Question: What is the definition of Petroleum Data Analytics?
Answer: Petroleum Data Analytics refers to the collection of Artificial Intelligence and Machine Learning tools, techniques, and methodologies that use data (actual field measurements) as the starting point, building blocks, and foundation of analysis, workflows, modeling, and decision making.
Question: What are the main components and technologies that form Petroleum Data Analytics?
Answer: The main technologies that are integrated to form Petroleum Data Analytics are Artificial Intelligence and Machine Learning algorithms. Some f these algorithms include (but are not limited to) traditional statistics, artificial neural networks, fuzzy set theory, and genetic algorithms.
Question: What is common among the technologies that form Petroleum Data Analytics?
Answer: The most common theme among the technologies that form Petroleum Data Analytics is their reliance on data (facts — actual field measurements) as the main source of information. These technologies start with observation of the behavior of a phenomena (through data) and then they try to make sense of these behavior (for example discover patterns in the data) and sometimes (when appropriate) to mimic their behavior in form of models.
Question: What is the objective of Petroleum Data Analytics?
Answer: The objective of Petroleum Data Analytics is to use data in order to perform: Descriptive Analytics (detail understanding of the data), Predictive Analytics (developing [training, calibration and validation] predictive models), and Prescriptive Analytics (forecasting and optimization of the processes in the oil and gas industry).
Question: Is Petroleum Data Analytics a new discipline in the oil and gas industry?
Answer: No. Petroleum Data Analytics does not introduce a new discipline in the oil and gas industry above and beyond the current disciplines such as Drilling, Geo-sciences, Reservoir, Completion, and Production engineering, Artificial Lift, and Surface Facilities. Petroleum Data Analytics is an enabler. It is a new technology used by the petroleum professional. Petroleum Data Analytics provide new approaches in solving the technical and engineering related problems that petroleum professionals deal with on a daily basis.
Question: Is Petroleum Data Analytics an IT related discipline?
Answer: Petroleum Data Analytics is NOT an IT related discipline. Just like other disciplines in the oil and gas industry, Petroleum Data Analytics is a user of the IT services. However, due to its nature, Petroleum Data Analytics is one of the major users and clients of the IT services. Being a Data-centric technology, Petroleum Data Analytics benefits from a well-designed, managed, and executed IT strategy that manages, standardizes, and delivers data to the entire organization.
Question: What types of expertise are necessary for Petroleum Data Analytics?
Answer: Those involved with Petroleum Data Analytics are first and foremost drilling, reservoir, completion, and production engineers as well as Geo-scientists. The most important technical proficiency required for a Petroleum Data Analytics expert in an organization is domain expertise in Petroleum Engineering. Furthermore, a Petroleum Data Analytics expert needs to understand, be verse, and eventually become an expert practitioner of Artificial Intelligence and Machine Learning. Ability to effectively use these technologies to solve everyday drilling, reservoir, completion and production engineering related problems defines a successful Petroleum Data Analytics Engineer (PDAE).
Question: In what ways Petroleum Data Analytics is different from other techniques currently used in the oil and gas industry?
Answer: Petroleum Engineering is a physics-based (and geology-based) area of science and engineering. As such, it has a long tradition and a track record of dealing with challenging problems. Until now, all of our “physics-based” models are created through mathematical equations to model the physical phenomena. Petroleum Data Analytics creates “physics-based” models and solutions not based on our today’s understanding of physics, but based on “facts” and “actual field measurements” (data) that represent the physical phenomena being modeled. In Petroleum Data Analytics, data is the starting point of the analysis, workflows, models, and solutions.
Question: If our current physics-base solutions that are based mathematical equations are working so well, then why would the industry need Petroleum Data Analytics?
Answer: We have been successfully producing oil and gas for the past several decades. However, many factors have made Petroleum Data Analytics a necessary technology in our industry as we embark on new challenges in the 21st century. One of the major characteristics of Petroleum Data Analytics is the avoidance of “Assumptions”, “Simplifications”, “Preconceived Notions”, and “Biases”. Complexity of many physical phenomena in petroleum engineering that are formulated through mathematical equations include major assumptions and simplifications. When it comes to unconventional resources (shale) “Preconceived Notions”, and “Biases” completely control the physics-base models that are used through mathematical equations. Being a fact-based technology, Petroleum Data Analytics becomes an undeniable necessity for organizations that want to stay at the top of their performance.
