Naval history
China's Carrier Programme: From a Towed Hull to Three Ships in Service
For most of the 2000s Chinese carrier aviation was a subject of reporting instead of a fact. An unfinished Soviet hull sat in a Chinese shipyard, intentions were denied and then not denied, and Western coverage cycled through rumours of imminent construction for years. What eventually happened is more instructive than any of the speculation, because it demonstrates which part of carrier aviation is actually difficult.
The hull that started it
The Soviet Union began building a carrier of the Kuznetsov class that was left unfinished when the Union dissolved, and the incomplete hull passed to Ukraine. It was sold in the late 1990s to a purchaser who stated it would become a floating entertainment complex, and towed: after a long and well-documented struggle to get transit permission through the Turkish Straits, to China.
It did not become an entertainment complex. It was refitted over roughly a decade and commissioned as Liaoning in 2012, initially and explicitly as a training ship.
That designation was accurate rather than modest. A navy with no carrier experience needs somewhere to learn, and learning on your only operational ship is how you lose it.
What came next
A second carrier, Shandong, was built domestically to a related design and commissioned in 2019: a significant step, because designing and building the ship yourself is a different capability from refitting one.
A third, Fujian, is substantially larger and, importantly, uses catapults and not a ski-jump ramp. That is the change that matters technically. A ski-jump launch limits the weight an aircraft can carry off the deck, which limits fuel and weapons, which limits range and payload. A catapult removes that constraint and permits heavier aircraft, including airborne early warning aircraft, which a ski-jump carrier essentially cannot operate and which are the single biggest gap in a ski-jump air wing’s capability.
Why the hull was never the hard part
Shipbuilding capacity was not China’s constraint; it has enormous shipbuilding capacity. The difficult parts of carrier aviation are the ones that cannot be bought or copied.
Deck operations. Launching and recovering aircraft from a moving deck at a high tempo, safely, is a choreography involving hundreds of people in defined roles. It is learned by doing it, and the learning is measured in years and paid for in accidents.
Aircrew qualification. Carrier landing is the hardest routine task in military aviation. Producing a pipeline of pilots who can do it at night, in weather, on a pitching deck, requires instructors who already can, which is a bootstrapping problem for a navy starting from zero.
Maintenance at sea. Sustaining an air wing on a deployment means depot-level capability afloat, spare parts logistics and a trained maintenance organisation.
Escorts and support. A carrier operates inside a group. Air-defence destroyers, anti-submarine capability and replenishment ships have to exist in sufficient numbers, and building them is a separate multi-decade effort.
Doctrine. What the ship is for (power projection at distance, sea control near home, or presence) determines everything about how it is used, and no other navy’s doctrine transfers unmodified.
What the episode says about the reporting
The years of speculation before Liaoning are a useful case in how defence reporting handles intentions. Satellite imagery of a hull in a yard is a fact. What it will become, when, and what it means are inferences, and they were repeatedly reported with more confidence than the evidence supported (in both directions) by people predicting imminent carriers and by people predicting the programme would never happen.
The accurate reading available at the time was narrower and duller: China had acquired a hull, had the industrial capacity to finish it, and would need a long time to learn to use it. That is roughly what occurred.
Where it stands as a capability
Three ships is a meaningful force and a limited one. Sustaining continuous carrier presence requires enough hulls to cover deployment, work-up and maintenance cycles simultaneously: the same arithmetic that constrains every navy. The catapult ship changes what the air wing can do and not how often the fleet can be somewhere.
The programme’s real significance is that the difficult, unglamorous competencies (deck cycles, aircrew pipelines, at-sea sustainment) are being built, and those are what take twenty years and cannot be shortcut.
Questions and answers
Why does a carrier programme take so long?
Because the ship is the smaller part of it. Hull construction is a decade-long industrial commitment; then an air wing has to be built, pilots trained in carrier landing (which is the hardest routine task in military aviation) deck crews trained, and escorts and replenishment ships made available. A navy has a carrier capability when all of that exists, not when a hull is launched.
What is the difference between a ski-jump and a catapult?
How much an aircraft can weigh at launch. A ramp lets an aircraft take off using its own thrust, which limits the fuel and weapons it can carry. A catapult accelerates it, which permits heavier loads and therefore longer range and more weapons. It is also considerably harder to engineer, and the step between the two is one of the clearest markers of a maturing programme.
Is a single carrier militarily useful?
Only in a limited sense, and this is consistently underweighted. One hull cannot be continuously available: it needs maintenance, its crew needs training cycles, and a fleet of one has no availability at all for part of any year. A navy that intends to keep a carrier deployed needs several, which is why hull counts are watched as a sequence as opposed to as a total.