Accurate characterization of the subsurface thermal environment (STE) is crucial for sustainable geothermal energy and urban heat management, yet it remains challenged by pronounced subsurface heterogeneity at the city scale. Conventional approaches, which primarily rely on deterministic numerical modeling based on sparse observations, fail to capture the heterogeneity and spatial distribution of the subsurface thermal environment (STE). Here, we introduce an integrated modeling framework that synergistically combines geostatistics, physics-based modeling, and Bayesian inference to reconstruct the STE under steady-state thermal conditions and quantify its uncertainty. Applied to Changzhou in the densely populated Yangtze River Delta, our framework probabilistically reveals the distribution of key factors (lithology, upper temperature, basal heat flow, thermal conductivity, Darcy flux, and subsurface temperature). A key finding from the sensitivity analysis identifies the upper temperature boundary as the dominant source of modeling uncertainty. We estimate the shallow geothermal potential of Changzhou to be approximately 2.7 × 104 GWh, sufficient to meet its winter heating demand and highlighting its potential as a sustainable energy source for the Yangtze River Delta. Our scalable and uncertainty-aware framework offers a pathway for the global assessment and management of subsurface thermal resources.

This work systematically examines the state-of-the-art in the modelling of pipe-ground heat interactions in buried pipes of thermal source networks (TSNs), i.e., low-temperature district heating and cooling networks that use decentralised heat pumps and often uninsulated distribution pipes. A structured database search and screening process (2010-present), complemented by natural-language search and citation tracing, identified 56 simulation studies. Most apply simplified assumptions (43), while fewer use transient ground representations (8) or 2D/3D and multiphysics models (5). The findings highlight a clear distinction between insulated and uninsulated TSNs. Insulated networks typically aim to minimise ground interaction and can therefore be represented using simplified methods. Uninsulated networks, in contrast, require a transient representation of the ground to capture dynamic thermal interaction with the surrounding ground, even at low operating temperatures. Reported comparisons indicate that simplifying ground dynamics can bias key effects by over 40%. Meanwhile direct validation of pipe-ground exchange remains rare. On this basis, the review provides practical guidance for model selection across design, operation, optimisation, and control, recommending transient ground models for most uninsulated networks. Additionally, the review highlights a recurring modelling inconsistency: advanced ground models are often applied for geothermal sources, while the connected distribution pipes are simplified, despite exhibiting a similar ground-coupled behaviour. Extending consistent ground modelling to distribution pipes would improve simulation accuracy. Overall, transient ground modelling should be treated as a baseline for credible simulation of uninsulated TSNs, while wider adoption is currently limited primarily by the scarcity of field-validated pipe-ground datasets.

ABSTRACT The ecological responses of urban groundwater ecosystems to anthropogenic pressures are poorly understood. Historic hand pumps offer a simple and promising access point to this hidden habitat and may facilitate the detection of groundwater fauna. This study provides a first overview of the stygofauna of Leipzig, Germany, and associated anthropogenic influences, based on investigations of historic hand pumps and groundwater monitoring wells. Statistical comparisons between sampling methods were conducted to assess potential selective effects. Leipzig's groundwater is characterized by distinct physicochemical conditions, notably elevated temperatures and reducing environments, indicating disturbed yet characteristic features of an urban aquifer in Central Europe. A total of 4500 stygofaunal individuals representing eight taxonomic groups were recorded, displaying pronounced spatial heterogeneity. Although three chemical parameters showed minor associations with faunal occurrence, these relationships should be interpreted with caution. Overall, the urban stygofauna reflects a complex interplay of natural and anthropogenic drivers. Significant differences in both physicochemical properties and faunal assemblages were observed between sampling methods. These differences likely arise from spatial factors and methodological constraints. Monitoring wells appear to alter in situ habitat conditions, limiting the representativeness of sampled fauna, while hand pumps exhibit taxon‐specific selectivity, resulting in biased community composition. Consequently, direct comparability between these sampling approaches appears to be limited.

