The Large Hadron Collider (LHC) at CERN

From Scholarpedia
This article has not yet been published; it may contain inaccuracies, unapproved changes, or be unfinished.
Jump to: navigation, search




LHC map
Figure 1: Location of the LHC

The Large Hadron Collider (LHC) is a high energy particle collider at the European Laboratory for Particle Physics (CERN). Situated underground it is approximately circular with a circumference of 26.7 km and it straddles the border between France and Switzerland just outside Geneva. Its primary goal is to deliver proton-proton collisions at a centre of mass energy of 13 to 14 TeV. It can also operate with ions and has delivered both lead-lead and proton-lead collisions. The LHC has four main experiments: ATLAS, CMS, ALICE and LHCb. Of these, ATLAS and CMS are general purpose detectors (GPD) designed for high luminosity and searches in a wide variety of channels.

The high energy is based on the large scale use of high field superconducting magnets. In particular the main bending magnets, described in more detail below, are a 2 beam-pipe design with a nominal dipole field of 8.3 T. The LHC produces the high collision rates required by using a large number of bunches with high bunch population, and by producing small beam sizes at the interaction points. CERN’s injector complex delivers the requisite high intensity bunches with exacting demands from the LHC on transverse beam size and longitudinal characteristics.

Project History

The feasibility study for a proton–proton collider in the LEP tunnel started at the end of 1982 and the presentation of this study at at the ECFA–CERN workshop "Large Hadron Collider in the LEP tunnel" held in Lausanne and Geneva 21 to 27 March 1984 marked the LHC epoch. The study took an in depth look at: machine parameters and performance; magnets; cryogenics; vacuum; RF; injection; beam transfer and the beam dumps; and radiation protection and safety aspects.

The results were present in the workshop proceedings as “A feasibility study of possible machine options" [1]. Similar effort was made in the consideration of options for the experiments and the physics motivation for the LHC. The machine part of the workshop was driven by acceptance of the key physics challenges of luminosity and energy. Luminosity requirements dictated the choice of proton-proton collisions with a target luminosity larger than 1x1033 cm-2s-1. The proposed centre-of-mass energy range was 10 to 20 TeV. This was essentially given by the size of the LEP tunnel and the maximum conceivable dipole field. Choice of ranges for luminosity and energy were motivated by consideration of the achievable physics and, importantly at the time, the need to compete with the planned superconducting super-collider (SSC). Comparing the initial parameter choices with what has eventually be achieved, one notes that the luminosity, and associated beam parameters have been exceeded all expectations. The foreseen operating beam energy, however, was an optimistic 8 to 9 TeV.

The initial phases outlined above paralleled two other important accelerator development, both in their own way having a large impact on the evolution of the LHC. The first was the SSC whose design and construction paralleled that of the LHC, and was undoubtedly a important competitor before its eventual cancellation in October 1993. The other major project was LEP whose construction, commissioning, and operation spanned 1983 to 2000, rivalling the LHC if only for resources.

Following the presentation of the preliminary feasibility studies in 1984, the following years were devoted to the basic technical choices and the exploration of the parameter space. From 1988 to 1991 magnet prototypes produced by European industry demonstrated the feasibility of the basic magnet design and in 1991 a magnet R&D facility was started at CERN. The project approved by the CERN council in 1994. By 1996 the main features of the dipole magnet design were frozen and process of transferring the associated to industry was started with the main contracts being signed between 1998 and 2001.

2003 to 2004 saw the start of tunnel installation and the arrival on site of hundreds of magnets and substantial progress on all other systems. Magnet installation took place between 2005 to 2007 followed by the sector by sector cool-down and testing of the superconducting magnets and associated systems.

The first circulating beam was established on 10th September 2008. However 9 days later a major fault developed in the interconnect between 2 dipole magnets. This caused extensive damage in one sector of the ring and the incident required the removal and repair of tens of dipoles and quadrupoles. In parallel a huge effort went into understanding the problem and improving the machine protection systems.

Beam was re-injected on November 29th 2009 and first collisions at a centre of mass energy of 7 TeV were delivered to the experiments on 30th March 2010.


The LHC has 8 arcs and 8 straight sections between the arcs. The straight sections are around 528 m long. Each straight section has associated with it surface and underground installations, lifts, and a wide variety of technical infrastructure. These locations are referred to as the LHC points.

  • The four main experiments are situated at point 1 (ATLAS), point 2 (ALICE), point 5 (CMS) and point 8 (LHCB).
  • Injection of clockwise beam (beam 1) takes place at point 2. Injection of anti-clockwise beam (beam 2) takes place at point 8.
  • The main collimator installations are at points 3 and 7.
  • The radio frequency (RF) system is situated at point 4.
  • The beam dump system is situated at point 6.

