Asynchronous Injection–Production Method in the High Water Cut Stage of Tight Oil Reservoirs

Abstract

Asynchronous injection–production cycle (AIPC) in a horizontal–vertical well pattern is an efficient strategy for enhancing water injection in tight reservoirs. However, current studies lack consideration of waterflood-induced fractures (WIFs) caused by long-term water injection. This paper takes block Z in the Ordos Basin, China, as the research object and first clarifies the formation conditions of WIFs considering the horizontal principal stress and flow line. Then, the pressure-sensitive permeability equations for the induce-fracture region between wells are derived. Finally, the WIFs characteristics in a horizontal–vertical well network with different injection modes are discussed by numerical simulation. The results show that WIFs preferentially form where flow aligns with the maximum principal stress, influencing permeability distribution. Controlling the injection rate of vertical wells on the maximum principal stress and flow line and cyclically adjusting the production rate of horizontal wells can regulate the appropriate propagation of WIFs and expand the swept areas. The parallel injection mode (PIM) and the half-production injection mode are superior to the full-production injection mode. This study can provide theoretical support for the effective development of tight oil reservoirs.

1. Introduction

Tight reservoirs have extremely low porosity and permeability, requiring techniques such as hydraulic fracturing for economic extraction [1,2,3]. In China, tight oil reservoirs, exemplified by the Ordos Basin, typically have permeability lower than 1 × 10⁻³ μm² [4], and can even be as low as 0.01 × 10⁻³ μm². These reservoirs are characterized by small pore throats, intricate pore structures, and insufficient natural energy and usually have weakly laminated interfaces and natural fractures [5,6,7]. These reservoirs do not have the ability to undergo depletion development, and their exploitation mainly relies on artificial stimulation techniques such as hydraulic fracturing. In the early stages of tight reservoir development, natural fractures are not the main transport channels for fluids, and fracture features are rarely identified in the data [8,9]. However, as development progressed, production performance data revealed characteristics indicative of fracture influence [10,11,12]. The main reason for the aforementioned phenomenon is that the buildup of formation pressure due to poor water absorbing capacity of low-permeability reservoirs leads to the generation and continuous extension of waterflood-induced fractures (WIFs) [13]. Unlike natural or artificial fractures, WIFs constantly change in shape and properties over time, and are significantly influenced by the development conditions such as water injection pressure, injection/production ratio, and so on [14,15].

Given the constraints posed by conventional water injection methods, there is an immediate imperative to explore effective water injection techniques that are appropriate for fractured formations [16]. These methods should consider the dynamic propagation of waterflood-induced fractures (WIFs) and mitigate issues such as one-way water sighting and flooding of highly water-cut wells [17,18]. The current state of fractured tight reservoir development is analyzed from two aspects: well pattern types and water injection strategies. Common well patterns for tight reservoirs include five-spot and seven-spot configurations, with the inverted five-spot pattern showing better performance [19]. The enhanced recovery observed in the mixed well pattern with vertical water-inject wells and horizontal oil-produce wells [20] is attributable to the greater contact area of the horizontal well [21,22]. The five-spot well pattern in the mixed well type has a higher water drive ripple coefficient [23], indicating better water drive efficiency. However, the above studies focus on the injection–production relationship and do not consider the phenomenon where water injection exacerbates reservoir heterogeneity, thus affecting development effectiveness. Zhang et al. [24] incorporated fracture parameters and stress field orientation, demonstrating the dynamic characteristics of water-induced fractures in a numerical simulation of diamond-shaped vertical well patterns. The simulated WIF propagation is constrained due to focusing only on the principal stress direction. The omission of fluid flow direction consideration restricts its applicability in optimizing various well patterns.

To address both energy supplementation and water control requirements, unstable water injection methods, such as huff-and-puff [25] and asynchronous injection [26], are applied to tight reservoirs. He [27] carried out asynchronous injection and recovery tests for multi-stage and multi-direction water-seeing horizontal wells. The method of “no production during injection and no injection during production” was adopted, and numerical simulation was combined with several rounds of field tests to optimize different injection and extraction parameters. The primary oil recovery mechanisms in AIPC technology are pressure perturbation and flow line deviation [28]. Oil saturation around the injection wells decreases more rapidly compared to cyclic water injection [29]. AIPC has proven effective in practical applications [30,31], including gas injection and other scenarios [32,33]. Besides, some studies provided improved AIPC to simulate inter-fracture injection–production in horizontal fractured wells [34,35]. However, it does not account for the potential natural fracture opening induced by pressure disturbances, nor do the fractures evolve dynamically throughout the simulation.

