Sandstone aquifers lacking mudstone interbeds are often considered homogeneous. However, depositional processes and diagenesis can generate significant permeability contrasts without notable variations in granulometry. Capturing this heterogeneity in flow models is essential but challenging. Outcrop studies provide valuable analogues for subsurface conditions, revealing sedimentary architectures and facies distributions not observable from borehole data alone. This study focuses on the Lower Gre`s Vosgien formation in eastern France, an important aquifer hosting lithium-rich geothermal brines in the Upper Rhine Graben, deposited by braided fluvial and aeolian systems during the Lower Triassic. This work aims to evaluate the influence of decimetre- to metre-scale sandstone features on advective transport modelling, through a two-dimensional conceptual model developed using a digital outcrop model and sedimentary facies analysis. Six progressively simplified hydrostratigraphic models are tested to assess how heterogeneity and its representation affect particle residence times, breakthrough curves, longitudinal macrodispersivity, and upscaled anisotropy. Results indicate that distinguishing fluvial- and aeolian-related sandstones significantly influences both horizontal and vertical advective flow. Failure to discretise individual fluvial facies and reliance on deterministic hydraulic conductivity values led to the underestimation of preferential flow pathways, delaying the prediction of first particle arrivals. Moreover, the discretisation between fluvial and aeolian-related deposits significantly affects macrodispersion results and yields considerable anisotropy when the conceptual model is upscaled. The strong coupling between advective flow with heat transfer and solute transport underscores the critical role of the observed sedimentary heterogeneity on the accurate understanding of lithium-rich geothermal brines circulation in the matrix porosity of the Lower Triassic sedimentary successions in the Upper Rhine Graben reservoirs.

Air relative permeability is a key parameter controlling gas transport in unsaturated porous media, yet its dependence on saturation history and pore structure remains insufficiently constrained by experimental data. In this study, air relative permeability was measured during controlled drainage and imbibition processes in homogeneous sands with contrasting grain-size distributions. Steady-state air-flow experiments were conducted in a laboratory column, and air permeability was determined from imposed gas-flow rates and measured pressure gradients. Water saturation was quantified independently from expelled-water volumes, allowing reproducible characterization of permeability–saturation relationships along defined hydraulic paths. The results reveal pronounced hysteresis in air relative permeability between drainage and imbibition, with strongly dependent behaviour on grain-size distribution. In the fine sand, air relative permeability decreases progressively with increasing water saturation, whereas in the coarse sand, it remains close to its maximum value over a wide saturation range during imbibition before collapsing abruptly near full saturation. Comparisons with permeability predictions derived from independently fitted water retention curves show significant discrepancies, particularly during imbibition, where saturation-based models fail to reproduce the observed permeability plateau and sharp decline. These findings demonstrate that air relative permeability in granular media cannot be described solely as a function of water saturation and that saturation history and pore structure exert a first-order control on gas transport. The experimental dataset provides a benchmark for evaluating and improving constitutive models for gas flow in unsaturated porous media.

Accurate determination of water relative permeability ( k rw ) is essential for modelling flow in porous media. Traditional methods often rely on bulk averaging and assumed unit hydraulic gradients, which limit spatial resolution and introduce uncertainties. This study presents a novel experimental column setup that integrates a PICO TDR probe for in situ water content measurement and dual pressure transducers for direct pressure gradient monitoring under steady-state flow conditions. Experiments were conducted on two water-wet quartz sands (P100 and P2040), which differed in grain sizes and porosities. A stepwise drainage and imbibition protocol enabled the derivation of both water retention curves and water relative permeability ( k rw – S w ) curves, using Darcy’s law and matching-point normalization. Initial saturation was achieved through upward imbibition using distilled water, preceded by a nitrogen injection to minimize air entrapment and accurately determine porosity. Each experiment was repeated under identical conditions to verify the reproducibility of the results. Results show minimal hysteresis between drainage and imbibition curves and excellent agreement with the Mualem–van Genuchten predicted k rw , using parameters independently derived from the water retention curves. The approach demonstrates high reproducibility and predictive accuracy, providing a strong platform for future studies involving pore-scale simulations, wettability alteration, or hysteresis modelling.

