Improved ERO modelling of beryllium erosion at ITER upper first wall panel using JET-ILW and PISCES-B experience

•The earlier ERO erosion modelling [1] for the beryllium ITER FW element was significantly revisited.•The revisit utilizes the recent experience of ERO application to the Be erosion experiments at JET-ILW tokamak and PISCES-B linear device.•The net erosion gets stronger by a factor of ∼2 mostly due to the new semi-analytic approach [10] for the sputtering ion trajectories in the surface sheath.•The paper shows the status of the underlying sputtering data including the most recent MD-OKMC simulations [11] and the role of CAPS. ERO is a 3D Monte-Carlo impurity transport and plasma-surface interaction code. In 2011 it was applied for the ITER first wall (FW) life time predictions [1] (critical blanket module BM11). After that the same code was significantly improved during its application to existing fusion-relevant plasma devices: the tokamak JET equipped with an ITER-like wall and linear plasma device PISCES-B. This has allowed testing the sputtering data for beryllium (Be) and showing that the “ERO-min” fit based on the large (50%) deuterium (D) surface content is well suitable for plasma-wetted areas (D plasma). The improved procedure for calculating of the effective sputtering yields for each location along the plasma-facing surface using the recently developed semi-analytical sheath approach was validated. The re-evaluation of the effective yields for BM11 following the similar revisit of the JET data has indicated significant increase of erosion and motivated the current re-visit of ERO simulations.