Complex Subsurface Systems Research Articles

Geomechanical assessment of a large-scale CO2 storage and insights from uncertainty analysis

Geological storage of CO2 provides a feasible solution to curb CO2 emissions to the atmosphere and relieve climate change. However, as CO2 is injected into the reservoir, the increased pore pressure may result in fault reactivations or damaging geomechanical changes to the caprock, which could be detrimental to the containment of injected CO2. On the other hand, uncertainty is inevitable due to the limited information of the complex subsurface system. In this study, we develop a workflow to assess the geomechanical impact of CO2 geological storage and include management of sources of uncertainties associated with the subsurface system. We utilize the hypothetical large-scale carbon sequestration situated in the Southern San Joaquin Valley in California, USA as a case study. We firstly set up a three-dimensional (3D) geomechanical model using a non-linear quasi-static finite element method. The applied workflow has features including: (1) Coulomb friction behavior to model fault slippage, (2) CAD representation of land surface and faults, (3) non-uniform pressure increase based on dynamic reservoir simulations, and (4) variable rock properties for formation layers. We perform uncertainty analyses using a response surface methodology and a Box-Behnken design to assess the significance of subsurface uncertainties to geomechanical responses. We find that the intact caprock will not have a shear failure even considering the subsurface uncertainties. However, the faults can reactivate and give potential flow pathways in the caprock. The presented workflow integrates advanced geomechanical models and uncertainty analysis, which can be readily applied to other CO2 storage projects to mitigate geomechanical sources of risk.

Local and Global Sensitivity Analysis of a Reactive Transport Model Simulating Floodplain Redox Cycling

AbstractReactive transport models (RTMs) are essential tools that simulate the coupling of advective, diffusive, and reactive processes in the subsurface, but their complexity makes them difficult to understand, develop and improve without accompanying statistical analyses. Although global sensitivity analysis (SA) can address these issues, the computational cost associated with most global SA techniques limits their use with RTMs. In this study, we apply distance‐based generalized sensitivity analysis (DGSA), a novel and computationally efficient method of global SA, to a floodplain‐scale RTM and compare DGSA results to those from local SA. Our test case focuses on the impact of 17 uncertain environmental parameters on spatially and temporally variable redox conditions within a floodplain aquifer. The input parameters considered include flow and diffusion rates, geochemical reaction rates, and the spatial distribution of sediment facies. Sensitivity was evaluated for three distinct components of the model response, encompassing both multidimensional and categorical output. Parameter rankings differ between local SA and DGSA, due to nonlinear effects of individual parameters and interaction effects between parameters. DGSA results show that fluid residence time, which is controlled by aquifer permeability, generally exerts a stronger control on redox conditions than do geochemical reaction rates. Sensitivity indices also demonstrate that sulfate reduction is key for establishing and maintaining reducing conditions throughout the aquifer. These results provide insights into the key drivers of heterogeneous redox processes within floodplain aquifers, as well as the main sources of uncertainty when modeling complex subsurface systems.

Towards a digital twin for characterising natural source zone depletion: A feasibility study based on the Bemidji site

Natural source zone depletion (NSZD) of light non-aqueous phase liquids (LNAPLs) may be a valid long-term management option at petroleum impacted sites. However, its future long-term reliability needs to be established. NSZD includes partitioning, biotic and abiotic degradation of LNAPL components plus multiphase fluid dynamics in the subsurface. Over time, LNAPL components are depleted and those partitioning to various phases change, as do those available for biodegradation. To accommodate these processes and predict trends and NSZD over decades to centuries, for the first time, we incorporated a multi-phase multi-component multi-microbe non-isothermal approach to representatively simulate NSZD at field scale. To validate the approach we successfully mimic data from the LNAPL release at the Bemidji site. We simulate the entire depth of saturated and unsaturated zones over the 27 years of post-release measurements. The study progresses the idea of creating a generic digital twin of NSZD processes and future trends. Outcomes show the feasibility and affordability of such detailed computational approaches to improve decision-making for site management and restoration strategies. The study provided a basis to progress a computational digital twin for complex subsurface systems.

Integrating Process‐Based Reactive Transport Modeling and Machine Learning for Electrokinetic Remediation of Contaminated Groundwater

AbstractAdvanced reactive transport models of fluid flow and solute transport in subsurface porous media are instrumental for the assessment of contaminant environmental fate and for the design of in situ remediation interventions. However, the increasing complexity of process‐based reactive transport simulators often leads to long runtimes, which poses severe restrictions for tasks that require numerous model evaluations. To overcome this limitation, we demonstrate how machine learning surrogate models, trained on the outputs of a limited number of process‐based reactive transport simulations, can predict the evolution of complex subsurface systems. We focus on electrokinetic enhanced bioremediation of chlorinated solvents in low‐permeability porous media, which is an in situ remediation technology entailing a suite of complex and coupled physical, chemical, and biological processes. A process‐based, multicomponent reactive transport model, capable of describing the key mechanisms of electrokinetic flow and transport, is setup in a two‐dimensional domain. The model accounts for electromigration and electroosmosis, the electrostatic interactions between charged species, the chemistry of the pore water solution, the microbially mediated degradation of the organic compounds, and the dynamics of different degraders. We develop a response surface surrogate framework using an artificial neural network as approximation function and we show that the surrogate model has the capability and the flexibility to capture the complex dynamics of electrokinetic remediation in subsurface porous media and allows computationally efficient model exploration, sensitivity analysis, and uncertainty quantification.

