But altitude is just one of many variables that define an orbit. Inclination, eccentricity, and period matter just as much. Two satellites at the same altitude can have completely different capabilities depending on how their orbits are tilted and shaped.

Instead of points on a map, orbits can be thought of as mission enablers. Each buys you a specific advantage, such as persistence, proximity, global access, or stability. Depending on what you’re trying to achieve, you can tailor the orbit to your mission.

Each comes with costs, too. Some orbits are hard to get to, while others are difficult to maintain. A few are prime real estate and get crowded.

The question isn’t just where to orbit. It’s what advantage do you need?

Persistence

A satellite above the equator at an altitude of 35,786 km circles the Earth at the same rate as the Earth’s rotation. It therefore hangs over one spot on Earth, earning it the name geostationary or GEO.

From this geostationary vantage, the satellite can see about 40% of Earth’s surface. But its usefulness degrades toward the edges of its field of view, where the satellite drops to roughly 5° to 10° above the horizon. Signals must pass through substantial atmosphere and are easily blocked by terrain and buildings. This makes GEO satellites less effective at servicing high latitudes.

But within the effective footprint, there are no handoffs, gaps, or timing windows. This creates a persistence advantage.

NOAA’s GOES (Geostationary Operational Environmental Satellite) weather satellites stare at the same hemisphere permanently, watching hurricanes develop in real time. Many missile warning systems need unblinking eyes over potential launch sites. Broadcast TV uses GEO because a fixed satellite enables fixed dish antennas, which are simple and cheap.

The cost is distance. Signal latency is about 240 milliseconds for round trip propagation, but around 500 to 600 milliseconds once you account for routing and processing. That’s noticeable in voice calls and painful for real-time applications.

Getting there is also expensive. As we’ll see below, orbital mechanics demands significant energy to get to GEO, which first requires traversing the geosynchronous transfer orbit (GTO), a Hohmann transfer as outlined below.

For small launchers, this orbit is often unachievable. For most larger rockets, the payload to GTO or GEO is a fraction of that to LEO. It’s often on the order of 30% to 40% to GTO, and lower to true GEO depending on the mission.

And the equatorial GEO belt has limited slots. Only so many satellites can park at useful longitudes before they start interfering with each other, making an orbital slot at this location a valued asset.

The image above is from NASA’s Orbital Debris Program Office and illustrates satellites in orbit from 2019, as seen from a viewpoint above the north pole. The well-populated equatorial GEO belt can be clearly observed, as well as the LEO and MEO orbital densities.

To expand on GEO, we can add inclination by tilting the orbit at an angle relative to the equator. Few satellites, except for those in GEO, orbit at 0° (equatorial) inclination. Most have some inclination to modify their ground track.

If you maintain the geosynchronous daily orbital period, but add inclination (and/or eccentricity), the satellite becomes geosynchronous (GSO) rather than geostationary (GEO). It no longer stays fixed. Instead, its ground track becomes an analemma (figure eight).

The relatively concentrated ground track is still useful, and we’ll see it deployed below for Tundra orbits.

Proximity

Low Earth orbit (LEO) offers the opposite advantage: closeness. Technically defined as 200 km to 2000 km, LEO is often most concentrated around 400 km to 600 km. At this altitude, a lap around the Earth takes about 90 minutes.

Here, you’re near the action. Round trip latency drops to 10 to 40 milliseconds. Imaging resolution improves dramatically. The closer you are, the smaller the objects you can resolve. Higher signal strength allows user terminals to be smaller and cheaper.

This makes LEO popular. SpaceX and Amazon operate megaconstellations for global satellite broadband here. As a point of reference, SpaceX’s Starlink constellation has just under 10,000 satellites in orbit, and Amazon is targeting a 3,000+ deployment.

Besides the telecom megaconstellations, the International Space Station orbits at 400 km, and many imaging and remote-sensing constellations fly at similar altitudes.

Here, the cost is the coverage footprint. A LEO satellite at 500 km sees only about 1-2% of Earth’s surface at a useful angle. This necessitates constant handoffs and complex coordination between orbiting satellites. Global coverage requires hundreds or thousands of satellites, usually in mid-inclination orbits of 30° to 60°.

This arrangement provides good coverage over populated areas, but not at high latitudes. Amazon’s megaconstellation below illustrates the limited coverage near the poles for an inclined LEO constellation.

At the lower end of the LEO altitude range, atmospheric drag is also a real factor. This eats away at a satellite’s velocity. Without periodic reboosting, satellites can fall out of orbit. Even with propulsion, many LEO platforms plan for lifetimes ranging in the single digits of years.

Proximity gets you close. But staying close takes work.

Balance

Medium Earth orbit — roughly 2,000 to 35,786 km — splits the difference between LEO and GEO.

For example, at 20,200 km, you’re high enough that each satellite sees about a quarter of Earth’s surface, but low enough for reasonable latency and signal strength. This is the GPS altitude. The constellation’s 31 satellites can cover the entire globe, and the geometry works for precise positioning because multiple satellites are always visible from any point on Earth.

Similarly, SES operates a communication constellation at 8,000 km. The pitch is high bandwidth with tolerable latency. It’s better than GEO’s delay, but simpler than LEO’s thousands of satellites and constant handoffs.

The downside is that MEO doesn’t necessarily optimize anything. It’s not as persistent as GEO, not as close as LEO. MEO is capable and flexible, but rarely the best choice for any single parameter.

Global Access

Maybe you want access — the ability to see anywhere and everywhere. Polar orbits help with this.

A polar orbit passes over both poles. In other words, the satellite circles the Earth from north to south. As Earth rotates beneath, each successive pass covers a new strip of ground. Wait long enough, and you’ve seen everything.

The advantage here is that nothing hides. The National Reconnaissance Office doesn’t need to choose which territory matters. Polar orbits give access to everything. Denied airspace, hostile nations, and polar regions are all within reach.

This fact is perhaps best embodied by the NRO’s Octopus emblazoned mission patches.

There’s also a tactical trick available. At the poles, all longitude lines converge. This means a small propulsive maneuver at the poles can shift subsequent ground tracks by hundreds of kilometers.

Satellites can be quickly readjusted to focus on emerging conflicts or areas of interest. Inclined orbits can’t do this easily; their ground tracks are constrained by geometry. Need to overfly a different target on the next orbit? Burn at the pole.

The cost is launch efficiency. An equatorial launch to the east gets a 460 m/s boost from Earth’s rotation. That’s about 5% of orbital velocity for free.

In practice, most missions launch from higher latitudes, such as 28.5° for Cape Canaveral, and go into mid-inclination orbits. This reduces the boost, but it still helps.

A polar launch from Vandenberg doesn’t get this benefit, as you’re launching nearly perpendicular to Earth’s rotation. As a result, the rocket has to supply all the orbital velocity, meaning polar access costs more propellant than equatorial orbits.

But for reconnaissance, global reach justifies the launch cost.

Static images only go so far in visualizing all the different orbits. NASA created a fantastic orbital fly-through video that illustrates the many dimensions of the satellites orbiting Earth.

High Latitude Loiter

The Soviet Union couldn’t use GEO effectively. Their territory stretches across high latitudes — 60°N and above — where geostationary satellites sit low on the horizon. Buildings block the signal. Terrain creates shadows. Trees get in the way.

They also lacked equatorial launch sites, making GEO expensive to reach from Baikonur.

The solution was Molniya orbits. These are highly elliptical paths with perigee around 500 km and apogee near the GEO radius. A satellite in this orbit screams through perigee in minutes but loiters around apogee for eight-plus hours. It hangs nearly stationary over high-latitude regions.

Molniya’s 63.4° inclination builds further advantage.

Earth is not a perfect sphere. It’s slightly squashed, with extra mass around the equator. That bulge pulls unevenly on an inclined, elliptical orbit, creating a steady twisting force that rotates the ellipse in space. The line connecting perigee and apogee slowly turns.

At 63.4° inclination, that twisting cancels out, and the torque sums to zero. The orbit still precesses in other ways, but the ellipse itself stops rotating. Apogee stays fixed relative to Earth.

From Moscow, the satellite appears to hover nearly overhead.

The cost is numbers. A single Molniya satellite only loiters for part of its orbit. Continuous coverage requires at least three spacecraft orbiting in relay.

Yet, three Molniyas beat zero GEOs.

Tundra orbits work similarly. The satellite traces a figure-eight pattern over the ground but spends most of its time over the target region.

Sirius XM originally used three satellites in geosynchronous Tundra orbits to serve North America. This keeps at least one satellite high in the sky at all times, which is critical for car radios that need a clear line of sight through windshields.

Consistent Lighting

Earth observation satellites need to compare images over time. Is the forest shrinking? Is the city growing? Did those trucks move?

This only works if the lighting is consistent. Shadows must fall the same way in January and July, morning and afternoon. Otherwise, you’re measuring sun angle, not ground truth.

Sun-synchronous orbits solve this. They exploit Earth’s equatorial bulge — the J2 perturbation to be specific — to precess the orbital plane at exactly the rate Earth orbits the sun. The result is that the satellite crosses every latitude at the same local solar time, every day, year-round.

The choice of local time depends on the mission. Mid-morning orbits around 10:30 AM balance illumination and shadow. They’re ideal for optical imagery where terrain contrast matters. Dawn-dusk orbits ride the terminator, keeping solar panels in near-constant sunlight and minimizing eclipses. Radar satellites love this. Noon orbits maximize brightness but flatten everything.

Landsat, the longest-running Earth imagery program, has used sun-synchronous orbits since 1972. Planet Labs relies on this orbit for modern commercial imagery. SSO is also popular among reconnaissance satellites focused on change detection. Consistent lighting is the advantage that makes comparison meaningful.

The physics requires an inclination of about 98° between 600 km and 800 km. That’s slightly retrograde, meaning the satellite actually orbits westward against Earth’s spin. You lose the rotation boost at launch and also pay a small penalty on top.

The Transfer Problem

Orbits are commitments. They are expensive to reach and difficult to change once you’re there.

When we think about adjusting an orbit, there’s raising the altitude and changing the inclination. Both cost delta-v — velocity change — and satellites carry limited fuel to perform such a maneuver.

Reaching LEO from Earth’s surface takes roughly 9.4 km/s. This factors in 7.8 km/s of orbital velocity and about 1.4 km/s to overcome gravity and aerodynamic losses.

To raise altitude from LEO to higher orbits, we use the highly efficient Hohmann transfer.

This maneuver moves a spacecraft from a circular orbit to an elliptical orbit and then back to a higher circular orbit. Starting in the first, low circular orbit, the spacecraft burns to raise apogee, entering an elliptical transfer orbit. It coasts to the high point — about five hours for LEO to GEO — then burns again to circularize. Moving from LEO to GEO costs about 4 km/s this way.

Plane changes are different. The cost scales with orbital velocity. In LEO, you’re moving fast — 7.8 km/s — so changing inclination is brutal. At GEO altitude, you’re moving under 2 km/s. The same angle change requires a fraction of the velocity. For this reason, plane changes are usually deferred to apogee and combined with circularization burns.

Once you’re in orbit, you’re largely locked in. A LEO satellite can’t just relocate to MEO without burning significant propellant. Space tugs and orbital transfer vehicles exist, but physics still applies.

Choose Your Advantage

The advantage space provides is real, but it isn’t automatic. It emerges from matching the right orbit to the right mission and accepting the trade-offs that follow. Persistence costs proximity. Coverage costs revisit time. Altitude costs delta-v.

The orbit you choose determines the satellite you build, the rocket you need, the ground segment you operate, and the business model you can sustain. Every decision downstream flows from this one.

Orbit isn’t where a satellite lives. It’s what a satellite can do.

Subscribe now

The breadth started to feel like a grab bag. A little bit of this, some of that. You do the natural thing. Bucket the topics logically, make a schedule that seems like it’ll flow. But it still didn’t feel right. The disparate topics raised a question: why does all this matter? Are these just interesting stories, or is there a there there?

That led me to think about what space is to me. The breadth of topics didn’t collapse. Instead, I came to understand it’s the very thing that makes space something special.

Space inspires.

I watched the Tom Hanks series From the Earth to the Moon when I was a kid. One episode profiled Gus Grissom, a famed Apollo astronaut who tragically died in the Apollo 1 fire.

He visited the factories that built Apollo hardware, walking the floor and meeting technicians.

One scene had him speaking with the toolmakers. They weren’t making the hardware. They were making the hardware that made the hardware, their work often unknown and overlooked.

Grissom spent the time with these guys, seeing their craft and talking about how their work enabled the bigger mission. I think about how proud they must have felt going home that night.

I probably watched this series in eighth grade, but thinking of this scene still chokes me up a bit. It represents so much of what I believe in. And I map it directly to working in space.

Launch is addictive.

The livestreams show the shiny side of launch. But in reality, it’s a brutal endeavour. To get to orbit, every system must push to the edge of its performance envelope.

The scope is also wildly broad, which means you have to keep track of many different systems and risks. Failures happen, and when they do, it’s often something that in and of itself feels stupid and embarrassing. But it was one of hundreds of risks being managed.

It’s a high profile game. The world can watch, and you get immediate feedback, usually in the form of a fireball, if it doesn’t work.

But people — myself included — just keep coming back for more.

I think because there’s something about it that you can’t quite get anywhere else.

It’s expeditionary. You’re out at a remote site running ops. There’s a ticking clock. If there’s an issue in the countdown, you’ll find yourself doing real math to resolve it. You have to make decisions. They have real, fast consequences. When the clock hits T0, all these super high performance systems need to turn on and work at exactly the same time.

It’s really terrible when they don’t. But that’s what makes it so good when they do.

Space is foundational to daily life.

We rely on space every day. We often think of it as the future. Something that’s coming. But space is here and been here for a long time.

GPS is widely understood. We use it to navigate home on Google Maps. But GPS doesn’t just give our position. It provides universal time too. This allows us to precisely synchronize clocks around the world.

We use GPS time for financial transactions from Venmo to credit cards to stock trades. It marks transactions to make sure each settles in the correct order. Bob paid Sally, then Sally paid Susan. If you look closely at the NYSE operations center, you’ll see an array of GPS antennas on the roof. They’re not for position. The building isn’t moving. They’re for time.

Similarly, the power grid needs more energy at 5 pm than at 5 am. But we can’t just hook a power plant up and flip a switch. The loads must be balanced and phases aligned.

Synchrophasors in substations measure the magnitude and phase angle of the alternating current running through the grid. Every measurement is timestamped using a GPS signal. This allows operators to compare measurements from a substation in New York with those from one in Florida at the exact same millisecond.

After the 2003 Northeast Blackout, synchophasors were rolled out to monitor the grid’s heartbeat. They’ve been attributed with a significantly more reliable grid in the years since.

Remove GPS, and we lessen our ability to transfer money and electricity — the two things that underpin an economy.

Space is the ultimate high ground.

We’ve wired our civilization through orbit. Now we need to defend it.

But it’s a unique domain.

It’s easy to find objects in space. Anyone with an amateur telescope can track spacecraft. So for the most part, everyone knows where everything is.

Defenses are difficult. You can’t armor something in orbit in the way you might on Earth. Even with plummeting launch costs, armor is too heavy to fly reasonably. Further, the velocity of objects in space is so high that it wouldn’t be particularly effective.

Attacks can be reversible. An adversary doesn’t need to destroy a satellite to make an impact. They can hold it hostage by jamming the signal or blocking the solar arrays. They can even kidnap it, dragging it to a different orbit. Are these acts of war? It’s unclear, and grey zone operations create a whole new dimension of conflict.

At the same time, space based assets are the enablers of nearly every operation on Earth. Satellites provide early warning of ballistic missile launches. We use GPS to direct troops, ships, planes, and weapons. Orders and intelligence travel the globe instantly via secure space based communications.

In any and every conflict in the future, space will play a role, and it’ll likely be a major one.

It’s also a geopolitical contradiction.

Russia invaded Ukraine in 2022. In the years since, Western states have decoupled from Russian energy, uranium, technology, and investments. Oligarchs’ superyachts have been impounded across the globe. The U.S. and NATO are supplying aid to Ukraine, propelling us into the most extreme proxy war since the Berlin Wall fell.

And yet: Russian and American astronauts still share meals on the International Space Station. They rely on each other for safety. They conduct joint experiments. The same countries that are arming opposite sides of a ground war cooperate daily in orbit.

Space has always held this tension. The Apollo-Soyuz handshake happened at the height of the Cold War. Former enemies built the ISS. The domain seems to exist slightly outside the standard rules of geopolitics.

Space is the next logical place for human exploration.

Humans have ventured nearly every inch of the world. Shackleton aimed at the South Pole. Hillary and Tenzing climbed Everest. Darwin mapped the Galapagos. On a macro scale, we’ve mapped and characterized the Earth. For true exploration, space is next.

Our current technology can get us to the Moon, Mars, and Venus. The first two offer more reasonable conditions for humans. Like on Earth, pure exploration is coupled with strategic and economic opportunity. There’s prime real estate on the Moon that could serve as an outpost, a leaping off point for resource mining, or further advantage for earth-based operations. Mars could serve as Planet B, a second home for humanity.

What comes after is more open ended. It requires new technology — either in propulsion for faster transit, life support for long duration missions, or a new understanding of physics entirely.

It holds the biggest question.

Is there other life out there? The best answer I’ve ever heard was “absolutely, but it doesn’t matter.” We’re separated by light years of distance and millennia of time. If advanced life is out there, it’s irrelevant because we’ll never cross paths.

There’s something settling about this answer, maybe akin to closure. But at the same time, you never know.

Space is all of this.

This breadth is what makes space special. What else brings together childhood ambition, the foundation of our economy, the ultimate battleground for modern war, and the wonder of if we’re alone out here in the universe?

In the last two decades, it’s been the arena for some of the most advanced engineering accomplishments. Vehicles the size of thirty-story buildings float in mid-air before softly landing.

More is coming. Year after year, exotic architectures are contemplated and then actualized, furthering the impact space has on all of our lives.

Sometimes this impact is obvious, while at other times it’s invisible. Either way, something this embedded is worth understanding, and that’s the reason to dig into space.

Subscribe now

It’s October 1944. A P-51 pilot over Europe spots a German Bf 109 fighter on his tail. He shoves the throttle forward, past the stop, and through the safety wire. As the wire snaps, engine power surges from 1,490 to 1,720 horsepower. He races away from danger, but has just a few minutes before the Merlin engine starts eating itself.

That broken wire represents a fundamental engineering reality: nominal systems fail in off-nominal conditions.

The most trustworthy designs don’t just meet the spec, they hold reserve capacity for moments when the spec becomes irrelevant. This reserve isn’t free. Its cost comes in weight, complexity, maintenance, and money. But it’s the difference between a system that survives the unexpected and one that doesn’t.

Three examples — spanning 80 years and three different failure modes — show how high performance systems can be built for the moments when everything goes wrong.

War Emergency Power

In WWII, engine power decided dogfights. More meant a pilot could outmaneuver or outrun the enemy. But what happens when a pilot encounters a threat while already at full throttle? Or stumbles across the enemy’s latest fighter with a bigger engine and higher top speed?

For these situations, American engineers created War Emergency Power.

WEP was a throttle setting beyond 100% of rated power. The throttle interface guarded the activation of this setting. Safety wire across a gate stop physically prevented the throttle lever from advancing. To access WEP, pilots had to sever the wire.

There was symbolism in this action. You were breaking something to get this power.

Boosting power meant cramming more air and fuel into each cylinder. But the challenge was detonation.

In a properly running engine, the spark plug ignites the fuel–air mixture, and a flame front propagates smoothly along the combustion chamber. Detonation occurs when the unburned mixture ahead of the flame front auto-ignites due to excessive heat and pressure.

This causes an extremely rapid pressure rise and shock waves within the cylinder, which can crack pistons and bend connecting rods. Pilots and auto enthusiasts know the tell-tale sound: engine knock.

The simplest way to boost was just opening the throttle wider, cramming more air into cylinders already running hot. This worked, but pushed the engine closer to detonation.

Injecting a water-methanol solution into the supercharger intake offered a clever solution. The water absorbed heat as it evaporated, cooling the intake charge and making it denser. This enabled more oxygen per cylinder. The methanol raised the fuel’s effective octane rating and also provided some cooling. Together, these effects raised the detonation threshold.

The U.S. military continuously developed this approach throughout WWII. By the time the P-51H took flight, WEP boosted power from 1,380 HP to 2,218 HP, an astounding 60% increase.

The Germans did the same with their MW 50 (Methanol Wasser 50) system. It could push a BMW 801 engine from 1,600 HP up past 2,000 HP.

They also fielded the GM-1 (Göring Mischung 1) system for high-altitude operations, where thin air starved engines of oxygen. This system injected liquid nitrous oxide directly into the supercharger intake. As the nitrous oxide decomposed in the heat, it released oxygen, chemically enriching the mixture where the atmosphere couldn’t.

The trade-off was explicit. Activating WEP triggered enhanced inspections and often a complete engine teardown. Technicians pulled the Merlin apart, looking for cracked pistons, scored cylinders, and damaged bearings. Every second of emergency power consumed engine life at a far accelerated rate.

But the designers weren’t trying to build a sustainable system. They were buying pilots a way out. The wire existed so that the cost of using it was clear. But if breaking a wire meant surviving, it was an obvious trade.

Subscribe now

Takeoff Contingency Power

Thirty years later, a similar principle appeared in a very different context. The Concorde carried 100 passengers at Mach 2, powered by four Olympus 593 turbojets. These engines each had a maximum continuous thrust of 28,800 lbf, which generated 736°C exhaust gas. At this power, the engines could run continuously for an unrestricted duration.

As with any aircraft, one of the most dangerous scenarios was losing an engine during takeoff. At this point, the plane was at maximum weight, minimum speed, and had no altitude to trade. With few options, this required a different kind of margin.

The Olympus 593 had two thrust ratings that mattered here: takeoff power and contingency power.

Takeoff power with reheat on (afterburners) generated 37,080 lbf. This was already 28% higher than cruise thrust and drove exhaust gas temperatures to 806°C. Under these conditions, the maximum duration was five minutes.

Contingency power pushed harder still.

This setting existed solely for engine-out scenarios during the takeoff roll. It drove the engines to 38,130 lbf and an exhaust gas temperature of 883 °C. At this pace, the engines had just two and a half minutes before thermal limits became structural limits.

Activation was automatic. When an engine’s N2 — the rotational speed of its high-pressure compressor — fell below 58.6 percent, the system reacted automatically. The takeoff monitor throttled the remaining engines up to contingency power to compensate. The pilots didn’t break a wire. The system detected and responded on its own.

This automation reflected a shift in how engineers thought about off-nominal operations. WEP assumed a pilot making tactical decisions under fire. Concorde’s contingency power assumed a crew managing a crisis with passengers aboard. The cognitive load had to stay low, and the response had to be immediate.

The trade was time-limited by thermal margins. The material selection in the Olympus 593 engines was tailored to handle high temperatures. Yet, at 883°C, the exhaust gas caused many engine parts to approach material limits.

Two and a half minutes were enough to clear the runway, establish a climb, and retract the gear. But they had to be done quickly and deliberately. The system provided immediate survival, not extended comfort.

In Flight Abort

Jumping ahead to modern day, the SpaceX Dragon capsule is the primary ride to space for U.S. astronauts. The vehicle carries two propulsion systems: Draco and SuperDraco. Both use monomethylhydrazine (MMH) and nitrogen tetroxide (NTO), which are hypergolic propellants. They combust on contact, with no spark required, making them highly reliable.

The sixteen Draco thrusters handle orbital maneuvering, with each generating 90 pounds of thrust. They’re designed for long-duration operations: attitude control, orbit adjustments, and deorbit burns.

The SuperDracos exist for a different purpose entirely. Eight engines are built into the capsule’s sidewalls, with each producing 16,000 pounds of thrust. All are oriented facing aft and together generate 128,000 pounds of thrust.

If the Falcon 9 beneath fails, the SuperDracos fire and accelerate Dragon away at about four g’s. The burn lasts roughly five seconds. In this time, the capsule travels nearly a mile, which is enough separation to be survivable. Once clear, Dragon’s trunk separates, the capsule reorients, and parachutes deploy.

Here, the engineering achievement wasn’t just in the engines, it was also in how they integrated into the vehicle.

Traditional launch escape systems on Mercury, Apollo, and Soyuz used solid-fuel towers mounted ahead of the capsule. They worked, but often had to be jettisoned partway through ascent because the tower’s aerodynamic fairing hit limits. This created blackout zones during ascent when the abort system was gone, but the rocket could still fail. In this window, the crew was unprotected.

Dragon’s integrated approach eliminates blackout zones. The SuperDracos are part of the capsule structure, not a jettisoned tower. The abort system remains continuously active from the pad straight through to orbit.

The time scales here are different from anything that came before. WEP gave pilots five minutes of emergency power before the engine needed a teardown. Concorde’s contingency rating allowed 2.5 minutes at elevated thrust. Both assumed human operators would recognize the problem, make a decision, and take action.

SuperDraco operates faster than human cognition. The transient from ignition to full thrust is 100 milliseconds. The system monitors Falcon 9 telemetry and, if parameters exceed limits, automatically fires. The entire chain of fault detection to separation happens faster than a pilot could process something going wrong.

Dragon carries the entire SuperDraco system purely for emergency off-nominal scenarios. At one point, SuperDracos were meant to double as propulsive landing engines. SpaceX dropped that in 2017, though recent filings suggest it may return for contingency landings. But generally, during a good mission, the system is never activated.

The SuperDraco system adds mass, complexity, and cost to every mission. But when you’re sitting on a million pounds of propellant, it’s the only way out.

The Trade

Every off-nominal system is a bet about which components you can afford to lose.

WEP destroyed engines. Concorde’s contingency rating shortened the time between overhauls. Dragon’s SuperDracos add mass to every mission. None of these reserves is free. They represent decisions to accept penalties during normal operation in exchange for capability in extraordinary times.

This is the hardest kind of engineering: paying every day for something you hope never to use.

Such systems are easy to cut. They add cost without adding capability — until the day they’re the only capability that matters. When written down and considered in detail, the value is obvious. Yet in design reviews, someone always asks: do we really need this margin? Every flight where the reserve goes unused looks like evidence that it wasn’t needed.

This works until it doesn’t.

The P-51 pilots, Concorde crews, and Dragon astronauts all know. When the failure cascade starts, reserve capacity is the only margin that matters.

Subscribe now

Photo Credits

• P-51 Throttle Quadrant / WEP Diagram: U.S. Army Air Forces, Pilot Training Manual for the P-51 Mustang (AAF Manual 51-127-5), 1945. Public Domain, via t6harvard.com

• P-51 Mustang: Courtesy of To Fly and Fight / CE “Bud” Anderson (toflyandfight.com)

• Concorde at Mach 2: Adrian Meredith / Royal Air Force, via PetaPixel

• Olympus 593 Engine: Heritage Concorde (heritageconcorde.com)

• SpaceX Draco and SuperDraco: SpaceX, via Spaceflight Now and SpaceNews

• Apollo Launch Escape System: NASA

A cut wire. A flooded connector. A misbehaving device that spams the network. Any one of these can bring down the system.

