Abstract
Long-term containment of CO2 and storage security depends mainly on physical and geochemical trapping mechanisms. Injected CO2 dissolves in the formation fluid and causes a sharp decrease in pH which in turn drives the dissolution and precipitation of minerals. Geochemical modelling is an important tool to understand and predict the behaviour of CO2 reactivity. In this paper three numerical codes, PHREEQC, GEM and TOUGHREACT, are compared with respect to brine – CO2 – rock reactions. Formation water compositions and mineralogies of three sandstone core samples from target CO2-storage formations were used as input for kinetic models. Mineral replacement reactions were observed when CO2 injection was initiated in all scenarios. While PHREEQC and GEM were generally in good agreement, TOUGHREACT gave diverging predictions on two models. It is considered that the discrepancies are caused primarily by the differences in the thermodynamic databases and activity models. The uncertainties in these calculations suggest that appropriate experimental tests should be performed to validate the models, so that they can be used to make predictions at the field scale.
Original language | English |
---|---|
Pages (from-to) | 1805-1816 |
Number of pages | 12 |
Journal | Chemical Engineering Research and Design |
Volume | 89 |
Issue number | 9 |
DOIs | |
Publication status | Published - Sept 2011 |
Keywords
- Chemical reactions
- CO 2 sequestration
- Dissolution
- Geochemical simulation
- North Sea
- Precipitation