Phase behavior of SCCO₂ sequestration and enhanced natural gas recovery

Abstract

Some noncommercial gas reservoirs with low reserves are feasible sites for CO₂ sequestration. Those gas reservoirs contain natural gas that can take up the potential pore space of SCCO₂ sequestration in the reservoir. The solution is to recover the natural gas by active CO₂ injection. This idea is carbon sequestration with enhancement gas recovery (CSEGR). In CSEGR, different zones of the formation fluid are formed during the gas migration. In this paper, the sequestration site is a PY gas reservoir. The pressure, volume and temperature properties of the formation fluid are tested by experiments or calculated by the program based on PR-EOS, using a Z-factor, Volume ratio in place ($V_{{{\text{r}} . {\text{scco}}_{ 2} }}$), density and viscosity. We discuss those experimental or simulation results to understand the fluid phase behavior in such a migration during CSEGR in a PY gas reservoir, and we give the suitable site (temperature) and the eligible pressure of the next core-flooding test.

Introduction

Some noncommercial gas reservoirs with low gas reserves are feasible sites for CO₂ geological sequestration. Many of them contain natural gas that can be potentially recovered. CO₂ sequestration in those natural gas reservoirs can be coupled with enhanced gas recovery by injecting CO₂. The added gas recovery can be used to offset the cost of CO₂ capture and storage (CCS). This idea was first planned for abandoned gas reservoirs and called carbon sequestration with enhanced gas recovery (CSEGR) (Oldenburg 2003). In reality, typical noncommercial gas reservoirs are similar. Although CSEGR has been discussed for more than 10 years (for example, Blok et al. 1997), the published field tests are only in Hungary (Kubus 2010), the Netherlands (Van der Meer et al. 2005) and the USA (Turta et al. 2008).

As estimated in the Joule II Non-nuclear Energy Research Program, for maximum storage capacity, CO₂ has to be stored as supercritical CO₂ (SCCO₂). Published basic research on CSEGR simplifies real natural gas as pure CH₄ (Mamora and Seo 2002; Seo and Mamora 2003; Oldenburg 2003; Nogueira and Mamora 2005; Turta et al. 2008). Such research suggests that SCCO₂ and natural gas should not completely mix in the reservoir during the gas migration. However, the mix is multi-contact and creates the SCCO₂–natural gas transitional zone. Thus, the formation fluid in the whole reservoir size could be simply divided into three zones on the swept region. Such areas are the SCCO₂ zone, SCCO₂–natural gas transitional zone and natural gas zone (Figs. 1, 2). In this paper, the region connecting the SCCO₂–natural gas transitional zone and the natural gas zone is called the “displacement front.” In addition, the region connecting the SCCO₂ zone and the SCCO₂–natural gas transitional zone is called the “storage front,” both of which are shown in Figs. 1 and 2.

Fig. 1

Schematic of CSEGR in the horizontal direction

Fig. 2

Schematic of CSEGR in the vertical direction

Some researchers now believe that gas condensate reservoirs are another possible site for CCS because the rise in formation pressure caused by a SCCO₂ injection can enhance condensate oil recovery (Sobers et al. 2004; Mbarrufet et al. 2009; Ramharack et al. 2010). That condition requires more SCCO₂ to mix with the condensate gas, which is very different from the CSEGR method discussed in this paper.

There are still other studies in the literature by famous research groups that delve into coupling oil recovery and carbon sequestration, such as “Micromodel investigations of CO₂ exsolution from carbonated water in sedimentary rocks” (Zuo et al. 2013), “Multi-scale experimental study of carbonated water injection” (Alizadeh et al. 2014), “CO₂ injection as an immiscible application for enhanced recovery in heavy oil reservoirs” (Khatib et al. 1981) and so on.

In this paper, the sequestration site is a PY gas reservoir. The pressure, volume and temperature (PVT) properties of the SCCO₂ zone, SCCO₂–natural gas transitional zone and natural gas zone are tested by experiments or calculated by a program based on the PR-EOS, using Z-factor, Volume ratio in place ($V_{{{\text{r}} . {\text{scco}}_{ 2} }}$), density and viscosity. We discuss those results to understand the phase behavior of each fluid zone during the gases migrations in gas reservoirs under the repressurization caused by the active CO₂ injection. We also attempt to assess the ideal injection site and several eligible pressures for CSEGR based on such PVT properties. The paper makes clear the necessity and feasibility of CO₂ sequestration in reservoirs and CO₂ injection for the improvement of gas recovery. It accordingly recommends the feasible injection depth of supercritical CO₂ and the practical gas production pressure range for enhancing the recovery of the PY gas reservoir.

