From bad to worse: Unknown Soyuz disaster
Hoping to recover quickly from major technical problems in the maiden mission of Soyuz on Nov. 28, 1966, Soviet engineers rushed to prepare the launch of the second spacecraft remaining from the aborted dual flight. With a much better understanding of technical and organizational challenges, leaders of the Soyuz project decided to send the 7K-OK No. 1 vehicle on a solo mission on December 14, 1966. However this time, the launch attempt brought a fatal disaster...
An upper composite configuration during the Dec. 14, 1966, attempt to launch the Soyuz 7K-OK No. 1 vehicle. Copyright © 2016 Anatoly Zak
Soyuz 7K-OK No. 1 mission at a glance:
Learning the lessons
Shortly after the botched landing of the Kosmos-133 mission, engineers established the root cause of the attitude control failures which had plagued the maiden flight of Soyuz. As it turned out, two similar mistakes had been made during the assembly of attitude control thrusters, DPO, and the backup propulsion system, DKD. Combined, they severely crippled the maneuvering capabilities of the 7K-OK No. 2 vehicle.
The problem with the DPO system was traced back to the re-positioning of the attitude control thrusters requested around March 1966, when the first pair of the flight-worthy Soyuz ships had already been undergoing assembly. Around that time, thermal dynamics engineers concluded that exhaust from some of the DPO thrusters could damage the ship's solar panels. To avoid any major re-configuration of the DPO system, leading architect of Soyuz Konstantin Feoktistov decided to simply turn the engine on its attachment fixture in the opposite direction. Because this change reversed the engine's thrust, it was necessary to make a corresponding switch in the flight control unit responsible for roll maneuvers. Feoktistov duly wrote a request for the change to the head of the flight control system development Viktor Legostaev, who, in turn, sent it down the chain of command to the corresponding department head, Shmyglevsky. His team evaluated the issue and recommended the head of the research lab Nevzorov to provide specifications to designers to change the avionics unit for the activation of the attitude control thrusters, BVDPO. Only around May, did the design department led by Chizhikov issue specifications to the production shop for the changes in the avionics unit, which had already been installed on the spacecraft.
The removal of the system required the permission of the Soyuz leading engineer Aleksei Topol. Dismayed by the "slow" process (by the Moon Race standards), Topol demanded his deputy Yuri Semenov to investigate why the change request spent so much time traveling through the system. As Boris Chertok honestly admitted years later, the Head of the lab Nevzorov, (who was near the "bottom of the food chain"), was picked to be the scapegoat and got an official reprimand. (466)
In the meantime, engineers set about replacing BVDPOs on already completed vehicles. However, according to one of the assembly technicians interviewed after the loss of the first Soyuz, that ship was not configured properly, even though, according to Chertok, the required upgrades were found to be implemented as requested on Vehicles No. 3 and No. 4, which were expected to carry cosmonauts.
Immediately, officials sent instructions to Site 31 in Tyuratam to carefully re-check the installation of the attitude control system on the second Soyuz (Vehicle No. 1) before processing it for launch. (466)
In his notes, Mishin cites his deputy Boris Dorofeev as reporting that Vehicle No. 1 had actually rolled out to the launch pad before its joint rendezvous mission with Vehicle No. 2 was cancelled sometimes in the early hours of Nov. 29, 1966. Now, processing teams in Tyuratam had to roll the spacecraft back to the assembly building, detach it from the rocket, pull it from its protective fairing and rotate it into vertical position, so that engineers could inspect the spacecraft and re-process it for another launch attempt. There was apparently some debate what to do with dangerous hydrogen peroxide and hydrazine fuel in the ship's tanks. Mishin deputy for testing Arkady Ostashev, whose brother had died in the Nedelin disaster in 1960, proposed to drain peroxide on the launch pad, but another deputy Boris Dorofeev and the head of the launch team Anatoly Kirillov argued for keeping the propellant onboard to save time. (774)
Planning for future launches
Despite the objections from the Igla rendezvous system developers, Mishin came out from the first Soyuz flight experience convinced that the docking of the two Soyuz spacecraft could be achieved much easier with a crew onboard. Therefore, he concluded that another unmanned dual test flight would be unnecessary. Not surprisingly, the head of cosmonaut training Nikolai Kamanin also strongly supported that opinion.
