Geothermal energy is the energy naturally present inside the earth crust. When a large volume of hot water and steam is trapped in subsurface porous and permeable rock structure and a convective circulating current is set up, it forms a geothermal reservoir. A geothermal system can be defined as - convective water in the upper crust of earth, which transfers heat from a heat source (in the reservoir) to a heat sink, usually the free surface. A geothermal system is made up of three main elements: a heat source, a reservoir and a fluid, which is the carrier that transfers the heat. As an alternative source of energy geothermal energy has been under attention of the researchers for quite some time. The reason behind this is the existence of several benefits like clean and renewable source of energy which has considerable environmental advantage, with no chemical pollutants or wastes are generated due to geothermal emissions, and the reliability of the power resource. Hence research has been directed in several directions like exploration of geothermal resources, modeling the characteristics of different types of geothermal reservoirs and technologies to extract energy from them. The target of these models has been the prediction of the production of the hot water and steam and thus the estimation of the electricity generating potential of a geothermal reservoir in future years. In a geothermal power plant reinjection of the heat depleted water extracted from the geothermal reservoir has been a common practice for quite some time. This started for safe wastewater disposal and later on the technology was employed to obtain higher efficiency of heat and energy extraction. In most of the cases a very small fraction of the thermal energy present in the reservoir can be recovered without the reinjection of geothermal fluid. Also maintaining the reservoir pressure is essential which gradually reduces due to continuous extraction of reservoir fluid without reinjection, especially for reservoirs with low permeabilities. Although reinjection of cold-water has several benefits, the possibility of premature breakthrough of the cold-water front, from injection well zone to production well zone, reduces the efficiency of the reservoir operation drastically. Hence for maintaining the reservoir efficiency and longer life of the reservoir, the injectionproduction well scheme is to be properly designed and injection and extraction rates are to be properly fixed. Modeling of flow and heat transport in a geothermal reservoir due to reinjection of coldwater has been attempted by several researchers analytically, numerically and experimentally. The analytical models which exist in this field deal mostly with a single injection well model injecting cold-water into a confined homogeneous porous-fractured geothermal reservoir. Often the thermal conductivity is neglected in the analytical study considering it to be negligible which is not always so, as proved in this study. Moreover heterogeneity in the reservoir is also a major factor which has not been considered in any such analytical study. In the field of numerical modeling there also exists a need of a general coupled three-dimensional thermo-hydrogeological model including all the modes of heat transport (advection and conduction), the heat loss to the confining rocks, the regional groundwater flow and the geothermal gradient. No study existing so far reported such a numerical model including those mentioned above. The present study is concerned about modeling the non-isothermal flow and heat transport in a geothermal reservoir due to reinjection of heat depleted water into a geothermal reservoir. Analytical and numerical models are developed here for the transient temperature distributions and advancement of the thermal front in a geothermal reservoir which is generated due to the cold-water injection. First homogeneous geothermal aquifers are considered and later heterogeneities of different kinds are brought into picture. Threedimensional numerical models are developed using a software code DuMux which solves flow and heat transport problems in porous media and can handle both single and multiphase flows. The results derived by the numerical models have been validated using the results from the analytical models derived in this study. Chapter 1 of the thesis gives a brief introduction about different types of geothermal reservoirs, followed by discussion on the governing differential equations, the conceptual model of a geothermal reservoir system, the efficiency of geothermal reservoirs, the modeling and simulation concepts (models construction, boundary conditions, model calibration etc.). Some problems related with geothermal reservoirs and geothermal power is also discussed. The scenario of India in the context having a huge geothermal power potential is described and different potential geothermal sites have been pointed out. In Chapter 2, the concept of reinjection of the heat depleted (cold) water into the geothermal reservoir is introduced. Starting with a brief history of the geothermal reinjection, the chapter describes the purpose and the need of reinjection of geothermal fluid giving examples of different geothermal fields over the world where reinjection has been in practice and benefitted by that. The chapter further discusses on the problems and obstacles faced by the geothermal projects resulting from the geothermal reinjection, most important of which is the thermal-breakthrough and cooling of production wells. Lastly