Embankment dams is the most common type of dams in operation inSweden today. Due to the nature of embankment dams, seepage throughthem will always occur. If the seepage velocity exceeds a critical velocity,internal erosion is initiated, which could lead to damage in form of pipingand sinkholes. To treat this problem, remedial grouting has beenperformed involving a combination of conventional grouts, i.e. cement andcement-bentonite as well as sodium silicate, which is a chemical grout thatalso known as water glass. Regarding the sodium silicate grout, there isconcern about the long-term permanence.The aim of this thesis has been to study the potential performance ofsodium silicate grouted in embankment dams. The first part of this thesisis a literature review of the general behavior of sodium silicate as a grout,its degradation processes and the factors that could induce degradation.The second part suggests monitoring methods to control and evaluate theperformance of the treated dam and the grout if degradation has occurred.Findings from literature generally indicates a high risk of instability andlow permanence of sodium silicate when grouted in an embankment dam.This type of grout will undergo degradation mainly in two forms: syneresisinduced shrinkage and leaching due to grout erosion or dissolution. As thedegradation has developed, an increase in permeability of the repaireddam core is a potential consequence.How the potential degradation of sodium silicate will affect the treateddams is suggested to be observed by monitoring the permeability of thegrouted core. Applicable monitoring methods for this purpose aremeasurements of pore pressure and temperature using piezometers. Thesecond direct method of monitoring a changed dam behavior is suggestedto be leakage analysis, in order to detect potentially increased leakagebecause of the grout degradation. An indirect way to investigate the damperformance is suggested to be monitoring of the grout state. Measurementof ion concentration of sodium and silicon respectively in leakage waterusing selective-ion electrodes will reveal any increase in ion concentrationdue to the potential grout dissolution or leaching.

As many dams were built around 1950, the expected life span of these dams are about tobe reached. With this, the need for monitoring and increased understanding of the damsstructural integrity increases. In order to prevent failures, two warning signals are defined;alert and alarm. The main difference being that the first indicates an unexpected changein behavior that needs to be addressed and evaluated in the near future, while the otherrequires that immediate action must be taken to ensure the safety of the dam.This report aims to evaluate the applicability of different models for designing alert values.In order to achieve this goal, two case studies have been performed. The first being onSchlegeis, an arch dam in Austria, and the second Storfinnforsen, a concrete buttress damin Sweden. The methods used are finite element modelling as well as data-based models.Data-based models work on the presumption that the dam behaviour is governed by variationsin environmental conditions such as temperature and water level. The report hasevaluated two commonly used data-based models, hydrostatic thermal time (HTT) and hydrostaticseasonal time (HST), as well as two machine learning based models artificial neuralnetworks (ANN) and boosted regression trees (BRT).The programs used in this report are BRIGADE plus for finite element method and MATLABfor the multi-linear regression analyses HTT and HST, as well as boosted regressiontrees. The neural networks were constructed in Python using TensorFlow and Keras API.The result from the case studies is that the commonly used data-based models HST andHTT perform well enough for creating predictions and alert levels when given a sufficientamount of historical data, approximately 3-5 years. Machine learning such as artificial neuralnetworks while comparable in prediction quality does not further increase the understandingof the dam behaviour and can due to the complexity of designing an appropriate networkstructure be less suited for this type of analysis. Finite element models can also capturethe behavior of the dam rather well. It is however not as accurate as data-based modelswhen sufficient data is available. An FE-model should be used for definition of alert valueswhen insufficient data exists after the dam conditions have been significantly altered, orwhen newly constructed. The main advantage that machine learning provides is that theyperform better for non-linear behavior than multi-linear regression.