As a result, on November 30, just two days after the failed landing of Vehicle No. 2, Mishin's deputy Arkady Ostashov delivered a plan for a solo flight of Vehicle No. 1. The second unmanned Soyuz was to test all the onboard systems and give ground control an opportunity to resolve numerous organizational issues uncovered in the first Soyuz mission. The success of the second solo flight would then clear the way for a rendezvous mission between Vehicles No. 4 and No. 3 with crews onboard launching as early as Jan. 29, 1967. (742)
In the first week of December 1966, Mishin already committed to flying the second Soyuz by December 18 of the same year, barely three weeks after the first.
A key meeting of the State Commission on the manned program convened on the evening of December 8. In their presentations, the top developers of the Soyuz spacecraft were unanimous that technical problems during the maiden flight had been stemming from assembly and testing errors and required no major design changes. Specialists admitted that the testing of the Emergency Escape System, SAS, for the Soyuz spacecraft was not expected to be completed until Jan. 10, 1967, and the Commission demanded to meet a new deadline of January 5. Still, the Commission saw no obstacles for the unmanned flight of Vehicle No. 1, which was penciled for a period from Dec. 15 to Dec. 18, 1966. (742)
On December 9, 1966, during a meeting of the OKB-1 leadership, the head of the production plant at OKB-1 Vakhtang Vachnadze admitted mistakes during the assembly of Vehicles No. 1, No. 2 and No. 4. Mishin listed three people at Hall 43 of the production plant and three others at Hall 44, where Soyuz undergoes its final assembly, as those responsible for the error.
In addition, systemic problems were apparently found in the assembly and the quality control process. Mishin quoted Chertok's report at the December 9 meeting as saying that there was no documentation that could properly trace the installation of the DO and DPO thrusters. Mishin also wrote that the upgrade request for the attachment bracket had failed to assign a new serial number to this piece of hardware, which apparently contributed to the confusion during the assembly. Moreover, the design of the device allowed to install the thrusters both ways. Quite unusually, the top managers also shared the blame for not bothering to make a visual inspection of the new thruster arrangement themselves. Finally, no "pumping" tests of the propulsion system had been conducted (using safe pressurized gas instead of fuel), even though it had used to be a routine procedure during the assembly of the Vostok spacecraft and other spacecraft at OKB-1. Had it been done, the wrong installation of the thrusters would have been easily detected.
On the morning of December 10, 1966, Mishin opened a meeting of the Chief Designer Council, which reviewed the updated flight manifest for the Soyuz 7K-OK ships and the 7K-L1 lunar flyby vehicles, aimed to sustain the competition in the Moon Race with the US. The urgency of the situation was underscored that very morning, when top military officials at the Ministry of Defense watched a movie compiled by the Soviet military intelligence, GRU, (probably from publicly available reels), documenting the grandiose NASA effort in the Apollo program, including the development of the Saturn-5 moon rocket.
At the meeting, Feoktistov announced that the Soyuz 7K-OK No. 1 vehicle was ready for launch and the liftoff was even advanced forward one day to Dec. 14, 1966.
On December 12, Mishin and his top associates departed Moscow for Tyuratam to witness the second attempt to launch the Soyuz. (774)
On the morning of December 14, 1966, in Tyuratam, the top managers gathered for the meeting of the State Commission, which oversaw the upcoming launch of Vehicle 7K-OK No. 1.
Head of launch personnel Anatoly Kirillov reported that all upgrades to the spacecraft had been completed and that all the hardware was ready for launch. Mishin's deputy Igor Yurasov reported that this time, the correct operation of the DKD propulsion system had been tested with the use of cold gas. The spacecraft was also suspended on a wire to test the proper operation of the flight control and attitude control systems.
Head of OKB MEI Aleksei Bogomolov also reported that special extensions had been added to the nozzles of the DKD engines to reduce their after-effect on the stability of the spacecraft. (774)
The Commission's chairman Kerim Kerimov insisted that immediately after the Soyuz reached orbit, all mission control functions had to be transferred to the NIP-16 ground station in Crimea. The move aimed to avoid major organizational problems uncovered during the first Soyuz flight. Because the 7K-OK No. 2 had run into trouble very early in its mission, the de-facto control over the flight had ended up in the hands of an improvised crisis team. The group included key officials, who watched the launch in Tyuratam and never had a chance to make it to the Crimean control site, which was much better equipped for handling the situation. As a result, instructions coming from the two locations duplicated and conflicted with each other, creating a situation bordering on chaos, which could have easily affected the outcome of the flight.
