WEBVTT 00:00:00.000 --> 00:00:03.000 Hello, Richard. 00:00:03.000 --> 00:00:08.000 Let me share my screen. Can everyone see that? 00:00:08.000 --> 00:00:09.000 Great. Yeah, that's much better. Thank you. 00:00:09.000 --> 00:00:21.000 And in full screen? Great. Yeah, so good afternoon, everyone. Thank you for the organizers for the opportunity to speak today. 00:00:21.000 --> 00:00:35.000 A few talks ago, Benno very eloquently outlined some of the thoughts that have gone into the sustainability efforts for other linear collider schemes that are far more advanced than what I'm going to talk about today, such as the ILC click. 00:00:35.000 --> 00:00:47.000 And what I am going to talk about today is a new concept that's only really been around for the last couple of years which implements plasma accelerator technology. 00:00:47.000 --> 00:00:53.000 In the hope that it can make a linear collider small and cheaper but also potentially greener. 00:00:53.000 --> 00:01:00.000 And that's hopefully what I'm going to focus on today and convince you that this could be the case. 00:01:00.000 --> 00:01:12.000 Okay, so a brief bit of background on particle physics and how plasmas could fit in. So we really are at a critical time for particle physics. We need to plan for the post LHC era. 00:01:12.000 --> 00:01:24.000 Fcc EE would be fantastic. I'm sure that that will be the outcome of the ESPPU process that everyone's very excited about, but it is extremely expensive. 00:01:24.000 --> 00:01:29.000 Linear colliders such as ILC and Qlik promised a reduced cost. 00:01:29.000 --> 00:01:35.000 At least for some part of the physics space person but they're still very expensive. 00:01:35.000 --> 00:01:44.000 And the key cost driver in many colitis is the accelerated gradient, which is limited to something like 100 megabolts per meter, often operating much lower than this. 00:01:44.000 --> 00:01:53.000 However, plasma acceleration promises much higher gradients. So something like potentially three orders of magnitude higher, but at least one order of magnitude higher. 00:01:53.000 --> 00:02:06.000 But when you're thinking about just the gradient, this sounds absolutely brilliant. You could reduce the cost of the facility by a factor of 1,000 as well, but that's actually not the only cost driver And importantly, as I'll explain, it's maybe not even the most important one. 00:02:06.000 --> 00:02:12.000 So you have to consider what are the cost drivers define the design choices of plasma accelerators. 00:02:12.000 --> 00:02:18.000 And I should point out that half is not new in terms of its idea. 00:02:18.000 --> 00:02:26.000 There have been several plasma-based collider designs proposed for the last 30 years And here's an example of one here on the bottom right. 00:02:26.000 --> 00:02:30.000 And typically they represent the state of the field at the time. 00:02:30.000 --> 00:02:37.000 So they're useful for identifying the remaining challenges, for focusing down where the field needs to go. 00:02:37.000 --> 00:02:41.000 For example where we are conceptually and where we need to be experimentally. 00:02:41.000 --> 00:02:51.000 So plasma accelerators is a developing technology. It's certainly not ready for applications of particle physics, certainly not high energy physics. 00:02:51.000 --> 00:02:57.000 But we're not far away. A lot of important results have come out in the last decade or so. 00:02:57.000 --> 00:03:23.000 We are maturing towards real life applications. For example, Strongfield QED, Some experiments were performed in 2018, which showed that you can actually perform this strong field QAD experiments by colliding a laser beam with an accelerated electron beam, which is accelerated in a laser weight of accelerator. But also free electron lasing, which has a lot of the same problems and challenges as a collider, but just on a smaller scale. 00:03:23.000 --> 00:03:31.000 So if we can demonstrate free electron laser, then it really is a motivator for the field. And these are some results that came out only a couple of years ago most published in Nature. 00:03:31.000 --> 00:03:40.000 And there are six key aspects of research that are relevant to colliders, and they're all tied up loosely in the equation for luminosity. 00:03:40.000 --> 00:03:59.000 So we need high energy transfer efficiency, high repetition rate and average power. We need to preserve the high beam quality for the interaction point, for focusing And also need to stage the very high energy. And the whole thing needs to be very stable. But there's also one outstanding challenge in our field, which is highlighted here in red, which I'll talk about in a bit. 00:03:59.000 --> 00:04:20.000 Positron acceleration. But what I wanted to just briefly touch on first was energy transfer efficiency because it has some relevance to sustainability So energy transfer efficiency in the plasma accelerators can in principle be extremely high it can be something like 80% from main beam to accelerate tube beam. 00:04:20.000 --> 00:04:26.000 And this is due to the fact that we can heavily beam load the electric fields in a plasma accelerator. 00:04:26.000 --> 00:04:31.000 And we can get away with this, whereas you can't really in more traditional radio frequency cavities. 