Workers prepare to install equipment at the LIGO facility in Hanford, Washington, in 2011.

Workers prepare to install equipment at the LIGO facility in Hanford, Washington, in 2011.

LIGO Laboratory

Got gravitational waves? Thank NSF’s approach to building big facilities

Nobody thinks of government bureaucrats when they hail a scientific breakthrough. But yesterday’s announcement that researchers working at the Laser Interferometer Gravitational-Wave Observatory (LIGO) detected gravitational waves might never have occurred without an earlier, nonscientific development. The National Science Foundation (NSF)—a U.S. agency historically focused on funding small-scale academic research—first had to figure out how to manage the financially perilous task of building a large and costly scientific facility.

More than a quarter-century ago, physicists asked NSF to spend nearly $300 million on an instrument to measure the gravitational waves that researchers had predicted would result from a cataclysmic event in deep space, like the merger of two black holes. That was no easy sell. The project would require technological leaps and bounds that even other scientists found incredible, recalls Rainer Weiss, a physicist at the Massachusetts Institute of Technology in Cambridge, who described the original idea for LIGO in paper in 1972. “People thought we were crazy,” he says.

But Weiss won the backing of long-time NSF program officer Richard Isaacson, and over the next decade the agency provided millions of dollars in research grants to advance the concept. NSF was OK with the fact that researchers didn’t expect to detect such waves for more than a decade, and only after the instrument had been upgraded. But they were very concerned that LIGO, the most costly project in the agency’s history, would blow a hole in the budget not just for astrophysics, but across all of NSF’s physical sciences directorate and, possibly, the agency itself.

It wasn’t an idle fear. Other federal agencies had seen the costs of big science facilities spiral out of control, wreaking havoc with their budgets, roiling researchers, and triggering harsh criticism from lawmakers.

A new account

Created in 1950, NSF had earned a stellar reputation funding basic research on university campuses. Unlike its larger government cousins, NSF didn’t manage big national laboratories (like the Department of Energy), nor build costly scientific payloads for spacecraft (like NASA). But by the late 1980s some academic disciplines had begun to assert that they needed one-of-a-kind large facilities like LIGO to remain globally competitive. And they turned to NSF for help.

With a budget that stood at $2 billion in 1990, NSF had little room—and scant experience—tackling projects like LIGO. “It was crazy,” then-NSF Director Neal Lane, now a university professor at Rice University in Houston, Texas, recalls about the agency’s traditional approach to supporting science. “You ended up cutting grants in order to fund a big new facility.”

So NSF officials conferred with their oversight body, the National Science Board, and came up with the idea of creating something new for the agency: a separate capital budget for brick-and-mortar projects. It would allow Congress to support multiyear projects without having to shrink the pot for bread-and-butter research grants. They called it the Major Research Equipment (MRE) account.

NSF sold the idea to Congress with the help of powerful legislators from states with a vested interest in LIGO—in particular Louisiana, one of two states that would host the facility, and Massachusetts, which had teamed with the California Institute of Technology (Caltech) in Pasadena to design the project.

Congress adopted the idea in 1994, just in time to handle the ramping-up of LIGO, which was being managed by Caltech. (Conflicts within the project’s upper management had led to a 1-year hiatus in construction, which had begun in 1992. NSF made two moves that righted the ship: approving the replacement of founding project director Rochus Vogt with Barry Barish, and appointing a new project manager at NSF to oversee operations.)

LIGO began doing science in 2002, 4 years later than initially scheduled. In 2008, NSF began building a $205 million update, called LIGO II, and completed it last year.

LIGO was the first project to benefit from the MRE account, Lane recalls. The new account also made it much easier for NSF to accommodate a 40% jump in LIGO’s original cost estimate, an increase that otherwise almost certainly would have required curtailing existing programs.

LIGO’s bureaucratic legacy

The MRE account has now become a fixture in NSF operations. In its early years it financed the twin 8-meter Gemini telescopes in Hawaii and Chile, a new South Pole station in Antarctica, and what eventually became a billion-dollar array of millimeter telescopes in Chile. Now, funded at a steady-state level of roughly $200 million annually, the MREFC (the words “facilities construction” were later added to its name) account this year supports work on the $344 million Daniel K. Inouye Solar Telescope, the $473 million Large Synoptic Survey Telescope, and the $434 million National Ecological Observatories Network, or NEON.

Although the account has been used mostly to build facilities in the physical sciences, the inclusion of NEON, the first project supported by NSF’s biology directorate, is no accident. “The MRE account was meant to encourage disciplines to take the long-range view of what facilities they needed to stay on the cutting-edge of science,” says Lane, a physicist. In 2014, for example, NSF completed construction of the $384 million Ocean Observatories Initiative, a global array of marine and seismic instruments. The account has also funded the recently commissioned Sikuliaq, an ice-capable research vessel now working in the Arctic.

Many of those projects have gone smoothly, whereas others have struggled. Just last week, a senior NSF official promised the science committee for the U.S. House of Representatives that the agency would follow recommendations in a new report by the independent National Academy of Public Administration detailing how NSF needs to improve its oversight of such large projects. At the same time, NSF is looking for a new team to manage NEON, which last summer was shrunk in size and scope after its costs soared and construction fell behind schedule.

NSF officials readily admit that building and managing large scientific facilities remains a challenge. “As we do more projects we learn,” says NSF Director France Cordóva, in Arlington, Virginia.

For the moment, however, she and other NSF officials are basking the glory of LIGO’s discovery. In their minds, the detection of a gravitational wave is a ringing endorsement of their decision more than 2 decades ago to get into the construction business. “The NSF likes to think of itself as the agency that can undertake high-risk, high-reward projects,” says Cordova, “and this is certainly one of them.”

With reporting by Adrian Cho.