In 1943, on a frozen lake in the Canadian wilderness, a group of conscientious objectors built a ship out of ice and sawdust—and it didn't sink. This wasn't the experiment of a madman, but part of a project discussed by Churchill and Roosevelt.
German submarines were sinking Allied convoys in the middle of the Atlantic—there, where for 1200 kilometers between Iceland and Newfoundland, the range of land-based fighters ended and dead water began. British pilots called it the Mid-Atlantic gap—an air pocket in the middle of the ocean where transports carrying food and tanks were left without cover. In 1942, the floor of the Atlantic turned into a graveyard: the Germans were sinking ships faster than shipyards could build them.
Churchill knew the numbers. Every sunken tanker meant a week less fuel for Spitfires, every cargo steamer—a month less bread for London. The convoy system worked near the coasts, but in the center of the Atlantic, ships went without cover for five days. Escort carriers were being built too slowly, long-range bombers couldn't yet cover the gap completely, and German "wolf packs" had learned to attack precisely in the blind zone.
At Combined Operations Headquarters—the staff responsible for amphibious operations and unconventional solutions—sat a man who proposed plugging the hole with a piece of ice six football fields long. His name was Geoffrey Pyke, and he was not an engineer.
Pyke invented weapons the way others write detective novels—through paradox. He proposed dropping saboteurs into Norwegian snows on motorized sleds with screw cylinders instead of wheels, covering landings with pykrete monitors armed with turret guns, and building ice ships for the Arctic. Most ideas were unworkable, but Lord Louis Mountbatten—head of Combined Operations—listened.
Mountbatten was a cousin of the king, an admiral, and a fan of unconventional solutions. He was the one who pushed through the idea of commandos—elite sabotage units—and he was the one who gave Pyke access to wartime resources. When Pyke proposed building aircraft carriers from frozen water, Mountbatten didn't laugh. He sent the idea to the scientific department.
There sat Max Perutz—future Nobel laureate in chemistry—and John Bernal, a communist crystallographer who studied ice structure between projects on amphibious operations. They took frozen water, added 14% wood pulp, and got a material that didn't shatter from a bullet, melted noticeably slower than ordinary ice, and withstood loads like low-grade concrete. The material was christened pykrete—after the inventor who couldn't write down the formula for strength but felt it intuitively.
Ice floats. Ice doesn't burn. Ice can be frozen in any bay where there's cold and fresh water. And if you add cellulose to it—it stops being brittle. By the logic of 1942, this looked not like madness but engineering elegance: take a free material, strengthen it with a cheap additive, and build what the steel industry doesn't have the capacity for. The project received the code name Habakkuk—after the Old Testament prophet who predicted miracles.
Churchill showed the idea to Roosevelt at the Quebec conference in the summer of 1943. Mountbatten brought two blocks into the conference hall: one—pure ice, the second—pykrete. He pulled out a pistol and shot at the first. The ice shattered into fragments. Shot at the second—the bullet stuck, and the block remained intact. And then—according to a version confirmed by Alanbrooke's diary—the ricochet from the pykrete block whistled between the generals, grazed Admiral Ernest King's trouser leg, and lodged in the wall.
The Americans were impressed, but not enough to allocate resources. Their shipyards were churning out escort carriers from steel—50 units in less than two years under the CVE program—and it worked predictably. Pykrete was impressive but unproven at scale. Churchill insisted: they needed at least one floating base that would close the Atlantic gap before the end of the war. The Canadians agreed to test it.
On Lake Patricia in Alberta province in the winter of 1943, they built a prototype: a box 60 by 30 feet, weighing 1000 tons. Walls—pykrete a meter thick, inside—a refrigeration system with a motor of one horsepower. It was built by conscientious objectors—pacifists whom the War Department couldn't send to the front but could force to work on construction. They sawed ice, poured forms, and waited for the water to freeze in the hard frost.
The prototype stood all summer. The refrigerator worked, the ice in the walls didn't melt, the structure held its shape. This proved that the idea of a frozen ship was viable—but didn't prove that a hull 600 meters long could be built from pykrete.
The final version of the project looked like this: 2000 feet long, 300 feet wide, displacement 2.2 million tons. For comparison: a modern Nimitz-class aircraft carrier—100,000 tons. Habakkuk was supposed to be twenty-two times heavier.
