Fiber Seismic Tomography for Geothermal Exploration
Proceedings Overview
Event: 49th Workshop on Geothermal Reservoir Engineering
Location: Stanford University, Stanford, California
Dates: February 12-14, 2024
Paper Reference: SGP-TR-227 1
Authors: Ettore Biondi, Jiaxuan Li, Valey Kamalov, Zhongwen Zhan
Affiliation: California Institute of Technology
Contact: ebiondi@caltech.edu
Keywords
Key Points: Seismic tomography, distributed acoustic sensing, Long Valley caldera
Abstract
Objective: To present a new seismic tomographic approach utilizing machine-learning on distributed acoustic sensing (DAS) data.
Location of Study: Long Valley caldera, California.
Methodology: Uses existing telecommunication fibers to detect subsurface anomalies associated with geothermal reservoirs.
Significance: Enhanced imaging methodology promising to lower exploration costs for geothermal energy.
Introduction
Need for Identification: Finding subsurface structures with high-temperature fluids is vital for geothermal energy strategies.
Existing Methods Limitations:
Magnetotellurics provide poor resolution for deeper systems.
Seismic imaging can characterize mesoscale velocity structures but relies on deployment density of devices.
Advancements with DAS:
DAS technology enables recording seismic data from telecommunication fibers with greatly improved spatial-temporal resolution.
The ability to use long fibers as dense sensor arrays enhances data collection significantly.
### Existing Methods Limitations: - **Magnetotellurics**: This electromagnetic technique is often limited by its ability to resolve subsurface features efficiently. While effective in mapping large geological structures, magnetotellurics tend to provide poor resolution for deeper geothermal systems, making it challenging to identify specific high-temperature fluid reservoirs that are crucial for geothermal energy exploration. - **Seismic Imaging**: Traditional seismic imaging methods can characterize mesoscale velocity structures, yet their effectiveness heavily depends on the density and strategic placement of measuring devices. Sparse deployment may lead to gaps in data resolution, impeding accurate imaging of complex geological formations. ### Advancements with DAS (Distributed Acoustic Sensing): - **Technological Breakthrough**: DAS technology has emerged as a revolutionary method for collecting seismic data by utilizing the optical fibers embedded in existing telecommunications infrastructure. This approach offers significantly enhanced spatial and temporal resolution compared to conventional methods, facilitating better detection of subsurface anomalies associated with geothermal reservoirs. - **Dense Sensor Arrays**: A key advantage of DAS is its capability to employ extensive lengths of fiber optic cables as dense sensor arrays. This feature substantially increases the spatial coverage of seismic data collection, allowing for more detailed imaging of geological structures. The high density of measurements captured in a continuous fashion enables researchers to observe intricate patterns within the subsurface, thereby improving the accuracy of geothermal exploration efforts. - **Cost-effectiveness**: By leveraging existing structures, DAS demonstrates a more economical approach to conducting seismic surveys. Instead of deploying numerous standalone sensors, the utilization of pre-installed fiber optics can lead to substantial reductions in operational costs while maintaining data quality and completeness.
Technology Description
DAS Deployment: DAS can be applied both on telecom fibers and on cables in boreholes, providing versatility in data collection environments.
High-Temperature Suitability: Fiber cables resist high temperatures, ensuring operational reliability in geothermal settings.
Spatial Resolution Capability: Achieves km-resolution which is critical for identifying high-temperature formations.
Imaging Methodology
Mathematical Framework: Uses Eikonal equation for modeling traveltimes in seismic tomography.
Double-Difference Approach: Employs matrix-free inversion and non-linear optimization schemes, facilitating calculations across extensive data sets.
Trade-Off Resolution: An alternate-direction method is used for optimizing event locations and velocity models iteratively.
Synthetic Example
Setup: A 2D synthetic test was performed with eleven seismic events.
True Model: Introduced a Gaussian low-velocity anomaly to assess the method's efficacy.
Results: Demonstrated successful imaging of the anomaly with a remarkable average traveltime precision of 0.1 ms after 26 iterations.
Field Data Application
Location: Long Valley caldera, previously identified for its hydrothermal energy production since 1984.