In this paper, we focus on a horizontal–vertical well pattern in the Z block that is in a high water-cut stage. To determine whether an appropriate asynchronous water injection method can constrain water-induced fractures and achieve higher oil recovery rates, we have developed a model and optimized the parameters of AIPC. Firstly, the production data and injection indication curves are used to identify whether induced fractures have developed in the reservoir. Then, we calculated the pressure threshold of WIF, which was updated during the model simulation. Finally, the asynchronous injection and extraction parameters of the horizontal well–straight well network are optimized by numerical simulation. AIPC is applied to the tight oil field with a horizontal–vertical well network to control the proper opening of waterflood-induced fractures, which improves the sweep efficiency and provides guidance for further tapping of remaining oil.

2. Methodology

The two mechanisms of water-induced fracture formation are the activation of NFs and the formation of new fractures. This study fully considers activation pressure, fracture pressure, and flow line relationships, providing a formula for induced fracture formation. Three assumptions are made for the method in this study:

• Fracture propagation behavior excludes the influence of chemical, biological, and other external factors.
• The two mechanisms can occur independently or simultaneously.
• In each simulation round, the principal stress direction and magnitude remain constant with fixed activation pressures and fracture pressures. Parameters can be adjusted for subsequent simulations if needed.

Figure 1 shows the flow of the process.

2.1. Waterflood-Induced Fracture

WIFs are dominant, highly permeable seepage channels centered on water wells that form in tight reservoirs after long-term water injection. The extension orientation of the WIF generally coincides with the direction of the main seepage fracture or the direction of the maximum horizontal principal stress. Consequently, the rapid flow of injected water along these fractures is prone to swiftly inundating wells positioned along the fracture direction. However, the sweep area on the two sides of the main streamline is small, resulting in low oil displacement efficiency [36].

WIFs are formed dynamically during water injection. In the tight oil reservoir with a relatively steep gradient of drive pressure, pre-existing natural fractures in the formation open when the water injection pressure exceeds the formation pressure [37]. This leads to a rapid increase in water content and liquid production. These newly opened natural fractures enhance the formation’s permeability. When the injection pressure has not reached formation fracture pressure, the production wells maintain stable production, and the water content and fluid production remain relatively stable. With the increase in water injection time, formation pressure gradually increases until it reaches the formation fracture pressure. At this point, fractures in the formation propagate and connect, forming a high-permeability channel, and the water content rises rapidly [38]. The formation of an interconnected high-permeability zone stabilizes the water content rate at a high value, leading to a significant reduction in the effects of sweep area.

The mechanism of WIFs is usually considered to be the activation of natural fractures and the formation of new fractures. The pressure conditions for formation are water injection pressure higher than natural fracture opening pressure (p_w > p_i), water injection pressure higher than closure pressure of hydraulic fractures (HFs) (p_w > p_c), or water injection pressure higher than formation fracture pressure (p_w > p_f) [39]. When the maximum horizontal principal stress and the direction of the fracture are consistent, the fracture closing pressure and the natural fracture opening pressure are approximately the same.

2.1.1. Single Fracture

(1) Natural fracture activation. When the water injection pressure exceeds the opening pressure of HFs, the HFs will open and propagate and eventually form large-scale WIFs [40,41]. The open pressure of a natural fracture is as follows:

$$p_{\mathrm{i}}={\displaystyle \frac{\nu }{1-\nu }}{p}_{\mathrm{o}}sin\theta +p_{\mathrm{o}}\cos \theta -p_{\mathrm{p}}+{\sigma }_{\mathrm{H}}sin\beta +{\sigma }_{\mathrm{h}}\cos \beta$$

(1)

where p_i is the open pressure of naturally fractures, ν is the Poisson ratio, p_o is overburden pressure, p_p is pore pressure, σ_H and σ_h are the maximum and minimum horizontal stress, respectively, θ is fracture dip angle, and β is the angle of maximum horizontal stress and fracture trend.