00:04:31.000 --> 00:04:35.000 Because the quality factor of these plasma wakes is effectively one. 00:04:35.000 --> 00:04:45.000 We preserve this plasma, we generate a single accelerated cavity, and it very quickly damps So we have to do all our acceleration in this one cavity. 00:04:45.000 --> 00:04:50.000 But that means that we can essentially do whatever we want to this cavity. We can perturb it in whatever we want. 00:04:50.000 --> 00:04:57.000 And we can see here that you can heavily beam load this and flatten out the electric fields. And this means that you can get uniform acceleration across the entire bunch. 00:04:57.000 --> 00:05:11.000 And because of this, we can get something like a factor of four increase in energy transfer efficiency over competitive schemes such as click And there has been a lot of research in this direction in the last few years. 00:05:11.000 --> 00:05:24.000 So, for example, if you break this down, you have to transfer energy from the dry beam to the the weight field and then from the Wakefield to the witness being the accelerating speed. And proof of principle results are showing you can get 60% 00:05:24.000 --> 00:05:30.000 From the driver beam to the wake and then you can extract 40% of that energy from the weight to the accelerating beam. 00:05:30.000 --> 00:05:43.000 And that gives a 20% beam to beam transfer, which is comparable with the number that I mentioned From ClickBot is still some far way away from the 80% that's theoretically possible. So there is more progress to be made. 00:05:43.000 --> 00:06:00.000 In this area. As I mentioned earlier, for all the benefits such as energy transfer efficiency and the progress they're in. Positron acceleration lags significantly behind was something like maybe a decade or two behind the progress that's been made for electron acceleration and plasma. 00:06:00.000 --> 00:06:09.000 And the main challenge here is that electron motion Which is essentially equivalent to iron motion for the positrons, but plasma electrons are much lighter. 00:06:09.000 --> 00:06:14.000 I should point out that there have been some experimental results in positrons. 00:06:14.000 --> 00:06:23.000 So for example. This is a result that was published in Nature about a decade ago, and they use the positron beam to drive a weight field. 00:06:23.000 --> 00:06:35.000 And also accelerate part of that positronomy towards the tail. So it has been experimentally demonstrated, but even if it were optimized very rapidly over the next few years. 00:06:35.000 --> 00:06:45.000 It would still be an issue, which is the luminosity for disorders amongst you below radio frequency and electron positron acceleration. 00:06:45.000 --> 00:06:52.000 So if we were to build a collider in this regard, we'd have a problem, which is positron acceleration and plasma. 00:06:52.000 --> 00:07:06.000 So we had a pragmatic idea, which was to accelerate the electrons to very high energy in plasma but then use radio frequency technology to accelerate positrons. 00:07:06.000 --> 00:07:16.000 But if you did this and you had it as a an even center of mass. So symmetric energies then the footprint would be dwarfed by the RF of the positrons. 00:07:16.000 --> 00:07:27.000 So the idea was, can we make it asymmetric? And essentially, if we go up in energy, let's just say arbitrarily four times higher in electronically and four times lower impositons. 00:07:27.000 --> 00:07:32.000 And you can make this whole thing more compact, but you make it less energy efficient. 00:07:32.000 --> 00:07:41.000 So to improve the energy efficiency, you have less charge at high energy in the electrons and more charge in the positron bunch. 00:07:41.000 --> 00:07:56.000 And then it turns out if you can make the admittances different so asymmetric due to asymmetrical the beta functions being therefore different at the at the interaction point, you can then improve tolerances for plasma accelerators. 00:07:56.000 --> 00:08:03.000 And this is how we've essentially progressed from what is effectively the ILC towards HALF. 00:08:03.000 --> 00:08:11.000 So we want to design part, but we want to do it properly and rigorously. And this send us back to the drawing board. How do we actually improve a particle collider? 00:08:11.000 --> 00:08:16.000 And I mentioned it earlier, it goes just beyond the accelerated radius. You have to optimize for the power source. 00:08:16.000 --> 00:08:21.000 The accelerator, the energy uses, and also importantly, put carbon tax on all of this. 00:08:21.000 --> 00:08:28.000 And you need to optimize this whole thing for cost. And I won't go into too much detail. This is just really poor posterity. 00:08:28.000 --> 00:08:37.000 But essentially, you can start looking at things like optimizing the beam power from the source to the beam itself. 00:08:37.000 --> 00:08:44.000 But you have to then also consider what the best type of driver technology is to reduce the cost of power delivered from that source. 00:08:44.000 --> 00:08:56.000 So once you factor in all these things, you have the benefits of plasma, but also the potential downsides of plasma. So you have to find a balance between how this could all look in a cost optimization scheme. 