Final Report Summary - SLHC-PP (Preparatory Phase of the Large Hadron Collider Upgrade)
Project context and objectives:
Proton collisions are all about creating new particles. Albert Einstein showed that energy and mass were interchangeable currencies, and the Large Hadron Collider (LHC) is a bureau de change of sorts. The CERN chain of particle accelerators packs dense beams of protons into two beam-pipes of the LHC, beneath the Swiss-French border. Travelling in opposite directions through the 27 km-long circular tunnel, the protons receive their final acceleration up to tera electronvolt (TeV) energies (3.5 TeV today, and 7 TeV after 2013).
The LHC then collides the protons together at the hearts of four particle detectors - ATLAS and CMS being the largest - that intersect the circular tunnel. When these collisions occur, the energy in the protons transforms into mass, generating a shower of new particles. The higher the energy of the collision, the heavier the particles that form in the debris. And they aren't just any particles - occasionally, the LHC will produce objects and interactions that have never before been seen by physicists, providing vital experimental evidence to back up the current theories of why the Universe is the way it is.
Though particle physics studies the very small, new observations could reveal the explanations for cosmic mysteries - what gives objects their mass, the nature of the 'invisible' matter that makes up most of the Universe, how the four forces that govern everyone and everything might unite into a super-force, and - maybe - why gravity is so much weaker than the other three forces.
The LHC is already the most powerful particle collider in the world, but for it to live up to its true potential, it must be upgraded to run at the highest possible luminosity. If the LHC is imagined as a torch probing the darkest recesses of the Universe, it's easy to see why a brighter beam would be more useful. A torch is made brighter by increasing concentration of photons in the beam, but the LHC beam will be brightened by making the proton bunches within it even denser.
Most of the proton collisions in the LHC will produce debris that is already familiar to physicists. The large or unusual particles they are most interested in are diamonds buried in heaps of common rubble. But while diamonds tend to stick around, the new particles are often extremely short lived: they disintegrate in fractions of a billionth of a second. To catch a glimpse of the fleeting particles, researchers take images of them through the four major detectors - recording particle 'signatures'.
Physicists rely on a complicated statistical process to separate the rare signatures of new physics from the huge number of similar signatures already expected to show up due to known physics. The larger the quantity of data physicists have to search through, the greater the chance of finding and confirming the signatures of rare particles.
The LHC upgrade will extend the lifetime of this scientific exploration into the unknown and produce ten times more data during the collider's second decade than during its first.
The main aim of the SLHC-PP was to prepare the first phase of this upgrade project for a formal proposal in 2011. Researchers needed to demonstrate new technologies, find solutions for radiation safety issues, form the collaborations that would carry out the work, and prepare the necessary management tools.
The SLHC-PP explored key upgrade pathways, demonstrating solutions to better focus the beams, design key components for the beam injectors, and power detectors more efficiently. A dedicated SLHC-PP team addressed radiation safety concerns by analysing the immediate and long-term effects of beam loss, in view of designing means to ensure safety and minimise the impact on the environment. On the collaboration side, the SLHC-PP contributed to the efforts in planning and initiating the upgrade projects of the ATLAS and CMS detectors. With the support of the SLHC-PP, existing tools for financial management, monitoring progress, and quality assurance have been improved, using the experience gained during the construction of the LHC.
The SLHC-PP activities already represent work from 18 institutions in 10 countries, but they also join with larger efforts coordinated through the detector collaborations and CERN's accelerator departments.
Work package (WP)1: Project management
The multifaceted SLHC-PP effort was managed by the WP1 team - comprising the project coordinator, the deputy project coordinator and the administrative manager. Roland Garoby and Duccio Abbaneo succeeded Lyndon Evans and Lucie Linssen as project coordinator and deputy, respectively, at the beginning of 2010. Mar Capeans has been the administrative manager throughout the project. The members of WP1 handled contracts and finances, monitored and reported progress in the other work packages, and took care of disseminating information both within the SLHC-PP and outside the collaboration.
In the first year, the management team set up a collaboration website (deliverable report 1.2.1) much of which is available for anyone to view. They also commissioned a website to address the general public, explaining the purpose of the SLHC-PP, the upgrades, and how they are funded.
The WP1 team arranged annual collaboration meetings, at which representatives from each partner institution presented their progress to the whole of the SLHC-PP collaboration. In addition, each WP group had the opportunity to discuss different challenges and approaches, bringing together collaborators from many institutions (milestone reports 1.1 1.2 1.3 and 1.4).
The project management team also served as a bridge between the SLHC-PP collaboration and the European Commission (EC). The meetings gave them a chance to share instructions from the EC about managing finances and preparing reports. Likewise, the collaboration prepared information to send back to the EC - the governing board of the SLHC-PP met to approve the deliverable and milestone reports, which demonstrated the progress of the WPs.
The WP leaders summarised each year's progress and financial information, and WP1 team edited these contributions into annual reports - always submitted to the EC on time (deliverable reports 1.1.1 1.1.2 and 1.1.3).
One of the SLHC-PP's most important roles was facilitating communication between accelerator and detector groups, and also between the two scientific collaborations working around the ATLAS and CMS detectors. Though these collaborations are rivals in the race for exciting discoveries, they face similar sets of problems as they prepare to deal with a luminosity upgrade that will deliver ten times as many collisions to their detectors and produce ten times as much data as at present.
Public events, organised by the project management and held at CERN, aimed at informing as many people as possible about the latest ideas and plans for the LHC upgrade. 'The meeting in 2010 was especially important for explaining the change of strategy resulting from the lessons learnt from the accident in 2008 and the repair in 2009,' says Roland Garoby.
As Garoby mentions, cutting-edge scientific research never goes precisely to plan, and it fell to the WP1 team to explain changes to the schedule and to make alternative proposals when some work packages ran into difficulty.
'The Commission was very open to our proposals for adapting the SLHC-PP Project to the changing LHC context,' says Garoby. He calls it a nice example of good and non-bureaucratic management, demonstrating an 'understanding of what pushing the performance of a state-of-the-art scientific instrument like the LHC is all about'.
WP 2: Accelerator upgrade coordination
Five accelerator sub-projects were underway as of April 2008 in preparation for the upgrade to the LHC. Many of the links in the chain of smaller accelerators that gradually speed up the protons before they are finally injected into the LHC's tunnel were scheduled for upgrades. New links would include Linac4, the Superconducting Proton Linac (SPL) and the Proton Synchrotron 2 (PS2). The penultimate accelerator, the Super Proton Synchrotron (SPS), also needed improvements to handle high luminosity beams. In the LHC itself, the Inner Triplet magnets that sharply focus the proton beam just before it enters each detector cavern were also due for replacement.
In building the LHC, CERN made individual agreements with each institution that provided components and personnel. However, because the LHC upgrade contains five different projects working towards a common goal, with funding provided by several bodies, keeping the efforts integrated requires attention. The WP2 team helped prepare this collaboration and improved communication and management tools.
'WP 2 could draw upon CERN's decades of experience in shaping international collaborations for building and operating particle physics detectors,' says Thomas Otto, coordinator of WP2. Such a collaboration has to assemble accelerator engineers and technicians 'who look for high-tech solutions which are operable 24 hours a day, seven days a week - and which may last half a century', he says.
The WP2 group contributed to drawing up membership procedures for the future accelerator collaborations, and created a schedule and cost plan. They also joined the effort to set up systems for managing finances, monitoring work done by collaborators, and ensuring the quality of contributed parts and installations (milestone report 2.1 deliverable report 2.1.1).
Communication
Previously, the main decision-making on accelerator projects happened at CERN, and because most collaborators worked at CERN, they could frequently meet in person.
In contrast, the collaborators on the LHC upgrade are spread across Europe, Asia and North America. Although they cannot simply walk to a specific CERN conference room, video conferencing through the web allows them to continue meeting regularly. The upgrade groups also assemble for week-long workshops at CERN several times each year (deliverable report 2.2.1). Between meetings, project websites connected with databases form common repositories of technical information, which are accessible worldwide at all times (deliverable report 2.2.2).
Quality assurance system
'The products that we need are quite complicated and depend on a lot of different suppliers and contributors,' says Otto. While comparing the quality standards for different accelerator components is often like comparing apples to oranges, the procedures for acquiring these devices can at least be standardised through a quality assurance (QA) system.
Orders for accelerator components begin with market surveys: private companies are asked to propose how they could produce the parts, and the LHC team then chooses the best offer. Finally, the delivered products must meet certain criteria. These standards were first introduced in the building of the LHC, and they now cover components from collaborating universities and laboratories as well as those from private companies (deliverable report 2.1.2).
Earned value management (EVM)
The WP2 team tailored the EVM software, which successfully managed the LHC construction after 2003, to the needs of the upgrade effort. EVM breaks down a large project into manageable chunks, achievable by one institution within a single year. The costs of material and manpower are fully incorporated into the project schedule as each step is defined in terms of money spent (milestone report 2.2 deliverable report 2.1.3). 'In its revamped version, EVM statistics are available as a handy 'dashboard' within CERN's planning software, avoiding the need to transfer unwieldy data files from one application to another,' says Otto.
With regular communication and close monitoring of progress through the EVM and QA systems, the many facets of the upgrade to the LHC combine in a single well-coordinated effort.
WPs 3 and 4: Coordinating the ATLAS and CMS upgrades
The present CMS and ATLAS collaborations each contain over 2000 people, scattered around the world. The upgrade effort, once it hits its stride, will be just as far-flung. In order to manage the upgrade research, design, construction, and installation, a management team needed to be in place from the beginning. These teams, defined as part of WPs 3 and 4 for ATLAS and CMS respectively, had two primary objectives: coordinating upgrade efforts among the groups in charge of various parts of the detector, and making sure any upgrade work is compatible with the existing detector.
Project management
The upgrade management teams assessed the scope of the upgrades, calculated their costs, and drafted schedules for how they might be completed (milestone reports 3.1 and 4.1 deliverable reports 3.2.2 and 4.2.3). In addition, they initiated the financial planning for the upgrade work. Teams to shoulder these responsibilities were appointed from among the ranks of the existing ATLAS and CMS collaborations (deliverable report 3.1.1 and 4.1.1). An arm of the management known as technical coordination ensures that upgrades are feasible and compatible (deliverable report 4.2.1).
ATLAS headed their upgrade project with an upgrade steering committee while the CMS team is led by an upgrade management board. These groups each drew up global agreements called the initial memoranda of understanding (deliverable reports 3.1.2 and 4.1.2) which are documents to be signed by the collaboration management and the national funding agencies contributing to the upgrade. The funding agencies then provide resources to universities and laboratories that join the ATLAS and CMS upgrade collaborations, allowing them to contribute components or money to support the upgrade.
The management of each experiment defined the decision-making structure, or how upgrades go from an initial idea, to a design, to an accepted project - including detailed specifications, installation procedures, and safety considerations (see WP5 for more on radiation risks). They also decided on reporting procedures (deliverable report 4.2.2 milestone report 3.2).
The CMS and ATLAS upgrade management teams are already making many important decisions about the scope and scheduling of the upgrades. They have assessed the cost of upgrading the detectors, estimated at about EUR 160 million each for ATLAS and CMS. These figures include materials, engineering, and staff effort (deliverable reports 3.1.3 and 4.1.2). The installations will be distributed among several LHC machine shutdowns over the next decade.
