 So, hello, everyone. Thank you all for joining us today in the second episode of this webinar series on breakthroughs in fusion research and development for those who missed the first episode, I'll put the link of the recording on the chat. So many thanks to our speakers for today. Jensen who Thomas St. Patterson and Ian Chapman for accepting our invitation to this event. I'm Matteo Barbarino for the International Atomic Energy Agency. This webinar gives an overview of the most recent groundbreaking results in fusion R&D to understand how such progress brings fusion energy closer to realization. Today's episode features the East Tocamac, the Wenderson 7x Stellarator and the Mass Upgrade Spherical Tocamac. As you probably know, on December 30, 2021, China's Experimental Advanced Superconducting Tocamac, or EAST, achieved stable 1,056 seconds steady-stake high-temperature plasma operation, setting a record long-pass operation. Wenderson 7x is the world's most advanced Stellarator and has been operating since 2015 in Germany. In these years, a number of breakthroughs have been achieved and more are expected in the upcoming operation phase. And the United Kingdom Atomic Energy Authority's new Mass-Tew experiment has demonstrated the effectiveness of an innovative heat exhaust system known as SuperX-Deverter. We're going to hear about these three facilities and their exceptional results. The format will be a sequence of three talks, 30 minutes each. Please type your questions and comments into the chat box. We will go through your questions during the 30-minutes Q&A at the end. So without further ado, please welcome Professor Johnson Hu, Deputy Director General, Institute of Plasma Physics, Chinese Academy of Sciences. Okay. Thank you, Chairman. I share my screen. Yes. Thank you. Okay. This is okay for you. Thank you. Ladies and gentlemen, it's my great honor to give a talk on the EAST Tocamac achieved a record longer-pass operation in thousands second scale on behalf of EAST team and collaborators. Sorry. So first, I would like to thank you all, collaborators from the world, contributed to EAST machine and plasma operations. Okay. This is the outline for this talk and these four parts. First, the challenge for longer-pass aesthetic operation. So longer-pass operation is essential for future fusion reactor. Static state Tocamac fusion reactor should be sustained. Fusion energy and output for sufficient longer time. And this fusion power is dependent on the three people product and the boiling time. First of all, conducting Tocamacs, no resistance in coils can be operated for longer-pass. EAST is also where we operate in longer-pass. So at this moment, development of supporting Tocamacs in the world where it includes AG7, EAST, QSTAR, WEST, GT6SA, and as well as AG1, as well as EAST. In the near future, EODEMO, CVTR, SPARC also will be designed as for superconducting facilities. High-temperature superconductor has attracted more attention for the superconducting Tocamacs. So this is key issues and challenges for longer-pass operations. First is plasma control and high-temperature performance maintenance in longer-time scale. And the second is heating and current driver for non-inductive stainless-state operation. The third one is plasma war interaction and its longer-time control. This includes recycling and particle exhaust, developed heat lot and power exhaust, impurity screen and control. The last one is called H integration for stainless-state reactors. It should be integrated of different multi-time scale physical process. So this is the second I will introduce the key to knowledge development for stainless-state operation on EAST. EAST superconducting Tocamacs have heat-like configurations, terminal, upper signal and low signal. And the major radio is about 1.7 to 1.85 meters and the magnetic field is about 3.5 tesla. It's my target is to get sensitive, for example, 1,000th second operation with about 20 to 30 megawatts, heating and current driver power and extensive advanced technology. My mission is to play the key role for understanding advanced stainless-state plasma physicals and provide valuable data-based basis for EAST and demo under the sensitive condition. So for an EAST have longer three days for longer parts operation, the first stage is the heating and current driver actions and relevant fundamental physical understanding and key to knowledge, key to knowledge. And the second step is to demonstrate longer parts more than 100 seconds at the mode plasma and develop for non-inductive hybrid scenarios. This is the first and second step almost finished and now is the third step. In this third step we would like to extend EAST operation regime to demonstrate sensitive high-performance plasma and deliver relevant physicals for future reactors. And in the last year EAST have made a bigger upgrade and we improve the heat flux capability for low-development and high-speed temperature resolution for diagnostics. And for the heating and current driver system, we arranged the parts for different heating systems and to improve the cobbling and injected power and also improve the god limiter heat flux removal capability. For the other system, we also improve the capability of the cooling system and also upgrade critical and pump systems. And for the main machine safety, we change the current leaders for the robust operation. This is for Tengsten-Teveto with higher heat load removal capability. We use either like Tengsten carbon structure to remove the graphite divider and the heat flux results in the capabilities increased from 2 MW per square meter to 10 MW per square meter. And for the flat-tap Tengsten-Teveto, the sample successfully tests heat flux removal capability to 20 MW per square meter. So this is the machine in 2014 we installed it like Apple Tengsten-Teveto. So in the last year, we installed upgraded lower Tengsten-Teveto. So now EAST have become four metal walls. So on EAST, we use very high efficiency work conditions including high-end current, SF, and routine intensive DCM work conditions and also use DCM injection during the plasma operation. This greatly reduces impurities, reduces power radiation loss and suppress recycling for good density control. We also control the hydrogen content in duty on plasma to lower than 10% for the SF heating. This increases the plasma stability and confinement. And we use good density control and particle loss, high palm speed working, and including the weather in weather wear pumps. This is about 210 kM per second for duty. We use the supersonic molecular beam injection and pallet injection for the core fueling. And we develop a variable heating and current driver system. On EAST, we use ARGCD, SF, ECH, and MBI. Totally the solar power is about 33 MW. In the last year, we arranged the pumps for different heating systems and reduce the merger interface in the furnace between the ARGCD and SF heating. And we also modified the SF antenna using a lower wave number about 7 per meter. And also ECH, we added our new CPI, Gertron in the last year. So totally for the ECH, power can be around for 200 seconds at 1.4 MW. So this is the heating and current driver for longer operation. So the Gertron ECH operation, as with negative cooling, is a durable launch. The ARGCD system with dedicated alternates and the gullible limiter, and double circle control to optimize the driving power, automatic restart control for Gertron and transmission. Plasma loop voltage control using 4.6 GHz ARGCD with optimized PID control in PCS. So on EAST, we developed a lot of diagnostics with the highest level and temporal resolutions. About 80 diagnostics have been installed. In the last year, 10 more new diagnostics developed. And these include MSE diagnostics for current profile control and the IR camera with large coverage. And the key diagnostics for longer operation include Thomson scaling per meter into barometer, motionless, dark effect diagnostics, and also less collective scattering diagnostics. We also developed a very stable plasma equilibrium magnetic control. Both is a feeble optic current sensor. Faraday effect with no signal drift and plasma and PF coils current measurement for plasma control is very good. And second is a very low zero drift integrated for longer plasma operation. The drift is lower than 200 megawatt valves, second for 100 second plasma operation. In the drift, deducation algorithm in PCS. And we use GPU area equilibrium reconstruction. This can be 100 times the computational acceleration efficiency, and extremely the PIFT routinely provides real-time magnetic equilibrium parameters. Now I give the talk introduction of achievements on 1000 second square plasmas. In the last 15 years, long plasma operation has been successfully extended on yeast. And this gives some mass storm. And in 2010, we get the first H-mode. And in 2012, we get 30 second H-mode, 400 second air mode. In 2016, we get 60 second H-mode. And in 2017, we get 100 second H-mode. In the last year, we get more than 100 with a high temperature, about 12 kV. And that's the operation, more than 1000 seconds. So this is the 4-mode, in target view, driving H-mode up to 100 seconds. And where's more? EIDB was found, and the confinement at 98 is larger than 1.015. And there is a 4-mode, or R-mode, driving and hitting, and there is a moment to input times the divider with the effective counter of heat and impurities. The times the impurities were low, and the maximum of the PFC temperature is lower than 600 degrees C. And the plasma density was well controlled with lower recycling, about 0.9. And we also get the 4-mode in target view, driving high beta scenario up to 60 seconds. This short H-mode is about 2.2, 2.1. And we extend the 4-mode in target view, high beta scenarios. The beta P was increased step-by-step in the last few years, and the highest beta P can reach to about 3, and the extended operation regimes, and for example high density. And the portrait for the ratio is about 50%, and this was achieved, dominant in electron hitting and zero pocket. And with small armies, where impurity was found, and the importer confinement, the highest, the edge 90-80 is about 1.4, and the EADP instead is the radio level of the smaller than 0.4. Okay, so in the last few years, we tried a lot of efforts to achieve the longer path operation with higher electron temperatures. From 2018, we get the plasma with