The utilization of seasonal thermal energy storage (sTES) systems is essential for balancing fluctuations between demand and surplus of heating/cooling in modern energy systems and to reduce overall greenhouse gas emissions from space heating. However, large storage volumes are required to store the heat over extended periods leading to a high demand for construction materials and processes. Yet, no comprehensive environmental evaluation compares sTES technologies across their life cycle phases. This study employs life cycle assessment to quantify the environmental impacts of three different type of sTES: a tank thermal energy storage (TTES), a water-gravel thermal energy storage (WGTES), and a pit thermal energy storage (PTES). Aquifer thermal energy storage (ATES) systems are also included as reference for evaluating the results. Greenhouse gas emissions from the construction phase vary between 1.4 (PTES) and 29.4 g CO2-eq/kWhth (WGTES), depending on the type of installation, storage size, and construction materials. Utilizing water as a filling material and large storage volumes with reduced surface-to-volume ratios enhance environmental performance. Controversely, materials such as concrete, steel, foam glass gravel, and related transport processes contribute significantly to the environmental impact. These should be replaced wherever possible by sustainable alternatives without compromising storage capacity and efficiency.

Aquifer heterogeneity and temperature-dependent fluid properties are key controls on heat transport in aquifer thermal energy storage systems, yet their combined influence on performance reliability in low-temperature doublets remains insufficiently quantified. We evaluate a three-dimensional confined-aquifer model using 50 stochastic hydraulic-conductivity realizations, four density–viscosity coupling schemes, and warm-well injection temperatures ranging from 14 to 30 °C, while the cold-well injection temperature is fixed at 6 °C. Performance is assessed using thermal-recovery percentiles, uncertainty range, and a robustness index that integrates mean performance, variability, and a minimum acceptable recovery threshold. Simulations reveal a pronounced asymmetry between the two wells. Warm-well recovery remains relatively stable across the investigated temperature range, whereas cold-well recovery deteriorates strongly with increasing warm-well injection temperature. The 10th-percentile cold-well thermal recovery decreases from approximately 61

Accurate prediction of heat transport is fundamental to the design and performance assessment of underground thermal energy storage (UTES) systems. Most groundwater heat-transport models assume local thermal equilibrium (LTE) between the solid matrix and pore water, implying instantaneous interfacial heat exchange and a single temperature field. However, recent experimental and modelling studies show that this assumption can break down under conditions commonly encountered in permeable and heterogeneous aquifers.Here we present a synthesis of a multi-scale research programme that identifies when and why local thermal non-equilibrium (LTNE) becomes relevant for subsurface heat transport, spanning grain-scale laboratory experiments and field-scale numerical modelling. At the grain scale, laboratory experiments resolving solid and fluid temperatures independently demonstrate that rate-limited interfacial heat exchange results in persistent solid–fluid temperature differences for coarse grains and elevated Darcy velocities representative of UTES operation. These effects are governed primarily by grain size and solid thermal properties and cannot be captured by LTE formulations.At the aquifer scale, three-dimensional stochastic simulations show that LTNE-like behaviour can emerge even when pore-scale LTE holds, solely due to hydraulic conductivity heterogeneity. Preferential advection along high-permeability pathways accelerates thermal fronts, while delayed heat diffusion into low-permeability domains leads to effective thermal retardation that deviates fundamentally from predictions based on volumetric heat capacity. This field-scale LTNE depends systematically on the variance and correlation length of hydraulic conductivity and the thermal Péclet number.Together, these results reveal a continuum of LTNE behaviour across scales: grain size controls interfacial heat exchange at the pore scale, while hydraulic conductivity heterogeneity governs delayed heat uptake at the aquifer scale. Ignoring either mechanism can potentially bias predictions of thermal plume migration, retardation, and heat recovery efficiency, with direct implications for UTES modelling, performance assessment, and design reliability.

