Greg Mello
April 14, 1996
 

• The nation does not need nuclear weapons laboratories as we know them in order to maintain a reliable nuclear arsenal.  Such a deterrent can instead be maintained by a remanufacturing and curatorship program, without most of what is now called the "stockpile stewardship" program.  That program is largely oriented toward sinecures at the weapons laboratories and toward retaining (and in some respects, enlarging) the capability to resume nuclear weapons development.  The planned program of above-ground testing facilities, and most of all the entirely unjustified NIF project, can be canceled without danger to U.S. nuclear weapons.



• Peer review between Los Alamos and Livermore is more about the mutual protection of programs and funding than it is about objective, mission-oriented, mutual critique.  Lawrence Livermore's weapons functions are essentially redundant and should be transferred to Los Alamos.  Livermore's civilian functions should be taken over by DOE Energy Programs and administered by Lawrence Berkeley Laboratory.  Los Alamos and Sandia should be downsized in place, emphasizing their stockpile management roles.



• Other sites in the DOE weapons complex are also funded excessively and should be trimmed, in some cases radically so, or closed.  The Nevada Test Site should be formally closed.



• The resulting savings -- on the order of $2 B/yr -- should be used to a) fund substantial new initiatives to aid dismantlement and nuclear security in Russia, parallel to but complementary to the Nunn-Lugar program, and b) renew energy programs within the DOE, emphasizing renewable energy, energy efficiency, transportation, industrial ecology, and related initiatives.



• These new Russian initiatives should be part of a U.S. fresh arms control and nonproliferation effort across a broad front, involving immediate unilateral measures such as an announcement of further dismantlements, a no first use policy, and other steps.



• Conversion of LLNL to civilian missions is an idealistic goal that cannot and will not occur without sweeping change within DOE and the nation.  Yet conversion is composed of a suite of decisions, all of which have independent merit regardless of their relative timing or coordination.  Those decisions can be found in greater detail in the RECOMMENDATIONS section below. 



• The stability of employment at LLNL should not be a factor in national laboratory policy.  Downsizing LLNL could in fact produce net jobs in the region.  Programs of great utility to the nation, however, like much of LLNL's EMATT program, should be shielded during downsizing and the transition to new management.  The final size of LLNL should be determined by the aggregate size of individual civilian projects and programs, established on a competitive basis.

Lawrence Livermore National Laboratory (LLNL) is located about 40 miles southeast of Oakland, California.  It is one of three nuclear weapons labs retained by the Department of Energy (DOE), the others being Los Alamos National Laboratory (LANL) and Sandia National Laboratory (SNL), both of which are in New Mexico. ⁽¹⁾   LLNL has been operated by the University of California since it was founded in 1953. 

Of the three labs, Livermore and Los Alamos both work on the actual nuclear explosive, or "physics package," which contains about 5% of the overall number of parts in a typical nuclear weapon.  Sandia handles the "weaponization" of that explosive, comprising the other 95% of parts in the weapon as a whole. 

LLNL is a large laboratory, with an annual budget of approximately $900 million (M) and about 7,200 full-time personnel, not including about two thousand contract employees.  While its defense mission commands the lion's share (about two-thirds) of its resources, LLNL is a multiprogram laboratory with diverse skills. 

LLNL's core competencies, as the lab describes them [and with comments in brackets like this], are:

• Multidisciplinary integration [this may simply mean that LLNL is accustomed to having large projects]
• Nuclear science and technology
• Lasers/electro-optics
• computer simulation
• sensors and instrumentation
• biotechnology [in selected areas]
• process/manufacturing technology [in selected areas] 

The logical place to begin our examination of the future of LLNL is with the normative future of the DOE weapons laboratories as a group.  Historically, the DOE nuclear weapons design labs could be said to have begun in 1943 with the construction of "Project Y" on a remote location in New Mexico called Los Alamos.  That lab was tasked with the design and assembly of the first nuclear weapons.  But fully 97% of the work, and the expense, of the Manhattan Project were elsewhere, as was the great bulk of the technical expertise.⁽²⁾