the electron temperature higher than 0.8% and 0.6 keV, and in the last year, we get the electron temperature higher than 10 keV, sustained for 101 seconds, and we get the electron temperature higher than 6 keV and sustained more than 1,000 seconds. Okay, for this higher temperature operation, and this shot. Sorry, we don't hear you anymore. Just a second. No, there's no audio. Not only if you change something with the settings, but the audio just disappeared. I see the microphone is on, but we don't hear you. Now? Now it's okay. Yes, thank you. We just lost you when you went on this slide, so you can restart from here. Thank you. Okay, I don't know why, but I have not muted it. So this is, okay, sorry, that's the moment, I don't know why. Okay, this is, in this shot, we found the reduction of the core turbulence by multi-scale instabilities in the action with the MIGD, and we found in intrinsic turbulence, turbulent current in high electron temperature with EATB and turbulence current in current direction and self-regulation system of turbulence and turbulence current was found. Exciting and counteroffensive 1-1 mode using 4.6 GHz ARGCD was also found. And this is 1-1 mode. So since the state is a longer path with a high billiard EO. So for this shot, the temperature on the divetter is less than 300 degrees C, and you can find this is very stable and is very good, with a very controlled. Okay, now is the 1,056 second long plus plasma with four metal walls in the last year. And this is also fully non-inductive plasmas with dominated electric heating and zero injected torque. And the injection power is 1.65 megawatts, and totally the injected energy is 1.73 GHz. And the confinement H89 is about 1.3, and the plasma equilibrium configuration is where a good contract and the particle recycling and a lot also was good contract. And so this is the density and the electron temperature prefers for this mode. And this is the i-mode. And the temperature is quite high, but the density is not so high. So this is the typical i-mode. And in this shot, the RF current driver is about 70 percent, and the 4-trip is about 30 percent. So this is higher, as you see, the efficiency due to the higher, the wave for grains and the lower cycling. And this is driving the current to send the discharge to longer duration that can be measured by the current. Okay, this shot is also a new high confinement and self-organizing regime. This is a super i-mode. This means this is the i-mode and plus the EADP under the double no configuration with ETRO and WSM. And the BIDAPI is about 1.5. And the magnetic and balance that the boundary and plasma perimeters and over the turbulence transition, self-organized circle, maybe the key mechanism to sustain stationary, so-called i-mode. In the action between the i-MIDAPI turbulence and the electron heat transport for the substance or sensory IDP, so this is very good for this long plasma discharge. Okay, this is also very good confinement with, so this is EADP with no detectable turbulence and turbulence present in lower collision and the lower torque and electron heating plasma. And we also use in this picture to move the point to more than one at this position. And also we found the same state where my mode plays a key physical role in this discharge, reduce the turbulence level, active control, and by the loop voltage. Okay, this shows the robust plasma control in these shots. And a very good shape control is a new favorable optic current sensors. And a very low zero drift indicator for long path with linear drift detection as shown in pieces. And this is a low voltage control is very good and for this non-inductive current drive. So for the heat flux results, this is significantly enhanced by the water-cooled diverter and the new 10,000-diverter with a very high heat removal capacity and make the, the temperature was countered where it's lower than 500 degrees C and the advanced double-node-diverter configuration helps to reduce the targeted heat flux maintaining the temperature of the components in connect with the plasma on the control. Much lower, 10,000-diverter in weight leads to the lower 10,000-diverter source from the diverter. Okay, this shows the embedded behavior in this shot and you can find the, the, the, the low the impurities is very stable and the 10,000-diverter concentration is estimated to be five or six minutes of fire and the frequent spartan of the meta-linear impurities sometimes occurred from time at the 7,000 seconds. Which does not, does not, does not impact this discharge and the particle recycling was also very good and controlled. Optimized configuration for this is good particle control and also very high efficient fueling by SMBI and normal poofing was used and enhanced particle source due to the enlarged diverter slot for the pumping and you can store more pumping by real-time lithium injection during this discharge. Okay, I give the summary of the, and the future plan. Key technologies for this operation have been developed on EAST including high-performance diverters, robust plasma control, SMBI feedback fueling, particle pumping, well-conditioning, continuous heating and current drive also include also high-resolution diagnostics. Where integrated control on plasma configurations, equilibrium, heating and current drive heat flux, particle exhaust impurities and recycling have been well integrated for extension of this longer plasma discharge. EAST achieves this process of tokamak operation within a path loss in a thousand second scale with a plasma temperature in the tens of million degrees and a self-driving current either will be as a force, forcing for the EAST to cure larger than five longer path and excessive operation. In the future, we will try to, we will use plowing, use plowing, excessive plasma with high density, high temperature include iron temperature. We will use the EAST mode in longer plasma discharge in support of EAST and every time future reactors towards, so this, we will continue development of additional key technologies include the glass and refreeze charge, stable detachment at mode, long path, quasi-snow flake and discharge and so on. And also use upgraded NBI analysis for iron heating and use more EAST for current and the fuel profile control and also develop effective co-edge modeling and prediction. Okay, that's all. Thank you for your attention. Thank you very much, Zhenzhen for sharing with us your exceptional results. As announced, we'll please type your questions on the chat and we'll take them at the end of the Q&A. I see that some of you are indeed already doing that, so thank you very much. Okay. Now please welcome Professor Thomas St. Pedersen, head of the Stellarator Edge and Director of Physics at the Max Plan Institute for Plasma Physics in Germany. Thank you very much, Matteo, for the invitation and let me share my screen. Oops, excuse me. Right, so it gives me real pleasure to talk about the key results from the W7X Stellarator and a little bit about a vision for a Stellarator power plant based on the results that you'll be seeing. Giving this behalf on the entire W7X team, which is truly multinational. So here's my overview. I'll talk about Stellarators and their main features. I'll introduce you to the W7X Stellarator, its goals, its size, its scope and a bit about the timeline. And then I'll give you some of our highlights from recent operation. Talk about near-term plans and at the end close up with a vision for a Stellarator power plant based on W7X. And then I'll summarize. So to introduce the Stellarator, I'd like to compare it to the Tokamak. They have a lot of similarities. It's the same topology for the magnetic field, nested flux surfaces, a combination of toroidal and poloidal magnetic field giving the twist of the field lines. But they're created in different ways. Also, the Tokamak, as you just saw, very impressive results from East. Many large experiments exist in the Tokamak realm, whereas only a few large experiments have been put into operation to date. Tokamaks have already achieved good plasma confinement in terms of energy confinement time and also at high temperatures and density. So high triple products have been achieved. And for the Stellarator, it is necessary to use computer optimization to maintain a good plasma energy confinement time at high temperature. Density is in some sense a free parameter that you can increase quite a lot in the Stellarator. One drawback of the way the twist and poloidal field is created in the Tokamak, namely that it has a very strong toroidal current that creates the poloidal field, is that the instabilities may damage the vessel walls significantly. And we saw before that you can actually operate for very long times, but you also saw that this was a relatively low plasma current. The advantage of the Stellarator is that it is more stable in this regard, right away electrons are not a concern. You can think of it as a rigid prescribed cage created by the coils without the need for large plasma currents. If you want to operate the Tokamak in steady state, you must, at the reactor scale, you must dedicate several hundreds of megawatts to the current drive. So they will be recirculating back into the plasma, lowering the plant efficiency. And this is not the case in the Stellarator because we don't need the current drive. It is intrinsically steady state. And hence, as a reactor, it will have low recirculating power. So there are reasons why we are pursuing this approach, even though it requires a complicated computer optimization and has a more complicated geometry. Another advantage that I just touched upon is that much higher densities are possible in a Stellarator than in an equivalent Tokamak. And this allows a reactor to operate and burn at a lower temperature and still get the same amount of power out as the equivalent Tokamak. And this has lots of advantages in particular for the plasma wall interaction. Let me illustrate the importance of Stellarator optimization by showing you a classical Stellarator, a previous generation Stellarator less optimized than the latest. And first I've launched a particle with a lot of parallel kinetic energy which passed easily through high field regions shown in red. And now, and it was well confined because of the flux surfaces and the rotational transform. What I showed here now was the same particle but just launched with