Placeholder Image

Subtitles section Play video

  • Okay, so thank you for that introduction, it's a real pleasure to be here and

  • be able to give you an update on our GCEP project on Solid Oxide Flow Batteries for

  • Grid-Energy Storage.

  • A couple years back, we, we gave a talk as well on the cell material

  • advancements we've been working on.

  • In today's talk we're going to focus more on the system concepts,

  • that will hopefully enable the technology to move it forward.

  • Before moving into that I just want to acknowledge our team members

  • PhD student Chris Wendel, Professor Bob Kee at the School of Mines.

  • Professor Scott Barnett and, and Doctors Gareth Hughes and, and

  • Zhan Gao at Northwestern are really working at advancing the Cell Technology.

  • So, in today's talk I'm going to briefly give you an overview of what, exactly,

  • is this technology.

  • And followed by, with I guess, I would say our view of some of the motivation and

  • the technology requirements that are needed for energy storage to move forward.

  • I'll then move in to some descriptions of reversible solid

  • oxide cells as flow batteries.

  • Here we'll look at, little bit at the theory of operation and

  • performance considerations.

  • As well as some performance estimates of really these large scale, megawatt size,

  • gigawatt hours, capacity systems that we would envision for, for bulk storage.

  • Brief, I'll give a brief update of some of the exciting developments in the South

  • development area where we're, we're really trying to push towards the 600 in sea

  • operation salutes and all this GM technology.

  • And we have some very interesting and encouraging results related to,

  • to cycling to show of these cells.

  • And that's very important how we're going to operate forward and

  • backward modes with this technology.

  • We don't want degradation there.

  • Lastly we'll, we'll touch on some of the economic projections for

  • these kind of large scale bulk energy storage systems.

  • I'll then briefly touch a little bit on what we've learned, in,

  • in featured reactions.

  • So in principle, a solid oxide flow battery really leverages

  • similarities to fuel cells where we're going to operate reversibly.

  • Here reversibly is not in the thermodynamic sense.

  • It's in the sense of reversing the current for the system to operate in a power

  • producing mode, and in an electrolysis or charging mode.

  • And we're, we're going to tank the reactants and, and

  • capture those in, in gaseous storage, and that's particular useful for

  • us because it gives us really the flow battery advantage.

  • We get the decouple power capacity from storage.

  • And so the power will scale with the size of the cell stack and

  • the energies will scale with the size of the storage tanks.

  • We also get the high efficiency advantage of solid oxide cell technology.

  • Which enables us to have really high round-trip

  • efficiencies as we move between modes.

  • We don't experience high polarization in electrolysis mode.

  • And the novel, relatively novel HCO chemistry that is experienced directly

  • within the cell, allows us to, to produce high energy dense fuels.

  • So shown here is a, is a, a real simple schematic of,

  • of a solid oxide cell and oxygen conducting one with some fuel storage.

  • Here we are showing methane and syngas and we are going to feed it with air, and

  • we're going to take the oxygen from there, reduce it, get those anions.

  • Moving and electrochemically oxidized those gaseous reactants into H2O and CO2.

  • We will capture that gas in a tank and essentially produce our power.

  • Now, in reverse mode, we can then accept ply voltage,

  • driver currents, and essentially put our power into the, the device.

  • And then move into the opposite mode, where we'll remove those previous

  • products of reaction out of storage back to our cell.

  • We'll strip out the oxygen.

  • Liberate some of that oxygen.

  • And in the meantime, directly within the cell, we will produce methane and, and

  • syngas.

  • In general, that'll give us favorable scaling this device, but

  • also something additionally unique is that it gives us really low-cost working

  • fluids compared to advanced, and other types of flow batteries.

  • In terms of motivation, certainly, the variability of renewable energy resources

  • is is well-known and motivates developing grid-energy solutions.

  • I like to at least see some picture of what that means?

  • Here are some minute by minute data shown from Hawaiian Electro Power

  • on a wind farm.

  • We can see really a ten exchanged within 30 minutes of, of power requirements.

  • And it's not just wind variability if we look at developing activities and

  • concentrating solar power and of course, PV penetration,

  • they've got power fall-off in, in the evening hours as well that would,

  • will need to be addressed to get high-capacity factors.

