WEBVTT 00:00.000 --> 00:23.000 Okay. Good morning, everyone. Thank you for coming to our talk. My name is Daniel Fernandes and I'm going to be talking about our project of building an open source battery for stationary storage. 00:23.000 --> 00:35.000 I am part of the Flowbattery Research Collective and we've been doing this project along with Utrecht University with their Fairbattery project with Josh series representing. 00:35.000 --> 00:46.000 We're also working with Kirk's Memphis also part of the Research Collective. He's currently in the US. He couldn't be here. 00:46.000 --> 01:06.000 Now, what is the Flowbattery Research Collective? These started last year. I have been working by myself in batteries. I'm a chemist and I started doing research on batteries at my house since like 2019. I started a blog and then Kirk found me. 01:06.000 --> 01:15.000 And he was like, hey, you're doing all these open source battery stuff. Let's start like a joint effort and do something more concrete. 01:15.000 --> 01:30.000 So we started the Flowbatter Research Collective to build an open source battery aiming to solar and we're in storage in the long term in this short term aiming to some academically oriented kids. 01:31.000 --> 01:40.000 Now, I'm first going to talk a little bit about lithium ion batteries because this is the battery technology that most of you are probably familiar with in lithium ion batteries. 01:40.000 --> 01:55.000 We're basically just moving lithium ions from graphite substrate where they are in a more reduced state to a normal metal oxide where they're in a more oxidized state. 01:55.000 --> 02:01.000 So basically we transport lithium ions and then we transport electrons and that's how these batteries work. 02:01.000 --> 02:09.000 What most people don't realize about these batteries is how thin this thing is. So when you see this, it looks pretty large. 02:09.000 --> 02:14.000 In reality, what you have in your phone is actually this. 02:14.000 --> 02:22.000 And you have these graphite, then you have these metal oxide in the middle. You have a separator. 02:22.000 --> 02:28.000 These entire thing, this is a scanning electron microscopy image, a cross section. 02:28.000 --> 02:33.000 And this entire thing is a fourth of a human hair in size. 02:33.000 --> 02:39.000 This is incredibly small. These batteries are like a miracle of modern nanotechnology. They're amazing. 02:39.000 --> 02:51.000 And what separates these batteries from shorting is a five micrometer separator that's there. It's incredibly, incredibly thin. 02:51.000 --> 02:57.000 And this is, I mean, issue with lithium batteries. You have thousands of these layers packed and rolled together. 02:57.000 --> 03:03.000 And that's why they are so amazingly energy dense. And then, you know, these happens. 03:03.000 --> 03:15.000 And this was most landing two weeks ago in California. They just catch fire because you get any, any sort of puncture in this like tiny, tiny, sliver of a layer and it all goes up in flames. 03:15.000 --> 03:20.000 These, these fires are not put out by the way. You have to allow them to burn out. 03:20.000 --> 03:26.000 So like, you cannot pull water on them because they just become worse. So you just leave them there. 03:26.000 --> 03:34.000 This was the same most landing fire. And now, this was also in the US, in California, in San Diego. 03:34.000 --> 03:41.000 These all tend to happen in California because they have some of the largest installations. 03:41.000 --> 03:54.000 And then, this was also in San Diego. And you can see how hot these fires are. These fires are quite, quite hot. That is, probably, you know, 03:54.000 --> 04:05.000 could set anything around it on fire as well. So this is why you have these huge clearance around these battery situations. 04:05.000 --> 04:21.000 And yeah, sorry, about the freedom units there. It's, it's 185, Josh, you as an American, you probably have better stats. 04:21.000 --> 04:32.000 And this is in Australia, in Germany, sorry. This is in Germany, so that you can't see that they also happen in Europe. They're not like only, only for Americans. 04:32.000 --> 04:42.000 So this is why we decided to go with working a solution that is more robust for large scale energy storage. 04:42.000 --> 04:49.000 So we have these flow batteries that are created that were created in the 1980s. 04:49.000 --> 04:54.000 And what you do with these flow batteries is that you don't have these solid state layers. 04:54.000 --> 05:01.000 But what we have is the cell in the middle, where the power conversion happens, where we actually flow the ions. 05:01.000 --> 05:07.000 And then we have these tanks where we store their reagents. 