Understanding the influence of sedimentary processes on reservoir architectures can be essential for improving the prediction of permeability heterogeneity. The Upper Rhine Graben offers an ideal geological context for using outcrops as reservoir analogues, as the rocks cropping out at the Graben's shoulders belong to the same formations that host lithium-rich geothermal brine in the Graben. The interval of interest in this study, the upper section of the Lower Gre`s Vosgien Formation (LGV), consists of fluvio-aeolian deposits considered preferential zones for fluid migration within the Buntsandstein Group. In this study, the main factors controlling permeability heterogeneity distribution of these fluvio-aeolian deposits are presented, in different scales, by combining permeability and petrographic characteristics with quantitative sedimentological characterisation of the LGV. The fluvial channel facies association (FA), predominantly composed of cross-bedded sandstones deposited in braided channels, exhibits higher permeability compared to the wind- and water-laid FA, made up of sandstones deposited in a sand-sheet-dominated aeolian system. Vertical profiles reveal permeability contrasts of up to four orders of magnitude between these FAs. At the facies scale, the fluvial channel FA shows a distinction in permeability associated with facies formed under different flow-regime conditions. Trough cross-bedded sandstones, formed under lower flow-regime conditions, exhibit higher permeability than low-angle cross-bedded sandstones, formed mainly under upper flow-regime conditions. In the wind- and water-laid FA, permeability distribution is directly influenced by palaeoclimatic variations. Low-permeability hybrid sand sheets (HSS) were deposited under relatively humid conditions, while high-permeability aeolian dunes (AD) formed during more arid periods. Despite the high permeability of the AD, the architecture of the wind- and water-laid FA suggests that the AD have low connectivity potential. The lateral continuity of the wind- and water-laid FA, extending over hundreds of metres, combined with the dominant occurrence of low-permeability HSS, indicates that this facies association may act as flow baffles or barriers. Compaction is the main process influencing the permeability and porosity of the LGV. Samples with a higher proportion of lithoclasts and infiltrated clay recorded a more advanced degree of compaction. In contrast, samples with a higher percentage of quartz overgrowth were less compacted. The diagenetic overprint affected the distinct facies differently, indicating that primary sedimentary processes and architecture govern the distribution of permeability heterogeneity in the LGV. Permeability data from outcrops exhibited median values up to three orders of magnitude higher than those from subsurface data. However, the same facies and similar trends of relative porosity-permeability responses to the distinct sedimentary features are observed in both outcrops and borehole samples. This highlights sedimentology an essential resource for reservoir heterogeneity studies.

Controlled laboratory experiments were carried out using the hanging column method. Prior to the experiments, three uniform silica sands, which were originally water-wet, were aged in contact with crude oil until they were moderately oil-wet. Five fractionally wet sands were obtained by mixing the water-wet sands with oil-wet sands containing 25, 50 and 75 vol% oil-wet sands. A total of 11 tests formed the basis for the present study. The measured water retention curves showed that the capillary pressure heads were greatly reduced in sands that were oil-wet or fractionally wet. Changes in the wettability of the sands also affected their irreducible water saturation: The higher the proportion of oil-wet sand in the sand mixtures, the lower the irreducible water saturation. To quantify the characteristics of the measured water retention curves, the Van Genuchten model was used. The two optimized parameters seem to indicate a general trend: The higher the volume fraction of oil-wet sand, the higher is α and the lower is n. For the three unaged sands and the aged medium-sized sand, it was found that each of the two branches of the measured retention curves can be suitably scaled to a unique curve if, in addition to the petrophysical parameters (intrinsic permeability, porosity, surface tension, gas-water contact angle), the irreducible residual water saturation and the residual air saturation are taken into account. To quantify the observed deviations of the other two aged sands from the unified Leverett J-function, a theoretical fit function was used to match the experimental data of the three unaged sands. The experimental data sets for P2040ag and P100ag were found to be overestimated overall by the fit function. However, when the petrophysical parameters of the unaged sands were used instead of the actual measured parameters, the individual experimental Leverett J-functions came closer to the uniform J-curve. Based on this, it could be concluded that the apparent differences in pore structure between aged and unaged sands in addition to wettability, expressed by the cosine of the contact angle, may have contributed to a further reduction in the capillary pressure plateaus of the aged sands, which was particularly visible and significant in the P2040 and P100 sand. Using the measured static contact angles for two-phase gas-oil and oil-water systems and the measured interfacial tensions when the porous medium is either water-wet or moderately oil-wet, it was shown that the Bartell-Osterhof equation overestimates the measured gas-water contact angles. Reasonable agreement was achieved when a calibrated roughness factor of the solid surface was considered in the Young's contact angle.