a lot of perpendicular kinetic energy so that it's magnetically trapped. And if the Stellarator has not been carefully optimized, such particles can tend to leave the device because of the gravity and curvature drifts illustrated down here. The Tokamak, of course, also has trapped particles as well as passing particles. The trapped particles are well confined because of the symmetry of the Tokamak whereas you need to apply essentially a hidden symmetry, hidden, not so obvious to the naked eye optimization of this. This orbit is just going to continue and be well confined. So let me start orbits in W7X. First, you see a particle launched with a lot of parallel kinetic energy passing through the high field and low field regions and tracing out a magnetic field line very closely. So let's go to the trapped particles which are the more complicated ones to confine. This is a deeply trapped particle. And as you can see with the optimization of W7X, the gravity and curvature drifts give you a colloidal procession that actually closes the orbits. So optimization and there are many different flavors of optimization one can pursue in Stellarators is a necessity to avoid prompt orbit losses and by that new classical transport losses which could be quite detrimental. And would dominate at the reactor scale to the point of making it difficult if not. So W7X's goal is to experimentally verify the reactor relevance of these optimized Stellarators and its status is that it has completed three operation phases since 2015. Its major radius is five and a half meters. The minor radius is half a meter so that across the plasma here is one meter and the plasma volume is 30 cubic meters. It uses superconducting coils that allow a magnetic field of 2.5 Tesla inside plasma. You see them appearing here in the CAD drawing. And as I just showed you, it has been optimized to give good energy confinement by reducing the new classical particle orbit losses. As mentioned, it's also been optimized so that it should have stability and a good equilibrium at relatively high plasma pressures volume average betas of up to 5%. It has furthermore been optimized to have an efficient power and particle exhaust solution using the island converter concept, which I will introduce. And in fact, it's been optimized for more things but these are particularly relevant and several of these I will show proof of that this optimization has in fact been possible and successful. I'm going to close up the cryostat in the CAD drawing here so that I can transition to a picture taken during construction. Obviously, we have been in operation since 2015 so this predates that but it's a nice picture that's showing the size of a person and you can look into the cryostat and also see the plasma vessel inside of that. Steady state operation will start in 2022 aiming in steps to go all the way up to 18 gigajoules per pulse. For example, that would be 30 minutes at 10 megawatts of heating power. Here is a summary of highlighted results. We have built the stellarator confinement cage, the magnetic flux surfaces and verified them with flux surface mapping with high precision, better than one in 100,000. We've verified the strong reduction of plasma generated parallel currents basically the bootstrap current and this improves the ability and assists efficiently exhaust using the island converter concept. And our latest result came out in nature about a year ago now proving that the orbit optimization works. In fact, the proof is that the confinement is so good also at high ion temperature that it would not have been possible in any previous stellarator and I will show you how that argument goes here. High performance discharge give us the chance of proving this statement I just made that the neoclassical optimization is successful. Given the measured density and temperature profiles of this discharge that was used in the nature paper the neoclassical transport can be calculated with high confidence given that it is a single particle orbit combined with binary collisions something that is quite tractable numerically. Our best performance has been with pellet fuel discharges so this one is also pellet fuel discharge with some density peaking which is important for some reduction of turbulence and this one applied central ECH heating of 4.5 megawatts and achieved an energy confinement time of 230 milliseconds 0.23 seconds. And most of the transport is actually due to turbulence it is about 70% of the heat flux the turbulent transport drives 70% of the heat flux at mid-radius and even more so at other radii and you see that here where we have calculated the neoclassical electron and ion losses and added them up and you can see that only 30% of the heat leaving the plasma at the mid-radius is by these neoclassical loss mechanisms so the rest is presumably turbulence but of course radiative losses which play a sub-dominant role and outside of that peak it is even more dominated you can see that temperatures are in the 2.5 to 3 kV range and the density is at 1.0 times 10 to the 20 that itself is not proof of optimization that proof comes here by comparing to earlier generations stellarator configurations. We can calculate the equivalent neoclassical losses and less optimized configurations by assuming the same density and temperature profiles and to make a one-to-one comparison so to speak we have scaled them to the plasma volume of W7X and the magnetic field strength of W7X and the result is that there are much higher neoclassical losses than these previous generation stellarator designs and the larger than the applied heating power in the W7X discharge we just showed. In other words they wouldn't have been possible those profiles that confinement time would not have been possible without the optimization in W7X as you just saw the neoclassical losses only amount to about a third and this is at the peak whereas in LHD we exceeded the total heating power the heat flux leaving the plasma at mid-radius going out to further radii would exceed it and that same would be 2 for W7AS which was already somewhat optimized and this one also has some optimization but not to the point of W7X and TJ2 one more generation back would have been very far from these results. So that's the proof that the optimization actually works and matters and can be measured. Let me move on to the island diverter which is a highly advantageous exhaust concept. The principle is that we're using an island chain at the edge of the plasma to create the indirect contact and X points for the exhaust and it works similar to a Tokamag in many ways just has many X points and this island chain works as a buffer region you have the heat flux going out of the plasma core going into the island along the field to the target plates and also across an X point here to the outside of the island also following field lines here because of course all of this is three dimensional and these diverter units which hug the plasma in the bean shape cross section, the thin cross section are obviously also three dimensionally shaped and you can see that there is evidence of a long interaction region here. This is also clear in infrared images that allow us to measure the heat flux in operation and the heat flux patterns are very much as expected. We did find some discrepancies which we in the meanwhile have been able to reconcile with the codes. These were minor but they were important to realize that one needs to add a little bit more physics to the code to get everything right. This is during attached operation where we have seen heat fluxes which are compatible with operation of the future WS-7X water cool diverter but are somewhat intense. But it gets better. We have achieved stable detachment for tens of seconds and we're showing a discharge here where the detached phase lasted about 28 seconds. During attached operation you get the heat flux going out as I described it before but the plasma can go into a state where it radiates away almost all of its heat and you can essentially not see on this color scale the heat flux during detachment. You see here time traces of important quantities. You can see that the radiation stays stable. You can see that the energy confinement time remains high. It has a little dip but it kind of recovers and you see the densities well controlled. You also see this little actually pre-programmed increase in density was exactly what triggered the detachment and the heat flux drop down to a much lower level. The diverter pressure, neutral pressure shows actually an increasing tendency and the particle exhaust remains efficient. The plasmas stay clean and this I believe is a discharge that could have been continued for much longer. It was essentially in steady state but let's not forget that the results I'm showing were with an uncool diverter and so the 30 seconds or so on the order a long discharge is 100 seconds because it was an uncool diverter. You saw this plot in the previous presentation also. Let me just point out where W7X discharges lie on this. This is a 5 megawatt discharge which had the record triple product for stellarators and it's very similar to the one that I showed you more details about as proof of neoclassical optimization. This is a 5 megawatt ECH discharge that I also showed you, namely the one that had detachment. It lasted for 32 seconds roughly of which the vast majority was in detachment and we had a low-powered discharge that allowed us to go to 100 seconds despite the diverter being uncooled. We are going back into operation and we have upgraded the device to actively cool components. Most importantly of course the diverter. This is a major undertaking to implement all this water cooling but this will allow us to operate for much, much longer pulses. The cooling water volume that we have is finite. That's what dictates the 18 gigajoule heating per pulse but we can certainly expect that for example the TETS discharge at 5 megawatt should be extendable up to the 18 gigajoule which would mean one hour and is 2 megawatt. One could operate for two and a half hours before reaching this limit. This high performance discharge was achieved only transiently and had some turbulent suppression which we will need to work on to extend over longer pulses but we believe that we have the tools to be able to do that because we will have major upgrades to both both heating and fueling. We will