05:07.000 --> 05:14.000 So we have something that gives electrons, something that takes electrons, and these are cycled in tanks. 05:14.000 --> 05:21.000 The great thing about this is that these tanks can be made as large as you want for storage capacity. 05:21.000 --> 05:25.000 And then if you need more power, then you get a bigger cell. 05:25.000 --> 05:31.000 So the power is determined by the cell, you have in the middle, your capacity is determined by the size of your tanks. 05:31.000 --> 05:38.000 So you decouple power and capacity, and that allows you to have a much more robust system. 05:38.000 --> 05:44.000 More, the more commonly studied systems are all achus, so they don't catch fire. 05:44.000 --> 05:56.000 And they are generally much more environmentally friendly than the chemistry, so the other of the lithium ion batteries. 05:56.000 --> 06:06.000 There is currently no open source battery initiative at all from either lithium ion or this field. 06:06.000 --> 06:11.000 And this is why we decided to create one. 06:11.000 --> 06:17.000 So a question is, how can these compete with the actual lithium ion batteries? 06:17.000 --> 06:24.000 Is there any way that they can actually be viable for large-scale storage? 06:24.000 --> 06:29.000 In reality, these flow batteries are nowhere nearly as dense as the lithium ion batteries. 06:29.000 --> 06:36.000 They are probably at 10th or even at 20th of the energy density of lithium ion batteries. 06:36.000 --> 06:45.000 However, they have great advantages in safety, in affordability, and in cycle life. 06:45.000 --> 06:50.000 Because you don't have their reagents as an intrinsic part of the system. 06:50.000 --> 06:55.000 If something breaks in the cell, you can take it apart and replace it. 06:55.000 --> 07:01.000 If some of their reagents stop working, you can replace their reagents. 07:01.000 --> 07:07.000 You can process their reagents and regenerate them if that's the case. 07:07.000 --> 07:12.000 So they can cycle for much longer than a lithium battery can. 07:12.000 --> 07:18.000 A lithium battery, any sort of small failure, will be critical and will destroy the battery in a flow battery. 07:18.000 --> 07:22.000 They are much more robust to the sort of failure. 07:22.000 --> 07:24.000 So why open source? 07:24.000 --> 07:28.000 So as I mentioned, there is no complete open source for battery architecture. 07:28.000 --> 07:33.000 There are some open source flow battery projects that aim at producing just the cell. 07:33.000 --> 07:36.000 So there are some open source papers that show you the cell. 07:36.000 --> 07:39.000 Like, we built the cell open source so that you can build it. 07:39.000 --> 07:45.000 But then when you actually go and try to work on this, you realize there's a bunch of other little things. 07:45.000 --> 07:50.000 You have pumps, you have the reservoirs, you have the electronics that you need. 07:50.000 --> 07:54.000 All these little things are needed that they don't tell you how to build. 07:54.000 --> 08:01.000 So if you are not already at a lab where you already have everything else, it is sort of very hard for you to do. 08:01.000 --> 08:06.000 So another problem we wanted to address is that flow battery research. 08:06.000 --> 08:11.000 Because of all these little things, suffers from a lot of reproducibility issues. 08:11.000 --> 08:16.000 A lot of these flow battery researchers will have completely different results. 08:16.000 --> 08:21.000 And there is no standard cell in the flow battery research community. 08:21.000 --> 08:27.000 So we also wanted to create a cell that could be used as a standard for researchers. 08:27.000 --> 08:39.000 That was also accessible to researchers in countries that do not have the resources to spend 20,000 euros to buy a single cell from a company that sells appropriator. 08:39.000 --> 08:52.000 So although we do want to sell kids ourselves as the flow battery research collective, we want to make them so that anybody can also build their own if they don't want to buy them from us. 08:52.000 --> 08:56.000 And also no large scale open source flow battery cell exists. 08:56.000 --> 09:06.000 Well there are these small projects that aim to give you an idea to build these small cells for lab research. 09:06.000 --> 09:18.000 There is really no open source cell design for like a large flow battery like what you would use in for large large scale storage. 09:18.000 --> 09:32.000 What we want to do so we have at the end of 2024 we completed our bench top cell which is small which is what will become our kit that we will be probably selling this year or next year. 09:32.000 --> 09:41.000 For anybody who wants to do this, this is all open source that are available at our website if you want to build your own or all the instructions are there. 