06 July 2023 |

From backup to game-changer?


I’m not going to use this email to convince you why energy storage matters. If you would like me to, read this piece from two months ago. Rather, this email is more of a resource to get up to speed on what energy storage looks like on power grids in the U.S. in 2023. 

Here’s what to know about everything pre-this decade: The majority of energy storage capacity connected to the grid in the U.S. is based on century-old concepts, like pumped hydro, rather than batteries. 

Pumped hydro is simple: Use excess energy to move water up a hill. When you want to harness energy again, let water flow downhill to turn a turbine.

According to the DOE, the U.S. has:

22 gigawatts of [pumped hydro] electricity-generating capacity and 550 gigawatt-hours of energy storage with facilities [across] the country.

For now, that dwarfs utility-scale battery energy storage capacity, of which less than 10 GW are operating at current:

Battery energy storage capacity in the U.S. may well exceed pumped hydro capacity by 2025

Pumped hydro is a form of gravitational energy storage. Its simplicity doesn’t make it less useful; pumped hydro works quite well, especially at large scales. With suitable reservoirs and siting, pumped hydro facilities can offer gigawatts of energy storage capacity

But there are many other ways to store energy. The main tech technology that probably comes to mind for you are batteries. Batteries store energy electrochemically and offer many advantages, especially at smaller scales (e.g., if you want to store energy in a car or electronic device).

A side by side of the 20th century’s leading energy storage technology vs. the 21st century’s 

Beyond gravitational energy storage and electrochemical energy storage, there are other chemical forms of energy storage (e.g., using energy to create hydrogen) and mechanical energy storage systems (e.g., compressed air energy storage). 

The present: Lithium-ion batteries break out

Increasingly, chemical energy storage is getting a lift. While gravitational and mechanical energy storage options offer scale, they’re often contingent on siting and geography. And while they can store a lot of energy for a long time, duration and scale aren’t the only things that matter. 

Power, i.e., the rate at which energy can be transferred, matters. The scale and duration of energy stored in giant water reservoirs is often desirable. But the ability to discharge energy quickly can be very important, too. 

Lithium-ion batteries, in particular, excel at discharging energy almost instantaneously. That’s part of why their inventor, John Goodenough, who passed away last month, won a Nobel in Chemistry. 

The past few decades have seen a renaissance in battery technology. Lithium-ion batteries revolutionized consumer electronics, like your phone and computer, first. But as they improved and people made more and more efficient, lithium-ion batteries matriculated into other, larger applications, like EVs:

Someday they may even moonlight in planes and ships. 

And they’re also migrating onto the grid to help balance renewable energy generation with the timing of when electricity demand is highest.

Most grid-connected battery energy storage systems are lithium-ion, partly because lithium-ion batteries are so good at quickly discharging energy when needed. Lithium-ion batteries represented more than 90% of total battery energy storage deployment in 2020 and 2021 in the U.S.

Why is the ability to discharge stored energy quickly theoretically so valuable? It’s easy to appreciate in the example of an EV; sometimes, drivers want to hit the gas pedal and feel the acceleration!

On the grid, an example might help. A few weeks ago in Texas, in the middle of a heatwave, a nuclear power plant (usually extremely reliable) tripped and went offline. Texas’ grid lost 1,235 MW of generation capacity from one minute to the next. Luckily, however, a new backup energy storage system had gone live on the grid only six days prior. And, you guessed it, that backup ‘reserve’ system is predominantly battery energy storage systems, which can discharge power faster than a backup peaker natural gas power plant can fire up. 

California, in particular, has led the charge in deploying utility-scale battery energy storage systems. As Matthew Zeitlin summarized nicely in Heatmap recently:

…in California, the recognition that renewables alone can’t power the grid 24 hours a day has led to a massive investment in energy storage, which can help approximate the on-demand nature of natural gas or coal without the carbon pollution.

California now has over 5 GW of battery energy storage capacity deployed on its grid, representing over half of the nation’s grid-scale battery energy storage and enough to meet ~10% of the state’s highest peak load.

