Archive for the category: Geomechanics for Energy and Resources

Mohammadpour, Mobarakeh¹, Arashpour, Mehrdad¹, Roshan, Hamid² and Masoumi, Hossein¹

¹Department of Civil Engineering, Monash University, Melbourne, Australia ²School of Mineral and Energy Resources Engineering, UNSW, Sydney, Australia

Geophysical logs have been routinely performed in coal mines for many years. Compressional velocity is one of the important characteristics which can be measured using sonic log. Despite importance of P-wave velocity and its application in geophysical and geomechanical studies; some boreholes in coal mines do not have P-wave velocity or sonic log in their log suits; as a result, empirical correlations were commonly used to estimate P-wave velocity. However, these models are mostly local correlations which were derived for specific areas. 

In this study Machine Learning based Gaussian Processes Regression was used to predict the P-wave velocity. Gamma, two density logs with different resolutions (Long Spaced Density and Short Spaced Density) and depth were applied as the input parameters. These three logs are the most common ones which are extracted from the coal mines. The model was generated using the data obtained from six boreholes in one of the Australian coal mines in Queensland. The data were divided into two groups including 35382 points for the training of the model and 11794 points for the testing. Root mean square deviation (RSME) and coefficient of determination (R²) were calculated to evaluate the accuracy of the proposed model.

Biography

Mobarakeh Mohammadpour is a PhD student at the at the Department of Civil Engineering, Monash University.

Sari Mustafa¹, Poulet Thomas¹, Alevizos Sotiris² and Veveakis Manolis³

¹CSIRO, Mineral Resources, WA, Australia, ²National Technical University of Athens, Athens, Greece, ³Civil and Environmental Engineering, Duke University, Durham, USA

Understanding and improving the performance as well as the productivity of reservoirs at extreme conditions (high pressure and high temperature is big challenge in Geomechanics. This is particularly important for the operation of unconventional shale gas reservoirs, but also applies to other areas like geothermal or deep conventional formations. Other applications include environmental remediation around nuclear waste disposal sites, as well as deep underground storage of energy resources. In order to meet such a formidable challenge, constitutive modelling is required that can span the temperature and pressure range of the Earth’s upper crust, and offer possibilities for long-term forward modelling at such extreme conditions. The main requirement is to formulate a plasticity theory which helps describe and even predict long-term behavior, at conditions where materials are frequently well beyond their initial yield, in environments where rocks may experience extreme temperatures, pressures, as well as internal transformations. To model such multi-physical processes, we suggested a multi-physics elasto-visco-plastic constitutive framework including the effect of interface processes, whereby the hardening law of plasticity is a function of the global and internal state variables of the problem (temperature, pressure, density, chemical potentials). The evolution of the laws of the state variables are therefore obtained from the governing laws of physics for mass and energy balance. We included the effect of interface processes at the grain contacts and surface, through an energy upscaling of the internal enthalpy of the system. The framework was validated against a suite of multi-physical tests in different materials, showing good agreement for a realistic range of material parameters. The analysis also showed that simulation results contain enough information to constrain the parameter space for the definition of the mechanical enthalpy, providing insights to develop further the underlying theoretical model and emphasizing the complementarity of data-driven and physics-based approaches.

Biography

Mustafa obtained his PhD in Geomechanics at, UNSW, to develope a new multiphysics framework for Geomaterials characterization with multiphysics feedback, through theoretical, numerical and experimental approaches. In 2019 he started with CSIRO in Mineral recourses department. In this role he is developing a novel non-destructive rock characterisation methodology.

