A joint Fermilab/SLAC publication

LHC smashes old collision records


The Large Hadron Collider is now producing about a billion proton-proton collisions per second.

The LHC is colliding protons at a faster rate than ever before, approximately 1 billion times per second. Those collisions are adding up: This year alone the LHC has produced roughly the same number of collisions as it did during all of the previous years of operation together.

This faster collision rate enables scientists to learn more about rare processes and particles such as Higgs bosons, which the LHC produces about once every billion collisions.

“Every time the protons collide, it’s like the spin of a roulette wheel with several billion possible outcomes,” says Jim Olsen, a professor of physics at Princeton University working on the CMS experiment. “From all these possible outcomes, only a few will teach us something new about the subatomic world. A high rate of collisions per second gives us a much better chance of seeing something rare or unexpected.”

Since April, the LHC has produced roughly 2.4 quadrillion particle collisions in both the ATLAS and CMS experiments. The unprecedented performance this year is the result of both the incremental increases in collision rate and the sheer amount of time the LHC is up and running.

“This year the LHC is stable and reliable,” says Jorg Wenninger, the head of LHC operations. “It is working like clockwork. We don’t have much downtime.”

Scientists predicted that the LHC would produce collisions around 30 percent of the time during its operation period. They expected to use the rest of the time for maintenance, rebooting, refilling and ramping the proton beams up to their collision energy. However, these numbers have flipped; the LHC is actually colliding protons 70 percent of the time.

“The LHC is like a juggernaut,” says Paul Laycock, a physicist from the University of Liverpool working on the ATLAS experiment. “We took around a factor of 10 more data compared to last year, and in total we already have more data in Run 2 than we took in the whole of Run 1. Of course the biggest difference between Run 1 and Run 2 is that the data is at twice the energy now, and that’s really important for our physics program.”

This unexpected performance comes after a slow start-up in 2015, when scientists and engineers still needed to learn how to operate the machine at that higher energy.

“With more energy, the machine is much more sensitive,” says Wenninger. “We decided not to push it too much in 2015 so that we could learn about the machine and how to operate at 13 [trillion electronvolts]. Last year we had good performance and no real show-stoppers, so now we are focusing on pushing up the luminosity.”

The increase in collision rate doesn’t come without its difficulties for the experiments.

“The number of hard drives that we buy and store the data on is determined years before we take the data, and it’s based on the projected LHC uptime and luminosity,” Olsen says. “Because the LHC is outperforming all estimates and even the best rosy scenarios, we started to run out of disk space. We had to quickly consolidate the old simulations and data to make room for the new collisions.”

The increased collision rate also increased the importance of vigilant detector monitoring and adjustments of experimental parameters in real time. All the LHC experiments are planning to update and upgrade their experimental infrastructure in winter 2017.

“Even though we were kept very busy by the deluge of data, we still managed to improve on the quality of that data,” says Laycock. “I think the challenges that arose thanks to the fantastic performance of the LHC really brought the best out of ATLAS, and we’re already looking forward to next year.”

Astonishingly, 2.4 quadrillion collisions represent just 1 percent of the total amount planned during the lifetime of the LHC research program. The LHC is scheduled to run through 2037 and will undergo several rounds of upgrades to further increase the collision rate.

“Do we know what we will find? Absolutely not,” Olsen says. “What we do know is that we have a scientific instrument that is unprecedented in human history, and if new particles are produced at the LHC, we will find them.”

Latest news articles

World’s biggest neutrino experiment moves one step closer

The startup of a 25-ton test detector at CERN advances technology for the Deep Underground Neutrino Experiment.

Scientific American

Machine learning is crucial to staying ahead of hackers trying to break into at CERN’s massive worldwide computing grid.

New York Times

Physicists monitoring the Large Hadron Collider are seeking clues to a theory that will answer deeper questions about the cosmos.