National Ignition Facility and Science-Based Stockpile Stewardship Resource Page
|To:||Energy Secretary Bill Richardson; Undersecretary Ernest Moniz|
|cc:||Members of SEAB Task Force Review of the National Ignition Facility|
|From:||C. E. Paine, M. B. McKinzie, T. B. Cochran, Nuclear Program, NRDC|
|Subject:||Analysis of the November 18, 1999 report issued by the University of California President's Council National Ignition Facility (NIF) Review Committee|
|Date:||December 14, 1999|
As you are no doubt aware, the President's Council of the University of California recently issued a report, dated November 18, 1999, on the status of the National Ignition Facility project that had been prepared by a review committee chaired by Steven Koonin, vice president and provost of Cal Tech, [hereafter referred to as "the UC Koonin Report, or "UC/KR-99"]. The findings and conclusions of this report are seriously flawed. It repeats errors contained in two previous NIF reviews, likewise chaired by Professor Koonin (NAS/NRC 1990 and 1997), that contributed to the present program debacle at Livermore. This memorandum provides a critique of the errors and omissions in the UC Koonin Report, and compares its findings and recommendations to those of the 1997 NAS Koonin Review (hereafter "NAS/KR-97").
Reverse Mission Creep
1. UC/KR-99: "The [National Ignition] Facility is designed to provide experiments in thermo-nuclear ignition, nuclear weapons effects testing, and high energy density science in support of the nuclear weapons Stockpile Stewardship Program (SSP)." (p. 1)
RESPONSE: Actually, at the time NIF Project was approved for the construction phase (May 1997), "nuclear weapons effects testing" was a subsidiary, distant, and somewhat implausible future mission for the NIF which DOE testified in court documents might or might not be pursued, depending on future modifications to the facility and/or the construction of an as yet unfunded second target chamber. Now, after the fact, nuclear effects testing has apparently been elevated to one of three major missions for the project, while the NIF project's primary mission -- achievement of fusion ignition in the laboratory -- has been subtly downgraded to "provid[ing] experiments in thermonuclear ignition." This "reverse mission creep" is symptomatic of a project in disarray. As discussed in greater detail below, there is now very little realistic prospect of "ignition" achieved by the "National Ignition Facility" project.
No Technical Issues?
2. UC/KR-99: The first charge to the UC Committee from UC President Atkinson: "Describe the technical issues involved in the NIF's projected cost and schedule overruns . . ." (p. 1)
RESPONSE: By its own admission, the UC Koonin panel made no serious attempt to fulfill this charge, apparently limiting its review to the very people responsible for leading the project to the brink of collapse:
"Thus, the Committee's investigation consisted of interviewing key people who have or had NIF management responsibility, as well as others who are or were leading reviews of the Project. The Committee did not conduct an ab initio technical or managerial review of the Project." (p. 2)
3. UC/KR-99: "The NIF is designed to focus 192 powerful laser beams onto a target and create a high-temperature environment. Among other applications, that radiation will be used to compress and heat a small capsule of deuterium to ignition and self-sustained fusion burn." (p. 2)
RESPONSE: This statement manages to be doubly misleading. On the one hand, attaining ignition and self-sustained fusion burn can scarcely be considered just one "among other applications," as it is the primary raison d'Ítre of the entire project! On the other hand, the statement that the 192 beam facility "will be used" to compress and heat a DT capsule to "ignition and self-sustained fusion burn" is just speculative cheer leading. The project now lacks even a computational design for a "baseline" cryogenic ignition target that can be fabricated, and it has already adopted and discarded at least three such designs since the "Critical Decision Two" -- approval for detailed engineering design -- was made in June 1994. Livermore now admits that expected levels of fluence damage to the laser optics will limit NIF operations to half power -- insufficient for ignition -- for an extended period, and perhaps permanently, depending on the outcome of additional R&D and laser design changes that cannot be validated until the first bundle of eight beams is constructed and operated.
To restore at least some of the performance levels specified in the baseline design, Livermore is proposing in essence to continue a major R&D program, including a redesign of the entire final optics assembly, on the very facility that it is simultaneously constructing. This is usually a sure fire formula for either massive technical failure, or a bottomless pit of spending, and sometimes it produces both.
If the DOE sponsors are now willing to acknowledge that the technical and scientific prospects for attaining ignition have receded -- NRDC has long argued that the NIF's scientific and technical underpinnings were grossly overstated to begin with -- then the NIF project must be judged likely to fail in the achievement of its primary mission, and therefore it should be terminated or significantly redirected before even more public funds are wasted.
4. UC/KR-99: "There are three major components of the NIF Project:
- the conventional facility construction: the laser building, target area, and optics assembly building;
- the laser and target system infrastructure that holds the laser equipment (including large support structures, vessels, beam path enclosures, cabling and utilities); and
- the laser equipment, including the laser glass." (p. 2)
RESPONSE: Actually, there are FOUR major components to the NIF Project. The UC Report omits entirely from its review the status of work on the ignition target assembly (hohlraum and capsule) target positioning system, and diagnostic instrumentation, all of which are in varying states of disarray. As noted, there is no computationally validated "baseline" ignition target that can be fabricated, and the capabilities needed to diagnose ignition experiments driven by NIF's longer (20ns) pulse do not exist. It's interesting to note that most of the shots conducted on both Beamlet and NOVA were short pulses (1-3 ns) with none longer than 10 ns. Livermore has essentially no experience, and no capabilities, for diagnosing Stockpile Stewardship experiments at longer pulse lengths. These "remain to be developed," we are told, yet such diagnostic capabilities are absolutely indispensable for "tuning" the laser beam, hohlraum, and target to perform together within the narrow temperature window of 250 - 350 eV that LLNL tells us is supposed to lead to ignition! It's a little like telling the pilot of a brand new 767, "Don't worry about those missing avionics, son. We can install all that when we're airborne."
5. UC/KR-99: "Laser and Target System Infrastructure and AssemblyÖ. Several additive factors have increased the projected costs of this segment of the project. These include the greater-than-previously-appreciated complexity, the unanticipated need for stringent cleanliness protocols during assembly, and the more complete design packages needed for outside procurement." (p. 3)
RESPONSE: This explanation amounts to a grotesque whitewash of the project's difficulties. To suggest at this late date, well into the construction phase, that the NIF project's problems have arisen from its "greater-than previously-appreciated complexity" or the "unanticipated need for stringent cleanliness protocols" is just plain bizarre. It's like a group of heart surgeons responding to a malpractice suit by arguing that the procedure they claimed could be performed safely was "more complex than they had previously appreciated" and besides, no one in medical school had ever taught them about the need for sterilization to guard against bacterial infection!