00:08:56.000 --> 00:09:14.000 And this is effectively what we did. We put together a detailed physics and cost model So we put lots of inputs that we could get from more advanced design such as that as in Qlik. And then we optimized all of this with Bayesian-based machine learning. 00:09:14.000 --> 00:09:28.000 And this is essentially the footprint that is spat out So you can see here we have this driver Linux Which is then being combined the bunches and then sent into this plasma accelerator of the neck. 00:09:28.000 --> 00:09:43.000 Which then takes the butchers from something like 4GV up to 380 gv And then we use some of those beams to generate the positron target these positrons are then sent all the way around here, put into a damping ring 00:09:43.000 --> 00:09:54.000 Put through a cool copper Linac and then sent together into this being delivery resistant to this interaction point here to produce 250 gb centers of match. 00:09:54.000 --> 00:09:59.000 And it starts looking very similar in some regards to existing designs. So for example. 00:09:59.000 --> 00:10:13.000 You end up with something like a click-like drive beam. And that makes sense because the plasma is essentially just a transformer. It's transforming low energy, high current beams into high energy, low calorie beans. 00:10:13.000 --> 00:10:21.000 And importantly, we have a plasma linac here which has 48 stages and 10 megawatts of heating. So these are outstanding challenges. I'll touch on these a little bit. 00:10:21.000 --> 00:10:35.000 Later. And we chose semi-conservative energy transfer efficiencies. So 40% beams of beam, so a factor of two higher than what's been shown experimentally, but still a factor of two below the theoretical value. 00:10:35.000 --> 00:10:43.000 And if you're interested in more details, this can be found in the submission that we put forward for the ESPP process recently. 00:10:43.000 --> 00:10:52.000 So importantly, this has a length of five kilometres And a cost of something like $3.4 billion in 2024 money. 00:10:52.000 --> 00:10:58.000 The emissions here are given in terms of construction and run costs. 00:10:58.000 --> 00:11:09.000 And the waterfall power is approximately 106 megawatts. And this is all assuming a 10-year period of operation producing an integrated litrosity to inverse atom. 00:11:09.000 --> 00:11:14.000 But importantly, if you can compare these numbers to other collider designs, such as ILC and Qlik. 00:11:14.000 --> 00:11:31.000 You can see here that the length is approximately 25% of the IRC. The cost is something similar. And the emissions from construction a percentage that's similar. And the same goes for click. If something like 50% the length and cost 00:11:31.000 --> 00:11:50.000 And emissions of click. So something that's interesting that's come out of this is that the emissions, the carbon emissions from construction approximately scale with the length and the cost And this makes sense given recent studies that have shown that a lot of the CO2 that's produced during the build is the concrete, for example. 00:11:50.000 --> 00:12:04.000 That's required for the tunnels. And the wall plug power here, I should say, is very similar to ILC and click, but this doesn't take account for the fact that we could potentially push the energy transfer efficiency even further in the future. 00:12:04.000 --> 00:12:14.000 Okay, so I won't touch too much on this slide. This is really just to highlight what the two major outstanding challenges are between where we are now and building half. 00:12:14.000 --> 00:12:21.000 They are essentially staging of plasma sources. We need to get some very high energy. We need 48 of these plasma stages for half. 00:12:21.000 --> 00:12:38.000 And the current state of the art is two stages. And that's only in a proof of principle way. And if we are going to go to very high luminosity, then we need a very high rep rate. And that requires much more average power in these plasma sources than we've ever had before, orders of magnitude, in fact. 00:12:38.000 --> 00:12:43.000 So we're going to need something that can handle multi keV plasma temperatures. 00:12:43.000 --> 00:12:52.000 And that's just from a single bunch. This could be hundreds of KV temperatures across a full bunch train, which is approximately the temperature of the sun. 00:12:52.000 --> 00:13:00.000 And we need to cool these. So we need much higher cooling rates than what are designed for click. 00:13:00.000 --> 00:13:07.000 Well, because we have this tangible design, it means that we can consider new optimizations, especially for sustainability. 00:13:07.000 --> 00:13:15.000 And one idea that we really just had only a week ago or so sketched out on the back of an envelope, which I'll just very briefly talk about now. 00:13:15.000 --> 00:13:33.000 Is how we essentially use this high temperature to do something? Can it give something back for us which couldn't be possible, for example, with a superconducting RF Linac because it's cold. So if you take the hearth example, you've got 23 megawatts of power going into this drive. 