Tork Qashqai, Mehdi¹ and Saygin, Erdinc¹  

¹Deep Earth Imaging Future Science Platform, CSIRO, Perth, Western Australia

Teleseismic body-waves coda recorded on the radial and the vertical components of a seismogram have been used for decades to image the local structures below a seismic station through the inversion of P-to-S receiver functions (RFs).  It has been shown that the observed response of the subsurface structure to a seismic wave, generated by a deep source, can be converted to a zero-offset reflection response by autocorrelating the recorded signals. Recently, the autocorrelation of the teleseismic P-wave coda or its inversion (Tork Qashqai et al., 2019) has emerged as a powerful tool to obtain complementary constraints on subsurface structures. Compared to the RFs, the autocorrelations of the teleseismic P-wave coda recorded on the radial and vertical components of a seismogram contain additional information. They include both P- and P-to-S converted phases, whereas the RFs mainly contain P-to-S phases as the P-waves are attenuated by the deconvolving the vertical component (P) from the radial (S_v) component. Therefore, one can account for the variability of both the V_p and V_s structures if the radial and vertical component autocorrelations are jointly inverted. Here, we present a new approach which can simultaneously estimate the crustal V_p, V_s, and V_p/V_s ratio structures below a seismic station by jointly inverting the vertical and radial component autocorrelations of the teleseismic P-wave coda. This has significant implications for characterizing the V_p/V_s ratio, which can be a good indication of the crustal composition. Our synthetic inversion tests showed substantial improvements in the estimation of the crustal properties (especially the V_p/V_s ratio) compared to the inversion of either the teleseismic RFs or the autocorrelation of the vertical component. The application of this method on passive seismic data recorded by a north-south oriented passive seismic experiment in central Australia (BILBY) provided the first comprehensive joint estimates of all crustal properties (V_p, V_s, and V_p/V_s ratio) for this experiment. We imaged crustal structures across the transition between the northern and southern Australian cratons which includes east-west trending geological domains of central Australia (the Gawler Craton, Eromanga and Officer Basins, Musgrave Province, Amadeus Basin, Arunta Block and the Georgina Basin). The comparison of the Moho estimates from the previous studies with our velocity and Moho models indicates that they might have imaged the top of a high-velocity lower crust in some regions. The overall trend of our Moho model follows the long-wavelength pattern of the Moho structure interpreted from the deep seismic reflection method along the GOMA seismic line that is parallel to the BILBY profile. It is also closer to the change of the reflectivity seen at the base of the crust in the GOMA migrated seismic section. Our approach is cost-effective and can be used in conjunction with the deep active seismic reflection profiling to obtain additional information, especially at depths where the deep seismic reflection method has penetration problems.

References:

Tork Qashqai, M., Saygin, E., & Kennett, B. L. N. (2019). Crustal imaging with Bayesian inversion of teleseismic P wave coda autocorrelation. Journal of Geophysical Research: Solid Earth, 124, 5888-5906.

Biography

Mehdi is a research scientist at the CSIRO Deep Earth Imaging Future Science Platform and graduated from Macquarie University in 2016. His main research interests lie in the developing and applications of the passive seismic methods for imaging the subsurface structure from the near-surface to the base of lithosphere.

Gammidge, Larissa¹; Xu, Min¹

¹Geological Survey of New South Wales, Department of Regional NSW, Maitland, Australia

The late Silurian to Devonian Darling Basin in western New South Wales is one of the least explored sedimentary basins in the state and is prospective for petroleum. It comprises many sub-basins that share, or partially share, a similar depositional history. The location of the boundaries of some of the sub-basins have recently been revised as more data have been acquired and new modelling completed.

Limited seismic and drilling data across the Darling Basin, particularly near the margins of the sub-basins, and extensive shallow cover required a reliance on gravity and aeromagnetic data to interpret the boundaries of the sub-basins. The most widely used dataset to interpret sub-basin boundaries is the 2006 Murray-Darling-Eromanga SEEBASE™ model, created by Frogtech.

The SEEBASE™ model for the Yathong and Ivanhoe troughs shows that they are contiguous, but they were interpreted as two troughs based on their different orientations and slightly different structural histories. However, the interpretation and modelling of seismic survey data acquired in 2009 and 2013 show that there is continuity of strata across the Yathong and Ivanhoe trough boundaries. Therefore, there is no justification for two separate troughs and a single trough is more consistent with the current data. The revised trough is named the Yathong–Ivanhoe Trough.

