The U.S. nuclear weapons complex: Pushing for a new production capability

On January 15, the Wall Street Journal published an op-ed by former secretaries of state George Shultz and Henry Kissinger, former Defense Secretary William Perry, and former Georgia Democratic Sen. Sam Nunn, which 37 other national security experts also endorsed. Entitled “Toward A Nuclear-Free World,” it was the second such essay in the Journal by these authors in as many years. (See also “A World Free of Nuclear Weapons.”) Both essays concerned the benefits–some immediate, others long-term–of specific nuclear policies the authors believe would be best advanced under the nuclear disarmament banner.

These authors do not mention that the United States and four other nuclear states (Russia, Britain, France, and China) are already legally bound to “pursue negotiations in good faith on effective measures relating to cessation of the nuclear arms race at an early date and to nuclear disarmament . . .” by Article VI of the Nuclear Non-Proliferation Treaty (NPT). The opinion of the World Court and subsequent U.S. diplomatic agreements has confirmed the binding character of these twin commitments to end the arms race and achieve nuclear disarmament. ¹ Most observers agree that the collective unwillingness of the five NPT nuclear weapons states to persuasively implement these Article VI obligations has harmed the NPT and the law-based nonproliferation regime it underpins. ²

If the disarmament aspiration expressed in these two essays means anything, it means refraining from long-term investments in the specialized, “responsive” infrastructure needed to make novel warheads. Nuclear weapons infrastructure investments that require large, long-term commitments of capital and skilled technical labor–scarce resources in any country–are good indicators of national nuclear intent. In other words, infrastructure investments make, and are, nuclear policy.

The U.S. government says as much. In 2006, Linton Brooks, the then administrator of the National Nuclear Security Administration (NNSA), emphasized the importance of long-term manufacturing investments as a foundation of more aggressive nuclear policies a “couple of decades” hence. “We can change our declaratory [nuclear] policy in a day,” he said during a speech to the East Tennessee Economic Council. “We can make operational and targeting changes in weeks or months. In a year or so we can improve integration of nuclear and non-nuclear offense. By contrast, the infrastructure and the stockpile it can support cannot change as quickly. Full infrastructure changes may take a couple of decades.”

Brooks is right. The factory complex at Los Alamos National Laboratory (LANL) needed to produce the fissile plutonium cores, or “pits,” for RRW or another new warhead isn’t expected to be completed until at least 2017. But as long as design and construction of these production facilities proceeds, Congress could “halt” RRW for a few more years, as it did in late 2007, without significantly affecting its final delivery schedule, assuming it were eventually approved.

Warhead design and engineering development are short-term activities compared with designing, constructing, equipping, and standing up operations in the facilities needed to actually build RRWs. The new buildings needed are orders of magnitude more complicated than the warheads and there is considerable managerial risk involved in acquiring them. ³ For example, the nuclear explosive portion in a warhead or bomb contains at most a few hundred components, nearly all of which are inert until use. By contrast, a typical automobile has more than 10,000 parts. A plutonium production complex contains millions of parts, and such a complex is anything but inert. To successfully operate it would require training and coordinating at least 1,000 people and would also require some success in meeting safety, security, and environment standards. Construction of the most recent large-scale U.S. pit production-related facility, Building 371 at Rocky Flats in Colorado began in 1973 and was completed in 1981 at a cost of $225 million ($524 million in today’s dollars). It operated for only one month before the Energy Department realized that the technology on which it was based would not work. The repair cost $400 million and took eight years. Energy called it a “fiasco.”

NNSA describes the proposed new factories at LANL as merely providing “capacity,” as if “capacity” could be created and then mothballed. One cannot build, equip, and stand up highly specialized factories that cost billions of dollars and hire and train hundreds of highly specialized technicians over many years without actually making the objects these costly and complex arrangements were meant to produce.

