Data Science in Oil and Gas

In the old days, modest data stores were centralized and easy to manage. Later, we have added tremendous number of sensors and devices measuring everything possible to measure leading to an exponential growth in data. Such data contain valuable information if we can make sense and do correlations. Artificial Intelligence and Machine Learning promises to unveils the potential opportunities empowered by the computational power and intelligent algorithms. Data is transformed into information, Information gives insights and subsequently actionable intelligence.

Artificial Intelligence (AI) and Machine Learning (ML) can be of a big value for upstream, midstream and downstream operations. In upstream, drilling and downhole activities different types of sensing devices bring tremendous amount of data that requires more of an intelligent model to analyze and conclude outcomes. Data Analytics techniques and Machine Learning modules are being utilized for analysis in such applications. On the other side, midstream and downstream have big potential applications that has not been tackled earlier in modifying, tuning and rectifying process control systems for optimal production and environmental impact.

Operations and Logics in process control systems in oil and gas and for decades are based only on physics equations that dictates parameters along with operators’ interactions based on experience and operating manuals. Artificial Intelligence and Machine Learning algorithms can look into the dynamic operational conditions, analyze them and suggest optimized parameters that can either directly tune logic parameters or give suggestion to operators. Interventions by such intelligent models leads to optimization in cost, production and safety.

Data Science Applications in Oil and Gas

As importance of introducing the intelligence aspect of utilizing data is realized, among the oil and gas industry several applications are identified in literature and by industrial experts to have big potential for artificial intelligence. There are an increasing number of areas in which machine learning and advance analytics are aiding oil and gas operating companies to improve their operational performance.

Drilling Systems

Solutions being developed and enhanced based on machine learning models to act as an advisor for rig operators empowering them with real time and quick options for operations. These closed-loop drilling solutions are based on machine learning models aiming to achieve greater operational performance while maintaining operational safety metrics.

Well Monitoring and Optimization

As normally oil and gas fields have multiple wells, it is always a target of operating companies to maximize and optimize the throughput of these fields. British Petroleum has been using advanced analytics over 600 of its wells analyzing flow data based on which ultimate recovery decisions are being taken. The pilot test has shown positive results leading to a plan to rollout the solution to over 4,000 wells that British petroleum operates worldwide.

Supply chain management and Optimization

There are a lot of supply chain dependencies in oil and gas industry such as trucking materials in and out, chemicals, chemical injections, pressure pumping, trucking in water and producing water. Greater visibility of the supply chain can reduce traffic jams, accidents and death. Traffic accidents are the main cause of death in the oil and gas industry.

Motivation of Utilizing Data Science in Oil and Gas

There are a number of factors that makes exploring Artificial Intelligence potential application in oil and gas on the top of the innovation and evolution lists of the industry especially in the digitalization era. The following are some points in this context.

1. Oil and Gas industry is adopting a “lower for longer” strategy in which companies are considering technologies the decreases cost, increase production, increase assets lifetime and achieve at the same time sustainability and environmental targets.
2. Industrial Firms believes that the total business value of new technologies such as Industrial IoT and Artificial Intelligence/Machine Learning sums to many billion dollars.
3. Industrial IoT and Evolution in sensing technologies is introducing tremendous amount of data that raise opportunities and infrastructure for AI and Machine Learning. Based on the energy research firm, zpryme, 82% of utilities in North America believe that IIoT is an important technology trend and 31% of them have a comprehensive plan for IIoT. For Machine Learning, the same study concludes that 83% believes that machine learning is practical for utilities.
4. In oil and gas plants, small percent increase in some projects can mean huge amount of dollars. One, two or three percent increase in a very large producing well could mean hundreds of thousands and even millions of dollars a day and savings and operational improvements.

Production Optimization, Planning and Monitoring

Improving uptime and performance of turbo machinery such as pumps, compressors and turbines helps companies to maximize return on an asset investment and increase productivity by reducing downtime and minimizing unnecessary maintenance that is very critical in a lot of cases because it allows companies to operate under leaner and meaner infrastructure.

Utilizing Machine Learning Models enhances and automates maintenance with the predictive or prescriptive analytical solutions contributing to a reduction in the total cost of ownership.

An example is the vibration analysis of machinery, a cognitive model can forecast equipment failure at least twenty times sooner than traditional analysis methodologies based on a pilot test conducted by SparkCognition®.

Another example of an application in this field is a model that reads tens of thousands of maintenance manuals and predict problems and prescribe solutions.

References

This article "Data Science in Oil and Gas" is from Wikipedia. The list of its authors can be seen in its historical and/or the page Edithistory:Data Science in Oil and Gas. Articles copied from Draft Namespace on Wikipedia could be seen on the Draft Namespace of Wikipedia and not main one.

---