CO2 storage efficiency: time, trap geometry and aquifer properties
Screening for CO2 storage locations in saline aquifers is a lot like exploring for oil and gas reservoirs. – But the dynamics of CO2 storage mean that we need to understand the dynamics of the whole aquifer system and not just the trap, says ExploCrowd’s Associated Specialist in Reservoir Engineering, Lex Rijkels.
– We don’t just need a trap that is big enough to hold all the CO2 we want to inject. We also need an aquifer that is big enough to absorb all the water we’re pushing out of the trap, to make room for the CO2.
The rock must also be permeable enough to allow these volumes of water to flow within a typical project life, Rijkels explains.
– Our experience is that the last factor is often limiting and that many commercially interesting cases don’t reach their ultimate storage potential within 25 years, as there is simply not enough time to push enough water out of the trap with wells that inject at the high rates required for a commercial project.
– Many of the commercially interesting cases don’t reach the ultimate storage potential within 25 years.
LEX RIJKELS
Screening and early prospects
While the ultimate aquifer capacity can be estimated arithmetically, addressing the critical time dependent factors requires an assessment of the dynamic performance of the whole aquifer system.
– Typically, such an assessment is performed using reservoir simulation models, which are a major undertaking to build and computationally expensive to run, Rijkels says.
The expense and resource intensity of such models means that they are rarely used in screening/early prospect evaluation, with the result that the dynamic aquifer performance is not quantified until late in the evaluation stage, leading to poor prioritization and missed opportunities.
BUILDING A PROXY MODEL
The model can be used to screen the optimal number of wells to achieve ultimate storage within a given project life.
The proxy model was built by running 1500 simulation models for various combinations of the input parameters, Rijkels explains.
– We then regressed functional descriptions for the initial rate, its decline, and late-time behaviour using the eight geological parameters as input. That gave us shape curves that describe the injection forecast per well for up to 50 years, which we found matched full-field forecasts to within 10-20% for a broad range of cases and time steps.
Evaluating traps – estimating potential
Output from the proxy model is shown below. ExploCrowd use the most promising aquifers and traps, and to make an informed first guess of the number of wells needed in a full-fields simulation model, of high-graded projects.
– We find it to be a huge advance on the traditional ‘storage efficiency factor’ approach which we believe to be inaccurate and occasionally misleading. The proxy model enables an early focus on the wider aquifer system and provides an accurate estimate of its potential to support CO2 injection over project lifetimes, Rijkels says.
– Of course, we continue to use full reservoir simulation models to incorporate the detail of the aquifer and trap properties, to understand the movement and dissolution of the CO2 plume and to optimise the number and placement of injector wells, but we can now focus this effort on assessing and optimising projects that we know have commercial potential, he adds.