The energy time machine

How big a deal for the grid are batteries?

Hi folks,

We’re about six weeks away from the end of the year. Can’t believe I just wrote that. That means it’s time for ‘year-in-review’ type content. 

So I figured I’d take this opportunity to circle back to a topic I’ve covered in depth this year, namely growth in battery energy storage in the U.S. Hope you enjoy – lmk. Lots of pretty charts along the way for the visually inclined.

Quick note: I used many charts that my friend Jeff at enersection creates here. Subscribe to Jeff’s work at enersection if you’re interested in energy data. 

In today’s email:

  • A battery bonanza?

  • A tale of two states

  • How big a deal is this really?

  • What’s hard

  • What’s the environmental impact?


Everyone, at one point or another, wishes they had a time machine. 

Perhaps you wish you knew next week’s lotto numbers. Maybe you want to zoom ahead to 2050 to see how much progress we’ve made in mitigating global warming. I want to look ahead a year to see whether I actually ran the New York Marathon like I’m now saying I will.

Fortunately, when it comes to energy, we have a flexible, effective, efficient time machine. I’m talking about batteries, as well as energy storage in general. 

To be sure, it has existed for a while. In many forms. And it can’t move energy back in time. But hey, my job is to get you excited about climate tech, so grant me a little creative license.

Batteries enable us all to have fun and fancy personal devices that connect to the internet. 

Now, they’re also increasingly adding a lot more touch points with the grid, another distributed, essential network (like the internet). 

The lithium-ion batteries in our phones, in EVs, and increasingly common on the grid, were first commercialized circa 1985

But energy storage has existed in many forms for longer. To this day, the lion’s share of energy storage on grids in the U.S. still comes from pumped hydro. These are simple systems: They pump water up a hill with excess energy and let it run down the hill later to spin a turbine. 

Five years ago, the picture was stark. 94% of grid storage was pumped hydro!

In March 2018, 94% of U.S. electricity storage capacity was pumped hydro. Batteries were ~3% (source).

Now, the picture is shifting. 

Installed U.S. battery storage capacity may exceed pumped hydro by 2025.

A battery bonanza?

Batteries are at the heart of 3 of the biggest U.S. climate tech / energy stories this year:

  • EV sales (across hybrids, plug-in hybrids, and fully electric cars)

  • Domestic manufacturing investments

  • Utility-scale energy storage deployment 

All are relevant. EV sales are skyrocketing, despite some bad press. Both domestic and international manufacturers are investing billions to produce batteries, especially for EVs, in the U.S. and Europe. 

Batteries aren’t just critical for EVs (or iPhones). One of the biggest stories in the U.S. power sector in 2023 has been growth in utility-scale, grid-connected battery energy storage deployment:

Chart via enersection

To be sure, 'generation' is a ‘generous’ term for what batteries do. They do not generate energy. Instead, like the time machine we daydreamed about earlier, they can move it ‘forward’ through time. 

Batteries, and energy storage more broadly, is not the only set of technologies designed to 'move' energy. Transmission and distribution networks that span the country (and globe), which, in my mind, looks kind of like a mycorrhizal network of fungi connecting trees, move energy (really it’s power, but I’m painting broad strokes here) through space.

Some argue that transmission and batteries are at odds with one another. Casey Handmer, whose perspective I respect and who is a good follow on Twitter, nay, X, wrote a piece a while back titled "Grid Storage: Batteries Will Win" on this topic. In his article, he opined:

Batteries and transmission are in direct competition. Both enable electricity arbitrage – the profitable repricing of a resource by matching different levels of supply and demand. Transmission moves power through space and batteries move power through time. And while batteries have a fixed cost per MWh delivered (that is falling about 10% per year), transmission lines get more expensive as they get longer.

To be clear, I think both transmission and more batteries will be vital to deep power sector decarbonization.

I agree that batteries are easier to deploy. Building new transmission lines has become a meme because of how long it takes to secure all necessary environmental, siting, right-of-way, and other approvals and to marshal all the stakeholders needed for that process. One important U.S. transmission line took 17 years to complete. Mind you, China's command economy doesn't seem to have these problems.

That's part of the reason why battery energy storage deployment is growing exponentially. Without sufficient transmission, other assets are needed to valorize renewable energy's potential more fully. 

A tale of two states

Before we explore batteries in more depth, it’s worth noting that most (almost all) grid-connected battery energy storage deployment is happening in Texas and California. It’s helpful to know why.

