Standard Thermal
Dirt cheap thermal energy.
Orca’s newest project is Standard Thermal. Austin Vernon, James Williams, and Brian Pal have been developing it for the last two years and will be launching it as a startup company this fall.
The thesis with Standard Thermal is that the cheapest practical way to store truly huge quantities of energy is to use the earth itself as a thermal battery. In other words, “rechargeable geothermal power”. It’s as simple as energy technology gets: embedded resistors heat up dirt when you have extra power, fluid channels remove the heat when you need it.
The trick is making it truly dirt-cheap. By digging into the dirt at a customer site, we access high-quality thermal mass for ~$1-2 per ton—an order of magnitude cheaper than anything on earth that is built on a foundation, mined and shipped, or made in a factory. The first time we heat it up past 600ºC, the soil transforms— almost manufacturing brick from clay right underground. With an unexpectedly cheap resistor material, ultra simplified electronics, on-site photovoltaics, and a thermal leak rate under 1% per month, we’re looking at sub $0.10/kWhthermal and $50/kWthermal. We think these numbers will let us supply renewable heat that competes with fossil fuels on a $/BTU basis.
.png)
The promise and peril of long duration energy storage
Anybody reading this will have thought about this stuff many times before so I’ll keep the context brief.
We have to find new ways to move energy in space and time. It's a meta-thermodynamic law: coordination costs increase exponentially as buffers decrease. As we bring seasonally variable energy sources online, we have to either build electric transmission capacity that will be economical despite being seldom used, or else we have to develop energy storage technology that makes money despite only selling energy a few times a year.
For anybody who’s been under a rock the last 5+ years, the writing is on the wall that Li batteries have won for hours-to-days and perhaps even days-to-weeks energy storage. Solar PV plus Li batteries are poised to change the world. The missing piece as everyone knows is in longer duration storage.
But long duration energy storage is a fundamentally weak business. The money you make in energy storage is proportional to the number of times you get to sell energy per year, so as durations increase, profits dry up. Imagine a theoretical dirt cheap electrochemical battery made entirely of pig iron. You can approximate cost as CapEx per unit of energy / cycles * discount factor. Even if the battery were made out of the cheapest and most abundant metal on our planet you’d still be out of the money: ($0.35/kWh energy content of iron) / (1 cycle per year for 20 years) * 4.5 (15% discount rate) = meaningful arbitrage only with seasonal electricity price swings > $0.10/kWh. And of course real batteries cost much[1] more than the constituent materials, don’t give back all of the energy you put in, and have other imperfections in proportion to how cheaply you make them… seasonal energy storage is hard!
Probably the most popular approach to long duration storage these days is thermal[2]. Storing sensible heat reduces material and system complexity requirements relative to electrochemical batteries, to the point where things almost pencil for long durations. It’s not a new idea—steelmakers have used refractory bricks to store heat between furnace runs for almost 200 years. But in recent years >30 startups have emerged proposing similar systems for longer duration storage, things like resistively heating big piles of bricks so you can use the heat later.
Bricks sound cheap. But look at the basic math again-- at $300-700/ton they’re still far too expensive to make money selling energy only a few times per year.
What could be cheap enough? Given that heat capacities of solids are all generally similar at ~3kb per atom, we’d need to store heat in a solid that costs ~$10-30/ton or less and tolerates very high temps. But nothing built on a foundation or made in a factory is sold for less than $100/ton. There’s almost nothing that even gets shipped in the dollar per ton range. If the material can’t even move to the site, that leaves just one medium that’s universally available to store thermal energy cost effectively: the very earth underfoot.
Our system
We embedded an array of dirt-cheap resistive elements in the soil. When the system is charging, a co-located solar PV array directly energizes the resistors, heating the dirt above 600ºC. When the system is discharging, cold fluid enters embedded pipes in the earth, heats up, and exits the system to either provide heat directly to an industrial user, or to boil water and generate electricity.
To give you a sense of the scale: heat travels about a meter through dirt in a month (δ ~ √(α t)), so a system designed for months-long discharge would involve hundreds of meters of pipe and resistor spaced about 2m apart. In that configuration, the system stores about 2MWhr thermal per meter of pipe+resistor. That’s fundamentally favorable—30 times the storage of a Tesla car battery for the cost of a meter of wire and pipe.
Still, getting cost down is everything in this project. If we want to enable intermittent renewable energy to compete with delivered natural gas costs, the storage facility itself can’t add more than ~$5-$8/MWhth to the delivered cost. That means resistors have to be almost as cheap as rebar, heat transfer has to be air or steam etc. Anything else is too expensive. Extra parts or systems—delete, delete, delete. Our 100kW test site in Oklahoma has been a brutal exercise in value engineering—no inverter, barely any MPPT, dead simple fault detection, as few electrical connections as possible, totally passive moisture management etc. Getting a super cheap and super-robust resistor material has been a particular enabler.
A power supply, not a battery
Ideally, an energy-system-stabilizing battery would be able to suck energy from the grid when electricity is cheap, and return it when prices are high. So why have we integrated our system with on-site generation?
