The recovery factor from tight gas reservoirs is typically less than 15%, even with multistage hydraulic fracturing stimulation. Such low recovery is exacerbated in tight gas condensate reservoirs, where the depletion of gas leaves the valuable condensate behind. In this paper, three enhanced gas recovery (EGR) methods including produced gas injection, CO2 injection and water injection are investigated to increase the well productivity for a tight gas condensate reservoir in the Montney Formation, Canada. The production performance of the three EGR methods is compared and their economic feasibility is evaluated. Chemical and Petroleum Engineering Department, University of Calgary, 2500 University Drive NW, Calgary, AB T2N1N4, Canada
Min Wang, Shengnan Chen, Menglu Lin
Edited by Yan-Hua Sun
Received: 23 September 2017 / Published online: 27 March 2018
© The Author(s) 2018
The recovery factor from tight gas reservoirs is typically less than 15%, even with multistage hydraulic fracturing stimulation. Such low recovery is exacerbated in tight gas condensate reservoirs, where the depletion of gas leaves the valuable condensate behind. In this paper, three enhanced gas recovery (EGR) methods including produced gas injection, CO2 injection and water injection are investigated to increase the well productivity for a tight gas condensate reservoir in the Montney Formation, Canada. The production performance of the three EGR methods is compared and their economic feasibility is evaluated.
Chemical and Petroleum Engineering Department, University of Calgary, 2500 University Drive NW, Calgary, AB T2N1N4, Canada
Sensitivity analysis of the key factors such as primary production duration, bottom-hole pressures, and fracture conductivity is conducted and their effects on the well production performance are analyzed. Results show that, compared with the simple depletion method, both the cumulative gas and condensate production increase with fluids injected. Produced gas injection leads to both a higher gas and condensate production compared with those of the CO2 injection, while waterflooding suffers from injection difficulty and the corresponding low sweep efficiency. Meanwhile, the injection cost is lower for the produced gas injection due to the on-site available gas source and minimal transport costs, gaining more economic benefits than the other EGR methods.
The successful application of horizontal drilling and multistage hydraulic fracturing technologies has boosted oil and gas production from tight reservoirs in the last decade. Although commercial development is enabled by the advanced technologies, estimated primary recovery factors remain to be as low as 5%–15%, owing to the ultra-low permeability (Hoffman 2012). Currently, a liquid-rich tight gas reservoir (e.g., Montney Formation) has attracted interest (Cui et al. 2013; Rivard et al. 2014). In a gas condensate reservoir, reservoir fluids appear as gas phase under initial conditions.
With the depressurization of the reservoir during the primary production, liquid condenses from the gas phase and builds up once the in situ reservoir pressure drops below the dew-point pressure, especially around the fractures and the well bottom hole. The condensate liquid will not flow until a critical condensate saturation is achieved. It is generally accepted that three zones are present in the formation from the wellbore to the reservoir boundary: (1) mobile gas and mobile condensate region near the wellbore, (2) transition zone including mobile gas and immobile oil, and (3) gas phase zone without condensate dropout (Penuela and Civan 2000).
The trapped condensate typically cannot flow, leaving a large amount of high-quality oil unproduced in the reservoir. Nevertheless, gas production will decrease as the presence of condensate restricts the gas flow toward the wellbore (Hinchman and Barree 1985; Moses and Donohoe 1987; Vo et al. 1989; Li and Firoozabadi 2000; Pope et al. 2000). Meanwhile, the condensate blockage problem is exacerbated significantly by the ultra-low reservoir permeability in a tight gas condensate reservoir, and the gas production rate could decrease by 50%–80% within the first 2 years (Ayyalasomayajula et al. 2005).
Thus, it is important to investigate the potential performance of the EGR methods to alleviate condensate blockage and further increase the recovery factor in tight gas condensate reservoirs. Extensive study has been performed to maintain the reservoir pressure above the dew-point pressure in conventional gas condensate reservoirs, using techniques such as water injection (Matthews et al. 1988), lean gas injection (Smith and Yarborough 1968; Abel et al. 1970; Sigmund and Cameron 1977; Abasov et al. 2000), CO2 injection (Goricnik et al. 1995; Narinesingh and Alexander 2014) and N2 injection (Aziz 1983; Abdulwahab and Belhaj 2010; Sadooni and Zonnouri 2015).