Efficient and data-driven prediction of water breakthrough in subsurface systems using deep long short-term memory machine learning

Water coning is one of the common issues in subsurface systems in which water flows into the production well through perforated zones. This phenomenon can cause severe problems in wellbore and surface facilities. Thus, accurate prediction of water breakthrough can help to adapt to the production mode and avoid such issues. Conducting flow simulations, as a conventional approach, can be very time demanding if one deals with large subsurface systems. Furthermore, several types of data are often collected during the life of a subsurface system each of which can help to predict the breakthrough and water coning. As such, it is very important to produce similar results using the time sequence data gathered from various geo-sensing tools. In this paper, a deep long short-term memory (LSTM) model is developed to predict the water cut and water breakthrough time for multiple production wells in a water flooding case. The dataset is generated by the Egg model with a multi-input-multi-output system. We found that the proposed model can capture the general trend of variation for the water cut time sequence data for a complex subsurface system. To evaluate the performance of our data-driven method, the results are compared with vanilla recurrent neural network (RNN), deep gated recurrent unit (GRU), and artificial neural network (ANN). The conducted comparison indicates that the proposed deep LSTM model outperforms the other three approaches when the results are compared with the numerical data.

Indicator-based data assimilation with multiple-point statistics for updating an ensemble of models with non-Gaussian parameter distributions

Ground-water flow and the transport of contaminants in groundwater systems are strongly controlled by geologic heterogeneities. Because, the data available to model these subsurface heterogeneities are generally sparse, the observed groundwater flow, piezometric head at well locations, or the concentration profile of solutes provide valuable additional information about these complex subsurface systems. Regardless, there is likely to be considerable uncertainty associated with the predictions of the subsurface heterogeneities. This has motivated the development of several ensemble-based schemes for assimilating flow response data into predictions of subsurface heterogeneities. Most ensemble-based data assimilation methods including ensemble Kalman filter (EnKF) or indicator-based data assimilation (InDA), used for assimilation of dynamic data into geologic models, utilize statistics in the form of covariances calculated using the ensemble of models to perform model updates. These covariance based updates do not preserve the complex spatial characteristics of geologic structures that are better represented using multiple-point statistics. Additionally, in EnKF, the mismatch between the observed and the simulated values are assumed to be linearly related to the parameter updates, and the distribution of the state variable is assumed to be multi-Gaussian. The spatial distribution of the primary variables like facies can be non-Gaussian because of the spatial continuity exhibited by channel facies (sands) and non-channel facies (clay), that have distinctly different properties. Also, the relationship between the mismatch and the parameter updates can be strongly non-linear. In indicator-based data assimilation (InDA) method the indicator transforms of parameter and mismatch variables are used, which affords the treatment of these variables as bi-variate non-Gaussian. Furthermore, in the indicator transformed space, restrictive linear assumption can also be lifted as the indicator transform is invariant under non-linear transformations. However, the updates are still governed by the bivariate interactions between the state variables. A multiple-point extension of InDA is proposed in this paper which utilizes existing multiple-point simulation algorithms like single normal equation simulation (SNESIM) in combination with InDA, to preserve the spatial characteristics of the geologic model while at the same time honor the observed flow or transport response. The proposed method is applied to a synthetic groundwater system with complex channel distributions. The channel features of the final updated ensemble of models is shown to converge towards the reference model and the ensemble flow responses are also shown to match the reference flow response.

Karakter dan Potensi Risiko Kerusakan Ekosistem Karst Cekungan Air Tanah Watuputih Kabupaten Rembang, Provinsi Jawa Tengah

