Page 229 - Fundamentals of Magnetic Thermonuclear Reactor Design

212 Fundamentals of Magnetic Thermonuclear Reactor Design

related to this criterion have been known as the ‘first-wall problem’ since as
early as the 1960s.
During a stationary discharge phase, high-density plasma occupies the
core of a plasma column. Because the plasma confinement by magnetic fields
is not ideal, H ions and α-particles drift from the core towards the wall,
cutting across the magnetic field lines, losing energy and getting recharged
(through charge-exchange reactions) on their way. Hitting the wall, they pro-
duce a double harmful effect of, first, thinning (by sputtering) the wall and,
second, cooling the plasma by radiation losses through poisoning the plasma
by wall erosion products.
To use the expensive superconducting magnetic system (that accounts for
20%–25% of the total tokamak capital cost) to the best advantage, the edge
plasma layer between the plasma column surface and the FW is made only
3–4 cm thick in large operating tokamaks and up to 10 cm in the international
thermonuclear experimental reactor (ITER). This ring-shaped layer is held in
place by the limiters—FW structural components located along the toroidal cir-
cumference of the chamber.
The particle and heat flows hitting the FW are extremely severe (see
Table 7.1 [1, 2]), especially near the limiters and divertor targets. Disruption of
the discharge current (when the instantaneous values of heat flow density may
2
reach 10 GW/m ) is a big hazard.
One important point is that a simple understanding of the physical pro-
cesses taking place in the plasma does not necessarily mean that you can
provide a detailed prediction of the plasma characteristics needed to de-
sign new, larger-scale machines. Existing scalings have an empirical base
and cannot ensure the quality of ‘long-shot’ extrapolations. One should
also bear in mind that in ITER, the plasma column size is several-fold,
and its volume is an order of magnitude greater than in the precursor JET
tokamak.
The prediction uncertainty is particularly relevant to cases where a predic-
tion of current disruptions, Edge Localized Mode (ELM) effects (instabilities
developing at the plasma edge) and runaway electron effects is to be made,
and where the number of prospective events, as well as heat load magnitudes,
lengths and locations are to be modelled.
Thus, data concerning the FW components’ operating environment should
be treated as tentative, and this assumption should be given precedence in the
design of the FW components.
To perform the protective function, the FW must not be transparent to
particle and heat flows coming from the plasma. In other words, it must
have optical integrity and use a cooling system. In this respect, the graph-
ite liner surrounding the plasma column and re-emitting heat coming from
the plasma can hardly be regarded as an absolute solution: while confining
the plasma spatially and shielding nearby structures from incident particle
fluxes, the liner offers no protection against the heat flows.