CO₂ plume geothermal systems offer a promising pathway for simultaneous carbon sequestration and renewable energy production, yet their optimization remains computationally prohibitive due to the complexity of coupled multi-phase flow, heat transport, and thermodynamic processes. This study presents a novel framework that integrates Non-isothermal Unsaturated-saturated Flow and Transport modeling with quantum neural network and hybrid quantum-classical ensemble regressors to accelerate CO₂ plume geothermal system design optimization. The methodology employs latin hypercube sampling to generate 1,000 Non-isothermal Unsaturated-saturated Flow and Transport simulations across several parameter spaces, extracting statistical features that undergo rigorous selection through Boruta, Chi-squared, and Pearson correlation algorithms with a standardized weight threshold of higher than 0.75. Two quantum architectures were developed to predict six geothermal variables, including system lifetime, injected, extracted, stored CO₂ mass, cumulative energy recovery and average heat extraction rate within lifetime. The quantum models achieved exceptional accuracy for most variables in the test section, with hybrid quantum-classical ensemble regressors architectures consistently outperforming quantum neural network variants, particularly when combined with boruta feature selection. Two optimization algorithms were employed for CO₂ plume geothermal system design, including moth flame optimization for single objectives and non-dominated sorting genetic algorithm Ⅱ for multi-objective scenarios to find robust optimal solutions based on developed surrogate models for injection overpressure, well spacing near and maximizing thermal energy extraction. The framework transformed a computationally intractable optimization requiring extensive simulation time into a rapid calculation while maintaining prediction accuracy comparable to full-physics models.

Using partial underground CO₂ storage as a working agent to harvest geothermal energy is a promising carbon capture, utilization, and storage (CCUS) method. It is particularly economically feasible to use or retrofit the existing infrastructure of a hydrocarbon field. Although technical advantages of integrated CO₂ sequestration and CO₂-circulated geothermal harvest using depleted hydrocarbon reservoirs have been reported, quantitative evaluations of economic benefits using existing wells of realistic reservoirs are rare. In this study, a 3-D hydrothermal flow model is built for the Triassic Argilo-Gréseux Supérieur (TAGS) Formation of the Toual gas field, Algeria using Schlumberger Petrel and CMG-STARS software. A three-phase operational scheme is proposed for sequential CO₂ sequestration and CO₂-circulated geothermal extraction over 100 years. The first phase is injecting CO₂ for 30 years, followed by concurrent cold CO₂ injection and hot CO₂ extraction in the developed CO₂ plume (circulation) for 40 years as the second phase. In the third phase, producing wells in the second phase are converted to injection wells while outer wells start to extract hot CO₂ for another 30 years. Scenario 1 is simulated using the selected nine existing wells of the field, while an optimized Scenario 2 is designed and simulated by adding seven newly drilled wells in addition to the existing wells. Scenario 3 shares the same numerical simulation of Scenario 1, but assumes the selected nine existing wells are newly drilled for the economic evaluation. Levelized Cost of Energy (LCOE), Net Present Value (NPV), and Return on Investment (ROI) are used as economic indicators. The results demonstrate that Scenario 2, which combines the use of existing and newly drilled wells, yields improved economic metrics compared to Scenario 1: 0.97 USD/MWh vs. 1.54 USD/MWh for LCOE and ＄2.9M vs. ＄1.1M for NPV. Both scenarios represent profitable endeavors, with ROI values of 1.3 % and 1.5 %, respectively. In contrast, Scenario 3 represents the worst-case scenario, with the highest LCOE at 2.90 USD/MWh and the lowest NPV and ROI at -＄0.4M and -0.2 %, respectively. The negative NPV and ROI in Scenario 3 indicates that CO₂-circulated geothermal harvesting in aquifers or giant depleted hydrocarbon fields, without leveraging existing infrastructure, is economically infeasible.

Industry-scale hydrogen is mainly produced by steam methane reforming (SMR), which uses natural gas as the feedstock and fuel and co-produces CO₂. This study aims to numerically evaluate hydrogen production by SMR under various reacting conditions. Unlike the previous studies with limited scenarios, the performance of SMR is continuously evaluated in a high-dimensional input-parameter space. The SMR plant including a combustor, a reformer, and a water-gas shifter is modeled in Aspen HYSYS software. The four key parameters, including methane fraction of the feedstock, reformer pressure and temperature, and shifter temperature, are treated uncertain and 50 samples are drawn from a four-dimensional parameter space defined by their ranges. Each sample is input to HYSYS model and mass ratio of each component in product streams is obtained as the output variables. Based on the 50 pairs of input-output data, response surfaces of the outputs are developed to surrogate HYSYS models. The fast response surface models are then used to calculate global sensitivity indices and evaluate SMR processes. Results show the reformer performance is controlled by temperature rather than pressure, and a temperature higher than 900 ${}^{\circ }$C can maximize the reaction rate. The water-gas shifting reaction is inhibited in the reformer but significantly enhanced in the shifter. Hydrogen is mainly produced in the reformer while the major function of the shifter is to convert CO to nontoxic CO₂.

In geothermal reservoir management, combined simulation-optimization is a practical approach to achieve the optimal well placement and operation that maximizes energy recovery and reservoir longevity. The use of machine learning models is often essential to make simulation-optimization computational feasible. Tools from machine learning can be used to construct data-driven and often physics-free approximations of the numerical model response, with computational times often several orders of magnitude smaller than those required by reservoir numerical models. In this short perspective, we explain the background and current status of machine learning based combined simulation-optimization in geothermal reservoir management, and discuss several key issues that will likely form future directions.

CO₂ circulated geothermal production can be integrated with CO₂ geological sequestration as a utilization method to offset cost. Investigation of heterogeneity impact is limited to CO₂ sequestration and its effect on CO₂ circulation and associated heat recovery is unclear. This study is aimed to improve the understanding of this problem by numerical experiments. A set of spatially correlated heterogeneous porosity fields is generated using a variety of geostatistical parameters, i.e., variance, correlation lengths, anisotropy and azimuth. Heterogeneous fields of intrinsic permeability and initial/residual water saturation are derived from porosity using equations regressed from a field dataset. Twenty combinations of injection pressure and well space obtained by Latin-Hypercube sampling are deployed in each heterogeneous field, generating a suite of numerical geothermal reservoir models. Performance indicators, including lifespan, net stored CO₂, produced heat flux, and total recovered heat energy in lifespan, are calculated from each model simulation. The simulation results suggest that geologic heterogeneity could develop high-permeable CO₂ flow paths, causing bypass of the hot low-permeable zones, shortened lifespan and reduced total recovered heat energy. Depending on the azimuth, anisotropy can create either flow barriers or preferential flow paths, increasing or decreasing heat sweeping efficiency. The relative angle between horizontal wells and the axis of maximum continuity of the heterogeneity can be optimized to maximize heat recovery efficiency. These finds provide useful insights of interplay between geological heterogeneity, well placement and operation of CO₂ circulated geothermal production.