The best network architectures make failures irrelevant. Critical functions keep running even when components have failed. This essay explores three strategies to make this happen:

1. Traffic Separation: Isolating critical and non-critical systems

2. Redundant Paths: Multiple channels carrying the same data

3. Self Healing Topologies: Networks that route around failures automatically

Combined, these approaches keep a system functioning through conditions that wouldn’t otherwise seem survivable.

Traffic Separation

The simplest network connects everything to everything else. A single bus carries all messages between all nodes. This works well until one component acts up.

Consider a car’s Controller Area Network (CAN) bus. Bosch introduced CAN in 1986 to simplify automotive wiring, and by the mid 1990s, it was standard equipment on most vehicles. The protocol allows electronic modules to communicate without a central computer. Any node can transmit, and an arbitration scheme establishes message priority.

Modern cars can have upwards of fifty control modules transmitting data.

But a single network has a weakness.

A faulty node can monopolize the bus by transmitting continuously, drowning out messages from others. This is known as the babbling idiot failure. One malfunctioning component can silence an entire network.

This can create real risk. If the controller that operates the door locks starts spamming, the brakes might not be able to communicate with the antilock system.

The solution is separation. Don’t put non-critical traffic on the same network as critical traffic.

Modern vehicles use multiple networks, each with a different mission:

• The powertrain network handles the engine and transmission.

• The chassis bus manages steering, brakes, and suspension.

• The body/comfort bus coordinates windows, locks, and seat controls.

• The infotainment network handles audio, video, and phone connectivity, often using MOST (Media Oriented Systems Transport), a protocol designed for high-bandwidth media.

A central gateway routes messages between buses when needed. But if a Bluetooth module goes haywire on the infotainment bus, the powertrain bus keeps running.

Some cars also include a reserve network. If any one network fails, its traffic can failover to the backup. I’ve often heard this called the party bus, where everyone’s invited when things go wrong.

Even More Separation

Aircraft take a different, more separated approach. Starting in the 1970s, ARINC 429 became the standard. Unlike CAN, it is unidirectional. Each bus has exactly one transmitter that can talk to up to twenty receivers. If you need bidirectional communication, you use two buses.

This architecture is deliberately simple. There’s no arbitration because there’s nothing to arbitrate. Only one device can ever transmit. A babbling idiot can only corrupt its own bus, and since each transmitter has a dedicated bus, a failure affects only the receivers on that specific wire.

ARINC 429’s cost was weight. All those dedicated buses added up to significant wiring. The standard remains onboard older aircraft, but newer designs have shifted to approaches such as AFDX, which combine determinism with network efficiency.

Redundant Channels

Separation protects critical functions from non-critical failures. But what about failures within the critical systems themselves?

For safety-critical applications — brakes, steering, fly-by-wire controls — a single bus isn’t enough. The solution is redundancy: run two or more independent channels carrying the same data.

FlexRay, introduced in 2006, was designed specifically for this purpose. Where CAN offers a single shared bus, FlexRay provides dual channels that can operate independently. The protocol was developed to enable “x-by-wire” systems: brake-by-wire, steer-by-wire, throttle-by-wire.

In a FlexRay brake-by-wire system, brake commands could travel on both Channel A and Channel B simultaneously. If Channel A fails — a cut wire, a shorted connector, a dead transceiver — Channel B continues operating. The brake controller receives identical data on both channels during regular operation and continues on the surviving channel after a failure.

The first production vehicle with FlexRay was the 2006 BMW X5, which used it for an adaptive damping system. Today, FlexRay remains common in premium vehicles for safety-critical applications, even as newer protocols emerge.

FlexRay also includes a bus guardian, which is dedicated hardware that monitors each node and disconnects it if it starts transmitting outside its assigned time slots. A faulty node can’t monopolize the bus because the guardian physically prevents it from transmitting.

Military aircraft have used this dual-channel approach for decades, and spacecraft do the same. These systems often go further with cross-strapping, which configures multiple redundant pathways between nodes so a backup link takes over if the primary fails.

Subscribe now

Self-Healing Rings

Dual channels protect against link failures. But if a node dies, everything downstream loses connectivity. Ring topologies can elegantly address this problem, creating self-healing networks that can survive a break in the chain.

Consider a linear four-node network. A break anywhere could sever communications.

Now, imagine four nodes connected in a circle, with data flowing clockwise. If the wire between Node A and Node B breaks, Node A can no longer reach Node B directly. But it can still reach Node B by going the long way around, through Nodes D and C.

A simple ring provides exactly one alternative path. But what if two wires break?

The solution is the dual counter-rotating ring. Two rings run through the same nodes, but data flows in opposite directions, clockwise on the primary ring, counterclockwise on the secondary.

During normal operation, you can use both rings for bandwidth, or keep one in reserve. When a single break occurs, traffic routes around it on either ring. When two breaks occur in different places, the rings wrap at each break point, forming a single loop that still connects all surviving nodes.

Fiber Distributed Data Interface (FDDI) pioneered the dual counter-rotating ring in the 1980s for local area networks. FDDI itself is largely obsolete, but the architecture lives on in telecommunications backbones, electrical grid networks, and automotive systems.

Tying it All Together

Modern aircraft such as the A380 and 787 have adopted AFDX (Avionics Full-Duplex Switched Ethernet), which combines many of the strategies explored above.

• Virtual links: Dedicated logical channels with guaranteed bandwidth and bounded latency, creating the determinism that safety-critical systems demand

• Redundancy: Dual networks with sequence numbering so receivers can detect and discard duplicates, similar to FlexRay’s dual-channel approach

• Traffic policing: Switches enforce bandwidth limits, acting as bus guardians that prevent any device from flooding the network

The result combines the best of both worlds. AFDX gets the wiring simplicity of a multi-transmitter network, while maintaining fault isolation. Switches contain babbling idiots to their local segment. Redundant networks ensure a single cable failure doesn’t bring down the system. And bounded latency guarantees that brake commands arrive when they need to, every time.

It seems straightforward on paper. But it took decades of careful systems engineering to create such an elegant approach.

Looking Ahead

So far, we’ve explored how systems absorb failures to stay nominal. But sometimes nominal isn’t enough, and a system needs to go beyond its design limits to survive an emergency.

Next, we’ll look at how systems can exceed their 100% operating point to overcome extraordinary conditions, sometimes at the cost of the hardware itself.

a sub-discipline of systems engineering that emphasizes the ability of equipment to function without failure.

This framing relegates it to a specialty. A box to check. A task for a dedicated team. But reliability engineering is just engineering.

It’s the entire point.

Complex systems operate with razor-thin margins, under demanding conditions, and for extended durations. They’re also designed under considerable uncertainty. This creates thousands of ways to fail.

Good designs are able to both avoid failure modes and tolerate them. Reliability means the part works. Redundancy means there’s an alternative if it doesn’t. Working these concepts directly into a system’s architecture can make it inherently trustworthy.

This series of As Built essays is about doing just that. We’ll explore a variety of approaches and architectures that enhance system confidence. The end product will be a menu of tried-and-true methods that can be adapted to any design.

Most operational systems carry a myriad of sensors. Some of these are engineering sensors to characterize performance and improve the design. Others are control sensors that the system actually uses to make decisions.

If a control sensor fails, it could end the mission. So we rarely rely on just one.

Are two better than one?

Consider the embarkation door on a cruise ship. It is relatively low on the hull, and passengers board through it at the dock. When the ship is at sea, this door must be sealed to protect from large waves.

The door has a contact switch that gives a binary reading: open or closed. The captain relies on this data to keep the ship safe. But what if the sensor fails?

We could add a second contact switch to achieve dual modular redundancy. On its surface, we've solved the problem—now we have a backup. If one switch completely fails and stops sending data, we could trust the remaining sensor.

But this architecture has limits. But what if one sensor reads open and the other says closed? Which do we believe?

While dual modular redundancy provides a backup if one sensor fails, it doesn’t enable error correction. If the two sensors contradict, the captain faces a difficult situation.

Three voters

By adding a third contact switch, we can accommodate contradiction through majority voting. Each sensor casts a vote: open or closed. A voting apparatus tallies these votes and outputs the majority value.

If two sensors read closed, but one reads open, then the voting system will output that the door is closed. Today, this is typically a small piece of software, but for Boolean values, it can also be created with analog circuits.

This is a triple-modular-redundant system, and it allows us to be one fault tolerant. In other words, one of the three sensors can fail, and we can still function normally.

For our ship door, this arrangement works well. But what if the data streams were more complicated than just open or closed?

Non-Boolean values

Let's consider another part of the ship: the steering system. As the vessel navigates the Pacific, it uses hydraulic rams to turn the rudder. The rudder weighs 100 tons and encounters significant hydrodynamic pressure as it turns. To overcome these loads, the piston rams are powered by hydraulic fluid at 2000 psi.

Unlike the door contact switch, there are two failure modes here: low and high. If the hydraulic pressure is low, the system can’t steer the ship. If the pressure is high, the system bursts.

Therefore, the captain must have an accurate understanding of the hydraulic pressure, which we collect with three redundant pressure sensors.

Our triple modular redundancy voting system works, but requires slightly more logic. Assume we have three pressure sensors. Our algorithm would be:

• First, establish which two sensors are in closer agreement.

• Second, take the average of the two sensors in close agreement.

For non-Boolean values, we need this second averaging step to address the variance between the two sensors that agree. They won’t read the exact same value due to calibrations, sensor drift, and general uncertainty. To achieve precision, we take the average.

Can it be simpler?

The frameworks listed so far achieve single fault tolerance, but require the sensor layer, the voting layer, and, for non-Boolean values, the final averaging action.

We can simplify by taking the median.

Consider the three hydraulic pressure sensors. Instead of voting, we read all three data streams and at each sample point, take the median. The composite median time series is displayed to the ship captain.

The beauty of this arrangement is how it passively handles a variety of fault conditions.

During nominal operation, we'd expect to see three pressure traces, tightly grouped. In this example, Sensor B is in the middle and is therefore reported as the median value.

However, maybe an exhaust leak starts to cause Sensor B to heat up. This results in the sensor drifting upward and taking a 200 psi offset. In this case, the median is Sensor B until it crosses above Sensor A, then Sensor A data is reported as the median. All of this happens automatically and transparently to the ship captain.

This arrangement also accommodates failed sensors. Maybe a big wave rocks the ship and breaks the wires to Sensor B, causing it to drop to 0 psi. The system seamlessly responds to now report Sensor C as the median.

In this setup, we don’t require voting or averaging. We simply always take the median of our three sensors, and in doing so, remain single fault tolerant. The best architectures don't detect failures. They absorb them.

Flagging faults

While the median operating system seamlessly handles failures, the issues still need to be reported to the ship captain. If we’re single fault tolerant and we’ve taken a fault, we can’t take another one.

It can be helpful to act on the median and flag any sensor that deviates from it. In the hydraulic pressure case, perhaps the system flags if any one pressure is more than 50 psi away from the median.

With this insight, the captain can take an action. This might be going down to repair the sensor, operating the ship more cautiously, or even returning to port.

Common mode failures

These setups all protect against one fault. We can architect systems to accommodate multiple faults by adding sensors and logic, but the effort can yield diminishing returns and increased complexity.

Common mode failures are also a consideration. This would be an event that damages more than one sensor. Maybe that exhaust leak in the engine compartment heats all three sensors, causing them to all drift high and trigger a false alarm.

For this reason, sometimes it’s beneficial to separate the sensors, moving them physically apart or routing the wire harnessing along different paths so it’s less likely that all get impacted by the same event.

Tying it all together

The most basic aspect of reliability is getting good data, and I’ve always viewed the triplicated median approach as one of the most elegant ways to do so.

The next part of this series will step up a level from sensors to networks. We’ll look at network redundancy and how to manage failovers when a component hits an issue.

In nearly every case study, we’ll see the same thing. Simplicity always wins—the fewer systems, the fewer possible failures. The simpler the logic, the easier it is to prove every case will work. And that’s the entire point.

I learned more while writing the piece than any other, and my thoughts on the nuclear fuel cycle have evolved significantly. Hopefully you find it interesting as well, and I'd love to hear your thoughts!

This is the second piece in a two-part series on nuclear fuel. While reactors garner significant attention, creating the fuel that powers them is equally challenging—if not more so.

In the last essay, we posed two questions:

1. How exactly do we fuel reactors?

2. Does the U.S. have nuclear fuel security?

We addressed the first, uncovering some striking insights:

• Except for Canada, every major nuclear power (the U.S., France, Russia, and China) depends on uranium imports.

• Russia controls 43% of global enrichment capacity.

• The U.S. continues to import enriched uranium from Russia; U.S. companies have secured waivers to do so through 2027.

• From 1993 to 2013, Russia turned warheads into reactor fuel, supplying 10% of U.S. electricity during that period.

We also began exploring fuel security. The U.S. is highly reliant on foreign supply chains, most acutely in mining and enrichment, where the annual needs are:

• 15,500 metric tons (MT) of uranium—99% imported

• 12 million separative work units (SWU) of enrichment—60% imported

But fuel security isn’t just about today’s supply. It’s about future options. How hard is it to get more reactor fuel? And how would we do it?

This essay examines eight opportunities. None is a silver bullet. Each chips away at the problem, some better than others. They also aren’t mutually exclusive, and in fact, a composite strategy is almost certainly required. Combined, however, they can chart paths to assured access to nuclear fuel.

Setting the Stage

This is a thought experiment. It explores the art of the possible, and doesn’t necessarily advocate for any one strategy. Assumptions are intended to be reasonable, and numbers are grounded in industry and government reporting.

We juggle many values—types of uranium, enrichment levels, etc.—so some are rounded for readability. Yet, no conclusions are altered from this.

Since any improvements to the fuel chain will play out over the next five to ten years, we set the baseline need in 2035 to a 33% increase in uranium and enrichment demand relative to today. This is a middle-of-the-road estimate, and exact growth depends on how quickly new plants come online.

We make two other key assumptions. First, the U.S. light water reactor fleet uses fuel enriched to an average of 4.4% U-235. Second, modern centrifuge enrichment produces tails at 0.3% U-235. Tails are the leftovers from enrichment—the part that didn’t get enriched. Their concentration represents U-235 that wasn’t extracted for nuclear fuel (as a reminder, natural uranium contains 0.7% U-235).

Our 2035 scenario therefore demands:

• 20,000 MT of 0.7% natural uranium (NatU)

• 16 million SWU of enrichment

• 0.3% centrifuge tails assay

• 2,000 MT of 4.4% low enriched uranium (LEU)

First, we’ll build a menu of opportunities. Then we’ll assemble them into a representative plan.

Opportunity 1: Import Foreign Uranium and Rely on Diplomacy

The first option is the least drastic—akin to doing nothing. Today, we leverage an extensive international nuclear fuel supply chain, buying natural uranium along with conversion and enrichment services. We can sustain and grow this by reinforcing global relationships.

As of 2023, the largest natural uranium suppliers to the U.S. are Canada (27%), Kazakhstan (22%), Australia (22%), Russia (12%), and Uzbekistan (10%). The rest comes from a long tail of smaller suppliers. Russia’s share has diminished since the Ukraine conflict began.

Canada and Australia are close U.S. allies, and both could increase uranium exports. Australia holds the world’s largest uranium reserves—about one third of the global total—yet accounts for only 8% of global output. Spare capacity exists.

The story in Central Asia is different.

In Kazakhstan, the U.S. faces fierce competition with Russia and China. In June 2025, Rosatom was selected to build the country’s first nuclear power plant; China is expected to build the second. Both countries are using reactor deals and entangled mining agreements to exert influence.

Similar Russian and Chinese conversations are happening in Uzbekistan, but at an earlier stage.

The current U.S. strategy is to pull these uranium stores into the Western supply chain. But as Central Asian nuclear value chains become tied to Russian and Chinese programs, more output could become politically or contractually locked to Eastern-aligned customers, leaving less flexibility in a crisis.

Opportunity 2: Mine Uranium Domestically

In 2022, the U.S. mined 75 MT of uranium. This jumped 4x to about 300 MT in 2024, but still represents a tiny fraction of the U.S. need.

This wasn’t always the case.

From 1950 to 1980, the U.S. was the world’s largest uranium producer. Early in this period, over 750 uranium mines operated across the country. At the peak in 1980, the U.S. mined 16,800 MT. This exceeds today’s annual uranium demand, highlighting the extent of uranium overproduction during this period.

The U.S. has modest uranium stores. As of 2022, an estimated 168,200 MT sits in reasonably accessible reserves—about 8.5 years of 2035 demand.

Tapping this would require scaling domestic production by over 50x. A tall order, but possible given our heritage from the 1980s. However, since then, environmental and political opposition has mounted.

Activation costs for new mines depend heavily on local factors—exploration success, geology, and logistics. Denison’s Midwest Main project in Saskatchewan offers a reference point. It’s planned to produce 2,700 MT annually for six years, with total capital costs projected at about $500M.

Opportunity 3: Expand Domestic Enrichment Capacity

The U.S. currently has about 5 million SWU of enrichment capacity at URENCO’s New Mexico facility. Centrus Energy also has limited enrichment capability at their demonstration plant, but it’s focused on producing HALEU (high-assay low-enriched uranium) at about 20% enrichment for next-generation reactors.

Two U.S. expansion projects are already planned:

• URENCO will add 1.4 million SWU by 2027

• Orano’s Project Ike at Oak Ridge will add “several million” SWU

Combined, these would bring U.S. enrichment capacity to around 10 million SWU.

To meet the 2035 scenario, another 6 million SWU must be installed. Based on recent projects, enrichment facilities cost $700 to $1,200 per SWU installed—a $4.2B to $7.2B investment.

Notably, URENCO and Orano are both European companies. To achieve full coverage of 2035 enrichment needs from domestic providers would cost $11.2B to $19.2B.

Beyond cost, centrifuge technology is a factor. Most commercial enrichment runs on gas centrifuges descended from European Zippe/URENCO designs. However, the Department of Energy is funding Centrus Energy to develop domestic centrifuges.

Subscribe now

Opportunity 4: Reenrich Depleted Uranium

When uranium is enriched, depleted uranium is left behind. Usually, the concentration is 0.2%–0.3%. Enriching to a lower tails concentration is more expensive, and is traded against the cost of procuring more natural uranium. Yet, if we’re supply constrained, it’s possible to reenrich the depleted uranium tails, extracting additional U-235.

The Department of Energy has 750,000 MT of depleted uranium stockpiled, which we’ll assume has an average U-235 concentration of 0.3%. If we can process down to 0.1% tails, this yields about 35,000 MT of 4.4% LEU—equivalent to enriching 350,000 MT of 0.7% NatU, or about 18 years of 2035 scenario demand.

We can also do the same for newly mined uranium.

Our baseline assumes 20,000 MT of 0.7% NatU is enriched to 2,000 MT of 4.4% LEU with 0.3% tails. If we drop tails to 0.1%, we produce 2,850 MT of 4.4% LEU.

That’s 42% more product. A substantial change. We could reduce our natural uranium needs by about 30% and still achieve our current fuel production levels.

Existing gas centrifuges or laser enrichment, an emerging method, could achieve this. SILEX (Separation of Isotopes by Laser Excitation) uses lasers to excite and separate U-235 from uranium hexafluoride selectively. It claims to require substantially less energy, potentially making lower tails assays more economical.

Opportunity 5: Downblend Uranium Weapons

As we saw with Megatons to Megawatts, nuclear fuel can be made from warheads or other highly enriched uranium.

As of 2023, the U.S. had 481 tonnes of highly enriched uranium (HEU), allocated as:

• 361 MT for weapons

• 87.5 MT for naval reserve

• 14.3 MT for research reactors

• 18.4 MT for downblending

Assuming reserves are enriched to 90%, then 18.4 MT of 90% HEU could be downblended to 375 MT of 4.4% LEU, which is equivalent to 3,750 MT of 0.7% NatU, or about two months of 2035 demand.

If all HEU—including every weapon in the arsenal—were converted to reactor fuel, it would yield 9,800 MT of 4.4% LEU, or five years of 2035 demand. But the U.S. isn’t going to abandon nuclear weapons.

Therefore, while downblending weapons uranium is possible, current stores don’t move the needle much.

Opportunity 6: Use Plutonium Weapons for MOX

Mixed Oxide Fuel (MOX) is a blend of 3% to 12% plutonium and depleted uranium. The resulting mixture is approximately equivalent to LEU and can power many reactors. In France, about 10% of electricity comes from MOX.

Besides uranium weapons, the U.S. also has plutonium stores, some of which could be converted to MOX.

The last detailed reporting on the U.S. plutonium stockpile was in 2012:

• 99.5 MT total plutonium

• 66 MT considered surplus, not slated for weapons

• Of the surplus, 34 MT is slated for conversion to MOX

Assuming the MOX uses 5% plutonium, we can calculate the reactor fuel produced and the natural uranium it would replace:

• 34 MT plutonium creates 680 MT MOX, replacing 6,800 MT of 0.7% NatU

• 66 MT plutonium creates 1,320 MT MOX, replacing 13,200 MT of 0.7% NatU

Utilizing plutonium for MOX isn’t a volume play. It represents less than a year of the 2035 need. Instead, it’s most useful to jumpstart fast reactor development, which requires MOX at higher concentrations. This is where DOE’s current commercial plutonium allocations are expected to go.

Opportunity 7: Reprocess Spent Reactor Fuel

In the U.S., this spent fuel is not reprocessed. Instead, it’s stored at plants in spent fuel pools or dry casks. The U.S. has accumulated about 90,000 MT of spent nuclear fuel over sixty years, adding about 2,000 MT annually.

In other countries, spent fuel is reprocessed to create additional fuel. Globally, about 400,000 MT of spent fuel have been produced, of which roughly 30% has been reprocessed.

Spent reactor fuel from light water reactors consists of roughly:

• 95% Uranium-238

• 1% (or less) Uranium-235

• 3% Fission products

• 1% Plutonium

• 1% Minor actinides (neptunium, americium, curium)

Reprocessing uses chemical and pyro processes to separate spent fuel constituents:

• Plutonium can be used to make MOX

• Uranium-235 can create traditional fuel

• Uranium-238 can be used in MOX or loaded into Pressurized Heavy Water Reactors

Separation creates a proliferation challenge. If plutonium is isolated, it could be stolen or diverted to weapons programs. This is one of the main reasons the U.S. has opposed domestic reprocessing. However, strategies exist to combat this risk. Plutonium can be converted into MOX almost immediately, limiting its time in pure form. Even better, some processes never completely separate it, leaving plutonium partially mixed with uranium before conversion.

The U.S. could use reprocessing to extract useful fuel from decades of spent fuel stores. The 90,000 MT of spent fuel contains about:

• 900 MT plutonium, which can create 18,000 MT MOX, replacing 180,000 MT NatU (~9 years of demand)

• 90,000 MT of 1% recovered uranium creates 18,800 MT of 4.4% LEU, replacing 188,000 MT NatU equivalent (~10 years of demand)

Combined, reprocessing the U.S. stockpile could net 19 years of reactor fuel supply.

Additionally, every batch of reactor fuel could be reprocessed going forward. Given 2,000 MT of annual spent fuel:

• 20 MT plutonium creates 400 MT MOX, replacing 4,000 MT NatU equivalent

• 2,000 MT of 1% recovered uranium creates 420 MT of 4.4% LEU, replacing 4,200 MT NatU equivalent

Combined, this yields about 820 MT of reactor fuel, which 41% of the original 2,000 MT fuel load. Note that the uranium portion would need to be further enriched, which is reflected in these numbers.

If we did once-through reprocessing, this would reduce our overall natural uranium needs by about 29% for the equivalent fuel production.

Reprocessing is a proven approach. Orano operates La Hague, a large reprocessing complex on the Cotentin Peninsula in northern France. Built in the 1960s for about $20B in modern dollars, the plant handles about 1,700 MT of spent fuel per year.

Opportunity 8: Deploy Fast Reactors

Fast reactors offer a fundamentally different approach to nuclear fuel. Rather than finding new uranium sources, they extract more value from what we already have. In a fast-neutron spectrum, abundant U-238 can be converted into fissile plutonium-239 while generating power.

These reactors typically use MOX or metallic alloy fuels with higher fissile content than light water reactors. More importantly, their fuel can reach significantly higher burnup levels, extracting far more of the energy contained initially in natural uranium.

Conventional light water reactors consume only a small fraction before fuel is discharged. Fast reactors with closed fuel cycles can, in principle, use a much larger share, improving resource efficiency by an order of magnitude or more.

This has significant implications for spent fuel. The stockpile at U.S. reactor sites contains an enormous amount of untapped energy. Fully exploiting it would require extensive deployment of fast reactors and mature recycling infrastructure, but the theoretical resource is vast.

The obstacle isn’t physics. It’s time, infrastructure, and cost. Fast reactors require an initial inventory of plutonium or high-assay uranium, along with reprocessing facilities to close the fuel cycle. Building this out would take decades, substantial capital, and the persistence required for long-horizon hardware programs.

Russia operates the only commercial fast reactor program today, and even there, the fleet is small—just two reactors. For the U.S., fast reactors could be a long-term strategic option, more so than a near-term supply chain fix.

Putting it all together

All of these opportunities can be combined in different ways. Below is what one path for the U.S. might look like. It’s somewhat idealized, and inefficiencies may erode certain numbers. Impacts should be considered on an order-of-magnitude basis.

Backdrop
We start with our 2035 scenario and assume Central Asian geopolitics have trended unfavorably.

• We need 20,000 MT of 0.7% NatU and 16 million SWU per year to create 2,000 MT of 4.4% LEU.

• Russia, Kazakhstan, and Uzbekistan stop exporting uranium to the U.S.

  • Impact: 13,600 MT of 0.7% NatU annual supply (68% of need)

Step 1
Step 1 actions are mainly operational and can be completed without significant new technology. They could fit within a 5-year horizon.

• Increase domestic uranium production 10x to 3,000 MT for $500M

  • Impact: Uranium supply increased to 16,600 MT (83% of need)

• Increase Canadian and Australian uranium imports by 35% (+3,400 MT)

  • Impact: Uranium supply increased to 20,000 MT (100% of need)

• Complete planned U.S. enrichment buildouts by URENCO and Orano

  • Impact: 10 million SWU (62% of need)

Step 2
Step 1 shores up the uranium supply but still requires significant imports of both uranium and enrichment services from abroad.

• Activate reenrichment (SILEX and centrifuge) and enrich down to 0.1% tails, reducing uranium need by 30%

  • Impact: Uranium need drops from 20,000 MT to 14,000 MT

• Activate a 2,500 MT reprocessing facility to increase fuel utilization by 29% using both recycled plutonium and uranium. Cost estimated at $20B.

  • Impact: Uranium need drops from 14,000 MT to 9,950 MT, 30% from the U.S.

• Build an additional 10 million SWU of domestic enrichment at a cost of $7B to $12B

  • Impact: Support for 100% domestic new uranium enrichment, reenrichment, and reprocessing

Reserves
Once SILEX and reprocessing are online, we can also tap existing stockpiles.

• Reenriching the U.S. stockpile of depleted uranium to 0.1% tails provides 18 years of reactor fuel.

• Reprocessing the existing stockpile of spent reactor fuel provides 19 years of reactor fuel.

Combined, this provides 37 years of reactor fuel, which is a sizable buffer.

