Flow Battery Research Collective

The cost per kWh of this is too high, the use of organic materials and ion exchange membranes puts this closer to what people do in academia and further from the cells that can be reasonably fabricated using DIY approaches. The use of an ion exchange membrane likely reduces the lifetime of this cell a lot. For a static cell, I think the sulfuric acid Cu/Mn battery we have discussed before is far more promising. That chemistry requires no exchange membranes, uses only commodity inorganic materials and cycles to 30-40Wh/L. For pure Fe batteries I am much more inclined to the Fe flow batteries.

@rowow said in Using cnc router cut PVC foam boards as cell frame: The long term goal and my personal next goal is to get into injection molding which of course is the best option. At scale, definitely! @rowow said in Using cnc router cut PVC foam boards as cell frame: Ill see about uploading the cell fusion 360 model A .step file would also be great so I could look at it in FreeCAD! I don't have a Fusion 360 license. I'd be curious to see if we could take a similar approach.

@doho said in Upcoming improvements to the dev kit: I have try-ed to download Your new files from github (.stl and .pdf), but the files all appeared as broken, even in github. The files to sprint for the new endplates are here: https://codeberg.org/FBRC/RFB-dev-kit/src/branch/clampable-cell/CAD/exports [image: 1771858819246-1c46bc80-7e99-4780-8178-46d24a289a32-image.png] I haven't updated the new You have to download them individually, unless you clone the repository with Git LFS set up---the large CAD files are handled with LFS and without LFS installed downloading the repository just downlaods a pointer/reference to the CAD file, not the actual file (at least that's how I understand it). Let me know if this works for you!

Very nice work @doho ! It's great to see your setup. @doho said in My Suction Luer Lock: But for getting reproducible resultants the flow through the cell should be down to up, but not horizontal (how much would be air isolating in the chamber when changing parts?) This is a good point, from what I've seen in other applications cells should generally clear air/produced gases by flowing against gravity from bottom to top, that is still possible with the new setup but the tubing from the pump outlet to cell inlet would have to be slightly longer. In my jig redesign I'll take this into account when writing the documentation.

@rowow We don't have the equipment or time to prepare these membranes but if you send us samples of 100-400um thick membranes we can test them out.

Video of the talk is now available here! https://video.fosdem.org/2026/aw1126/3EEZZB-open-source-batteries.av1.webm

Did another leak test today with water, correctly with 2x ~12 mm plywood endplates each side. Saw no leaks through the edges which was great news, but the barbed connections on the cell showed signs. Also, the MP-6R pumps struggle with the current flow frame design, which has 0.8 mm wall thickness and a 1 mm internal channel (electrode area therefore 2x0.8 + 1 = 2.6 mm thick). I remade (and pushed to the repo) the flow frame with a 3 mm internal thickness, in order to alleviate this pressure drop. Here is the test setup, I ran out of tubing (ordered 2m but they sent 1 m ), so the connections aren't ideal but this time no kinks in the flow path. Note, I put these drain valves in, of course they are pointing the wrong way for now, will need to elevate the setup so they can point down in the future. [image: 1769764493660-9fafda86-7f70-439f-9ecb-eeb9a7316215-img_20260129_153901-resized.jpg] Because all the connections are on one side (in anticipation of stacking these cells), I also made "front" and "rear" versions of endplates, inner/outer current collectors, and gaskets in the FreeCAD files. This will make low-volume prototyping a bit more expensive but more robust against leaks, which no one wants! [image: 1769763690358-037a7741-97dc-4012-8e95-cf8ab8760653-image.png] This is the dimension that went from 1 mm to 3 mm to facilitate using MP-6R pumps. [image: 1769763764115-d4ac6efb-4694-4d13-a941-8ffb85995001-image.png] We have the MP-6R now. It is the 6W high-flow version. the MP-10 is also 6W but lower flow / 50 % higher max pressure, then the MP-15R can do almost 3x higher pressure than the MP-6R but at 10 W. @danielfp248 can hopefully print the 3mm flow frames and I can get them at FOSDEM, then try them out. If it turns out we need the bigger pumps, I'll order them from AliExpress: [image: 1769763852108-f14b9379-cbc5-4542-940c-b33c0bacdb14-image.png]

