Jeffrey LewisFull Cost Accounting & the B53

Workers at PANTEX dismantled the last B53 nuclear gravity bomb, a nine-megaton behemoth affectionately known as the “Crowd Pleaser.”

It sheer size and unfathomable explosive power — 600 Hiroshimas! — are so difficult to comprehend that one can’t help think about the Cold War and ask: What the hell were they thinking?

That’s not what interests me, though.  What really interests me is not that the US weaponeers built the B53 during the Cold War, but they put so little thought into the day when, sooner or later, someone would have to take it apart.

The B53 was, apparently, a nightmare to disassemble.

NNSA had to develop new “tooling” and processes to dismantle the B53.  NNSA Administrator Tom D’Agostino conveyed the idea when he said “You don’t attack these things with a screwdriver and a crescent wrench.” The picture at the top of post shows one element of the specialized tooling — the bright orange object fastened to the outside of the bomb. If you look at some of the reports from PANTEX, posted on the Defense Nuclear Facilities Safety Board Website, you can see descriptions of many tooling issues. (Your search string is: “B53 SS-21”  This was no mean feat, not least because the US withdrew the bomb from the stockpile in 1997 because it was not safe.

(An aside: There is a whole bunch of insipid commentary about this dismantlement being part of the “Prague Agenda,” whether for good or ill.  The dismantlement of the B53 has nothing to do with the Prague Agenda, Global Zero, or anything else.  The weapons was retired in 1997 because it was unsafe. Whether you think the US should have zero, 1,000, or 10,000 nuclear weapons, no one thinks the United States should have B53s.)

A significant part of the problem is that the US did not, during the Cold War, spend very much time or money on the problem of how to make nuclear weapons that were easy to dismantle.  This was a major talking point for Bush Administration officials making the case for the Reliable Replacement Warhead, who were quick to point out that our legacy warheads “were not designed to be taken apart.”

The B53, given the sheer size and hazard of the thing, was unusually problematic.  When the NNSA announced in October 2010 that it was finally authorizing the dismantlement of the B53 after several years of delays, this was a big news.

This was a very expensive dismantlement.

The specific figure is, unfortunately, classified, but the FY2010 DOE budget for dismantlement jumped from $53 million to $96 million in FY2010. When asked why there was so much money in the FY2010 budget request, NNSA Administrator Tom D’Agostino explained “We needed to do our most complex weapons system dismantlement — work on the W84, the B53, for example — particularly the 53. It’s taking a long time. And so we feel that by the time FY ’10 is done we’ll have finished the authorization basis work, the tooling, the methods and approaches needed to take apart that warhead. And so by FY ’11, by the time FY ’11 starts, we’ll be in the business of actually taking apart that warhead in and of itself.” [Emphasis mine.]

NNSA probably has a cost estimate for the B53 dismantlement.  I’d love to see it.  Sadly, the official “dismantlement” budget that DOE releases seems incomplete. I plotted the number of warhead dismantlements by year against the total spending in the official dismantlement budget — the result is a cloud, even with time lags and other forms of data water-boarding.  There just isn’t any correlation.  Senior NNSA officials, in a conference call, “drew a distinction between the funding level and the number of warheads dismantled.” I see why.  The fact that dismantlement rate does not correlate to funding levels suggests to me that the bulk of the costs are not visible (perhaps to prevent clever people from inferring classified information about dismantlement rates from unclassified budget figures).

Before the mid-1990s, there wasn’t even a separate budget line for dismantlement activities.

When the Office of Technology Assessment asked about dismantlement costs in the early 1990s, they were told the direct costs were about $25-30 million dollars.  (That is about $40-50 million in 2010 dollars, which is line with the 2000-2009 average of $47 million/year for dismantlement activities. How is that for consistent?) OTA was skeptical about that figure, however, noting that “The total operating budgets for the two sites most engaged in dismantlement— Pantex and Y-12—are $240 and $460 million, respectively, for FY 1993. Managers at each of these sites have said that at least two-thirds of their current efforts are devoted to the dismantlement mission or related work.”Today, the operating budgets for PANTEX and Y-12 are about $650 million and $935 million.  If the two-thirds rule still holds, then dismantling warheads is a much more expensive process than the official budget suggests.

Nuclear weapons mean never having to clean-up after yourself.

