the home of online investigations

You can support the work of Bellingcat by donating through the following link:

Is this Ukrainian Buk a Clue in the MH17 Investigation or a Red Herring?

January 5, 2015

By Aric Toler

On July 19, two days after the tragic shoot down of Malaysian Airlines flight 17, the Security Service of Ukraine (SBU) publishedan image in a press release that showed two Buk-M1 anti-aircraft systems. One of them was the Buk-M1 that likely shot down MH17, transported on a truck eastward through Luhansk on July 18. The other was a Buk numbered 312 operated by the Ukrainian military that was videotaped on March 19 at the Yasinovatsky post just north of Donetsk, Ukraine. While the SBU did not explicitly claim that these two Buks were one in the same, it is very easy to assume this from the juxtaposition of the two images, and much proverbial ink has been spilled establishing the significance of Ukrainian Buk 312.


Recently, photographs of Ukrainian Buk 312 uploaded on the Russian social media site VKontakte (VK) on August 17, 2014  have surfaced, allowing us to finally settle the significance—or insignificance—of this anti-aircraft system.

A number of Russian social media users and bloggers have recently shared their findings of Buk 312 and established that it was at the Kramatorsk airfield in August 2014. This reaction from the Runet has been intensified due to a recent interview between Ukrainian journalist Anatolii Sharii and a Ukrainian soldier attesting that Ukrainian Buk 312 was redeployed from Luhansk to Kramatorsk. You can read a previous Bellingcat article on the (in)significance of the new information from this interview here.

Two images of Buk 312 were first posted on the VK page of Ukrainian soldier Igor Polulakh. The original images have been deleted, but you can find archived images here, here, and, thanks to Ukraine@War, here.

JJTLj0LGifc 43264_original

The blogger “militarizm” geolocated the image of Buk 312 to government-controlled Kramatorsk airfield, particularly to the coordinates of 48.7045442, 37.6280022. This geolocation is likely correct. The Google Map imagery from August 23 shows a number of helicopters and two heavy military vehicles around these coordinates. While the helicopters in the Google imagery are not present in the images, many other features are, including the similarly colored path (red lines), lines separating the concrete tiles (purple and blue lines), two particular structures (yellow/green circles), two trees (black 1 and 2), and a tree to the left of the camera (orange arrow):


1-5-2015 11-05-25 AM 1-5-2015 11-04-09 AM

As noted by the blogger Ukraine@War, what appears to be a Buk with four white-tipped missiles is located at roughly the same site as the Buk in the VK images. It is impossible to tell if the particular vehicle in the satellite imagery is the same as Buk 312, but considering the position of the vehicle and the four white tips, it is likely a Buk of some kind, if not 312 itself.

Polulakh posted these two images on August 17, 2014 along with a batch of other photographs from various other locations. Therefore, it is difficult to be certain when these photographs were taken without additional metadata or indicators of time. However, the Kramatorsk airfield was under heavy attack throughout the spring of 2014 (especially in April and May) and Polulakh posted a previous batch of images on June 27, 2014, making it most likely that the photo was taken in July or August. Additionally, a series of historical imagery of the Kramatorsk airbase shows us the difference between the base on July 12 (above) and August 23 (below), indicating rapid mobilization at the site.


From this, it is extremely likely that the photographs were indeed taken in late July or early/mid August.

Some have claimed that pro-Russian separatists seized this particular Buk in June in Donetsk, following a report from the Zvezda channel on June 29 and a comment from a Ukrainian official that a “non-operating” Buk was left behind.  Others, including the “militarizm” blog referenced at the beginning of this report, have accused Ukraine of using its Buk 312 of downing MH17.

The latter scenario—laying the blame on Ukraine—is extremely unlikely to be true for a number of reasons that are immediately evident: the 35km range of the Buk while Kramatorsk is located over 80km from the crash site, the angle at which MH17 was struck by fragments of an anti-aircraft missile indicates that the missile came from the southeast, and the improbable logistics of transporting a government-controlled anti-aircraft system through the heart of separatist territory, firing it, then extracting it without even a whisper on social media.

The former scenario—separatists seized Ukrainian Buk 312 and used it to down MH17—can now be considered extremely unlikely. As documented in Bellingcat’s earlier report “Origin of the Separatist Buk,” the “side-skirt damage fingerprint” of Ukrainian Buk 312 has little resemblance to that of the Buk being transported out of Donetsk on the morning of July 17. Additionally, with firm confirmation that Ukrainian Buk 312 was photographed in late July/August 2014 in a government-controlled airfield, we can safely say that pro-Russian separatists never seized, let alone used, this particular anti-aircraft system.

Considering the confirmation of the presence of Ukrainian Buk 312 in government-controlled Kramatorsk in the summer of 2014 and the lack of evidence that this particular Buk was anywhere near Snizhne on July 17, we can safely say that 312 had nothing to do with the downing of MH17 and a different Buk was used.

Aric Toler

Aric Toler started volunteering for Bellingcat in 2014 and has been on staff since 2015. He currently heads up Bellingcat's training efforts and its Eastern Europe/Eurasia research.