  • So currently there is no battery technology that really serves

  • most of our energy storage.

  • Worldwide is predominantly pumped hydro.

  • In, that's, but this problem still exists, and those who are facing this,

  • primarily, often island nations for example, are,

  • are already trying to develop solutions, and I'll call them poor solutions.

  • Taking high grade electrical energy, and storing it in low-grade hot water for

  • example, so called thermal battery.

  • That's being done by Hawaiian Electro Power to, to manage this,

  • this variability.

  • It's also being done in an Electricity Arbitrage Models in Minnesota,

  • for example.

  • I've called them the, the dubious honor of having the largest thermal battery in

  • perhaps in the country at one-gigawatt hours.

  • High-grade electrical energy,

  • low-grade hot water is essentially thermodynamic syn.

  • But in, on the other hand you know,

  • good economics doesn't necessarily always mean good thermodynamics.

  • In general,

  • though, in order to enable that technology we've got to reach some certain targets.

  • We've been keeping our eye on these as we look at this technology.

  • Certainly capital cost and round-trip efficiency but perhaps most importantly

  • some levelized cost of electricity storage around a dime per kilowatt hour cycle.

  • We need cycle capability, and depending on the application,

  • you'll need various modes various duration of storage.

  • If we now turn to looking at the technology itself.

  • Just operationally, we can take a look at a voltage current

  • plot which is a representation of the cells performance characteristic.

  • And shown here, we can see that in power producing mode or fuel cell mode

  • the voltage will decrease as you increase the current density or, or produce more

  • power in response to, to overpotentials and irreversibilities within the cell.

  • The slope of this curve represent the overall resistance.

  • In fuel cell mode, the higher the voltage the higher the efficiency.

  • And in electrolysis mode we can see a rel, a relatively smooth transition shown here

  • in this cartoon, but that's actually what we see experimentally as well.

  • There isn't a large over potential that gives us good electrolysis efficiencies

  • low-applied voltage needed there.

  • But here you want low voltage equals high electricity in electrolysis mode.

  • So, if we look at the round-trip stack efficiency, which is not shown here, okay.

  • It's basically the voltage of the fuel cell divided by the voltage

  • of the electrolysis device.

  • That's the ratio.

  • So you want high fuel cell voltage, low electrolysis voltage,

  • that will give you a high round-trip efficiency.

  • At the system level we not only need to be mindful of the stack but.

  • We're moving these reactants back and forth between the tank and the stack and

  • so there's an auxiliary power component that enters into this ratio.

  • So, in the end, how we can improve system efficiency.

  • We can improve the cell by reducing over potential and at the system level,

  • we got to mindful of the balance of plant and thermal management.

  • And when we look at thermal management,

  • one of the unique attributes here is by doing methanation locally,

  • within the cell and electrolysis mode, we are able to obtain low

  • electrolysis voltages, get towards a thermal neutral operation, as well.

  • So, when we look at a fuel cell, it requires heat rejection.

  • We're air cooled we're operating it at relatively high temperatures.

  • But in electrolysis this is of course and endothermic process.

  • It requires a heat source.

  • As, and we can see that when we reduce H2O that's certainly the case.

  • We're going to leverage HCO chemistry here and because of the nickel in, in,

  • in the fuel electrode we can also do heterogeneous chemical reactions and

  • reduce CO2 as well through H2 and

  • provide us with some CO which can then be combined with hydrogen to methanate.

  • Which is highly exothermic, okay?

  • And that's very nice for us.

  • Because we have a exothermic local source where we have an endothermic process.

  • We've got good matching of sources and sinks there.

  • And ultimately, low temperature is what we would want in relatively high pressure

  • to achieve that methanation.

  • One of the considerations we're faced with as well is.

  • If we're going to design one of these systems, what do we charge the tanks with?

  • What is the composition we want?

  • And what are the considerations therein?

  • So in these systems, we have to be concerned about carbon deposition.

  • This is a deleterious effect on, on, on, solid oxide cells.

  • And it degrades their performance rapidly should that happen.

  • So shown here is in the right is essentially

  • a compositional space used in a so called Gibs diagram or ternary diagram.

  • Where the shaded area above the rad indicates the thermodynamically

  • favorable region for carbon deposition to occur.