Functionally the LHC is divided into 8 sectors (namely sectors 12, 23, 34, 45, 56, 67, 78, 81). A sector spans the underground installation between 2 LHC points - thus sector 12 lies between point 1 to point 2 etc. Importantly a sector can be cooled and powered independently.

Description of the LHC and its main components

The energy of the LHC was essentially determined by the bending radius of the LEP tunnel and the achievable dipole field strength. High luminosity is achieved by a large number of bunches; high bunch population; and small beam size at the interaction points.

The magnet lattice is a standard separated function, strong focusing design.

The four experiments' long straight sections are equipped with high gradient quadrupoles that allow the production of small beams at the interaction points. In ATLAS and CMS, in particular, the very small beam sizes produced are key to obtaining very high collision rates.

It was clear from the start that superconducting magnet technology would have to be used and choice of cooling the magnets with superfluid helium at 1.9 K was early decision. Two separate beam pipes allow for a large number of bunches and a "2 in 1" magnet design was quickly adopted.



The LHC has a veritable zoo of magnets. The principle elements are the main dipoles, main quadrupoles, and the insertion region quadrupoles and the inner triplet quadrupole magnets. The inner triplets are in fact a set of 4 high gradient quadrupole magnets symmetrically placed around each experiment. Besides these there are hundreds of magnets ranging through closed orbit dipole correctors, trim quadrupoles, sextupoles, octupoles, decapoles and even duodecpole corrector magnets.

The main dipoles are a "2-in-1" design using niobium-titanium Rutherford cable cooled to 1.9 K.

Number of dipoles Dipole length Dipole field at 450 GeV Dipole field at 7 TeV
1232 14.3 m 0.53 T 8.33 T


The LHC is fully dependent on the industrial scale cryogenics system. There are 5 cryogenics plants situated around the ring with surface installations feeding an extensive underground system.

  • The final systems cools 24 km of superconducting magnets at 1.9 K. This represents 1800 superconducting magnets with a cold mass of around 37,000 tonnes.
  • Fully operational the system has 88 tons of superfluid helium at 1.9 K out of total inventory of 150 tonnes of helium.
  • The cooling power at 4.5 K is 8 x 18kW.

The overall system represents the installation of a leak free superfluid helium system with tens of thousands of welds, and effective thermal insulation and heat-load management. The final numbers are impressive and the operation of the total system with good overall availability must be regarded as a major success.

Beam vacuum

The so-called cryogenics vacuum is that in the beam pipes in the cold part of the ring. The challenge is good beam lifetime in a cryogenic system where heat input into the 1.9 K helium circuit must be minimized and where significant quantities of gas can be condensed on the vacuum chamber. In cold sectors pumping is insured by cold surfaces for all gases except helium. To avoid subsequent desorption, low initial pressures are required before cool-down, and this is ensured by turbo molecular pumps etc.

Potential heat input from a variety of sources (electron cloud, synchrotron radiation, image currents) was clearly a concern, and lead to development of the beam screen. The beam screen is used to minimize the impact of these dynamics effects on the pressure and heat load to the cold magnets. Related concerns include the need for low resistivity (hence a copper layer on the beam screen); and beam stability (via a drive for low impedance).

There are major mechanical challenges in the cold interconnects where flexibility is required to allow for thermal and mechanical offsets during alignment and operation at 1.9 K, while ensuring good electrical continuity to ensure low impedance. This led to the development of the so-called plug-in modules (PIMs).

In the warm sections non-evaporable getter (NEG) provides most of the pumping capacity, with additional ion pumps for the noble gases which are not pumped by the NEG.

Power converters

The LHC power converters that supply the superconducting magnet circuits need to produce high current, at low voltage, with very high precision, accuracy, reproducibility, stability and resolution over a large dynamic current range. The exquisitely stringent demands arise from beam based constraints on, for example, ripple and the induced optics imperfections arising from quadrupole gradient errors.


The key role of the collimation system is to stop high amplitude particles (both in betatron and momentum space) impacting the cold mass. A collimator typically consists of two meter long jaws made of carbon fibre-reinforced carbon (carbon-carbon). The jaws are carefully manufactured to present a flat surface to the beam and allow micron level set-up accuracy. Over 100 individual collimators are set-up in a strict hierarchy with respect to the beams. Closest to the beam are the so-called primary collimators with the secondary and tertiary collimators positioned further away from the beam.