Question: What is the best way to include and implement Petroleum Data Analytics in an organization (NOCs, IOCs, Independents, Service Companies …)?
Answer: Implementation and inclusion of Petroleum Data Analytics in organizations require the support of the top management. However, it is important that the top management does not get exposed to fictitious and incorrect advertisements on the engineering application of Artificial Intelligence and Machine Learning, which currently, is such a big characteristics of this technology in our industry. Petroleum Data Analytics should be integrated in the organization in an organic manner. The top management must support the professionals in their organization to get exposed and trained on the engineering application of Artificial Intelligence and Machine Learning.
Although Artificial Intelligence ans Machine Learning is currently taught in many universities, none of the current petroleum engineering departments include any courses or training of such technologies in the context of oil and gas related problems (hopefully this will soon change as Petroleum Data Analytics becomes a potential purely online Masters Degree at West Virginia University). Therefore, lack of knowledge on these subjects and some skepticism on their applicability can be viewed as natural. It should be remembered that Petroleum Data Analytics is a revolutionary (non-traditional) way of approaching physical problems and as such may not be easy for some to embrace in a rapid pace. Change is always tough. Inclusion of Petroleum Data Analytics in an organization should include the following six steps:
1. Generate success stories
Perform projects based on data-driven analytics, generate a portfolio of successful projects and document all the details. Use these success stories in order to generate the required momentum in the organization. Make sure they are well communicated within the organization.
2. Train professionals
Enlist the help of Petroleum Data Analytics experts in the industry and train professionals in your organization on the fundamentals of the technology so that it is not perceived as a non-technical magic wand.
3. Create a need for infrastructure
Use the success stories and the training in order to support the argument that there is a need for solid infrastructure for data driven analytics. This infrastructure can have many forms including hiring of new blood, purchasing required software tools, etc.
4. Implement across the organization
Implement Petroleum Data Analytics across multiple disciplines in the organization generating a competition for more success stories and communicate them within the organization.
5. Provide IT support
By now there should be plenty of the momentum for a solid IT infrastructure that can support all data related activities in the organization. Develop the required IT for the entire organization.
6. Make a culture out of it.
Make Petroleum Data Analytics an everyday technology, for drilling, formation evaluation, well testing, reservoir engineering, reservoir modeling, reservoir management, completion optimization, production optimization, artificial lift analysis, modeling and optimization, surface facility, financial analysis, Health, Environment and Safety, etc.
Question: What are the main concerns of Petroleum Data Analytics?
Answer: Petroleum Data Analytics experts may also be called Petroleum Data Analytics Engineers (PDAE). In order for a petroleum professional to fit this title, she/he must first and foremost be a petroleum engineer with specific expertise in one of the traditional disciplines in the oil and gas industry. In other words, she/he must be a Geo-scientist, or a Drilling, Reservoir, Completion, or Production engineer. Furthermore, Petroleum Data Analytics Engineer must have expertise and skills that distinguishes her/him from the traditional petroleum engineer and Geo-scientist. These expertise include a working knowledge of statistics, data mining, machine learning and artificial intelligence.
Petroleum Data Analytics Engineer solves everyday petroleum engineering related problems that serves the organization in producing more hydrocarbon and increasing recovery. They do this by using data from the drilling operation to reduce NPT, increase ROP, and increase safety of the drilling operation. Reservoir Data Analytics Engineer develop data-driven technologies to better characterize complex reservoirs, build data-driven reservoir models, smart proxies of numerical models, data-driven reservoir management tools, and Production Data Analytics Engineer use data-driven technology to analyze, model and optimize completions (hydraulic fracturing) and increase artificial lift efficiencies (these are only a small set of technologies developed and or used by Petroleum Data Analytics Engineer).
Petroleum Data Analytics Engineer address the main technical concerns encountered in the exploration and production industry by combining their petroleum engineering know-how with state of the art data-driven technologies in order to increase efficiency and build completely new solutions using data as the main building blocks of their solutions.