This two-step scenario allows the U.S. to reinforce supply through mining and allied partnerships while building out infrastructure for laser enrichment and reprocessing. While it enables complete detachment from Kazakhstan and Uzbekistan, maintaining interests in Central Asia is geopolitically beneficial.

Outlook

The opportunities in this scenario are just one way to achieve nuclear fuel security. There are many others. In this case, we maintained some allied imports to avoid exhausting domestic uranium stocks too quickly. A 100% domestic solution is also possible, but would rely on finding or accessing increased uranium stores. We could also consider creating natural uranium reserves through stockpiling.

The main takeaway: pathways exist. They’re not short—perhaps 10 to 20 years. And they’re not cheap—tens of billions of dollars. But they’re there and growing. While SILEX scaling still needs to be demonstrated, reprocessing has a 60-year heritage.

There are risks. SILEX and reprocessing both present proliferation challenges. Laser enrichment units are expected to be relatively small and consume little power. This creates a risk that they will walk away into the wrong hands. Reprocessing can isolate plutonium, which also makes it potentially mobile. Even a small amount could accelerate a weapons program.

Fast reactors represent another paradigm. The best time to ramp up these efforts would be after reprocessing comes online. To be maximally effective, the U.S. would want a tailored reactor fleet—light water reactors, fast reactors, and potentially some CANDU plants that don’t require enriched fuel. Many studies have explored such arrangements, but implementation requires a cohesive, state-led strategy.

The needs could also change. We didn’t explore HALEU, which is enriched to ~20%. Many new reactors are designed around this fuel type, and broad deployment would significantly increase enrichment demand.

Ultimately, all this must be considered as the U.S. seeks to jumpstart the nuclear sector. There’s much focus on the reactor fleet and emerging technologies—but without fuel, no electricity gets generated.

Subscribe now

General References

Congressional Budget Office, Nuclear Power’s Role in Generating Electricity (Washington, DC: CBO, 2008), https://www.cbo.gov/sites/default/files/cbofiles/ftpdocs/88xx/doc8808/11-14-nuclearfuel.pdf.

Reja Younis, “Pitting Nuclear Modernization Against Powering AI: Trump’s Plans for the US Plutonium Stockpile,” Center for Strategic and International Studies, 2025, accessed November 28, 2025, https://www.csis.org/analysis/pitting-nuclear-modernization-against-powering-ai-trumps-plans-us-plutonium-stockpile.

Denison Mines, “Denison Announces Results from Midwest ISR Preliminary Economic Assessment,” news release, October 2024, accessed November 28, 2025, https://denisonmines.com/news/denison-announces-results-from-midwest-isr-prelimi-122828/.

World Nuclear Association, “Australia,” accessed November 28, 2025, https://world-nuclear.org/information-library/country-profiles/countries-a-f/australia.

“Kazakhstan Selects Rosatom for First Nuclear Power Plant,” World Nuclear News, June 2025, accessed November 28, 2025, https://www.world-nuclear-news.org/articles/kazakhstan-selects-rosatom-for-first-nuclear-power-plant.

Eugene Rumer, “Back to the Future? Kazakhstan’s Nuclear Choice,” United States Institute of Peace, October 2024, accessed November 28, 2025, https://www.usip.org/publications/2024/10/back-future-kazakhstans-nuclear-choice.

World Nuclear Association, “Kazakhstan,” accessed November 28, 2025, https://world-nuclear.org/information-library/country-profiles/countries-g-n/kazakhstan.

U.S. Energy Information Administration, “U.S. Uranium Production Increased in 2023,” Today in Energy, June 2024, accessed November 28, 2025, https://www.eia.gov/todayinenergy/detail.php?id=62744.

“Kazakhstan to Begin Producing Its Own Nuclear Fuel,” Reuters, August 8, 2025, accessed November 28, 2025, https://www.reuters.com/business/energy/kazakhstan-begin-producing-its-own-nuclear-fuel-2025-08-08/.

William Blair, SWUning Over Nuclear Deregulation (Chicago: William Blair, 2025), https://www.williamblair.com/-/media/downloads/eqr/2025/williamblair-swuning-over-nuclear-deregulation.pdf.

GRS, “La Hague Reprocessing Plant: Expansion and Continued Operation Until at Least 2100,” accessed November 28, 2025, https://www.grs.de/en/news/la-hague-reprocessing-plant-expansion-and-continued-operation-until-least-2100.

Sam Meredith, “Nuclear Fuel Enrichers Race to Cut Dependence on Russia as AI Drives Power Demand,” CNBC, October 4, 2025, accessed November 28, 2025, https://www.cnbc.com/2025/10/04/urenco-centrus-orano-enriched-uranium-nuclear-russia-ai-data-center.html.

“Uranium Mining in Michigan’s Upper Peninsula,” MLive, accessed November 28, 2025, https://www.mlive.com/galleries/POB25Q6Z7BELZJ2HL7M3ZJMZTY/.

“Energy Resources Australia Pulls Plug on Ranger 3 Deeps Expansion,” ABC News (Australia), June 11, 2015, accessed November 28, 2025, https://www.abc.net.au/news/2015-06-11/energy-resources-australia-pulls-plug-on-3-deeps-expansion/6540046.

Complex hardware systems occupy a unique category of technology development. Think satellites, reactors, autonomous boats, and launch vehicles. Unlike software that ships in sprints or simpler hardware that iterates in months, these systems have multi-year development horizons.

Such extended timelines come with a fundamental problem: the world refuses to stand still while you build.

The market assumptions that justify starting a complex hardware program rarely survive to its completion. Customer needs evolve, emerging technologies present better offerings, and new entrants leverage resources that didn’t exist when you began.

Meanwhile, regulations shift, investor appetite can swing from euphoria to terror (and sometimes back), and public opinion can turn big accomplishments into political liabilities. What looked like a goood fit in year one can be eroded by year five.

The challenge extends beyond just timing.

Hardware programs must maintain alignment across multiple stakeholder groups: customers, regulators, investors, government administrations, the public, and others. This goes beyond product-market fit, to a complete program-market fit. Lose support from any critical stakeholder, particularly around funding milestones, and the program stumbles, regardless of technical merit.

While the market can shift quickly, hardware programs tend to accumulate substantial inertia. Changing course is a process. Teams lock into architectures, supply chains crystallize, and regulatory pathways narrow.

Success requires anticipating not where the market is, but where it will be throughout the course of development. While perfect foresight is impossible, the best programs come close—as the hockey adage goes, skating to where the puck will be.

Multiple Cycles Matter

Consider a six-year development program. Along the way, there might be three key financing milestones. At each, you must demonstrate market alignment and a path to profitability. At year six, the hardware is operational and it’s time to start scaling up actual profits.

Yet, the challenge is that each stakeholder group operates on its own cycle, and they all rarely synchronize.

Customers

First, is our offering relevant? In other words, do customers want to buy what we’re selling? At the start of the project, hopefully the answer to this question is yes. If you waved a magic wand and the product appeared, there would be people waiting to use it.

However, maybe you’ve peered into the future and recognized that the market today doesn’t yet know what it needs. You’re betting that customer need will intensify—or emerge entirely—by the time you deliver.

Or, you’ll forge the fit yourself. Shape market expectations during development, educate customers on problems they don’t yet see, create demand where none exists today.

In the worst case, deep into your development program the offering becomes irrelevant. For example, the best Loran receivers became markedly less useful as GPS become more available (though they could be making a comeback).

Technology Stack

Is your technology stack state-of-the-art? Again, it most likely is at the beginning of the project. But your stack could take so long to develop that newer, more advanced technologies can emerge and maybe even mature faster.

I saw this happen with metallic 3D printed parts. When we started printing, Inconel 718 was the most mature high-performance alloy used for space-grade engine parts. But by the time the engines were developed, GrCop84, a copper alloy, had matured to the state-of-the-art. Just as printable, but higher performance.

We’d optimized our architecture around yesterday’s material.

Subscribe now

Competitors

Do you have a durable advantage over your competitors? This might come in the form of team, technology, relationships, readiness or other moat.

If there are mature incumbents, you likely start with no advantage. You might have a potential edge—speed, technology, or cost structure—but that potential isn’t real yet.

In the best case, this advantage grows throughout your development program, hitting an all time high when you enter the market.

But new entrants that come behind you present a unique challenge: they’re shiny and new.

While you’re in year five—battle scarred from setbacks and delayed by regulatory hurdles—a new competitor launches with fresh funding and zero baggage. They haven’t accomplished much yet, but their unblemished story looks more attractive. Your hard won experience becomes indistinguishable from failure.

The longer you take to develop, the more vulnerable you become to someone starting the same journey with newer tools and a cleaner narrative.

Capital Markets

Is what you’re doing investable? This is a fundamentally different question than whether the technology works, or even if there’s customer demand.

At the most basic level, the potential value must justify the investment required, given total risk exposure. It also must look attractive on a relative basis. If an investor doesn’t invest in your company, but instead in another company doing something else (or similar), can they get a better return for less risk?

Entry point also matters. The price an investor buys at must be sufficiently low to achieve their target return. For early funds, this might be 25x or 50x. At the later stage, maybe 5x-10x. This creates a trap. Raise at too high a valuation early, and the company needs to grow into it before becoming investable again.

Additionally, general market cycles and sentiment matter. If public markets are at local lows when you fundraise, it’s a headwind (depending on the investor you’re targeting). If sentiment for your sector is high, you can ride the tailwind.

Regulatory Framework

Generally, regulatory frameworks are slow moving, if not fully static. This is both good and bad.

If the regulatory environment supports your business model at the onset, you’re in good shape. Unless something drastic happens, you should be okay by the time you come to market. And if something is going to change, you’ll see it coming over an extended period of time.

It’s not uncommon to seek regulatory changes during your development path, but it can add risk. For example, currently, the Nuclear Regulatory Commission takes 6-10 years to certify a new reactor. Their target for microreactors is 1.5-2 years, which is likely a key factor in many nuclear startups’ business models.

Political Will

Separate from regulators, many high profile projects benefit from favorable political will. This can change every four years with different administrations. Projects that take longer than a single presidential term can face real political risk.

The higher profile and politically polarizing an endeavor is, the more care must be taken if the project is going to cross administration cycles. Further, it’s not possible to predict the next administration four years out. At best, insights can start to be drawn one year out from a potential shift.

For example, in the current administration, electric vehicle and renewable energy investments are more challenged than just eighteen months ago.

Public Support

Big projects live in public view. Rocket launches close beaches and roads. Nuclear reactors create safety risks—part perceived and part real. Factories create jobs.

All of these nudge the public’s perception of the program in different ways.

Public support operates as a force multiplier. Strong public backing accelerates permits, attracts talent, and provides political cover. Public opposition does the opposite—every milestone becomes a battle, every setback becomes ammunition, every delay compounds.

Public opinion forms early and hardens fast. A single early incident can define perception for years. Programs must invest in public trust from day one, not as an afterthought when opposition emerges.

The Composite Challenge

All of these alignment profiles combine into a composite curve of program market fit. It’s easiest to consider this qualitatively, as exactly how they combine mathematically isn’t straightforward.

Yet by mapping these cycles together, we can start to predict what the future might look like, when to time critical milestones, and whether we’ll have momentum when our product reaches market.

Moreover, these cycles don’t operate independently. They interact and amplify.

When capital markets turn bearish, investors scrutinize technology risk more harshly. When public opinion sours, political will evaporates. When new competitors emerge with shinier stories, customer excitement wanes. A misalignment in one dimension cascades into others.

The opportunity lies in seeing this clearly. While others react to each cycle independently, the best programs anticipate the interactions, time their moves to catch multiple tailwinds, and build resilience where headwinds are inevitable. They don’t just survive the cycles. They use them.

Subscribe now

However, global supply chain for reactor fuel is both complex and thin.

Seventeen countries mine uranium, six convert it to enrichable gas, and nine possess commercial enrichment technology. These activities often take place in different countries. Complicating matters, Russia plays a dominant role throughout.

This raises two questions:

1. How exactly do we make fuel for reactors?

2. Does the U.S. have nuclear fuel security?

We’ll explore these questions in a two part essay. This first installment follows uranium’s path from mine to reactor. In doing so, it reveals the degree of U.S. foreign dependence for a variety of fuel services.

This is a big deal today. Nuclear makes up 20% of U.S. electricity generation. But it’ll be an even bigger deal tomorrow, as energy demands surge.

In the second part, we’ll examine U.S. access to nuclear fuel and the tools in our toolkit to assure it. Interestingly, not all of them involve mining more uranium.

The Basics of Reactor Fuel

Nuclear reactors generate energy through fission. A neutron strikes a heavy atom, splitting it into two and releasing enormous energy. The breakup releases additional neutrons that collide with other nuclei, creating a self-sustaining chain reaction.

Uranium is the most common nuclear fuel. In its natural form, it contains two isotopes: uranium-235 and uranium-238. The former can fission, the latter cannot.

However, the natural U-235 concentration is just 0.7%. In most reactors, this isn’t high enough to sustain a chain reaction. Instead the concentration must be enriched to 3-5% for power plants or above 90% for weapons.

Plutonium can also serve as fuel in many reactors.

Unlike uranium, plutonium doesn’t occur naturally. Instead, it’s produced as a byproduct in fission reactors. When neutrons bombard U-238, it doesn’t fission, but instead absorbs a neutron. This converts it to plutonium-239, which can fission and begins doing so immediately.

Over the course of a standard 18-month fuel cycle, plutonium levels continually increase. By the end, about 30% of the power generated is coming from plutonium that didn’t exist at the start.

In this sense, nuclear reactors achieve what no other power source can. While generating electricity, they also manufacture new fuel.

The Uranium Supply Chain

There are five major steps to creating uranium reactor fuel:

1. Mining and Milling of uranium ore into uranium oxide (U₃O₈), known as yellowcake

2. Conversion of yellowcake into gaseous uranium hexafluoride (UF₆), known as hex gas

3. Enrichment of hex gas to increase the concentration of U-235 relative to U-238

4. Deconversion of hex gas back to solid uranium dioxide (UO₂)

5. Fuel Fabrication of uranium dioxide into reactor-specific fuel assemblies

Mining, Milling and the Supply Deficit

While uranium deposits exist worldwide, only seventeen countries mine it commercially. Kazakhstan leads with 43% of the global supply, followed by Canada (15%), Namibia (11%), and Australia (9%).

This is where fuel security risk starts.

Except for Canada, every major nuclear power depends on uranium imports. In 2022, the United States required 15,500 metric tons of uranium, 99% of which was imported.

Mining produces solid uranium ore or, in some cases uranium-bearing fluids. Milling then crushes, leaches, and chemically processes this material to produce yellowcake.

There’s also a peculiar supply and demand relationship in the uranium market.

For the last thirty-five years, more uranium has been consumed than produced. In 2022, the global uranium need was 59,000 metric tons. Yet only 49,500 metric tons were produced, covering about 85% of the demand.

Secondary sources fill the production gap. This includes stockpiles from decades of overproduction, reprocessed spent fuel, and downblended weapons-grade material. The strategic implication is that nuclear fuel security doesn’t depend on mining alone.

Uranium currently trades for about $176/kg ($80/pound), which implies the 2022 global annual production of 49,500 metric tons was worth about $8.7B.

Subscribe now

Conversion’s Hidden Bottleneck

Conversion transforms yellowcake into hex gas, which the only form suitable for enrichment. Often this step is overlooked for the higher profile mining and enrichment activities that bookend it.

However, conversion is the first global bottleneck and there are only six facilities in five countries with this capability.

The Enrichment Turning Point

Once in gaseous form, uranium can be enriched. This marks the sharp transition from globally traded commodity to tightly guarded asset.

The same equipment that enriches uranium for peaceful power generation can also produce weapons-grade material. For this reason, enrichment facilities are highly restricted, even more so than reactors themselves.

Exemplifying this dynamic, there are numerous global programs that provide countries with reactor fuel in exchange for their commitment to not pursue enrichment technology themselves.

Nine signatories to the Treaty on the Non-Proliferation of Nuclear Weapons (NPT) commercially enrich uranium. Notably, Russia makes up 44% of the global capacity. Non-NPT countries such as Iran, North Korea and Pakistan also enrich, but do not do so at a globally commercial scale.

Modern enrichment mostly relies on gas centrifuges spinning in excess of 100,000 RPM. The heavier U-238 molecules migrate outward, while lighter U-235 stays near the center where it’s collected. Thousands of centrifuges are connected in series to create cascades, where each step enriches the product of the last.

About 90% of global enrichment happens this way. Other approaches include electromagnetic and thermal separation, but these are less efficient and generally more dated approaches.

Look to the future, an emerging laser enrichment process has the potential to rewrite the economics of uranium separation. SILEX (Separation of Isotopes by Laser Excitation) uses precisely tuned lasers to selectively excite U-235 hexafluoride molecules, consuming a reported 75% less energy than centrifuges.

With this efficiency gain, SILEX may be able to further enrich depleted uranium — the leftover tails from previous enrichment efforts. Currently, it is not economic to access these stockpiles via centrifuge enrichment.

SWU Economics

Enrichment is measured in Separative Work Units (SWU). This framework quantifies both the technical difficulty of enrichment and serves as the commercial unit for trading enrichment services.

SWU accounts for the complex thermodynamic work needed to separate uranium isotopes. The calculation considers three streams: the uranium feedstock, the enriched product, and the depleted uranium tails. The concentration left in tails—usually 0.2-0.3%—represents an economic optimization between the cost of natural uranium and enrichment services.

The exact global demand for enrichment is difficult to isolate because fuel can be fabricated from secondary sources. However, some estimates place at around 37.6 million SWU per year, which is likely a lower bound.

The price per SWU fluctuates significantly, epecially since Russia invaded Ukraine. The current average market rate is roughly $190 per SWU.

Deconversion and Fabrication

After enrichment, hex gas is converted back into a stable solid. The gas is chemically reduced to uranium dioxide powder, pressed into pellets, and sintered at high temperature. These pellets are loaded into zirconium alloy tubes to form fuel rods, which are welded and inspected.

The rods are then assembled into reactor-specific fuel assemblies—17×17 lattices for most PWRs, 10×10 channel boxes for BWRs, or hexagonal bundles for VVERs. Each design has its own geometry, materials, moderating strategy, and quality controls. Many newer reactors aim to use fuel in pelleted TRISO form for safety considerations.

In some cases fuel fabrication is performed by the reactor vendor and in others by independent fuel companies. Some nations import entire assemblies, while others maintain their own fabrication lines as part of their energy security strategy.

Plutonium and MOX Fuel

We discussed earlier how plutonium created in-situ can power reactors. But it can also be loaded directly into a reactor at the start of the fuel cycle.

Reactors can be loaded with Mixed Oxide Fuel (MOX), which is a blend of 3-12% plutonium oxide and depleted uranium. MOX has powered forty four reactors globally, with France operating twenty two that combined generate 10% of French electricity.

In general, fast reactors can use MOX far more efficiently. These units use unmoderated fast neutrons that convert U-238 into plutonium while burning existing plutonium. Some designs achieve breeding ratios above 1.0, producing more fissile material than they consume.

Generating Power from Weapons

An alternate source of reactor fuel comes in the form of former weapons.

Weapons-grade uranium is highly enriched to 90%+. It can be combined with depleted uranium (<0.3%) to create reactor fuel at 3%-5%.

The Megatons to Megawatts Program demonstrated this at scale. From 1993 to 2013, Russia downblended 500 metric tons of weapons-grade uranium into reactor fuel—roughly equivalent to 20,000 Soviet warheads. This was sold to the U.S. and generated approximately 10% of U.S. electricity over the period.

In another example, known a Project Sapphire, the U.S. airlifted 600 kilograms of highly enriched uranium from Kazakhstan in 1994. The weapons grade material has been found barely guarded in a former Soviet facility. The U.S. procured and secured the material, transporting it to Tennessee to become reactor fuel.

The U.S. has also downblended part of its own nuclear arsenal. Beginning in the late 1990s, the Department of Energy converted roughly 160 metric tons of highly enriched uranium—drawn from retired warheads, naval fuel, and other military stocks—into low-enriched uranium suitable for civilian reactors.

Plutonium from dismantled nuclear weapons can also be turned into reactor fuel. Weapons often contain metallic plutonium, which is chemically converted into plutonium oxide and blended with depleted uranium to form MOX.

Russian Involvement

Russia holds a large position in the global reactor fuel market. The country has broad capabilities, and the world became accustomed to procuring fuel from Russia through Megawatts to Megatons and other programs.

As a point of reference, Russia historically accounted for 25% - 30% of uranium imports into the U.S.

The invasion of Ukraine in 2022 complicated this. Yet, while the world wanted to decouple from Russian uranium, it was easier said than done.

The U.S. quickly issued broad sanctions against Russia, but uranium was exempt from them. Later in August 2024, new legislation prohibited uranium imports from Russia, but allowed for waivers through 2028. Russia retaliated with a ban on uranium exports to the U.S.

Towards the end of 2024, Russia-U.S. uranium trade had nearly halted. However, it resumed on a case-by-case basis in early 2025. In February, the vessel Atlantic Navigator II delivered 100 tons of enriched uranium to the port of Baltimore.

Given bilateral restrictions, future imports are uncertain. However, one U.S. company, Centrus, has announced they’ve secured waivers for Russian imports through 2027.

Where does this leave us?

Annually, the U.S. requires approximately:

• 15,500 metric tons of uranium of which 99% is imported.

• 12 million SWU of enrichment of which 60% is imported.

In other words, we have a high dependence on foreign sources for commercial nuclear fuel.

As the U.S. refocuses on nuclear reactor development the needs will only increase. As a result, there is heightened focus on assuring access to nuclear fuel for the U.S.

In May 2025, the DOE moved to reshape its surplus plutonium strategy. An executive order directed the DOE to stop its longstanding disposal program and instead establish pathways to make surplus plutonium available as advanced reactor fuel.

Under the resulting initiative, about 19 metric tons of plutonium are being offered to U.S. reactor developers, particularly those working on fast reactors. The effort is significant, but remains at the pilot stage rather than full commercial scale.

This brings us to the core question: Does America have nuclear fuel security? And if not, what tools exist to achieve it?

Part 2 examines seven such tools. These range from mining more uranium domestically and bolstering enrichment capacity to re-enriching depleted uranium, downblending weapons and activating advanced reactors.

Pathways exist, and we’ll consider the political will and capital backing required to pursue them.

General References

OECD Nuclear Energy Agency. Uranium 2024: Resources, Production and Demand. Paris: OECD Publishing, April 2025. https://www.oecd-nea.org/upload/docs/application/pdf/2025-04/7683_uranium_2024_-_resources_production_and_demand_2025-04-22_14-29-2_928.pdf.

Bellona Foundation. “How Will the US Ban Russian Enriched Uranium Impact Both Countries?” Bellona.org, May 2024. https://bellona.org/news/nuclear-issues/2024-05-how-will-the-us-ban-russian-enriched-uranium-impact-both-countries.

Financial Times Markets Data. “Company Announcement.” PR Newswire, November 5, 2025. https://markets.ft.com/data/announce/detail?dockey=600-202511051704PR_NEWS_USPRX____PH16561-1.

International Panel on Fissile Materials. “United States.” Fissile Materials. Accessed November 12, 2025. https://fissilematerials.org/countries/united_states.html.

OECD Nuclear Energy Agency. “Uranium: Resources, Production and Demand (Red Book).” OECD-NEA. Accessed November 12, 2025. https://www.oecd-nea.org/jcms/pl_28569/uranium-resources-production-and-demand-red-book.

World Nuclear Association. “Conversion and Deconversion.” Information Library: Nuclear Fuel Cycle. Accessed November 12, 2025. https://world-nuclear.org/information-library/nuclear-fuel-cycle/conversion-enrichment-and-fabrication/conversion-and-deconversion.

William Blair & Company. “SWUning Over Nuclear Deregulation and Expanding Enrichment Capacity.” William Blair Insights. Accessed November 12, 2025. https://www.williamblair.com/Insights/SWUning-Over-Nuclear-Deregulation-and-Expanding-Enrichment-Capacity.

International Atomic Energy Agency. Country Nuclear Fuel Cycle Profiles. Technical Reports Series No. 425. Vienna: IAEA, 2005. https://www-pub.iaea.org/MTCD/Publications/PDF/TRS425_web.pdf.

I always started with a snowfort. I’d spend hours building a perfect one. I’d turn the snowpile at the end of the driveway into a fortress equipped with tunnels, walls, and hidden snowball stockpiles.

But by the time I was fully ready to fight, my dad had long gone back inside. 

The risks of being unprepared are easy to understand. We all know what that failure mode looks like.

But being overprepared—as I learned every snowfall—can be just as damaging. The risks are just better hidden. 

The trap is that overpreparing feels like making progress. You’re doing lots of work (building the fort), but you’re not advancing toward your goal (having the fight). You can get so caught up in crafting the perfect plan that you miss your chance to act. Worse, you can get stuck building the wrong thing.

Maybe you need a shield, not a fort. But you’ll only find out by getting into the fight

Running a business or hardware program demands momentum. You have to stay in motion, make decisions, and move while the window’s open. 

Early in the company, we began using “snowforting” as a codeword for when we were too deep in planning mode — a quick, shared cue to shift towards execution.

Of course, certain things demand extensive preparation. Its all a balance and sometimes you do need a great fort. 

But frequently, you don’t lose because your fort wasn’t perfect. It’s because you never made it into the fight.

The U.S. is reindustrializing. This brings visions of factories and steel mills, hot sparks flying, and durable goods rolling off production lines. But manufacturing is just one — albeit very important — dimension of industrial capability. Instead, industrialization requires three fundamental capabilities: the ability to design complex systems, manufacture them at scale, and operate them effectively.

Each of these three categories has many tiers. In manufacturing, we naturally think of the Original Equipment Manufacturer (OEM) and their Tier 1-4 suppliers. Design has the same depth. A company might perform overall system design while outsourcing subsystem and component work to specialist firms.

Operations spawns its own ecosystem. The final user might be the manufacturer itself, or an entirely separate company using the hardware to deliver services. Additionally, someone must maintain these systems, whether it be the operator, manufacturer, or third-party specialists.

The pattern of who owns what across these three phases—what we’ll call the “shape” of industrialization—determines everything from development speed to capital requirements to competitive resilience. Mapping the shapes can illuminate where a company fits, what role it can play, and what challenges it will face as it scales.

Making a Map

We can arrange the three phases — design, manufacturing, and operations — on a map, along with their respective subgroups. In doing so, we start to see how the industrialized ecosystem works, noting that the mission activity under operations (top right) is ultimately where end-user value is delivered.

Besides these three phases, testing and logistics also play critical supporting roles. Every part that’s built must be validated through via test, which can have significant infrastructure requirements and other costs. Likewise, transporting and warehousing the hardware can drive substantial operations. While these activities are critical, for this discussion, we’ll consider them as elements within the three main phases above.

The Shapes of Industry

This map becomes powerful when we use it to visualize real organizations. The cells each company occupies — and avoids — tell a story about strategy, capability, and competitive positioning. We’ll consider five examples to illustrate how different programs position themselves through different shapes.

Boeing 737 Aircraft

Consider the Boeing 737. Its mission is to deliver transportation services to traveling customers.