Cavern thermal energy storage (CTES) is a technological variant of underground thermal energy storage that relies on flooding of subsurface cavities or tunnels for long-term heat storage. Such installations are crucial, particularly for conserving excessive solar and waste heat from the warm season to be used during the cold season of the year. This review provides, for the first time, a comprehensive synopsis of different types of these large installations, which are most prominent in Scandinavian countries. It is revealed that CTES can be distinguished between those that are pre-designed and those that are re-engineered former infrastructures (commonly oil reserves), whereas single and multi-cavern systems exist. Based on existing and planned CTES implementations, characteristic design parameters are identified, and a structured insight into common criteria for an optimal layout is provided. While it is most cost-efficient to reuse existing facilities and apply storage volumes of far more than 100,000 m3, a perfect geometric layout needs to account for controlled thermal stratification, attuned aspect ratio, and optimal area-tovolume ratio. One crucial factor is the long-term geo-mechanical stability of the ambient rock mass, which ideally represents compact crystalline rock with negligible groundwater flow. Our study summarizes the findings of existing installations that may serve as blueprints for planning, constructing, operating, and monitoring new CTES, including hot water and pressurized storage concepts.

ABSTRACT Subterranean ecosystems in urban environments are exposed to multiple anthropogenic stressors that compromise groundwater quality and biodiversity. Rising groundwater temperatures, oxygen depletion, hydraulic disturbance and chemical contamination are among the main pressures. Complex geological settings further modulate subsurface conditions and habitat suitability. Yet the combined effects of urban and geological factors on groundwater fauna remain poorly understood. This study examines the relationships between environmental conditions and the diversity and abundance of major faunal groups along gradients of urbanisation and geological gradients in 91 monitoring wells of the city of Halle (Germany) and its surroundings. Repeated measurements revealed an urban–rural gradient with rising groundwater temperature and dissolved organic carbon (DOC) concentrations towards more urbanised areas. Despite generally low dissolved oxygen (DO) levels, fauna occurred even under hypoxic conditions. Faunal abundance and diversity were primarily governed by redox‐sensitive variables. Groundwater from wells lacking fauna showed elevated DOC concentrations, increased groundwater temperature, greater depth to the groundwater table and lower DO concentrations. Urbanisation indicators, that is, elevated DOC, temperature and potassium, are negatively correlated with faunal metrics, particularly under the combined effects of stressors. The aquifer type alone exerts only a minor influence. All major taxonomic groups were present in both porous and fractured aquifers. Spatial variability in faunal composition was more pronounced between eastern and western sectors than along the urban–rural gradient, highlighting complex interactions between geological and urban factors in shaping groundwater biodiversity.

In this study, we investigate the geothermal potential of the shallow subsurface in Dresden, Germany. The analysis considers the status quo scenario in which accumulated heat can be recycled. Installing all possible geothermal systems based on the available space, this heat could supply Dresden's residents for 3 years with energy for space heating. However, a fair CO2 price would have to be implemented to improve economic value. Next, a near-future scenario is studied, in which accumulated heat has been recycled and, considering all technical constraints, the annual heat input provides a sustainable potential that can provide up to 4.5% of annual heating demands (HDs). However, there is a very high spatial variability that is studied in regard to its socio-economic implications. Finally, two far-future scenarios (SSP245 and SSP585) are studied to understand the effect of climate change on the suitability of geothermal systems. Depending on the scenario and circumstances, up to 82% of the city's climate neutrality targets might be reached.This article is part of the theme issue 'Urban heat spreading above and below ground'.

Accurate temperature prediction is crucial for optimizing the performance of borehole heat exchanger (BHE) fields. This study introduces an efficient Bayesian approach for improving the forecast of temperature changes in the ground caused by the operation of BHEs. The framework addresses the complexities of multi-layer subsurface structures and groundwater flow. By utilizing an affine invariant ensemble sampler, the framework estimates the distribution of key parameters, including heat extraction rate, thermal conductivity, and Darcy velocity. Validation of the proposed methodology is conducted through a synthetic case involving four active and one inactive BHE over five years, using monthly temperature changes around BHEs from a detailed numerical model as a reference. The moving finite line source model with anisotropy is employed as the forward model for efficient temperature approximations. Applying the proposed methodology at a monthly resolution for less than three years reduces uncertainty in long-term predictions by over 90%. Additionally, it enhances the applicability of the employed analytical forward model in real field conditions. Thus, this advancement offers a robust tool for stochastic prediction of thermal behavior and decision-making in BHE systems, particularly in scenarios with complex subsurface conditions and limited prior knowledge.