  • And the, the open, the white zone really is is unfavorable for

  • that, and that's where we want to operate.

  • So in doing so, you can see the red dot up here is where we might start on

  • a hydrogen carbon ratio oxygen ratio for, for fuel cell mode.

  • As we oxidize the fuel, we'll move towards this fully oxidized region shown in

  • the light blue, and we don't really want to be fully oxidized.

  • In this system, we want to be not fully oxidized and not fully reduced.

  • This is our operating window, if you will, to move back and forth.

  • If we look at the bottom graph, we can see basically on the left hand side

  • equilibrium gas constitution on a molar basis.

  • It's, it's a wet basis shown here versus oxygen content.

  • And we can basically move back and forth between, shown here between 4 and

  • 40% oxygen conversion, which will allow us to have fairly high storage capacity.

  • We can produce methane in a 60-40 ratio with hydrogen here, on a dry basis.

  • And at this end of the cell, so basically as you produce a fuel cell mode,

  • you'll see us reducing the CH4, producing H2O.

  • These, of course, will be tanked for electrolysis mode.

  • So one of the proposed applications we've been looking at is really bulk storage.

  • To, in order to get there, we need very large tanks.

  • And very large tanks can be realized with pressurized underground

  • gaseous storage of our reactants.

  • Using salt caverns, for example, natural gas reservoirs, saline aquifers.

  • And so we're, this concept is actually being very seriously considered.

  • Particularly in Europe in, in, in Germany.

  • And we're looking, in collaboration with the Danish Technical University, at,

  • at designing this so-called surface system which will, will convert and

  • store our reactants.

  • Using, survey data on natural gas, reservoirs in Denmark, for example.

  • We can estimate 500 gigawatt hours of storage would be available for

  • one plant that has a 250 megawatt capacity.

  • And the punch line here is, we'll get to this later.

  • But in the end, these storage costs can reach $0.03 to

  • $0.04 per kilowatt hour with storage durations of months.

  • Which is particularly important.

  • Germany in particular, very interested in month long duration storage.

  • Because of the low PV insulation during the winter in particular.

  • Because we produce methane we find it really interesting if the technology is

  • also suitable to support the so called power to gas platforms.

  • They're very of, of increasing interest and particularly by Europe and

  • getting off of Russian natural gas and using renewable green electricity,

  • if you will, to make SNG.

  • This technology is perfectly applicable to that.

  • In the end, though, we need this top surface system.

  • And that involves systems integration and thermal management strategies

  • in moving essentially, between the caverns and the stack.

  • And so, just very briefly.

  • We have to pressurize and preheat the reactants that get over to the stack.

  • We can recover some of that energy from fuel cell exothermic operation

  • to reduce our balance of plant parasitics.

  • From the cavern and we'll take our CH4, preheat it, and

  • expand it because it's operating at say 160 bar and the stack is at 20 bar.

  • We'll recuperate some power, and we'll introduce steam and

  • use the tail gas, if you will, of that

  • process to meet our process heating needs before dumping it into the CO2 cavern.

  • We'll get DC power out and we go to electrolysis mode,

  • we basically reverse and store in the CO2 cavern, store in the CH4 caverns.

  • Importantly, in order to make this viable we want to use the same equipment.

  • Okay, so, that means they have to be sized and operated and

  • designed such that that can be done.

  • We also have to carefully manage water in these systems.

  • We're going to knock it out and

  • generate it because we can't really easily store it in these caverns and extract it.

  • When we look at performance trades, clearly a key issue is what pressure and

  • temperature should this stack be operated at?

  • One of the things we like about this project is we get cell material

  • development, we get systems aspects going on, and

  • the two get to talk to each other we can say from a systems view.

  • I don't really need very low temperature or I need a different pressure for

  • you guys that focus on perhaps, depending on the application.

  • So here we show a plot of, of roundtrip efficiency for the stack and the system,

  • we'll just focus on that verses stack pressure.

  • And we see an optima is here, and

  • that optima is basically is the interplay between the,

  • the auxiliary power, depending on what the stack pressure is.

  • So if the stack pressure is relatively low,

  • we can get net power out of our system in fuel cell mode.

  • And that can offset our electrolysis pumping requirements.