Today, it is not at all clear that the nation still needs nuclear weapons laboratories as we know them in order to maintain a reliable nuclear arsenal.  If the manufacturing plants of the weapons complex, with LANL and Sandia already among them, retained or in some cases acquired the physical, metallurgical, and engineering expertise necessary to remanufacture nuclear weapons to replace those we have, the very large scientific enterprises that define the three weapons labs today would not be necessary.  With fully tested weapons in the arsenal and an enormous experimental database, with downsized remanufacturing facilities ready to replace parts as needed, and with weapons physicists and other scientists organized in small teams to act as curators of weapons lore, stockpile stewardship tasks could be easily absorbed in what is today called the stockpile management program.  This approach has been described in more detail elsewhere.  ⁽³⁾

For reasons that will shortly become more clear, among the three labs it is Sandia that has the best reasons to continue as a full-fledged, albeit downsized, lab.  In brief, this is because Sandia's work is more closely tied to weapon safety, security, and reliability than is the work of the other two labs. 

Unfortunately, the weapons labs, which are more politically powerful than the DOE which funds them, can be relied upon to intentionally magnify their importance and misrepresent the facts in order to preserve their status and funding.  One LANL weapons manager put it bluntly to the DenverPost last year:  "We have to scare the politicians into doing the right thing." The right thing, of course, is to keep delivering about $3.5 billion every year to the three labs. 

Instead of what could be a relatively simple and inexpensive curatorship approach, the DOE, captive of its labs, has opted for what it calls its "science-based stockpile stewardship" program.  This option just happens to be about the most expensive approach to maintaining a nuclear deterrent that can be devised, costing, together with the rest of the nuclear weapons program, more than twice as much as was spent on nuclear weapons in the middle 1970s, during which time there was an active program of materials production, active design, and nuclear testing. ⁽⁴⁾   Despite its name, DOE's program is less "science-based" than it is science debased, because it is driven by program needs, in this case weapons program needs, and not scientific inquiry.  The vast bulk of the expense is directed into very narrow fields of investigation. 

When World War II was over, the scientists at Project Y remarked how little science they had done on the mesa as compared to their former work in universities, despite the unprecedented collection of stellar minds and the essentially unlimited funding.  How much more so is this true today, when, as we shall see, the main purpose of the program is little more than maintaining sinecures and skills until the hoped-for day when active nuclear development is politically fashionable again.

LLNL was founded because of the perceived need to speed the development of thermonuclear explosives, to create a competitive environment in the development of nuclear weapons in general, and to provide peer review in nuclear weapons design.  Although after decades of competition the nuclear weapons programs in Livermore and Los Alamos have evolved to differ in some details, the overall functions and program areas in each lab's nuclear weapons program are essentially identical.  But does the complexity of maintaining nuclear arsenals in the absence of nuclear testing require that redundancy? 

U.S. nuclear explosives are based on very sophisticated analysis and thorough testing.  But they are very robust, and not very complicated, devices.  As noted above, just about 5% of the parts in the weapon are contained in the explosive device itself.  While there have been several serious design errors -- errors which destroy confidence in the two-lab peer review system -- extensive stockpile surveillance data shows that nuclear weapons have been kept in a highly reliable condition with a very small effort from the two design labs. ⁽⁵⁾   Once a weapon was fully tested, nuclear testing played essentially no role in maintaining the stockpile. 

Within the weapon as a whole, the nuclear explosives themselves have accounted for a very small fraction of the problems encountered in routine surveillance, and a small proportion of the number of these problems that required fixes.  The upshot is that the modern weapons selected to remain in the U.S. stockpile into the next century are extremely safe ⁽⁶⁾  and reliable, and the explosive parts of those weapons are in fact the most reliable parts of them. 

