMS conducted by Royal Institute of Technology (KTH) in Sweden. The JASMINE analyses reproduced most of the experimental results well with a common parameter set, suggesting that the primary phenomena are appropriately modelled.