The proposed Chemistry and Metallurgy Research Replacement (CMRR) Facility at Los Alamos

The United States has now begun to heavily invest in the specialized manufacturing infrastructure needed for new nuclear weapons, pivotally at LANL. The flagship of this complex is the CMRR project to be built at LANL’s Technical Area (TA)-55. NNSA describes the current cost for CMRR as at least $2.2 billion. But if completed, it would probably cost more. ⁴

The CMRR consists of two buildings–the Nuclear Facility (NF), comprising roughly nine-tenths of the project in dollar terms, and the Radiological Laboratory, Utility, and Office Building (RLUOB). Together, the two buildings would comprise some 400,000 square feet of new interior space, and the NF’s 6-metric ton vault would approximately triple LANL’s plutonium storage capacity. ⁵ If completed, the CMRR would be the largest construction project in the history of LANL in inflation-corrected dollars.

The two CMRR buildings would be linked by tunnels and connect to LANL’s existing 30-year-old plutonium facility (PF-4), which has been modified for production using operational funds over the last decade or more. NNSA has now begun a more extensive renovation of PF-4 in an open-ended, long-term construction line item called the “TA-55 Reinvestment Project.”

At present, pit production utilizes approximately one-quarter of PF-4’s 59,600 square feet of nuclear floor space; the CMRR NF would add at least 22,500 additional square feet of this type, some with greater ceiling height, providing greater operational flexibility. Ceiling height has been a limiting factor regarding manufacturing equipment and production processes in PF-4.

RLUOB construction is approximately 40 percent complete, while after four years, the Nuclear Facility is still in preliminary design and it’s unclear when, or if, it will be completed or when construction might begin if approved. Physically, the 90,000-cubic-yard pit dug at the NF site, ostensibly to investigate seismic conditions, is now the staging yard for RLUOB construction. Therefore, the earliest possible construction start date for the NF is spring 2009–the earliest RLUOB could be completed. ⁶

Such a schedule seems optimistic, as a number of significant NF design issues remain unresolved, including seismic design, overall safety design, and building size. (See the summary of the “Draft Complex Transformation Supplemental Programmatic Environmental Impact Statement”.) As of March 2007, conceptual vault design, including provisions for fail-safe cooling of plutonium stores, hadn’t been finalized. ⁷

It’s difficult to predict the ultimate capacity of a LANL pit production complex anchored by a renovated PF-4 and the two CMRR buildings–especially if additional production space or an additional two production buildings were subsequently added, as NNSA suggests might happen. ⁸

Whether built with just the RLUOB, the RLUOB and the NF as planned, the RLUOB plus a “supersized” NF, or with the whole project doubled in size by subsequent construction, the CMRR is unnecessary to maintain the present nuclear arsenal or any subset of it for several decades. The CMRR is needed, however, to manufacture significant quantities of pits for novel nuclear explosives. ⁹

How many pits could LANL make–with and without CMRR?

LANL has possessed the capability to make pits since 1945. But until last year–when it produced 11 new pits, some or all of which were assembled into W88 Trident warheads at the Pantex nuclear weapons plant near Amarillo, Texas–LANL hasn’t made pits for the stockpile since 1949, with one or two possible exceptions. ¹⁰

LANL’s current pit manufacturing capacity is uncertain and open to interpretation. On the one hand, NNSA could choose to displace or terminate certain programs currently housed in PF-4; on the other hand, some of those programs are likely needed for new-design nuclear explosive package certification, without which pit production has no reason to proceed.

At a minimum, successful certification of new-design nuclear explosives requires the use of extensive design, testing, and simulation capabilities. These might not be sufficient; nuclear testing might also be required. So any decision to resume pit production has long coattails, tasking most of PF-4 and much of the nuclear weapons complex as a whole.

In February 1996, Energy said LANL’s pit production capacity, prior to any investment, was “10 to 20 pits per year.” ¹¹ Later that year, Energy stated that LANL pit production of “up to 50 [pits] per year” is “inherent with the facilities and equipment required to manufacture one component [pit] for any stockpile system.” ¹² In 2005, the Secretary of Energy’s Advisory Board (SEAB) Nuclear Weapons Complex Infrastructure Task Force said LANL’s existing pit production capacity could (and should) be increased by a ratio of “1:20.” This twentyfold increase wasn’t a rhetorical flourish; rather, it was predicated on producing an RRW or RRW-like pit designed for mass production involving simpler design, broader tolerances, robotic production technologies in some steps, and fewer toxic materials, which would allow greater ease, flexibility, and speed of production. ¹³