According to Equation (1), the opening sequence of cracks at different inclination angles is different. As the angle between the direction of flow and the maximum horizontal principal stress decreases, the opening pressure of natural cracks decreases. Directional enhancement of water absorption occurs around the well after the natural fracture opens, and the injected water advances along the natural fracture in the direction of the maximum principal stress.

(2) New fracture formation: When the pressure is higher than the crack pressure, a new crack is formed. The fracture pressure of the reservoir is calculated by the Hubbert-Willis empirical equation [42]:

$$p_{\mathrm{f}}=3{\sigma }_{\mathrm{h}}-{\sigma }_{\mathrm{H}}+{\sigma }_{\mathrm{f}}-p_{\mathrm{o}}$$

(2)

where p_f is the fracture pressure and σ_f is tensile strength.

When neither natural nor artificial fractures are developed in the formation around an injection well and water is continuously injected with an injection pressure close to or above the formation rupture pressure, an induced seam along the maximum horizontal principal stress will form from around the injection well.

2.1.2. Fracture Area

The formation conditions of a single fracture can be determined by Equations (1) and (2), while the water injection-induced fracture is a high-permeability area where multiple fractures communicate with each other. In this paper, different fracture formation mechanisms are combined to represent the calculation formula of regionally induced fracture pressure threshold and equivalent permeability in a unified form.

We define the fracture state index for area (A_f). When there are no natural fractures and no hydraulic fractures, the A_f is 0. The closer A_f is to 1, the higher the development degree of natural fractures or hydraulic fractures. The pressure threshold function is as follows:

$$p_{\mathrm{s}}=(A_{\mathrm{f}}{p}_{\mathrm{i}}+(1-A_{\mathrm{f}})p_{\mathrm{f}})e^{1-\cos \gamma }$$

(3)

where γ is the angle between the injection–production connection line (main flow line) and the maximum horizontal principal stress. In formations with the same fracture development status, WIF is more likely to form between injection and production wells.

WIFs are a highly permeable fracture zone connecting oil and water wells.

$$K_{\mathrm{f}}(p)=\left\{\begin{array}{cc}{K}_{\mathrm{i}}& p<p_{\mathrm{s}}\\ c_1{K}_{\mathrm{i}}{e}^{c2(p-p_{\mathrm{s}})}& p\ge p_{\mathrm{s}}\end{array}\right.$$

(4)

where K_f is the permeability of the WIF-influenced area, K_i is the initial permeability, p is pressure, and c₁ and c₂ are the coefficients that adjust the pressure-sensitive relationship.

The process of WIFs is influenced by both geological and engineering factors, and its direction is related to the maximum horizontal stress and the direction of injection streamlines. The smaller the angle between the injection and extraction well connecting line and the direction of the maximum horizontal principal stress, the lower the threshold of injection pressure for multi-directional WIF opening.

2.2. Asynchronous Cyclic Waterflooding for the Horizontal–Vertical Well Pattern

Horizontal well–straight well asynchronous injection and recovery development methods can change the direction of injected water seepage, play the role of seepage suction and differential pressure unblocking in low-permeability reservoirs, improve the efficiency of oil repulsion, and reduce the water content rate [43].

Cyclic water injection offers an enhanced displacement effect by alternating between periods of high-intensity injection and shut-in well percolation [44,45]. Figure 2a depicts a mixed five-point well unit with four straight wells and a horizontal well. Conventional continuous injection creates areas between injection wells where residual oil is difficult to recover, such as the yellow portion of Figure 2a. Figure 2b represents 1/4 of the unit. The background figure illustrates the water-driven ripple area during conventional continuous injection. The red solid line indicates the displacement front for each of the three cycles, and the blue dashed line marks the change in displacement front during the shut-in periods in the current cycles. Compared to conventional continuous water injection, cyclic water injection can delay the onset of water breakthrough and increase the waterflood-swept volume. This water injection method avoids the occurrence of a single dominant channel and ultimately improves the displacement efficiency of the reservoir.