Grandes estruturas, sejam elas naturais ou artificiais, estão sujeitas a variações, em suas dimensões e posição, no espaço e no tempo. O monitoramento de estruturas está diretamente ligado com a segurança das mesmas, pois o colapso de uma estrutura artificial ou movimentação de estruturas naturais podem causar perdas econômicas, impactar o meio ambiente e ceifar vidas. Neste sentido o foco deste trabalho é aplicar a tecnologia GNSS (Global Navigation Satellite System) e redes geodésicas no monitoramento de estruturas. Para confirmar a ocorrência de deslocamentos utilizou-se o Teste de Congruência Global. Durante o desenvolvimento do trabalho foi proposto um método designado por Análise Geométrica, para fornecer indicação de possíveis deslocamentos. A metodologia proposta nesta pesquisa foi aplicada no monitoramento da barragem da Usina Hidrelétrica de Itaipu e pilares da sua rede de trilateração. Para o desenvolvimento da pesquisa foram realizadas quatro campanhas de observações. Os dados de cada campanha foram processados para determinar vetores, que posteriormente participaram de ajustamento vetorial pelo Método dos Mínimos Quadrados, para cada época de levantamento. O ajustamento forneceu coordenadas dos pontos da rede e a matriz variânvia-covariância, para cada época de observação. Estas informações foram combinadas duas a duas para aplicar a Análise Geométrica e o Teste de Congruência Global. Os resultados dos dois métodos de teste apresentaram boa correlação entre si e indicaram possíveis deslocamentos em pontos da rede de referência de monitoramento por trilateração e pontos da barragem. / Natural or artificial large structures are subjected to variations in their dimensions and position, in space and time. Structures monitoring is directly attended with their security, since the collapse of an artificial structure or displacement of natural structures can cause economic loss, impact the environment and cause the death of people. Hence, the aim of this study is to apply GNSS technologie (Global Navigation Satellite System) and geodetic networks in structures monitoring. In order to confirm the occurrence of the displacements the Global Congruence Test was employed. During the development of this study it was proposed a method so-called Geometric Analisys, to indicate possible displacements. The methodologie proposed in this research was applied to Itaipu hydro-electric power station and its trilateration networks. GNSS data was surveyed in four campaings. Data from each campaing were processed to determine vectors, which were posteriorly adjusted by means of the least squares method, for each survey epoch. The adjustment provided the coordinates of the network vertices and the covariance matrix, for each observation epoch. These informations were combined two by two to apply the Geometrical Analisys and the Global Congruence Test. Results from both test methods show good agreement and indicate possible displacements in vertices of the monitoring reference network and in object points of the dam.

[pt] Neste trabalho empregou-se a técnica de redes neurais artificiais e modelos de Box & Jenkins (1970) para análise, modelagem e previsão dos valores de vazão e de cargas de pressão na barragem Corumbá I, do sistema Furnas Centrais Elétricas, a partir dos dados de instrumentação disponíveis desde 1997. A previsão de valores prováveis pode auxiliar em tomadas de decisão durante a operação da barragem. A utilização de métodos estatísticos e de redes neurais artificiais é especialmente recomendado em situações onde a solução através de métodos determinísticos, analíticos ou numéricos, torna-se difícil por envolver modelagens tridimensionais, com condições de contorno complexas e incertezas na variação espacial e temporal das propriedades dos materiais que constituem a barragem e sua fundação. Tradicionalmente, as análises de séries temporais são normalmente abordadas sob a perspectiva de métodos estatísticos, como os modelos de Box & Jenkins. No entanto, redes neurais artificiais têm-se constituído ultimamente em uma alternativa atraente para investigações de séries temporais por sua capacidade de análise de problemas de natureza não-linear e não-estacionários. Neste trabalho são apresentadas três aplicações envolvendo o comportamento da barragem Corumbá I: previsão das vazões através da fundação junto à ombreira esquerda, previsão das cargas de pressão em piezômetros instalados no núcleo central da barragem e no solo residual de fundação e, finalmente, a previsão dos valores das leituras em um piezômetro supostamente danificado em determinado instante de tempo. Em todos estes casos, os resultados obtidos pelos modelos de Box & Jenkins e redes neurais artificiais foram bastante satisfatórios. / [en] In this work, artificial neural networks and the Box & Jenkins models (1970) were used for analysis, modeling and forecasts of water discharges and pressure head development in the Corumbá-I dam, owned by Furnas Centrais Elétricas, from the instrumentation data recorded since 1997. Prediction of the probable values can be a powerful tool for early detection of abnormal conditions during the dam operation. The use of statistical methods and artificial neural network techniques are specially recommend in situations where a solution with a deterministic approach, analytical or numerical, is difficult for involving three- dimensional modeling, complex boundary conditions and uncertainty with respect to the spatial and temporal variation of the material properties of the dam and its foundation. Time series analyses are traditionally carried out using a statistical approach, such as the Box & Jenkins models. However, artificial neural networks have become in the recent years an attractive alternative for time series problems due to their inherent ability to analyze nonlinear and non-stationary phenomena. Three applications of time series analysis, related to the instrumentation data collected from Corumba-I dam, are presented and discussed in this thesis: forecast of water discharges through the foundation near the dam left abutment, prediction of pressure heads in piezometers installed in the impermeable central core and the residual soil foundation and, finally, prediction of the pressure heads that would be read in a piezometer that, at a given instant of time, stops working being supposedly damaged. In all these cases, the results obtained from the Box & Jenkins models as well as the artificial neural networks are quite satisfactory.