California needs this battery energy storage as it deploys more solar. As we’ve analyzed previously, without energy storage to ‘shift’ electricity, the timing of when solar produces most of its electricity is often mismatched with when electricity demand is highest on power grids. This creates a curve that looks a bit like a duck; electricity is readily available during the day when the sun is shining, but less so at night when solar power generation dwindles and demand skyrockets: 

Energy storage systems are needed to help balance this out. With energy storage, you can take solar power and shift it from noon, when it’s often least needed, to 7:30 pm, when it’s most needed.

Another factor that points to the importance of storage is curtailment. Curtailment refers to when renewable energy resources like wind and solar can produce more power than the grid needs at a given point in time. In these scenarios, the resources are curtailed, a fancy word for ‘turned off.’ 

Even in California, where battery energy storage system deployment is taking off, ~700,000 MW hours of solar and wind power were curtailed in April alone. That’s equivalent to 0.29% of the total load California’s grid served across all of 2022. To summarize, battery energy storage isn’t just backup for the grid. It’s now an integral technology to balance the grid as renewables accelerate.

The future: Multidimensional battery development

The CEO of an independent power producer and energy storage developer/operator I spoke with recently noted he believes they’re ~18 months away from deploying “a different type of technology,” meaning an energy storage system using a non-lithium ion battery chemistry. 

Primarily, he cited cost and the longer-duration storage as driving factors.

As we’ve discussed, lithium-ion works great from a response time perspective. But other solutions may work better if you’re focused on energy storage and cost duration. Companies like Form Energy are pioneering chemistries that are more cost-effective than lithium-ion as they use more abundant materials; Form Energy’s batteries will primarily be made of iron. That said, they just broke ground on their factory and don’t expect to ship product until 2028.

Part of what’s most interesting about this conversation to me is how it reveals how multidimensional battery design is across chemistries and configurations.

There are many variables you can optimize for in battery energy storage:

  1. Power capacity: The amount of electricity a system can discharge at ‘full throttle.’
  2. Storage duration: The amount of time a storage system can discharge at its power capacity before depleting its energy capacity
  3. Energy density: How much physical battery you need to store a set amount of energy
  4. Response time: How instantaneously the system can jump into action 
  5. Cycle life: How many times can you charge and discharge the system before it degrades
  6. Material constraints: Scarcity of the core materials/supply chain concentration 
  7. Safety: Will it catch on fire, like lithium-ion batteries notoriously do?
  8. Cost: What’s the all-in cost to get a system deployed, especially considering material inputs?

The list goes on. It makes sense that lithium-ion batteries won’t be the only chemistry used to balance grids in the future; other chemistries will lend themselves well to different applications and sites. Especially considering how much IRA incentives emphasize domestic battery content, chemistries primarily made of more abundant materials should see a lift in the future.

Still, out to 2025, most of the planned energy storage capacity additions in the U.S. are lithium-ion-based systems (see below). For one, that’s what’s available right now. And secondly, from a bankability perspective, the lenders that work with developers to underwrite these systems and navigate tax equity financing like what’s tried and true, which, right now, is lithium-ion. 

That’s a lot of planned lithium-ion battery energy storage (green dots) capacity additions (data and chart via enersection)

The net-net

This email has primarily been an overview of the grid-connected energy storage space in the U.S.; there isn’t necessarily a trenchant ‘opinion’ for me to add here at the end other than reinforcing that I think this is one of the most pivotal and interesting areas in climate tech right now. 

Next week, we’ll hone in on how battery energy storage projects actually get deployed, the key challenges they face in 2023, and how various stakeholders can work together to alleviate said challenges.

Beyond that, let’s finish with a question. As we go deeper into energy storage this month, what would you most like me to spend time covering and unpacking? Some options are below, though you’re welcome to go ‘off road’ too:

  • Market map of early-stage companies innovating in energy storage
  • Tech breakdowns on new, innovative energy storage concepts
  • Company deep dive on one to two players in the space
  • More on the policy front, i.e., what’s being done to support energy storage deployment?