Those Magnificent Men and their Atomic Machines
The Nation's Cockpit: The DUCC and Decision-Making Under Nuclear Attack
In 1957, the Soviet Union launched Sputnik, and the world changed forever.
Sputnik was not just a technological and political triumph; it was a military threat. A rocket that can carry a satellite into orbit can also carry a nuclear weapon to the United States. The American military and public had known of the theoretical possibility of a nuclear-armed intercontinental rocket since 1945, but the threat had been abstract, unreal, until the Soviet's bleeping aluminum ball passed overhead.
Before the ICBM, the United States could rely on the Distant Early Warning RADAR line to provide at least two hours of warning time of an attack. There would be enough time for the president to be woken from sleep, briefed, and evacuated. More importantly, there would be enough time for the president to decide – to decide if the US was going to war. In two hours, equipment can be checked and errors corrected; explanations can be demanded of the Soviet embassy. Even if a mistake was made, the Strategic Air Command B-52s took hours to reach its targets and could be recalled.
An ICBM would arrive 15 to 25 minutes after being spotted on RADAR. Fifteen minutes to decide the fate of millions. And ICBMs cannot be called back – once the president ordered a counterstrike, it could not be rescinded.
This was wholly unacceptable. What if there was a mistake? What if the launch was accidental, or a rogue general, or even a third party trying to provoke a war? The president had to survive long enough to reach a measured decision. The issue was not the personal survival of the president. The issue was the continuity of the national decision-making system, ensuring that SAC would be neither paralyzed nor forced into automatic retaliation.
Two solutions were considered: mobility and hardness. Mobility meant keeping the president on the move, on plane or train or ship, so that the Soviets could not find and kill him. Hardness meant burying the president, deep under ground, deeper than even a nuclear weapon could reach.
The ultimate “hard” proposal was the Deep Underground Command Center, or DUCC. Buried a full 3,500 feet under Washington, D.C., the DUCC was designed to survive multiple direct hits from 300 megaton nuclear weapons. Deep under the charred remains of the nation's capitol, the president and his advisors could assess the situation, communicate with our allies, and direct our military forces to an appropriate response.
Presidential Survival Before the ICBM
The first study of presidential survival during nuclear war started by the Joint Chiefs of Staff in 1946 and finished in 1948, a year before the first Soviet atomic test. Taking the study as a starting point, the Joint Chiefs proposed building a Joint Command Post outside Washington, D. C., where military and political decision-makers could move to during a nuclear war.
At the time, atomic bombs were relatively few in number and difficult to deliver. Planners expected war to begin with an “atomic blitz” against Soviet cities and factories, which would last about a month. Given their rarity, it was thought the only cost-effective targets of an atomic weapon would be cities and industrial centers. The atomic blitz would be followed by a ground war lasting years.
The JCP – later renamed the Alternate Joint Communications Center, AJCC, to distinguish it from the Pentagon's Joint Command Center – would hold enough staff to allow the president to plan and organize the war. Since it was located well outside Washington, it would be relatively safe from atomic attack; it would be too small a target to justify using an atomic bomb. Disagreements over service responsibilities and other issues delayed construction, but ground was broken in 1950 at Fort Ritchie, and the facility was finished in 1953.
Figure 1: The AJCC Today
The AJCC was initially only hardened to withstand 30 psi of overpressure. Overpressure is a measure of blast strength – 1 psi will shatter glass, 3 psi will destroy wood frame housing, and 5 psi will destroy anything short of reinforced concrete. 30 psi may sound like a lot, but modern missile silos are hardened to withstand up to 2,000 psi. Since the Distant Early Warning RADAR line provided two hours of warning time, and it took thirty minutes to reach the AJCC by helicopter, there would be plenty of time for the president to evacuate if an attack was detected.
Besides the AJCC, other hardened facilities were built in the 50s. The military built hardened bomb shelters at the White House (70 psi) and at Camp David (100 psi), and a civilian equivalent to the AJCC, called High Point, at Mount Weather (50 psi). But the AJCC was expected to house the president in the event of war.
However, it was clear that these were only temporary solutions. The development of the hydrogen bomb in 1952 increased nuclear weapons yield from kilotons to megatons; a near-miss by a high-yield hydrogen bomb could do as much damage as a direct hit by an atomic bomb. Not only were weapons getting bigger, but the number of weapons was expanding as well, allowing the Soviets and Americans to expand target lists beyond cities to include leadership bunkers.
