Issues in Science and Technology, Online A Publication of National Academy of Sciences National Academy of Engineering Institute of Medicine University of Texas at Dallas
  • Home
  • Back Issues
  • Search

Limiting the Tools of War


The Case against New Nuclear Weapons

New tactical bombs would have little military value and would undercut U.S. nonproliferation efforts.

Does the United States need nuclear bombs to destroy enemy bunkers and chemical or biological weapons? For some people, the answer is clear. Strong proponents of nuclear weapons speak of the need to give the president every possible military option, and the Bush administration's 2002 Nuclear Posture Review reflects this affirmative response. On the other side, committed opponents maintain that no potential military capability could justify designing--let alone building or using--new nuclear bombs. For both camps, the details of the proposed weapons are irrelevant.

Yet neither of the simple arguments for or against new nuclear weapons is broadly accepted. The United States does not develop every possible weapon simply to provide the president with all options; policymakers have, for example, judged the military value of chemical weapons insufficient to outweigh the political benefits of forgoing them. On the other hand, the nation has never rejected nuclear use outright and has always reserved the possibility of using tactical nuclear weapons. Indeed, until the end of the Cold War, such weapons were central to U.S. military thinking.

Despite their disagreements, the people engaged in debate over new nuclear weapons have tacitly agreed on one thing: that these weapons would deliver substantial military benefits. Thus, they have cast the dilemma over new nuclear weapons as one of military necessity versus diplomatic restraint. But this is a false tension: New nuclear weapons would, in fact, produce few important military advances. Yet their development would severely undercut U.S. authority in its fight against proliferation.

Advocates of new tactical nuclear weapons have tended to focus shortsightedly on simple destructive power. In particular, most arguments for bunker-busting nuclear weapons ignore the difficulty of locating threatening bunkers in the first place. During the Gulf War of 1991, military planners painstakingly assessed the potential consequences of bombing Iraqi chemical weapons facilities, debating nuclear and nonnuclear weapons, as well as the option of leaving the bunkers alone. Ultimately, the military used conventional weapons to bomb every known facility. Subsequently, however, international weapons inspectors, aided by Iraqi defectors, discovered that those targets had been the mere tip of a vast Iraqi system for producing and storing weapons of mass destruction. Had the military used nuclear weapons to bomb all known chemical facilities during the Gulf War, the United States would have made barely a dent in Iraq's deadly capability while incurring massive political backlash as people died from the accompanying nuclear fallout.

The challenge of finding hidden targets is the norm, not an exception. In Afghanistan, U.S. efforts to eliminate the Taliban and Al Qaeda were hindered by the difficulty of tracking down their underground hideouts. Intelligence technology, which relied heavily on detecting mechanical equipment, power lines, and communications systems to identify hidden facilities, floundered in the face of a backward enemy who employed none of the technologies being searched for. Osama bin Laden is still alive not because the United States lacked powerful weaponry, but because U.S. intelligence could not find him in the caves of Tora Bora.

Still, an inability to locate all enemy weapons stockpiles and underground leadership targets is not an argument for leaving alone those that can be found. But proponents of nuclear weapons have overstated the capability of the nuclear option even in cases where targets can be located, while underestimating nonnuclear potential. In particular, proponents have contended that nuclear weapons are needed to compensate for difficulties in precisely locating underground targets; that they are needed to neutralize chemical and biological agents and thus prevent their deadly use; and that only with nuclear weapons will there be no "safe havens" (no depth below which enemies are safe). However, each of these arguments can be debunked, as illustrated in the following examples.

Inadequate intelligence

Libya has been suspected of producing chemical weapons at its Tarhunah complex, located 60 kilometers southeast of the capital city of Tripoli and hidden in tunnels and bunkers under roughly 20 meters of earth. The problem is that U.S. analysts have not been able to produce an exact blueprint of the underground chambers. This lack of precision leads some observers to argue that although the facility is, in theory, shallow enough to be destroyed with conventional arms, uncertainty concerning its location may require the large destructive radius of a nuclear weapon to compensate.

