Search

Identification of preferential flow‐paths, such as fractures, is required for various issues in geosciences. When chemicals are injected into the subsurface, monitoring the resulting structural and chemical changes remains a challenge. The ability of geophysical tomography to tackle this problem is not fully explored due to the lack of numerical methods suitable for modeling narrow structures. We explore how discrete representation of preferential flow‐paths provides innovative ways to invert electrical resistivity data collected during reagent injection at a contaminated site. The data set is inverted with a scheme where a new fracture is added at every iteration. This allows identifying newly created narrow conductive structures from the field data collected before and after injection. Fracture location remains overall consistent despite using different starting points for the fracture search. A prior constraint on fracture length improves convergence. These results show the potential of discrete inversion for identifying narrow structures from electrical resistivity monitoring. Plain Language Summary: In some geoscience activities, such as water extraction, geological storage, and contaminant fate, chemical solutions are injected into wells, which potentially leads to modify the system properties. Identifying preferential flow‐paths is therefore necessary even though monitoring structural and chemical changes in the subsurface remains a challenge. Electrical geophysical imaging techniques may be adapted to provide information for this purpose. However, their potential is not fully explored due to the limitations of numerical models to represent small‐scale structures. In this work, we explore the representation of preferential flow‐paths by discrete elements for the inversion of electrical resistivity data collected during a reagent injection at a contaminated site. The inversion scheme based on discrete forward simulations successively identifies small‐scale structures and focuses on evaluating their localization and length. The results show the need for further efforts to use discrete electrical resistivity inversion in the presence of narrow structures. Key Points: Standard modeling strategies are not well suited to identify narrow geological structures from cross‐borehole electrical data inversionInversion scheme based on discrete‐dual‐porosity forward simulations enable to image narrow preferential flow‐paths from electrical dataPrior assumptions are needed to invert the position and length of narrow preferential flow‐paths [ABSTRACT FROM AUTHOR]

The Valgarður database is a compilation of data describing the physical and geochemical properties of Icelandic rocks. The dataset comprises 1072 samples obtained from fossil and active geothermal systems, as well as relatively fresh volcanic rocks erupted in sub-aerial or sub-aqueous environments. The database includes petrophysical properties (effective and total porosity, grain density, permeability, electrical resistivity, acoustic velocities), as well as mineralogical and geochemical data obtained by point-counting, X-ray Fluorescence (XRF), quantitative X-ray Diffraction (XRD), and Cation Exchange Capacity (CEC) analyses. The motivation behind this database is threefold: (i) aid in the interpretation of geophysical data including uncertainty estimations, (ii) facilitate the parameterization of numerical reservoir models, and (iii) improve our understanding of the relationship between rock type, hydrothermal alteration and petrophysical properties [ABSTRACT FROM AUTHOR]

Processing of geophysical data is a time consuming task involving many different steps. One approach for accelerating and automating processing of geophysical data is to look towards machine learning (ML). ML encompasses a wide range of tools, which can be used to automate complicated and/or tedious tasks. We present strategies for automating the processing of time-domain induced polarization (IP) data using ML. An IP data set from Grindsted in Denmark is used to investigate the applicability of neural networks for processing such data. The Grindsted data set consists of eight profiles, with approximately 2000 data curves per profile, on average. Each curve needs to be processed, which, using the manual approach, can take 1–2 hr per profile. Around 20 per cent of the curves were manually processed and used to train and validate an artificial neural network. Once trained, the network could process all curves, in 6–15 s for each profile. The accuracy of the neural network, when considering the manual processing as a reference, is 90.8 per cent. At first, the network could not detect outlier curves, that is where entire chargeability curves were significantly different from their spatial neighbours. Therefore, an outlier curve detection algorithm was developed and implemented to work in tandem with the network. The automatic processing approach developed here, involving the neural network and the outlier curve detection, leads to similar inversion results as the manual processing, with the two significant advantages of reduced processing times and enhanced processing consistency. [ABSTRACT FROM AUTHOR]

Hydrogen sulfide (H 2 S), an environmentally harmful pollutant, is a byproduct of geothermal energy production. To reduce the H 2 S emissions, H 2 S-charged water is injected into the basaltic subsurface, where it mineralizes to iron sulfides. Here, we couple geophysical induced polarization (IP) measurements in H 2 S injection wells and geochemical reactive transport models (RTM) to monitor the H 2 S storage efforts in the subsurface of Nesjavellir, one of Iceland's most productive geothermal fields. An increase in the IP response after 40 days of injection indicates iron-sulfide formation near the injection well. Likewise, the RTM shows that iron sulfides readily form at circumneutral to alkaline pH conditions, and the iron supply from basalt dissolution limits its formation. Agreement in the trends of the magnitude and distribution of iron-sulfide formation between IP and RTM suggests that coupling the methods can improve the monitoring of H 2 S mineralization by providing insight into the parameters influencing iron-sulfide formation. In particular, accurate fluid flow parameters in RTMs are critical to validate the predictions of the spatial distribution of subsurface iron-sulfide formation over time obtained through IP observations. This work establishes a foundation for expanding H 2 S sequestration monitoring efforts and a framework for coupling geophysical and geochemical site evaluations in environmental studies.