This time, the core team of experts and officials, including Agadzhanov, Chertok, Ryazansky, Tregub, Feoktistov, Raushenbakh, Levin, Pavlov, Anokhin, was instructed to be in Crimea at least 24 hours before launch. Most of its members departed Moscow for a military airfield in Saki, (nearest to NIP-16) on the morning of Dec. 13, 1966. Before that, General Agadzhanov and Tregub organized joint training sessions aimed to hammer out coordination problems between civilian spacecraft engineers and the military personnel, which operated the Soviet network of ground stations. The ballistics expert team led by Vladimir Yastrebov also prepared to shift all operational calculations during the mission from the military center at NII-4 in Bolshevo (outside Moscow) to NIP-16 in Crimea. That move promised to avoid long waits for the transmission of large volumes of calculation data through the bottlenecks of inadequate long-distance communications channels. As a result, ground controllers could have much earlier predictions for time windows, when the spacecraft flew within range of ground control. (466)
The meeting of the State Commission was concluded with a formal "go" to the fueling of the launch vehicle. (774)
Upon reaching space, the Soyuz 7K-OK No. 1 spacecraft was expected to conduct an orbit correction on the second day of the flight, during the 18th orbit of the mission. The commands for the maneuver would be uplinked to the ship's flight control system during the 16th orbit, based on measurements of its actual trajectory conducted during the flight over the USSR during the 13th and 14th orbits. The orbit correction would bring the vehicle to an imaginary rendezvous region.
During the 50th, 51st and 61st orbits, mission control would conduct more measurements of the ship's orbit, which would be used during the 62nd orbit to upload settings for the braking maneuver. The spacecraft would then deorbit during the 65th orbit of the mission, after a more than four-day flight and land within the Turkestan Military District, in the Southern USSR.
The liftoff of the 7K-OK No. 1 vehicle, the second in the Soyuz program, was scheduled for Dec. 14, 1966, at 14:00 Moscow Time, from Site 31 in Tyuratam (16:00 local time). At the time of the planned liftoff, it was a clear winter day at the launch site, with the sun hanging low over the horizon.
As the rocket was brought to a 15-minute readiness for launch, key officials, including the veteran head of the launch team Anatoly Kirillov, the Chairman of the State Commission Kerim Kerimov, Vasily Mishin and his associates Shabarov, Dorofeev and the head of OKB-1's Kuibyshev branch Dmitry Kozlov went down the stairs of a protective bunker located just 300 meters from the launch complex. Kirillov took charge of launch operations, watching the rocket through a periscope.
The final preparations proceeded normally and all ground stations received an announcement about a one-minute readiness for liftoff. The launch officials then issued the traditional "Key to Launch" command, initiating the fully automated launch sequence. At ignition, the familiar flash of light and billowing smoke appeared from the flame trench below the rocket, but suddenly the flames died down and the rocket remained motionless on the pad.
At the launch control bunker, specialists saw that right at the time of the ignition, the diagnostics system had detected a problem with the Block B or G strap-on booster of the rocket and aborted the liftoff. The fire-supression system was activated and extinguished the flames of the core-stage engine after it had been urgently cut off. Onboard systems of the rocket were also reset and the vehicle was powered off.
Propulsion system specialists estimated that the rocket could need minor refurbishment, which could take between two and three days. (774) The rocket seemed in stable condition and Kirillov ordered his officers to return to the launch pad and secure the vapor-enveloped vehicle. Kirillov, along with Mishin, Kerimov and their entourage also emerged from the bunker and headed to the pad, as the day was turning to dusk in Tyuratam. Mishin remembers coming close enough to the rocket to actually touch the frost-covered vehicle. (774)
To access the engines of the first stage, the lower service platform rolled out from its protective niche below the launch pad under the rocket, so the personnel could check ignition sticks and their cables. (466)
After the initial inspection, the personnel began preparations for defueling. On the main level of the launch pad, the giant arms of the service gantry began to rise from horizontal into vertical position to give technicians access to the rest of the vehicle. According to Kamanin, the decision to raise the gantry was also dictated by the need to protect the rocket from the unusually high wind. (742)
Fortunately, the arms were still moving into position, when, suddenly, a bright flash of light and a loud bang shook the facility. For a moment, numerous witnesses watched in amazement, as the Emergency Escape System at the top of the rocket fired and pulled the upper portion of the Soyuz spacecraft high into the air, followed by a parachute descent back to the ground between 300 meters and a kilometer away, according to different accounts. (231)
In the meantime, back at the launch complex, the two halves of the released payload fairing crashed loudly on the concrete and tiny streams of burning liquid run down the sides of the headless rocket. Moments later, flames emerged at the top of the vehicle.
