Field notes The Energy Transition for the Rest of Us

Catalyst N° 122 of 125 27 Aug 2026

The rise of metal fuels

with Richard Wang, co-founder and CEO, Voya Energy

In this note
  1. 01The question
  2. 02The answer
  3. 03The argument
  4. 04What you need to know first
  5. 05Details worth keeping
  6. 06Claims worth citing
  7. 07Where it’s contested

The question

Diesel generators are universally disliked and nobody has managed to replace them. Could aluminum do it?

The answer

Voya’s bet is that it can, because aluminum beats diesel on the exact dimensions that made diesel impossible to displace: it is denser than diesel, it stockpiles for years, and it cannot catch fire. The most valuable part of the episode is not the product itself but the diagnosis underneath it, which finally explains why a technology everyone complains about has survived every attempt to replace it.

03The argument

The setup is the strongest part, because it explains a failure everyone has observed without understanding. Diesel generators are terrible and ubiquitous: roughly 170 gigawatts installed in the US alone, over 10,000 in Northern Virginia, a $30-40 billion annual global market. Users dislike them. Communities fight them. And nothing has displaced them.

The reason is a combination of properties no alternative matches at once. A diesel generator is a microgrid in a box needing nothing else. It is remarkably cheap, roughly $500 to $800 per kilowatt for hardware and about $1,000 with emissions controls, against far more for turbines or fuel cells. And the fuel stores enormous energy in a small, stable, unpressurized space, so 48 or 96 hours of runtime sits in a plastic tank. What the battery world calls very long duration is trivial for diesel.

That is why the obvious substitutes fail. Natural gas introduces a pipeline dependency, and pipelines are stressed at precisely the moments the grid is stressed, so a winter storm threatens both at once. That correlated failure means gas can never be as provably reliable as fuel physically sitting on site. Batteries have excellent power quality and instant start but die on duration, since backup wants 48 hours and batteries get sized for four to eight. And for equipment running perhaps 50 hours a year, capital cost dominates, which is why over-engineering the generator makes no economic sense.

Aluminum enters as a way to beat diesel on its own terms. Aluminum exists in nature bonded to oxygen, and smelting uses large amounts of electricity to break that bond. The metal therefore carries that energy, and it “wants” to return to oxide. Let it, in a controlled way, and you recover the energy. Voya does this electrochemically rather than by burning, using millimeter-scale pellets fed into cells where ambient oxygen meets an alkaline water-based electrolyte, dissolving the aluminum and driving electrons through an external circuit. Burning the metal would produce high-grade heat requiring a steam turbine, which is expensive and hard to scale. The electrochemical path also inherits two decades of cost reduction and talent from the battery and fuel cell industries.

The properties that follow are what matter. Aluminum is claimed at roughly twice diesel’s energy density by volume on an electricity-out basis and slightly better by weight. It is a solid that does not burn, rust or corrode, so it stockpiles for years and ships on ordinary rail, trucks and bulk carriers with no special handling. And because nothing in the system is flammable, including the aluminum hydroxide byproduct, the fire-separation distances and buried tanks that make diesel installations mostly empty space disappear.

04What you need to know first

Energy density
Energy per unit volume or weight. The reason diesel won and the reason batteries cannot do multi-day backup in a reasonable footprint.
Metal-air reaction
Metal reacting with oxygen from ambient air to release energy, here captured as electricity directly rather than as heat.
Power quality
How clean and stable the electrical output is. Diesel is poor at it, which matters enormously for GPUs and requires extra batteries or capacitors to buffer.
Common-mode failure
When your backup fails for the same underlying reason your primary system failed. The core argument against pipeline-fed gas backup.

05Details worth keeping

  • Diesel’s maintenance burden is worse than its reputation suggests. The fuel attracts atmospheric moisture and can grow algae, so it must be “polished” every few months, and the generators need test runs once or twice a quarter. That is substantial overhead for equipment used a handful of hours a year.
  • The grid-asset argument is the most interesting secondary claim. That 170 gigawatts of installed diesel would be enormously valuable as peak capacity, and the DOE has pursued emergency rulings to let it run during grid stress. What blocks it is air quality runtime limits and community objection. Clean backup with no runtime cap would convert an idle fleet into a dispatchable resource already paid for by data center operators.
  • The claimed two-in-one benefit: because it responds like a battery, it could replace both the diesel stack and the lithium buffer that hyperscalers install for power quality, in a smaller footprint than either alone.
  • The defense case is vivid. Military generators idle badly, running far off peak efficiency, and run hot and loud. Wang relays that in Ukraine a diesel generator is treated as a homing beacon for drones, spottable on infrared from miles off.
  • The fuel strategy deliberately avoids clean scrap. Beverage cans recycle in a tight closed loop and displace primary aluminum, which is genuinely useful, so Voya says it will not touch those streams. It targets contaminated mixed-alloy scrap such as shredded end-of-life vehicles, which is discounted, largely decoupled from primary aluminum pricing, and today often landfilled or exported cheaply to Asia.
  • The long-term vision is the more radical idea: recycle the aluminum hydroxide byproduct back through smelting, siting smelters where electricity is extraordinarily cheap, and ship the metal to energy-importing regions. That makes aluminum a globally transportable energy carrier, a way of moving cheap electricity across oceans in solid form.

06Claims worth citing

As of 2026-08-27. Figures about Voya’s own product come from the CEO and describe designed rather than deployed performance, so attribute them to the company. The diesel market figures are the more broadly quotable ones.

  • Roughly 170 gigawatts of diesel generation installed in the US; more than 10,000 generators in Northern Virginia. Kann
  • Global market of roughly $30-40 billion annually for generators plus fuel. Wang
  • Diesel generator hardware roughly $500-800 per kilowatt, about $1,000 with emissions controls. Wang
  • Aluminum roughly 2x diesel energy density by volume on an as-converted basis, slightly better by weight. Wang
  • Voya at approximately 100 megawatts per acre in high-density configuration, about 4x a diesel installation. Wang
  • Instant cold start versus 5-10 seconds for diesel. Wang
  • Defense logistics: roughly 20-30% less mass and about 70% less volume shipped versus a typical military diesel generator. Wang
  • Claimed fuel cost parity with diesel in the US on fully delivered cost, and cheaper in Europe and Asia. Explicitly no green premium. Wang
  • Backup generators at data centers may run only about 100 hours a year. Wang
  • Only about one operational aluminum smelter remains in the US. Wang
  • Target of capturing roughly 1-cent-per-kilowatt-hour electricity in metal form at favorable global sites. Wang

07Where it’s contested

  • It’s a portfolio company conversation, disclosed up front, as is routine on this show. Kann does put the two standard objections to Wang, why this is possible now and where the fuel comes from, and gets substantive answers to both.
  • The product claims are design targets, not deployment results. No third-party testing, customer data, installed base or manufacturing timeline comes up. That is normal for a company at this stage and worth remembering before repeating the cost parity figure as established.
  • The long-term smelting vision is explicitly aspirational. Closing the loop depends on breakthroughs described as things Voya is “working on” and believes it can leverage. The near-term scrap strategy and the long-term energy-carrier vision should be judged separately, since the second is far more speculative.
  • The diesel diagnosis is the durable part. Why diesel resists displacement, why gas backup carries correlated risk, and why batteries cannot reach multi-day duration are all independently useful and not contingent on Voya succeeding. That analysis would survive even if this particular company does not.

Cite as: “The rise of metal fuels,” The Energy Transition for the Rest of Us, note on Catalyst with Shayle Kann, August 27, 2026. CC BY 4.0. View the Markdown