India Built Its Own Cryogenic Rocket Engine After Russia Froze a Sale in 1993, and It Took Until 5 January 2014 to Fly Successfully

India’s first home-built cryogenic engine flew on 15 April 2010 and could not keep burning for more than one second after ignition, according to the government’s own reply in the Lok Sabha. The same stage worked on 5 January 2014, and the reason ISRO had to build it at all goes back to a Russian contract that Moscow froze in July 1993.

A cryogenic stage burns liquid hydrogen and liquid oxygen, which stay liquid only at minus 253 and minus 183 degrees Celsius. ISRO’s brochure for the 2014 flight says the pumps feeding such an engine turn at around 40,000 rpm, and that these propellants give far more thrust per kilogram burned than solid or storable liquid fuels. India wanted one for the GSLV, a rocket designed to put communication satellites of about two tonnes into geosynchronous transfer orbit.

A January 1991 contract and an American objection

ISRO and the Soviet space agency Glavkosmos signed an agreement on 18 January 1991, SpaceTech Asia reports in its 2017 history, covering cryogenic engines and the technology to make them. The engines Russia eventually supplied were known as the KVD-1. A UPI report from August 1993 dated the contract to 1990 and valued it at $350 million, so the accounts differ on both the year and the price.

Washington objected on missile grounds. UPI reported that the United States imposed limited trade sanctions on ISRO and Glavkosmos in May 1992. The American concern, under the Missile Technology Control Regime (a 1987 export-control arrangement among Western countries), was that the technology could be diverted to intercontinental ballistic missiles. India and Russia replied that cryogenic engines are unsuited to that use and that no country uses them for it. Neither was a member of the regime, though both had said they would follow its guidelines.

How the sale was frozen in 1993

Boris Yeltsin visited New Delhi in January 1993 and pledged to go ahead, UPI reported. By July the position had reversed. On 16 July 1993 the US State Department announced an agreement with Russia, and the United States Information Agency reported that Russia had agreed to freeze the proposed engine and technology sale to India. Russia would follow the regime’s rules closely from 1 November 1993, and the two countries would cooperate on civilian satellite launches and a space station. Under Secretary of State Lynn Davis, who led the American team, said the United States was not compensating Russia for the lost Indian contract.

UPI tied Russia’s announcement to President Bill Clinton raising the issue with Yeltsin at the G-7 summit in Tokyo. On 15 August 1993, in his Independence Day speech, Prime Minister P. V. Narasimha Rao said India would develop its own cryogenic technology within two years. The first successful flight of that technology came in January 2014, more than twenty years later.

Seven stages ordered, with no technology transfer

Russia did not walk away entirely. In late July 1994, India’s Minister of State for External Affairs, Raghunandan Lal Bhatia, told Parliament that Russia would sell seven cryogenic engines, four at first with an option on three more. “This will be without transfer of technology,” he said, according to The Moscow Times. The paper noted that this contradicted claims by Indian space scientists in September 1993, when one of them said ISRO had even received drawings of the engine.

Those Russian stages carried the GSLV for years. ISRO’s chairman, K. Radhakrishnan, said in October 2010 that ISRO had ordered seven and used five, Spaceflight Now reported from the International Astronautical Congress in Prague. The same report says the Russian stage was blamed for minor orbit-placement problems in 2001 and 2007.

ISRO’s own Cryogenic Upper Stage Project was meant to replace those imported stages. SpaceTech Asia dates ISRO’s decision to build its own engine to 1994, with the Liquid Propulsion Systems Centre leading the work.

The Nambi Narayanan case

In 1994, Kerala Police arrested S. Nambi Narayanan under the Official Secrets Act, Bar and Bench reports. LiveLaw describes him as ISRO’s project director for cryogenic technology and says he was detained for over 50 days. The Central Bureau of Investigation took over, found the case baseless and filed a closure report. In 1998 the Supreme Court quashed a Kerala government notification ordering a further investigation.