What has caused serious problems in the past, and can be expected to do so in the future, is the substitution of untested designs for the fully-tested ones now present in the stockpile.  As an independent panel of nuclear scientists warned DOE in 1995: 

However greatest care in the form of self-discipline will be required to avoid system modifications, even if aimed at "improvements," which may compromise reliability. ⁽⁷⁾ 

The most reliable weapons are thus the ones we have tested.  No new nuclear designs are needed or wanted.  The implication of this is that maintaining the nuclear arsenal requires skills and capabilities, but only a small fraction of the design personnel now employed at the weapons labs. 

These realities raise serious questions about retaining two huge design laboratories with essentially redundant missions.

In early 1992, then-Chairman George Brown of the House Science, Space, and Technology Committee wrote DOE Secretary Watkins urging him to establish LLNL as a civilian technology laboratory, transferring LLNL's nuclear defense missions to Los Alamos within a context of radical downsizing of DOE's nuclear programs. 

Although Brown's idea didn't take, a bill was subsequently introduced by one Republican in the House for a DOE lab closure commission, modelled after the military base closure process and motivated by a desire to eliminate the redundancy and waste embodied in the DOE's large laboratory system. 

Watkins' successor, Secretary Hazel O'Leary, brought George Brown's proposal into the Clinton administration when she spoke of the possibility of a "green lab" in California.  No one had to ask what lab she was talking about. 

Her Advisory Board Task Force on the Futures of the DOE National Laboratories, usually called the Galvin Panel, recommended after a full year of careful study that LLNL's defense missions gradually be transferred to Los Alamos. 

The Congressional Budget Office, writing in 1994, suggested that closure of LLNL was an option worth considering, along with downscaling of DOE's stockpile stewardship plans.  They were critical of the National Ignition Facility (see below) and of DOE's ambitious program to advance nuclear weapons physics. 

All these threats to LLNL's nuclear weapons future were successfully defeated.  But the underlying reasoning behind them remains sound, and as budget pressures continue to mount, LLNL's raison d'etre can be expected to undergo close scrutiny again.

NIF-ty Science?

Enter the National Ignition Facility (NIF), a giant laser device that will be as big as a small stadium, costing about $1.2 billion dollars for design and construction alone.  Given the above questions about LLNL's future, NIF is widely recognized as a necessary -- though perhaps not sufficient -- bulwark for the lab's long-term stability.  NIF's 192 laser beams, if all goes well, will create X-rays that focus on a millimeter-scale pellet of tritium and deuterium to induce a tiny thermonuclear explosion in a process called "inertial confinement fusion", or ICF. 

Ostensibly, the NIF project has three primary goals:

• To help maintain the U.S. nuclear stockpile
• To advance fusion energy technology
• To contribute to science 
• Consider first NIF's defense role.  Politically, it is the only salient one. 

The relative simplicity and high reliability of the physics package has already been discussed above.  Within the nuclear explosive in turn, the so-called "secondary" has been the most reliable part, having experienced only one or two trivial problems that were very easily fixed.  These secondaries are sealed and have no high explosives in them.  They have a long life, and are relatively inexpensive to remanufacture as needed.  The U.S. now possesses an enormous manufacturing overcapacity in this area, which can be maintained in a standby condition at "minimal" cost. ⁽⁸⁾ 

These facts are important because NIF's contribution to the stockpile stewardship program lies in its proposed role in studying nuclear weapons secondaries -- components which have had essentially no problems and need little study.  If there were problems in the future (which is not expected), NIF could not really discover or solve them in any case.  Those are tasks for routine surveillance and remanufacturing, which are necessary and sufficient, as well as far simpler and cheaper. 

NIF's link to stockpile maintenance is thus indirect, through its role in maintaining and training staff.  The actual data it would provide is admitted to be secondary -- at least as far maintaining existing weapons is concerned. 

