This paper develops threat models and mitigation techniques for mutually consistent satellite faults that are potentially hazardous to aviation users. Of immediate interest among the possible threats in this class are faults in the generation of Earth Orientation Parameters (EOPs), which cause an effective rotation of the satellite constellation. The EOP threat can be mitigated using multi-constellation Advanced Receiver Autonomous Integrity Monitoring (ARAIM). However, this approach requires independence across core constellations which can be difficult to validate. In response, this work provides a new fault detection algorithm that operates at the aircraft in a single constellation mode. Using the fact that adjacent ephemerides are guaranteed to be valid over a two-hour overlap period, the airborne algorithm generates two estimates of the satellite position vector. The difference of these estimates is used as the measurement input to a weighted least squares estimation algorithm that estimates the parameter vector characterizing the EOP fault. In addition, it is shown that EOP faults can be accurately modeled as an initial bias plus a linear time variation, which enables the determination of their impact at any point in time. A fundamental assumption of the airborne algorithm is that a set of validated (i.e. fault-free) ephemerides is available at aircraft dispatch. In order to provide this information, a civil ground monitor is proposed that requires at least two ground stations for fault detection. Ephemerides that pass civil ground monitor validation are broadcast to the aircraft through the Integrity Support Message (ISM); a key component in all postulated ARAIM architectures. Global availability analysis shows that the inclusion of the EOP fault detection algorithm does not cause a significant loss of availability for LPV-200 aircraft precision approach applications.