Publications
[Go to Proceedings Refereed] [Go to Proceedings Non-Refereed] [Go to Theses]
Underlined names are links to current or past GEG members
REFEREED PUBLICATIONS IN JOURNALS
8.
Bondarenko, N., C. Goldberg, S. Williams-Stroud, and R.Y. Makhnenko, Machine learning enhanced interpretation of wellbore data for underground CO2 storage in Illinois Basin, Rock Mechanics and Rock Engineering, 2026. https://doi.org/10.1007/s00603-026-05324-2 [Download] [View Abstract]This study explores the application of data-driven approaches to enhance the interpretation of geophysical wellbore data for
carbon capture and storage (CCS) in the Illinois Basin. Extensive data from exploration and pilot-scale projects are analyzed
to assess their potential for generating synthetic datasets that can effectively replicate some missing information for ongoing
projects. Specifically, various machine learning (ML) models, including random forest, gradient boosting, feed-forward neural
networks, and others, are employed to predict sonic velocities based on the mineralogical composition of participating rock
and to identify fracture locations using petrophysical logs. ML-based approaches significantly outperform the traditional
effective media interpretation, exhibiting improved accuracy in sonic velocity prediction and capturing small-scale heterogeneity of the formations. However, the prediction of fracture presence remains challenging due to data imbalance and the
complex interplay between fractures and inherent heterogeneity of geologic formations. Application of t-distributed stochastic
neighbor embedding (t-SNE)—an advanced technique for multidimensional data visualization—reveals that spatial heterogeneity strongly influences geophysical properties, limiting ML models’ performance in detecting fractures. Despite these
challenges, ML approaches show promise in complementing traditional methodologies, enabling faster and more informed
decision-making processes during early project stages. Future research is needed to address data limitations and enhance
the reliability of ML models in diverse geological settings.
7.
Bondarenko, N., H. Kim, K. Kim, and R.Y. Makhnenko, Experimental insights into CO2 flow in fractured crystalline rock, International Journal of Rock Mechanics and Mining Sciences , 200, pp. 106443, 2026. https://doi.org/10.1016/j.ijrmms.2026.106443 [Download] [View Abstract]The injection of carbon dioxide (CO2) or the non-wetting fluid intrusion into crystalline rock remains being a poorly understood process due to the complexities in characterizing very low permeable and stiff geomaterials. This knowledge gap is critical for enhanced geothermal systems and in-situ carbon mineralization projects where CO2 may serve as a mobile fluid within fractured crystalline formations. Most of the existing studies rely on numerical simulations with limited experimental validations and do not fully consider the complexity of multi-phase flow. Hereby, we adopt a novel method to evaluate the degree of saturation of the non-wetting fluid in a tight rock with nanometer scale pore sizes from accurate poromechanical and hydraulic measurements, as well as wetting and non-wetting fluid characteristics. We select thermally damaged granite and naturally fractured rhyolite as representative crystalline rock, fully saturate them with water, and perform simultaneous injection of water and liquid CO2. The flow properties are measured using the core flooding device that allows observation of multiple fluid flow at controlled rates. CO2 breakthrough pressures for pressurized fractured crystalline rock are measured to be on the order of 0.1–1 MPa. The exponent values for relative water permeabilities are 1.6 for granite and 1.9 for rhyolite – significantly lower than those typically reported for tight rock, meaning that the fluid flow is mainly governed by the fractures. The exponent values for relative CO2 permeability are above 5.5, indicating high sensitivity to the degree of CO2 saturation. Moreover, CO2 saturation appears to remain below 50%, even when CO2 is the only injected fluid and its overpressures exceeds 6 MPa. Overall, this study highlights significant limitations in using CO2 as a working fluid for geoenergy projects in crystalline rock.
6.