The whole idea of lasers is for understanding the physics of secondaries, but also more particularly, for maintaining that cadre of scientists who both understand the fusion process and all the things that go along with that...The stewards really are more important than the equipment...the purpose of the Stockpile Stewardship program is in fact to maintain the stewards, and the right type of experiments. ⁽⁹⁾ 

In the final analysis, then, the entire so-called "science-based stockpile stewardship" program, and NIF especially, is really more about maintaining scientists than it is about maintaining warheads. ⁽¹⁰⁾ 

As for the second justification for the project -- energy production -- NIF is not on the critical path to any practical fusion energy source, if indeed "practical" and "fusion energy" can be used in the same sentence.  One reason is that NIF uses glass lasers; glass lasers cannot be repeatedly fired because the huge pieces of laser glass heat with each firing and lose the precise alignment required.  Heavy-ion drivers, not lasers, appear to be best route to a higher energy gains in the ICF program. 

And ICF itself is intrinsically a poor choice of fusion technologies, because its energy release is discontinuous, as well as for other reasons. 

More generally still, fusion energy itself is a poor bet, for even if fusion should become technically feasible, (i.e.  if net overall life-cycle energy production ever becomes positive and the materials and nuclear waste issues are ever settled) it is unlikely to ever be economically competitive.  Widespread renewable energy sources are already much cheaper than fusion will ever be.  Simply put, the least expensive and most practical fusion energy source is the sun.  Pork-barrel and special interest politics, embroidered by pseudo-scientific myths and driven by career investments, have provided most of the push for large-scale fusion research in the U.S and continue to do so.  This is privately understood within the DOE labs, where the history and prospects for fusion research are the butt of many cynical jokes. 

There is no doubt that NIF could be an interesting tool for physicists; this last justification for the project is the only one that makes sense.   Yet, were it forced to compete for scarce funds with other experimental priorities within the physics community, it is doubtful that it would be funded. 

NIF is beset by other problems.  Serious proliferation issues continue to swirl around the project, ranging from the direct dissemination of valuable weapons design expertise to other countries, to the indirect diplomatic effect of investing more than a billion dollars in facility whose utility for improving weapons design codes, and hence for facilitating new designs, far exceeds its vanishingly small utility for maintaining existing weapons. 

Whether NIF will even work is an open question; DOE has admitted that expert opinion is guarded on the subject. 

All in all, NIF is a perfect example of the kind of big-ticket science the labs have come to love:  huge appropriations (DOE estimates its life-cycle cost at $4.5 billion), with absolutely no required products or deliverables. 

The final question in the present context is:  will NIF and its associated programs be large enough to serve as an anchor for LLNL's future?  The answer is, probably not.  The facility is expected to create fewer than 300 long-term jobs at LLNL, about 4% of the current lab population.  That's about $15 M for each permanent laboratory job created.

Given the uncertainty surrounding LLNL's future as well as that of the giant new project designed to keep the lab open, LLNL might well consider exploring some new missions.  In theory at least, LLNL has the skills to contribute much more than it does on a variety of valuable scientific and technological fronts.  Some of these are:

• Renewable energy and energy conservation R&D
• Advanced transportation technologies and underlying sciences
• Environmental R&D
• Materials science and manufacturing technologies
• Basic science, pure and applied, in LLNL's traditional specialties
• Advanced computation 
• Cooperative work for industry in these and other areas. 

Thematically, these missions could be linked by the common purpose of advancing technology that would support, and assist transition to, a sustainable society.  These missions could thus be largely described by the phrase "sustainable technology."  LLNL already has projects in all the above areas.  Some of these programs, particularly those grouped under LLNL's Energy, Manufacturing, and Transportation (EMATT) Directorate, have excellent track records. 

In addition to this list, LLNL's work on nonproliferation could continue to provide a valuable service to the national security community.  It would be of benefit to institutionally divorce such nonproliferation work from the work involved in maintaining nuclear weapons, however, and thus give LLNL a focus on arms control and disarmament and Los Alamos the stockpile management role. 

So there are alternatives for LLNL other than redundancy and fiscal irresponsibility on the one hand, and closure on the other.  But there are some serious complications involved in making such a shift.

