Impact/context

Deep geothermal energy and geological CO₂ sequestration (carbon capture and storage, CCS) are key technologies for achieving global decarbonisation, but both depend on accurately simulating complex fluid and heat transport processes within subsurface porous geological formations, processes which are currently slow and computationally expensive when including models for uncertainty quantification. This limitation delays project development, increases costs, and reduces investor and regulatory confidence. CCS is essential for reducing emissions in hard-to-abate industries. Meanwhile geothermal energy offers a stable, renewable baseload, particularly in regions with limited solar or wind resources.

These technologies support multiple sustainability goals, including clean energy (SDG 7), climate action (SDG 13), and infrastructure development (SDG 9), and are especially relevant for industrial communities and countries with strong geothermal or CCS potential. Improving simulation of these subsurface reservoir systems capabilities could accelerate deployment, reduce risks, and play a crucial role in advancing the global transition to net-zero emissions.

How quantum could help

Quantum computing could offer potential acceleration in three critical subproblems: PDE solving, inversion and uncertainty quantification. In the first case, the complex multiphase flow equations of subsurface porous media systems could, in the future, benefit from quantum linear system algorithms such as variants of the Harrow–Hassidim–Lloyd (HHL) algorithm in an FTQC regime under specific sparsity/conditioning assumptions. While NISQ devices are insufficient for full-scale PDE solvers at the moment, future fault-tolerant hardware may enable large-scale reservoir simulations not currently tractable on classical HPC. In the second case, Quantum Monte Carlo and Quantum Amplitude Estimations could be used to reduce sampling-based overheads in uncertainty quantification, particularly in Bayesian inversion tasks for reservoir parameter estimation (i.e. permeability fields, probability estimation). In principle these methods could provide a quadratic speedup, but in reality any gains in the near-term will be problem specific and on toy-scale sized models, given the complexity of these reservoir systems. In the third case, quantum optimization approaches could be explored for well placement and injection/extraction strategy design, which are combinatorial optimization problems that could potentially be formulated on NISQ hardware today.