Assessment of Refracturing Potential of Low Permeability Reservoirs Based on Different Development Approaches

Abstract

The technique of refracturing is an effective method to solve the rapid decline in oil well production caused by factors such as severe reservoir energy loss and fracture failure after the initial hydraulic fracturing of low-permeability reservoirs. The key to designing refracturing lies in establishing a model for evaluating the potential fracturing layers. Based on the geological characteristics of the low-permeability conglomerate reservoir in the Lower Wuerhe area of the Eig District of the Xinjiang Oilfield, this paper studies the influence of different development approaches on the distribution pattern of remaining oil in the reservoir. A coupled model of remaining oil distribution and the in situ stress field is established and discusses the characteristics of the four-dimensional in situ stress field under different development modes. This paper analyzes the influence of geological factors and well network factors on the distribution of residual oil, and analyzes the influence of various factors, such as reservoir properties and injection and extraction parameters, on ground stress. Based on the residual oil distribution and ground stress changes, an evaluation method for screening potential fractured layers in reservoirs with different development modes (water injection development and depletion development) is developed.

1. Introduction

With the continuous advancement of unconventional oil and gas development in China, the proportion of low-permeability oil and gas reservoirs in oil and gas extraction has been increasing year by year. However, conventional development methods using volumetric fracturing for low-permeability oil and gas reservoirs often exhibit characteristics of rapid production decline, low initial recovery rates, and low ultimate recovery factors [1]. In recent years, the refracturing technique has gradually become an important means to improve the development effectiveness of low-permeability oil and gas reservoirs and enhance the recovery rates of mature oilfields.

The concept of refracturing was initially introduced in the United States. Although hydraulic fracturing of small-scale horizontal wells can effectively increase production rates during the initial stages, the gradual depletion of reservoir energy and the diminishing abilities of artificial fracture diversion and fluid flow can lead to declining production rates. Additionally, difficulties in water injection arise. Thus, to reinvigorate the production capacity of mature wells, exploit remaining oil, and enhance recovery rates, the technique of refracturing was introduced. This technique has been applied in numerous oil and gas fields in the United States, such as the Westbrook oilfield and the Barnett gas field in Texas. Following the implementation of refracturing measures, the production capacity of mature wells experienced significant improvements. Practice has indeed demonstrated that these measures can lead to substantial returns on investment [2].

In recent years, both domestic and international scholars have conducted extensive research in the field of refracturing, focusing mainly on mechanisms related to fracture propagation after refracturing [3,4], mechanisms of imbibition [5,6], and predictions of post-modification production capacity [7,8,9]. R. Jayakumar et al. [10] found that using the “refracturing new fractures and modifying old fractures” approach can increase the modified volume of low-permeability reservoirs, enhance well production, and yield economic benefits. Furthermore, Huang Jixiang et al. [11] employed finite element analysis to solve a numerical model that couples reservoir permeability with geomechanics. They considered the changes in fracture diversion capacity and elucidated the variations in stress and flow fields during the depletion production of wells.

Within the permissible range of fracture spacing, it is advisable to prioritize the strategy of refracturing new fractures in order to enhance reservoir modification volume and subsequently increase individual well production. Additionally, Ren Jiawei et al. [12] established models for predicting the production capacity of refracturing and calculating the stress field of fractured horizontal wells using numerical simulation. They determined the optimal combination of fracture parameters under the best modification method. Furthermore, Huang Ting et al. [13] investigated production enhancement mechanisms during pre-frac energizing and post-frac shut-in processes. Combining these with the refracturing modification technique, they proposed an energy storage refracturing technology and optimized the design of fracturing parameters. Zhang Wanchun et al. [14] took the Shizhuang South coalbed methane reservoir as an example, using the G-function and net pressure history to analyze the characteristics of refracturing fractures. They inverted the form of these fractures and further explored the relationship between fracture morphology and fracturing effects. Results indicated that fractures in high-productivity wells were mostly long and narrow, followed by short and wide. When fractures took a short and wide form, gas production was lower, providing references for optimizing subsequent designs of refracturing.

Kang Shaofei et al. [15], using the FX block in the Ordos Basin as a case study, analyzed the influencing factors of reservoir modification volume for fracturing. They selected 12 factors as input parameters for a BP neural network and established a predictive model for fracture length and fracture zone width using the BP neural network. Moreover, He Haonan et al. [16], based on Darcy’s law and the principles of fluid flow mechanics, considered the influences of initiation pressure gradient, gravity, and capillary forces. They developed a new model for predicting the post-refracturing production capacity of mature wells in tight oil reservoirs that incorporates imbibition effects. They selected 10 refracturing wells in a certain area of tight oil reservoirs for single-well production calculations, demonstrating that imbibition effects have a significant impact on well production. Wang Zhonghui et al. [17] conducted experiments applying fracturing diversion-blocking agents to enhance the effective propagation volume of fractures. Through on-site trials, they developed a set of fracturing methods suitable for practical field applications. They utilized the advantages of high-displacement fracturing to address issues, such as poor refracturing effects and short effective periods of measures involving fracturing and blocking removal in ternary composite flooding. Liu Jinjun et al. [18] proposed a methodology for addressing the relevant issues concerning fracture types and temporary blocking agent selection in the context of reoriented hydraulic fracturing. They demonstrated the effectiveness of the reoriented fracturing technique and conducted on-site reoriented fracturing experiments in the Quandong Oilfield. Qi Zhu [19], based on a comprehensive wellsite database, established a machine intelligence theory. They employed non-dimensional parameter methods for well and formation selection, along with an excitation evaluation model. This model identified the parameters affecting re-fracturing excitation efficiency, including elasticity–plasticity, permeability, porosity, completion parameters, declining production factors, and skin factor. Combining artificial neural networks and the Backpropagation (BP) algorithm, they calculated weight indices for different reservoir lithologies and analyzed the final evaluation value of fracturing effects. Ma Huibo et al. [20], utilizing data on previous fracturing conditions and dynamic development production data from the Karamay Gas Field, identified well selection decision factors. They established a re-fracturing well selection model using the Analytic Hierarchy Process and the entropy value method, integrated with the Capitalized Grey theory. This model provided a solution approach for ranking candidate re-fracturing wells from best to worst, allowing for the quantitative determination of suitable candidates. Huang Bo et al. [21], building upon conventional directional hydraulic fracturing, introduced the concept of “intra-fracture reoriented fracturing”. They analyzed four fracture patterns associated with intra-fracture reoriented fracturing, delineated suitable reservoir conditions, discussed implementation strategies, and proposed a preliminary conceptual design methodology.