The Arborea plain in Sardinia (Italy) is classified as a nitrate vulnerable zone (NVZ). In the present study, the individual work steps that are necessary to progress from the existing 3D hydrogeological model to a 3D numerical groundwater model using the interactive finite-element simulation system FEFLOW 7.4 are shown. The results of the transient flow model highlight the influence of the drainage network on the overall groundwater management: the total water volume drained by the ditches accounted for approximately 58% of the annual outflow volume. The numerical transport simulations conducted from 2012 to 2020 using hypothetical field-based nitrate input scenarios globally underestimated the high concentrations that were observed in the NVZ. However, as observed in the field, the computed nitrate concentrations in December 2020 still varied strongly in space, from several mg L−1 to several hundreds of mg L−1. The origin of these remaining local hotspots is not yet known. The modeling of rainfall fluctuations under the influence of climate change revealed a general long-term decline in the groundwater level of several tens of centimeters in the long term and, in conjunction with a zero-nitrate scenario, led to a significant decrease in nitrate pollution. Although hotspots were attenuated, the concentrations at several monitoring wells still exceeded the limit value of 50 mg L−1.

The Grombalia aquifer constitutes a complex aquifer system formed by shallow, unconfined, semi-deep, and deep aquifers at different exploitation levels. In this study, we focused on the upper aquifer, the Wadi El Bey coastal aquifer. To assess natural aquifer recharge, we used a novel physiography-based method that uses soil texture-dependent potential infiltration coefficients and monthly rainfall data. The developed transient flow model was then applied to compute the temporal variation in the groundwater level in 34 observation wells from 1973 to 2020, taking into account the time series of spatially variable groundwater recharge, artificial groundwater recharge from 5 surface infiltration basins, pumping rates on 740 wells, and internal prescribed head cells to mimic water exchange between the wadis and aquifer. The quantified deviations in the computed hydraulic heads from measured water levels are acceptable because the database used to construct a scientifically sound and reliable groundwater model was limited. Further work is required to collect field data to quantitatively assess the local inflow and outflow rates between surface water and groundwater. The simulation of 12 climate scenarios highlighted a bi-structured north—south behaviour in the hydraulic heads: an increase in the north and a depletion in the south. A further increase in the pumping rate would, thus, be severe for the southern part of the Wadi El Bey aquifer.

Heat transfer fluid mining represents a thermodynamic perturbation for geothermal reservoirs: The pumping of hot water coupled with the re-injection of colder water at depth favors the dissolution of some rock-forming minerals of the deep reservoir (e.g. feldspars), while promoting the precipitation of secondary phases, resulting in a possible change in the permeability and porosity of the reservoir. Such an impact is even greater when one considers the acid stimulations aimed at increasing the injectivity of the geothermal system. In that respect, no consensus exists in the literature regarding the impact of secondary phases on the dissolution rate of primary phases and therefore, on the sustained modification of pore structure. The present study aimed at shedding new light on these questions. Hydrothermal experiments of K-feldspar alteration were conducted at conditions relevant for the geothermal reservoir of Soultz-sous-Forêts (T = 180 °C, acidic pH domain). Measurements of cation release rates were combined with characterizations of secondary coatings (mineralogy, extent of coverage, thickness and porosity) to determine the reactivity of submillimeter K-feldspar powders with and without secondary precipitates. The formation of µm-thick boehmite coatings on K-feldspar grains was found to result in a modest decrease in its reactivity, which might be better explained by the presence of dissolved Al in the bulk solution. This result was independently confirmed by reactive transport simulations, which revealed that the impact of secondary coatings may become significant only when their thickness exceeds a few tens of microns, or if the dissolution rate of the primary phase is significantly greater (106 times) than that of orthoclase. Taken together, this study offers new constraints on the intricate interplay between dissolution and precipitation reactions, of prime importance for modeling more accurately the impact of mass transfer and porosity generation resulting from fluid circulation in geothermal reservoirs.

