A Redox Flow Battery Made From Iron Industry Waste

Researchers at the University of Southern California have found a way to make an effective and competitive redox flow battery out of the iron industry’s waste products.  Luckily for us, the results of the paper were posted on an open journal and we could take a look into the tech behind this battery.

As electric utilization, adoption of electric cars, and the use of renewable power continues to rise, engineers all over are searching for the perfect utility scale battery. We have all heard about Tesla’s 100MW lithium battery pack in South Australia. The system is a massive success and has already paid itself back. However, engineers all over were quick to point out that, until we have a breakthrough, Lithium cells are just not the right choice for a utility system in the long run. There has to be a better solution.

What should a good battery look like?

A Vanadium Redox flow battery located at the University of New South Wales. Credit: Radiotrefoil

Grid scale storage is a important and difficult problem to solve. If we don’t have batteries, we can’t move away from fossil fuel power generation. Without grid storage every power surge, such as England turning on its kettles for tea, would knock it out.

Currently power plants are spooled up and down all day to match the demand. Even if we didn’t have renewable energy on the horizon, better batteries would allow us to run smaller power plants more efficiently to service demand. You could run one plant at maximum efficiency all day and the battery could level the load. It’s the difference between your car’s fuel economy if you slowdown and accelerate in the city or stay at cruising speed on the highway.

The ideal battery has a few requirements. It has to take charge at a reasonable rate. It also needs to be able to delivery that charge instantly. Most importantly, the battery must be cheap and last a long time. It’s literally the impossible fast, good, cheap dilemma. Engineers have tried everything from molten salt to stacking rocks. A hydroelectric dam, for example, is nothing more than a battery made out of water and gravity. Tesla’s large batteries are made out of lithium-ion cells.

The primary economic decision driver in picking a system like this is the Levelized Cost of Energy Storage (LCOS). This is the sum of the capital and operating costs of the system by the total energy stored and delivered over the life of a system. The paper mentions that the US Department of Energy specifies an LCOS target of 2.5 cents / KWh. At an installation cost of $200 / kWh, a battery must deliver 8000 kWh of energy over its lifetime. As the authors point out, if you consider one discharge and charge cycle a day, this means the battery should last 22 years. We know that lithium cells are not up to that challenge.

Redox Flow Batteries

A diagram from the paper showing the operation of a typical flow battery.
A diagram from the paper showing the operation of a typical flow battery.

There is another type of battery out there, which we’ve covered before, called a redox flow battery. This battery has the potential for nearly unlimited life, low operating cost, and might even be ecologically friendly.

The chemicals can be two separate chemicals, the same one, or any arbitrary blend. This affects how efficient the battery is and how hard it is to recycle/refresh the chemicals used during operation.
The chemicals can be two separate chemicals, the same one, or any arbitrary blend. This affects how efficient the battery is and how hard it is to recycle/refresh the chemicals used during operation.

In this battery two chemicals flow past each other separated by a membrane.  Depending on how you place the electrodes in the chemicals, protons pass through the membrane from one vat to another building up a potential difference between the two. The substances used can be the same chemicals, different ones, or even mixes of the two.

The current state of the art redox battery is based on vanadium. The largest install is in Japan at 60 MWh. However, China’s truly impressive power-infrastructure super-corporations are currently building a 800 MWh install. However, vanadium isn’t cheap, and the material is more than a little bit toxic.

Iron Batteries

There’s another element that works wonderfully in these batteries, iron. It’s just waiting on a breakthrough that would let it operate at scale, and that’s where the researchers have had their breakthrough. They found that they can use iron sulfate, a by-product of the iron industry, and anthraquinone disulfonic acid as the chemical on both sides of the vat and get a very effective redox battery.

It has every advantage over the vanadium battery aside from a lower cell voltage. This means that the battery installation has to be a bit larger and more complex, but in the end it stands to be cheaper because the primary electrolytic is a relatively safe industrial offshoot, and the battery has the simplicity of it being a symmetric system: the chemical mix on both sides of the membrane is the same, making refreshing the solution easy. The authors estimate that this battery will cost $54 / kWh while the current state of the art vanadium batteries are hovering between $160 / kWh to 180 / kWh. That’s a nice reduction.

