What does it take to build a commercial hot fusion reactor?
Nuclear fusion is cool or hot — pick your favourite adjective. This article is part of our fusion series. Read the main fusion article here. Keen to know more about the fundamentals behind hot fusion technology? Get all the details here.
By Eshel Lipman, Torben Schreiter & Iris ten Have
You might be surprised to learn that the step-by-step plan for commercialising fusion energy isn’t all that different from other hardware-based climate solutions. It is a matter of researching, optimizing, and eventually reaching a commercial solution. And yet, we don’t have any commercial fusion system.
Let’s deep dive and see what the detailed research and development process of building a hot fusion reactor looks like:
- The research phase. It all starts with understanding the fundamentals that are essential in plasma physics. This is the experimental phase, where a startup should be looking into heating systems, confinement topology, reactor design, and working on successfully creating plasma. Theoretical simulations of magnetic confinement topology and reactor design are essential, as they will decrease the initial CAPEX costs significantly: instead of building a physical machine, calculations can tell which design is the most promising. With the increased computation power of nowadays, a full setup can be simulated before being put into experimentation.
- The optimisation phase. In this phase, the relevant parameters identified in the research phase will be optimised. The optimisation should enable testing materials, such as high-temperature superconductors (HTS) if needed, other magnets, a lithium blanket on the reactor wall, the confinement topology, etc.
- The self-sustained plasma phase. Here is where it gets a bit more specific for commercialising fusion energy. This phase should be focused on achieving a self-sustained plasma leveraging the right topology as identified in the optimisation phase.
- The net energy demonstration phase. This phase might be the most challenging. So far, getting more energy out of fusion than was put in has only been recently demonstrated for a few seconds in academic research facilities.
- The commercialisation phase. This is the stage at which an actual device for the commercial production of energy from nuclear fusion is commercialised. Similar to any other hardware-based climate tech solution, this phase could be sped up by adopting a modular approach. Don’t forget, supportive legislation will be a key requirement for the design and operation of commercial fusion reactors as well.
So where do we stand with the hot fusion technologies in relation to the five stages? To answer these questions we need to talk about the Q-factor.
“Q” — the key to gaining net energy from fusion
How much energy goes into a fusion reactor and how much comes out of the reactor?
Typically, when looking into fusion, you will hear about the term “breakeven” or the Q factor. Q is the ratio between the input power and the output power. For example, Q=1 means that the system is giving exactly the same amount of energy as it has absorbed from the outside.
However, aiming for Q>1 is not straightforward as there is more than one Q. This is very relevant, but mostly neglected in mainstream coverage. Let’s go over the definitions:
Q_scientific (sometimes called Q_physics) refers to the ratio of energy (heat) produced by a fusion reaction to the energy consumed by the fusion reaction. In more technical terms, Q_scientific is the ratio of the thermal power produced by a fusion plasma to the injected thermal power used to heat that plasma. Whenever evaluating any current Q_scientific achievement today, it’s important to understand that even though this factor is great for scientific experiments, it doesn’t consider the inevitable energy conversion losses when producing that heating power, which are significant. Nor does it consider the energy required to operate the rest of the reactor.
When you read about “Q” in mainstream media, in almost all cases, it will be Q_scientific. Thinking back to TRLs (that we discussed in our main article of the series), any experiments that manage to demonstrate Q_scientific>1 would be comparable to reaching TRL 3, the final stages of the research phase.
Q_engineering considers real-world constraints such as materials limitations, geometry, and operating conditions. This factor provides a more realistic assessment of achievable performance levels. This means energy in versus energy out at the total system level. In other words, energy is drawn from the outlet versus energy is exported back to the grid.
Any Q_engineering value greater than 1 means the reactor can produce net electrical power. Q_engineering=1 would be comparable to TRL 4, a technology demonstrated in a controlled environment.
Q_commercial is probably the most relevant Q for investors. Reaching a Q_commercial=1 would be comparable to TRL 5, a technology validated in a relevant environment. Once more energy is produced than consumed on a technical level, we need to consider initial capital investment (CAPEX) and any operational expenditure (OPEX, e.g fuel cost). In other words, this is the energy breakeven that will ensure the financial viability of the entire project (including the profit for the off-taker, standardisation, and creating the supply chains and others).
Since no fusion company is close to this stage, any discussion about the viability of such a project is purely hypothetical at this point in time.
Now that we have an understanding of the Q factor, we should look at the relationship between the different Qs. This is perhaps even more important for understanding the economics of fusion. Here we see that the relationship between the three different Qs is not linear. See the graph below.
To make it clear, let’s analyse the last NIF’s Lawrence Livermore National Laboratory successful experiment, which announced a net gain:
Q_scientific. NIF produced 3.15 megajoules (MJ) of fusion energy output from the 2.05 MJ of laser light that reached the small cylindrical chamber. Thus, Q_scientific=1.53. Q>1, hence a gain of energy, but is that really the case?
Q_engineering. The calculation didn’t involve about 300MJ that was needed to power up the 192-beam laser in the first place. In this case, Q would be Q_engineering=0.0105. This means we need to see a 100x improvement before the energy balance of the system will be positive; or before Q_engineering>1.
