Monitoring techniques for embankment dams: A study on ERT and IP measurements, and seepage modelling of the Älvkarleby test dam : A multi-method approach to investigating internal anomalies in an embankment dam

Abstract

Modern society depends on dams for renewable hydropower, water supply, and flood control. As climate change increases the frequency of extreme rainfall and flood events, and as the demand for renewable energy sources continues to grow, ensuring the safety of these critical structures becomes more urgent than ever. One of the most serious hidden threats in embankment dams is internal erosion – a slow process where fine particles are washed out from within the dam, potentially leading to leaks, sinkholes, or even failure. This process is difficult to detect using traditional visual inspections, which is why engineers are seeking innovative ways to monitor dam health.
The study is part of a Swedish research project exploring advanced geophysical techniques to detect internal erosion in embankment dams. The methods we tested are called Electrical Resistivity Tomography (ERT) and Induced Polarization (IP). These are non-invasive techniques that work like a kind of “medical scan” for dams. Electrodes placed in and around the structure measure how easily electric currents pass through the dam materials. By analysing both resistivity and chargeability (how materials temporarily store electric charge), we can generate 3D images of the dam's interior and gain insights into both moisture content and soil grain size—factors that are critical for assessing erosion and material integrity.
To test the effectiveness of ERT and IP in real-world conditions and compare them with other monitoring techniques, a 4 m high test embankment dam was constructed in Älvkarleby, Sweden. This dam contained six engineered defects (such as crushed rock zones and foreign material blocks) deliberately hidden within its core and filter zones. Their locations were unknown to the monitoring team, simulating the challenge of detecting unknown damage in an operational dam.
Between 7 500 and 14 000 ERT and IP data points were collected daily. These were processed using 3D inversion techniques to produce evolving models of resistivity and chargeability over time. The combined approach proved more powerful than either method alone.
The results were promising. Both ERT and IP successfully detected two of the five core defects – a horizontal and a vertical crushed-rock zone. A third defect, a concrete block in the core, was weakly indicated by ERT but with some positional error. One defect, a wooden block in the core, became visible only after three years of monitoring. The fifth core defect – a crushed-rock zone at the abutment – remained undetected, likely due to its small size and poor resolution in that specific area. This highlights a key challenge: resolution and electrode coverage are critical for successful detection, especially in full-scale applications. The fine filter defect was detected after two years of dam operation through both ERT and IP, appearing as a large anomalous zone with high resistivity and low chargeability.
In addition, the combined use of ERT and IP helped detect anomalous zones unrelated to the engineered defects. After the dam was dismantled, some of these anomalies were confirmed to be zones of internal erosion, further demonstrating the value of these methods.
Another important contribution of this research was the integration of geophysical data into seepage modelling. The ERT data significantly improved the characterization of dam materials and allowed for more precise delineation of water pathways. This, in turn, enhanced the accuracy of seepage models, leading to better predictions of potential leakage zones and critical weak points within the structure. By combining geophysical imaging with traditional seepage modelling, the study demonstrated how ERT can support more reliable dam safety evaluations.
Although not all defects were detected during the initial monitoring phase—and some only became apparent after several years—the results demonstrate that ERT, IP, and the integration of geophysical methods with seepage modelling are powerful tools for long-term dam monitoring and enhancing characterization of dam materials.

• author: Norooz, Reyhaneh
• supervisor: Torleif Dahlin; Sara Johansson
• opponent: Prof. Butler, Karl, University of New Brunswick, Canada.
• organization: Division of Engineering Geology
• publishing date: 2025-09-06
• type: Thesis
• publication status: published
• subject: Geotechnical Engineering and Engineering Geology; Geophysics
• keywords: Embankment dam, Monitoring techniques, Internal erosion, electrical resistivity tomography, Induced Polarization, Seepage modelling, Älvkarleby test dam
• defense location: Lecture Hall A:C, Building A, Klas Anshelms väg 16, Faculty of Engineering LTH, Lund University, Lund. The dissertation will be live streamed, but part of the premises is to be excluded from the live stream. Zoom: https://lu-se.zoom.us/s/61135986481
• defense date: 2025-11-06 09:00:00
• ISBN: 978-91-8104-671-7
• ISBN: 978-91-8104-672-4
• language: English
• LU publication?: yes
• id: 2092bc1a-c1ea-4c35-aeda-8417404fe352
• date added to LUP: 2025-09-30 14:08:06
• date last changed: 2025-10-07 03:22:01