The research area is a burgeoning karst area which is part of the Kendeng Kars area in Rembang Regency, where the karstification process continues and has a complex subsurface system. In addition to the karst area of the Watuputih Groundwater Basin (CAT) written in Presidential Decree Number 26 of 2011 point 124 is included in the CAT B classification, which is located across the regency between Rembang and Blora Regencies. The morphological changes caused by human beings and the sharing of mining activities and plans for the construction of a cement factory by PT SI and its subsidiaries such as PT SG are also investors such as PT SIR, PT RSI, PT GMM, PT BA, PT RBP, PT SAF, PT ICCI which have reached an area 493 hectares in Tegaldowo, Gunem District. These activities will directly or indirectly damage the ecosystem of caves, underground rivers and parental springs that are scattered in the Watuputih CAT area. The Bengawan Solo River, the Lusi River and the Tuyuhan River are large rivers that tipped in the Watuputih CAT area. There are 74 caves, 29 springs, 44 ponors in the Tegaldowo and Timbrangan areas, 10 caves and 3 springs in the Perhutani area. Morphological changes due to mining can affect the pattern of water distribution, where caves, springs and ponors have a very important role in controlling water supply, the degradation of the amount of water stored in the Watuputih CAT results in changes in the composition of the base flow (diffuse flow) compared to the total flow. Limestone mining will reduce the absorption function and the amount of diffuse water storage, on the contrary will increase conduit flow when it rains. If an area has criteria as karst and CAT areas then it must be a major concern in the sustainability of its ecosystem. Provide protection and stop all activities that can damage geological protected areas that have karst landscapes and water recharge, especially from small scale or massive mining activities. Provides an understanding of karst areas in the perspective of disaster risk reduction. It needs to be developed and informed to various parties across sectors and across administrations so that appropriate policies can be made as an ecological function. Keywords: Watuputih Groundwater Basin, Water Distribution, Ecosystems, Ecological Function, Protection

A Multiscale Approach for Geologically and Flow Consistent Modeling

Subsurface geological models are usually constructed on high-resolution grids in a way that various complexities and heterogeneities are depicted properly. Such models, however, cannot be used directly in the current flow simulators, as they are tied with high computational cost. Thus, using upscaling, by which one can produce flow consistent models that can alleviate the computational burden of flow simulators, is inevitable. Although the upscaling methods are able to reproduce the flow responses, they might not retain the initial geological assumptions. The reservoir models are initially constructed on uniform and high-resolution grids and then, if necessary, are upscaled to be used for flow simulations. A subsurface modeling approach that not only preserves the geological heterogeneity but also provides models that can be used, straight or with a small level of upscaling, in the flow simulators is desirable. In this paper, a new multiresolution method based on (1) the importance of conditioning well data and (2) being geologically and flow consistent is presented. This method discretizes the initial model into several regions based on the available data. Then, the initial assumed geological model is converted into, for example, various high- and low-resolution models. Next, the high-resolution model is used for regions with high-quality data (e.g., well locations), while the low-resolution model is used for the remaining areas. Finally, the patterns of these areas are interlocked, which result in a multiresolution geologically and flow consistent subsurface model. The accuracy of this method is demonstrated using two-phase flow simulation on four complex subsurface systems. The results indicate that the same flow responses, in a much less time, are reproduced using the multiscale models. The speed-up factor gained using the proposed method is also several orders of magnitude.

Influence of flow velocity and spatial heterogeneity on DNAPL migration in porous media: insights from laboratory experiments and numerical modelling

Understanding the migration of dense non-aqueous phase liquids (DNAPLs) in complex subsurface systems is important for evaluating contamination source zones and designing remediation schemes after spill events. Six sandbox experiments were performed to explore the individual effect of flow velocity, and the combined effect of flow velocity and layered lenses on a DNAPL (PCE) migration in porous media. DNAPL saturation was measured using a light transmission system, and saturation distribution was quantified by spatial moments. The experimental results show that large flow velocity significantly promotes lateral and vertical migration of the low-viscosity DNAPL, while when layered lenses exist, the infiltration rate decreases and horizontal spread increases. Migration processes were numerically simulated, and the modelling results tested against experimental results. Furthermore, migration of DNAPLs with different viscosities was simulated to explore the combined effects of flow velocity and geological heterogeneity. Simulation results show that enhanced heterogeneity makes low-viscosity DNAPLs migrate along preferential pathways, resulting in irregular DNAPL morphology. Layered lenses combined with heterogeneity complicate the effect of flow velocity on the migration of low-viscosity DNAPLs by changing percolation paths. Results also demonstrate that flow velocity exhibits relatively little influence on the migration of medium/high-viscosity DNAPLs, which is predominantly controlled by viscosity and heterogeneity. Enhanced heterogeneity has a larger effect on migration behavior. Findings indicate that the migration paths and position of the source zone could change significantly, due to the combined effect of groundwater flow velocity and geological heterogeneity; thus, comprehensive hydrogeological investigation is needed to characterize the source zone.