The extreme Rube Goldberg complexity of the proposed 192-beam NIF design was painfully evident even to untutored outside observers of the project from the very beginning. And how can the world's allegedly foremost institution for solid-state laser research seriously maintain that they somehow "failed to anticipate" that sufficiently protecting the amplifiers, laser cavities and optics from contamination and impurities of all kinds was of the utmost importance?
There are only two possible responses. Either LLNL is not technically competent to build and operate the NIF, in which case the project should be terminated, or somebody is lying, and these are not the real reasons why the project is in trouble. While we have never been ones to discount evidence of national laboratory incompetence, based on what we have learned about this project and national lab behavior in general over years, NRDC is inclined toward the "somebody-is-lying" theory. And we believe that the fundamental problem lies not in "mismanagement" but in the misrepresentation by LLNL and its supporters of the real state of the project scientists' understanding of all the physical processes involved in achieving laser fusion in the laboratory. In short, what ails this project is institutional self-aggrandizement and technological hubris, and the observed "management deficiencies" are merely derivatives of this underlying condition.
6. UC/KR-99: " Laser Equipment Design. Inadequate operating budget allocations in the early years of the Project led to an underfunding of the technology development activities. As a result, prototyping of some laser subsystems was delayed, causing a 12-18-month delay in the final design of these elements and a corresponding increase in the estimated costs." (p. 3)
RESPONSE: This explanation strains credulity. If alleged "underfunding" of the "prototyping of some laser subsystems" during "the early years of the project" was a problem, why wasn't this problem mentioned in the NAS/KR-97 report. On the contrary, this report stated:
"The Beamlet [NIF beamline prototype] has been operational since 1994 and has provided a test bed to gain confidence in the NIF laser architecture and design, operating at the fluence levels planned for the NIF laser. The Beamlet laser has validated almost all aspects of the design of the NIF laser. The exceptions are the final focusing optics and the lower damage threshold of the rapidly grown potassium dihydrogen phosphate (KDP) crystals; there are adequate plans for dealing with these two remaining issues (emphasis added)"
"No Technological Showstoppers?"
7. UC/KR-99: "Two remaining technology issues involve the large-scale production of laser glass within specification and cost goals and the operational lifetime of the final optics. The Project goal was to scale the production rate of laser glass by a factor of ten while decreasing the unit cost by a factor of three. While both glass vendors have met these goals, the details of the manufacturing process must be refined to meet all the required performance specifications. Both the Project and its reviewers expect that this will be accomplished." (p. 3)
RESPONSE: What does it mean to say that the glass vendors (Schott and Hoya) have "met the goals for large-scale production" when the laser glass still FAILS to meet the needed performance specifications? Would you buy a set of tires at a 66% discount -- the "factor of three" reduction in unit cost -- if the fine print on the treadwear warranty read, "Guaranteed for only 4,500 miles," an effective tenfold reduction from typical levels of endurance, or "Unsafe at speeds over 35 miles per hour," when you need to drive at 65 mph. This is no minor problem.
Tons and Tons of Glass, But None of It Meets Spec
UC/KR-99 mentions in passing only the problem of "excess water in the glass," and that this is being "addressed adequately" -- whatever that means -- but makes no mention of the volume inhomogeneities, surface ripples, and platinum inclusions that have also affected the amplifier glass produced by the more economical "continuous pour" technique. Why is the problem of producing specification-compliant amplifier glass only now coming to the surface, when the project is well into the construction phase?
One would have thought this project milestone to have been an essential pre-condition for transition to the construction phase in May 1997, as laser glass performance strongly affects the performance levels for the facility attainable at tolerable cost, and conceivably the cost-benefit calculus for the entire NIF project. In its "Executive Summary," the NAS/KR-97 report concluded, under the heading "Technological Readiness," that the "NIF baseline uses a laser architecture selected to meet the required performance levels at acceptable cost while respecting the limits imposed by laser fluence damage." (p. 3) If this is all you read, you might easily have concluded that the fluence damage/cost problem had been taken care of.
But buried in the body of the report (p. 21) the authors of NAS/KR-97 sought to cover themselves, if not the taxpayers, by noting in passing, "Peak fluence from a 4-pass amplifier through the optical beam path to the KDP harmonic-conversion crystal remains below the damage fluence of the optical elements, if they are procured from the highest quality fabrication. The baseline design uses lower-cost components that can be more readily mass produced but have yet to meet the required damage thresholds (emphasis added)." Now, two years into the construction phase, and with Beamlet dismantled, we learn that Beamlet never met the peak fluence requirement even with the good glass!
In yet a third place in the NAS/KR-97 report (p. 25), one finds a passing reference to the strong dependence (in principle) of cost "on the maximum allowed optical fluence of components along the laser optical beam path," but this caveat was immediately followed by an assurance that the NIF laser "is designed to operate at the maximum fluence consistent with reliability and projected performance." The NAS/KR-97 report further stated that the NIF laser "is expected to operate at a fluence level approximately twice that of NOVA" but "below proven damage levels of the optical components in the beamline path (emphasis added)." It also stated that a program had been instituted "to work with optical component vendors to further increase damage-fluence values by improving finishing methods," leaving the clear impression that these values were already minimally acceptable. But nowhere did the NAS/KR-97 report highlight or quantify the sensitivity of the overall project cost to the then (and still ongoing) failure to obtain acceptable glass components using the cheaper production methods included in the "baseline" cost estimate. (Indeed, the problem of damage fluence levels cannot be solved by better glass alone, and is prompting a major redesign of the final optics assembly midway through the construction phase.)
Under the heading "Remaining Hurdles," and the subheading "Optics damage thresholds and cost," (p. 35), the only "hurdle" mentioned in the NAS/KR-97 report was procuring affordable harmonic conversion crystals, as follows:
"While the current cost estimate contains funds to purchase slow-growth KDP crystals if necessary, the effect on schedules must be considered if the fast growth crystals fail to meet damage threshold requirements." (p. 35)
In other words, the NAS Koonin panel either never fully understood, or perhaps elected not to reveal, the difficulties in procuring affordable amplifier glass that meets minimally required damage fluence levels, and the sensitivity of the total project costs -- including future O&M costs -- to this particular factor.
