For immediate release March 22, 2023

Schedule for Nuclear Warhead Core ("Pit") Production Slipping, Costs Increasing: NNSA's Strategy is Failing

Analyst: Ambitious crash plan for two pit factories is failing, as independent review warned could happen

Contact: Greg Mello, 505-265-1200 office, 505-577-8563 cell

Permalink * Prior press releases

Albuquerque, NM -- The National Nuclear Security Administration's (NNSA's) fiscal year 2024 (FY24) funding request portrays NNSA's effort to prepare for production of plutonium cores ("pits") for nuclear warheads as falling further behind the congressionally-imposed schedule, as construction and start-up costs keep increasing.

NNSA decisions in 2020 ratified the agency's 2018 strategy to divide pit production efforts between two main sites, Los Alamos National Laboratory (LANL) in New Mexico, where NNSA hoped to begin producing at least 30 high-quality ("war reserve," WR) pits per year (ppy) by 2026, and the Savannah River Site (SRS) in South Carolina, where NNSA hoped to begin producing at least 50 WR ppy by 2030.

From statements in NNSA's FY24 budget request released last week, and from public remarks by NNSA Administrator Jill Hruby (54:28 to 56:19) and others, it now appears that LANL will not be able to make start making 30 WR ppy until some time after 2030 and possibly not until 2036 (pp. 210-211), while SRS will not be able to make at least 50 WR ppy until at least 2036 (pp. 2156-2157) and possibly not until 2039 (pp. 236-241).

These newly-expected delays are greater than those which became public in December 2022 (pp. 2156-2157) and early February ("Installation of "Base" Capability to Produce 30 Plutonium Warhead Cores ("Pits") at Los Alamos To Be Delayed 4 Years, to 2030," Feb 3, 2023).

Billions of dollars in new construction, new manufacturing equipment, and additional operational costs (not least, those associated with hiring and training new staff) are required at both sites.

Last year, NNSA's FY23 budget request listed some $21 billion (B) in pit production costs through FY27, of which $7 B had already been spent. Extrapolating from NNSA's projections, we estimated the spending required to bring both sites to full production to be about $14 B at each site or $28 B in all, including sunk costs. At LANL, this meant full production in roughly FY28, two years late, and at SRS in roughly FY33, three years late and one-third of the way through NNSA's then-current estimate of "2032-2035." [1][2]

NNSA's new budget request implies combined start-up costs of at least $34 B, $6 B (21%) more than last year, assuming full LANL production could begin as early as 2031 and full SRS production could begin by 2036. Of this, $10 B has already been appropriated.

NNSA now warns (pp. 210 -213, 236-241) that estimated capital costs in the main pit production projects at each site may rise even further as designs mature, by an estimated 30-40% at LANL and 20-40% at SRS. As noted above, NNSA also warns that its construction schedules are likely to slip by a further 2-4 years at LANL and a further 1-3 years at SRS, incurring additional operational costs prior to full production [3] for these years as well.

In the worst case, these cost increases and schedule delays imply that the start-up costs for pit production under the current strategy (including $10 B in sunk costs), could grow to as much as $46 B before the goal of reliably producing 80 war reserve pits per year is achieved.

Last year, pits from LANL were already quite expensive, in the most optimistic case costing at least $50 million (M) apiece through the middle 2030s. The combined effect of the new delays and cost increases, even in the most optimistic cases, now puts the per-pit cost of these early-time LANL pits north of $90 M each. Per-pit costs at the new, larger SRS facility will be roughly $20 M each, lower than LANL's costs by roughly a factor of 4 or 5, as best as can be estimated at this time. 

The total number of pits that can be produced at LANL by the mid-2030s or late 2030s is also considerably reduced, given these delays.

Reliable pit production from LANL at any level in the 2020s now appears to be off the table entirely, although a few war reserve pits may be produced, if NNSA's plans for LANL proceed without further hitches.

In March 2022, NNSA estimated (p. 211) the total cost of the W87-1 warhead program in then-year dollars, excluding pit production, as $13.0 B for NNSA and $1.2 B for DoD, or $14.2 B in all. As can be now seen, the cost of new pits for W87-1 warheads, none of which are to be made with re-used pits, will dwarf all other program costs.

A year ago, the W87-1 First Production Unit (FPU) ("Phase 6.5" in NNSA parlance) was to occur (p. 43) in 2030 , with quantity production ("Phase 6.6") ensuing the following year. The new delays in pit production imply this schedule cannot be met. Indeed on February 14 of this year, Administrator Hruby said that W87-1 FPU would (now) occur (sometime) "in the early 2030s." The initial warheads to be deployed on the new "Sentinel" silo-based intercontinental missiles, which are to enter service beginning in 2029, are to be W87-0 warheads.

In December, the FY23 National Defense Authorization Act required the Nuclear Weapons Council and NNSA Administrator to review pit production strategy and the associated stockpile plans and provide a plan and any alternatives for the achievement of 80 war reserve pits per year as soon as possible to Congress, by the end of this month, together with other reporting requirements (pp. 2156-2157). The timeline for this review -- just three months --is short.

