Costly Projects Cast Science in a Harsh Light

The effort to construct the world’s most powerful laser is at least a billion dollars over budget and four years behind schedule, raising disquieting questions about the way large science projects in the United States are managed.

Easily the country’s largest science construction project, the stadium-sized National Ignition Facility (NIF) at the Lawrence Livermore National Laboratory is the keystone of the effort to ensure the safety and reliability of the nation’s nuclear weapons.

Sixty times more powerful than any laser ever built, the NIF laser is designed to focus the energy of 192 beams on a target the size of a rice grain in the hope of igniting fusion--thus allowing scientists to study safely the raging heart of a thermonuclear reaction. It offers a way to experiment with the forces in a nuclear explosion without detonating a bomb.

Not so long ago, a formal review judged the colossal laser effort “by far the best-managed of any U.S. government project.”

Federal auditors, however, recently reached a different conclusion. In all, the laser project could cost as much as $4 billion, twice what was expected, and could be up to six years late. So many scientific uncertainties remain that, even with more money and time, there is no guarantee that the laser will ever perform as promised, they said.

Now some in Congress are angry enough to kill it. While it is unlikely that they will succeed, their ire underscores a disenchantment with the way that large science projects are sometimes managed and the half-truths that scientists tell to keep them funded.

“They sold me a bill of goods and I am not happy about it,” said U.S. Sen. Harry Reid (D-Nev.). A former project supporter, he is now among those arguing to kill it.

“I cannot tell you how angry I am that [the Department of Energy] and the national laboratories consistently do this sort of thing to Congress. They over-promise and underdeliver at a vastly inflated price.

“Enough is enough.”

In that sense, the problems of the NIF laser are emblematic of a number of scientific mega-projects in recent years, from the space station to the ill-fated Superconducting Supercollider: A bold idea captures the official imagination and wins funding, only to have its proposed solutions to scientific conundrums grow increasingly complex as money starts to flow, deadlines loom and second thoughts set in.

Costs soar. Schedules stretch like Silly Putty. High expectations conceived with what one analyst calls “giddy techno-optimism” are scaled back.

Indeed, when management experts at the Rand Corp. examined 52 major civilian projects ranging in cost from $500 million to more than $10 billion, they found that the average cost growth was 88%. Only four of them were completed for the promised cost. The average schedule slip was 17%.

The bigger and more innovative the project, the more likely it was to go awry.

Since it was proposed a decade ago, the National Ignition Facility has been controversial, embroiled in arms control politics.

But what unsettles even some project proponents is that despite numerous contract compliance reports and outside peer review panels, so many of its problems went unnoticed for so long. They worry that the need to find new funding for the laser project will mean that other, perhaps equally important, nuclear weapons safety projects will be curtailed.

Ask Livermore director Bruce Tarter how so many bright, talented scientists got into so much trouble, and his answer is simple: “We screwed up. We missed it. Every review missed it. We all missed it.”

U.S. Sen. Tom Harkin (D-Iowa), who seeks to withhold extra funding for the project, puts it differently. “We’ve seen this pattern before: massive cost overruns, schedule delays, poor management, unresolved technical problems,” he said.

Frontier of Knowledge Is Uncertain Terrain

The frontier of knowledge is uncertain terrain for any scientific enterprise.

Cost overruns, delays and technical snafus are occupational hazards for anyone steering a major project toward completion, whether it is a mountaintop observatory, a space probe or a sprawling particle accelerator. This is especially true when the elements most crucial to the effort’s success have yet to be invented when the project begins.

When Livermore researchers conceived the laser facility, they did not know how to make the tons of ultra-pure glass it would require, how to grow the huge, 700-pound crystals for its lenses or how to protect its optics from the fierce light that would assault them. Only recently did they recognize that they also did not even know how to properly assemble its maze of super-clean laser tubing.

In just this way, several experts said, the laser project grew in a decade from a $400-million proposal to a multibillion-dollar project.

“The problem for large scientific projects is to do something that is being done for the first time, balanced against cost, schedule and promises to the government,” said Caltech physicist Barry Barish.

He is overseeing construction of a $271-million Laser Interferometer Gravitational Wave Observatory (LIGO) in Hanford, Wash., and Livingston, La., meant to detect ripples in the fabric of space and time.

“That is a hard balancing act,” he said.

Barish knows firsthand just how hard. When he took over the LIGO project in 1994, it was in so much trouble that Congress balked at funding it anymore. Costs had grown by 40% and it looked to be three years behind schedule.

