Jun 5, 2007

Science and Global Security

  • Jun. 6th, 2007 at 1:38 AM

[then head of Los Alamos, Pete Nanos, with head of research and applied science, Brian Bowsher, Atomic Weapons Establishment Aldermaston UK]

There is a good entry on the Reliable Replacement Warhead at Wiki that, from what I can see, was written by Carey Sublette of the Nuclear Weapons Archive. It is argued that RRW-1, to replace the W-76, is going to be based on the W-89. We know that the design chosen for RRW-1, the Lawrence Livermore design, is based on a previously tested warhead that never made it to the active stockpile. The main source used at Wiki is this line on the LLNL effort,

…One candidate under consideration as a starting point is the W89, a 200-kiloton warhead designed for a short-range attack missile. It is well-tested, plus it comes from a long line of well-understood designs and uses every safety and security feature available at the time…

Wiki also states,

…The W89 was also reportedly designed using recycled pits from the earlier W68 nuclear weapon program, recoated in Vanadium to provide the temperature resistance…

The source for this is provided at the following, highly useful discussion on Pu pits,

… Vanadium is another cladding element, but it is unknown whether it is just experimental or in use. Vanadium was used in 1993 during the W89 pit re-use program at Pantex as a fire resistant cladding on W68 pits being converted for use as W89 pits…

It is also pointed out, and it's an important point so far as RRW is concerned, at Wiki that,

…The W89 design was already equipped with all then-current safety features, including insensitive high explosives, fire-resistant pits, and advanced detonator safety systems…

The author goes on to speculate that the modifications to the W89 for the RRW-1 are,

…Modifications for the RRW design would probably have included replacing Beryllium neutron reflector layers with another material, and increased performance margins throughout the design, possibly including more fissile material in the pit and a thicker radiation case or Hohlraum…

The first point is highly likely. I have seen official statements on RRW pointing out that it would use less "exotic materials" to enable ease of re-manufacture. The first point fits into these statements quite nicely.

A point missed from this is the question of yield. The W89 had a yield of 200Kt. The W-76, that RRW-1 is supposed to replace, has a yield of 100Kt. We have seen previously here a statement by the CinC of US Strategic Command that RRW-1 is to have the same strategic military capabilities as the W-76 and that expressly included having the same yield. If so, a very important modification in the design of the W-89 would be reducing its yield from 200Kt to 100Kt whilst using more fissile material in the pit. Perhaps the alternative neutron reflector enables this or perhaps the amount of tritium used for RRW-1 will vary with that employed for the W89. Both of these may be a factor.

If so, what about the issue of testing? Will come back that later.

This emphasis on the W89 can be confusing however. Let me show you what I mean. There is a separate entry on the W89 at Wiki, clearly written by the same author, that states that the W89 was to be used as the warhead of choice for the SRAM II. This was a short range air to surface missile. However, I have come across reports on the RRW design arguing that the LLNL design chosen was based on a warhead designed for the Navy.

I'm confused. We will deal with later on.

Before we solve this little problem there is a good transcript of a 1991 Congressional hearing on nuclear weapons. There is some good stuff on the W89,

…DOE and DOD continue to give very high priority to modernizing the SRAM A (W-69) warhead which has the potential for burn and/or detonation during an aircraft accident or fire. The result is dispersal of plutonium over a wide area…


Modernizing the SRAM A (W-69) with modern nuclear detonation and plutonium dispersal safety features would be the equivalent of developing a new warhead. The DOD has addressed the primary safety concern by removing the SRAM A from active alert. The long term solution will be achieved by replacing the SRAM A with the SRAM II (W-89), which includes insensitive high explosive and a fire resistant pit…


The Department of Energy (DOE) has placed high priority on modernizing or replacing the W69 with a warhead that incorporates all of the modern safety features. A cost tradeoff study performed while the W89 was in Phase 2 development indicated that it was more effective to replace the W69 warhead on SRAM A with the new W89 warhead, rather than use a Stockpile Improvement Program (SIP) to make safety changes in the W69. Congressional appropriation language mandated that the W89 warhead be compatible with the SRAM A missile…

The bit about the development phase of the W89 is interesting. The Wiki article uses another source to demonstrate that it was envisaged that the W89 would replace the W88. So it could be that the bit about RRW-1 being based on a previous Navy design was a reference to the W89 as possible replacement for the W88. Could be but I don't think so as I will show. Anyway the source on the W89/W88 matter, a 1991 LLNL paper, is interesting for other stuff which should have been discussed.