Boeing designs the top-level aircraft. They do some subsystem design (structures) but outsource much of it. For example, the engines come from CFM International, which is a joint venture between GE and Safran. At the component level, most design is outsourced to vendors that make commercial-off-the-shelf (COTS) parts, such as pressure sensors or connectors.

When it comes time to manufacture the plane, Boeing does the final aircraft integration as the OEM. Tier 1 systems are mostly outsourced. Engines come from CFM. Spirit AeroSystems makes the fuselage. Down at Tier 2, Collins Aerospace provides components and mechanisms. In total, the 737 reportedly has 600 suppliers across the Tier 1 & 2 levels, and even more are expected at the lower levels. Tier 4 materials are mostly outsourced, with companies such as Hexcel and Toray Industries supplying composites and other materials for the structural assemblies.

This arrangement positions Boeing on top of the design and manufacturing phases, performing the most integrated work, while outsourcing much of the rest.

The operations phase reveals a key distinction. Boeing is no longer at the top. Instead, airlines such as Southwest and Alaska Airlines provide mission services by flying customers around the world.

Some support activities are outsourced. Refueling services are procured from airports. Food and beverages restocking often comes from vendors. Pilot training is a mix. Boeing runs a training division that some airlines rely on. Conversely, Southwest operates its own flight training center.

When it comes to maintenance, refurbishment, and overhaul (MRO), the industry also diverges. Southwest operates its own MRO facilities and performs heavy maintenance on its 737 fleet, which is one of the largest in the world. At the other end of the spectrum, Alaska Airlines performs light repairs but outsources heavy MRO activities to companies such as AAR Corporation.

SpaceX Falcon 9 Launch Vehicle

The shape of SpaceX’s Falcon 9 program is dramatically different. SpaceX is both highly verticalized from a design and manufacturing standpoint, and they are the end service provider. In Boeing’s case, it builds the 737 and sells it to airlines to operate. In SpaceX’s case, they build Falcon 9 and operate it themselves, transporting customer satellites to space.

In the design phase, SpaceX performs most engineering work in-house. This starts with the vehicle-level design for Falcon 9. It extends to the subsystem design of the Merlin engines, the avionics suite, and the primary structure (equivalent to an airplane’s fuselage). At the component level, SpaceX also performs significant design of mechanisms, valves, and sensors, though some of these are procured as COTS.

The manufacturing phase mimics the verticalization of design. For most hardware, SpaceX serves as the OEM and the Tier 1 and Tier 2 suppliers. At Tier 3, it’s a mix. Machined parts are split between in-house and outsourced. Some COTS items are procured. At Tier 4, SpaceX generally procures raw materials but has also done some in-house metallurgical work.

In the operations phase, SpaceX again is verticalized. They launch the rocket, and for several years, I was fortunate to be an operator in mission control for Falcon 9 launches. They also manage their own launch sites and handle refueling and other support activities. Finally, SpaceX performs all refurbishment for boosters that have landed before they’re relaunched.

United Launch Alliance Vulcan Launch Vehicle

United Launch Alliance (ULA) provides a good contrast to Falcon 9. ULA operates the Vulcan launch vehicle and performs the top-level design, build, and launch operations.

However, they tend to procure more Tier 1-3 components from vendors. In particular, ULA buys engines from Blue Origin to power Vulcan. As a result, their shape takes on a different form in the mid tiers of design and manufacturing.

U.S. Navy Virginia-class Submarine

The Virginia-class submarine program creates yet another distinct shape: distributed industrial capability integrated by government oversight, with exclusive military operations.

In the design phase, the submarine program splits responsibilities. General Dynamics Electric Boat and Huntington Ingalls Industries Newport News Shipbuilding share the system-level design under Navy oversight. Major subsystems are designed by specialists — nuclear reactors by BWX Technologies, combat systems by Lockheed Martin, and sonar by Northrop Grumman.

Ultimately, the Navy maintains final design authority, even if through contractors. For this reason, here we’ll categorize them as the system design owner.

Manufacturing follows a similar distributed pattern. Electric Boat builds the bow, engine rooms, and command and control systems, while Newport News builds the stern, habitability, and machinery spaces. The yards alternate responsibility for the final assembly of each submarine.

Tier 1 suppliers span many of the traditional defense contractors, such as Lockheed Martin, Raytheon, Northrop Grumman, as well as specialists like BWX Technologies for the reactor components and Kollmorgen for the mast. This industrial base represents thousands of suppliers across all tiers, carefully managed to maintain the strategic capability.

The operations phase is where the shape becomes unique. Unlike commercial models, the Navy operates these submarines exclusively. The Navy trains all crews and conducts all missions. Contractors support maintenance and modernization, but under Navy supervision. This shape exists to ensure production of a key national security system, that has little commercial overlap.

Omega Aerial Refueling Services

Omega Aerial Refueling Services is a unique, lesser known case study. During combat operations, the military provides aerial refueling. However, Omega has carved out a niche in providing aerial refueling for training, exercises, flight testing, and long cross-country or overwater movements.

The firm operates a fleet of six KC-707 and KDC-10 aircraft that can provide contract aerial refueling to military aviation. When we look at Omega’s industrialization map, they occupy just the mission support cell under the operations phase.

Note that shapes shift with perspective. We could map this differently, showing the Navy operating fighters that Lockheed Martin designs and builds. But viewing Omega alone best illustrates this specialized model.

Likewise, we could even draw an organization’s domain-specific shape — its map for propulsion or avionics. Often these shapes diverge for the same company.

How Shape Determines Performance

Why do these shapes matter? The form they take influences how the organization operates across a variety of dimensions. Six of the most important are considered below.

Control, Responsiveness, and Insulation

A bigger shape means greater control. For example, a team can design the exact system they need without compromising for vendor availability.

It also enhances responsiveness. During development, there will be learnings. A part might break or not fit as expected. Verticalized groups can respond rapidly. Designs can be quickly modified, and new parts machined.

During heavy engine development at ABL, this was a big enabler for the program. We could run an engine test, identify the need for a new turbopump impeller, design and machine that new impeller, and get the engine back on the stand all within two days. If this were outsourced, it would take weeks, if not months.

Verticalized businesses also often have a degree of insulation that protects them from things like short-term shocks. If there aren’t many vendors in the supply chain, then global disruptions won’t have a big impact.

Speed

Many often cite verticalization as a speed enabler. While true in some scenarios, it can be detrimental in others.

As illustrated above, vertical integration can mean more control and responsiveness, accelerating the program. The designers of the Sten gun carefully prescribed every feature of every component in the weapon. This extreme degree of design for manufacturing led to legendary build times.

However, verticalization also means that there’s more that must be done. If we want to machine parts in-house, we need to build a machine shop, hire a team, and get it running smoothly, all before we can machine your first part. If we outsource, parts arrive in weeks. Both paths have their benefits, and speed must be considered case by case.

Intellectual Property and Competitive Moats

Vertical integration can create stronger intellectual property protection and deeper competitive moats. When organizations build in-house, innovations remain proprietary, and competitors cannot access the same capabilities simply by working with shared suppliers.

This dynamic explains why traditional aerospace companies often compete on system integration rather than component innovation.

Capital Requirements

Capital requirements generally increase with the size of the shape. Every activity performed requires facilities and infrastructure, so highly verticalized businesses face large upfront capital requirements. They also maintain larger teams, driving operational expenses.

For example, cutting new turbopump impellers meant we had to build and staff an entire machine shop, operating nearly around the clock.

However, in the longer term, vertically integrated entities should achieve better margins, resulting in lower long-term capital needs.

Financial Opportunity

Larger shapes typically create higher financial opportunity. If this weren’t true, expansion wouldn’t make sense.

Increased financial opportunities come in several forms. Performing more activities enables the company to capture what would otherwise be its vendors’ profit margins. Additionally, operational efficiencies often reduce costs across the operations. Finally, a larger shape may create new revenue streams. Some verticalized capabilities can themselves become business units.

Risk

Industrial efforts face a variety of different types of risks, and a company’s shape determines its risk exposure. There’s engineering risk: does the design work? Build risk: can we make parts to the required quality? Team risk: can we hire and develop the right people? Capital risk: can we finance it all? The more a program does in-house, the more risks fall into its portfolio.

When we outsource, these risks don’t disappear—they transfer to our vendors. These risks consolidate to become supply chain risk from our perspective. Vendor delivery becomes the critical uncertainty.

In many cases, a smaller shape reduces risk exposure. Airlines carry fuel cost risk. When aviation fuel prices spike, their margins suffer. Boeing avoids this immediate exposure, though sustained high fuel costs eventually impact aircraft sales.

What Shape is Right?

There’s no universal answer. The optimal shape depends entirely on an organization’s specific context and constraints. If there’s time and resources to run an in-house spiral development program, that could make sense. If not, it might be best to start by buying COTS hardware.

Most founders coming out of SpaceX highly verticalize their startups. This has clear and proven advantages, but strictly pattern-matching can be a fatal flaw.

SpaceX’s shape works for SpaceX because of Elon, access to capital, first mover momentum, and many other SpaceX-specific factors. Trying to emulate SpaceX’s exact shape without the same foundation doesn’t work, as many have discovered.

Instead, it’s helpful to consider the current market dynamics, the resources the company can bring to bear, and the program’s end goal — to maximize market share, achieve profitability quickly, limit technical risk, or something else entirely.

Many shapes can succeed. Ultimately, what these case studies show is that intentionality matters the most.

Subscribe now

It reveals distinct defensive challenges. For drones, sheer volume dominates. For ballistic missiles, interceptor capability, battery coverage, and interceptor availability all drive outcomes.

Drones

First, we’ll take a look at how drone launches have scaled and their corresponding intercept rate.

• Attack volume has increased nearly 12x since the beginning of 2024. In Q1 2024, an average of 452 drones were launched each month. In Q3 2025, this reached 5,360 per month.

• Most drones are intercepted. Monthly intercept rates have varied between 80% and 97%, with an average of 89%. Defensive systems have scaled proportionally to attacks.

• But many still break through. While the intercept rates remain high, the rising launch rate has led to a higher hit rate. In Q1 2024, just 71 drones per month hit their targets. In Q3 2025, 725 per month did so. That’s more drones breaking through than were even launched in early 2024.

• Three (potentially diplomatic) events interrupted the ramping trend. Attack volume dropped notably in three months: April 2025 (-42% versus the prior month, coinciding with ceasefire discussions), August 2025 (-34%, aligned with the Trump-Putin Alaska summit), and December 2024 (-23%, cause unclear). While we can’t establish a direct link between these diplomatic events and the reduction in attacks, it seems Russia may have reduced attack volumes around these engagements.

• Defenses must keep up. Excluding the three diplomatic outliers, attack volume is accelerating. Defenses have largely matched the scaling. However, if attack scaling outpaces defensive measures by just a few percentage points, it translates to hundreds more successful strikes monthly.

Ballistic Missiles

Next, we’ll contrast the drone dynamics with those of ballistic missiles. Markedly different trends emerge.

• Lower volume, higher stakes. Since January 2024, Russia has launched between 17 and 97 ballistic missiles per month. The average attack volume over the period is 45 per month, but with significant monthly variability.

• Most are not intercepted. Only 8% of ballistic missiles were intercepted in Q1 2024. This improved to 36% between June and August 2025. However, the intercept rate plummeted to 6% in September 2025. The drop is partly attributed to enhanced terminal maneuvering by the Russian munitions.

• More than one ballistic missile hits Ukraine every day. In Q3 2025, an average of 42 ballistic missiles evaded defenses per month. This equates to an average of just under 1.5 missiles per day.

• Maneuvering defeats current interceptors. Russia is primarily using short-range ballistic missiles: the Iskander and Kinzhal systems. These weapons recently implemented enhanced terminal maneuvering specifically designed to evade Patriot interceptors. The warheads steer just before impact, causing interceptors to miss.

• Limited coverage is a fundamental factor. Ukraine operates seven Patriot batteries. With each battery defending approximately 100-200 square kilometers, total coverage represents less than 1% of Ukraine’s territory. Practically, seven batteries mean seven defended locations, such as major cities or critical infrastructure sites. Ballistic missiles launched into uncovered areas cannot be intercepted.

• Changing the calculus requires more batteries and interceptors, with enhanced capabilities. Additional batteries expand defended areas. Improved interceptors’ counter maneuvering systems. More of them enable Ukraine to keep up with a potential ramp of ballistic missile attacks.

Implications

This dataset illustrates divergent defensive challenges. Ukrainian data scientist Petro Ivaniuk maintains a more detailed dataset breaking down strikes by platform. It’s updated nearly daily and provides the foundation for more detailed analyses.

One key data gap is the number of intercepts attempted. This would reveal how many defensive munitions are expended per successful intercept and to what degree the low ballistic missile interception rates stem from interceptor shortages or failures. In other words, how many missiles go unengaged simply because no interceptors are available?

The contrast is stark between the situation in Ukraine and the performance of Israel’s missile defense. During conflicts with Iran and Hamas, Israeli defenses routinely intercepted 75%+ of cruise and ballistic missiles. This is accomplished using multilayered defense systems with numerous interceptor platforms and complete territorial coverage.

For defense programs under development, these patterns define focus areas. Drone defense demands high-volume, cost-effective interceptors that can scale elastically. Ballistic missile defense requires both expanded coverage (more batteries), enhanced capability against maneuvering threats (better interceptors), and increased magazine depth (more interceptors).

Subscribe now

Note: My company, Long Wall, develops missile defense technology, but it is not related to the systems or statistics presented here.

Then we’ll shift gears. We’ll explore how Russia’s nuclear exports create diplomatic leverage, how passive safety systems underpin new fission plants, how one metallurgist was behind nuclear proliferation to Pakistan, Libya, Iran, and North Korea, and why the 1970s nuclear boom collapsed and how today’s renaissance might be different.

But for now, how nuclear reactors work!

Subscribe now

Nuclear reactors come in many shapes and sizes. Of the 416 operating worldwide today, some are cooled with water, others with gas or liquid metal. Some need enriched uranium, others don’t. All harness fission, but their engineering diverges.

The stories behind these differences reveal as much as the physics:

• The United States pioneered pressurized water reactors (PWRs) for submarines, then scaled that military technology for civilian power.

• France initially built gas cooled reactors to sidestep uranium enrichment, but pivoted to PWRs as operational maturity accumulated.

• Canada also avoided enrichment by developing heavy water reactors that run on natural uranium, allowing the country to capitalize on its extensive uranium resources.

• Russia operates the world’s only commercial fast reactors that can breed fuel while generating power. But there are only two in operation.

These aren’t just engineering variations. They’re national strategies.

Each architecture reflects its host nation’s technology readiness, supply chains, resource access, and economic reality. What makes sense for one country might be nonsensical for another. These choices are sticky. Many were made decades ago, but still influence which designs dominate, who builds reactors, and who controls the fuel.

Understanding the dynamics behind them will position us to predict where the industry will head next, as many new programs come online.

The Core Elements

Every nuclear reactor follows the same basic process: sustain a fission chain reaction, generate heat, and convert it to electricity. They’re simple goals that require complex execution.

Five subsystems make it possible:

• Fuel: Undergoes fission, releases neutrons, and generates heat

• Moderator: Slows neutrons to speeds where fission becomes probable

• Coolant: Carries heat from the core to the turbines

• Power conversion: Transforms thermal energy into electricity

• Control: Increases or decreases the rate of fission

These five components can combine to make many design permutations. For example, water can serve as both a moderator and a coolant simultaneously. Conversely, we can use gas as a coolant and graphite as a moderator. Similarly, fuel enrichment levels can vary dramatically.

Surveying the Landscape

We’ll consider twelve different nuclear reactor architectures, and we’ll start by segmenting the reactor landscape by neutron speed into two categories:

• Thermal reactors slow neutrons down using moderators like water or graphite, through many collisions between the neutrons and the fluid’s atoms.

• Fast reactors operate without moderators, keeping neutrons at the high speeds they’re born with through fission.

Thermal Reactors

About 99.5% of nuclear electricity worldwide comes from thermal reactors, making this category the dominant approach. By slowing the neutrons, thermal reactors enable high uranium-235 fission cross sections, which make chain reactions sustainable with fuel enriched to just 3-5% uranium-235. However, thermal reactors can only burn fuel “once through,” which means that any fuel recycling must happen at a different facility.

Within thermal reactors, we can subdivide by moderator and coolant combination. Some architectures use the same substance for both functions, while others separate them.

• Light Water Reactors use ordinary light water (H₂O) as both moderator and coolant. Pressurized Water Reactors (PWR) and Boiling Water Reactors (BWR) are the two standard types of light water reactors. Combined, they represent 90% of all nuclear power generated globally on a per Watt basis.

• Light Water, Graphite Moderated Reactors use light water as a coolant and graphite as a moderator. Famously, this is the architecture of the Russian RBMK reactor, which was used at Chernobyl. Certain nuances of this design helped contribute to the catastrophic accident.

• Heavy Water Reactors use deuterium oxide (D₂O) as both moderator and coolant. Deuterium is a hydrogen isotope with one proton and one neutron, making it a heavier variant of standard hydrogen. This design enables the use of natural uranium fuel without the need for enrichment.

• Gas Cooled Reactors use either carbon dioxide or helium as a coolant with solid graphite blocks for moderation. These reactors can achieve higher operating temperatures than water cooled designs.

• Molten Salt Reactors use molten fluoride salts as both a fuel carrier and coolant. The fuel is dissolved directly in the liquid salt rather than formed into solid fuel rods.

Fast Reactors

Fast reactors do not use a moderator, allowing neutrons to remain at high energy. This unlocks two key capabilities. First, they can breed fuel. Fast neutrons convert fertile uranium-238 into fissile plutonium-239. Second, they can consume nuclear waste. These reactors can fission actinides that would otherwise remain radioactive for thousands of years.

The tradeoff comes in fuel requirements. Since fission is less probable with fast neutrons, we need a higher density of potential fission targets. Fast reactors need either highly enriched uranium (15-20% or higher) or plutonium to achieve criticality. Once operating, they can breed enough plutonium to refuel themselves and even fuel other reactors.

There are two primary types of fast reactors, and both are liquid cooled:

• Sodium Fast Reactors use liquid sodium as a coolant, which has high heat transfer capability and minimal neutron moderating properties. However, liquid sodium is highly reactive with air and water, requiring significant safety systems.

• Lead Fast Reactors use liquid lead/lead-bismuth as coolants. These also have high heat transfer and low neutron moderation. Lead-based coolants are mostly inert, making them safer options for plants. However, the substance is highly corrosive, requiring special materials.

There are a variety of gas cooled fast reactors in research and development. They operate similarly to gas cooled thermal reactors, but without the use of moderators. These designs can theoretically achieve very high temperatures with improved efficiency, but face challenges in maintaining adequate heat removal with low density gas. Since none are commercially deployed, they are not considered in detail here.

Building a Nuclear Reactor

Now that we’ve characterized the landscape, let’s dive into the details of the various subsystems.

Of all architectures, PWRs are the most prevalent. They make up 74% of the global reactor fleet and generate 78% of worldwide nuclear power on a per Watt basis.

Their widespread deployment makes PWRs the logical example for this case study. PWR designs vary plant by plant, and here we’ll consider a generalized reactor for illustrative purposes.

Fuel

At the center of a nuclear reactor is the fuel assembly. A typical PWR core consists of approximately 200 assemblies arranged in a precise geometric pattern. Each fuel assembly is a 17 x 17 square array of long, narrow cylindrical rods. In this 289-position matrix, 264 positions contain fuel rods, 24 house guide tubes for control rods, and one is used for instrumentation.

Fuel rods are hollow zirconium alloy tubes about half an inch in diameter and roughly 12 feet long. Inside each rod, about 400 uranium dioxide pellets are stacked to form the fissile material. The uranium dioxide is enriched to 3-5% uranium-235, with the remainder being uranium-238.

Zirconium has an extremely low neutron absorption cross-section, so it steals very few neutrons from the fission process. The fuel rods are densely packed with roughly an eighth of an inch between them, creating channels for water to flow.

Moderator

PWRs use water as both a moderator and a coolant. It is pressurized to about 2,200 psi to prevent boiling. As the water flows around the fuel rod assemblies, it extracts heat, reaching temperatures of about 650°F. At the same time, its hydrogen atoms slow neutrons through repeated collisions.

Temperature creates a negative feedback loop in PWRs, providing inherent safety. As water heats up, its density decreases. This creates more space between hydrogen atoms, making neutron impacts less likely to occur. Fewer neutrons are slowed, and as a result, the fission rate decreases.

This decrease in power due to increase in temperature is a built-in safety feature that has helped PWRs gain regulatory approval and public acceptance worldwide.

Coolant

PWRs use a two-loop coolant design. The water that cools the fuel rod assemblies and doubles as a moderator is known as the primary loop. It becomes mildly radioactive as neutrons interact with the water molecules.

The primary loop carries heat from the fuel rod assemblies to a series of heat exchangers, known as steam generators. Here, the primary loop transfers heat to the secondary loop, which also uses water coolant. The secondary loop operates at lower pressure, around 1,000 psia, allowing the water to boil and create steam for turbine generators.

The two-loop design isolates radioactive materials from turbine systems and provides multiple barriers against the release of radioactive materials.

Power Generation

The secondary loop carries steam to conventional Rankine steam cycle generators. The hot steam flows over turbine blades, causing them to spin a generator and create electricity. After this, the steam passes through a condenser where it’s cooled back to liquid form before returning to the steam generator.

A third coolant loop provides condenser cooling, interfacing with the environment through seawater, cooling ponds, or cooling towers. This condenser loop is twice removed from the reactor core, ensuring no radioactive materials escape. The iconic nuclear cooling towers are part of this final heat rejection system.

Control

Lastly, reactors need to be controlled. We need ways to ramp power up, ramp it down, and shut it off in an emergency. We achieve this by controlling the number of neutrons that move through the core. More neutrons mean more fission and more power. Fewer neutrons mean the opposite. To adjust the neutron levels, we use neutron absorbers, which are materials that capture neutrons and prevent further fission.

PWRs use control rods that slide in and out of the 25 guide tubes in each fuel assembly. The control rods are hollow stainless steel tubes filled with a neutron absorbing material. Often, an 80% silver, 15% indium, and 5% cadmium alloy is used. This mixture is tuned to have a high absorption cross-section across a wide range of neutron energies.

Control rods are inserted from above by a drive mechanism that moves them vertically. The deeper the insertion, the more neutrons absorbed and the lower the reactor power.

The control rods are connected to the drive assembly via an electromagnetic coupling. It requires an electrical current to remain latched. If power is lost or operators perform an emergency shutdown, the coupling is released and the rods drop into the core within seconds. This gravity-assisted, failsafe shutdown is called a SCRAM, and is a cornerstone of nuclear safety.

PWRs can also dissolve boric acid directly in the coolant for fine reactivity control. Operators adjust boron concentration throughout the fuel cycle. It’s high at the beginning when fuel is fresh, and gradually reduced as the fuel depletes. This allows precise power adjustments without moving control rods.

Examining the Other Architectures

Building off the PWR example, we’ll examine other architectures to understand their differences and tradeoffs, starting with the other thermal reactors.

Light Water Reactors

As noted above, there are three types of light water reactors. PWRs and boiling water reactors (BWRs) use ordinary light water for both neutron moderation and heat removal. The Russian RBMK reactor uses light water as a coolant and graphite as a moderator. We’ve already covered PWRs extensively, so we’ll dive right into BWRs.

Boiling Water Reactors

BWRs embrace what PWRs avoid. They allow water to boil in the core. This fundamental choice cascades through every aspect of the design. BWRs operate at 1,000 psi, about half the pressure of PWRs. At this pressure, water boils at approximately 545°F, converting to steam right in the reactor core.

As a result, BWRs don’t have the massive heat exchangers that define PWRs. The simplification eliminates an entire cooling loop worth of pumps, piping, and potential failure points, reducing costs and improving efficiency.

Yet this elegance brings unique challenges:

• Turbine hall radiation: The steam carries radioactive fluids directly to the turbine, making it a radiation area. This complicates maintenance and requires extensive shielding.

• Chemistry challenges: Without a secondary loop barrier, corrosion products from turbines flow directly through the core, depositing on fuel rods. BWRs also cannot use dissolved boron for reactivity control since it would corrode downstream equipment.

• Control complexity: Boiling creates power instabilities as steam voids form and collapse unpredictably. This requires sophisticated monitoring to prevent dangerous oscillations.

• Bottom mounted control rods: Control rods insert from below using hydraulics rather than gravity. If hydraulic systems fail, the rods cannot drop into shut down the reactor.

• Larger pressure vessel: The reactor vessel must house not just the core but steam separators and dryers, making it significantly larger and more complex than PWR vessels.

Despite these challenges, forty three BWRs are operating worldwide, concentrated in Japan, the United States, and Sweden. Modern BWR designs address these vulnerabilities with passive safety systems. Natural circulation designs eliminate pumps entirely. Isolation condensers provide cooling without power. But the fundamental trade-off remains: mechanical simplicity in exchange for operational complexity.

RBMK Reactors

RBMKs represent a uniquely Soviet approach to reactor design, combining light water cooling with graphite moderation. RBMK stands for Reaktor Bolshoy Moshchnosti Kanalnyy, which translates to High Powered Channel Reactor.

Unlike PWRs and BWRs that contain fuel in a single pressure vessel, RBMKs use thousands of individual pressure tubes running vertically through a massive graphite block. Each tube contains water cooled fuel assemblies, while the surrounding graphite moderates neutrons.

This architecture enables two strategic advantages. First, operators can refuel individual channels without shutting down the reactor, maintaining high availability and continuous plutonium production. Second, the design uses relatively low enriched uranium, with graphite’s low neutron absorption enabling criticality.

However, this design harbors a fatal flaw: a positive void coefficient. In most reactors, steam voids slow the reaction because water moderates neutrons. But RBMKs use graphite for moderation, while water mainly absorbs neutrons. When water turns to steam, fewer neutrons are absorbed, but moderation continues unchanged. This accelerates the reaction and can trigger runaway power surges.

This characteristic, combined with control rod design flaws and operator errors, led to the Chernobyl disaster. During a 1986 safety test, operators withdrew too many control rods to overcome xenon poisoning from low power operation. When they attempted an emergency shutdown, a design flaw in the control rods briefly increased reactivity. The positive void coefficient amplified this spike into a steam explosion that destroyed the reactor.

Eight RBMK reactors still operate, all in Russia, with significant safety modifications including modified control rods and operating restrictions to prevent the conditions that enabled Chernobyl. No new RBMKs are under construction.

Heavy Water Reactors

Pressurized Heavy Water Reactors (PHWR) operate in the thermal neutron spectrum, but approach neutron management differently. They use heavy water as a moderator and coolant. Heavy water is deuterium oxide (D₂O), which is water made with deuterium, a heavy isotope of hydrogen containing both a proton and a neutron in its nucleus.

The key to PHWR design lies in deuterium’s superior neutron economy. While ordinary hydrogen is an excellent neutron moderator due to its low mass, it also absorbs neutrons, removing them from the fission process. Deuterium still slows neutrons effectively but absorbs far fewer. It is about 600 times less likely to capture a neutron than regular hydrogen. This efficiency enables PHWRs to achieve criticality with natural uranium (0.7% uranium-235), eliminating the need for enrichment.

