The RD-250 engine at the center of an international storm

In 2017, North Korea stunned the world with a series of test launches of long-range ballistic missiles. One popular explanation for the rogue state's remarkable progress in rocketry essentially blamed Ukraine for providing North Korea with know-how on the powerful RD-250 engine which bore some superficial resemblance to a North-Korean engine. But was it really possible, given the scale of effort required to reproduce and drastically redesign a complex rocket engine?

Bookmark and Share


A cluster of three two-chamber RD-250 (8D518) engines formed a six-chamber RD-251 (8D723) propulsion system of the R-36 rocket. archive

Known specifications of the RD-250 (8D518) engine:

OKB-456 (now NPO Energomash)
R-36 ICBM, Stage I, Main propulsion
Development period
Thrust at sea level
80.4 tons (788 kilonewtons)
Thrust in vacuum
89.9 tons (881 kilonewtons)
Specific impulse at sea level
270 seconds
Specific impulse in vacuum
301 seconds
Duration of burn
120 seconds
Combustion chamber pressure
85 kilograms per square centimeter / 8.83 megapascales
Nozzle extension ratio
Dry mass
728 kilograms
Engine dimensions (height/diameter)
2,600/1,000 millimeters
Unsymmetrical dimethyl hydrazine, UDMH
Nitrogen tetroxide, N2O4


From the publisher: Pace of our development depends primarily on the level of support from our readers!


RD-250: The top in its class

In 1961, the OKB-456 propulsion bureau (now NPO Energomash) led by Valentin Glushko formulated a design of a new engine for the R-36 (8K67) ballistic missile conceived around a year earlier at OKB-586 in Dnepropetrovsk. Like most Soviet military rockets of this era, R-36 would use hypergolic (or storable) liquid propellant, which would allow keeping the missile fueled and ready for launch for prolonged periods of time.

However, for the Soviet rocketry, the development of the new engine, designated RD-250, marked the transition of hypergolic propulsion systems to nitrogen tetroxide as an oxidizer. It is less corrosive than the AK-27I nitric oxide previously used on the R-16, R-14 and R-12 rockets. (Both oxidizers trigger self-ignition on contact with dimethyl hydrazine fuel.)

Along with a more advanced combustion process developed for the RD-250 series, the new oxidizer helped significantly increase the thrust of the new engine. The switch to nitrogen tetroxide, which demands stricter thermal conditions than its predecessor, became possible thanks to an early decision to put all R-36 missiles in underground silos. The previous experience with silo-based missiles proved that their temperatures could be maintained within a range from +5C to +35C degrees. Older Soviet missiles deployed on open pads across the USSR were required to withstand much harsher conditions with temperatures ranging from -50C to +50C degrees. The same factor also allowed to switch the start-up system for RD-250's turbopump from liquid bottles to simpler solid powder charges.

Other improvements in the RD-250 series included the use of expendable membranes not only for storing the engine on the operational rocket, but also as a part of the ignition system. The ignition process was modified to take place with unfilled pumps. The engine shutoff system was simplified to rely exclusively on pyrotechnic valves without the need to vacuum all the propellant lines in flight. The new engine also used more advanced and less corrosion-prone materials. (424)

Inside OKB-456, the work on the engine was delegated to a team of engineers led by Mikhail Gnesin and Yuri Tkachenko, who had previously overseen the development of the propulsion system for the R-16 ICBM. Like the RD-218/219 engines on the R-16, the new RD-250 series developed for the R-36 featured the so-called opened-cycle design, where gas used to power the engine's main turbopump is then exhausted overboard rather than being directed into the combustion chamber. However the new engine, while being 35 centimeters longer and 20 percent heavier in relative mass, nevertheless exceeded its predecessor by 22 seconds in specific impulse — a key performance characteristic.

Like R-16, the first stage of the new rocket would use a cluster of three engines with two combustion chambers each. The combined cluster of three RD-250 modules received the designation RD-251. (113) The two combustion chambers on the RD-250 shared a turbopump and a gas generator welded to the static part of the turbine. It would be activated with a powder-fueled starter firing through three nozzles in the static body of the turbine. The two chambers on each RD-250 also shared avionics and pipelines, however each pair was now made fully autonomous from the other pairs in the three-pair cluster to simplify their production and testing. The only feature still shared by the six-chamber assembly was the propellant drainage system, which would be needed in case of an aborted launch.