extend our ECRH facility from 10 to 12 gyrotrons and we have a program over time to operate 1 megawatt gyrotrons to 1.5 megawatt. The ECRH system is operating at the second harmonic at 140 gigahertz and it's laid out for a steady state in this case meaning 30 minutes. We'll be doubling our NBI power and adding ICRH and the increased heating will be combined with a state-of-the-art continuous pellet fueling system that you see having arrived together with the main scientist Jürgen Balsum which will lead to significant further performance improvements. We've seen that pellet discharges are the ones that have the highest performance and show some reduced turbulence. Looking towards the future W7X has confirmed a lot of expectations of ours so we should be asking ourselves how a stellar reactor would look based on this. The high field capability of high temperature superconductors as demonstrated so powerfully recently by commonwealth fusion systems with their 20 Tesla demonstration coil is a game changer in fusion. The triple product scales like B cubed at the beta limit or possibly even B to the fourth power if you believe that the transport is gyro boom. So you will get a very substantial boost if you add more magnetic field to your configuration. So we can scale up the W7X results that have been achieved to an ignited relatively compact one gigawatt electric steady state stellarator by doing a factor of two in linear size of the plasma and a factor of three in B field strength. This gives us round numbers to work with and predictions are that that would actually be an operation point that is ignited assuming that the ISSO4 scaling continues and that the that the confinement continues to improve with magnetic field as expected. Let me compare to either and to to the arc very compact high field Tokamak vision and plans for CFS. What I've done here is taken an illustration from a previous paper and taken into account the ability to go up to higher magnetic fields than this one and scaled it down accordingly. This allows us to come up with a vision of a power plant that you can see optically is actually on the size or even smaller than either in particular if you think about what is the largest coil dimension the largest coils and Tokamaks are the poloidal field coils which are necessary for position control of the Tokamak plasma and this is also you see in green here the largest coils in in arc the largest coils of a stellar even though this device is clearly much larger than this one also producing more energy the largest coil dimension which will be the non planar coils similar to what was the case in W7X will be actually substantially smaller and this will allow us to transport these more easily along infrastructure as many know it was necessary to wind the PF coils on site on either. So this very much mitigates the let's say disadvantage of the stellar being less compact than an equivalent Tokamak so to summarize W7X provides proof that optimization makes the stellar and attractive fusion concept the optimal optimization to reduce neoclassical transport is successful and I should add to this that we have new numerical results pouring out in also in a peer reviewed literature that show that the neoclassical losses can be further reduced in fact down to a level that is equivalent to those in a Tokamak or even the alpha confinement can be on par with the either or better those are new results that have come out in the literature in the last six months or so so W7X provides experimental proof that neoclassical transport reduction through optimization is successful but actually it's just the beginning of that journey we demonstrated stable complete diverted with good exhaust efficiency and this shows that one can one can alleviate the heat flux challenge at least we've shown this up to up to five megawatts and we expected to be further extended to higher heating powers also because we've seen that that radiating even a hundred percent of the injected power through an edge radiation layer is stable in W7X the preparations are well underway for steady state operation the in vessel installation has ended in December 2021 pump down and cool down was achieved in May we're still on track for first plasma operation for the fall of 2022 so there is much more to come in the next few years and you already achieved results extrapolate to an attractive robust flexible reactor concept so I believe that the future for accelerators is very bright and I thank you for your attention and we'll now stop sharing screen thank you Thomas for sharing with us so as before tap the questions in the chat and we'll take them at the end I see that you all are putting questions and monitoring the chat and we'll make sure to go through those at the end so now we go to the last presentation talk of today please welcome Professor Ian Chapman from the University Authority you're on mute I shared my slides too early and couldn't work out how to unmute so good morning, evening, afternoon wherever it is for you in the world I'm going to give you a brief rundown of a few results from the first campaign of massed upgrade and then talked a little bit about the implications for the massed upgrade Ian, sorry, you went to mute accidentally I don't know how, thank you I have no idea how I achieve that this is WebEx, I'm not used to it I thought I'd start with the UK government's position on fusion they published a fusion strategy last year for the first time ever they had two goals to demonstrate the commercial viability of fusion by building a prototype in the UK and to establish an industry base which can actually build power plants so we obviously as the government national lab in fusion are supporting that strategy as we go towards fusion power plants I think it's worth stepping back which ultimately will be whether fusion gets adopted by the market or not, can it compete in the market the levelised cost of electricity has a very strong exponent on the availability so how much of the year you run, this really matters somebody once said to me about Jet Jet is a maintenance facility which occasionally operates rather than the other way around and as people in the fusion community operate our machines most of the time rather than upgrading them and maintaining them most of the time the other things which we care about as plasma physicists, as fusion scientists are things like the thermal performance which we've just heard lots about from Thomas but actually the biggest drivers to the cost of electricity, to the consumer are the cost capital so how much does it cost to build the plant overnight and then how much does it cost to finance that build if you look at where that money goes about a third of it is what I would call fusion capital so things which are specific to fusion magnets, cryogenics about a third of it is conventional capital and then our operations costs we need hundreds of people, maybe thousands of people to operate power plants and so they're significant the replacement costs of replacing the blanket and disposable items like that are again significant but really you're dominated by this two-thirds overnight costs in the build so what costs money when you're building power plants if we look at what we've learned from ETA and also what we would predict for power plants from systems code studies you can see it's dominated by the magnets and the buildings and to be honest this is true whether it's a stellarator, a spherical tokamaka, conventional aspect ratio machine in all cases this is true so you see a very large fraction of the cost is driven by the magnets and the buildings that you put the magnets in so this is really what drives the cost to the consumer and so we have to be cost conscious about how we strip out cost and that essentially is the genesis of the spherical tokamaka people were saying well can we use the magnetic field more efficiently can we drive down the scale can we put it in a smaller building, a smaller bioshield and there will strip out costs from these very two these two large substantial items so the one of the many challenges around spherical tokamaks is how you might exhaust the heat so the boundary condition of course for fusion doesn't change you still need the fuel to be hot enough to fuse if you put that fuel into a smaller volume in the spherical tokamaka then the chances of damaging the wall are obviously much larger so you need a clever way of extracting the heat from a compact design and that is the primary reason that we build this machine last upgrade is to test a novel way of exhausting the heat from compact designs I'll tell you more about that in a moment we're also using this machine to do some experiments which are of relevance to the conventional line and relevant to ETA which is the basis for future spherical tokamaks and I'll talk about that as well so what's going on at the heart of the device we obviously have the core of the plasma is fusing in the middle and the particles are moving around the device but the confinement is not perfect so some escape and a swept down using the colloidal field coils to the diverter at the bottom and this is a conventional diverter configuration what we've done in the upgrade is to include an additional seven colloidal field coils in the diverter regions top and bottom which expands the diverter leg out in radial space this is about 50 centimetres but in toroidal space the connection length increases by about 20 metres and the whole time the particles move in that diverter chamber as I say mirror to top and bottom of course they radiate energy and so that when they impinge upon the wall and reach the plasma fusing components the hope is that they have reduced in energy sufficiently that we could conceive of a way of exhausting power when we go to gigawatt level plants and how do I go to the next slide the other benefit of mass upgrade is that we have huge flexibility by the introduction of all of these coils there are 19 colloidal field coils in the device that allows us to study a real wide range of configurations going from a conventional double null to an X-diverter in double null, a super X-diverter so extending this leg around to the new diverter surface here and flaying the leg we can look at a disconnected double null with an extended leg on the outboard side different configurations on the inboard side as well looking at X-diverter and even looking