Radiation Damage Reduced To Tolerable Levels?
8. UC/KR-99: "The second technological issue is damage to the final optics from the high intensity ultraviolet light to which they will be exposed, which will shorten their lifetime. This damage issue is not a technological "show-stopper," but rather a potentially serious limit on operational cost and efficiency. More work will be needed to better understand the underlying damage phenomenology so that the facility will be able to operate economically at full power." (p. 3-4)
RESPONSE: What does it mean to say "radiation damage" is not a potential "technological show-stopper," but rather "a potentially serious limit on operational cost and efficiency?" Obviously, at some expected mean-time-between-failure (MTBF), the latter approximates the former, either because the final optics are breaking all the time, or because the performance limits imposed so compromise what can be achieved with the NIF facility that it is no longer worth further expenditure of public funds. In place of the detailed analysis needed to honestly review this question, UC/KR-99 offers only an unsupported disclaimer (see next point) that technical issues are being "addressed adequately."
Are Technical Issues Being "Successfully Handled"?
9. UC/KR-99: ". . . the Project reports that a number of the technical issues have been solved or are being addressed adequately. These involve amplifier performance, contamination control in the facility, excess water in the glass, and concerns regarding radiation damage to the optics. Again, the [UC President's] Committee did not engage these technical issues, but talked to those who are reviewing them. The technical peer review groups believe that the technical issues can be successfully handled by the Laboratory, and the Committee has no reason to believe otherwise. The Laboratory has an excellent track record of resolving technical issues of this sort (emphasis added )." (p. 5)
RESPONSE: "NO REASON TO BELIEVE OTHERWISE?" Surely the Koonin panel must be jesting. A Cabinet official of the United States government was induced by LLNL briefings to provide false and misleading testimony to Congress, and the Koonin panel can find no reason to question LLNL's assurances regarding the technical status of the NIF project. In reality, LLNL has a terrible track record of "solving problems of this sort." The expected performance of every ICF machine ever constructed at Livermore has been wildly overestimated, leading to the construction of ever more powerful glass laser drivers based on a seemingly endless trail of erroneous performance projections.
A Reality Check for the Koonin Panel
LLNL's Mirror Fusion Test Facility (MFTF) of the late 1970's - about a $500 million dollar facility in today's dollars -- was so poorly designed that it never operated, not even for a single day! In March 1985, the $300 million "Cottage" underground test of LLNL's x-ray laser focusing concept was so poorly designed that the experimental results could not be meaningfully interpreted. Several other underground x-ray laser experiments - including Diablo Hawk (1978), Cabra (1983), Correo (1984) and Goldstone (1985) are known to have been partial or complete failures. A number of the veterans from this failed program now work on the NIF project.
After 26 years and the expenditure of some two billion dollars of the public's money, the Laser Directorate's uranium Atomic Vapor Laser Isotope Separation (AVLIS) project still failed to demonstrate a commercially viable production scale enrichment process, and last July the funding for the project was terminated by its newly privatized sponsor, the U.S. Enrichment Corporation (USEC), which lacks the very deep pockets of the American taxpayer that LLNL has drained for so long. (Incidentally, some of the management from this failed AVLIS demonstration program has now been brought in to help run the NIF Project!) In the late 1980's, LLNL's production scale AVLIS demonstrator for plutonium met a similar fate, falling two years behind schedule (despite numerous reprogrammings and the expenditure of some $90 million per year on the project) before both DOE and Congress pulled the plug in 1990.
Lest the SEAB conclude that LLNL is uniquely incompetent, between 1980 and 1990 Los Alamos expended on the order of $1 billion dollars -- to this day no one really knows exactly how much money was wasted -- on two large ICF projects, the Antares and Aurora chemical lasers, that (for quite different reasons) utterly failed to meet their scientific objectives. Antares was preemptively, secretly, and illegally destroyed at huge cost to the taxpayer before completing its authorized experimental program, and the funds secretly diverted to the Aurora KrF laser, an unqualified technical failure that was abruptly terminated when its inherent conceptual flaws were finally subjected to outside scrutiny in 1990. In fact, it was only the unsolicited intervention of a dissident Los Alamos ICF expert -- Leo Mascheroni -- that saved the 1990 Koonin Panel from an embarrassing public endorsement of KrF Aurora.
As for LLNL's success in "controlling contamination," the lab's main site was designated by the EPA as a Superfund cleanup site in 1987, in order to cope with the tritium, plutonium, and volatile organic chemicals contaminating the soil and groundwater. Site 300 received Superfund status a few years later. The Defense Nuclear Facilities Safety Board recommended, and DOE ordered LLNL's plutonium facility shut down for 5 months in 1995 and again, when criticality and other safety violations persisted, for about 18 months in 1997 - 1998.
Throughout this sorry record of scientific hubris and technical failure, there is one constant -- a pervasive lack of accountability. At DOE's weapons laboratories, apparently, nothing succeeds like failure.
Where is the "Baseline Target"?
10. UC/KR-99: "Advances in target physics suggest that many of the same experiments might now be possible at less than full laser energy, although the degree of technical risk associated with many of these experiments increases as the available energy decreases." (p. 4)
RESPONSE: What "advances" in target physics? The situation facing the National Ignition Facility today is that, after more than half a billion dollars appropriated for NIF construction, and several billion dollars of cumulative investment in the ICF program, there are no viable ignition targets. Simply put, all the ignition target capsule designs for which LLNL has periodically claimed "ignition" in computer simulations have either subsequently (and quietly) been discarded due to crippling hydrodynamic instability problems, or cannot be fabricated with any known precision manufacturing technique, leaving a null set of "baseline" ignition targets for the NIF.
How Much Did "Beamlet" Prove?
11. UC/KR-99: "The Beamlet, a full-scale set of all the required optical components (apart from the final optics) for one of the NIF's 192 beamlines, was tested in 1994. It demonstrated the required laser performance and was built within the estimated cost. This success gave the Laboratory confidence that it could build the full NIF using its in-house capabilities. (p. 5).