Now, six years later, the sophisticated gravity wave detectors, each with five miles of underground vacuum pipes, are nearing completion. The LIGO met its revised budget and schedule in part by stepping back from the cutting edge to more modest goals.

In the same period, NASA’s space station project blossomed from an eight-year, $3-billion U.S. flagship to a $60-billion, 16-nation project that has been in the works for 16 years.

The first of many crews to live aboard the 14-story orbital outpost on a long-duration basis isn’t expected until later this fall, almost a decade later than originally planned. It could be five more years before the station is finished.

In any project, the quest for technical perfection can be unmanageable.

Over the past decade, for example, the Gravity Probe B project at Stanford University has spent about $500 million to build and orbit the world’s most perfect gyroscopes in a test of Einstein’s theory of relativity. Now university project managers need $72 million more to fix last-minute technical problems before the experiment can get off the ground. No one is sure where the extra money will come from.

Even when costs and schedules are tightly controlled, the complexity of today’s technical endeavors is such that a single misstep can be fatal, as a recent string of failed Mars probes demonstrated. They came in on time, on budget, and were lost in space one after the other, because of simple math errors and programming glitches.

But not since Congress canceled an $8-billion particle accelerator called the Superconducting Supercollider in 1993 has a large science project like the laser facility been mired in such financial and engineering difficulties.

Certainly, none has been so central to the future of the nation’s nuclear weapons program.

Controlling Extreme Conditions

Inside the National Ignition Facility, the 75-ton neutron shield is ajar, suspended on its massive hinges like the door of a vacant bank vault.

In hard hat and safety glasses, Bruce Warner strolls across the threshold to face an immense, 500-ton target chamber designed to contain conditions that normally exist only in the interior of stars or in an exploding nuclear weapon.

Warner, an experienced laser physicist at Livermore, is one of several newly appointed program managers on the NIF laser project. He has been on this job nine months, and he bristles at the suggestion that it won’t meet its goals.

“Given the opportunity to finish NIF, we are more optimistic now than when we started that we will get [fusion] ignition,” Warner said.

Slathered with a thick coating of blue gunite like the shell of a swimming pool, the 32-foot-high spherical chamber is studded with holes for the 192 laser tubes that will converge there.

One day, if the project is completed, it will be the heart of the largest and most complex optical system ever built, pieced together from about 7,500 optical elements, including 3,000 slabs of special optical glass, 2,500 fused silica lenses and 1,600 mirrors and polarizers. Other parts of the system will require about 25,000 other optical elements.

Its thousands of tubes and other components will have to be put together and maintained at a level of cleanliness as strict as a computer chip fabrication plant.

This federal beam machine will focus all its combined energy inside this sphere onto a BB-sized pellet of deuterium and tritium, with peak energies equal to 1,000 times the electrical generating power of the United States.

In the instant of ignition, as temperatures flare to several million degrees, the tiny target will be compressed to a plasma six times denser than at the center of the sun.

If all goes as planned, the result will be brief bursts of self-sustaining fusion reactions similar to those occurring in a star.

But for the moment, the sphere in front of Warner is simply the center of a cavernous construction site the length of two football fields, ripe with the smell of fresh concrete and new paint. It is the unfinished foundation of what its builders hope will be the showplace of 21st century laser physics.

To make that all happen, the project needs an additional $1 billion and about four years more than originally planned, even by the laboratory’s own accounting.

Madelyn Creedon, deputy administrator for defense programs at the National Nuclear Security Administration, part of the Department of Energy, said Friday that total project-related costs are now estimated to be $3.5 billion, compared to an original cost estimate of $2.1 billion. Even that figure, she said, does not include the cost of developing the laser’s special fusion target, estimated at $491 million.

The extra costs, Livermore managers said, include about $50 million for additional project management, $252 million for contingencies, and $200 million for manufacturing costs. About $250 million is the price in salaries and other fixed costs of taking so much longer to do it all.

Project managers said they expect to have all 192 laser beams operational by 2008, with initial experiments beginning as early as 2004.

When the full laser facility is up and running, they expect to conduct about 800 “shots” a year--about two a day--one-third as many as the laboratory promised in 1995.

As they began to revamp the project recently, one of the first things that Livermore’s new project managers realized is that they don’t know how to assemble the laser properly.

So, earlier this month, they announced they had hired an outside engineering company to do it for them at a contract cost of $230 million.

