Forcing A Breakthrough In Energy Storage: An Interview With Dr. Steve Griffiths
Looking Toward Long Duration Storage
OP: I assume that in your discussion of CSP with storage, and demand side technology, you’re talking about what is needed now to take us up to 20% intermittent renewables in the grid. Jumping up to 20 to 40%, we’re probably looking out 10 to 20 years in the future. What are you seeing in this longer time frame?
SG: Yes, what I’ve spoken about so far is looking at 6 to 12 hour storage, and making the grid more robust. These are all things that have to happen today as we’re moving in a direction of having higher renewables.
Now, regarding storage, this is where I think things get interesting. Because for storage right now, what you’re seeing is a proliferation of lithium ion technology. That’s happening because lithium ion is obviously the choice for battery storage in vehicles. The automotive sector is pulling forward a lot of the battery innovations in the lithium ion space.
But lithium ion batteries, to be quite honest, are probably going to be good for 4 to 6 hours of grid storage and up to a couple hundred megawatts in scale. When you get beyond that, the challenges of lithium ion scalability become apparent. If you think about this, with lithium ion batteries, we have battery cells arranged into battery packs that are simply multiplied in number to achieve more storage capacity. So it’s almost like a linear progression in cost, because you’re not generating any more electricity from batteries. Rather, you’re just storing it for later use. When you want to shift the use of electricity in time, you’re adding cost to achieve the flexibility. So at some point it really doesn’t make a lot of sense to go with lithium ion if you’re talking about long-term storage that needs to have some very significant economies of scale to be economical.
Moreover, taking a look at what’s happening with lithium ion today, and considering that cobalt is now a key material in advanced lithium ion technology, there’s a concern that we’re going to have a cobalt shortage given the projections for use of cobalt-based lithium ion electrodes in electric vehicles. So that may limit the deployment of lithium ion batteries, particularly for grid-scale storage.
OP: Higher shares of renewables will require longer term storage?
SG: Yes, long duration storage. Let’s say we have a couple days with limited generation from renewables but we have very high shares of renewables on the grid. We have to be able to have some way in which energy is stored and dispatched to the grid over a period of days.
This requires moving to a new class of technologies that allow you to achieve scalability for cost-effective storage when you have just a few charge and discharge cycles each year and discharges for, potentially, for very long periods.
If you look at technologies like flow batteries, you have the ability to completely decouple the power generation from energy storage and so scalability becomes more achievable. With flow batteries, you have tanks of electrolytes that run through cells with positive and negative electrodes and a membrane and allow for an ion exchange to occur in order to generate electricity. With this type of technology, you want to have very cheap tanks, electrolytes and other systems components that allow for cheaper unit production of electricity as size increases.
The challenge, though, is that today the incumbent flow battery technology is vanadium redox. It’s good except the vanadium, which is both the catholyte and the anolyte in different charge states, is expensive and energy density is modest at best. However, there’s some breakthrough research going on now, especially that’s come out of MIT, that’s looking at using sulfur-based flow batteries. The value here is that Sulphur is abundant and cheap and so can be used in large systems with improved energy density and so scale pretty well. When this happens, all of a sudden you can start talking about, 10 hours, 20 hours, 30 hours of dispatchable storage that’s very cost effective.
Now long-duration storage doesn’t always have to be via flow batteries. It can be via whatever technology you can conceive that allows you to dispatch electricity economically for long durations and with few cycles. ARPA-E has launched a new program for this kind of long-term energy storage R&D.
So in short we can have flow batteries that are working with very cheap electrolytes, or any other technology in which you don’t use expensive metals or commodities and can achieve sufficient energy density for scalability. The key is to achieve very low cost per charge and discharge cycle.
So that’s long duration storage, which is critical when you get to very high shares of renewables. Lots of research now is going toward what’s anticipated to come in 10 to 15 to 20 years.
Carrying Across Seasons
OP: And beyond 20 years?
SG: In many locations we’ll move beyond 50% renewable electricity and in most locations this much renewable energy will not be fully accommodated within a given season. You’re going to have to curtail excess energy, which means that you won’t be able to use all that you have and the economics will be negatively impacted.
What I’ve seen as far as seasonal storage interest is largely from European countries. Some studies there have focused on 50%-plus renewable electricity as the point where seasonal storage becomes very relevant. The major seasonal storage interest now is in hydrogen. Hydrogen has actually become one of the new programs of Mission Innovation, which is the major global initiative to accelerate public and private clean energy innovation.