The second stage of the R-36 received a similar two-chamber engine called RD-252, which had been optimized for operation beyond the atmosphere in near-vacuum conditions. The RD-252 had a nozzle extension and its combustion chamber pressure was increased from 84 to 91 atmospheres. The engine also had a different truss structure, which attached it to the stage. RD-252 boasted a 25-second advantage in specific impulse over its predecessor — the RD-219 on the second stage of the R-16, despite being 1.5 times taller and having the same relative mass.

Troubled development

The development of the RD-250 series was formally approved by a government decree on April 16, 1962, (424) and work on the R-36 and its engines was put on fast track in the wake of the Cuban missile crisis the same year. In the rushed atmosphere of the arms race, OKB-452 had sent design documentation for the engine to the production plant at OKB-586 in Dnepropetrovsk, Ukraine, before it completed its initial test firing program in Moscow. This allowed to quickly refurbish the production line in Ukraine for the new engine, while it was still in development. (113) The dedicated propulsion division at OKB-586, called KB-4, oversaw the upgrades of the manufacturing line in Dnepropetrovsk, which had previously produced engines for the R-16. (809)

From May 1962 to August 1964, OKB-456 made 145 test firings of RD-252 and RD-250 engines at its test stand in Khimki near Moscow. In the meantime, the production line in Dnepropetrovsk began churning up the serially produced engines, and engineers from KB-4 took random copies from the manufactured batches and fired them at their test stand No. 3 in Dnepropetrovsk. During the first phase of trials, known as "technical and selective control tests" or KTI and KVI, the RD-250 series logged 174 firings.

Finally, 18 R-36 rockets also made test flights. In total, RD-252 and RD-250 engines made 391 firings during this period.

However, around 1964 routine tests of serially produced engines in Dnepropetrovsk began revealing fatal high-frequency vibrations in their combustion chmabers, something unusually frequent for engines burning hypergolic propellants. Mysteriously, the RD-250 first-stage engine mostly displayed this destructive flaw at ignition, while its cousin -- RD-252 -- would self-destruct during the main mode of operation, even though both engines used identical injection nozzles to initiate combustion. (113) The RD-252 also showed that its specific impulse was around 3.5 seconds less than required and its gas generator also displayed instability caused by high frequency vibrations when its propellant had a temperature higher than +20C degrees. Engineers also discovered a mismatch in the dynamics of the ignition between bench testing and real flights. (424)

These massive problems threatened to halt the already running production line. OKB-452 enlisted help from the NII TP and NII KhM research institutes and even from its rivals at Isaev's design bureau in a bid to solve this multidimensional puzzle. (424, 113)

Initially, the high frequency problem was dealt with by increasing the variation in the supply rate of the propellant injectors. Also, the gas generator on RD-252 was equipped with a new acoustic filter at its exist to mitigate its high-frequency problem.

The profile of the RD-252 engine's nozzle was also modified to make up for the lack of specific impulse, while preserving the engine's dimensions.

Also, step by step, the dynamic conditions at Stand No. 3 in Dnepropetrovsk were tuned up to match what had been seen in real launches. (424)

Finally, the ignition profile was changed to promote a more stable combustion process.

All the upgrades cleared the way to the resumption of test firings in September 1964. By that time, the test facility at NPO Energomash was already refurbished for firing the RD-253 engine featuring closed-cycle combustion and intended for the first stage of the UR-500 (Proton) rocket. As a result, all further testing of the RD-250 series moved to Stand No. 3 in Dnepropetrovsk and it was managed by the local KB-4 propulsion division within the OKB-586 design bureau.

To support KB-4, Glushko ordered formation of a special task brigade which moved to Dnepropetrovsk and participated in the planning of test programs, operational management of firings and post-firing analysis of the tests. The group was headed by I. A. Klepikov. Other managers and specialists were shuttling between Moscow and Dnepropetrovsk as needed.

Problems persist

From Sept. 15, 1964, to May 31, 1965, the joint team conducted 220 tests of RD-252 and RD-250 engines. The final certification tests, known as MVI for Inter-agency Verifications Tests, were conducted on three RD-250 and three RD-252 engines. (424)

In April 1966, an interagency commission supervising the re-design effort approved the modified production documentation for the RD-250 engine.