ATLAS and CMS each have teams of technical experts that, among other tasks, check for compatibility between the present experiment and proposed upgrades. These teams, called the project office in ATLAS and the review office in CMS, are arms of the technical coordination groups. As the upgrade work integrated into the existing experiments, the project and review offices became increasingly central bridges between the upgrade teams and the existing technical coordination for the two collaborations.
Technical coordination faces a complex working environment. These teams are under significant pressure to remove old components and install new ones without damaging nearby parts of the detector - a difficult task when some subdetector components, such as the pixel detectors, allow only a few millimetres' clearance on each side.
'The technical coordination plays a very important role,' says Steinar Stapnes, leader of WP3. 'In some ways even more important than in the original experiment, as the constraints and complexity of the experimental environment are higher.'
From idea to upgrade
For each proposed upgrade to an existing subdetector component within the ATLAS or CMS detectors, an important step in the process is a simulation to establish that the new device will bring the improvements to detector performance that it promises. At the same time, researchers begin developing the prototypes needed to demonstrate the feasibility of the proposed solutions. All this happens within the group in charge of upgrading that particular subdetector. Once these steps have been cleared, the final idea can be presented to the relevant steering committees or management boards.
At this point, the idea goes through the management of the existing experiment: the project and review offices ensure compatibility with other subdetector components, and the resource management team ensures that funding exists to take the idea further.
If it all checks out, a formal proposal including schedule, delegation of work, and cost is submitted to the LHC committee (LHCC) and resource review board (RRB). These bodies provide oversight for CERN activities. The LHCC is composed of respected experts who review CERN science and technology proposals, selecting the most scientifically promising projects and ensuring that they are sound. Representatives of the funding agencies make up the RRB, and this group oversees the finances.
CMS upgrade project leader Jordan Nash calls the system 'federal' as it gives the group in charge of each subdetector component a lot of latitude to choose its own best upgrades. At the same time, a second tier of management can coordinate the upgrades of each group to make sure they are compatible and serve the broader performance goals of the overall detector.
Mapping the present and future of the detectors
In upgrading ATLAS and CMS, the collaborations are taking state-of-the-art technology and pushing it further. Each detector, which is essentially a digital camera with hundreds of millions of pixels, fills a space half the size of the Notre Dame cathedral in Paris and contains as much as 13 thousand tonnes of material. In order to effectively upgrade the detectors, the collaborations need to keep close tabs on current and future equipment specifications.
To do this, the ATLAS' project office and CMS's technical coordination team have set up central databases, accessible through the Web, containing the designs and locations of each detector component. Dozens of users are already adding the next generation designs for upgraded parts to these databases as they become available, including new designs for sensors, electronics, support, and cooling devices.
Many of the original drawings were made with previous generations of computer aided design programs. These old drawings have been converted to the new standard programs, ensuring that the descriptions and specifications for even the oldest parts of the detectors - developed in the 1990s - would not be lost.
Deputy project leader of the SLHC-PP Duccio Abbaneo says these drawings are crucial. 'If you want to plan an upgrade of a detector that is inaccessible and in a radiation environment, as a starting point you must have a very detailed and accurate model of what you have built, because you can't go there and check.'
The databases aren't just for detector components - they also include equipment for taking apart the present detectors and installing the new components as well as drawings to show how installations could proceed (deliverable reports 3.2.3 and 4.2.2).
WP 5: Radiation safety
Radiation strong enough to damage the LHC machinery deep below ground is an unavoidable side effect of accelerating and colliding particles. At CERN, assessing and managing this risk is a top priority. While the beam is running, physicists need regular access to auxiliary caverns, not far from the LHC tunnel and detector caverns, which contain electronics and cooling pumps. And during downtime, when the beams are no longer circulating and colliding, they need access to the LHC tunnel and the detector caverns in order to carry out any repairs and upgrades. But the detectors and accelerator components continue to emit radiation for some time after the beam has stopped circulating, much like an oven takes time to cool down once it is turned off.
To gauge the increased radiation hazards associated with the luminosity upgrade, the WP5 team has simulated the expected radiation in areas where people are expected to work during beam. It has also assessed the degree to which magnets and detector components will be 'activated', continuing to radiate after the beam is stopped.
Key results:
- safety upgrades were implemented to protect Linac4 construction workers without need of a radiation area;
- upon replacement, collimators and inner triplet magnets will be classified and disposed of as radioactive waste;
- the ATLAS service cavern must be a controlled radiation area while the upgraded LHC is running;
- the shafts delivering power from the surface to an SPL-like machine will need protection against escaping radiation;
- the beam dump of a PS2-like machine can be made safe for access without cool-down periods;
- the upgraded PS will not make significant amounts of air radioactive.
WP 6: Final focusing magnets
Just before the two beams enter the detector caverns from opposite sides, a set of focusing magnets known as the 'Inner Triplets' squeezes them down. By leaving less empty space between the protons in a bunch, these focusing magnets increase the number of protons that actually collide with those from an oncoming bunch.
The inner triplets will suffer significant radiation damage, the accelerator version of wear and tear, over the first decade that the LHC is running. By around 2022, they will need to be replaced. 'If we replace them, we might as well make use of later technology and knowledge we have gained with the LHC project to increase the luminosity,' says Stephan Russenschuck, leader of WP6.
With over a decade's head-start, the WP6 team is already developing new magnets that can squeeze the protons into more tightly packed bunches at the point where they collide inside the detectors.
Key results:
- the suite of final focusing magnets were designed;
- new equipment was developed to make and test the new magnets;
- test coils for the main magnets have been made; these passed electrical tests;
- two corrector magnets are completed, with the last finished by the end of 2011;
- two 2 m-long prototypes of the main magnets are under construction.
WP 7: Towards a new accelerator
If the Superconducting Proton Linac was to be built, it would have followed on from Linac4, accelerating negative hydrogen ions from 160 mega electronvolts (MeV) to 4 giga electronvolts (GeV). As it was designed to accelerate longer, more frequent pulses of ions, the SPL project team needed to design a new negative hydrogen source to be placed at the beginning of Linac4.
These two linear accelerators would endow the beams with almost three times as much energy as the current Linac2 and Proton Synchrotron Booster do, leading to brighter beams entering the new synchrotron, PS2. The improved brightness relates to a peculiarity of Einstein's theory of special relativity. Particles effectively get more massive as they approach the speed of light. Since objects of greater mass are harder to move, the repulsion between the particles - brought on by their negative charges - seems weaker.
In 2007, it appeared affordable to design and build the SPL and PS2, but in 2010, after a detailed analysis and first cost estimate, the CERN management decided against it as part of the baseline upgrade to the LHC. Still, the WP7 team continued with the development of the negative hydrogen source and studies for developing control systems for the accelerator's radiofrequency cavities. This work is expected to benefit the nuclear and particle physics community worldwide in other ways.
'The construction of the SPL remains only as a back-up plan for the LHC,' says SLHC-PP project coordinator Roland Garoby, 'but even technologies developed through the SLHC-PP that are not of direct use in the new plans have important spin-offs'.
Key results:
- a plasma chamber was designed to cope with high driving rates and constructed;
- two more prototypes study different magnet setups;
- the first measurement of optical hydrogen emission lines in a pulsed plasma was made;
- two existing superconducting radiofrequency (RF) cavities were studied in depth, providing information for modeling;
- computer models were developed, showing that it is possible to power two RF cavities with a single amplifier;
- control boards for this system were developed and will be tested.
WP 8: Power delivery at the trackers
The tracking detectors at the centres of the ATLAS and CMS experiments presently contain 86 million and 75 million individual sensor channels, respectively, to trace the paths taken by each particle created in the collisions that occur at the detector's core. The sensor channels are arranged in tens of thousands of modules, and each of these has its own 100 m-long cable to supply it with power. With the upgrade work in the ATLAS and CMS detectors (see WP3 and WP4), the number of sensor channels in the tracking detectors will rise by a factor of ten.
There are three good reasons to find a new strategy to power the three-quarters of a billion sensor channels in each upgraded tracker, detailed in the science and technology section. The most obvious is space: there simply isn't enough room for all those new cables.
The WP8 team has contributed to two new detector-powering schemes: DC-DC conversion and serial powering. The DC-DC conversion scheme allows a small current to run through the long cables at a high voltage, so the cables can be much thinner, making room for additional cables. A DC-DC converter then provides the small voltage and high current needed by the chips on the sensor modules that export data. Serial powering, meanwhile, delivers power to several sensor modules through a single set of cables - again helping to free up space and reduce the power dissipated in the cables.
Key results:
- a DC-DC converter design was selected and radiation resistant control systems were developed;
- the converter board was optimised to avoid interference with the sensors;
- prototype converters were constructed for the future tracking detectors;
- DC-DC converters were integrated with present detector modules;
- four strategies for serially powering detector modules were developed;
- timing signals coming into the detector module and data signals coming out were standardised;
- a scheme was implemented to bypass any problematic module;
- the method was successfully tested in ATLAS tracker assemblies.
Plan for the future
The High Luminosity LHC (HL-LHC) and LHC Injector Upgrade (LIU) projects will implement the upgrade plans developed in part through the SLHC-PP. The present running plans for the LHC are shown in the figure to the left, superimposed on graphs which map the expected accumulation of data. Though the precise dates may change, engineers will make the 'splice repairs' which will allow the LHC to run safely at its full design energy of 7 TeV during the next long shutdown, starting early in 2013. ATLAS has also scheduled its first major upgrade for this time. Around 2017, accelerator engineers will turn their attention to installing new equipment for the upgrade of the accelerators. The CMS collaboration will also begin making major improvements to their detector - some making use of the new powering schemes.
After 2020, both detectors will finalise their upgrades, and new inner triplets will be installed in the LHC ring. The accelerator and detectors should run into the 2030s.
Project results:
Science and technology results
A variety of technologies have been planned, designed, and prototyped through the SLHC-PP, and scientific studies have been undertaken. Results of this work include identifying key detector upgrades, the findings of the radiation studies, the designs and prototyping of the final focusing magnets, the research towards a new accelerator and source, and the development of two new schemes for powering the central subdetectors within the ATLAS and CMS detectors.
The futures of the detectors
Through the management bodies and procedures developed with contributions from the SLHC-PP, the CMS (WP4) and ATLAS (WP3) upgrade teams have identified key improvements needed to make the most of the extra particle collisions in the upgraded LHC.
WP4: Upgrades for CMS
The CMS collaboration will install the first major upgrades during the shutdown scheduled for around 2017. At this time, CMS will receive a new pixel detector - the subdetector component that sits at the core of CMS, closest to the path of the LHC particle beam. CMS collaborators will also replace the photodetectors in the hadronic calorimeter, a subdetector further from the particle beam that measures the energy of particles containing quarks. Then, in a later shutdown after 2020, the entire tracker subdetector will be replaced, and nearly every other system will receive some improvement. Of note, the system for recording the most interesting proton collisions will be upgraded.
Pixels
CMS's silicon pixel subdetector is the highest-resolution detector in CMS, tracking particles fresh from the collisions in three dimensions. When the LHC is operational, 10 million particles/cm2 pass through the pixel subdetector each second. This rate is expected to increase by almost a factor of ten when the LHC is upgraded.
Added sensing layers in the silicon pixel subdetector will provide an additional handle for teasing apart the particle tracks streaming through the detector. An additional layer of pixels will be added, together with extra pixel-filled 'caps' on each end of the cylindrical pixel subdetector.
