External maintenance rate prediction and design concepts for high reliability and availability on space station Freedom

As envisioned in 1990∗∗The study reported in this paper dealt with the space station Freedom design as it was visualized in early 1990. NASA has made many changes since then (some in response to the findings of this work), so many of our observations no longer apply to the current ‘Space Station Alpha’ design. Nevertheless the lessons learned from the evaluation of the 1990 design are generally applicable. rnational space station Freedom (SSF) was to be a large, complex, multi-purpose manned facility with a 30-year design life. The presence of a crew and the size of the investment would require high reliability and availability for safety- and mission-critical systems, but even with these requirements satisfied, long life and complexity would make equipment failure and repair inevitable. Recognizing this fact, in 1990 NASA and its contractors performed a study of the maintainability of the then-current space station configuration. Based on well-founded predictions of equipment failure rates, life limits, maintenance rate factors, and repair times, maintenance on equipment outside the pressurized modules of the station alone would have absorbed more than 3000 crew hours per year, which was clearly prohibitive. The chief reason for this maintenance overload was not the unreliability of equipment; it was the sheer number of components present. This paper summarizes this study, and presents a number of design and operational concepts which promise to alleviate the maintenance overload, many of which have been adopted in the current design. ntral recommendation was to think of Freedom as a long-term ‘facility’ analogous to a ground-based research laboratory or manufacturing plant, rather than as a ‘space mission’. This philosophy leads to placing as much emphasis on maintainability as reliability in design; rethinking conventional manned-spacecraft design rules such as ‘fail-op-fail-op-fail-safe,’ which contributed heavily to the complexity of the baseline configuration; prioritizing maintenance by time-to-effect as well as ‘criticality,’ and zero-basing systems to wring out excessive complexity while retaining acceptable performance and safety.