Chart via enersection

For one, Texas and California are among the largest energy markets in the U.S., which means there are more opportunities for battery energy storage developers to deploy projects and make money on repricing and arbitrage. These two states also have a significant share of solar and wind power in their grid mixes. This is important because batteries are solar and wind’s best friend. 

As the oil lobby and fossil-fuel-fired Twitter bros will gladly remind you, the sun doesn’t always shine, and the wind doesn’t always blow. By storing solar and wind energy, batteries make those assets more similar to other generating assets with higher uptime. 

In other words, batteries smooth out fluctuations in energy production by storing excess energy when generation is high and releasing it when demand is high or generation is low.

There are other reasons Texas and California are leading the charge here. For one, Texas faced a severe power outage in February 2021 during a winter storm, highlighting the need for a more resilient grid. Batteries can help with that. 

Both states have also developed active policies to support energy storage deployment. California has set targets for energy storage capacity deployment and offers incentives to battery storage system developers. Texas has also implemented programs and incentives to promote energy storage.

Other states should (and will) follow suit. 

As an aside, another reason driving the battery boom is also pure economics. Solar development has boomed in the past decade. It’s crowded now. Battery energy storage project development isn’t as crowded (yet). I know developers who started in solar and have now fully pivoted to batteries.

How big a deal is this really?

While the growth in battery energy storage’s contribution to the grid, when charted, is exponential, the total amount of energy that batteries are ‘generating’ in the U.S. is still small. 

Chart via enersection

For perspective, one power plant, the Diablo Canyon nuclear power plant in California, is not that far from where I'm currently writing this newsletter. It alone will generate ~3x more energy than all utility-scale, grid-connected batteries in the U.S. will this year. Pretty remarkable!

Here's my back-of-the-envelope math:

  • Diablo Canyon produces ~16 TWh of energy annually.

  • The chart we opened the newsletter with showed 4.5M MWh of 'generation' so far. One million megawatts make a terawatt, so that's 4.5 TWh of 'generation.' By year-end, it'll be between 5 and 6 TWh.

However, I'm not here to go all nuclear bro on you. Yes, we should use nuclear power plants, especially ones that have already been built. The Diablo Canyon plant should operate as long as feasible. But the comparison between nuclear and batteries is in many ways unfair. Batteries' value isn't all about 'generation.' They exist to make all other energy assets more valuable. They're the sixth man of the energy generation basketball team.

For one, batteries make renewable energy more valuable. They effectively enhance renewable capacity factors, making it more attractive to develop renewables in the first place. That’s good for existing assets, for developers who want to develop future projects, and for the grids (more on this in a sec).

We can ‘sense’ that batteries are helpful in the increasinge rate they’re being deployed alongside solar (as well as in the form of standalone projects).

Batteries also help ease peak load, i.e., times of intense electricity demand, on the grid. Importantly, they can respond to need and discharge (or ‘ramp’) quickly. That’s key because traditionally, the generators designed to ramp up fastest are natural gas peaker plants (or on smaller scales, diesel generators).

In sum, batteries make many of the other assets around them better. And they help the grid, whether from a peak load response or resilience perspective. 

What’s hard

Even just 20 years ago, the grid was highly centralized, especially at the level of actual generation. If you lived in a city, most of your electricity probably came from a handful of large coal, natural gas, or nuclear power plants as well as hydroelectric dams and pumped hydro energy storage reservoirs.

Now, we're heading towards a future where there are millions of assets on the grid. 

This is what 'distributed energy resources' ('DERs') and virtual power plants are all about. Virtual power plants refer to a network of decentralized energy resources, including rooftop solar, batteries, EVs with bidirectional charging (also a battery), smart thermostats, and more, which, all taken together and scaled across many homes in a city or community, can be managed like a power plant. 

DERs recently had a party in Brooklyn. I missed it because I was hiking in Boulder. I was bummed! 

The DER folks are a bit of a climate 'tribe,' and I don't mean that in a bad way. I see the appeal. DERs offer an exciting vision for the future: a future with fewer carbon emissions, cheaper electricity bills, and more ownership, independence, and control for citizens.

Still, moving from a model with, say, five core 'nuclei' (power plants) generating energy for a grid to one in which millions of new, smaller nuclei are connected is no small feat. Harmonizing all the new assets coming online will be a significant challenge. 

It will be challenging for grid operators, who are used to the centralized model, and for the grid itself. Grid infrastructure wasn't built to support the new model. Considering how old much of the grid is, what’s happening with DERs is kind of like trying to run ChatGPT 3.5 on 1990s internet infrastructure.