But the pixel upgrade involves more than adding extra sensors. The CMS pixel group is also cutting out detector dead space - things like cables, cooling, and structural elements. These components don't sense the particles from the collision, but may absorb particles or break up clean, high-energy tracks into a cascade of smaller particles that make interpreting those high-energy tracks more difficult.
Part of the material savings come from new power management schemes, detailed in WP8. Large cables and connectors will be replaced with streamlined versions that take up less space. The carbon fibre support structure for the pixel detector will also be lighter. By using carbon dioxide to cool the detector rather than the present fluorocarbon coolant, the CMS team can use thinner cooling pipes as well as smaller heat exchanger contacts.
This design is scheduled to be implemented in 2016. A third iteration of the pixel subdetector may be installed in the CMS detector in the early 2020s.
Photodetectors
The hadronic calorimeter in CMS is composed of brass or steel 'absorber' tiles, layered with scintillating plastic tiles. The calorimeter tries to absorb all of a particle's energy - forcing it to create cascades of smaller particles as it passes through the brass or steel - and then measures the resulting particles with the scintillating plastic. Scintillators give off light when charged particles pass through, and this light is collected in optical fibres, which take it to a photodetector to be converted to an electrical signal for analysis.
Photodetectors are specially designed photodiodes that can operate in the exceptionally strong (4 Tesla) magnetic field that exists inside the CMS detector. However, the photodetectors currently installed in CMS sometimes spark when exposed to that strong magnetic field: charges build up inside them, and when discharged, the signals looks confusingly like that of a high-energy particle. 'It makes you think something happened and it didn't. You get a big spike of energy that isn't real,' says Nash. The detector may waste time by exporting data from these 'phantom' collisions.
Silicon photomultipliers could avoid the sparking problem. This technology wasn't around when CMS was being designed - they were first developed in Russia in the late 1990s. Silicon photomultipliers operate at a low voltage and are resistant to the charge build-up seen in the photodetectors now deployed in CMS. Moreover, silicon photomultipliers have a signal-to-noise ratio ten times better than the present photodetectors, making the data they produce more easily interpreted.
Finally, the silicon multipliers can be installed in a way that gives information about the rate at which a particle deposits its energy in the calorimeter. In the current design, the particles essentially travel 'up' a tower of alternating scintillator and absorber tiles. The photodetectors have to sum up the total signal from the scintillators in the tower to give the original particle's energy. But the silicon photomultipliers can differentiate between individual scintillator tiles, providing information about how much energy is deposited in each 'floor' of the tower.
Trigger and data acquisition - recording the important particles
Most digital camera users don't print every photo they take, and nor do particle detectors record every collision they see. In fact, the vast majority of collisions - over 99 % - are discarded the moment they are measured. This is because they contain only well established particles and interactions, which are uninteresting in the search for new physics.
The first port of call for data is the trigger, a system which decides which particle collisions to keep and which ones to discard. Presently, the electronics that export data from the detector are connected to the trigger through electrical wires. In order to cope with higher data rates, the CMS team will replace these with an optical system. 'You can send a higher volume of data through an optical fibre,' says Nash - one reason why internet broadband networks around the world increasingly rely on optical fibres rather than traditional copper wires.
By increasing the processing power of the data acquisition system, CMS will be able to look at the finer details of the data coming in. The detector systems currently assess whether a particle has passed through a sensor or not, but they can squeeze more information out of the signal by looking at nuances of the particle's energy and location. To harness the power of the upgraded processors, the CMS team is developing new software.
For more information on the scope of the upgrade see this draft of the plans: over 45 MB.
WP3: Upgrades for ATLAS
The silicon sensors, CMOS chips and controllers within the ATLAS detector, and the optical fibres for exporting data, are all designed to be 'radiation hard' - they can withstand the flux of high-energy charged particles flying out of the proton and lead ion collisions. However, even these durable devices eventually break down after a few years to a decade of exposure, so replacing them will be a necessary part of any upgrade work. At the same time, the electronic systems for sending data out of the detector and choosing interesting events may be improved, streamlining data analysis. The 'end cap' regions of the detector, which close off the barrel-shaped central parts, will also need upgrades to cope with higher radiation rates.
As in CMS, most ATLAS subdetectors are expected to eventually receive some upgrades. While the Insertable B-Layer is the current main project, planned for the shutdown period 2013 - 2014, others are under various stages of planning and consideration.
Insertable B-Layer
The pixel detector contains silicon sensors, each not much larger than a hair's breadth to a side. Eighty million of these, distributed across three cylindrical layers and four end-cap discs, precisely track the paths of particles as they leave the collision point.
The innermost layer of the pixel detector, known as the B-layer, is the first port of call for particle debris leaving the beam pipe and entering the detector, and the concentration of particles endured by this first layer of sensors is higher than anywhere else in the detector. Originally, this layer was to be removed and replaced, but removing the old layer was too risky, and the operation would take too long.
It is easier to instead remove the beam pipe and replace it with a slightly smaller version. This leaves enough space to add a fourth cylindrical layer, nesting inside the current B-layer. ATLAS upgrade project leader Steinar Stapnes says this new pixel layer will 'add a very important early tracking point of particles leaving the interaction region and increase the robustness of the entire pixel system'.
Further upgrades
The ATLAS collaboration is planning several other upgrades, in various stages of study and approval:
- The trigger team, which develops the automated systems for choosing which particle collisions to record, is developing an improved trigger which draws on data from the particle tracking systems. 'It will help select interesting events both faster and more efficiently,' says Stapnes.
- In general, the trigger team is trying to combine information from many parts of ATLAS as early as possible. The trigger runs through a chain of selection criteria to decide whether to keep or discard a proton collision, and combining multiple detectors earlier can allow ATLAS to adjust the selection system to more effectively seek out or ignore specific kinds of collisions.
- The subdetector devoted to muons, electron-like particles that often herald interesting collisions, will need upgrades for its end cap 'wheels'. The revamped LHC will produce muons in such abundance that more than one will regularly come through a sensor at the same time. An extensive research and development effort is looking into the best type of replacement.
- The upgraded LHC will expose the end caps of the calorimeter system, which stops particles and measures their energies, to the maximum barrage of particles that they were designed to withstand. 'Two solutions are being considered, replacing it with version more robust with respect to high rates, or putting another calorimeter in front of it where a significant part of the energy of the incoming particles is deposited,' says Stapnes.
- To keep the proton beams from colliding with air molecules, they are kept inside pipes which have been emptied of air. To make this 'beam pipe' more transparent to particles passing through, ATLAS is looking to replace it with something made of lighter materials. And, so that sensors can be placed as close to the collisions as possible, the upgrade team is also trying to reduce the beam pipe's radius near the centre of the detector.
The largest and most costly upgrade needed is the replacement of the entire ATLAS Inner Detector, expected to happen in a decade's time. 'A very comprehensive research and development program is underway,' says Stapnes. This program of research and development (R&D) will produce new designs for silicon sensors, data export systems that can withstand high levels of radiation, cooling systems, and support structures. The upgrade will comprise both the silicon strip tracker and pixel detector, for which new powering schemes were developed in WP8.
'The planning and detailed research and development for all these upgrades are well underway,' says Stapnes. 'The formal steps needed to move them one by one into realisation will follow the procedures established for the Insertable B-Layer project, and these have been developed with the support of the SLHC-PP WP3 funding and activities.'
WP5: Radiation safety
Higher luminosity beams and collisions have one unfortunate side-effect: they create more radiation. Some of it is along the accelerator chain as protons occasionally stray from the beam, and some of it is in and around the detector caverns, where ten times as many protons are colliding in tiny explosions of new particles. The increase in radiation needs to be predicted ahead of time so that additional safeguards can be implemented to ensure that neither personnel nor the environment will be harmed.
Radiation studies also looked into the radiation risks associated with implementing upgrades, such as the construction of Linac4 and the replacement of the accelerator components nearest the collision point. These studies are ongoing at CERN, and will eventually cover all radiation concerns connected to the upgrades.
Originally, WP5 intended to study six areas: the point where the Linac4 construction site approaches the running Linac2 accelerator, the radiation that might escape the hypothetical SPL and PS2 accelerators, the radiation that enters the service caverns of ATLAS and CMS, and the activation of the Inner Triplet magnets, which are slated for replacement after 2020.
However, because the PS2 and SPL are no longer part of the luminosity upgrade at the LHC, in 2010 WP5 instead began to consider how much radiation is likely to be emitted by the older accelerators running at higher luminosity. Specifically, focus turned to the Proton Synchrotron (PS), which WP5 project leader Thomas Otto calls, 'the more than 50 year old centrepiece of CERN's accelerator chain.'
A good starting point for radiation safety is the law: anyone expected to come into contact with ionising radiation through their work must not be exposed to more than a 20 milliSievert (mSv) radiation dose per year. This applies to personnel at hospitals or nuclear power plants, and to those at CERN working near the running accelerators and particle detectors. For comparison, the French Institute of Radioprotection and Nuclear Safety estimates that the annual dose for flight personnel is 5 mSv on the routes with the highest exposure to radiation from space, for example New York to Tokyo or Paris to Tokyo.
CERN then takes radiation safety further, following the ALARA principle, which ensures that radiation doses are as low as reasonably achievable. The upshot: it's preferable to avoid exposing personnel to radiation by shielding even those areas that pose a very low risk of contributing significantly to an annual radiation dose.
One of the first steps for the WP5 team was to identify the most important factors that contribute to radiation risk. In accessible areas while beam is running, these are: the energy of the beam, the number of particles accelerated with each pulse, and the average amount of beam passing through an area at any given time. After beam has stopped, activated materials can also emit radiation.
The team also identified the locations in the accelerator chain where particles from the beam were likely to escape and pose a radiation risk. One such point is at the collimators, blocks of heavy metal (often tungsten) that strip straying particles off the beam by passing it through a relatively narrow hole, which are spaced at regular intervals around the accelerator.
The beam is less stable at the points where it enters and leaves accelerators, so these areas are also more likely to be exposed to escaping particles. Finally, beam dumps - massive blocks of tungsten, iron, concrete, and other materials designed to absorb the beam when it is no longer of use - are liable to become activated (milestone report 5.1).
Once the WP5 team had identified the risk factors, they assessed the radiation risks on a project-by-project basis.
Activation
When charged particles run into atomic nuclei, they can change the identity of a nucleus, making it into a radioactive isotope. These atomic nuclei have unusual neutron numbers for the element, and the unstable forms eventually decay by emitting other particles. Those with short half-lives will disappear quickly, settling back to a non-radioactive isotope after emitting an energetic photon, known as a 'gamma ray', and a particle - most often an electron or its antimatter partner, the positron. However, the longer-lived isotopes are more problematic, and can require equipment to be stored for tens to hundreds of years underground or behind thick walls of concrete.
Linac4
The new Linac4, currently under construction at CERN, will need to feed into the PSB through the same injection line used at present by Linac2. The construction workers digging the new tunnel were present on the site from December 2008 until October 2010, overlapping with the LHC's first run at the end of 2009 and its second from March to December 2010.
The question for WP5 was: how can CERN ensure that the construction workers excavating the earth and building the housing for Linac4 will not be exposed to radiation from the running Linac2?
Initially, CERN planned to mitigate the risk by building the link between Linac4 and Linac 2 in winter, when the LHC was not running. But the civil engineers quickly pointed out that digging in frozen ground is impractical. Building work would have to occur during the warmer months, with the LHC operational. At the onset of excavations, the workers were definitely safe - 20 m of earth separated them from Linac2.