TOUGH: Model Use, Calibration, and Validation

Simulation models for subsurface flow and transport require the estimation of many disparate parameters to capture properties and processes that cannot be explicitly represented in the model. The main challenges for model calibration reside in the formulation of the forward model, the identification of influential parameters, the selection of appropriate calibration data with sufficient information content about these parameters, and the performance of a robust and efficient minimization algorithm capable of handling complex topologies of the resulting objective function. The large variety of subsurface systems and related engineering problems that are addressed by numerical simulation prevents us from compiling definite lists of critical input parameters and desirable observation data to be used for model calibration. Instead, we describe the approach that we consider necessary to arrive at a defensible model for predictive simulations. This process includes sensitivity analyses and synthetic inversions prior to data collection, and detailed residual, error, and uncertainty analyses after automatic model calibration. We illustrate the approach by discussing applications of the iTOUGH2 simulation-optimization code, which supports the development of TOUGH2 models for nonisothermal, multiphase, multicomponent flow and transport processes in fractured porous geological media. We conclude that the calibration of models representing complex subsurface systems remains challenging and requires research that is targeted at the needs of scientists and engineers who address simulation problems of practical relevance.

Advanced Vadose Zone Simulations Using TOUGH

The vadose zone can be characterized as a complex subsurface system in which intricate physical and biogeochemical processes occur in response to a variety of natural forces and human activities. This makes it difficult to describe, understand, and predict the behavior of this specific subsurface system. The TOUGH nonisothermal multiphase flow simulators are well suited to perform advanced vadose zone studies. The conceptual models underlying the TOUGH simulators are capable of representing features specific to the vadose zone, and of addressing a variety of coupled phenomena. Moreover, the simulators are integrated into software tools that enable advanced data analysis, optimization, and system‐level modeling. We discuss fundamental and computational challenges in simulating vadose zone processes, review recent advances in modeling such systems, and demonstrate some capabilities of the TOUGH suite of codes using illustrative examples.

Analyzing unsaturated flow patterns in fractured rock using an integrated modeling approach

Characterizing percolation patterns in unsaturated fractured rock has posed a greater challenge to modeling investigations than comparable saturated zone studies, because of the heterogeneous nature of unsaturated media and the great number of variables impacting unsaturated flow. This paper presents an integrated modeling methodology for quantitatively characterizing percolation patterns in the unsaturated zone of Yucca Mountain, Nevada, a proposed underground repository site for storing high-level radioactive waste. The modeling approach integrates a wide variety of moisture, pneumatic, thermal, and isotopic geochemical field data into a comprehensive three-dimensional numerical model for modeling analyses. It takes into account the coupled processes of fluid and heat flow and chemical isotopic transport in Yucca Mountain's highly heterogeneous, unsaturated fractured tuffs. Modeling results are examined against different types of field-measured data and then used to evaluate different hydrogeological conceptualizations and their results of flow patterns in the unsaturated zone. In particular, this model provides a much clearer understanding of percolation patterns and flow behavior through the unsaturated zone, both crucial issues in assessing repository performance. The integrated approach for quantifying Yucca Mountain's flow system is demonstrated to provide a practical modeling tool for characterizing flow and transport processes in complex subsurface systems.

Real-Time, in Situ Monitoring of Bioactive Zone Dynamics in Heterogeneous Systems

Successful implementation of in situ bioremediation is contingent upon understanding how physicochemical and microbial factors affect the formation and dynamics of microbially active regions known as bioactive zones (BAZs). This study demonstrates how a novel fiber optic detection system can be used to test hypotheses concerning real-time, in situ BAZ formation and dynamics. This study focuses on naphthalene transport in saturated porous media containing defined physicochemical and microbial heterogeneities. Biological activity was measured using a lux reporter bacterium, Pseudomonas putida RB1353, that bioluminesces during naphthalene catabolism. Results show that the presence of defined heterogeneities drives the development of BAZs at material-property interfaces where the confluence of naphthalene, dissolved oxygen, and sufficient microbial density is optimal. Thus, despite successful transport of P. putida RB1353 into a sterile low-permeability region containing substrate, BAZ formation in this region was limited by local physicochemical conditions (e.g., naphthalene and dissolved oxygen bioavailability). In another instance, transport of P. putida RB1353 occurred against advective flow, resulting in BAZ formation upgradient of inoculated regions. Defined systems such as this can be used as a basis for predicting localization of activity in complex subsurface systems.

Sorption phenomena in subsurface systems: Concepts, models and effects on contaminant fate and transport

The behavior, transport and ultimate fate of contaminants in subsurface environments may be affected significantly by their participation in sorption reactions and related phenomena. The degree to which the resulting effects can be quantified and predicted depends upon the extent to which certain fundamental aspects of sorption are understood, and upon the accuracy with which these phenomena can be characterized and modeled in complex subsurface systems. Current levels of understanding of the reactions and processes comprising sorption phenomena are discussed in this paper, as are the forms and utilities of different models used to describe them. Emphasis is placed on concept development, on the translation of these concepts into functional models for characterizing sorption rates and equilibria, and on the application of these concepts and models for explaining contaminant behavior in subsurface systems. Examples are provided to illustrate the impacts of sorption phenomena on contaminant transport.