The trade-off is heavy water itself. Deuterium is naturally occurring, but it makes up just 0.016% of seawater. The process of concentrating it is both expensive and time consuming. A typical 700 MW PHWR requires about 500 tons of heavy water, which represents about 10%-20% of the initial plant capital cost. However, if managed properly, the heavy water inventory lasts for decades.

CANDU Reactors

The most widely deployed PHWR design is the Canada Deuterium Uranium (CANDU) reactor. At its heart sits the calandria, which is a large horizontal cylindrical tank filled with low-pressure heavy water to serve as a moderator.

Between 380 and 480 horizontal channels run through the calandria, each consisting of two concentric tubes. The inner pressure tube, often made of zirconium-niobium alloy, contains the fuel bundles and high-pressure coolant at 600°F. The outer calandria tube isolates the hot pressure tube from the cool moderator. A carbon dioxide-filled gap between the tubes provides thermal insulation.

This unique architecture enables online refueling, which is a defining CANDU advantage. While the reactor operates at full power, a refueling machine latches onto individual pressure tubes, removes spent fuel bundles from one end, and inserts fresh ones from the other.

Forty six PHWR reactors currently operate worldwide. Canada initially developed the technology to leverage its uranium resources without enrichment capability. Later, India adopted it after being excluded from enrichment technology following its 1974 nuclear test. Both nations have since become major CANDU operators.

Gas Cooled Reactors

Moving beyond water based systems entirely, gas cooled reactors enable higher operating temperatures and improved thermodynamic efficiency. Carbon dioxide and helium are common coolant candidates. Neither provides neutron moderation. Instead, that is accomplished via solid graphite blocks in the reactor core.

Magnox and Advanced Gas Cooled Reactors The United Kingdom pioneered commercial gas cooled reactors in the 1950s with Magnox reactors, which were named for their magnesium-aluminum alloy fuel cladding. These reactors arranged natural uranium fuel rods within vertical channels drilled through a massive graphite core. Carbon dioxide coolant, pressurized to 100-300 psi, flowed through these channels, extracting heat while the surrounding graphite moderated neutrons. The low neutron absorption of both carbon dioxide and graphite enabled criticality with natural uranium.

Twenty-six Magnox reactors operated successfully for decades, but their low operating temperature limited efficiency. The UK addressed this with Advanced Gas Cooled Reactors (AGRs) in the 1960s. AGRs retained the graphite core and CO₂ coolant but switched to 2.5-3.5% enriched uranium. This enabled outlet temperatures of 1200°F and improved thermal efficiency.

However, this performance came with challenges. Hotter carbon dioxide becomes corrosive, attacking steel components. Additionally, decades of neutron bombardment cause the graphite moderator to shrink and crack. As channels distort, fuel assemblies can jam, and control rod insertion becomes uncertain. Of the fourteen AGRs built, eight continue operating in the UK, but graphite degradation limits their remaining lifespans.

High Temperature Gas Reactors (HTGR) Modern HTGR designs use helium as a coolant, with a target temperature of 1380°F. This temperature is hot enough for direct use as a process gas in some industrial applications, such as hydrogen production or steelmaking.

These reactors are often planned to use TRi-structural ISOtropic (TRISO) fuel. This type of fuel consists of tiny uranium kernels wrapped in multiple ceramic layers that contain fission products even at extreme temperatures.

In some pebble bed designs, tennis-ball-sized graphite spheres containing thousands of TRISO particles flow continuously through the core like gumballs in a machine. Fresh fuel is added at the top, and spent fuel is removed at the bottom. This enables refueling without shutdown and inherent safety.

The trade-off is power density. Helium’s low density means it removes heat poorly compared to water, requiring larger cores for the same power output. An HTGR producing 600 MW might have a core volume three times larger than an equivalent PWR. This size penalty, combined with TRISO fuel fabrication complexity, has limited commercial deployment despite the technology’s elegance. China operates the world’s only HTGR.

Molten Salt Reactors

Molten salt reactors (MSR) radically depart from conventional reactor design. Instead of solid fuel rods, they dissolve fuel directly in molten liquid fluoride salts. This is typically a mixture of lithium fluoride, beryllium fluoride, and uranium tetrafluoride. This liquid serves as both fuel and coolant, flowing through the reactor core at 1200°F-1300°F while remaining at atmospheric pressure. Oak Ridge National Laboratory (ORNL) proved the concept with the 8 MW Molten Salt Reactor Experiment (MSRE) from 1965 to 1969.

The architecture offers compelling advantages. Without pressure vessels or water that can flash to steam, MSRs achieve walk-away safety through physics alone. A freeze plug sits at the reactor’s lowest point. This is essentially a section of salt pipe actively cooled to keep it solid. If power fails or temperatures exceed limits, the plug melts and gravity drains the fuel into subcritical geometry tanks where passive cooling prevents damage. No pumps, valves, or operator action are required.

Liquid fuel also enables continuous processing. Gaseous fission products like xenon and krypton bubble out naturally, eliminating the neutron poisoning that plagues solid-fuel reactors. Operators can adjust fuel composition on the fly, adding fresh uranium, removing fission products, or changing the uranium-plutonium ratio. Theoretically, an MSR could run for decades without shutdown, achieving near complete fuel burnup.

Yet, materials challenges are limiting. The fuel salt, while excellent at retaining most fission products, is viciously corrosive. Fluoride salts attack virtually every known alloy, leaching chromium from stainless steel and causing intergranular cracking. ORNL developed Hastelloy-N specifically for MSRE, but even this specialized nickel superalloy suffered embrittlement after prolonged exposure.

There are also unique control and safeguard challenges. Without fuel assemblies to count, proliferation monitoring becomes complex.

No commercial MSR operates today. China is developing a pilot plant, while ventures in Canada and the United States pursue various designs.

Fast Reactors

Fast reactors operate at the opposite end of the neutron spectrum. Without moderators, neutrons maintain their high energies. As described above, this enables breeding by converting uranium-238 to plutonium-239 and waste burning. But it requires high fissile content: 15-20% enriched uranium or plutonium to achieve criticality.

Only two commercial fast reactors operate worldwide, both are in Russia, and both are sodium cooled.

Sodium Fast Reactors (SFR)

Liquid sodium is widely used in fast reactors for good reasons. At atmospheric pressure, it remains liquid from about 208°F up to 1621°F, enabling reactor coolant circuits to operate at low pressure. Its high thermal conductivity and favorable heat transfer properties support high power densities and efficient heat removal. Sodium is a weak neutron moderator and has a low neutron absorption cross section, helping maintain a fast neutron spectrum essential for breeding.

The fundamental challenge is sodium’s violent reactivity with water and air. A pinhole leak sprays sodium that ignites on contact with air, burning at 1,832°F. Water contact triggers explosive reactions, producing hydrogen gas and caustic sodium hydroxide. This drives complex three-loop designs: radioactive primary sodium transfers heat to non-radioactive intermediate sodium, which then generates steam. Each interface is a potential failure point.

Engineering solutions exist, but they add complexity. Argon cover gas prevents air contact. Double-walled piping with leak detection catches failures before they escalate. Dump tanks allow rapid draining during emergencies.

Russia’s BN-600 SFR has operated since 1980, surviving multiple sodium leaks and fires. Its successor, BN-800, began operation in 2014 with improved safety features. France’s Superphénix and Japan’s Monju SFRs both shut down after repeated sodium leaks and operational problems.

Lead Fast Reactors (LFR)

Lead reactors avoid sodium’s fire hazard, using either lead or lead-bismuth as the coolant. Lead doesn’t burn and remains liquid from 621°F to 3,160°F. The Soviet Alfa-class submarines demonstrated the potential. Compact lead-bismuth reactors propelled the submarines to a remarkable 42 knots underwater.

However, lead-bismuth creates polonium-210 through neutron capture, which is one of the most toxic substances known. Just one millionth of a gram is lethal if inhaled. At standard operating temperatures, it volatilizes, contaminating the entire system.

When maintenance was necessary, crews in special suits worked in shifts measured in minutes. Lead also corrodes steel aggressively, dissolving nickel and redepositing it elsewhere, plugging channels. Several Soviet submarines were lost when coolant solidified in port after heating failures.

Russia is building the BREST-300 plant, which uses pure lead to avoid polonium, and the plant is slated to start operating in 2026.

Small Modular Reactors

Small Modular Reactors (SMRs) are a unique approach to reactor design and production. Nearly any of the architectures discussed above can be adapted to this format. SMRs bet on manufacturing over construction. The reactors are typically smaller (under 300 MW), built in a factory, and then shipped to the site for installation.

From a manufacturing perspective, the method promises more controlled manufacturing environments, parallel production, and higher quality control.

Operationally, deploying modular units also has benefits. Plants can be incrementally constructed through the installation of a series of smaller reactors. This enables more gradual capital expenditure, mitigating financial and operational risk. Additionally, small units can be co-located directly with the infrastructure they power, such as airports, datacenters, or military installations.

However, the unit cost of smaller reactors may suffer from reduced economies of scale, and it’s not yet clear what production volume is required to overcome this dynamic. Additionally, there may be a heightened proliferation risk associated with a more distributed footprint.

A substantial number of SMR programs are underway, particularly in the U.S. These range from kilowatt-level units for defense installations to grid-scale arrays of modular units that provide commercial power.

What Actually Gets Built

The nuclear industry’s architecture selections tell two stories. The first is convergence: 306 of 416 operating reactors are PWRs, a 74% dominance that traces directly to submarine hulls. The second is persistence: despite this overwhelming convergence, alternative architectures refuse to disappear.

PWRs won through compound advantages. Admiral Rickover’s submarine program debugged the technology with naval discipline and unlimited budgets. When nations needed civilian power, they adapted what worked underwater. Each plant built deepened the moat: more operational data, refined supply chains, experienced workers, proven regulations.

Yet the alternatives endure where strategic value exceeds economic penalty. Canada’s CANDU reactors trade complexity for fuel independence. Russia’s fast reactors enabled closed fuel cycles. Britain’s aging AGRs soldier on despite graphite degradation. These aren’t mistakes. They’re different answers to the same question, shaped by unique constraints.

The pattern reveals a fundamental tension in nuclear technology. Standardization brings efficiency, learning curves, and cost reduction. But it also creates dependencies on enrichment infrastructure, foreign fuel supplies, and technologies optimized for other nations’ needs.

This tension isn’t a flaw to be fixed but a reality to navigate. The winner won’t be the best reactor physics can produce, but the one that best balances technical merit, economic viability, and strategic value for the nation that builds it. In nuclear power, these three don’t always align. And that’s why, after seven decades, we still see such radically different machines splitting atoms for the same fundamental purpose.

Subscribe now

General References

International Atomic Energy Agency. PRIS: Power Reactor Information System (Overview). Vienna: IAEA, n.d. Accessed August 25, 2025. https://www-pris.iaea.org/

World Nuclear Association. “Nuclear Power in France.” Updated 2025. Accessed August 25, 2025. https://world-nuclear.org/information-library/country-profiles/countries-a-f/france.aspx.

World Nuclear Association. “Nuclear Power in the United Kingdom.” Updated 2025. Accessed August 25, 2025. https://world-nuclear.org/information-library/country-profiles/countries-t-z/united-kingdom.aspx.

World Nuclear Association. “High-Temperature Gas-Cooled Reactors.” Accessed August 25, 2025. https://world-nuclear.org/information-library/nuclear-fuel-cycle/nuclear-power-reactors/advanced-nuclear-power-reactors.aspx#HTGR.

World Nuclear News. “China’s HTR-PM Achieves Commercial Operation.” December 2023. Accessed September 23, 25, 2025. https://www.world-nuclear-news.org/Articles/Chinas-HTR-PM-achieves-commercial-operation.

Westinghouse Electric Company. “RFA-2 Fuel Design.” Data sheet, ca. 2016. Accessed August 25, 2025. https://westinghousenuclear.com/media/yaln2un5/rfa-2-fuel-design.pdf.

University of Wisconsin. “Pressurized Water Reactor Pressure Vessels (Lecture 37).” 1997. Accessed August 25, 2025. https://fti.neep.wisc.edu/fti.neep.wisc.edu/neep423/FALL97/lecture37.pdf.

Massachusetts Institute of Technology (OpenCourseWare). “PWR Description.” Engineering of Nuclear Systems (22.06), lecture slides, 2010. Accessed August 25, 2025. https://mitocw.ups.edu.ec/courses/nuclear-engineering/22-06-engineering-of-nuclear-systems-fall-2010/lectures-and-readings/MIT22_06F10_lec06a.pdf.

International Atomic Energy Agency. Pressurized Water Reactor Simulator (Training Course Series No. 22). Vienna: IAEA, 2003. Accessed August 25, 2025. https://www-pub.iaea.org/MTCD/Publications/PDF/TCS-22_2nd_web.pdf.

U.S. Nuclear Regulatory Commission. “Moderator Temperature Coefficient of Reactivity (Glossary).” Accessed August 25, 2025. https://www.nrc.gov/reading-rm/basic-ref/glossary/moderator-temperature-coefficient-of-reactivity.html.

U.S. Nuclear Regulatory Commission. “SCRAM (Glossary).” Accessed August 25, 2025. https://www.nrc.gov/reading-rm/basic-ref/glossary/scram.html.

Mitsubishi Heavy Industries. “PWR Fuel.” Mitsubishi Nuclear Fuel Co., Ltd. Accessed August 25, 2025. https://www.mhi.com/group/mnf/products/pwr.html.

World Nuclear Association. “Nuclear Power in Canada.” Updated 2025. Accessed August 25, 2025. https://world-nuclear.org/information-library/country-profiles/countries-a-f/canada.aspx.

Oak Ridge National Laboratory. “History | Molten Salt Reactor Experiment (MSRE).” Accessed August 25, 2025. https://www.ornl.gov/molten-salt-reactor/history.

Oak Ridge National Laboratory. “MSRE Operating History (ORNL-TM-3229).” 1970. Accessed August 25, 2025. https://moltensalt.org/references/static/downloads/pdf/ORNL-TM-3229.pdf.

World Nuclear Association. “Lead-cooled Fast Reactors (LFR).” Accessed August 25, 2025. https://world-nuclear.org/information-library/nuclear-fuel-cycle/nuclear-power-reactors/advanced-nuclear-power-reactors.aspx#LFR.

Generation IV International Forum (via OECD/NEA). Sodium-Cooled Fast Reactor (SFR) Systems. Paris: OECD/NEA, various. Accessed August 25, 2025. https://www.gen-4.org/gif/jcms/c_9350/sodium-cooled-fast-reactor-sfr.

World Nuclear News. “Beloyarsk BN-600 Fast Neutron Reactor Gets 15-Year Extension.” April 2025. Accessed August 25, 2025. https://www.world-nuclear-news.org/articles/beloyarsk-bn-600-fast-neutron-reactor-gets-15-year-extension.

World Nuclear News. “BN-600 Licensed to Operate Until 2025.” April 2020. Accessed August 25, 2025. https://www.world-nuclear-news.org/Articles/BN-600-licensed-to-operate-until-2025.

International Panel on Fissile Materials (Hui Zhang). “China Started Operation of its First CFR-600 Breeder Reactor.” December 15, 2023. Accessed August 25, 2025. https://www.belfercenter.org/research-analysis/china-started-operation-its-first-cfr-600-breeder-reactor.

International Atomic Energy Agency. “Small Modular Reactors (SMRs).” Accessed August 25, 2025. https://www.iaea.org/topics/small-modular-reactors.

Elsevier (ScienceDirect). “Experiments on Silver-Indium-Cadmium Control Rod Failure during Severe Accidents.” Annals of Nuclear Energy 94 (2016): 268–284. Accessed August 25, 2025. https://www.sciencedirect.com/science/article/pii/S0306454916304455.

U.S. Nuclear Regulatory Commission. “Safety Evaluation—Control Rod Assemblies: Use of Hafnium or Ag-In-Cd.” ML20247B696, 2020. Accessed August 25, 2025. https://www.nrc.gov/docs/ML2024/ML20247B696.pdf.

Before we begin, a quick request. As Built is still young, and I’m working to grow awareness. I would be forever appreciative if you could share the blog with one person you think might find it interesting. Thanks so much, and I hope you enjoy!

Share As Built

The Sten gun exemplifies rapid engineering and design for manufacturing. With limited resources, this World War II era program accomplished what many modern defense production efforts strive to achieve. In about a month, British engineers created a low cost, mass producible weapon that would eventually arm millions.

The story offers many insights for modern efforts. But to truly understand how the British made this happen, we first need to contextualize the crisis that made it necessary.

Slow Beginnings

Britain’s involvement in what would become World War II began as a gradual diplomatic effort before a sharp inflection transformed it into a wartime crisis. The origin traced back to the 1919 Treaty of Versailles, which ended World War I. It was highly punitive, requiring Germany to disarm, pay reparations, and surrender territory.

As a result, Germany was forced to grapple with an extreme sense of defeat and humiliation. This was inherently destabilizing. Separately, political and social unrest presented across Europe, which was further exacerbated by a highly strained international economy. This led to the period between the world wars being unstable and insecure.

In 1933, Adolf Hitler became Chancellor of Germany and quickly began to rearm the country. While such development of the German military was prohibited under the Treaty of Versailles, the action drew relatively little international resistance. This set the stage for German territorial expansion.

In March 1938, Germany annexed Austria. Six months later, Hitler demanded possession of the Sudetenland, a border region of Czechoslovakia. Britain and France appeased and compelled Czechoslovakia to cede the Sudetenland in an effort to achieve peace in Europe. Yet, in March 1939, Hitler violated this agreement and moved to occupy the rest of Czechoslovakia.

Alarms sounded across Europe, and Britain shifted its policy. In an attempt to curb German expansion, Britain and France provided security guarantees to Poland, which were promptly invoked. On September 1st, 1939, Germany invaded Poland. Two days later, Britain and France declared war on Germany. Seven years of diplomacy had failed.

Phoney War

Britain deployed an expeditionary force to mainland Europe, but the war’s initial phase left observers bewildered. From September 1939 to May 1940, virtually no Allied military land operations occurred on the Western Front. There was so little military activity that this period became known as the “Phoney War.” Years of slow diplomacy gave way to slow battle.

In sharp contrast, at sea, German U-boats were active. Over the course of 1940, they would sink 563 ships in what was referred to as a “happy time” for the German submarine force. This prowess would later serve as a significant factor behind the Sten’s development.

Blitzkrieg Inflection

The slow Phoney War came to an abrupt end on May 10th, 1940. German troops, tanks, and aircraft poured into Belgium, Luxembourg, and the Netherlands. Within days, German forces advanced over a hundred miles. The Blitzkrieg — lightning war — had arrived in Western Europe with devastating effect.

The Allied response crumbled. Forces rushed north to meet what they expected would be the main German thrust through Belgium, as had happened in World War I. However, the German attack traversed the Ardennes Forest in southeastern Belgium and northern Luxembourg — terrain the French had previously considered impassable for armor. German tanks broke through French defensive lines and raced toward the coast.

Over the next two weeks, Allied forces were forced backwards at an astonishing rate. Quickly, the retreat turned catastrophic. British and French troops found themselves trapped against the English Channel. Rapidly advancing German forces led to the famous Dunkirk evacuation, where British civilians crossed the Channel to aid the Navy in rescuing the trapped soldiers.

Seven years of failed diplomacy had evolved into eight months of largely inactive war, which then collapsed into three weeks of devastating retreat. Britain was stunned.

Outgunned

While the May 1940 Dunkirk evacuation successfully rescued 338,000 Allied soldiers from the shores of France, most of their equipment was left behind. Britain lost an estimated 2,500 artillery pieces, 85,000 vehicles, 75,000 tons of ammunition, and thousands of machine guns on French beaches.

These equipment losses were only half the problem. Even before Dunkirk, British forces had been outgunned. Their bolt-action rifles were no match for German submachine guns that could deliver 500 rounds per minute. The British had neither the right weapons nor enough of them.

The German forces’ lightning speed, coupled with persistent air raids, made an invasion of Britain seem inevitable. Britain prepared by mobilizing 1.5 million volunteers into the Home Guard. Their first mission was to interfere with and slow down invading forces until the regular army could respond. Their second mission was to continue the resistance in the event of an occupation.

But while Britain was long on militia volunteers, it was short on weapons to arm them.

Three Bets

To address this problem, Britain launched three parallel efforts, which is itself a case study in parallel pathing and risk management.

Emergency Weapons
The first effort was to get weapons — any weapons — into soldiers’ hands. Britain pulled together about 300,000 Lee Enfield rifles and 59,000 Lewis guns from World War I stores. The regular army took most of the rifles, while the Home Guard received the Lewis guns, as well as about 50,000 shotguns collected from the civilian population. This approach had minimal cost and lead time, but the weapons were outdated and the quantities were relatively low.

Imports
Second, Britain bought whatever it could find. Canada sent 75,000 Ross Mark III rifles — a World War I era weapon that had been withdrawn from the front lines in 1916 for jamming. The United States sold 785,000 M1917 rifles and 108,000 Thompson submachine guns.

The Tommy guns were a step up in terms of technology, but this strategy was expensive and uncertain. At this point, each Tommy gun cost $4,600, and the total purchase topped $800 million, both in modern dollars. Notably, the Lend-Lease program had not yet been established, so this was a cash-and-carry deal structure, which drained British resources.

Even after the purchase, the weapons still had to cross the Atlantic, and this was the U-boat “happy time”. The German Navy was now operating from French ports, further extending its maritime dominance. Every shipment risked sending desperately needed weapons to the ocean floor.

Domestic Production
Third, Britain sought to develop its own weapons domestically. Some designs were in production, but they weren’t quite what was needed. For example, the Bren was a highly favored British machine gun that was developed throughout the late 1930s. It was reliable and accurate. But it was a large squad-level weapon, expensive, and required precision manufacturing. By the time of the Dunkirk evacuation, only about 30,000 had been produced, and the majority had been lost in France.

Britain needed something different: individual weapons for every soldier, rapidly mass produced at low cost. As the British Small Arms Committee framed the problem, “The most important consideration at the moment seems to be to get some form of machine carbine acceptable to all three services into production as quickly as possible.”

The first attempt was to reverse engineer and copy German weapons. The British designed the Lanchester submachine gun, which was a direct adaptation of the German MP 28. It was considered to be well made but required skilled craftsmen and precision parts. While cheaper than the Bren and the Tommy gun, it was still expensive. This was a step in the right direction, but still not quite what Britain needed.

36 Days to Save the Country

In December 1940, Lanchester manufacturing was scaling up at the Royal Small Arms Factory. There, a draftsman named Harold Turpin was supposed to be creating production drawings for the Lanchester gun.

However, he saw another way. Working after hours with Major Reginald Shepherd, a World War I veteran recalled from retirement, Turpin started sketching a different kind of weapon. Not better. Simpler.

The design philosophy was fundamentally different than other weapons programs. What if you designed a gun specifically for people who had never built guns before? What would the program look like with a focus on production over precision?

Turpin’s first insight came with the trigger mechanism. Where the Lanchester had complex machined parts, he designed a trigger that needed just two pieces of bent metal. This thinking spread to every component. The grip was a bent steel tube. The barrel’s protective shroud was made of stamped metal with holes cut into it.

By January 1941, just 36 days after starting, Turpin had built the first prototype. By many accounts, he made most of the parts himself. Some were even done in his home workshop. The weapon looked as if it had been assembled in someone’s garage, because, essentially, it had.

Looks aside, the Sten embodied three key design principles:

1. Minimal precision machining: Of the Sten’s 59 parts, only the bolt and barrel required precision work. Semi-skilled workers could stamp or weld everything else from stock metal. This eliminated the bottleneck of specialized machine tools and expert machinists.

2. Distributed, low infrastructure production: The simple components could be manufactured nearly anywhere. Britain set up production lines in converted furniture workshops, bicycle factories, and even private homes. Famously, the Tri-Ang toy factory was converted into a 24-hour Sten production line. Hundreds of small operations emerged across the country. Bomb one, and production continued elsewhere.

3. Widely available ammunition: The Sten fired 9mm rounds, the same ammunition used by German forces. This decision was strategic. Captured enemy ammunition could be used immediately, meaning resistance fighters across occupied Europe could resupply from German sources. This reduced the pressure to scale ammunition production.

Throughout early 1941, Britain tested a variety of prototypes. By March, less than ten months after Dunkirk, they placed an order for 100,000 Sten guns. Yet, the designers kept simplifying. Version after version, they chopped every non-essential feature: flash hiders, wooden handguards, folding foregrips. If it didn’t help the gun shoot, it disappeared.

These decisions drove serious results. Eventually, the Sten could be built in five and a half hours and for $10 ( $220 in modern dollars). This was a stark contrast to the Tommy gun, which took days to make and cost thousands of dollars. Over the course of World War II, Britain would produce 4.5 million Stens.

The Gun They Could Build

The Sten delivered submachine gun capability at rifle-like prices and was churned out by the thousands in bicycle shops and furniture factories. By many measures, this was a success.

However, the design choices that enabled mass production created predictable problems. A simple blowback operation led to frequent jams. Minimal safety mechanisms caused accidental discharges. While easy to manufacture, the side-mounted magazine made handling awkward. Accuracy beyond 100 yards was negligible. Critiques vocally pointed all this out.

Yet, none of this mattered strategically. The Sten was designed for urban combat, resistance operations, and last ditch defense. Two factors matter in these scenarios. First, having a gun at all, and second, maximizing the volume of fire. The Sten checked both boxes.

High rate production also meant commanders could issue Stens more liberally than other weapons. Resistance groups across occupied Europe received them via airdrop. The weapon’s compatibility with captured German ammunition made it ideal for guerrilla operations behind enemy lines. Flaws existed, but in the calculus of total war, they were acceptable.

Decades of Service

Some of the most telling validation came decades later. While it was designed in 36 days as an emergency stopgap, the Sten remained in service with the British military until the 1960s and with other forces worldwide until the 1980s. The “proper” weapons it was meant to temporarily replace were retired decades earlier.

Britain’s 1940 crisis forced a fundamental question: do you want the best weapon or the weapon you can actually build? Britain did the latter, and they weren’t the last to face this choice.

Today, Ukraine is building drones in distributed Kyiv apartments using 3D printers and commercial components. At first, they crashed, were jammed, and had a limited range. However, Ukraine has built thousands of them while rapidly iterating on the technology. Now these systems have proven formidable against a better armed adversary.

The lesson of 1940 keeps teaching itself. In existential conflicts, given the resources on hand, the weapon you can build will always beat the weapon you can’t.

Subscribe now

General References

“Treaty of Versailles - Reparations, Military, Limitations,” Encyclopædia Britannica, last updated August 26th, 2025, https://www.britannica.com/event/Treaty-of-Versailles-1919/German-reparations-and-military-limitations.

“Rearming Germany,” Facing History & Ourselves, accessed September 21st, 2025, https://www.facinghistory.org/resource-library/rearming-germany.

“Appeasement and ‘Peace for Our Time,’” The National WWII Museum, October 14th, 2024, https://www.nationalww2museum.org/war/articles/appeasement-and-peace-our-time.

