by Colonel Cassad


Translated by Seva

One of the readers sent a comparative analysis of UA losses in the war in Donbass based on the data of the Military Balance for 2013 and 2016, with the emphasis on the lists of official personnel of the UA (Ukrainian Army) units.

The most authoritative in the world military analytical agency International Institute for Strategic Studies issued a report Military Balance for 2016, which, when compared to the report for 2013, allows to estimate irrecoverable losses of military hardware in the UA during the active phase of ATO at ~50%, taking into account vehicles repaired after Petro Poroshenko had officially announced the loss of ~60%. These reports also allow one to offer an estimate of the losses in the UA personnel during ATO. In vehicles that were damaged beyond repair there were 29,700 solders that likely were at least wounded.

2016 2013 Losses Losses, % Crew Personnel
Total 4,974 9,986 5,012 50% 29,690
Artillery 1,862 3,351 1,489 44% 6,870
Self-propelled artillery 541 1226 685 56% 2,349
122 mm 2C1 “Gvozdika” (Carnation) 247 600 353 59% 4 1,412
152 mm 2C3 “Akatsiya” (Acacia) 235 463 228 49% 4 912
152 mm 2C19 “Msta-C” 35 40 5 13% 5 25
Towed artillery 515 1065 550 52% 3,740
122 mm D-30 75 369 294 80% 6 1,764
152 mm 2A36 “Giacint-B” 180 287 107 37% 8 856
152 mm 2A65 “Msta-B” 130 185 55 30% 8 440
152 mm D-20 130 215 85 40% 8 680
Multiple launch rocket systems (MLRS) 348 554 206 37% 781
MLRS BM-12 “Grad” (Hail) 185 351 166 47% 3 498
MLRS “Uragan” (Hurricane) 70 137 67 49% 4 268
MLRS “Smerch” (Tornado) 75 80 5 6% 3 15
Tanks 788 1110 322 29% 1,170
T-64, including “Bulat” (Dagger) 710 1100 390 35% 3 1,170
T-72 from storage 70
T-80 from storage 8
Armored personnel carriers (APC) 1,319 3,028 1,709 56% 12,580
APC-2 960 1,434 474 33% 10 4,740
APC-1 210 994 784 79% 10 7,840
Armored fighting vehicle (AFV) 490 1,432 942 66% 9,070
AFV-70 235 857 622 73% 10 6,220
AFV-80 110 395 285 72% 10 2,850

How do analysts count military hardware and personnel

Below we present a few comments regarding the methods analysts use to count military hardware and personnel, to give you an idea how reliable these data are.

Naturally, analysts do not use websites with pictures of demolished hardware, such as, because only a fraction of destroyed hardware is photographed and only a fraction of that can be identified.

Professional analysts use a different very efficient method. They take “official staffing” of the armed forces. Because standard armaments of military units are no secret, one can calculate military hardware by multiplying units by the hardware pieces they have. This method somewhat overestimates the amount of serviceable military hardware in the UA, as pieces in repairs or non-functional ones are counted as active units. The advantage of this method is that it takes into account the capacity of the military industry to repair damaged hardware: damaged tank or APC would not be counted as lost if they were repaired.

This method allowed evaluation how many units with standard armament the UA had in 2013 before ATO, and how many units are there in 2016. The losses of different arms vary, but overall are at about 50%. This matches Petro Poroshenko’s statements about the 60% losses and very modest numbers of purchased new arms and repairs of damaged ones. Kharkov tank-building plant stated that the plans for new production and repairs totally failed In fact, in terms of repairs and maintenance of tanks, the Ukrainian military industry can mostly repaint them. The capacity of the Ukrainian military industry for serious repairs of severely damaged vehicles and new production is a few dozens a year, as is clear from the report of Military Balance.

We’d like to note that this method requires taking into account that the numbers of military hardware can be reduced not only by losses in warfare, but also by sales to other countries. However, during this period the UA sold abroad relatively few tanks (see, essentially only 11 tanks T-72 to Nigeria. These sales do not change much, as compared to losses of thousands in the war. Besides, the Ukrainian military mostly sell abroad the equipment not used in ATO.

Estimates of personnel losses and the paradox of mobilizations without demobilization

One must mention that during ATO military analysts ironically noted that the first three waves of mobilization were not matched by demobilization, but the UA did not create any new units, even though more than 30,000 men were drafted. That black humor was caused by the impression that mobilizations just replenished losses of killed and wounded. There is a possibility, though, that some of the mobilized solders were sent to semi-regular units of the National Guard, which could explain this paradox.

