Overview
This research evaluated infrastructure to assess the offshore production platform integrity, and identified technologies that could aid in infrastructure life extension, as well as inform risk mitigation and maintenance needs.
More than 120,000 wells have been drilled in U.S. waters since the 1890s. These wells connect to a complex network of infrastructure that includes pipelines, platforms, subsea installations, ports, and terminals. The aging of this infrastructure, coupled with varying maintenance and monitoring strategies and extreme operational conditions, can impact the rate and severity of failure. As a result, industry is looking for technologies, analyses, and tools to help extend the life of the existing infrastructure and guide development of future installations to meet a lifespan of more than 50 years. This project utilized data, big data computing, and advanced analytics to evaluate the condition of the production infrastructure, assess potential hazards, and optimize the development and deployment of existing and new infrastructure technologies in the offshore environment. The data and analytical results produced from this task offer additional information to regulators and industry regarding the integrity of offshore infrastructure, as well as new insights that could enhance or establish new practices, policies, and regulations to improve safety, minimize equipment failure, reduce costs, and mitigate potential hazards.
Approach
This three-year project utilized data and models from NETL’s Offshore Risk Modeling (ORM) suite to perform advanced analyses on data pertinent to existing infrastructure. The findings provide novel analytical insight regarding the integrity, hazard trends, and methods to extend the life of existing infrastructure using various mitigation and maintenance methods.
During the first year of this project, researchers developed a robust database that incorporates information on the location, use, design, and condition of offshore production infrastructure, as well as trends in environmental conditions and experimental results detailing optimal operational conditions for current and new technologies. These data, along with data and tools available from ORM, were used to assess the current state of offshore infrastructure. In addition, the Advanced Infrastructure Integrity Modelling (AIIM) analytical framework was designed to support the assessment of platform infrastructure and associated hazards to help evaluate the optimal development and use of existing and new infrastructure technologies.
In year two, NETL researchers continued to collect, integrate, and assess additional infrastructure, environmental, and technology information. Researchers developed multiple machine learning and advanced models to assess the remaining lifespan of infrastructure, which included a Gradient Boosted Regression Tree, Artificial Neural Network, and a Geographically Weighted Regression Model. Documentation on the preliminary analytics and the more advanced models also began. In addition, development started on the NETL Common Operating Platform, which enables users to access, visualize, and analyze offshore structural, geologic, environmental, and incident data.
In the third year of this project, researchers maintained and released an integrated platform dataset through the Energy Data Exchange (EDX®) (Romeo et al., 2021). A technical report on this project’s data development, findings during preliminary analytics, and an overview of the advanced analyses was released (Nelson et al., 2021a). Machine learning and advanced analytical models were applied to the integrated platform dataset, and validation within the multi-model AIIM framework resulted in high-accuracy results of 95–97% when predicting the remaining lifespan of offshore platforms. This project was extended one quarter into year four for the release of the NETL Common Operating Platform, as well as associated data, model results, and advanced analytics.
Outcome
This project leveraged big data, big data computing, spatio-temporal tools from the Offshore Phase 1 ORM suite and coupled them with new data and advanced, spatial and machine learning algorithms to support intelligent analytics via an interactive, online platform to help identify and address hazards related to offshore infrastructure. Analytical findings from this effort offer a region-wide evaluation of platform integrity, insights from which could inform infrastructure maintenance and inspection needs and identify potential hazards for DOE, regulators, industry, and other stakeholders.
The AIIM framework offers novel, intelligent models that leverage techniques from machine learning, spatial analyses, and big data computing. Internally validated analytical outputs offer new insights on the current state of offshore infrastructure and potential hazards, and support predictive analytics for optimized offshore infrastructure development.
The NETL Common Operating Platform offers access to key data and analytical tools from this project. This interactive platform allows DOE, regulators, industry, and other stakeholders to interact with data and analytics to better inform decisions, such as those related to reducing infrastructure hazards, costs, and extend infrastructure life for current and alternative uses.
