KiOR and the Iceberg We Failed to See: Reading the Ocean in FOAK Projects
It’s just past 11:30 p.m. on a cold April night, and the North Atlantic is flat as glass.
On the bridge of the Titanic, the warnings have already come in—ice reports, coordinates, positions. Nothing dramatic. Nothing definitive. Just enough to know that something is out there, somewhere ahead.
Below deck, the engines are still turning hard. The ship is making speed. No one thing is wrong. That’s the problem.
A lookout squints into the dark. There’s no moon. No wind. No waves breaking at the base of ice to give it away. Just a shape where a shape shouldn’t be.
“Iceberg, right ahead.”
What follows isn’t a single failure. It’s a sequence. A delay. A turn that’s almost enough. Bulkheads that almost hold. Lifeboats that almost suffice. A system that, for a moment, is still recoverable. And then it isn’t.
—
The gates are locked now at the KiOR plant in Columbus, Mississippi. The reactors are silent, the pipes cold. But for a time this was one of the most ambitious bets in the bioeconomy—a “magic catalyst,” 67 gallons per ton, a billion-dollar vision backed by serious money, serious people, and a world that desperately wanted it to work. And like the ship that couldn’t turn quite fast enough, it didn’t.
We’ve spent the better part of a decade explaining why. We’ve pointed to the catalyst, the yields, the scale-up, the management. Those are valid vectors to consider. And yet, the critique has never quite felt complete. Because each explanation treated KiOR as if it existed on its own—like a machine that either worked or didn’t. Based on public reporting, industry accounts, and our prior coverage of the KiOR story, a clearer pattern emerges in hindsight.
But KiOR wasn’t just a machine. It was a ship. And it wasn’t sailing in a spreadsheet. It was sailing in an ocean.
The Software Ghost in the Industrial Machine
The tragedy of KiOR began with a fundamental mismatch of cultures. Backed early on by Silicon Valley investors, the project was managed (by others, an experienced management team) with what was an impressive velocity, a “blitzscaling”, that brought to mind the rhythms we know at the Digest from our software developments and daily publishing. The company pushed for fast innovation cycles—why not? Many do — but there was a fatal arrhythmia when applied to the multi-year steel-and-concrete of a biorefinery, using the tools of the time.
Software could iterate in a vacuum, even then. First of a Kind (FOAK) plants struggle with this, even now. Then, inside KiOR, the signals we might recognize now, were there, early. Pilot results didn’t match the story. Internal voices argued for a different path—baseline the system, split the reactions, slow down, learn.
Instead, the project accelerated. The story pointed one way. The data pointed another.
That gap has a shape. Later, we’ll call it Symbolic Compression Divergence (SCD)—the moment where the narrative of a drop-in fuel becomes physically decoupled from the oxygen-laden acidic reality of the output.
The Refinery Wall
The promise was refinery-ready fuel. The reality was a bio-oil with roughly 17% oxygen.
KiOR hit what the industry now calls the Refinery Wall. Partners like Catchlight Energy (Chevron/Weyerhaeuser) and ExxonMobil could not process the product. It wasn’t “drop-in.” It was drop-kick.
Instead of aligning with partners and their experience, as we related in our long narrative look back at the KiOR story, the company continued along its original development path. They spent another $90 million bolting on a hydrotreater and hydrogen plant to make their product usable.
That’s something, but it is not partner alignment. In the end, it strikes us as adding weight to a boat already pointed the wrong way.
The Drain and the “Dr. Doom” Factor
Then there was the slow leak. The plant was designed for 67 gallons per ton. It produced 22 at a critical stage of its deployment.
Not a miss—a different universe. When KiOR slipped from 67 to 22, it didn’t just miss a target. It crossed a threshold—where uncertainty was being generated faster than capital markets could absorb it.
At that point, the plant isn’t producing fuel. It’s producing doubt—and doubt is the one thing capital markets price instantly. This is a Reflexive Risk Regime, where every attempt to fix the physics only deepens the financial hole.
So, KiOR wasn’t just building a plant. It was sitting in a capital and validation sink. This is the moment in the North Atlantic when the bow is already flooding. You’re still afloat. But physics has taken over. And belief, eventually, runs out—especially when critical warnings fail to change the course.
In our KiOR retrospective, we reported from our sources that engineers like Mark Ross—dubbed “Dr. Doom” for accurately calculating the 22-gallon yields—and Agnes Dydak raised concerns that did not ultimately change the project’s direction. When adaptive capacity is constrained, the system begins to degrade from the inside.
