Physics in Safety and Safety in Physics: Lessons from the LHC and its Underground Experiments

Deep underground, beneath the rolling hills of the Swiss-French border, lies a machine built to answer the most fundamental questions about the universe. It is a ring of superconducting magnets, vacuum pipes, and detectors stretching twenty-seven kilometers in circumference, designed to smash protons together at speeds approaching that of light. This is the Large Hadron Collider, a feat of engineering that generates immense amounts of energy, creates powerful magnetic fields, and operates at temperatures colder than deep space. To the outside observer, the goal is simple: collide particles to discover new laws of nature. But for the scientists and engineers who run the machine, the path to discovery is paved with a different kind of challenge. The very forces that make the experiment possible—the stored energy of the beams, the magnetic fields holding them in place, the liquid helium keeping the magnets cold—are also the sources of the most significant dangers. If something goes wrong, the energy released could damage the machine, harm the people working nearby, or alter the environment in the tunnels.

For decades, the scientific community has treated the physics of the experiment and the safety of the facility as two separate tracks. One team designs the collisions to maximize data; another team builds the fences, alarms, and emergency stops to keep everyone safe. Alexander Sakharov, a physicist with experience at both Manhattan University and CERN, proposes that this separation is an illusion. In a new analysis, he argues that safety is not an external layer added to the machine after the design is finished. Instead, safety is woven into the physical fabric of the experiment itself. The laws of physics that allow the collider to work also dictate exactly how it can fail, and the systems built to prevent those failures define the very boundaries within which science can happen.

Sakharov's work reframes the relationship between the machine and its protection systems. He suggests that we should think of the collider not just as a collection of parts, but as a single, complex system moving through a vast space of possible states. Some of these states are safe, while others lead to disaster. The goal of safety engineering is not merely to stop the machine when it breaks, but to map out the safe paths through which the machine can travel. A safe path is not just about being in a safe place at a specific moment; it is about how the machine got there and how it moves to the next step. Just as a driver must follow a specific route to avoid a cliff, the collider must follow a specific sequence of actions to avoid releasing its stored energy in a way that causes damage. If the machine tries to jump from one state to another too quickly, or if it passes through a configuration that hasn't been validated, the safety systems will stop it.

The paper uses the Large Hadron Collider and its two massive detectors, ATLAS and CMS, as the primary examples to illustrate this idea. These detectors are not passive instruments; they are active, energetic systems in their own right. They contain their own superconducting magnets, which store gigajoules of energy, and complex networks of cooling and electrical systems. When a magnet suddenly loses its superconducting ability, a phenomenon known as a quench, the stored magnetic energy must be dissipated safely, or it can melt the equipment. Similarly, the liquid helium used to cool the magnets expands by a factor of a thousand if it warms up and turns into gas. If this happens in a confined underground tunnel, it can displace the oxygen, creating a hazard for anyone working there. Sakharov shows that the physical models used to predict how a beam of protons behaves are the same models needed to predict how a helium leak will spread. The science of the collision and the science of the hazard are one and the same.

One of the most striking insights in the paper comes from looking at the history of the facility. In 2008, shortly after the collider began operating, an electrical fault in a specific section of the tunnel caused a massive release of helium. The event damaged the machine and forced a long shutdown. Sakharov points out that this accident was not just a failure of a component; it was a failure of the safety model. The engineers had predicted how much helium might be released, but the actual event exceeded those predictions. The lesson learned was that safety knowledge must evolve alongside the machine. As the facility operates, tests, and encounters near-misses, the understanding of how hazards propagate changes. The safety systems are then updated to reflect this new knowledge, creating a feedback loop where the experience of the machine improves the protection of the machine.

The paper also highlights that different parts of the facility have different safety needs, even if they seem similar. The ATLAS and CMS experiments, while both located in the same underground tunnel, have different designs for their magnets and detectors. Consequently, their safety systems are not identical. ATLAS uses a specific arrangement of magnets that requires one type of protection, while CMS uses a different arrangement that requires another. This proves that safety cannot be a one-size-fits-all rulebook. It must be tailored to the specific physical architecture of each experiment. The way a detector is built determines how it can fail, and therefore determines how it must be protected.

A crucial part of Sakharov's argument is the concept of the "safe trajectory." It is not enough for a machine to be in a safe state; the path it takes to get there must also be safe. For example, ramping up the current in a magnet to reach its operating level is a process that takes time. If the current is increased too fast, the magnet might quench. If it is increased too slowly, it might not reach the necessary state before other systems fail. The safety system must monitor not just the final value of the current, but the speed at which it changes and the sequence of other events happening at the same time. This means that safety is a dynamic property. It depends on the timing of actions, the order of operations, and the ability of the system to react to changes before they become irreversible.

The paper also addresses the idea that safety is often seen as a limitation on science. Engineers might think that safety rules prevent them from pushing the machine to its limits. Sakharov argues the opposite. He suggests that robust safety systems actually enable more ambitious science. By having reliable protection systems that can detect and stop a problem before it causes damage, scientists can operate the machine at higher energies and with more intense beams than they could otherwise. Without these systems, the risk of catastrophic failure would be too high to justify the operation. In this way, safety is not a barrier to discovery; it is the foundation that makes high-performance operation possible. It allows the facility to explore dangerous physical regimes with confidence, knowing that the boundaries are well-defined and the exits are clear.

The discussion extends to the future of particle physics, including the proposed High-Luminosity LHC and even more ambitious projects like the Future Circular Collider. As these machines become larger and more powerful, the stored energy and the potential hazards will increase. Sakharov argues that the design of these future facilities must integrate safety from the very beginning. Instead of designing the machine first and adding safety later, the engineers must design the machine and its safety systems together. This approach, which he calls "safety by design," ensures that the physical layout of the facility supports the protection systems. For example, the placement of ventilation shafts, the segmentation of electrical systems, and the accessibility of components for maintenance are all decisions that affect safety. If these are considered during the initial design phase, the facility will be easier to operate and safer to maintain throughout its lifetime.

The paper concludes by emphasizing that safety is a continuous process, not a one-time achievement. A facility is safe only as long as its protection systems are working, its operators are trained, and its models are up to date. As the machine ages, as components are replaced, and as new technologies are introduced, the safety requirements must be re-evaluated. The history of the Large Hadron Collider shows that even the most carefully planned systems can encounter unexpected problems. The key is to have a framework that allows for learning and adaptation. By treating safety as an intrinsic part of the physics, rather than an external constraint, the scientific community can build facilities that are not only capable of great discoveries but also capable of sustaining those discoveries over decades of operation.

In the end, the paper offers a unified view of the collider. It is a place where the laws of physics create both the opportunity for discovery and the potential for danger. The systems that keep the machine running and the systems that keep the people safe are two sides of the same coin. They are bound together by the same physical principles, the same energy flows, and the same need for precision. Understanding this connection is essential for the future of high-energy physics. It reminds us that to push the boundaries of what is known, we must first understand the boundaries of what is safe. And in doing so, we find that the path to the unknown is paved with the careful, deliberate, and deeply physical work of protection.