What a Demilitarized Zone Means for Liability and Responsibility for Transboundary Harm Emanating from Zaporizhzhia: Some Preliminary Thoughts

Written by

‘We cannot continue this situation where we are one step away from a nuclear accident. The safety of Zaporizhzhia Nuclear Power Plant is hanging by a thread.’

This was the alarming observation made by Rafael Mariano Grossi, Director General of the International Atomic Energy Agency (“IAEA”), at a Board of Governors meeting taking place on 12 September 2022. These comments came at the aftermath of findings published by the IAEA Support and Assistance Mission to Ukraine’s Zaporizhzhia NPP (“ISAMZ”) that a shelling operation from 5 – 6 August 2022, had ‘compromised all of the [IAEA’s indispensable] Seven Pillars [of nuclear safety and security]’ (2nd Summary Report by the Director General, ¶8). On 11 August 2022, Grossi told the United Nations Security Council that ‘IAEA experts have preliminarily assessed that there is no immediate threat to nuclear safety as a result of the shelling or other recent military actions. However, this could change at any moment’. These sentiments were echoed by António Guterres, Secretary-General to the United Nations, who expressly called for the establishment of a “demilitarized zone” enclosing the NPP. In furtherance of this objective, on 15 September 2022, the IAEA Board of Governors passed a resolution (26 in favour, two against, seven abstentions) whereby they called on Russia to: 

immediately cease all actions against, and at, the Zaporizhzhya Nuclear Power Plant […] in order for the competent Ukrainian authorities to regain full control over all nuclear facilities within Ukraine’s internationally recognized borders, including the Zaporizhzhya Nuclear Power Plant, to ensure their safe and secure operation, and in order for the Agency to fully and safely conduct its safeguards verification activities.

In a Joint Statement dated 23 September 2022, ‘the ministers of Foreign Affairs of Canada, France, Germany, Italy, Japan, the United Kingdom, Ukraine and the United States of America, and senior officials from the Republic of Korea and Switzerland’, underscored the pertinence of a demilitarized zone by cautioning ‘that the heightened risks of a nuclear incident will remain dangerously high as long as Russia remains present on the site of ZNPP’.

Alas, it appears that these warnings have fallen on deaf ears. Russian authorities recently elected to illegitimately detain the Director General to the Zaporizhzhia NPP, thus further undermining the safe operation of the plant. Moreover, on 8 October 2022, the IAEA reported that the ‘nuclear safety and security situation […] has deteriorated further, with the facility losing all access to external power due to renewed shelling overnight’. In response, President Volodymyr Zelensky emphasized the importance of the demilitarization efforts, claiming that ‘[t]he protection zone “is an absolute and urgent imperative” in order to prevent a nuclear accident’.

The law governing demilitarized zones is codified in Article 60 of Protocol I to the Geneva Conventions of 12 August 1949, to which both Russia and Ukraine are States Parties. Municipally, demilitarized zones are recognised by Ukraine’s 2004 IHL Manual and Russia’s 2001 Regulations on the Application of IHL. The institution of a demilitarized zone is contingent on the conclusion of an express agreement, and unilateral declarations are void (Article 60(2)). The signatories are entitled to tailor the conditions operational within the zone, and traditionally will elect to evacuate all military personnel and equipment from the area, as well as, call for the cessation of all hostilities therein (Article 60(3)).

In this post, I will discuss how (if at all) the declaration of a demilitarized zone around the site of the NPP would alter principles governing international liability and responsibility for transboundary harm under the 1963 Vienna Convention on Civil Liability for Nuclear Damage (“Vienna Convention”) and IHL, respectively. My earlier co-authored piece contains analysis and contextual information that may be a helpful accompaniment to understanding the analysis in this post.

Civil Liability Under the Vienna Convention

One will recall that in the event of nuclear damage, Article IV(1) legislates a regime of “strict” and “absolute” liability, which is channeled to an Operator, irrespective of any fault on its part, provided that the damage ‘has been caused by a nuclear incident […] in his nuclear installation’. The “Operator” is defined as the ‘person designated or recognized by the Installation State as the operator of that installation’ (Article I(1)(c)), and Article I(1)(d) defines “Installation State” in two alternative limbs as:

[(i)] the Contracting Party within whose territory that installation is situated or, [(ii)] if it is not situated within the territory of any State, the Contracting Party by which or under the authority of which the nuclear installation is operated.

