This thesis investigates the effect of perforation strategies on CO2 injectivity in depleted gas fields used for Carbon Capture and Storage (CCS). Explored components of these strategies include the state of the wellbore upon perforation, the combination of shaped charge and gun system (perforation technique) employed, and perforating after reservoir recharge. The injectivity of CO2 was assessed by SLB through perforation experiments, which showed that different perforation strategies significantly alter the performance of perforations in cores. However, in conventional reservoir models, these effects are not captured. To bridge the gap between experiments at the core scale and perforation effects at the reservoir-scale, a set of numerical models was developed. Firstly, a high-resolution model that reproduced the experimental results was created. Its output was applied in a well-scale model, in which perforation geometries and properties were embedded in the mesh to find perforation effects in a section of the reservoir. Finally, these results were upscaled into a coarser model by adapting the well index with the corresponding skin factor.

Generally, it was found that the inclusion of perforation effects in the well-scale models, reduced the injectivity of the well. For the wellbore state upon perforation, an empty wellbore was found to be effective in contrast to a fluid-filled wellbore. Moreover, among the perforation techniques, the following trends were observed: technique 1 attains the highest injectivity due to its improved performance in high-permeability layers, and technique 3 has the most consistent performance in all layers of the model. Lastly, reservoir recharge results in enhanced perforation clean-up in combination with technique 1, providing the only strategy where the associated skin is negative.

Furthermore, an analysis on flow in sections of the reservoir was conducted, which reveals that while crossflow between layers takes place, the performance of the perforation in the high-permeability layer is dominant in determining the total injectivity of the well. This finding puts an emphasis on targeting high-permeability layers and choosing the perforation strategy that is suitable for these layers. Thus, the perforation strategy that is found to be most suitable employs an empty wellbore, perforation technique 1, and potentially reservoir recharge if permitted by logistic and economic constraints.

Though the framework described in this study demonstrated that perforation effects can be efficiently upscaled to the reservoir-scale, a number of assumptions were made to yield numerical stability and convergence in simulations. The assumption of Darcy flow in the clean tunnel may be the most significant, as it could have resulted in an overestimation of the found crushed zone permeability.