These batteries are fascinating bits of technology. They’re also surprisingly hackable. Of course some of the steps to get maximum efficiency, such as doping the membrane with carbon nano-tubes may require [Ben Krasnow] levels of hacking.  But rudimentary batteries can be made out of scrap, which is a pretty good indication of the technology’s viability. We’re curious to see how energy storage revolutions will change the world in the future. What do you think?

54 thoughts on “A Redox Flow Battery Made From Iron Industry Waste

  1. “Grid scale storage is a important and difficult problem to solve. If we don’t have batteries, we can’t move away from fossil fuel power generation.” is only somewhat true…

    Though it depends on how one uses the word “battery”.

    Alternative energy storage solutions do exist and are already in use on most grids.
    A few notable examples would be:

    Pumped water energy storage.
    That simply involves having a fairly large elevation difference, and pumping water from one reservoir to another. And then letting gravity take it back down through a turbine. These systems do have limits, they aren’t that practical in the middle of sub Saharan Africa for an example due to a lack of water. But one such system is used in England to solve their kettle issue. And other systems exist in in a long list of countries. Wikipedia has a list of countries making use of this technology: https://en.wikipedia.org/wiki/Pumped-storage_hydroelectricity

    There is also flywheel based energy storage.
    A notable facility is one in New York state: https://en.wikipedia.org/wiki/Flywheel_storage_power_system
    But other than that, large scale facilities are not that common, but smaller facilities are used for regenerative breaking on things like trains/trams. Though, synchronous condensers do have the ability to store some energy and mitigate brown outs. Though their main job is to facilitate power factor correction, and if one starts searching for them, they aren’t all that uncommon, though still a rare sight. https://en.wikipedia.org/wiki/Synchronous_condenser

    There is also compressed air energy storage.
    These can make use of disused natural gas wells. Though, power efficiency isn’t all that great at current, since it is only about 60%. But improvements are steadily being made and all one needs is an efficient compressor and air engine, and a place to store the compressed air. Though, high pressures will be needed for it to reach an energy density comparable to batteries. (Like the 100+ bar region…)

    Then there is the good old idea of just lifting a large weight and letting it drive a generator on the way down.
    But I haven’t seen any actual systems being built. Other then clocks…

    Another energy storage system in use is molten salt.
    In short, one has a large very well insulated container containing liquid salt at about 600-1200 degrees C, we can cool it down by letting it boil water for us. And the energy needed to be release for it to transition from liquid to solid is typically the main way to store energy.
    These systems are used in concentrated solar power stations to provide sufficient warmth to run the steam turbines during night. (Since apparently, people need power into the nights and early mornings….)

    So “batteries” are far from the only solution in existence.
    Though, batteries do have their own advantages compared to most of these other systems that only starts making sense at large scales.

    I for one don’t want a pumped water UPS for my house, nor a flywheel one, or a molten salt one… Compressed air on the other hand at least is something that can sever dual use fairly easily. Though, even then, compressed air has a very low energy density at usual “shop pressures” (8-15 bar)…

    1. Most viable pumped storage locations world wide are already in use as such.
      Flywheel storage has massive drawbacks in that it’s not very energy dense, the maintenance cost to keep it running is pretty high and the results of failure can be nothing short of catastrophic. (I’ve seen photo’s of the aftermath of an energy storage flywheel prototype breaking loose in a test facility in the Netherlands. Thankfully no-one died but several people where injured as the several dozen kilogram flywheel went pinballing through the hall at initially over 20k rpm. To say the damage was impressive is an understatement).
      Compressed air has similar problems, and it’s inherently a very lossy process and extremely not energy dense. Compressing a gas like air is just very inefficient, and there isn’t much you can do about that. The best you can hope to improve is compressing the gas enough that the residual heat has enough available enthalpy to be able to extract it and re-use it for something in another lossy process
      Molten salt is useable is certain circumstances, but due to the high temperatures involved coupled with (usually corrosive) salts, the price per kWh is pretty high.

      If we can crack the lifetime issue batteries are still the most viable option for grid scale storage.

      1. Flywheels do have some major downsides. Their main advantage is their fairly high peak power output. I have heard of a few industrial facilities that spin up a flywheel before starting some high power demanding equipment, also starting to dump the energy from the flywheel into said equipment. Turning a few seconds of rapid increase in power demand into a much smoother ramp lasting many minutes. (Though, I guess companies running such equipment contacts their local power providers regardless…)

        Power factor correction is another area where flywheels are still in use. But here the RPMs aren’t as crazy, and the main goal isn’t really long term energy storage.

