The experimental rock sample is from the core in the injection production wells of X UGS, and the reservoir is carbonate fracture-pore type. According to the analysis of logging and coring data, the porosity of the reservoir is 0.39%–16.96%, and the permeability is 0.001–223.69 mD. The core is processed into a column sample with a diameter of 2.5 cm and a length of 5 cm (Figure 1), testing porosity and permeability, select core samples similar to reservoir physical properties as experimental rock samples.

Permeability is measured by core damage evaluation instrument, porosity is measured by overburden porosity-permeability instrument, and microstructure is observed by scanning electron microscope. The core damage evaluation instrument mainly consists of injection system, temperature control system, pressure control system, back pressure system, metering system, data acquisition and control system, the injection system can adjust the pressure by 0–50 MPa, the temperature control system room temperature ∼150°C, the maximum confining pressure of the pressure control system is 90 MPa, and the accuracy of the metering system is 0.001 g (Figure 2A). The overburden porosity-permeability instrument is mainly composed of pressure regulating system, sampling system, automatic control system, data acquisition and analysis system, etc., the porosity measurement range: 0.01%–40%, the permeability measurement range: 0.00005–15,000 mD, the pressure sensor: 0–250 psi, 0–10,000 psi (Figure 2B).

For the pertinence and accuracy of experimental research, the experimental temperature 82 C (X UGS reservoir temperature), confining pressure 71 MPa (X UGS reservoir overlying formation pressure). Conduct triaxial rock mechanics experiments on experimental rock samples, under the experimental confining pressure of 71 MPa, the average compressive strength is 314.75 MPa, which means that under the experimental conditions of adding confining pressure, the experimental rock samples undergo elastic deformation when the pressure is less than 314.75 MPa; when the pressure reaches 314.75 MPa, plastic deformation occurs to breakage (Figure 3). The critical pressure for displacement deviation of the experimental rock sample is 314.75 MPa. Injection pressure is 13.20 MPa–28.06 MPa (X UGS operation of upper and lower limit pressure). The injection pressure is refined and decomposed in the experiment, the injection pressure with 3 MPa as the pressure interval from the lower limit 13 MPa to the upper limit 31 MPa, which are 13 MPa, 16 MPa, 19 MPa, 22 MPa, 25 MPa, 28 MPa, and 31 MPa. The effective stress is the difference between the confining pressure and the injection pressure, which are 58 MPa, 55 MPa, 52 MPa, 49 MPa, 46 MPa, 43 MPa, and 40 MPa (Table 1).

The experimental medium is 8% potassium chloride solution, in order to eliminate the effect of velocity sensitivity on the experimental results, the experimental displacement flow should be determined. The specific experimental steps are as follows: 1)Put the core sample into the core holder, increase the confining pressure, and let it stand for 4 h to eliminate the stress sensitivity; 2)Keep the confining pressure always higher than the inlet pressure 2 MPa, and test the permeability under the flow rates of 0.10 cm³/min, 0.25 cm³/min, 0.5 cm³/min, 0.75 cm³/min, 1.00 cm³/min, 1.5 cm³/min, and 2.0 cm³/min. When the permeability change rate is greater than 20%, it indicates that velocity sensitivity has occurred, and the flow corresponding previous point is the displacement flow. The permeability change rate can be expressed using Eq. (1). Permeability   change   rate : D w = K i − K w K i × 100 % (1) Where, D _w is the permeability change rate under different flow rates; K _i is the permeability of core under different flow rates, respectively (mD); K _w is the initial permeability of core, respectively (mD).

Test the displacement flow of different experimental samples, which shows that the flow rate when the permeability change rate is greater than 20% increases with the increase of initial permeability, but when the initial permeability of the sample is within a certain range, the flow rate when the permeability change rate is greater than 20% is a constant value (Table 2). The reservoir itself contains fine particles, particles migrate under the scouring of high-speed fluid, which may block the pore throat or accompany fluid discharge from the reservoir (Muecke, 1979; Gruesbeck and Collins, 1982; Civan, 2015; Li et al., 2017; Hou et al., 2015). When the initial permeability of the sample is within a certain range, its corresponding pore structure is consistent, and the fluid exhibits the same properties when passing through the pore. Therefore, different displacement flow rates are used for different experimental samples.

