Published on June 27th, 2014 | by Christopher DeMorro

Video: Isentropic Wants To Store Energy In Gravel

June 27th, 2014 by  

We’re finally figuring out ways to efficiently and affordably generate green power…now the problem we have to solve is how to store that power. A UK company called Isentropic wants to store green energy via a heat pump system that uses the thermal storage of small rocks to cheaply and efficiently capture green energy.

You’ve no doubt noticed that on hot days, gravel tends to get, well, hot. Isentropic exploits that natural heat storage with a Pumped Heat Electricity Storage (PHES), which pumps gases between cold and hot storage tanks, compressing and expanding the gases at a 72 to 80% round-trip efficiency.


That’s even better than the 65% efficiency of hydropower, the current efficiency benchmark. The video does a great job of breaking down the simplicity of this green energy storage system, and it doesn’t take a genius to see why it would be a cheaper storage solution then traditional batteries.

The PHES system could become the go-to means of storing excess energy generated by solar or wind power. More importantly, it does so with a cheap and abundant natural resource (rocks!) rather than expensive lithium-ion batteries, dangerous liquid chemicals, or terrain-specific hydro power. It certainly seems simple enough to work, but will it ever become more than just a good idea?


Check out our new 93-page EV report, based on over 2,000 surveys collected from EV drivers in 49 of 50 US states, 26 European countries, and 9 Canadian provinces.

Tags: , , , , ,

About the Author

A writer and gearhead who loves all things automotive, from hybrids to HEMIs, can be found wrenching or writing- or else, he's running, because he's one of those crazy people who gets enjoyment from running insane distances.

  • Excoriator

    Unfortunately,. they seem to have been quiet for rather a long time now. The latest news update is Jan 2015. It is now September. Well over the ‘won’t pan out’ period granted them by Bob Wallace (see this thread) a year or so ago.

    It would be nice to hear how they are getting on with this pilot plant.

  • JamesWimberley

    If this really works as advertised – especially the cost; the design principles are of Victorian simplicity – expect the company to be snapped up by a major player like Siemens or ABB. On the face of it, there should not be any safety issues to prevent the units being deployed in cities, or as part of large commercial self-consumption installations.

    • Wouldn’t large companies like Siemens and ABB have been contacted already? At this stage, what’s keeping them from building it themselves? Unless they trademarked various thermodynamics terms and patented the entire process. I’m honestly curious. I find this kind of stuff interesting.

      • Bob_Wallace

        This company claimed sometime back a new heat pump design which they patented.

    • Doug Cutler

      Yes, put them almost anywhere within reason, just not residential or scenic but that would be easy to avoid. Large factories or institutions yes. But I’ve inquired of Isentropic: these systems do not scale down for domestic use. No problem. I’d be happy to drive past one and smile.

  • I think these ideas are awesome. I haven’t taken thermo in about 25 or so years and forgot what isentropic means. Oh, yeah something about not losing entropy in or out of the system. Or the process is adiabatic and reversible. Which means heat isn’t transferred from the system (I guess the rocks) and the surroundings (I guess the fluid maybe or outside the vessels?). Nerds, have at it.

    I’m guessing this system is not just water and rocks placed into big vats or enclosed insulated vessels. That system would probably work for about 17 days before plugging from biomass accumulation and becoming inoperable. Even with enormous amounts of chlorine bleach or hydrogen peroxide added to kill the bacteria. Bacteria growth is the bane of fluid flow through porous media. Ranging from water filtration and treatment to oil and gas extraction from fractured wells.

    I’m assuming the fluid may be something like glycol/water. The rocks would be engineered and sterilized. Big rocks tend to have limited heat transfer potential. Surface area is key. After that it may work.

    • Bob_Wallace

      Gravel. Lots of surface area.

      Argon (according to previous reports). They use a gas, not water.

      • Gravel doesn’t have much surface area compared to internal volume, if that is important for the process to work. The system would need a lot of residence time (the time fluid is in contact with the media) between solid and fluid.

        On a weight/volume basis – the smaller the particle, the greater the surface area, i.e. clay more than sand, sand more than gravel. On the other hand, gravel increases permeability or the ability for the fluid to flow, compared to smaller particles. It seems that’s why they may be using gravel. Permeability (pressure loss) is why refinery operations like absorbers use engineered media to maximize the area available for heat or mass transfer. This would be beds packed with ceramic saddle packed media. It allows for both efficient heat transfer and fluid flow.

        Gas has a much lower heat capacity than liquid. This may not be important. However the interchange of heat may be the critical mechanism. The goal is to get the solid at a consistent temperature, since they are kind of promoting isentropy.