This shift will be and already is difficult for utilities, which don't make money for electricity they don't produce and, therefore, have less money to invest in, say, maintaining, improving, and expanding the grid.

Battery storage at a solar farm with switchgear in background (Shutterstock)

What’s the environmental impact?

Another challenge is that batteries, even when connected to the grid and sited alongside renewables, don't always reduce greenhouse gas emissions. At least not right now. 

That might sound paradoxical, but battery energy storage operators, first and foremost, are trying to make money. I asked Emma Konet, the CEO of Tierra Climate, a startup working to solve this problem through corporate offtake structures for carbon abatement, to help me explain why. 

Here’s what she had to say:

We studied this and found that 80% of operating battery storage caused emissions to increase in 2022. Grid operators currently dispatch generators based on lowest cost, not lowest carbon. There are many factors that can impact a fossil fuel generator’s cost. Combine those with the opportunity costs of providing other services to the grid and round trip energy efficiency losses, and a battery can actually “induce” carbon emissions. With carbon offtake structures in place, however, most Texas BESS assets could be incentivized to change their behavior from carbon inducing to carbon abating.

I asked Emma to expand further on what she meant by 'inducing' carbon emissions. If you love the nitty-gritty of power markets, this is for you. If you don't, skip ahead. Here's more from her:

There are three main scenarios where "induced" emissions could occur.

  1. For example, if a battery charges from $20 power produced by natural gas and then discharges to displace $60 power that is also natural gas (but maybe comes from a combustion turbine instead of a combined cycle plant), it makes $40 on its MWhrs, but the carbon arbitrage isn't wide enough to actually abate carbon. The battery loses 15% of the energy it charges in the form of heat, contributing to the non-positive carbon abatement. Basically, a fossil fuel plant ran to charge the battery when it wouldn't have otherwise because that generation wouldn't have been necessary. 

  1. A battery provides ancillary services to the grid and may charge or discharge uneconomically to maintain a state of charge necessary to provide that service reliably.

  1. Prices may be uncorrelated to emissions. For example, if LNG demand causes natural gas prices to rise above coal, then coal is dispatched first before natural gas is in the stack. So, a battery operator who purely follows price signals could charge off of coal and displace natural gas, substituting a 'cleaner' resource for a dirty resource. 

In summary (Emma’s words, still), power markets fail to account for these issues because carbon isn't priced into dispatch optimization engines. Prices are. Battery operators try to make as much money as possible and may inadvertently increase emissions if the conditions aren't 'right' on the grid in the process. As grids become more highly renewable, this problem should go away on its own, but that will take a while (like, decades). 

Said differently, now back in my (Nick's) words, because battery storage developers and operators are focused on the bottom line first and foremost, there can be scenarios where they charge batteries with energy from fossil fuel-fired generation that then drives more fossil-fuel-fired generation later or doesn't pencil from a carbon perspective for other reasons (e.g., because batteries aren't 100% efficient at storing the energy they're charged with). 

All of this attunes us to a fact that’s worth remembering. As batteries come online on grids, they don't just service renewables. Depending on the project, they may service the entire grid and all its assets. Until the grid is 100% low-carbon, batteries can raise emissions under the wrong conditions.

The net-net (tl;dr)

Batteries – and other forms of energy storage – are essential to deep power sector decarbonization. Renewables alone, even with significant transmission, won't get us there. The only places in the world with über-clean grids have a lot of hydroelectric power, geothermal power, nuclear power, or some combination of all those. Hydroelectric and geothermal energy are highly location-specific; if you live in Quebec or Washington, you're lucky. If you live in Japan, you’re not.

To reiterate, the role of batteries isn't fundamentally generation. Measuring how many MWh batteries have moved through time is a cool headline stat. However, it doesn't capture their primary role, which is to make renewables better for everyone, including project developers, asset operators, utilities, and individuals (in the case of DERs).

There's a call for startups here. We met Tierra Climate, a great example of an early-stage startup working on an under-discussed issue with battery deployment, namely that batteries don't automatically reduce emissions when you plop them down and connect them to the grid.

The need for both hardware and software to harmonize energy generation, delivery, and management in a future where grids have millions of potential generation and storage assets rather than a handful will be huge. The software will look different for EV charging, for homes, for larger buildings, etc… Many teams are already working on this. I can point you to some.  

Plus, once we ‘solve’ all this in developed markets, like the U.S., we'll get to do it all in developing markets, too. No shortage of opportunity!


– Nick