The limits and challenges of error field correction for ITER
Significant progress has been made in interpreting the effects of non-axisymmetric 'error' fields on a plasma through ideal MHD stability and a dominant 'least stable' ideal mode through which the fields couple to the tearing resonant surface. However, in contrast to expectations from such theories, experiments have found limited success in correcting error fields, with single correction coil arrays giving benefits of between 0% and {approx}50% correction (in terms of improvement to a low density locked mode limit), dependent on the structure of the error and correcting fields. With additional coils up to {approx}70% is possible. It was unclear whether this represented an intrinsic stability or control limit, or higher order toroidal or poloidal harmonic effects. Thus, studies on the DIII-D tokamak explored correction of a proxy error field, using two differently structured coil arrays. This enabled the principles of error correction to be tested at high amplitudes and operational densities, with known pure n=1 fields. Results showed substantial residual effects from the corrected n=1 field, with improvements of only {approx}50% in the low density locked mode limit. This suggests that n=1 error fields must couple to more than one surface in the plasma, and this is conjectured to be through more than one ideal mode, thereby requiring precise correction. For ITER, updated predictions of field error have been obtained and compared with revised scalings for tearing mode thresholds, indicating 50% or better error field correction will be needed. This will likely require more than one well coupled correction coil array and sets a challenge for theory to model the behavior, in order to clarify the plasma response and braking mechanisms, and so the effectiveness of ITER's correction coils and the possible need for support from its edge localized mode control coils.