However, the data from Military Balance show that the black humor of military analysts was not far from the mark. The UA simply did not have the hardware to create new units. Also, even Anton Geraschenko (advisor to the Minister of Internal affairs of Ukraine Arsen Avakov; member of Rada; gain notoriety with the creation of the site “Mirotvorets” (Peacemaker) containing personal information about various opposition figures and journalists – translator’s note), who oversaw the creation of the National Guard units, said that one should not exaggerate the personnel of the National Guard, which never exceeded 5,000. Because of that, Anton Geraschenko complained about low numbers of “patriots”.

Since the analysts know the exact numbers of military hardware lost by the UA in ATO, as well as the crew of each vehicle, one can calculate the number of personnel in destroyed vehicles – 29,690 solders.

Let me remind you that the analysts only count hardware that cannot be restored, i.e., very seriously damaged. In most cases, this means damage to the body with subsequent fire, so that repairs with spare parts make no sense – the vehicle has burnt out. It is obvious that members of the crew of vehicles with this kind of damage were at least wounded. Thus, the number of killed and wounded UA solders is about 29,690, and roughly corresponds to the numbers mobilized in the first three waves, which just replaced lost solders. This did not leave any for demobilization.

The problem with spare parts for Russian tanks and APCs/AFBs

We also need to comment on tanks T-72 and T-80 produced by Uralvagonzavod (a Russian machine building company located in Nizhny Tagil, Ural Region, Russia. It is the largest tank producer in the world manufacturing, among other things, tank “Armata” – translator’s note). Some of these tanks were in storage since the times of the USSR. But, according to Military Balance, they were not used by the UA until 2015. Moreover, functional T-72 were sold abroad even in 2014. Thus, the losses of T-72 and very few battle-ready T-72 tanks in the UA in 2015 were not so much the result of war losses, as the lack of spare parts from Russian Uralvagonzavod.

The UA had similar problems with Soviet APCs and AFVs. Although the Ukrainian military industry can produce a limited number of APCs and AFVs, Ukraine needed serial production of spare parts for Soviet machines. Spare parts for AFV-70 and AFV-80 are produced by the Gorki car plant. Parts for APC-1 and APC-2 are produced by the Kurgan machine-building plant. The arms for these AFVs and APCs are produced by the Tula machine-building plant. All these Russian suppliers are unavailable for Ukraine, which explains dramatic losses of APCs and AFVs. This might mean that personnel loses were lower – AFVs and APCs were written off after moderate damage. On the other hand, with the losses of AFVs and APCs at 56-66%, even with new production, the UA cannot form large motorized infantry units it had in 2013 and take advantage of its large mobilization resources. This explains reduced mobilization plans, as the UA does not have necessary vehicles. The war on school buses is a thing of the past, especially considering the capacity of modern artillery, such as self-propelled artillery pieces and MRLS, discussed below.

Shadow fighters of anti-battery units

What strikes one in Military Balance data are huge losses of light howitzers D-30 – ~80%. Partially this can be explained by the fact that many howitzers were transferred to semi-regular units of the National Guard that mostly shelled cities and townships. But this does not explain the losses of self-propelled artillery pieces Gvozdika (~60%) and 152 mm Akatsia (~50%). It certainly does not explain the losses of ~50% of MRLS Grad and Uragan. The destruction of self-propelled artillery (SPA) is hard due to its mobility and armor. The destruction of MRLS requires an incredibly short time of reaction, as they leave the positions within a minute. Thus, these losses of self-propelled artillery and MRLS and incredible losses of howitzers D-30 suggest that the UA was confronted by modern artillery, very possibly Russian.

Tymchuk (Ukrainian military commenter – translator’s note) wrote many times about the presence of Artillery Radio-location system (RLS) Zoopark-1 in Donbass ( This RLS can observe missile trajectories and calculate the location of launchers even before the missiles hit the ground. Then Zoopark, using communication channels protected against radio-electronic interference, transmits the coordinates of the launchers or artillery pieces for targeting. However, only a modern SPA or MRLS using global positioning information Glonass can act on this to immediately hit SPA, MRLS, or a howitzer. Indeed, there was a battery of “on-leave” servicemen using Msta-C, possibly modernized to Msta-CM. It is doubtful that these Msta-Cs are war trophies, as the UA lost only 5 units of heavy Msta-C. Also note that the UA has only 35 heavy Msta-C, which is too few to win artillery duels on such an extended front. Another candidate for fighting with Zoopark-1 is modern Russian MRLS Tornado-G, but it is hard to distinguish it from usual MRLS-12 Grad: the difference is in electronic systems of control, navigation, and communications.

Either way, these huge losses of the UA artillery cannot be explained without supposing that its adversaries possessed more powerful and modern artillery, at least at the time of artillery duels. It is likely that reduced shelling of Donbass towns by the UA is not so much the result of Misk-2 agreement as the result of anti-artillery actions, which might have been conducted by “on-leave” Russians.