DAS Configuration: Utilized two DAS arrays across a 100-km transect with more than 9000 channels.
Data Collection Span: Captured over 6000 seismic events to construct noise-filtered datasets, with a selection of 843 earthquakes based on adequate SNR.
Initial Models
Depth Specifics: Subsurface properties examined at two depths relative to mean sea level, providing velocity anomalies and VP/VS ratios.
Machine Learning Integration: Employed PhaseNet-DAS for precise arrival time captures of seismic waves, significantly scaling the size of datasets.
Results and Discussion
Tomographic Output: High-resolution imaging identified significant low-velocity anomalies correlated with geothermal systems.
Continuity with Other Methods: Results validate findings from magnetotelluric studies and surface hydrothermal spring data.
Implications: The findings underscore the method's promise for localizing geothermal resources while potentially minimizing exploration costs.
### Results and Discussion - **Tomographic Output:** High-resolution imaging utilizing advanced seismic tomography techniques revealed significant low-velocity anomalies in the subsurface. These anomalies are indicative of the presence of high-temperature fluids, which are essential components of geothermal systems. The detailed imaging not only highlights these features but also provides insights into the geological structures and compositions aligned with geothermal reservoirs. - **Continuity with Other Methods:** The results obtained through this methodology corroborate findings from previous magnetotelluric studies, which have often indicated similar low-velocity zones in geothermal contexts. Additionally, data derived from surface hydrothermal spring observations align with these tomographic results, reinforcing the reliability of the method. This agreement with established techniques validates the innovative approach employed in this study, showcasing its robustness in geothermal exploration. - **Implications:** The implications of these findings are significant for the field of geothermal energy exploration. The enhanced imaging ability promises more precise localization of geothermal resources, allowing for more targeted drilling efforts. Given the lower operational costs associated with using existing telecommunications fiber infrastructure for data collection, this method offers the potential to reduce overall exploration expenditures. Thus, leveraging this technology may lead to more sustainable and economically viable geothermal energy projects in the future
Conclusion
Innovative Workflow: Demonstrated a new matrix-free adjoint traveltime tomography for geothermal exploration.
Potential Benefits: Method shows promise for improved geophysical insights, aiding in geothermal energy efficiency.
Broader Applicability: The technology could transform existing fiber networks into effective geothermal sensors, reducing operational costs.
### Innovative Workflow - The innovative workflow focuses on the application of a **matrix-free adjoint traveltime tomography** method tailored specifically for geothermal exploration. This approach streamlines the data processing phase, allowing for enhanced accuracy and faster computation times. By avoiding the complexities and resource intensity of traditional matrix-based approaches, the methodology provides a more efficient framework for analyzing seismic data related to geothermal reservoirs. ### Potential Benefits - The proposed method offers significant advantages in terms of **geophysical insights**. By utilizing advanced algorithms and machine learning techniques, this tomographic approach enhances the resolution of subsurface imaging. This increased clarity can lead to a more comprehensive understanding of geothermal systems, thereby improving the overall efficiency of geothermal energy extraction. - In addition, the methodology can lower costs associated with data acquisition and processing, making geothermal energy projects more feasible and attractive to investors and stakeholders. ### Broader Applicability - The implications of this technology extend beyond just geothermal exploration. With the capability to repurpose existing fiber optic networks, this technology paves the way for widespread deployment in various geoscientific applications, including monitoring natural hazards and environmental changes. By converting standard telecommunication fibers into dense arrays of seismic sensors, operational costs can be substantially reduced, which may encourage more extensive geological surveys and lead to improved management of geothermal resources. - Additionally, this approach holds potential for adaptive monitoring systems that can continuously assess geothermal site conditions in real-time, thus facilitating timely interventions and decision-making in geothermal energy management.
Acknowledgments
Support and Contributions: Thanked OptaSense, California Broadband Cooperative, and NSF for their support of the research efforts.
References
Key Studies and Authors Cited:
Aizawa et al., 2009 | Ajo-Franklin et al., 2019 | Bailey et al., 1976 | Biondi et al., 2021 | and more.
Technical Background Sources: Reviews and studies on seismic tomography applications and methodologies.