We note that one of the main issues and main concerns that Congress and others have about RRW is the point about testing. NNSA claims that the RRW-1 (based on W89?) can be certified without testing. But notice that the 1991 LLNL paper actually shows that RRW type manufacturing reforms were already on the pipeline for the W76. I.E. we get,

…Develop a new warhead incorporating IHE to replace the W76 warhead…

Sounds a bit like RRW, no? According to the document we get the following little code

…4; p+3d [4; p+3d]…

This refers to the

…Estimate of the number and kinds of nuclear tests required to accomplish various warhead safety upgrades and other stockpile actions…

It also states,

…Incorporating IHE into the Navy's W76 Mk4 warhead to upgrade its safety is problematical if Rocky Flats
is not in operation. If Rocky Flats is in operation, the W76 could be replaced with a new design that incorporates both IHE and a FRP…

Confusion here is that in the one go the doc talks about this being a W76 "upgrade" (which constitutes a "new" warhead) but then speaks of a "new design" in fact. The RRW issue is full of semantic quicksand like this.


There was a good story on the RRW at Scientific American that suggests something different to the take on RRW-1 that Sublette (?) has written at Wiki. We get this from SciAm,

...In fact, the reason the Livermore design triumphed is because it is based on a former design, one detonated underground before the U.S. moratorium on such experiments in 1992. "[The pit] was nuclear tested four times," says Bruce Goodwin, Livermore's associate director for defense and nuclear technologies. "It's the exquisite test pedigree of the baseline for this design that gives very high confidence that it will work as expected."...


Only a limited number of such primaries have been tested. "It's the SKUA9 design," Goodwin says, one of a series of primaries created by Livermore during the nuclear testing program simply to test the viability of secondaries, and never produced as a weapon. As a result of this prior testing, this first Reliable Replacement Warhead (RRW1), if built, would require no further detonations, according to the NNSA and Livermore...

That's a pretty good source. According to this basically RRW-1 is based on a pit that was developed merely to test secondaries and was never meant to become a weapon, so preseumably we are not talking about the W89 but I would not be so hasty.

But notice the bit about 4 tests. Is it a coincidence that in 1991 the LLNL paper states that replacing the W76 with a new warhead incorporating the features of what is now called the Reliable Replacement Warhead required 4 tests?

So the interesting thing here to expand a little, is that means that in order to develop a "new warhead" incorporating IHE to "replace" the W76 would require, so LLNL estimates, 4 nuclear tests one of these being a production verification test and 3 new warhead development tests.

So given that RRW-1 is a LLNL design what has changed since then that enables LLNL to state with confidence that the development and deployment of RRW-1 as replacement for W-76 incorporating precisely these features can happen without testing? I'm not saying that they can't do it. Perhaps advances in hydrodynamic testing and supercomputer simulations of nuclear explosions enables this but then again maybe not. I think this document suggests that LLNL should provide an explanation. This business about SKUA9 drops a pithy label that means nothing without further discussion.

Ok, we know that the W89 itself was tested (it entered development engineering, phase 3 of a 6 phase process, in 1988 with phase 3 being the last of the development phases prior to the production phases) but this document shows that, at best, the RRW concept without testing is a stunt that could be pulled only in recent times.

Recall that Los Alamos conducted hydrodynamic tests in support of its more revolutionary design,

… LANL announced last year that it had fired its first hydrodynamic shot in support of its design on Sept. 6, 2006, and that "early data analyses indicate that these features will perform as LANL's weapons codes had predicted."…

But it is interesting to observe that, according to the following entry at the Los Alamos Blog, that both respective labs in independent evaluation of the other lab design that,

… LANL's new design, while not tested in its entirety in an underground explosion at the Nevada Test Site, is far from being "untested." In fact, a number of experiments were performed on various facets of the design, including a non-nuclear implosion, diagnosed by radiography. Both the LANL and LLNL teams carried out independent computer simulations of each other's RRW designs. Los Alamos's computer simulations correctly predicted the marginal behavior of the Livermore design, and the successful behavior of their own; Livermore's simulations erroneously predicted the "failure" of the Los Alamos design. On the other hand, the LANL team's calculation of the implosion experiment, carried out prior to the actual experiment, correctly predicted the results, while the LLNL team's calculation did not. This calls into question not only the capabilities of the Livermore designers, but the computational tools they use…

