Less than 10% of oil is usually recovered from liquid-rich shales and this leaves much room for improvement, while water injection into shale formation is virtually impossible because of the extremely low permeability of the formation matrix. Injecting carbon dioxide (CO2) into oil shale formations can potentially improve oil recovery. Furthermore, the large surface area in organic-rich shale could permanently store CO2 without jeopardizing the formation integrity. This work is a mechanism study of evaluating the effectiveness of CO2-enhanced oil shale recovery and shale formation CO2 sequestration capacity using numerical simulation.
Sumeer Kalra1, Wei Tian2, Xingru Wu2
1Warwick Energy Group, 900 W Wilshire Blvd, Oklahoma City, OK 73116, USA. 2University of Oklahoma, 100 E Boyd St. Sec 1210, Norman, OK 73019-1003, USA
Received: 8 June 2017
© The Author(s) 2017.
Petrophysical and fluid properties similar to the Bakken Formation are used to set up the base model for simulation. Result shows that the CO2 injection could increase the oil recovery factor from 7.4% to 53%. In addition, petrophysical characteristics such as in situ stress changes and presence of a natural fracture network in the shale formation are proven to have impacts on subsurface CO2 flow. A response surface modeling approach was applied to investigate the interaction between parameters and generate a proxy model for optimizing oil recovery and CO2 injectivity.
An unconventional reservoir is a hydrocarbon resource that could not be economically recovered without stimulation because of its extreme low permeability. Even with current stimulation and completion technologies, many fields are expected to produce less than 10% of initial oil in place. CO2-enhanced oil recovery (EOR) for liquid-rich unconventional reservoirs has been investigated through core flooding (Wang et al. 2017) and field ‘huff-n-puff’ pilot tests (Hoffman and John 2016). Our literature survey shows that field applications of enhanced oil recovery (EOR) for unconventional oil reservoirs are limited.
The rising level of CO2 in the atmosphere has caused concerns about temperature increases. Carbon capture and storage (CCS) in geological formations deep beneath the surface is an effective approach to reduce CO2 level in the atmosphere (Sorensen et al. 2009; Shen et al. 2015). Geological properties of the prospective CO2 storage formations are analogous to the ones in oil- and gas-producing fields. It signifies the necessary geological conditions that support hydrocarbon accumulation are also conducive for CO2 storage.
CO2 injection and displacement processes are significantly different from each other in tight/shale formations as flow through fractures dominates CO2 displacement. For tight formations, the injected CO2 flows through hydraulic fractures first and then migrates into the rock matrix to swell and displace in-depth oil. In the displacement process, CO2 in the reservoir condition behaves as a super critical fluid, which has the viscosity of a gas and liquid-like density. Figure 1 shows the volume change of CO2 with respect to depth. The CO2 property calculation is based on IUPAC correlations (Augus et al. 1973). It is valid from 220 to 1100 K, and the maximum pressure is 1000 bars below 700 K, 600 bars from 700 to 1100 K.
Fig. 1 Volume and density changes with depth for CO2 by considering a geopressure gradient of 9794 Pa/m and geothermal gradient of 0.027 K/m for a volume of 2.83 m3 of CO2 at surface conditions (Kalra and Wu 2014).
A study of the mechanism of CO2 injection in liquid shale reservoirs
A compositional model was set up to study CO2 injection and migration in a synthetic liquid shale reservoir for EOR and CO2 sequestration. To focus on the flow mechanisms of CO2 inside the reservoir, a zone of study (ZoS) was created with one injector and one producer. Both wells were horizontal wells within the 12.2 m of pay zone and had a lateral length of 122 m each, located in the middle of the pay zone. The laterals of the two wells were parallel and 304.8 m apart, with fracture perforations in opposite directions, as shown in Fig. 2. We assumed that the flow pattern between each hydraulic fracture pair is similar to the others. Therefore, considering such symmetry, we only included one hydraulic fracture for each well in the ZoS to simplify the simulation (Fig. 2).
Fig. 2 X-Y cross section of the base reservoir model showing the injection and production wells and hydraulic fracture (i.e., zone of study). Blue lines represent the horizontal wellbores.
Table 1 summarizes the Middle Bakken reservoir properties from the literature, and the averaged values were implemented in our reservoir model. Table 2 gives the reservoir grid dimensions used in the base model and the volume estimated from the simulated reservoir. The gridding of ZoS is shown in Fig. 2. Grid blocks along the Y- and Z-directions had equal dimension of 12.2 and 1.22 m, respectively.
Table 1 Range and average reservoir properties used to build the base model.
The grid blocks in the X-direction consist of local grid refinement near hydraulic fractures. Each hydraulic fracture had a grid dimension of 0.0089 m as width in the X-direction, with logarithmically increasing grid size toward the center to avoid convergence issues. The half-length and height of the injector hydraulic fracture were assumed to be 122 m in the Y-direction and 12.2 m in the Z-direction, respectively. The injector hydraulic fracture had an enhanced permeability of 2.5E−13 m2. The hydraulic fracture connected to the producer at the other side had the same fracture dimensions as the injector fracture but a different fracture permeability of 6.9E−14 m2. The hydraulic fracture properties defined in the reservoir model are summarized in Table 3. Matrix grid blocks have permeability of 4.9E−18 m2.
Table 2 Grid dimensions and volume calculations for the base reservoir model.
Table 3 Hydraulic fracture properties in the reservoir model (Kurtoglu 2014; Kumar et al. 2013).
The oil from the Middle Bakken Formation is light crude with an average density of 42° API. The fluid characterization in our model was based on the Nojabaei et al. (2013) study, and the input parameters for the phase behavior study are shown in Tables 4 and 5. The modified Pederson correlation was applied to estimate the viscosity of the reservoir fluid. The calculated bubble point pressure for the fluid was about 1.9E+07 Pa. The CO2 flooding is a miscible flooding.
Table 4 Compositional data of reservoir fluid (Nojabaei et al. 2013).
Two separate sets of relative permeability curves were used in this study. One defined the matrix, and the other was for hydraulic fractures. The initial water saturation is 0.30. The relative permeability curves closely matched the curves obtained by history matching for the Elm Coulee field in North Dakota (Shoaib and Hoffman 2009).
Table 5 Binary interaction parameters for the fluid component (Nojabaei et al. 2013).
Continuous injection of pure CO2 was modeled. The simulation time was 30 years, starting from January 1, 2010, with an injection bottom-hole pressure of 6.9E+07 Pa, which was below the formation fracturing pressure of from 7.6E+07 to 7.9E+09 Pa observed for Middle Bakken Formation. The injection and production constraints are summarized in Table 6.
Table 6 Well properties and operation restrictions used in the reservoir model (Kumar et al. 2013; Kurtoglu 2014; Shoaib and Hoffman 2009).
Oil shale/tight reservoir characteristics
To accurately simulate flow in a tight formation with ultra-low permeability, a complex fracture network is a crucial part of model. In our study model, hydraulic fractures and natural fractures were explicitly created for simulation.
Reservoir heterogeneity modeling
Reservoir heterogeneity depends on formation lithology, depositional environment and fracture development due to stress changes. Studies of the lithology and depositional environment of the Parshall field in North Dakota showed that the Middle Bakken Formation could be divided into eight different litho-facies with depth, as observed in various studies (Simenson 2010; LeFever 2011). The three integrated litho-facies were divided into 10 layers of the model (Table 7).
Table 7 Heterogeneous thickness layer in reservoir model clubbed in 3 types to consider the 10 different lithology layers.