When scientists at Lawrence Livermore National Laboratory successfully demonstrated the Beamlet in 1994, they were confident they could duplicate the single laser, multiplied 192 times, to produce the world's largest laser. But the Beamlet lacked the complexity to be a true prototype for the National Ignition Facility, a laser designed to test nuclear explosions, theorize on the inner workings of stars and be a model for cheap fusion energy. Officials now concede the facility is likely to cost $350 million more than its estimated $1.2 billion price tag, if built as planned, and could be 18 months behind its original completion date of 2003. Blame for the project's shortcomings belongs to management, a University of California committee reported in November. The university operates the project, with the bulk of the lab's funding and direction coming from the US Department of Energy (DoE). The committee said managers set the project contingency at 15 percent, too low for a technically complex endeavor. The baseline for the project -- the set of technical, scope and cost parameters that define it -- was set too early in the design process to be realistic. A new baseline is due in June. Scientists at the National Ignition Facility discovered that aerosols are formed with each flashlamp pulse of the laser. Thermal currents carry the clusters around in the laser amplifier environment and diminish the laser's effectiveness. The committee also stated that the laboratory failed to understand the complexity of the project, especially its need for stringent cleanliness protocols during assembly. About 4500 clean connections are needed to prevent or reduce flashlamp-induced damage to the laser slabs, antireflective coatings and optics. There is no quick fix for such a flawed project, said Steven Koonin, the California Institute of Technology provost who headed the committee. The DoE will deny the university at least $2 million of its $5.6 million contract for substandard performance. Secretary of Energy Bill Richardson ordered that major laser assembly work be conducted by an independent contractor. A DoE task force, scheduled to meet twice in November and once in December, was due to report to Richardson in January. Robin Staffin, Richardson's science adviser, said scrubbing the project isn't likely because it is important to the DoE's Stockpile Stewardship Program, which recognizes the US ban on underground nuclear testing while keeping the nation's nuclear capability in a state of readiness. Project opponent Michael McKinzie, a senior scientist with the National Resources Defense Council, speculated that the government would support a scaled-down version of the facility with perhaps as few as 48 or 96 beams. One cluster of beams, run at half energy, is sufficient to test the nation's nuclear stockpile, he said.