It also states,

… both designs that were submitted were equally far from any Cold War nuclear warhead that was tested before the moratorium imposed by the first President Bush in 1992…


The process by which the RRW design was chosen was deeply flawed, since the members of the committee (five from the military and two from NNSA) that performed hours of in-depth technical reviews over 18 months, voted overwhelmingly for the LANL design. Since the RRW is intended for placement aboard submarines, which carry by far the largest number of nuclear weapons in the arsenal, the Navy's wishes ought to have been paramount, but they were overruled by NNSA's political, rather than technical considerations…

Sour grapes or insight from an insider?

Ok. Back to the Navy and Air Force confusion thing. The waters are further muddied by Chuck Hansen's excellent US Nuclear Weapons: The Secret History. Here Hansen states that the W89 is actually a warhead for the Navy, basically a successor to SUBROC and that the W90 is the warhead for the SRAM II. There is a table that is taken from Hansen's book that shows this.

According to this the W89 was a Navy warhead, that is a nuclear depth charge for ASW delivered by missile from a submarine. In fact there should be no need for confusion. The replacement for SUBROC that Hansen refers to was known as the RUM/UUM-125 Sea Lance and according to that detailed discussion on RUM/UUM-125,

…The payload of the UUM-125A, as well as the RUM-125A, was planned to be a nuclear depth charge with a W-89 thermonuclear warhead. The W-89 was the warhead of the projected AGM-131A SRAM II with a nominal yield of 200 kT…

In other words the W89 was going to be both a Navy and Air Force gig. That deals with our confusion. But Hansen in his book states that the replacement of the W69 (produced in 1971), i.e. the W89, "reportedly" makes better use of fissile material whilst re-producing the same yield. That is, the W89 was to be more efficent than the W69. How so? That would be nice to know given that the Pu pits for the W89 are re-cycled W69 pits. This shows the good test pedigree of RRW-1, by the way. But notice also that according to Sublette (?) at Wiki that RRW-1 will use more fissile material than the W69, at the same yield. So can we be fair and say that RRW-1 might, repeat might be more "safe" and "reliable" than the W76 (the W76 in the docs reveiwed above do give the W89 higher rating on this than the W76 but remember that RRW-1 might be based on the W89 but then does not mean that is is the W89) but that it is less efficient than the W89. If indeed RRW-1 uses more fissile material to produce the same yield as the W89 then RRW-1 is less efficient than the W89.

If critics on the actual safety and reliability of the RRW are correct vis a vis the W76 and Stockpile Stewardship then it might very well be the case that the US is being sold a new warhead that could very well be less "reliable" (given the testing issue) and less efficient than the warhead its design is based on if not the warhead it actually replaces. Surely the W76 would not use as much Pu as the W89, after all the W76 was supposed to be a MIRV based warhead that place a premium on good weight to yield ratios. Just thinkin' out loud.

So is RRW-1 all about;

1.) Design based on W89 and re-cycled W69 pits
2.) Non Berrylium Neutron reflector
3.) Fire resistant pit
4.) More fissile material than the design basis (W89)

OR, SKUA9 pit...if so, what's the deal?

But there is one more thing not discussed.

I have had a previous post showing that the RV for RRW-1 will not be the same as the RV for the W-76, the Mk4. Rather the RV for RRW-1 looks to be the Mk5.

So, in conclusion, we have some, pretty tentative mind you, grounds to believe that the foundation of RRW-1 is the W89 warhead and the Mk5 Re-entry Vehicle.

So we add (5) Mk5 RV.

I'd love to see what the replacement for the W88 is supposed to look like. Remember that the W88 is supposed to have a non traditional geometric basis for the plutonium pit, i.e. non spherical.

...The calculations for a nonspherical primary are apparently orders of magnitude harder than for a spherical primary (a spherically symmetric simulation is one dimensional, while an axially symmetric simulation is two dimensional),...