  • In the end, that interplay gives us an optimum of around 20 bar which we like,

  • because that matches a lot of the high pressure turbine spools that are available

  • that might be integrated with the system.

  • Similar trades are present when we look at temperature and reacted utilization.

  • And, and those optima are shown here.

  • If we just quickly move into now looking at some of the cell

  • technology advancements that, that have been ongoing with this project.

  • We're really focused on these next generation material sets leveraging

  • really LSGM technology to push toward 600 C and with high cycle durability.

  • Briefly, here's an SEM image of the microstructure of one of the cells.

  • And you could see the thick LSGM electrolyte layer.

  • The dense layer that's, that's right here.

  • Overall the, the sum of these layers is quite thin.

  • But you can see here there is on, on, on your,

  • your air electrode, we have our gas diffusion support.

  • It's LSF.

  • We have a nickel infiltrated LSGM fuel electrode that allows us to get high

  • current densities for high triple phase boundary area, if you will.

  • This is on an SLT support which gives it strength and

  • one of it, the unique pieces of this is, is the nano-particle nickel infiltration.

  • In the fuel electrode.

  • If we look at the performance characteristics,

  • we can get the high performance.

  • High performance here demonstrates at the power density of 1.6 watts

  • per square centimeter at 650 C.

  • As far as we know, that's, that's one of the records.

  • It works in both modes very well.

  • The area specific resistance is .18.

  • We've been targeting 0.2 Ohm's square centimeter for the system.

  • And we've demonstrated that at the, at, at really button cell level.

  • We have to do better on the 600C

  • polarization curve if you will that's getting slightly higher.

  • And we still need better performance there.

  • But most interestingly,

  • I think one of the tests that we've been running is on cycle durability.

  • We, we need to cycle these things forward and backward and

  • no one has really tested this kind of technology in this mode.

  • So we've looked here at really one and 12 hour cycles.

  • You can see a 30 minute operation on one mode, 30 minute operation on the other,

  • switching back and forth between these modes.

  • For different cycle times.

  • So here's is a one hour show, but we've also done 12 hour cycles in as well.

  • So six hours in one mode,

  • six hours in the other mode at different operating current densities.

  • And what you'll notice here is on this light blue curve,

  • if you're just operating electrolysis mode, you get fairly rapid degradation.

  • But as we change into cyclic mode,

  • we get reduced degradation as exhibited by the change in total resistance over time.

  • And we've tested this for 1000 hours, and as you can see, as,

  • once you get below a certain threshold, operating current density,

  • the degradation mechanisms have turned off, essentially, or interrupted.

  • And we find that that actually happens around .8 amps per square centimeter.

  • Which is at least twice as high, or about twice as high as we think is

  • economically needed, to develop the technology.

  • So we're really encouraged, by these results in particular.

  • In the remaining minutes, I'd like to give you a little snapshot of the economics,

  • when we first presented a couple of years back.

  • There are a lot of questions on that.

  • We had no data.

  • I can report some data.

  • On this, at this time.

  • And, that's unfortunate.

  • This okay I'm in IBM PC and, and these equations aren't showing up.

  • But.

  • What I would say is briefly there is a simple calculation that basically takes

  • the it takes the investment cost and

  • divides by the energy storage in a round trip efficiency in the number of cycles.

  • And you get essentially a simple storage cost metric.

  • In sense for kilowatt hour.

  • The challenge with this method is,

  • it assumes a hundred percent capacity factor, in doing so.

  • In order to, be able to perform this, we need to cost out the plant.

  • So we've done some bottom up plant costing using some of these parameter values.

  • Here briefly, highlighted here.

  • 250 megawatt rating.

  • We've shown we can get higher round trip efficiency but we just put in 70% here.

  • Mature life projections for solid oxide cell technology.

  • Again, we're using costs from solid oxide fuel cells that are very applicable here.

  • But perhaps not exactly applicable depending on the cell material sets.

  • The storage there's a fair amount of good data here.

  • We've been leveraging existing natural gas reservoir data from Lille Torup facility

  • in Denmark, 120 million cubic meter natural gas reservoir facility.

  • We make use of 70 million cubic meters of that.

  • We need a 50 million cubic meter push in gas to support that activity.

  • And we've priced out that cost based on the existing cost that we know for

  • that, that.

  • And we've extrapolated for CO2 caverns.

