On October 25, 2024, NPEC’s Executive Director, Henry Sokolski, gave remarks before “Nuclear Safety and Security in an Insecure World,” a workshop organized by American University and the International Atomic Energy Agency.
Proportionate Targeting of Nuclear Plants and Secrecy
Remarks by Henry Sokolski
Given the nuclear drama playing out in the Middle East and Ukraine, clarifying what harm military targeting of civil nuclear facilities could produce could not be more timely. Equally important is determining what proportionate strikes entail.
The legal treatment of these matters for the last half century has been all too binary. You target these plants and either “hit” or “miss” them. They either release hazardous forces and inflict severe harm or they don’t.
This treatment may have seemed reasonable when bombing accuracies were measured in thousands of feet and there were only a few types of reactors, dumb bombs, and inaccurate weather forecasts. Given such limitations, effective aerial assaults against nuclear plants had to be relatively massive and either “destroyed” the plant or not.
Modern technology, however, has changed that.
Today, targeting inaccuracies are measured in feet (i.e. close to zero). Precision targeting and tailored munitions make it possible to hit a variety of individual plant nodes (or combination of nodes) with differing effects. These nuclear plant nodes include the spent fuel pond storage building, the reactor control room, and exposed steam turbine pipes. They also include the plant’s electricity supply needed to assure the plant’s secure operation, such as emergency diesel generators and power lines coming in and out of the substation transformers on and off the reactor site. Disabling these nodes with cyber warfare is yet another promising (albeit poorly understood) offensive military capability.
The growing number of reactor designs complicates matters further. Two years ago, the IAEA identified more than 80 new small modular reactor designs. Add to this new types of reactor fuels and fuel processing and fabrication plants and you have yet another additional layer of complication.
Some designs have positive void coefficients, such as the RMBK reactors at Chernobyl and Kursk (which lack serious containment structures). Others are fast reactors, including those under construction by TerraPower in Wyoming, the reactors Google intends to use to power AI, and those China, Russia, and India are currently building or operating. If any of these facilities suffer a loss of coolant, their nuclear reactions will speed up and the plants could explode as Chernobyl did.
The fast reactors noted and a good number of others use liquid sodium as a coolant. Sodium reacts violently upon contact with the oxygen in water, air, and even in concrete. In contrast, most thermal reactors have negative void coefficients. If these thermal reactors lose coolant, they may release radioactivity but their reactivity declines; they are not prone to explode. However, some boiling water reactors, such as those at Fukushima, can experience fuel failures that result in hydrogen explosions and, like RMBKs, have almost no containment.
Exploiting the unique vulnerabilities of different types of reactors with precision targeting can vary the effects and radiological releases significantly. At one extreme, hitting a spent fuel pond can induce a spent fuel fire and produce Chernobyl-like effects. At the other, hitting a car in the plant’s employee parking lot would kill no one directly but would likely cause precautionary plant shutdowns, grid instability, and public alarm.
Today’s military planners also have much better modeling tools for calibrating their attacks. A six-day weather forecast today is as accurate as a one-day forecast was in 1970. The number and accuracy of publicly available radiological dispersal models have also increased.
All of these factors complicate the targeting of nuclear plants. On the other hand, they also afford an opportunity for military planners and operators to disable these plants without releasing harmful forces that could do severe harm to innocents.
That is why this forum is so important. More public forums of this sort at the IAEA, the International Red Cross, and elsewhere are needed. Certainly, excessive government secrecy about the consequences of different military strikes against nuclear plants and what they are doing specifically to avoid disproportionate attacks is misguided.
Proportionality might not be litigated in national courts but it is an essential principle that must be practiced in war planning and operations to prevent the loss of good order and discipline within any fighting force. Experienced officers understand that the self-restraint and self-discipline proportionate military operations require are as essential to assuring victory as superior weaponry and strategy.
Ultimately, the only way to assure such good order and discipline is by offering the public sufficient examples of what proportionate targeting looks like against specific nuclear plants. With these examples, the public can properly hold the military accountable for grossly disproportionate assaults against what are, after all, public utilities.
At a minimum, the United States and its military allies should encourage unclassified military case study-based instruction on what disproportionate and proportionate military strikes against nuclear plants look like. This instruction should be coordinated with top nuclear experts in and out of industry. Most important, U.S. and allied military top brass need to tie military operators’ and planners’ promotions to the proper mastery and application of such instruction.
With the increased targeting of critical civilian infrastructure, including nuclear plants, doing anything less is a prescription for unnecessary and possibly dramatic harm.