However, the construction workers needed to dig right up to the side of the Linac2 building, which meant that as the excavation progressed, eventually just two 2.5 m-thick concrete walls separated them from Linac2. Only the last metre of excavation was likely to pose any significant risk, so the WP5 team simulated the radiation levels likely to come through the two walls.
The simulation showed that without additional shielding, radiation levels would be higher than the 0.5 µSv/h limit for sites accessible by the general public. To bring the radiation down to acceptable levels, CERN added a third concrete wall, 40 cm thick.
With this addition in place, Otto says: 'A number of calculations with the same Monte Carlo code were done in order to demonstrate that the radiation levels would be so low that there is never the need of defining a limited access radiation area'. The excavation workers were at such low risk that they were not required to wear dosimeters to measure radiation levels.
The WP5 team also added a radiation monitor inside the Linac2 tunnel, which can shut down the accelerator if the radiation level there gets too high. The sensor was set for a limit 25 % lower than that prescribed by the simulations in order to provide an ample safety margin. These precautions ensured that the radiation dose received by the construction workers was well under 1 mSv per year, the legal dose limit for non-radiation workers - a success for the WP5 team.
The CERN technicians and engineers who will install the new accelerator in the finished Linac4 tunnel will spend more time near the wall approaching Linac2, and so will still be required to wear dosimeters - but for these workers, such precautions are a standard aspect of their daily work (deliverable report 5.2.1).
Simulations
To forecast the amount of radiation in various parts of the accelerator complex, the WP5 team modelled the effects that the beam and collisions would have on equipment and work spaces. The code used to simulate the radiation and activation of materials, FLUKA 2008, is a type of Monte Carlo code made for the purpose of estimating the effects of ionising radiation and how it moves through materials. The name Monte Carlo comes from gambling algorithms, which attach probabilities to a number of outcomes.
In the FLUKA code, for example, researchers can evaluate the probability that a negative hydrogen ion collides with a nucleus after travelling a certain distance through a magnet. Provided a collision occurs, probability also rules the fate of the outcome: if the starting nucleus is copper it may become radioactive cobalt after the collision, through the removal of a few protons and neutrons. The code runs through these scenarios for millions of particles, randomly selecting the outcome of each step along the way in accordance with the probabilities set through experiments and theoretical models. It simulates the effect of years of radiation exposure.
Inner triplets
The sets of magnets which provide the final focusing of the LHC beams before they enter one of the detectors are on the front line for absorbing the shrapnel of the particle collisions that occur inside the detectors. When high energy particles strike the magnets, they can turn ordinary atoms into radioactive isotopes. For a while, these isotopes act like tiny time-bombs, ready to release radiation when they decay into ordinary atoms again.
Provided that the level of radioactivity in materials becomes negligible within 30 years, CERN will be able to store them and then dispose of them as ordinary waste. However, WP5 showed that the inner triplets will stay radioactive for a longer period.
Ideally, these magnets could be reused, but workers can't afford to spend long hours carefully dismantling them inside the tunnels, where the radiation levels are likely to be high. With the right strategy for disconnecting the magnets, though, Paolo Fessia of CERN's technology department reckons that half of them could be saved and used as spares for the ALICE and LHCb experiments. The rest of the material will be scrapped at a special site for radioactive materials.
The WP5 team recommends that considering how to dismantle the radioactive accelerator parts should become a consideration in the designs of future triplet magnets. Components should fit together 'like bricks in a Lego game' rather than being bolted to one another, says Otto. Alternatively, robots could be used to take the magnets apart remotely, ensuring worker safety no matter how long the dismantling procedure (deliverable report 5.2.1).
Service caverns and surface buildings
A quirk of the local geology required CMS to leave a 6 m-wide pillar of rock between the experimental cavern and the service cavern. The corridor between the two winds around it. This natural shielding is much better than that installed around the ATLAS experiment. The ATLAS service cavern, which contains electronics and cooling equipment, is protected by just 2 m of concrete.
Presently, while the LHC is running, the dose rate inside the ATLAS service cavern is 2-5 µSv/h. With ambient dose equivalents ranging between 0.5 and 15 µSv/h at the highest, the area can be classified as a supervised radiation area. This means that only those certified as radiation workers have access, but they may take their time accomplishing their tasks as it is impossible to exceed annual dose limits with the quoted radiation levels.
When the LHC is upgraded for high luminosity, the dose rates should increase by ten times, reaching 20-50 µSv/h. Because there is no space to add extra shielding, the cavern will have to become a controlled radiation area, which means that the personnel who enter will need special training and closer monitoring. At that point, their work in the service caverns will be subject to time restrictions because in the absence of additional shielding, working more rapidly is the only option to minimise the radiation dose. The workers are unlikely ever to approach the legal limit for exposure of 20 mSv per year, but it is worthwhile to train more people to repair or replace components in the electronics, cryogenics, and other services.
At the surface, the two shafts leading down to the ATLAS experimental hall are each covered with a 1 m-thick 'plug' made of concrete. The narrower of the shafts emits 1 µSv/h while the wider emits 3 µSv/h. An additional metre of concrete - a little more for the wider shaft - will be added to the plugs after the upgrade. The extra shielding will be enough to keep the radiation rates below 2.5 µSv/h, which is the limit for a public area that is not permanently occupied. However, safely setting that much concrete over a hole up to 18 m in diameter is not an easy task, and engineering work will be needed to design the structures that will hold and move the concrete (deliverable report 5.1.2).
Superconducting proton Linac (SPL)
The SPL would be built 20-30 m underground, allowing the soil to provide effective shielding against any lost negative hydrogen ions from the beam, which will have energies of up to 5 GeV. The accelerator needs about 16 shafts, each 2.8 m wide, to run bundles of cable-like RF waveguides from the klystron RF amplifiers at the surface to the RF cavities that accelerate the beam. 'These vertical shafts are also a way out for neutrons which are produced in the accelerator,' says Otto. Neutrons may be generated if stray negative hydrogen ions hit an RF cavity or other equipment.
Simulations of beam lost in the tunnel revealed that more work is necessary to minimise the radiation exposure inside the klystron buildings. Since technicians need access to the klystron building while the accelerator runs, how to protect them from radiation leaking up from below will be a concern for the future developers of SPL-like machines, such as the team working on the European Spallation Source (ESS) in Lund, Sweden (deliverable report 5.2.1).
Proton synchrotron 2 (PS2)
The WP5 team also simulated the radiation risk expected for the beam dump near the PS2, before plans to build the PS2 were shelved. When negative hydrogen from the long, straight SPL entered the oblong ring of the PS2, it would have run through a thin piece of carbon which strips away the two electrons, leaving the bare proton. The PS2 would then alter the trajectory of these protons to travel around its accelerator ring. However, in some cases only one or neither of the electrons would be removed, and the hydrogen would not make it into the PS2 ring. The negative hydrogen would bend the wrong way in the magnetic field, and the neutral hydrogen would go straight.
To catch these wayward particles, accelerator physicists would need to add a beam absorber. Ideally, the absorber would be able to catch both the negative and neutral hydrogen. To capture the diverging streams of particles in a single, relatively small absorber, the beam dump would have to lie close to the accelerator.
The dump near the PS2 would need to catch about 6.4 x 1019 particles per year, with each particle at an energy of 4 GeV. It would have a design reminiscent of the 'layers of an onion', says Otto. A carbon core would be enveloped by aluminium and then tungsten. This would then be surrounded by a block of iron, 1 m high and 3.2 m long. Finally, the entire structure would be encased in a 20 cm layer of concrete.
If this was to go into the PS2 accelerator vault, it would need to give off less than 50 µSv/h per hour so as not to pose a radiation risk to the workers making repairs to the accelerator during shutdowns. The beam dump would breach this limit only at the opening where the beam enters the block of absorbers. A simple mechanism to block that beam entrance would be enough to keep the radiation level sufficiently low (deliverable report 5.2.1).
Although the PS2 is no longer planned for construction, this study of the PS2 beam dump can serve the ESS in Lund, as the accelerator there will employ an identical injection scheme, requiring similar equipment.
Validating the simulations
Simulations are a valuable tool for forecasting the activation of materials at the LHC and, eventually, the radiation emitted by the LHC after upgrade work has been completed. However, to rely on this information for safety purposes, WP5 needs to give evidence that the simulations are accurate. 'The best way of proving that you are not making a big error somewhere is to conduct some experiments, some validation,' says Otto.
One way to do this is to see whether the simulations give accurate predictions about the LHC. Before the accelerator started up in 2008, WP5 placed hundreds of radiation detectors in the LHC tunnel, the ATLAS and CMS experimental halls, and the service caverns. Most of these are passive detectors, which record the total radiation to which they are exposed, to be assessed after collection. A few send out a live feed of information. One kind, the MPX detectors placed in the hottest positions of ATLAS experiment, have been already able to register a short-term activation induced during some relatively high luminosity proton collisions in the LHC.
The plan was to collect the passive detectors after the LHC's first year of running to make sure that the simulations were accurate. Unfortunately, the first year of LHC beam came later than expected. At the end of 2010, the Paul Scherrer Institut (PSI) team removed and evaluated their neutron detectors, indicating radiation levels in the same order of magnitude as expected for the luminosity accumulated so far. Another set of detectors, placed by different collaborators coordinated by the CERN radiation protection group, is still underground, waiting until the end of 2011 to be removed.
These detectors will eventually give vital feedback about the amount of radiation that will reach accessible areas during beam as well as the amount of activation that materials will pick up (deliverable report 5.1.1).
Impact study
The WP5 team also needed to explore ways that radiation may leave CERN and possibly impact the environment. There are three ways this could happen: through the release of radioactive air or water, and through the inappropriate disposal of radioactive waste.
The inner triplet magnets are set for replacement, as are the collimators. Because simulations show that both types of equipment will be significant sources of radiation for over 30 years, they will need to be sent to a special site for radioactive waste storage and disposal.
During the operation of an accelerator, it is also possible for the oxygen, nitrogen, and other components of ordinary air to become radioactive, and so the WP5 team studied this activation process. Of special concern is the chance that the 52-year-old PS accelerator may activate the air inside the tunnel and then release it into the environment.
Even though the PS will handle beam intensities 1000 times higher than it was originally designed for, the simulations of WP5 show that the release of radioactive air should be 100 times smaller than the legal limit. Nevertheless, the ventilation system will be renovated to control airflow from the PS and measure the actual amount of radioactivity released through the air.
On the other hand, the simulations did reveal weak points in the PS shielding that will need to be reinforced with concrete or iron to reduce the likelihood of stray radiation escaping into the CERN campus. This is hardly surprising - although scarcely any part of the actual accelerator is original, its circular tunnel and earthen hill of shielding date back to the late 1950s.
CERN is generally fortunate in its location when it comes to water risks - the tunnels are far from large stores of ground water. However, rainwater may percolate through the ground near the accelerators and so pick up radiation. Water which reaches the present accelerator tunnels at CERN is not activated beyond the levels of natural radioactivity. Even so, Otto says: 'In the vicinity of future beam dumps, built for higher beam intensities, a sensible precaution would be to separate the evacuation paths of water ingress from the ordinary gutters. In the case of a slight elevation of radiation levels in the water, this would allow CERN to treat it separately.'
As a result of the radiation studies in WP5, CERN has taken steps to reduce the radiation exposure of personnel, the upgrade planners know how much the radiation from the LHC and its accelerator chain will increase, and safety officers have recommendations for how to mitigate this future risk.