Bondarenko, N., Y. Podladchikov, S. Williams-Stroud, and R. Makhnenko, Stratigraphy-Induced Localization of Microseismicity During CO2 Injection in Illinois Basin, Journal of Geophysical Research: Solid Earth, 130, pp. e2024JB029526, 2025. https://doi.org/10.1029/2024JB029526 [Download] [View Abstract]Subsurface fluid injection stimulates complex hydromechanical interaction, necessitating the integration of geomechanical data across spatial and temporal scales to consider the sophisticated behavior. Induced seismic response is usually associated with the complex reservoir architecture and pre-existing features that are three-dimensional, such as local stratigraphy, fractures, faults, and other discontinuities. This study encompasses laboratory characterization of the coupled hydromechanical response of cores extracted from rock formations in Illinois Basin: reservoir - Mt. Simon sandstone, basal seal - Argenta sandstone, and crystalline basement - Precambrian rhyolite. High-resolution numerical modeling allows considering the three-dimensional complexity of the Illinois Basin Decatur Project with spatial resolution comparable to one of the active seismic surveys. A detailed reconstruction of the evolving state of stress in formations lacking direct stress measurements is achieved by numerical modeling that integrated laboratory-derived hydromechanical properties, a porosity-permeability relationship, active seismic data, and an inverted three-dimensional porosity distribution. It appears that the microseismic clusters, mainly observed in the crystalline basement during the injection, are linked to zones experiencing more critically stressed conditions prior to injection. These zones have a potential for reactivation during the injection and are attributed to the specific local stratigraphy of the injection site, as well as transfer of triggering perturbations during the injection.
5.
Kim, H., S. Ding, N. Bondarenko, D.C. Willette, S. Salahshoor, and R.Y. Makhnenko, Experimental approaches for characterization of water-hydrogen flow in reservoir rock, Journal of Energy Storage, 114(B), 2025. https://doi.org/10.1016/j.est.2025.115785 [Download] [View Abstract]Subsurface hydrogen (H2) storage has emerged as a promising solution for overcoming challenges in renewable energy generation. The feasibility of geologic hydrogen storage in saline aquifers requires a comprehensive analysis of multiphase fluid flow within reservoir formations. This study investigates the water-hydrogen transport properties of homogeneous quartz-arenite Berea sandstone and heterogeneous Ironton/Galesville containing clay-rich bedding planes. An experimental setup is introduced to measure the intrinsic and water-hydrogen relative permeability under representative in-situ stress conditions. Single-phase flow tests reveal the bedding-normal intrinsic permeability of Ironton/Galesville to be ⁓10−17 m2 – three to four orders of magnitude lower than the one measured for bedding-parallel orientation and Berea sandstone. In the two-phase flow tests, hydrogen exhibits significantly lower relative permeability than water primarily due to its lower viscosity. Hysteresis in relative permeability is observed only in primary episode, disappearing for the consecutive drainage and imbibition cycles. The strongest hysteresis is identified in vertical Ironton/Galesville, attributed to pore structure complexity and variation in capillary responses. Comparison with relative permeability estimation based on pore structure analysis underscores the limitation of such methods and highlights the importance of direct measurements. These findings provide critical insights into water-hydrogen flow mechanisms and offer valuable data for evaluating subsurface hydrogen potential.
4.
Bondarenko, N., Y. Podladchikov, and R. Makhnenko, Hydromechanical impact of basement rock on injection-induced seismicity in Illinois Basin, Scientific Reports, 12, pp. 15639, 2022. https://doi.org/10.1038/s41598-022-19775-4 [Download] [View Abstract]The common explanation of observed injection-induced microseismicity is based on the measured stress state at the injection interval and the assumption that it remains the same in the vicinity. We argue here that representing the stress state in different geologic formations over the injection site with the single Mohr’s circle is insufficient due to local stratigraphic features and contrast in compressibilities of the involved formations. The role of hydromechanical coupling in the microseismic response is also crucial for the proper assessment of the problem. Thoroughly monitored Illinois Basin Decatur Project revealed the majority of CO2 injection-associated microseismic events being originated in the crystalline basement. Even though basement faults can serve as the conduits for fluid flow—the predicted pressure increase seems to be insufficient to trigger seismicity. To address this issue, accurate laboratory measurements of rock properties from the involved formations are conducted. The pre-injection stress state and its evolution are evaluated with the hydromechanically coupled numerical model. It appears that the presence of an offset in a stiff competent layer affects the stress state in its vicinity. Therefore, both the pre-injection stress state and its evolution during the fluid injection should be addressed during the induced seismicity assessment.
3.