“The thing we really stubbed our toe on was a classic engineering of a big, complex thing,” said George Miller, Livermore’s most experienced weapons designer, who is now the laboratory’s associate director for NIF programs. Miller oversaw the development of the 150-kiloton W84 thermonuclear warhead used for ground-launched cruise missiles.

“We thought we knew how to put it together,” Miller said. “The fact of the matter is what we did not appreciate is that NIF is a project whose scale is significantly beyond the experience of the laboratory. Putting this thing together is like putting together a Rubik’s Cube.”

Outside Reviews Called Flawed

There were many outside reviews of the laser project before construction began. But most were compromised by conflicts of interest or a lack of independence, according to Stephen Bodner, former head of the laser fusion program at the U.S. Naval Laboratory, and Christopher Paine, a senior researcher at the Natural Resources Defense Council, which has long opposed the project.

Caltech Provost Steve Koonin was chairman of a panel that evaluated the project for the National Academy of Sciences in 1990 and again in 1997. So many members of the 1997 panel were project supporters, consultants or advisors that the Resources Defense Council obtained a federal court injunction barring the Energy Department from using the report or disseminating it for several years.

After the laser’s problems came to light recently, Koonin chaired a review of the project for the University of California, which manages the three national weapons laboratories. The Department of Energy had fined the university $2 million of its normal management fee for its failure to properly oversee the laser project.

Until this year, Koonin had no serious doubts about its feasibility or the ability of those in charge to manage it.

“Frankly, there was not a hint we should have been concerned,” Koonin said. However, he said, the project “should have been watched more carefully.”

Now, after months of management reviews and official recriminations, the people in charge of the laser project at Livermore are contrite.

“NIF had some significant problems,” said laser project manager Edward Moses, who took over last year. “They had to be faced and fixed. We expected to take our lumps and we did take our lumps for the past year.”

The visionaries have been reassigned or found new jobs, Livermore officials say. There is a new team in charge: practical people with a more realistic attitude born of decades managing complex projects.

“Essentially, the whole top management team of the project is different,” Miller said.

Ask Miller, Moses and their colleagues at Livermore today about issues raised by the federal auditors, and they say they are confident that they have fixed their most serious problems. Unlike their predecessors, they are more open to outside advice, and optimistic that they can solve any remaining scientific difficulties.

They cite a stringent review by 40 Energy Department scientists and consultants, which was released Friday, to demonstrate that they are back on the right track. That review concludes that Livermore’s new project plan is “reasonable, self-consistent and credible.” The laser can be built on time and on budget under the new schedule and cost estimates, they said.

The problem now is that no one may believe them.

Whatever the laser project ultimately costs the taxpayer, Livermore itself has already paid a high price in its credibility, several congressional analysts and independent critics of the program said.

Not so many months ago, Energy Secretary Bill Richardson stood before an audience of Livermore employees to laud their skill in keeping the country’s largest science construction project within budget and on time. They had earned their government’s “admiration and gratitude,” he said.

Listening in the audience that day in June 1999, however, were as many as two dozen laboratory scientists who knew that the laser project was in serious trouble. Some had known for months. Lab director Bruce Tarter had known formally for three days before the secretary’s visit, he later acknowledged.

Even so, they all kept their concerns to themselves, allowing Richardson to make his speech unchallenged by what they knew and to then pass on his misleading assessment to Congress a few weeks after the ceremony.

It was another three months before the full scope of the project’s troubles began to reveal themselves.

“They lied to us,” Harkin said. “They simply lied to us.”

(BEGIN TEXT OF INFOBOX / INFOGRAPHIC)

Star Chamber

By igniting a fusion reaction, the 192 lasers of the National Ignition Facility at the Lawrence Livermore National Laboratory could allow researchers to study thermonuclear reactions, like those inside a star or in an exploding nuclear weapon, without actually detonating a bomb or a warhead in a test.

*

The laser beam path

The laser beam travels hundreds of yards from the point where an optical pulse is generated to the target chamber. When the light reaches the target, it has been amplified more than 10 trillion times and changed from infrared light to ultraviolet light.

*

What is fusion?

Just as the large nucleus of an atom can split, releasing the energy of fission, so the neutrons and protons at the center of some atoms can fuse together to release energy. The Livermore facility will use lasers to fuse atoms of deuterium and tritium.

*

Source: Lawrence Livermore National Laboratory.

Researched by LEE HOTZ and LYNN MEERSMAN/Los Angeles Times