In August, Russia accused Ukraine of attacking an auxiliary transformer that forced a 50 percent reduction in the operating capacity of its Kursk nuclear plant. No radiation was released. It was a textbook example of how to disable a nuclear plant without directly assaulting it.
In the attached article for the Army War College’s journal Parameters, I make the case that such restrained targeting should be required in the case of nuclear plants. Although it has become fashionable to downplay proportionality concerns about harming civilians and civilian objects in war—and even to treat such concerns as obstacles to military effectiveness—I argue that with nuclear plants it’s just the opposite.
As I note, Clausewitz, Eisenhower, Churchill, Truman, and even Rommel understood that needlessly killing civilians and destroying infrastructure critical to their welfare only complicates the prospects for military success. It can increase local resistance, jeopardize secure logistical lines of communication, deprive one’s forces of the option to co-opt critical facilities for their own use, and complicate these facilities’ eventual repatriation. In the case of nuclear plants, it also risks contaminating combat zones, thereby jeopardizing both offensive and defensive operations.
Fortunately, with precision-guided munitions, it is possible to disable nuclear facilities without risking these hazards. I highlight some of these options in the article. I also recommend that the best way to ensure our military heeds such technical insights when targeting critical civilian infrastructure is to make their professional promotions dependent upon mastery of these matters through instruction or in practice.
As always, I’m curious to get your take on the attached.
As Ever,
Henry S.
September 16, 2025
By: Henry Sokolski
A Case for Military Proportionality: Disabling Nuclear Plants
For nearly a decade, protecting civilians and civil objects from disproportionate military assaults has been a top priority of the Pentagon. Two Department of Defense secretaries from the first Donald Trump administration championed quantifying and reducing harm to civilians and civil objects. Under the Joe Biden administration, the Pentagon further focused on protecting civilians and civil objects, and, in 2023, Congress created a Civilian Protection Center of Excellence within the Department of Defense. This center, consisting of a staff of 30 people with an annual budget of $7 million, helped military commands execute their missions while minimizing collateral damage.1
In early 2025, however, the Pentagon cut the funding and eliminated almost all the staff in the Civilian Harm Mitigation and Response office and the Center and asked Congress to eliminate the legal requirement for its continued operation. Rattled, some wondered if the Department of War was rescinding its previous guidance on limiting civilian harm. The answer to the question was unclear.2
Trump administration officials stated the Civilian Protection Center of Excellence jeopardized war fighters’ abilities to do their jobs. But those officials did not discuss a deeper set of developments: Hamas’s October 7 attack against Israeli citizens; Israel’s crushing response, which killed thousands of noncombatants; and Russia’s attacks against Ukrainian civilians and civil infrastructure. Each development challenged many experts’ previous beliefs about what proportionality should prohibit.
Both Vladimir Putin and Benjamin Netanyahu insist their military operations are proportionate. These claims, in turn, rely on an American view of proportionality Abraham Lincoln’s top military and legal adviser, Francis Lieber, promulgated in the 1860s. The Lieber Code (General Orders No. 100) championed avoiding attacks on civilians and civilian objects. But the code also allowed, if a compelling military objective emerged whose achievement incidentally entailed harming civilian people and objects, that attacks were permissible. Commanders on the front lines should decide what actions are militarily justified or not, according to the code.3
Some have argued Lieber’s view renders proportionality hopelessly subjective. If commanders were free to determine what actions are justified, proportionality would seem to be little more than a standard of behavior the weak may demand of the strong, but the strong can effectively ignore. Victorious nations rarely litigate against their own officials or officers for disproportionate military actions (that is, for ignoring or violating the requirements of proportionality).4
Therefore, enforcing proportionality against defeated foreign nations might be attractive, but demanding one’s own military enforce proportionality is less realistic or practical. At best, realists argue, limiting harm to civil persons and objects is advisory; institutionalizing or promoting proportionality by creating Pentagon centers goes too far.