Bondarenko, N., S. Williams-Stroud, J. Freiburg, and R. Makhnenko, Geomechanical aspects of induced seismicity during CO2 injection in Illinois Basin., The Leading Edge, 40(10), pp. 823-830, 2021. https://doi.org/10.1190/tle40110823.1 [Download] [View Abstract]Carbon sequestration activities are increasing in a global effort to mitigate the effects of greenhouse gas emissions on the climate. Injection of wastewater and oil-field fluids is known to induce seismic activity. This makes it important to understand how that risk relates to CO2 injection. Injection of supercritical CO2 into the Cambrian Mt. Simon sandstone in Illinois Basin induced microseismicity that was observed below the reservoir, primarily in the Precambrian crystalline basement. Geomechanical and flow properties of rock samples from the involved formations were measured in the laboratory and compared with geophysical log data and petrographic analysis. The controlling factors for induced microseismicity in the basement seem to be the hydraulic connection between the reservoir and basement rock and reactivation of pre-existing faults or fractures in the basement. Additionally, the presence of a laterally continuous low-permeability layer between reservoir and basement may have prevented downward migration of pore pressure and reactivation of critically stressed planes of weakness in the basement. Results of the geomechanical characterization of this intermediate layer indicate that it may act as an effective barrier for fluid penetration into the basement and that induced microseismicity is likely to be controlled by the pre-existing system of faults. This is because the intact material is not expected to fail under the reservoir stress conditions.
2.
Smirnov, V.B., A.V. Ponomarev, A.V. Isaeva, N. Bondarenko, A.V. Patonin, P.A. Kaznacheev, S.M. Stroganova, M.G. Potanina, R.K. Chadha, and K. Arora, Fluid initiation of fracture in dry and water saturated rocks, Izvestiya, Physics of the Solid Earth, 56(6), pp. 808-826, 2020. https://doi.org/10.1134/S1069351320060099 [Download] [View Abstract]We present the results of the laboratory studies on fluid initiated fracture in the samples of porous-fractured rocks that have been initially saturated with a pressure-injected fluid and then tested under increasing fluid pressure in saturated rocks. The tests were conducted at the Geophysical observatory “Borok” of Schmidt Institute of Physics of the Earth of the Russian Academy of Sciences. The laboratory is equipped with electrohydraulic press INOVA-1000. The experiments were conducted on the rock samples with substantially different porosity. The tested samples were made of Buffalo sandstones, granites from the well drilled in a seismically active region and of granites from the well in the Voronezh crystalline massif. The permeability of granite samples was varied by their controlled artificial cracking by successive heating and cooling. The experimental procedure was set up in the following way. A preliminarily dried sample was initially subjected to uniaxial loading in uniform compression (confining pressure). Loading was performed at a constant strain rate until the moment when the growth rate of acoustic emission (AE) activity began to accelerate which indicated that the stress level approaches ultimate strength. Since that, the loading rate was decreased by an order of magnitude, and water was infused into a sample from its top face. The bottom end of a sample was tightly sealed and impermeable to water. After this, the pore pressure in the sample that had got saturated with water to that moment was raised in steps whose amplitudes were varied. The obtained results of the laboratory studies show that the character and intensity of fluid initiation of fracture markedly differ under primary fluid injection into the porous-fractured samples and under the subsequent increases of the pore pressure in the saturated samples. The time delay of acoustic response relative to fluid initiation and the amplitude of the response proved to be larger in the case of water injection into dry samples than in the case of raising the pore pressure in saturated samples. The AE response to the decrease in the pore pressure was also detected in the experiments. The theoretical analysis of fluid propagation in a pore space of an air-filled sample in the model with piston-type air displacement has shown that in the case of water injection into a dry sample, the fluid pressure front propagates more slowly than in the saturated sample.
1.
Kaznacheev, P.A., Z.-Yu.Ya. Maibuk, A.V. Ponomarev, V.B. Smirnov, and N. Bondarenko, On the issue of analysis of acoustic emission event statistic on data of single sensor in rock thermal fracture experiments, Geophysical Research, 20(1), pp. 52-64, 2019. https://doi.org/10.21455/gr2019.1-5 [Download] [View Abstract]Authors examine the problem of estimation of b-value for energy distribution of thermal acoustic emission (TAE) events on basis of amplitude distribution of TAE impulses. Impulses are registered by single TAE sensor. Authors have analyzed the effect of factors, associated with elastic waves propagation, on the energy of impulses. The analysis shows, that effect of elastic waves absorption in heated sample is the most significant from these factors. Two events of energy distribution are considered – with one and two sub-functions. It has shown, that the same b-value of registered impulse amplitude distribution and initial event distribution is observed if only b-value is stable in all TAE-event energy range (one sub-function). In this situation, there is one characteristic generation mechanism of events in all sample volume. But if b-value is not stable in different energy ranges (two sub-functions), then elastic waves absorption in the sample distorts initial distribution. Authors propose technique of “true” b-value estimation on basis of distribution analysis of registered TAE impulses in several amplitude subranges.