On the capabilities of the "CFD-like" code GOTHIC for addressing nuclear containment thermal-hydraulics issues

by Dr Michele Andreani

Thursday 26 Mar 2020, 14:00 → 15:00 Europe/Zurich

OHSA/B17 (Paul Scherrer Institut)

The first part of the presentation is dedicated to the illustration of the capabilities of the GOTHIC code, which is developed by NAI in USA. The GOTHIC containment code is mostly used by American utilities for their daily operation uses, and by other organisations for addressing operational and licensing issues. Typically, the use is based on a multi-compartment representation of the containment, including all equipment and structures, but with very coarse subdivision of the containment. For this reason, a large part of the technical community working in the area of containment analysis tends to consider GOTHIC a “lumped-parameter” (LP) code.

However, in GOTHIC the hydraulics is represented with a network of connected volumes, where each volume can be represented by a 3-D mesh consisting of a very large number of cells. The mathematical structure is based on a multi-field, multi-component, two-fluid formulation (liquid, vapour consisting of steam and several gases, and droplets), with a turbulence model used for properly simulating momentum transport. Therefore, GOTHIC has many of the capabilities of CFD codes, and it is thus appropriate to refer to it as to a “CFD-like” code.

GOTHIC can thus be used to “bridge” the coarse information produced by LP or integral codes to the detailed simulations using CFD codes. In fact, the code can produce detailed 3-D information at a fraction of the computational cost of CFD codes. This makes GOTHIC a very useful tool, for instance, for analysing long transients, where the 3-D distribution of gases is required and sensitivity studies are required for addressing safety issues. In the second part of the presentation, an example of this kind of application to a Swiss NPP is briefly discussed to show this capability.

Several aspects of the modeling strategy have been based on very extensive validation over two decades, which has also included the participation in international benchmarks. In these exercises, the accuracy obtained with the GOTHIC code proved to be comparable to that obtained with CFD codes. In the last part of the presentation, the outcome of the most recent validation exercise will be discussed in some detail, comparing the results obtained with GOTHIC with various CFD predictions.