During the change of effective stress, test the permeability and porosity changes at different effective stress points. The experimental steps are as follows: 1)Put the core samples into the core holder of the overburden porosity-permeability instrument, test the porosity and permeability, select two core samples with similar porosity and permeability as a group, and select the cores representing the physical properties of different reservoirs as different groups; 2)A core sample of the same group is put into the core holder of the Core damage evaluation instrument, linearly load the confining pressure at a loading rate of 0.69 MPa/min, load the confining pressure to 71 MPa and keep it constant, and stabilize for 4 h; the displacement flow is used to pressurize to the injection pressure, maintain the pressure for 15 min, measure the permeability of this point; At the same time, another core sample of the same group is put into the core holder of the overburden porosity-permeability instrument, and the effective stress corresponding to the injection pressure is added as the confining pressure, maintain the pressure for 15 min, measure the pore of this point; 3)According to the injection pressure interval, increase the injection pressure in turn to the maximum, and test the permeability of core samples at different injection pressure points; The effective stress corresponding to the injection pressure is taken as the confining pressure to test the porosity of core samples at different effective stress points, this process simulates the gas injection process of UGS; 4)Change the displacement direction, according to the injection pressure interval, and then start from the maximum injection pressure, reduce the injection pressure to the minimum injection pressure by the displacement flow, and test the permeability of core samples at different injection pressure points; The effective stress corresponding to the injection pressure is taken as the confining pressure to test the porosity of core samples at different effective stress points, this process simulates the gas production process of UGS; 5)Repeat steps (2)–(4) to complete the experiment until to complete multi cycle injection and production; 6)Replace the core sample and repeat steps (2)–(5) until the experiment is completed. Finally, based on the change characteristics of permeability and porosity under different effective stress, define the effect of effective stress change on the injection production capacity of the reservoir.

There is a correlation between porosity, permeability with effective stress. The pore volume and permeability of rocks decrease with the increase of the cycle of pressure reciprocating changes (Ranjith et al., 2013; Song et al., 2022). The natural gas in the UGS is stored and migrated in the fractures and pores of the reservoir, with the injection production of the UGS, the internal characteristics of the reservoir are also changing (Dake, 1983; Gobran et al., 1987; Economides and Nolte, 1989; Cosentino, 2001; Lucia et al., 2003; Jaeger et al., 2009; Lake, 2012; Ahmed, 2018) (Figure 4). In order to clarify the influence of the change of the internal characteristics of the reservoir on the injection production capacity of the reservoir, use the product of porosity and permeability to characterize the internal characteristic value of the reservoir. The ratio of the internal characteristic value of the reservoir at different injection and production stages of the UGS to the internal characteristic value of the reservoir before the construction UGS, called injection-production coefficient, which can be expressed using Eq. 2. Before the construction of the UGS, the maximum injection-production coefficient is 1, As the UGS is built and operated, the injection production capacity gradually changes. According to the operation status of the UGS, the injection-production coefficient is divided into four ranges, analyze the change of the injection-production coefficient with the periodic injection production of the UGS, indicates the injection production capacity of the UGS (Table 3). Injection − production   coefficient : S z c = K n ⋅ Q n K i ⋅ Q i (2) Where, S _zc is the injection-production coefficient of UGS at different injection-production stages, dimensionless; K _n is the permeability of UGS at different stages of injection and production, respectively (mD); Q _n is the porosity of UGS in different injection production stages, respectively (%); K _i is the initial permeability of UGS, respectively (mD); Q _i is the initial porosity of UGS, respectively (%).

Select three groups of samples representing small, medium and large internal characteristic values of the UGS reservoir, namely, 1, 2 and 3 groups, to conduct single cycle injection and production, and test the change of injection-production coefficient under different effective stress (Figure 5).

For the same sample, under the same effective stress, Szc in the gas injection stage is greater than that in the gas production stage, which indicates that gas injection and production will damage the reservoir’s injection production capacity and is irreversible. Different samples have different Szc after injection and production, Group 3 (0.65) > Group 2 (0.55) > Group 1 (0.48), which shows that Szc is positively with the internal characteristic value of the reservoir, that is the larger the internal characteristic value of the reservoir after injection production, the higher Szc is, indicating that the higher the permeability and porosity of the reservoir, the better the injection production capacity; after the completion of injection and production, the variation value of Szc (Eq. 3) is different, Group 1 (0.08) > Group 2 (0.07) > Group 3 (0.05), indicating that the variation value of Szc is inversely correlated with the internal characteristic value of the reservoir, which shows that the greater the internal characteristic value of the reservoir, the smaller the change value of Szc, indicating that the higher the initial permeability and porosity of the reservoir, the less damage to the injection production capacity of the reservoir caused by gas injection and production. Therefore, reservoir reconstruction after completion of the injection production wells in UGS can improve the permeability and porosity of the reservoir and play a role in reservoir protection. Variation   value   of   injection − production   coefficient : Δ S zc = S zc ′ − S zc ″ (3) Where, △S_zc is the variation value of injection-production coefficient; S _zc ^’ is the injection-production coefficient in gas injection stage; S _zc ^’’ is the injection production coefficient in gas production stage.

Select 2nd group of internal characteristic values of representative UGS reservoir for multi cycle injection and production, and test the change of Szc under different effective stresses (Figure 6).