Theory

Z-factor

CSEGR depends on the supercritical phase behavior of CO₂ and the multi-contact between SCCO₂ and the natural gas. The degree of nonideality and supercriticality shown by the gases can be expressed by the Z-factor. Z-factor is also the key to gain other PVT properties. We have generated Z-factors for the SCCO₂ zone, SCCO₂–natural gas transitional zone and natural gas zone of the target gas reservoir by experimental and phase calculation methods.

Laboratory measurements are taken with the use of a PVT cell. The basic operation consists of pressurizing a known volume of gas in a PVT cell within a temperature-controlled oven (Sobers et al. 2004). In this paper, the Z-factors of pure CO₂, a 23.33 % CO₂–natural gas mixture and the pure natural gas are determined by experimental testing. Such Z-factors are tested by the DBR company’s JEFFRI PVT instrument, which can be used under high temperatures and pressures. When $Z_{\text{f}}$ is defined as the Z-factor for PVT cell conditions, the experimental testing method is given by:

$$Z_{\text{f}} = \frac{{V_{\text{f}} P_{\text{f}} T_{\text{sc}} }}{{V_{\text{sc}} T_{\text{f}} P_{\text{sc}} }}$$

(1)

where $V_{\text{sc}}$ (m³) is the gas volume at standard temperature, $T_{\text{sc}}$ (°C), and standard pressure, $P_{\text{sc}}$ (Pa). $V_{\text{f}}$ is the gas volume at the temperature and pressure in the PVT cell. $Z_{\text{f}}$ is the Z-factor under cell PVT conditions. The standard condition in China is 20 °C and 1.10e5 MPa.

Then, we select the suitable calculation method for the Z-factor based on the measured values for a 23.33 % CO₂ (volume fraction)–natural gas mixture under different conditions. The calculation method options are the Soave–Redlich–Kwong EOS (Soave 1972), Peng–Robinson EOS (Peng and Robinson 1976) and experience formulas such as the Hall–Yarborough method (Hall and Yarborough 1973), Dranchuk–Purvis–Robinson method (Dranchuk et al. 1974), Dranchuk–Abu–Kassem method (Dranchuk and Abou-Kassem 1975), Hankinson–Thomas–Phillips method (Hankinson et al. 1969), Li method (Li and Gang 2001) and Zhang method (Zhang et al. 2005). Experience formulas need to be combined with non-hydrocarbon correction methods to gain the higher accuracy of the acid gas prediction. We choose the Guo correction (Guo et al. 2000). The PR-EOS has the greatest accuracy and fits with the Chinese oil/gas engineering standard to predict the Z-factors of a CO₂–natural gas mixture. The relative average deviation for different conditions is 0.94 %. Therefore, we select the PR-EOS to predict the Z-factors of the formation fluid.

Volume ratio in the place ($V_{{{\text{r}} . {\text{scco}}_{ 2} }}$)

CSEGR, as a development of CCS, should also account for the effect of carbon sequestration. To do this, the Volume ratio in place ($V_{{{\text{r}} . {\text{scco}}_{ 2} }}$) as the volume ratio between the formation fluid and CO₂ of the same moles on a certain reservoir condition is defined. If this parameter is less than 1, the volume of the formation fluid is less than the same moles of CO₂. Such a condition will be helpful to CO₂ storage. On the other hand, if this parameter is greater than 1, the fluid squeezes the SCCO₂ storage space and is more useful to EGR than SCCO₂ under the formation conditions. The Volume ratio in place ($V_{{{\text{r}} . {\text{scco}}_{ 2} }}$) is given by:

$$V_{{{\text{r}} . {\text{ scco}}_{ 2} }} = \frac{{V_{i} }}{{V_{{{\text{CO}}_{2} }} }} = \frac{{Z_{i} T_{i} P_{\text{f}} }}{{Z_{{{\text{CO}}_{2} }} P_{i} T_{\text{f}} }}$$