RESPONSE: This statement is seriously inaccurate. Beamlet's optical components were not "full scale," and saying that Beamlet demonstrated all the required optical components "apart from the final optics" is a little like saying the Army has "demonstrated" the performance of a National Missile Defense System "apart from the homing kill vehicle." Some of the differences -- and more are coming to light each day -- are as follows:
- Beamlet operated at the intensity/output fluence levels projected for a NIF beamline for pulses from 1 ns to 10 ns. But Livermore has calculated that ignition can only be achieved at the NIF using a longer pulse: 20 ns in total with about a 3 ns high intensity component. In order to accommodate such a long pulse, the NIF spatial filters must stay open for about 20 ns. LLNL claims to have devised a means to keep NIF's spatial filters open for the required 20 ns, but this solution was never fully tested on the prototype Beamlet laser.
- Most of the Beamlet tests were at one micron without frequency conversion, and the third harmonic experiments were primarily with short pulse, not NIF-like shaped pulses.
- Beamlet's adaptive optics improved beam quality using one micron light, but we suspect SEAB will find the results less impressive after factoring in the inevitable deterioration associated with the process of frequency conversion.
- Beamlet used primarily a 35 cm aperture beam, while the NIF design uses a 40 cm beam, which translates to a 20% increase in the illuminated area, a significant difference. Blocking the edges of the laser could have led LLNL to overly optimistic predictions of laser beam performance, by failing to consider the effects of gain distortion near the edges.
- Beamlet used glass that was not produced by the less costly continuous melt process planned for the NIF components.
- Even with the more expensive glass, damage to Beamlet's optical components from frequency tripled light began at 5 joules/cm2 ("J/cm2"), not at the higher design fluence of 8 J/cm2 needed to accommodate ignition experiments. Since the safe operating point for the NIF is now probably closer to 4 J/cm2, the laser output will be limited to about 0.9 MJ, well below the requirement for any credible ignition target.
- With the Beamlet project, there existed the opportunity to do a full systems test of the modular NIF design concept. Unfortunately no Beamlet shot ever simultaneously tested the full set of NIF performance goals: a full-energy, full-pulse-length shot converted to the third harmonic and focused onto a target.
The Koonin panel's assurances that the Beamlet data underpins a high degree of confidence in NIF performance are misplaced, and should be discounted in whatever conclusions SEAB draws in this area. Given that the aforementioned serious difficulties were known to Project leaders prior to the dismantlement of Beamlet, why did LLNL (and DOE?) permit Beamlet to be shut down in FY 1998, dismantled, and sent to Sandia as a "backlighter" diagnostic for the Z facility, and who were the parties most responsible for this irrational decision?
For proponents, the NIF Project now faces a convenient Catch-22: without Beamlet, the only way left to resolve the major technical uncertainties in the performance of the NIF laser is to build it (or at least a portion of it). The years and years of science and engineering effort and billions of ICF program dollars expended have effectively been reduced to a huge open-ended gamble with public funds. It did not have to happen this way. There is absolutely no compelling need to rush deployment of the NIF in order to enable some critical aspect of Stockpile Stewardship, much less any energy or basic science application. In fact, LLNL's exaggerated claims regarding NIF's chances for ignition and its role in Stockpile Stewardship, followed by the abrupt disclosures of major problems and lying to senior officials in Washington, only reinforced Congressional concerns about the ability of the Stockpile Stewardship Program to ensure future stockpile reliability and safety, and helped persuade even moderate GOP senators that "now was not the time" to ratify a CTBT.
Are Management Deficiencies the "Root Cause"?
12. UC/KR-99: "As previously stated, management deficiencies, rather than technical problems, are the root cause of the cost and schedule overruns."
RESPONSE: No serious disinterested observer of the DOE's ICF program since the early 1980's could possibly make this claim. It amounts to an obvious -- even pathetic -- case of special pleading by some of the very persons who are complicit in the decision to press ahead with construction of the NIF without first obtaining the necessary (thoroughly peer-reviewed) scientific and technical consensus on the feasibility of the project. Persistent and potentially crippling scientific uncertainties have pervaded the project since its origins in the 1990 NAS Review, likewise chaired by the ever confident Professor Koonin.
These fundamental uncertainties extend even to the basic choice of the 1 - 2 megajoule energy regime for NIF, which may or may not be based on a correct scaling of the results from the still classified Halite-Centurion ignition capsule experiments of the 1980's. (These were "driven" by the X-radiation from underground tests, and constituted a proof in principle of DT capsule compression and heating to fusion conditions, via ablation of the capsule surface and the "hot spot" ignition concept.) Even many NIF proponents concede that the results from these tests suggested that the NIF planned operating regime is at best marginal for the attainment of ignition conditions, and other scientists with detailed knowledge of the Halite-Centurion experiments incline to the belief that NIF can't and won't work, and that much more energy is required to overcome the problem of hydrodynamic instabilities. Even if by some as yet unforeseen technological miracle -- and that is what ignition will require at this point -- LLNL manages to stumble upon the design for a DT capsule that can be fabricated and driven to a fusion energy yield marginally greater than the laser energy incident on the target assembly -- i.e., sufficient to claim "ignition" -- the mode of ignition is so different from the operation of nuclear weapons that LANL scientists continue to doubt its relevance to Stockpile Stewardship.
The problem of marginal energy for ignition has been compounded by NIF's specific developmental problems, such as the post-baseline addition of beam smoothing by spectral dispersion and random phase plates, which will likely result in at least a 7% reduction in beam energy delivered to the target. Further reductions will likely be required in order to sustain an economically tolerable rate of damage to the laser optics, thereby jeopardizing not only achievement of ignition but also the utility of NIF for other types of Stockpile Stewardship experiments.
The supposed utility of NIF's indirect-drive capabilities for weapons physics experiments has also been questioned, due to the production of high energy lines in the hohlraum x-ray spectrum above 1 keV that penetrate ahead of the Plankian "drive" portion of the radiation spectrum to preheat, perturb, or destroy the target. Perhaps the best known NOVA experiment, which changed understanding of the high pressure EOS of deuterium, was actually done in a direct drive configuration. To our knowledge, Livermore has never published a complete x-ray spectrum from a NOVA hohlraum, and yet much is made of the importance of NIF's predicted indirect drive capabilities for conducting Stockpile Stewardship experiments.
These technical issues deserve the most intensive independent peer review and analysis before the NIF project is restructured and allowed to proceed.