On 14 September 2018 a three-judge bench led by Chief Justice Dipak Misra gave its judgment. Bar and Bench quotes it: “The criminal law was set in motion without any basis.” The court ordered the State of Kerala to pay him ₹50 lakh (5 million rupees) within eight weeks, and set up a committee under former Supreme Court judge D. K. Jain to consider steps against the officials involved. The court noted that he could still pursue his civil suit for more compensation.

Two failures in 2010

GSLV-D3 lifted off on 15 April 2010 with the first Indian-built cryogenic stage. The Lok Sabha was told on 9 March 2011 that the engine could not sustain combustion beyond one second after ignition. Spaceflight Now reported the cause as an anomaly in the liquid hydrogen turbopump. The government appointed a Failure Analysis Committee under G. Madhavan Nair, with eleven experts.

GSLV-F06 failed on 25 December 2010, but that flight carried a Russian-supplied stage, not the Indian one. A Lok Sabha answer of 23 February 2011 said the rocket behaved normally until 47.5 seconds, then developed growing orientation errors and broke up at 53.8 seconds. The primary cause was ten connectors at the bottom of the Russian cryogenic stage that snapped early. They had been carrying control commands to the first stage electronics. The launch cost about ₹300 crore (3 billion rupees).

What changed before 5 January 2014

ISRO’s brochure for GSLV-D5 lists the fixes. The fuel booster turbopump was redesigned to allow for the way its bearings and casing expand and contract at cryogenic temperatures, the ignition sequence for the main engine, steering engines and gas generator was modified, and the lower shroud that covers the engine in atmospheric flight and the stage’s wire tunnel were redesigned. The engine went through acceptance tests at ISRO’s main-engine and high-altitude test facilities.

GSLV-D5 lifted off at 16:18 IST on 5 January 2014 after a 29-hour countdown. After 17 minutes 5 seconds it put the 1,982 kg GSAT-14 into its intended transfer orbit, ISRO’s Space India reports. It was the second flight of the indigenous stage. ISRO says the main engine and two steering engines together give a nominal 73.55 kN of thrust in vacuum, and that the stage fires for a nominal 720 seconds.

The bigger engine behind Chandrayaan-3

The CE-20 is a separate engine, built for the larger LVM3, which ISRO first called GSLV Mk III. ISRO gives it a nominal vacuum thrust of 186.36 kN and a specific impulse of 442 seconds (ISRO’s measure of an engine’s efficiency). Its 2014 notes describe it as a new Indian engine running on the gas generator cycle.

The first LVM3 test, on 18 December 2014, carried a passive, non-working cryogenic stage. The first flight with a working one, GSLV Mk III-D1, left Sriharikota at 5:28 pm IST on 5 June 2017 with GSAT-19, which at 3,136 kg was then the heaviest satellite launched from Indian soil. ISRO says the 28-ton stage performed as predicted.

An ISRO release of 7 July 2026 says the CE-20 has worked in eight successive LVM3 missions, including Chandrayaan-2, Chandrayaan-3 and three commercial missions, and that it has met human-rating requirements for Gaganyaan. Chandrayaan-3 left on LVM3 M4 on 14 July 2023 at 14:35 IST, and the landing it enabled is covered in our Chandrayaan-3 article. The wider programme is in the piece on ISRO’s frugal engineering, and the crew rocket plans are in the Gaganyaan article.

What India still lacks

India’s semi-cryogenic engine is still in testing. The SE2000 burns liquid oxygen and kerosene at 2,000 kN of thrust and is meant to power a stage called SC120, which would replace the LVM3’s L110 core stage. ISRO says that with an uprated cryogenic stage this would lift LVM3’s capacity to geosynchronous transfer orbit from 4 tonnes to 5 tonnes. The first hot test of the engine’s power head, which holds every system except the thrust chamber, took place on 28 March 2025 and lasted 2.5 seconds. On 5 September 2026 the ninth test in the series ran at the full 200 tonnes of thrust for 5 seconds, inside a 35-second firing. ISRO’s release says the SC120 is still being developed and that the switch from a start tank to a run tank was validated to allow longer power head tests later.