        Also, “Compressing a gas like air is just very inefficient, and there isn’t much you can do about that.”

        Well, in fact, we can do a lot about that.
        Energy efficiency of air compressors is largely proportional to how much the air inside the compressor heats up during compression.
        Using inter-coolers in multi stage compressors can radically increase power efficiency. And each compressor stage doesn’t need to compress the air as much, resulting in a smaller temperature increase. We can also run the compressor slower, giving the heat in the air more time to soak into the walls of the compressor.

        Some compressor systems aiming at efficiency puts a water mist into the air to increase its thermal mass so that it can’t as easily get warm. While others use heat sink fins to get better contact between the air and the cylinder walls. (Since air conducts heat fairly poorly.)

        The typical shop/home compressor isn’t even remotely built for efficiency, but rather speed. Industrial compressors largely follow the same trend. (Until one gets up to literal buildings providing compressed air to whole factory districts then inter-coolers are rather common.)

        Same thing applies to air engines, except the air gets colder instead.

        Energy storage wise, compressed air at 200 bar stores about 0.1 MJ/L, lithium ion batteries are up at 0.90–2.43 MJ/L
        But the gigantic difference is the cost for the storage volume, compressed air storage just needs an air tight space, like a disused natural gas well, that can contain millions of cubic meters of empty space. While batteries fill their volume with a lot more expensive materials.

        But compressed air isn’t an ideal solution in all cases, its mostly low cost and low leakage.
        Peak power and quick response times is where batteries and flywheels are superior.

        Molten salt on the other hand is kinda only useful if one has a concentrated solar plant at hand. Otherwise it really isn’t worth it.

      1. Even on a large scale, it isn’t all that practical.
        And dangling large weights in old mine shafts can have questionable safety. So a rigorous safety factor must be upheld, not to mention inspections of the wires holding the weight in the air.
        And I don’t want to imagine the forces the gear reduction would be under….

        Though, holding lets say we have a cubic meter of solid lead 300 meters in the air.
        Then we have 11300 kg roughly, able to give us about 9.2 kwh of energy.
        9.2 kwh is though on par with a bunch of lead acid batteries. So in this case, it would likely be less work to just use the lead for batteries…

        Its no surprise this technology hasn’t caught on outside of clocks.

        Maybe a large coiled up building sized spring is a better idea? (I am joking….)

        1. There was a HaD article a few months ago about an energy storage system based on lifting and then stacking weights. One of the benefits versus an electro-chemical solution would be the low self discharge rate (depends how strong the wind is).

        2. “Maybe a large coiled up building sized spring is a better idea? (I am joking….)”

          You had to be joking huh? :-|) Seriously, however, makes me wondering about the mass energy balance and long term cost effectiveness of an underground spring mechanism included.

          Speaking of underground, I do wonder about the pumped hydroelectric being implemented underground more also. Guessing, most caves have already been flooded for use and where there may be locations… there isn’t so much for water to pump back and forth along with being in a strata that won’t dissolve (unless they maybe pump in salt water and line the mines critically). Maybe more critically designed and re-enforced underground locations can be developed.

          Fun with the k value of a huge spring maybe best to coin as Mainspring Energy” to add to the Gravity Battery method(s).

          I didn’t realize there was a Gravity Light on the market… or even DIY versions. Why not add a spring or springs for more stored energy?

          1. Underground caves could theoretically be used as pumped water energy storage facilities indeed.

            Though, most caves/mines tends to leak sufficiently to need pumps to stay reasonably dry as is.
            So if one has a very dry mine/cave, then it could be useful for such a purpose. And some mines can be very very deep.

            Not to mention that one doesn’t need a mountain for it to work.

    2. Then there’s thermal storage, which while not practical to turn the stored energy back to electricity, it does cover the biggest use of electricity pretty much spot on – HVAC. Where things get interesting is that pound for pound, making ice for air conditioning nets a similar energy density to running an air conditioner from lithium batteries, but the thermal storage tanks can be cycled an unlimited number of times, can be left at any state of charge for extended periods with no impact on lifespan, has no inherent limit on lifespan, and is orders of magnitude cheaper.