Usually, two schemes of fluid injection are employed for pressure maintenance in a condensate reservoir. One is full pressure maintenance where the fluid is continuously injected into the reservoir, while at the same time, the condensate is produced. The other is the partial pressure maintenance where gas is injected into the reservoir after the primary depletion to slow pressure decline and re-vaporize the condensate (Abel et al. 1970; Meng and Sheng 2016). Currently, researchers mainly rely on laboratory studies or reservoir simulations to investigate the performance of EGR methods in unconventional tight reservoirs due to the lack of field test data.
Yu et al. (2014) studied the efficiency of CO2 injection to enhance the gas recovery in a shale gas reservoir, considering the adsorption of CO2 in the shales with a high total organic content. An experimental design method was employed to search for the best operational scenario for the CO2 injection. Sheng (2015a, b) investigated the huff-n-puff performance of the produced gas in a shale gas condensate reservoir via a simplified simulation model containing only one fracture stage. They concluded that huff-n-puff methane injection is an effective option to enhance the gas and oil recovery for a shale gas reservoir with a permeability of 100 nD (i.e., 0.0001 mD).
Haghshenas et al. (2017) simulated CO2 huff-n-puff in a liquid-rich Canadian unconventional reservoir accounting for the fluid adsorption and the compositional heterogeneity. CO2 huff-n-puff results were only positive for certain operating conditions, and additional sensitivity study to EGR operations was needed. The mechanism and feasibility of the EGR is still not very clear, and it is necessary to investigate the key factors affecting the effectiveness in the tight gas condensate reservoir.
In this paper, we focus on evaluating different EGR methods after the reservoir has been depleted for several years and the flowing bottom-hole pressure of producers remains below the dew-point pressure. A sensitivity study was further conducted on the key factors that affect the performance of EGR methods. More specifically, a geological model which contains 27 horizontal wells was built and a sub-model containing 3 wells was cut out for the reservoir simulation, each well containing nearly 30 stages of hydraulic fractures.
Three EGR methods including produced gas flooding, CO2 flooding, and waterflooding were then applied and their performances were evaluated. Sensitivity studies of the key operational and geological parameters were conducted to investigate their effects on the gas and condensate production. Economic feasibility of EGR methods was also analyzed. This work can advance the understanding of the mechanisms of enhancing recovery in unconventional tight condensate gas reservoirs and provide a reference for the future EGR field application in the Montney Formation.
2 Geological model
The target reservoir, located in the Montney play, is situated at the border of Alberta and British Columbia, in the Western Canadian Sedimentary Basin, Canada. The Montney Formation is composed of two production zones: the upper Montney and the lower Montney, containing 449 trillion cubic feet of marketable natural gas, 14,521 million barrels of marketable natural gas liquids and 1125 million barrels of oil, as estimated by Canada’s National Energy Board (NEB 2013). Multistage hydraulic fractures placed along a horizontal well are the main completion method in the Montney Formation to achieve commercial well production rates (Kuppe et al. 2012).
Fig. 1 Well locations in Accumap (numbers in black indicating the formation thickness at the well location) (a) and the geological model (b)
In this study, a geologic model, covering an area of 34,000 m long and 18,000 m wide, is built for a liquid-rich gas play in the Montney Formation. Reservoir depth ranges from 2800 to 3500 m with a thickness of 200 m. Twenty-seven horizontal wells have been drilled and fractured in the simulated area, and their locations are shown in Fig. 1a, and Fig. 1b shows their positions in the geological model. Reservoir properties such as matrix permeability, porosity and water saturation are derived using laboratory measurements and well-logging data (Ghanizadeh et al. 2014). The reservoir properties such as permeability and porosity are listed in Table 1.
Table 1 Reservoir model properties
3 Reservoir simulation model
3.1 Model description
The geological model was upscaled, and a section of the model containing three horizontal wells was selected and history matched for the reservoir simulation studies. All three wells are hydraulically fractured. Well 1 and Well 3 are both 3000 meters long with 31 stages, while Well 2, which is located in the middle, is 2100 meters with 27 stages. Properties of the hydraulic fractures are shown in Table 2, and the perforation type is open hole. The reservoir model dimension is 1050 m wide with 21 grids in the I direction, 3800 m long with 76 grids in the J direction and 60 m thick with 7 grids in the K direction.