Research Products
This project, completed in FY21, resulted in the following research products:
· AIIM: The Advanced Infrastructure Integrity Modelling (AIIM) analytical framework offers a multi-model, big data-driven approach to evaluate infrastructure integrity. While built to assess the current state of platforms in the Gulf of Mexico, AIIM is adaptable to forecast alternative features for a range of infrastructure types.
· NETL’s Common Operating Platform: An interactive online platform, securely hosted through EDX®, offers access to the data and analytics from this project. This platform enables users to identify and assess critical hazards related to offshore infrastructure and help inform decisions to reduce potential hazards, costs, and extend infrastructure life for DOE, regulators, industry, and other stakeholders.
· Comprehensive platform dataset has been released on EDX® and through the NETL Common Operating Platform.
Applied Machine Learning Model Comparison: Predicting Offshore Platform Integrity with Gradient Boosting Algorithms and Neural Networks
Dyer, A., Zaengle, D., Duran, R., Nelson, J., Wenzlick, M., Wingo, P., Bauer, J., Rose, K., and L. Romeo. (In Review, 2021). Applied Machine Learning Model Comparison: Predicting Offshore Platform Integrity with Gradient Boosting Algorithms and Neural Networks. Marine Structures.
Forecasting Platform Integrity with Machine Learning & Advanced Analytics for Reuse Optimization Strategies and Risk Prevention
Romeo, L. , and Bauer, J., “Forecasting Platform Integrity with Machine Learning & Advanced Analytics for Reuse Optimization Strategies and Risk Prevention,” Carbon Management and Oil and Gas Research Project Review Meeting Aug. 26, 2021. https://edx.netl.doe.gov/offshore/wp-content/uploads/2021/12/Forecasting-Platform-Integrity-with-Machine-Learning_NETL_08262021.pdf
Forecasting Offshore Platform Integrity: Applying Machine Learning Algorithms to Quantify Lifespan and Mitigate Risk
Romeo, L., Dyer, A., Bauer, J., Barkhurst, A., Duran, R., Nelson, J., Sabbatino, M., Wenzlick, M., Wingo, P., Zaengle, D. and Rose, K. 2021. Forecasting Offshore Platform Integrity: Applying Machine Learning Algorithms to Quantify Lifespan and Mitigate Risk. Machine Learning in Oil & Gas. April 15, 2021. Virtual.
A Geostatistical Analysis of Offshore Oil Platform Lifespan
Nelson, J., Romeo, L., Duran, R., and Bauer, J., “A Geostatistical Analysis of Offshore Oil Platform Lifespan,” NETL-ORISE Powerful Poster Series (online), March 23, 2021.
Forecasting Offshore Platform Integrity and Lifespan – Improving Safety and Reliability
Romeo, L., Dyer, A., Zaengle, D., Nelson, J., Wenzlick, M., Duran, R., Sabbatino, M., Wingo, P., Barkhurst, A., Bauer, J., and Rose, K., “Forecasting Offshore Platform Integrity and Lifespan – Improving Safety and Reliability,” invited presentation at the ICCOPR Quarterly Meeting (virtual), December 9, 2020.