      1. Stanford University used to have a olympic size pool of ice that they froze each night on cheaper electricity and then used for air conditioning purposes for their chilled water loop for the whole campus

    3. I don’t think I’d want a compressed air system in my house either. The damage an old 80G tank can do if it explodes at 100PSI is life-threatening, but at least usually confined to the building that it’s contained within.

      I read an MIT paper on RedOx Flow Batteries about a year ago, and ran some napkin math, working on the premise that I’d want to store about 3 days worth of power. I came up with tank sizes of 1m^3 each, with a 1m^2 membrane, ie about the size of 2 standard IBC totes with a membrane in between. It occurred to me that the world has lots of experience with the engineering, manufacturing and movement of those totes, and that could be used to great advantage to lower the barrier cost of a flow battery.

  2. >The authors estimate that this battery will cost $54 / kWh

    That depends on the power of the battery. The most expensive part of a flow battery is the cell stack, and the low voltage of the cells means it costs a whole lot more.

    1. Not sure what you mean by this. Power is included in the estimate, and the authors have written that the voltage is lower. Are you talking about the initial cost of building a site?

      1. Yes. Flow batteries can be built with a large tank and a small electrolytic stack, which makes for a lower price per kWh, but the charge rate will obviously be low. If you want a faster charging/discharging battery, then the cell stack has to be bigger, which puts the cost up. Having lower voltage means the power of the stack is lower, so it has to be built that much bigger still.

        I’ve seen estimates that this thing costs several million dollars per MW just for the electrolytic membranes and electrodes, or billions for a gigawatt-scale system. The storage capacity can be many megawatt-hours just by having bigger electrolyte tanks, but if you want to respond to grid variations, you also need the power.

        So it’s like nuclear power – you have to build it at a huge scale to make it economical.

      1. The cells are limited in how much current AND voltage they produce – how many milliamps per square cm at what voltage. Getting more power out means exactly adding more nafion to the stack – more cells in series.

          1. Yeah, because for such small power demands, primary cells offer much greater energy density and lower price. You can have lithium-thionyl-chloride batteries that give 20,000 mAh in a standard D-cell with a shelf-life of 10 years and practically no self-discharge whatsoever, and that’s 3.6 Volts with a peak current capacity around 2-4 Amps so you also get crazy power out of them.

            If you had a power cable to the device to charge the cell, then you’d be using a wall-wart instead.

          2. Thionyl has the disadvantage that you aren’t really supposed to use them in consumer applications, and lithium is a finite critical resource, and I don’t think anyone actually recycles this stuff.

            The energy is almost unbeatable (Except maybe metal-air cells packed in dry nitrogen till they’re needed), but for anything not safety critical, we really shouldn’t just be tossing lithium around.

  3. “A hydroelectric dam, for example, is nothing more than a battery made out of water and gravity. ”

    Yeah, uh, no, that’s not true for a start. Might as well say that the diesel engine in a diesel electric train is a battery.

    Enjoy being battery! Enjoy providing power for! Nine volts power! Last very long! Keep providing power until die! Give power and power and more power until cannot give power anymore! Enjoy very much giving power!

    full version here: https://www.theonion.com/i-enjoy-being-a-battery-1819583258

    1. You have the putzes on the other side calling the Lion battery based power inverters “generators”. They do the same job as a generator until they run down. You have a hard time topping off the fuel and starting them back up again. Yet they insist on calling them generators.

      1. Well that’s a peeing contest I never engaged in. To call them that hurts no one, unless one looks for any reason, to feel hurt. In general they deliver what they are said to deliver. Would I purchase on? No. Would I assemble one? N. I have, only to meet a particular need.

      1. Most hydroelectric dams are built in rivers. Most of the capacity is actually upstream, in the small streams and waters that take days and months to reach the dam, and the dam can only regulate the flow on the scale of a few hours to a couple days. If you want to pump it back up, you need to flood a very large area for the reservoir, or build a really long pipe to get it back upstream.

          1. Hydroelectric dams also suffer from two major issues: silting and methane production. The varying water level washes down organic matter which sediments at the bottom and starts to break down consuming all the oxygen there, which makes hydroelectric dams like big biogas reactors.

            They’re not exactly “CO2” neutral, and you always have to mind the possibility of a dam failure which tends to kill a lot of people and destroy whole cities and towns when it happens.

  4. Seem to remeber reading on this site(couldn’t find it again) about NiFe cells aka Edison batterys and even heard a few storys about people using them fairly resently for off grid, yes heavy and large but they do work right? (Really not my field of expertise)

    1. They do work, they’re just very heavy. But I believe there are still electric cars from Edison’s time with the original batteries out there, they last incredibly long times.