Q_commercial. This is not easy to calculate, but you need to factor in CAPEX, OPEX, and profit margins as well. Therefore, in NIF’s example, Q_commercial would have to be much bigger than 1000 MJ — which is several orders of magnitude away. In other words: a commercially viable (laser) fusion system is many years away from us.
The following illustration shows the relationship between Q_scientific and Q_engineering, as the total amount of energy drawn from the grid was almost 300 times greater than the net energy produced by the fusion reaction.
With mainstream media currently covering the topic extensively, it is easy to see that great progress is made in fusion research all the time. And while that has certainly allowed us to move past Q_scientific>1 and somewhat closer to Q_engineering=1, we are actually still two orders of magnitude away from it. That is a lot. Needless to say, we are still very far from the ultimate goal: Q_commercial>1.
And why is that? Let’s discuss the biggest hurdles fusion researchers and entrepreneurs are facing along their journey.
The challenges that are delaying fusion energy’s commercialisation
Below we will discuss 6 key challenges that need to be overcome by researchers, engineers and policymakers if fusion is to become a controllable, viable energy source.
Technology & engineering challenges
1. Physical prototyping costs and duration are high. In fusion but also in other companies with a substantial hardware footprint, every iteration of the prototype is time- and resource-consuming. Confinement of the plasma requires very specific reactor configurations that are extremely complex to fine-tune. Therefore, timeframes may be way long and can accumulate to anything from a few months to years per device. For example, it took 10 years to design the Wendelstein 7-X and another 10 years to build. That means the Wendelstein 7-X required 20 years and €1B from kickoff to the first lab experiment. That is not to say that commercial operations might not be faster, but it would be surprising if they were orders of magnitude faster.
Through the research phase, small and cheap setups can be used to test the heating systems and other technological parameters, however, the breakeven state and the reactor state require a final design. There is a disproportionate relationship between reactor surface area and plasma volume. If you build too small, the heat losses through the surface will render the experiment useless.
2. The initial energy required to kickstart nuclear fusion is high. The issue with kickstarting nuclear fusion is that it requires very high temperatures and densities as well as long confinement times. Therefore, fusion reactors will require a significant amount of energy to start up the process. Once started, any commercial fusion reactor would be able to sustain itself through the process of nuclear fusion, but during the prototyping phase, the energy supply burden on the facility is not an easy one. Kickstarting the reactor requires high currents and thus complicated and expensive power electronics.
3. Plasma instabilities stop the nuclear fusion reaction. The magnets’ physical configuration must be extremely precise to successfully sustain the plasma (in Wendelstein the precision tolerance is 1 millimetre!). Plasma magnetic hydrodynamic instabilities are a type of instability occurring when the magnetic field is not uniform. This can happen if the plasma is moving or if there is a tiny change in the magnetic field, causing a positive feedback loop, which will completely destabilise the plasma.
Additionally, it is a major challenge to control the plasma’s turbulence. The helium and neutrons formed as a result of nuclear fusion should move away from the centre of the plasma towards the outside of the reactor into the lithium blanket but at a very specific rate. The lithium blanket is mainly there to absorb the neutrons formed and to “breed” more tritium fuel, while the heat produced by the fusion reaction is extracted by a cooled divertor. However, if the particles leave the centre of the plasma too quickly, they essentially take out too much of the heat with them and the plasma gets too cold. This stalls the self-sustaining fusion reaction and the reactor efficiency with it.
Commercialisation & scale challenges
4. There are supply chain issues. The supply chain of fusion fuels, such as deuterium and tritium (the latter being a rare and volatile radioactive element with a half-life of only 12 years), should be developed further. In addition, the supply chains related to high-temperature superconductors (HTS) as well as other rare earth metals and hard-to-source materials that are required both for building the prototypes as well as during ongoing operations should be strengthened.
5. The world needs more nuclear physicists, mathematicians and plasma physics researchers. To create a well-established industry around it, we need to build the next generation of innovators in the space. One may say that with the increase of the fusion trend, more professionals will come, but we need to consider the role of additional ‘artificial’ support, like enhanced governmental intervention in endorsing relevant courses in academia, grants for relevant research and more.
6. There is no regulatory framework for operating nuclear fusion power plants. Fusion reactors will be deployed first in countries with a regulatory framework that allows deployment while ensuring safety. If we learned anything from the 70-year-old fission industry, it won’t be an easy ride and will face a lot of NIMBY resistance. To speed up the process of scaling fusion (if and when this happens), regulators need to start creating standards and regulations now.
Some of the startups like Commonwealth Fusion Systems and Tokamak Energy have already made progress in some of the areas mentioned above, particularly in creating a stable plasma, with some optimization already completed.
Most startups, however, are focusing solely on progressing towards a Q_scientific>1 as a first milestone, while kicking some other substantial problems down the road, which will obviously take a lot of time and a lot of funding. While the engineering and scientific challenges have to be solved to get Q_engineering>1 and practically allow fusion-based energy, we do believe that any regulatory and scale challenges elaborated earlier will be solved quickly after achieving the relevant technological milestone.
Can fusion be a viable energy source one day? Definitely. Does anyone know when the first commercial reactor will be up and running? Not yet.
Are we developing an appetite for a scientific breakthrough that will accelerate the energy source of the future? Yes! But many stars need to align for (hot) fusion to take off.
This is why we invite you to head to our main article of the series that explains our views towards the industry as a VC case.
