kth.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Performance of data-based models for early detection of damage in concrete dams
KTH, School of Architecture and the Built Environment (ABE), Civil and Architectural Engineering, Concrete Structures.ORCID iD: 0000-0003-2594-4107
KTH, School of Architecture and the Built Environment (ABE), Civil and Architectural Engineering, Concrete Structures.ORCID iD: 0000-0003-3586-8988
KTH, School of Architecture and the Built Environment (ABE), Civil and Architectural Engineering, Concrete Structures.ORCID iD: 0000-0001-8336-1247
2020 (English)In: Structure and Infrastructure Engineering, ISSN 1573-2479, E-ISSN 1744-8980Article in journal (Refereed) Published
Abstract [en]

A failure of a massive concrete dam could cause catastrophic consequences. The purpose of monitoring is to detect anomalies and damage at an early stage to prevent failure. Data-based models for anomaly detection are based on the hypothesis that the behaviour of an undamaged dam will follow an expected pattern, and deviation from this pattern is an indication of damage. In this study, simulations were used to create time series for an undamaged dam and three different damage scenarios at three different locations in the dam body. Three common data-based models were used to predict a dams crest displacements, both on the generated artificial data and the corresponding measurements from the dam. Prediction bands for future measurements were created, and the ten time-series were used to test the ability to detect damage. All models could detect instantaneous damage but struggle to detect progressive damage; the Neural network outperforms the two regression models. The choice of the mathematically optimal threshold limit leads to a large number of false alerts. Requiring three consecutive values outside the threshold before an alert is issued, increases the possibility to receive an early alert compared to the standard approach where observations are classified individually.

Place, publisher, year, edition, pages
Taylor and Francis Ltd. , 2020.
Keywords [en]
anomaly detection, concrete dam, Dam safety, data analysis, false alarms, monitoring, statistical models, Concrete dams, Concretes, Data reduction, Regression analysis, Time series, Catastrophic consequences, Different damages, Massive concrete, Optimal threshold, Progressive damage, Regression model, Damage detection
National Category
Civil Engineering
Identifiers
URN: urn:nbn:se:kth:diva-274273DOI: 10.1080/15732479.2020.1737146ISI: 000519329200001Scopus ID: 2-s2.0-85081265852OAI: oai:DiVA.org:kth-274273DiVA, id: diva2:1452703
Note

QC 20200707

Available from: 2020-07-07 Created: 2020-07-07 Last updated: 2024-01-10Bibliographically approved
In thesis
1. A systems approach to ice loads on concrete dams
Open this publication in new window or tab >>A systems approach to ice loads on concrete dams
2022 (English)Doctoral thesis, comprehensive summary (Other academic)
Abstract [en]

Dams are mainly used for the storage of water to electricity production and irrigation, or for river regulation. Continuous work to assure a high dam safety level is a prerequisite to minimize the risk for the uncontrolled release of water. An essential part of the safety evaluation of concrete dams is to understand the loads they are exposed to and the expected response of the dam. Under normal conditions, the behaviour of concrete dams is, to a great extent, governed by the ambient variation in temperature and water level. For concrete dams in cold climates, the large variation in ambient temperatures between summer and winter is particularly significant. In addition, these dams may be subjected to a pressure load from the expansion or movement of an ice sheet on the reservoir. The current guidelines for these ice loads are based on the dam's location and state that concrete dams must be designed for a line load of 50-250 kN/m. Thus, the ice load constitutes a significant part of the total load, especially for small dams. Despite its relatively significant impact, the knowledge about ice loads is insufficient, and the magnitude and return period of ice loads constitute one of the greatest uncertainties during stability evaluations of concrete dams. Furthermore, an apparent contradiction is that measurements and models indicate that ice loads are higher than the recommended values. Simultaneously, there are no reported dam failures where the ice-load has been addressed as the reason for the breach.

To increase the knowledge about ice loads and the structural behaviour of concrete dams, this thesis applies an approach where the ice and the dam are parts of a structural system. The thesis contains six studies investigating the dam's, the ice's or the system's response to external loads.  Studies of the dam are aimed at increasing the understanding of the normal behaviour of concrete dams. Studies of ice loads include measurements, and a major contribution from this project is the development of a 1$\times$3 m$^2$ ice load panel, the installation of the panel on a concrete dam, and subsequent measurements during six winters. In addition, a systematic review and meta-analysis of previous measurements have been performed. The studies of the different parts have been connected through two studies of the entire system. The first is a parameter study where the static interaction between ice and dam is simulated to quantify how geometric variations of the reservoir and ice affect the mechanical ice loads. The second study searches for detectable influence from the ice load in the measured behaviour of concrete dams.

The results show that the structural behaviour of concrete dams under normal conditions is primarily governed by the variation in water level and temperature. For the studied dams, these effects are significantly larger than the influence from damage and degradation. Ice load measurements and simulations show that ice loads varies significantly along the dam. This variation makes it difficult to quantify the impact of external factors on the magnitude of ice loads. Despite this difficulty, three independent analyses show that ice thickness, water level change, and the dam's properties have an evident effect on the magnitude of ice loads. If current guidelines are to be updated to consider local conditions at the dam, these three parameters should be included. Ice loads of the magnitudes measured and specified in the current guidelines should have a notable impact on the behaviour of a dam during normal operation. However, such an impact has not been found in the eight dams studied within this project. This result indicates that the ice loads measured locally do not necessarily represent the global ice load that acts on the entire structure.