        • Matt

          Water is a non-compressible fluid. While 12 atm pressure is not where the are storage the power. They want the temp change they get from compressing/expanding the gas. Use or gravel is likely because it is cheap and you can get it where ever you build this. Using a material with phase change in the 300-500 range on the hot side and -100 to -160 on the cold side, might sound tempting. But but that adds to cost. Plus you have to worry about a fluid moving into your pumps. This way the gravel stays in its tank. And expansion is a much smaller issue.

          • Water’s incompressible and wet, I believe. Thanks.

            OK, I just watched the video. Wow! This is not an inexpensive system. Or simple. Or mobile.

            The issue is not gravel as indicating in the headline.

            Let’s assume the video is to scale. That 12 bar tank would have to be stainless and well engineered. Let’s assume its about 200,000 gallons (by eyeballing the workers for scale) – The purchased cost may range around $250,000 to $300,000. This doesn’t include insulation, etc. The other tank is at 1 atm, but needs to handle cold. Let’s say $75,000 to $150,000.

            Those are pretty complex pumps. Let’s say it’s a skid unit of four (or three). That’s about $200,000 for the turnkey skid. I’m kind of thinking of a bunch of skid fluid handling systems and lumping them all together.

            The thing needs to be pulled together with engineering, construction management, financing, site install, piping, wiring, controls, housing, power feed and whatnot – The installed cost could be about $10,000,000 ranging from $5,000,000 to $20,000,000. (this is way SWAG and only to compare gravel cost.

            They’ll need about 2,000 cubic yards (2,500 tons) of gravel – give or take – at a delivered cost of $25 per ton – that’s about $50,000 for gravel. Gravel is like a rounding error.

            Since solid media is key to the success of this thing, I really don’t see skimping on heat storage and transfer media.

            The fluid is going to have to be pretty fancy and thermodynamically and mechanically stable across a wide temperature range.

          • Doug Cutler

            Apparently Isentropic now has brand new third party testing from a very well established engineering firm Parsons Brinckerhoff:


            Here’s an excerpt:

            “In the report’s detailed example, the cost for a plant with six hours of storage is estimated to be $618/kW, giving a breakthrough per hour storage cost of $103/kWh. The report also estimates a round-trip efficiency in excess of 90%.”

            I’m not clear how the cited figure of $103/kWh relates to Isentropics’s own touted target of $35/MWh or even if its a units error of the kind I routinely make. (Musician here.)

            I also wonder the scale of the system upon which Parsons Brinckerhoff did their analysis or if they are already extrapolating to the kind of scale system under construction.

            I once emailed Isentropic to inquire if their system could ever be scaled for household use. Apparently no, they need the large scale heat storage to achieve their efficiencies.

          • At this stage of development, cost estimating is order of magnitude.

            Here’s a really cool report by dept. of energy on storage technologies: status and research.


            It looks like the range of grid level storage batteries go from 10s of KW to 10s of MWs. Given the size of this thing, I’ll assume something around 20 MW (20,000KW) or $618/KW * 20,000KW = $12,360,000. The $618 from your comment above.

            This is the installed cost. It seems like a cost range should be used until further development. As long as it’s in the range of other flow battery technologies for grid scale.

            Anyway, since other flow battery groups are still in the early phases of development, maybe they should not worry too much about cost, and more on performance demonstration.

            It sounds like they’ve done some engineering feasibility work (the P-B engineer). It’s probably best to do some harder design, process simulation modeling (there are canned software packages for this), and a small pilot test.

            Anyway, I’m totally curious about this and other flow batteries. Good stuff.

          • A Real Libertarian

            I’m not clear how the cited figure of $103/kWh relates to Isentropics’s own touted target of $35/MWh or even if its a units error of the kind I routinely make. (Musician here.)

            $103/kWh is the price of storage capacity, $35/MWh is the price of stored electricity.

            $103/kWh of capacity x 1,000 = $103,000/MWh of capacity.

            $103,000/MWh of capacity ÷ $35/MWh of stored electricity = 2,943 charges before replacement.

        • JamesWimberley

          Scientists have designed microporous materials with surface areas orders of magnitude higher (link) than lumps of gravel. Of course, they are also orders of magnitude more expensive.

          • Microporous materials, like granular activated carbon (GAC) and catalyst substrate aren’t really necessary for this application of heat transfer. For instance, GAC has about a football field of surface area compared to the same mass of gravel. Same with packed beds for catalytic converters. The packed bed would be simple ceramic shapes – nothing too fancy. Maybe assume the price would triple. Still not a dent in the overall capital cost.

            Research engineers do packed beds and flow through porous media more than scientists. Scientists don’t have the math to model transport phenomena (that was an engineer v. scientist slight, if it wasn’t obvious)

            Anyway, this thing is cool. The post text could have been a bit more investigative and thorough. Unless, uninformed, but enthusiastic, investors are the best investors. Which is why snake oil gets sold so well.