But the responsibilities of the president in a nuclear war had also shrunk. The atomic blitz had grown into Massive Retaliation, which promised the destruction of Soviet society from the air, making a lengthy ground war unnecessary. The president only had to survive long enough to give the order. In fact, the SIOP – the official war plan – assumed an unlimited attack on the enemy's society and military. Even if the president ordered a limited response, it might not be possible to launch anything less than a total attack – there simply would not be time to draw up the target lists.
No Warning: Hardness versus Mobility
The ICBM changed everything. The new Ballistic Missile Early Warning System gave only 15 to 25 minutes of warning of a ballistic missile strike – nowhere near enough time for the president to reach shelter. And BMEWS couldn't detect submarine-launched ballistic missiles (SLBMs) at all. Even after the new infrared detection satellites were deployed, warning times for SLBMs might be less than fifteen minutes. If the president was killed, it would take time to determine who the new president was, find them, brief them, and transmit their instructions – time that would not be available in the chaos of war.
One possible solution to this was to make nuclear response automatic – to authorize the strategic nuclear forces ahead of time to retaliate against an enemy attack. This was Eisenhower's solution – if SAC commanders found themselves cut off and under attack, they were authorized to execute their war plan without further orders.
Figure 2: Excerpt from Eisenhower Nuclear Predelegation Letter
This was not acceptable to the new Kennedy administration. Supremacy of civilian authority was too important. Besides, reflexive responses left no room for flexibility, only automatic devastation. And flexibility was one of the buzzwords of the new Secretary of Defense, Robert McNamara.
Flexible Response was the new strategy, replacing Massive Retaliation. The rigid SIOP was to be upgraded to a menu of options, giving the president the ability to tailor the American response to the level of attack. Instead of an atomic blitz, we would try to fight a Limited War, restraining escalation at a level below city bombing.
But Flexible Response required the president to survive long enough to decide what the response would be – and not just our response to the initial attack, but to the enemy's response to our response, and to their response to our response to their response, and so on. Not only did the president have to survive, he had to remain in contact with the armed forces and the allies – and with the enemy. He needed to be able to propose a ceasefire, to make threats, to negotiate. He needed a staff to interpret and analyze and display information, and to convert his decisions into orders for the military. In short, wherever the president ended up, he would need hundreds of staffers and secure, uninterruptable communications links.
The military proposed the National Military Command System (NMCS) in 1961 as a solution. NMCS had four nodes. The first, used in ordinary conditions, would be the National Military Command Center (NMCC), an upgraded, unhardened Pentagon war room. The second node would be the Alternate National Military Command Center (ANMCC), which would be the AJCC upgraded to withstand 140 psi overpressure. The third would be the National Emergency Airborne Command Post (NEACP), a trio of Boeing EC-135Js –
747s 707s fitted with enhanced endurance and advanced communications
equipment. The EC-135Js were later replaced with E-4Bs.
Figure 3: NEACP Today
And the fourth would be the National Emergency Command Post Afloat (NECPA), a pair of old warships, the cruiser Northampton and the light carrier Wright, fitted out as command centers.
Figure 4: USS Wright in 1963
The Army also proposed a train-based command post, but it was decided this did not offer any advantage over NEACP and NECPA. The NMCS contained a mix of hardened and mobile facilities, covering all the bases... Except that it took fifteen minutes for the president to reach the NEACP, and ten minutes for the KC-135 to escape the blast radius of a nuclear airburst over the airfield.
The US could count on only fifteen minutes warning at most of a missile attack via submarine. NECPA was even further away via aircraft. ANMCS might or might not be survivable against existing ICBMs, but Soviet missile accuracy was only going to improve, and it would be vulnerable sooner or later.
NEACP and NECPA could ensure the president's survival, but only if he evacuated Washington before tactical warning of an attack. And presidents had shown no inclination to leave the capital during a period of tension, a point driven home hard by the Cuban Missile Crisis in 1962.
DUCC and Cover
In 1963, the Defense Department proposed a solution: the Deep Underground Command Center, or DUCC.
Studies of a DUCC had been percolating since at least 1962. But it was in 1963 that the proposal reached the president's desk, with the approval of both Secretary of Defense McNamara and Secretary of State Dean Rusk. The DUCC would be a capsule buried 3,500 feet under the Pentagon. Two versions were proposed, a “Moderate” version offering space for 300 people, and an “Austere” version with space for 50, and which could be expanded to the Moderate version if necessary. Elevators would descend from the White House, Pentagon, and State Department to the facility depth, where horizontal tunnels would lead to the capsule.
