Putting Big Data To Work For Offshore Insights

The Deepwater Horizon disaster in 2010 released more than 130 MMgal of oil into the Gulf of Mexico (GoM) over 87 days and is the largest accidental marine oil spill to date. Teams of experts assembled to mitigate the catastrophe, but rapid emergency response and control efforts were hindered by inadequate tools and large uncertainties associated with this extreme offshore play.

The tragedy emphasized the need for tools to assess risks and predict behaviors of a range of offshore engineered-natural systems from the reservoir to the shore. Offshore U.S. oil production in the GoM accounts for 16% of total U.S. crude production, and globally, offshore drilling accounts for less than 5% of the world’s wells, yet accounts for 30% of total global oil production. The efficiency and production from offshore wells are notable; however, as offshore development continues to expand and venture into new territories, methods and resources to improve systemwide knowledge and support planning, prevention and rapid response needs for future spills are required to protect human life and the environment.

The National Energy Technology Laboratory (NETL) is one of 17 U.S. Department of Energy (DOE) national laboratories and the only laboratory dedicated to fossil energy research. NETL’s researchers work to develop, integrate and mature technology solutions to enhance the nation’s energy foundation and protect the environment for future generations. Ensuring the safe, effective development of the nation’s oil and gas resources is a key priority in this mission. The advances taking place within NETL’s offshore research portfolio are available to help reduce the likelihood of damaging geohazard-induced events that are associated with offshore hydrocarbon E&P. Additionally, these technologies can reveal the economic potential of these domestic resources in greater detail.

Modeling offshore risk
Developed by a team of geodata science researchers at NETL, the Offshore Risk Modeling (ORM) suite improves the prediction and evaluation of offshore systems by addressing key gaps in knowledge and the need for spill prevention aligned to data-computing solutions across time, space or regions. Fundamentally, many of the tools offered in the ORM suite leverage Big Data and Big Data computing to help users assess local questions (e.g., rig placement risks, geohazard and reservoir property predictions) using a “big picture” approach. Working with existing commercial tools and software systems, this first-of-its-kind suite provides a comprehensive framework for future predictions, analyses and visualizations to better inform offshore hydrocarbon E&P, predict the fate and transport of potential spills, improve prediction of subsurface properties for areas with no prior data and identify regional vulnerabilities.

The NETL team designed the tools and models in the ORM suite to be easily accessible to the industry, regulators and scientists for a range of commercial-and research-related efforts. The ORM suite comprises digital applications that can be deployed directly on a personal device, such as a desktop computer, laptop or tablet, or run through a virtual cloud-computing framework. All tools are available for download or web use. In addition, there are instances of the ORM suite being developed for specific users in an online, Dockerized computing platform to offer greater portability and utility, as the models and tools are maintained on one server while data are hosted and/or streamed in real time from authoritative sources.

ORM suite tools can handle large amounts of data in the range of gigabytes to petabytes depending on the scale of the analyses being performed, efficiently utilizing millions of data from numerous sources. This makes the ORM suite a comprehensive solution for providing predictions of potential hazards and mitigating challenges faced during common offshore oil and gas activities.

When the tools and models are used either individually or together, the ORM suite can help the offshore energy industry with worst-case discharge planning and preparedness, evaluation of site-specific metocean geohazards and improved prediction of subsurface properties for reserves and drilling calculations.

Components
The ORM comprises two data management and curation tools and six analytical components that allow end users to assess and combine data surrounding many environmental, socioeconomic and geological factors. When used together, the analytical tools provide more holistic analyses that aid in predicting and preventing risks and operational costs. The analytical tools included in the ORM at present are noted below.

The Climatological Isolation and Attraction Model (CIAM) is a quick prediction and response tool that applies mathematical theories of dynamical systems and metocean data, including real-time ocean current patterns, to determine where oil and other particles in the ocean (e.g., debris, hazardous waste and plankton) are likely to be attracted or repulsed. CIAM offers offshore commercial and scientific communities a novel and efficient way to summarize big ocean current and/or large wind data and generalizes previous approaches by calculating transport patterns independent of where a spill may originate.