“Munich Agreement,” Encyclopædia Britannica, last updated July 20th, 1998, https://www.britannica.com/event/Munich-Agreement.

“Invasion of Poland, Fall 1939,” Holocaust Encyclopedia, United States Holocaust Memorial Museum, accessed September 21st, 2025, https://encyclopedia.ushmm.org/content/en/article/invasion-of-poland-fall-1939.

“How Europe Went To War In 1939,” Imperial War Museums, accessed September 21st, 2025, https://www.iwm.org.uk/history/how-europe-went-to-war-in-1939.

“The Phoney War,” Legion Magazine, February 15th, 2020, https://legionmagazine.com/the-phoney-war/.

“Battle of France,” Encyclopædia Britannica, last updated May 10th, 2019, https://www.britannica.com/event/Battle-of-France-World-War-II.

“Blitzkrieg Explained,” Imperial War Museums, accessed September 21st, 2025, https://www.iwm.org.uk/history/blitzkrieg-explained.

“Cheap Shot: How the Sten Gun Saved Britain,” HistoryNet, October 30th, 2019, https://www.historynet.com/cheap-shot-how-the-sten-gun-saved-britain/

“The Sten Gun, Its Name and the Men Behind It,” The Armourers Bench, May 18th, 2020, https://armourersbench.com/2020/05/17/the-sten-gun-its-name-and-the-men-behind-it/

“British Infantry Weapons of World War II,” Infantry Weapons, archived May 22nd, 2009, https://web.archive.org/web/20090522121134/http://infantry-weapons.org/smg_lmgs.htm

“Tri-ang: From Toys to Stens,” Royal Armouries, https://royalarmouries.org/objects-and-stories/stories/tri-ang-from-toys-to-stens

Carter, David M. “Sten Mk3: Britain’s Cheapest Submachine Gun of WWII.” Firearms News, August 1st, 2022. https://www.firearmsnews.com/editorial/sten-mk3-submachinegun-ww2/481344

McCollum, Ian. “British Submachine Gun Overview: Lanchester, Sten, Sterling, and More.” Forgotten Weapons, August 7th, 2014. https://www.forgottenweapons.com/british-submachine-gun-overview-lanchester-sten-sterling-and-more/

McCollum, Ian. “Sten MkIII: A Children’s Toy Company Makes SMGs.” Forgotten Weapons, April 4th, 2011. https://www.forgottenweapons.com/sten-mkiii-a-childrens-toy-company-makes-smgs/

Kay, Garry James. “The Weapon That Saved the British.” Recoil, April 10th, 2017. https://www.recoilweb.com/the-weapon-that-saved-the-british-106527.html

“Equipment the British Lost at Dunkirk that the Germans Reused,” Military Matters, September 14th, 2021, https://militarymatters.online/military-history/equipment-the-british-lost-at-dunkirk-that-the-germans-reused/

“US Aid For The WW2 British Home Guard,” M. Watkin, accessed December 2024, https://www.mwatkin.com/us-aid-to-britain

“Money & Guns in WWII, the Cost of SMGs and Sidearms,” War History Online, April 24th, 2019, https://www.warhistoryonline.com/instant-articles/money-and-guns-in-wwii.html

“Ship losses by month,” Uboat.net, https://uboat.net/allies/merchants/losses_year.html

“Wartime Changes: The Bren MkI Modified and Bren MkII,” Forgotten Weapons, https://www.forgottenweapons.com/wartime-changes-the-bren-mki-modified-and-bren-mkii/

“Ugly But Effective: How the Sten Gun Became a WWII Workhorse,” ITS Tactical, May 7th, 2018, https://www.itstactical.com/warcom/firearms/ugly-but-effective-why-the-sten-gun-became-a-wwii-workhorse/

“The Sten - Meet the $10 Submachine Gun That Helped the Allies Win WW2,” MilitaryHistoryNow.com, May 3rd, 2018, https://militaryhistorynow.com/2013/12/09/the-venerable-sten-britains-10-dollar-submachine-gun/

“How the STEN Submachine Gun Helped the British Army During WWII,” War History Online, October 31st, 2022, https://www.warhistoryonline.com/instant-articles/sten-submachine-gun-british-army-wwii.html

Power is estimated to have hit 30,000 MW within seconds, but the true peak is unknown since instrumentation failed quickly. Driven by a control rod design flaw and a powerful positive feedback loop, this spike initiated one of the most infamous nuclear accidents in history.

To understand why this happened, we follow the neutron — the invisible particle that determines whether reactors generate power or catastrophe. In doing this, we'll look at how engineers harness nuclear fission, making the process self-sustaining and energy-generating, while also safe and controllable.

This piece builds on How nuclear physics works. Check it out if anything is unfamiliar.

Subscribe now

Following the Neutron

Recall from our last piece on nuclear physics that during fission, a uranium atom splits into smaller nuclei, releasing enormous energy and 2-3 new neutrons. While impressive, a single fission event is not enough to generate meaningful energy. We need trillions of fission events to occur every second. We achieve this by using the neutrons from one fission to initiate more. But there can't be too many. Otherwise, the chain reaction could run away, which is what happens in a meltdown or a bomb.

Unfortunately, we can't control individual neutrons. They're too small, too fast, and too numerous. Instead, we control their environment. By choosing the right materials and geometry they encounter, we can tip the odds, making fission more likely here or absorption more likely there. This approach forms the basis of modern reactor engineering.

Neutron Strikes

Our primary point of influence is when a neutron collides with an atom. Once released during fission, neutrons race through any material in their path — fuel, coolant, and reactor structure. Eventually, they collide with the atomic nuclei of these materials.

When a neutron strikes an atom, one of three things happens: it bounces off (scattering), gets captured (absorption), or splits the atom (fission).

Scattering can take two forms:

• Elastic scattering conserves the total kinetic energy while redistributing it between particles. A collision with a small nucleus, like hydrogen, can stop a neutron completely. Impacts with larger nuclei slow them down much less. For this reason, light atoms like hydrogen serve as good neutron moderators, slowing them down.

• Inelastic scattering occurs when fast neutrons carry enough energy to excite the nuclei of heavy atoms. The neutron loses kinetic energy by lifting the nucleus to a higher energy state. The excited nucleus then emits gamma rays to return to the ground state, therefore converting the neutron's motion into radiation.

Absorption permanently captures the neutron. The nucleus becomes heavier by one neutron and often releases gamma rays to shed excess energy. For example, when uranium-238 absorbs a neutron, it transforms into uranium-239, initiating a decay chain that ultimately produces plutonium-239.

Fission is a special type of absorption. The captured neutron tips a delicately balanced nucleus past its breaking point. Uranium-235 absorbs a neutron to become uranium-236, but this heavier nucleus is unstable. It splits violently into two fragments, ejecting 2-3 new neutrons and releasing 200 MeV of energy.

These interactions occur probabilistically. Every collision has some chance of each outcome, and the two biggest influencers are the target material and the neutron's energy at the point of collision.

Measuring Probability: Cross-Sections

To control these collisions, we first need a way to characterize their probabilities. To do this, we use a metric called cross-section. This isn't necessarily a measure of physical area, but a representation of the likelihood of an interaction occurring. It's an effective target area. If nuclei were dartboards, cross-sections would measure how big they appear to incoming neutrons.

Each type of interaction has its cross-section:

• Scattering: the chance of bouncing off

• Absorption: the chance of being captured

• Fission: the chance of splitting the nucleus

We measure these phantom targets in units of barns. A barn is equivalent to 10^-28 square meters or roughly the size of a uranium nucleus. The unit's name comes from physicists' surprise at discovering certain cross-sections so large they were "as easy as hitting the broad side of a barn."

Some interactions are indeed that easy: with a 30,000-barn absorption cross-section, cadmium-113 voraciously captures thermal neutrons, making it ideal for use in control rods.

The cross-sections don't just depend on the target material. They shift dramatically with neutron speed, turning the same nucleus from nearly invisible to unmissable.

The Speed of Neutrons

Neutron speed and energy are directly related, and we generally categorize neutrons by their energy levels as:

• Fast Neutrons: above 100 keV

• Epithermal Neutrons: 0.1 eV to 100 keV

• Thermal Neutrons: below 0.1 eV

When a neutron is first released from fission, it has about 2 MeV of energy and is moving at about 7% of the speed of light. This makes it a fast neutron. At those speeds, the probability of fission with most relevant materials is low.

However, as the neutron impacts many different atoms, it quickly loses energy through elastic scattering. After a few dozen collisions, the neutron has lost most of its initial kinetic energy and reaches thermal equilibrium with its surroundings.

In other words, it has about the same energy as the surrounding atoms due to their random thermal motion. At room temperature, this is about 0.025 eV. When the neutron hits this point, we consider it a thermal neutron.

Thermal neutrons behave dramatically differently. For example, uranium-235's fission cross sections are:

• 1.3 barns for fast neutrons

• 585 barns for thermal neutrons

This means that a thermal neutron is 450 times more likely to induce fission in U-235 than a fast neutron.

The process of slowing a neutron down is known as moderation, and is a cornerstone of reactor engineering. Materials like water and graphite are great moderators, since they contain light elements — hydrogen, carbon, oxygen — that collect neutron energy during elastic collisions.

Yet even with perfect moderation, controlling a chain reaction would be impossible without a unique quirk of nuclear physics: not all neutrons appear at once after a fission event.

The Time Delay That Saves Us

When a uranium atom splits by fission, about 99.3% of neutrons appear almost instantly (within nanoseconds) and are called prompt neutrons. The remaining 0.7% are known as delayed neutrons and are released a few milliseconds to a few minutes later. They come from the decay of certain fission products.

This small fraction of delayed neutrons is crucial for reactor control. Without them, the timescale from neutron-induced fission to a new batch of neutrons would be extremely short. Small increases in neutron quantity could cause reactor power to increase exponentially on a microsecond timescale, which is too fast to reasonably control with mechanical systems.

Delayed neutrons stretch that timescale from microseconds to tenths of a second. This is within the response time of modern control systems, giving operators time to adjust control rods and manage the changing reaction.

Criticality

If we combine all the neutron factors covered so far, we can achieve the most fundamental balance in a reactor: criticality. This is the point at which a fission chain reaction is self-sustaining.

Consider a uranium-235 fission that releases three neutrons. If all three go on to induce additional fissions, the reaction accelerates. One fission causes three fissions. If none cause additional fissions, then the reaction dies out. One fission causes zero fissions.

The sweet spot is exactly one neutron induces one new fission. The other two neutrons get absorbed or escape. One fission causes one fission, creating the perfect balance of criticality.

We use this threshold to characterize the reactor's operating state and describe whether the neutron population is stable, growing, or shrinking:

• Subcritical: Each fission leads to less than one additional fission. The reaction diminishes and eventually stops.

• Critical: Each fission event leads to exactly one more fission event. The reactor maintains steady power.

• Supercritical: Each fission leads to more than one additional fission. Power increases.

Critical and Prompt Critical

Generally, criticality is considered to include both the prompt and delayed neutrons. Alternatively, if only prompt neutrons sustained the chain reaction, then it is called prompt critical. Such a scenario can be uncontrollable. The prompt neutrons would sustain the reaction, and every additional delayed neutron that presented would further increase power levels.

For example, consider a system with 1,000 total neutrons, comprised of 993 prompt and 7 delayed neutrons.

• If all 1000 neutrons are required to sustain the chain reaction, then the system is critical.

• If only the 993 prompt neutrons are needed, then the system is prompt critical.

Nearly all reactors are designed to operate with delayed neutrons factored in. Weapons, on the other hand, do not. They typically rely on prompt neutrons alone for their explosive power release.

Reactivity

One challenge with criticality is that it's difficult to measure it directly. Instead, we use the concept of reactivity to characterize how far a reactor is from criticality.

First, let's think of neutrons in generations. Each generation is born and either creates another generation or is lost to the system. By comparing one generation to the next, we can determine whether the neutron population is growing, shrinking, or staying constant.

We use the neutron multiplication factor, k-effective (keff), to make this comparison.

This measurement is centered around 1, which is the threshold for reactor behavior:

• keff < 1: Subcritical (power falling)

• keff = 1: Critical (power steady)

• keff > 1: Supercritical (power rising)

Next, we rearrange the equation to magnify changes and define reactivity as:

This approach centers reactivity around zero:

• ρ<0: Negative reactivity, fewer neutrons in each generation, power falling

• ρ=0: Zero reactivity, steady neutron population, reactor critical

• ρ>0: Positive reactivity, more neutrons in each generation, power rising

Dollars and Cents

Although reactivity is dimensionless, scientists use various units to reference it. In the US, reactivity is often measured in dollars and cents. One dollar of reactivity is the amount between normal criticality and prompt criticality. A cent is one-hundredth of a dollar. At zero cents, we're critical. At 100 cents (one dollar), we're prompt critical.

• < 0$: Reactor is subcritical

• = 0$: Reactor is critical, using both prompt and delayed neutrons (typical steady state operation)

• 0$ to 1$: Reactor is supercritical, but still depends on delayed neutrons, and power rise is slow enough to be safely controlled

• = 1$: Reactor is prompt critical, and power increases faster than conventional control mechanisms can respond

Every action affects reactivity. Insert control rods: negative. Add fuel: positive. Increase temperature: negative (in properly designed reactors).

With these fundamental concepts, we can now understand how engineers select materials and design reactors.

Engineering a Reactor

In designing a reactor, the core objective is to achieve and control criticality. However, there are many solutions to this problem. Each comes from a different combination of architectural, design, and operational decisions. The choices have significant path dependency, and all must be considered in unison as the reactor is scoped.

Thermal vs. Fast Architecture

The top-most decision is whether the reactor operates in the fast or thermal regime. So far, we've highlighted the need to moderate neutrons to thermal energies, but reactors can also be built to operate with fast neutrons.

Known as fast reactors, these plants are designed to work with unmoderated fast neutrons. Since fission cross sections are smaller with fast neutrons, these plants require more highly enriched uranium (15-20% or more) or plutonium-based fuel.

The approach brings compelling advantages. Fast reactors breed more fuel than they consume by converting U-238 to plutonium. Their higher energy density enables smaller cores, while liquid metal coolants allow high-temperature operation without pressurization.

Yet fast reactors remain rare commercially. Sodium coolant burns on contact with air and reacts violently with water. Higher enrichment raises proliferation concerns. High construction costs have limited adoption.

Fuel Considerations

Fission reactors can be powered by uranium and plutonium, while thorium can also be used indirectly as a fuel.

Uranium

Uranium is the most common fission fuel. It is a naturally occurring resource and typically found as 99.3% uranium-238 and 0.07% uranium-235. Of these isotopes, only uranium-235 is fissile and can be used for fission. Some reactors are engineered to operate with natural uranium and low concentrations of uranium-235.

However, most require a higher concentration of fissile uranium-235. The process of enrichment increases the relative concentration of uranium-235 atoms to those of uranium-238. Commercial reactors commonly use 3-5% enriched uranium, but advanced designs consider levels up to 20%.

Plutonium

For all practical purposes, plutonium is not a naturally occurring element. Instead, it's created as a byproduct inside uranium reactors. Uranium-238 absorbs a neutron and becomes uranium-239, which is unstable. It undergoes beta decay, transforming into neptunium-239, which is also unstable. After one more beta decay, plutonium-239 emerges. This process happens anytime uranium-238 is exposed to neutrons, which occurs in all commercial reactors.

U-238 + neutron -> U-239 -> Np-239 + beta emission -> Pu-239 + beta emission

Interestingly, plutonium itself is fissionable. Therefore, a reactor begins burning pure uranium and gradually transitions to a mixture of uranium and plutonium. The breeding is significant. Though no plutonium is loaded initially, by the end of a fuel cycle, it contributes roughly one-third of the total energy output.

Plutonium-239 is an exceptional reactor fuel. Its thermal fission cross-section is 750 barns—even higher than uranium-235's 585 barns—and it requires a smaller critical mass. These same properties, however, make it suitable for nuclear weapons, creating a proliferation risk.

Thorium

Thorium has long been contemplated as an alternate nuclear fuel. Thorium-232 can breed into fissile uranium-233, which has a 530-barn cross-section, rivaling that of uranium-235. However, the process also creates uranium-232, which emits fierce gamma radiation. This makes fuel handling particularly hazardous.

Thorium is attractive due to its abundance, which is approximately three times that of uranium. However, while several countries pursue thorium research, only India has made it central to their nuclear strategy.

Fuel Mass, Shape, and Enrichment

Once a fuel is selected, mass and geometry influence how many neutrons stay inside the core versus escaping. A sphere minimizes surface area and is the most efficient at retaining neutrons. Depending on the shape, a certain amount of mass is required. Below critical mass, too many neutrons flee before causing fissions, and the chain reaction cannot sustain itself.

Additionally, in the case of uranium, the enrichment level determines the number of fissile targets. By increasing the concentration of uranium-235 targets through enrichment, fission probabilities can be improved.

Operational Factors

Beyond design, specific passive and active operational considerations also influence criticality. Temperature can provide inherent safety. Most modern reactors have a negative temperature coefficient of reactivity, meaning that rising temperature naturally reduces the rate of fission. This is a deliberate strategy to prevent runaway reactions. This is achieved through mechanisms such as:

• Doppler Broadening: As fuel temperature rises, uranium-238 nuclei vibrate more vigorously. This broadens their neutron absorption resonances, increasing the likelihood that they will capture neutrons instead of allowing them to cause fission. The effect is instantaneous — occurring in microseconds — and acts as the first, fastest line of defense against sudden power surges.

• Void Formation: In thermal reactors that use water as a moderator, higher temperatures reduce water density, creating steam bubbles (voids). Steam is far less effective at slowing down neutrons than liquid water, resulting in fewer neutrons in the thermal energy range, which in turn reduces the fission rate. This feedback is critical in boiling water reactors and pressurized water reactors, but does not occur in gas-cooled reactors or liquid sodium fast reactors.

• Thermal Expansion: As the fuel and core structures heat up, the atoms move farther apart, slightly reducing fuel density and increasing neutron leakage. This decreases the probability of neutron–fuel interactions that cause fission. While slower than Doppler broadening — acting over seconds rather than microseconds — it provides an additional layer of stabilizing feedback.

These negative feedback mechanisms create inherent stability. If a reactor begins to overheat, physics itself applies the brakes—no control rods, no operators, no power required. This self-regulating behavior underpins modern reactor safety.

Chernobyl’s RBMK design violated this principle catastrophically. Unlike water-moderated reactors, the RBMK used graphite for moderation and water primarily as coolant and neutron absorber. When water boiled into steam, its ability to absorb neutrons decreased, while graphite continued moderating. This left more neutrons available for the chain reaction, increasing reactivity — a positive void coefficient. Under the unstable conditions during the accident, rising power produced more steam, which in turn increased power further, creating a runaway feedback loop. Within seconds, this led to a massive steam explosion and subsequent core destruction.

For active control, we use control rods. These are made from neutron-hungry materials like boron or cadmium; they're inserted to absorb excess neutrons or withdrawn to increase reactivity. They're the throttle and brake of nuclear reactions, giving operators direct plant control.

Selecting Materials

Lastly, we must pick the materials we use throughout the plant. Since our primary method of controlling neutrons involves collisions, the type of material they impact determines everything. Looking at the problem through this lens, we can use the various cross-sections to consider which materials are best for which role.

• Fissile fuels have large fission cross-sections for sustained chain reactions, such as uranium-235 and plutonium-239.

• Fertile materials absorb neutrons without fissioning and instead breed new fissile isotopes. Uranium-238 can capture a neutron to create plutonium-239 while leaving enough neutrons for the chain reaction.

• Control materials feature enormous absorption cross-sections. Examples include cadmium and boron. Natural boron's 3,840-barn absorption cross-section makes it ideal for control rods and emergency shutdown.

• Structural materials must have minimal absorption to avoid influencing the nuclear reactor. Zirconium is nearly transparent to neutrons, perfect for cladding for fuel rods without changing the reaction dynamics.

• Moderators combine high scattering with low absorption. We want to slow a neutron down, but not capture any. Water, heavy water, and graphite are all effective moderators. Water's hydrogen atoms have a scattering cross section of 20 barns, but an absorption cross section of only 0.33 barns. They slow down neutrons with minimal losses.

• Coolants vary by reactor type. Water often serves dual purposes in thermal reactors as both moderator and coolant. In fast reactors, sodium and lead are best as they have low absorption and no moderation.

Harnessing the Neutron

Ultimately, nuclear reactors come down to controlling neutrons, which is an exercise in exposing them to the right materials at the right times. As we saw in the last piece, the nuclear fission reaction enables the release of immense energy by splitting atoms. The strategies outlined in this piece represent the ways we can harness, control, and sustain this phenomenon. They're the tools in the toolkit of a reactor designer.

In the following piece, we'll examine the various types of reactors that can be created using these tools. We'll see radically different reactor designs. Pressurized water reactors combine moderators and coolants into one fluid. CANDU plants use heavy water to burn natural uranium without enrichment, while fast reactors can breed fuel for other reactors.

While the core physics always remain the same, this toolkit enables a variety of different goals to be achieved by reactor designers around the world.

Subscribe now

General References

Benedict, Manson, Thomas H. Pigford, and Hans Wolfgang Levi. Nuclear Chemical Engineering. 2nd ed. New York: McGraw-Hill, 1981.

Bickel, John H. "Module 3: Neutron Induced Reactions." Department of Nuclear Engineering, North Carolina State University, 2012. https://www.nrc.gov/docs/ML1214/ML12142A051.pdf.

International Atomic Energy Agency. INSAG-7: The Chernobyl Accident: Updating of INSAG-1. Vienna: IAEA, 1992. https://www-pub.iaea.org/MTCD/Publications/PDF/Pub913e_web.pdf

International Atomic Energy Agency. Thorium Fuel Cycle — Potential Benefits and Challenges. IAEA-TECDOC-1450. Vienna: IAEA, 2005. https://www-pub.iaea.org/MTCD/Publications/PDF/TE_1450_web.pdf

Keepin, G. R. Physics of Nuclear Kinetics. Reading, MA: Addison-Wesley, 1965.

Knoll, Glenn F. Radiation Detection and Measurement. 4th ed. Hoboken, NJ: John Wiley & Sons, 2010.

Lamarsh, John R., and Anthony J. Baratta. Introduction to Nuclear Engineering. 3rd ed. Upper Saddle River, NJ: Prentice Hall, 2001.

National Nuclear Data Center. "Chart of Nuclides." Brookhaven National Laboratory. Accessed August 2025. https://www.nndc.bnl.gov/chart/.

OECD Nuclear Energy Agency. Technology Roadmap Update for Generation IV Nuclear Energy Systems. Paris: OECD-NEA, 2014. https://www.gen-4.org/gif/upload/docs/application/pdf/2014-03/gif-tru2014.pdf

Stacey, Weston M. Nuclear Reactor Physics. 3rd ed. Weinheim: Wiley-VCH, 2018.

Tucker, Colin. How to Drive a Nuclear Reactor. Cham: Springer International Publishing, 2020.

World Nuclear Association. "Fast Neutron Reactors." Last modified March 2024. https://world-nuclear.org/information-library/current-and-future-generation/fast-neutron-reactors.aspx.

World Nuclear Association. "Physics of Uranium and Nuclear Energy." Accessed May 2025. https://world-nuclear.org/information-library/nuclear-fuel-cycle/introduction/physics-of-nuclear-energy.

World Nuclear Association. "Plutonium." Last modified August 16, 2023. https://world-nuclear.org/information-library/nuclear-fuel-cycle/fuel-recycling/plutonium.

A few days ago, SpaceX scrubbed a Starship launch. The culprit wasn't engine trouble or a technical malfunction. It was a violation of the anvil cloud rule. To most observers, the Texas sky looked clear enough for flight. But somewhere nearby, a flat-topped anvil cloud lurked.

These far-reaching cloud structures can deliver devastating lightning blows from miles away, threatening vehicles in flight. Understanding the invisible risks presented by these weather features contextualizes why some launches go and others scrub — even on clear sunny days.

Subscribe now

Why lightning matters

First, why do lightning strikes matter? When lightning hits a vehicle in flight — such as a rocket or an airplane — it delivers a significant charge. Ideally, the vehicle's skin and structure act as a Faraday cage, keeping the electrical charge on the exterior. And for this reason, flight vehicles often use conductive paint and take care to ground all components.

If electrical bonds are poor or the magnitude of the charge is simply high enough to overcome the protections, damage can occur. Gaps in bonding can become arc paths, which pose an ignition risk if they’re near fuel tanks. Additionally, electronics can be upset, impacting the mission.

Aviation lightning strikes are relatively common. A commercial aircraft is typically struck about once every 3000 flight hours, or about twice a year. These vehicles are highly engineered to direct the current to specific exit points on the aircraft and down to the ground.

Launch vehicle strikes are less frequent, largely because operators go to great lengths to avoid them after a string of notorious accidents. In 1969, Apollo 12 was famously struck by lightning, temporarily knocking out power and threatening what would become the second moon landing mission.

Later, in 1987, a lightning strike on the Atlas Centaur 67 mission disabled the control system, resulting in the complete loss of the mission — one of the worst possible outcomes.

In 2019, a Soyuz vehicle was struck on ascent. It continued with its mission, but provided some of the most stunning imagery of a launch vehicle lightning strike.

Most of the strikes discussed above came from "triggered lightning," which is a separate phenomenon where a plane or rocket induces lightning. However, they're illustrative of the impact any lightning can have on a vehicle in flight. As a result, mission planners take significant precautions to avoid all types of lightning, some of which can originate from anvil clouds.

How clouds get charged

We often take for granted that electrical strikes can come from clouds. However, it's helpful to examine how exactly they become charged in the first place.

Clouds start with transparent water vapor in the air. As relative humidity approaches 100%, water vapor condenses and droplets form around tiny dust or aerosol particles, called cloud condensation nuclei. These water droplets are opaque and create clouds.

As temperatures drop below freezing, some water droplets freeze. However, temperature alone isn't enough to turn water to ice. Similar to droplet formation, ice also requires an ice nuclei particle to seed and initiate growth. These are not as common, which means not every water droplet will turn into an ice crystal. Instead, as temperatures drop, some water droplets become supercooled and can be as cold as -40°F while still in liquid form.

When the supercooled water droplets touch the surface of an ice crystal, they will freeze to it and expand in size. This creates soft hail balls, known as graupel.

At this point, we have three types of moisture floating around in the clouds:

• Small ice crystals

• Supercooled water droplets

• Soft hail known as graupel

Winds are constantly blowing these components around in the cloud, causing them to bump into each other. When ice crystals and graupel collide, they build an electrical charge via friction. At temperatures below approximately 5°F, the ice crystals become positively charged, while the graupel acquires a negative charge.

Usually, in cumulonimbus clouds, there is a strong updraft in the central part of the cloud base, driving air upward. The small ice crystals (positively charged) get caught in this updraft and are carried upwards. The larger and denser graupel (negatively charged) tends to fall or stay suspended in place.

As a result, the cloud begins to take on a charge structure. The upper portion is positively charged, and the central lower region is negatively charged. This dipole charge structure is the simplest way to understand cloud electrification. However, in reality, the charging process can be more complex, sometimes with other small portions of the cloud also acquiring a positive charge.

