Risk-Based Comparison of Consolidated and Distributed Satellite Systems

Abstract

Governments and large private satellite operators face a strategic decision for multipayload satellite programs: whether to use large consolidated satellites with multiple payloads or constellations of a few smaller spacecraft (here a “distributed” system). One advantage of the latter is that it allows faster replacement of failed payloads and more efficient insertion of new technologies. The costs and the benefits of each option are a function of a myriad of uncertain factors, including the sequence of decisions that have to be made during the lifetime of the program. We present a general model based on decision and risk analysis to support the choice between these two satellite architectures for a long-term program in Earth's orbit. This model is built on probabilistic risk analysis of satellite vehicles and includes the operational decisions that may have to be made later, such as the construction and launch of new satellites, either because one fails or because a new technology has become available. The value of each architecture is based on that of the data that the satellite system provides when it is operated according to an optimal policy. We compute the value of each satellite architecture, reflecting the uncertainties involved and the decision maker's risk attitude in optimal management decisions. This model thus yields the sensitivity of both satellite architecture and operational decisions to factors such as risk aversion and the anticipated rate of technology advancement. We find that in many circumstances, more risk-averse decision makers may prefer distributed satellite architectures and the flexibility that they provide, even at a potentially higher initial cost. We also show that distributed systems tend to be preferred when the payload technology is expected to improve rapidly over time.