at snowflakes and we can do balanced up and down snowflakes as well so it's a hugely flexible device which allows us to look at all sorts of diverter configurations the main ways that this works you sort of saw this already in the video is that the power is deposited over a much larger area so you have a combination of increasing the radius for the target and broadening the scrape-off layer width and the second primary way that we reduce the heat which is impingent on the wall is by just increasing the field line length so the connection between the outboard mid-plane and the diverter targets as I say goes up about a factor of three in mast upgrade compared to mast so these are really significant changes it's not a factor of 10%, it's a factor of 200% and over that sort of huge increase which really plays to the benefit of the spherical geometry we hope that you can get a very large dissipation of the heat before through radiation and neutral particles before they reach the diverter so this is what we were attempting to achieve we made this prediction before we built the device it was nearly 10 years ago comparing a conventional diverter as will be used in IETA to the superax diverter with a long leg and all of those dissipated losses that I just talked about this shows you the peak heat flux on the diverter plate so looking across the diverter plate in a narrow channel and this is just normalised compared to the superax where we were expecting at least 10 times reduction in the peak heat flux incident on the wall in fact nearly 20 times reduction and these were the first results that we got within a month actually of starting operation with long legs and you can see we really do see a significant reduction in the heat impingent on the wall using this long leg diverter design which gives us a lot of confidence that we can move forward with this as a basis for spherical document power plants using the benefit of this long leg the other things I said that we would be attempting to do in mast upgrade is to further our knowledge basis for IETA the key to that of course is that we must have plasmas which last for much longer than they did in mast upgrade and get to steady state so here you can see now we're well over one second long discharges almost kiesent stored energy and the green wall fraction beginning to saturate at a high level in this case so we're getting to stationary conditions and in the next campaign we'll be hoping to extend those pulse lengths significantly further in principle we should be able to go to five second pulses in due course as we continue to upgrade the machine we are looking at how the pedestal confinement and L behavior changes you can see in the plot on the right that now that we have this improvement in the closure of the diverter we get a much hotter temperature pedestal so already in mast upgrade even with relatively low power we now get hotter temperature pedestals than we did in all of the different scenarios that we looked at in mast so you can see here a 400 EV pedestal and we can also using our in vessel coils look at mitigation of elms so this is very early shot but you see the application of the RMP coil currents is causing a change in the behavior so that's something that we'll be looking at in great more detail with our range of diagnostics in the next campaign we're also looking at how we might translate this to the future for spherical talk about power plants extending significantly the breadth of parameters which are achievable in spherical talk max particularly looking at the impact of stronger shaping and higher beta as we would want in power plants on confinement so this is a big issue for us and we will be more than doubling the heating power over the next two or three years we currently have 5 megawatts of NBI power we are currently installing another 5 megawatts of NBI power taking that to 10 and 2 megawatts of EC power as well on top so we're installing two gyrotrons at the moment both of which should be available in three or four years time we're also looking at access to density limits and how far we can push the density we would like to be operating at a high greenwalled fraction in a power plant so how far can we push and you can see the change from good confinement to degrading H mode and ultimately transitioning back to L mode as we continue to ramp the density so we get to sort of 60, between 60 and 80 percent greenwalled and we're trying to understand more about what sets those limits in spherical tokamaks and also we care significantly about the alpha and ikon modes of course at lower field we activate alpha and ikon modes significantly and that causes redistribution of particles and current in the plasma and ultimately fusion power this is a big deal in spherical tokamaks and we need a way to mitigate that we have moved one of our beams to being an off axis neutral beam and the change in the distribution of the fast ions born from the beam definitely we can already see that we can mitigate some of the on axis alpha and ikon mode activity by the introduction of this off axis distribution and in the upgraded heating power we will have an intermediate beam position between the on axis and off axis which again will allow us to tailor the distribution function of the particles with greater flexibility which will really allow us to optimise this behaviour all of those capabilities are adding into our work to think about what follows that and the potential for spherical tokamak based power plants as I said in the thesis right at the start the aim of the spherical tokamak is to drive down the capital cost the overnight costs of the plant which we hope would then result in a lower cost of electricity to the consumer there are of course many challenges from having a compact device one of which is the heat exhaust although the early results from massed upgrade give us confidence that we may be able to manage the heat exhaust in a power plant but the compact nature of the centre column as well and the access for various services cause other problems which we will need to find solutions to as part of finding solutions to them we are embarking on a prototype power plant design study for a machine called STEP the UK government have invested about 300 million pounds into this design work this is now a national endeavour it's not just UKAA we have about 300 companies working with us on this detailed design the internal team is about 300 people now and continues to grow so this is a serious endeavour where we are doing a proper sort of concept prospecting around the spherical tokamak possible space now and the early results show that we are really reducing size if I compare the emergent design for STEP with a radial build of about 8 metres to the European demo design as one of very many conventional aspect ratio tokamak designs where you're up somewhere between 14 to 20 metres those designs the European demo is nearly 17 metres radial build and all of that build costs money so you have very large toroidal field magnets which require substantial cryogenic cooling a very large cryo plant and lots of strand and big vessel segments all of which increase the cost so the beauty of the spherical tokamak is that you can take out a lot of that cost and as you can see the radial build is less than half the size we are very much in concept design phase at the moment we will complete that concept design by March 2024 and we are on track for that we will then move into a detailed engineering design phase that will not be quick and we are expecting that detailed engineering design phase to be best part of 10 years overlapping with some of the early long lead item procurement and the early infrastructure on site then the main construction happening over the 2030s aiming to complete the build by 2040 before moving into commissioning this of course is a very aggressive timeline but it is always good to have ambitious aims as a sign of that the UK government are finding a site upon which step will be built and we will begin mobilising before the end of this year on that site and start clearing the site we had 15 nominations which really span from the top of the country right to the bottom and east to west we down selected to these five we made a recommendation to our department in the government two months ago and we expect an announcement towards the end of this year where one of these sites will be chosen as the site for step and as I say before the end of the year we will already begin mobilising on that site we are also thinking about the future target operating model UKAA as you know is a research organisation we are not well disposed to delivering very large programmes of this nature as well as us providing the fusion expertise we will be partnering with a large technology and engineering partner and a facility construction partner to create a new organisation which in the first instance will be a subsidiary of UKAA a company limited by shares with two new partners and a very wide industrial base using a large number of SMEs in delivering discrete work packages but alongside two very large engineering primes this we hope to stand up within the next two years so by 2024 and finally a word on regulation as we think about how power plants can be enacted in the country in the future we also need a proportionate regulatory framework the UK government have been doing quite a lot of thinking about how fusion should be regulated in the UK initially sparked by a body we have called the regulatory framework to look at how novel and disruptive technologies should be regulated so for instance at the moment they are doing a study of drones and how drones should be regulated or genomics they did one on fusion and suggested that we should take a proportionate pro innovation approach which is different to conventional fission regulation the government then ran a consultation process on that and following the consultation process that they will be creating a new regulatory environment bespoke for fusion so not using the office of nuclear regulation in the UK but actually using the environment agency so treating fusion very differently to fission and having a bespoke tailored framework rather than trying to use what already exists for fission and I think this is a very important step that's been taken so I'll close there and say we have a new growth we published our first ever fusion strategy