      Over more reasonable timescales like 30 years, LTO will be better as soon as we figure out how to recycle it, if we haven’t already.

      1. They’re also relatively inefficient. Nickel-iron cells need some overpotential to charge, and the difference between the charging voltage and the voltage the cell ultimately gives out is lost as heat.

      1. Fission could get cheaper, if we get around that “OMG those types of reactors could make bomb material.” when they can run off the waste of other reactors and make it safer and less expensive to dispose of.

        1. My understanding is that constructing nuclear power plants is 98.5% or thereabouts of the cost, and fuel cost is so small an overall operational cost as to be insignificant. Reprocessing is slightly more expensive than digging up new uranium, so nobody wants to pay for it. The reason to reprocess is it almost totally solves the nuclear waste problem, if someone is willing to subsidize it. I could be wrong…

          1. 98% of the cost of constructing nuclear power is the cost of cutting through the red tape, because it takes a decade to get permits and lobby for the thing, pay every NIMBY to shut up, and all the while keep paying people and keeping up your business. Then when you’re building it, people are constantly stopping you for extra special inspections because of complaints from political organizations that are simply trying to sabotage the effort.

            Also the fact that every nuclear power plant is an individual – there’s no standard parts or designs because nobody’s building enough to warrant volume production and re-using old plans is not permitted. It’s like cars in 1900: only a handful exist and every part, every nut and bolt, is made for that car only, so there’s no efficiencies of scale.

            SMRs would change that, but it goes back to the first problem: you can only get permits to build one every 10-20 years, at best, so you have to make it count. You have to make them pay all the billions it cost you to wait that long.

          2. The decommissioning costs and time are very high, do not forget that. If they were factored in like the maintenance costs from the beginning, they would make electricity from nuclear power plants way more costly than from solar or hydroelectric.
            And the costs in case of severe accident (can´t say it never happened, and that it won´t happen again) are absolutely disastrous.

          3. > If they were factored in

            They are. Nuclear power plant operators have to pay into a trust fund set up by the NRC. This was made so the operators couldn’t dump the decommissioning costs on the society if the company goes belly up.

            >they would make electricity from nuclear power plants way more costly

            The decommissioning costs are estimated at around $800 million per site and already completed decommissioning projects have cost around $1 billion, which is some few percentage points on the cost of the typical plant and its lifetime of operation. Since the money has to be paid in advance, it’s already in the cost of the electricity.

            The high decommissioning cost is just one of those myths that anti-nuclear folk like to push for propaganda. It’s complete bunk.

          4. >And the costs in case of severe accident (can´t say it never happened, and that it won´t happen again) are absolutely disastrous.

            Hydroelectricity kills 35 times more people per energy produced compared to nuclear. That’s because when a hydroelectric dam bursts, it tends to wash away every little town and hamlet downstream. Even if you don’t count China and other developing countries into the statistics, hydroelectricity has killed and destroyed twice as much.

          5. https://www.world-nuclear.org/information-library/nuclear-fuel-cycle/nuclear-wastes/decommissioning-nuclear-facilities.aspx

            >In the USA, utilities are collecting 0.1 to 0.2 cents/kWh to fund decommissioning. They must then report regularly to the NRC on the status of their decommissioning funds. About two-thirds of the total estimated cost of decommissioning all US nuclear power reactors has already been collected, leaving a liability of about $9 billion to be covered over the remaining operating lives of about 100 reactors

        2. Not true, fission is mostly expensive because nuclear power plants are expensive. Those shouldn’t be built to lower standards, otherwise you get Tepco-grade power plants.

          1. The standards you need to adhere to depend on the size of the plant.

            Smaller reactors require less safety, but they cannot be built because of political issues.

  5. I question the cost figures of traditional flow batteries. Vanadium flow batteries have been around for quite some time, but the reason we don’t see news of them being installed is they are more expensive than even lithium batteries (check current prices).

    What looks interesting is the announcement of Form Energy to make a 150hr storage battery for a Minnesota coop utility wind farm. https://pv-magazine-usa.com/2020/05/08/form-energy-claims-its-aqueous-air-battery-provides-150-hour-duration-storage/

Leave a Reply to Alexander WikströmCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.