However, even after all was said and done, some engines still continued showing rare high-frequency problems, which prompted officials to retroactively qualify final certification tests as Phase I, while seeking further solution to the problem before Phase II tests. Two possible solutions were tried -- the development of a new injector head and an attempt to develop an ignition mode which would eliminate the high-frequency conditions.

During 1965 and 1966, four variations of injector heads were tried fruitlessly, after which the effort was abandoned. Also, six models of special anti-pulsation filters were tried during tests of individual combustion chambers in NIIKhimmash test center, again, without any positive results.

Finally a solution

All further efforts were focused on changing the ignition process itself. Various ignition modes were first modeled with the help of the Ural-2 computer, which was used to calculate 62 simulation models of the process. That work finally helped narrow down the most effective method of changing the ignition parameters. Extensive test statistics allowed engineers to detect minuscule variations in pressure inside the combustion chambers at ignition. It gave rise to a hypothesis about different conditions in the formation of the ignition pressure in the two engines. Thus, engineers focused on an effort to negate these differences and bring the ignition process in RD-250 and RD-252 to the same timeline.

For that purpose, the hydraulics lab at NPO Energomash built a water simulator of the engine's propellant feeding system and used photo-cameras capturing 1,000 frames per second to visualize the early moments of the fuel-injection process. These tests showed some differences in the character of the propellant supply for RD-250 and RD-252. (424)

The analysis showed that propellant filled the injectors of the combustion chambers differently, because it traveled different distances in the cooling loops of the regular nozzle for the RD-250 and in the extended nozzle on the RD-252. As a result, the fuel, which serves as cooling fluid, heated up more in the RD-252. (113)

Using the water simulator, engineers began adjusting the process to synchronize the ignition in the two engines. It was shown that a lack of synchronization of just 0.02 seconds could make a difference between the success and failure of the firing. (424)

Engineers then developed special thresholds in RD-250 to slow down the ignition process. (113) The method was first implemented on the oxidizer line. (424)

In the meantime, the internal nozzle section of the combustion chamber for the RD-252 received thermal shielding made of zirconium oxide, which reduced propellant heating by 14 degrees. The effectiveness of this measure was then confirmed in 59 firing tests. (113, 424)

In February 1966, the RD-250 entered a new series of tests with the ignition process modified on the oxidizer line. Out of 219 tests, four engines still showed high-frequency vibrations. Then, the similar changes were implemented on the fuel lines, followed by 15 tests of experimental RD-250 engines. The water simulator was used again with special monitoring of the fuel injection. Another 22 KTI and KVI firing tests had followed, this time, with both fuel and oxidizer lines modified. (424) These efforts were finally successful and in July 1967, the R-36 ICBM was formally accepted into armaments. (113)

By Sept. 1, 1967, RD-250 logged 372 tests, including 33 firings in 11 flights of the 8K67 and 8K69 missiles. None of them showed problems. On the decision of the interagency commission, 12 firings tests were conducted to simulate the most adverse conditions at launch, but the engines went through them with flying colors.

The official Phase 2 certification tests of the RD-252 and RD-250 engines were conducted at the end of 1967. The interagency commission could finally declare that the engines matched the required specifications.

As of March 15, 1968, the RD-250 series had accumulated the following test statistics:

Total number of tests...
...Including KTI, KVI tests
...Launches on rockets
310 (in nearly 80 flights)


Space-grade RD-261 engine

In 1965, the Soviet government approved the development of the space launch vehicle based on the R-36 missile. Specifically for that project, OKB-456 design bureau upgraded the RD-250 series under the names RD-261 (11D69) and RD-262 (11D26). The main goal of the program was adapting the engine for wider operational temperatures, because launch vehicles were expected to lift off from open launch pads rather than climate-controlled silos.