00:13:33.000 --> 00:13:36.000 And only about 10 megawatts make it to the accelerating beam. 00:13:36.000 --> 00:13:48.000 And then of all of that energy, essentially 20 megawatts is lost in the RF. So you could potentially put a water cooling loop In amongst all of this. 00:13:48.000 --> 00:13:54.000 Heat the cool water from call to lukewarm with the RF in Act. 00:13:54.000 --> 00:13:59.000 Then part of the energy in these drive beams is lost in the dumps. 00:13:59.000 --> 00:14:05.000 You could then take that lukewarm and make it warm and then you can put it through the extremely hot plasma cells. 00:14:05.000 --> 00:14:11.000 So on the previous slide, I showed a scheme for a cryochords plasma source, this may be what's required. 00:14:11.000 --> 00:14:30.000 But if we could get away with water, then you could cool with water and then take what's lost in the plasma and turn it into very hot steam. And if you just look back at the envelope calculations, assuming the costs for the ventilation and cooling that we scale from click and also very rough estimates of how much it would cost 00:14:30.000 --> 00:14:35.000 To operate a steam turbine how much they would be efficient. 00:14:35.000 --> 00:14:39.000 You get a very rough idea of how you could go through this scheme. 00:14:39.000 --> 00:14:53.000 And I won't go through the numbers. I'll leave them on the slides online so you can take a look at them in more detail. But potentially you can generate something or say something like five to 15 megawatts of power This saves something like 10 to 40 megawatts upfront in these ventilation and cooling costs. 00:14:53.000 --> 00:14:59.000 And that then saves you something like 5 to $15 million per year in the electricity bill. 00:14:59.000 --> 00:15:07.000 And this is just the heart. If you want to go to a 10 TV collider, which is one of the goals in the US, You could be looking at 100 million per year. 00:15:07.000 --> 00:15:12.000 And this is only possible with hot technology. So to conclude. 00:15:12.000 --> 00:15:19.000 Plasma has the potential to increase sustainability in many ways. The obvious way is the build cost and the operation cost. 00:15:19.000 --> 00:15:36.000 And half represents the most tangible design in this regard. But the tangibility means that we can start thinking about real world problems and producing real world solutions, such as, for example, using the plasma heating to return some energy back to the system. 00:15:36.000 --> 00:15:45.000 And we're likely structured the surface of sustainability benefits of plasma accelerators. Thank you. 00:15:45.000 --> 00:15:51.000 Thank you so much, Richard, for the very interesting talk. And we are ready for any questions. 00:15:51.000 --> 00:15:55.000 Join, please. 00:15:55.000 --> 00:16:05.000 Thank you, Richard. This is actually a fantastic talk. The slides where you had the changes from ILC to health Beautiful. 00:16:05.000 --> 00:16:16.000 Yeah, this one is a very nice one. Can you explain to me again why you want to go for the asymmetric energies to charges to remittances? What does that do in terms of sustainability? 00:16:16.000 --> 00:16:27.000 So from the point of sustainability it doesn't do too much. So it doesn't necessarily save you in power, for example. 00:16:27.000 --> 00:16:36.000 So one of the problems of going from in energy, from symmetry to asymmetric energy is that you actually then make this whole thing less energy efficient. 00:16:36.000 --> 00:16:39.000 And we could have just left it at that. We could have just said, right. 00:16:39.000 --> 00:16:55.000 We'll leave it as it is, same charges, different energies. And we'll just take a hit in the power. But then really the asymmetric charge was to deal with that from a sustainability point of view. We didn't want to come in saying that we don't care about how much this thing is going to produce like in terms of CO2 00:16:55.000 --> 00:17:03.000 So I guess the answer is really that we've considered this and we've tried to just make it as stable as possible with regards to ILC. 00:17:03.000 --> 00:17:22.000 So sustainability wise, it works out similar in terms of how this thing or runs, but really it's much cheaper and lower in the carbon tax to build this thing. 00:17:22.000 --> 00:17:26.000 And there is one more question from Chandra Mulli. 00:17:26.000 --> 00:17:34.000 Hey, I might have missed it, but what's the current status of the project? 00:17:34.000 --> 00:17:42.000 So we are developing a pre-CVR. So we've published a 10-page submission to the ESPPU. 00:17:42.000 --> 00:17:57.000 That's online. There's a link in the slides to archive. And we've made progress essentially in individual elements of R&D for this project, essentially subsystems. 00:17:57.000 --> 00:18:07.000 But they've kind of operated in a siloed way so far. We've put together this, which is Bayesian optimized, but it takes a lot of leaps of faith. 00:18:07.000 --> 00:18:17.000 So the next step over the next year is to take what we've done for the ESPPU and turn it into a cradle to grave simulation, which covers the entire concept. 00:18:17.000 --> 00:18:31.000 And then publish that in a pre-CDR. And then from there, a CDR and a TDR over the next three to five years. 00:18:31.000 --> 00:18:37.000 Well, I hope that sort of answers the question. Thanks a lot, Richard. This was a very interesting talk. 00:18:37.000 --> 00:18:44.000 I think we can move on to the last one of the day, if I'm not wrong, hopefully not wrong.