A number of responses to Weisz's article and another (very pessimistic) article on population appeared in the November issue.
The February issue carries a cover illustration of a simulation of a magnetically confined plasma, from an article (subscription required) by Donald Batchelor of Oak Ridge Lab, on the current state of computer modeling of toroidal fusion plasmas. There's some urgency to getting this right, as the fusion community prepares for the ITER project - if the site selection process ever comes up with a location, that is...
The purpose of ITER is to reach the point where the plasma is actually "burning" - that is, the energy processes are dominated by the fusion itself, rather than external heating, and the plasma becomes "self-organized". What are the ideal external shapes and magnetic fields and injected RF waves to guide the formation of a stable burning plasma?
Batchelor's article discusses the emerging need for "integrated" simulations of plasma physics - rather than subdividing the problem into plasma microturbulence theory, magnetohydrodynamics (MHD), and transport theory as has been done in the past, there's a need to integrate all this modeling to account for couplings between the different phenomena in a realistic model of a whole reactor.
The immense range of time-scales involved means even the biggest supercomputers can't really simulate with reasonable detail for more than a millisecond or so. What's needed are models that integrate the real physics at short timescales and the immense energy stores and flows to allow simulation of long-term behavior - and those models will need to be validated with real experiments, not just simulations. In the work they're doing now, Batchelor and associates involved in the Fusion Simulation Project "will provide publicly released simulation tools that will include a collection of experimentally validated models of limited scope as developed in the research component." and they "will also create an evolving, comprehensive model called the integrated plasma simulator (IPS)."
The science challenges listed are informative:
For example, we in the plasma physics community need to understand how MHD stability is affected by plasma currents that can be driven by both transport processes and external sources such as waves. We need to understand the interaction between microturbulence and the global structure of density, temperature, and plasma flow. And we especially need to understand the physics of the plasma-edge region, where large- and small-scale stability and transport are all complicated by a number of factors: Those include proximity to the material containment walls, magnetic field structures that no longer form closed flux surfaces but make contact with solid material, and fluxes of neutral particles that are unaffected by the magnetic field. Conditions at the plasma edge, as experiments have confirmed, have an enormous effect on plasma confinement.
How close is practical fusion today? Would an infusion of significantly more research funds bring it closer? From a timeline of fusion projects, it's clear progress has been slow. The record for fusion energy production is still the 16 MW produced by tritium-deuterium fusion at the Joint European Torus in 1997 (Princeton's TFTR reached about 10 MW a few years earlier). US contributions in recent years seem haphazard, with initial support for the ITER project withdrawn and then later restored, but with conditions that have already resulted in a smaller-scale project. Nevertheless, once ITER actually does move forward, we may see some major strides toward commercial production starting in about 10 years (and after another $5 billion invested).
Fusion has great potential, but it still has a long way to go.