Figure 5: Diagram of DUCC System
Officials could descend to the DUCC without leaving their buildings, so there would be no external signs of evacuation – the president could take shelter without the political consequences of visibly leaving Washington, D. C. It was even suggested that officials on the presidential succession list might spend one day a week in the DUCC. That way, no matter what, at least one successor would survive, and be in a position to quickly reestablish control of the military.
It was claimed that the system could withstand multiple direct hits by 200 to 300 megaton nuclear weapons, or by 100 megaton weapons that penetrated to a 70 to 100 foot depth. For comparison's sake, the largest nuclear weapon ever detonated had a yield of about 50 megaton, and the largest ever produced in numbers was about 25 megatons. But, in the early 60s, nuclear weapon yields had been steadily growing since their introduction, and 300 megatons seemed like a pessimistic but reasonable extrapolation of Soviet capabilities in the early 70s.
Few details are available on the capsule itself, but some extrapolation is possible based on Army engineering manuals and similar but less extreme facilities. We know the austere version would offer only 5,000 square feet of space, equivalent to a 10' x 10' square for each occupant. The moderate version would be slightly better at 50,000 square feet, or a 13' x 13' square per occupant. The occupied area would be contained within a larger chamber of double the area, and would probably be mounted on gigantic springs to ride out ground shock, which would be the main threat to the system.
It would be theoretically possible to blast out enough dirt to physically breach the DUCC. But a 300 megaton weapon digs only a 967 foot deep crater in granite, requiring four such bombs landing precisely on top of each other to dig out a breach. This sort of accuracy would be difficult even for modern ballistic missiles, although not impossible.
The main damaging mechanism would be the shock wave that is generated in the rock, which would act similarly to an earthquake. Ground shock could directly injure or kill the DUCC's occupants – hence the springs – or it could cause spalling, in which fragments of the chamber roof fall off. To prevent this, the tunnels would probably be lined with cast iron or even stronger materials.
Supplies would be stashed in the capsule for 30 days of “buttoned-up” occupancy, which would hopefully be enough time for surface radiation to cool to survivable levels. Although the main elevator access shafts would probably be collapsed by bombing, multiple tunnels would provide hardened exits outside the likely attack area. In addition, unspecified “hardened communications” would be provided.
In the event of a missile warning, the president and other key officials would reach the protected depths via elevator in only ten minutes, and the capsule in another five. This would be ten minutes less than the time to reach safe distance aboard NEACP. Nonetheless, the Joint Chiefs of Staff were, at best, unenthusiastic about the plan.
In the view of the JCS, the main failing of the DUCC was that it was simply too small. Even the moderate version did not have enough space for an adequate staff. While the president might survive, he would not have the personnel with him to properly analyze the situation and disseminate orders. The JCS estimated that, of the 300 people that could be crammed into the moderate DUCC, at least 175 slots would be filled with personnel for maintenance, communications, housekeeping, and otherwise just keeping the shelter running. The JCS themselves would require a staff of 50 to execute orders received from the president. That left only 75 slots for the president, his advisors, and civilian personnel from the Defense Department, State Department, and other key organizations.
Aside from space constraints, the analysis of the DUCC's survivability was based on theoretical calculations that, while consistent with what was known about nuclear weapons effects, had obviously never been tested. Not only that, but the JCS had serious doubts about the survivability of the DUCC's communications antennas, which would have to be close to the surface. The Chiefs concluded that, all in all, the austere version might be worth building as a last-ditch survival shelter, but there was little point to protecting the president if he would no longer be in command of the armed forces.
There was more. The DUCC would be expensive. Even the austere version would cost $110 million over a period of almost four years, while the moderate version would cost $310 million over five and a half years. And that was only construction costs – even more would be needed to operate the system, to staff it and maintain it and keep it stocked with supplies. The budget for the NMCS program, for NEACP and NECPA and so on, was only $850 million over five years. The DUCC would be competition for the JCS' preferred systems, and it seemed unlikely that Congress could be persuaded to add money to the budget for presidential survival. In particular, the ANMCS would be phased out once the DUCC was completed, even though the new system offered far fewer seats at a much higher cost.
The JCS also had, perhaps, one more objection. The military had grown used to ordering its own affairs in the 50s. Secretary of Defense McNamara and his team of civilian “Whiz Kids” had intruded into what they saw as military business, canceling favored development programs and ordering changes in the war plans. McNamara felt that, since the president was responsible for the military's actions, the president should be the one who gave the orders – and that held in time of war as well as peace. Rather than just issuing the “go code” and allowing the air force to fight as it saw fit, the president should control the nature and degree of the military retaliation. Harold Brown, one of McNamara's Whiz Kids and, at the time, head of Defense Research & Engineering, actually suggested the JCS' opposition might be based on a desire that the civilian leadership perish after giving approval for a nuclear strike. “Basically,” he said, “the Chiefs probably aren't interested in having the civilian command survive. If we were to come to a war, they would only get in the way.”