the problem of this thesis is discussed which is to model the transient temperature distribution and the movement of the cold-water thermal front generated due to the reinjection. The need of this modeling is elaborated which represents the motivation of taking up the problem of the thesis. Chapter 3 describes an analytical model developed for the transient temperature in a porous geothermal reservoir due to injection of cold-water. The reservoir is composed of a confined aquifer, sandwiched between rocks of different thermo-geological properties. The heat transport processes considered are advection, longitudinal conduction in the geothermal aquifer, and the conductive heat transfer to the underlying and overlying rocks of different geological properties. The one-dimensional heat transfer equation has been solved using the Laplace transform with the assumption of constant density and thermal properties of both rock and fluid. Two simple solutions are derived afterwards, first neglecting the longitudinal conductive heat transport and then heat transport to confining rocks. The analytical solutions represent the transient temperature distribution in the geothermal aquifer and the confining rocks and model the movement of the cold-water thermal front in them. The results show that the heat transport to the confining rocks plays an influential role in the transient heat transport here. The influence of some parameters, e.g. the volumetric injection rate, the longitudinal thermal conductivity and the porosity of the porous media, on the transient heat transport phenomenon is judged by observing the variation of the transient temperature distribution with different values of the parameters. The effects of injection rate and thermal conductivity have been found to be high on the results. Chapter 4 represents another analytical model for transient temperature distribution in a heterogeneous geothermal reservoir underlain and overlain by impermeable rocks due to injection of cold-water. The heterogeneity of the porous medium is expressed by the spatial variation of the flow velocity and the longitudinal effective thermal conductivity of the medium. Simpler solutions are also derived afterwards first neglecting the longitudinal conduction, then the heat loss to the confining rocks depending on the situation where the contribution of them to the transient heat transport phenomenon in the porous media is negligible. Solution for a homogeneous aquifer with constant values of the rock and fluid parameters is also derived with an aim to compare the results with that of the heterogeneous one. The effect of heat loss to the confining rocks in this case is also determined and the influence of some of the parameters involved, on the transient heat transport phenomenon is assessed by observing the variation of the results with different magnitudes of those parameters. Results show that the heterogeneity plays a major role in controlling the cold-water thermal front movement. The transient temperature distribution in the geothermal reservoir depends on the type of heterogeneity. The heat loss to the confining rocks of the geothermal aquifer also has influence on the heat transport phenomenon. In Chapter 5 another analytical model is derived for a heterogeneous reservoir where the heterogeneous geothermal aquifer considered is a confined aquifer consisted of homogeneous layers of finite length and overlain and underlain by impermeable rock media. All the different layers in the aquifer and the overlying and underlying rocks are of different thermo-hydrogeological properties. Results show that the advancement of the cold-water thermal front is highly influenced by the layered heterogeneity of the aquifer. As the cold-water thermal front encounters layers of different thermo-hydrogeological properties the movement of it changes accordingly. The analytical solution derived here has been compared with a numerical model developed by the multiphysics software code COMSOL which shows excellent agreement with each other. Lastly it is shown that approximation of the properties of a geothermal aquifer by taking mean of the properties of all the layers present will lead to erroneous estimation of the temperature distribution. Chapter 6 represents a coupled three-dimensional thermo-hydrogeological numerical model for transient temperature distribution in a confined porous geothermal aquifer due to cold-water injection. This 3D numerical model is developed for solving more practical problems which eliminate the assumptions taken into account in analytical models. The numerical modeling is performed using a software code DuMux as mentioned before. Besides modeling the three-dimensional transient temperature distribution in the model domain, the chapter investigates the regional groundwater flow has been found to be a very important parameter to consider. The movement of the thermal front accelerates or decelerates depending on the direction of the flow. Influence of a few parameters involved in the study on the transient heat transport phenomenon in the geothermal reservoir domain, namely the injection rate, the permeability of the confining rocks and the thermal conductivity of the geothermal aquifer is also evaluated in this chapter. The models have been validated using analytical solutions derived in this thesis. The results are in very good agreement with each other. In Chapter 7 the main conclusions drawn from the study have been enlisted and the scope of further research is also pointed out.

