The Russian Siege of the Azovstal Steel Plant in Ukraine: An International Humanitarian Law Perspective Features
Odessicus/ Wikimedia Commons
The Russian Siege of the Azovstal Steel Plant in Ukraine: An International Humanitarian Law Perspective

On 21 April 2022, Vladimir Putin, in a televised meeting with his defence minister Sergei Shoigu, ordered Russian armed forces to seal off the Azovstal steel plant in Mariupol from the outside world.  The intention, according to Putin, was to spare Russian armed forces from the very high casualty rate that would almost certainly stem from a direct attack on the facility which is serving as a last-stand bunker for thousands of Ukrainians.  The trouble for Putin is that this tactic amounts to ‘siege warfare’ which, while not banned outright, is heavily regulated by international humanitarian law (IHL).  This piece will explain why the siege is illegal.    

The Azovstal Steel Plant 

The city of Mariupol lies in the Donbas region to the southeast of Ukraine.  It sits on the coast of the Sea of Azov and, to its misfortune, lies right in the middle of Russia’s intended ‘land bridge’ between Crimea in the south and Donetsk and Luhansk in the east.  The city has suffered ferocious Russian bombardment with the effect that, in recent weeks, Ukrainian forces have withdrawn to the Azovstal steel plant which doubles-up well as a military base due to its huge size (four square miles), numerous basement levels (ordinarily used for storing raw materials) and thick walls (built to handle extreme heat from furnaces).  Media reports suggest that around 2,000 Ukrainian troops are based at the plant.  From a Russian perspective, those troops need to be neutralised before a full victory in Mariupol can be claimed.

Naturally, storming a labyrinthine industrial complex that has been used by Ukrainian forces as a base for a prolonged period (with the effect that they know it well and have likely set booby-traps throughout) is not an attractive prospect.  It would likely result in the deaths of hundreds of Russian soldiers.  Indeed, Shoigu acknowledged that it would be difficult and take several more days to defeat the Ukrainians in any direct fighting at the steelworks.  Consequently, Putin declared a direct assault to be ‘impractical’.  Instead, he ordered Shoigu to ‘[b]lock off this industrial area so that a fly cannot pass through’.

Siege Warfare and Starvation in International Humanitarian Law

The Russian tactic of denying any entry to, or egress from, the Azovstal steel plant amounts to siege warfare.  As Van Schaak put it, ‘[s]iege warfare is a tactic … that involves surrounding a garrison or a populated area with the goal of driving out the enemy forces by deteriorating their defenses and cutting them off from reinforcements and vital supplies’.  Siege warfare has been used since antiquity and is permitted under IHL; there are no provisions that seek to ban its use outright.  However, in practice, the use of siege warfare is heavily limited by other aspects of IHL that are designed to protect civilian populations from the scourge of war.

Perhaps the key protection for civilians in the context of a siege is that starvation must not be weaponised.  The Geneva Conventions of 1949 are the core international treaties in the context of armed conflict and Additional Protocol I to those treaties (which added important updates in 1977) provides that ‘[s]tarvation of civilians as a method of warfare is prohibited’.  This expression of treaty law (binding only the parties to the treaty) has since been confirmed by the International Committee of the Red Cross (ICRC) in its customary international humanitarian law study (CIL Study) to have crystallised into customary international law (binding all States).  Naturally, when civilians are caught up in conflict, the prohibition against starving civilians has the potential to clash with the general permissibility of siege warfare. This results in the siege/starvation conundrum.

Resolving the Siege/Starvation Conundrum

The resolution to the siege/starvation conundrum hinges on the notion of ‘purpose’.  What is the purpose of the siege (and the resulting constriction of food supplies)?  The ICRC has said ‘the prohibition of starvation as a method of warfare does not prohibit siege warfare as long as the purpose is to achieve a military objective and not to starve a civilian population.  Unfortunately, on this critical issue, there is ambiguity in treaty law and in the literature over how the term ‘purpose’ ought to be interpreted.  In theory, a broad spectrum of possible interpretations is available.  At one end of the spectrum, ‘purpose’ could be interpreted narrowly so that, for a siege to be unlawful, the sole purpose of the siege must be to starve civilians.  At the other end of the spectrum, ‘purpose’ could be interpreted broadly so that, for a siege to be unlawful, it would suffice if even a small part of its purpose was to starve civilians.