13. UC/KR-99: "Since the NIF is a large and complex project that requires new technology and manufacturing development on an unprecedented scale, development and implementation following accepted sound management principles is the most certain route to success. That various aspects of the installation are novel or require further development need not, and should not, compromise an orderly and cost-effective implementation of the work." (p. 6)
RESPONSE: This statement strains credibility. No amount of received managerial wisdom can be expected to overcome deficits caused by a weak scientific and technical foundation, or a schedule driven by a "rush to failure" deployment strategy. Certainly, there can be no "certain route to success" in such circumstances, as LLNL has amply proven on numerous occasions over the last several decades, and continues to demonstrate to his day.
14. UC/KR-99: "Quarterly reviews organized by DOE and other reviews initiated by the Laboratory and UC were ineffective in identifying problems in a timely manner. To the credit of the Project, the problems were eventually identified internally."
RESPONSE: The UC reviewers really are on bended knee here. This statement completely begs the question of why DOE's and other reviews -- including Koonin's own -- were so "ineffective in identifying problems in a timely manner." The only answer that presents itself is that these reviewers either conspired with LLNL to cover up the problems, or that LLNL officials mislead the reviewers regarding the true status of the project. In all likelihood it was both, since the major external reviews that we are aware of were neither thorough nor balanced, but indeed obviously tilted in favor of the project and stacked with cronies and consultants of LLNL and other contractors to the DOE laser fusion program.
"Generic Deficiencies," or Conscious Deception?
15. UC/KR-99: "We note that LLNL management deficiencies should have been identified through proper oversight by the DOE and the University. The National Research Council report, Improving Project Management in the Department of Energy, notes several generic deficiencies that, in the case of NIF, essentially guaranteed over-budget cost at the Project's inception. Three illustrative examples are:
"The NIF Project contingency was set at 15%. A contingency that was based on a risk assessment for the Project would have been set at a much higher level (~30%) for such a technologically complex endeavor.
"The baseline (i.e., a set of technical, scope, cost, and schedule parameters that describe the expected capabilities, cost, and duration of a project ) was set too early in the design process. It should have been set later when the technical definition and implementation plan were complete.
"Project activities in the TPC, but outside the TEC, suffered shortfalls in funding that negatively affected the Project." (p. 6)
RESPONSE: This statement verges on brazen hypocrisy. Professor Koonin chaired the NAS 1996-97 Review of the DOE ICF program that was tasked with "determining the scientific and technological readiness of the NIF project" to proceed to the construction phase. Two other members of the current UC Panel, Professor Roger Falcone and Dr. Hermann Grunder, likewise served on the earlier NAS panel. They had ample opportunity then to assess the risk of the NIF project, to recommend a delay to further refine the design baseline prior to construction, and to urge adoption of a larger contingency that more accurately reflected the project's real level of complexity and risk. Instead, here is what they concluded in 1997:
"Most of the start-up costs have been moved to the LLNL-projected operating budget for NIF and NOVA operations. In addition, there are [additional] NIF-related non-TPC costs [i.e., costs not included in the "Total Project Cost" estimate] of $397 million in LLNL-projected operating funds in FY98-FY02. The contingency has also been reduced from $144 million (20.6%) to $127 million (14.6%). In addition to the TPC, non-TPC funds of $120 million (FY98-FY02) are planned for R&D in optics and laser technology. LLNL plans to use these funds to perform R&D and to qualify vendors to the point that a 14.6% contingency becomes acceptable. (p. 32)
"These additional funds [outside the TPC] projected by LLNL will advance the technologies and qualify vendors to the extent that a 14.6% contingency is self consistent. (p. 33)
"With regard to NIF project management, the committee finds that there is a competent, well supported project team and a well-aligned project organization. The project has undergone a stringent Title I Review, and an Independent Cost Estimate review has verified the TEC cost estimates. Although the contingency currently provided is unusually tight for such a project, the availability of support from the ICF program during the construction period provides the margin to systems and components necessary for the NIF to the point that a low contingency might be acceptable.
"In sum, the committee believes that the NIF can be delivered to specifications within the stated TPC [Total Project Cost], as augmented by LLNL-projected operating funds, allowing the high-energy-density and ignition experimental programs to proceed; there are no identifiable "show stoppers." (p. 37)
Hiding NIF's Risk Premium
In other words, Professor Koonin and his NAS/UC panel colleagues first tacitly endorsed a strategy of concealing the NIF Project's real risk premium -- which they put at an additional $397 million over five years -- within Livermore's regular operating budget, but now that the NIF bandwagon is in the ditch, they suggest that the wiser course would have been to set a contingency of 30% based on a more realistic assessment of the project's risks! (In reality, an additional $180 million won't begin to get the NIF project "back on track," if by that phrase one means ensuring that the project will achieve its original design objectives at anything close to its previously estimated cost.)
Needless to say, it was just such a realistic assessment of the NIF project's real risks that Professor Koonin, et al., assiduously resisted, to the point that NRDC resorted to litigation, resulting in a federal court injunction against dissemination or reliance by DOE on the 1997 NAS report. For the record, here is how NRDC summarized the project's risks in an affidavit for the Court in April 1997:
"United States national security would be enhanced if the construction of the proposed NIF were deferred until, at a minimum, the DOE completes further inertial confinement fusion research and development showing that the NIF project has achieved:
(a) a degree of control over the delivery of the laser beam energy to the target, as shown in a series of 10-beam beam smoothing experiments underway using the existing NOVA laser, that is sufficient to demonstrate, via peer reviewed calculations, the radiation temperature and time-dependent uniformity of radiation illuminating the capsule that is needed to ignite the NIF baseline indirect-drive target design;
(b) a degree of understanding of the capsule implosion physics sufficient to demonstrate, via planned experiments on the University of Rochester's OMEGA laser and computer simulations, that the chosen baseline cryogenic target capsule design has a high probability of being driven to ignition by the realistically achievable drive parameters as demonstrated in (a);
(c) a high degree of confidence that the damage threshold for the laser optics actually achievable in manufacture at the budgeted cost will support normal operation of the NIF laser at the power levels required to achieve (a) and (b); [emphasis in original]
(d) proven capabilities for fabricating, emplacing, cooling, and characterizing cryogenic ignition targets of the type required for the NIF baseline target design; and
(e) diagnostic capabilities for the NIF beamline and target chamber with a degree of resolution and accuracy needed for: pointing, timing, and pulse shaping of the laser energy required for ignition; characterizing the time dependent radiation flux within the hohlraum volume; imaging the capsule implosion with better spatial and temporal resolution; and post-ignition characterization of capsule radiation outputs.
