PSI operates a cyclotron based high intensity proton accelerator routinely at an average beam power of 1.3MW. With this power the facility is at the worldwide forefront of high intensity proton accelerators. The beam current is practically limited by losses at extraction and the resulting activation of accelerator components. Further intensity upgrades and new projects aiming at an even higher average beam power, are only possible if the relative losses can be lowered in proportion, thus keeping absolute losses at a constant level. Maintaining beam losses at levels allowing hands-on maintenance is a primary challenge in any high power proton machine design and operation. In consequence, predicting beam halo at these levels is a great challenge and will be addressed in this paper. High power hadron driver have being used in many disciplines of science and, a growing interest in the cyclotron technology for high power hadron drivers are being observed very recently. This report will briefly introduce OPAL, a tool for precise beam dynamics simulations including 3D space charge. One of OPAL's flavors (OPAL-cycl) is dedicated to high power cyclotron modeling and is explained in greater detail. We then explain how to obtain initial conditions for our PSI Ring cyclotron which still delivers the world record in beam power of 1.3 MW continuous wave (cw). Several crucial steps are explained necessary to be able to predict tails at the level of 3\sigma ... 4\sigma in the PSI Ring cyclotron. We compare our results at the extraction with measurements, obtained with a 1.18 MW cw production beam. Based on measurement data, we develop a simple linear model to predict beam sizes of the extracted beam as a function of intensities and confirm the model with simulations.

With an average beam power of 1.3 MW the PSI proton accelerator facility is presently at the worldwide forefront of high intensity accelerators. This talk describes critical aspects and recent improvements related to generation and transport of the high intensity beam in a cyclotron based facility. The installation of new accelerating resonators in the second of two cyclotrons led to a significant improvement in view of beam intensity but also the reliability of the facility. Besides the overall performance and further upgrade plans the discussed topics include: space charge dominated beam dynamics, beam loss handling, activation and specialized technical interlock systems.

A new type of ultra-cold neutron (UCN) source based on the spallation process is under construction at PSI. The essential elements are a pulsed proton beam with highest intensity (Ip≥2mA) and a low duty cycle (1%), a lead spallation target, a large D2O moderator and a solid deuterium (sD2) converter system. Spallation neutrons are thermalized in the D2O, further cooled and partially downscattered into the ultra-cold neutron range in the sD2. The expected UCN density is higher than 1000UCN/cm3 in typical experiments, an increase of almost two orders of magnitude over the best source currently available (at ILL). The new UCN source is expected to come into operation in 2009.

The new protontherapy facility PROSCAN at PSI consists of a dedicated 250MeV superconducting cyclotron, new beam lines, two gantries (one new) and a new area for eye treatments. A degrader and laminated magnets in the beam line enable fast energy changes to modulate the range of the scanning pencil beam in the patient. In the design of systems and components a lot of emphasis has been put on reliability, easy maintenance and a clear separation of responsibilities.

A new type of ultracold neutron source based on the spallation process is under construction at PSI. The essential elements of this source are a pulsed proton beam with a highest intensity of (Ip > 2mA) and a low duty cycle (~1%), a lead spallation target, a large D2O moderator and a solid D2 (sD2) converter system. Spallation neutrons are thermalized in the D2O, further cooled in the sD2 and, finally, some of them are down-scattered into the ultra-cold neutron range (Tkin < 250 neV). The expected UCN density in the new source is 3000 UCN/cm-3, almost two orders of magnitude higher than with the best source currently available (at ILL).