Samira Mohammadkhani

Direct air capture (DAC), as a complementary strategy to carbon capture and storage (CCS), offers a scalable and sustainable pathway to remove CO₂ directly from the ambient air. This study presents a detailed evaluation of the amine-functionalised metal-organic framework (MOF) sorbent, mmen-Mg₂(dobpdc), for DAC using a temperature–vacuum swing adsorption (TVSA) process. While this sorbent has demonstrated promising performance in point-source CO₂ capture, this is the first dynamic simulation-based study to rigorously assess its effectiveness for low-concentration atmospheric CO₂ removal. A transient one-dimensional TVSA model was developed in Aspen Adsorption and validated against experimental breakthrough data to ensure accuracy in capturing both the sharp and gradual adsorption kinetics. To enhance process efficiency and sustainability, this work provides a comprehensive parametric analysis of key operational factors, including air flow rate, temperature, adsorption/desorption durations, vacuum pressure, and heat exchanger temperature, on process performance, including CO₂ purity, recovery, productivity, and specific energy consumption. Under optimal conditions for this sorbent (vacuum pressure lower than 0.15 bar and feed temperature below 15 °C), the TVSA process achieved ~98% CO₂ purity, recovery over 70%, and specific energy consumption of about 3.5 MJ/KgCO₂. These findings demonstrate that mmen-Mg₂(dobpdc) can achieve performance comparable to benchmark DAC sorbents in terms of CO₂ purity and recovery, underscoring its potential for scalable DAC applications. This work advances the development of energy-efficient carbon removal technologies and highlights the value of step-shape isotherm adsorbents in supporting global carbon-neutrality goals.

Efficient management of a CO2 storage site requires an uncertainty assessment for storage capacity and injectivity. For the case of a saline aquifer onshore Denmark, an unsteady-state core-flooding experiment was conducted in order to assess a potential risk of CO2 injectivity impairment. The experimental results indicate that the rock mineralogy, samples permeability, and samples dimensions were essentially unaffected by CO2 flooding, but precipitated salt and some fines were detected after the end of experiment.The experimentally obtained differential pressure readings and the brine effluent volume are history-matched to estimate the uncertainty range of the saturation functions for the two-phase water-scCO2 system. The Brooks-Corey (1964) parametrization is used for the relative permeability functions, and the Skjaeveland et al. (1998) parametrization – for the capillary pressure function. The fluid system is modelled with the Peng-Robinson equation of state for the aqueous phase with CO2 dissolution, geochemical reactions, and with salt precipitation. The model is implemented in a commercial simulator GEM (CMG, 2022).A global search method with multiple starting guesses is used to obtain a solution to the inverse model. However, the GEM model appears to be computationally expensive for multiple forward solutions; a faster simplified two-phase immiscible model, built using MRST (Lie, 2019), is used as a proxy to the CMG model.The convergence history of the global search method consists of approximately 50,000 records contains the residuals between the numerical solution and the experimental results, and the corresponding 10 optimization parameters (critical saturations, Corey exponents, endpoint permeabilities, and the Skjaeveland et al. (1998) parameters of the capillary pressure function). Out of these 50,000 records, approximately 800 correspond to small residuals, which yield a good match to the experimental data. For the available computational setup (40 cores, 128 Gb RAM), it takes approximately a week to obtain the convergence data.The converged non-unique MRST solutions, which lie withing a prescribed threshold from the experimental data, are correspond to certain ranges of the Brooks-Corey-Skjaeveland parameters. However, the number of converged solutions is not sufficient to characterize the corresponding distributions.A feedforward neural network is set up to approximate the mapping between the tuples of Brooks-Corey-Skjaeveland parameters and the values of the residual between the experimental results and the MRST predictions. The network, implemented with Keras (Chollet, 2015) and PyTorch (Paszke et al., 2019) backend, consists of 4 layers with 1,000 neurons each, connected with the RELU activation functions. A threshold-based weighting is used to prioritize samples with small residuals. The neural network is trained on the whole convergence history of the global search method, using 10 epochs and the Adam optimizer (Kingma, Ba, 2014). The trained network reaches good accuracy overall with Pearson’s correlation coefficient close to unity. The training takes approximately a minute on a single GPU. The forward solution with the neural network is several orders of magnitude faster as compared to an MRST forward run. Uncertainty quantification of the converged Brooks-Corey-Skjaeveland parameters is done in a Bayesian framework, implemented using CUQIpy (Alghamdi et al., 2023), (Riis et al., 2023). The history-matched values of the optimization parameters are interpreted as samples of a Gaussian prior distribution. It is shown that the parameters of the capillary function are subject to larger uncertainty, as compared to the parameters of the relative permeabilities. The posterior distribution is obtained by evaluating the joint distribution for small values of the residual. Using the neural network as a forward problem solver allows to sample the posterior distribution and to obtain the maximum a posteriori estimation of the sought optimization parameters.

We present a systematic study of crude oil–brine–rock interactions in tight chalk cores at reservoir conditions. Flooding experiments are performed on outcrops (Stevns Klint) as well as on reservoir core plugs from Dan field, the Ekofisk and Tor formations. These studies are carried out in core plugs with reduced pore volumes, i.e., short core samples and aged with a dynamic ageing method. The method was evaluated by three different oil compositions. A series of synthetic multicomponent brines and designed fluid injection scenarios are investigated; injection flow rates are optimized to ensure that a capillary-dominant regime is maintained. Changes in brine compositions and fluid distribution in the core plugs are characterized using ion chromatography and X-ray computed tomography, respectively. First, we show that polar components in the oil phase play a major role in wettability alteration during ageing; this controls the oil production behavior. We also show that, compared to seawater, both formation water and ten-times-diluted seawater are better candidates for enhanced oil recovery in the Dan field. Finally, we show that the modified flow zone indicator, a measure of rock quality, is likely the main variable responsible for the higher oil recoveries observed in Tor core samples.