On June 1, the US Department of Energy sent to Congress an "interim certification for the revised cost and schedule baseline" for the National Ignition Facility (NIF), including "a preliminary estimate of the total project cost and funding profile." The plan is based on having "the full National Ignition Facility capability consisting of 192 beams" completed in fiscal year 2008, with some beams operational in 2004. The cost and schedule for NIF has been under review for about nine months (FPR, September/October 1999, p. 57), since it became known that final assembly of the unprecedentedly large laser would be much more complex that anticipated.

The Lawrence Livermore National Laboratory (LLNL), where the laser is being constructed, initially estimated the anticipated cost increase to about 30% higher than the original estimate of $1.2 billion. The report sent to Congress, however, estimates the cost increase to be about 85% higher than the planned number.

In its report, the DOE acknowledges that "the most expeditious schedule to complete the National Ignition Facility has shown that it could be completed as early as the end of fiscal year 2006," and that the slower schedule, which was mandated by DOE, would result in a higher total cost. The Livermore laboratory had completed a detailed cost estimate for the faster schedule of $1.95 billion, compared to the original estimate of $1.2 billion. DOE's construction cost estimate for the slower schedule is $2.25 billion. DOE chose the slower, higher cost plan in order to cap the annual costs over the next several years. Energy Secretary Richardson reportedly wanted to limit expenditures to increments of no more than $100 million per year, but reluctantly agreed to provide $150 million per year when told that the project could not be properly managed at the lower figure. DOE indicated they planned to provide an additional $95 million in FY 2001; "no more than $150 million in FY 2002 and 2003; $140 million in FY 2004; $130 million in FY 2005; and a declining profile thereafter." DOE's insistence on a cost-profile-based schedule (FPN00-24) resulted in the need for another new rebaselining of the project. DOE promises to provide a firm estimate by mid September. DOE had originally been told by Congress to submit a definitive rebaseline by June 1.

DOE also estimated that "Other related (non-construction) costs" associated with NIF would rise from $0.8 billion to $1.4 billion.

Despite the cost increases, Secretary Richardson asserted that a detailed review by several groups, including the Directors of the three DOE weapons laboratories, has concluded that "The National Ignition Facility remains a cornerstone of the Stockpile Stewardship Program." He said that "the project is technically sound and based on good engineering design." He cited reviews by the Secretary of Energy Advisory Board Task Force on the National Ignition Facility, the University of California President's Council National Ignition Facility Review Committee, the National Ignition Facility Programs Review Committee, the Director of the Lawrence Livermore National Laboratory, and the Department of Energy's Energy Systems Acquisition Advisory Board. Richardson said that "Project management has been restructured and has demonstrated over the last six months that it is capable of managing a project of this scope. In particular, it has engaged and effectively utilized industrial experts with relevant experience to both review and participate in the project."

In a May 31 memo to the Energy Secretary, Acting Deputy Administrator for Defense Programs, Brigadier General Thomas Gioconda told the Secretary that NIF project management within the Department of Energy had also been restructured to support the NIF rebaselining activities. These included setting up a NIF Project Office reporting to the Deputy Administrator. This Office is being headed by James Anderson, who previously was a senior staff member and director of the Tritium Systems Test Facility at the Los Alamos National Laboratory. The Office Director will be located in DOE's Washington Headquarters, but his Deputy will located on site in Livermore. Gioconda noted that, at LLNL, the new NIF Project Manager, Edward Moses, now reports directly to the Associate Director for NIF Programs, also a newly established function, and that that Associate Director, George Miller, reports directly to LLNL Director Bruce Tartar. Gioconda also noted that the Directors of all three DOE weapon laboratories have recently completed a NIF "White Paper" concluding that "NIF is one of a set of essential capabilities that are needed to address the significant technical challenges associated with developing a science-based understanding of the nuclear stockpile."

In the package of materials sent to Congress, DOE also released a copy of a May 30, 2000 letter from LLNL Director Tartar to Secretary Richardson in which Tartar states, "I am confident that we can bring the NIF project to a successful completion within the new cost and schedule guidelines provided to us by the Secretary." Tartar notes that DOE had previously (last fall) asked LLNL to rebaseline the project for fastest completion and lowest total cost. "In accordance with the DOE guidance last fall, we developed a detailed plan for NIF for the most aggressive path to completion. This plan is one that is project optimized and is based on many months of effort working with NIF vendors and industrial experts to determine a reliable forecast of cost and schedule by doing a "bottom-up" sum of more than 25,000 individual items," Tartar said. Tartar said that the total project cost under the "project-optimized plan was estimated to be $1.95 billion" and that first beams would be operational at the end of FY2003 and the full 192 beams by the end of FY2006. Tartar noted that these dates were "both two years later than originally planned." He said that implementation of that plan "would have required funding increased over the original plan of $150 million in FY2001 and $240 million in FY2002." He noted that DOE had later "requested development of several stretched options that would reduce the early peak funding requirements," citing the potential for "a negative impact on other elements of the Stockpile Stewardship Program" from the project-optimized plan. He said that the cost and schedule data for the stretched plans "were derived from the database of the project- optimized plan by stretching certain efforts and costs so that peak funding rates would be lower." He noted that the NIF Program Review Committee (NPRC), chaired by Hermann Grunder, Director of the Jefferson National Accelerator Laboratory, "was not satisfied with our method of estimating the stretched options by manipulating the estimates derived for the project-optimized plan." He quotes the committee as saying, "In the judgment of the NPRC, this method is an entirely inadequate approach to arriving at sound estimates for such a megaproject as NIF." Tartar says the committee recommended "a detailed analysis to determine the new, self-consistent baseline cost and complete funding profile required." Tartar, however, states that, in his view, the estimate of $2.25 billion for the DOE-selected option is "an upper bound on the total project cost." Tartar says that LLNL will complete the new baseline estimate by August 1.

