Monday, February 19
Data-driven analytics is transforming Oil and Gas – from more accurate interpretation of geological and seismic data to more sophisticated reservoir modeling, to improved drilling, completion, and production decisions. Analytics provide more robust and timely solutions to business problems. In this session, we will discuss the landscape of data-driven analytics, the promises, the challenges, and the solutions from both a technological and an organizational point of view.
This session discusses recent developments in the application of machine learning and advanced analytics in reservoir characterization. Speakers will cover a broad range of challenges including interpretation of seismic, well logs and tests, rock and fluid properties and integration of these data to build reliable G&G models, faster and cheaper. Practical tips will also be shared to select appropriate machine learning algorithms and how to avoid their common pitfalls when applied to reservoir characterization problems.
This session will explore applications of statistical inference, a fusion of geological and operations data, pattern recognition, and the real-time analysis of streaming data to drilling engineering. The most important application areas include improving drilling efficiency, detecting and mitigating drilling problems, reducing well construction costs, and minimizing non-productive time. In addition, we welcome discussions of drilling automation, which has historically faced skepticism from the Oil & Gas industry even as drilling activity has steadily increased and automation has become pervasive in other sectors.
In the past several years, there has been significant research and technological advancement in the area of field development and operation through the use of data mining and statistical modeling. Examples include better well rates estimation, more insightful and timely understanding of well behaviors, more effective completion operations, predicting failures before they happen, etc. This session will investigate the use of data analytics in the areas of production, completion, and operations to optimize production and reduce operating expenses.
The Oil and Gas industry has historically made extensive use of physics-based models for development and production. Such models have been difficult to apply and correctly interpret for unconventional resources. However, the scale of most unconventional plays is enormous, generating huge amounts of data and with it the hope that data-driven methods can successfully improve business results. This session will focus on applications of data analytics to the optimization of unconventional resource development and operation. We will explore how statistical models, machine learning, and artificial intelligence can use hard data (i.e. field measurements) to improve upon the “analysis by anecdote”, preconceived notions, and biases from conventional development that have dominated this segment of the petroleum industry.
Tuesday, February 20
Addressing everyday issues associated with Reservoir Management, Surveillance, and Optimization require models that are accurate and execute in seconds/minutes with meaningful results. So far our industry has been forced to sacrifice one in order to accomplish the other. Accurate models have proven to be unbearably slow, and fast techniques (e.g., reduced physics, reduced order, statistics-based models) approximate the problem to the edge of irrelevance. This session will focus on technologies that address both accuracy and speed, simultaneously.
The Oil and Gas industry continues to face increased regulations and the need for responsible environmental stewardship to maintain our license to operate. Data analytics is being used to improve safety and environmental performance through such techniques as text mining, pattern recognition, and real-time analysis of streaming sensor data. This session will explore some of the analytical opportunities identified in the safety and environmental arena.
Subsurface and surface facilities monitoring and operations continue to pose severe technical and cost challenges in Oil and Gas. Recent advancements of Internet-of-Things such as cheap sensors and ubiquitous video surveillance, data analytics and visualization technologies, and easy access to computing make it possible to provide real-time actionable information to operators. The effective adoption of these technologies will improve operation (e.g., prevention instead of reactive intervention) as well as efficiency. This session will present case studies for applying these new technologies in the Oil and Gas subsurface, facilities and maintenance space.
In the past few years, we have witnessed impactful and rapid advancement of AI technologies largely driven by business opportunities in the Internet industries. To harvest the benefit of these technologies to Oil and Gas, however, is non-trivial given our industry’s unique challenges such as datasets with rich features, complex physical systems and dynamic processes, data and model uncertainties, a strong requirement of domain knowledge, lack of workforce familiar with these technologies, etc. As a result, such advancement generates excitement as well as hypes and confusion in our industry. In this session, we would like to discuss emerging opportunities such as technical themes applicable to Oil & Gas applications and ways to address the challenges.