Aquifers under urban areas are highly impacted by human activity and altered in terms of thermal, chemical, and also ecological conditions. In particular for ecological conditions, the causes and implications of changes in faunal communities for groundwater management and use are not yet fully understood. At the same time, large and dense urban clusters, such as the city of Berlin, Germany, rely on water supply from groundwater and other sources within their city limits. The aim of the CHARMANT project is therefore to develop a groundwater management approach specifically designed for the complex, multifaceted conditions in the urban underground that incorporates assessment of groundwater ecosystems and thermal management of the subsurface. Long-term changes in the thermal subsurface conditions are evaluated based on repeated measurements of temperature-depth profiles, which show an increase in warming down to 100m. Likewise, warming near the surface (20 m below ground level) is spreading from the city centre towards the suburban areas, due to increased surface sealing, subsurface infrastructure and climate change. Frequent occurrence of groundwater fauna, i.e. stygophile and stygobiont species, is found to be limited to locations in the Berlin-Warsaw glacial valley in central Berlin or in the vicinity of surface waters (approx. 11 % of all measurement wells). At the same time, some of the regularly sampled wells exhibit rare mass events with hundreds or even thousands of fauna individuals, which are not linked to changes in abiotic groundwater parameters. Also, for the specific case of Berlin, occurrence groundwater fauna appears to be constraint mostly due to low contents of dissolved oxygen linked to natural hydrogeological conditions. Overall, these heterogeneous conditions make quantitative assessment of the ecological status based on existing approaches difficult.The thermal state of the subsurface of Berlin is further assessed by thermo-hydraulic modelling that aims at identifying areas with similar groundwater conditions, so-called archetypes, whilst taking groundwater temperature as a proxy for overall anthropogenic impact. In the future, these groundwater archetypes will be linked to chemical conditions, e.g. presence of typical urban contaminants, as well as the ecological status, e.g. presence of specific groundwater fauna, in order to obtain groundwater use types. These use types represent 3D, spatially-resolved conceptual models, that facilitate the integration of aspects of spatial planning above the surface as well as different regulatory frameworks. Furthermore, the project aims at using the simplified representation of complex subsurface processes in these archetypes for communicating groundwater management strategies and enhancing awareness and active participation of citizen and other stakeholders with the aim of minimizing conflicts of groundwater use.

Heat transport in porous media is crucial for gaining Earth science process understanding and for engineering applications such as geothermal system design. While heat transport models are commonly simplified by assuming local thermal equilibrium (LTE; solid and fluid phases are averaged) or local thermal non-equilibrium (LTNE; solid and fluid phases are considered separately), heat transport has long been hypothesized, and reports have emerged. However, experiments with realistic grain sizes and flow conditions are still lacking in the literature. To detect LTNE effects, we conducted comprehensive laboratory heat transport experiments at Darcy velocities ranging from 3 to 23 m d−1 and measured the temperatures of fluid and solid phases separately for glass spheres with diameters of 5, 10, 15, 20, 25, and 30 mm. Four replicas of each size were embedded at discrete distances along the flow path in small glass beads to stabilize the flow field. Our sensors were meticulously calibrated, and measurements were post-processed to reveal LTNE, expressed as the difference between solid and fluid temperature during the passing of a thermal step input. To gain insight into the heat transport properties and processes, we simulated our experimental results in 1D using commonly accepted analytical solutions for LTE equations and a numerical solution for LTNE equations. Our results demonstrate significant LTNE effects with increasing grain size and water flow velocity. Surprisingly, the temperature differences between fluid and solid phases at the same depth were inconsistent, indicating non-uniform heat propagation likely caused by spatial variations in the flow field. The fluid temperature simulated by the LTE and LTNE models for small grain sizes (5–15 mm) showed similar fits to the experimental data, with the RMSE values differing by less than 0.01. However, for larger grain sizes (20–30 mm), the temperature difference between fluid and solid phases exceeded 5 % of the system's temperature gradient at flow velocities ≥17 m d−1, which falls outside the criteria for the LTE assumption. Additionally, for larger grain sizes (≥20 mm), the LTNE model failed to predict the magnitude of LTNE (i.e., temperature difference between fluid and solid phase in time series) for all tested flow velocities due to experimental conditions being inadequately represented by the 1D model with ideal step input. Future studies should employ more sophisticated numerical models to examine the heat transport processes and accurately analyze LTNE effects, considering non-uniform flow effects and multi-dimensional solutions. This is essential to determine the validity limits of LTE conditions for heat transport in natural systems such as gravel aquifers with grain sizes larger than 20 mm.