MESHWR / Department of Hydrology and Water Resources / Lebalelo area of Sekhukhune district is one of many areas in South Africa experiencing portable water scarcity, especially during prolonged dry season. Due to the dominance of low yielding aquifers in South Africa, it is essential to manage groundwater resources in these low yielding aquifers. However, the management of low yielding aquifer is difficult in areas like Labelelo where the hydrogeological characteristics of the aquifers are understudied. This study investigated the hydrogeological characteristics of the aquifers in the area using combined geophysical method and analytical groundwater models. Four newly drilled borehole and five existing boreholes were used for this study. Geophysical survey was carried out using magnetic and electromagnetic methods. The magnetic survey was used to locate the position of magnetic bodies such as dolerite dykes and different lithologies with different magnetic properties. The electromagnetic survey however, was used to determine zones of high permeability associated with the intrusive bodies as well as high permeability zones in fault planes. Step test, constant discharge test and recovery tests were conducted on all the boreholes to stress the borehole. This was used to determine a suitable and sustainable pumping rate of the aquifer. Pumping test data from the pumping period and recovery was evaluated and interpreted using AQTESOLVE. Aquifer transmissivity, storativity, internal and external hydraulic boundaries were determined from the data. The transmissivity in the area ranges from 0.08 to 124.7 m2/day. The aquifer types in the area are double porosity aquifer, radial flow aquifer with single porosity. Inductive Coupled Plasma (ICP-MS) was used to measure heavy metals, trace metals and cations while Ion Chromatography (IC) was used to determine anions in groundwater of the study area. The groundwater in the area is dominated by calcium carbonate as a result of long residence time with dolomite. The hydrochemistry of the water indicates that the chemistry of the groundwater in the area is mainly controlled by rock-water interaction. / NRF

Hydrogeological hazards are quite diffuse rainfall-induced phenomena that affect mountain regions and can severely impact these territories, producing damages and sometimes casualties. For this reason, hydrogeological risk reduction is crucial. Mitigation strategies aim to reduce hydrogeological risk to an acceptable level and can be classified into structural and non-structural measures. This work focuses on enhancing some non-structural risk mitigation measures for mountain areas: debris-flow rainfall thresholds, as a part of an Early Warning System (EWS), multivariate rainfall scenarios with multi-hazard mapping purpose and public awareness. Regarding debris-flow rainfall thresholds, an innovative calibration method, a suitable uncertainty analysis and a proper validation process are developed. The Backward Dynamical Approach (BDA), a physical-based calibration method, is introduced and a threshold is obtained for a study area. The BDA robustness is then tested by assessing the uncertainty in the threshold estimate. Finally, the calibrated threshold's reliability and its possible forecast use are assessed using a proper validation process. The findings set the stage for using the BDA approach to calibrate debris-flow rainfall thresholds usable in operational EWS. Regarding hazard mapping, a multivariate statistical model is developed to construct multivariate rainfall scenarios with a multi-hazards mapping purpose. A confluence between a debris-flow-prone creek and a flood-prone river is considered. The multivariate statistical model is built by combining the Simplified Metastatistical Extreme Value approach and a copula approach. The obtained rainfall scenarios are promising to be used to build multi-hazard maps. Finally, the public awareness within the LIFE FRANCA (Flood Risk ANticipation and Communication in the Alps) European project is briefly considered. The project action considered in this work focuses on training and communication activities aimed at providing a multidisciplinary view of hydrogeological risk through the holding of courses and seminars.

