Methodology for Implementing Monitoring Data into Probabilistic Analysis of Existing Embankment Dams

Monitoring data provide valuable information on embankment dam behavior but are typically not integrated into a classical probabilistic safety assessment. This paper introduces a Bayesian-inspired methodology to directly integrate actual dam monitoring records into a Monte Carlo probabilistic safety...

Full description

Saved in:
Bibliographic Details
Main Authors: Ljubo Divac, Veljko Pujević, Dejan Divac, Miloš Marjanović
Format: Article
Language:English
Published: MDPI AG 2025-06-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/15/12/6786
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Monitoring data provide valuable information on embankment dam behavior but are typically not integrated into a classical probabilistic safety assessment. This paper introduces a Bayesian-inspired methodology to directly integrate actual dam monitoring records into a Monte Carlo probabilistic safety assessment using a finite element framework, without recalibrating the original input parameters ‘distributions. After the baseline (unweighted) set of simulations is generated, the method assigns a weight coefficient to each simulation outcome based on the likelihood of matching monitoring data, effectively updating the baseline probabilistic analysis results. Therefore, such “weighted” analysis produces an updated probability distribution of the dam’s factor of safety (FS) that reflects both prior uncertainty of model parameters and actual monitoring data. To illustrate the approach, a case study of a rockfill triaxial test specimen is analyzed: a baseline probabilistic analysis yields a mean FS ~1.7, whereas the weighted analysis incorporating monitoring data reduces the mean FS to ~1.5 and narrows the variability. The weighted analysis suggests less favorable conditions than the baseline projections. This methodology offers a transparent, computationally tractable route for embedding monitoring evidence into reliability calculations, producing more reflective safety estimates of actual dam behavior.
ISSN:2076-3417