Information about the WP6, WP7 and WP8 and their expected results can be found in the documents attached.
Proton collisions are all about creating new particles. Albert Einstein showed that energy and mass were interchangeable currencies, and the Large Hadron Collider (LHC) is a bureau de change of sorts. The CERN chain of particle accelerators packs dense beams of protons into two beam-pipes of the LHC, beneath the Swiss-French border. Travelling in opposite directions through the 27 km-long circular tunnel, the protons receive their final acceleration up to tera electronvolt (TeV) energies (3.5 TeV today, and 7 TeV after 2013).
The LHC then collides the protons together at the hearts of four particle detectors - ATLAS and CMS being the largest - that intersect the circular tunnel. When these collisions occur, the energy in the protons transforms into mass, generating a shower of new particles. The higher the energy of the collision, the heavier the particles that form in the debris. And they aren't just any particles - occasionally, the LHC will produce objects and interactions that have never before been seen by physicists, providing vital experimental evidence to back up the current theories of why the Universe is the way it is.
Though particle physics studies the very small, new observations could reveal the explanations for cosmic mysteries - what gives objects their mass, the nature of the 'invisible' matter that makes up most of the Universe, how the four forces that govern everyone and everything might unite into a super-force, and - maybe - why gravity is so much weaker than the other three forces.
The LHC is already the most powerful particle collider in the world, but for it to live up to its true potential, it must be upgraded to run at the highest possible luminosity. If the LHC is imagined as a torch probing the darkest recesses of the Universe, it's easy to see why a brighter beam would be more useful. A torch is made brighter by increasing concentration of photons in the beam, but the LHC beam will be brightened by making the proton bunches within it even denser.
Most of the proton collisions in the LHC will produce debris that is already familiar to physicists. The large or unusual particles they are most interested in are diamonds buried in heaps of common rubble. But while diamonds tend to stick around, the new particles are often extremely short lived: they disintegrate in fractions of a billionth of a second. To catch a glimpse of the fleeting particles, researchers take images of them through the four major detectors - recording particle 'signatures'.
Physicists rely on a complicated statistical process to separate the rare signatures of new physics from the huge number of similar signatures already expected to show up due to known physics. The larger the quantity of data physicists have to search through, the greater the chance of finding and confirming the signatures of rare particles.
The LHC upgrade will extend the lifetime of this scientific exploration into the unknown and produce ten times more data during the collider's second decade than during its first.
The main aim of the SLHC-PP was to prepare the first phase of this upgrade project for a formal proposal in 2011. Researchers needed to demonstrate new technologies, find solutions for radiation safety issues, form the collaborations that would carry out the work, and prepare the necessary management tools.
The SLHC-PP explored key upgrade pathways, demonstrating solutions to better focus the beams, design key components for the beam injectors, and power detectors more efficiently. A dedicated SLHC-PP team addressed radiation safety concerns by analysing the immediate and long-term effects of beam loss, in view of designing means to ensure safety and minimise the impact on the environment. On the collaboration side, the SLHC-PP contributed to the efforts in planning and initiating the upgrade projects of the ATLAS and CMS detectors. With the support of the SLHC-PP, existing tools for financial management, monitoring progress, and quality assurance have been improved, using the experience gained during the construction of the LHC.
The SLHC-PP activities already represent work from 18 institutions in 10 countries, but they also join with larger efforts coordinated through the detector collaborations and CERN's accelerator departments.
Work package (WP)1: Project management
The multifaceted SLHC-PP effort was managed by the WP1 team - comprising the project coordinator, the deputy project coordinator and the administrative manager. Roland Garoby and Duccio Abbaneo succeeded Lyndon Evans and Lucie Linssen as project coordinator and deputy, respectively, at the beginning of 2010. Mar Capeans has been the administrative manager throughout the project. The members of WP1 handled contracts and finances, monitored and reported progress in the other work packages, and took care of disseminating information both within the SLHC-PP and outside the collaboration.
In the first year, the management team set up a collaboration website (deliverable report 1.2.1) much of which is available for anyone to view. They also commissioned a website to address the general public, explaining the purpose of the SLHC-PP, the upgrades, and how they are funded.
The WP1 team arranged annual collaboration meetings, at which representatives from each partner institution presented their progress to the whole of the SLHC-PP collaboration. In addition, each WP group had the opportunity to discuss different challenges and approaches, bringing together collaborators from many institutions (milestone reports 1.1 1.2 1.3 and 1.4).
The project management team also served as a bridge between the SLHC-PP collaboration and the European Commission (EC). The meetings gave them a chance to share instructions from the EC about managing finances and preparing reports. Likewise, the collaboration prepared information to send back to the EC - the governing board of the SLHC-PP met to approve the deliverable and milestone reports, which demonstrated the progress of the WPs.
The WP leaders summarised each year's progress and financial information, and WP1 team edited these contributions into annual reports - always submitted to the EC on time (deliverable reports 1.1.1 1.1.2 and 1.1.3).
One of the SLHC-PP's most important roles was facilitating communication between accelerator and detector groups, and also between the two scientific collaborations working around the ATLAS and CMS detectors. Though these collaborations are rivals in the race for exciting discoveries, they face similar sets of problems as they prepare to deal with a luminosity upgrade that will deliver ten times as many collisions to their detectors and produce ten times as much data as at present.
Public events, organised by the project management and held at CERN, aimed at informing as many people as possible about the latest ideas and plans for the LHC upgrade. 'The meeting in 2010 was especially important for explaining the change of strategy resulting from the lessons learnt from the accident in 2008 and the repair in 2009,' says Roland Garoby.
As Garoby mentions, cutting-edge scientific research never goes precisely to plan, and it fell to the WP1 team to explain changes to the schedule and to make alternative proposals when some work packages ran into difficulty.
'The Commission was very open to our proposals for adapting the SLHC-PP Project to the changing LHC context,' says Garoby. He calls it a nice example of good and non-bureaucratic management, demonstrating an 'understanding of what pushing the performance of a state-of-the-art scientific instrument like the LHC is all about'.
WP 2: Accelerator upgrade coordination
Five accelerator sub-projects were underway as of April 2008 in preparation for the upgrade to the LHC. Many of the links in the chain of smaller accelerators that gradually speed up the protons before they are finally injected into the LHC's tunnel were scheduled for upgrades. New links would include Linac4, the Superconducting Proton Linac (SPL) and the Proton Synchrotron 2 (PS2). The penultimate accelerator, the Super Proton Synchrotron (SPS), also needed improvements to handle high luminosity beams. In the LHC itself, the Inner Triplet magnets that sharply focus the proton beam just before it enters each detector cavern were also due for replacement.
In building the LHC, CERN made individual agreements with each institution that provided components and personnel. However, because the LHC upgrade contains five different projects working towards a common goal, with funding provided by several bodies, keeping the efforts integrated requires attention. The WP2 team helped prepare this collaboration and improved communication and management tools.
'WP 2 could draw upon CERN's decades of experience in shaping international collaborations for building and operating particle physics detectors,' says Thomas Otto, coordinator of WP2. Such a collaboration has to assemble accelerator engineers and technicians 'who look for high-tech solutions which are operable 24 hours a day, seven days a week - and which may last half a century', he says.
The WP2 group contributed to drawing up membership procedures for the future accelerator collaborations, and created a schedule and cost plan. They also joined the effort to set up systems for managing finances, monitoring work done by collaborators, and ensuring the quality of contributed parts and installations (milestone report 2.1 deliverable report 2.1.1).
Communication
Previously, the main decision-making on accelerator projects happened at CERN, and because most collaborators worked at CERN, they could frequently meet in person.
In contrast, the collaborators on the LHC upgrade are spread across Europe, Asia and North America. Although they cannot simply walk to a specific CERN conference room, video conferencing through the web allows them to continue meeting regularly. The upgrade groups also assemble for week-long workshops at CERN several times each year (deliverable report 2.2.1). Between meetings, project websites connected with databases form common repositories of technical information, which are accessible worldwide at all times (deliverable report 2.2.2).
Quality assurance system
'The products that we need are quite complicated and depend on a lot of different suppliers and contributors,' says Otto. While comparing the quality standards for different accelerator components is often like comparing apples to oranges, the procedures for acquiring these devices can at least be standardised through a quality assurance (QA) system.
Orders for accelerator components begin with market surveys: private companies are asked to propose how they could produce the parts, and the LHC team then chooses the best offer. Finally, the delivered products must meet certain criteria. These standards were first introduced in the building of the LHC, and they now cover components from collaborating universities and laboratories as well as those from private companies (deliverable report 2.1.2).
Earned value management (EVM)
The WP2 team tailored the EVM software, which successfully managed the LHC construction after 2003, to the needs of the upgrade effort. EVM breaks down a large project into manageable chunks, achievable by one institution within a single year. The costs of material and manpower are fully incorporated into the project schedule as each step is defined in terms of money spent (milestone report 2.2 deliverable report 2.1.3). 'In its revamped version, EVM statistics are available as a handy 'dashboard' within CERN's planning software, avoiding the need to transfer unwieldy data files from one application to another,' says Otto.
With regular communication and close monitoring of progress through the EVM and QA systems, the many facets of the upgrade to the LHC combine in a single well-coordinated effort.
WPs 3 and 4: Coordinating the ATLAS and CMS upgrades
The present CMS and ATLAS collaborations each contain over 2000 people, scattered around the world. The upgrade effort, once it hits its stride, will be just as far-flung. In order to manage the upgrade research, design, construction, and installation, a management team needed to be in place from the beginning. These teams, defined as part of WPs 3 and 4 for ATLAS and CMS respectively, had two primary objectives: coordinating upgrade efforts among the groups in charge of various parts of the detector, and making sure any upgrade work is compatible with the existing detector.
Project management
The upgrade management teams assessed the scope of the upgrades, calculated their costs, and drafted schedules for how they might be completed (milestone reports 3.1 and 4.1 deliverable reports 3.2.2 and 4.2.3). In addition, they initiated the financial planning for the upgrade work. Teams to shoulder these responsibilities were appointed from among the ranks of the existing ATLAS and CMS collaborations (deliverable report 3.1.1 and 4.1.1). An arm of the management known as technical coordination ensures that upgrades are feasible and compatible (deliverable report 4.2.1).
ATLAS headed their upgrade project with an upgrade steering committee while the CMS team is led by an upgrade management board. These groups each drew up global agreements called the initial memoranda of understanding (deliverable reports 3.1.2 and 4.1.2) which are documents to be signed by the collaboration management and the national funding agencies contributing to the upgrade. The funding agencies then provide resources to universities and laboratories that join the ATLAS and CMS upgrade collaborations, allowing them to contribute components or money to support the upgrade.
The management of each experiment defined the decision-making structure, or how upgrades go from an initial idea, to a design, to an accepted project - including detailed specifications, installation procedures, and safety considerations (see WP5 for more on radiation risks). They also decided on reporting procedures (deliverable report 4.2.2 milestone report 3.2).
The CMS and ATLAS upgrade management teams are already making many important decisions about the scope and scheduling of the upgrades. They have assessed the cost of upgrading the detectors, estimated at about EUR 160 million each for ATLAS and CMS. These figures include materials, engineering, and staff effort (deliverable reports 3.1.3 and 4.1.2). The installations will be distributed among several LHC machine shutdowns over the next decade.