But, whatever the JCS thought, the Secretary of Defense, the Secretary of State, and the President liked the idea. Tentative approval was given in December of 1963, pending funding from Congress.
Would it Have Worked?
This is obviously a question that can't really be answered, short of building a DUCC and dropping some hydrogen bombs on it. But we can make some extrapolations.
The Atomic Energy Commission conducted two underground nuclear tests in granite as tests of superhard structure construction, HARDHAT in 1962 and PILEDRIVER in 1966. Granite is also present in the Washington area, and would be the logical stone to build the DUCC in. Pressure sensors were placed in the ground near the bombs hardened tunnels similar to those used in superhard facilities were dug nearby.
The HARDHAT test showed that tunnels generally collapse when subjected to peak shockwave pressure equal to the unconfined compression strength of the rock. The peak shockwave pressure is proportional to the yield of the bomb times a coupling factor, raised to the power of 0.3. The coupling factor determines how much of the bomb's energy is converted into groundshock. In the underground HARDHAT test, the coupling factor was about 100; in a surface burst, the coupling factor will be 1.
Most of the tunnels in the HARDHAT and PILEDRIVER tests collapsed under 1 to 1.5 kilobars of peak stress; however, some extremely hardened tunnels survived 3 to 4 kilobars. Given the importance of the DUCC, it is likely the tunnels would be hardened to the higher standard.
The HARDHAT test shot created peak pressure of 1 kilobar at a radius of about 450 feet from the point of detonation. HARDHAT was a 5 kiloton detonation with a coupling factor of 100. A hypothetical attacker using a 300,000 kiloton detonation with a coupling factor of 1 would create 1 kilobar pressure about 3,066 feet below the missile impact point.
We can attempt a second calculation using data from the WinGS code for a 5.6 megaton surface burst. 1 kilobar of peak stress is induced by a 5.6 megaton burst at a depth of 350 feet. Adjusting for yield, this suggests the boundary of 1 kilobar will be at a depth of 3789 ft. Although this is below the DUCC's depth, the DUCC would presumably be engineered to withstand 3 to 4 kilobars, allowing it to ride out a 1 to 1.5 kilobar stress.
The wide variation between these two calculations should show just how extremely tentative these figures are. Nonetheless, it appears not implausible that, with proper tunnel lining and shock absorbers, the DUCC could indeed survive a 300 megaton surface burst.
But could it survive more than one? Each burst will excavate a crater 967 feet deep. A second 300 megaton bomb landing in the crater of the first will create higher stress at the DUCC's depth. At this point we are getting well beyond the realm of napkin calculations, however. Nonetheless, on the face of it, the DUCC concept does not appear impossible.
In early 1964, the Defense Department requested funds to build the DUCC as part of the army's military construction appropriations. The House Armed Services committee was aware of the Joint Chiefs' dislike of the plan, and decided not to allocate new funds to the program. The House also rejected the diversion of other authorized funds to feasibility studies for the concept.
The Pentagon continued to study the concept, and in early 1965 the Office of the Director of Defense Research & Engineering developed a list of possible sites for the DUCC. While the House continued to reject fully funding the program, they did agree to allocate $4 million in for FY1966 to “more fully develop plans and to again present the actual construction authorization request” the next year.
With the House continuing to reject full funding of the program, the civilian officials in the Defense Department began to lose interest in the project. In March of 1966, McNamara decided to pull the plug, and ordered the DUCC planners reallocated to other tasks.
There is very little documentary evidence about the end of the DUCC, but we can make certain deductions about McNamara's motives in addition to Congressional intransigence. However, these are somewhat speculative.
By 1966, Flexible Response was falling out of favor, and Assured Destruction was the new buzzword. The problems of limited war seemed far less solvable to the Whiz Kids when in office than they had seemed while ensconced at the RAND corporation. In particular, it took two sides to fight a limited war, and it seemed unlikely that a sort of Marquis of Queensbury rules could actually hold in a nuclear war.
Rather than try to limit a nuclear war, Assured Destruction focused on preventing one from breaking out, by ensuring both sides had the capability to launch devastating counterattacks even after absorbing a nuclear first strike. The new approach placed less emphasis on the president making decisions while under attack – in fact, it placed very little emphasis on how to fight a nuclear war at all. Strategy would no longer be an analysis of how to fight a nuclear war; it be an analysis of how to prevent one.