These complications include:

• LLNL managers will fight conversion with every tool available.  This is because LLNL's nuclear weapons mission is perceived as a unique national mission and funding niche, while the above new missions are not.  These managers could, in theory at least, get behind a program that had long-term potential to maintain LLNL's budget at a very substantial level or even to expand it.  In fact, however, this is unlikely to happen unless large political shifts clearly threaten to extinguish LLNL's nuclear weapons programs.  The mere threat of cutbacks has, historically, has only redoubled LLNL's advocacy of its nuclear mission.



• The large-scale funding implied by conversion really comprises national conversion as well as lab conversion, and this is at present politically unrealistic.  There is now no visible support in either Congress or the Executive to make this transition. 



• The culture of secrecy, the lack of support for new civilian programs by weapons-obsessed managers, and the competition for funds between existing weapons programs and new initiatives all make an extremely challenging environment in which to attempt new missions.  It may well be that in most cases scarce federal funds would be better spent in environments without these drawbacks.  In short and overall, the corporate culture of LLNL is antithetical to conversion efforts and may well doom them.



• Trauma will probably therefore be needed at LLNL to bring about cultural transformation before any new civilian missions can succeed at any but a small scale.  Conversion, by definition, involves inadequate trauma.



• The scale of the new programs, the type and range of projects pursued, and the effectiveness of these technologies in meeting societal goals are all threatened by the new congressional requirement that technology transfer funds (the bulk of the funds now available for civilian missions at LLNL, including the EMATT program) be used only in ways that benefit the military mission of the lab.



• Costs at LLNL remain high, despite some improvements since 1990.  It now costs some $124,000 per year to maintain each LLNL employee, or about $250,000 per year to maintain each scientist, engineer, or manager with all supporting staff and costs.  Many other fine institutions, e.g.  universities, are available to do advanced scientific and engineering work at lower costs than this. 

These "complications" amount to a near impossibility for what is conventionally called "conversion" at LLNL.  But conversion is only one option for change.

Option Zero is the DOE's current plan.  While, by definition, it is politically realistic in the short run, it has the serious problems discussed in the text so far.  These problems will mount over time, either causing distress for LLNL, distress for the world, or both. 

Option One involves nearly full-scale conversion of LLNL to a new primary mission, probably "sustainable technology" as discussed above.  It has the complications discussed above.  The advantages, if such a transition were possible, are also plain: 

• The scale of work, if it is managed well, is large enough to accomplish significant technical advancement in some or possibly several relevant technical areas.  Such advancement could permanently alter, and for the better, the economic feasibility of these sustainable technologies, allowing society to more easily make critically important watershed choices. 



• The scale of work constitutes a commitment that implies other synergistic elements of a national policy, elements that could complement R&D by, for instance, creating new markets through government incentives and restrictions, demonstration programs, and other actions.  One such program might be a "Sustainable Communities" initiative, where an initial government subsidy would be used to provide incentives to demonstrate a variety of new sustainable technologies, including perhaps in Livermore. 



• Economic trauma to the Livermore community is largely avoided. 



• Talent is kept in place at LLNL and not allowed to disperse; this is of particular concern in the current EMATT program, which could in many ways be the "seed" of a converted LLNL if it is not disbanded.  (Current budget cuts are already threatening to do this, in substantial part.) 



• Providing ample new work can, one theory goes, ease the political transition of LLNL in Congress.  Such ample new work, this theory continues, may in fact be the only way to politically deal with the power now vested in LLNL.⁽¹¹⁾
  
  It may be that no political sponsors, and no political momentum, can develop behind new programs unless they be of a certain scale.  "Make no small plans." 



• Option One, if it were possible, might aim in a lab downsized somewhat from present, comparable to the period of relative stability LLNL enjoyed in the 1965 to 1975 decade, i.e. in the $700 M range.  This option was developed by Tri-Valley CAREs in 1994. ⁽¹²⁾   Further analyses of specific expanded missions are now in development by Tri-Valley CAREs. 