It would suggest that the RRW-W88 primary would be spherical, if RRW means that the emphasis on yield to weight ratios can be relaxed. That's how the weapons complex will sell it as being more "reliable" than the W88.

Of course, we note that the US House Committee that overlooks nuclear weapons development funding, the House Energy and Water Appropriations Subcommittee, has gutted funding for the RRW until a nuclear posture review is developed that demonstrates how RRW fits into broader US nuclear strategy.

This is interesting because it suggests that on the RRW Team Bush has shot themselves in the foot. From what I have seen, as blogged before, the RRW is actually related to Team Bush's 2001 Nuclear Posture Review. The 2001 NPR introduced the concept of strategic dissuasion. The RRW concept, particularly when combined with Complex 2030, really does fit into strategic dissuasion. Now Team Bush is very secretive. It has only produced two National Security Strategy of the United States documents, for Clinton this was basically an annual affair, and what we know of the 2001 NPR is based on leaks. Clinton had an NPR as well and this was discussed in the context of an Annual Defense Department report. Ok, the Clinton NPR was crap and most of it was classified but the token official release of its findings contrasts with what we know about the 2001 NPR.

By being so secretive about its nuclear weapons strategy the Bush Administration has effectively undercut the RRW in Congress.


We know that the UK has formerly decided to replace the warhead for its SLBMs, the W76, precisely as a part of Trident "replacement". We know that Atomic Weapons Establishment, Aldermaston UK has a relationship with the US weapons labs. So, despite congress, will now RRW-1 work go down to Aldermaston?

I also notice at the AWE website that they celebrated the 100th year since Einstein came up with the special theory of relativity (in 2005). They called it Einstein Year or something. I wonder if any of the smart boys and girls at Aldermaston appreciated the irony that they were celebrating "Einstein Year" when Einstein was such a vocal opponent of nuclear weapons?

Probably not.


Anonymous said...

Given the way NNSA handled RRW, the only thing the Nation is getting that it doesn't already have is a reason for LLNL to remain in the nuclear weapon design business. This is the true reason LLNL "won" the RRW (admitted to if you dig through all the NNSA excuses). Remember NIF? Congratulations America.

Anonymous said...

Great RRW analysis. Not encouraging, but very enlightening.

Sorry for deviating now, but after reading the following article I couldn’t help wonder how many SFPD and FBI skaggs we have in our midst embracing the worst the LAPD has to offer? Probably more than we’d want to believe. In any profession, including science, politics, self interest and bias can brings out the worst.


Report: Complaints up against LAPD

The number of complaints filed with the Los Angeles Police Department alleging misconduct by officers and civilian workers increased last year, a study found.

There were 6,716 complaints filed in 2006, up 3 percent from the previous year, according to an annual report to the five-member Police Commission, which sets policy for the department.

The increase does not account for an immigration rally last month in which officers were accused of using excessive force.

Police officials said the increase showed citizens were more willing to file complaints against department employees. But a police watchdog group said it raises concerns that the department's standards might be slipping.

"The average citizen should be concerned that complaints against the police are up because it signals that something is wrong," said Pete White of the Los Angeles Community Action Network.

According to the report, the number of LAPD employees facing complaint investigations who quit under pressure, were fired or were suspended increased from 421 in 2005 to 451 last year.

Commission President John Mack said only 2.7 percent of the 2,822 complaints of "discourtesy" were sustained last year, and said he asked the department to investigate the handling of such complaints, the Los Angeles Times reported.

The most frequent allegation, made 16.3 percent of the time, was "neglect of duty," the report found. It said the department investigated 391 complaints of preventable traffic collisions and 135 complaints of discrimination.

Anonymous said...

No need to appologize. The LAPD article, as you point out, says something about human nature. The RRW decision was, unfortunately, influenced by similar forces. The Los Alamos National Laboratory was victimized because of it. The same forces are likely to blame for the plight of those who blow the whistle these days. Without accountability, abusive behaviors will thrive and flawed decision making, like the RRW decision, continue. Let's just hope the damage being done isn't irreparable.

Terry Goldman said...

2005 was the International Year celebration of Einstein's "Annus Mirabilis" , 1905, when he published papers on relativity, Brownian motion, light quanta, ... It was celebrated by APS, DOE, and physics and scientific societies around the world, not just AWE. How did you miss it?