This year, NNSA stated, “A reasonable judgment of the inherent capacity of a production line for nuclear components exceeds 50 per year. A modern factory-style layout could result in a minimum [emphasis added] inherent capacity in the range of 125 components per year.” ¹⁴

Existing LANL pit production capacity is somewhat predicated on the nine-wing Chemistry and Metallurgy Research (CMR) building in TA-3. Despite extensive recent upgrades, much of the CMR may be nearing the end of its usefulness for this purpose. According to NNSA Administrator Tom D’Agostino, pit production could continue at LANL without either the CMRR or CMR, or possibly with part of the CMR, as NNSA wrote in response to congressional questions in 2007. ¹⁵

How many pits per year LANL could produce if CMRR were built is even less clear, as the uncertainties–including uncertainties in CMRR’s size and the number of facilities ultimately available at TA-55–are compounded. In addition, as a senior Energy official explained to me in 2002, the achievable production rate in a given number of square feet of plutonium space is a sensitive function of the technology used. It is also a function of the complexity and tolerances required in the type of pits produced. Any capacity cited today isn’t necessarily what might be available 10 years from now if technology development were to continue–and RRW were approved.

Production capacity is also a function of flexibility, e.g. whether two or more kinds of pits are to be produced simultaneously or in rapid succession.

The lowest capacity is governed by what might be called the “fiasco factor.” Accidents and malicious acts, previously unknown or undisclosed infrastructure or management inadequacies, enforcement actions, and preventive stand-downs have all occurred at LANL and are real possibilities. A production capacity of zero could easily result from any of them, possibly for a long time.

The highest capacity achievable could be significantly greater than the advertised maximum of 200 pits per year.

CMRR’s congressional funding

CMRR appeared in 2003 as a “project engineering and development” line item, becoming a standalone construction project the following year. Since then the Senate, thanks to New Mexico Republican Sen. Pete Domenici, has reliably backed CMRR funding. The House of Representatives, however, has zeroed out the CMRR in three of the past five years and proposed cuts of more than one-half in the other years. The Senate has largely won these battles.

In its most recent markup (for the fiscal year 2008 appropriation), the House Appropriations Committee zeroed out the project and wrote: “Proceeding with the CMRR project as currently designed will strongly prejudice any nuclear complex transformation plan. The CMRR facility has no coherent mission to justify it unless the decision is made to begin an aggressive new nuclear warhead design and pit production mission at Los Alamos National Laboratory.” The House as a whole agreed with this assessment by a wide margin, rebuffing an amendment introduced by New Mexico Democratic Rep. Tom Udall to restore funding for the CMRR, pit production operations, and nuclear weapons overall.

But Senate appropriators had fully funded the project. When the omnibus appropriations bill finally passed in mid-December, the CMRR was funded at $75 million for fiscal year 2008, about 86 percent of NNSA’s request. Neither the bill nor the report contain specific guidance as to which parts of the CMRR project are to receive the abridged funding; project management is privileging RLUOB construction. ¹⁶

What dire consequences would occur if the CMRR Nuclear Facility wasn’t built?

None. Halting the CMRR would not even remotely threaten any existing U.S. nuclear capability–not now and not for many decades to come. But such a step could reflect an aspiration toward disarmament, depending on other policies adopted. In that case, it would express the spirit of the Shultz, Perry, Kissinger and Nunn editorials.

If the United States isn’t prepared to take even this kind of baby step toward fulfilling its NPT obligations, it’s difficult to see how Washington could ever play a constructive role in the international cooperation necessary to prevent nuclear proliferation.

This article has been adapted from a larger piece entitled “Build Warhead Factories Now, Worry About Weapons Policy Later: Will Congress Take Back the Reins?”, available at the Los Alamos Study Group website.