Assured Destruction made the dread “nuclear Pearl Harbor” surprise attack seem less and less likely. The danger of Mutually Assured Destruction would act as an impenetrable barrier to a sneak attack. Instead of a “bolt from the blue,” a nuclear war would probably begin only after a period of rising tension.
But, if nuclear war would only come during a crisis, then there was no need for the DUCC. NEACP and NECPA were far more survivable than the DUCC if the president was on them. The advantage of the DUCC was that the president could enter it in the time between the detection of a ballistic missile and its impact. If the president had even thirty minutes warning, he could reach NEACP and be safely in the air before the missile hit. With the fading of the surprise attack threat, the DUCC lost its primary justification.
There was probably another reason as well besides the shift in strategic thinking. The Vietnam War was heating up, and more and more focus and money was going into the jungles of southeast Asia. Schemes for presidential survival in a future nuclear war seemed less important compared to the demands of a war being fought in the present.
So, rather than spend the money to build a DUCC, the Pentagon invested in upgrading its existing systems. The communications and endurance of the NEACP and NECPA were improved, and the ANMCC was hardened to 400 psi overpressure capability. And the DUCC faded into history.
The Last Chapter
But that's not, quite, the end of the story. The idea of a superhard command post, surviving nuclear war through sheer depth rather than mobility, continued to simmer in the defense community.
The specifics are still classified, but we know that in 1977 – twelve years after the end of the DUCC – the Protective Design Center of the US Army Corps of Engineers was involved in something called the “Alternate Military Command Center Improvement Program.” Unfortunately, the entirety of our knowledge can be reduced to two paragraphs and one picture. This was posted on the PDC's website, and hastily removed after an email was sent to the center inquiring about the project:
“Special Projects Office (later to become the Protective Design Center) was created in 1977 to work on a classified Department of Defense program. The Alternate National Military Command and Control Center Improvement Program involved criteria development, studies, and preliminary design of a deep underground highly hardened and survivable command and control center. The center included separate structures for command personnel, power, fuel, and water. Over 3 miles of air entrainment tunnels were required as well as access shafts to the surface.
“Although canceled in 1979, the experience, expertise, and leadership in protective design and classified programs that Special Projects gained from this work brought other unique projects and major programs to the District.”
Figure 6: Model of Capsule on Shock Springs
Literally nothing else is known about this program. Probably there were other, even more deeply buried studies, which we still do not know of. The DUCC may have died, but the idea didn't.
Freedman, Lawrence. The Evolution of Nuclear Strategy. International Institute for Strategic Studies, 1981, 1983.
Sturm, Thomas A. The Air Force and the Worldwide Military Command and Control System. USAF Historical Division Liaison Office, 1966.
Wainstein, L., C. D. Cremeans, J. K. Moriarty, and J. Ponturo. The Evolution of U. S. Strategic Command and Control and Warning, 1945–1972. IDA Study S-467.
“The Deep Underground Command Center (Proposed).” A Secret Landscape: America's Cold War Infrastructure.
Foreign Relations of the United States: Johnson Administration, 1964-1968, Volume X: National Security Policy. Documents No. 3, 4, 52, 77, 110. (Link excludes doc. no. 110).
Interview with Carl Kaysen by Marc Trachtenberg and David Rosenberg. First session, August 3 1988.
Effects of Nuclear Weapons
Butkovich, Theodore R. “Calculation of the Shock Wave from an Underground Nuclear Explosion in Granite.” Third Plowshare Symposium: Engineering with Nuclear Explosives, 1964.
Effects of Nuclear Earth-Penetrator and Other Weapons. Committee on the Effects of Nuclear Earth-Penetrator and Other Weapons of the National Research Council. The National Academies Press.
Figure 1: Center for Land Use Interpretation. From “Bunkers Beyond the Beltway." Used under an Attribution-Noncommercial-Share Alike 3.0 Creative Commons License. The image was cropped and shrunk, and the resulting work is also under an Attribution-Noncommercial-Share Alike 3.0 Creative Commons License.
Figure 2: Excerpt from document released by National Security Archive. Found at link.
Figure 3: US Government. Found at link.
Figure 4: US Government. Found at link.
Figure 5: Author. Figure 5 is hereby released under an Attribution-Noncommercial-Share Alike 3.0 Creative Commons License, on the condition that any page displaying this picture must contain a link to this page.
Figure 6: US Government. Found at link.