Kirillov rushed to the loud speaker system and yelled: "Everybody off the pad and into bunker immediately! The lower service cabin (personnel) evacuate through the pathway to underground oxygen plant! Water to the launch pad!" (466)
Seeing the impending catastrophe, people now ran for their lives. Many of them had personal memories of the Nedelin disaster six years earlier. As the veteran of the mission Vladimir Syromyatnikov later wrote with his famous black humor: "a few world records were broken that day in running and high jumping." (201)
Witnesses could joke afterwards because, this time, the fire was spreading slowly at first, without doubt helping to prevent a much greater tragedy. Still, after around two minutes, flames spread down the rocket and smaller explosions triggered larger blasts, followed by a total conflagration at Site 31. The largest of five blasts blew off windows in buildings located as far as a kilometer away from the epicenter, as the pad was burning in winter dusk.
Kamanin who was watching the rocket from a building around 700 meters from the pad, remembers sharp shards of glass from broken windows piercing walls in the rooms facing the pad. Miraculously, nobody was injured. (742)
Throughout the night, initial head counts indicated that everybody was OK, but the body of a Major Leonid Korostylev, from Kirillov's launch personnel team (2nd Detachment of Unit No. 33797), was later found in one of the caverns under the launch pad. He was pronounced dead of suffocation. Korostylev had been seen running from the pad with a group of soldiers, but apparently decided it was enough to hide behind the concrete wall at the edge of the facility.
Many other soldiers and officers were reportedly injured, but survived. (774) However two conscripts, who were sent to check underground passages below the pad, got overwhelmed by still lingering oxygen vapors and unable to use their gas masks, succumbed as well the next day. (201)
On the morning of December 15, Mishin convened a meeting at Site 2 in Tyuratam. His deputy Dorofeev and propulsion expert Anatoly Iyudin reported the preliminary information about the accident. The aborted ignition was attributed to a faulty oxygen valve on the first stage. Several investigative groups were formed: one dealt with misfiring of the ignition system, another group had to assess the damage to the launch facility, but the main investigative team led by Igor Yurasov naturally focused on the Emergency Escape System, SAS.
The previous night, as the fire was still raging on the pad, engineers feverishly sifted through their documentation, trying to figure out how the SAS rocket could be inadvertently triggered. Kamanin quoted what was probably an initial and rather improbable hypothesis that one of the arms of the service gantry could tilt the rocket beyond seven degrees from vertical to trigger SAS. (742) Boris Chertok gave credit to Nikolai Khlybov, a gyroscope expert, for nailing down the real culprit.
During the ascent of the Soyuz rocket, the emergency escape process and the automated engine cutoff can be triggered by the core stage sensors, which measure the deviation of the rocket's flight control gyroscopes from the prescribed direction within the so-called Inertial Coordinate System, which is tied to fixed stars in the sky. If the rotor of the gyroscope deviates from the prescribed position in space above a certain angle, those sensors trigger emergency procedures, including the activation of escape rockets.
Following the launch abort on December 14, the flight control system of the Soyuz rocket was deactivated and most of its other systems were switched off, however the gyroscopes continued spinning on inertia, which could keep them going for as long as 40 minutes. All that time, the rotor of the gyro continued "moving" within the Inertial Coordinate System due to the Earth's rotation. (466) Around 27 minutes after the aborted launch, the angle of the rotor "exceeded" the allowable limit within the reference system, sending an activation signal to the Emergency Escape System, SAS. In a tragic coincidence, SAS had remained under power, because it was designed to save the crew not only in flight but also on the pad.
As a little consolation to the Soyuz developers, the Emergency Escape System did demonstrate its reliable operation, though in process, it "saved" an empty capsule at a cost of three lives. "While everybody was grieving, the SAS people were celebrating," Syromyatnikov remarked.
The next obvious question for engineers was why the activation of the SAS had led to a catastrophic fire on the pad? Initially, it was believed that the instrument compartment of the spacecraft, which is left behind during the emergency, had been ignited by the exhaust of the escape rockets. (774) However, it was later established that the 32 pyrotechnic charges, which split the upper section of the Soyuz from the rest of the rocket at blastoff had ignited a highly flammable coolant spilling from severed pipes leading to the instrument compartment. To exacerbate the situation, the pumps of the thermal control system continued sending the coolant down the severed lines, literally adding fuel to the fire. The flames then ignited tanks of the hydrogen peroxide propellant in the instrument compartment of the Soyuz and then spread to the huge propellant tanks of the rocket below.