The reality check is that even with the lowest storage costs, we need captive power on site. While some entrepreneurs are anticipating that extremely cheap or negative-cost grid electricity will be intermittently available, actual wholesale electricity costs are difficult to disentangle from interconnection, frequency and reliability services, and prices may never get that cheap for any meaningful period. Li batteries could easily beat us on shorter arbitrage opportunities; with the rise of stationary Li-ion, negative midday pricing or the $9000/MWhr prices of winter storm Uri may never re-occur. There's also the fact that energizing the system with DC power on-site saves cost and complexity--deleting inverter, rectifier, separate arc and ground fault protection systems etc.
This all gets to the PV-centric framing of this project: we're selling a buffered supply of solar heat, not just a battery.
Who’s the customer?
In most near near-term cases, we also want to integrate our system with on-site consumption:
1) (near term) The economics are best if we combine captive renewable generation with a customer who wants 24/7 high-quality heat. That means the customer could be anyone who uses steam: industrial drying, food processors, factories, paper mills, chemical plants etc.. If our combined solar + storage cost is cheaper than the delivered cost of fuel, then we should have a good chance to win customers. The comparisons already look favorable for industries dependent on propane, LNG or coal, but even within the US there is spread in delivered natural gas prices that could allow us to compete even though trading hub prices can be absurdly low.
2) (near term) A second no-brainer case is to combine electricity generation and heating for higher latitude customers. Imagine a school in Scotland that wants renewable power on-site. They build enough PV to power their school for the winter months. That’s serious overcapacity for the summer months. We use that overcapacity to heat our installation all summer. It’s then used to heat the school building all through the winter.
3) (medium term) The world-conquering dream is for our PV-based steam to replace fossil-generated steam at conventional power plants. That will let us feed electricity back into the grid using otherwise stranded generating assets (e.g. a coal plant). You might see this as a way to combine an existing, uncompetitive coal plant with thermal energy storage and captive renewables to give it economics more similar to a natural gas power plant.
Fuel replacement at industrial facilities is a straightforward case. People want their products, they need hot steam to make them. You just gotta beat their fuel price by a big enough margin to make the hassle of switching to solar steam worthwhile. But repowering old coal plants is more complex to analyze. To start wrapping our heads around that case, we kicked off a collaboration with Alicia Wongel and Ken Caldeira at Stanford. They looked at how our system would work on an idealized electrical grid, in competition with other technologies with very different economics
Unsurprisingly, Alicia found that the earth-based thermal battery (here in red) plays a bigger role in an energy system with imported LNG prices (>$30/MWhr) than in one with with U.S. natural gas prices, and an even bigger role in hypothetical all-renewables scenarios.
Alicia also found that the thermal storage might play a role even if our system costs creep more than we expect. Such is the stickiness of the need for long duration storage.
Closing thoughts
This project may not have exactly the ‘weird factor’ of most Orca projects. It’s not addressing a high-leverage unasked question like making butter out of carbon or rerouting planes to brighten the sky. Rather, it’s a better answer to a question many others have been reaching for. That kind of thing has hazards, but we see strong reasons to lean in:
- Lowest energy storage costs we’ve ever projected or dared imagine.
- Lower complexity than electrochemical, hydrogen, or pressure storage.
- Round trip (electricity to electricity) efficiency better than hydrogen.
- Captive renewables are now sometimes cheaper than fossil fuels on a BTU basis — now is the moment for the brute-force approach that lets them win on a levelized basis.
Paraphrasing Austin: underground gas storage, oil tanks, and coal piles are nearly free, 1000x cheaper than batteries per kilowatt-hour. Solar PV is ascendent but to win it needs storage almost as cheap. We think this is the ticket.
Anyway that’s the quick and dirty take. All of this is laid out with more care, rigor, and nuance elsewhere:
Footnotes
[1] And the cheaper your constituent materials, the harder it is to get your "idiot index" close to one. Approx indices for batteries these days:
Leclanche/alkaline ~13
Lead acid ~7
LFP ~1.5NMC, LCO ~1.5
[2] Some people always want to mention gravity storage in this conversation. But I see nuclear, thermal or chemical energy as the only worthwhile games in town when it comes to energy storage.
This prejudice comes from looking at the fundamentals: When you store energy, you either set something moving/vibrating (e.g. thermal or flywheel) or you stretch a little “spring”.
The four “springs” that nature provides are the strong nuclear force, the weak nuclear force, the electromagnetic force, and the gravity force. You store energy in these "springs" by stretching or remaking a nuclear or chemical bond, or moving a mass in a gravity field.
Gravity is vastly weaker than the other forces. 1038 times weaker than the strong nuclear force, 1028 times weaker than the weak nuclear force, and 1036 times weaker than the electromagnetic force.
And the gravitational energy stored by a system of particles across ordinary distances (say 1km) is more than 40 orders of magnitude less than the thermal energy of those particles at ordinary temperatures.
So you are just vastly better off dealing with energy in the form of motion, (e.g. thermal, flywheel), electromagnetics (batteries, fuels, springs), or nuclear stuff than you are trying to rig up something to store energy with gravity.
Where gravity has worked, it generally overcomes that 30-40 order disadvantage only by messing with something at an almost planetary scale, like pumping water into a giant artificial lake or aquifer—stuff that only works in select situations and can have big downsides.