Join the Bellingcat Mailing List:

Enter your email address to receive a weekly digest of Bellingcat posts, links to open source research articles, and more.


  1. EM

    You are really stupid. And also a big liar.

    “The 13:20 timestamp is also the timestamp when external air traffic systems lost the plane.” – You’re lying. The plane was visible on the radar after the missile hit. full report
    “The GKOVD radar screen video replay from 13.20:47
    – 13.21:08 and 13.21:18 – 13.25:57 (15.20:47
    – 15.21:08 and 15.21:18 – 15.25:57 CET) showed targets which are considered to be aeroplane debris falling down. ”

    “Though it is strange they have found nothing” – Nothing strange, moron. All was fine until until 13.20: 03 when MH17 was hit by a Buk missile.
    “A missile hit most certainly does not shut down flight recorders, deliberately so. They are customarily placed into tail section BTW”
    The Cockpit Voice Recorder and Flight Data Recorder stopped abruptly at 13.20:03 (15.20:03 CET) because the power supply was interrupted”. The missile hit “shut down” the power supply of the aircraft. Of cource, flight recorders dont work without elektricity. Nothing strange, moron. It is simple.

    “I think the autopilot takes over and does something somewhat meaningful if the pilots suddenly let go of steering.” – You’re too stupid to “think” Everything was fine with the pilots until 13:20:03. The crew was alive, and communicated with the air traffic controller: “The last radio transmission made by flight MH17 began at 13.19:56 (15.19:56 CET) and ended at 13.19:59 (15.19:59 CET).

    “Given that there is only 8.5 km between Rozsypne and Hrabove, the missile hit so late is impossible” – This is possible and this is the confirmed fact.
    “777 cruises at around 250 m/s, covering the distance involved in less than 40 secs” – yep, normally functioning Boeing 777 (or another aircraft) that flies with constant speed and direction, but not a bunch of falling parts.
    “From the ballistic trajectory analysis it can be concluded that all the pieces of wreckage from the fuselage parts in front of STA888/909 departed the aeroplane immediately after the last Flight Data Recorder recording.”

    • muchandr

      I looked for the final report and found it here

      if you don’t mind. They do seem to be saying the same BS you do (See map page 6) plane parts falling all the way around last FDR point 13:20 which is the same external telemetry got, last CVR point shortly before 13:19. That requires the plane disintegration into small parts that lost all forward inertia and fell like a sack of bricks!?! This is not what happened and directly contradicts their own statement of some passenger maintaining consciousness for some 90 secs (in the hull that was merely depressurized) During that time the dead hulk of the airplane would’ve covered twice the distance they show on the map. Yet there is no description on how the plane is supposed to be torn into small pieces immediately. Needless to say, locations significantly to the N of FDR point are not at all possible unless there was an internal explosion, as they are not at all along inertial path.

      Most importantly, flight recorders do work without power! It is a US Federal regulation, so it is at the very true for recent American-built planes. Flight recorders must have enough battery capacity to operate a powerful underwater beacon for at least 30 days

      Thus, the report is certainly doctored, thanks for the hint. They do say about power being severed. I actually found the exact schematics for 777. The recorders are 18 VDC devices, but have two redundant 115 VAC connections to power. You do know that all 115/230 VAC power on planes comes from battery banks also?

      Presumably they are specifically protecting the FDR point actually corresponding to an earlier point en route which makes a shot from a Ukranian Buk possible.

      • EM

        “Flight recorders must have enough battery capacity to operate a powerful underwater beacon for at least 30 days” – yep. to operate a powerful underwater beacon but not to operate recorder.
        “Thus, the report is certainly doctored” – in your sick head.
        “You do know that all 115/230 VAC power on planes comes from battery banks also” – so what? All the same, do you have batteries or not, if the electrical circuit is broken.

        “That requires the plane disintegration into small parts that lost all forward inertia and fell like a sack of bricks!?!” – you lied. nobody said this.
        “the centre and rear parts of the aeroplane were estimated to have taken about 1-1.5 minutes to reach the ground. Other, lighter parts, will have taken longer” – it was said in reality

        “which makes a shot from a Ukranian Buk possible” – you lied.
        “The last radio transmission made by flight MH17 began at 13.19:56 (15.19:56 CET) and ended at 13.19:59 (15.19:59 CET) – it is 4 (four) seconds before 13:20:03 it was near Rozsipne. Ukrainian Buk missiles range is 35 kilometers. Kramatorsk airfield is located at the distance more than 80 kilometers from the Rozsipne, Avdiivka more than 50 km . Ukrainian buk have can’t shot down MH 17 it is absolutelly imposible!

  2. EM

    “So, your scenario involves a missile hitting a plane immediately overhead near instantaneously and plane dropping to the ground also near instantaneously”
    – You lie. Not “my” and not “scenario”, only facts. And not “instantaneously” – “the centre and rear parts of the aeroplane were estimated to have taken about 1-1.5 minutes to reach the ground. Other, lighter parts, will have taken longer…paragraphs 3.4.1 and 3.4.3 it was shown that shortly before the Cockpit Voice Recorder stopped recording at 13.20:03 (15.20:03 CET), a high-frequency sound wave was detected, originating outside the aeroplane from a position above the left hand side of the cockpit propagating from front to aft. Shortly after the Cockpit Voice Recorder and Flight Data Recorder stopped recording the Ukrainian and Russian Federation radar data, SATCOM data and ELT activation data all show that the aeroplane suffered structural 117 of 279 failure and lost electrical power, experienced a deceleration (described in paragraph 3.4.3), and started to break up.”

