Assessing Unaccounted Events and Their Possible Impact on the Maeslant barrier's Non-Closure Probability
W.J. Waasdorp (TU Delft - Civil Engineering & Geosciences)
A.M.R. Hoffmans – Mentor (TU Delft - Hydraulic Structures and Flood Risk)
Jose A.A. Antolinez – Graduation committee member (TU Delft - Coastal Engineering)
Gina Alexandra Torres Alves – Graduation committee member (TU Delft - Hydraulic Structures and Flood Risk)
R. D.J.M. Steenbergen – Graduation committee member (Universiteit Gent)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
The Maeslant barrier is a storm surge barrier and a critical component of the Dutch coastal flood defense system. Its reliability is formally assessed through a Reliability and Availability (RA) analysis, which estimates the probability of non-closure during storm events. However, concerns have been raised regarding the completeness and transparency of this analysis, particularly the potential omission of relevant failure events. This thesis investigates whether a selected set of previously unaccounted for events can be systematically identified and quantified to improve the accuracy of the non-closure probability.
A three-stage methodology was developed. First, a structured inventory of unaccounted-for events was constructed using HAZOP, FMEA, What-If, and external event screening techniques, mapped across four analytical dimensions. Second, the list was filtered based on estimated occurrence probability and quantifiability, resulting in a shortlist of three events: epistemically uncertain events, non-stationary component degradation, and the unverified reliability of human interventions. Third, these events were quantified using structured expert judgment, research into time-dependent fault tree modeling, and human reliability assessment.
Results indicate that these unaccounted-for events can alter the estimated non-closure probability, either increasing it by an order of magnitude or reducing it by up to 50%. Moreover, the analysis revealed limitations in the current RA analysis, including outdated reliability assumptions, a fragmented integration of human interventions, and a lack of empirical data. These findings support the need for a more transparent and adaptable RA framework. The discussion highlights that while completeness in risk assessment is theoretically unattainable, similar to the limitations of physical laws, models should strive for an optimal balance between complexity, traceability, and applicability.
Recommendations include developing a centralized component lifecycle database, maintaining a registry of previously unaccounted-for events, formally integrating the OPSCHEP model into the fault tree structure, and adopting structured human reliability verification. These changes can improve the accuracy, transparency, and credibility of the Maeslant barrier’s non-closure probability and serve as a blueprint for other critical infrastructure systems.