Option Two is like Option One, but smaller.  It thereby gains some fiscal feasibility, and deals more effectively with the cultural transformation required.  Many of the disadvantages of Option One remain, however, coupled with the new problem that political advocacy for small new programs may itself be small. 

Option Three involves downsizing LLNL, then converting to new missions.  It addresses the hard truth that the policy decisions of downsizing and offering new missions are not, from a national perspective, intrinsically or logically linked in any way.  This fact is repeated frequently within the lab when the subject of conversion is broached.  To attempt to link them is likely to mangle both decisions, leading to failure to successfully implement both the downsizing and the creation and assignment of new R&D tasks.  Put another way, the rhetoric of conversion tends to set value on maintaining LLNL as an end in itself, which effectively undercuts change.  The rhetoric of conversion is often more paralyzing than creative.  Option Three avoids this trap. 

Some of the disadvantages of the previous options are avoided:  new civilian programs need not be grown in a weapons culture, and the restrictions on funding that tie civilian projects to the weapons program will be largely gone -- because the weapons money will be largely gone. 

Certainly LLNL's leadership and political sponsors will be more eager in Option Three to embrace new programs of whatever size because that is all that will be available.  Lack of size, uniqueness, and permanence in the new programs will loom smaller as obstacles. 

Studies of defense downsizing in other communities, e.g. Boeing in Seattle, have shown that the forced release of highly-skilled knowledge workers into a community with an existing high-technology employment base can generate new technology products, a large number of business start-ups, and economic diversification.  ⁽¹³⁾  Conversion, in other words, need not happen at a single site. 

That is, given the economically-isolated nature of nuclear defense work, a telling number of highly-trained and capable former workers are likely to enter the "real" economy (regionally, for the most part) and begin producing goods or services.  Many will be considerably more economically productive than they were at LLNL.  It is quite possible that greater net regional jobs and greater net regional income could be the result, although most of the displaced workers will themselves lose income in the process, and this will probably be especially true for those who can least afford it. 

Given a reasonably healthy economy, however, if one believes that LLNL's scientific and professional staff could not find productive work, it is then reasonable to also question whether this talent is really, in fact, all that talented. 

The principal disadvantage of this option is that there will be essentially no technology transfer funds to ease the transition into civilian research within the lab.  Growth in new missions must occur after many of the excellent seed projects are disbanded.  This disadvantage is somewhat diluted by the relatively high concentration of high-technology jobs in the LLNL region, which could serve as a skills buffer. 

Some facilities at the site could, under this option and the next, be operated by Lawrence Berkeley Lab (LBL), out of which LLNL was originally formed.  It might well make sense to consolidate the remnants of LLNL under its original sponsor, for a variety of reasons. 

Option Four is one of closure or near-closure, partial privatization, and subsequent incremental conversion.  It is the logical extension of Option Three.  The principal advantage of this option is clarity.  There would be no LLNL managers to fight conversion, and most economic uses of the site will be greeted with enthusiasm and weighed against each other with fresh eyes. The converted site, whatever institutional ownership it may have, and whatever ownership and funding mosaic may come, would represent a close partnership between government and industry.  This augurs well for creating regional employment, quite likely more so on a dollar-for-dollar basis than was LLNL.

Recommendations 

Common ground emerges from a close examination of the last four options above.  That common ground comes from the realization that conversion of LLNL is not one decision but a large suite of decisions, which will or will not be individually made by a large number of actors.  There is, gratefully, no superintending entity that can centrally and logically plan the intricately-linked decisions that would comprise conversion without trauma.  These decisions -- the decisions to end some programs and to begin others -- are independently desirable regardless of their timing, scale, or linkage.  Those qualities depend on political circumstances which are impossible to understand, predict, or control, and which are certain to change in unknown ways.  We therefore offer the following mutually-supporting set of recommendations, all of which we believe have merit independent from all from the others.  Taken as a whole, however, they are even better than the sum of the parts.  While a few of them may seem far afield, actually all of them bear centrally on the future of LLNL.