  • There that's relatively unproven storage CO2.

  • We've essentially taken CH4 costs and more or less double them for the risk.

  • In the end, we get a capital cost at this scale of around,

  • less than $1,100 per kilowatt.

  • If you look at the total expense breakdown up here, it's not just capital costs.

  • We got operating maintenance cost here and staff and so forth to operate.

  • But in the end we're about 30% on, on the stack.

  • In less than 15% in the storage.

  • With this simple costing method then allows us to get us around $0.03

  • per kilowatt hour on storage costs with this method, which if you look,

  • compare favorably against compressed air, air, hydrogen and,

  • and pumped hydro in these other bulk storage categories.

  • We think that's, perhaps a little too simple, and, more we can leverage instead,

  • the resources of this storage facility, using electricity,

  • spot market prices and essentially, using supply and demand characteristics.

  • Of the grid market and

  • do essentially market arbitrage to buy and sell power essentially.

  • Buy power cheap, charge your system and

  • sell it when the price of electricity is high.

  • So the cautionary note here is.

  • In making these calculations, of course, we knew what the prices were.

  • It's historic prices and we could optimize the sell-buy strategy

  • which then means this is really a maximum annual income estimate, okay?

  • So if we look at 2008 electricity spot mark, market prices,

  • our colleagues in DTU really performed the study.

  • They use the Danish market because that's what they were interested at the time with

  • our system.

  • And we don't get a capacity factor of 100% in this scenario, we get 61%.

  • When you look at the life cycle cost,

  • that raises it from almost $0.03 to almost $0.08.

  • But you do get revenue from this and you can drop that by 4% to a net

  • overall storage cost of just under $0.04 at $0.037 per kilowatt hour.

  • There is lots of considerations that in the future, increasing

  • renewable energy penetration will mean higher electricity price volatility.

  • And you could essentially do more arbitrage under those scenarios.

  • With those scenarios then, there has been because eh,

  • Denmark in particular is interested in 100% renewables integration by 2035.

  • They are very seriously looking at then the price impact on their markets, and

  • they have done scenario forecasting reviews.

  • Those forecasts with the 2008 buy/sell hour, strategy.

  • And, we show that under that.

  • And shown here in the red curve is the buy/sell strategy and, and, and the price,

  • spot market prices that might be expected in the future.

  • With high penetration,

  • you could actually make money with electricity electricity storage.

  • Again, this is maximum.

  • And of course there's lots of uncertainties here, but it does suggest

  • that if you, even if you weren't perfect, you might end up at zero cost on storage. Okay.

  • So to wrap up here we see

  • that there are a lot of markets that we could enter within this technology.

  • Not only the so called power gas platform.

  • We can do both storage and more recently within the project confines.

  • We're turning now our attention to distributed scale storage that will

  • compete with the Vance flow batteries and sodium sulfur batteries in,

  • in the kilowatt hour to low megawatt hour ranges.

  • There's a lot of work that certainly needs to be done yet.

  • We need, really need to push the envelope on the operating temperature further with

  • the LSGM technology of results shown here are for small scale cells, okay?

  • Cell scale up is always a challenge and that needs to be done.

  • Long term stability and durability testing.

  • We have to operate in cyclic modes with the actual reacting gases we envision.

  • And of course if you're going to run this thing up and down, you need to know

  • something about the dynamics of the capability of, of the system.

  • So, with that I'd say we learned fair amount.

  • We believe we can get fairly high round trip efficiencies.

  • We can even get above 80% if you can integrate formally with nuclear CSP,

  • for example.

  • And regardless of, of how we, we estimate the economics, we think they're much

  • they're very attractive and can meet or exceed the DOE targets.

  • Now with that I'd like to thank some of my collaborators and open it up for

  • questions.

  • [APPLAUSE].

  • >> Questions for Long?

  • Okay, do the back first so yeah that's easier.

  • >> Yeah, I have a question about if you guys have any problems with cell ac,

  • sorry, here.

  • >> Okay. Sorry. >> With cell activity when you're running in electrolysis mode

  • to converting the CO2 to methane?

  • Do you have any issues with making C2s or, you know, products that you don't want?

  • >> No, actually, the, the electrodes are catalytically active enough that they

  • reach equilibrium rapidly with, with without even pulling out oxygen.