A nuclear weapon detonated at or near the surface produces a large crater and sends a massive shock wave into the ground. Underground facilities within this crater are destroyed, as are facilities slightly outside the zone by strong stresses that rupture the earth. Based on the intelligence community's knowledge (even given its uncertainty) about the Tarhunah facility, it is apparent that a five-kiloton ground-penetrating nuclear weapon could destroy it. This attack would produce a moderate amount of nuclear fallout, the precise nature of which would depend on whether the weapon was detonated inside the facility or in the surrounding earth. To be conservative, military planners would have to assume the latter. Such a blast would kill every human being within approximately 15 square kilometers, according to calculations by Robert Nelson of Princeton University. Although this zone would not reach Tripoli, concerns about fallout would require medical monitoring for civilians as far as 20 kilometers downwind from the facility. U.S. troops in the zone would have to halt operations or risk being exposed to fallout. Troops could not enter the immediate facility area to inspect damage or collect intelligence, even with protective gear, which is ineffective against nuclear fallout.

Alternatively, there are a number of nonnuclear approaches that are already available or could be developed for destroying or neutralizing this type of complex. If the main bunker could be more precisely located, then a single earth-penetrating conventional bomb could reach it. A missile the length of the current GBU-28 penetrator, modified to strike the surface at twice the GBU-28's current impact speed, could smash through the cover of earth and reinforced concrete and destroy the facility with conventional explosives. This suggests that the military should focus on improving intelligence capabilities, particularly the ability to precisely map underground targets that have already been located, rather than on devising ever more powerful weapons.

Even if the facility cannot be precisely localized, several conventional penetrator missiles used simultaneously could mimic the effect of a small nuclear weapon. One scenario would be to mount multiple sorties to cover the entire suspected facility area. In a more sophisticated approach, the military is now developing a "small-diameter bomb" that packs several penetrating missiles into the payload of a single aircraft--essentially, an underground version of the ubiquitous cluster bomb. Extending the small-diameter-bomb concept to missiles the length of the GBU-28 would enable simultaneous delivery of as many as 24 penetrating missiles, at least several of which would be expected to penetrate the facility.

Still other options are available. If the facility were operating, then conventional electromagnetic pulse weapons--recently added to the U.S. arsenal--might be applied to destroy or disable equipment inside. Because an electromagnetic pulse can easily travel down a bunker's power and ventilation ducts, equipment inside would be vulnerable to attack. Such weapons could be delivered by cruise missile.

Advocates of new tactical nuclear weapons have tended to focus shortsightedly on simple destructive power.

In an indirect approach to rendering the facility useless, cruise missiles could be used to temporarily block its entrances. It also would be possible to establish a "no-personnel zone" or "no-vehicle zone" around the facility. A range of intelligence assets, such as spy satellites, would be trained on the area surrounding the complex, and any attempt to move material into or out of the facility would be stopped. Although the facility itself might continue to produce weapons, those weapons could not be removed and used on the battlefield. These approaches would be limited by the need to continually devote assets to a single facility or to mount repeated attacks; if there were many simultaneous targets of concern, the method might not prove feasible.

In each case of applying conventional weapons, collateral damage due to chemical dispersal would be minimal outside the facility. Inside, chemical agents would be dispersed, but U.S. troops inspecting the area could mitigate the dangers from these by wearing protective gear.

Agent defeat

Proponents of nuclear weapons for attacking stockpiles of chemical and biological agents, called "agent defeat weapons," typically argue that the biological or chemical fallout produced by a conventional explosive attack can be more deadly than the fallout produced by a nuclear weapon. This argument misses two crucial points: In many cases, nonnuclear agent defeat payloads can avoid spreading chemical and biological fallout; and the fallout from a nuclear attack, though perhaps smaller than the potential biological or chemical fallout, is still prohibitive.

Consider a hypothetical example from Iraq, which is suspected of retaining stockpiles of weaponized anthrax and is known to use hardened bunkers extensively. A typical bunker might be 20 meters in height and cover an area measuring 400 square meters, have walls that are five meters thick and a roof of reinforced concrete, and be buried under five meters of earth. Built during the absence of United Nations weapons inspections, the bunker's existence has become known to U.S. intelligence through satellite imagery captured during its construction. It is believed to contain several tons of anthrax in storage barrels, though in the absence of a continuing ground presence, this cannot be confirmed.

A 20-ton penetrating nuclear weapon (if it were developed) detonated at the floor of the facility would incinerate its contents, preventing the dispersal of anthrax. But it would also spread nuclear fallout. Deaths from acute radiation poisoning would be expected as far as one kilometer downwind. People nearer than four kilometers downwind would, if not evacuated quickly, receive a radiation dose greater than that received by a nuclear worker during an entire year.