The standard liquid transport processes in porous media happens through a network of interconnected pore bodies and pore throats (here called the primary network). When a non-wetting phase displaces a wetting phase from a porous sample (drainage), thin films of the wetting phase are bound to be left on the surface of the constituting grains (for example when air displaces water from a porous rock, thin films of water are left behind, covering the rock grains). Under certain conditions, isolated liquid films can eventually merge, forming a secondary network of interconnected films and capillary bridges (see red arrows in the figure) that can effectively enhance the overall connectivity of the medium and act as a new pathway for fluid transport. We have performed experiments using transparent networks with the objective of studying transport processes that are enhanced by film flow. Our setup allow us to directly visualize the secondary network in the sample and we have shown how fluid bodies that are not linked via the primary network can actually be connected via the secondary network. This connection has important consequences for processes such as the dispersion of pollutants in soils and the transport of nutrients to plants in arid regions. ReferencesMoura, E. G. Flekkøy, K. J. Måløy, G. Schäfer and R. Toussaint, “Connectivity enhancement due to film flow in porous media,” Phys. Rev. Fluids 4, 094102 (2019).Moura, K. J. Måløy, E. G. Flekkøy, and R. Toussaint, “Intermittent dynamics of slow drainage experiments in porous media: Characterization under different boundary conditions,” Front. Phys. 7, 217 (2020).

A rare dataset of in-situ Be-10 from high-resolution depth profiles, soils, rock outcrops, and stream sediments is combined with geochemical analysis and modelling of regolith evolution to understand the variability of denudation rates in a mountain watershed (Strengbach critical zone observatory). High-resolution depth profiles are key to detect the presence of mobile regolith and to highlight how it affects the critical zone evolution. The modelling of regolith evolution and Be-10 concentrations along depth profiles allow us to estimate both the cosmic ray exposure age (19 kyr) and the mean denudation rate (22 mm kyr(-1)) of the regolith without any steady-state assumption on Be-10 concentrations. Comparison with maximum denudation rates inferred from topsoil samples collected from the surface of the depth profiles and calculated using the temporal steady-state assumption of Be-10 concentrations highlights an overestimation of denudation by a factor of two. Maximum spatially averaged denudation rates determined from stream sediment samples also likely overestimate denudation rates by a factor of two. These biases are significant for investigating the geomorphological evolution and we propose a method to correct denudation rates using the inherited Be-10 concentrations and the cosmic ray exposure age deduced from the high-resolution depth profiles. A key result is also that a steady state of Be-10 concentrations and a steady state of regolith thickness are two different equilibrium states that do not necessarily coincide. The comparison between locally corrected and spatially averaged denudation rates indicates that the watershed geomorphology is not in a topographic steady state but is modulated by regressive fluvial erosion. Nonetheless, our study demonstrates that even in a watershed where the steady-state assumption of Be-10 concentrations is not verified, the spatial variations of in-situ Be-10 concentrations in sediments still carry qualitatively relevant information on the geomorphological evolution of landscapes.

The PRIMA Sustain-COAST European project aims at exploring innovative governance for sustainable coastal groundwater management and pollution reduction in the context of a changing climate by involving researchers, local citizens, water stakeholders, and policy makers in interactive dialogue. Four study sites have been selected, among them the Wadi El-Bey watershed in Tunisia, located about 40 km south of Tunis. The study area is the Grombalia aquifer whose size is approximately 391 km². It is boarded to the north by the Gulf of Tunis and the Tekelsa Hills, to the east by the Abderrahman Mountain and the oriental coastal highlands, to the south by the Hammamet Hills, and to the west by the Bou Choucha and the Halloufa mountains. The Grombalia aquifer is bounded northward by the Mediterranean Sea and westward by the Gulf of Tunis. It constitutes a complex aquifer system formed by shallow unconfined, semi deep, and deep aquifers with different exploitation levels. The interest of the study relies on the upper aquifer. Surface flow occurs mainly in 5 wadis toward the north, reflecting regional topographic gradients. During the last few decades, the Grombalia shallow unconfined aquifer had been under stress by groundwater pumping due to the increasing population and development of agricultural and industrial activities. Recently, it has been noticed in some wells a rise in the level of the water table due to the abandonment of the exploitation of surface wells and to the irrigation by the water transferred from the north of the country, and considerable deterioration of groundwater quality due to saltwater intrusion and increased nitrate contamination as well as the organic matter in terms of COD. A groundwater numerical model for the Grombalia aquifer has been developed using Feflow 7.4 to simulate groundwater level changes under steady state and transient conditions. The steady state flow calibration was carried out using the water levels measured 1972 in 35 observation wells and then used as initial state of the Grombalia aquifer system. To show the influence of groundwater management, especially for agricultural activities, and interaction with surface water, measurements of water level, water temperature, pH, electric conductivity and water quality data (e.g., nitrate concentration) have been conducted during the 2020 field campaign, at selected monitoring wells and in neighbouring transects of surface water. The groundwater model constitutes a solid basis for further studies under transient flow and transport conditions to compare different water management, climate change and contamination scenarios, and is part of the calibrated multi-criteria decision supporting system developed in the PRIMA Sustain-COAST project context. References The project is funded by the General Secretariat for Research and Technology of the Ministry of Development and Investments under the PRIMA Programme. PRIMA is an Art.185 initiative supported and co-funded under Horizon 2020, the European Union’s Programme for Research and Innovation.