The oil enhancement mechanism of asynchronous cycle water injection in a 5-point horizontal–vertical well pattern is similar to the cycle water injection described. In each cycle, by alternately turning on and off the wells in each cycle, the dead area between injection wells is reduced, which would be caused by continuous injection. Asynchronous cyclic waterflooding can be divided into the following four stages.

Stage 1—Water injection: The two straight wells near the toe end of the horizontal wells are opened and the two straight wells near the heel end of the horizontal wells are shut down. The horizontal production wells are shut down. Injected water enters the low-permeability reservoir along the microfractures. Because the external pressure of the injected water in the crack is much greater than the external extrusion force of the matrix pores, the seepage on the injection side is strengthened, the remaining oil in the matrix pores is displaced, and the flow distance of the injected water is expanded.

Stage 2—Oil production: All four injection wells are closed and production commences in half of the horizontal sections. The unidirectional valves associated with the hydraulic fractures in the horizontal sections proximal to the toe end are closed while the unidirectional valves in the hydraulic fractures of the horizontal sections near the heel end are opened. This stage enhances seepage on one side of the horizontal well, so that residual oil can flow from the matrix pore into the horizontal wellbore through the microfractures.

Stage 3—Water injection: Unlike Stage 1, the strategy for water injection wells is completely reversed. The two straight wells near the toe end of the horizontal wells are shut down and the two straight wells near the heel end of the horizontal wells are opened. The horizontal well is completely closed. This stage allows the residual oil to flow on the other side of the horizontal well. The alternate injection of water injection wells can avoid the possible flow line disturbance between injection wells and reduce the dead oil area.

Stage 4—Oil production: All four injection wells are closed and production commences in half of the horizontal sections. The open horizontal section and closed horizontal section of the unidirectional valve are opposite to stage 2. At this stage, the seepage is strengthened from the toe-end side to the low-permeability direction to drive out the residual oil in the pore space of the matrix in the low-permeability direction.

Repeating the above four-stage working system, the asynchronous injection and production method changes the direction of seepage of injected water, inhibits the flow of high-permeability hydraulic fracture, promotes the flow in the low-permeability direction, and drives out the crude oil endowed in the channels with lower pore permeability. At the stage of well closure, the injected water invades the matrix system along the small pore throats under imbibition effect and drives out the remaining oil along the large pore space, which improves the oil-driving efficiency. Periodic alteration of the seepage field in the reservoir causes an unstable pressure drop in the formation, which provides power to the crude oil attached to the surface of the rock particles and promotes the flow of crude oil through the multi-scale pore structures of the formation, and finally flows into the wellbore through the fractures to be extracted.

3. Results and Discussion

3.1. Reservoir Model

Block Z of the Heshui area is located in the southwestern part of the Ordos Basin, NW China, and the main reservoir is Chang6 member of the Triassic Yanchang Formation (Figure 3a). The study area develops natural fractures with a NE 70° direction. The maximum principal stress is NEE-SWW [46]. The main cause of WIFs in this area is that the water injection pressure is higher than the fracture opening pressure for a long time. Referring to the Chang 6 rupture pressure gradient of 0.022~0.023 MPa/m in other oil fields, the fracture pressure in this area is calculated to be at 35.2 MPa. The theoretical fracture pressure calculated from Equation (2) is 37 MPa. By injecting water for a long time at close to the fracture pressure, the induced fracture of a small area may be formed around the compression fracture.

In this paper, we focus on the typical well pattern type in block Z, characterized by a mixed 5-point well pattern of vertical and horizontal wells. Two units within the study area, currently in the high water cut stage (Figure 3b), were selected and modeled with tnavigator. This simulation platform provides advanced tools for geoscience, reservoir, and production engineering. The basic information is as follows: the horizontal wells have a length of 700 m, the dimension of the established model is 2250 m × 1950 m × 53 m, permeability ranges from 0.01 × 10⁻³ μm² to 0.48 × 10⁻³ μm², porosity ranges from 0.18% to 12.42%, the total number of grids is 63,180, of which 39,663 are effective. The simulation parameters are detailed in

Table 1. Simulation parameters.