Lightning strikes

The ground beneath a storm is induced to be positive by the cloud's negative charge center. The air gap between the cloud and the Earth serves as an electrical insulator, allowing a significant charge to accumulate.

If the potential exceeds the insulating capacity of the air, a lightning strike happens. Usually, the threshold is about 3,000,000 Volts/meter. When a strike does occur, it's known as a "negative strike," given that the lightning originated in negatively charged clouds.

This is your everyday lightning variety that accompanies thunderstorms. Rockets typically won't launch with an active lightning storm like this overhead.

How anvils take shape

But the launch scrub of interest occurred with a clear blue sky. Few clouds were in sight, except for a large anvil formation. To understand its threat, let's first figure out what it is.

In the center of a thunderstorm, the warm updraft continues to carry positively charged ice crystals upward, higher and higher. The trend continues until the updraft reaches the tropopause, a layer of air located just above the troposphere. At Texas latitudes, this threshold presents at 35,000 to 40,000 feet. At this point, air temperature stops decreasing with altitude, which causes the tropopause to act as a ceiling.

As a result, a small portion of the cloud will rise vertically above the tropopause, but most of the cloud structure will begin spreading out horizontally. Positively charged clouds spread out expansively and can reach 50 to 100 miles from the original thunderstorm.

Anvils can persist long after their parent storm has weakened, drifting for one to three hours in strong upper winds. Even as the storm below fades, the anvil remains highly charged and capable of producing dangerous lightning.

Anvil lightning strikes

Anvil lightning is a special variety. Recall, the tops of the clouds are positively charged, so this is positive lightning. While much of the ground is induced positive, pockets of negative charge do exist. This lightning finds and strikes them with intensity.

While it only accounts for 5-10% of all strikes, positive lightning tends to be more potent than traditional negative lightning. It exists higher in the atmosphere, which means more charge can accumulate. These strikes are known to travel up to 30 miles away from the cloud and transmit two to ten times more energy than common negative lightning.

This means that not only can an anvil cloud float far away from its parent thunderstorm, it can also strike far away from itself. As a result, if anvil clouds are present, even at a distance, caution is exercised with launch operations.

Some anvil lightning comes downward from the cloud to strike its target. However, other lightning extends out the side of the cloud structure horizontally, before turning down to the Earth. From the ground, this can make it look like the lightning is coming out of a clear blue sky, and these strikes are often referred to as "bolts from the blue."

The anvil rule

Since anvil clouds can strike both far and with fury, they pose a pronounced hazard to launch vehicles. To manage this risk. NASA, the FAA, and launch vehicle operators establish strict rules that dictate when launches can and can't proceed.

These are known as Launch Commit Criteria (LCC), and are constantly evaluated throughout the countdown to determine GO/NO-GO for launch. For anvil clouds, the weather LCCs typically take into account:

• If the anvil cloud is attached or detached from a thunderstorm, and the time since detachment

• Distance to the anvil cloud

• Time since last lightning discharge

The exact distances and thresholds vary by launch operator, location, and mission type. In some cases, more detailed considerations are made for temperature and time-associated distance. However, the overarching goal is to keep the rocket grounded if an anvil could strike it in flight.

And for that reason, sometimes launches scrub under clear blue skies.

Subscribe now

General References

NASA. Falcon 9 Crew Dragon Launch Weather Criteria Fact Sheet. July 2021. https://www.nasa.gov/wp-content/uploads/2021/07/falcon9_crewdragon_launch_weather_criteria_fact_sheet.pdf

NASA. NASA Technical Standard: Lightning Protection Requirements for Space Launch Vehicles, NASA-STD-4010. Baseline Edition. https://standards.nasa.gov/sites/default/files/standards/NASA/Baseline/0/nasa-std-4010.pdf

Federal Aviation Administration. Launch Safety: Lightning Criteria for Expendable Launch Vehicles. Federal Register 76, no. 110 (June 8, 2011). https://www.federalregister.gov/documents/2011/06/08/2011-14146/launch-safety-lightning-criteria-for-expendable-launch-vehicles

National Weather Service. “Lightning Science: How Thunderstorms Electrify and Cause Lightning." NOAA Weather-Ready Nation. https://www.weather.gov/safety/lightning-science-electrification

National Severe Storms Laboratory. “Lightning Frequently Asked Questions." NOAA NSSL Education. https://www.nssl.noaa.gov/education/svrwx101/lightning/faq/

National Oceanic and Atmospheric Administration. “The Positive and Negative Side of Lightning." NOAA JetStream. https://www.noaa.gov/jetstream/lightning/positive-and-negative-side-of-lightning

University of Arizona Department of Atmospheric Sciences. "Cloud Electrification and Lightning." ATMO 589 Lecture Notes, February 16, 2015. http://www.atmo.arizona.edu/students/courselinks/spring15/atmo589/lecture_notes/feb16_2015.html.

ZT Research. "Electrification of Clouds." ZT Research Blog (blog). Accessed August 27, 2025. https://ztresearch.blog/education/electrification-of-clouds/.

"Lightning Strike on Emirates A380 During Flight Caught on Video," The National, May 10, 2016, accessed August 26, 2025, https://www.thenationalnews.com/uae/transport/lightning-strike-on-emirates-a380-during-flight-caught-on-video-1.370276/.

"Lightning Strikes Airplane in Flight," YouTube video, accessed August 26, 2025, https://www.youtube.com/watch?v=-jQVsI7erv8.

It's natural to focus on the technical challenges: the math, models, and mechanical design. They are the core foundation of our solution. Yet, making that solution reliable, delivering it on a relevant timeframe, and at a competitive cost depends on something different: program management.

Running a strong program can be as tricky as the technical solution itself. Many interdependent pieces move simultaneously, and we must operate with incomplete information over long feedback cycles. There’s path dependency, and small changes cascade into big impacts. This all creates fundamental uncertainty.

Trying to eliminate this uncertainty through brute force—throwing people and resources at the problem—works briefly. But as those who've done it know, it's not sustainable or scalable. Instead, we need a framework that thrives on uncertainty rather than fighting it.

If we had infinite time and resources, none of this would be a problem. For decades, the waterfall approach assumed we did. Activities were serialized, and requirements were fully defined before execution started. This worked during the Cold War, when programs were government-led with unlimited budgets. But the world has shifted.

Modern hardware programs face new realities. Many are privately funded. Markets move faster than development cycles, and capital is selective. Even government programs feel these pressures as threats evolve faster than traditional development can counter.

These conditions demand high speed, high quality, low cost, and small team execution. Yet, achieving all four simultaneously is a significant challenge. Combined with the hardware's inherent uncertainty, teams often resort to brute force.

The solution hides in the problem. Rather than fight uncertainty, spiral development makes it the organizing principle. By treating risk as the primary unit of progress—not schedule or features—spiral development transforms hardware's greatest liability into its engine of improvement. Each loop teaches us what matters and what doesn't, replacing guesswork with evidence.

Subscribe now

Barry Boehm introduced spiral development in 1986 for software programs. He asserted that scope cannot be fully defined at the start of a project, and risk must be incrementally reduced through iterative cycles. The approach proved powerful. Teams studied it, creating an abundance of papers and frameworks to implement it successfully.

When applied to hardware programs, the method has had success, but also bumped into limitations.

The first key issue is that the details of spiral development often aren’t fully explained to the team. This is a pitfall I’ve fallen into and the motivation behind this entire essay.

Without practical guidance on how spiral development will be implemented, confusion ensues. Engineers express concern about moving ahead without precise requirements. Managers see uncoordinated iteration. Hazardous spiral look-alikes emerge, such as multiple waterfalls linked in a circle, or the same work repeated endlessly until solutions emerge or teams burn out.

Compounding the issue, hardware faces unique constraints that software doesn't:

• Physical considerations: Complex machined parts might take weeks to cut on the mill, driven by the physical metal removal rate. Bad weather can bring outdoor operations to a standstill.

• Stakeholder expectations: Customers, regulators, and employees often still expect detailed upfront requirements, well-structured design reviews, predictable milestones, and comprehensive documentation.

• One-way doors: Large purchase orders commit a program to a supplier. Architectural decisions might take years to prove flawed, often discovered only during integration or testing.

While real, these issues are surmountable. To do so, we'll start with Boehm's core spiral concept and overlay hardware-specific constraints, milestones, and artifacts. Next, we'll create a common language for hardware teams to speak about spiral development.

The playbook we create will be most applicable for programs that are mission oriented, high performance, cross domain, safety critical and highly regulated—for example, satellites, reactors, ships and similar systems where the stakes justify the rigor. However, elements can be adapted to any program.

The spiral map

Spiral development needs a visual framework, and the spiral map provides it.

Our path starts near the center of the spiral at the Project Kickoff Review (PKR). We move clockwise and outward through four different quadrants:

• Scoping (upper right): Define what we're building and why.

• Actualization (lower right): Transform ideas into physical reality.

• Confidence Building (lower left): Systematically reduce risk through analysis and testing.

• Documentation (upper left): Capture decisions, enable reviews, and facilitate handoffs.

As we move around the spiral, the primary goal of each revolution is to get us to the next one. Each lap has clear goals that progressively increase the program's maturity. Formal reviews serve as checkpoints that define the start and end of each revolution:

• Project Kickoff Review (PKR): Initiates the program, introduces the problem space, organizes the team, and establishes a preliminary plan.

• Conceptual Design Review (CoDR): Reaffirms a refined understanding of the problem, reviews early derisking work, and presents multiple conceptual solutions for evaluation.

• Preliminary Design Review (PDR): Examines detailed engineering, including mechanical and electrical design, system sizing, and analysis. Complex projects often have tiered PDRs to review subsystems. PDRs are the workhorse of spiral development.

• Critical Design Review (CDR): Validates the design in a near-final state after significant iteration, confidence, and independent review.

• Initial Operating Capability Review (IOCR): Certifies the completed system and concludes development, marking the transition to initial operating capability. This often overlooked review appropriately concludes the development phase of the program.

Our illustration shows just one pass through each level, but it’s common to take multiple passes. We may discover issues during testing or modify the design after the PDR. In these instances, we repeat laps around the spiral, conducting a second or third delta-PDR until we're ready to proceed.

Key elements

To illustrate the spiral approach in action, we’ll zoom in on ten key elements. The first half of these focuses on rapidly shaping the project in its early phases, while the second half increases rigor as we approach field deployment.

This is just a select list that serves to illustrate the philosophy. We deliberately stop short of prescribing every detail. Ultimately, each program must adapt the spiral to its specific needs. Taking some elements and leaving others. This essay presents plans for framing a house, but the particular color of the interior wall paint is up to the program runner.

1. Early scope

Programs begin with a simple directive.

We need a new isolation valve for an electric tugboat.

That's enough to kick things off and assign an owner. But the job isn't to immediately start engineering. It's to scope.

Begin by developing a comprehensive understanding of the big picture. Take an outside perspective and question why this hardware matters and how it fits into the world.

• Understand the system: What does this valve do? Why is it needed? What happens if it fails? Map all interfaces, upstream and downstream.

• Validate the prompt: Will a new design actually solve the problem? Are we treating symptoms or the root cause? Sometimes the problem could be solved procedurally instead of with hardware.

• Consider alternatives: Could we modify existing components? Find an off-the-shelf solution? Address it operationally? Even if we proceed as planned, exploring alternatives sharpens our understanding.

This early scoping work establishes the foundation for everything that follows.

2. Requirements and constraints

Understanding the requirements comes next, but flexibility is important at the very beginning of the program.

The valve must open within 50 milliseconds after a command while the ship is underway.

This example illustrates two common missteps. First, preliminary performance requirements often default to theoretical maximums without considering cost, reliability, or actual operational needs. In other words, since we're not sure what we need, we set the requirement aggressively.

Second, the requirement focuses on the primary operation while overlooking other critical states. Hardware often operates, or at least exists, in many different states. The valve might be transported to the ship by train. The shock loads from train cars colliding during coupling could exceed anything seen at sea. These requirements often drive the design as much as the primary ones.

Instead of listing every requirement at this point, map out all the states and environments the hardware will experience. This creates a framework we can fill in as we gain more information, ensuring we design for the complete lifecycle.

Immediately adjacent to requirements are constraints, which can be both hard and soft. Hard constraints come from physical limitations, such as interface geometries, mass budgets, and vibration or radiation environments.

The valve must be less than 6 inches wide to fit on the engine.

Regulatory regimes also impose hard constraints.

The valve must be one fault tolerant to meet the Coast Guard's safety criteria.

Meet with regulators early to develop a clear understanding of the frameworks and to what extent they're tailorable. They often shape how we structure the entire program, and we need early visibility into the processes.

Soft constraints reflect operational realities and organizational preferences. To start, consider the hardware operations. Who builds it and with what tools? Will it be serviced in the field during winter? If so, will the technicians be wearing heavy gloves?

The valve will be serviced by the ship's crew, who have basic mechanical training.

Institutional preferences matter just as much, and often come from difficult lessons learned. The team might favor in-house builds over COTS solutions, or there could be legacy vendor relationships to honor. Scar tissue from previous issues is also always a significant factor.

The team avoids Philips-head screws after a stripped head turned a two-hour maintenance operation into a two-day fiasco.

3. Architecture and supply chain

Architecture and supply chain decisions can be one-way doors. These choices have a high degree of path dependency and can be difficult to change in the future.

The valve actuation architecture can be electronic, pneumatic, or hydraulic.

Each architecture has different benefits and drawbacks. Pneumatic systems need compressed gas infrastructure. Hydraulics requires fluid pumping systems. Electronic architectures rely on reliable power and control circuits. Once chosen, the architecture drives countless downstream decisions, including which suppliers can support it.

There was only one hydraulic valve vendor, but many pneumatic valve vendors. Going pneumatic eliminated sole-source risk.

Look deeper than just primary vendors. What's their supply chain look like? Do they have sufficient access to raw materials? Is their equipment specialized or commodity? What’s their capacity, and how easy is it to increase? Map the entire chain from raw materials to finished parts, asking these questions along the way.

Vendor visits are particularly powerful. Watch the actual processes. Talk with machinists and shop managers. Show them sketches and they'll spot manufacturing challenges immediately. In these discussions, spend time on two key future scenarios. First, what are the ways in which the supply chain can be disrupted? Second, how quickly can production scale, and what are the limits? Expect disruptions and ramp-ups to both present at some point, and have plans for each scenario.

The valve body has a twelve-week lead time but a low unit cost. We'll hold three ship sets in inventory to buffer disruptions and enable quick scaling.

4. Seeing is believing

Hardware teams that don't see the same system don't build the same system. Maintain up-to-date system visuals accessible to everyone in the program. These should convey both how the system looks and how it operates.

Start this practice early. As programs progress, design work concentrates in CAD models that only a few engineers see daily. Without deliberate effort, technicians, program managers, and even other engineers lose sight of the latest and greatest design details.

Consider this canal lock system from 1800. One page of illustrations conveys both the mechanism and its operation. This type of visual becomes the program's North Star—a shared reference that keeps everyone aligned.

5. Street fighting math

Before firing up simulation software, use Street Fighting Math, MIT Professor Sanjoy Mahajan's toolkit for rapid estimation. The linked text and its examples are worth a read, and key elements include:

• Dimensional analysis to infer relationships between quantities and approximate solutions.

• Boundary cases to expose what we expect to happen at the extremes of the operating envelope.

• Successive approximation to make a rough estimate and refine it step by step.

• Reasoning by analogy to relate an unfamiliar problem to something we know well.

• Picture proofs to use diagrams to reason through a complex analysis visually.

Beyond delivering early insights, this method helps us to develop a sense of the problem and its sensitivities. If we jump straight to a detailed simulation, we bypass the critical step of refining our intuition. This can be problematic when larger models fail to perform as expected.

When the valve opens instantly, what does the pressure spike to? When it creeps open over minutes, how does flow establish? Sketch both cases before modeling.

6. Boxy monsters

Boxy monsters are my favorite. These are hardware prototypes designed for rapid manufacturing and iteration. They're usually ugly and heavy, but deliver high impact insights by getting us into test quickly.

They're not mockups. The key parts of the system being validated need precision and full engineering attention. But everything else is optimized for simple manufacturing and ease of future modification. When changes are identified during the test, the boxy monster can be retrofitted on the fly without needing to make a new one from scratch.

Our first valve prototype uses a boxy aluminum housing that any shop can machine in a day. But the internal flow passages and sealing surfaces are production quality.

Even with non-final geometry, these prototypes provide crucial build experience. How do parts fit? What tools do technicians need? Does the test infrastructure work? This knowledge feeds directly into production design and work instructions.

The best boxy monsters are built to be rebuilt. They're not pretty, but they derisk the program faster than any other investment.

7. Independent review

Programs develop blind spots, and deviation becomes normal. Problems hide in plain sight. An Independent Review Team (IRT) breaks this cycle by bringing fresh eyes and hard questions with deep and relevant experience.

The valve IRT: two retired valve designers who've seen every failure mode, one former regulator who knows what matters most, and the Head of Engine Maintenance who runs field repair operations daily.

A strong IRT stays engaged throughout, not just at milestones. They can embed with the team to tackle specific challenges or conduct formal reviews at natural inflection points. Even preparing for their questions raises the bar. Teams find and fix issues rather than defend them.

At ABL, we conducted a multi-day Independent Review Board after our first launch failed. This was a company-wide event, held at an off-site conference center. Ten external experts, including customer representatives, spent three days dissecting decisions and designs.

The process helped us address the specific failure mode, as well as a variety of proximal issues. It surfaced a variety of other improvements and drove transparency into the program. It was a process I wish we'd started years prior.

8. First article build

The first article build marks a critical transition from prototype to operational hardware. This isn't a boxy monster anymore. It's fully engineered hardware that closely resembles production units. Treat this build with production-level rigor, as it becomes the template for all subsequent builds.

• Write comprehensive work instructions before starting, but plan for significant updates during and after the build, as assembly sequences are refined.

• Perform complete acceptance testing on the unit, subjecting it to the same standards that production hardware must pass. This validates the design, as well as our test procedures and infrastructure.

• Document everything with copious notes, photos, and videos throughout the build process. Record assembly sequences, tool requirements, fit-up issues, and technician feedback.

Most importantly, finish the build. When parts don't fit or features are missing, find temporary workarounds. Document the issues, but keep going. It's critical to complete all operations; otherwise, problems hide until the next unit is built.

The valve cap was too short to fit on the body, so we machined temporary spacers to complete the first article build.

9. Development, confidence, and qualification testing

Testing is the ultimate confidence builder, but not all tests are equal. Success requires deliberate choices about what we're testing and why.

First, think about configuration. Some tests characterize specific phenomena on simplified test articles. Others qualify complete hardware in exact flight configuration (test-like-you-fly, in aerospace terms).

To measure valve timing, we built a test rig with just the internal poppet and sensors—no need for the whole assembly.

Next, consider the overall test objective. Some tests validate the overall design (i.e. qualification). Other tests verify that a specific unit is healthy, which is essentially a screen for workmanship or other issues (i.e. acceptance testing).

With our configuration and objective set, we need to consider our approach. I tend to think about this in two ways: spear fishing and casting a net.

• Spear fishing targets known objectives—validating seal integrity, confirming pressure ratings, and measuring actuation speed. We test specific requirements and previously identified risks.

• Casting a net hunts for all possible problems, known and unknown, through broad environmental tests, extended duration runs, and extreme edge cases. This is intended to find the failure modes we didn't anticipate—the unknown unknown mission killers.

Finally, decide whether to validate or destroy. Validation tests to analytical limits while preserving hardware. Destruction tests push beyond predicted failure points to find actual physical limits.

Analysis predicted the valve would burst at 3000 psi. We tested one unit to 2500 psi, and it survived. We pushed another until it failed at 3200 psi, indicating about 200 psi of conservatism in our analysis.

These considerations combine to create three main categories of tests.

Development testing characterizes specific behavior and targets particular risks. It's inherently spear fishing. The goal is to gather specific information that we can use to refine the design and inform decisions on the next spiral cycle. Does the actuation force match our calculations? How does seal performance degrade with temperature? What are the actual frequency response characteristics? Often, the test plan evolves as results reveal where to focus.

Confidence testing casts wide nets using complete systems. The hardware should be in near-final configuration so we're building confidence in something relevant. Push environmental limits, run extended durations, and test multiple units for variation. Destroy some units to find true margins. Keep one life leader unit available for future testing. This pseudo-qualification campaign builds internal confidence before official testing.

Qualification testing is the official run-for-record campaign to meet internal standards and regulatory requirements. Test articles must be in the exact production configuration. Plan for multiple units, again keeping one as a life leader for delta qualification if design changes emerge. If confidence testing was done well, qualification should feel routine. We should be verifying that the design works as intended and documenting this proof, rather than discovering new failure modes.

10. Documentation artifacts

Good documentation is the institutional memory that enables spiral development. Unlike in waterfall approaches, where documentation captures final decisions, spiral artifacts must evolve continuously while maintaining traceability.

Effective documentation falls into four essential categories:

• Work completed: Hardware history logs, test reports, failure analyses, lessons learned

• Technical state-of-affairs: Active risks, open issues, current designs, bills of materials

• Program state-of-affairs: Inventory levels, material planning, unit costs, build status

• Future plans: Goals, schedules, staffing, budgets, resource allocation

Implementation varies by program, but the principle remains constant: documentation should serve the program, not bureaucracy. Each artifact must enable better decisions, faster iteration, or reduced risk. If it doesn't actively support spiral development, eliminate or simplify it.

Select strategies

Along the spiral path, hardware programs encounter predictable challenges. Teams struggle with incomplete information, evolving requirements, and unexpected setbacks. These select strategies help to navigate such recurring situations.

1. Make the best decision we can

Decisions can't wait for perfect information. We'll never know everything—the supplier's actual lead time, the exact vibration loads, or how regulatory frameworks will evolve.

At each decision point, make the best decision we can with the information on hand. Document reasoning—what we knew, what we assumed, and what remained uncertain. When new information emerges, this record enables intelligent updates. The spiral framework expects this. We'll make thousands of decisions as we progress, with opportunities to revisit them on future laps.

2. Eyes inward

Previous work becomes stale without revalidation. On each revolution, look inward to verify earlier work and ensure it remains applicable in light of new knowledge. This isn't about second-guessing previous decisions. It's about systematically assessing earlier work as our information set evolves. This ensures the program doesn't move forward with outdated assumptions.

As an example, hardware programs can take years to execute, during which the world changes. Events may render requirements that were relevant two years ago no longer applicable.

Since the PKR, a new engine configuration was introduced, which changed valve timing requirements and drove program-wide updates.

3. Taking a lap

When problems hit, go forward, not backward. Often, we need to take an extra lap around the spiral. This shouldn't be seen as a failure, but rather a natural part of the program. We might have a qualification test failure, uncover a design flaw, or encounter a supplier issue. When these happen, resist the urge to reverse course. Going backward is disorienting and can create configuration chaos.

Instead, chart a path forward from your current position around the spiral and back to the last major review. Pass through every documentation gate. Update all artifacts, and conduct a repeat review to synchronize the team. This maintains stability while incorporating new learning.

After confidence testing revealed valve seal degradation, we conducted a delta-PDR to update the design with improved materials.

4. Recognizing convergence

The spiral is working when we see convergence and metrics trend in the right direction. Issue tickets close faster than they open. The risk register shrinks at each review. Design changes decrease with each lap. Test failures become rare exceptions.

When these trends sustain across multiple spiral cycles, we're narrowing in on mission-ready hardware.

Applying the framework

This framework is a starting point. Every program must tailor it to its specific context. When elements don't fit, modify or eliminate them. When gaps emerge, fill them with program-specific approaches. The framework itself should spiral, growing with each application.

This approach requires teams embrace uncertainty rather than fight it. Change is inherent, and rework is expected. Refining our understanding and reengineering our solution to match is forward progress, not toil.

These ideas emerged from years of hardware development, watching various approaches succeed and fail. I've been fortunate to learn from steely-eyed operators, meticulous engineers, and skilled technicians. Yet, many others have deeper experience and hold wisdom that this framework doesn't yet capture.

Every application of the spiral helps to mature and refine the approach. Each program brings together people with deep, specialized knowledge that can strengthen the framework. We're still on an early revolution of understanding hardware spirals. The best insights are yet to come.

Subscribe now

General References

1. Civil engineering: diagrams of canal locks. Coloured engraving by J. Pass, 1 March 1800. Wellcome Collection. https://wellcomecollection.org/works/vexamm82/images?id=bj3naadn

2. Barry W. Boehm, "A Spiral Model of Software Development and Enhancement," Computer 21, no. 5 (May 1988): 61-72, accessed July 15, 2025, https://www.cse.msu.edu/~cse435/Homework/HW3/boehm.pdf.

Yet scientists do it. They know the mass, spin, and age of a black hole with remarkable precision. They achieve this not by studying the object, but rather its effects. They point telescopes just outside the black hole's event horizon—the point of no return where objects are permanently drawn in.

They capture superheated gas spiraling into oblivion, stars whipping around at millions of miles per hour, and gravitational waves rippling through spacetime. Every observation on the outside helps to build an understanding of the inside.

Complex problems share this same opacity. The production bottleneck that defies identification. The teams that inexplicably can't collaborate. The component that behaves differently on the bench than on the vehicle.

Staring directly at these problems yields nothing but frustration. Add in a bit of pressure, and the situation can feel overwhelming.

When I hit this point, I've learned to stop staring at the void. Instead, shift focus to the periphery, just like astrophysicists do. A playbook emerges.

1. Find observable effects. What characteristics show up at the boundaries? What might we infer about the inside from these?

2. Shock the system. Change the inputs. Slow the production rate. Replace a component. Shift team composition. Watch how the system responds—or doesn't.

3. Track the patterns. What makes things better? What makes them worse? Map the trends to the stimuli. Be patient. The downstream effects often take time to present.

Step by step, patterns will emerge. It may not be possible to solve the problem fully from the outside. However, we can at least reduce the opacity enough to start seeing what we’re working with. At that point, we have options — to solve the problem or navigate around it.

Very quickly, the situation will feel surmountable. We can, in fact, address problems that we can’t see fully. Just don't get too close to that event horizon in the process.

Subscribe now

Photo: Event Horizon Telescope Collaboration

Further, nuclear technology isn't one thing. Development approaches have diverged dramatically over the last eighty years. Different countries have adopted vastly different strategies driven by technical breakthroughs, resource constraints, sanctions, and public support — or lack thereof. Some countries even halted their nuclear efforts, while others steadily accelerated them.

Today, the industry is undergoing what may be a renaissance. Exploding datacenter demand is driving power needs skyward, while billions are pouring into fusion research. At the same time, fission plants are being exported around the world, creating sophisticated international dependencies and spheres of influence. Yet, the perceived and real dangers of nuclear persist. Chernobyl, Three Mile Island, and Fukushima are household names, and the possibility of weapons-grade uranium recently prompted airstrikes against Iran.

No other technology can both decarbonize entire electrical grids and destroy the cities it serves. It is this fundamental duality that makes nuclear technology uniquely complex to evaluate.

