Optimisation of out-vessel magnetic diagnostics for plasma boundary reconstruction in tokamaks

Abstract

To improve the low frequency spectrum of magnetic field measurements of future tokamak reactors such as ITER, several steady state magnetic sensor technologies have been considered. For all the studied technologies it is always advantageous to place the sensors outside the vacuum vessel and as far away from the reactor core to minimize radiation damage and temperature effects, but not so far as to compromise the accuracy of the equilibrium reconstruction. We have studied to what extent increasing the distance between out-vessel sensors and plasma can be compensated for sensor accuracy and/or density before the limit imposed by the degeneracy of the problem is reached. The study is particularized for the Swiss TCV tokamak, due to the quality of its magnetic data and its ability to operate with a wide range of plasma shapes and divertor configurations. We have scanned the plasma boundary reconstruction error as function of out-vessel sensor density, accuracy and distance to the plasma. The study is performed for both the transient and steady state phases of the tokamak discharge. We find that, in general, there is a broad region in the parameter space where sensor accuracy, density and proximity to the plasma can be traded for one another to obtain a desired level of accuracy in the reconstructed boundary, up to some limit. Extrapolation of the results to a tokamak reactor suggests that a hybrid configuration with sensors inside and outside the vacuum vessel could be used to obtain a good boundary reconstruction during both the transient and the flat-top of the discharges, if out-vessel magnetic sensors of sufficient density and accuracy can be placed sufficiently far outside the vessel to minimize radiation damage.