In the same injection production cycle, during gas injection reservoir pore pressure increases, while the overlying formation pressure remains unchanged, the increased pore pressure offsets part of the overlying formation pressure, and the effective stress acting on the reservoir decreases, making the reservoir internal fractures and pores space of the reservoir partially recovered, which is shown as Szc increases with the decrease of the effective stress; Contrary to gas injection during gas production, the reservoir pore pressure decreases, while the overlying formation pressure remains unchanged, and the effective stress acting on the reservoir increases, making the internal fractures and pores space compression of the reservoir smaller, which is shown as Szc decreases with the increase of effective stress. At the same effective stress point, Szc during gas injection is greater than Szc during gas production, and the difference increases with the increase of effective stress, indicating that the change of effective stress is irreversible to the damage of reservoir injection production capacity. In different injection production cycles, Szc decreases with the increase of injection production cycle, from 0.75 at the beginning of the first injection production cycle to 0.43 at the end of the fifth injection production cycle, but the decrease amplitude becomes smaller. After the third injection production cycle, the decrease value of injection-production coefficient tends to be stable, indicating that the damage to the reservoir injection production ability becomes weaker and stable due to the change of effective stress with the increase of injection production cycle. The injection-production coefficient of the 1st∼3rd injection production cycle is 0.75–0.51, and the injection production capacity of the reservoir is moderately strong; 0.49–0.42 after the fourth injection production cycle, the injection production capacity of the reservoir is moderately weak. Therefore, in order to maintain the injection production capacity of the UGS reservoir, the UGS capacity will be expanded after the completion of the 3rd injection and production cycle.

Select 3 samples representing the internal characteristic value of the gas reservoir, the effective stress is constant, With the increase of the effective stress acting time, real time test of injection-production coefficient changes under different effective stresses (Figure 7).

The overburden pressure of the gas reservoir is 71 MPa, and the UGS operates repeatedly at the upper and lower pressure of 28.06 MPa–13.20 MPa (the effective stress is 42.94 MPa–57.8 MPa), so the effective stress is 42.94 MPa, 57.8 MPa and the average value is 50.37 MPa. Under the same effective stress, Szc shows three stages of rapidly decrease, stable and further decrease with the increase of action time. The rapid decrease of Szc occurred within 1.5 h, with a decrease range of 2%–5%; Szc occurs stably in 1.5 h–3.0 h; Szc decreased slightly after 3.0 h. For different effective stresses, the smaller the effective stress, the greater the Szc, under the same action time, the greater the effective stress, the smaller the Szc. It shows that the damage of effective stress to the injection production capacity of UGS reservoir occurs at the beginning of injection production, and the damage tends to be stable with the injection production.

According to the scanning electron microscope pictures of the experimental samples, it can be observed that the pore morphology in the sample is mainly circular pore, and at the same time, there are long and narrow natural fractures (Figure 8). Carbonate fracture-pore type reservoir rock is hard in compactness, and the space for skeleton particle deformation is very small, therefore, the effective stress first causes the deformation of the fracture pore space, with the increase of effective stress, skeleton particle deformation and skeleton particle arrangement change successively (Figure 9). When the effective stress is between 40 MPa and 49 MPa (injection production pressure 31 MPa–22 MPa), Szc is larger and changes rapidly, the reason is that the reservoir rock with the increase of the effective pressure has obvious compaction, and the porosity and permeability decrease rapidly, at this stage, Szc changes are mainly caused by the deformation of the fracture porosity space. When the effective stress changes between 58 MPa and 49 MPa (injection production pressure 13 MPa–22 MPa), Szc is relatively small and basically unchanged, indicating that the injection production capacity is relatively weak and stable, the reason is that the space available for compression in the rock sample is reduced, and the decrease range of porosity and permeability is reduced, at this stage, Szc changes are mainly caused by the deformation of skeleton particles and the deformation caused by the change of skeleton particle arrangement.

Multi cycle injection and production reduces the injection production capacity of the UGS, but with the increase of injection and production cycle, the injection production capacity tends to be stable, indicating that the fractures and pores of the reservoir rock are reduced to a certain extent due to multi cycle effective stress, but still have injection production capacity. The scanning electron microscope results show that there are dense and a large number of microconvex bodies on the fracture and pore wall of the sample, when the fracture and pore are closed to a certain extent, these dense and large number of microconvex bodies contact the other wall of the fracture and pore, forming a supporting role for the fracture and pore (Figures 10, 11).

Figure 4 shows, Szc can be divided into three stages of rapid decrease, stability and redecrease with the increase of effective stress acting time. At the beginning, fractures and pores are closed to a certain extent under the effect of effective stress, resulting in a rapid reduction of Szc. After the effective stress is applied for a period of time, Szc is relatively stable due to the support of the microconvex body on the fracture and pore wall (Figure 12A). As the wall microconvex bodies, especially those in point contact, are prone to stress concentration, as the effective stress action time continues to increase, the microconvex body will break (Figure 12B), and the fractures and pores will be closed to a certain extent again, eventually causing Szc to decrease again (Figure 12C).