Tritium Issues in Next Step Devices

Abstract

Tritium issues will play a central role in the performance and operation of next-step deuterium-tritium (DT) burning plasma tokamaks and the safety aspects associated with tritium will attract intense public scrutiny. The orders-of-magnitude increase in duty cycle and stored energy will be a much larger change than the increase in plasma performance necessary to achieve high fusion gain and ignition. Erosion of plasma-facing components will scale up with the pulse length from being barely measurable on existing machines to centimeter scale. Magnetic Fusion Energy (MFE) devices with carbon plasma-facing components will accumulate tritium by co-deposition with the eroded carbon and this will strongly constrain plasma operations. We report on a novel laser-based method to remove co-deposited tritium from carbon plasma-facing components in tokamaks. A major fraction of the tritium trapped in a co-deposited layer during the deuterium-tritium (DT) campaign on the Tokamak Fusion Test Reactor (TFTR) was released by heating with a scanning laser beam. This technique offers the potential for tritium removal in a next-step DT device without the use of oxidation and the associated deconditioning of the plasma-facing surfaces and expense of processing large quantities of tritium oxide. The operational lifetime of alternative materials such as tungsten has significant uncertainties due to melt layer loss during disruptions. Production of dust and flakes will need careful monitoring and minimization, and control and accountancy of the tritium inventory will be critical issues. Many of the tritium issues in Inertial Fusion Energy (IFE) are similar to MFE, but some, for example those associated with the target factory, are unique to IFE. The plasma-edge region in a tokamak has greater complexity than the core due to lack of poloidal symmetry and nonlinear feedback between the plasma and wall. Sparse diagnostic coverage and low dedicated experimental run time has hampered the development of predictive models. Diagnostic advances are urgently needed to better characterize the plasma edge and wall and improve our predictive capability.