Artificial intelligence (AI), automation and machine learning all promise to redefine business operations in complex sectors – and it doesn’t get more complex than upstream operations for oil and gas.
According to a recent Ernst & Young survey, 92% of oil and gas companies worldwide are investing in AI or planning to do so in the next 5 years. Importantly, 50% of oil and gas executives say they have already begun using AI to help solve challenges at their organizations. Even The World Economic Forum suggests that, by 2025, large-scale adoption of AI in the oil and gas industry could account for 10–20% in cost savings.
In this post, I explore the significant role that Artificial Intelligence can play in optimizing critical business processes and unifying strategies around inventory, procurement, and maintenance by removing data silos.
Master data catalogues and governance: rightsizing critical inventory
Despite the trillions of dollars tied up in drilling, exploration and production, organizations are still planning poorly. The root cause of this? Poor data governance of master data.
Master data catalogs describe wells, production facilities, the stores and parts inventories, and the assets and data that identify them. Typically, a major producer’s global catalog might contain 2-3 million items with an inventory value worth of 3-4 billion USD. Yet, while master data management is critical for maintaining system integrity and enabling efficient business processes and analytics, its importance is often overlooked by the sector. This problem is compounded by siloed information dispersed across multiple business units. Each entity or region has adopted its own processes and nomenclature, potentially resulting in inconsistent formats, data duplication, missing information and more.
In tandem, well maintenance is typically managed regionally, with spares procurement on a per-incident basis. There is no planning across the fleet for critical parts held in inventory. For instance, a suitable specification bearing may well be stocked locally, but another part is flown in because cataloging naming/format inconsistencies removes the local option.
Here, IFS’s AI-enabled data analytics capabilities can rapidly help standardize component naming conventions, providing a global, holistic view of the inventory data. For example, IFS.ai capabilities autonomously pull data, enabling users to leverage insights at the right time for continuous improvement. While ensuring the health of assets, it can also correlate historic operating temperature, pressure and maintenance data with production outages, revealing the most uptime-critical assets and helping to plan appropriate condition-based maintenance.
Greater visibility leverages buying power
By using AI-analysis to reveal a global view of critical parts requirements, purchasing power is increased. Procurement decisions can be standardized and consolidated. Producers can secure enough components for all sites in one order transaction, maximizing a volume discount, before shipping them out to regional hubs. A 10 or 15 percent saving on a global inventory spend of $5-6m USD is certainly worth chasing. This approach also removes the internal cost of raising multiple regional purchase orders, which could be as much as $150-175 USD per order. With a clear understanding of needs, annual global purchases, on pre-agreed terms with select suppliers, can even become automated.
The Aberdeen Group estimates 50% of annual unscheduled asset downtime can be attributed to the lack of spare parts and stock outs. AI has a role to play in optimizing and rightsizing the type and level of inventory held. Maintenance organizations typically set a stocking strategy for critical spare parts at the conception of the asset. But this min/max safety buffer is an initial, often arbitrary projection – not yet based on real-world operational insight. This could result in safety stock levels 2-3x higher than needed, for the lifetime of the asset, consuming working capital that, at the end of the asset’s life, will not be recouped.
By applying AI, however, service and maintenance companies can analyze the real-world usage of components in similar operating scenarios. The result is a far more informed, and realistic stocking level recommendation, significantly reducing inventory stockholding and costs. Case study data from Deloitte suggests inventory optimization can reduce inventory carrying costs by 80%, raise material availability from 93 to 97+%, and generate savings of 20%.
AI in exploration: optimizing drilling campaigns
Any offshore drilling campaign project requires an extensive supply of specialist piping (oil field tubulars) alongside warehoused drilling spares. These may represent stock worth trillions of dollars. However, because there is typically no holistic visibility or analysis of the existing drilling spares stockholding, the potential saving available through re-deployment is not factored into the start of another new project.
AI offers a way to analyze drilling spares and warehousing inventory and match it to new project requirements. For example, AI and machine learning could undertake a cost analysis to establish if transferring existing inventory or buying new would be more cost-effective. Once again, AI promises better planning for supply chains to meet production targets efficiently.
It is also conceivable that we could use AI to orchestrate the purchasing and inventory management for multiple drilling campaigns as a capital project – for example, executing purchasing for an entire year’s activity in advance with a single, unified AI-driven system that seamlessly unites procurement project management with inventory management and warehousing.
ESG targets: minimizing the carbon footprint
Applying AI to help optimize production operations, both in terms of maintenance and oil and gas extraction, will become increasingly important in minimizing emissions caused by well downtime or overproduction.
Fields typically yield both natural gas and oil, and controlling the balance during extraction is critical. AI can assist by forecasting and planning these highly complex extraction strategy decisions.
In the case of land-based wells, water and chemical tanks need to be regularly emptied to prevent the need for flaring if gas is overproduced. By using IFS Cloud’s AI-enabled Planning, Scheduling and Optimization, and GPS data, water truck collections can be prioritized and routed to the most productive wells first, based on real-time needs.
Conclusion
AI has a transformative role in enabling exploration and production companies and drilling contractors alike to achieve a holistic view of their operations. By removing silos, AI promises to optimize buying power and stocking strategies while driving maximum uptime and productivity.
Interested in learning more? Find out how IFS Cloud and its AI-powered capabilities are digitally transforming the Oil & Gas industry here.
