MIT framework gives fusion developers a way to test viability

The next generation of fusion devices is exactly where this framework gets tested. Tokamak, inertial, or magnetized-target, the framework doesn’t discriminate.

(Image courtesy of EUROfusion)

After decades of fundamental research, fusion is moving towards practical energy applications. A new generalised framework from researchers affiliated with MIT and Rutherford Energy Ventures, led by D.G. Whyte and A. Lo, asks a more commercial question: how should a fusion power plant’s economic gain be assessed?

Its economic gain factor, Q_econ, is the ratio of a fusion power plant’s economic gains to its costs over the plant’s lifetime. The framework takes inspiration from the Lawson criterion, the physics threshold first derived in 1957, applying a similar universal logic to a plant’s economic gains and costs rather than to plasma performance alone.

Necessary, not sufficient

The rule at the centre of the framework is direct. Q_econ must be at least 1 for a design to offer any prospect of net returns under the model. Meeting the threshold does not guarantee commercial success. Most but not all costs are included in the model, so reaching it does not by itself establish real-world commercial viability. Missing it means the design offers no prospect of net returns within the model’s assumptions.

One surface, many concepts

The framework is designed to apply across different fusion plant concepts. Instead of modelling a specific reactor, it works from a control surface, denoted S, surrounding the fusion fuel. This is the boundary through which the model assumes all fusion energy must be extracted before it is converted into useful energy for sale. Costs and gains are normalised to that surface rather than to absolute power output, making the resulting criteria independent of absolute power production and impartial to fuel cycle or confinement method. The engineering and cost inputs still have to come from the individual design.

Utilisation, the fraction of calendar time a plant is operating and producing energy to sell, depends on how long the control surface can operate before it needs replacing and how long replacement takes. X_S, the surface’s most optimistic lifetime limit, is modelled as an energy-fluence limit, though other failure modes may impose earlier limits in practice. For the same normalised fusion power density and otherwise identical inputs, a lower fluence limit produces shorter operating periods, lower utilisation and lower modelled economic gain.

Ten dials, not one design

Ten controlling normalised design parameters make up the full model. These include fusion power density, the surface’s energy-fluence limit, replacement time, net price of energy, systems-wide net-energy conversion efficiency, target cost per unit of energy yield, and the normalised construction-and-delivery and fixed O&M costs of the plant. The authors vary these parameters across wide ranges to examine how design, financing and operational choices affect the prospects for economic viability.

The approach applies to steady-state and pulsed concepts alike, and the paper illustrates how it could be applied across magnetic, inertial and magnetised-target fusion. Turning it into numerical results for a specific design still requires cost, performance and finance data particular to that concept and marketplace.

Why the timing

The paper places the framework in the context of fusion’s movement from fundamental research towards practical energy applications. It points to growing demand for dispatchable, sustainable, carbon-free energy with high power density, including from emergent energy-intensive sectors such as artificial intelligence. It also cites recent advances including an indirect laser-drive fusion implosion with net energy gain and significant self-heating from fusion products, and high-temperature-superconductor fusion magnets demonstrated at very high magnetic field.

The paper’s question is what comes next. How should these technical achievements be assessed in economic and commercial terms, rather than only in scientific ones? Lo told MIT News that it’s challenging to reduce complex scientific and engineering requirements to economic consequences.

What it could mean for industry

For developers, the framework could provide a way to stress-test a design’s modelled economics. Which of the ten parameters is the weakest link, and how much would it need to change for Q_econ = 1?

For investors, it could provide a common analytical yardstick for examining technologies that are otherwise difficult to compare directly, giving a way to ask a tokamak company and an inertial-fusion company the same economic question using the same framework.

For suppliers, the fluence-limit parameter could be particularly consequential. It links the model’s energy-fluence limit for S to plant utilisation and replacement economics, making material lifetime and replacement intervals relevant not just to engineering performance but also to whether a design clears the model’s economic threshold.

The paper

The authors are affiliated with Rutherford Energy Ventures. Whyte is also affiliated with MIT’s Plasma Science and Fusion Center, while Lo is affiliated with MIT’s Sloan School of Management, CSAIL, EECS, the Operations Research Center and the Laboratory for Financial Engineering. The paper is available as an open-access CC BY 4.0 preprint on arXiv and is identified as submitted to the Journal of Fusion Energy.