Certain sources favour the narrow interpretation of ‘purpose’.  For example, the San Remo Manual (which concerns naval warfare) provides that starvation of civilians must be the ‘sole’ or ‘primary’ purpose of a sea blockade for it to be considered unlawful (Doswald-Beck, San Remo Manual, 1995, para 102(a)).  Likewise, the Air and Missile Warfare Manual provides that starvation of civilians must be the ‘sole or primary purpose’ of an aerial blockade for it to be considered unlawful (Program on Humanitarian Policy and Conflict Research, HPCR Manual, 2013, rule 157(a)).  Notwithstanding these declarations, it would be unwise to attempt to extrapolate a general rule on ‘purpose’ based on the relatively niche rules that have been created for naval and aerial warfare.  This is because battles at sea and in the air are generally devoid of civilians, in stark contrast with land warfare which has grown increasingly urbanised and features much more intermingling of combatants with civilians (Melzer in Clapham and Gaeta, The Oxford Handbook of International Law in Armed Conflict, 2014, page 298).  Moreover, as Gloria Gaggioli points out, taking a narrow approach to ‘purpose’ in the context of land warfare ‘would reduce considerably the value of the provisions on the prohibition of starvation’.  Such a devaluation of humanitarian protection could surely not be in alignment with the object and purpose of Additional Protocol I which is intended, inter alia, to ‘reaffirm and develop the provisions protecting the victims of armed conflicts’ (Additional Protocol I to the Geneva Conventions, 1977, Preamble).  It is probably due to considerations such as these that the ICRC’s CIL Study does not affix any adjectives to the work ‘purpose’ and does not comment on whether the word should be interpreted narrowly or broadly.

If customary international law is ambivalent on the proper interpretation of ‘purpose’, how should the term be construed?  To answer this, we must refer to the fundamental basis of IHL.  For Kolb, ‘the whole body of IHL is the result of a careful and constant balancing of … military necessity and humanitarian need’ (Kolb, Advanced Introduction to International Humanitarian Law, 2014, page 77).  Likewise, for Dinstein, IHL is ‘predicated on a subtle equilibrium between the two diametrically opposed stimulants of military necessity and humanitarian considerations’ (Dinstein, The Conduct of Hostilities under the Law of International Armed Conflict, 2016, page 9).  In short, the law ‘relies on a balance between humanity and military necessity and it is only when they are evenly matched that the correct legal decision can be arrived at in any given scenario’ (Winter, Journal of Conflict and Security Law, 2020, Volume 25 Issue 1, page 6).  The consequence of this in the context of siege warfare is that, as Blank put it, there must be ‘balance between the humanitarian imperative of preventing starvation and the effectiveness of siege as a military tactic’.

The Legality of the Siege of the Azovstal Steel Plant

Returning to the specific case of the Azovstal steel plant, we are now equipped to reflect on whether the purpose of the siege strikes an appropriate balance between military expediency and humanitarian protection.  The answer to this is simple, it does not.  This is because Putin’s order to ‘[b]lock off this industrial area so that a fly cannot pass through’ is entirely one-sided.  It pays no regard whatsoever to the existence, nature or extent of civilian suffering that it might precipitate.   Further, the order does not consider any measures that may be taken to mitigate civilian suffering.  As the ICRC noted, the prohibition on starvation ‘clearly implies that [civilians] must be allowed to leave … during a siege’ or, alternatively, that the besieging party ‘must allow the free passage of foodstuffs and other essential supplies’ (Henckaerts and Doswald-Beck, Customary International Humanitarian Law, 2005, rule 53).  The order to wholly seal-off the steel plant does not contemplate any such measures and so the ‘purpose’ of the siege is flawed.  As Gaggioli put it, ‘failing to attempt an evacuation of civilians, or at least the most vulnerable among them, coupled with a denial of humanitarian assistance should suffice to indicate that the purpose of the siege is to starve civilians’.

For the present analysis, it is sufficient that the purpose of the siege was flawed by the failure to consider the potential for (and possible mitigations against) civilian suffering: those factors alone render it incompatible with IHL.  Nonetheless, it is worth noting that the siege has indeed gone on to cause a great deal of civilian suffering.  Media estimates place between 1,000 and 2,000 civilians in the Azovstal steel plant.  Moreover, conditions are reported to be dreadful.  According to the BBC, women and children trapped in the plant were running out of food and water.  Indeed, one of the Ukrainian personnel at the scene claimed that they were ‘desperately low on supplies’.  Chillingly, a report noted that a child had ‘seen no daylight for weeks, having been there from February’.  In other words, the consequences that IHL is intended to preclude – or at least mitigate by requiring a sound purpose for a siege (one that strikes a balance between humanitarian considerations and military necessity) are, sadly, precisely the consequences that have arisen in the Azovstal steel plant.  

There is no blanket ban on siege warfare in IHL.  However, this tactic of subduing an opponent chafes against numerous aspects of the law that are designed to spare civilians from some of the worst excesses of war.  Whether a given siege is legal depends on its purpose.  The term ‘purpose’ is undefined and so its interpretation must reflect the fundamental equilibrium in IHL between military expediency and the notion of humanity.  In the case of the Azovstal steel plant, there was (quite explicitly) no attempt to accommodate humanitarian considerations when the decision to besiege the plant was made.  Moreover, no discernible flexibility has emerged since the onset of the siege to suggest that humanitarian considerations have been taken into account.  Consequently, the purpose of the siege was, and remains, grounded solely in military expediency.  Thus, the siege itself is in violation of IHL.

Elliot Winter is a lecturer in international law at Newcastle University in the United Kingdom.