I mention this for two reasons.  First, it is sort of crazy that we don’t really know how much it costs to dismantle a warhead. (And if we do, that’s a recent phenomenon.)  Second, it is really crazy that the costs associated with dismantlement aren’t part of the original funding.

It turns out that the nuclear weapons complex simply doesn’t do “full cost accounting.”  If you build a municipal solid waste facility, for example, the “back end” costs are part of the consideration.  After all, part of the cost of any activity is cleaning up after one’s self (or one’s generation).  That’s not, apparently, true for nuclear weapons or nuclear weapons facilities.  The cost estimates for RNEP and RRW, for instance, only described what it might cost to make the weapons.  Not what it might cost to take them apart someday.  The same goes for all the big ticket infrastructure items in the Obama Administration’s modernization of the nuclear weapons complex like the Uranium Processing Facility. The clean-up costs get passed along to the future, quite possibly including individuals born after the weapons or facilities were retired.

Of course, that’s bad management and, from a generational perspective, a little selfish. It is irresponsible for policymakers to simply pay for the construction of nuclear weapons and related facilities, while leaving dismantlement and clean-up costs to future generations.

Accounting for the full cost of nuclear weapons is the responsible thing to do.

My friend Joe Cirincione has taken a lot of guff for his claim that the Obama Administration plans to spend $700 billion on nuclear weapons over ten years.  Some people argue that Joe should exclude from his estimate all the “good” money spent on environmental clean-up and nonproliferation. I agree with Joe — environmental remediation and securing nuclear materials are part of the cost of having nuclear weapons.  It wasn’t wrong for Joe to include those costs in his estimate; it was wrong for Cold War policymakers to exclude them and let future generations foot the bill.

Walter Pincus, in October 2010, wrote a long, interesting article about the B53 as a cautionary tale.  “It’s a tale that should have lessons for today as the Obama administration considers spending more than $180 billion over the next decade modernizing the nation’s nuclear stockpile and replacing the intercontinental ballistic missiles, strategic submarines and aircraft that would deliver them,” Pincus wrote. “There are steps being taken to win Republican votes for a treaty with Russia designed to reduce nuclear weapons. The lesson: Don’t build more weapons than you need or could use.”

I, too, think the B53 is a cautionary tale but I draw a slightly different lesson. Today, Capitol Hill is in the midst of a furious debate over the same “bargain” that the Administration struck to win Senate ratification of the New START Treaty.  Some people say,”Hey we had a deal!”  Others say, “Hey, we had a lot of deals before you refused to raise the debt ceiling to pay for them.” The discussion gets a little testy.  Members of Congress employ epithets like “irrational” and “unilateral disarmament.”

I don’t intend to wade into this discussion just yet, but I would suggest that  — whether the Administration builds new weapons or new facilities — future  appropriations should include funds set aside to pay for dismantlement and clean-up.  The amount of these funds should, over the course of the lifetime of the asset, sum to the expected liability associated with dismantlement, remediation and so-on.

Of course, that makes nuclear weapons programs more expensive in the short-term, but that is little more than special pleading by one generation.  Over the long-term, such an approach is almost certainly more cost-effective since all costs are considered in the planning phase.  Had policymakers worried about the cost of dismantling the B53 when it was being built, they almost certainly would have made different design choices that would have made dismantling the B53 less expensive and much safer.

That is the lesson I take from the final dismantlement of the B53.


  1. bradley laing (History)

    Could we get some context? The government was covering up the truth about above ground tests before 1963, and then, in 1968, covered up the truth about below ground tests.

    I’d read that the FBI went around to newspapers in the early 1960s with a tape recording that was supposed to discredit Martin Luther King. Those papers did not take the FBI bait, but they also refused to ask if the FBI was spying on anyone else for their perfectly legal political views.

    I want to know more about the times the decisions were being made.

  2. Peter Burt (History)

    It’s not just nuclear warheads that add to the legacy costs of the military nuclear programme. Here in the UK the Ministry of Defence is just about to launch consultation on what to do with the radioactively contaminated hulks of its out-of-service nuclear powered submarines, some of which have been rusting away in dockyards for nearly 40 years while the government works out what to do with them ( The costs for the dismantling programme are unquantifiable at present, but given that the programme will stretch over around 60 years, will be huge.

    And that’s just for disposing of the submarine hulls – we’re not even beginning to think about what to do with the spent reactor fuel.

  3. Allen Thomson (History)

    I listened to an NPR story about this yesterday afternoon that revealed that the B53 was as big as a bus and had six times the explosive power of the bomb that leveled Hiroshima. I considered sending them a note, but had a failure of will.