Parameter	Unit	Value
Maximum principal stress, σh	MPa	28
Minimum principal stress, σH	MPa	36
Tensile strength σf	MPa	5
Overburden pressure, po	MPa	21
Poisson ratio, ν	-	0.25
Pore pressure, pp	MPa	14.5
Angle of maximum horizontal stress and fracture, β	°	[0, 5]
Fracture state index for area, Af	-	[0, 1]
Initial permeability, Ki	10−3 μm2	0.36

.

3.2. Result Analysis

This study focused on well patterns in the high water cut stage. Referring to historical data, we initially simulated a 10-year period with an injection rate of 3 m³/d and a production rate of 6 m³/d. At the end of the simulation, the bottom hole pressure of the water injection well is 30.8~32.7 MPa, and the water saturation in wells P1-1 and P1-2 reached 91% and 96%, respectively. The bottom-hole pressure of the injection well and the water saturation of the production well exhibited deviations of less than 5% from the actual data, confirming the method’s validity.

The development of waterflood-induced fractures was then assessed. Figure 4a presents the field map of threshold pressure. The figures indicate that lower pressure levels are needed to form WIFs along the flow lines of the injection and production wells. Notably, the pressure required in the region between the two fractured horizontal wells is slightly higher than in the heel and toe sections of the horizontal wells. This is attributed to the fluid’s preferential flow along the hydraulic fracture, leading to limited fluid washout in this area. Consequently, the threshold pressure is higher here than in other fractured areas but lower than in less fractured areas. Using Equation (4), we obtained the distribution of high-permeability channels at this stage (Figure 4b). These high-permeability strips exhibit directionality consistent with the maximum principal stress and the orientation of hydraulic fractures.

3.2.1. Asynchronous Mode of the Full-Production-Injection Well

Different asynchronous modes affect the flow field in various ways. The asynchronous modes of the five-point horizontal–vertical well pattern can be classified into parallel injection mode (PIM) and diagonal injection mode (DIM). In the PIM, water invades from one side of the horizontal well; in the DIM, water invades from both corner sides of the horizontal well. Each cycle consists of four stages: injection–production–injection–production, with different wells working during the two injection stages. Based on these asynchronous modes, we categorize the various classes of injection wells into Group 1 and Group 2, as shown in

Table 2. Injection wells groups in different modes.

Mode	Group 1	Group 2
PIM	I1-1, I1-2, I1-3	I2-1, I2-2, I2-3
DIM	I1-1, I2-2, I1-3	I2-1, I1-2, I2-3

.

The injection duration (D_i) was fixed, while only the production duration (D_p) was varied to explore the effect of different ratios of injection–production duration (r_ipd) on the performance of enhanced oil recovery. The cumulative oil production was simulated for six r_ipd values (0.25, 0.5, 0.75, 1, 1.25, 1.5). Meanwhile, the injection/production ratio was maintained at 1:1, and the total injection volume remained constant. In these scenarios, we simulated 20 cycles. The cumulative oil production for the different cases is shown in Figure 5. With the increase in the injection time interval, the cumulative oil production first increased and then decreased.

AIPC objectively alters the flow direction of injected water, effectively preventing it from channeling along fractures towards the production well. For both methods, a too high r_ipd is detrimental to recovery. Due to the reservoir’s high density, pressure and fluid propagation are slow. Long injection cycles tend to maintain high pressure at the bottom of the injection well leading to the formation of induced fractures. In addition, short production cycles with high recovery rates hinder fluid supply and exacerbate reservoir heterogeneity. After multiple cycles, the WIFs and hydraulic fractures facilitate flow communication, making it difficult to utilize the residual oil in the matrix.

The cumulative oil production increases and then decreases as the injection interval increases. PIM has the best program effect at r_ipd = 0.75, and DIM has the best program effect at r_ipd = 1. Overall, PIM is more effective than DIM. Due to the maximum horizontal principal stress direction being NE 72°, diagonal injection leads to a significant pressure increase in the direction of the maximum principal stress, resulting in the formation of seepage channels. Although this enhances the water absorption capacity of the injection wells, the swept area remains small, leading to low cumulative oil production.

