Artificial Intelligence in Oil & Gas – Upstream

Artificial intelligence (AI) is a diverse scientific field, but within the Oil and Gas industry there are two primary applications of the technology: machine learning and big data analytics. Let’s start off with defining what these terms are for the sake of clarity of curious minds outside the Data Science community.

Artificial intelligence is a branch of computer science that uses any device that aims to enable the software to analyze its environment using either predetermined rules and search algorithms, or pattern recognizing machine learning models, and then make decisions based on those analyses. It is often used to describe machines (or computers) that mimic “cognitive” functions that humans associate with the human mind, such as learning and problem-solving. Siri on the iPhone is a good example of voice recognition AI software at work.

Photo Cred: Theguardian.ng

Machine learning involves feeding historical data into mathematical algorithms for iterative trend analysis which models that data with a line of best fit. The result of this process is the algorithm determining what mathematical function that is best defines that large historical data-set. Machine learning can also be used to run simulations, using predictive data models to discover patterns based on a variety of inputs.

 

How do we know we’ve got the equation that best replicates the data?

A portion of the historical data is set aside as a test set. Once the algorithm has found the equation, it can be tested using known and unknowns from the variables from the test set. The test of the model is satisfactory when the deviation between the same data points from the historical & modelled data is as close to zero as possible. Artificial intelligence is globally perceived as the fundamental driver of future development and productivity, innovation, competitiveness and job creation for the 21st century.

Big Data analytics uses these above techniques for problem-solving and optimization to analyze, systematically extract information from, or otherwise deal with data sets that are too large or complex to be dealt with by traditional data-processing application. AI software uses algorithms to process the volume, variety, velocity, veracity, value, and complexity of the large data-sets.

 

AI in UPSTREAM OIL & GAS

The industry is trying to harness a greater percentage of its own data to generate unforeseen cost savings via more efficient operations and greater precision in E&P operations for far less money. Big Data Analytics provides real-time insights for decision support to gather and transmit information more effectively. With the increased use of IOT compliant devices, meters and sensors, real-time data capture is possible from the wells to the terminal.

Combined with the decades of historical exploration, drilling, production, commercial, financial and seismic data, the industry is sitting on a massive data bank. Surveys revealed that companies with better Big Data and Analytics capabilities were twice as likely to be on top of financial performance in the field, five times more likely to make decisions quicker than their peers and three times likely to execute decisions as planned.

Photo Cred: IBM Big Data & Analytics Hub

Like their counterparts in other industries, energy companies have complex, legacy Information Technology systems which have evolved over decades and now contain various islands of disparate data sets.

Oil and gas organizations have dealt with huge amounts of data for decades in their quest to learn what lies below the surface. Big Data makes it possible to gather and transmit information more effectively.

 

Global Adoption & Impact Of AI

Photo Cred: Forbes.com

In-depth industry reports from McKinsey, IBM and GE on the global impact of AI on various industries revealed insights pointing to a bright future for AI. Let’s get into some of the most interesting observations:

  1. AI has the potential to create between $3-$6 trillion in value annually across nine business functions in 19 industries. It is expected to gradually add around $13 trillion by 2030 to current worldwide economic output.
  2. AI will generate up to $2.6 trillion in additional value in Marketing and Sales and as much as $2 trillion in Supply Chain Management and Manufacturing. This number could go up in the post-Covid economy in the Healthcare industry especially.
  3. AI will add $200 billion in value to Pricing & Promotion and $100 billion to Customer Service Management in Retail.
  4. McKinsey predicts AI will have up to an 11.6% impact on Travel industry revenues and up to 10.2% on High Tech.
  5. A European trucking company reduced fuel costs by 15% using AI to optimize the routing of delivery traffic, improving fuel efficiency and reducing delivery times.
  6. In 69% of the use cases in the study, AI and deep neural networks improved performance beyond what existing analytic techniques were able to deliver.
  7. IBM’s Center for Applied Insights reported that companies with successful analytics programs are proven to outperform their competitors. Every $1 they spent on AI programs yielded $10.66 in operations costs & unrealized revenue. These users were also 5 times as likely to make key decisions faster than their peers.
  8. The Big Data revolution is expected to transform industries globally on a scale similar to that of electricity revolutionized in the early 20th Century.
  9. Global spend on Big Data hardware, software and services grew to $114 Billion in 2018.
  10. 73% of US companies spend a fifth of their overall technology budget on Big Data analytics.
  11. Globally adopted applications of AI technologies: self-driven cars, computer vision; natural language; virtual assistants, robotic process automation, and advanced machine learning.
  12. Global oil industry has hitherto only harnessed 1% of its data using AI. Transformation to digital oil fields could add almost $1 trillion to the world’s oil economy by 2025 (Oxford Economics and Cisco Consulting Services, 2015)

 

Artificial Intelligence Applications in Upstream Oil & Gas

Critical decisions can now be made much faster and more accurately. Real-time, interactive dashboards and reports allow oil company personnel to quickly get different views and drill into information about a specific well or asset. Thus, when key variable changes, they know immediately and can take the preemptive, prescriptive action.