Coastal areas around the Mediterranean basin concentrate population, multi-sector economic activities and agricultural activities. This induces an important need in fresh water and high solicitation of coastal aquifers, which can lead to salt water intrusion. This issue, added to contaminated surface water percolating towards the aquifer, and along with climate change show the urge for innovative groundwater management, especially in coastal areas. The PRIMA Sustain-COAST European project aims at exploring innovative governance for sustainable coastal groundwater management and pollution reduction in the context of a changing climate by involving researchers, local populations, water stakeholders and policy makers. The Arborea plain in Sardinia (Italy) is characterized by an intense agricultural activity based on dairy cattle farming (approximately 31.000 livestock units in the district). The area, reclaimed from a lagoon in the 1920s, is intensely used for fodder crops to feed the cattle. Thus, an important drainage network has been developed to maintain the soil in suitable conditions for agriculture. Heterogeneous nitrates contamination of the aquifer system has been highlighted through soil sampling and groundwater monitoring in the Arborea plain in previous studies and the zone is classified as a Nitrates Vulnerable Zone (following Directive 91/676/CEE). The hydrogeology of the study site is characterized by two main aquifers: the upper one, unconfined, hosted in a sandy unit (SHU), separated from the second aquifer, hosted in an alluvial formation (AHU), by lagoon deposits aquitard. In the present study, we show the individual work steps to get from the existing 3D hydrogeological model to a 3D numerical groundwater model using the interactive finite-element simulation system Feflow 7.4. The developed partially unstructured steady-state flow model takes into account the recharge of the aquifer system by surface water, the drainage and irrigation network and the seasonal variation of water volumes drained and spread on the land. Also accounted for are water pumped by farms for technical use and livestock, groundwater flow between the different units and interactions with seawater. Results show the influence of groundwater management, especially for agricultural activities, and interaction with surface water, which is highly impacted by anthropic networks (irrigation and drainage). Ongoing research is aimed at quantifying the spatio-temporal distribution of nitrate in the SHU aquifer under transient groundwater flow conditions to compare different water management, climate change and contamination scenarios. References The project is funded by the General Secretariat for Research and Technology of the Ministry of Development and Investments under the PRIMA Programme. PRIMA is an Art.185 initiative supported and co-funded under Horizon 2020, the European Union’s Programme for Research and Innovation. We also acknowledge funding from the Italian Ministry of University and Research CUP no. J84D18000180005.

Little is known about the deep water circulation and the role of deep weathering processes on the Critical Zone (CZ) evolution. In this study, main fractures from four deep boreholes (0-120 m) and water collected from fractures were analyzed to improve our understanding of the deep CZ in a granitic headwater watershed (Strengbach watershed, France). Geochemical analysis indicates that chemical composition of deep water is clearly different than subsurface waters, with higher mean concentrations and stronger spatial variability across the watershed. Hydrogeochemical modeling results highlight the difference of functioning between subsurface system (springs, piezometers, and stream waters) and deep system (deep water from borehole fractures). The subsurface system is characterized by fairly homogeneous water flow, high mean pore velocities, and water-rock interactions in a porous regolith, while the deep water circulation behaves much more as independent systems along fractures, with low mean pore velocities, and variable mineralogy and hydrodynamic conditions. A state of chemical equilibrium can also be reached along fractures in the deep CZ for some primary minerals (biotite and K-feldspar), a feature never observed so far in the simulations of subsurface water chemical composition. Even if the deep water exhibits higher solute concentrations, significantly lower mean pore velocities inferred in the fractured zones imply that the deep water is responsible for a negligible part of the total hydrologic and weathering fluxes in the watershed (<1%). The limited role of the deep weathering processes compared to the subsurface processes suggests an up to bottom control of the CZ evolution at a millennial timescale. Our results also indicate that the majority of fresh water available for water supply at a human timescale originates from subsurface waters, as the deep water circulation is too low for being critical at a timescale relevant for societal needs in this type of geological context.