Hi all, and apologies for the delay! This year has started off with quite a lot of administrative burden for me and I haven't had as much time for research as I anticipated. @Santiago-Eduardo said in Life Cycle Assessment (LCA) for the FBRC redox-flow battery: Elektrolyte: The group noticed that two slightly different electrolyte compositions are mentioned. Sorry for the confusion, the correct mass composition can be found in the documentation here: https://fbrc.codeberg.page/rfb-dev-kit/electrolyte.html, the masses listed will prepare approximately 10 mL of electrolyte. @Santiago-Eduardo said in Life Cycle Assessment (LCA) for the FBRC redox-flow battery: They assume an 880 ml volume for one single cell. Do you estimate this volume to maintain the obtained results until now? Is this the volume foreseen to achieve the 22 Wh/single large-format cell? Yes this would be correct volume scaling for the large-format cell, although of course still a lot smaller than an eventual life-size system! It is basically t]e volume that we will end up using for our tests of the large-format cell (still to come). @Santiago-Eduardo said in Life Cycle Assessment (LCA) for the FBRC redox-flow battery: The group is assuming the EE value to estimate this. This value does not include energy demand from pumps and electronics, correct? Correct, these losses are often summed up in RFB literature as "balance-of-plant" or BoP if you want to search for some values. @Santiago-Eduardo said in Life Cycle Assessment (LCA) for the FBRC redox-flow battery: Meaning: to store 1 Wh, 1.56 Wh needs to be taken from the grid (excluding electronics and pumps). Does this make sense, or are we oversimplifying here? You've got it exactly! @Santiago-Eduardo said in Life Cycle Assessment (LCA) for the FBRC redox-flow battery: Used electricity For the same purpose of modelling the use phase, it is important to define from which country and what type of energy/electricity is being used to charge the electrolyte. Since the users of the FBRC battery can be anywhere, but are currently mostly centered in Europe, the group has decided to choose the European electricity grid mix data to represent the current FBRC reality. This makes sense to me. @Santiago-Eduardo said in Life Cycle Assessment (LCA) for the FBRC redox-flow battery: The separator, for instance, would be one of these peripheral impacts, since it would need to be replaced after some cycles (probably faster than the electrolyte). Since cycle durability of photo paper is still unknown, the group will model different scenarios from 10–100 cycles in steps of 30 cycles. Do you feel this is a reasonable range? Do you already now conditions such as density and flow rate the larger cell will work with? The group will assume 4 layers for the larger cell although in some parts are 3 layers stated. While separators can be replaced, I am doubtful in an industrial system that they would, due to the labor costs. From my understanding, Li-ion lifetimes are often given as 2,000 cycles to 80% of initial capacity; for flow batteries, the data isn't as solid, but for VRFB the lifetime claims are more on the scale of 20,000 cycles or 20 years, whichever comes sooner (taken with a grain of salt...). We haven't done any testing that long-term, so don't have much for your to extrapolate, but RFB technoeconomic papers with operation and maintenance (O&M) costs incorporated would give you a good idea of membrane/pump replacement frequency (if ever). I would increase the cycle range to much longer terms, with the upper end in the 1,000s at least. We aren't yet locked-in on flow rates for the large cell as we are still settling on choice of pumps and flow field design. @Santiago-Eduardo said in Life Cycle Assessment (LCA) for the FBRC redox-flow battery: The current BOM and building instructions do not provide specific links to purchase the necessary chemicals. To model the electrolyte production, including the transportation of each chemical, the group has assumed the following production locations based on market data and worldwide production trends. If your own experience differs in this, please do not hesitate to comment. These assumptions all make sense to me; Daramic separator (which we also use in addition to paper depending on the test) can/is produced in the EU though not exclusively. I hope this clears things up for you all somewhat, and again, sorry for the delay!

@danielfp248 I always performed more cycles before starting 40 mA / 100 mAh charging (10 half-cycles at 20 mA / 10 mAh + 4–10 half-cycles at 30 mA / 10 mAh). The cell was also wet with demineralized water (leakage test). Could this also be the cause? I was also always using the membrane frame. Wouldn't the electrolyte leak through the paper membrane? @kirk Apart from solving the electrolyte leakage issue, does this "pulling-through configuration" improve the total capacity of the system in any other way? Also, I am wondering whether you are using the default flow frame from the documentation (2 mm thick, right?) or a different one with another thickness and the 0.1 mm silicone gaskets.

Hi sepi, your refactor-modular works on my PCs: my mainPC (Kubuntut 2404) and two Notebooks for measurements (Kubuntu 2404 and Win11), thanks. I'm waiting for your calibration wizard. My hardware is waiting incl. Mystat, but till now I did not do tests.

@sepi Without the four-wire control the pump speeds are definitely harder to manage, sorry to hear that, though the PWM regulator may help. Can you visually see how different the rotation speeds are? That should be rough proxy for volumetric flowrate. I am not surprised they won't spin below 11 V. @sepi said in My build (very slowly progressing): Is this 20-40ml/min measured with the cell in line or just the pump. Ideally best to measure with the cell inline, to be closer to the actual conditions. However, as the tubing wears, this can drift, just FYI. Not a huge deal but something to be aware of. @sepi said in My build (very slowly progressing): Is the figure obtained with water or electrolyte? Water for now, just as a proxy, minimize exposure to electrolyte. This is kind of a 1, maybe 2 significant digit measurement, no more (the volumetric flowrate). @sepi said in My build (very slowly progressing): What would you recommend doing against the excess flow? @sepi said in My build (very slowly progressing): the left passing up to 40% more water at the same voltage. At the same voltage, is the left pump spinning at (nearly) the same rate as the right? if they are spinning at the same rotational speed and the left is passing 40% more water, it sounds like there is some flow restriction in the right pump's flow loop. If they have identical flow loops (in terms of pressure drop), and are set at the same voltage, I would guess that they should rotate at similar speeds, but this may not be the case for the brushed motors (e.g. variance in motor properties). What is important for testing, is that both pumps are above a minimum flowrate to ensure good mass transfer, and that the pressure imbalance between the two sides isn't too great as to cause transfer of electrolyte from one side to another through the separator. If the flowrates aren't perfectly matched but those two previous conditions are met, it's not a big issue, although it's just better for repeatability for them to be the same.

The tubing is gone now, thanks for buying at my store