The biggest, most expensive physics machine in the world is riddled with thousands of bad electrical connections.

Many of the magnets meant to whiz high-energy subatomic particles around a 17-mile underground racetrack have mysteriously lost their ability to operate at high energies.

Some physicists are deserting the European project, at least temporarily, to work at a smaller, rival machine across the ocean.

After 15 years and $9 billion, and a showy “;switch-on”; ceremony last September, the Large Hadron Collider, the giant particle accelerator outside Geneva, has to yet collide any particles at all.

But soon?

This week, scientists and engineers at the European Center for Nuclear Research, or CERN, are to announce how and when their machine will start running this winter.

That will be a Champagne moment. But scientists say it could be years, if ever, before the collider runs at full strength, stretching out the time it should take to achieve the collider's main goals, like producing a particle known as the Higgs boson thought to be responsible for imbuing other elementary particles with mass, or identifying the dark matter that astronomers say makes up 25 percent of the cosmos.

The energy shortfall could also limit the collider's ability to test more exotic ideas, like the existence of extra dimensions beyond the three of space and one of time that characterize life.

“;The fact is, it's likely to take a while to get the results we really want,”; said Lisa Randall, a Harvard physicist who is an architect of the extra-dimension theory.

The collider was built to accelerate protons to energies of 7 trillion electron volts and smash them together in search of particles and forces that reigned earlier than the first trillionth of a second of time, but the machine could run as low as 4 trillion electron volts for its first year. Upgrades would come a year or two later.

Physicists on both sides of the Atlantic say they are confident that the European machine will produce groundbreaking science—eventually—and quickly catch up to an American rival, even at the lower energy.

All big accelerators have gone through painful beginnings. “;These are baby problems,”; said Peter Limon, a physicist at the Fermi National Accelerator Laboratory in Batavia, Ill., who helped build the collider.

But some physicists admit to being impatient. “;I've waited 15 years,”; said Nima Arkani-Hamed, a leading particle theorist at the Institute for Advanced Study in Princeton. “;I want it to get up running. We can't tolerate another disaster. It has to run smoothly from now.”;

The delays are hardest on younger scientists, who may need data to complete a thesis or work toward tenure. Slowing a recent physics brain drain from the United States to Europe, some have gone to work at Fermilab, where the rival Tevatron accelerator has been smashing together protons and antiprotons for the last decade.

Colliders get their oomph from Einstein's equivalence of mass and energy, both expressed in the currency of electron volts. The CERN collider was designed to investigate what happens at energies and distances where the current reigning theory, known as the Standard Model, breaks down and gives nonsense answers. It got the nod in 1994 after the Superconducting SuperCollider, which would have been 54 miles around and 20 trillion electron volts, was canceled by Congress.

The collider's own prodigious energies are in some way its worst enemy. At full strength, the energy stored in its superconducting magnets would equal that of an Airbus A380 flying at 450 miles an hour, and the proton beam itself could pierce 100 feet of solid copper.

In order to carry enough current, the collider's magnets are cooled by liquid helium to a temperature of 1.9 degrees above absolute zero, at which point the niobium-titanium cables in them lose all electrical resistance and become superconducting.

Any perturbation, however, such as a bad soldering job on a splice, can cause resistance and heat the cable and cause it to lose its superconductivity in what physicists call a “;quench.”; Which is what happened on Sept. 19, when the junction between two magnets vaporized in a shower of sparks, soot and liberated helium.

Technicians have spent most of the last year cleaning up and inspecting thousands of splices in the collider. About 5,000 will have to be redone, Steve Myers, head of CERN's accelerator division, said in an interview.

The exploding splices have diverted engineers' attention from the mystery of the underperforming magnets. Before the superconducting magnets are installed, engineers “;train”; each one by ramping up its electrical current until the magnet fails, or “;quenches.”; Thus the magnet gradually grows comfortable with higher and higher current.

All of the magnets for the collider were trained to an energy above 7 trillion electron volts before being installed, Myers said, but when engineers tried to take one of the rings' eight sectors to a higher energy last year, some magnets unexpectedly failed.

In an e-mail exchange, Lucio Rossi, head of magnets for CERN, said that 49 magnets had lost their training in the sectors tested and that it was impossible to estimate how many in the entire collider had gone bad. He said the magnets in question had all met specifications and that the problem might stem from having sat outside for a year before they could be installed.

Retraining magnets is costly and time consuming, experts say, and it might not be worth the wait to get all the way to the original target energy. “;It looks like we can get to 6.5 relatively easily,”; Myers said, but 7 trillion electron volts would require “;a lot of training.”;

Many physicists say they would be perfectly happy if the collider never got above five trillion electron volts. If that were the case, said Joe Lykken, a Fermilab theorist who is on one of the CERN collider teams, “;It's not the end of the world. I am not pessimistic at all.”;

For the immediate future, however, physicists are not even going to get that. Myers said he thought the splices as they are could handle 4 trillion electron volts.

“;We could be doing physics at the end of November,”; he said in July, before new vacuum leaks pushed the schedule back a few additional weeks.

“;It's not the design energy of the machine, but it's 4 times higher than the Tevatron,”; he said.

Pauline Gagnon, an Indiana University physicist who works at CERN, said she would happily take that energy level. “;The public pays for this,”; she said in an e-mail message, “;and we need to start delivering.”;