In the twenty-first century, Artificial Intelligence (AI) techniques have experienced a boom following concurrent advances in computer power and capabilities, availability of large amounts of data and better theoretical understanding; and that is why AI techniques have become today an essential part of a variety of technological industries, helping to solve many challenging problems in computer science, software engineering and robotics. Compared with humans, existing AI lacks several features of human reasoning; most notably, people have powerful mechanisms for reasoning, such as space, time, and physical interactions (Wikipedia).
With growing challenges and low prices, faced by the oil and gas industry over the last 15 years, the Big Companies feel the necessity of review and improve all the critical processes. The use of artificial intelligence (AI) specifically in the oil and gas industry has gained major importance from at least 10 years. AI will for sure be used in the near future in every part of exploration and production procedures like geology, geophysics, and reservoir engineering, producing a major technological impact.
It needs to be said that a lot of the people who talk about AI enthusiastically, do not mean artificial intelligence literally, as an autonomous system capable of making decisions on its own. What they most commonly mean is predictive and analytic algorithms, and the process that allows for the deployment in a huge variety of tasks in the upstream oil industry: machine learning.
AI systems consists of various tools: machine learning, fuzzy logic, artificial neutral networks, mathematical optimization, statistics, probability and expert systems. These systems transform data into valuable insight that can be applied across various stages of the E&P life cycle, including seismic, geology, petrophysics, reservoir, drilling and production.
Ranganathan (2016) point out that Geoscientists working for the oil industry, have the knowledge and experience to locate reserves, but as resources become more and more scarce, AI systems may hold the key to boost the discovery of new exploration or appraisal areas. AI systems may play in the future, a crucial role in enabling Geoscientists to be more productive irrespective of their experience.
Fuzzy logic is an AI mechanism to drive decisions when data is incomplete or partially unreliable. If an algorithm is designed with a certain number of inputs, fuzzy logic can help overcome deficiencies when the data is spurious or inconsistent. Fuzzy logic can also help with well simulations and reservoir characterization where data needs to be extrapolated (Slowikowski, 2016) in order to produce a forecast.
Today the usage of machine learning in digital subsurface workflows is increasing progress towards truly integrated reservoir studies. Collaboration across all involved disciplines (geophysics geology, petrophysics, and reservoir engineering) will increase and improve as workflows become more and more automated, visualization options become more powerful and integrated; and the prediction and decision making process will become more robust, supported for the best available combination of human intelligence/experience; and AI at every step of each workflow (Breeland, 2018).
In the near future, AI will offer a robust data interpretation process that can assist Geoscientists, with critical knowledge transfer and decision making. It will enable higher productivity and create opportunities for a faster advancement. While AI has been in use in some of the upstream processes over the last decade, there are many people with the skeptical belief that there is no substitute for the human brain.
Velda (2016) quoted Naveen Rao, the CEO of Nervana Systems who thinks that in some way the economic pressure pushes companies to be more forward-looking in terms of new technologies like AI, and by the contrary, “when profits are very high, the tendency is to see less innovation because there is less competitive pressure to do so.” In today’s market added Rao, “companies are facing reality and seeking ways to make money in the field during difficult times, using techniques like this as part of that solution”.
In 2018, the oil company Total announced a partnership with Google Cloud to develop AI technologies that will be applied to exploring and assessing oil and gas fields. Their teams are currently working in a wide range of AI applications, including analysis of satellite images and rock sample images to enhance exploration efforts of the company (Twentyman, 2019).
Chowdhury (2019a, 2019b) cited in two articles of The Telegraph, digital version, that BP’s venture capital division has recently invested $5M in an AI company (Belmont Technology) in an aggressive move to booster its digital capabilities in many activities around oil and gas projects. This company apparently has developed a cloud-base, AI platform that BP specialists can feed with actual and historical data from different sources as geology, geophysics, reservoir information.
This AI platform, which has been nicknamed Sandy, added Chowdhury (2019b), is expected for example to map out its subsurface assets and make better-informed decisions based on data processed through machine learning. Also, Sandy may help to unlock critical data for the subsurface engineers at a much faster pace, and more importantly, make use of neural networks to perform simulations of BP’s different projects.
Maybe in a too futuristic scenery, BP may consider that instead of using a team of geologists, sequence stratigraphers, geophysicists, etc., to apply their particular knowledge in a new exploratory area, they are going to apply algorithms against all the data (AI algorithms are capable of learning from data), that BP has been gathering through the years in all continents and in all kind of sedimentary basins, to help them figure out where are the best places to explore, getting rid of human bias. The final goal for BP cited Chowdhury (2019b), maybe is to find new technological ways of accelerating its projects timeframes, ranging from exploration projects to reservoir modelling in its production assets.
According to Habibi (2018), AI in the oil and gas industry is expected to grow from 2017 to 2022 and reach a market size of USD 2.85 Billion by 2022. This expected growth is due to the adoption of big data technology, digitalizing the oil and gas industry through adopting a variety of analytic predictive algorithms and automation systems.
Patrick Pouyanne, CEO at Total wrote on march 4th 2019, in a LinkedIn article: “We shall let the machines do what they do better than us: manage and analyze data, propose options, and devise solutions. If done this way, women and men stay at the heart of the production process, ready to make the decisions that cannot be automated. AI may become history’s more powerful tool, a precious accessory to the human mind, never its rival.”
• Breeland, M. (2018) The melding of artificial and human intelligence in digital subsurface workflows: a historical perspective. First Break, Vol. 36, pp. 85-89.
• Chowdhury, H. (2019a) BP invests $5M in AI firm as it looks to technology to boost its exploration activities. The Telegraph, January 28th.
• Chowdhury, H. (2019b) How artificial intelligence is shaking up the oil and gas industry. The Telegraph, February 21st.
• Habibi, A. (2018) Artificial Intelligence (“AI”) and the Oil & Gas Industry. Data Driven Investor, December 5th.
• Ranganathan, A. (2016) Artificial Intelligence in Upstream Oil and Gas. Oil + Gas Monitor, June 24th.
• Slowikowski, M. (2016) How Artificial Intelligence Could Help Transform The Oil Industry. Oil Price, August 21st.
• Velda, A. (2016) Artificial Intelligence Takes Shape In Oil, Gas Sector. HartEnergy, April 4th.