FAQ on Liquid Metals (LM)

on liquid metal walls’ benefits

  • Plasma-facing liquids can handle neutrons and energetic charged particles better than solid walls. For instance, they suffer no structural damage by neutrons, because they have no crystal structure. Erosion and sputtering are much reduced. Finally, the first wall can tolerate higher heat fluxes, if it flows (compared to a static solid wall with embedded cooling conduits). Important bonuses are reduced maintenance and more continuous operations.

  • Li-based liquid walls, even thin ones (< 1 mm), have benign erosion, sputtering and recycling properties. They basically act like sponges that absorb impurities and do not release them back in the plasma. This keeps the plasma clean, which is essential for fusion. One reason is the high chemical activity of Li -whether solid or liquid. Other reasons are physical (low sputtering and erosion) and they benefit from the wall being liquid.

  • Sufficiently thick walls (35 cm, for our choice of materials) stop most neutrons. This serves two purposes: preventing neutron damage and activation of the rest of the reactor, including HTS, and harvesting fusion energy, 80% of which is indeed carried by neutrons. The remaining 20%, in the form of alpha particles, remains in the plasma and keeps it fusion-hot.

    Our Li-based wall is thick enough to breed all the Tritium (one of the fusion fuels) needed by the reactor. Effectively, our liquid wall is a “naked” breeding blanket, directly exposed to the plasma.

  • Flowing walls exhaust the extreme heat and particle fluxes from the plasma, thus increasing the lifetime of plasma-facing components.

    In this approach, the flowing, plasma-facing liquid metal is the first working fluid of the power plant. In our specific design, heat is then transferred to supercritical CO2 and finally water.

on electromagnetic control

  • Liquid metals are electrically conductive. When an electric current flows through them, its combination with the ambient magnetic field used to confine the plasma generates a force, known as Lorentz force. Strategically placed electrodes result in currents and thus forces of proper direction, as to counteract gravity.

  • No: thanks to the high magnetic field generated by HTS to confine the plasma, the Lorentz force on the liquid metal can easily exceed gravity, even for small currents applied.

    Said otherwise, parallel currents attract each other. In this case, the currents in the rigid HTS attract the deformable, current-carrying liquid metal. Since the attractive force scales like the current product, and currents in the HTS are very high (tens of MA), relatively low currents in the liquid metal suffice for levitation (few kA).

  • There will be sensors of thickness and velocity. In feedback with them, actuators will apply local currents, hence forces, and ultimately control the local thickness and velocity of the liquid metal. In fact, electrodes will act as both sensors and actuators, and will be aided by other types of sensors. An array of electrodes ensures good, uniform thickness. Experiments with arrays of sensors and arrays of actuators have already begun. Next, we will interface them and “close the feedback loop”.

  • Yes, there will be ports for plasma heating, diagnostics, pumping and pellet fueling. Initial experiments showed that the liquid metal just flows around the ports like a river flows around the pillars of a bridge: the flow is depleted in the wake of the port, and turbulence is enhanced. Both effects are slight; the former and, to some extent, the latter, can be electromagnetically counteracted by electrodes (see previous question).

on materials and corrosion

  • The liquid metal will be in contact with solid parts both in-vessel and outside the reactor, in the plant that will clean it from impurities and other byproducts, extract heat, and extract tritium. For minimum maintenance and replacement, solid parts must thus have good resistance to corrosion by liquid metals. We deal with this by testing and implementing corrosion-resistant materials and coatings. Generally, other corrosion reduction strategies would be to reduce the temperature and flowrate. Those are not possible here, due to heat extraction from the reactor.

  • Instead of lithium or a lithium alloy, we will use a solution of lithium and lithium hydride. Since lithium hydride has a higher melting point (688 C) than lithium (180 C), the Li-LiH solution will have a lower tendency to evaporate. Consequently, it can operate at a much higher temperature than liquid Li (200 C higher), for the same evaporation rate. Note that the ultimate limit lies not in the evaporation rate “at the source” (the liquid wall), but in how much Li and H effectively reach the diverted plasma. This is a complex transport problem that we are currently modeling and will experimentally characterize and optimize. For now, based on peer-reviewed tolerable evaporation rates for Li, we estimate that the +200 C “advantage” will allow operations at ~700 C, possibly higher. The higher the temperature the merrier, for the thermodynamic efficiency of the reactor. Our goal is to approach 900 C as closely as we can, but even 700 C would be far hotter and far more efficient than any competitor.

on liquid metal droplets

  • In a tokamak, a liquid metal droplet or solid wall “flake” or dust cause a disruption: the stored energy and plasma current are lost in few ms. Its consequences (runaway electrons, halo currents, electromagnetic stresses etc.) can damage the tokamak. Interestingly, if a disruption is inevitable for other reasons, “killer pellets” or liquid jets can be deliberately injected to trigger the disruption in a controlled, mitigated manner.

    Stellarators, on the other hand, have no plasma current. Solid or liquid objects falling into them cause a “radiative collapse”: the stored energy is lost in few ms in the form of a big flash of light. Massless photons cause practically no damage to the device. Nevertheless, the event is still a nuisance, because the plasma needs to be re-initiated and brought to fusion-relevant densities and temperatures again, which takes several seconds. During that time there is no electricity black-out because thermal power plants have a “thermal inertia” and keep producing electricity even if the heat source stops for several seconds.

  • We continuously reduce the frequency of these events and are currently at approximately one droplet per minute in a reduced-scale, cylindrical prototype.

  • The ‘hard’ or essential limit is one droplet every many minutes (10’?) in a full-scale, stellarator reactor - the exact requirement is under discussion with utilities. The reason is that a droplet causes a radiative collapse; after that, the plasma must be re-started and ramped again to fusion densities and temperatures, which takes several seconds. That is, a droplet causes an interruption of less than 1’ in the fusion heat supply. One such interruption every 10’ translates in a duty cycle of 90% or higher. While not taking full advantage of stellarators’ steady-state nature, this duty cycle is still much higher than tokamaks or other fusion devices, and perfectly acceptable for an energy supplier.

    It would be even better if droplets were so infrequent that the plasma is only interrupted when planned (for example to remove Helium ashes and impurities and start a new, cleaner plasma) and there are no unplanned interruptions due to droplets. For that, the average time between subsequent droplets falling in the plasma must exceed the plasma-discharge duration, which for stellarators can last hours. Longer than that is possible but unnecessary and counter-productive from an ash, impurity accumulation and other operational standpoints.

  • Droplets form when enough energy is supplied to the fluid to break the surface tension that holds the layer/stream together.

    To that end, one must avoid discontinuities in the solid “riverbed” on the side opposite to the plasma, as well as sharp bends and sharp obstacles amidst the flow.

    (Magneto)hydrodynamic oscillations, waves, instabilities, and turbulence also perturb the free surface. The good news is that these disturbances are damped by the strong magnetic field and rarely become large enough to break into droplets (“passive stabilization”). Deliberately applied e.m. forces provide additional “active stabilization” or even “feedback stabilization”, in feedback with sensors.

Have a question and don’t see it here?

Send it to contact@renfusion.eu. FAQs will be answered here, less frequent questions will be answered by individual e-mails.