including as I've just said how fusion will be regulated power plants will be regulated the first results from master and I must stress this was a short first campaign I've already shown some significant reduction in heat flux using alternative diverter configurations which are very promising the next campaign is about to start actually so it will start over the summer and we already have a series of upgrades planned so nearly troubling the heating power and cryogenically cooling the diverter and various other upgrades and in parallel we're progressing well with the concept design for what a power plant might look like we hope to complete that in about two years time before we move into a detailed engineering phase with some larger engineering partners okay I'll finish there thank you thank you very much Ian and we are ready to move to the Q&A and panel discussion I wait for Zhen Zeng to turn on his camera if possible great okay so Ian there is one question for you already but we wait a few minutes you get a few more so I'll work my way to you starting from east again so we go to Zhen Zeng and the question you had on the chat I hope you had all the time to look at him and maybe you went through those I'll go chronologically so I'll start from the first one that I saw there was a question on what is the limiting factor to make passes longer than 1,000 seconds in east so this for this 1,000 second discharge in soft land discharge in a well we could say we can extend the pass for longer and the plasma control might be a problem due to the integration non-linear drift so the drift problem is also one limit before pass to extend the plasma division so so far we tested the 1,000 seconds for integration it shows a linear drift it's a problem at the beginning but I have to say in my talk we find the problem and people and this linear drift was very low for long duration so so according to our experiments this non-linear drift will be affected by magnetic measurement and causing problems for real time reconstruction and plasma control so and even in the future if we send the plasma division I think this drift problem maybe also needed to make more some more jobs to dedicate the drift thank you there was a question if you could try to explain the difference between the conventional i-mode and the super i-mode yeah this super i-mode it means general i-mode and plus the electron IDP so this mode this discharge we can find that near the center gradient is very high and this means it's a very good electron internal transport barrier with high temperature and plus the age turbulence barrier as normally for the biannual i-mode just the age temperature barrier so this we think is super i-mode due to the center electron IDP okay thank you I think you partially answered this question but on what is the ultimately limiting what is you ultimately limiting the fully non-inductive discharge duration okay in effect our this is the thousand second i-mode has a lower power injection has the heat heat load and recycling or current travel was not a key however in separate experiments we run high beta P with a higher power injection and this normally is quite a little difficult to get for non-inductive discharge mostly due to the recycling and the heat load so in the future for the next ago we increase the heating power and maybe send the lung plasma to charge duration and this this problem will be also will be met and should be resolved okay thank you very much I'll come back to you I'll go now to Thomas Thomas there was a question on what kind of active control of the magnetic field is required for W7X to maintain the plasma position shape right so given that the plasma flux surface that the flux surfaces themselves are created from the coils you don't need vertical control or control of that sort it really is a very stiff cage for the plasma now it's advantageous as we do have we've minimized the boot strap current but there is a little bit of it left it is potentially advantageous for diverter operation to change currents in we have a planar coil set where we can slowly over time tens of seconds change the IOTA the the poloidal versus tooidal components of the magnetic field to adjust slightly for what's created from the boot strap current but fundamentally there isn't need in a stellar for active control of the plasma shape the only thing that does change a little bit is the edge topology and that can be then controlled with you can also do some current drive or changing coil current slowly the amount that you need is really minimal thank you so what is the nature of turbulent transport in a stellar and how does this scale to a bigger device so turbulence is a very big topic lots of answers to that we believe that the iron temperature gradient turbulence will be dominant in a reactor and it is also dominant in W7X so how it scales to a reactor very good if we go by the empirical results we have we can use something like the ISSO4 scaling which is somewhat equivalent to the the ITER 98 scaling which is basically a scaling for turbulent transport confinement due to due to the fact that turbulence dominates the transport so how these scale I think it would be good to have more data points on that to have a higher field the stellarator would be built and I should also say that the stellarator because of the optimization and the many degrees of freedom in the optimization for over 10 years there are ideas out for how to optimize the stellarator to reduce turbulent transport so next generation stellarators will have that feature in them that optimization includes lower turbulent transport next what are the W7X plans for deuterium operation and what are the expectations on the isotope effect in optimized stellarators so that's a that's a great question so W7X was laid out for deuterium operation it's been conceived from the beginning as being able to go to deuterium we don't have the final allowance for deuterium operation yet this will come within let's say the next five years or so hopefully and we will then be able to address the isotope effect what I can say about that is that LHD has operated with deuterium and W7AS operated with deuterium and they find that the confinement is essentially the same in LHD between hydrogen and deuterium which you can argue is a little bit of an isotope effect because the gyro radii are different or you could argue it's not because the turbulence appears to be the same whether you have deuterium or hydrogen so the jury is a little bit out on whether there is going to be an isotope effect I think it has a lot to do with things like zonal flows and how the turbulence saturates and this might well lead to the result that you don't get an isotope effect in the sense of what I just said for LHD that you will get roughly the same confinement with deuterium as you do with hydrogen but W7X will be able to address that Thank you Thomas I'm moved to Ian but I'll come back to you too so Ian two questions there was a comment first I've noticed a significant smaller breathing blanket in step so the question is can you estimate the tritium breathing difference between EU demo versus step and then there was a question about is tritium self-sufficiency a requirement for step Yes so we're working on our design basis that we need a tritium breathing ratio of 1.2 because anything less than that doesn't give us enough margin frankly because that's pretty optimistic anyway so we're taking that as a design basis one of the benefits of the spherical design is that because the centre column is so narrow you don't need to breathe on the inboard side so you don't have to worry about getting coolants and things down the inboard side which is quite hard so we can do it all on the outboard side yet our preliminary designs for the blankets are of order a metre thick they have to breathe all the tritium that we need for the machine of course the volume of the plasma is much smaller than it is in a conventional aspect ratio so in absolute terms but yes this is something we're very cognisant of and working hard on exactly what that blanket might look like and what multipliers will need what level of enrichment of lithium-6 we need all the same challenges in the conventional aspect ratio to be honest thank you and a question on because step will be smaller in size compared to the other demo type device being planned do you see this as a limitation in terms of electricity production yes and no so step as a comparison as a prototype for step we're designing a bit over a gigawatt thermal which we expect to produce we're aiming for confidently a hundred megawatt electric but probably a couple of hundred megawatt electric if you compare that with demo demo is a bit over two gigawatt thermal aiming for 500 megawatt electric so it's twice the electric output but it is a pulse machine whereas a spherical document would have to operate as a continuous machine we just don't have the solenoid swing and so the megawatt hours will be pretty comparable in the two to be honest thank you so there was a question which was with Thomas Ian but I think it can be posed to all of you so we'll start from Zhen Zheng and then Thomas and Ian you can all answer and the question was about these different configurations which were presented and so someone in the audience was asking how important do you think is it for the world fusion program to have this diversity of ideas being pursued at the same time okay so I have not finally this question which one so the question is how important is to have the diverse so many diverse configurations like you know the one which are featured today to be pursued so research and developed at the same time how important is this for the fusion community and then future fusion industry so I look at this question so yes do you mean the 1002nd charge? No I think so I think the question is more about the configuration of the device as a tokamak spherical tokamak conventional conventional tokamak accelerator how important is that the community keeps on research and developing different confinement configuration lines Thomas do you want to go? Sure sure I can say something to that I think it's very important that we have more options open obviously I think the stellarator has lots of answers to the challenges and that's why I'm working on stellarators and Ian has another approach that will allow the power plant to be more compact