With the oil and gas industry in the midst of the Great Crew Change and low oil prices demanding increased operational efficiency, petroleum engineers who dabble in digital and tech-driven millennials interested in geology are in high demand. Remote pipeline maintenance, 3D digital reservoir mapping and big data are making petroleum data scientists a necessary addition to even the smallest of companies, and training for these up-and-comers is rapidly opening up across Silicon Valley and Houston, Texas alike.
Digital Oil and Gas Improves Industry across the Board
Novel approaches to data analytics offer insights on cost reduction during down markets and enable production adjustments based on real-time market demand. Improvements in data collection and analysis enable the oil and gas industry to reduce costs, minimize risks, discover new resources, improve operational efficiency, optimize investments, curtail environmental hazards and streamline maintenance.
Instead of pulling a damaged well offline for three days for repairs, wireless pump-mounted sensors and data science can predict the damage before it occurs or reduce downtime to mere hours, saving the company tens of thousands of dollars in revenue. Real-time data on pressure, vibration, resistance and temperature combined with instant analytics and global communication will allow oil companies to make highly-informed decisions in the blink of an eye.
A recent Microsoft survey found that 86% to 90% of oil and gas executives say leveraging more analytics capabilities, increasing mobile technologies, and leveraging more internet-of-things and automation would boost company value. Chevron estimates it could increase production rates by 8% and increase recovery rates by 6% in a fully optimized digital oilfield.
High Demand for Data Organization, Interpretation and Machine Learning
The increase in demand for petroleum data science continues to attract venture capital funding for data science startups. Petroleum data scientists are needed, first and foremost, to convert the mass quantities of geological, production, completion and maintenance data into a format usable for active decision making.
Today, oil and gas companies are struggling to sort and interpret existing drilling, seismic and economic data stored and organized in a tangle of tagged formats, Oracle databases, and XML-based standards that cannot intercommunicate. Petroleum data scientists develop tools that allow complete interactive interpretation of these data sets.
In addition, oil and gas companies currently need data scientists to apply machine learning technologies - training software to identify useful data versus junk data. Data scientists must also monitor the learned behaviors and continually update the machine learning process. San Mateo, California-based Tachyus developed its Data Physics machine learning software to help oil and gas companies decrease operation costs and increase production. The software is currently used in over 10,000 domestic wells. The Silicon Valley startup, Kaggle, works with machine learning using geological data from shale plays to predict the best locations to lease and drill and to examine the best drilling and completion parameters for different reservoir types.
With the surge of innovative technologies in equipment sensors, masses of new data will need to be processed and analyzed. GE is working with British Petroleum (BP) on efficient operation of rotating machinery by analyzing data on rotor position, vibration, pressure, temperature and other variables. GE’s data analysis allows detection of slight alterations in operation indicating suboptimal performance.
In April, 2016, Baker Hughes launched its FieldPulse model-based predictive analytics software that enables operators to proactively optimize production across entire fields by communicating asset performance in real-time. The software uses real-time data on a number of variables to alert well operators of problems in advance, identify highest production potential and manage assets accordingly.
Houston and Silicon Valley Training Future Petroleum Data Scientists
Universities and other institutions are currently working on applying data science to the oilfield. In Silicon Valley, the Energy Resources Engineering Department at Stanford University hosts a Smart Field Consortium for researchers studying reservoir model order reduction and well placement using models created from geological, reservoir and production data.
The University of Southern California offers a master’s degree in Smart Oilfield Technologies. USC’s Chevron-sponsored Center for Interactive Smart Oilfield Technologies, CiSoft, provides research and training facilities.
Houston, Texas is demonstrating the largest growth in petroleum data science training next to Silicon Valley. BP recently founded the Center for High-Performance Computing in Houston, Texas to interact with global industry leaders at facilities like the Center for Petroleum and Geosystems Engineering, the Institute for Petroleum Research, the Berg-Hughes Center for Petroleum and Sedimentary Systems and the Global Petroleum Research Institute. British Petroleum’s Well Advisor Program currently provides real-time data on wellsite operations for over 100 offshore wells.
In addition to the numerous online training programs and degrees offered by the University of Houston, Rice University, and Texas A&M, several Houston-based STEM education programs focusing on the petroleum industry are already in place for high school students and younger.
As Houston, Texas houses the world’s leading petroleum engineering and geology programs, working with Silicon Valley tech experts, the city is poised to provide a central hub in which venture capital funding, incubators and training programs collaborate to make great strides in petroleum data science.