The myths of huge amounts of stored USSR military equipment that the UA has

The Military Balance report specifically notes that the amount of military hardware in storage that the UA has is not just very small but zero in many categories. Everything that could be taken out of storage in terms of SPA, MRLS, APCs, AFVs had already been taken, and the UA has nothing in storage. The UA has some 203 mm howitzers Pion, outdated “nuclear artillery” pieces, which, without nuclear missiles, such as ”Perforator”, are not particularly useful because of their unacceptably large scatter of fired missiles. That’s why the UA does not use these self-propelled howitzers, except for 7 units, which is more for PR actions than real military tasks.

The other significant stored source for the UA – nearly 700 tanks. However, most of them are T-72 and T-80, which cannot be repaired without Russian spare parts. That is why the UA uses only 78 of these tanks, the rest are likely to be “cannibalized” as a source of spare parts. The condition of the tanks in storage is also in doubt. While the irrevocable loss of tanks T-64 in ATO was ~35%, but, according to Military Balance, the UA did not use a single T-64 from storage. As Poroshenko advisor Bityukov notes, the tanks in storage are “dead bodies” ( that can be used at best as the source of guns, as artillery pieces of necessary quality are not produced in Ukraine.

In fact, stored tanks made possible to reduce the tank losses to 29%. At the same time much more critical losses of APCs, AFVs, self-propelled artillery, and MRLS were not replaced from storage or purchases from the Ukrainian military industry, as the pieces of these types are either not produced in Ukraine, or produced in minuscule quantities, as compared to huge losses in ATO.


Poroshenko’s statement that the UA lost up to 60% of military hardware during the war is likely true.

The Ukrainian military industry has no capacity to compensate for losses of that magnitude and managed to repair only 10% of damaged hardware.

The Ukrainian military industry does not produce self-propelled artillery and MRLS, and produces very few APCs and AFVs. Repairs depend on Russian spare parts, so the losses of these units are irrevocable.

Judging by the losses of light self-propelled artillery pieces (59%), light howitzers (80%), and MRLS “Uragan”, in Donbass the UA was faced by more professional and modern artillery of the adversaries and was hopelessly outgunned in counter-artillery battles. Since the Ukrainian military industry does not produce self-propelled artillery and MRLS, the UA artillery cannot be replenished.

Large losses of APCs and AFVs make it impossible for the UA to take advantage of its superior mobilizational resources and create big motorized infantry units. In fact, the personnel of motorized infantry was reduced by ~60%.

Most likely, the number of killed and seriously wounded UA solders is ~30,000, as this number is the sum total of crews of seriously damaged hardware, and the same number was mobilized in the first two waves without the creation of new units.

To summarize, as far as the most common military hardware, such as APCs, AFVs, and light self-propelled artillery, are concerned, the UA losses are ~60% and cannot be replaced by the Ukrainian military industry. The UA capacity for offensive at this level of losses is doubtful. In addition, one can expect a manifold increase in the official numbers of the UA personnel losses, as the data on lost hardware and mobilization suggest the loss of ~30,000 solders.

We might want to consider the report lostarmour on lost armored vehicles by the Ukrainian and Novorossian armies in Donbass in 2014-2016.

Generally speaking, both methods have their advantages and disadvantages, as it is obvious that lostarmour did not count everything destroyed, as well as that the loss of hardware (counted based on staffing standards) in some cases did not mean that it was destroyed. For example, some hardware lost after 2013 was left in Crimea and returned to Ukraine only partially. Some hardware could have existed only on paper and even before the war could have been non-repairable. This suggests that the real losses of the UA still need to be further researched to make the conclusions more precise.

The comparison of Military Balance reports for 2013 and 2014 can be found here:

PS. Regarding “on-leave” Russian personnel and “Zoopark” systems, it is worth noting that a Kiev court refused to acknowledge “Russian aggression”

As I wrote here almost two years ago:

One can argue anything in the Internet, but diplomats would play ping-pong for weeks and months. There is certain job to be done, and there is its informational and diplomatic cover. Just like in Crimea. This game continues, and is accepted by all sides. One would think that the masters of the junta have many “proofs” of the work of “Voentorg” (literally “Military Store”; the name stands for the for Russia sending supplies to the rebels – translator’s note), but with amazing constancy they repeat the same line “But the rebels certainly could not have that” and receive the standard answers like “Whoever understands whatever, and whatever do the rebels have, you have no proof”. As this arrangement is cyclical, it can last indefinitely, even until Kiev or Lvov is taken: Lavrov will step up to the microphone and state that Russia does not participate in anything. And officially this will be so. Whereas anything that any blogger or commenter wrote in his blog would remain personal opinions of individuals. I dare you to say that this scheme does not work.

As we see, the scheme, where officially nothing exists, can work not just for weeks and months but for years, and these rules of the game are accepted by all participants, with some stipulations.

The Essential Saker: from the trenches of the emerging multipolar world