Samenvatting

Increased activity of the European beaver (Castor fiber) presents a growing structural threat to regional flood defenses in the Netherlands due to extensive burrowing. Flemish Waerways Beaver Experts did not expect this high of beaver activity and the only sustainable solution against them so far was reactive maintenance instead of preventive and predictive. This research introduces a predictive, risk-based framework that integrates Extreme Value Theory via the Gumbel distribution, a temporal Potential-Functional Failure (P-F) time-curve, and a weighted Multi-Criteria Analysis (MCA) to evaluate a 250-meter case study reach of Canal Beverlo, divided into ten 25-meter segments. This analysis is based on Ultimate Limit State (ULS) calculations using Lane’s Weighted Creep Theory and Limit Equilibrium Method (LEM) used in D-Geo Stability calculations. The ULS Calculations have resulted in max allowable beaver tunnel length of 13.6m, as a factor against piping, and the most dangerous case of burrowing for embankment’s Factor of Safety. The spatial Gumbel model established a baseline extreme signature μ= 5.28m β= 1.43m), identifying recorded 9.0-meter burrows as severe statistical anomalies in the 92.8th percentile. Furthermore, the P-F time-curve demonstrated a staggered asset degradation pathway, showing that while 9.0-meter burrows trigger immediate threat windows of within a season, remaining segments fail across a distributed two-month window afterwards. The developed MCA framework successfully functions as a operational guide, allowing water boards to transition from crisis management to a phased, proactive maintenance schedule while optimizing the deployment of deterministic software like D-Geo Stability.