This line of thinking is intuitive and appealing. But it ignores a critical point: Sparing civilians and civilian objects unnecessary harm is often essential to achieving military victory.
Carl von Clausewitz, known for championing the necessity of violence in battle, was just as emphatic that wars could only be won by reaching political solutions the enemy’s military and leadership—and the enemy’s population—could accept. Needlessly killing civilians and destroying infrastructure critical to their welfare only complicates reaching lasting political solutions. For Clausewitz, the need to inflict violence in war had to be measured against the war’s ultimate objective, which is always political. Violence against civilians is self-defeating if it undermines the achievement of the war’s ultimate political objective.5
Thus, Winston Churchill and Dwight D. Eisenhower resisted calls in 1944 for the indiscriminate bombing of French cities and infrastructure during World War II because though such bombings would weaken German defenses, they would also dramatically undermine French political support of the Allied powers and the Allies’ resistance to the Nazis. Indiscriminate bombing would also complicate the reconstruction of the French economy after the Allies won the war.6
For similar reasons, President Harry S. Truman rejected the advice of his commander in the field, General Douglas MacArthur, who wanted to use nuclear weapons on North Korea and China. Truman feared attacking these states with nuclear weapons would escalate the conflict, cause unnecessary destruction, and turn international public opinion against the United States. Truman understood maintaining international support was essential to containing China and deterring Russia’s use of nuclear weapons after the end of the Korean War.7
One of Adolf Hitler’s best generals—Erwin Rommel—also refrained from using excessive force against civilians to protect his communications and supply lines from local disruption. Rommel understood that, in some cases, good military discipline and order required restraint, as did pacific relations with the local population (for example, in Northern Africa). Rommel’s attention to these points helped secure supply lines and reduced local resistance to his forces’ operations.8
Nazi troops terrorized enemy populations, but General Walther Wever, who served as the Luftwaffe’s chief of staff in the mid-1930s, argued such actions. Responsible for formulating Germany’s military air doctrine, Wever rejected the idea of bombing cities to break the will of the people. Wever believed such attacks were, at best, distractions from the Luftwaffe’s main mission: destroying the enemy’s armed forces. Wever also believed terror bombing was militarily self-defeating because it increased, rather than reduced, local resistance, jeopardizing the achievement of the Luftwaffe’s prime military missions.9
Besides these arguments, there are additional reasons for not hitting certain civilian facilities. Attacking chemical plants and nuclear facilities can poison the theater of operations with dangerous contaminants and hamper military operations (for example, if a dam is attacked, flooding the terrain). Such attacks can also prompt major evacuations which, in turn, retard military movements.
However, another advantage of avoiding conducting military assaults on civilian objects relates to military cohesion. As I noted in a previous Parameters article, Protocol I of the Geneva Conventions specifically discourages nations from attacking civilian objects, especially if doing so would risk releasing “hazardous forces” that could inflict “severe harm” on innocent civilians. Although the United States has signed the protocol, 174 nations took the additional step of ratifying it. The United States chose not to do so. As such, the United States is at odds with most of its NATO Allies.10
Thus, in 2022, foreign and military ministers in the European Union, the United Kingdom, and Germany declared Russian strikes against Ukrainian infrastructure and the Zaporizhzhya nuclear power plant were prosecutable war crimes. The United States took no position. In a war game conducted in 2022, close US Allies that have ratified Protocol I were at odds with Washington regarding how to respond to Russian attacks on Allied reactors. The United States’ Allies wanted to respond strongly to what they saw as a war crime, whereas the United States did not. In the game, the other NATO members were concerned NATO would be drawn into a larger conflict if Poland and Ukraine jointly attacked Russia. These concerns held up war operations and resulted in the United States using Article 5 of the North Atlantic Treaty to keep Poland from participating in a Ukrainian strike against Russia.11
Finally, temporarily disabling civilian infrastructure (for example, water, gas, and oil pumps; energy pipelines; telecommunications lines; and electrical-supply systems) can afford clear military advantages over physically obliterating civilian infrastructure, even if no hazardous forces are released. The temporary disablement of civilian infrastructure deprives one’s enemy of the ability to use infrastructure facilities, facilitates their subsequent use by one’s own forces in war, and allows for their speedy repatriation once the war is over.12
All of these points recommend fostering effective military applications of proportionality against civilian objects. The question is how.
