Artificial Intelligence, Machine Learning Applied to Oil Exploration | Sci-Tech

A rig at Vietnam Oil and Gas Group’s Bach Ho oilfield. (Photo: VNA)

Hanoi (VNA) – The
Vietnam Petroleum Institute (VPI), member of the Vietnamese oil and gas group
(PetroVietnam), applied artificial intelligence (AI) and machine learning to quickly identify the presence of fractured granite subsoil, with an accuracy rate of over 80%, saving exploration time and costs.

The discovery and successful exploitation of fractured granite basement reservoirs at Bach Ho (White Tiger) and other oilfields in Vietnam has changed the traditional view of oil and gas exploration and contributed to the formation of a new in the region and in the world.

However, identifying fractured granite subsoil using traditional methods requires specialized equipment which may require suspension of oil well drilling, lengthening the lease term of the rig (current lease is around 65,000 to 68,000 USD per day) and increase expenses.

To optimize this process, VPI has developed a fracture prediction solution that applies artificial intelligence and machine learning, helping to accurately detect the presence of fracture systems based on real-time data.

The relevant data will be used as input to the machine learning algorithms and then the models will be ranked, ranked and evaluated to find the best for fracture prediction.

VPI tested the accuracy of this prediction model on drill data from 12 wells in selected fields with similar geological structures, showing a prediction accuracy rate of over 80%.

The results are given on the Machine Learning Operations Platform (MLOps), which is expected to help improve drilling efficiency, ensure process safety and shorten the rental time of the rig. Depending on each contractor’s drilling plan, the cost savings can reach hundreds of thousands of US dollars.

Le Ngoc Anh, Chief Data Officer of VPI, said the fracture prediction model uses existing data from the oil and gas industry to develop optimization algorithms to create new values ​​and ensure safety and security.

Applying machine learning and artificial intelligence to oil and gas exploration is the next step in implementing PetroVietnam’s digital transformation strategy to promote business model change, optimize operations and improve governance capacity.

In early July, VPI successfully built Oilgas AI, an ecosystem using artificial intelligence and machine learning to collect, demonstrate and analyze specialized data on commodities such as crude oil, gasoline, liquefied petroleum gas and natural gas to offer solutions that help businesses make smarter decisions faster and more efficient in their day-to-day operations, as well as building long-term solutions. long-term plans and strategies.

PetroVietnam develops and perfects appropriate and effective solutions, while coordinating with socio-political organizations to promote digital transformation in its subsidiaries.

In a resolution on digital transformation by 2025, with a vision for 2030, issued in February this year, PetroVietnam saw digital transformation as a necessity to increase productivity, quality, efficiency and competitiveness.

Digital transformation is seen as a strategic task of
PetroVietnamand is of particular importance for the group’s businesses and units.

The work is expected to help the company promote innovation and improve its business performance and competitiveness, as well as its leading position in the oil and gas industry, thereby contributing to the country’s digital economy.

In the first half of 2022, PetroVietnam pumped 5.48 million tons of crude oil, exceeding the first half target by 23%.

Meanwhile, gasoline production exceeded plan by 14% and fertilizer production exceeded target by 8%. The production and supply of gas, electricity and other energy products have also reached a high level.

First-half revenue was estimated at VND468.3 trillion (USD20 billion), up 55% year-on-year and double the six-month target. The group contributed VND66.1 trillion to the state budget, 2.2 times more than the half-year plan and up 41% from the same period last year, according to statistics./.

Sherry J. Basler