Jeffrey LewisWSJ On Reliable Replacement Warhead

Carla Anne Robbins, writing in the Wall Street Journal (subscription required) has a long article summarizing the Reliable Replacement Warhead program (A blog has posted the full text of Robbins’ article).

Unfortunately, the article suffers from the shallow “he said, she said” style of journalism that simply presents unsupported assertions in matched pairs (The RRW is awesome! Critics doubt awesomeness.)

Bob Peurifoy sent a note to Ms. Robbins, asking a series of uncomfortable questions about some of these unsupported assertions. I reprint it here:

December 16, 2005

Carla Anne Robbins:

I read with interest your front-page article on nuclear warheads in the December 14, 2005, issue of the Wall Street Journal.

I believe that this nation must maintain a credible nuclear weapon system deterrent as a component of national security. I assert that national security policies and strategies, rather than NNSA/Lab self-interests, should direct the future maintenance of the nuclear weapon program.

I suggest that your article lacks balance and that you have been misled. I cite examples:

Example 1

I find in your article phrases such as

“…sturdier, more reliable warhead…”
“…more-reliable warhead”
“…a more dependable warhead…”
“…building a new, sturdier replacement warhead”
“…The reliable warhead…”
“…the Reliable Replacement Warhead”
“…a new stockpile that was more reliable”
“The reliable-warhead program…”
“…hedge against technical failures…”

My response to the above phrases:

The antonym for reliable is unreliable. Are the warheads now in the stockpile unreliable? Have the design labs misled the country? What does sturdier mean?

Are the warheads now in the stockpile frail, e.g., don’t meet military requirements? How can warheads of the same type, built at the same time, offer a hedge against technical failures? How will these technical failures be found?

Example 2

I find in your article phrases such as:

“…the aging arsenal is becoming increasingly difficult and costly to maintain.”
“…the U.S. spends billions of dollars each year to monitor its stockpile and extend the weapons’ life.”
“…new technology should permit crafting one that is easier to build, cheaper to maintain and safer to store…”
:…still less experience with the effects of aging on warheads…”
“…to begin replacing certain aging components of a 30-year-old warhead called the W-76…The plan is to extend its life another 30 years.

My response to the above phrases:

Why is the aging arsenal becoming “increasingly difficult and costly to maintain”? The only scheduled maintenance involves the periodic replacement of components containing tritium. Will the new warheads not use tritium? Occasional non-scheduled repair or replacement of hardware found defective by surveillance activities has been necessary. If finding such defects is important, why is the stockpile surveillance program being scaled back?

I do not believe the U.S. spends billions of dollars each year to monitor its stockpile.

[Ed. note: Peurifoy is correct. In FY 2005, the US spent $1.3 billion on Directed Stockpile Work including $ 278 M for routine maintenance and surveillance, as well as $363 M on “service life-extension programs”. The remaining $806 M was appropriated for warhead retirement and research and development programs such as the RRW.]

What surveillance data justifies spending billions to extend a weapon type’s life?

What is their expected life? No weapon type has ever been retired because it died.

Tell me about this new technology. “Easier to build”? Perhaps at the margin. “Cheaper to maintain”? Please explain. “Safer to store”? What does this mean?

What aging effects have been found? The general character of the independent surveillance program was established in the late 1950s when sealed pit/thermal battery-powered warheads first entered the stockpile. Improvements to the program have been made as deemed desirable. Perhaps 15,000 weapons have been examined.

Concerns about aging seem to focus on plutonium pits, now said by NNSA and the labs to be okay for 60 years. What components in the W-76 are suffering unacceptable aging that will void their value before 2040? Why can’t components be remanufactured to original specifications?

Example 3

I find in your article two statements:

“…was designed to deter a foe far different from those the U.S. now faces. ‘You would not create the current stockpile if you were starting now’”; and
“A 2001 review identified a new set of potential adversaries, including Iraq, Iran, North Korea, Syria and China. It called for a broader array of nuclear capabilities, including weapons that could go after hardened, deeply buried targets, and less-powerful warheads to reduce ‘collateral damage.’”

My response to the above two statements:

Is it realistic to believe that the U.S. needs different weapon designs in order to atom-bomb such adversaries? What would the new stockpile consist of? What is a “broader array of nuclear capabilities”? The active stockpile now contains warheads with available yields ranging from subkilotons to megatons. They can be delivered by aircraft, ICBMs and SLBMs. Careful design and manufacturing have provided good reliability, as attested to by the surveillance program. They all meet DOD safety requirements.

Regarding weapons that could go after hardened, deeply buried targets, see Effects of Nuclear Earth-Penetrator and Other Weapons, by a committee chaired by Dr. John Ahearne and sponsored by the National Research Council of the National Academies.