09:41.000 --> 09:54.000 And we're currently testing chemistries and different materials to you know when we start selling the kit be able to sell something that has some very reproducible results that anybody could will reproduce. 09:54.000 --> 10:06.000 Then by 2025 we want to have a larger cell format single cell but a larger something that would be scalable for large scale energy storage. 10:06.000 --> 10:19.000 And then by the end of this year hopefully we can have a stacking of these cells so that we can start moving towards something that would be more at the scale of powering a house. 10:19.000 --> 10:29.000 Rather than you know the first large scale we are aiming for is this which is something that might power your house rather than something that might power you know was small town. 10:29.000 --> 10:37.000 So this is an image from Redflow which is a company that recently went bankrupt. 10:38.000 --> 10:54.000 This is also why we're so interested in these being open source and it's because Redflow invested a bunch of time into the cell that is actually pretty good but they just couldn't make the capital work you know there. 10:54.000 --> 11:05.000 And if the technology has been open source maybe a lot of people could have you know either started other businesses or built built them for themselves. 11:05.000 --> 11:12.000 So we are interested in making it such that this scenario with a. 11:12.000 --> 11:33.000 And then yeah there are very large flowboring installations right now like they want you see there which you know you can 200 mega watt hours so flowbatteries are currently being used for large scale installations but we obviously are not going to go there first our aim is to go to something like this first. 11:33.000 --> 11:49.000 And then we'll see if we're able to go to even larger scales once you get to the scale getting to that scale is way easier the most difficult step is between the scale we're at at a the lab scale and these like intermediate scale. 11:49.000 --> 12:04.000 So this is this was our first prototype for flowbatteries cell and you can see these these cells have a lot of different layers we started by building. 12:04.000 --> 12:11.000 And these cells out of um polypropylene but. 12:11.000 --> 12:31.000 These initial design we used silicon gaskets we used the you can see you have your current collecting electrodes there at the ends and this is then where the flow frame is where the solution goes through and then you have a membrane at the center separating both reagents and. 12:31.000 --> 12:59.000 And it is at the same where there's a lot of possibilities for leakage and problems and this was the first one that we ever built this was a recent printed so it wasn't polypropylene because we never got to print polypropylene well enough because it worked too much and these are a small diaphragm pumps none of these worked the tooling didn't work the cell didn't work the pumps didn't work nothing. 13:01.000 --> 13:17.000 But what whatever works the first time you tried right and this is what we use for controlling these pumps and so we use an Arduino and then we use these a very cheap motor drivers. 13:17.000 --> 13:26.000 At the entire cost of these is you know a very low sprawl less than you know 20 years or something for the electronic part. 13:26.000 --> 13:33.000 And this was our second try where we like had these. 13:34.000 --> 13:51.000 Polypropylene bodies machine and we built those we used now per static pumps instead of diaphragm pumps several improvements from like the first design and this is how it looked while it was running. 13:51.000 --> 14:03.000 As you can see nothing too exciting just to pumps but it still had problems you can see that pump there is no supposed to be orange so. 14:03.000 --> 14:14.000 So leaks do happen so this this was also all in my apartment and I'm still here so it's not that dangerous. 14:14.000 --> 14:29.000 So then we changed to this second design which is was made to you know prevent all those leak problems and we decoupled the flowing part of the battery from the. 14:30.000 --> 14:40.000 Like face plates of the battery in our previous design these plates were exposed to the chemistry to some extent so they needed to be made from an inert plastic. 14:40.000 --> 15:00.000 Now we decoupled them and that way we can actually have those plates built from PLA or PETG things that are easier to print and then our flow frames are made out of polypropylene but that geometry is just way easier to print the flow frame is just the one thing with the barps that's there in the middle. 15:00.000 --> 15:20.000 And this is how the new design looks like you can see that the pumps have been changing and it's because we started with the cheapest possible pump we could get and I broke it and then we moved to the next one and we're doing that I've been doing that iteratively until I get to the cheapest thing that works. 