IBM has developed Big Data & Analytics capabilities enable drilling and production assets to be assessed individually but rolled up into an enterprise-wide operations view by collecting, managing and analyzing large volumes of data.

These capabilities represent the most cost-effective, efficient means of ingesting, analyzing and visualizing all the data needed to accurately assess what’s happening in the oil field in real-time.

Combined with the decades of historical exploration, drilling, production, commercial, financial and seismic data, the industry is sitting on a massive data bank. Producers deploying Big Data Analytics capabilities were twice as likely to see top of financial performance in each asset (reportedly seen improvements of ranging from 15% – 40%) in the following areas:

Seismic Data Management

The upstream analytics begins with the acquisition of seismic data (collected with sensors) across a potential area of interest in search of petroleum sources. Once the data is gathered, it is processed and analyzed to determine a location for drilling. Seismic data can further be combined with other data sets (a company’s historical data on former drilling operations, research data, etc.) to determine the amount of oil and gas in oil reservoirs.

Optimization Of Drilling Processes

One way to optimize drilling processes is to customize predictive models that forecast potential equipment failures. As a starting point, the equipment is fitted out with sensors to collect data during drilling operations. This data, together with the equipment metadata (model, operational settings, etc.) is run through machine learning algorithms to identify usage patterns that are likely to end in breakdowns.

Improve Reservoir Engineering

The variety of downhole sensors (temperature sensors, acoustic sensors, pressure sensors, etc.) can gather data needed for companies to improve reservoir production. For example, with big data analytics, companies can develop reservoir management applications to get timely and actionable information about changes in reservoir pressure, temperature, flow and acoustics to increase insight and control over their operations to drive reservoir performance and profitability.

To Survey And Monitor Oil Exploration Areas

The company employs a seismic analysis to survey the area and indicate whether the given area contains oil and gas deposits. The more sophisticated big data analysis allows understanding the nuances of a particular drilling site before deciding to drill.

To Forecast Production

Shell installs optical fibre cables with sensors within the wells to measure seismic data. This data is further analyzed using artificial intelligence technologies to create 3D and 4D maps of the oil reservoirs to find out how much oil and gas is still left in the reservoir.

To Extend Equipment Lifespan

Generating tons of sensor data, Shell runs advanced analysis on drilling sites machinery to improve its performance and proactively understand what equipment requires maintenance. This stimulates a longer drilling time with fewer maintenance stops and. Solely in Nigeria, Shell has managed to save over $1 million by leveraging sensor analytics.

To Increase Logistics Efficiency

Shell utilizes complex algorithms to analyze transportation and production costs, economic factors that drive demand as well as weather patterns to determine how and where to move refined products and how to set the prices.

To Cut Net Carbon Footprint

According to Shell’s latest sustainability report, the company “supports the vision of a transition towards a net-zero emissions energy system”. One way the company plans to reduce emissions is to use carbon capture and storage technology empowered by big data software. This is of supreme interest to governments and industry regulators in today’s industry and could be the key to making zero flare more profitable.

 

Big Tech Providing AI Applications In Oil & Gas

Tech leaders Amazon and Google are providing services with the global oil giants to catalyze the rise of the digital oilfield. France’s Total and Google have signed an agreement to use artificial intelligence (AI), enabling the oil major to assess oil and gas fields faster. The agreement with Google will allow Total’s engineers to speed up the assessment of oil fields. Under the agreement, the companies will develop AI programs to interpret subsurface images from seismic studies and automate the analysis of technical documents.

Google’s Cloud Oil, Gas & Energy team is led by Obama’s former Chief Sustainability Officer and a 25-year BP expert. Their objective is to bring sustainability and AI solutions for data-driven drilling and production, automation of new reserve discovery and streamline contractor cost optimization.

Amazon Web Services (AWS) provides energy companies with the foundation to transform complex business and operational systems and accelerate the transition to a more sustainable energy future. AWS delivers the broadest and deepest cloud platform and industry solutions for energy companies to revamp legacy operations. This makes you less carbon-intensive and accelerates your development of innovative renewable energy businesses and business models.

These companies are industry leaders in AI deployment having developed capacity in the last 10 years. Valuable insights can be drawn from AI’s successes in the transportation, telecommunications, biomedical and manufacturing industries.

We are sure in for some interesting times in next decade of the energy industry.

0 Points


One thought on “Artificial Intelligence in Oil & Gas – Upstream”

  1. https://waterfallmagazine.com
    I truly love your site.. Great colors & theme. Did you develop this website yourself?
    Please reply back as I’m wanting to create my
    own website and want to learn where you got this from or exactly
    what the theme is named. Appreciate it!