The Redwood vs. The Kelp
Then there was the spinning. Internal teams pulling in different directions. Conflicting R&D paths. Call it the Churn.
Looked at one way, KiOR’s fatal mistake was what you might call its Redwood Architecture: a massive, rigid, single-point-of-failure plant in Columbus that required everything to work perfectly at once. In that sense, it was brittle. And, fracking was coming along at the same time, at scale — oil prices were adjusting, and oil majors were looking at their options in a different way. When the environment rotates, a Redwood can’t bend. It snaps.
The tragedy is that a Turbine actually existed for the company to escape its fateful ‘iceberg encounter’. A “Stealth Team”—including Dennis Stamires, Robert Bartek, and Mike Brady—had already designed a two-pot reactor system. It was what we call in our FOAK analyses Kelp Architecture: resilient, modular, able to survive great storms, and designed to solve the KiOR catalyst deactivation problems threatening the company’s future.
If the Churn is high and the system is rigid, you don’t get resistance. You get shear failure.
The FOAK Fragility Index
If you were standing on the Columbus site in the early 2010s, here’s what you would have been measuring—without knowing it:
- Narrative Load
The weight of promises versus physical reality. - Structural Friction
The rigidity of the system versus the need for flexibility. - Entropy Export Delta (EED)
The speed at which bad news becomes better design.
Seen this way, KiOR didn’t fail because of one thing. It failed because the ship and the ocean were out of sync.
TEA can tell you whether the boat should float. TRL can tell you whether it’s been built before. But neither tells you whether you’re launching into a rising tide, a rip current, or a storm. Or an iceberg.
Coda: If We Could Have Read the KiOR Ocean
Our goal today is not to revisit a painful episode for no reason. Let’s ask, what would it have looked like to navigate using a FOAK Fragility Index in real time? KiOR failed gradually—then suddenly. A framework like FOAK Fragility would have changed the signals management acted on.
Signal 1: Pull Misalignment (SCD)
- Indicator: Product requires $90M downstream workarounds (The Refinery Wall)
- Meaning: The system is misaligned with the value chain
- Action: Pause scale-up. Re-baseline product-market fit
Signal 2: Divergence Threshold Breach
- Indicator: Actual yields hit 22 vs. 67 gallons
- Meaning: You’ve entered a Reflexive Risk Regime; uncertainty is outpacing capital
- Action: Trigger a capital discipline gate. Reset external guidance to physical reality
Signal 3: Arrhythmia (Churn Mismatch)
- Indicator: Operational issues evolve in hours; board decisions take months
- Meaning: Software-speed thinking is colliding with industrial-speed physics
- Action: Increase decision frequency. Reduce approval layers
Signal 4: Suppressed Adaptation (Low EED)
- Indicator: The ‘Dr. Doom’ warnings are not acted upon; the two-pot reactor approach is not pursued
- Meaning: The system’s metabolism is failing; information is not becoming improvement
- Action: Establish protected channels for dissent. Fund the alternative path
Fast Forward to Now: Reading Today’s Ocean
Let’s pivot to you. Where the logos are different. The decks are cleaner. The ocean hasn’t changed. Let’s look at some problems that companies face today and how FOAK Fragility can assist.
Scenario 1: Capital Tide Goes Out (The Drain)
What you see: Interest rates spike. DevCap disappears.
Signal: You’re consuming runway to buy time. You’ve entered a Drain.
Move: Don’t row harder. Get buoyant. Shift to a Minimum Viable Plant. Pull forward revenue through tolling or interim products.
Scenario 2: The Catalyst Keeps Dying (Fighting the Churn)
What you see: Catalyst life collapses under real feedstock.
Signal: You are a Redwood in a hurricane.
Move: Stop defending the design. Build a Turbine. Shift to modular, staged systems.
Scenario 3: Policy Slips, Offtake Walks (Pull Misalignment)
What you see: 45Z is delayed. Anchor customer exits.
Signal: The Pull moved. You’re pointed at yesterday’s market.
Move: Re-anchor to today’s demand. Diversify offtake. Adjust product slate.
Scenario 4: Costs Blow Out Post-SPAC (Narrative Load)
What you see: Equipment costs surge. Capital no longer closes.
Signal: Narrative Load exceeds system metabolism.
Move: Right-size. Strip scope. Protect the learning loop.
Bottom line, once you launch. The iceberg sheds water, not tears. The ocean doesn’t care about your model. It only cares whether your structure can stay coherent under pressure. You don’t avoid failure by building a bigger ship—or even a better one.
You avoid it by reading the ocean early enough to turn.
Category: Top Stories
