Senior researcher at NORCE, Rodica G. Mihai, leads the drilling automation and autonomy work package at DigiWells, and has been working on artificial intelligence (AI) in drilling for years.

  • We have shown that artificial intelligence techniques can be used to achieve higher levels of automation in drilling and consequently reduce costs and make drilling for oil and gas safer and more efficient. The AI techniques we have used in drilling automation and particularly, autonomous drilling can have big potential within other fields as well, says Mihai.

Several factors influence how relevant a field is for technology transfer from drilling automation. According to the EU classification of AI systems[1], autonomous drilling comes under high-risk AI (critical infrastructure).

  • When addressing autonomous decision making in drilling, one needs to account for the complexity and high-risk nature of the environment. The drilling process deals with uncertainties and observability challenges. It is a fragmented process where many different suppliers are involved and need to collaborate to ensure a safe drilling process. This raises several challenges when it comes to stability of such a complex distributed system, tells Mihai.

Other subsurface and drilling fields

An obvious usage of the AI techniques from autonomous drilling in oil and gas, is within other drilling operations, such as geothermal drilling, drilling for CO2 injection wells and drilling for hydrogen storage or production. Mihai particularly points to geothermal drilling.

  • Our previous research on autonomous drilling is a good basis for further research that can make Norway a winner within deep geothermal drilling. The transfer of knowledge from autonomous drilling, already demonstrated for oil and gas and currently in the industrialization phase, can lead to a substantial cut in costs and increased efficiency in geothermal drilling, particularly within deep drilling for district heating. By building on existing research on autonomous drilling with further research enhancements to include specific challenges of the geothermal case, it may be feasible to scale up the geothermal operations through boosting the number of wells that the same team can manage, explains Mihai.

Geothermal district heating is when geothermal energy is used for space and water heating, for both private households and for industry, through a distributed network.

Environmental monitoring and energy systems

Also in different sectors, there are great possibilities for the usage of the AI techniques developed in DigiWells. Two relevant examples are environmental monitoring and monitoring of energy systems. In a newly awarded EU project, currently in the GAP Phase, under HORIZON-MISS-2023-OCEAN-01-03 call, Atlantic and Arctic sea basin lighthouse.

Mihai will work together with colleagues from NORCE Environment and a large European team on addressing climate change & human threats to marine biodiversity. NORCE will contribute with methods and knowledge from drilling such as digital twin, data modeling and other AI techniques. Going forward, energy systems will be even more variable, characterized by production at different sites, at different times and involving multiple providers. Hence, managing the overall energy system will increase in complexity. The AI techniques from autonomous drilling can come to good use when monitoring and steering the energy system at large under variable and uncertain conditions, says Mihai

Subcommittee on verification and validation of DSATS

In drilling automation, several automation solutions from different providers need to seamlessly work together to ensure safe drilling operations. This raises several challenges when it comes to the stability of such a multi-agent system. SPE- Drilling Systems Automation Technical Section (DSATS) recognizes these challenges and started a subcommittee to advance collaboration on these aspects.

  • The Verification and Validation subcommittee of DSATS is focusing on how to address the verification and validation of the interplay in a multi-agent architecture of drilling automation systems, with multiple providers, says Mihai who leads the subcommittee.

Rodica spoke on the use of AI for autonomous decision making at a live SPE event this February. Almost 600 participants followed the event live to talk about AI in Drilling.