¹The United States reiterated its commitment to nuclear abolition in the consensus statement of the 2000 Nuclear Non-Proliferation Treaty (NPT) Review Conference, agreeing to a set of 13 detailed, “practical steps for the systematic and progressive efforts to implement Article VI.” Prior to this, the World Court unanimously ruled in 1996 that “there exists an obligation to pursue in good faith and bring to a conclusion [Emphasis added.] negotiations leading to nuclear disarmament in all its aspects under strict and effective international control.”
²The author speaks from personal observations at several NPT preparatory and review conferences but also see the formal conclusions of Lewis Dunn et al, Science Applications International Corporation, “Foreign Perspectives on U.S. Nuclear Policy and Posture”, December 4, 2006, prepared for the Defense Threat Reduction Agency (DTRA). Another recent testimony to this view is a speech delivered by IAEA Secretary-General Mohamed ElBaradei on February 11, 2008.
³Keith Schneider, “U.S. Spent Billions on Atom Projects That Have Failed,” New York Times, December 11, 1988, p. A1.
⁴ Energy Department Congressional Budget Request for FY2009, Vol. 1, National Nuclear Security Administration (NNSA), p. 298. The cost of more than $2.2 billion for the Chemistry and Metallurgy Research Replacement Facility (CMRR) is derived from NNSA’s estimate of “above” $2 billion for the CMRR Nuclear Facility (NF), its estimate of $164 million for the Radiological, Utility, and Office Building (RLUOB), and an allowance in the low tens of millions for specialized RLUOB equipment and furnishings–carried now in a separate CMRR project account, “Phase B”–bringing the total to “above” $2.2 billion. Construction costs for even ordinary construction are inflating rapidly and can be expected to continue to increase for the next decade. The CMRR NF is a complex project that involves large quantities of concrete and steel. For these reasons, the CMRR can be expected to increase in cost significantly over the nine years NNSA allots for further design and construction. These CMRR costs don’t include the required new $240 million Technical Area (TA)-55 security perimeter, which must in part be built twice to accommodate construction, the new Pit Radiography Facility ($47 million), the TA-55 Reinvestment Project (at least $200 million), the Radioactive Liquid Waste Treatment Facility Upgrade ($80 million), or the TA-54 nuclear waste disposal expansion project (at least $60 million). Nor do they include demolition and disposal of the existing Chemistry and Metallurgy Research (CMR) facility ($400 million). All of these projects (save for CMR demolition and disposal) are functionally required for CMRR operation.
⁵Los Alamos National Laboratory (LANL), CMRR briefing slides, p. 8, no date.
⁶Personal communication with Steve Fong, NNSA CMRR project staff, January 18, 2007.
⁷Oral response to author’s questions, CMRR public meeting, Fuller Lodge, Los Alamos, New Mexico, March 2007.
⁸See NNSA, “Draft Complex Transformation Supplemental Programmatic Environmental Impact Statement,” pp. S34, 35. Similar plans have been internally available at LANL since at least 2001, e.g., LANL 2001 Comprehensive Site Plan, “TA-55 Preconceptual Plan,” Los Alamos Study Group files.
⁹Neither the CMRR nor Technical Area (TA)-55 as a whole is needed to produce nuclear explosives made with uranium.
¹⁰According to a personal communication with Ken Silver at East Tennessee State University, there are indications LANL’s TA-21 site may have briefly resumed quantity pit production in the immediate aftermath of the disastrous 1969 fire at Rocky Flats.
¹¹Energy Department, Draft Stockpile Stewardship and Management Programmatic Environmental Impact Statement (SSM PEIS), Los Alamos Study Group.
¹²Energy Department, Final SSM PEIS, Volume 1, pp. 3-4, Table 3.1.1.2-1, note “A,” September 1996. Note: “A” is note “1” there.
¹³Anonymous congressional source.
¹⁴NNSA, “Complex Transformation Supplemental Programmatic Environmental Impact Statement (CTSPEIS),” pp. 2-22, December 2007.
¹⁵House Energy and Water Development Appropriations Subcommittee, March 29, 2007, supplemental questions for the record, p. 584 in printed version of “Energy and Water Development Appropriations for 2008.” The use of CMR as solely a radiological laboratory rather than a nuclear facility, to my knowledge, hasn’t been investigated. Neither to my knowledge has there been any comprehensive study of current and planned mission requirements for LANL’s nuclear facilities or radiological facilities.
¹⁶Personal communication with Steve Fong.