With increased oil price volatility and depleting crude oil reserves, oil and gas companies are progressively turning towards the implementation of AI to improve efficiency and revenues. Energy companies may not be generally regarded as an AI pioneer, however, they have been consistently investing in digital technology, with global investment growing at 20% annually since 2014. Digital services solely for the oil and gas industry are predicted to be over $30 billion by 2025, totaling nearly $150 billion in annual savings.

Energy Optimization and Downtime

Research shows that production lags in barrels of oil per day caused by accidental damage have significantly reduced since 2017. However, unplanned production loss due to downtime has been on the rise. Annual losses from unplanned downtime are pegged at USD 38 million to upwards of USD 88 million. Studies have also shown that an average offshore installation, experiences about 27 days of unplanned downtime a year.

This is where Artificial Intelligence along with Data Science and IoT based predictive analysis help in yielding substantial savings. AI-enabled predictions augment traditional predictive methodology and flag costly downtime early on. With better monitoring and data science-led predictive analysis, early detection of potential equipment failure directly benefits the company’s bottom line. Reducing downtime and maximizing production capacity is at the core of operations and asset capability, and, AI-augmented operations help reduce the business impact of unplanned production loss.

How Artificial Intelligence and IoT helps

AI isn’t a lone technology but comprises of a set of systems that help understand and improve operations and predictive capabilities. It includes natural language processing (NLP), machine learning and deep learning, neural networks, virtual agents, and autonomics. Artificial Intelligence links chunks of information together, recognizing and establishing connections and work processes. Experts and operators use this data to conduct simulations and clarify data and results.

Whilst AI has a rather extensive scientific implication, within the oil and gas industry, it is usually employed in machine learning and data science. Machine learning (ML) uses external sensors, reading, and identifying patterns that indicate a failure or outage. ML helps build an intelligent workflow, facilitating automation, and improving asset capability. Offshore installations of oil and gas companies can oversee complex operations and act quickly to resolve issues that may have otherwise been undetected by human operators.

AI platforms can be employed to analyze subsurface geophysical data, accurately mapping underground oil deposits. This helps understand the value of the reservoir, and also makes drilling methods more efficient. Machine learning is also used to run simulations such as to assess the impact of a new site and appraise the environmental impact before commencing a project. AI also plays a significant role in making offshore installations safer.

Data Science wields AI to glean information and data and connects related pieces of data together to produce meaningful and critical insights from existing information. This analysis is used in the industry to make oil and gas exploration and production more efficient and helps utilize existing infrastructure more effectively.

IoT leverages data sourced from intelligent devices such as sensors, including its temperature, vibration, flow rate sensors, and other data points. It identifies anomalies in equipment behavior, predicting a failure of equipment within a timeframe. Along with AI-enabled systems, IoT provides a more holistic analysis and better decision-making capacity for operators and experts.

AI, IoT, and Asset Reliability

Nearly 42% of offshore equipment is estimated to be more than 15 years old with an average of 13% downtime, much of which is unplanned. AI and IoT enabled systems can be applied to monitor exploration, development, and drilling. Components and potential failure points are connected to various sensors monitoring pressure, torque, vibration, temperature, and flow rate. Historical and contextual data is augmented with real-time data points from sensors, running it through Machine Learning algorithms creating predictive models. These algorithms identify and flag early warnings for equipment failures reducing downtime and enhancing predictive maintenance.
According to the US Dept. of Energy, AI and IoT driven systems can drive a 45% reduction in equipment downtime and a 30% reduction in maintenance costs. Oil & gas companies stand to benefit substantially with investments in AI including

  • Enhancing asset reliability and reducing costs
  •  Improving operational efficiency
  •  Improving asset utilization
  •  Reducing environmental impact

Challenges Ahead

Oil & Gas companies have aggressively invested and adopted AI-driven preventive maintenance systems. However, the energy sector faces inherent challenges and unprecedented situations today. For successful implementation of AI and IoT systems, companies need to integrate their legacy equipment to newer protocols. 80% of legacy equipment is connected to local networks, and are unable to function across TCP/IP networks. Uninterrupted data transmission and sufficiency in data collection also poses a challenge. With assets in remote locations, poor communications coverage could lead to disruptions in data collection, resulting in the failure in predictive warnings and ultimately leading to equipment downtime. Further, it may take up to a year to collect enough data for a reliable prediction. It is ideal to have real-time data from sensors throughout the asset lifecycle. However, this lag in data collection could impede actual results on the field.