The Zaporizhzhia NPP is situated at the outskirts of Enerhodar in the territory of Ukraine. As aforementioned, Energoatom is the designated Operator for the purposes of the Vienna Convention. In my opinion, this would not change in the event that a demilitarized zone is declared over the region.

The commentaries to Article 60 tacitly recognize that demilitarized zones do not exist within a vacuum, nor does the declaration of one result in a normative change to title over the land. Rather they are protected institutions operating within the sovereign territory of a state. Inter alia, this characterization is observed in the commentary to paragraph 1, which prescribes that ‘the Party to the conflict in whose territory the demilitarized zone is located cannot set up military installations there’ (1987 Commentary, ¶2305). Thus, any area over which the demilitarized zone extends would remain part of Ukrainian territory. Consequently, for the purposes of Article I(1)(d), the Zaporizhzhia NPP would continue to qualify as a “Nuclear Installation” (Article I(1)(j)) situated within the territory of a Contracting Party, namely Ukraine.

In my opinion, the declaration of a demilitarized zone falls short of engaging the second limb. Admittedly, the Explanatory Texts to the Vienna Convention fail to provide any guidance as to when an installation shall be deemed as ‘not situated within the territory of any State’. Nonetheless, reading this provision in its ordinary meaning, I hypothesize that this limb was intended to incapsulate those installations which are, say, positioned in areas beyond the limits of national jurisdictions (e.g., although not a party to the Vienna Convention, the nuclear reactor operated by the United States of America at its McMurdo Research Station in Antarctica from 1962 to 1979).

It is equally important to recall that pursuant to Article IV(3)(a), ‘[n]o liability under this Convention shall attach to an operator for nuclear damage caused by a nuclear incident directly due to an act of armed conflict, hostilities, civil war or insurrection’. The Explanatory Texts fail to provide an exhaustive or non-exhaustive list of acts which may qualify as ‘an act of armed conflict’. They do however reveal that the term, ‘an act of’, may be read interchangeably with, ‘an act committed in the context of’ (Explanatory Text, p.48). This expansive interpretation gives leeway to argue that complications at the NPP resulting from the Russian occupation thereof may fall within the ambit of this provision (so far as they also satisfy the direct cause condition). However, even then, due to the limited exposition regarding this provision, it is impossible to determine definitively whether, in the event that a demilitarized zone is declared and control of the plant is readmitted to Energoatom (either exclusively, or in unison with the ISAMZ), any nuclear incident which is caused partially or solely due to residual complications stemming from the former occupation of the plant by Russian authorities (which has been reported as having undermined the IAEA’s seven indispensable pillars of nuclear safety and security) would be sufficient to satisfy this ambiguous threshold and exonerate Energoatom from liability.

Would the Declaration of a Demilitarized Zone Modify Russia’s Status and/or Obligations as the Occupying Power Under IHL?

The key questions herein are, firstly, whether the declaration of a demilitarized zone would constitute the end of Russia’s occupation of the Zaporizhzhia NPP?, and secondly, how (if it all) would this then alter the obligations prescribed to it under the law of belligerent occupation?

As of this date, it is reported that the Russian forces remain in control of approximately three quarters of the Oblast, including the premises of the NPP.

Publicists such as Sir Adam Roberts, Lassa Oppenheim and Gerhard von Glahn endorse the view that occupation terminates once the foreign army withdraws from the relevant territory (Roberts, ¶20). Von Gahn believes this to be the case ‘regardless of the source or cause of such dispossession’ (Ibid.). This is a practically sound approach, as nothing else could more meaningfully undermine or annul the belligerent’s ability to exercise effective control over a region. Article 60(3)(a) provides a non-exhaustive list of characteristics (four in total) commonly prescribed to a demilitarized zone. It reads, ‘[t]he subject of such an agreement shall normally be any zone which fulfils the following conditions: (a) all combatants, as well as mobile weapons and mobile military equipment, must have been evacuated’. Removal of military personnel can be observed in sate practice in relation to the First Kashmir War (‘[n]o troops shall be stationed or operate in the area of the BRUZIL Nullah from south of MINIMARG to the cease fire line’, Agreement Regarding the Establishment of a Cease-Fire Line in the State of Jammu and Kashmir, ¶D). However, the commentary to Article 60(3) confirms that ‘the word “normally” in the introductory sentence was introduced to permit the Parties to agree about zones which do not fulfil all the conditions of this paragraph’ (1987 Commentary, ¶2310). Indeed, the authorised presence of military personnel is reflected in state practice in relation to the Korean War (Panmunjom Armistice Agreement of 1953, Article 1(10)).