Multi-purpose project requirements lead to construction of large dams. In order to maintain the desired safety level of such dams, comprehensive inspections based on use of a number of precise instruments are needed. The ideal dam instrumentation system should provide time-dependent information about critical parameters so that possible future behavior of the structure can be predicted. New dams are normally equipped with adequate instrumentation systems. Most of the existing dams, however, do not have adequate instruments or current instrumentation systems may not be in good condition. By implementing the modern equipment to existing dams, the uncertainty associated with the impacts of aging or unexpected severe external events will be reduced and possible remedial measures can be taken accordingly. This study summarizes the major causes of dam failures and introduces the instruments to be used to monitor the key parameters of a dam. The concept of the instrument retrofitting to an unmonitored dam is highlighted through a case study. A sample system is proposed for G&ouml / k&ccedil / ekaya Dam, with reference to an investigation of the current condition of the structure. The deficiencies observed during a site visit are listed and the corresponding rehabilitative repair measures are suggested. Finally, different alternatives of a new instrumentation system are introduced and compared in terms of technical and economical aspects.

Grandes estruturas, sejam elas naturais ou artificiais, estão sujeitas a variações, em suas dimensões e posição, no espaço e no tempo. O monitoramento de estruturas está diretamente ligado com a segurança das mesmas, pois o colapso de uma estrutura artificial ou movimentação de estruturas naturais podem causar perdas econômicas, impactar o meio ambiente e ceifar vidas. Neste sentido o foco deste trabalho é aplicar a tecnologia GNSS (Global Navigation Satellite System) e redes geodésicas no monitoramento de estruturas. Para confirmar a ocorrência de deslocamentos utilizou-se o Teste de Congruência Global. Durante o desenvolvimento do trabalho foi proposto um método designado por Análise Geométrica, para fornecer indicação de possíveis deslocamentos. A metodologia proposta nesta pesquisa foi aplicada no monitoramento da barragem da Usina Hidrelétrica de Itaipu e pilares da sua rede de trilateração. Para o desenvolvimento da pesquisa foram realizadas quatro campanhas de observações. Os dados de cada campanha foram processados para determinar vetores, que posteriormente participaram de ajustamento vetorial pelo Método dos Mínimos Quadrados, para cada época de levantamento. O ajustamento forneceu coordenadas dos pontos da rede e a matriz variânvia-covariância, para cada época de observação. Estas informações foram combinadas duas a duas para aplicar a Análise Geométrica e o Teste de Congruência Global. Os resultados dos dois métodos de teste apresentaram boa correlação entre si e indicaram possíveis deslocamentos em pontos da rede de referência de monitoramento por trilateração e pontos da barragem. / Natural or artificial large structures are subjected to variations in their dimensions and position, in space and time. Structures monitoring is directly attended with their security, since the collapse of an artificial structure or displacement of natural structures can cause economic loss, impact the environment and cause the death of people. Hence, the aim of this study is to apply GNSS technologie (Global Navigation Satellite System) and geodetic networks in structures monitoring. In order to confirm the occurrence of the displacements the Global Congruence Test was employed. During the development of this study it was proposed a method so-called Geometric Analisys, to indicate possible displacements. The methodologie proposed in this research was applied to Itaipu hydro-electric power station and its trilateration networks. GNSS data was surveyed in four campaings. Data from each campaing were processed to determine vectors, which were posteriorly adjusted by means of the least squares method, for each survey epoch. The adjustment provided the coordinates of the network vertices and the covariance matrix, for each observation epoch. These informations were combined two by two to apply the Geometrical Analisys and the Global Congruence Test. Results from both test methods show good agreement and indicate possible displacements in vertices of the monitoring reference network and in object points of the dam.

