The Importance of Subsurface Characterization and Monitoring During Development and Operation of Underground Gas Storage Facilities

Natural gas consumption is expected to grow significantly in coming decades in response to cleaner energy initiatives. Underground gas storage (UGS) will be key to addressing supply and demand dynamics for this transition to be successful. This technical paper will demonstrate the importance of an integrated subsurface characterization and monitoring approach not only for the construction of UGS, but also to guarantee safe and efficient operation over many decades. Key to long-term success of UGS is maximizing working capacity with respect to volume and pressure and maintaining well injection and withdrawal capabilities. Initial assessment steps involve determination of maximum storage capacity and an estimation of required cushion gas volumes. In similar manner to conventional field evaluation, we perform an integrated geological, geophysical, petrophysical and geomechanical characterization of the subsurface. However, for UGS facilities, the impact of cyclic variations of reservoir pressures on subsurface behavior and cap rock integrity also needs to be evaluated to determine safe operating limits at every point in time during the life of the UGS project. The holistic approach described above allows the operator to optimize the number of wells, well placement, completion design, etc. to ensure long-term safe and efficient operations. Furthermore, close integration of subsurface understanding with optimization of surface facilities, such as the compression system, is another critical component to ensure optimum UGS performance and deliverability. Moreover, another important task of the final phase of UGS facilities design involves enablement of sustainable operation through an asset integrity management plan. This phase is articulated around reservoir surveillance plans that monitor pressure, rock deformation and seismicity, in addition to regular wellbore inspection. Through close operations monitoring and the utilization of advanced data analytics, observations are compared to existing models for validation and operation optimization. Importantly we show that adapted monitoring programs provide critical long-term insight regarding the field response during successive cycles, leading to significant improvement in working gas capacity. A key consideration of this integrated UGS development strategy is based on the seamless integration of subsurface characterization, wellbore construction and well completions to ensure technical and commercial flexibility. The approach also emphasizes the integration with surface facilities design to ensure a true "Storage to Consumer" view for effective de-bottlenecking. Coupled with integrated subsurface integrity monitoring, this ensures a faster, cost efficient and safe response to the construction and operation of UGS facilities.

Subsurface Characterization Using Ensemble Machine Learning

Reservoir characterization is an ambitious challenge that aims to predict variations within the subsurface using fit-for-purpose information that follows physical and geological sense. To properly achieve subsurface characterization, artificial intelligence (AI) algorithms may be used. Machine learning, a subset of AI, is a data-driven approach that has exploded in popularity during the past decades in industries such as healthcare, banking and finance, cryptocurrency, data security, and e-commerce. An advantage of machine learning methods is that they can be implemented to produce results without the need to have first established a complete theoretical scientific model for a problem – with a set of complex model equations to be solved analytically or numerically. The principal challenge of machine learning lies in attaining enough training information, which is essential in obtaining an adequate model that allows for a prediction with a high level of accuracy. Ensemble machine learning in reservoir characterization studies is a candidate to reduce subsurface uncertainty by integrating seismic and well data. In this article, a bootstrap aggregating algorithm is evaluated to determine its potential as a subsurface discriminator. The algorithm fits decision trees on various sub-samples of a dataset and uses averaging to improve the accuracy of the prediction without over-fitting. The gamma ray results from our test dataset show a high correlation with the measured logs, giving confidence in our workflow applied to subsurface characterization.

Technology Focus: Formation Evaluation (August 2021)

Leading into the third quarter of this year, I am honored to be able to highlight and share three impactful SPE papers that demonstrate integration at its best. In reviewing the papers, five main technical themes emerged. These include * Machine learning and artificial intelligence as applied to formation evaluation * Production analysis methodologies and their effect on understanding rock characterization and behavior * Subsurface characterization primarily focused on rock typing and permeability * Tool advancements (openhole, cased-hole, or laboratory-based tools) * Subsurface-to-production integration across subdisciplines (e.g., geology, geochemistry, petrophysics, and engineering) The latter is the common thread between the three papers recommended and discussed here. In this new decade, the prevalence of integration is at the forefront of the scientific community. Every discipline, scientist, or company has a way in which they define the term “integration.” Regardless of how you define the effort that links disciplines quantitatively, the importance of constraining subsurface characterization to link it to production results and drive toward a predictive model is a critical accomplishment for our industry. As such, I’d like to highlight three papers in this feature (OTC 30644, SPE 201417, and SPE 202683) and the knowledge and workflow applications they define and demonstrate. Sharing these integrated work flows with the community aids in teaching and leads to best-practice components of integrative studies. These efforts also share and demonstrate how to bridge the gap between in-situ characterization and wellhead performance prediction and results—in other words, the static-to-dynamic link between rock and fluid properties as quantified and how they will inevitably produce hydrocarbon through the rock and fluid interactions. Recommended additional reading at OnePetro: www.onepetro.org. SPE 201334 Combined Experimental and Well-Log Evaluation of Anisotropic Mechanical Properties of Shales: An Application to Wellbore Stability in the Bakken Formation by Saeed Rafieepour, The University of Tulsa, et al. SPE 201486 A New Safe and Cost-Effective Approach to Large-Scale Formation Testing by Fluid Injection on a Wireline Formation Tester by Christopher Michael Jones, Halliburton, et al. SPE 201735 Integrated Reservoir Characterization With Spectroscopy, Dielectric, and Nuclear Magnetic Resonance T1-T2 Maps in a Freshwater Environment: Case Studies From Alaska by ZhanGuo Shi, Schlumberger, et al.