Overview: steering the path to practical fusion power
Fusion energy promises clean, abundant electricity, but turning that promise into reliable power requires exquisite control of the plasma and the reactor environment. Researchers at the UK’s MAST Upgrade and the Dutch institute DIFFER are making key strides in two critical areas: distributing heat evenly inside the reactor and regulating the core fuel density. Their work lays important groundwork for future facilities like STEP in the UK and the international ITER project, where stability and predictability are non-negotiable for sustained operation.
These advances show that even tiny shifts in magnetic fields can instantly alter where heat and particles travel inside a fusion device. At the same time, smarter fuel-management strategies—ranging from gas injection schemes to the use of frozen pellets and advanced predictive models—are enabling more reliable density control. Together, they address the twin challenges of preventing local hot spots and maintaining the right fusion power output.
Power-sharing dynamics in a double-null configuration
One strategy to spread vast heat loads is the double-null configuration, where an extra divertor helps distribute power more evenly across the vessel wall. In the UK’s STEP project, this layout directs most heat to the outer divertors, reducing the stress on inner targets and enhancing the machine’s resilience. The double-null arrangement is formed by overlapping upper and lower magnetic geometries; when aligned, power-sharing is balanced, but even small misalignments can bias the system toward the upper or lower divertor.
Experiments in MAST Upgrade, conducted with researchers from DIFFER, investigated the dynamics of double-null power sharing for the first time. The results were striking: fluctuations in power-sharing directly and immediately altered divertor target conditions. Even at modulation frequencies up to 200 hertz, the impacts appeared at the targets without delay, implying there is little time to respond to imbalances. This finding underscores the need for continuous, real-time control to maintain balance, a requirement that will shape design and operation of future reactors with similar configurations.
Core density control and density-feedback strategies
The core challenge for fusion power is to sustain a high density of fuel ions to maximize reaction rates. In deuterium–tritium (DT) plasmas, the rate of fusion scales with particle density in the core. At MAST Upgrade, a long-running core density controller faced limitations, sometimes allowing the density to drift toward unstable values that terminated discharges or triggered protective sensor events when density dropped too low.
Researchers developed a density-control approach that uses simple, robust feedback: injecting gas via a valve driven by a sinusoidal voltage pattern, while measuring the resulting density with a laser interferometer. The key insight was to extract dynamic behavior from the resulting time-series data, allowing the team to infer how long it takes for the core density to rise and by how much when the valve opens by a given amount. This enabled a cruise-control style approach—an offline, model-informed set of settings that could be used to guide real-time operation without additional experiments.
In a later demonstration, the cruise-control followed target densities, visibly integrating the density trajectory with the magnetic confinement. This work contributed to a substantial reduction in disruptive terminations, increasing the success rate of experiments by roughly 20%. The density-control developments also provided essential groundwork for more advanced exhaust control strategies, where coupling between core fueling and edge behavior must be carefully managed.
The injection of frozen pellets: a high-precision percussion for core fueling
Gas-valve fueling works well at the device edge, but for a full-scale reactor the outer edge is too hot for gas to penetrate effectively. Frozen pellets, injected through the edge, are the approach of choice for delivering fuel directly into the core, increasing local density where fusion power is produced. However, pellets introduce discrete, sizable density changes that can stress the edge and trigger disruptions if not carefully managed. In ITER, each pellet can represent a non-trivial fraction of the plasma particle count, making real-time control a complex problem.
To address this, scientists developed a predictive control framework that can optimally time and size pellet injections. The model predictive control (MPC) approach evaluates high-fidelity simulations of ITER-like plasmas to determine the best pellet strategy while keeping edge densities within safe limits. The results show a path to robust pellet-driven fueling that can be adapted to other devices, balancing rapid response with disruption avoidance.
Connecting to STEP and broader fusion programs
All these control pieces—balanced power-sharing in double-null geometry, dynamic core-density regulation, and pellet-based fueling with predictive control—converge in STEP, the UK program that seeks to demonstrate practical fusion power. The work emphasizes two crucial ideas for a viable reactor: (1) deploying robust, continuous control loops capable of countering fast disturbances in the divertor and exhaust systems, and (2) using predictive, physics-informed models to anticipate plasma behavior and plan actuator actions accordingly.
Beyond STEP, the methods and findings feed into the ITER project and other international efforts, where the scale and complexity of the plasmas demand high-fidelity control architectures. By combining real-time sensing, fast-acting actuators, and predictive models, researchers are moving closer to stable, controllable fusion power plants that can deliver electricity with minimal environmental impact.
Implications for the future energy landscape
As fusion devices grow in size and sophistication, the ability to regulate heat loads, control core density, and schedule fuel injections with high precision will be a defining capability. The cross-border collaboration between MAST Upgrade and DIFFER illustrates how targeted experiments and modeling can uncover practical control strategies that reduce risk and improve performance. If these advances continue, they will accelerate the realization of fusion as a reliable and clean energy source for future generations.