Hull—pykrete 12 meters thick, reinforced with steel pipes carrying refrigerant. Inside—a labyrinth of refrigeration chambers, each with its own compressor, to maintain a temperature of minus 16 degrees Celsius. Without cooling, pykrete began to "flow"—not melt, but deform under its own weight, like a glacier. Engineers called this cold flow, and it meant the ship needed not just to be frozen but continuously kept in a freezer even in tropical waters.
On top of the pykrete—a wooden deck, on top of the deck—a steel frame for the flight deck. Armament: 40 twin gun mounts of 4.5-inch caliber, placed in turrets around the perimeter. Air group: 150 aircraft—fighters, torpedo bombers, reconnaissance planes. Propulsion: 26 electric motors, mounted externally on brackets like oars on an ancient Roman galley, because a traditional propeller shaft would have to pass through the pykrete hull, creating a thermal bridge—a zone where heat from outside would penetrate inside and begin destroying the material.
This was not an "ice ship" but a hybrid of a refrigeration unit, wooden platform, and steel aerodrome, wrapped in a shell of frozen cellulose. Every element depended on the others, and if even one compressor failed—a chain reaction would begin: the pykrete would heat up, deform, crush adjacent compartments, break pipes, and refrigerant would leak into the ocean.
Construction would require 300,000 tons of wood pulp—Canadians would have to compete for cellulose with newspaper printing presses—and the steel for the refrigeration plant alone would exceed what went into the entire hull of an ordinary steel carrier. Refrigeration capacity in 1943 was barely enough for preserving meat for the army. Even if they managed to build one Habakkuk, it couldn't be towed to the Atlantic—speed of 7 knots meant a couple of weeks' journey from Canada to the middle of the ocean. Which meant it had to be built on site—in the open sea, in winter, at temperatures below zero.
The Admiralty calculated and refused. Not because the project was technically impossible—with sufficient resources it could be realized. But by mid-1943, the Atlantic gap was beginning to close without pykrete.
The Americans launched a conveyor belt of escort carriers: Casablanca-class ships were built in three months—from keel laying to launch. They were small, slow, with minimal armor, but there were many of them. By year's end, dozens of CVEs were patrolling the Atlantic, and each carried 20-30 aircraft that covered a chunk of the dead zone.
The British got the Azores—Portugal agreed to lease them, and the RAF deployed a base for long-range Liberators there. These bombers flew 3500 kilometers and could hang over a convoy for eight hours straight. The Mid-Atlantic gap shrank to a narrow strip and then disappeared completely.
German submarines began dying faster than they could be built. In "Black May" 1943, Dönitz lost 41 submarines—more than in all of 1941, and by December the loss count reached 244 boats versus 87 for all of 1942. New-generation centimetric radars detected conning towers in fog, sonar picked up boats underwater, and "hedgehogs" bombed not astern but ahead of the bow—where the boat hadn't yet turned away. The Battle of the Atlantic turned from hunting convoys into hunting the hunters.
Habakkuk was closed in December 1943 without official announcement. Documents were classified, the prototype on Lake Patricia was abandoned—it melted for another whole year before it went to the bottom, leaving behind sawdust and rusty refrigeration pipes. Mountbatten left to command the campaign in Southeast Asia, Perutz returned to studying protein structure, Pyke proposed several more ideas—all were rejected.
The project didn't fail—it became unnecessary. Industry and tactics solved the problem faster than engineering could perfect the experiment with frozen cellulose. War is a race between invention and production, and in 1943 production won.
Habakkuk was not madness. It was a rational response to an extreme threat, formalized into a project by people who weren't afraid to look ridiculous. In 1942, when Pyke proposed the idea, the Allies were losing 700,000 tons of merchant shipping per month—faster than they could build. If that pace had continued another year, Britain would have run out of fuel and food before Germany ran out of army.
The project died not from absurdity but from changing context. By summer 1943, the volume of escort carrier construction exceeded losses from submarines, the range of patrol aircraft closed the last blind zones, and radar technology made hunting submarines almost predictable. Habakkuk turned out to be a weapon for a war that was already ending—too expensive, too slow, too dependent on a unique combination of cold, wood, and time.
But on Lake Patricia there still lie the remains of a pykrete block that didn't sink and didn't fall apart. The material worked. It's just that the world around it managed to change faster than the ice managed to melt.