Consequences of thermal fracture developments due to injection cold CO2 into depleted gas fields

CO2 storage is planned in a depleted gas field called P18, which is located offshore in the vicinity of the Dutch coast. This project is also known as the ROAD project, which is the Rotterdam capture and storage demonstration project. In the P18-4 compartment, cold CO2 will be injected into a formation with a temperature of 120°C. Cooling of the reservoir with the cold CO2 will induce thermal stresses, similar to those resulting from injection of cold water. The thermal stresses around the well can promote the propagation of fractures into the reservoir and possibly the caprock. This is usually called thermally induced fracturing. In contrast to cold water injection, for CO2 injection thermally induced fracturing is usually not taken into account. In this paper, we are exploring the potential development of thermal fractures and their impact on the reservoir rock, caprock and well injectivity of the P18-4 reservoir. The P18-4 reservoir was modelled by a modified TOUGH2/ECO2M module. The module is designed for geological storage of CO2 in a wide range of temperature, pressure conditions and therefore different CO2 phase conditions, but was not able to model phase transitions in low pressure – high temperature reservoir. We modified the ECO2M module to be able to model the transition of gaseous to liquid CO2, needed for the pressure temperature conditions investigated in P18-4. To estimate the maximal extent of a fracture into the reservoir we started with 2D simulations of cold CO2 injection without any fracture development in TOUGH2/ECO2M module. The resulting pressure and temperature fields were used in a geomechanical analysis to determine the thermo-elastic and poro-elastic stresses, and to assess the fracture development into the reservoir (using the finite-element software tool DIANA). After completing a number of timesteps, these fractures were introduced back into the TOUGH2 model and the potential injectivity enhancement was analysed. The TOUGH2-DIANA coupling was made semiautomatic, therefore an efficient exchange of data between the two software packages was possible. The DIANA model was initialised with the corresponding stresses at a depth of 2500m, rather than the actual depth of 3200m at which the minimum in-situ stress was too high for any (thermal) fracture to occur. On one side of the cylinder the fracture development was modelled. We employed a layered model representing a quarter cylinder of the reservoir for the TOUGH2 simulations. The model was initialised at 20 bar and 100°C. Injection of 1.1 Mton/yr was assumed, which resulted in a temperature front up to 250 meters into the reservoir after 5 years of injection. The quarter cylinder was mapped upon a 2D plane strain geometry for the geomechanical simulations. The large temperature changes resulted in a fracture development in “the shallow P18-4 reservoir” (2500m depth) of up to 140 m after 5 years of injection. The fracture development was translated into a permeability change in the fractured part of the reservoir and in this way we are able to predict the new injectivity of the reservoir. The injectivity increased from 0.7 Kg/s bar to 2.0 Kg/s bar. In the 2D simulations, we have seen the extent of a fracture in a single-layer reservoir, however this does not completely reflect the vertical heterogeneity in “the shallow P18-4 reservoir”. Also, it was not possible to examine the effect on the caprock with this model. Therefore a 3D model was constructed in both TOUGH2 and DIANA, which resulted in even more shallow P18 reservoir in order to show the ability to simulate the development of a crack inside the caprock. Not only horizontal propagation of the fracture was considered, but also to allow vertical fracture propagation of the fracture into the caprock. The stresses due to the cold injection have to overcome not only the horizontal minimum stress, but also the vertical stress in order to propagate into vertical direction. From the 2D simulations and 3D simulations, it was clear that for the true depth (3200m) of the P18-4 reservoir, no fracturing as a result of the cold CO2 is expected, which is in line with the earlier results. However in “the shallow P18-4 reservoir” (2500m) there is fracture development into the reservoir. The 3D simulation results are very sensitive for the values of the following properties: permeability in the fracture, brine salinity and the reduction of permeability due to salt precipitation. The main impact of thermal fractures are: · Well injectivity increases. · The shape of the CO2 front changes, which means that the cold CO2 can extent much further away from the well than without fracturing. · The probability of the development of fractures in both the reservoir and the caprock increases. · CO2 flow patterns become complicated due to salt precipitation and blocking in the reservoir and/or caprock · The result is very sensitive for the values of the following properties permeability in the fracture, brine salinity and the reduction of permeability due to salt precipitation