RAMS engineering

Introduction

In case you are not familiar with the fundamentals of RAMS engineering, you are referred to Fundamentals of RAMS engineering. Below, we will deal with the more specific aspects of RAMS applied to spacecraft.

Spacecraft reliability

Spacecraft failure rate data

An overview of typical spacecraft (S/C) failures can be obtained from Satellite Outages and Failures.

In studies by [Frost & Sullivan] and [Sultan] of past satellite outages and failures it has been determined that the majority of failures can be attributed to the bus (i.e. the payload platform), rather than the payload, see the next table.

The difference between the data from Frost & Sullivan and Sultan is that the former uses a more limited set of data focussing on on-orbit failures of Commercial Communications Satellites only, whereas Sultan also included other satellites, including Earth Observation (EO) satellites. Another reason for this difference might be differences in the anomalies counted. Frost & Sullivan counted only those anomalies that have resulted in insurance claims, in order to avoid problems of biased and incomplete data as well as issues of data comparability. The type of data used by Sullivan is not known.

Using the data from Sultan and combining this with data from [Jane's], it is estimated that over the period 1995 upto and including 2000 (a 6-year period) roughly 560-570 satellite on-orbit failures have occurred (provided that all failures that occurred have been reported and taken into account in the study of Sultan). This than translates into about 95 satellite on-orbit failures per year. Assuming that we have on average 1000 active (operational) S/C this than translates in a failure rate:

Lambda = 95/1000 = 0.095 failure per spacecraft/yr.

This means that on average every year 1 out of every 10 S/C will experience a failure. Whether these failures are catastrophic failures or not is still to be determined. 

Subsystem failure rates

Using the data from Sultan we can estimate the percentage number of  orbital failures associated with the spacecraft bus subsystems. Results are given in the figure below.

^{RCS = Reaction Control System, ACS = Attitude Control System, AKM = Apogee Kick Motor, EPS = Electric Power System, Mechanism = Structures and Mechanism, TT&C = Communications and Command & Data Handling, PKM = Perigee Kick Motor, Thermal = Thermal system}

The figure shows that the two S/C subsystems most prone to failure are the propulsion (RCS + AKM + PKM) and the ACS system.

The percentage number of orbital failures per S/C subsystem can also be estimated from the Frost & Sullivan data. In that case, we find that the power subsystem is responsible for over 65% of all insurance claims made. A good second is the propulsion system being responsible for ~21% of all claims made.

The different results may be attributed to different categories of spacecraft being considered, differences in the (calendar) period considered, differences in the definition of the various S/C subsystems, or because of differences in failures being considered (for instance, not taking into account those failures that only have benign consequences).

References

1. Commercial Communications Satellite Bus Reliability Analysis, By Frost & Sullivan
2. Cheaper, faster is better only if safer: Need for “FIRST”®: “Failure Information Reliability Space Tool ” To retrieve & analyze orbital failures, by N. Sultan, CSAT.
3. Jane's Space Directory.