Accelerator Driven Systems (ADS) for transmutation of nuclear waste typically require 350 MeV–1 GeV accelerators delivering proton fluxes of 5–10 mA for demonstrators, and 20–50 mA for large industrial systems. Thus, such machines belong to the category of the so-called HPPA (High-Power Proton Accelerators), with multi-megawatt beam power. HPPA are presently developed and constructed at great pace for their broad utility in fundamental or applied science. Compared to other HPPA, many features and requirements are similar for the ADS driver. However, there is a need for exceptional reliability: because of the induced thermal stress to the sub-critical core, the number of unwanted “beam-trips” should not exceed a few per year, a requirement that is several orders of magnitude above usual performance. Consecutive to the work of the European Technical Working Group (ETWG) on ADS, the Preliminary Design Study of an Experimental ADS (PDS-XADS) was launched in 2001 as a 5th Framework Program EC project.1 A special Working Package (WP3) was dedicated to the accelerator design, taking especially into account that the issue of “beam-trips” could be a potential “show-stopper” for ADS technology in general. A reference solution, based on a linear superconducting accelerator with its associated doubly achromatic beam line, has been worked out to some detail. For high reliability, the proposed design is intrinsically fault tolerant, relying on highly modular “de-rated” components associated to a fast digital feedback system. The proposed solution also appears to be robust concerning operational aspects like maintenance and radioprotection. A roadmap for construction as well as the required consolidated budget was elaborated. A program for the remaining R&D, focused on experimental reliability demonstration of prototypical components has been elaborated. This R&D will be performed in the 6th Framework Program EC project EUROTRANS,2 which presently is just starting.