Worldwide shallow groundwater is increasingly exposed to anthropogenic impacts. The thermal state of this important resource is affected not only by global warming but also by various local structures that release heat into the subsurface. This additional heat can accumulate and lead to local hotspots or - mostly urban - areas of elevated groundwater temperatures. The consequences of this warming for groundwater quality and ecology are widely unknown. Groundwater ecosystems are embedded in a naturally relatively stable environment, where temperature changes can affect the highly specialized, cold-stenotherm invertebrate community and meso- to psychrophilic microorganisms. In this study, we examine whether and how a groundwater temperature hotspot impacts groundwater ecology. We identified such a thermal anomaly in Hockenheim, Germany, caused by a water park with heated swimming pools and basements. The thermal impact was monitored over the course of a year by temperature data loggers in nine wells – four upstream and downstream of the structure each and one inside the basement. The same wells were sampled for chemical and microbiological parameters, such as the microbial total cell count and the cellular ATP content, as well as groundwater fauna. We additionally tested three wells in a nearby forest to obtain reference values that are mostly unaffected by anthropogenic interference. The measurements were repeated every three months in order to account for seasonal variations. The preliminary results show a local heat plume and an increase in groundwater temperatures by up to 8 K. However, there is no significant deterioration in the ecological parameters. Regarding the fauna, which generally shows low abundance due to oxygen depletion in the study area, we observed only a minor decrease within the thermally affected zone. Finally, the outcome of this study will improve our understanding of the vulnerability of groundwater ecosystems in the context of subsurface warming.

Urban areas are expanding worldwide. Consequent changes to the environment, such as surface sealing, underground structures, and alterations to vegetation, do not only affect the surface but also impact the subsurface. Groundwater, a crucial natural resource for humankind, an essential water resource to plants, and a huge aquatic ecosystem, is potentially threatened by these changes. In particular, the hydraulic conditions and the water quality in shallow urban aquifers are being altered significantly. This study focuses on the shallow aquifer of the city of Munich, Germany, examining it in terms of urban land use classes (dense and discontinuous sealed surfaces, parks, forests, and agricultural sites) and typical physical and chemical measures used in standardised groundwater monitoring. In addition, the groundwater quality assessment is complemented by selected microbiological indicators. The objective of the study was to evaluate if physical, chemical and microbiological variables are suited for a routine groundwater monitoring in an ecological manner and if they show the same distribution patterns in an urban environment. Groundwater below the five land use categories distinguished showed differences in variables such as chloride concentration, temperature, and total bacterial cell counts. However, the considerable natural variation in the depth of the groundwater table across the city partly masked the effect of urban land use on groundwater hydrochemistry. Bacterial activity in shallow urban groundwater, measured as cellular ATP concentrations, on average, was in the range of clean surface waters rather than nearnatural groundwater. Concentrations of dissolved organic carbon (DOC) and nutrients, on the other hand, were overall low. In summary, Munich's shallow gravel aquifer mirrors an energy-limited, oligotrophic ecosystem. Strong correlations were observed between bacterial cell counts and DOC concentrations, with groundwater temperature being a significant influencing factor alongside the concentration of major ions. Some of the physical-chemical variables and microbiological measures exhibited variations between two sampling seasons, as well as between well water and pumped groundwater. In conclusion, the consideration of land use classes provided useful information on the impact of urbanisation on groundwater hydrochemistry and microbiology. Both sets of criteria sensitively indicated deviations of the urban to rural groundwater characteristics. Finally, seasonal effects and differences in the type of water sampled need to be considered for the setup of a routine integrative monitoring scheme.

