Can the ITER plant explode?
The amount of fuel in the plasma at any time is about 1 gramme - enough for a few seconds' reactions. The reaction process itself is a fine balance of making an ultra-high vacuum, injecting only fuel into it, driving adequate currents within it and in surrounding coils to contain it, and reaching sufficient density to be able to heat it enough to raise fusion power to peak levels. If anything goes wrong in any of these processes, the plasma hits the surrounding material walls, and the resulting flood of impurities extinguishes the plasma. That is the challenge of fusion. There can be local problems with all the energy concentrating on certain surfaces, causing melting and damaging the investment, but there is no nuclear-driven thermal runaway possible.
Certain accidents can release water coolant into the reaction chamber. If there are hot surfaces nearby these can dissociate water into hydrogen and oxygen, and explosive mixtures can result if care is not taken. Again, the energies are enough to damage the investment, but not to cause sufficient damage to affect the general public. The plant is designed with various lines of defence to avoid the possibility of contamination outside the plant in the case of such hypothetical accidents.
How big a leak of radioactivity can occur?
A large number of accident sequences have been analysed in ITER, and this has led to the identification of limiting cases which envelope the worst case scenarios. These limiting cases, when combined with the most damaging weather conditions, lead to estimates of the amount of radioactivity that can be dispersed outside the plant under the worst cases. Assuming no evacuation, this then leads to estimates of the dose received by members of the surrounding population.
These analyses show that under normal operation the extra annual dose to the most exposed individual is about 1% more than that of natural background radiation. Under the worst case accident scenario, the extra annual dose the most exposed individual would receive is about double that of natural radiation (which itself varies by more than this factor from place to place). Even under the worst accumulation of accidents (for which no chain of events can be envisaged), evacuation of the most exposed individual would not be required according to ICRP and IAEA guidelines.