Abstract [sv]

Dammar används främst för att lagra vatten för elproduktion och bevattning eller för flödesreglering. Ett kontinuerligt arbete med att säkerställa en hög dammsäkerhetsnivå är en förutsättning för att minimera risken för ett okontrollerat utsläpp av vatten från dessa dammar. En viktig del av säkerhetsutvärdering av betongdammar är förståelse för de laster som dessa utsätts för samt kunskap om dammens förväntade respons. Under normala förhållanden påverkas en betongdamms strukturella beteende främst av variationer i vattennivån och de omgivande temperaturerna. För betongdammar i kallt klimat är den stora variationen i temperatur mellan sommar och vinter särskilt betydande. Dessutom kan dessa dammar utsättas för ett tryck från expansion och rörelse hos det istäcke som kan bildas på magasinet. För betongdammar utgör storleken på denna islast en betydande osäkerhet. De nuvarande riktlinjerna är geografiskt baserade och anger att betongdammar ska dimensioneras för en linjelast med storleken 50-250 kN/m. En sådan storlek innebär att islasten, särskilt för små dammar, utgör en betydande del av den totala lasten. Trots dess relativt stora påverkan är kunskapen om islaster låg och utgör för tillfället en av de största osäkerheterna vid utvärdering av betongdammars stabilitet. Det finns dessutom en motsättning där mätningar och modeller visar att islasten är högre än de rekommenderade värdena, samtidigt som det inte finns några rapporterade dammhaverier orsakade av islast.

I denna avhandling behandlas isen och dammen som ett system för att på sätt öka förståelse kring både islaster och betongdammars beteende under normal drift. Avhandlingen innehåller sex enskilda studier som studerar dammens, isens eller systemets respons till yttre variationer. Dammen har studerats genom transienta analyser i syfte att öka förståelsen kring, och upptäcka avvikelser i betongdammars normalbeteende. Isen har studerats genom mätningar där ett stort bidrag från detta projekt är utvecklingen av en 1$\times$3m$^2$ islastpanel, installationen av denna panel på en betongdamm, och efterföljande mätningar under sex vintrar. Dessutom har en systematisk sammanställning och analys av tidigare mätningar genomförts. Dessa studier har sammanlänkats genom två studier av hela systemet. Den första av dessa är en parameterstudie där den statiska interaktionen mellan is och damm simuleras i syfte att kvantifiera hur geometriska variationer hos vattenmagasinet och isen påverkar den mekaniska islasten. Den andra studien innefattar transienta analyser av betongdammars beteende med syfte att detektera inverkan från islast i det uppmätta beteende hos betongdammar.  

Variation i vattennivå och temperatur har större påverkan på det strukturella beteendet under normaldrift hos de studerade dammarna än irreversibla förändringar orsakade av nedbrytningsprocesser och skador. Islastmätningar och simuleringar visar att islasten varierar betydande längs med dammen. Detta försvårar kvantifierandet av inverkan från yttre faktorer på islastens storlek. Trots detta visar tre oberoende analyser att istjocklek, vattennivåförändring och dammens egenskaper har en signifikant påverkan på islastens storlek. Om nuvarande riktlinjer ska uppdateras för att beakta lokala förhållanden vid dammen bör dessa parametrar inkluderas. Islaster av de storlekar som uppmätts och anges i de nuvarande riktlinjerna bör ha en tydlig påverkan på dammens beteende vid normaldrift. En sådan påverkan har dock inte observerats hos de åtta dammar som studerats i detta projekt. Det ger en indikation om att de islaster som mäts lokalt inte nödvändigtvis är representativa för den globala islasten som verkar på hela dammen.

Place, publisher, year, edition, pages
Stockholm: KTH Royal Institute of Technology, 2022. p. 136
Series
TRITA-ABE-DLT ; 221
Keywords
Ice loads, Ice pressure, Dam safety, Concrete dams, Cold regions, Monitoring, Measurements
National Category
Infrastructure Engineering
Research subject
Civil and Architectural Engineering, Concrete Structures
Identifiers
urn:nbn:se:kth:diva-306937 (URN)978-91-8040-115-9 (ISBN)
Public defence
2022-02-09, F3, Lindstedtsvägen 26, KTH Campus, Videolink - Password 627271 - https://kth-se.zoom.us/j/66770434327?pwd=NkZRMk9KdWU2S0U5dmtTVERsL05rZz09, Stockholm, 13:00 (English)
Opponent
Supervisors
Note

QC 20220111

Available from: 2022-01-11 Created: 2022-01-11 Last updated: 2022-06-25Bibliographically approved

Open Access in DiVA

No full text in DiVA

Other links

Publisher's full textScopushttps://www.tandfonline.com/doi/full/10.1080/15732479.2020.1737146

Authority records

Hellgren, RikardMalm, RichardAnsell, Anders

Search in DiVA

By author/editor
Hellgren, RikardMalm, RichardAnsell, Anders
By organisation
Concrete Structures
In the same journal
Structure and Infrastructure Engineering
Civil Engineering

Search outside of DiVA

GoogleGoogle Scholar

doi
urn-nbn

Altmetric score

doi
urn-nbn
Total: 418 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf