Grid Scale Batteries for 4, 10, and 100 hours!

How to store our abundant and cheap renewable energy...

Grid-Scale Storage!

The LCOE (Levelized Cost of Energy) of renewables (Photovoltaics and Wind) is already the lowest of all energy sources but they are intermittent energy so we want to store it to dispatch it when it is needed. This note captures some of the recent progress in Energy Storage, with a focus on grid-scale and a US and California bias.

How long? (4 hours, 10 hours, 100 hours, a season?)

An Energy Storage System (ESS) has several key properties:

• How much energy can be stored in the facility? (MWh)

• How much power can the facility deliver?  (MW)

• For how long can the power be delivered? (Hours/Days)

• Cost! ($$)

The use of storage in the grid has changed over time as the cost of the storage changes and as the percentage of renewables in the grid increases. For example, early on ESS was mostly used for frequency regulation.

The majority of today’s ESS deployments address the “duck curve”, the difference between consumption and solar generation during each day. These deployments are configured for 4 hours of power.

• Short-duration weather patterns, like storms, may impact solar generation beyond 4 hours. A common target for these is 100 hours, 4 days, as this will help with most storms. We are beginning to see ESS deployments targeting 100 hours.

• Beyond that, Solar, and Wind are impacted by seasons and this requires seasonal storage. We are just starting to talk about these cases.

ESS Growth

As of February 2024, EIA reports that 81% of the new power in the grid involves Solar and Batteries, and that ratio is only increasing. A more recent report from July 2024 shows that the US added 20.2 GW of generation power in the first half of the year, and of those, PV totaled 12 GW, 59% of all, and battery storage added 21% (4.2 GW).

California

California has seen a huge growth in Grid storage. Starting from 500 MW in 2020, we reached 5,6 GW in 2023. In late 2023, Moss Landing was the world biggest storage project with 3GWh. As this writing, California has 10,383 GW online with 3,8GW more planned before end of 2024 (CEC survey).

California is not the only state adding storage…

New York

.New York is now poised to go big on energy storage.

Texas

Texas’s main grid operator, ERCOT, has been adding renewables and storage very fast in recent times. Below is a dashboard from gridstatus.io showing that the maximum power storage in ERCOT was accomplished a few days ago, storing 3.927 MW - check out how fast it is growing!

As a comparison, the record in CAISO is 6.2 MW but ERCOT is growing even faster. This is in part because ERCOT is an energy-only market.

BTW, if you wonder, gridstatus.io does not list data on storage in NY; but do check their page.

Lithium-Ion Storage

Lithium-Ion is the type of battery used in pretty much all of todays BEVs, residential batteries, and grid-scale ESS.  Most of today’s ESS target 4h duration; there is no intrinsic reason for the 4 hour emphasis, just economics. Earlier Li-ion was based on NMC chemistry and, while the cost of batteries based on NMC continues to go down, new deployments have moved to a different chemistry.

LFP

Lithium Iron Phosphate (LFP) are cobalt-free, are very safe, and are cheaper than NMC batteries. All new residential batteries are moving to LFP (check my note on the topic) for these reasons. LFP are heavier than NMC, but not substantially so and are increasingly showing up in EVs too.

Tesla’s Megapack switched to LFP for Megapack 2.0 around 2021 and comprise a large percentage of the grid-scale battery deployments in the US, and it is actually the largest producer worldwide.

Sodium-ion (Na-ion)

A chemistry even cheaper than LFP is Sodium-ion (Na-ion).  Na-ion is beginning to appear in some entry-level EVs, but the best fit may be in stationary batteries. Natron is a startup that recently announced the plans to build a new 1.4 Billion GigaScale manufacturing facility.

The industries listed in the Natron site do not currently list grid-scale storage; but Na-ion is already being used in China.

Na-Ion looks great but even cheaper than Na-ion is … iron rust!

Iron Air Storage

Form Energy is using Iron Rust to store electricity. Turns out that rusting is reversible and can be used to store electricity (see Wikipedia). The technology is very promising; it uses abundant minerals, it is cheap, it is thermally safe, and it can be produced at large scale. Iron is heavy but that is not an issue for an ESS.

Na-ion seems a great match for 100 hour storage, and Form Energy has been signing a fairly large number of contracts:

An early contract was in 2020 with Great River Energy, in Cambridge. That contract is for 100 hours: 1.5MW/120MWh. In 2023, Form Energy signed a contract in California for 5MW/500MWh, and then another in NY: 10MW/1000MWh. And in Aug 2024, a bigger one in Maine for 85MW/8500MWh

Added: California’s CPUC just approved 1GW of multi-day storage.

Not a single-solution problem

A basic theme across all these technologies is that storage is not a single problem.  Iron is not a good solution for an EV: too heavy, and not enough power.  But it is a good technology for grid-storage because the basic materials are cheap, the construction is “easy”, and it can scale very well.

Flow Storage

A Flow battery is yet another type of battery: “The fundamental difference between conventional and flow batteries is that energy is stored in the electrode material in conventional batteries, while in flow batteries it is stored in the electrolyte”.  The cost of a Flow Battery is lower than Li-ion or Na-ion and this opens new use cases (caveat: I don’t know how Flow compares in cost with Iron-Air).

ESS is the leading vendor and California, again, is leading in testing out this technology. ESS signed a 10 hour storage deal with SMUD: 200 MW/2GWh.

Hot Rocks … CO2 … and More!

There are other ways to store “renewable energy”; for example, it can be stored as heat to be used directly as such, or.. not, or it can be stored as CO2, or … I will write a separate post on some of those alternatives.

Takeaway

Grid-scale storage is a key companion to very cheap renewable energy.  Storage enables additional deployment of renewables and expands its applicability into additional use cases.   The regulatory infrastructure (in California, Texas, and other places) is fairly favorable to storage - no need to convince PUCs to change regulation - and we see adoption by power generator companies, and even by our IOUs.  Adoption of storage will continue to grow!

Extra References

Collecting some useful links here:

• The 2024 LCOE report from Lazzard. And the 2024 report from NREL.

• EIA is the U.S. Energy Information Administration and it has a wealth of Electricity datasets. Download and explore at your leisure. EIA also has many interesting reports like this one.

• Our CEC has a section on Electricity Data including a nice dashboard on Storage (the data seems to match what is in the EIA datasets).

• Energy-Storage.News reports on news in the ESS industry.