(2)

where $V_{{{\text{r}} . {\text{scco}}_{ 2} }}$ is the Volume ratio in place, $V_{i}$ (m³) is the gas volume at a certain temperature, $T_{i}$ (°C), and certain pressure, $P_{i}$ (Pa), and $Z_{i}$ is the Z-factor under the same conditions. “i” can be the pure CO₂ at another temperature or pressure. The $V_{{{\text{r}} . {\text{scco}}_{ 2} }}$ can be helpful to estimate the ideal injection site for CO₂. “i” can be the CO₂–natural gas mixture, or the natural gas. Then, the $V_{{{\text{r}} . {\text{scco}}_{ 2} }}$ suggests the ability of EGR with SCCO₂. $V_{{{\text{CO}}_{2} }}$ (m³) is the volume of the pure CO₂ system under certain reservoir conditions, and $Z_{{{\text{CO}}_{ 2} }}$ is the Z-factor for such conditions.

We can plot the $V_{{{\text{r}} . {\text{scco}}_{ 2} }}$–pressure ($V_{{{\text{r}} . {\text{scco}}_{ 2} }}$–p) curves of the SCCO₂ zone, SCCO₂–natural gas transitional zone and natural gas zone of the target gas reservoir by Eq. (2) based on Z-factors.

Density and viscosity

Density and viscosity are important PVT properties affecting the gases migrations in the reservoir. However, traditional experiments for these two-phase properties are usually costly or time-consuming. Many experts used novel correlations to study the density and viscosity in PVT experiments. Hemmati-Sarapardeh et al. (2013) studied reservoir oil viscosity correlations. Naseri et al. (2014) found a correlation approach for predicting the PVT properties of reservoir oils. We have made a program mainly based on the PR-EOS to predict viscosity and density together, and the viscosity model of a program presented by Guo (Guo et al. 1999) and based on the PR-EOS. Compared with the two above predictions, the calculated results are credible and within the acceptable range.

The benefits of CSEGR

Target reservoir and the natural gas

PY gas reservoirs are located in the high point of the TYY structure of EHD fault-salient in a LC rifted basin (Fig. 3). Its depth is 900–1028 m. Geological properties and the natural gas hydrocarbon composition of the TQ layer are shown in Table 1. It is estimated as a low permeability and low porosity reservoir with low dry gas reserves abundance. In addition, the reservoir has a tight cap rock without bulk porosity and bulk permeability above it (Fig. 4). It is a possible site to perform CSEGR.

Fig. 3

PY arch structure reservoir

Table 1 Reservoir properties and natural gas composition of the QT layer

Fig. 4

Gas diagram of the PY

Phase behavior of natural gas

Figure 5 graphs the Z-factors–pressure (Z-factors–p) curves for natural gas. Figure 6 graphs the $V_{{{\text{r}} . {\text{scco}}_{ 2} }}$–p curves of the gas in the TQ layer. Figure 7 graphs the phase diagram for the gas with iso-density lines. Moreover, Fig. 8 graphs the phase diagram for the gas with iso-viscosity curves. These figures have a typical pressure–temperature profile from the wellhead to the bottom of PY reservoir (p–T profile). Figure 5 suggests that the gas shows nearly ideal gas behavior and supercriticality is not obvious. Figure 6 indicates that the volume of the gas is over 2 times that of SCCO₂ in the reservoir. Figure 7 shows that the density of the gas is less than 100 kg/m³. Figure 8 indicates that the viscosity of the gas is almost 0.01 cp under the reservoir conditions. The natural gas is light and thin.

Fig. 5

Z-factors of natural gas

Fig. 6

Volume ratio in place ($V_{{{\text{r}} . {\text{scco}}_{ 2} }}$) of natural gas

Fig. 7

Phase diagram of natural gas with iso-density lines

Fig. 8

Phase diagram of natural gas with iso-viscosity curves

The necessity of ESEGR in PY gas reservoir

Figures 5 and 6 suggest that natural gas will take up the pore space of reservoir, reducing the potential space of SCCO₂ storage. Thus, effective carbon storage in a PY gas reservoir should be combined with the production of the gas. It is not only good for the stable sequestration of the SCCO_2, but also the repressurization caused by active CO₂ injection will enhance natural gas recovery.