Also, when the R-36 rockets reached retirement in the mid-1970s, OKB-452, then renamed KB Energomash, developed a plan for refurbishing their RD-250 and RD-252 engines into RD-261 and RD-262, respectively, so they could be used in space launches. (113)

Post-Soviet history


An RD-261 engine during the assembly of Tsyklon rockets at the Yuzhmash factory in Dnepropetrovsk in 2010. Credit: Roskosmos

According to the Ukrainian government, the production of the RD-250 series and its derivatives stopped in Dnepropetrovsk in 1991 and three years later, the engine's production line was dismantled. As a result, all production activities related to RD-250 stopped in Yuzhmash in 1994. By that time, 30 already manufactured first-stage engines and 10 second-stage engines had been used to assemble a total of 10 Tsyklon-3 rockets, which were delivered to Russia between 1992 and 2008. The final Tsyklon-3 rocket flew from Plesetsk in 2009, delivering the Koronas-Foton satellite.

In the meantime, Ukraine embarked on a commercial space venture which envisioned the light-weight Tsyklon-4 rocket based at the equatorial launch site of Alcantara in Brazil. The proposed launch vehicle was expected to use the first and second stages of the older Tsyklon rockets and a newly developed third stage.

However, KB Yuzhnoe faced insurmountable problems trying to restore the production of the RD-250 series, which were needed for the first and second stage of Tsyklon-4. Unable to restore the production of RD-250, the management at KB Yuzhnoe hoped buy back old Tsyklon vehicles stored in Russian arsenals. Due to lack of payloads for these rockets in Russia, the unused vehicles could theoretically be purchased almost at a price of scrap metal. However, when Ukrainian officials had finally got around to acquiring the rockets, big politics intervened.

The initial Ukrainian offer to buy back Tsyklons coincided with overtures to the West made by the Ukrainian president Viktor Yanukovych at the beginning of his reign in 2010. In March 2012, Yanukovych approved an association agreement with the European Union which angered the Kremlin. As a result, Moscow refused to supply Tsyklons. The Ukrainian space officials then ordered an urgent audit of the remaining hardware at the Yuzhmash production plant to see if any remaining inventory could be used for the Tsyklon-4 project. However, the review failed to identify even a single flight-worthy combustion chamber from the RD-250 series.

Only after very difficult negotiations and the re-alignment of Yanukovych to Moscow, was KB Yuzhnoe finally able to secure the buyback of three Russian Tsyklon-2 rockets, which had been manufactured from 1983 to 1986. In November 2013, when Ukraine had postponed its moves toward Europe, Moscow gave the green light to the Tsyklon deal. Other Russian-Ukrainian space agreements were also negotiated during this period, including Ukraine's participation in the Russian super-heavy launcher program along the lines proposed within the Sodruzhestvo project.

It is still unclear how much Ukraine paid for these three Tsyklon vehicles, but according to unconfirmed rumors, the originally allocated sum had mysteriously disappeared and the funds had to be disbursed for a second time.

The three Tsyklons were delivered to Dnepropetrovsk just in time before the popular uprising in Kiev at the beginning of 2014 toppled Yanukovych's government and put Ukraine back on a path toward European integration. The subsequent Russian annexation of Crimea burned the last economic bridges between the two former Soviet republics.

In the meantime, the Ukrainian-Brazilian venture collapsed within the following year. The total price tag for the fruitless Tsyklon-4 project was estimated at more than $900 million, with as much as $400 million spent by Ukraine. However, KB Yuzhnoe hoped to re-use much of the engineering experience gained in the failed enterprise in the newly proposed Tsyklon-4M rocket.

The North-Korean issue


A launch of North Korea's Hwasong-14 missile.

On Aug. 14, 2017, the International Institute for Strategic Studies, IISS, published a paper called “The secret to North Korea’s ICBM success.” The study suggested that North Korea had relied on a heavily modified version of the RD-250 engine. After the IISS report had made headlines in the New York Times and many other general media publications, KB Yuzhnoe issued a series of official denials alleging political motives behind the publication.

The editor of this site reached to two veteran experts in rocket propulsion in Ukraine who had proved to be very reliable in the past. The following info is based on their testimonies.

In a rare agreement with the official stance, the Ukrainian engineers told that a close look at the current capabilities in the field of large rocket engines at KB Yuzhnoe, including the above-described experience with the RD-250, eliminates practically all suspicions about the company's involvement.

The development of a rocket engine normally goes through three major phases: namely, the design on the drawing board, followed by the production and testing of experimental prototypes and concluding with the serial manufacturing of the flight-worthy hardware. None of these phases, when applied to the North Korean engine development could see any major Ukrainian involvement, KB Yuzhnoe veterans said.

