For CO2 capture and storage to play a relevant role in the reduction of European emissions, activities need to be scaled up significantly from what we are doing today. This will require good estimates on how much CO2 can be practically and safely stored at each site. To minimize risk of leakage, it is also necessary to have a good understanding of the long-term fate of the injected CO2, thousands of years into the future.
Commonly used simulation tools today are of limited use to address these questions as they are primarily developed to meet the needs of the petroleum industry. Another challenge is the lack of measured data needed to calibrate and validate theoretical models. Approximately twenty years of measured data related to CO2 storage exists today, but this covers only a tiny fraction of the long-time perspective that needs to be considered.
In this project, these problems were addressed by developing specialized software for in-depth analysis of potential storage sites in terms of storage capacity, optimal use, and long-time containment. The methods used are based on mathematical formulations especially developed for modelling CO2-storage, and are therefore able to produce better and more rapid results than existing 3D-simulators. In particular, we have focused on methods which allow for efficient computation of parameter sensitivities. These methods are well suited for use in automatic simulation frameworks either for finding optimal site utilization or for integrating monitoring data. We have used the developed software to estimate maximal storage potential, identify optimal strategies for injection and to estimate model parameters from observed data.
The project was carried out in cooperation with a research group at the University of Texas, that has studied natural geological CO2 deposits for many years. The Bravo Dome in New Mexico is one such deposit, where CO2 of volcanic origin has been trapped underground for over a million years. This site is well known and has been studied for a long time, and very good and extensive measurements are available. Such natural CO2 deposits are (and will remain) the only direct sources of information about long-term geological storage of CO2. The software developed in this project has been used to try to understand the data acquired at Bravo Dome.
We put large effort into making the source code developed in this project well documented and usable for others. The code has been distributed as open source software under the GPL license as part of the Matlab Reservoir Simulation Toolbox (MRST)
1) Getting the CO2 specific simulation stage to a level where real aquifer complexities can be taken into account.
2) In practice show how simulation-based storage estimates can be achieved by novel work flows and communicated the need for such estimates for large scale utilization of CO2 storage.
3) Introducing adjoint-based sensitives for model matching and evaluation of monitoring strategies.
How to inject hundreds of megatonnes of CO2 in large saline aquifers is a key question if geological carbon storage is to mitigate climate changes. In the project, we will build a set of computational tools that can be used to study and optimize such operations. The project is a natural continuation of the CLIMIT projects MatMoRA, MatMoRA-II, IGEMS-CO2, and Numerical CO2 laboratory. Building on the accumulated knowledge and open-source software developed in these projects, we now take our research one step further to investigate the large-scale problem of CO2 storage in a socioeconomic perspective. To achieve this, we continue our philosophy of having a chain of modeling tools ranging from light-weight, approximate methods to advanced 3D modeling. In this multifaceted tool-chain, different types of approximations are made to address questions like: How much CO2 can be injected? Where does the CO2 go during the injection period? What is the large-scale pressure response? How much CO2 will leak? Where should injection hubs be placed to maximize utilization of pore space? And so on. In addition, we will use a natural CO2 field as an analogue to obtain new knowledge of processes relevant to long-term CO2 storage and to validate and verify our computational methods. This will be achieved through a close cooperation with Prof. Marc Hesse and the Geological Porous Media Group at The University of Texas at Austin.
In brief, the project will contribute new computational methods and workflows for evaluating large-scale CO2 storage potential in an aquifer-wide setting. Model parameters relevant to long-term storage will be obtain from a natural CO2 field. And last, but not least, the project will produce workflows and open-source, open-data tool-chains for evaluating large-scale storage scenarios particularly relevant to saline aquifers from the North Sea.