Ulricke,
I feel that the safest approach to the use of the IAEA primary standards
is to make thorough calibrations of your internal standards upon first
opening and then not attempt to store them. Reality is that we always
want to try and extend their lifetime.

We have three natural abundance-range water standards which we calibrate
at two yearly intervals against the primary stds and two enriched stds
produced by dilution of 10%+5% 2H and 18O enriched water to get 100 and
500 permil for D. These are our daily calibration materials for both D
and O18. The natural abundance waters are each 50 litres of Antarctic
ice, local tap water (45 degrees South) and sea water, distilled and
stored under silicone oil in opaque (s.s.) tanks. With these three we
cover the range -230 to 0 for D and -30 to 0 for O18

WE supplement our calibration statistics by sending the waters out to
other laboratories too.

In my view, the Gehre method is really good as published

Cheers

Robert Van Hale