  • When you pull out oxygen, obviously we'll drive the equilibrium forward but

  • we make methane and C O and

  • H 2 as exactly as you might predict thermodynamically.

  • >> [INAUDIBLE] >> Nice work Rob.

  • Can you hear me?

  • >> Yeah.

  • >> Nice work.

  • Can you care about comment about the the caulking problem and whether you

  • see it more in the, in the electrolysis mode than in the fuel cell mode?

  • >> That's a good question because as you move from electrolysis mode,

  • you're moving towards the caulking boundary.

  • Certainly one of the questions we have is, you know the thermodynamics you know,

  • is nice, and it provides insight and guidance on how to select conditions.

  • But you really are dealing with local phenomena when you're flowing these

  • reactant gases through the passages of the cell, and if you don't have a, a good

  • distribution, you could have locally rich zone, so to speak, which could,

  • could produce carbon deposition, which would degrade performance.

  • So what's not so well known is what we would call the safety margin

  • that would be required to, to push you away from that thermodynamic boundary.

  • So what hydrogen and carbon ratio, and

  • what operating conditions would give you sufficient safety margin to not coke up.

  • So that, that will be revealed more in the cell testing.

  • As a part of the project we've built a pressurized rig at Northwestern and

  • they're going to be operating under

  • sine gas conditions at pressure and temperature.

  • And that'll give us some better insight.

  • Nevertheless, there's still fairly well mixed conditions under the lab,

  • the lab environment.

  • >> The back, right there.

  • >> So how important it is to lower the operating temperature of this devices,

  • and what do you think is the main idea of in that direction of research?

  • Or how, how to, how you think you can achieve that goal?

  • >> Okay. So, lowering the,

  • the operating temperature really makes more sense.

  • At the large, bulk scale we don't think we need that lower temperature at this point

  • but we look at, we start turning towards distributed scale systems.

  • You know, tens of kilowatts, to hundreds of kilowatts or megawatt.

  • We, we think we will have, we basically, in order to keep the cost low we want to

  • strip out a lot of the DOP equipment that we can.

  • So, we think we can get relatively simple and elegant designs.

  • However we'd like to avoid pressurization in those situations, as well, and

  • so shown here for example is a round trip efficiency versus stack temperature.

  • You do have an expander included, but

  • you can see that as we lower the stack temperature, we can get close to

  • 78% round trip efficiency at 600 C for one of these small scale systems.

  • And we really think we need, you know, depending on whether or

  • not you have the expander you're going to be closer to 70% efficiency if you don't,

  • but you really need the 600 C.

  • The barriers are really the polarization resistances that are occur.

  • The, the resistances go up as you reduce temperature because the ionic connectivity

  • of, of the cell goes down.

  • One of the strategies could be to reduce the air electrode polarization resistance.

  • We think they, they might be able to do that by doing more

  • nanoparticle infiltration on that electrode.

  • >> Just like has been doing on the fuel electrode with nickel,

  • except it might be done with Samaria, for example.

  • I have a quick question you mentioned in your cost analysis

  • that your stack probably should last for at least five years.

  • So could you explore a little bit and

  • say why you believe it will be as long as five years?

  • >> Yeah. We don't know how long it will be.

  • Right now what we see is,

  • after 1,000 hours in these small scale cell test virtually no degradation.

  • The challenge is, of course, we have to operate on the carbonaceous

  • fuel feed stocks we envision, and that hasn't been done over hours.

  • The cycling doesn't seem,

  • at this point, okay it seems like it, it, it does have promise.

  • Other solid oxide cell technology has been demonstrated well past 20,000 hours.

  • All the developers of that traditional focus and

  • technology development are, are focused on increasing endurance.

  • It's going to take certainly several years,

  • I'm sure, to achieve that, but that's economically what the target has been.

  • Some cells, like see, old seaman's tubular cells, they lasted 70,000 hours.

  • But we think 40,000 is where your going to have to start to enter the market place.

  • That's real consistent with fuel cell technology.

  • >> Okay, let's thank Bob again.

  • [APPLAUSE]

Okay, so thank you for that introduction, it's a real pleasure to be here and

Subtitles and vocabulary

Click the word to look it up Click the word to find further inforamtion about it