June 10, 2025
How Equinor used AIM to save $12M in CAPEX
Discover how Equinor used WhiteSpace’s AIM solution to analyze millions of scenarios and save millions in CAPEX on the Johan Sverdrup Phase 3 project.
The challenge
In late 2022, Equinor’s development team was gearing up for the next phase of one of the most important oil fields on the Norwegian Continental Shelf: Johan Sverdrup Phase 3.
After two successful development phases, Phase 3 focuses on unlocking additional reserves in the southern portion of the field. But the task ahead was far from simple.
Multiple conflicting objectives had to be resolved:
How many wells would be required to reach key targets?
Where should templates be placed to balance drilling cost and infrastructure constraints?
Could longer, more complex wells reduce subsea infrastructure?
What was the best combination of layout, routing, and reservoir access?
This wasn’t just a complex puzzle — it was a planning problem with millions of possible permutations.
And timelines were tight. Equinor needed to move from screening to concept select rapidly while maintaining engineering integrity, cost control, and operational feasibility.
Using AI to explore all options
Enter AIM (Artificial Intelligence for Maturation) — a joint solution developed by WhiteSpace Solutions in collaboration with Equinor’s Subsurface Excellence and Digital unit.
Unlike traditional field development planning tools that require planners to manually iterate through a handful of predefined concepts, AIM was designed to do something radically different:
Explore every possible combination of well targets & trajectories, template configurations, and pipeline routing options.
Present a Pareto front of options, allowing teams to clearly see the trade-offs between cost, complexity, and risk.
And most critically, AIM embedded a human-in-the-loop design, giving engineers full control over design rules, constraints, and the final decision.
AIM in practice
In Phase 3, Equinor’s development team used AIM to integrate the full 3D subsurface and seabed of the field and apply constraints such as:
Avoiding shallow hazards and pockmarks
Adhering to anti-collision margins
Cost rules for well architecture, rig moves, SURF routing and template sizes
Engineering preferences from Phases 1 and 2
Once constraints were defined, AIM began generating well paths using evolutionary algorithms. These were evaluated not just on cost, but also on trajectory complexity, feasibility, and flow assurance implications. The next layer used mixed-integer linear programming (MILP) to identify optimal template locations and routing—essentially assembling a full-field layout from the bottom up.
Hundreds of thousands of scenarios were created, evaluated, filtered, and clustered. Then the most promising ones were presented for detailed engineering review.
Human-in-the-loop: AIM provides engineers with full control over design rules, constraints, and the final decision.
The Result: $12mln CAPEX saving
Among the most powerful outcomes was AIM’s validation of the project team’s conclusion that fewer, larger templates combined with longer, more complex wells could significantly reduce overall CAPEX.
AIM validated that larger size drilling templates reduced subsea cost
AIM confirmed that accepting long-step out well design allowed a CAPEX reduction by 5-18% relative to a layout minimizing drilling risk
A previously overlooked layout — found by AIM — was adopted as the final concept as it improved the production potential. This lead to direct CAPEX savings of $12mln through that improved concept.
Just as critically, AIM gave Equinor the ability to:
Rapidly and rigorously screen millions of options instead of manually testing a few dozen
Build a digital audit trail of trade-off decisions across Drilling & Wells, Subsurface, and SURF (Subsea Umbilicals, Risers, and Flowlines)
Validate or challenge assumptions with data-backed evidence
The result wasn’t just cost savings—it was confidence in the plan and alignment across disciplines in a shorter timeline.
What's next?
The success of AIM in Johan Sverdrup Phase 3 and other projects has sparked broader adoption across Equinor and interest from other operators. Future AIM deployments will include:
Integration with flow assurance and economic modeling tools
Expanded optimization capabilities for brownfield tiebacks
Further automation of detailed engineering handover
For WhiteSpace Solutions, this project was another validation of the power of AI-assisted planning — not as a black-box solution, but as a strategic co-pilot that empowers engineers to make better, faster, and more defensible decisions.
Better planning. Less guessing.
Johan Sverdrup Phase 3 proved what’s possible when field development planning is reimagined with AI. AIM didn’t just help Equinor make a better decision—it helped them explore every option worth considering, understand the trade-offs, and choose with confidence.
In today’s world of complex, capital-intensive offshore development, that’s a competitive edge that few can afford to ignore.
Want to learn more or read the detailed SPE Paper on our work for Equinor? Contact us below.