The predictions of two recent and two classical mathematical models are compared with experimentally measured three-phase relative permeabilities. Experimentally determined constitutive relationships in two-phase systems were used as model input parameters to numerically predict relative permeabilities (kr) in three-phase systems. Then the estimated results were compared with experimental three-phase permeabilities measured along decreasing water saturation/decreasing oil saturation/increasing gas saturation (DDI) paths. The results of the individual models to each of the three fluids involved (water, oil, and gas) were analyzed in detail. The simulated results showed that the Total Differential (TD) compatible model overestimates significantly both the global mobilities as well as the relative phase permeabilities in the three-phase system. There was improvement in the prediction with the TD compatible model when experimental data were used to locally impose the global mobility and fractional water and gas fluxes in the ternary diagram. Globally, the best prediction of the measured kr values was obtained with the so-called mechanistic model. However, its numerical implementation requires a preliminary calibration of the relative phase permeabilities in a three-phase system against experimental data along one DDI path to quantify the required six characteristic coefficients. In contrast to the TD compatible model, which by construction does not exhibit any numerical instabilities, elliptic zones in the water-oil (NAPL)-gas ternary diagram were identified in the mechanistic model.

This report is done in the framework of EU project Destress related to sustainability of geothermal exploitation. It is divided in three distinct parts. Part A done by UoS provides an overview of the experimental pre-stimulation tests done on an analog two micas granite to assess the permeability evolution of the Soultz GPK4 well during long term acid treatments with or without thermal treatments. A complementary modelling of acid stimulation has been performed using the KIRMAT code that uses single and double porosity models and account for a wide range of mineralogy. The model evaluates changes in porosity and permeability in the vicinity of the geothermal wells (up to 6m) based on changes in the amount of primary and secondary minerals and as a function of time. Part B done by ESG provides an update of the monitoring and investigations made onsite and especially on the injection well GPK-4 from the Soultz-sous-Forêts power plant (France), confirming MS28. The well hydraulic monitoring is presented. The injectivity index varies between 0.54 kg/s/bar and 0.65 kg/s/bar during the reporting period, from early 2017 to February 2020. In parallel to the hydraulic performance during exploitation, a chemical monitoring of the geothermal brine has been carried out since January 2018 by collecting geochemical data from the production well GPK-2 and the second reinjection well GPK-3. Several geochemical analyses have been done and confirm the high TDS and the gas content of the native brine. Part C done by GFZ and ESG, describes the soft chemical stimulation of the Soultz-sous-Forêts injection well GPK-4 from the concept to the operation (executed in December 2019), with observed environmental and hydraulic consequences and results. DESTRESS Demonstration of soft stimulation treatments of geothermal reservoirs 2 28.02.2020 Authors Part A from UoS: Patrick Baud, Jamie Farquharson, Alexandra Kushnir, Yann Lucas, Viet V. Ngo, Alain Clément, Bertrand Fritz, Gerhard Schäfer, Jean Schmittbuhl Part B & C from ESG: Régis Hehn, Nicolas Cuenot, Justine Mouchot, Vincent Maurer, Albert Genter, Olivier Seibel, Abigaelle Peterschmitt, Benoit Imbs Part C from GFZ: Thomas Reinsch, Ernst Huenges Lead Beneficiary UoS Type R report, document etc. OTHER software, technical diagram etc. DEM demonstrator, pilot etc. E ethics DEC website, patent filing etc. Status Draft WP manager accepted Project coordinator accepted Dissemination level PU Public CO Confidential: only for members of the consortium Contributors 1-GFZ 5-GES 9-GTL 13-SNU 2-ENB 6-TNO 10-UoS 14-KIC 3-ESG 7-ETH 11-TUD 15-ECW 4-UoG 8-GTN 12-NEX 16-WES Creation date 1.10.2019 Last change 25.02.2020 Version final Due date 29.02.2020 Submission date 28.02.2020 DESTRESS Demonstration of soft stimulation treatments of geothermal reservoirs Context and objectives DESTRESS is aimed at creating EGS (Enhanced Geothermal Systems) reservoirs with sufficient permeability, fracture orientation and spacing for economic use of underground heat. The concepts are based on experience in previous projects, on scientific progress and developments in other fields, mainly the oil & gas sector. Recently developed stimulation methods are adapted to geothermal needs, applied to new geothermal sites and prepared for the market uptake. The DESTRESS concept takes into account the common and specific issues of different sites, representative for large parts of Europe, and will provide a generally applicable workflow for productivity enhancement measures. The project mainly focuses on stimulation treatments with minimized environmental hazard (“soft stimulation”), to enhance the reservoir in several geological settings covering granites, sandstones, and other rock types. This deliverable was written in the framework of Task 4.3 (“Realization of chemical injection tests during long-term circulation (GPK-2/GPK-4 Soultz-sous-Forêts) & verification of injectivity index and permeability enhancement with low seismic nuisance (GRT-1 Rittershoffen)) of WP4 (“Demonstration of combined hydraulic-thermal-chemical treatments in sandstones, carbonate rocks and granites”). It compiles all the work done for the Soultz-sous-Forêts site