  4. anon (History)

    While I don’t doubt that the B53 got placed on the list of weapons to be retired because it was unsafe, I don’t think this was the reason why it was finally retired when it was retired. There was a BRAC round in the mid-1990s that recommended closing the only air base that could support B-52s equipped with B53s. The decision was made to retire the bomb, in part, because it would have been too costly and complicated to relocate the mission to another airbase.

  5. anon (History)

    “Had policymakers worried about the cost of dismantling the B53 when it was being built, they almost certainly would have made different design choices that would have made dismantling the B53 less expensive and much safer.
    That is the lesson I take from the final dismantlement of the B53.”

    I totally disagree. Procedures & tooling for disassembly were designed, fab’d & approved before the first WR 53’s were even built. PX PTs have safely dismantled several hundred W/B53 over the past 40+ years. The tooling, procedures & personnel qualifications were approved via the Nuclear Explosive Studies prior to each cycle by LASL/LANL, AEC/DOE, SNL, etc., etc.

    Just because the B53 is considered large by today’s standards does not automatically mean it’s difficult to disassemble. Over the years there have been many other weapons that were much more difficult to disassemble.

    In the past, this small number of 53’s disassembled over the past year would have represented no more than 1-2 months workload. Just because DOE/ current contractor at PX have taken 2+ yrs to completely redesign the process, tooling & procedures spending millions$$$ just to disassemble these last few 53”s does not mean there was no concern for disassembly when it was being designed. The 1st NES would never have been approved if approved tooling, procedures & qualified technicians were not demonstrated at that time.

    What the hell were you thinking?

  6. Cheryl Rofer (History)

    Some disarmament advocates are using the administration’s request for decreased funding of dismantlement for FY 2012 as an indication of lack of commitment. But when the added costs for B53 dismantlement are considered, that budget is simply returning to normal.

  7. Mark Lincoln (History)

    I have been astounded with the coverage given the demise of the last B-53.

    It has made NPR and the press world-wide.

    Who has been pimping this story?


    • Stephen Schwartz (History)

      Well, the NNSA has, and pretty successfully (though I’d say flogging, not pimping). Look, it’s not the most important story in the world, but it is important, and it’s good to see the news media take notice of the fact, even if only for a day, that thousands of nuclear weapons still exist.

      But if you want to blame someone, you can blame me, maybe. A few years ago, after D’Agostino gave a talk about the nuclear weapons complex at CSIS (around the time the agency was getting a lot of heat over the RRW proposal), I went up to him and said that the NNSA needed to break free of the reflexive mentality that everything it does is secret and start touting some success stories, principally that disarmament is happening all the time (albeit not as quickly as many of us would like). Holding press conferences and discussing important milestones (like this one), I said, could help the public better understand what’s happening with US nuclear weapons and where some of the nuclear budget dollars are going, as well as help dispel the widespread notion at NPT conferences and elsewhere that the United States is not doing anything to eliminate its weapons. There is some awareness that treaties have been signed, I noted, but no actual confirmation or coverage when the weapons themselves get taken apart.

      And while you’re at it, I said, how about releasing the disarmament totals each year, as was done during the Clinton administration? There’s no reason for that information to be classified, and its release would provide hard evidence that the United States is working to meet, however slowly, its Article VI commitment. D’Agostino seemed genuinely appreciative and said he’d take those recommendations back to his staff.

      If I’m disappointed in anything about this story, apart from Jeffrey’s excellent points above, it’s that those figures are still kept secret. Given that the Obama administration finished the job Clinton started in the mid-1990s and got the DOD in May 2009 to declassify the total stockpile numbers by year from 1962 onward, this is particularly inexplicable, as is the fact that the NNSA still won’t say how many B53s were built or what its yield was. Secrecy for secrecy’s sake is neither necessary nor helpful.

    • Jeffrey (History)

      The May 2009 declassification included dismantlement numbers. It will be interesting to see if this Administration ever releases another stockpile or dismantlement number again.

    • Stephen Schwartz (History)

      Right. Sorry, I meant to say that. Of course, no one at NNSA seems to have mentioned those numbers the other day. The press didn’t appear to use them either to get the agency to discuss this year’s dismantlement rate or even to shed some light on the past (although a few reporters did try to find out how many B53s were in the arsenal after 1997).

