Linear accelerator bunkers are designed to prevent unintentional exposure of staff, patients and the public.
The bunker is typically designed with a single entry point. This makes it easier to protect the entrance from both radiation scatter out of the bunker as well as people unintentionally entering the bunker. Bunkers usually consist of the entrance (which may have a neutron door), a maze, a primary and a secondary barrier. Ideally, rooms that adjoin the linac bunker should be for minimal use (eg. storerooms); the exception is the control room which is always manned.
For linear accelerators that use high megavoltage energy photons (over 10 MV), it is vital to have adequate ventilation to remove radioactive gases that are generated by neutrons (released through photodisintegration). Air within the bunker should be cycled at least 6 times per hour.
The barriers are designed to attenuate almost all of the radiation produced by the linear accelerator. The barrier thickness is dependent on the position of the barrier relative to the linear accelerator gantry.
The primary barrier is designed to attenuate the primary radiation beam that is produced by the linear accelerator. The thickness of this barrier is dependent on the most penetrating energy produced by the linear accelerator - typically 10 MV, 18 MV or 20 MV depending on the model. For 18 MV, a barrier thickness of 260 cm of concrete is sufficient; this thickness can be reduced if more attenuating materials are used (eg. dense concrete or metals).
The secondary barrier is designed to attenuate scattered radiation from the primary beam as well as leakage radiation from the treatment head. Leakage radiation is typically of higher energy than scattered radiation, and therefore the thickness of the secondary barrier is typically determined by the leakage radiation. The secondary barrier thickness is typically half of the primary barrier thickness - eg. 130 cm of concrete for an 18 MV beam.
Photons may be scattered off walls or other components within the bunker. This scattered radiation could exit through a door in the room unless that door is 'hidden' by a maze. The maze is typically designed to prevent scattered photons from exiting the bunker directly - they must interact with a wall before they reach the door. This reduces the amount of radiation that escapes from the bunker by a significant amount.
The exception to this is when neutrons are generated through photodisintegration (photon energies over 10 MV). A maze that is sufficient to attenuate scattered photons may not be sufficient for scattered neutrons, which may be scattered multiple times. This is overcome through the use of a larger maze (which causes neutrons to interact an increased number of times before reaching the exit) or through the use of a neutron door.
Neutrons may be scattered more frequently than photons, and are also attenuated through different mechanisms. The ideal neutron attenuator is dense and of low atomic number; in general neutrons doors contain a 12 cm panel of boranated polyethylene which is a good neutron absorber. Photons may be released through neutron absorption, and therefore an additional lead barrier of 2.5 cm thickness is added on the outer side of the door to attenuate these photons.
A neutron door allows a bunker to be more compact, but unfortunately leads to increased treatment time as staff must wait for the door to close before activating the linear accelerator. A more complex maze takes up more space but treatment time is faster. An alternative is to simply use photon energies under 10 MV, which prevents the formation of neutrons and the need for a neutron door.