with a focus on the chemical stimulation executed in December 2019. DESTRESS Demonstration of soft stimulation treatments of geothermal reservoirs A. Soultz-sous-Forêts: pre-stimulation laboratory tests and modeling 1 Laboratory-scale acid stimulation of Soultz-sous-Forêts granite 1.1 Background and sample selection The GPK-4 well was completed in 2004 and has been used variously for both production and injection. It reaches a vertical depth of 4982 m, but strong deviation means that the overall length of the well is ~5260 m. The open-hole section of the well is estimated to be around 200 °C, and intersects a finegrained two mica granite. The material recovered from the drilling operation is insufficient for the purposes of our experiments; instead, blocks of granite were collected from a site near Forbach in Germany, where the surface expression of the granite basement is accessible. The variably-altered granites contain muscovite and biotite—determined by X-ray powder diffraction (XRD) and energydispersive X-ray microanalysis (EDX), and appear to be an ideal analogue for the deep reservoir material at Soultz-sous-Forêts. Many of the granites were found to host macroscopic fractures. Although previous studies have identified fractures in the Soultz basement to host illite and calcite, we were unable to confirm the presence of either mineral through XRD or EDX analyses. Samples were divided into three suites for experimentation (shown in Figure 1): a fine-grained leucocratic 2-mica granite (G2M-A): a slightly hydrothermally altered granite containing additional secondary minerals such as apatite (G2M-B): and an unaltered granite identical to G2M-A, but containing abundant macroscopic fractures (G2M-F). DESTRESS Demonstration of soft stimulation treatments of geothermal reservoirs Figure 1: Examples of three classes of granite collected at Schwarzenbach. [above] sample photographs. [below] Scanning Electron Microscope images. Mineral phases identified using XRD and EDX: Qz = quartz; or = orthoclase; ab = albite; bt = biotite; ms = muscovite; ap = apatite. 1.2 Equipment and experiment design 1.2.1 Acid permeameter In order to explore the influence of aggressive permeants treatments on the two mica granite, over a range of relevant pressures and temperatures, a new permeameter has been custom-designed and constructed at Université de Strasbourg. The apparatus is comprised of a pressure vessel connected to a two-cylinder fluid pump. Confining pressure is applied by silicone thermofluid using a two-stage manual hand pump, up to a maximum of 70 MPa. The pressure vessel is wrapped in high temperature AMOXTM fabric tapes, and the whole ensemble is enclosed in a bespoke clamshell jacket. This jacket is composed of layers of Tempmat (mechanically bonded glass fiber matting), fiberglass cloth, and silicone-impregnated fiberglass cloth, which insulates the pressure vessel effectively due to the low thermal conductivity of the constituent materials. The AMOXTM tapes are connected to a programmable PID temperature controller. The controller reads temperature from a thermocouple embedded beneath the insulating jacket, and the power output to the heat tapes is adjusted accordingly in order to heat the pressure vessel to the desired (user-set) temperature. Within the pressure vessel, thermofluid surrounds a flanged seal, custom-moulded from temperatureand acidresistant rubber. The sample and two spacers are inserted into the seal. Another k-type thermocouple is in contact with the sample, in this case a mineral insulated k-type probe coated in Halar®, an acidresistant copolymer of ethylene and chlorotrifluoroethylene. Temperatures of the sample and vessel exterior are recorded using a National Instruments voltage input module integrated into a custombuilt data acquisition hub. The acquisition hub communicates with a programme written in LabVIEW, which is also used to send commands and receive flowrate, pore pressure, and fluid volume data from the Quizix pump. Figure 2 presents the details of our new permeameter designed and built for this project. DESTRESS Demonstration of soft stimulation treatments of geothermal reservoirs Figure 2: Schematic of acid permeameter. [a] Technical diagram of permeameter circuit. [b] Sketch of primary components. [c] Diagram of pressure vessel interior. (1): pressure vessel; (2): confining pressure inlet/outlet; (3) pore fluid outlet; (4): vent valve; (5): pore fluid inlet/outlet ports; (6) Quizix pore fluid pump; (7): permanent reservoir; (8) pressure gauge; (9): analogue pressure transducer; (10): confining pressure outlet valve; (11): confining pressure pump/oil reservoir; (12): effluent reservoir. (a): upstream endcap; (b): downstream endcap; (c): silicone gasket; (d): confining pressure inlet/outlet; (e) silicone gasket; (d) confining pressure inlet/outlet; (e) steel spacers; (f): sample; (g): annular sleeve. 1.2.2 Batch reaction tests Alongside the development of the acid permeameter, sample suites were immersed in acid solutions for varying periods of time. For the purposes of this study, two different concentrations of hydrochloric acid were prepared (0.2 N and 2.0 N HCl), by combining concentrated HCl with distilled and deionised water. Mass, porosity, and permeability of all samples were measured prior to immersion. Periodically, samples were removed and re-characterised in order to monitor the evolution of these physical properties. Additionally, some samples underwent thermal stressing in a furnace, whereby they were heated at 1 °C min-1 until a target temperature, left to dwell at that temperature for 2 hours, then cooled again at the same rate. A control suite of granites was set aside for mechanical testing (uniaxial compressive strength), against which to compare