Starting with the design of the RD-250, experts say that KB Yuzhnoe does have a heavily guarded hard copy of production documentation on the engine, but it had never been digitized and its transfer to North Korea would not be authorized by the Ukrainian Space Agency, which reviews all the company’s dealings with outside customers. While admitting a number of commercial contracts with various countries, the Ukrainian engineers said that in their positions at the bureau they would have certainly known about any official contacts of their organization or its employees with the North Koreans, while there were none. Nobody at the bureau ever traveled to North Korea either, they said. Even with the production documentation at hand, the Ukrainians constantly needed the assistance from NPO Energomash in Moscow on various aspects of manufacturing.

However, the most serious charge in the IISS publication claims that a one-chamber version of the RD-250 had been produced in Russia or Ukraine:

“One has to conclude that the modified engines were made in those factories…In addition, Western experts who visited KB Yuzhnoe (in) Ukraine within the past year told the author that a single-chamber version was on display at a nearby university and that a local engineer boasted about producing it.”

However, as described above, the Ukrainian space agency spent more than a decade trying to obtain the original RD-250 engine, badly needed for the nation’s rocket program. After spending a decade and almost $400 million, the Ukrainians were still unable to simply reproduce the Russian RD-250 with two combustion chambers, let alone develop and build a brand-new new, heavily modified one-chamber version, which now appears on the North Korean ICBM.

If any such hardware would reach production stage, numerous people would know about it and, certainly, such a major advance would not be possible to conceal, one veteran propulsion engineer told As was described above, a thorough audit of the remaining hardware at the Yuzhmash production plant had failed to reveal even a single flight-worthy combustion chamber for the RD-250.

Cynics could argue that rogue workers at the Yuzhmash production plant could have sold the hardware illegally before that audit, but all the sources familiar with the matter and the author’s own observations over many years following the Ukrainian space industry and during a recent visit to Ukraine indicate that popular stories often cultivated by the Russian press about the collapse of the Ukranian space industry are greatly exaggerated. Obviously, espionage or theft could not be completely ruled out in any country, but the espionage alone would not likely produce such a complex system for North Korea and it is beyond the scope of allegations in the IISS study.

The failure to reproduce the RD-250 was one of the major reasons that Ukrainian engineers conceived a drastically new Tsyklon-4M rocket around 2016, which would avoid the use of Russian engines.

The newly proposed Tsyklon-4M aimed to fly from a new spaceport of Canso in Canada, would rely on a newly developed Ukrainian engine — the RD-870 — burning non-toxic cryogenic oxidizer (which also makes the engine largely useless for military purposes). The RD-870 will employ all the key components of the Soviet-era RD-120 engine, which was also mass-produced in Ukraine, for the exception of its combustion chamber made in Russia. The newly proposed Ukrainian version of the engine would use an available cache of 50 combustion chambers still retained at the KB Yuzhnoe design bureau. These combustion chambers were intended for Soviet-era ballistic missiles, but their production line was dismantled after the end of the Cold War. According to Ukrainian specialists, these combustion chambers were designed for the so-called closed-cycle engines and they would not be compatible with North Korean rockets.

Even with available combustion chambers, Ukrainian engineers are still facing an uphill battle in manufacturing the RD-870, especially after a complete breakdown of cooperation with Moscow during the Crimean crisis in 2014. Without a supply of key construction materials from Russia, the RD-870 program has huge challenges to transition from paper to metal and the problem has remained unresolved until now.

This current situation with the RD-870 highlights the improbability of a claim in the IISS report that North Koreans could steal or buy newly produced engines and then somehow fashion a super-complex system into what is essentially an entirely new engine.

Finally, Ukrainian experts reject the IISS contention that just because similar engines are not known to be produced in China, India, Iran or France, they must have come from Russia or Ukraine. “No such engines exist in Ukraine or Russia either,” a Ukrainian propulsion expert said.

Ironically, the latest allegations reportedly prompted KB Yuzhnoe to make an initial estimate of what would it take to build the engine seen on the North Korean rocket.

Although Ukrainians admit a superficial resemblance of some components on the North Korean propulsion system to those on RD-250, they see a much simpler explanation. “North Koreans could simply be inspired by the same photos of RD-250 (found in the IISS report),” one expert said. A full-scale copy of an RD-250-derived engine is also available at the limited-access demo room at NPO Energomash in Moscow and was seen by the author of this article (see photo above).