A future demilitarization agreement which abides to the specificities suggested by Secretary-General Guterres would require the withdrawal of all military personnel, indiscriminately. In such an event, under the foregoing general principles, one could reasonably argue that Russia’s occupation of the Zaporizhzhia NPP would be terminated. Alas, here one begins to encounter a lacuna. IHL is unclear as to the extent (if any), of so called “post-occupation” obligations, which may be prescribed to a (former) belligerent occupant (Benvenisti, p.86). Eyal Benvenisti interprets Article 43 of the Hague Regulations expansively (Ronen, p.435), believing that the obligation to ‘ensure, as far as possible, public order and safety’ applies equally to the ‘continuation of “public order and civil life” during and immediately after the termination of the occupation and transition to indigenous rule’ (Benvenisti, p.87). Substantively, the principle should presumably extend to the obligation to prevent transboundary environmental harm, which Benvenisti deems as falling within the ambit of Article 43 (Ibid., pp.18 and 84) and the corpus of international law generally as affirmed by the ICJ in its Namibia Advisory Opinion (Ibid., pp.84-85). However, Benvenisti does not specify whether the extension would apply spatially to a temporary cessation of occupation, or to those which do not necessarily guarantee a return to indigenous administration. The practical elements of abiding by this purported extension are likewise ambiguous and perhaps best assessed on a case-by-case basis. It would be uncontroversial to suggest that in surrendering the premises to Energoatom, Russian authorities would need to release and reinstate the detained Director General of the Zaporizhzhia NPP and disclose any and all existing deficiencies with the site’s operations. However, whether Russia would also need to provide additional support through, say, the provision of materials, capital or additional scientific personnel is less clear.

However, in light of Russia’s recent declaration of annexation over the Zaporizhzhia Oblast, it is highly unlikely that any future demilitarized zone will see Russia relinquishing its military presence in or around this plant. Thus, realistically, in order for any prospective demilitarized zone agreement to be finalized, it may need to provide for the regulated joint presence of some military personnel on the premises. Professor Roberts suggests that, in accordance with state practice, the termination of an occupation shall not be offset by the residual presence of belligerent forces so far as there is a fundamental change in the status of their presence (Roberts, ¶27). He alludes for instance to various peace treaties which, while expressly terminating occupation, nonetheless provided for the continual (albeit mostly conditional) presence of foreign troops. It is unclear if a formal agreement recognizing the end of the occupation is a mandatory condition precedent, or whether an agreement whereby the majority of the belligerent forces are expunged and the residual personnel are deprived of any meaningful administrative powers of effective control (thus more subtly bringing the occupation to an end), would equally suffice. Either way, it appears that there are legitimate avenues for Russia to absolve itself of its status as an Occupying Power (and potentially the flowing obligations under Article 43 of the Hague Regulations) while strategically maintaining military personnel at the premises enclosed by the demilitarized zone. It shall perhaps serve as some comfort that, in the event that the change in status is not sufficiently meaningful, or the troops fail to adhere to their limited roles, thus essentially rendering the ‘new situation an occupation by another name’, Professor Roberts writes that this would invariably result in the reassertion of the law of occupation (Roberts, p.618 and Roberts, ¶33) and the obligations prescribed to the Occupying Power.


It is evident that the declaration of a demilitarized zone would give rise to even further unexplored questions.

Now more than ever, it has become imperative for the IAEA, under whose auspices the Vienna Convention was concluded, to provide further elucidation on the interpretation of these ambiguous provisions.

The lacunae regarding the post-occupation obligations of belligerents are quite unfortunate in the given circumstances. Perhaps some of the lacunae could be covered through bespoke regulations and conditions incorporated into the agreement underlying and establishing a demilitarized zone. Unfortunately, it is unlikely that Russia will be amenable to consenting to any such additional obligations, given that currently it is characteristically reluctant to even take a seat at the negotiation table.

Print Friendly, PDF & Email

Leave a Comment

Your comment will be revised by the site if needed.