ATLAS and CMS each have teams of technical experts that, among other tasks, check for compatibility between the present experiment and proposed upgrades. These teams, called the project office in ATLAS and the review office in CMS, are arms of the technical coordination groups. As the upgrade work integrated into the existing experiments, the project and review offices became increasingly central bridges between the upgrade teams and the existing technical coordination for the two collaborations.
Technical coordination faces a complex working environment. These teams are under significant pressure to remove old components and install new ones without damaging nearby parts of the detector - a difficult task when some subdetector components, such as the pixel detectors, allow only a few millimetres' clearance on each side.
'The technical coordination plays a very important role,' says Steinar Stapnes, leader of WP3. 'In some ways even more important than in the original experiment, as the constraints and complexity of the experimental environment are higher.'
From idea to upgrade
For each proposed upgrade to an existing subdetector component within the ATLAS or CMS detectors, an important step in the process is a simulation to establish that the new device will bring the improvements to detector performance that it promises. At the same time, researchers begin developing the prototypes needed to demonstrate the feasibility of the proposed solutions. All this happens within the group in charge of upgrading that particular subdetector. Once these steps have been cleared, the final idea can be presented to the relevant steering committees or management boards.
At this point, the idea goes through the management of the existing experiment: the project and review offices ensure compatibility with other subdetector components, and the resource management team ensures that funding exists to take the idea further.
If it all checks out, a formal proposal including schedule, delegation of work, and cost is submitted to the LHC committee (LHCC) and resource review board (RRB). These bodies provide oversight for CERN activities. The LHCC is composed of respected experts who review CERN science and technology proposals, selecting the most scientifically promising projects and ensuring that they are sound. Representatives of the funding agencies make up the RRB, and this group oversees the finances.
CMS upgrade project leader Jordan Nash calls the system 'federal' as it gives the group in charge of each subdetector component a lot of latitude to choose its own best upgrades. At the same time, a second tier of management can coordinate the upgrades of each group to make sure they are compatible and serve the broader performance goals of the overall detector.
Mapping the present and future of the detectors
In upgrading ATLAS and CMS, the collaborations are taking state-of-the-art technology and pushing it further. Each detector, which is essentially a digital camera with hundreds of millions of pixels, fills a space half the size of the Notre Dame cathedral in Paris and contains as much as 13 thousand tonnes of material. In order to effectively upgrade the detectors, the collaborations need to keep close tabs on current and future equipment specifications.
To do this, the ATLAS' project office and CMS's technical coordination team have set up central databases, accessible through the Web, containing the designs and locations of each detector component. Dozens of users are already adding the next generation designs for upgraded parts to these databases as they become available, including new designs for sensors, electronics, support, and cooling devices.
Many of the original drawings were made with previous generations of computer aided design programs. These old drawings have been converted to the new standard programs, ensuring that the descriptions and specifications for even the oldest parts of the detectors - developed in the 1990s - would not be lost.
Deputy project leader of the SLHC-PP Duccio Abbaneo says these drawings are crucial. 'If you want to plan an upgrade of a detector that is inaccessible and in a radiation environment, as a starting point you must have a very detailed and accurate model of what you have built, because you can't go there and check.'
The databases aren't just for detector components - they also include equipment for taking apart the present detectors and installing the new components as well as drawings to show how installations could proceed (deliverable reports 3.2.3 and 4.2.2).
WP 5: Radiation safety
Radiation strong enough to damage the LHC machinery deep below ground is an unavoidable side effect of accelerating and colliding particles. At CERN, assessing and managing this risk is a top priority. While the beam is running, physicists need regular access to auxiliary caverns, not far from the LHC tunnel and detector caverns, which contain electronics and cooling pumps. And during downtime, when the beams are no longer circulating and colliding, they need access to the LHC tunnel and the detector caverns in order to carry out any repairs and upgrades. But the detectors and accelerator components continue to emit radiation for some time after the beam has stopped circulating, much like an oven takes time to cool down once it is turned off.
To gauge the increased radiation hazards associated with the luminosity upgrade, the WP5 team has simulated the expected radiation in areas where people are expected to work during beam. It has also assessed the degree to which magnets and detector components will be 'activated', continuing to radiate after the beam is stopped.
Key results:
- safety upgrades were implemented to protect Linac4 construction workers without need of a radiation area;
- upon replacement, collimators and inner triplet magnets will be classified and disposed of as radioactive waste;
- the ATLAS service cavern must be a controlled radiation area while the upgraded LHC is running;
- the shafts delivering power from the surface to an SPL-like machine will need protection against escaping radiation;
- the beam dump of a PS2-like machine can be made safe for access without cool-down periods;
- the upgraded PS will not make significant amounts of air radioactive.
WP 6: Final focusing magnets
Just before the two beams enter the detector caverns from opposite sides, a set of focusing magnets known as the 'Inner Triplets' squeezes them down. By leaving less empty space between the protons in a bunch, these focusing magnets increase the number of protons that actually collide with those from an oncoming bunch.
The inner triplets will suffer significant radiation damage, the accelerator version of wear and tear, over the first decade that the LHC is running. By around 2022, they will need to be replaced. 'If we replace them, we might as well make use of later technology and knowledge we have gained with the LHC project to increase the luminosity,' says Stephan Russenschuck, leader of WP6.
With over a decade's head-start, the WP6 team is already developing new magnets that can squeeze the protons into more tightly packed bunches at the point where they collide inside the detectors.
Key results:
- the suite of final focusing magnets were designed;
- new equipment was developed to make and test the new magnets;
- test coils for the main magnets have been made; these passed electrical tests;
- two corrector magnets are completed, with the last finished by the end of 2011;
- two 2 m-long prototypes of the main magnets are under construction.
WP 7: Towards a new accelerator
If the Superconducting Proton Linac was to be built, it would have followed on from Linac4, accelerating negative hydrogen ions from 160 mega electronvolts (MeV) to 4 giga electronvolts (GeV). As it was designed to accelerate longer, more frequent pulses of ions, the SPL project team needed to design a new negative hydrogen source to be placed at the beginning of Linac4.
These two linear accelerators would endow the beams with almost three times as much energy as the current Linac2 and Proton Synchrotron Booster do, leading to brighter beams entering the new synchrotron, PS2. The improved brightness relates to a peculiarity of Einstein's theory of special relativity. Particles effectively get more massive as they approach the speed of light. Since objects of greater mass are harder to move, the repulsion between the particles - brought on by their negative charges - seems weaker.
In 2007, it appeared affordable to design and build the SPL and PS2, but in 2010, after a detailed analysis and first cost estimate, the CERN management decided against it as part of the baseline upgrade to the LHC. Still, the WP7 team continued with the development of the negative hydrogen source and studies for developing control systems for the accelerator's radiofrequency cavities. This work is expected to benefit the nuclear and particle physics community worldwide in other ways.
'The construction of the SPL remains only as a back-up plan for the LHC,' says SLHC-PP project coordinator Roland Garoby, 'but even technologies developed through the SLHC-PP that are not of direct use in the new plans have important spin-offs'.
Key results:
- a plasma chamber was designed to cope with high driving rates and constructed;
- two more prototypes study different magnet setups;
- the first measurement of optical hydrogen emission lines in a pulsed plasma was made;
- two existing superconducting radiofrequency (RF) cavities were studied in depth, providing information for modeling;
- computer models were developed, showing that it is possible to power two RF cavities with a single amplifier;
- control boards for this system were developed and will be tested.
WP 8: Power delivery at the trackers
The tracking detectors at the centres of the ATLAS and CMS experiments presently contain 86 million and 75 million individual sensor channels, respectively, to trace the paths taken by each particle created in the collisions that occur at the detector's core. The sensor channels are arranged in tens of thousands of modules, and each of these has its own 100 m-long cable to supply it with power. With the upgrade work in the ATLAS and CMS detectors (see WP3 and WP4), the number of sensor channels in the tracking detectors will rise by a factor of ten.
There are three good reasons to find a new strategy to power the three-quarters of a billion sensor channels in each upgraded tracker, detailed in the science and technology section. The most obvious is space: there simply isn't enough room for all those new cables.
The WP8 team has contributed to two new detector-powering schemes: DC-DC conversion and serial powering. The DC-DC conversion scheme allows a small current to run through the long cables at a high voltage, so the cables can be much thinner, making room for additional cables. A DC-DC converter then provides the small voltage and high current needed by the chips on the sensor modules that export data. Serial powering, meanwhile, delivers power to several sensor modules through a single set of cables - again helping to free up space and reduce the power dissipated in the cables.
Key results:
- a DC-DC converter design was selected and radiation resistant control systems were developed;
- the converter board was optimised to avoid interference with the sensors;
- prototype converters were constructed for the future tracking detectors;
- DC-DC converters were integrated with present detector modules;
- four strategies for serially powering detector modules were developed;
- timing signals coming into the detector module and data signals coming out were standardised;
- a scheme was implemented to bypass any problematic module;
- the method was successfully tested in ATLAS tracker assemblies.
Plan for the future
The High Luminosity LHC (HL-LHC) and LHC Injector Upgrade (LIU) projects will implement the upgrade plans developed in part through the SLHC-PP. The present running plans for the LHC are shown in the figure to the left, superimposed on graphs which map the expected accumulation of data. Though the precise dates may change, engineers will make the 'splice repairs' which will allow the LHC to run safely at its full design energy of 7 TeV during the next long shutdown, starting early in 2013. ATLAS has also scheduled its first major upgrade for this time. Around 2017, accelerator engineers will turn their attention to installing new equipment for the upgrade of the accelerators. The CMS collaboration will also begin making major improvements to their detector - some making use of the new powering schemes.
After 2020, both detectors will finalise their upgrades, and new inner triplets will be installed in the LHC ring. The accelerator and detectors should run into the 2030s.
Project results:
Science and technology results
A variety of technologies have been planned, designed, and prototyped through the SLHC-PP, and scientific studies have been undertaken. Results of this work include identifying key detector upgrades, the findings of the radiation studies, the designs and prototyping of the final focusing magnets, the research towards a new accelerator and source, and the development of two new schemes for powering the central subdetectors within the ATLAS and CMS detectors.
The futures of the detectors
Through the management bodies and procedures developed with contributions from the SLHC-PP, the CMS (WP4) and ATLAS (WP3) upgrade teams have identified key improvements needed to make the most of the extra particle collisions in the upgraded LHC.
WP4: Upgrades for CMS
The CMS collaboration will install the first major upgrades during the shutdown scheduled for around 2017. At this time, CMS will receive a new pixel detector - the subdetector component that sits at the core of CMS, closest to the path of the LHC particle beam. CMS collaborators will also replace the photodetectors in the hadronic calorimeter, a subdetector further from the particle beam that measures the energy of particles containing quarks. Then, in a later shutdown after 2020, the entire tracker subdetector will be replaced, and nearly every other system will receive some improvement. Of note, the system for recording the most interesting proton collisions will be upgraded.
Pixels
CMS's silicon pixel subdetector is the highest-resolution detector in CMS, tracking particles fresh from the collisions in three dimensions. When the LHC is operational, 10 million particles/cm2 pass through the pixel subdetector each second. This rate is expected to increase by almost a factor of ten when the LHC is upgraded.
Added sensing layers in the silicon pixel subdetector will provide an additional handle for teasing apart the particle tracks streaming through the detector. An additional layer of pixels will be added, together with extra pixel-filled 'caps' on each end of the cylindrical pixel subdetector.
But the pixel upgrade involves more than adding extra sensors. The CMS pixel group is also cutting out detector dead space - things like cables, cooling, and structural elements. These components don't sense the particles from the collision, but may absorb particles or break up clean, high-energy tracks into a cascade of smaller particles that make interpreting those high-energy tracks more difficult.
