Isentropic - Storing Energy In Gravel

One of the outstanding problems in energy provision is how to store it in such a way that it can be accessed rapidly and efficiently on demand. Now a Cambridge based company,...
07 December 2010

Interview with 

Jonathan Howes, Isentropic


Chris -  Now one of the outstanding problems in energy provision is how to store it in such a way that the energy can be accessed rapidly and efficiently on demand - in other words, when you want it.  But now a Cambridge based company called 'Isentropic' has developed a cutting edge solution using gravel pits as enormous batteries where they store energy thermally.  To explain how it works, Technical Director Jonathan Howes is with us.  Hello, Jonathan.

Jonathan -   Hello, Chris.

Chris -   So first of all Jonathan, tell us what actually is the problem that you're grappling with here?

Jonathan Howes, Technical Director, IsentropicJonathan -   This came about as a result of some thinking I was doing in the mid to late 1990s when, like many engineers, I was quite keen on clean power generation methods.  The more I looked at it though, the more I realised that we were actually trying to solve the wrong problem because there were plenty of people working on clean tech ways of making power.  But the big problem with all of these is that they're either very inflexible, like for example nuclear, or very extremely erratic such as wind or solar.  Tidal is less erratic, but it still comes and goes with the tides inevitably.  So, the more important problem to solve was to buffer energy so that the intermittency problem or the inflexibility of nuclear would then go away.  This would actually enable the widespread adoption of these other technologies.  The real concern for me was always sustainability rather than clean tech because the opposite of sustainability is unsustainability which is self-evidently a silly way to go.  So, this struck me as being the bigger problem.  In solving the issues of making sustainability achievable, you inevitably end up cleaning power supply anyway because sustainable almost by definition means clean.

Chris -    To put this into perspective and to give an everyday context to it, one of the big problems that power supply companies grapple with is Eastenders finishes or some other hot television program finishes, everyone gets up, whips on the kettle, turns the boiler and the lights on, and so on, and there's a big surge.  You've got to have capacity there to cope with that and that means you've got to run stations beyond demand in order to have the spare capacity in the system to meet that demand.

Jonathan -   Yes.  This is the load levelling problem which is the definition of the short term storage issue.  There are two primary storage problems.  One is long term or seasonal storage.  Perhaps storing energy during the summer when it's plentiful for use in the winter.  The other one is the short term load levelling as you've just described for the local peaks and troughs in demand.

Chris -   So tell us about Isentropic's solution then.  What are you doing and how does it work?

Jonathan -   Okay.  What we are doing - this is a thermodynamic approach.  If you consider a heat engine, as you'll find under the bonnet of a car, it produces a temperature difference, usually by burning a fuel.  They don't have to do it by burning a fuel, you can use solar energy to create the temperature difference.  That temperature difference then passes through a device and produces mechanical power.

Isentropic schematicNow there's an opposite of this called a 'heat pump' which is again, well-known but I think perhaps it's not quite so widely appreciated that heat pumps are in fact exact inverse of an engine.  And a heat pump takes mechanical power through the shaft and converts it into a temperature difference by pumping heat from one place to another.  If you can make a heat pump sufficiently reversible, thermodynamically reversible, you have the possibility of using it to pump heat from one volume of thermal mass to another volume of thermal mass, and then allowing it to discharge back through a similar or the same device, back to its original state, and then releasing the power that was used to create the temperature difference.

Chris -   So, to explain practically how this might work, I would consume electricity from the grid.  I would use this to do some work, moving something from one place to another place where there'll be a temperature difference which I can store.  And then when I want the energy back, I can move the thing from the place where it's at one temperature to the other temperature, getting the energy that I used in the process back out - at some degree of loss obviously, there's no such thing as 100% efficiency yet - and that means that then you've stored the energy somewhere, until such time that you'll need it, but you can recover it from that thermal difference.

The upper reservoir (Llyn Stwlan) and dam of the Ffestiniog Pumped Storage Scheme in north Wales.Jonathan -   That's exactly correct.  It's exactly analogous to pumped hydro where water is pumped from one level and one lake to another lake, and then it's consuming energy, and then allow it to flow back down again through a turbine to release the energy again.  Where they pump water up and down a hill, so altitude represents temperature, we pump heat up a thermal hill to a high temperature reservoir.

Chris -   And what are you heating up?  What are you cooling down to make your thermal difference?

Jonathan -   We have an engine circuit which runs an inert gas, namely argon, through a compression and expansion process.  If argon is compressed, or any gas is compressed, the temperature rises.  Having raised the temperature of the gas, it's then passed through a particulate store, a mineral particulate store which in its simplest form could be regarded as gravel.  Although of course, you could use a ceramic or something more sophisticated, there's every reason not to for reasons of cost.  It leaves the heat behind in the particulate and that cools back down to its original temperature, but remains at the higher pressure.  If you then expand it back to its original pressure, then the temperature now falls to significantly below ambient, and it's passed through another store, leaving the cold behind in that store.  If the heat exchange takes place efficiently and slowly, it can be highly reversible.  If the compression and expansion take place in very, very well designed insulated spaces at high speed, that can also be highly reversible.  So what we do is string four highly reversible processes together so we can now run it backwards as an engine, allowing the heat to run back through the machine, releasing a mechanical power to drive a generator to get the electricity back again.

Chris -   What sort of efficiency can you manage?

Jonathan -   Okay.  Round-trip efficiencies, based on our earliest testing, we estimated with fairly cautious assumptions that we could achieve about 72 per cent.  With more realistic assumptions, we're in the range of about 75-80 per cent.

Chris -   And in terms of space, how would this compare with the example you gave earlier which is pumping water from a low-lying lake to higher-lying lake so that you can then recover that energy by running the water back down to turbines later?  How much energy can they make and how much space do they take up compared with you?

Jonathan -   They are typically fairly sizeable plants because if you've got a suitable mountain range or a range of hills and an appropriate valley to use, it tends to get used to the limit.  So, they tend to be quite large capacity plants.  Bath County in Virginia in the USA for example has a 30-gigawatt storage capacity.  If we did our equipments with the same 30-gigawatt capacity, it would consume approximately 1/300th of the land area.  Another way of looking at it, we scoped out a prototype utility scale machine of 2 megawatts power capacity and 16 megawatts hours storage capability.  That consumes a footprint of about 8 metres x 16 metres x 7 metres high.

Chris -   So very, very small and efficient.  You've got the prototypes running, but when can we see this plugging into the grid, and giving us a surge capacity and long term storage that we need for the future?

Jonathan -   We are initiating the design of the utility storage demonstrated prototype which is going to take around about 2 to 2 ½ years.



Add a comment