The cleaning process is multi-turn. The protons that impact the primary collimators are scattered to larger amplitudes and impact the secondaries on subsequent turns. The showers produced in the secondaries are absorbed by down-stream absorbers. Much work has gone into the development of an understanding of the behaviour of scattered protons and the development of showers.

The solution deployed uses robust carbon-carbon solutions for primary and secondary collimators and tungsten for tertiary and absorbers. The principle cleaning sections are situated in point 7 (betatron) and 3 (momentum). The betatron cleaning section, for example, consists of 3 primary collimators, 11 secondary collimators, and 5 absorbers for each beam. Tertiary collimators are situated on the incoming beam in the experimental interaction regions to intercept the tertiary halo and offer protection to the inner triplets.

The collimation system also has to withstand the potential impact of an asynchronous beam dump and in this regard provides an important passive protection role, safely catching the swept beam and preventing the impact of primary protons on the cold mass. Additional protection devices are positioned down stream of the beam dump system to fully protect against the associated risks. There are also a series of protection devices at the injection regions to protect the downstream elements in case of injection kicker misfires etc.

Beam dump system

The beam circulating in one of the two LHC rings has a stored energy of up to 360 MJ when the LHC operates at 7 TeV with its nominal intensity. The local loss of only a very small fraction of this beam is sufficient to induce a quench in one of the superconducting magnets; the loss of the whole beam in an uncontrolled way could cause very serious damage to equipment.

The damage potential of the LHC beams make the LHC beam dump system (LBDS) the most critical subsystem in the machine. There is, by design, tight coupling with the Beam Interlock System (BIS).

The LBDS is designed to perform fast extraction of beam from the LHC in a loss free way []. For each beam a system of 15 horizontal kicker magnets (MKD), 15 vertically deflecting magnetic septa (MSD) and 10 diluter magnets (MKB) is installed. After the kickers the beam sees an additional deflection when traversing the down-stream Q4 quadrupole. The MSD deflect the beam vertically before it is further swept in the horizontal and vertical planes in a spiral shape by the MKB kickers. After several 100 m of beam dump line the beam is absorbed by the dump block (TDE).

To protect the septa from mis-kicked beam a special fixed 8 m long graphite protection device (TCDS) is placed just in front of the MSD. For nominal operations the MKD rise time should always be accurately synchronised with the 3 micro-second abort gap in the beam, so that no beam is swept across the aperture. However some failures can occur which lead to an asynchronous dump. In addition stray particles may also be present in the abort gap. To protect the LHC aperture from these eventualities, a movable single-jawed 6 m long graphite protection device (TCDQ) is installed upstream of Q4, supplemented by a two-jaw 1 m long graphite secondary collimator (TCSG) and a 2 m long fixed iron mask (TCDQM).


Injection of beam 1 takes place in IR2 following transfer from the SPS down transfer line TI2, injection of beam 2 takes place in IR8 following transfer from the SPS down transfer line TIT. Each injection system consist of 5 horizontal septa followed by 4 vertical kicker magnets. Considerable care is required to meet the challenge of injecting potentially damaging intensity into a superconducting environment through limited aperture.

Injection projection devices include transfer line collimators to protect against extraction and transfer line problems. Downstream of the each set of kickers is movable absorber (TDI) consisting of two vertical 5 m long jaws designed to intercept the beam in case of mis-kick of the incoming or circulating beam.

Machine Protection System

Machine protection is vitally important for LHC operation over the safe beam limit. In essence it comprises the beam interlock system (BIS) and the safe machine parameter system (SMP) []. The BIS takes inputs from a large multitude of users and will trigger a beam dump if any signal a problem by breaking the interlock loop.

The beam drives a subtle interplay of the LBDS, the collimation system and protection devices, which rely on a well-defined aperture, orbit and optics for guaranteed safe operation.

Progress with beam

2008 to 2009

Beam was first introduced into the LHC during a series of injection tests in August and September 2008 [2]. Even at this early stage it appeared that the magnetic machine and optics were in good shape, the aperture was clear and that key beam instrumentation was functional. On 10th September 2008, in a very public display, beams were threaded around the full ring. Beams were captured by the RF system shortly after. Initial progress was good but commissioning with beam was cut brutally short by the incident on the 19th September [3].

The majority of 2009 was spent in the extensive repair of sector 34. Importantly a more robust and sensitive quench protection system (nQPS) system was developed and deployed. A splice measurements campaign on both warm and cold sectors established a good understanding of the issues that caused the incident and indeed revealed hidden dangers. The splices in the interconnects remained a concern and the decision was made to limit the beam energy. Initially to to 3.5 TeV in 2010 and 2011, and subsequently to 4 TeV in 2012.