    “It is a higher source of authority than whatever Dutch Safety Board is”
    – Analysis of flight recorders was made by Branch’s laboratory at Farnborough, United Kingdom, accompanied by an international team of air safety, representatives of ICAO and investigator of the Interstate Aviation Committee, not by Dutch Safety Board. But of course the stupid troll with the nickname “muchandr” can interpret the data of the flight recorders much better. Everyone should believe him. LOL
    “the report is sloppy or deliberately covering up Ukranian ass” – No, comrade. Everything is fine with the report. Thats you, by all means trying to cover Russia’s ass. You spread manipulation and misinformation.

    Mh 17 was shot down by BUK 332 from 53rd Anti Aircraft Missile brigade, Kursk, Russian Federation. This is strongly proved fact.

    • muchandr

      In the UK? The friendliest of all jurisdictions. Let me include this here for you again. A flight recorder is DC-powered through a sizeable battery that is included in the protected box

      You can clearly see the battery compartment on all the pictures of the things too. They are also in the tail section, because it is usually the last part to break off. You need a compelling argument to demonstrate in excess of 3400G the thing can handle without anything unusual happening to the tail. The plane was hit above the cockpit, disabling the pilots and maybe controls, but no critical system of the plane itself.

      The spread of parts within only 10 km suggests low altitude they broke off. The speed of a plane falling uncontrollably is always too fast when it hits the ground. You somehow assume that the stalling plane slows down when it falls. It does not, quite on the contrary. You can use the following calculator to see how fast that plane would fall if indeed it fell to the ground like a bag of brick under the force of gravity

      I tried 245 tonnes for loaded 777, 9 square meter and 0.5 drag coefficient (sphere in turbulent flow) to get 843 m/s, some 3.2 times faster that its aerodynamic cruising speed! This calculation is most sensitive to the last parameter, but even coefficient of 1 for something awfully inconvenient shaped like a brick still gives 576 m/s This makes it blatantly obvious that a stalling airplane of this size hits the sound barrier of 343 m/s rather quickly, which it is unlikely to survive

      If you are interested in doing it properly, you are supposed to use volumetric drag coefficient and a special area of volume to the power 2/3 for airliners and surfaces of revolution like their cylindrical bodies

      These objects are considered highly streamlined and simple terminal velocity calculation for bag of bricks that I gave you actually never applies. I haven’t done this because it will obviously produce an even more stupendous terminal velocity no airliner will survive intact.

      Math is highest of all truths. Physics maybe less so, but still more than some biased bastards in UK who’s got not enough skill to even frame Russia properly. Obviously Russia won’t accept this investigation until “intergovernmental” something includes their representative.

  3. muchandr

    I did not lie, they did. The battery is part of every recorder device per NTSB regulation. It is not so easily disconnected. You can actually see the devises having two parts, one of which is the battery. The 115 VAC feeds only charges the battery through inverters, the recorder is a 18 VDC device in case of 777. This was supposed to be an argument that there is nowhere externally to even get a 18 VDC powerline, it always uses the battery that is simultaneously recharged on the other end. When you realize that this is a fact, you’ll have to explain how the plane decelerated to 111 km/s in downward direction and strictly less than this in forward direction. The actual speed of the plane in its own frame of reference is something called a cross product if you have any interest in linear algebra. If the plane decelerated somehow, it would not have disintegrated due to several small holes in sheet metal and possibly broken windows. When the pilots are alive, they routinely land planes missing cockpit windows. The plane disintegrates because it accelerates to free fall, probably close to critical velocity. It is not possible to find parts of the plane falling from 10 km high right underneath the place it was hit, nor only 8.5 km apart, suggesting even smaller forward velocity than downward. Because it is less than 10 km? This violates laws of conservation of momentum.

    What are they covering up? Obviously the fact that there were whole battalions of Ukrainian Buks in range for that shot along the flight path, but only if the plane took several minutes to descend. Look at this again

    Nobody is disputing the 13:20 FDR timestamp because it corresponds to external telemetry (that ELT thing did it) and 13:19 timestamp of last communication with tower. I am saying that it cannot somehow be simultaneously the timestamp when the plane was hit and when it hit the ground. It is not an instantaneous thing.
    And hence it happened either crash happened much later and there is further flight recorder data or it was never over Rozsvyrne place. Both at the same time are possible, only one not really.

    You repeating the 53rd BS does not make it any truer. Even though no 3×2 was actually sighted in Ukraine, what was sighted is a single M1-series TELAR with no number and everybody seems to agree the separatists had no other. In this case, I don’t think it could have taken that shot and hit anything. Chicken out they did and scram, but so did the other suspects.


Leave a Reply

  • (will not be published)

You can support the work of Bellingcat by donating through the following link: