Preliminary evaluation of the expansion system size for a pressurized gas loop: application to a fusion reactor based on a helium-cooled blanket

Abstract

Some considerations to preliminarily design the size of the Expansion Volume (EV) and the relief pipes for a Vacuum Vessel Pressure Suppression System, to be adopted in a fusion reactor based on a helium cooled blanket, are presented. The volume of the EV depends on the total energy of the cooling system and it can be sized based on a required final pressure at equilibrium, by a simple energy balance. Two different EV solutions have been analysed: a "dry" EV and a "wet" EV. In this last, a certain amount of water could be mixed (by spraying or discharging in a pool) with the discharged helium, to reduce its temperature and allowing a lower size of the EV with respect to the "dry" solution. The pressure peak in vacuum vessel (VV) depends mainly on break area and flow area of the relief pipes and a simple formula to be used to size these pipes is suggested. The computer code CONSEN has been used to perform sensitivity analyses and to verify the methodology.

Economic analysis in risk management

“…The model described in the previous section is calibrated and benchmarked here against the results obtained from the validated CONSEN code on the same system, described in (Caruso, 2016) and characterized by the parameter values in Table 1. In the referenced report, the large toroidal break previously described is assumed to happen in the EU DEMO tokamak, caused by the melting of 10 m² of FW and leading to the release of all the BB coolant inventory to the VV.…”

“…All results from the CONSEN analyses reported in the following are taken from the figures in (Caruso, 2016). The main difference between the CONSEN model of the VVPSS and the GETTHEM one is the absence in the latter of the piping system in the RLs; while these pipes can actually be modelled in GETTHEM, the results of the simulation performed including a 1D model of the pipes showed negligible differences, but consistently higher computational cost (due to the necessity to discretize the pipes with a moderate number of nodes) with respect to the simulation without the piping model.…”