Beam was circulated again on 29th November 2009. Once beam was back there was rapid progress in the three and half weeks available in November to December. Collisions with stable beam conditions were established at 450 GeV, and the ramp to the maximum energy at the time of 1.18 TeV was successfully attempted.


2010 was devoted to commissioning and establishing confidence in operational procedures and the machine protection system. At this stage the basics were sorted out, laying the foundation for what followed. Ramp commissioning to 3.5 TeV was smooth and led to first collisions at 3.5 TeV un-squeezed on the 30th March 2010. Squeeze commissioning successfully reduced the beta* to 2.0 m in all four experiments. After the squeeze was commissioned there was a period of stable beams punctuated with continued system commissioning. In June the decision was taken to go for bunches with nominal intensity. This involved another extended commissioning period which included the need to stabilize single beam instabilities. There was a halting push through nominal intensity commissioning to a total stored beam energy of around 1 to 3 MJ.

To increase the number of bunches the move to 150 ns bunch trains was made and the crossing angles in the experimental IRs brought on. A phased increase in total intensity was then performed. The proton run finished with beams of 368 bunches of around 1.2x1011 protons per bunch, and a peak luminosity of 2.1x1032 cm-2s-1. The operational year ended with a 4 week lead-lead ion run.


The beam energy remained at 3.5 TeV in 2011 and the year saw combined exploitation and the exploration of performance limits. Re-commissioning with beam after the Christmas technical stop took around 3 weeks. There was a ramp-up to around 200 bunches (75 ns) taking about 2 weeks. There was then a scrubbing run of 10 days which included 50 ns injection commissioning. After an encouraging performance the decision was taken to go with 50 ns bunch spacing. A staged ramp-up in the number of bunches then took place with 50 ns bunch spacing up to a maximum of 1380 bunches.

Having raised the number of bunches to 1380, performance was further increased by reducing the emittances of the beams delivered by the injectors and by gently increasing the bunch intensity. The result was a peak luminosity of 2.4x1033 cm-2s-1 and some healthy delivery rates which topped 90 pb-1 in 24 hours.

The next step up in peak luminosity followed a reduction in beta* in ATLAS and CMS from 1.5 m to 1 m. This was made possible by careful measurements of the available aperture in the interaction regions concerned. These measurements revealed excellent aperture consistent with a very good alignment and close to design mechanical tolerances. The reduction in beta* and further gentle increases in bunch intensity produced a peak luminosity of 3.8x1033 cm-2s-1.

2012 and 2013

2012 was a production year at an increased beam energy of 4 TeV. The choice was made to continue to exploit 50 ns and run with a total number of bunches of around 1380. Based on the experience of 2011, the decision was taken to operated with tight collimator settings. The tighter collimator hierarchy shadows the inner triplet magnets more effectively allowing a more aggressive squeeze to a beta* of 0.6 m.

The price to pay was increased sensitivity to orbit movements, particularly in the squeeze, and increased impedance. The latter having a clear effect on beam stability as expected. There was the a determined and long running campaign to improve peak performance. This was successful to a certain extent, and revealed some interesting issues at high bunch and total beam intensity, but had little effect on integrated rates. Instabilities, although never debilitating, were a reoccurring problem and there were phases when they cut into operational efficiency.

It was a long operational year and included the extension of the proton-proton run until December resulting in the shift of a four week proton-lead run to 2013. Integrated rates were healthy at around the 1 fb-1 per week level and this allowed a total for the year of about 23 fb-1 to be delivered to ATLAS and CMS.

Other experiments

Besides the delivery of high instantaneous and integrated proton-proton luminosity to ATLAS and CMS, the LHC was also able to fulfil a number of other physics programs.

  • 2010 and 2011 saw lead-lead ion runs which delivered 9.7 and 166 ub-1 respectively at an energy of 3.5Z TeV. Here the clients were ALICE, ATLAS and CMS.
  • Luminosity levelling at around 4x1032 cm-2s-1 via transverse separation enabled LHCb to collect 1.2 and 2.2 fb-1 in 2011 and 2012 respectively.
  • ALICE enjoyed some sustained proton-proton running in 2012 at around 5x1030 cm-2s-1 with collisions between enhanced satellite bunches and the main bunches.
  • There was a successful beta* = 1 km run for TOTEM and ALFA in 2102. With tmin of approximately 0.0004 GeV2 this was the first LHC measurement in Coulomb-Nuclear Interference region.
  • The three years operational period of Run 1 culminated in successful proton-lead run at the start of 2013. ALICE, ATLAS, CMS and LHCb all took data.
Personal tools

Focal areas