Example 4

I find in your article this statement:

“They’ve been told to design a warhead that has the explosive power of the W-76, but inside the larger body of a more powerful warhead, the W-88.”

My response:

What does this mean? If you want a warhead that has the explosive power of the W-76, then use the W-76. If the ‘larger body’ means the Navy MK5 reentry body and you wish to use the W-88 at a lower yield, detune the warhead. Keep in mind that the major costs of such a modification fall on the Navy.

Example 5

I find in your article a paragraph which I paraphrase: Los Alamos’s lead engineer at the [Nevada Test] site, says, “We’ve worked with the State Department” to explain to embassies there’s no plan to violate the test moratorium.

My response:

I am overjoyed to learn that Los Alamos engineers at the Nevada Test Site have promised the State Department that Los Alamos has no plan to violate the CTBT.


By separate e-mail I am sending you attachments to a paper I prepared dealing with the possible use of uranium 235 pits rather than plutonium pits in some of our primaries. My objective is to eliminate the need for a costly modern pit facility. Also, I’m sending you a critique I wrote and distributed in response to the draft SEAB Report concerning the nuclear weapon complex of the future.

Bob Peurifoy

I’ve also posted the attachments to Peurifoy’s paper on the possible use of Uranium 235 pits in the URRW.

Keep in mind that Peurifoy’s first preference is to keep the current arsenal and that his goal is to start a much-needed discussion, not provide the last word.

Peurifoy notes some previous tests of U-235 pits and amends his previous suggestion, by concluding that the diameter of plutonium pits would be preferable for certain naval systems. Pu pit production for these warheads could be handled at Plutonium Facility 4 (PF4) at Los Alamos Technical Area 55—eliminating the need for a Modern Pit Facility.


  1. John Field (History)

    This uranium RRW is really my favorite topic. I wish there were an open source collaboration I could join to compete against the LANL and LLNL teams. Sort of like Space/X or something. I need a blog.

    My concern[/complaint] has always been that the uranium primaries are going to be heavier and therefore less efficient at a given yield which would lead to problems in the radiative efficiency.

    From Bob’s attachments, I see that Stokes test was 19 kt in 317 lbs in an all oralloy pit.

    OK, I guess that’s good enough. Franklin prime seems a little thin at 5 kt in 300 lbs though.

    Everything kind of stops working when you get down to less than about [roughly speaking] 40 tons TNT yield per pound of primary weight. This is about the point where there is as much energy in photons as in hot gas. Below this point, the reductions almost all come out of the photons and radiative efficiency falls like a missile out of the sky.

    I’ve always figured that the US fancy primaries today are probably in the range of several tens of kt and maybe 200-300 lbs so therefore north of this threshold. At least that’s what the rumors seem to be.

  2. Greg Mello

    Thank you for the Wall Street Journal/blog link regarding the Reliable Replacement Warhead (RRW) and the excellent comments by Bob Peurifoy.

    Having spent a lot of time studying pit production at LANL, I think your statement, “Pu pit production for these warheads could be handled at Plutonium Facility 4 (PF4) at Los Alamos Technical Area 55—eliminating the need for a Modern Pit Facility” is incorrect. As time goes by I am learning more about why this idea won’t work; one snapshot of the problems was included in a 2004 article: [Greg Mello, “U.S. Nuclear Warhead Research and Production With No Real Debate, Diminishing Prospects for Control, _INESAP Bulletin_ 23].

    I know that Frank von Hippel and Steve Fetter have judged otherwise, but their paper basically just counts pits in an very idealized fashion as physicists will do, without taking into account the very real institutional, production engineering, cultural, and political problems involved. [see: Steve Fetter and Frank von Hippel, “Does the United States Need a New Plutonium-Pit Facility?” Arms Control Today May 2004.] Their cost estimates are ludicrously low as well as I recall, and have been left behind by subsequent events (i.e. current cost estimates for the Chemistiry and Metallurgy Replacement facility, or CMRR, essential to the project).

    LANL has more severe seismic problems than decisionmakers have fully absorbed as well; LANL’s seismic team has found stratigraphic records of 3 events of magnitude “~ 7” in the Holocene on the fault that comprises the western boundary of the lab.

    As you know, the LANL facility has been judged by the SEAB Task Force on complex infrastructure to be running at “5%” efficiency; this is just a symptom of many underlying problems which have no easy solutions that are also SAFE solutions from an engineering perspective. For example, there is a deep disagreement between the DNFSB and NNSA over the safety systems for PF-4 and the proposed adjacent CMRR as well. As PF-4 ages further it is getting more expensive; by the time it is supposed to be up and running as a production facility it will be about as old as Rocky Flats was when it closed.

    Best, Greg Mello