Part of the material savings come from new power management schemes, detailed in WP8. Large cables and connectors will be replaced with streamlined versions that take up less space. The carbon fibre support structure for the pixel detector will also be lighter. By using carbon dioxide to cool the detector rather than the present fluorocarbon coolant, the CMS team can use thinner cooling pipes as well as smaller heat exchanger contacts.
This design is scheduled to be implemented in 2016. A third iteration of the pixel subdetector may be installed in the CMS detector in the early 2020s.
Photodetectors
The hadronic calorimeter in CMS is composed of brass or steel 'absorber' tiles, layered with scintillating plastic tiles. The calorimeter tries to absorb all of a particle's energy - forcing it to create cascades of smaller particles as it passes through the brass or steel - and then measures the resulting particles with the scintillating plastic. Scintillators give off light when charged particles pass through, and this light is collected in optical fibres, which take it to a photodetector to be converted to an electrical signal for analysis.
Photodetectors are specially designed photodiodes that can operate in the exceptionally strong (4 Tesla) magnetic field that exists inside the CMS detector. However, the photodetectors currently installed in CMS sometimes spark when exposed to that strong magnetic field: charges build up inside them, and when discharged, the signals looks confusingly like that of a high-energy particle. 'It makes you think something happened and it didn't. You get a big spike of energy that isn't real,' says Nash. The detector may waste time by exporting data from these 'phantom' collisions.
Silicon photomultipliers could avoid the sparking problem. This technology wasn't around when CMS was being designed - they were first developed in Russia in the late 1990s. Silicon photomultipliers operate at a low voltage and are resistant to the charge build-up seen in the photodetectors now deployed in CMS. Moreover, silicon photomultipliers have a signal-to-noise ratio ten times better than the present photodetectors, making the data they produce more easily interpreted.
Finally, the silicon multipliers can be installed in a way that gives information about the rate at which a particle deposits its energy in the calorimeter. In the current design, the particles essentially travel 'up' a tower of alternating scintillator and absorber tiles. The photodetectors have to sum up the total signal from the scintillators in the tower to give the original particle's energy. But the silicon photomultipliers can differentiate between individual scintillator tiles, providing information about how much energy is deposited in each 'floor' of the tower.
Trigger and data acquisition - recording the important particles
Most digital camera users don't print every photo they take, and nor do particle detectors record every collision they see. In fact, the vast majority of collisions - over 99 % - are discarded the moment they are measured. This is because they contain only well established particles and interactions, which are uninteresting in the search for new physics.
The first port of call for data is the trigger, a system which decides which particle collisions to keep and which ones to discard. Presently, the electronics that export data from the detector are connected to the trigger through electrical wires. In order to cope with higher data rates, the CMS team will replace these with an optical system. 'You can send a higher volume of data through an optical fibre,' says Nash - one reason why internet broadband networks around the world increasingly rely on optical fibres rather than traditional copper wires.
By increasing the processing power of the data acquisition system, CMS will be able to look at the finer details of the data coming in. The detector systems currently assess whether a particle has passed through a sensor or not, but they can squeeze more information out of the signal by looking at nuances of the particle's energy and location. To harness the power of the upgraded processors, the CMS team is developing new software.
For more information on the scope of the upgrade see this draft of the plans: over 45 MB.
WP3: Upgrades for ATLAS
The silicon sensors, CMOS chips and controllers within the ATLAS detector, and the optical fibres for exporting data, are all designed to be 'radiation hard' - they can withstand the flux of high-energy charged particles flying out of the proton and lead ion collisions. However, even these durable devices eventually break down after a few years to a decade of exposure, so replacing them will be a necessary part of any upgrade work. At the same time, the electronic systems for sending data out of the detector and choosing interesting events may be improved, streamlining data analysis. The 'end cap' regions of the detector, which close off the barrel-shaped central parts, will also need upgrades to cope with higher radiation rates.
As in CMS, most ATLAS subdetectors are expected to eventually receive some upgrades. While the Insertable B-Layer is the current main project, planned for the shutdown period 2013 - 2014, others are under various stages of planning and consideration.
Insertable B-Layer
The pixel detector contains silicon sensors, each not much larger than a hair's breadth to a side. Eighty million of these, distributed across three cylindrical layers and four end-cap discs, precisely track the paths of particles as they leave the collision point.
The innermost layer of the pixel detector, known as the B-layer, is the first port of call for particle debris leaving the beam pipe and entering the detector, and the concentration of particles endured by this first layer of sensors is higher than anywhere else in the detector. Originally, this layer was to be removed and replaced, but removing the old layer was too risky, and the operation would take too long.
It is easier to instead remove the beam pipe and replace it with a slightly smaller version. This leaves enough space to add a fourth cylindrical layer, nesting inside the current B-layer. ATLAS upgrade project leader Steinar Stapnes says this new pixel layer will 'add a very important early tracking point of particles leaving the interaction region and increase the robustness of the entire pixel system'.
Further upgrades
The ATLAS collaboration is planning several other upgrades, in various stages of study and approval:
- The trigger team, which develops the automated systems for choosing which particle collisions to record, is developing an improved trigger which draws on data from the particle tracking systems. 'It will help select interesting events both faster and more efficiently,' says Stapnes.
- In general, the trigger team is trying to combine information from many parts of ATLAS as early as possible. The trigger runs through a chain of selection criteria to decide whether to keep or discard a proton collision, and combining multiple detectors earlier can allow ATLAS to adjust the selection system to more effectively seek out or ignore specific kinds of collisions.
- The subdetector devoted to muons, electron-like particles that often herald interesting collisions, will need upgrades for its end cap 'wheels'. The revamped LHC will produce muons in such abundance that more than one will regularly come through a sensor at the same time. An extensive research and development effort is looking into the best type of replacement.
- The upgraded LHC will expose the end caps of the calorimeter system, which stops particles and measures their energies, to the maximum barrage of particles that they were designed to withstand. 'Two solutions are being considered, replacing it with version more robust with respect to high rates, or putting another calorimeter in front of it where a significant part of the energy of the incoming particles is deposited,' says Stapnes.
- To keep the proton beams from colliding with air molecules, they are kept inside pipes which have been emptied of air. To make this 'beam pipe' more transparent to particles passing through, ATLAS is looking to replace it with something made of lighter materials. And, so that sensors can be placed as close to the collisions as possible, the upgrade team is also trying to reduce the beam pipe's radius near the centre of the detector.
The largest and most costly upgrade needed is the replacement of the entire ATLAS Inner Detector, expected to happen in a decade's time. 'A very comprehensive research and development program is underway,' says Stapnes. This program of research and development (R&D) will produce new designs for silicon sensors, data export systems that can withstand high levels of radiation, cooling systems, and support structures. The upgrade will comprise both the silicon strip tracker and pixel detector, for which new powering schemes were developed in WP8.
'The planning and detailed research and development for all these upgrades are well underway,' says Stapnes. 'The formal steps needed to move them one by one into realisation will follow the procedures established for the Insertable B-Layer project, and these have been developed with the support of the SLHC-PP WP3 funding and activities.'
WP5: Radiation safety
Higher luminosity beams and collisions have one unfortunate side-effect: they create more radiation. Some of it is along the accelerator chain as protons occasionally stray from the beam, and some of it is in and around the detector caverns, where ten times as many protons are colliding in tiny explosions of new particles. The increase in radiation needs to be predicted ahead of time so that additional safeguards can be implemented to ensure that neither personnel nor the environment will be harmed.
Radiation studies also looked into the radiation risks associated with implementing upgrades, such as the construction of Linac4 and the replacement of the accelerator components nearest the collision point. These studies are ongoing at CERN, and will eventually cover all radiation concerns connected to the upgrades.
Originally, WP5 intended to study six areas: the point where the Linac4 construction site approaches the running Linac2 accelerator, the radiation that might escape the hypothetical SPL and PS2 accelerators, the radiation that enters the service caverns of ATLAS and CMS, and the activation of the Inner Triplet magnets, which are slated for replacement after 2020.
However, because the PS2 and SPL are no longer part of the luminosity upgrade at the LHC, in 2010 WP5 instead began to consider how much radiation is likely to be emitted by the older accelerators running at higher luminosity. Specifically, focus turned to the Proton Synchrotron (PS), which WP5 project leader Thomas Otto calls, 'the more than 50 year old centrepiece of CERN's accelerator chain.'
A good starting point for radiation safety is the law: anyone expected to come into contact with ionising radiation through their work must not be exposed to more than a 20 milliSievert (mSv) radiation dose per year. This applies to personnel at hospitals or nuclear power plants, and to those at CERN working near the running accelerators and particle detectors. For comparison, the French Institute of Radioprotection and Nuclear Safety estimates that the annual dose for flight personnel is 5 mSv on the routes with the highest exposure to radiation from space, for example New York to Tokyo or Paris to Tokyo.
CERN then takes radiation safety further, following the ALARA principle, which ensures that radiation doses are as low as reasonably achievable. The upshot: it's preferable to avoid exposing personnel to radiation by shielding even those areas that pose a very low risk of contributing significantly to an annual radiation dose.
One of the first steps for the WP5 team was to identify the most important factors that contribute to radiation risk. In accessible areas while beam is running, these are: the energy of the beam, the number of particles accelerated with each pulse, and the average amount of beam passing through an area at any given time. After beam has stopped, activated materials can also emit radiation.
The team also identified the locations in the accelerator chain where particles from the beam were likely to escape and pose a radiation risk. One such point is at the collimators, blocks of heavy metal (often tungsten) that strip straying particles off the beam by passing it through a relatively narrow hole, which are spaced at regular intervals around the accelerator.
The beam is less stable at the points where it enters and leaves accelerators, so these areas are also more likely to be exposed to escaping particles. Finally, beam dumps - massive blocks of tungsten, iron, concrete, and other materials designed to absorb the beam when it is no longer of use - are liable to become activated (milestone report 5.1).
Once the WP5 team had identified the risk factors, they assessed the radiation risks on a project-by-project basis.
Activation
When charged particles run into atomic nuclei, they can change the identity of a nucleus, making it into a radioactive isotope. These atomic nuclei have unusual neutron numbers for the element, and the unstable forms eventually decay by emitting other particles. Those with short half-lives will disappear quickly, settling back to a non-radioactive isotope after emitting an energetic photon, known as a 'gamma ray', and a particle - most often an electron or its antimatter partner, the positron. However, the longer-lived isotopes are more problematic, and can require equipment to be stored for tens to hundreds of years underground or behind thick walls of concrete.
Linac4
The new Linac4, currently under construction at CERN, will need to feed into the PSB through the same injection line used at present by Linac2. The construction workers digging the new tunnel were present on the site from December 2008 until October 2010, overlapping with the LHC's first run at the end of 2009 and its second from March to December 2010.
The question for WP5 was: how can CERN ensure that the construction workers excavating the earth and building the housing for Linac4 will not be exposed to radiation from the running Linac2?
Initially, CERN planned to mitigate the risk by building the link between Linac4 and Linac 2 in winter, when the LHC was not running. But the civil engineers quickly pointed out that digging in frozen ground is impractical. Building work would have to occur during the warmer months, with the LHC operational. At the onset of excavations, the workers were definitely safe - 20 m of earth separated them from Linac2.