As of possible scenarios for the origin of the North Korean engine, Ukrainian experts suggest an indigenous effort, but point at China as the most likely source of assistance in propulsion know-how, with Russia being a distant second possibility. Unlike Ukraine, the former two countries have at least some clear political motivation to help North Korea advance its missile program as a tool against the United States.


Known specifications of the RD-250 family:

RD-250 (8D518)
RD-251 (8D723)
RD-252 (8D724)
Number of combustion chambers
Two-chamber single engine
Three-engine cluster
Two-chamber single engine
Thrust at sea level
80.4 tons
241 tons
Thrust in vacuum
89.9 tons
270 tons
96 tons
Specific impulse at sea level
270 seconds
270 seconds
Specific impulse in vacuum
301 seconds
301 seconds
317.6 seconds
Combustion chamber pressure
85 kilograms per square centimeter
85 kilograms per square centimeter
91 kilograms per square centimeter
Dry mass
728 kilograms
1,729 kilograms
715 kilograms
Fueled mass
1,980 kilograms
810 kilograms
Engine height
2,600 millimeters
1,760 millimeters
2,190 millimeters
Engine diameter
1,000 millimeters
2,520 millimeters
2,590 millimeters


The RD-250 family variants:

Development period
RD-250 (8D518)
Three RD-250 comprise RD-251
R-36 ICBM, Stage I, main propulsion
RD-250P (8D518P)
Three RD-250P comprise RD-251P
R-36 ICBM, Stage I, main propulsion
RD-250M (8D518M)
Three RD-250M comprise RD-251M
R-36 ICBM, Stage I, main propulsion
RD-250PM (8D518PM)
Three RD-250PM comprise RD-261
R-36 ICBM, Stage I, main propulsion
RD-251 (8D723)
Three-engine cluster
Stage I (R-36/8K67; Tsyklon)
RD-251P (8D723P)
Three-engine cluster
Stage I (R-36P/8K68
RD-251M (8D723M)
Three-engine cluster
Stage I (R-360/8K69
RD-252 (8D724)
Two-chamber engine
Stage II (R-36, R-36P, R-36O, Tsyklon
RD-261 (11D69)
Modified RD-251
Stage I for Tsyklon-2 and Tsyklon-3
RD-262 (11D26)
Modified RD-252
Stage II for Tsyklon-3 and Tsyklon-3


Next chapter: RD-870


Read (and see) much more about Angara rockets and many other space projects in Russia
in a richly illustrated, large-format glossy edition:


Bookmark and Share


Page author: Anatoly Zak; Last update: March 3, 2020

Page editor: Alain Chabot; Last edit: August 31, 2017

All rights reserved




Mikhail Gnesin led a department responsible for the development of the RD-250 series at the OKB-456 design bureau. Credit: NPO Energomash


Yuri Tkachenko worked as deputy department head responsible for the development of the RD-250 series at the OKB-456 design bureau. Credit: NPO Energomash


The RD-219 (8D713) engine originally used on the second stage of the R-16 ICBM served as a basis for RD-252. Copyright © 2017 Anatoly Zak



RD-251 engine. Credit: NPO Energomash


The RD-251 (8D723) engine cluster on the R-36 missile. Copyright © 2017 Anatoly Zak


A cutaway view of the combustion chamber for RD-250 engine. archive


The T 270-000 turbopump, which drove the RD-251 (8D723) engine. Click to enlarge. Copyright © 2017 Anatoly Zak


The RD-252 (8D724) engine, which powered the second stage of the R-36 missile, deferred from its first-stage version by a longer nozzle optimized for operation in vacuum. Copyright © 2002 Anatoly Zak


A combustion chamber of the RD-252 engine. Copyright © 2017 Anatoly Zak


The dual Aerojet General LR-87 engine used on the first stage of the American Titan-2 rocket. The rocket second stage used a single LR-91 engine. Click to enlarge. Copyright © 2017 Anatoly Zak


Click to enlarge. Credit: Roskosmos


An RD-261 engine during the assembly of Tsyklon rockets at the Yuzhmash factory in Dnepropetrovsk in 2010. Click to enlarge. Credit: Roskosmos