Modeling heat transport in saturated porous media typically assumes local thermal equilibrium (LTE) conditions, though this assumption lacks justification. Recent work has revealed local thermal non-equilibrium (LTNE) effects for groundwater flow conditions, which standard one-dimensional analytical and numerical models fail to capture accurately. In this study, we develop and validate a 2D numerical model for two-phase heat transport at the granular scale to describe experimental LTNE effects previously observed, by coupling heat fluxes in both phases with a heat transfer term. Our results show that LTNE and non-uniform flow effects are superimposed and challenging to disentangle. However, the experimental results, expressed as temperature difference between solid and fluid phase , best match the case where the heat transfer coefficient (maximal efficiency), showing that is insensitive for flow rates of 3-23 m and grain sizes of 5-30 mm. The model further confirms that different and negative for same grain sizes are caused by non-uniform flow where arrival of the thermal front varies at the grain scale. Overall, our findings reveal multiple different heat transport concepts: (a) LTE, which is widely used; (b) Baseline LTNE, arises from different thermal properties between phases; (c) Phase transfer LTNE, which involves a limited heat transfer rate between phases; (d) Non-uniform flow LTNE, which is caused by pore-scale flow variations. This detailed concept of LTNE effects suggests that interphase heat transfer plays a negligible role toward LTNE effects at the granular scale under conditions relevant for hydrogeology.

The subsurface urban heat island (UHI) effect can provide latent clean geothermal potentials for cities. Understanding the city-wide subsurface temperature evolution under different land surfaces is significant in making better use of geothermal energy. This research presents a study of Nanjing to identify the city-wide temperature distribution and evolution characteristics and further estimates the geothermal potential in Nanjing. Low-cost satellite-measured temperatures were used to derive the subsurface temperatures through a liner regression correction method, with higher accuracy verified by measured borehole data. The simulation results indicate that the concrete surface exhibits higher average temperatures than the grassland surface, resulting in relatively higher subsurface temperatures. The deviations of simulated subsurface temperatures are attributed to many factors, including the influence of complex atmospheric conditions on satellite-measured temperature accuracy, land surface heat absorption, and infiltration in the shallower layer. Furthermore, it reveals that the urban areas have 14.7% greater geothermal potential compared to rural areas, due to the subsurface UHI effect. This study provides a potentially efficient and convenient method for the estimation of potential urban geothermal energy.

Precise prediction of heat transport in porous media holds crucial significance in Earth Sciences for diverse applications, ranging from the design of geothermal systems to utilizing heat as a tracer in aquifers. Traditionally, the description of heat transport has been simplified by assuming local thermal equilibrium (LTE), where the temperature within the fluid and solid phases in the representative elementary volume is presumed to reach an instant equilibrium. In reality, assuming two distinct phases coupled by heat transfer across their surface area describes the complete physics and is referred to as the local thermal non-equilibrium (LTNE) model. While earlier research delved into the theoretical aspects of LTNE effects, a notable gap exists due to the absence of experimental data to elucidate the heat transport mechanism in porous aquifers containing natural grains. To address this gap and investigate LTNE on a granular scale, we conducted systematic flow-through experiments employing porous media containing glass spheres with distinct grain sizes of 5, 10, 15, 20, 25 and 30 mm. Each sphere contained a small temperature sensor for the solid temperature, accompanied by sensors in the adjacent pore space to measure the fluid temperature. Our findings revealed that the temperature difference between two phases grows with increasing grain size and flow velocity ranging from 9 to 61 m d-1, thereby highlighting qualitative LTNE effects in relation to grain size and flow velocity. To further enhance our understanding, we used a numerical model to investigate the heat transfer coefficient, fitting the LTNE model to the experimental data. These simulation results indicated evidence of non-uniform flow which we included into our model to estimate its effects on heat transport. This comprehensive approach contributes valuable insights into the intricate interplay of LTNE effects, grain sizes, and flow velocity, advancing our understanding of heat transport in natural porous media.