The pressure disturbances caused by asynchronous injection–production promote the penetration of injected water into deeper pore throats within the matrix. Figure 6 illustrates the bottomhole pressures in the production wells, showing that P1-2 is significantly more effective at recharging than P1-1. Analyzing the well network and fracturing patterns, P1-2 is situated lower compared to P1-1, and the fractures and injection wells (I2-1 and I1-2) are closer together. This proximity gives P1-2 an advantage in pressure retention. However, it also increases the likelihood of water breakthrough. The DIM exacerbates this phenomenon. Figure 7 shows a field map of the high permeability region at five cycle intervals with PIM. The high permeability region around well P1-2 is much more connected to the hydraulic fracture.

In this section, it is shown that the best production performance can be achieved by using the PIM, considering the principal stress direction and natural fracture distribution in the region. The optimal asynchronous mode and r_ipd are parallel injecting mode and 0.75, respectively.

3.2.2. Asynchronous Mode of the Half-Production-Injection Well

Horizontal wells with part of the well section open allow for a greater flow line overlap. This section simulates the asynchronous mode of the half-production-injection well. Each cycle consists of four stages of injection–production–injection–production, where the grouping of wells working in the two injection stages is the same as in the previous series of simulations. The two production stages open only half of the horizontal wells near the heel or near the toe, respectively.

Figure 8 shows the cumulative oil production at the end of 20 cycles for different cases. Comparing the simulation results of the asynchronous mode of the production-injection well, the cumulative oil production of the half-production-injection well is higher under each r_ipd. Since the horizontal well section opened in the production stage is the well section in the opposite direction of the previous injection stage, the swept area of the injected fluid is larger compared to the asynchronous mode of production-injection wells. The oil between the injection wells is driven closer to the production wells, ultimately increasing cumulative oil production.

The present scheme also demonstrates superior performance in terms of pressure maintenance. As illustrated in Figure 9, there is an increase in bottomhole pressure of approximately 2 MPa per well, indicating more effective pressure maintenance. The pressure at P1-1 is essentially the same for both injection modes. However, the DIM simulation results in a slightly higher bottomhole pressure for P1-2 compared to PIM, aligning with previous simulation findings. Notably, the optimal ripd for DIM is 0.5, whereas for PIM it is 0.75. The AIPC for half-production-injection well shortens the length of individual cycles, achieving a higher recovery degree within a shorter simulation duration.

Figure 10 shows a field map of the high permeability region at five cycle intervals with PIM. In contrast to Figure 7, the morphology of the high permeability zones with the half-production–injection well method between the fractured joints of the production wells is more diffuse and does not communicate directly with the injection wells. This avoids directional seepage of injected water and controls the water content of the production wells.

4. Conclusions

In this paper, we consider the WIFs caused by long-term water injection, take the high water cut, typical horizontal–vertical well pattern unit of Block Z as an example and propose to control the proper openings of WIFs by the asynchronous injection–production cycle method. This method provides an effective adjustment method for the development of fractured tight reservoirs at the late stage of development, improves the sweeping efficiency, and guides the further exploitation of the remaining oil.

• The use of a 5-point horizontal–vertical well pattern enhances recovery through an asynchronous injection–production method. The main recovery mechanism involves periodically altering the pressure distribution to mobilize residual oil in the matrix through pressure perturbation and changes in flow lines, thereby expanding the swept area of the injected water.
• WIFs are more likely to form where the angle between the flow line and the maximum horizontal principal stress is small. In the full-production injection mode, high permeability regions are concentrated between the injection wells and the fractures. In the half-production injection mode, high permeability regions are more dispersed and appear between the fractures.
• Cumulative oil production initially increases and then decreases with the rise in the ratios of injection–production duration (r_ipd). The optimal r_ipd ratios were 0.75 and 0.5 for the parallel injection mode in the full-production injection case and half-production injection case, respectively, and 1 and 0.75 for the diagonal injection mode, respectively.
• The parallel mode (r_ipd = 0.5) of the half-production injection well fully utilized the imbibition mechanism and maximized oil recovery. It limited the expansion of WIFs and prevented water breakthroughs.
• This method is suitable for reservoirs where WIFs have already been identified; otherwise, the lack of reference pressure-sensitive permeability may reduce the accuracy of predictions.