In hindsight, it turned out that just a few days before, on Dec. 11, 1966, a live test of the emergency escape system in Vladimirovka, near the Kapustin Yar test range, had revealed that very problem. During the liftoff of the escape rocket prototype, the thermal control lines were cut and the coolant spillage occurred, causing a fire on the remaining part of the spacecraft. However, thanks to empty hydrogen peroxide tanks and inoperable thermal control pumps on the prototype Soyuz, the fire could not spread too far during the test. Moreover, the probability of firing escape rockets on the pad was considered very low and the deadline pressure of the upcoming launch left no time for a further analysis of potential problems. As a result, no alarm bells rang in the runup to the second Soyuz launch attempt.
The post-accident analysis showed the need for the deactivation of the SAS by remote control, before the umbilical mast of the launch pad has a chance to reconnect to the rocket after an aborted launch attempt. Other planned upgrades also included an embargo on all flight emergency commands before the confirmation of the liftoff, the addition of special lines for arming and disarming the emergency escape rocket via internal cables of the launcher, bypassing the umbilical mast of the launch pad.
To reduce the fire hazard, the lines of the thermal-control system were retrofitted with self-locking valves and wrapped in fire-proof insulation, while the payload fairing was equipped with fire extinguishers. Beginning with Vehicle 7K-OK No. 8, the flammable coolant on the Soyuz was to be replaced with a fire-safe anti-freeze fluid.
All the upgrades were certified during a live firing of the SAS system at a test bench in Vladimirovka on April 12, 1967. (52)
The State Commission met again on Dec. 16, 1966. It made a decision to urgently prepare the original launch pad at Site 1 for operations, because it was estimated that the devastated Site 31 would be out of commission until at least May 1, 1967. (774)
The head of the test range in Tyuratam General Kurushin promised to get Site 1 back in business in a month, enabling the tentative scheduling of another solo launch of the Soyuz spacecraft for Jan. 15, 1967. The follow-on rendezvous mission of two manned Soyuz spacecraft was now pushed back to March 1967. (466)
On December 17, Mishin discussed repair efforts at Site 31 with Marshall Nikolai Krylov, the Commander of the Strategic Missile Forces, who gave permission to "cannibalize" available launch structures from one of the pads for the R-7 ICBM at the military launch site in Plesetsk.
Mishin finally returned from Tyuratam to Moscow on Dec. 20, 1966. Two days later, at a meeting of the State Commission, Boris Chertok presented results of the sub-commission work on the December 14 accident. His report said that the OKB-1 design bureau had been lacking a list of criteria, which would allow the firing of the emergency rockets remotely by a radio command. As a result, the KRL command-radio channel was out of order at the time of the accident. The bureau also had no specific procedures for protecting emergency rockets from an accidental activation command and had no plan for the development of such a system. Finally, the requirements for the fire safety of the emergency rockets were also absent, Chertok said.
As a result, it was decided not to rely on the activation of the emergency escape rockets via the KRL radio channel even during the next attempt to launch the Soyuz spacecraft (Vehicle 7K-OK No. 3). However all other upgrades in the wake of the December 14 accident had to be completed by January 10 of the next year. (774) In the meantime, the OKB-1 management had to switch its attention to the upcoming lunar fly-by missions of the 7K-L1 spacecraft.
Unknown for years
The Soyuz launch disaster was never officially announced. On the day after the accident, Kamanin watched the film footage of the fire at the launch pad, but it had been recorded from a very remote vantage point and the camera had begun rolling only after the activation of the emergency escape system and the following explosions. (742)
In 1994, the Russian cosmonaut Gennady Strekalov made one of the earliest disclosures about the disaster of Dec. 14, 1966, during a meeting in Houston and US space journalist James Oberg recorded his account. The incident was then described in the official history of RKK Energia, and in the memoirs of Boris Chertok, with further details coming from the diaries of Nikolai Kamanin and Vasily Mishin.
A Soyuz 7K-OK spacecraft with a "passive" docking mechanism undergoes pre-launch processing in Tyuratam. Image courtesy Amy Collins
An original emergency escape system, SAS, is being integrated with a 11A511 (Soyuz) rocket. Image courtesy Amy Collins
The emergency escape system of the original Soyuz spacecraft is ready for a live test.
An original Emergnecy Escape System for the Soyuz 7K-OK spacecraft. Image courtesy Amy Collins
Arkady Ostashev led the testing of the first Soyuz spacecraft before launch. Credit: Roskosmos
Boris Dorofeev oversaw the Soyuz program as Assistant to Chief Designer Vasily Mishin. Credit: Roskosmos
Anatoly Kirillov led the military personnel of the launch complex in Tyuratam during early tests of the Soyuz spacecraft. Credit: Roskosmos
During the second flight of Soyuz, the spacecraft was expected to conduct a solo flight.