Column: How deep a ‘pit’ will the new administration dig at Los Alamos before stopping?

• By Greg Mello
• September 16, 2021

Every U.S. nuclear warhead and bomb contains a plutonium core, or “pit.” These slowly age. If for some reason the U.S. is still building nuclear warheads in the late 2030s and 2040s – which we can be sure will be an era of deepening climate collapse, global famine, and other severe crises – those in charge of building those new warheads may require new pits.

Yet for reasons ultimately based on an ideology of global dominance, and on corporate self-interest, the U.S. now has a fabulously-expensive, crash program to make new pits not then, but now. Such a program is exactly what in 2019 the Institute for Defense Analyses warned defense agencies against. A rushed program, they said, was more likely to fail.

The pits are “needed” for a new Air Force warhead, to be made in sufficient quantity to place up to three warheads on each missile, a practice the U.S. abandoned a decade ago. Right now, there are enough highly-accurate, modern warheads to put one on each missile, without new pits. Alternatively, if a super-duper new warhead were for some reason required (e.g. to give the nuclear labs and plants something expensive to do), there are plenty of “young” pits of just the right kind – at one warhead per missile.

It gets worse. In 2017 the National Nuclear Security Administration, then under the direction of former Air Force general Frank Klotz, formally determined that the plutonium facility at Los Alamos, built for R&D in the 1970s, was too old, too small, and too otherwise important to be a production facility. To build enough pits, and to have a factory that would last until it was truly needed, either a new facility would have to be built at Los Alamos National Laboratory, or a partially-completed facility in South Carolina could be repurposed. Unsurprisingly, NNSA found that the SC facility would be cheaper, faster, and less risky.

The New Mexico senators and their allies had a cow.

By 2018 Klotz, an Obama holdover, was gone. His successor, Lisa Gordon-Hagerty, made the non-Solomonic decision to build two pit factories, to satisfy everybody. Until then, nobody had seriously thought of building two such beasts.

One factory would be built inside LANL’s old R&D facility – exactly what NNSA had said should never happen. This one would start production in the mid-2020s. The SC factory would be bigger, permanent, and more resilient, but would take longer to complete.

Building two factories was going to be expensive, but it wasn’t clear just HOW expensive. Meanwhile Gordon-Hagerty and others outmaneuvered OMB and DOE to extract from Trump, then under threat of impeachment, an unprecedented 25% increase in warhead spending.

As a result, money isn’t limiting in this program right now. Starting soon however, the Biden Administration will have to make some mature decisions, because costs are exploding.

Apparently LANL promised more than it could deliver. Early last year NNSA belatedly revealed that LANL’s cramped factory would require 24/7 operations to make just 30 pits per year, implying an extra 2,000 staff members, more infrastructure, and an unprecedented ballet of complexity. Operational costs would be a billion per year.

Then this year, projected capital costs in SC came in very high. LANL’s nuclear construction costs also rose by billions.

Start-up costs for the two sites now lie in the range of $32-39 billion through 2033, far more than anyone imagined, with $27-$34 billion still to go, more than half at LANL. LANL pits will cost north of $50 million apiece, at least tripling the cost of any new warheads, again assuming production goes perfectly. It won’t. Even assuming perfect reliability, LANL by itself can’t make enough pits.

Meanwhile the brand-new SC plant, five times the size of LANL’s decrepit facility and much safer, is being designed to make all the pits DoD says it needs, with one production shift.

The new Administration inherited Gordon-Hagerty’s folly. What will they do?

The obvious answer, which could have bipartisan support, is to limit LANL’s role to process prove-in and training, or to single-shift operations. Such a step would cut the overall program cost nearly in half while still meeting the 80 pit per year goal no later than in the current plan.