Nonnuclear payloads might, however, spread less collateral damage while avoiding political problems. A penetrating bomb carrying a fragmenting warhead and incendiary materials could be used. The warhead would break the anthrax out of any exposed containers, and the heat from the incendiary materials would neutralize the anthrax. Containers that were heavily shielded might not break open, but although the anthrax would not be destroyed, neither would it be released. The bunker would remain intact.

Alternatively, a penetrating bomb carrying submunitions and neutralizing chemicals could be used. The submunitions would spread throughout the bunker and release the anthrax from its containers, even if it were stored behind barriers, and the neutralizing chemicals would render the anthrax inert. The bunker would probably remain intact, although it could be breached if it had been poorly constructed.

U.S. planners may not want to directly attack the bunker. Instead, a watch could be placed on the facility using satellite imagery coupled with armed unmanned aerial vehicles. Anyone or anything attempting to enter or leave the bunker would be destroyed, making the anthrax inside unusable.

Deep burial

Among proponents of new nuclear weapons, the most consistent error is the assumption that they would be silver bullets, leaving no underground facilities invulnerable to their effects. But such is not the case. Even the two-megaton B-83 bomb, the highest-yield weapon in the U.S. arsenal, would leave unscathed any facilities buried under more than 200 meters of hard rock. In contrast, functional defeat approaches--sealing off entrances rather than directly destroying the bunker--have no depth limitations.

To better understand this, consider North Korea's Kumchangri underground complex, which was once suspected of housing illicit nuclear weapons activities. The depth of the facility, built into the side of a mountain, is not publicly known, but its main chamber may quite possibly be deeper than 200 meters, putting it out of the range of even megaton-sized, earth-penetrating nuclear weapons. Even if the facility were only 150 meters underground, a one-megaton penetrating nuclear weapon would be required to destroy it, and the resulting nuclear fallout would have enormous consequences. If the wind were blowing southwest, then the North Korean capital of Pyongyang, 80 miles away, would have to be evacuated within hours of detonation to prevent the death of more than 50 percent of its residents from radiation poisoning. If the wind were blowing north or northwest, then residents of several large cities in China would have to be evacuated immediately. And if the wind were blowing south, then residents of several large cities in South Korea, as well as U.S. troops stationed in the DMZ, would have to be evacuated within hours to avoid numerous radiation deaths.

Alternatively, regardless of the facility's depth, military planners could seek to disable rather than destroy the facility. Cruise missiles could be used to collapse entrances to the bunker. Entrances, however, might be reopened quickly, requiring repeated sorties to keep the facility closed. Thermobaric weapons, which debuted in Afghanistan, could be used to send high-pressure shock waves down the tunnels, possibly destroying equipment inside the facility.

An "information umbrella" approach also might be applied. The United States, possibly together with allies, would declare that no North Korean vehicles would be allowed to come near the facility. This curfew would be monitored using surveillance assets, and any vehicle attempting to enter or leave the facility would be destroyed.

Misguided federal efforts

Despite the limitations of nuclear capabilities, some policymakers are marching ahead. For the past year, Congress has been focused on a new weapon system called the robust nuclear earth penetrator (RNEP), a modification of either the B-61 or B-83 bomb that would have improved earth-penetration ability. The Department of Energy (DOE), in its 2003 budget request, asked for funding to begin a three-year, $45 million "feasibility and engineering" project that would include "paper studies" of the RNEP and might possibly "proceed beyond the mere paper stage and include a combination of component and subassembly test and simulation," according to John Gordon, then the administrator of the DOE's National Nuclear Security Administration.

This effort would be misguided. Misunderstanding of weapons technology and engineering has consistently marked congressional debate over the RNEP, and any further discussion first requires setting the record straight. Some observers have incorrectly characterized the RNEP as a low-yield "mini-nuke," implying that it is more usable than other nuclear weapons. But as Rep. Curt Weldon (R-Penn.) pointed out correctly during House debate, the RNEP is not a mini-nuke--indeed, it is a very large, clumsy weapon.

Yet confronted with the observation that nuclear weapons are militarily useless, many people embrace this as a virtue, arguing that these weapons are for deterrence, not for warfighting. Their claim, however, is based on dubious deterrence arguments left over from the Cold War. At the core of this claim is the contention that deterrence works only when the United States threatens what the enemy values most; that many enemies are so foreign that it is impossible to reliably judge what they value; and that the United States therefore must be able to robustly threaten every asset of theirs.