As noted, taking legal action against disproportionate military actions is a proposition. Attempting to conduct legal proceedings against the military of a defeated nation may be politically useful after a war has ended. But attempting such proceedings seldom, if ever, significantly impacts the execution or planning of military operations before or during a given operation.
Only a handful of cases exist in which US officials have applied US criminal law against soldiers who have harmed noncombatants. As a result, mining national case law for guidance is unlikely to produce many useful insights. That said, the US military has discouraged inflicting unnecessary harm on noncombatants.13
But what of the disproportionate targeting of civilian infrastructure and objects? Unfortunately, few (if any) specific national, military, or international legal cases regarding such indiscretions have been decided. To discourage disproportionate attacks against civilian infrastructure, the United States instead relies upon using written military doctrine, training, planning, and reviewing previous operations.14
The Pentagon’s most useful specific guidance, however, is the result of classified analysis. This guidance presents problems.
First, the guidance is at odds with the demands of domestic civil defense. In terms of domestic civil defense, the civil objects of greatest interest—oil, gas, water lines, electrical-supply systems, nuclear power plants, transportation systems, and public communications systems in the United States and in allied nations—are all public utilities. Therefore, protecting civil objects requires public engagement and sound, convincing narratives about how the objects might be attacked and what will or should be done to protect them. This domestic requirement cannot be met without sharing some of the very information the military keeps from the public.
Second, the failure to produce a public narrative about what civil objects should not be hit, and why, discourages informed civilian oversight of the military. Freed from public scrutiny, the military is under little pressure, if any, in planning operations to avoid disabling civilian objects.
Finally, excessive secrecy regarding the targeting of civil objects allows for overly permissive legal guidance. Currently, US legal guidance is all too similar to the guidance Russian and Israeli officials use. American, Russian, and Israeli officials have all argued in different ways that if a military commander determines striking an object is militarily necessary or important enough, it becomes a critical military object that can be attacked, even if the strike inflicts severe harm on noncombatants. Such permissive legal views can open the door to making disproportionate attacks.15
The point here is hardly academic. Current US Air Force targeting guidance and classified collateral-damage algorithms from the Pentagon (which discourage strikes against civilian objects capable of releasing harmful forces) clearly discourage the targeting of civilian objects capable of releasing harmful forces (for example, dams and power reactors or toxic chemical plants).16
Yet, The New York Times reported, in 2017, Air Force targeting guidance was short-circuited when a highly classified US special operations unit—Task Force 9—hit Syria’s Sadd al Furāt dam. The dam was on a “no-strike” list, and attacking it required top-level authorization. But Task Force 9 bombed the dam because the attack was deemed a military emergency. Although United States Central Command officials admitted the dam had been bombed, they insisted only the dam’s towers had been hit.17
Exactly what happened when the dam was hit and the government’s specific reasons for launching the attack have not yet been shared with the public. Yet, the Lieber Institute at West Point published a legal defense of the attack. The author argued that the dam was a “military objective.” As such, it was a legitimate target, even though attacking it might release harmful forces that could inflict severe harm to civilians. Although logical, this line of reasoning risks justifying almost any strike against civilian objects.18
This predicament raises the question: Is there a sounder line of reasoning that would recommend more restrictive guidance?
If the Pentagon is serious about promoting effective military operations against civil objects, looking at the past can help. For example, one could critique Russia’s attacks against Ukraine’s electrical-supply system. But, some might see those critiques as litigating the current war and use them to support making legal filings with the International Criminal Court. Given its fear of being tried in the International Criminal Court, the Pentagon would likely be wary of making an analysis of Russia’s attacks public.