Assessing Current and Future Infrastructure Hazards: Forecasting Integrity using Machine Learning and Advanced Analytics
Romeo, L., Dyer, A., Zaengle, D., Nelson, J., Wenzlick, M., Duran, R., Sabbatino, M., Wingo, P., Barkhurst, A., Bauer, J., and Rose, K. 2020. Assessing Current and Future Infrastructure Hazards: Forecasting Integrity using Machine Learning and Advanced Analytics. Oil and Gas Project Review Meeting. October 26, 2020. Virtual. https://www.osti.gov/servlets/purl/1768713
Advanced Geospatial Analytics and Machine Learning for Offshore and Onshore Oil & Natural Gas Infrastructure
Justman D., Romeo, L., Barkhurst, A., Bauer, J., Duran, R., Dyer, A., Nelson, J., Sabbatino, M., Wingo, P., Wenzlick, M., Zaengle, D., Rose, K. invited talk. Advanced Geospatial Analytics and Machine Learning for Offshore and Onshore Oil & Natural Gas Infrastructure. GIS Week 2020. October 6-7, 2020. Virtual. https://www.osti.gov/servlets/purl/1767074
Machine Learning Driven Forecasting of Offshore Infrastructure Integrity
Romeo, L., Rose, K., Barkhurst, A., Duran, R., Dyer, A., Nelson, J., Sabbatino, M., Wenzlick, M., Wingo, P., and Zaengle, D., “Machine Learning Driven Forecasting of Offshore Infrastructure Integrity,” in preparation.
Building Big Data Geospatial Tools for a Common Operating Platform: Cumulative Spatial Impact Layers
Romeo, L. and Barkhurst, A. Building Big Data Geospatial Tools for a Common Operating Platform: Cumulative Spatial Impact Layers. DOE GIS Users Group Meeting. September 10, 2020. Virtual. Invited presentation.
Building an Analytical Framework to Measure Offshore Infrastructure Integrity, Identify Risk, and Strategize Future Use for Oil and Gas
Dyer, A., Romeo, L., Wenzlick, M., Bauer, J., Nelson, J., Duran, R., Zaengle, D., Wingo, P., and Sabbatino, M., “Building an Analytical Framework to Measure Offshore Infrastructure Integrity, Identify Risk, and Strategize Future Use for Oil and Gas,” accepted to the Esri User Conference, San Diego, CA, July 13–15, 2020, https://www.esri.com/en-us/about/events/uc/overview.
Harnessing the Power of DOE Data Computing for End-User Analytics
Rose, K., Barkhurst, A., Mark-Moser, M., Romeo, L., and Wingo, P., “Harnessing the Power of DOE Data Computing for End-User Analytics,” SMART Webinar June 25, 2020, https://www.youtube.com/watch?v=G5oUWSb-kHc&feature=youtu.be.
Geospatial Machine Learning to Mitigate Offshore Infrastructure Hazards
Dyer, A., Zaengle, D., Romeo, L., Wingo, P., Wenzlick, M., Sabbatino, M., Bauer, J., Nelson, J., and Duran, R., “Geospatial Machine Learning to Mitigate Offshore Infrastructure Hazards,” accepted, Esri User Conference, San Diego, CA, July 13–17, 2020, https://www.esri.com/en-us/about/events/uc/overview.
Intelligent Risk Modeling for Offshore: Assessing Current and Future Infrastructure Hazards
Romeo, L., Rose, K., Barkhurst, A., Duran, R., Dyer, A., Nelson, J., Sabbatino, M., Wenzlick, M., Wingo, P., and Zaengle, D., “Intelligent Risk Modeling for Offshore: Assessing Current and Future Infrastructure Hazards,” June 2020.
Explore research products that are related to this project.
See related Phase 2 projects.
*Image Source: NETL
Number of platform incidents across the Gulf of Mexico, between 2006–2017, by water depth (data from BOEM, BSEE).
*Image Source: NETL
Maximum velocity magnitude (range 0 to 2.75 m/s) recorded for each month (each column is roughly a season) from twelve years (2003-2014) of sea-surface velocity, presented as a function of longitude and latitude.
Contacts
Jennifer Bauer
Co-Principal Investigator
Lucy Romeo
Research Geospatial Scientist
Co-Principal Investigator
Kelly Rose
Offshore Portfolio Lead
Co-Principal Investigator
Roy Long
Offshore Portfolio Technical Manager
Effective Resource Development
Alexandra Hakala
Senior Fellow (Detail)
Geological & Environmental Systems
Philip Reppert
Associate Director
Geological & Environmental Systems