Table 2 Properties of hydraulic fractures
Porosity and permeability distributions of the numerical model are shown in Fig. 2. Local refining grids were generated to represent the hydraulic fractures in the reservoir model. Relative permeability curves of the reservoir matrix are shown in Fig. 3 (Lan et al. 2015). For the hydraulic fractures, the relative permeability curves are assumed to be two straight lines. For the multiphase fluid flow in the reservoir, the three phase permeabilites are calculated by Stone’s second model (CMG 2016).
Fig. 2 Permeability and porosity variation of the simulation model. a Permeability distribution. b Porosity distribution
Fig. 3 Relative permeability curves for the Montney Formation. a Oil and water relative permeability curves (krw: water relative permeability; krow: oil relative permeability) b Liquid and gas relative permeability curves (krg: gas relative permeability; krog: oil relative permeatility)
3.2 Reservoir fluid properties
The area of interest is located in a gas condensate zone. Figure 4 depicts the calculated phase envelope of the recombined fluid at a production gas–oil ratio of 1200 m3/m3. It can be seen that the dew-point temperature is 64 °C and the dew-point pressure is 23.2 MPa. Reservoir conditions (98 °C, 30.5 MPa) belong to the retrograde condensation area of the generated phase envelope, as seen in the figure.
Fig. 4 Phase behavior diagram of the reservoir fluids
Reservoir pressure decreases as the well production proceeds, while reservoir temperature keeps constant. When the pressure drops below the dew-point pressure, the liquid condensate begins to condense from the gas phase and remains immobile till its saturation reaches a critical value. The newly formed liquid will not only reduce the amount of condensate (i.e., oil) production at the wellhead but also block the gas from flowing toward the wellbore, which may reduce the gas production rate at the same time. Thus, it is essential to maintain average reservoir pressure above the dew-point pressure and slow further pressure decline when developing gas condensate reservoirs.
3.3 History matching studies
History matching was performed to further tune the reservoir simulation model to better represent the formation rock and fluid properties. In this model, the bottom-hole pressures of the producers were applied as constraints, while the gas and condensate production rates were matched. Reasonable history matching results were achieved for all three wells, and Fig. 5 depicts the history matching results for Well 3. As seen, a production history of 450 days has been history matched, and the tuned model was reliable for reservoir simulations and production predictions.
Fig. 5 History matching results for Well 3. a Gas rates. b Condensate (oil) rates
4 Results and discussion
Figure 6 depicts the schematic diagram demonstrating the three hydraulic fractured horizontal wells that are distributed in the simulation model. The well spacing is 300 m, the fracture spacing is 80 m, and the half-length of the hydraulic fracture is 125 m. Primary production continues for about 5 years (from Day 450 to Day 2200) and results suggest that the average reservoir pressure drops to 22.9 MPa, which is slightly lower than the dew-point pressure at 23.2 MPa (See Fig. 8). The aforementioned three EGR methods are then applied on Day 2200 to prevent a large amount of liquid being condensed from the gas phase.
Fig. 6 Schematic diagram of the three fractured horizontal wells and the simulation area
Only primary production is applied in the base case. Scenario 1 represents the produced gas flooding scenario, where Well 2 is converted to a produced gas injector, while Well 1 and Well 3 remain producers after 5 years of primary production. It should be noted that Well 2 is converted back to a producer after injecting produced gas for ten years. Scenarios 2 and 3 are the CO2 flooding scenario and the waterflooding scenario. Similarly, Well 2 is converted to a CO2 injector or water injector, while Well 1 and Well 3 still remain producers after 5 years of depletion. The water injection scenario is included in the simulations only for the comparison with gas injection scenarios.
4.1 Reservoir pressure
As mentioned above, pressure maintenance is essential for a gas condensate reservoir. The flooding characteristics in tight reservoirs are different from those in the conventional reservoirs. Figure 7 shows the pressure distribution of the CO2 flooding scenario after 1 month and 1 year’s injection, respectively. We can observe that the injection gas first flows in the hydraulic fractures and then penetrates into the surrounding matrix under the high injection pressure. High pressure is still mainly limited in the areas near the hydraulic fractures after 1 year injection due to the low matrix permeability.