A safer alternative would be to consider future targets which, if hit indiscriminately, could release hazardous forces and render vast areas inhospitable or toxic. An obvious poster child for releasing hazardous forces is nuclear plants. Whereas Israel, the United States, Iraq, and Iran have all previously targeted nuclear reactors, none were large plants that were actually operating. As a result, none of the attacks produced major radiological releases. In contrast, indiscriminately hitting a large, operating nuclear reactor could release significant radiation.19
If a reactor’s spent fuel pond storage system was hit, the likely radiological releases could force millions of people to evacuate, as confirmed by US Nuclear Regulatory Commission, government-sponsored, and private studies. The areas rendered uninhabitable could also be quite large: from 30,000 to 100,000 square kilometers (the latter area is larger than the entire state of New Jersey). In an attack against a spent fuel storage facility, US Nuclear Regulatory Commission staff conservatively estimate the radiological release could be 100 times greater than that of the Fukushima accident.20
The case of an attack against a spent fuel storage facility is extreme. A less dramatic scenario is the radiological release attendant to a loss of coolant induced by a military assault. Still, a wholesale, indiscriminate attack against Iran’s Bushehr power reactor could release significant radiation and force the evacuation of hundreds of thousands to millions of nearby civilians.21
Wholesale, indiscriminate attacks are precisely the kind of assault diplomats and lawyers aimed to prevent when they crafted Protocol I to the Geneva Conventions a half century ago. This international framework has several provisions that focus on the most likely type of military assault against nuclear power plants at the time: wholesale aerial attacks, which were almost certain to trigger massive releases of radioactivity. Today, things are different. With precision targeting and tailored munitions, nuclear power plants can be disabled in many ways without releasing radiation.22
Oddly, this transition to precision is still not fully reflected in the Pentagon’s legal guidance on targeting nuclear plants. The Pentagon’s Department of Defense Law of War Manual does suggest that “in light of the increased potential magnitude of incidental harm, additional precautions, such as weaponeering or timing the attack such that weather conditions would minimize dispersion of dangerous materials, may be appropriate to reduce the risk that the release of these dangerous forces may pose to the civilian population.” This counsel, however, ultimately is only advisory. As the manual notes, “the intended effects on the civilian population were secondary to the military advantage gained by attacking the electrical power infrastructure.” As such, they are legitimate targets. A commander must determine whether launching an attack against a nuclear plant that might result in severe harm to innocent noncombatants is proportionate. But it is entirely his or her call.23
The objective behind this legal allowance is unclear. Might the United States still have contingency plans to target adversaries’ nuclear plants with US nuclear weapons? If so, the manual’s legal position would help preserve this option. Yet, considering the massive amounts of radioactivity a nuclear strike against an adversary’s plant would release (contaminating vast swaths of territory), such a strike today would hardly make military sense.
A half century ago, indiscriminate attacks against nuclear plants were necessary. Given missile and aerial-bombing inaccuracies (only half the unguided bombs dropped in the Vietnam War got within roughly one-tenth of a mile of the intended target), bombing was more of an all-or-nothing proposition: One either had to obliterate the target (and risk releasing radiation if the target was a large, operating nuclear reactor or spent fuel facility) or miss the nuclear facility entirely. To disable a plant, one had to decimate it.