None of the above capabilities has been demonstrated, and some of the capabilities have not even been designed. It is not clear that all of these capabilities above can ever be achieved . . ."
This was an accurate depiction of the situation in April 1997, and it remains so today, only more so, since ignition via indirect drive looks less, not more, achievable than it did 2.5 years ago, and the NIF laser looks less, not more, likely to supply the drive conditions needed for ignition experiments.
16. UC/KR-99: "While the goal of doing things "smarter, faster, cheaper" is admirable, it can lead to denial and delay in correcting problems if shortcuts are taken that weaken the necessary project management discipline."(p. 6)
RESPONSE: We have never encountered a more euphemistic and attenuated description for the acts of withholding information from legally cognizant authorities, and lying.
17. UC/KR-99: "Specifically, there were shortfalls in funding R&D, which cascaded into design delays with significant cost implications." (p. 7)
RESPONSE: What shortfalls? The NIF project and supporting ICF research program budget requests have been fully supported by the Clinton Administration and the Congress. In FY 1999, DOE spent $503.4 million on the ICF program, of which the NIF Project and the LLNL Laser Directorate received $291 million and $102 million, respectively, for a total of $393 million flowing to Livermore for NIF. Through FY 1999, the nation has spent $660 million on NIF design and construction, and another $200 million or so per year on the ICF base "program," which for at least the last five years has been largely directed to supporting the NIF indirect drive program. So the nation has already invested on the order of $1.6 billion in the NIF project. Where is the "shortfall" in funding? Only those reared in the gold-plated culture of the weapons laboratories could possibly regard this rich funding environment as constituting a "shortfall."
A Lack of "Effective System Integration?"
18. UC/KR-99: "The Project suffered from a lack of effective system integration. An important example was the insufficient appreciation of the difficulty of integrating all the lasers and their infrastructures to fit within a tightly constrained building while preserving maintenance accessibility over the long-term operation of the facility. As a result, the building was scaled back to reduce its cost without understanding the impact on the rest of the Project." (p. 7)
RESPONSE: Again, this explanation by itself is hard to accept, as careful control and documentation of the design evolution -- and "deconflicting" of design changes over the life of a project -- is the basic threshold management problem for any large-scale project. Professor Koonin's 1997 NAS panel report failed to identify Livermore's capabilities for successful project integration as an issue for concern. On the contrary, it noted that "the work breakdown structures and project organization were in place and aligned to provide clear technical interfaces" and that "interface control, and a Baseline Change Control Board are in place to maintain the scope, cost, and time line." It really is astonishing how this reportedly "well aligned" organization fell apart within two years. Perhaps the real explanation is that it was never there to begin with.
Wanted: "A Team of Tough Outside Reviewers?"
19. UC/KR-99: "None of the relevant management entities, the DOE, the University, nor the Laboratory, received a critical, independent review of the Project after its initiation. Instead, the NIF Council functioned as more of an advisory group rather than as a team of tough outside reviewers." (p. 8)
RESPONSE: More crocodile tears from Professor Koonin, who had his chance to perform just such a "critical independent review" with an NAS Committee in 1996-97, and blew it, violating the Federal Advisory Committee Act in the process. Nor did he or any of his NAS/UC committee colleagues audibly call for closer scrutiny of the project after the construction started in June 1997. On the contrary, the current UC Review does nothing to alter Professor Koonin's reputation as a captive apologist for the project. The UC President's Panel was neither "independent" nor "balanced." Given that UC is the Management and Operating (M&O) contractor ostensibly responsible for "managing" LLNL, and hence "mismanaging" the NIF Project, the ability of the Koonin's UC Panel to perform an objective and unbiased review was severely compromised from the outset. In addition to their prior association with both the biased 1997 NAS Panel, Koonin and Falcone have both been paid consultants to LLNL, and Grunder has been a member of the LLNL Laser Directorate's "NIF Council." Panel member William R. Frazer is the interim chair of the UC President's Council on the National Laboratories and a former Vice President of UC tasked with overseeing the UC's lab management contract.
Other panel members included: William Friend, a retired senior executive and consultant for Bechtel, Inc., which operates the Nevada Test Site for DOE; Robert McCrory, director of the University of Rochester's Laboratory for Laser Energetics, a NIF program contractor, and Prof. Raymond Jeanloz, professor of geophysics at UC Berkeley, a frequent participant in the deliberations of the MITRE Corp.'s JASON Division study groups, which have repeatedly endorsed NIF. Jeanloz is also a Senior Technical Adviser to the DOE's Stockpile Stewardship Program and participant in a recent SSP "30-Day Review" trumpeting the technical accomplishments of the NIF project, including the misleading claims that "pilot optics are on schedule and meeting their goals" (p.43) and that "advances in ICF target physics are reducing the technical risk in achieving ignition on NIF" (p.38) We are not acquainted with the other three members of the UC President's Panel, but two (Pickar and Sanders) are lower ranking colleagues of Koonin's at Cal Tech, and William Simmons is a retired executive from TRW, a major national security systems contractor with many ties to LLNL.
We find it noteworthy that, while the Secretary of Energy is withholding $2 million of UC's "at risk" contractor performance fee, and may withhold more as a result of the NIF Project's cost overruns and LLNL misconduct, the DOE Defense Programs' Website has seen fit to post UC President's Council Report on the NIF without explanation or disclaimer of any kind.
A "Natural Inclination" to Mislead?
20. UC/KR-99: "The Project had a history of solving problems without raising them externally. Among these were the technology breakthroughs described above; the recovery in cost and schedule for the conventional construction following early setbacks; and the utilization of noncommercial entities for some of the optical components (e.g., polarizers and deformable mirrors). Because of these successes, there was a natural inclination to `tough it out' and deal with the matter internally." (p. 8)
RESPONSE: More inexcusable excuses for withholding information from, and lying to, responsible authorities, and more evidence that Professor Koonin takes his role as senior project apologist very seriously indeed.
Inadequate Tracking Tools?