However, the construction workers needed to dig right up to the side of the Linac2 building, which meant that as the excavation progressed, eventually just two 2.5 m-thick concrete walls separated them from Linac2. Only the last metre of excavation was likely to pose any significant risk, so the WP5 team simulated the radiation levels likely to come through the two walls.
The simulation showed that without additional shielding, radiation levels would be higher than the 0.5 µSv/h limit for sites accessible by the general public. To bring the radiation down to acceptable levels, CERN added a third concrete wall, 40 cm thick.
With this addition in place, Otto says: 'A number of calculations with the same Monte Carlo code were done in order to demonstrate that the radiation levels would be so low that there is never the need of defining a limited access radiation area'. The excavation workers were at such low risk that they were not required to wear dosimeters to measure radiation levels.
The WP5 team also added a radiation monitor inside the Linac2 tunnel, which can shut down the accelerator if the radiation level there gets too high. The sensor was set for a limit 25 % lower than that prescribed by the simulations in order to provide an ample safety margin. These precautions ensured that the radiation dose received by the construction workers was well under 1 mSv per year, the legal dose limit for non-radiation workers - a success for the WP5 team.
The CERN technicians and engineers who will install the new accelerator in the finished Linac4 tunnel will spend more time near the wall approaching Linac2, and so will still be required to wear dosimeters - but for these workers, such precautions are a standard aspect of their daily work (deliverable report 5.2.1).
Simulations
To forecast the amount of radiation in various parts of the accelerator complex, the WP5 team modelled the effects that the beam and collisions would have on equipment and work spaces. The code used to simulate the radiation and activation of materials, FLUKA 2008, is a type of Monte Carlo code made for the purpose of estimating the effects of ionising radiation and how it moves through materials. The name Monte Carlo comes from gambling algorithms, which attach probabilities to a number of outcomes.
In the FLUKA code, for example, researchers can evaluate the probability that a negative hydrogen ion collides with a nucleus after travelling a certain distance through a magnet. Provided a collision occurs, probability also rules the fate of the outcome: if the starting nucleus is copper it may become radioactive cobalt after the collision, through the removal of a few protons and neutrons. The code runs through these scenarios for millions of particles, randomly selecting the outcome of each step along the way in accordance with the probabilities set through experiments and theoretical models. It simulates the effect of years of radiation exposure.
Inner triplets
The sets of magnets which provide the final focusing of the LHC beams before they enter one of the detectors are on the front line for absorbing the shrapnel of the particle collisions that occur inside the detectors. When high energy particles strike the magnets, they can turn ordinary atoms into radioactive isotopes. For a while, these isotopes act like tiny time-bombs, ready to release radiation when they decay into ordinary atoms again.
Provided that the level of radioactivity in materials becomes negligible within 30 years, CERN will be able to store them and then dispose of them as ordinary waste. However, WP5 showed that the inner triplets will stay radioactive for a longer period.
Ideally, these magnets could be reused, but workers can't afford to spend long hours carefully dismantling them inside the tunnels, where the radiation levels are likely to be high. With the right strategy for disconnecting the magnets, though, Paolo Fessia of CERN's technology department reckons that half of them could be saved and used as spares for the ALICE and LHCb experiments. The rest of the material will be scrapped at a special site for radioactive materials.
The WP5 team recommends that considering how to dismantle the radioactive accelerator parts should become a consideration in the designs of future triplet magnets. Components should fit together 'like bricks in a Lego game' rather than being bolted to one another, says Otto. Alternatively, robots could be used to take the magnets apart remotely, ensuring worker safety no matter how long the dismantling procedure (deliverable report 5.2.1).
Service caverns and surface buildings
A quirk of the local geology required CMS to leave a 6 m-wide pillar of rock between the experimental cavern and the service cavern. The corridor between the two winds around it. This natural shielding is much better than that installed around the ATLAS experiment. The ATLAS service cavern, which contains electronics and cooling equipment, is protected by just 2 m of concrete.
Presently, while the LHC is running, the dose rate inside the ATLAS service cavern is 2-5 µSv/h. With ambient dose equivalents ranging between 0.5 and 15 µSv/h at the highest, the area can be classified as a supervised radiation area. This means that only those certified as radiation workers have access, but they may take their time accomplishing their tasks as it is impossible to exceed annual dose limits with the quoted radiation levels.
When the LHC is upgraded for high luminosity, the dose rates should increase by ten times, reaching 20-50 µSv/h. Because there is no space to add extra shielding, the cavern will have to become a controlled radiation area, which means that the personnel who enter will need special training and closer monitoring. At that point, their work in the service caverns will be subject to time restrictions because in the absence of additional shielding, working more rapidly is the only option to minimise the radiation dose. The workers are unlikely ever to approach the legal limit for exposure of 20 mSv per year, but it is worthwhile to train more people to repair or replace components in the electronics, cryogenics, and other services.
At the surface, the two shafts leading down to the ATLAS experimental hall are each covered with a 1 m-thick 'plug' made of concrete. The narrower of the shafts emits 1 µSv/h while the wider emits 3 µSv/h. An additional metre of concrete - a little more for the wider shaft - will be added to the plugs after the upgrade. The extra shielding will be enough to keep the radiation rates below 2.5 µSv/h, which is the limit for a public area that is not permanently occupied. However, safely setting that much concrete over a hole up to 18 m in diameter is not an easy task, and engineering work will be needed to design the structures that will hold and move the concrete (deliverable report 5.1.2).
Superconducting proton Linac (SPL)
The SPL would be built 20-30 m underground, allowing the soil to provide effective shielding against any lost negative hydrogen ions from the beam, which will have energies of up to 5 GeV. The accelerator needs about 16 shafts, each 2.8 m wide, to run bundles of cable-like RF waveguides from the klystron RF amplifiers at the surface to the RF cavities that accelerate the beam. 'These vertical shafts are also a way out for neutrons which are produced in the accelerator,' says Otto. Neutrons may be generated if stray negative hydrogen ions hit an RF cavity or other equipment.
Simulations of beam lost in the tunnel revealed that more work is necessary to minimise the radiation exposure inside the klystron buildings. Since technicians need access to the klystron building while the accelerator runs, how to protect them from radiation leaking up from below will be a concern for the future developers of SPL-like machines, such as the team working on the European Spallation Source (ESS) in Lund, Sweden (deliverable report 5.2.1).
Proton synchrotron 2 (PS2)
The WP5 team also simulated the radiation risk expected for the beam dump near the PS2, before plans to build the PS2 were shelved. When negative hydrogen from the long, straight SPL entered the oblong ring of the PS2, it would have run through a thin piece of carbon which strips away the two electrons, leaving the bare proton. The PS2 would then alter the trajectory of these protons to travel around its accelerator ring. However, in some cases only one or neither of the electrons would be removed, and the hydrogen would not make it into the PS2 ring. The negative hydrogen would bend the wrong way in the magnetic field, and the neutral hydrogen would go straight.
To catch these wayward particles, accelerator physicists would need to add a beam absorber. Ideally, the absorber would be able to catch both the negative and neutral hydrogen. To capture the diverging streams of particles in a single, relatively small absorber, the beam dump would have to lie close to the accelerator.
The dump near the PS2 would need to catch about 6.4 x 1019 particles per year, with each particle at an energy of 4 GeV. It would have a design reminiscent of the 'layers of an onion', says Otto. A carbon core would be enveloped by aluminium and then tungsten. This would then be surrounded by a block of iron, 1 m high and 3.2 m long. Finally, the entire structure would be encased in a 20 cm layer of concrete.
If this was to go into the PS2 accelerator vault, it would need to give off less than 50 µSv/h per hour so as not to pose a radiation risk to the workers making repairs to the accelerator during shutdowns. The beam dump would breach this limit only at the opening where the beam enters the block of absorbers. A simple mechanism to block that beam entrance would be enough to keep the radiation level sufficiently low (deliverable report 5.2.1).
Although the PS2 is no longer planned for construction, this study of the PS2 beam dump can serve the ESS in Lund, as the accelerator there will employ an identical injection scheme, requiring similar equipment.
Validating the simulations
Simulations are a valuable tool for forecasting the activation of materials at the LHC and, eventually, the radiation emitted by the LHC after upgrade work has been completed. However, to rely on this information for safety purposes, WP5 needs to give evidence that the simulations are accurate. 'The best way of proving that you are not making a big error somewhere is to conduct some experiments, some validation,' says Otto.
One way to do this is to see whether the simulations give accurate predictions about the LHC. Before the accelerator started up in 2008, WP5 placed hundreds of radiation detectors in the LHC tunnel, the ATLAS and CMS experimental halls, and the service caverns. Most of these are passive detectors, which record the total radiation to which they are exposed, to be assessed after collection. A few send out a live feed of information. One kind, the MPX detectors placed in the hottest positions of ATLAS experiment, have been already able to register a short-term activation induced during some relatively high luminosity proton collisions in the LHC.
The plan was to collect the passive detectors after the LHC's first year of running to make sure that the simulations were accurate. Unfortunately, the first year of LHC beam came later than expected. At the end of 2010, the Paul Scherrer Institut (PSI) team removed and evaluated their neutron detectors, indicating radiation levels in the same order of magnitude as expected for the luminosity accumulated so far. Another set of detectors, placed by different collaborators coordinated by the CERN radiation protection group, is still underground, waiting until the end of 2011 to be removed.
These detectors will eventually give vital feedback about the amount of radiation that will reach accessible areas during beam as well as the amount of activation that materials will pick up (deliverable report 5.1.1).
Impact study
The WP5 team also needed to explore ways that radiation may leave CERN and possibly impact the environment. There are three ways this could happen: through the release of radioactive air or water, and through the inappropriate disposal of radioactive waste.
The inner triplet magnets are set for replacement, as are the collimators. Because simulations show that both types of equipment will be significant sources of radiation for over 30 years, they will need to be sent to a special site for radioactive waste storage and disposal.
During the operation of an accelerator, it is also possible for the oxygen, nitrogen, and other components of ordinary air to become radioactive, and so the WP5 team studied this activation process. Of special concern is the chance that the 52-year-old PS accelerator may activate the air inside the tunnel and then release it into the environment.
Even though the PS will handle beam intensities 1000 times higher than it was originally designed for, the simulations of WP5 show that the release of radioactive air should be 100 times smaller than the legal limit. Nevertheless, the ventilation system will be renovated to control airflow from the PS and measure the actual amount of radioactivity released through the air.
On the other hand, the simulations did reveal weak points in the PS shielding that will need to be reinforced with concrete or iron to reduce the likelihood of stray radiation escaping into the CERN campus. This is hardly surprising - although scarcely any part of the actual accelerator is original, its circular tunnel and earthen hill of shielding date back to the late 1950s.
CERN is generally fortunate in its location when it comes to water risks - the tunnels are far from large stores of ground water. However, rainwater may percolate through the ground near the accelerators and so pick up radiation. Water which reaches the present accelerator tunnels at CERN is not activated beyond the levels of natural radioactivity. Even so, Otto says: 'In the vicinity of future beam dumps, built for higher beam intensities, a sensible precaution would be to separate the evacuation paths of water ingress from the ordinary gutters. In the case of a slight elevation of radiation levels in the water, this would allow CERN to treat it separately.'
As a result of the radiation studies in WP5, CERN has taken steps to reduce the radiation exposure of personnel, the upgrade planners know how much the radiation from the LHC and its accelerator chain will increase, and safety officers have recommendations for how to mitigate this future risk.
Information about the WP6, WP7 and WP8 and their expected results can be found in the documents attached.