Contrary to what NNSA says in its budget request (p. 210), the present pit production strategy is not supported by prior studies. The 2017 Analysis of Alternatives (AoA) formally rejected using LANL's aging and relatively small main plutonium building (PF-4) as an enduring pit production facility. It also rejected splitting pit production between two sites, as grossly inefficient. The 2018 pit production Engineering Assessment neither analyzed nor endorsed the present strategy.

In fact NNSA has not conducted any study supporting its present strategy, which was proposed in May 2018 after the Chair and Ranking Member of the Senate Armed Services Committee wrote to the secretaries of DoD and the Department of Energy expressing displeasure at the outcome of the 2017 AoA, favoring LANL as the sole pit production site, as members of the New Mexico congressional delegation also did. Both letters advocated underground production "modules" at LANL, which had already been reviewed by the Government Accountability Office (GAO) and found to be grossly inadequate and ill-considered ("Congressional auditors flay Los Alamos plutonium plans," Aug 10, 2016).

An independent assessment by the Institute for Defense Analyses advised that "[t]rying to increase production at PF-4 by installing additional equipment and operating a second shift is very high risk." While NNSA is not trying to achieve production levels of more than 30 WR ppy at PF-4, the agency has found that the only way to produce even at this level reliably is precisely by "installing additional equipment and operating a second shift." This is the purpose of the Los Alamos Plutonium Pit Production Project (LAP4) (pp. 210-235).

Comparison of the causes NNSA lists for delays of its two main pit production capital projects, LAP4 (pp. 210 -213) and the Savannah River Plutonium Processing Project (SRPPF, pp. 236-241) reveals that shortages of design engineers in particular are central to these delays. These two projects compete for the same small pool of experienced, capable designers with the appropriate security clearances -- and, though it is not stated in the budget request, for staff trained in other plutonium specialties. NNSA struggles to attract and retain the growing workforce needed for its current mission at these and its other sites (see for example "Evolving the Nuclear Security Enterprise," Sep 2022), as Administrator Hruby discussed in February. The two projects also compete for scarce federal management attention.

In January, prior to these latest delays, we provided a precis of what we believed to be a better pit policy here, which also happens to align with arms control considerations.

Study Group director Greg Mello:

"NNSA's pit strategy, inherited from the Trump Administration when it was crafted in response to congressional pressure, is apparently failing. I don't know how else to describe it.

"The New Mexico delegation could not stand losing any part of the tens of billions of dollars represented by plutonium pit production. It was to be the ultimate gravy train for northern New Mexico, a permanent fountain of money and jobs. The present fiasco is the direct result of that greed.

"About $5 billion has been spent at LANL so far, preparing for pit production. Where's the economic and social development? The New Mexico delegation still thinks LANL's plutonium pits are golden eggs.

"Both main pit projects are delayed by several years. They are far over their original budgets. A decade or more before they can produce pits in any quantity, they are competing with each other for scarce personnel, specialty equipment, and funding. As the Institute for Defense Analyses said in 2019, "eventual success is far from certain." How true that is.

"LANL pit production is going to take years more, and cost billions of dollars more, than previously thought. In 2017, NNSA thought LANL could make 30 pits per year for $3 billion, and do it by 2026 -- at least temporarily. That was the plan. Now the reliable production of 30 pits per year at LANL  is going to cost at least $18 billion -- six times as much -- and possibly as much as $25 billion, just to get going.

"We now know that reliable LANL production can't be set up prior to about 2031 and it might take until 2036. That's just too late and costs too much for too few pits to matter.

"The Nuclear Weapons Council, and Congress, need to read between the lines in NNSA's budget request. Pit production at LANL was never a good idea, for any number of reasons. Now it's done. Cooked. Finished. NNSA thought using PF-4 for a factory was a bad idea in 2017 and it's taken six years of wasted effort to make this blindingly clear to anyone who looks.

"NNSA should terminate the LAP4 project. That is the off-ramp. Pit production demonstration and training is not part of this project, so they could continue. Neither are the basic safety improvements needed at PF-4."

Further details are available upon request. A spreadsheet with NNSA's budget projections and their extrapolation through pit production start-up will be published shortly here.

Notes:

1. For the two main capital projects supporting pit production, we used the high end of NNSA's latest cost estimates at "Critical Decision 1," a very early point in these design processes. Final project costs are nearly always at least as high as the highest CD-1 estimates. We omitted many smaller capital projects at LANL as well as any share of supporting site-wide infrastructure.

2. The GAO, using optimistic schedules and costs provided by NNSA in mid-2022, wrote that combined construction and start-up costs might be held to $24 B or less (p. 52), while pointing out that NNSA had not provided an integrated master schedule, or detailed costs, for this huge effort. See "GAO: NNSA's Huge Program to Build New Warhead Cores ("Pits") Lacks Detailed Schedule, Budget, and Scope of Work," press release (2 of 2), Jan 12, 2023.

3. At LANL, NNSA projects that its operational costs for pit production -- costs exclusive of line-item construction projects -- will begin to exceed $1 billion annually by FY28, anywhere from three to eight years before full production could begin. LANL is already running its main plutonium facility around the clock, and expects to hire and train close to 2,000 additional workers to staff and support two production shifts, requiring roughly 4,000 people in all. At SRS, operational costs are expected to be much lower, both in the 2020s and after production begins, as the larger SRS facilities will require only a single production shift to operate at baseline capacity.

***ENDS***