We experimentally and numerically study the influence of gravity and finite‐size effects on the pressure‐saturation relationship in a given porous medium during slow drainage. The effect of gravity is systematically varied by tilting the system relative to the horizontal configuration. The use of a quasi two‐dimensional porous media allows for direct spatial monitoring of the saturation. Exploiting the fractal nature of the invasion structure, we obtain a relationship between the final saturation and the Bond number SnwF=Bo0.097 using percolation theory. Moreover, the saturation, pressure, and Bond number are functionally related, allowing for pressure‐saturation curves to collapse onto a single master curve, parameterized by the representative elementary volume size and by the Bond and capillary numbers. This allows to upscale the pressure‐saturation curves measured in a laboratory to large representative elementary volumes used in reservoir simulations. The large‐scale behavior of these curves follows a simple relationship, depending on Bond and capillary numbers, and on the flow direction. The size distribution of trapped defending fluid clusters is also shown to contain information on past fluid flow and can be used as a marker of past flow speed and direction.

This paper presents a new modelling approach to quantify the hydraulic diffusivity of low‐permeability unconsolidated porous media under confined saturated‐flow conditions in the laboratory. The derived analytical solution for the transient variation of the hydraulic head after flow interruption was applied to experimental data obtained from continuous measurements of the water pressure at two locations in the soil column. Three soil samples made of a mixture of natural bentonite (at different mass fractions) and medium sand were studied during a series of stepwise constant flow rates of water. The numerical results well fit the experimentally measured decrease of the dimensionless hydraulic head. The study shows that the increase of the mass fraction of bentonite in the soil sample from 10 to 30% is accompanied by a strong decrease of the hydraulic diffusivity from 2.4 × 10−2 to 1.1 × 10−3 m2 s−1, which is clearly due to the decrease of the hydraulic conductivity of the soil sample. The specific storages obtained for each of the three samples are in the same order of magnitude and seem to decrease with the increase of mass fraction of bentonite. However, they clearly reflect the predominant portion of the compressibility of the porous medium compared with that of water. Compared with reported literature values for clayey soils, the specific storage values in this study are slightly higher, varying within the range of 2 × 10−3 to 8.1 × 10−3 m−1.. The experimental results also give insight into the limitations of the modelling approach. In the case of low‐permeability soils (K < 2 × 10−6 ms−1) and steady‐flow conditions with low Reynolds numbers, for example, Re < 0.003, it is recommended to choose a time duration for flow interruption between subsequent flow rate steps of longer than 5 s. For high‐permeability porous media, to increase the precision of the quantified hydraulic diffusivity, it might be useful to select a measuring frequency significantly higher than 1 Hz.