Beryllium has been used as a plasma-facing material to effect substantial i
mprovements in plasma performance in the Joint European Torus (JET), and it
is planned as a plasma-facing material for the first wall (FW) and other c
omponents of the International Thermonuclear Experimental Reactor (ITER). T
he interaction of hydrogenic ions, and charge-exchange neutral atoms from p
lasmas, with beryllium has been studied in recent years with widely varying
interpretations of results. In this paper we review experimental data rega
rding hydrogenic atom inventories in experiments pertinent to tokamak appli
cations and show that with some very plausible assumptions, the experimenta
l data appear to exhibit rather predictable trends. A phenomenon observed i
n high ion-flux experiments is the saturation of the beryllium surface such
that inventories of implanted particles become insensitive to increased fl
ux and to continued implantation fluence. Methods for modeling retention an
d release of implanted hydrogen in beryllium are reviewed and an adaptation
is suggested for modeling the saturation effects. The TMAP4 code used with
these modifications has succeeded in simulating experimental data taken un
der saturation conditions where codes without this feature have not. That i
mplementation also works well under more routine conditions where the conve
ntional recombination-limited release model is applicable. Calculations of
tritium inventory and permeation in the ITER FW during the basic performanc
e phase (BPP) using both the conventional recombination model and the satur
ation effects assumptions show a difference of several orders of magnitude
in both inventory and permeation rate to the coolant. (C) 1999 Elsevier Sci
ence B.V. All rights reserved.