Reasoning For Maintenance Assessment in Early Design Phases

Uncertainty Quantification (UQ) is becoming increasingly relevant in the reliability assessment of complex technical systems, particularly in early design phases where empirical data is limited. This paper introduces an integrated methodology that combines UQ techniques with Life Cycle Cost (LCC) analysis and dynamic Fault Tree Analysis (FTA) to support decision-making in maintenance strategy development. A key focus lies on the modeling of maintenance-related failures and their impact on system availability, especially under conditions of incomplete or uncertain failure rate information. A central contribution of this work is the explicit integration of maintenance actions into static fault tree analysis by modeling maintenance as a functional prerequisite rather than a repair process. This allows maintenance-related risks to be consistently represented in early-phase static FTA when empirical failure data are unavailable. To address this challenge, the paper proposes a practical approximation, which enables realistic estimation of system availability by incorporating uncertainty into early-stage reliability models. This approximation facilitates the representation of maintenance errors and their propagation within fault tree structures, enhancing the predictive power of the analysis. Uncertainty is propagated using Monte Carlo simulation to support early-phase reliability assessment under incomplete failure information. These simulations allow for the exploration of a wide range of scenarios and parameter sensitivities, contributing to a deeper understanding of system behavior under uncertainty. The applicability and benefits of the proposed approach are demonstrated through a representative case study. Results highlight the potential of the methodology to improve reliability predictions, optimize maintenance planning, and reduce lifecycle costs in the conceptual design phase of safety-critical systems.