Big science. Big bucks. Big questions.

In the eight years since the United States stopped detonating thermonuclear warheads beneath the scrub and sand of the Nevada desert, federal scientists have been developing less earthshaking ways to ensure that the country's nuclear arsenal remains reliable and safe.

That's what the enormous laser being built at the Lawrence Livermore National Laboratory in California was designed for. The idea: Mimic on a tiny scale the conditions in exploding nuclear weapons by shooting 192 light beams at a tiny cylinder of hydrogen. Then gather the information and use it in computer models, which help predict the behavior of aging weapons in the stockpile.

But the flagship facility has rapidly become the Big Dig of Big Science.

Since its kickoff in 1993, costs for the National Ignition Facility (NIF) have nearly doubled to between $3.5 billion and $3.9 billion, depending on whose figures are used, and project managers at the lab have covered up cost overruns. It is running at least six years behind schedule. And critics, who argue that the facility could also be used to design new weapons, doubt whether the laser ever will work as advertised.

Now, the US Department of Energy is asking Congress to tack an extra $95 million to NIF's budget for the new fiscal year. And the agency wants to shift $40 million from another of its accounts.

As the price tag for the International Space Station nears $100 billion, the NIF request comes at a time of renewed scrutiny for some of America's most ambitious science projects. The Energy Department says it remains generally positive about the project's prospects for success, but patience on Capitol Hill is ebbing as mistakes and cost overruns mount.

If NIF fails to get the extra cash, the project "will be seriously undermined," noted US Energy Secretary Bill Richardson in a letter to Congress that accompanied a revised budget and review.

When it went wrong

Support for the project among lawmakers began to falter last year. First, project chief Michael Campbell admitted that he had never received a doctorate from Princeton University, as he had claimed he had.

Shortly after Mr. Campbell resigned, Mr. Richardson disclosed that officials at Lawrence Livermore had covered up grave technical and managerial problems.

Agency officials reviewed and reorganized the project.

Then, last June, department officials acknowledged that the cost of the laser had grown from $2 billion to $3.3 billion and that it would be completed in 2008, not 2002 as originally planned.

In August, the US General Accounting Office (GAO) released its own assessment of the project, claiming that costs were closer to $3.9 billion. The GAO report also cited significant technical uncertainties that, while not fatal to the project, could further boost costs and stretch the completion schedule.

To Rep. Floyd Spence (R) of South Carolina, chairman of the House Armed Services Committee, the GAO "raised legitimate questions about the ability of [Department of Energy's] science-based approach to stockpile stewardship to adequately replace nuclear testing" as a way to ensure the stockpile remains effective.

Is the worst behind?

For their part, senior Energy officials contend that the worst may be over.

The officials, who asked not to be quoted directly, suggest that all the technical hurdles that the GAO mentioned have now been resolved. They anticipate that the first few beams will be fired up in 2004, the first stockpile experiments will begin in 2006, and the full complement of lasers will be ready in 2008.

But program critics aren't as sanguine about the laser's operational prospects.

In the past, the project has gone ahead - even pouring concrete to build the structure to house the laser - despite the fact that technologies crucial to the experiments hadn't been fully developed yet, says Christopher Paine, who tracks the US nuclear-weapons program at the Natural Resources Defense Council in Washington.

One question that has not been resolved, for example, centers on the glass needed to perform the tests. "There are a host of issues on whether the laser can achieve [the appropriate energy level] without exploding the glass," says Mr. Paine.

Indeed, the most recent review panel underscored the importance of spending an additional $30 million on optics research and development to make sure that the right kind of lenses and other components are available in 2008 and 2009.

The laser's importance

Others doubt the need for NIF at all to safeguard the stockpile.

NIF "is not that critical to establish the safety and reliability of the stockpile," says Charles Ferguson, a physicist at the Federation of American Scientists, echoing a sentiment expressed even by some weapons designers at other DOE weapons labs.

They worry that efforts to cover NIF's rising costs will come at the expense of other stockpile-stewardship projects.

Some members of Congress reportedly would like to kill the project. But Paine and Stephen Bodner, who retired last year as the head of the US Naval Research Laboratory's laser-fusion program, propose that Congress cap further spending on buying and assembling parts.

They suggest holding back on more money for NIF until the project completes a tough set of tests on one complete "bundle" of eight laser beams.

"Congress should not press ahead before Livermore has shown that the NIF laser system can meet all the project's technical requirements as tolerable cost," they write in a commentary in a current issue of the journal Nature.

(c) Copyright 2000. The Christian Science Publishing Society