thermal-hydraulic codes for cores and thermal-hydraulic codes for systems

gsyou

I am quite confused for the thermal-hydraulic codes for cores and the thermal-hydraulic codes for systems. I know COBRA is a thermal-hydraulic code for cores and RELAP is a thermal-hydraulic code for systems. For the reactor transient analysis, Does it need to couple the neutron kinetics codes (like PARCS) with the thermal-hydraulic codes for cores(like COBRA) and then couple with the thermal-hydraulic codes for systems (like RELAP)? Or, only one kind of thermal-hydraulic codes is needed to be coupled?
Some code like SIMULATE and SIMULATE-3k should have thermal-hydraulic feedback themselves. Is that thermal-hydraulic feedback just like the thermal-hydraulic codes for cores? And, does it still need to couple the thermal-hydraulic codes for systems (RELAP) to do the reactor transient analysis? Does PARCS have such thermal-hydraulic feedback inside the code?
Thank you!

Astronuc

Traditionally, core simulation codes had simpler or idealized TH models and TH codes had simplified power inputs. Coupling has been done in the form of chaining, i.e., feeding the output of one code to the input of the other, e.g., SIMULATE output into COBRA (or one of its derivatives). Meanwhile, Studsvik has been improving the TH model in SIMULATE.

At the moment, there are efforts to develop 'fully coupled' neutronics and TH/CFD systems. However, that requires substantial computing resources, particularly as the resolution of the problem increases, i.e., becomes finer. For example, it's relatively easy to run a single channel with idealized flow resistance to grids, but quite another matter to incorporate a detailed subchannel mode that captures the rotational flow of a mid-grid with mixing vanes.

COBRA/VIPRE are subchannel TH codes. RELAP/RETRAN are larger system (primary) TH/codes.

Beyond coupling neutronics/TH is the goal of coupling neutronics/TH/FTM systems, where FTM is fuel thermo-mechanical codes. Coupling the three systems is the goal of the CASL program, and to some extent NEAMS.

gsyou

Astronuc,
Thank you for your reply.
So in what situation, the neutron kinetic codes need to be coupled with sub-channel TH codes, and in what situation coupled with system TH codes? Is it necessary to couple the neutron codes with all of these two kinds of TH codes together for accident analysis?
Thank you!

Astronuc

For steady-state operation, it seems the simplistic models in current neutronics codes (core simulators) are fine. For AOOs and transients, one might want a more mechanistic (CFD) code. The system codes like RELAP or RETRAN handle the primary system, as opposed to the details of the core.

One has to realize that the legacy codes have evolved from earlier models that were necessarily simplistic because of the limits on computational systems back in the 60s and 70s. Now we have computing power that is orders of magnitudes greater, and so now we are developing multiphysics simulation system to take advantage of the high performance computing (HPC) systems.

With respect to coupling codes, the need depends on the goal. In safety analysis, one only needs to show that one will not exceed specified limits. Over the decades, we've increased the energy generated per unit mass of fuel (burnup), and that pushes us closer to technical and safety limits of the fuel, but that doesn't necessarily mean increased risk, as long as we have good simulation capability that can quantify the margins to those limits. It turns out that the older methods had lots of margin because they had so much uncertainty. With better simulation technology, we can get more out of the fuel, nuclear reactors and plants without increasing risk.