will be easier to evolve the configurations because it's not as much capital cost has some other other risks than the stellarator so I think it's very important that we pursue more than one concept eventually we might end up saying there is only one solution that will out-compete the others but it could also well be that the market will want big power plants for certain markets and smaller more modular ones for others and a stellarator is unlikely to be a very compact reactor but it's going to be a very reliable reactor and it's something that you can scale up if you want and so it has features and it's very has high efficiency as was also pointed out by Ian in the chat so it will have its features and it will be competitive in the market but that doesn't mean that the other concepts wouldn't have a chance and I think this is it is important that we continue to evolve we need fusion urgently and we shouldn't just have one shot on goal we should have several parallel activities and they will all they will also all to some degree benefit from each other lots and lots of the fundamental engineering and plant components are going to be similar between the different the different concepts so we can also benefit from each other I would I would completely echo that as well the energy market is huge and it's not that there's one design of gas plant and one design of photovoltaics or one design of wind turbines there are many in all different bits of the energy market and I fully expect that there will be a fusion breaks into it many designs of fusion plants that should be nurtured at the same time I just point out I think we really should do more to reach in as the public sector and we're all public sector people we should do more to reach into the private sector and try and support them because the private sector is able to have a very different approach to risk than we are they have to have a much higher risk appetite than we do with taxpayers money and that can be a good thing and we should do a very good thing Okay Chen Zeng if you want you can compliment or you want to add anything or we can I have a couple of questions for you you can also try to answer there was a question on why do we need why do you need low recycling conditions in east so first I would like to define the low recycling in our case the low recycling means recycling should be than one. So this means normally the recycling is higher than one. This means the plasma density is difficult to be controlled. So in our case, the recycling is about 0.9. It's not so low, but it's currently lower than one. So this is a low recycling power set. So for the other reasons, during our limited energy, low-hybrid current driver power, so this also need a stable density and also make a good for non-inductive current drive. Okay. Thank you very much. And that was a question I think here is more asking if you could elaborate on what is keeping you from running essentially always 1000 seconds long pulses experiments. So this 1000 second plasma discharge was obtained from step-by-step. We tried a lot of attempts before. So we found 10 seconds, 100, 2000, 1000. Okay. This is also with a solution of several issues from plasma control, coupling, and so on. Also included are getting flux removed. So in the experiments, several issues sort of gradually. So and also for the control system, we remove the linear zero drift by PFCs. And also plasma loop voltage is well controlled by the law about power. So of course, this is for this discharge, this is 1000 second discharge with a lower power injection. And in the future, and some other, as of course, for example, to into more power and so on. And this is something should be also needed to be resolved. Anyway, we come to this 1000 second charge should be repeatable at the moment. Okay. Thank you. Thomas, there are lots of questions for you and the accelerator. There's a, I think this is a very quick one. Tritium is not part of W7X operation, correct? That is correct. Yes. Okay. Move along. So how do you choose the right optimization for the standard editor to make it a power plant? How do you go? Yeah, that's a great question. There are many different flavors of optimization. And even within an optimization, you can you can optimize in different directions within that. So in the end, it's going to come come down to which which one will give you the most economic power plant. So we need to build into our optimizations a real cost function, if you will, so that we have a basis on which we can we can choose which is the better concept. Now, they also have different risks associated with them, that various stellarator flavors. And one needs to quantify that. And as Ian was saying, it's also a question how much risk is is acceptable to those who fund the activity. So, you know, I think that the that the quasi homogenous configurations that W7X line has a lot going for it. I think quasi helical has a lot going for it. And quasi exosymmetry has some advantages. Now, in the end, one one will either need to evolve more than one stellarator configuration, just like we have the the various aspect ratios of the token max. And or you say there is one that we will pursue because it has the best outlook right now. I think it's actually an advantage that there are so many possibilities for optimization in the stellarator space that we can find solutions for the challenges and have in the end we need an integrated solution for for all of those problems. And the fact that you have many different possibilities for optimizing the actual magnetic cage for the plasma is an advantage. And if you find one that's good enough, that will get us going. And maybe next generation will be a different flavor and be even better. What's important is that we we take this step as soon as we can and move towards net energy production. And not not to say yes, we found the very, very best stellar optimized stellarator. It just needs to be so good that we can produce net electricity with it in an economic fashion. Thank you, Thomas. Ian, there was a there's a question for you. It just came on the chat about compact fusion devices attractive, but neutron damage of the plasma facing material seems to be more severe. What about lifetime of plasma facing components? And if the lifetime of the material is shorter as compared to, I guess, EU demo? Yeah, it's a very good question. And yes, the answer is it is it is a it's a complication of the spherical tokamak. In the past, that has been seen as an inhibitor, actually, to a power plant based on a compact geometry. I would say that our philosophy is to think again about what are lifetime components and what are also what I'm looking for disposable components. So if you if you set up your geometry in such a way that you optimize the design for maintenance and removal of components, then the fact that they might last three or four years doesn't actually matter, because you can you can get them out and replace them quickly. When we built mast, we built it as a cassette, essentially, so I had a bottom lid, then an outer sleeve, we craned in the lower diverter, we craned in the upper diverter, we put down the center column and put the lid on. So it's not built as segments of an orange as you have to with a conventional aspect ratio or a stellarator. So the fact that you can turn it into a cylindrical geometry makes the maintenance much easier. And you don't have big lifts going through complicated ports, as you would do in either of the other two geometries. Indeed, that's one of the big big hurdles with the conventional European demo design is how do you get blank? It's very difficult to do. If you can make that cylindrical, then it's much easier. And then the fact that your components, your placervacing components for sure, they will get damaged and they won't last for 30 years. But if you replace them every three years or five years, it doesn't matter because the maintenance is easier. So it's a very different philosophical approach, not saying the materials have to last for 40 years. They don't have to. Thank you. And Thomas, since we're on this subject, there was a very long question for you. Let me try to go through it. So the question was, what would be the heat loads? And I did the diverter first wall for the one gigabyte electricity accelerator power plant, which you showed. And so how did heat loads compare with a tokamak compact machine like ARC? And then the question was, what are the current material considerations? Okay, that's a great question. And it's right on the money because in fact, such a compact, powerful device would be right at the limits of what you can. In fact, I think its output would be limited by what the walls can handle. Now, in the stellarator, we have shown in W7X that we can radiate away all of the energy at the edge or essentially all of it. And this allows us to speculate that we could do that in a reactor. And if you can do that, you would actually be distributing the plasma heat onto your plasma-facing components rather uniformly. And you would end up down at one to two megawatt per square meter of loading. Now, anywhere where that gets concentrated, you may want to need to concentrate this for efficient particle exhaust. That will then get higher than that. But indeed, the 1000 megawatt electric is one that is going to be, we might need to dial that down to simply because of power exhaust issues. And to answer the question of arc, I'm not an expert on arc, but this vision that I showed is in many ways inspired by arc. And the same, I think it will be roughly the same situation in arc that they will need a radiating mantle to get the heat distributed uniformly. And if it really gets concentrated, one will have to operate it with a lower output. And you were asking for materials. I think the main material we consider is tungsten. We also have plans to upgrade W7X eventually with a tungsten diverter and get rid of carbon in the machine entirely. So tungsten is a plasma-facing component, but there are other possibilities. There are people who think about liquid metals, a thin layer of liquid metals. There are possibilities that can be explored. But I think tungsten is the most qualified material for a first wall at the moment. Is that the case for step two, Ian? Yeah. Jinxing, there's a question. I think you mentioned what's coming up for you, so maybe can you talk about