This argument is not compelling. Consider, for example, the recent debate over the value of deterrence in confronting Iraq, in which analysts and politicians split into two camps. One faction claimed that Saddam Hussein is fundamentally undeterrable, and thus the United States must disarm him. The other argued that Saddam is rational and deterrable, and thus the United States should not attack. No camp argued that it is impossible to deter Saddam only because the United States currently has no earth-penetrating nuclear weapons that place his underground bunkers at risk--and that Iraq therefore should be attacked.

Some proponents of the RNEP have sought to dodge detailed debate over its utility, arguing that without the proposed feasibility study it is impossible to determine whether the weapon would be useful. But to understand what a feasibility study can and cannot accomplish, consider one of the leading RNEP proposals: modifying the two-megaton B-83 bomb to add ground-penetrating capability. Using basic physics, it is possible to estimate the weapon's potential penetration depth and, in turn, place upper and lower bounds on fallout hazards and destructive capability. Indeed, many people, including myself, have made such estimates--and doing so certainly does not require a multiyear, multimillion-dollar feasibility study.

The proposed study would build on these basic calculations, looking more carefully at engineering limitations and narrowing the basic estimates. For example, laboratory scientists might conclude that although it may be scientifically possible to build a missile capable of achieving a 900-meters-per-second earth impact, it would be impossible to engineer the missile to withstand such an impact shock. This failure would reduce the estimate of earth penetration, decreasing the projected destructive potential while increasing the expected fallout hazard.

Congress must more seriously consider the advantages of nonnuclear options.

Given preliminary scientific estimates, military and political decisionmakers--before initiating the research project--should be able to come to one of three conclusions regarding the RNEP:

First, either its maximum destructive capability is too small, or the minimum fallout hazard too large, to make further development worthwhile. In this case, the proposed study is unnecessary.

Second, its minimum destructive capability is large enough, and the maximum fallout hazard small enough, to warrant development. In this case, the argument for a feasibility study as a preliminary step is a ruse; the more honest position would be to immediately endorse full development and deployment of the new weapon.

Third, depending on where within the preliminary estimates the destructive capability and fallout hazards fall, the weapon may or may not be useful. In this case, a feasibility study would be essential to refining the estimates and making a decision on proceeding.

The third possibility, however, is unlikely. The fallout hazard that such a bomb would produce is enormous. It could quickly kill people hundreds of kilometers downwind, and would contaminate cities even further away. There are only two reasonable conclusions from this exercise: Either one can decide that this is excessive collateral damage and oppose the RNEP, or one can decide that no collateral damage is excessive for this mission and support the RNEP. To claim a need for further engineering study of the robust nuclear earth penetrator is disingenuous.

Broader discussion needed

Though many people now maintain that the military has little interest in tactical nuclear weapons, policymakers continue to contemplate developing and deploying them. This will, unfortunately, remain the natural state unless political decisionmakers force a change. Although designers of nuclear weapons have a built-in imperative to seek nuclear solutions to military problems, there is little to be gained by the uniformed military from pushing back. It falls to Congress to actively solicit the advice of military thinkers on the utility or lack thereof of new tactical nuclear weapons.

To date, only the Senate Committee on Foreign Relations has devoted substantial hearing time to tactical nuclear weapons. But these weapons have not only political but military liabilities. To explore these issues, the House and Senate Armed Services Committee should convene hearings on the robust nuclear earth penetrator and on tactical nuclear weapons more broadly. The committee should solicit input from retired military officers and from individuals who have spent time understanding both the nuclear and nonnuclear options. Only by making direct comparisons will policymakers be able to find agreement on a way forward.

Recommended reading

Thomas Dower and Joseph Howard II, "Countering the Threat of the Well-Armed Tyrant: A Modest Proposal for Small Nuclear Weapons," Strategic Review, Fall 1991.

Geoffrey Forded, "USA Looks at Nuclear Role in Bunker Busting," Jane's Intelligence Review, January 2002.

House Policy Committee, Subcommittee on National Security and Foreign Affairs, An Agenda for the Nuclear Weapons Program, 2003.

Robert Nelson, "Low-Yield Earth Penetrating Nuclear Weapons," Science and Global Security 10, 2002.

U.S. Department of Energy and Department of Defense, United States Nuclear Posture Review, 2002.

Stephen Younger, Nuclear Weapons in the Twenty-First Century (Los Alamos National Laboratory, 2000).

Michael A. Levi ( is director of the Strategic Security Project at the Federation of American Scientists in Washington, D.C. This article is adapted in part from his paper Fire in the Hole: Nuclear and Non-Nuclear Options for Counterproliferation (Carnegie Endowment for International Peace, 2002).