The Yathong–Ivanhoe Trough boundary has been defined based on the extent of outcropping Devonian sedimentary rocks and the interpreted pre-Devonian basement rocks. Additionally, the SEEBASE™ and aeromagnetic data were used to define the boundary in areas of no outcrop. The former Yathong Trough has been shortened in the north–south direction by almost half. In contrast, the former Ivanhoe Trough is revised to be larger.

The revised boundary for a combined Yathong–Ivanhoe Trough is a more geologically robust interpretation than previous ones. This has implications for petroleum exploration targets within the trough. Recent work has highlighted areas where petroleum systems may be located and gives a greater understanding of the development of the Darling Basin over time.

Biography

Larissa is part of the Geological Survey of NSW, in the Petroleum and Renewables team. She has helped produce data packages for the Darling Basin, soil gas surveys and validated location information. Prior work includes supporting Coal Innovation NSW’s CO2 storage program, drilling two wells in the Darling Basin.

Musolino, Matthew¹ Holford, Simon¹ King, Rosalind¹, Hillis, Richard¹

¹The University of Adelaide, Australian School of Petroleum and Energy Resources, Adelaide, Australia

Accurate estimates of in-situ stresses and rock strengths are required for multiple practical applications during petroleum exploration and development, such as ensuring wellbore stability, minimising breakouts, and the design of hydraulic fracturing treatments. Issues regarding the aforementioned practices cost operators about US$8 billion globally each year. However, when predictive geomechanical models are constructed, both rock stress and strength are typically estimates rather than fully quantified, and their uncertainties are rarely examined. Based on data from the Cooper Basin in South Australia, this presentation examines uncertainty relating to estimates of the three principal stresses (vertical stress and minimum and maximum horizontal stress) and rock strength as a result of (a) available data inputs and (b) methodological approaches. We show that the magnitude of vertical stress at depths of ~3 km can vary up to 6 MPa depending on the methodologies employed when integrating and processing density logs and sonic transit time data. Minimum horizontal stress magnitudes are commonly estimated using leak-off tests (LOTs). We demonstrate that accurate interpretation of LOTs is challenging due to factors such as the presence of pre-existing fractures, cement channelling, non-linear pressure build-up, and plastic and elastic leak off. Based on the choice of interpretation technique and calculations based on the assumption of tensile or shear failure, estimates of minimum horizontal stress may also vary by up to 6 MPa at depths of 3 km. The maximum horizontal stress magnitude is typically considered to be the most difficult to quantify, and we compare five methods (frictional limits, presence of drilling induces tensile fractures, presence of wellbore breakouts, wellbore breakout width, and shear plane failure). We show that at depths of ~2.5 km, estimates of maximum horizontal stress magnitude determined using these approaches can vary between 18-40 MPa. Finally, we compared empirical approaches for estimating rock strength using sonic velocity data with laboratory uniaxial compressive tests for a variety of stratigraphic units in the Cooper Basin. We show that, depending on the empirical correlation being used, rock strengths may be underestimated by 25-43% when comparing sonic velocity log derived rock strength to physical compression testing. In summary, our investigation into the uncertainty in principal stresses and rock strength estimation leads us to proposes enhancements to methodologies concerning density log preparation and filtering, check-shot calibration, low data density leak-off interpretation, and rock strength-sonic transit time relationships. Our results provide end-users with a better understanding of the uncertainties associated with stress and strength estimates that can be factored into geomechanical risk assessments.

Biography

Matthew is a final year PhD student at the Australian School of Petroleum and Energy Resources. Matthew is a 3-year ASEG research grant awardee with experience in potential field geophysics exploration. Matthews’ current goal is to see practical applications of his research to industry through optimised geomechanical workflows.