          • Bob_Wallace

            When you talk about gravel, remember that one option is ~1/4″ or smaller pea gravel. Pretty small stuff but still large enough to keep from clogging.

          • Good point, but I’m only talking about gravel since it was talked about in the original post. I truly thought, after a quick skim, this was much simpler. After watching the video, there’s some bells and whistles.

            The reason media or gravel or packing is important is the process relies on the establishment of thermal layers. Any inconsistency in packing or preferential flow paths, the system may breakdown. I’ve done enough bench, pilot and full scale work to have realized it’s always the simple things that can be a problem. Fixable. But lends to headaches and late nights.

          • Bob_Wallace

            ” Any inconsistency in packing or preferential flow paths, the system may breakdown”


            Wonder if they could put grids of highly heat conductive material in between layers of gravel in order to spread the heat. Even make them perforated tubes to create horizontal pathways.

  • mds

    Those are some big pressure tanks, a lot of steel. Another very high cycle-life energy storage technique for storage of solar to be used in the evening and at night. This does not have the ability to store power seasonally like pumped hydro. Their competition will be EOS, Aquion, Ambri, Tesla Li, and others looking to provide solar power in the evening (neck of the duck curve) and throughout the night. These companies are targeting similar and even lower cost than $35/MWh = 3.5c/kWh. Nice to see more competition and nice solution here. Elegantly simple.

    Seasonal storage of solar and wind still needs low-cost solutions, but this is like range-extension in EREVs. It can be solved for now by using conventional generation. Most of the world’s population lives closer to the equator where seasonal power storage is not really an issue.

    • Bob_Wallace

      I’m just not seeing a reason for seasonal storage.

      We may find it desirable to have some >”>5 but <8" day storage. (Not sure what the 5 and 8 numbers should be.) Liquid fuel for turbines might be best for that. Cheap long term storage.

      Most battery storage is more expensive. Ambri's liquid metal battery may be in the PuHS price range but it's too early to tell.

      • JamesWimberley

        The only cheap method of seasonal storage is biological. Grow sugarcane or something in the summer, and burn or ferment it in the winter. But I agree with Bob that we will almost certainly work round the seasonal issue simply by overbuilding wind and solar. In 15-20 years, which is when these issues will become operational in the USA, both will be below 2c/kwh LCOE, so massive overbuild is affordable.

        The Germans (one of the Fraunhofer institutes) see a need in a 100% renewables scenario for 2 weeks’ cover for windless and sunless periods in winter. This looks about right. It may be cheapest, for a given carbon bang, to leave the last 10% of electricity from natural gas and offset with large-scale sequestration.

        • Bob_Wallace

          Overbuilding – some people don’t realize that we greatly overbuild current/traditional generation. In 2011 and 2012 coal had a CF of less than 60%. Over 40% coal plants were not in use. Natural gas, even more so. NG had a CF <30%. More than 70% of the time gas plants weren't running.

          Overbuilding cheap (< 4c) wind and solar is likely to pencil out very nicely as opposed to long term storage.

          Biomass, perhaps, But we're likely to end up with a lot of paid off gas plants. Cheap to leave sitting idle most of the time.

      • Guest

        We certainly need multi-day (or a week of) storage for northern climates, because there is not enough solar insolation in winter and wind does not come regularly every day.

        The need for multi-day or a week of storage in winter is very realistic for most of European countries.

        • Bob_Wallace

          Multi-day is not seasonal.

          • Grad

            Well maybe I was not very clear. Storage must provide multi-day (or a week) of energy. Because every winter there is a week when wind isn’t blowing.

            This means that storage most probably has to be seasonal, because surplus is spreadout through whole year.

          • JamesWimberley

            No. Look at the ERCOT charts of wind output in Texas in this recent post (link). The peaks and troughs follow a cycle of less than a week, corresponding to the movement of anticyclonic weather systems from W to E across the globe. The cycle is sometimes interrupted by high-pressure systems that can stay in place for a fortnight but no more. Your proposition only holds for solar. Winter storage will be of wind energy.

          • eveee

            Yes. Midwest has strong winter winds, less so midsummer. California is the opposite. Offshore winds are also different. Besides weather predictions, the seasonal and diurnal natures are predictable. Its all about probabilities. Its not a random source.

          • Bob_Wallace

            I’ve mostly seen people using “seasonal” in terms of storing electricity from summer solar for winter use. I really can’t see that happening.

            Let me show you an interesting graph. It’s from a paper by Budischak, et al. – the link is in the “Is 100% Renewable Possible” section on the upper right side of the page.

            The top line is four years of wind and solar output – real world – for the largest wholesale grid in the US, the PJM that serves all or part of 16 NE US states. The Budischak group was answering the question of whether it would be possible/feasible to supply a major grid with nothing but wind and solar. They used wind and solar data from NOAA and actual minute to minute demand from the grid.