[pt] Em obras de maior porte e grande responsabilidade (portos, barragens, usinas nucleares, etc.), a quantidade de instrumentações pode se tornar suficiente para permitir a construção de modelos de variabilidade temporal das propriedades de interesse com base em redes neurais artificiais. No caso de barragens, o monitoramento através da instalação de um sistema de instrumentação desempenha um papel fundamental na avaliação do comportamento destas estruturas, tanto durante o período de construção quanto no período de operação. Neste trabalho empregou-se a técnica de redes neurais temporais (RNT) para análise, modelagem e previsão dos valores de vazão na barragem Funil, do sistema Furnas Centrais Elétricas, a partir dos dados de instrumentações disponíveis no período compreendido entre 02/09/1985 e 25/02/2002. As redes neurais temporais empregadas foram: RNT com arquitetura feedforward associada a técnica de janelamento, RNT recorrente Elman, RNT FIR e RNT Jordan. Adicionalmente, foram utilizadas duas técnicas para análise das séries temporais: os modelos de Box & Jenkins (1970) e métodos geoestatísticos, com a finalidade de comparar com o desempenho das RNT´s. Nesta pesquisa estuda-se ainda a geração de intervalos de confiança para RNT e para métodos geoestatísticos. As previsões de vazão analisadas neste trabalho, envolvendo o comportamento da barragem Funil, apresentaram resultados satisfatórios tanto os obtidos pelos modelos de redes neurais temporais como pelos de Box & Jenkins e métodos geoestatísticos. / [en] In works of great responsibility (ports, dams, nuclear power, etc.), the amount of instrumentation data may allow the construction of models for the temporary variability of the properties of interest based on neural network techniques. In case of dams, the monitoring through the installation of an instrumentation system plays a fundamental part in the evaluation of the behavior of these structures, during the construction period as well as in the operation period. In this work the technique of temporal neural networks (TNN) was used for analysis, modeling and forecast of the water discharges values in the Funil dam, from Furnas Centrais Elétricas system, starting from the data of available instrumentation in the period between 02/09/1985 and 25/02/2002. The temporal neural networks used in this research were the following: TNN with feedforward architecture and the windowing technique, recursive TNN Elman, TNN FIR and TNN Jordan. Two additional techniques (Box & Jenkins and geostatistical models) were employed for analysis of the time series with the purpose to compare the results obtained with neural networks. In this research the generation of confidence intervals for TNN and geostatistical methods were also investigated. The discharge values forecasts analyzed in this work for the Funil dam presented satisfactory results, with respect to the neural network, Box & Jenkins and geostatistical methods.

Hydropower is an important energy source since it is fossil-free, renewable, and controllable. Characteristics that become especially important as the reliance on intermittent energy sources increases. However, the dams for the hydropower plants are also associated with large risks as a dam failure could have fatal consequences. Dams are therefore monitored by several sensors, to follow and evaluate any changes in the dam. One of the most important dam surveillance measurements is seepage since it can examine internal erosion. Seepage is affected by several different parameters such as reservoir water level, temperature, and precipitation. Studies also indicate the existence of a time lag between the reservoir water level and the seepage flow, meaning that when there is a change in the reservoir level there is a delay before these changes are reflected in the seepage behaviour. Recent years have seen increased use of AI in dam monitoring, enabling more dynamic warning systems.  This master’s thesis aims to develop a model for dynamic warning signals by predicting seepage using reservoir water level, temperature, and precipitation. Furthermore, a snowmelt variable was introduced to account for the impact of increased water flows during the spring season. The occurrence of a time lag and its possible influence on the model’s performance is also examined. To predict the seepage, three models with different complexity are used – linear regression, support vector regression, and long short-term memory. To investigate the time lag, the linear regression and support vector regression models incorporate a static time lag by shifting the reservoir water level data up to 14 days. The time lag was further investigated using the long short-term memory model as well.  The results show that reservoir water level, temperature, and the snowmelt variable are the combination of input parameters that generate the best results for all three models. Although a one-day time lag between reservoir water level and seepage slightly improved the predictions, the exact duration and nature of the time lag remain unclear. The more complex models (support vector regression and long short-term memory) generated better predictions than the linear regression but performed similarly when evaluated based on the dynamic warning signals. Therefore, linear regression is deemed a suitable model for dynamic warning signals by seepage prediction.