      I’ve heard that the 2009 historical stockpile figure release was a one-time deal. If true (and there have been no new numbers forthcoming in 15 months), that’s very unfortunate. As for the dismantlement figures, the hitch back in 1994 was the the DOD would only allow numbers to be released for systems that were no longer in the arsenal (hence the 1962 cutoff). Without knowing more about exactly which systems were dismantled between 2000-2009, it’s hard to say if DOD changed its tune or made a one-time exception. Since the figures encompass a fiscal year, Tuesday would have been an ideal opportunity to release the 2010 number. Of course, if that was significantly below 356 (the 2009 rate, which was itself far below the rate of 648 for 2008), that might have raised questions the NNSA would have preferred not to answer.

  8. Alex (History)

    I love the idea that a 9-megaton nuclear bomb can be considered “safe” or “unsafe”.

    Mind you, coming from the country that gave the world Violet Club…

  9. thermopile (History)

    As a contrast to the nuclear weapons decommissioning costs, a good counter example is the nuclear power industry. Decommissioning costs must be considered from the outset:

    (scroll down about 1/2 page for the “Decommissioning Funds” part)

    While the plants don’t have to have all the funds available for decommissioning on Day One, they do have to have a plan for making those funds available. The formulas for estimating decommissioning costs are established by law, and each licensee must report to the NRC every two years on how their funds are doing.

    Great post. Life-cycle costs are just something most budgeteers don’t grasp well, largely because it’s inconvenient.

  10. Stephen Young (History)

    You write: “Today, the operating budgets for PANTEX and Y-12 are about $650 million and $935 million. If the two-thirds rule still holds, then dismantling warheads is a much more expensive process than the official budget suggests.”

    But we know that’s not true. In Pantex and Y-12 in 1993 were not doing Life Extension Programs, at least not as such. Today, most of their time, energy and budget is focused on major, expensive, complicated LEPs. Dismantlement is squeezed in around the edges.

    • Jeffrey (History)

      We may suspect, but I don’t think we know for any given year. There were certainly years where NNSA argued that dismantlement slowed because the ongoing LEPs, but then there have been years with relatively little LEP work, too.

  11. Nick (History)

    One of the largest?! Has there been anything bigger than B53? Shouldn’t it be the largest.

    ….”The dismantlement of the B53 bomb – the oldest weapon in America’s arsenal and one of the largest in US history – is a major accomplishment that has made the world safer and for which everyone involved should be proud,” said Deputy Secretary of Energy Daniel Poneman…

  12. Anon (History)

    The Mark 17 Bomb, in the stockpile from 1955 to 1957 was much larger.

    • Captain Ned (History)

      As were the Mk 24 and the Mk 41. The Mk 36 was of equivalent size.

  13. Pedro (History)

    What the hell were they thinking??

    What were Edward Teller’s first words after monitoring Ivy Mike from his bunker?:

    “Was it big enough?” – If my memory of “Dark Sun” is correct.

    That’s what they were thinking.

  14. Bill (History)

    Wasn’t the B-41 supposed to have a capability of up to 25 megatons? A very dirty bomb! I remember reading that it was produced in early 1960s and not fully retired from stockpile until mid-1970s but am not sure. Stan Norris, can you help out on this?

  15. Ara Barsamian (History)

    Why all the fuss about the B53? It wasn’t the biggest US bomb (MK17, 24, etc), or the most complicated.

    300 lbs of Cyclotol HE, 37 Lbs of HEU pit and dual channel neutron initiators are not a big deal to dismantle. The HE crumbles rather easily because of age (all plasticizer gone)but with careful removal procedures is not that dangerous.

    The problem is with contemporary penchant for exagerations. In the “old” days we used a lot of common sense and good engineering practices in lieu of today’s multimillion dollar simulations and esoteric tools.

    By the way, B53 was usable for these 40 years without any significant maintenance because it was a conservative HEU-based rather than Pu-based weapon. The new Pu-pit facility is a big waste of taxpayers money (5+billions and counting…).

    As far as reliability, 9MT of brute force beats all the smart microelectronics in todays AFAF’s with or w/o GPS…but we like high tech and like to play with the stuff rather than focus on tasks on hand…

    Finally, whether one gets vaporized by 1kT or 9MT is immaterial to the recipient, and if we ever use nukes, collateral damage is not going to be our highest priority…

    • Jeffrey (History)

      Writing every day is an interesting experience. It makes one very aware of how a particular phrasing can misfire.