Today, however, median miss distances for precision weapons are measured in meters or in smaller units. As a result, nuclear plants can be disabled in many ways without risking harmful releases of radiation. The Russians, in the Russia-Ukraine War, have demonstrated several disabling techniques. These techniques exploited the nuclear-safety requirement for irradiated reactor fuel to be cooled continuously to prevent it from overheating, failing, and releasing dangerous, radioactive by-products.24
Rather than prompting such failures, analysis suggests Russia has been careful to target the electrical power–supply systems needed to keep the nuclear plants’ cooling and safety systems running. Russia’s aim is twofold: first, to force the plants’ operators to shut them down for safety reasons, and second, to increase the credibility of making follow-on strikes that might risk a significant release of radiation.25
The power-system components Russia has targeted include on- and off-site electrical transformers; high-voltage lines running in and out of the plants; cooling water supply systems; a major dam critical to supplying water to the Zaporizhzhya nuclear power plant; and major, off-site electrical power–generating plants needed to stabilize the electrical-supply grid supporting the nuclear plant’s safe operation.
Through repeated strikes on these nonnuclear components, Russia has succeeded in shutting down Europe’s largest nuclear power plant—the Zaporizhzhya nuclear power plant. In addition, Putin can now collapse Ukraine’s entire electrical-supply system at a time of his choosing. Meanwhile, Russia says it could restart the Zaporizhzhya nuclear power plant to supply electricity to territories occupied by Russia in a matter of months.
More could be said about Russia’s studied targeting of Ukraine’s nuclear power plants and electrical-power systems. But Russia’s attacks afford a clear example of disabling critical civilian objects (reactors) to its military advantage without releasing hazardous radiation.26
Of course, other nuclear examples should be considered. Some states use portions of their civilian nuclear programs to make nuclear-weapons materials—for example, China, India, and North Korea. Disabling the facilities used to make nuclear-weapons materials would be a worthy military objective. Physically, obliterating those facilities and risking the widespread dispersal of harmful radiation, however, could be militarily counterproductive.
In addition to nuclear plants, indiscriminate attacks against civil objects such as dams, petrochemical plants, and biolabs could release harmful forces that inflict severe harm on large numbers of civilians. Can such civil objects be disabled without running the risk of civilian harm? Military prudence suggests it is worth finding out, especially so in light of the increasing threats cyberattacks, locally lofted drones, and long-range missiles pose to civilian infrastructure.
Recommendations
What steps can the US military take to update its plans and operations for targeting and protecting civil infrastructure?
First, the Pentagon should publicly share much more information about its thinking than it has to date, which would allow for greater civilian oversight, sharpen military planning, and increase the clarity of current policy and legal guidance.
Second, the Pentagon should work with private industry and other government departments focused on civil-infrastructure protection—the US Department of Homeland Security and the US Nuclear Regulatory Commission—to produce convincing public narratives about why and how civil objects should be protected and to improve existing protection schemes. Planning to protect this infrastructure has long been underway, but under the protection of the Federal Energy Regulatory Commission’s Critical Energy / Electric Infrastructure Information, which keeps these plans from the public. What’s needed is a sensible tear sheet for public consumption.27
Third, the Department of War should offer Congress routine public reports about matters related to protecting civil infrastructure. The US government must prepare the public for a future in which the United States’ electrical-supply systems, energy pipelines, biological research facilities, potentially dangerous petrochemical plants, telecommunications systems, and civil nuclear facilities may come under attack. Setting the public’s expectations about what can and should be done, actively and passively, to defend these systems should not wait until an attack occurs.
Finally, training is critical. The Department of War’s military education training institutions should offer dedicated, unclassified courses that provide technical and historical instruction on the targeting and defense of civil objects. The instruction should be fortified by unclassified government simulations for civilians and military officials, which play out alternative targeting plans against civil objects that could release hazardous forces.
How will the US government accomplish these objectives? The first step is to make mastering these matters a requirement for military promotion. This step could be done quietly, without top-down scolding, legal hectoring, or creating centers. The best US military operators and planners already know civil objects and nuclear facilities are becoming increasingly significant military targets. The Pentagon should reward and support efforts to clarify what should be done to disable and protect civil objects and nuclear facilities.
Acknowledgments: I would like to thank Caitlyn Collett for providing essential assistance in the production and editing of this special commentary.
To read the full piece, click here.