21. UC/KR-99: "Because of shortfalls in R&D funding and inadequate tracking tools, the Project had been forced into a situation of "informal austerity," in which there was little receptivity to raising problems that would require additional funds." (p. 8)
RESPONSE: Inadequate tracking tools? According to Koonin's 1997 NAS report, "this tracking system is quite detailed, with the project broken into procurement packages. There are cost accounting plans with schedule milestones that allow close monitoring of progress based on completed activities. The relatively large number of milestones to be measured allows an objective assessment of progress." (p. 31)
Did Prior "Success" Lead to Senior Management "Complacency?"
22. UC/KR-99: "There was a lack of proactive intrusion into the management of the Project. While the Laboratory has a long history of rigorously challenging scientific program leaders, such review was not applied to the management of the NIF Project. It is possible that the Directorate's previous record of scientific and project achievement and success led the rest of the Laboratory, the University, and the DOE toward a complacent attitude." (p. 9)
RESPONSE: More convoluted apologetics for egregious misconduct by a major publicly funded scientific laboratory. As noted above, the Laser Directorate's recent record apart from NIF is hardly one of unblemished "achievement and success," given the recent $2 billion failure of the U-AVLIS project. Even the much touted "success" of NOVA was more apparent than real, given that NOVA was originally planned to be a 20-beam, 300kJ laser that (former director) John Nuckolls calculated could reach ignition. When a review chaired by John Foster showed that there was no way Nova could reach ignition, a deal was worked out that provided the originally requested funding for the project ($186 million) but scaled it back to a "science-oriented" facility with only 10 beams. A highly speculative physics analysis was then thrown together (by J. Lindl) to show how ignition issues would easily scale from NOVA to the next facility (NIF) in a way that answered all fundamental questions about achieving ignition with glass lasers.
23. UC/KR-99: "The culture of the Project was such that the airing of problems to the external world was strongly discouraged. The Project, Directorate, and Laboratory leadership did not consider it valuable or appropriate to involve either DOE or the University in the early definition and discussion of the problem. Until the problem and its implications were better understood and options for dealing with it explored, there was a reluctance to simply bring forward bad news." (p. 9)
RESPONSE: This account merely begs all major questions, such as WHO was responsible for establishing a "project culture" that "strongly discouraged" the "airing of problems to the external world"? What is the "external world" in this case: senior lab management? the university? DOE? Congress? scientific colleagues?
Were formal or informal policies in place that abridged the academic rights and responsibilities of UC employees, DOE regulations, or civil protections under Federal law?
Did NIF project or senior lab managers establish a work environment in which employees felt that their jobs and potential for advancement could be threatened if they told the truth about the project's actual status "to the external world"?
Who was responsible for lying to the Secretary of Energy and other senior officials last June regarding the status of the project, thereby inducing a Cabinet official to provide false and misleading testimony to Congress? The UC/KR-99 Report answers none of these questions.
Time to "Move Forward?"
24. UC/KR-99: "As detailed above, most of the NIF problems are not technical. The Project faces serious issues because the DOE and the Laboratory violated some basic principles of sound project management." (p. 10, emphasis added)
RESPONSE: As detailed above, this statement is the rhetorical equivalent of trying to nail Jello to the wall. But note that it contains a subtle concession from the earlier (silly) assertion that "management deficiencies, rather than technical problems, are the root cause of the cost and schedule overruns." Now we are told that "most" of the NIF problems are not technical - implying that at least some surely are. This is a huge difference. Even one unsolved "technical problem" can wreck a project of this nature, particularly if the inability to solve the problem in question has a cascading effect on other sensitive parameters of device performance; and these effects can no longer be offset or remedied at tolerable cost, because large portions of the facility are already in place. With systems integration in the NIF Project as half-baked as it appears to be, the occurrence of this kind of expensive and essentially irremediable defect seems increasingly likely if the project stays on its current course and schedule.
"A New Champion for the NIF?"
25. UC/KR-99: "The Laboratory should appoint an Associate Director (or equivalent) for the NIF. This must be a person who can devote the necessary time to serve as a champion for NIF as a defined Project in its own right . . . The Associate Director (or equivalent) must be the executive-level sponsor of the NIF Project who can, and will, run interference and interact with all the necessary internal and external constituencies to once again get NIF back into a healthy situation and environment." (p. 10)
RESPONSE: Following the recent rejection of the CTBT by the United States Senate -- partly on the strength of Republican doubts about the efficacy of the science-based "virtual testing" strategy epitomized by NIF -- the last thing DOE or the national security needs is another uncritical cheerleader for the NIF Project.
26. UC/KR-99: "The Laboratory Director must establish a critical, intrusive, independent review of the Project. . . . The review report should be disseminated to the NIF Associate Director, the Laboratory Director, and to the University Office of the President." (p. 10)
RESPONSE: Interestingly, UC/KR-99 omits DOE, the Congress, the National Academy, the media, and public interest organizations from the list of potential recipients of a new independent review! Is this merely an oversight, or will the lockout of the "external world" essentially continue, only this time with the active (rather than supposedly passive) connivance of the University of California.
27. UC/KR-99: "The very importance of NIF, an integral part of the Stockpile Stewardship Program, places its own constraints and pressures on the Project . . . The Committee believes that the NIF Project MUST be given time to develop a valid and realistic baseline and implementation plan. With these in hand, the Project budget and control schedule would then be properly reviewed, adjusted, and ultimately approved by the DOE and appropriate funding authorities. These steps, adequate financial support, and the implementation of our recommendations would give the Committee confidence that the Laboratory can successfully complete and operate the NIF." (p. 12)
RESPONSE: Many persons with long experience in how stockpile weapons work and how they fail do not believe that NIF is critically important to maintaining the safety and reliability of the U.S. nuclear stockpile. Nor is deployment of the NIF on any particular schedule absolutely vital to the retention and education of stockpile stewards. NRDC agrees, however, that any decision to proceed with NIF requires more time, but not for the reasons stated. Once again, Professor Koonin is assuming facts not in evidence. The first-order question before DOE and the nation is not WHEN, but WHETHER, and then IN WHAT FORM the NIF Project should be continued. Answering this question will require a no-holds barred, in-depth review of all the long festering technical and scientific issues surrounding the NIF.