SCCO₂ zone and the ideal injection site

Phase behavior of SCCO₂ zone

Shown in Fig. 9 are the Z-factors of pure CO₂. Figure 10 is the $V_{{{\text{r}} . {\text{scco}}_{ 2} }}$–p curves of pure CO₂. Figure 11 graphs the phase diagram for CO₂ with iso-density lines. Moreover, Fig. 12 graphs the phase diagram for CO₂ with iso-viscosity curves. Figure 9 indicates that the supercriticality of SCCO₂ is obvious for reservoir conditions. Figure 10 shows that the underground volume of SCCO₂ will self-contract quickly and then remain constant during an ongoing CO₂ injection. Figure 11 shows that the density of SCCO₂ will increase by 100 kg/m³ under a 1–2 MPa pressure increase if the temperature is near the critical temperature. Figure 12 indicates that the viscosity of SCCO₂ is at the level of the gases and higher than the natural gas viscosity.

Fig. 9

Experimental Z-factors of pure CO₂

Fig. 10

$V_{{{\text{r}} . {\text{scco}}_{ 2} }}$ of pure CO₂

Fig. 11

Phase diagram of CO₂ with iso-density lines

Fig. 12

Phase diagram of CO₂ with iso-viscosity curves

The ideal injected site

The ideal injection site of CSEGR must have the right depth with the right temperature to keep CO₂ in a supercritical state. Figures 9 and  10 suggest that a too high formation temperature should prevent the self-contraction of SCCO₂ for a maximum storage capacity in place. So deep gas reservoirs are not suitable for CSEGR. When 32 °C < T < 50 °C and 7.4 MPa < p < 20 MPa, Z-factor–p curves sag down acutely, and $V_{{{\text{r}} . {\text{scco}}_{ 2} }}$ quickly decreases to 1. The density (over 600 kg/m³, Fig. 11) is heavy enough to allow for CO₂ to migrate to the lower part of the reservoir. Based on such data, we believe 1000 m below (42 °C, 10 MPa; the relevant data are in Table 1.) the PY gas reservoir is available to both the effective SCCO₂ sequestration and CSEGR.

SCCO₂–natural gas transitional zone and the suitable pressures

The multi-contact mix during the gas migration makes the natural gas concentration (Cn.g) decrease successively from the displacement front to the storage front. The phase properties of SCCO₂–natural gas mixtures with different Cn.g can reveal the supercriticality of the SCCO₂–natural gas transitional zone.

Phase behavior of SCCO₂–natural gas transitional zone

Figure 13 graphs the Z-factor–p curves for the 4 Cn.g profiles (5, 30, 50 and 76.67 %) of the SCCO₂–natural gas transitional zone under the ideal SCCO₂ injection site temperature. Figure 14 shows the Z-factor–Cn.g curves. Figure 15 is the $V_{{{\text{r}} . {\text{scco}}_{ 2} }}$–p curves. Figure 16 shows the density–Cn.g curves under reservoir conditions. Moreover, Fig. 17 shows the viscosity–Cn.g curves under reservoir conditions. Figures 14, 15, 16 and 17 demonstrate that the diffusion of the gas will weaken the supercriticality of the SCCO₂–natural gas transitional zone. In addition, supercriticality decreases from the storage front to the displacement front, while the $V_{{{\text{r}} . {\text{scco}}_{ 2} }}$ increases. This demonstrates that the SCCO₂–natural gas transition zone is a “mechanical spring” in the natural gas zone, protecting the SCCO₂ storage space in the storage front and allowing for continuous CO₂ injection. At 10, 15 and 20 MPa, phase properties change faster than other pressures. It indicates that supercriticality is outstanding in such a pressure region.