The Blowout Spill Occurrence Model (BLOSOM) is an open-source, comprehensive modeling program that predicts the fate and transport of oil for hypothetical and actual offshore spill events to support planning and spill response preparedness needs. BLOSOM is the first open-source oil spill and blowout model in 4-D, which has been compared to and validated against traditional and industry-applied spill models. Built upon a flexible framework, BLOSOM consists of several modules to help visualize and predict the scope of environmental damage in the event of a blowout, which include behavior in high-pressure environments, subsea dispersants, gas and hydrate dynamics and other features as well as utilizing knowledge of how particulates move throughout all levels of the water column.

Cumulative Spatial Impact Layers (CSIL) is a geographic information system-based tool that rapidly identifies and quantifies potential socioeconomic and environmental risk. The CSIL tool is capable of handling multiple disparate datasets, can measure data density and produce multivariable layers that identify vulnerabilities within a given area. It can be used in concert with other ORM tools to support planning, such as ingesting BLOSOM simulation outputs and summarizing the potential risks or response availability associated with hydrocarbon events. The CSIL tool can be used to assess onshore response capabilities to determine what areas are at risk and how prepared emergency officials are in the event of a spill.

The Spatially Weighted Impact Model (SWIM) is a decision-support tool that incorporates relationships among oil spill information, response availability and potential risk. Users can apply weights to evaluate modeled spill events based on potential impacts, the magnitude of the oil spill and response preparedness against a “baseline” scenario, which might consist of what different users consider to be worst-case events. This feature allows users to rank and compare different scenarios and come up with varying plans of action for planning and response needs.

The Subsurface Trend Analysis improves the predictions of subsurface property values using a combination of geologic knowledge and advanced spatio-temporal statistical methods. The approach leverages information about geologic systems to improve prediction of subsurface properties critical for reserves calculations, exploration and resource identification, geohazard prediction, drilling safety and improved well design.

The Variable Grid Method communicates uncertainty for data and model results, which is critical when utilizing multiple tools and approaches. The Variable Grid Method tool provides the flexibility to use different data types and uncertainty qualifications, and it preserves overall trends and patterns observed within the data while enabling users to customize the analysis and final product to meet their needs and best communicate results in an intuitive manner.

Two ORM applications support data acquisition, curation and collaboration useful to both the offshore and broader fossil energy community. There has been growing recognition of the value and challenges associated with data preservation, curation, management and reuse, and the ORM team has been addressing these needs for almost a decade. This resulted in the building of the Energy Data eXchange (EDX) and the GeoCube to address data access, curation and reuse challenges specific to fossil energy and ORM users.

EDX is the DOE’s Office of Fossil Energy’s (FE) virtual data library and laboratory. Launched in 2011 and based on data access challenges identified during the Deepwater Horizon spill response, FE’s EDX platform is an online data curation and collaboration platform. EDX fosters and supports the life cycle of data for FE users and includes public and private resources for FE teams. Presently, EDX curates products from fossil energy research, including data, tools and models, and it supports select virtual analytical needs of FE users, such as those performed by the virtual ORM suite.

GeoCube is a custom web-mapping application hosted via EDX that allows users to quickly view and visualize spatial data, download resources, identify overall trends and patterns in the data and share these discoveries with others. The application allows users to upload and visualize their own data in a user-friendly manner as well as spatial datasets served via the EDX platform.

Capabilities
The ORM suite can be applied to improve safe and efficient operations in offshore systems for a range of stakeholder needs.

Simulating 4-D oil spill and blowout scenarios: BLOSOM can model multiple hypothetical and historic oil spill simulations at various locations and times. BLOSOM simulations can be coupled with the CIAM to enhance and validate the predicted fate and transport of oil spills. BLOSOM does not presently pull data through EDX, although one could host BLOSOM-compatible data on EDX for download, which can be visualized through GeoCube. Together, BLOSOM and CIAMcoupled models offer an improved prediction of the fate of oil spill particles, which offers critical information to inform oil spill prevention, response and preparedness efforts.

