Flux Energy Solutions (Türkiye) used Seequent’s Volsung platform to cut geothermal simulation time by 80% and costs by 76%, providing an AI digital asset for renewable energy in Türkiye.
By Paul Gorman
Türkiye’s complex geology and fault systems have a positive side, boosting the nation’s efforts to adopt renewable energy and further harness its geothermal potential.
Geothermal energy generates about 3% of Türkiye’s electricity. The country’s use of geothermal power for heating puts it third in the world behind China and the United States, with significant possibilities to expand from the domestic to the larger-scale market.
This is a country which has been using hot underground water and steam since ancient times. Close to 2 gigawatts of geothermal power plants are now operational, with a potential total of 4.5GW in future.
Flux Energy was employed to maximise the sustainable geothermal reservoir performance of a 69.5-megawatt field in the volcanically active Salihli district of western Türkiye, using advanced numerical simulation and AI-driven proxy modelling.
3D temperature contour view of numerical view
The size, complexity, and fractured nature of the reservoir was going to be challenging for traditional modelling and data processing, and manual methods and conventional software were unable to be used. Flux needed a fast, scalable, and geothermal-focused digital platform to accurately simulate 3000 scenarios and integrate results with AI models.
Flux chose Volsung for its robust processing and automation, and fully integrated geothermal simulation, allowing for the management of thousands of simulation scenarios with high reliability and efficiency.
Conquering shortcomings
Traditional reservoir-modelling platforms have had significant limitations, according to Flux Energy Co-Founder and Reservoir Simulation Expert Ali Başer.
‘Conventional tools often lacked seamless integration between the reservoir, wellbore, and surface systems, making it difficult to simulate field-wide geothermal scenarios efficiently.
‘Managing multiple simulation cases was manual, error-prone, and time-consuming. Moreover, the absence of GPU acceleration capabilities in these tools significantly limited our ability to conduct high-resolution simulations or generate large datasets necessary for machine learning.
‘These shortcomings made it nearly impossible to work on complex subsurface systems involving natural fractures, especially when a robust and accurate history match was critical for sustainable geothermal reservoir management.’
Volsung’s intuitive user interface enables the fast and precise modelling of such fractured reservoirs, he says. In planning to use that, Flux carried out natural-state modelling and flow simulations, prepared, cleaned and analysed extensive datasets including production histories, and developed and trained AI-based proxy models to simulate millions of production and injection scenarios in a fraction of the time required by conventional methods.
Flux Energy Solutions’ conceptual model of the field
Careful sustainable management
At the core of Flux Energy’s efforts was the need to manage the reservoir sustainably.
‘Geothermal energy is only renewable if carefully managed,’ Başer says.
‘Improper production can lead to thermal drawdown and reservoir pressure loss. Our objective was to create a reservoir model capable of guiding long-term decisions to mitigate these issues and enhance field longevity.
‘This project is of strategic importance to Türkiye’s renewable energy goals. As the country accelerates geothermal development, robust reservoir management becomes critical for both energy security and environmental sustainability.
‘The innovative use of AI to extend and enhance conventional simulation workflows can transform how geothermal fields are managed, enabling near real-time scenario testing and operational optimisation.’
There were several significant challenges to using legacy approaches: the reservoir was highly heterogeneous and naturally fractured, making it difficult to achieve reliable history matches; the size and complexity of the dataset was problematic; and it was tricky to integrate AI seamlessly with numerical simulation while maintaining accuracy.
Volsung, with GPU acceleration and integrated workflow tools, enabled significant performance gains, Başer says.
‘By using AI, we reduced scenario evaluation time for a massive 10 million scenarios from an estimated 380 years to under 10 days, unlocking new possibilities for real-time geothermal reservoir management,’ he says.
A natural state model
Energy security
Improvements through using Volsung can help extend the field’s operational life, minimise thermal drawdown, and reduce the need for additional drilling.
‘Volsung empowered us to efficiently run 3,000 simulations, reducing modelling time by over 90%. This breakthrough allowed us to compress a project that would typically take 60 months using legacy methods, which would not have been feasible with the required quality, down to just 12 months,’ Başer says.
When it comes to social obligations, the project promotes renewable and locally sourced energy, and supports energy independence.
‘The efficiency gains also allowed more time to engage stakeholders, ensuring the model met local operational and environmental needs.
‘There were no significant environmental or social risks. By minimizing trial-and-error development and avoiding unnecessary well interventions, the model contributed to responsible field operations with fewer disruptions and a smaller physical footprint on the land.’
Key metrics
- Reducing 60-month project into 12 months without sacrificing quality
- Cutting project costs from about $US1 million with legacy tools to $240,000
- Slashing evaluation for 10 million scenarios from 380 years to less than 10 days
Flux Energy Solutions’ groundbreaking application of AI earned recognition as one of three finalists in the 2025 Going Digital Awards in Infrastructure – Subsurface Modelling and Analysis category.