As part of Safety-of-Life (SoL) civil aviation GNSS augmentation systems, the GNSS receiver algorithms determine Protection Levels (PL) that binds the Positioning Error (PE). The capability to compute these Protection Levels depends on the availability of information that characterizes and bounds the value for the signal different error sources. These error contributions include the satellite environmental noises, the atmospheric delays and, significantly, the satellite orbit and clock errors. For this reason, the continuous monitoring and evaluation of the accuracy of GNSS constellations orbits and clock broadcast data is a crucial element in the current operation and future implementation of GNSS based Safety-of-Life (SoL) navigation services such as SBAS, ARAIM, RT-PPP Integrity, or the Safety-of-Life commitments for next-generation GPS and Galileo. In this study, a platform and a methodology for orbits and clock data long-term error characterization developed by GMV are presented. The magicARAIM platform allows the evaluation and characterization of satellite navigation errors from multiple navigation services, including GPS and Galileo ephemeris (with especial relevance for future ARAIM services). Also, it supports the evaluation of the long-term error of other sources of satellite orbits and clocks data including SBAS services, as WAAS or EGNOS, or even RTCM corrections or precise products generated by different entities, assessing the accuracy of the reference products itself. The magicARAIM platform can be used both to identify instantaneous integrity faults, as well as to continuously determine whether the statistical behavior is consistent with the statistical assumptions supporting the various integrity concepts. In particular, the platform has been used to analyze the presence of biased errors in the GPS Ephemeris radial direction when compared to the IGS reference products due to a different definition of antenna offsets. The source of those biases is determined, and a methodology to correctly manage those biases is presented. The first section of this paper introduces the problem in the satellite error evaluation derived from the diversity of antenna offset calibration by different entities. The second section describes the magicARAIM platform high-level architecture and key methodologies. The third section provides the evaluation of long-term satellite orbit and clock information as provided by three types of services: broadcast Ephemeris, including GPS and Galileo; SBAS corrections (including EGNOS, WAAS, MSAS, GAGAN, SDCM and the Australian and New Zealand SBAS Testbed), and RT-RTCM services (including Australian and New Zealand SBAS Testbed PPP service).