CFETR? CFETR, yeah. So at this moment, it's just the finish that you saw and maybe and although I have not yet pulled back government, so this CFETR is what I say is also we want to get the cure from about 10 to 30. And so it's a big machine and I'll say. So at the moment, we are pushing this project to be built. Okay, thank you. Thomas, I think there's a question about how the geometry complexity of the accelerator and how much would this complexity affect the use design of diagnostic system? So I think I should start by saying that if you look to the reactor scale, the stellarator being let's say passively stable in almost every regard will require less diagnostic access than the more driven feedback stabilized concepts. And now it's possible, we found on W7X, we have more than 50 diagnostics on the machine. There are ways you can optimize this stellarator coil solution, which actually has many degrees of freedom, so that you can open up space for ports and access for diagnostics. For diagnostics. So it's actually quite compatible with diagnostics. Now, diagnostics in general for a fusion power plant is a challenge all on its own, but I really don't think it's worse for the stellarator. In some sense, it could well be better because you don't need as much active control. There was this question about you showed us all these great advantages of the stellarator line or the talk mark. So the question was, do you believe that the results produced by W7X will bring more attention to stellarators line in the future? And I would add, if you could talk about the Simon's collaboration on hidden symmetries for fusion energy. Yes, it certainly deserves more attention in my opinion. It should be thought of as domain or a main approach to fusion. I'm really very, very optimistic about stellarators, I have to say. Now the Simons Foundation has allowed us to make an international collaboration that has been spending over four years and has just been extended another three years. That collaboration has brought together people from different institutions and from different skill sets working together on making better optimization codes with higher numerical efficiency and more fidelity as well. And this has brought a lot of these results I mentioned at the end of my talk. For example, bringing the optimization of the neoclassical transport down to the level where you can even argue below that of a Tokamak. And it's really had a revolutionary nature in what we thought earlier possible for stellarators. There were things that we thought like the stellarators are always going to have more alpha losses than the Tokamak because of the inevitable helical ripple that doesn't look like that's the case anymore. And a bunch of other things that, for example, it was also said that the optimization used on W7X is only really possible at high fidelity if it's a large aspect ratio. We have new designs, they're not completely mature, they haven't been optimized completely for like MHD stability, but there are new designs that are much more compact and have very good particle confinement, very good neoclassical optimization. So that endeavor has really been key in making enormous progress and you can find papers from the collaboration in the peer-reviewed literature. It's really amazing the results that are coming out of it. There was a question about lower aspect ratios to stellarators. The question was to what degree we can make lower aspect ratios. Okay, yes. So I had mentioned already that the lower aspect ratio seems much more possible than we thought a few years ago where it seemed that it has to be somewhat large aspect ratio to optimize it properly and that does not seem to be the case anymore. Now you may not want to go to extremely compact stellarators because you will get some of the challenges that Ian's facing with his design, that you just don't have a lot of space for blankets and your heat exhaust might be more difficult. Of course Ian was showing us that there are game plans for all of this and so maybe we can pick up some of those. But it's not clear to me that at very low aspect, there will be some optimum aspect ratio and it is going to be a lower aspect ratio than W7X. Maybe it's five, maybe it's four because you simply will be able to produce the next step device or a power plant at lower capital cost. The same arguments that Ian was bringing forward also are true for the stellarator. But to have it so compact as masked etc, that remains to be seen whether that is really the optimum for the stellarator. I kind of doubt it. I think the aspect ratio five, six is probably the sweet spot but it remains to be seen. This just now like a follow-up question from Lorenzo Boccicini, if you're in this optimization process you're also taking into account technological requirements such as sufficient place for a building blanket, solution for the vacuum vessel or the vessel components replacement. So this is a vision I should say and it does not have a whole lot of details behind it. It has though been made with the thought that the arc design solutions can be transferred to a stellarator. That the thinner blankets that they're going to be using will be used also in the stellarator. So in terms of vacuum vessel access, stellarators are built, W7X in particular, built out of identical periods which consist of mirror image half periods and so you can imagine a maintenance scheme where you could swap out a half period for maintenance and pulling another one. There are other concepts that don't require you to have that capability and they still need to be evolved and fleshed out to allow for access. Those are all things that need to be developed before we can say something but there are good ideas and I think we will find solutions. Okay, that's encouraging. Again there is a question about tritium availability but I heard you said you're about to start the concept of design phase for step. Is it something you already have a tritium field cycle task force or are you looking into this? About tritium availability, as I said before, as a design constraint we must provide enough tritium for our own purposes so we're aiming for a TBR of 1.2. I'm not very, I'm not especially worried about startup charge, I mean there were sort of stuff in the press recently that was largely right but not wholly right. I would point you to, I mean there are a number of scientific journals on the availability of the tritium stock in the world. We published one, Michael Kovari from UK, published one, it was best part of a decade ago now but it's a really good article which explains how much tritium there is and the decay times and how it would be used on potential projects and what sort of startup charges will be required. So it's not something that, look we have to pay attention to it as a community but it doesn't worry me overtly but we do have to get to the point where the prototypes breed enough. We can't just all be consuming and not producing right so breeding is an important thing and we've got to design prototypes to breed. Thank you Ian. So I will take just one last question because it was about AI for fusion and I should mention that we just launched a couple of weeks ago a new coordinated research projects on AI for fusion, I'm putting the link on the chat and it's open for applications and so the question was, it could be for all of you, what role at the moment has AI if you're applying these methods in your machines and we can start from Ian, Thomas and then Zhenzhen. Sure so yes is the short answer and multiple different ways so as we get to systems code design and one of the challenges of fusion is that to have a high fidelity systems code which includes all the multi-physics which is required for fusion is just hugely computationally expensive so an alternative approach is to go to a reduced fidelity emulator but then put a machine learning wrapper over that so that you can do lots of simulations and not require exascale machines and have a machine learning wrapper over the top. So we're looking into that and pursuing that as a potential way of being able to produce holistic systems code design. Then in more niche areas like disruption control or something like that you could well imagine applications for artificial intelligence or our robotics control similarly so we do quite a lot of AI for application in robotics and in control systems. Thank you Ian Thomas. Yes so I think actually AI can play a very substantial role for stellarators in certainly also for operation we have first results showing that the infrared images can be used to extract information about the magnetic topology with AI but I think much more importantly we will be able to as Ian was also saying we'll be able to proxy complicated calculations and speed up and that is useful for the optimization and will allow us to bring in things like turbulence optimization. The turbulence codes of today are really too slow but using either proxies from analytic theory or proxies with AI is also under development and this will allow us to have things so fast. There are only proxies perhaps but they can be quite accurate ones enough to guide an optimization. This we also have the examples of MHD equilibrium calculations that are being proxied now with AI so it just has an enormous potential for speed up. Thank you. Shenzhen. Okay so I'll agree in 12 of you also so I think the AI is very important for the tokumak simulation and the prediction and also the trainings for example the disruption there are a lot of dates and no way we use AI maybe it's easy to know the interaction between the turbulence and so on and a lot of physical process so I think for the plasma good control and to use AI to to make a very very very process modeling prediction and also use for the perception of control and feedback control and so on is very important yes. Great thank you very much. Thank you all for joining us today and special thanks to the speakers. This closes this session and I should mention that we're already planning session three and of course we'll keep on having a diversity of configurations as as already done in these two episodes. This third session will try to be we'll try we'll try to have it sometime in September, October. If you want to present results from your facility send me an email and we'll try to try to accommodate your request so thank you again and see you next time. Okay thank you.