Dear [congressional and executive branch] colleagues --
The short editorial below is pending publication in a trade newsletter but will not be published before December 4 [now December 11, if then]. It may be important to send it to you sooner, hence this note.
It concerns pending appropriations and authorization legislation for the rapidly-growing program to produce the plutonium cores ("pits") for nuclear warheads, found in the FY21 Energy and Water Development appropriations bill and the FY21 National Defense Authorization Act (NDAA).
If you are not familiar with the challenging pit production program, a brief overview was provided by the Department of Energy (DOE) Inspector General on pp. 5-7 in Management Challenges at the Department of Energy – Fiscal Year 2021. In a nutshell, contractors for the National Nuclear Security Administration (NNSA) are to manufacture pits at two sites, the Los Alamos National Laboratory (LANL) in New Mexico, and the Savannah River Site (SRS) in South Carolina, both of which require billions of dollars in retooled infrastructure and about two thousand additional trained employees at each site before quantity production can begin in the late 2020s to early 2030s timeframe. In principal, and only if long-standing safety issues can be solved, LANL can start production sooner, but only at higher cost, lower capacity, and for an uncertain period of time, given LANL's present aged facilities.
We think this program, in its present form and on its present schedule, is quite poorly justified. As others have noted, the risk of failure is great (Independent Assessment of the Plutonium Strategy of the NNSA, IDA, for DoD, Mar 2019).
The cost will be large.
- NNSA estimate*: $0.21 M to $0.94 M/pit (median: $0.57 M/pit)
- At SRS, Congressional Budget Office estimate (p. 14): $6 M/pit
- At LANL, Los Alamos Study Group estimate based on FY21 NNSA budget request: $38-$60 M/pit (slides 29-31; see discussion below)
(*NNSA asserts that the total cost of pits for the W87-1 program will be $300-$750 million (M) (Government Accountability Office, p. 22). The number of W87-1 warheads to be built is classified but is likely to lie in the range of 800-1,400 warheads. This gives an implied cost range of $0.21 M to $0.94 M per pit.)
Fifty-year life-cycle program costs appear to fall somewhere in the range of $50-$100 billion.
Pending a (badly-needed) fundamental review, the common-sense requirements of the House Appropriations Committee, which would also trim runaway spending, appears to be the most responsible approach.
Further information and details are available in this recent workshop presentation: "NNSA pit production strategy: no clear goals, plans, or likelihood of success; Production at LANL has high risks and costs, few or no program benefits," Oct 1, 2020 (updated Oct 3, 2020).
A comprehensive collection of open-source analysis and resources on this topic can be found here.
The guest editorial mentioned above:
Plutonium pit production: no rush required
NOVEMBER 18, 2020
By Greg Mello, Los Alamos Study Group
NNSA’s plutonium pit production strategy lacks clear goals, resourced plans, and has little likelihood of on-time, on-budget success. No credible contingency plans have been prepared. Congress should take steps toward remedying these problems before it adjourns in December.
NNSA’s current plan is to build and operate two pit factories to achieve an enduring production capacity of ≥80 pits per year (ppy) beginning in 2030. Supposedly, the Savannah River Plutonium Processing Facility (SRPPF) would produce ≥50 ppy (or ≥80 ppy) by 2030, and LANL’s 1978-vintage plutonium facility (PF-4) would produce ≥30 ppy by 2026 and if necessary ≥80 ppy by 2030 – which even former NNSA Administrator Lisa Gordon Hagerty has said is impossible.
Pit production is highly unlikely to meet this deadline, especially if rushed (Institute for Defense Analyses). No master plan and schedule, or credible backup plan, have been prepared (GAO, e-p. 2; pp. 29-38).
Prior to the present Continuing Resolution the House passed an appropriations bill (H.R. 7617) providing $1.08 billion (B) for pit production in FY21, $0.29 B less than requested but $0.37 B more than FY20. The bill cuts $135 million (M) from SRS’s and $157 M from LANL’s pit production accounts while allowing most (at LANL) or all (at SRS) funds requested for pit infrastructure (pp. 170-171).
In its report, House Appropriations notes NNSA has no detailed pit production plan:
The Committee remains concerned that NNSA has not prioritized the development of a resource-loaded integrated master schedule that includes all pit production-related, project-related and program activities as recommended by the GAO and does not appear to have plans to complete such a schedule until after it would have had to achieve certain pit production milestones. (p. 140)
NNSA is directed to submit such a plan within 30 days of enactment. Noting that the risk of failure is high, House Appropriations also requires that NNSA submit, by 120 days after enactment, a contingency plan for managing the stockpile in the event the 2030 deadline is not met, including
options to ramp up pit production that extend the current need dates for pit production; how the hedge and fielded stockpile could be configured to serve as an interim solution; and an estimate of how many years current pit production need dates could be extended by advancing pit reuse concepts. (p. 140)
These are straightforward, common-sense requirements. Given NNSA’s poor track record of timely, adequate reporting, they should be strengthened by withholding funds in the interim.
CBO estimates SRPPF’s marginal pit cost at $6 M (p. 14), about one-third the total cost of each warhead. LANL’s costs are much higher. In 2017 the cost of establishing an average 30 ppy production at LANL was thought to be $3 B (s. 2). The reality is quite different.
NNSA is requesting $7.6 B for pit production startup at LANL through FY25 (s. 23) and apparently will incur costs through 2030 in the $14 B range (s. 29). One of the reasons for this is the newly-revealed necessity of running PF-4 and surrounding facilities on a 24/7 basis to achieve just 20 ppy (p. 15), necessitating a much larger staff and more support facilities. Production at LANL will cost $38-60 M per pit (slides 29-31), almost an order of magnitude more than at SRS and roughly two orders of magnitude beyond what can be inferred from NNSA’s stated total cost for producing all new pits for the W87-1 ($300-$750 M, p. 22).
The enormous cost premium for what appears to be an arbitrary W87-1 deadline also indicates higher risk. It is not yet possible to credibly predict future pit production at LANL, given the unresolved problems flagged by the DNFSB (slides 6-15; for dates of possible resolution see s. 21).
Another House initiative makes less sense. In its markup of the FY21 NDAA, the House Armed Services Committee added Section 3115 (pp. 2284-2287), which seems intended to halt investments at SRPPF by requiring an impossibly-accurate cost certification during the next administration. SRS production would be replaced by “temporarily surging the production of such pits at [LANL] and other [unnamed] mitigation strategies.”
Setting aside the obvious political factors, the feasibility of any LANL “surge” is uncertain at best. NNSA’s 2017 Analysis of Alternatives found that a PF-4 “surge” was infeasible (p. 2); its 2018 Engineering Analysis rated the alternative involving a surge with by far the highest risk (Alt. 2c, s. 8). In 2019 IDA found that any PF-4 surge would be “very high risk” (p. vii). GAO cites a classified LANL study saying LANL was only “marginally capable” of reaching 30 ppy by 2026 “and sustaining that rate thereafter” (p. 35).
Congress needs to look much more actively at why all these – or any -- pits are supposedly needed by 2030. There are several alternatives, as the House report hints. The next Congress can and should take up this and related topics in more fundamental ways. |