Electromagnetic Interference during the Installation of a New Linear Accelerator
Abstract
Purpose
The interaction between static magnetic fields and charged particle beams necessitates isolating linacs from the magnetic fields from MRI systems and other sources. In the absence of known sources of strong magnetic fields, it has not typically been observed that magnetic field shielding is necessary for routine installations. We report on the novel problem, and attempts at mitigation, of a linac installation affected by magnetic fields over twice the level of background in the absence of known sources of magnetic fields.
Methods
Prior to magnetic shielding, linac performance was characterized in terms of loss of dose rate vs gantry angle, and a magnetic field survey was performed in the vault. A prototype magnetic shield composed of M36 steel was built to allow for continued testing without demolition of the existing space, and the linac performance measurement and magnetic field survey were repeated.
Results
Prior to shielding, dose rate variation exceeded 10% (tolerance) for 40% of photon energies, and 60% of electron energies. In the area designated for prototype shielding, gauss readings averaged 0.94 G on the ground, 1.12 G at 1 m, and 2.08 G at 2 m (tolerance 1 G). After shielding, dose rate deviations for all energies fell within 10%, and average gauss readings fell to 0.79 G, 0.79 G, and 0.74 G on the ground, at 1 m, and at 2 m, respectively.
Conclusion
Magnetic field mitigation after accelerator installation is possible but can double the construction costs associated with installation of a new machine due to the need to demolish and then refinish a previously prepared space, and the increased difficulty in performing construction around a linear accelerator rather than in an empty room. Some of the additional expenses and delays can be avoided with simple measurements taken prior to installation, during site survey.