Just saw the first Hydrogen plasma at the Wendelstein reactor. I haven't found anywhere that gives a good explanation as to why they first used a Helium plasma - can anyone shed any light?
Converting my comment to an answer:
Well, I'd start up a new plasma system using an inert gas - I'd rather find out any problems with a non-explosive gas mixture. Once I know that everything works (vacuum, valves, no leaks, plasma actually ignites, ...) I would move on to the more exciting mixtures.
Remember - for any non-trivial system (and even many trivial ones!), the chance that it will all work perfectly the first time is very slim. It is always a good idea to test with as low a hazard level as possible, and design your system to be 'rescue-able' if something goes wrong (e.g. main vacuum pumps will always fail when the chamber is full of a pyrophoric gas - what are you going to do now?).
In my understanding the reason for using helium first is slightly different: It is correct that helium provides a safe way to start an experiment but safe in that sense that when you use hydrogen you usually have a problem with recycling and refuelling from the wall (hydrogen being bounded to some wall materials, leaves the wall-material, called desorption). This leads to an increase in the overall density which leads to a decrease of the temperature and you want to have high temperature.
In addition, the walls will be dirty for the initial experiments and bombardment with helium helps to clean them (hydrogen instead would tend to stick to the wall material, being adsorbed there).
update: As asked by @Emilio Pisanty, I found a reference in which it is stated:
It is known from previous fusion experiments that the walls can absorb and subsequently release significant amounts of hydrogen when hydrogen plasmas are used. Without a divertor, there is a risk that the shot-to-shot reproducibility of the plasma density evolution will be poor in this early campaign. Helium, being a noble gas, is much less likely to be loaded to the walls than hydrogen, and the shot-to-shot reproducibility of helium plasma density evolution is expected to be better.