Statistical study of TCV disruptivity and H-mode accessibility

Optimising tokamak operation consists of finding a path, in a multidimensional parameter space, which leads to the desired plasma characteristics and avoids hazards regions. Typically the desirable regions are the domain where an L-mode to H-mode transition can occur, and then, in the H-mode, where ELMs and the required high density<y can be maintained. The regions to avoid are those with a high rate of disruptivity. On TCV, learning the safe and successful paths is achieved empirically. This will no longer be possible in a machine like ITER, since only a small percentage of disrupted discharges will be tolerable. An a priori knowledge of the hazardous regions in ITER is therefore mandatory. This paper presents the results of a statistical analysis of the occurrence of disruptions in TCV. (author) 4 figs