The Invisible
Famine Machine

The Haber-Bosch process — the industrial method for fixing atmospheric nitrogen into ammonia, developed by Fritz Haber and Carl Bosch between 1909 and 1913 — is the most consequential chemical process in human history. It is also, per unit of output, one of the most energy-intensive.

The process requires two inputs: atmospheric nitrogen (pulled directly from the air via cryogenic separation) and hydrogen. The hydrogen is the critical and expensive input. It is almost exclusively produced by steam methane reforming (SMR) — a process where natural gas (primarily methane) is reacted with steam at 700–1,000°C over a nickel catalyst, producing hydrogen and carbon dioxide as a byproduct. This is why Haber-Bosch is inseparable from natural gas: not because gas is burned for heat (though it is), but because gas is the hydrogen source. Approximately 75% of the natural gas consumed in ammonia production becomes feedstock hydrogen; the remaining 25% is burned for process heat.

In the synthesis loop, the purified nitrogen and hydrogen gases are compressed to 150–300 times atmospheric pressure (150–300 bar). At this pressure, they are passed over an iron catalyst at temperatures of 400–500°C. Under these extreme conditions, the nitrogen triple bond breaks, and nitrogen and hydrogen recombine as ammonia (NH₃). The conversion per pass is only 15–25%, meaning the gas mixture must be recycled multiple times through the catalyst bed, with ammonia continuously extracted by cooling and condensation.

The plant runs as a sealed, continuous loop, 24 hours a day, 365 days a year. Stopping and restarting is enormously expensive — thermal cycling damages the catalyst and the pressure vessels. These plants are not designed to be switched off.

About half the nitrogen atoms in your body right now passed through an industrial reactor running at 200 times atmospheric pressure. You are, in a meaningful chemical sense, partly a product of a fertilizer plant.

Approximately 55% of global ammonia production is then converted to urea, the world's most widely used solid nitrogen fertilizer. The conversion process reacts ammonia with carbon dioxide (the byproduct of the SMR hydrogen production step) at high pressure, producing urea and water. Urea contains 46% nitrogen by mass — the highest of any solid fertilizer. To understand the dependency: without the standard application of 200 pounds of urea per acre, US corn yields would fall from approximately 180 bushels per acre to roughly 40–70 bushels.

The energy intensity is staggering. The Haber-Bosch process consumes approximately 1–2% of global energy supply and produces approximately 1.3% of global CO₂ emissions. To produce one tonne of ammonia requires 28–35 gigajoules of energy — roughly equivalent to the energy in 800 litres of diesel. With global ammonia production at approximately 185 million tonnes per year, the total energy consumption is immense.

Existing plants run at 95–100% utilisation permanently. Unlike most industries, where spare capacity provides a buffer against demand spikes or supply disruptions, ammonia synthesis plants are baseload infrastructure. They are designed to run continuously at maximum output because the economics only work at scale and the startup costs are prohibitive. There is no global strategic reserve of ammonia or urea equivalent to the Strategic Petroleum Reserve for oil.

The repair and rebuild timelines are equally daunting. A major ammonia plant — the synthesis reactor, the compressor trains, the reformer — represents approximately $800 million to $1.5 billion in capital equipment. The primary pressure vessels and catalyst beds are custom-engineered components with lead times of 18–36 months from specialist manufacturers, primarily in Germany, Italy, and Japan. A catastrophic failure — an explosion, a prolonged fire, a military strike — takes 3–5 years to repair if the structural shell is intact, and 7–10 years to rebuild from scratch. There are fewer than 10 engineering firms globally capable of designing and constructing a world-scale ammonia plant, including KBR, ThyssenKrupp, Haldor Topsoe, Casale, and Johnson Matthey.

There is no strategic reserve of nitrogen fertilizer. There is no spare capacity in the system. And the factories that make it take seven years to build. If the Strait of Hormuz closes for a year, the food shortage doesn't arrive in 30 days — it arrives in 18 months, and it doesn't leave.

The food storage buffer is often misunderstood. The "less than 30 days of food stored" figure requires careful unpacking. It refers to global grain stocks as a proportion of annual consumption, measured by the FAO's stocks-to-use ratio. As of 2023, global cereal stocks represent approximately 30–33% of annual use — meaning roughly 120 days of supply in absolute terms. However, this figure is misleading in two ways: a significant portion is held by China (which maintains strategic reserves and restricts exports), and the figure represents stocks in storage, not distributed food.

The more accurate framing is this: if nitrogen fertilizer supply was disrupted at planting season, the first harvest failure would arrive 9–12 months later, and the second compounding failure 9–12 months after that. By year two, accessible global grain stocks would be effectively exhausted. This is the actual timeline of a fertilizer crisis — not 30 days, but 18–24 months to catastrophic shortage.

The 2021 Sri Lanka case study is the most recent real-world demonstration of what rapid nitrogen fertilizer removal does to an agricultural economy. In April 2021, Sri Lanka's government banned all synthetic fertilizer imports, attempting a forced transition to organic agriculture. Within six months, rice yields had fallen 30–40%, and tea production (the country's primary export) had collapsed. The government reversed the policy in November 2021, but the damage was done. By 2022, Sri Lanka had no foreign currency to purchase the fertilizer or the food it could no longer grow, contributing directly to the economic and political collapse that saw the President's residence stormed.

The Haber-Bosch process is the invisible machine that keeps half the world alive. It is a triumph of industrial chemistry that has allowed the human population to expand far beyond the natural carrying capacity of the planet. But that triumph has created a profound structural vulnerability.

We have built a global food system that is entirely dependent on a single, energy-intensive industrial process, concentrated in the world's most volatile region, with no spare capacity, no strategic reserves, and no viable alternatives at scale. The machine must run continuously, 24 hours a day, 365 days a year, fed by an uninterrupted supply of natural gas. If it stops, the consequences are not measured in economic terms, but in calories.

The famine machine is invisible until it stops working. By then, it is too late to build another one.