Absent a thorough and complete scientific and technical review, it is simply not possible to arrive at fully informed conclusions regarding the tradeoffs involved in terminating, restructuring, or continuing the NIF Project as it is currently planned. Obviously, if such a review determines that the NIF "baseline" strategy for achieving ignition via cryogenic indirect-drive targets is no longer -- or indeed never was -- sufficiently viable to warrant the expenditure of billions of taxpayer dollars on an "ignition facility," this conclusion could throw into question the whole rationale for proceeding with the NIF project
A Dozen Questions to Guide a Review of the NIF
Among the many questions to be weighed in the SEAB's review, or a more detailed and extensive review that hopefully might follow, are the following:
- What are the cumulative impacts of the currently achieved levels for beam smoothing, amplifier glass impurities, final optics radiation damage, spatial filtering, harmonic conversion, insufficient focusing, etc. on the NIF laser's actual, "as-built" capacity for driving a target to ignition?
- What is the actual state of beam pointing and target assembly (hohlraum and target) diagnostic capability today for the NIF long-pulse environment, which of these postulated capabilities are essentially "vapor ware," and which can reasonably be relied upon to be available by 2005 to diagnose and control both ignition and non-ignition hohlraum experiments?
- In light of a sensitivity analysis covering the "most likely to be achieved" rather than the postulated "ideal" range of NIF capsule drive conditions, which baseline ignition target designs still clearly fall within a region where one may confidently state that sufficient energy in the appropriate pulse shape can be delivered to the target? Which candidate target designs are now revealed to be marginal or unworkable within the range of beam energies, pulse shapes, and diagnostic "tuning" capabilities that are now considered "demonstrated" or highly likely for the NIF?
- Which, if any, of the class of "plausibly ignitable" targets can be fabricated now, and which can confidently be projected as susceptible to fabrication within the next five years? -- ten years? What is a realistic assessment of the technical prospects and costs for fabricating indirect drive targets that can be driven to ignition in light of the most likely set of drive parameters achievable at tolerable cost on the NIF?
- In the light of the actual state of affairs as outline in 1-4 above, what are the merits of the technical case offered by those experts who now conclude that NIF has poor to marginal prospects for achieving ignition?
- What is the technical case presented by those experts who have always believed that NIF, even in its ideal baseline design configuration, can never drive a target to ignition because it lacks the energy needed to forestall evolution of hydrodynamic instabilities that disrupt the symmetry of the capsule implosion?
- In the context of DOE's revised strategy for setting priorities among Stockpile Stewardship missions -- to ensure that the stockpile refurbishment and remanufacturing missions are given highest priority -- what is the relative importance of NIF compared to other SSP facilities that are more directly tied to these highest priority missions?
- If the achievement of ignition on any particular timetable is relatively unimportant to the SSP, and/or there is low confidence that ignition can be achieved with the present NIF design, what alternative configuration of NIF is minimally required to carry out planned non-ignition stewardship experiments deemed important to the program?
- How important are these non-ignition experiments relative to other funding priorities within the overall stewardship program, and could less expensive facilities -- including a scaled down NIF -- be used to carry out (all? some?) of these experiments?
- Are there important but as yet undisclosed limitations on the ability of indirect drive hohlraums to supply the type of drive and diagnostic conditions needed for non-ignition SBSS experiments, e.g. to determine opacities and EOS?
- What is the balance of benefits and costs involved in restructuring the NIF Project along the following lines:
(a) stop and defer all work on the current construction effort not connected with procurement, assembly and deployment of the first bundle of 8 beams;
(b) conduct prolonged testing and experimentation with this bundle to validate all currently proposed and any additional changes in the baseline design needed to restore NIF's operating parameters, or establish new ones;
(c) concurrently, conduct a painstaking bottom-up review, involving the broadest possible cross section of informed scientific and technical opinion, on the scientific and technical requirements for achieving fusion via inertial confinement, and seek expert consensus on the most appropriate strategy for continued exploration of ICF; and
(d) restructure, redirect, or terminate the NIF project, based on the results of this comprehensive review.
What are the costs and benefits of terminating the NIF Construction project and redirecting a portion of the funds saved into an enhanced exploratory research and computations effort, and Omega experimental campaign, geared to improving our understanding of the fundamental physical phenomena involved in ICF capsule implosion in both direct and indirect drive configurations, and involving evaluation of a wide range of plausible "drivers".
ERRATA SHEET FOR NRDC MEMORANDUM RE:
UC PRESIDENT'S ADVISORY COUNCIL REVIEW OF NIF
In Item #4, page 3, the statement is made that the Beamlet never operated at a pulse length greater than 10 ns. This is based on dated information from early 1997 and is incorrect. Beamlet was operated with 20 ns pulses of two types: shaped with a high intensity component, and unshaped at lower intensity. The shaped pulses did reach a fluence in joules/cm2 on the best shot that is appropriate to a 1.8 MJ NIF, but this was demonstrated without simultaneously demonstrating the other necessary laser conditions, including wider aperture, frequency conversion, focusing, and measurements of time dependent imaging from the pinhole.
- In Item number #7, page 4 the statement is made the Beamlet never met the peak fluence requirement even with the "good" [i.e. more expensive] glass. Again, Beamlet did meet those fluence requirements on one shot using one micron light. More accurately, it never met the requirements using one-third micron light and without incurring optical damage at levels significantly below the NIF design point of 8 J/cm2.
The above technical errors do not alter the conclusions of the memorandum regarding the highly uncertain scientific and technical status of the NIF Project.
Sign up for NRDC's online newsletter
- Latin America Green News: green tax in Chile, blue carbon in Costa Rica, transparent processes requested for Cabo Dorado in Mexico
- posted by Den√©e Reaves, 4/14/14
- Jonathan Schell, Climate Change, and the (Still Pending) Fate of the Earth
- posted by Mark Izeman, 4/4/14
- Energy Efficiency Progress by America's Utilities: An Update
- posted by Ralph Cavanagh, 3/25/14
Nuclear Document Bank
- NRDC’s Perspectives on the Economics of Small Modular Reactors
- NRDC's Amicus Response in Support of Friends of the Earth
- An Evaluation of the NRC Response to the Fukushima Nuclear Accident
Related NRDC Press Releases
- NRDC Senior Nuclear Scientist Unveils Radiation Dosage Data at Senate Hearing
- Nuclear Security Summit Fosters an Illusion of Achievement, Says NRDC
- NRDC Reacts to the Obama Administration's Nuclear Posture Review
NRDC Gets Top Ratings from the Charity Watchdogs
- Charity Navigator awards NRDC its 4-star top rating.
- Worth magazine named NRDC one of America's 100 best charities.
- NRDC meets the highest standards of the Wise Giving Alliance of the Better Business Bureau.