Fig. 13

Z–p profiles of the SCCO₂–natural gas transitional zone

Fig. 14

Z–Cn.g curves of the SCCO₂–natural gas transitional zone

Fig. 15

The $V_{{{\text{r}} . {\text{scco}}_{ 2} }}$ –Cn.g curves of the SCCO₂–natural gas transitional zone

Fig. 16

Density–Cn.g curves of the SCCO₂–natural gas transitional zone

Fig. 17

Viscosity–Cn.g curves of the SCCO₂–natural gas transitional zone

The eligibly pressures

Assessing the feasible pressure of CSEGR in the field involves considering many controlling factors. However, we can obtain the eligible pressures for the next core-flooding test based on phase behavior research. This involves repressurization by continuous CO₂ injection to squeeze all fluid zones and the volume that the SCCO₂ zone can decrease to most quickly to maintain safe CO₂ storage. For ESEGR, the average $V_{{{\text{r}} . {\text{scco}}_{ 2} }}$ of the SCCO₂–natural gas transition zone and the displacement front should be greater than 1 for EGR. In addition, the $V_{{{\text{r}} . {\text{scco}}_{ 2} }}$ of the storage front should be less than 1 to protect the SCCO₂ zone and SCCO₂ storage.

We propose 3 schemes for CSEGR in the PY gas reservoir to estimate several eligible pressures. Scheme 1 is producing gas and keeping the ideal injection site pressure (10 MPa) by continuous SCCO₂ injection. Scheme 2 raises the formation pressure to 15 MPa. Moreover, Scheme 3 increases the formation pressure to 20 MPa. Table 2 lists the different $V_{{{\text{r}} . {\text{scco}}_{ 2} }}$ for the schemes’ SCCO₂–natural gas transitional zones. In Table 2, the average $V_{{{\text{r}} . {\text{scco}}_{ 2} }}$ of the transitional zone and displacement front of both scheme 1 and scheme 2 are greater than 1. This suggests that transitional zones under 10 and 15 MPa would benefit from EGR. However, the $V_{{{\text{r}} . {\text{scco}}_{ 2} }}$ of the storage front at these pressures is less than 1. This suggests that the transitional zones under 10 and 15 MPa would benefit from CO₂ storage in storage front and EGR in the displacement front. As the average $V_{{{\text{r}} . {\text{scco}}_{ 2} }}$ is less than 1 in scheme 3, it suggests that transitional zones under 20 MPa only benefit from CO₂ sequestration. Therefore, scheme 1 and scheme 2 are the eligible schemes for CSEGR in a PY reservoir.

Table 2 Volume ratio in place ($V_{{{\text{r}} . {\text{scco}}_{ 2} }}$) of the transitional zone at 10, 15 and 10 MPa

Conclusion

Natural gas will take up the pore space of a reservoir and reduce the potential space for SCCO₂ sequestration. Therefore, CSEGR is necessary if we conduct CO₂ sequestration in a PY gas reservoir. Multi-contact during the gas migration in CSEGR forms the SCCO₂–natural gas transitional zone. Thus, the formation fluid in the whole reservoir could be simply divided into three zones on the swept region. Such areas are the SCCO₂ zone, SCCO₂–natural gas transitional zone and natural gas zone. The PVT properties of the formation fluid may be summarized as follows:

An ideal CO₂ injection place is significant to the ideal gases migration during CSEGR. The ideal CO₂ injection site should have the right temperature to keep the gravitational differentiation between the SCCO₂ and natural gas large enough. In addition, the SCCO₂ zone should be at a stable volume for SCCO₂ storage. Thus, the SCCO₂ zone will stay in the lower part of reservoir. The natural gas zone will rise to the higher part of reservoir for gas production. The SCCO₂–natural gas transitional zone can separate the other two fluid zones into certain regions. Thousand meters beneath the PY gas reservoir is available to both effective SCCO₂ sequestration and CSEGR.

Repressurization by continuous CO₂ injection squeezes all of the fluid zones. The volume of the SCCO₂ zone can decrease quickly to maintain safe CO₂ sequestration under a suitable pressure for CSEGR. Thus, the SCCO₂–natural gas transition zone should be more useful to EGR under a suitable pressure for CSEGR than SCCO₂ under original formation conditions. In addition, the transition zone is a “mechanical spring” in the natural gas zone, protecting the SCCO₂ storage space in the storage front and allowing for the continuous CO₂ injection. Based on phase behavior research, the ideal injection site pressure (10 MPa) and 15 MPa pressure are the eligible pressures for CSEGR in a PY reservoir.