Identifying critical subsurface characteristics: Various subsurface conditions can negatively impact hydrocarbon exploration safety and costs. The Subsurface Trend Analysis can be used to define potential hazards based on geologic data and expertise to improve predictions of subsurface characteristics that may impact drilling, such as areas of overpressure. The uncertainty associated with these predictions can be quantified and visualized using the Variable Grid Method. Together, the Subsurface Trend Analysis and Variable Grid Method help decision-makers pinpoint areas where overpressure might be present and make confident decisions regarding mitigation strategies that ensure safe operating conditions.

Evaluating response preparedness: CSIL can be used to summarize and visualize oil spill fate and transport from BLOSOM with the spatial and temporal distribution of socioeconomic and environmental variables. Overlapping these data with CSIL helps identify gaps in response infrastructure readiness. SWIM can compare and rank multiple scenarios to strategically identify regions where additional prevention equipment is needed to improve oil spill preparedness.

Assessing offshore infrastructure integrity: Data from EDX and GeoCube have been leveraged to identify and  visualize historic pipeline and platform failure incidents. CSIL has summarized metocean data (e.g., wave height, current velocity and wind speed) to statistically evaluate the effect of extreme environmental conditions on infrastructure integrity over time.

Implications
The approach of the ORM suite is a paradigm shift from how the industry and regulators traditionally evaluate the offshore environment. In the past, local analyses and smaller-scale datasets were the primary focus in informing decisions. Industry and regulators have conventionally based their models on limited data or information; however, because of the advanced, data-computing nature of the tools in the ORM suite, end users now have access to streamlined, analytical results that incorporate larger and more diverse spatiotemporal systems than conventional models or methods can afford.

Response to the Deepwater Horizon blowout met challenges in critical areas. Traditional response resources, such as booms, oil skimming methods and dispersants, were used in cleanup efforts, but oil traveling underneath the sea surface at such extreme water depths posed new, unforeseen and lasting challenges to both containment and fate analyses. Predictions of oil transport were inadequate due to limitations of spill models available at the time, leading to oil materials such as tarballs washing up in unforeseen locations. This had immediate and long-lasting impacts on tourism and other ocean-dependent industries in the area. The efficiency of response operations also was affected by a lack of knowledge about the geohazards, subsurface reservoir properties and Deepwater Horizon well infrastructure pre-and post-blowout, which hampered efficiency and confidence in subsea and subsurface spill control efforts.

The tools of the ORM suite were designed to aid in preventing these and other environmental hazards from occurring in similar spill scenarios through better prediction methods and risk assessment. With more information on oil fate and transportation, as well as emergency response preparedness and geologic properties from drilling operations, many of the challenges from the Deepwater Horizon spill may have been prevented or had their impact drastically reduced.

Outside of spill events, similar data and modeling challenges impact daily operations, planning and decisions by regulators and the industry alike. While the capabilities of the ORM suite were influenced and informed by the lessons learned from the catastrophic Deepwater Horizon event, they also address daily, “normal” challenges, leveraging Big Data to inform smallscale needs and preventing vulnerabilities.

While initially developed with a focus on the GoM, the capabilities of the ORM suite are not limited to that region and have been extended to other U.S. and international waters. The ORM suite is flexible and can be adapted and integrated into other models to suit end users’ needs and account for differences in location. In practice, several components of the ORM have been leveraged to the DOE and external users to assess operational and environmental risks for regions around the U.S., including offshore Southern California and the Gulf of Alaska. Components of the ORM also have been utilized beyond the U.S., including users in Australia, Brazil, India, Mexico, Spain and the U.K.

By applying novel and efficient methods to cross-examine data across space and time, the ORM suite supports predictive analyses for safer, more prudent efforts and rapid-response, real-time assessments. The ORM suite is able to adapt from one need to the next, filling knowledge gaps, reducing resource uncertainty, assessing geohazard potential and supporting decision-making, thereby improving operational efficiency and safety.