This essay is the first in a series that maps the nuclear landscape across physics, engineering, economics, and geopolitics. Beyond diving into nuclear technology, this series also serves a second purpose. It is a prototype for future As Built landscape series. I deliberately chose nuclear for the maiden voyage. It’s a topic I care deeply about, but to which I have no professional ties, which allows me to write independently.

Our main goal is to characterize the global fission market. To do this, we'll start with the core physics of nuclear reactions and how engineers harness them. We'll then examine the history of nuclear power and the policy decisions that shaped it, finally arriving at today's market. After all that, we'll repeat the process for fusion.

Subscribe now

There are three types of nuclear reactions: fission, fusion, and decay. They're all fundamentally different processes, united by one fact: each transforms atoms into different elements.

Fission and fusion are common terms. Fission reactors power cities worldwide. Fusion research promises unlimited clean energy. Decay is less well known by name, but it's what makes materials radioactive. This essay unpacks how these three reactions work, forming the foundation of the series.

Atomic Structure

Since nuclear reactions transform atoms at their core, we first must understand the structure of an atom.

Atoms are the fundamental building blocks of all matter. At the center is a dense nucleus that packs together positively charged protons and neutral neutrons. Negatively charged electrons orbit the nucleus like satellites around Earth.

Each element on the periodic table is defined by a unique number of protons. Copper has 29 protons, hydrogen has one, and uranium has 92. Conversely, the number of neutrons can vary, and an atom's mass number is the total of its protons and neutrons.

Isotopes are atoms with the same number of protons but a different number of neutrons. Some isotopes are stable, while others are unstable and radioactively decay.

For example, carbon-12 has six protons and six neutrons, and is stable. Carbon-14 has six protons and eight neutrons, and is unstable. It undergoes radioactive decay with a half-life of approximately 5,730 years, making it helpful in determining the age of organic matter.

Chemical vs. Nuclear Reactions

We create more complex materials by combining atoms into molecules. Water (H₂O) consists of two hydrogen atoms and one oxygen atom. Steel is a mixture of iron, carbon, chromium, nickel, and other atoms. To make these materials, or break them apart, we use two types of reactions: chemical and nuclear.

Chemical reactions happen outside the nucleus. Atoms are rearranged by breaking and forming electron bonds. When gasoline burns, Carbon-hydrogen bonds break, and the atoms recombine with oxygen to form CO₂ and H₂O. Importantly, the atoms themselves don't change. Carbon stays carbon. Hydrogen stays hydrogen. Only the electron's configuration changes.

Nuclear reactions happen inside the nucleus. We don't rearrange atoms. Instead, we change the nucleus of the atom itself. That means changing the number or arrangement of protons and neutrons. These reactions include:

• One nucleus splitting into smaller pieces (fission)

• Two nuclei combining into one (fusion)

• A nucleus emitting particles and transforming (decay)

In all three, the atom's identity can change. In other words, it becomes something else. Going back to the carbon example, when carbon-14 decays, it emits an electron, and one neutron becomes a proton. Now with seven protons and seven neutrons, the atom has transformed into nitrogen-14.

As we'll see during fission, a uranium atom also changes its identity, splitting into different smaller atoms and releasing immense energy in the process.

Chemical Reactions

Before we examine nuclear reactions, chemical reactions help to establish a baseline for comparison, and let's consider how they generate energy.

As electrons orbit the nucleus, they do so in different defined energy levels. Known as orbitals, each has a specific amount of potential energy. Generally, electrons closer to the nucleus are more tightly bound, while those in higher orbitals are more available for bonding.

When atoms form a chemical bond, electrons from different atoms interact, often by occupying a shared orbital between the nuclei. This lowers the overall energy of the system and creates a stable bond.

• Forming a bond releases energy, as electrons drop into a lower-energy orbital.

• Breaking a bond requires energy since we're pulling electrons away from a low-energy, shared state.

• Net energy is generated by breaking high-energy bonds and then forming low-energy bonds. The difference in energy is released as heat.

For example, in gasoline molecules, carbon and hydrogen atoms share electrons in stable C-H and C-C bonds, which store chemical energy. During combustion, these bonds break and the atoms form new, lower-energy bonds with oxygen, forming CO₂ and H₂O and releasing the energy difference as heat.

Chemical bonds vary in strength, but most lie in a range of 1–5 electron-volts (eV) per bond. For example, it takes about 4.3 eV to break a hydrocarbon (C-H) bond.

In reactions like combustion, dozens of bonds are broken and reformed per molecule. One kilogram of gasoline roughly contains 44 megajoules of chemical energy. Keep this baseline in mind as we consider nuclear reactions.

Nuclear Reactions and Binding Energy

Nuclear reactions alter the nucleus of the atom, rather than rearranging the electrons. This releases immense energy. Counterintuitively, we can release energy both by splitting heavy nuclei and combining light ones. The concept of binding energy will explain this phenomenon.

Our starting point is Einstein's famous E = mc² equation, which demonstrates that mass and energy are interchangeable. A small amount of mass (m) can be converted into an enormous amount of energy (E) because the speed of light (c) is a large number, and it's squared in the equation.

Applying this practically, when protons and neutrons bind together to form a nucleus, the resulting nucleus has slightly less mass than the sum of its individual parts. This missing mass has been converted to energy. Specifically, it is the binding energy that holds the nucleus together. It’s the energy required to pull a nucleus apart into individual protons and neutrons. It's also the energy released when a nucleus is assembled from those constituents.

For example, when two protons and two neutrons come together to form a helium nucleus, the individual masses add up to 4.0331 atomic mass units (amu), but a helium nucleus has a mass of only 4.0026 amu. The difference, 0.0301 amu, has been converted to 28.3 million electron volts (MeV) of binding energy.

Compare this 28.3 MeV to the 4.3 eV from a hydrocarbon chemical bond. It's 6.5 million times more energy. This comparison isn’t quite apples-to-apples, but it illustrates the different orders of magnitude of energy levels between the chemical and nuclear regimes.

The Binding Energy Curve

Here's where nuclear physics gets interesting. The amount of binding energy isn't the same for all nuclei. Some are bound much more tightly than others. This variation explains why fission and fusion can both release energy.

We usually think about binding energy per nucleon — that is, how much energy is required (or released) per proton or neutron in the nucleus. The more tightly bound the nucleons are, the lower the potential energy and the more stable the nucleus.

Plotting binding energy per nucleon versus mass number reveals a distinct trend. The binding energy per nucleon increases rapidly for light elements, peaks around iron-56 and nickel-62, then gradually decreases for heavier atoms. Elements at the peak are most tightly bound. It takes the most energy to pull them apart. The most stable nuclei sit at the peak, and everything else wants to get there.

• The fusion regime is to the left of the peak. Light nuclei, like hydrogen, are weakly bound. They fuse together to create heavier nuclei, like helium, and move up the curve. The resulting nucleus has a higher binding energy per nucleon, and that difference is released as energy.

• The fission regime is to the right of the peak. Heavy nuclei, like uranium-235, are also relatively loosely bound. When they split into smaller nuclei that are closer to the peak, these products are more tightly bound. The difference in binding energy per nucleon is released as energy.

In both cases, we're moving towards a more tightly bound, stable state. Energy can be released both by building up small nuclei or breaking down big ones, as long as we're trending towards the peak of the curve.

Fission: Splitting the Atom

In fission, a heavy nucleus splits into multiple lighter nuclei, releases several neutrons, and produces a burst of energy. Naturally occurring fission is rare. In most cases, we need to induce it by bombarding a nucleus with a free neutron.

The most common fission process involves uranium-235 (U-235), which is a naturally occurring isotope with 92 protons and 143 neutrons. When a neutron collides with a U-235 atom, it's absorbed, forming uranium-236 (U-236). It's still uranium, but now with 236 total nucleons.

U-236 is unstable and rapidly undergoes fission, usually breaking into two smaller nuclei called fission products, plus additional neutrons, gamma radiation, and heat. The exact fission products vary probabilistically on a fission-by-fission basis. However, the quantity of nucleons and energy are conserved. Products can include barium, krypton, strontium, cesium, iodine, and xenon, among others.

Let's consider one possible example. Uranium-235 absorbs one neutron and becomes Uranium-236. This undergoes fission and splits into Xenon-140 and Strontium-93, while also releasing three neutrons and about 200 MeV of energy.

U-235 + neutron -> U-236 -> Xe-140 + Sr-93 + 3 neutrons + 200 MeV energy

Let's break down what happened here:

• We created fission products. Xenon-140 and strontium-93 are both radioactive isotopes. They're unstable and will decay over time. These fission fragments are commonly referred to as nuclear waste.

• We created three new free neutrons. Uranium-236 starts with 236 nucleons. Our fission products, xenon-140 and strontium-93, total 233 nucleons, leaving three free neutrons. These neutrons are essential. They can go on to cause further fission events, creating a self-sustaining chain reaction, which makes nuclear power possible.

• We released energy. Each fission event releases ~200 MeV of energy. It comes in a variety of forms, with an average breakdown of:

  • 80-85% kinetic energy from fission fragments

  • 2-3% kinetic energy from prompt neutrons

  • 3-4% prompt gamma radiation

  • 5-7% delayed beta/gamma from decay

  • ~5% lost to antineutrinos

Focusing on the kinetic energy, fast-moving fission products and neutrons travel through the solid fuel. They only make it microscopic distances as they collide with surrounding atoms, converting kinetic energy into heat. It is this heat that we subsequently capture and use to produce electricity.

Trillions of fission events happen per second. The sum of these represents the total amount of energy that is stored in uranium fuel. Theoretically, if all of our fuel were fissionable U-235 (which it's not), one kilogram would hold 82,000,000 MJ of energy.

However, pure U-235 fuel is not practical. Real uranium fuel contains mostly non-fissile U-238 isotopes, which significantly changes the energy density. Therefore, while commonly cited, the above value is theoretical and can be misleading when used for comparison purposes. Later, we’ll establish a framework for a more balanced approach to energy density comparisons.

Fusion: Building Bigger Atoms

Fusion is the second type of nuclear reaction. Two light nuclei combine to form a single, heavier nucleus and release energy in the process. Fusion can occur between many different nuclei. However, the most promising fusion reaction for energy generation involves combining two hydrogen isotopes:

• Deuterium (H-2): Hydrogen with one proton and one neutron

• Tritium (H-3): Hydrogen with one proton and two neutrons

Fusion's challenge is overcoming nature's fundamental forces. Recall that protons are positively charged, which means any two atomic nuclei naturally repel each other. Known as the Coulomb barrier, this is one of the primary obstacles to fusion. For two nuclei to get close enough for the strong nuclear force to take over and bind them together, they must collide with tremendous energy.

Creating the Conditions

Achieving such energy requires extreme conditions. Depending on the approach, this might be temperatures that exceed 100 million degrees Fahrenheit or pressures above 100 billion atmospheres. These conditions create plasma, which is a fourth state of matter where nuclei and electrons exist separately. Only in this plasma state can nuclei move fast enough to overcome their mutual repulsion.

These conditions don't exist naturally on Earth, and creating them is challenging. The plasma must be contained and controlled without any material walls, since no known material can withstand prolonged direct contact. This constraint has driven the development of two main approaches:

• Magnetic confinement traps plasma in invisible cages. Powerful magnetic fields, arranged in doughnut-shaped tokamaks or twisted stellarator ribbons, suspend plasma away from any physical walls. The challenge is maintaining this magnetic prison long enough for sustained fusion reactions, keeping the plasma hot and dense while it tries to escape in every direction.

• Inertial confinement compresses fuel faster than it can explode. Arrays of lasers focus enormous power onto fuel pellets smaller than peppercorns, heating and crushing them so rapidly that fusion ignites before the plasma can expand. The challenge is achieving uniformity and symmetry. Small timing or alignment mismatches let the pellet blow apart before fusion can run its course.

Energy Creation and Capture

When deuterium and tritium nuclei collide with sufficient energy, they combine to create Helium-4. In the process, they expel one neutron and release about 17.6 MeV of energy.

H-2 + H-3 -> He-4 + 1 Neutron + 17.6 MeV

This might seem modest compared to the ~200 MeV from fission, but when normalized by nucleon mass, the power of fusion becomes apparent. The fusion reactants are much lighter (hydrogen instead of uranium). This results in high energy release on a per nucleon basis:

• Fusion: ~3.5 MeV per nucleon

• Fission: ~0.85 MeV per nucleon

On this basis, fusion releases about four times more energy than fission.

About 20% of the energy appears as kinetic energy of the helium nucleus. In magnetic confinement systems, this remains trapped by the magnetic fields used to contain the plasma.

The remaining 80% becomes kinetic energy of the fast neutron, which escapes the magnetic confinement entirely. These high-energy neutrons carry most of the fusion energy out of the plasma, where they similarly create heat, which can be captured for electricity production.

When scaled up to practical quantities, deuterium-tritium fuel has a theoretical energy density of roughly 340,000,000 MJ/kg. However, like with fission, inefficiencies and other considerations must be taken into account to compare energy densities.

Breeding Fuel

The neutrons also serve another crucial function. They breed more tritium fuel. Many fusion reactors are lined with lithium-6 blankets. When a neutron collides with a lithium-6 nucleus, it causes it to split into a tritium (H-3) atom and an alpha particle (equivalent to helium-4). Some energy is also released in the process.

Li-6 + neutron -> He-4 + H-3 + 4.8 MeV

This breeding process is essential because tritium doesn't exist naturally in significant quantities.

Deuterium, on the other hand, is naturally occurring, though in small quantities. Approximately 0.015% of seawater is composed of deuterium atoms, which can be concentrated and extracted. The ability to breed tritium in situ and general deuterium access gives fusion a sustainable fuel supply.

Fusion Today and Its Advantages

While fusion occurs naturally in stars, controlling it on Earth has proven extraordinarily difficult. Fusion has been achieved operationally in thermonuclear weapons, where a fission bomb creates the extreme conditions needed to trigger fusion reactions.

Harnessing fusion for power generation remains a development effort. International megaprojects like ITER compete with dozens of private fusion companies racing toward commercial viability. In the past few years, experiments have demonstrated that achieving net energy gain is possible, but scaling fusion to supply the electrical grid remains a work in progress.

Fusion's promise extends beyond just an energy advantage, and includes:

• Waste Stream: Fusion's primary byproduct is helium, an inert gas posing no storage concerns. Neutron activation of reactor components produces some radioactive waste, but this typically decays to safe levels within decades.

• Fuel Supply: Fusion fuels (deuterium and lithium for breeding tritium) are widely available.

• Safety: Fusion cannot experience a runaway reaction or meltdown. The extreme conditions required for fusion are inherently self-limiting. Any disturbance instantly stops the process.

Comparing Energy Density

It's common to compare the energy densities of chemical, fission, and fusion reactions. Using the theoretical values above, fission and fusion are about 1.8 million and 7.7 million times more energy-dense than gasoline, respectively.

However, as mentioned, this comparison is misleading. Uranium isn't pure fissile U-235. Fusion has a variety of losses that may reduce the output. All three processes have inefficiencies when converting from thermal to electrical energy.

To conduct a more practical comparison, we establish the "useful" energy output from each reaction. To do this, we consider real-world constraints of fuel composition and system efficiencies.1

• Gasoline Combustion: We account for a 20% efficiency in converting stored energy to practical work. This captures the thermal and mechanical losses commonly found in an automobile.

• U-235 Fission: We account for an industry standard fuel burnup value of 45 GWd/tHM, which is a measure of energy produced per input fuel mass. The units are GigaWatt-days per ton of heavy metal (U-235, U-238, plutonium, etc.). This metric is helpful as it inherently accounts for enrichment. We also assume 33% efficiency in thermal to electrical energy conversion.

• D-T Fusion: Since commercial fusion has not yet been achieved, we use illustrative assumptions. Our assumed efficiencies are 75% to 85% for radiation and confinement, 80% to 95% for neutron capture. Additionally, we assume 10% to 20% of input fuel is consumed (recognizing that unburned fuel can be recycled through the reactor multiple times). The fusion assumptions should be considered the least mature of the inputs.

Given this approach, we get the following adjusted values for energy density.

On a practical basis, when compared to gasoline,

• Fission is approximately 145,000 times more energy-dense

• Fusion is between 695,000 and 1,870,000 times more energy-dense

While lower than the theoretical numbers, these energy densities remain astounding.

However, perhaps the best comparison is a simple anecdotal example. While naval reactor specifications are classified, the enriched uranium core is estimated to be in the thousands of pounds. This powers an aircraft carrier for twenty-five years between refueling. By comparison, a large container ship burns 3-6 million pounds of fuel oil in just a two-week crossing of the Pacific.

Decay: Nature's Nuclear Transformation

Unlike fission and fusion, nuclear decay commonly occurs on Earth. When a nucleus is unstable (i.e., has the wrong ratio of protons to neutrons, or is too heavy), it will eventually decay through one of three processes. These release particles or radiation and often transform the nucleus into a completely different element.

Nuclear decay doesn't play a significant role in energy production, but it is central to radiation safety, waste management, medical imaging, and overall nuclear stability.

Decay also powers some of our most ambitious space missions through radioisotope thermoelectric generators (RTGs), more commonly referred to as "nuclear batteries." These units directly convert heat from radioactive decay into electricity. For example, the Voyager spacecraft used plutonium decay to transmit data to Earth from billions of miles away.

Alpha Decay (α)

In alpha decay, an unstable nucleus emits an alpha particle, which is made of two protons and two neutrons. Essentially, it is a helium-4 nucleus. This reduces the original atom's mass number by four and its atomic number by two, turning it into a different element. Alpha decay typically occurs in very heavy elements, such as uranium or radium, which are too large to remain stable.

Alpha particles have low penetration power and can be stopped by something as thin as a sheet of paper. However, they pose a significant health risk if inhaled or ingested.

Beta Decay (β)

Beta decay involves a change in the internal structure of a nucleus, where a neutron turns into a proton or vice versa. In beta-minus decay, a neutron converts to a proton, releasing an electron (beta particle) and an antineutrino. In beta-plus decay, which is a rarer process, a proton is converted into a neutron, accompanied by the emission of a positron and a neutrino.

This process changes the element by shifting the atomic number up or down by one. Beta particles have greater penetrating power than alpha particles and require thicker shielding, like plastic or glass, to stop them.

Gamma Decay (γ)

Gamma decay doesn't change the number of protons or neutrons. Instead, the nucleus releases excess energy in the form of light photons. Gamma decay often follows alpha or beta decay, as the nucleus rearranges into a more stable energy state. Because gamma rays are highly penetrating, they require dense shielding materials, such as lead or concrete, to contain them.

Half-Life

The half-life of a radioactive isotope is the time it takes for half of the sample to decay. Some isotopes have a short half-life and rapidly decay over seconds or minutes. Others have a long half-life and might take thousands or millions of years to decay.

A few example half-lives are:

• Hydrogen-3 (Tritium): 12.32 years

• Carbon-14: 5,730 years

• Uranium-238: 4.5 billion years

Decay matters for a wide range of applications. In medicine, certain isotopes are used in cancer therapy and diagnostic imaging. As mentioned earlier, carbon-14 decay is used to estimate the age of ancient organic materials.

In nuclear power generation, the half-lives of fission products define the duration for which nuclear waste remains hazardous. Additionally, heat from decay products continues even after reactor shutdown, which means active cooling is often needed both during and after operation. Managing this is a key factor in establishing passively safe nuclear reactors.

Tie Up

These physical laws have always held true, but our knowledge of them is remarkably new. Henry Becquerel discovered radioactivity from uranium salts in 1896, and Marie Curie famously furthered the research. Ernest Rutherford identified the atomic nucleus in 1911. James Chadwick discovered the neutron in 1932.

Just seven years later, in 1939, Leo Szilard and Albert Einstein wrote a letter to President Roosevelt warning that advances in nuclear fission made it probable this technology could be used for weapons, and that Nazi Germany might be developing such weapons. The Manhattan Project began three years later in 1942.

These three nuclear reactions now shape global power structures. Fission determines which nations can project naval force across oceans and which companies can provide carbon-free baseload power. Fusion could define the energy landscape of the next century. Decay governs nuclear waste policies and plant safety systems.

From here, the nuclear series will focus on fission. We’ll next examine how engineers harness fission to create sustained controlled chain reactions that enable electricity generation.

Subscribe now

General References

Chemistry LibreTexts. "11.6: Bond Energy." Accessed May 23, 2025. https://chem.libretexts.org/Courses/WidenerUniversity/CHEM_175-General_Chemistry_I(Van_Bramer)/11:_Thermochemistry/11.06:_Bond_Energy.

Engineering Toolbox. "Fossil vs. Alternative Fuels - Energy Content." Accessed May 23, 2025. https://www.engineeringtoolbox.com/fossil-fuels-energy-content-d_1298.html.

OpenStax. "10.3: Nuclear Binding Energy." In University Physics III - Optics and Modern Physics (OpenStax). Accessed May 23, 2025. [https://phys.libretexts.org/Bookshelves/University_Physics/University_Physics_(OpenStax)/University_Physics_III_-_Optics_and_Modern_Physics

World Nuclear Association. "Physics of Uranium and Nuclear Energy." Accessed May 23, 2025. https://world-nuclear.org/information-library/nuclear-fuel-cycle/introduction/physics-of-nuclear-energy.

Peter Lux. "Energy Density of Uranium." PLUX. Accessed May 23, 2025. https://www.plux.co.uk/energy-density-of-uranium/.

World Nuclear Association. "Nuclear Fusion Power." Accessed May 23, 2025. https://world-nuclear.org/information-library/current-and-future-generation/nuclear-fusion-power.

Andlinger Center for Energy and the Environment. "Fusion Energy." Princeton University, 2016. https://acee.princeton.edu/wp-content/uploads/2016/05/ACEE-Fusion-Distillate.pdf.

International Atomic Energy Agency. "What is Deuterium?" Last modified January 13, 2023. Accessed May 23, 2025. https://www.iaea.org/newscenter/news/what-is-deuterium.

Richard J. Pearson. "The Availability & Supply of Critical Natural Resources for the Realization of a Fusion Pilot Plant: Fuels for Fusion." Presentation at U.S. DOE Fusion Workshop, Washington, DC, June 1–3, 2022.
https://science.osti.gov/-/media/fes/pdf/fes-presentations/2022/Pearson_resource-availability-and-supply_presentation.pdf.

NASA Science. "Power: Radioisotope Thermoelectric Generators." Accessed May 23, 2025. https://science.nasa.gov/planetary-science/programs/radioisotope-power-systems/power-radioisotope-thermoelectric-generators/.

ChemTalk. "Radioactive Decay | Alpha, Beta, and Gamma." Last modified October 11, 2024. Accessed May 23, 2025. https://chemistrytalk.org/radioactive-decay-alpha-beta-gamma/.

National Institute of Standards and Technology. "Atomic Masses and Fundamental Physical Constants." NIST Physical Measurement Laboratory, 2020. https://physics.nist.gov/cuu/Constants/f

Footnotes

1

Basic assumptions and calculations for “useful” energy density comparison.

NFL training camps mark the beginning of each season. Teams reconvene after months off, bolstered by draft picks and trades. They spend weeks conditioning and practicing. Players learn to communicate and work together as a cohesive unit. Coaches evaluate talent to set the roster.

Under Bill Belichick, the New England Patriots did something different. They stripped names and numbers off jerseys during early camp sessions. No identifiers. Just plain uniforms running drills.

This leveled the playing field.

The moment something is named, it becomes anchored. Reputation and expectation follow. That's the senior guy. The first-round pick. The Alabama product. The one who fumbled in the playoffs.

Our brains crave shortcuts—easy ways to categorize and pattern match. But these shortcuts blind us to current reality. We focus on pedigree instead of performance.

Blank jerseys forced raw evaluation. Every player existed only as today's effort. Not college stats. Not contract value. Not fan expectations. You may have finished last season at the Pro Bowl, but today, you're just another guy in a uniform.

It wasn’t just about the coaches’ evaluation. This approach was also targeted at on-field communication. Offensive and defensive units coalesced around performance, rather than pedigree. The best players revealed themselves.

When names and numbers returned later in camp, the right players were in the right positions. Many were unknowns who emerged as stars—and went on to build the most dominant dynasty in NFL history.

Subscribe now

The line between healthy subcultures and toxic cliques is razor-thin.

Subculture is an integral part of a strong organization. At the beginning, when companies are small, there's typically one culture. This is the foundation as the organization grows. But growth creates specialization. Not everyone has the same job. Some folks work in the field. Some in the shop. Others on corporate teams. Everyone's in the trenches working to make the company successful, but the trenches are all different.

This is where subculture comes in.

Each group both needs to and will naturally develop its own subculture. This stems from shared experiences and common work scope. It coalesces the team. Ideally, there's a culture-carrying leader that spearheads the subculture and unites the group. The team’s ties to this leader often match or exceed their ties to the overall company.

Healthy subculture is prosocial. It creates something bigger than any one person, that all can be a part of. It includes common language, team mottos, and swag. Trust and pride underpin it all. Subculture enables cultural reach that a top-down whole company culture can't achieve without feeling forced.

Teams use inclusive language when discussing challenges. Lots of "we" and "us." Groups actively collaborate across domain boundaries with other teams. Much work gets done. Subculture is good for business.

On the bad side of the line are cliques.

Cliques in organizations are basically the same as in high school. In their most basic form: exclusionary, critical of others, and ripe with gossip. At their worst: disrespectful and immature. They focus more on internal colleagues than external challenges. Cliques draw borders between teams and build identity through shared enemies. They demotivate and destabilize the organization.

The bad news is that once they form, cliques are hard to reverse. After a team flips from subculture to clique, usually the only option is a surgical removal of the entire group. You can try to address it with just the foremost clique leaders, but often the behavior and mindset quickly become deeply rooted. Leaving even the slightest bit can cause it to grow back.

This poses a practical challenge. It's understandably difficult to remove an entire team without losing functionality in the business.

The good news is that cliques aren't usually contagious. If Team A makes the unfortunate swap from subculture to clique, it doesn't necessarily trigger a chain reaction that causes Team B and Team C to follow. Team A might be an acute problem impacting the entire organization, but at least it's isolated.

Maintaining the balance is both straightforward and difficult.

First, don't let fear of clique ever hold back the healthy variant. Encourage subculture. It makes the entire organization stronger if everyone has deep, direct ties to their immediate peers. Find the right culture carriers, give them resources, and watch teams coalesce around them.

The best way to address cliques is to prevent them from forming in the first place. Create opportunities for cross-team interaction. Perspective is an excellent defense against clique. Bring groups together that wouldn't normally cross paths in their usual work.

It’s also essential to ensure the foundational organizational structure is strong and present. This can be difficult with distributed teams and during periods of high tempo. However, the organization’s core culture and values must always be present as a North Star for all subcultures.

At the same time, watch for the warning signs: inside jokes that become weapons, team meetings that devolve into gripe sessions about other departments, or culture carriers who build loyalty through shared negativity rather than shared mission. In summary, lots of "us vs. them."

If they do form, cut out cliques faster than you think you should. If you're inclined to try coaching first, set a firm deadline. In every instance I've dealt with this, I wish I'd moved to the scalpel faster. The loss of functionality we all so deeply fear rarely materializes, and if it does, it's never that bad.

Someone always rises to the occasion, especially when the clique is gone.

Subscribe now