Dans les Landes de Gascogne, le fonctionnement de la nappe phréatique superficielle plio-quaternaire est très liée au fonctionnement de la végétation subjacente. Pour comprendre la variabilité à l'échelle du paysage de la productivité de ces forêts, il est important de passer de l'échelle de la parcelle à celle du bassin versant. La mise en place d'un dispositif de suivi de type bassins versants jumelés à l'échelle de petits bassins versants forestiers situés dans le Massif Landais : les bassins versants du Tagon (Biganos) et du Bouron (Belin-Beliet) a permis le suivi hydrologique et piézométrique de la nappe plio-quaternaire. Le dispositif inclut une modification du couvert végétal par l'occurrence de deux tempêtes successives à 10 ans d'intervalle. La collecte de données cartographiques, numériques et statistiques a permis une reconstitution de l'usage des terres et des classes d’âges des peuplements forestiers, pour la période 1984- 2010 sur les deux bassins versants. Cette reconstitution a ainsi permis de retrouver l'évolution des stocks de carbone et de biomasse des peuplements pour les deux sites expérimentaux et de quantifier l'impact d'une tempête sur l'organisation paysagère des peuplements (âge et variabilité). Le lien entre les forêts de Pin maritime et la nappe phréatique plio-quaternaire a été mis en évidence par : (1) l'étude de la croissance des arbres pour des sites où la profondeur de la nappe était contrastée, au travers d'une étude dendrologique et (2) l'analyse d'oscillations journalières sur les séries chronologiques des fluctuations demi-horaires du niveau phréatique, qui peuvent être attribuées à l’évapotranspiration journalière. La compréhension des liens entre variables hydroclimatiques et quantité de bois produite, i.e. largeur de cerne, de manière rétrospective, a permis d'identifier les facteurs de variabilité et les facteurs limitants pour chaque type de site, en lien avec la profondeur de la nappe superficielle. L’analyse des chroniques piézométriques a mis en lumière l'importance des prélèvements journaliers dans la vidange de l'aquifère. La dernière partie de ce travail pose les bases d'une future modélisation couplée végétation-nappe pour le bassin versant du Bouron. La reconstitution spatiale du fonctionnement de la nappe phréatique a été entreprise en régime stationnaire puis en régime dynamique avec le modèle NEWSAM de l'Ecole des Mines de Paris. Les facteurs expliquant le fonctionnement de l'aquifère et les flux d'eau à l'échelle de la petite région sont identifiés. L'effet de la recharge, issue du modèle de végétation pour le Pin maritime GRAECO, est en particulier testé, via une étude de sensibilité, en relation avec les différences de recharge induites par les différents types de couverts, et en particulier, l'âge des peuplements. Les points clefs à améliorer sont identifiés afin de perfectionner les modèles utilisés, et de reproduire de façon réaliste le fonctionnement de petits bassins versants dans la région. / In the Landes de Gascogne area, the functioning of the shallow plio-quaternary water-table is tightly linked with the functioning of the top vegetation. In order to understand the variability of forest productivity at the landscape scale, there is a need to move forward the stand scale to the watershed scale. Part of this study was to setup hydrological equipment of paired-watershed, representative of the forested plain in the Leyre watershed, considering water-table level and hydrologic measurement. The selected sites were the Bouron watershed (Belin-Beliet) and the Tagon watershed (Biganos), which were contrastingly impacted by two storms within the last 10 years. The land-cover cartography was obtained with satellite, in-situ and statistical data, on a retrospective basis, accross the 1984-2010 period. Those maps give the evolution of the land use and the Pinus pinaster stand age. The reconstitution of the age structure thought time, converted to carbon and biomass stock, are showing the impact of a storm, comparing the control watershed to the damaged one. The link between forests and water-table was explored on : (1) the growth of trees developing in sites with contrasted water-table levels, by way of dendrological measurement and (2) the analysis of diurnal oscillation on the water-table level signal, attributed to vegetation removal. The understanding of the link between hydroclimatic data and wood production, i.e., ring width, shows that limitating factors should not be the same across different sites. In particular, water-table deeper than 3 meter is not a determinant resource for growth, but nutriment may be more limitating at those sites. On the other side, discharge of the water-table cannot be dissociated to evapotranspiration reprise. Oscillations in the diurnal signal of water-table level shows that this removal can be determinant, and deeper than expected according to the vegetation cover. The last part of this work build the basis of a future model, representing the coupling between vegetation and water-table at the watershed scale. The dynamic of the aquifer was simulated in permanent and transient state with the NEWSAM hydrogeological model (Ecole des Mines, Paris). Different processes were identified as key factor to explain water-table functioning, such as recharge effect, zone vadose impact and surface drainage. The recharge effect, coming from GRAECO, was particularly tested, thought a sensitivity analysis, in relation with growing proportion of young forest stand (less than 5 years old). Points to improve and refine are identified in order to perfect the two models used here.