Fig. 7 Pressure distribution for the CO2 injection scenario (unit kPa)
The reservoir pressure for each scenario following the primary production is also depicted in Fig. 8. As seen, the average reservoir pressure keeps decreasing in the base case where no fluid injection is applied. The average reservoir pressures for the produced gas injection and CO2 injection scenarios are both above the dew-point pressure while that of the waterflooding scenario slightly increases, but fails to stay above the dew-point pressure. This is because the ultra-low permeability of the reservoir matrix (0.004–0.009 mD) restricts the water penetration into the formation rocks, leading to a low water sweep efficiency.
Fig. 8 Impact of the fluid injection on the average reservoir pressure
In addition, the higher sweep efficiency of the gas injection can also be attributed to the following aspects: (1) the condensate oil swells and its viscosity decreases due to the gas dissolution in oil; and (2) the interfacial tension could be reduced or eliminated if a miscible condition is reached. However, due to the low permeability, the high pressure area only remains near injectors and pressure around producers is still low, which lowers the positive effect of viscosity and interfacial tension reduction. In the produced gas injection scenario, much injected gas accumulates near the injector during the flooding process. In order to recover the large amount of injected gas for better revenue, Well 2 is converted back to a producer after 10 years of produced gas flooding. The reservoir pressure then drops significantly.
4.2 Production performance of EGR methods
The injection pressure for the three scenarios is set the same at 45 MPa. Figure 9 shows the cumulative gas and condensate production of the produced gas injection, CO2 injection and water injection for the target formation. It should be noted that for the produced gas injection scenario the amount of produced gas that is injected into the formation needs to be subtracted from the total gas production in order to calculate the net natural gas production, which is shown in Table 3. It can be seen from Fig. 9a and Table 3 that the base case leads to the highest gas production, followed by CO2 injection, while the produced gas injection and water injection scenarios share a low cumulative gas production.
Fig. 9 Cumulative gas and condensate production of the three EGR methods. a Cumulative gas production measured under standard condition (sm3). b Cumulative condensate (oil) production
However, the produced gas injection and CO2 injection display a significant increase in the cumulative condensate production. The cumulative condensate production of the produced gas injection is 52.7% higher than that of the base case and CO2 injection indicates a 40.0% improvement in cumulative condensate production (see Fig. 9b and Table 3). Although its gas production is reduced, the water injection scenario also demonstrates a slightly higher cumulative condensate production than those of the base case scenario. This is because the injected water reduces the relative permeability of the gas phase and thus decreases the gas ability to flow to the wellbore. However, the reservoir pressure increases, preventing liquid being condensed in the reservoir.
Table 3 Cumulative production for different enhanced/improved gas methods
As aforementioned, Well 2 is converted back into a producer and is put into production on Day 5850. It can be seen that a large amount of gas is produced during this stage shown as sharp increases in the gas production for the produced gas injection scenario in Fig. 9a. It is worth pointing out that the cumulative gas condensate production (i.e., cumulative oil production) only slightly increases during such process due to a low percentage of the heavy components in the injected gas near the injector.
The barrel of oil equivalent (BOE) is adopted to assess the gas and condensate productivity for the different scenarios. The BOE is an industrial unit of energy equivalent to the amount of energy released by burning one barrel of crude oil. The calculated BOE results are shown in Table 3. The produced gas injection displays the highest BOE amount, followed by the CO2 injection, base case and waterflooding, respectively. In addition, the BOE of the waterflooding is lower than that of the base case.
This is because the injected water has decreased the effective gas permeability in the formation, leading to a lower gas production rate. In other words, the increase in condensate production due to a higher reservoir pressure during waterflooding cannot compensate for the loss of gas production compared with the base case. The cumulative condensate production of the base case is the lowest among all scenarios, as the low reservoir pressure of the base case promotes condensate to be condensed and left unproduced in the reservoir.
4.3 Phase envelop change
The phase diagram changes during the produced gas injection and CO2 injection processes as a result of the compositional change of the reservoir fluids. Figure 10 demonstrates the new phase diagram with the production gas–oil ratio of 1500 with produced gas or CO2 injection. It can be seen that both the critical pressure and temperature decrease, and the two-phase region shifts to the left side, compared to the phase envelope shown in Fig. 4. Such changes will help prevent the oil condensation in the formation under the reservoir conditions and further increase the condensate production at the wellhead.
Fig. 10 Phase diagrams of the produced gas injection and CO2 injection scenarios. a P–T diagram with produced gas injection. b P–T diagram with CO2 injection