15:20.000 --> 15:27.000 So basically breaking every tier of AliExpress pump I can I can find. 15:28.000 --> 15:38.000 Yeah yeah so like I've broken them in many ways you know corrosion like just the electronics fry I mean. 15:39.000 --> 15:50.000 And this sell this sell work pretty well and these are some of our first testing results these are charged this charge curves for a sell. 15:51.000 --> 16:09.000 And we are using initially sink iodide as a chemistry because it is very these materials are easy to source like the reagents are quite easy to source in the EU without any need for special licensees to buy chemical materials so. 16:09.000 --> 16:32.000 We are testing sink iodide and it's not extremely toxic it's not it's not like you can drink it but you know it's it's fine and we we got we cycle these for more than five days and these energy density to give you an idea is similar. 16:32.000 --> 16:45.000 This is not like half the energy density of what like a commercial flow battery achieves and and then you can see on the other side our efficiencies through time are pretty stable. 16:45.000 --> 16:55.000 And the research the normally in research the membrane that you can see here in the middle so these these membrane is usually a very expensive. 16:55.000 --> 17:09.000 Caria next change membrane that is it's called napion and it's expensive and we don't use that we use a micro Paris membrane that's very fancy it's called photo paper. 17:09.000 --> 17:31.000 So we use photo paper as our membrane and this is why our energy efficiencies are under our coolant efficiencies are not super high we're in like 65% energy efficiency because we do get some leakage of stuff and some additional resistance but it's very very cheap to achieve. 17:31.000 --> 17:41.000 And then we can also push it to higher energy densities so they see at the level of a commercial flow battery. 17:41.000 --> 18:08.000 But you can see there's only two cycles here and it's because the other stuff fails it's not like the chemistry that failed at this point it was too being it was palms it was like these things become so concentrated as things become more energy dense they are usually more reactive right because you're in states were things you know are holding a lot of chemical energy they want to release so they start reacting with other stuff in the system. 18:08.000 --> 18:18.000 which includes two being which includes the the pump which includes like other stuff around so this we only have two cycles here because of the. 18:19.000 --> 18:46.000 And now this is the pump that we got for the next step this is like a beefy pump this is like the size of like a face this is an impeller pump and to give you an idea our previous pumps are going like maybe 60 milliliters per minute this is like 6000 milliliters per minute so this is for the next scale up that we're doing and. 18:47.000 --> 19:02.000 Yeah, so this is a little bit about how you guys can get involved if you want to help always trying to build a kit with our instructions is useful because nobody has tried so we don't know. 19:03.000 --> 19:09.000 So we want to make the documentation better and want to make sure that people like can do that. 19:10.000 --> 19:31.000 Also, if you don't want to build a full kit just testing 3d printing of our 3d printing pieces works because we need to make sure that they can be printed in a lot of different printers and with different materials like for the flow frames we can only use polypropylene it's the only chemically compatible like widely available 3d printed material that can be used. 19:33.000 --> 19:41.000 You can also help us test different to be materials or pumps and also. 19:42.000 --> 19:52.000 Building on testing larger scale sales are designed so still in progress there, but when we have them helping us test that with only water because we don't want anyone to die helping us test. 19:52.000 --> 20:14.000 Then that would be very useful and then from an electronic perspective we don't have a BMS for managing these batteries at scale like in in the small lab scale they batteries don't even produce you know power to run the very static pumps they're they're very static pumps are very inefficient and at a large scale the aim is to have. 20:15.000 --> 20:24.000 At least well 2% of the energy go to the pumps at most or and then I would like to acknowledge will. 20:25.000 --> 20:34.000 Josh and the for battery team Kirk or chief engineer the organizers of the conference and then the entire initiative is financed by and on its. 20:34.000 --> 20:40.000 And the Io and trust fund so thanks to an on that for.