            The second line down is when they had to dip into storage. At least in that part of the country there’s not a lot of need to pull in a lot of stored power during the winter. There might have been days, or even a week, when the wind did’t blow in some part of the grid coverage area but it’s pretty clear that there was either plenty of solar or the wind was blowing in a different part.

            When they ran the shortest was in the summer. Look at the red spikes. That’s when they found it cheapest to use a bit of NG rather than install enough wind/solar/storage to meet those rare (~7 hours per year) demands.

            Now they didn’t include hydro, load shifting or power trades with adjacent grids so it could be (it probably is the case) that much less storage is needed than what they needed for a wind and solar only grid.

          • Matt

            Hear, hear. Over build (as is done today with fossil) is the way it will be handle for while. Today the biggest “storage market” is load shifting within a 24 hour period. Multi-day storage is still small (those few or grid people), I expect this to grow a lot as micro and mini grids spring up. Both in the developed world (hospital campus, base); any portion of the grid that want to know it will always have power. And in developing work, all those little villages that are current off grid or on a portion of the grid that isn’t stable. All the talk here of cost is a little off. You can cost the $/MW (capacity) but to price a $/kWh you need # cycles it will last and #cycles/week it will be used (location dependent). If you only cycle it once the $/kWh is much higher. Which is why true seasonal storage (summer to winter) is cost so much more. The limit on this design looks to be time, the argon gas does not react with the gravel; not # of cycles (like a chemical battery). I say time since a those pressure tanks (likely biggest cost) will not last 10k years.

          • Bob_Wallace

            I want:

            A program that would let the user model a ~100% grid using real world wind/solar/demand data like Budischak used.

            Allowed the amount and cost of wind, solar and storage to be varied so that demand was met and the cost of electricity determined.

            Budischak used prices for wind and solar in 2030 that we’ve already dropped below (at least for solar). We’re seeing new storage being developed. It would be fun to put in the most recent costs (even hypothetical costs) and see how the mix and overall cost of electricity would change.

            BTW, I just found that the link to the Budischak paper isn’t in the “100%” section. I’ll post it here for anyone who wants to give it a read….


          • Bob_Wallace

            Let’s assume a thin stainless steel tank surrounded by a thicker concrete shell.

            Exterior foam insulation and housed in a building that protected from the weather.

            Wonder what the expected tank lifetime would be?

            Insulation might have to be replaced as well as the building skin, but the tanks should go on and on and on?

  • Guest

    Big claims there.

    Any real world demonstration project up and running?

    • mds

      See Doug Cutler’s comment.

  • Doug Cutler

    Isentropic is in the middle of a trial scale build aided by a $20M UK grant. We presume somebody looked into it pretty carefully and saw the potential but it will still be two or three years before we know if the numbers bear out at scale.

    The touted cost of $35/wmh is crazy low and way below any of the other battery grid storage initiatives. We will have to see. Worth keeping an eye on though.

    • Bob_Wallace

      Seems like we should know sooner. How long does it take to build a heat pump? A couple of insulated rock chamber shouldn’t take long. A month’s worth of running should determine efficiency. Scaling up numbers is just math.

      If we don’t hear info in six months or so then I’d say this one won’t pan out.

      • Doug Cutler

        Yes, why should it take so long for proof of concept on such a simple system? Whatever, someone in UK government saw clear to ante up a big chunk of change.

        I’ve had a little email chat with an Isentropic executive complaining about their lack of news. Seems they’re happy to keep a low profile until the current scale build is completed and they have some solid real world numbers. There’s an executive BBC radio interview in the news section of the Isentropic website with a little more info on the build:

        • Guest

          And when is current scale build supposed to be completed?

          • Doug Cutler

            Either 2016 or 2017. I don’t remember exactly. Its buried in a radio interview with executive James Mcnaghten listed on the Isentropic site under “News”.

            BUT . . . same Isentropic news page is now showing a very recent third party verification. See my second reply to Bob Wallace.

        • This is not a simple system. The fluid needs to be thermo and mechanically stable over a wide temp range. The vessels are big and complex. And the process is simple only in concept – until the bugs are worked out. It will probably need to go from bench to small pilot to full scale pilot to market ready – with a lot of testing. And that takes time. I’m sure they need investors to get it ready.

          I like it, nonetheless. It’s just not a couple of beds of gravel with fluid passing through – which kind of was what was described. There’s a good mix of chemical engineering, mechanical and electrical involved.

      • Doug Cutler

        Here’s an update: looking at Isentropic’s news section they now appear to have independent corroboration of their claims. Now claiming 30% cost of pumped hydro. Maybe its overreach or scam. I hardly have the expertise to tell. Here’s the link:

    • mds

      Minor correction: $35/MWh for better clarity. Thanks for that info!

      • Doug Cutler

        Right you are. Thanks

Back to Top ↑