      I should say that when I wrote “What the hell were they thinking?” I was trying to express why the B53 dismantlement seemed to catch public attention. Of course, I knew what the hell they were thinking, which is why I then indicated that I was more interested in why it was so hard to dismantle the B53.

      Several comments, however, suggest to me that the transition I chose conveyed a sort of irritation at the B53 designers, which wasn’t really my point at all. Rather, my point is that we have a policy process that did not, and still does not, factor in the full cost of any new build, whether it is a weapon or a building, and simply defers too many important costs to future generations.

      Had weapons designers been told that ease of dismantlement was an important goal, I had no doubt they would have done so. The problem is short-sighted policymaking, not competence of weapons design.

    • Jason (History)

      Just as an aside, the Army produced its M55 chemical rockets without thinking of how to disarm/dispose of them. I believe the theory was that they would be used, and therefore, who cared about the demil phase? As a result, the past 25 years or so of chemical destruction efforts have always centered on the M55 demil as the most critical and dangerous. Live and learn.

      Not saying that the designers of the B53 were thinking in this mode, but who knows?

    • George William Herbert (History)

      Ara writes:

      300 lbs of Cyclotol HE, 37 Lbs of HEU pit and dual channel neutron initiators are not a big deal to dismantle. The HE crumbles rather easily because of age (all plasticizer gone)but with careful removal procedures is not that dangerous.

      The 300 lb and Cyclotol are well known from prior public statements. Can you explain where you got 37 lb HEU pit weight from?

      37 lb is credible in one sense – “16 kg” is the standard quoted in non-classified sources for HEU implosion weapons, and is supported by modeling; that’s 16.8 kg. It seems somewhat of a stretch in another. At 50% fissile efficiency with good boosting that’s something like 140 kt. That would imply roughly 32x secondary scaleup for the clean variant and 64x for the fissile tamper variant. Those figures seem to exceed published (and modeled) values, by about a factor of two.

      And … *what plasticizer* ??? Cyclotol is a RDX/TNT castable explosive, not a PBX. TNT ages (and to a much lesser degree RDX), so it physically decaying with age is not unheard of. But plasticizer decay seems unlikely unless there’s a Cyclotol variant I haven’t heard of before. I just rechecked the LASL databook and Cooper and Meyer and no, they list ingredients as TNT and RDX, period. So… *what plasticizer* ???

  16. John Olsen (History)

    It’s touching that they have little squares mounted on the walls to buffer the shock wave if the darned thing goes off accidentally. Do they have any concept? Nice article though.

    • anon (History)

      “It’s touching that they have little squares mounted on the walls to buffer the shock wave if the darned thing goes off accidentally. Do they have any concept? Nice article though.”

      You’ve apparently never worked in a solid concrete room like an assembly cell/bay. The acoustics are terrible.

      “buffer the shock wave”??? – now that’s a hoot – LMAOROTF!!!

    • Allen Thomson (History)

      Actually, I was wondering about the little squares: What are those things?

  17. Ara Barsamian (History)

    Although many don’t like to hear this, the B53 was in retrospect, one of the safest designs
    B53 was based strictly on HEU which is not poisonous, selfheating and unstable like Plutonium. In addition, it did not use any Beryllium, whether as a pit reflector, or secondary ablator.

    So taking it apart, the hardest thing is the safe removal of the crumbling HE around the pit and the old detonators.

    This accounts, besides its yield for bunker-busting, for the longevity of this design without billion dollar LEPs.

    Have you tried/seen the nightmare to take a W88 apart?

    So excuse me if I take a jaundiced view of the unjust criticism of B53 and difficulty of dismantling it…

  18. anon (History)

    “Have you tried/seen the nightmare to take a W88 apart?”

    Gulp! – I hope these folks have not tried taking apart a w88.

    Actually, the 88’s a pussycat compared to others. How about the 48 – now there’s one most wouldn’t have enjoyed seeing the NE being disassembled.

  19. Robert Merkel (History)

    I don’t think that kind of accounting provision is used anywhere else in government.

    Furthermore, if you assume a multi-decade life for weapons, the cost of disposal is much smaller than you might think. Let’s say disposal costs are 60% of construction costs (ignoring inflation). Let’s further assume that you can earn a risk-free real return of 3%. You’d only need to invest 25% or so of your construction costs to cover disposal 30 years hence.