Modelling heat transport in porous aquifers is generally based on the assumption of an instantaneous local thermal equilibrium (LTE) between the solid and the fluid phase. Previous studies have revealed that this assumption can be violated, e.g. in the presence of fast or preferential flow causing delayed heat diffusion into the grain structure matrix, referred to as local thermal non-equilibrium (LTNE). However, conditions and scales at which LTNE effects should be taken into account in natural heterogeneous sediments are almost unexplored. We study the relation between macro-scale heterogeneity, thermal dispersion and LTNE through numerical simulations of heat transport in three-dimensional heterogeneous hydraulic conductivity fields. The advection-diffusion equation is solved using the Multiphysics Object-Oriented Simulation Environment (MOOSE), an open-source, parallel finite element framework. The spatial and temporal evolution of the heat plume generated by a line source under steady-state flow conditions is examined. For understanding the propagation of heat plumes, the role of delayed diffusion caused by LTNE effects needs to be distinguished from hydro-mechanical dispersion. Therefore, we estimate the thermal dispersion and the effective thermal retardation for each time step using a stochastic approach. LTNE effects are present, when the effective thermal retardation deviates from the predicted, apparent thermal retardation. Simulations show good agreement between the effective and the apparent thermal retardation for homogeneous hydraulic conductivity. With increasing heterogeneity, characterized by a higher variance of the log-conductivity, the effective retardation becomes lower than the apparent retardation at early times. Furthermore, we estimate the effective thermal retardation for a homogeneous flow field with added thermal dispersion based on the dispersion coefficients resulting from the heterogeneous simulations. We find that there is a significant difference in the evolution of effective retardation between the homogeneous and the heterogeneous case both of the same thermal dispersion, which we associate to LTNE effects. Our modelling approach thus allows to quantify LTNE induced by field-scale heterogeneity.

In the safety assessment of a nuclear waste repository, it is crucial to identify and isolate possible pathways for radionuclides from the waste canister to the biosphere and vice versa. Despite the potential self-sealing properties depending on the host rock and the decrease of the permeability over time, the migration depends on the presence of a fracture network, the connectivity of the fractures, and their connectivity to tunnels or drifts. To describe the initial state of said pathways, we demonstrate the inversion of transient pressure measurements to delineate the structural and pneumatic properties of an excavation-induced fracture network at the Meuse/Haute-Marne Underground Research Laboratory (URL) operated by the French radioactive waste management agency ANDRA. The tomographic tests were carried out as sequential constant-rate injections of nitrogen and the resulting pressure perturbations were recorded in nearby borehole intervals. In total, nine boreholes with two injection/monitoring intervals each are available on a volume of approximately 3m times 3m times 5m. Therefore, the joint inversion of more than 300 signals allows unique insights into the excavation-induced fracture network. A discrete fracture network (DFN) model is applied for the forward and inverse modeling. Thereby, the strong heterogeneity of the distribution of hydraulic or pneumatic properties caused by the fracture network can be described. A numerical model is used to simulate the transient pressure diffusion in the DFN. The structural properties and the permeability of the DFN model are characterized by solving the inverse problem. The inversion relies on a stochastic model of the DFN parameters based on the Bayesian equation. The posterior distribution, i.e., the distribution of the DFN parameters given the measured data, is the product of likelihood and prior distribution. The likelihood function compares the error between the measured data and the simulated outcome of the tomography experiments for a given DFN model. The prior distribution includes information about the fracture properties obtained in previous studies. The posterior distribution is characterized by generating samples from the posterior with Markov chain Monte Carlo (MCMC) methods. Due to the unknown number of fractures, the insertion and deletion of fractures are possible according to the reversible jump MCMC algorithm. The inversion approach results in several DFN realizations that are approximately equally likely which is illustrated in a fracture probability map and a map of the permeability distribution. Thereby, preferential flow paths can be characterized.