Tartar's letter also addresses the question of whether the new schedule for NIF was consistent with the needs of the Stockpile Stewardship (SSP) and whether a facility with fewer beams could meet the Stewardship requirements. He said that three external reviews had addressed these questions and concluded that "NIF remains extremely important to SSP." He said that the NPRC committee had considered the benefits to SSP of 48, 96 and 192 beams on NIF and concluded that "while some good weapon physics can be done with fewer than 192 beamlines, NIF should be completed to its full 192 configuration, with ignition as a primary objective."

Tartar says, "A full capability, 192-beam NIF remains at present as the only facility that can provide the extremely high temperatures, densities and fusion ignition conditions unique and critical to our effort to understand and interpret many of the processes that occur in exploding nuclear weapons. NIF also is extremely important in attracting, training and judging the quality of the weapon scientists who will have to make future decisions about the certifiability of weapons in an aging stockpile." He says, "Data from the full 192-beam NIF will be necessary to enable our weapon scientists to make the certification decisions of the future. The data from NIF, including those from experiments in energy-density regimes achievable only with ignition, will also be necessary to test the validity of the computer models being developed in the Accelerated Strategic Computing Initiative (ASCI), another essential element of the SSP."

Tartar said that reviews were also undertaken to reconsider the question, "Is there a reasonable likelihood that continued and increased funding for NIF will lead to a technically successful conclusion?" He said that all external reviews had concluded that "If completed as originally conceived, NIF is likely to produce ignition and provide the necessary data to the SSP so that sound certification decisions can continue to be made."

Tartar cited several examples of "major success" in the NIF project to date, but acknowledged that "One technical issue that has not yet been completely resolved is the damage produced in the last few optical components by ultraviolet light." He said that "with present technology, some damage would be observed in the last few optical elements in each beamline if experiments were run at the highest intended operating fluences. Eventually the damage would be severe enough that the damaged optics must be removed, refurbished and reinserted." He said that LLNL would plan to address this problem by "choosing appropriate operating limits during initial experiments" and then, through continued R&D, "to increase these operating limits as results from our technology development program become available and as we gain operating experience on NIF itself." He said that "many non-ignition weapon physics experiments can be done within the initial operating limits."

Tartar addressed issues associated with industrial involvement in NIF. He said, One of the major criticisms of both the internal and external reviews of NIF over the last year has been that the Project was not effectively taking advantage of industrial expertise in areas in which the laboratories do not have relevant experience. The Project received much praise for its effective partnerships with industry in constructing the conventional facilities and in jointly developing and installing state-of the art production-line techniques for manufacturing the thousands of large-aperture optical components needed. However, it was found that the size and scope of the effort needed to assemble, install and integrate the beampath infrastructure for 192 laser beams within the NIF building and with the necessary cleanliness was greatly underestimated." Tartar acknowledged that, "Perhaps because the Laboratory had successfully assembled, installed and integrated the laser beamlines for six previous state-of-the-art laser systems, we fell into the trap of assuming we had the expertise to do the same for NIF." He said, "In each of those previous systems, we had to meet the stringent cleanliness requirements inside each beamline that we do for NIF. However, those previous systems had at most twenty beamlines and they were assembled inside buildings that were clean rooms of class 10,000 or cleaner. NIF's building is so large that it was not cost-effective to make it a clean room of a classification similar to our previous laser facilities. Therefore, the large difference in cleanliness between that required inside the beamline and what exists in the laser bay contributed to judgment error. We should have but did not recognize that the size and complexity of assembling a clean infrastructure for 192 beamlines in a tight space bears more resemblance to building a fabrication plant for producing semiconductor chips than to building our previous laser research facilities." Tartar said that, "Acknowledging our underappreciation of this point, we immediately took several actions. We consulted with experts in the semiconductor and aerospace industries (both have much experience assembling large clean structures). Going beyond consulting, we also contracted several experts and companies to help us develop better strategies to take advantage of industrial expertise."

Tartar said that, "as a result, we are pursuing two major acquisition strategies that are very different from our original plan: (1) We will procure components from industry at a higher level of integration than previously planned. For example, the 48 Pre-Amplifier Modules (PAMs) required (each containing 2736 optical components) will be produced as integrated units rather than buying the components and relying on in-house assembly and integration.(2) With DOE's recent approval, we are awarding a single Integration Management and Installation (IMI) contract to put the mechanical and electrical beampath infrastructure in place for all 192 beams. Patterned on our successful industrial partnership for construction of conventional facilities, this will provide us with the industrial expertise to do the job we previously assumed would be done with in-house labor. The IMI will likely be NIF's largest single procurement."

Tartar concluded, "This has been a difficult and stressful year for all of us in the NIF Project at LLNL, at UC and at the DOE. It is never pleasant to admit deficiencies and shortcomings, but to be ultimately successful we must face unpleasant facts objectively, learn from them and do the best job we can to recover from them. I believe we have done just that." The report to Congress can be viewed at http://www.dp.doe.gov/dp_web/news_f.htm