In summary, scholars both domestically and internationally have conducted a series of studies in the field of refracturing, with a focus on mechanisms of fracture propagation after refracturing, and predictions of post-modification production capacity. However, there is currently no comprehensive method for selecting potential layers for refracturing tailored to different reservoir development approaches. This paper takes the development of the low-permeability conglomerate reservoir in the Lower Wuerhe area of the Eig District of Xinjiang Oilfield as an example. By thoroughly analyzing the reservoir characteristics and development features of the related areas of the Xinjiang Oilfield, the paper investigates the distribution of remaining oil and the four-dimensional stress field in reservoirs under different development approaches. This study aims to select potential layers for refracturing in different development modes, providing theoretical guidance for the selection of techniques, such as temporary plugging and energy storage, in refracturing.

2. Numerical Simulation Model

The studied reservoir is the low-permeability conglomerate reservoir in the Lower Wuerhe area of the Eig District of Xinjiang Oilfield. It is situated at depths ranging from 2300 to 3300 m and covers an oil-bearing area of 47.47 square kilometers. The reservoir has an average effective thickness of 62.2 m, with geological reserves of 96.25 million tons. It was initially developed in 1979 and has undergone four density adjustments and localized five-fold density tests. There are a total of 1304 wells, consisting of 985 oil wells and 319 water wells. The comprehensive water cut is 76.6%, and the recovery factor is 27.79%. The average recovery factor of the Lower Wuerhe Formation in the Eig District is 23.27%, with the highest recovery factor found reaching 28.68%. Refracturing has been conducted over the years, with over 400 measures taken annually. Due to the vast number of older wells in the area, there are significant differences in geological characteristics, development challenges, and wellbore conditions.

For this study, the Petrel-Re + Eclipse integrated research platform was employed. A total of 1413 injection and production data wells are present within the entire area. The structural arrangement of the target block is illustrated in Figure 1. Based on data collected from relevant wells, drilling tests, and dynamic monitoring data, the following reservoir property models were established:

• Net-to-Gross Ratio (NTG) was calculated using an averaging algorithm.
• Porosity distribution was determined using the arithmetic mean algorithm after NTG weighting.
• Permeability was calculated using a geometric mean algorithm weighted by porosity and NTG.
• Saturation was determined using an arithmetic mean algorithm weighted by porosity and NTG.

The attribute distributions are displayed in Figure 2.

The static geomechanical model is built by using Petrel 2020 software, after collecting basic data, drilling data, and well test data, the seismic data are used to model the geological structure of the target area, including fault modeling, initial stratigraphic stratification with logging data, structural lattice modeling, and the porosity, permeability, and saturation models are input into the model to initialize the calculation of the geological storage capacity. The initial calculation of geological reserves is performed, and the static geomechanical model is established by combining the well-completion data and the PVT analysis data. The enumeration method is used to initialize the model saturation field and pressure field to establish a numerical model, which is combined with the production history fitting. The numerical model is initialized for calculation, coupled with rock physical properties to form an initialized static stress field.

Based on the aforementioned structural and reservoir property models, a geological reserve calculation was conducted. Incorporating well completion and downhole measures data, geological model water saturation data were introduced. The model was initialized using an enumeration method for saturation and pressure fields. After initialization, the numerical model yielded a reserve of 1.12 × 10⁴ tons, with an error of 1.7% compared to the geological model total reserve, satisfying the standard requirement of an error less than 5%. The initial pressure field and saturation field are illustrated in Figure 3. Following the establishment of an initial coarse model, adjustments were made to phase permeability and reservoir permeability models. Liquid quantity was fitted to match the entire area’s oil production and water cut, while the reservoir pressure was fitted using conversion standards. Adjustment coefficients for various types of cumulative injection volumes are listed in

Table 1. Adjustment factors for each type of cumulative injection.

Wells	Water Injection Volume Range	Adjustment Factor
35	>500,000 m3	0.2
21	200,000 m3–500,000 m3	0.3
113	100,000 m3–200,000 m3	0.5
105	<100,000 m3	0.9

. The final fitting results are depicted in Figure 4, with the overall trend of water cut aligning consistently and the variation in reservoir pressure closely mirroring measured data.

3. Distribution Pattern of Remaining Oil

This section discusses the distribution pattern of remaining oil. Through analysis of production data from various wells combined with the established numerical simulation model, the distribution pattern of remaining oil can be categorized into four types.

The first type is fault-controlled: the remaining oil is enriched in the corners of faults. Faults in the target block are developed, and most of them are multi-oil and water systems, there are few control wells in the tectonic high part near the faults, and most of them are depleted development wells, with a small water injection wave, so it is an important place for the distribution of residual oil.

The second type is micro-tectonic control: micro-tectonics is a slight undulation change of the top surface (bottom surface) of the oil layer on the background of the trap structure, mainly for the positive tectonics, negative tectonics, and oblique tectonics. The theoretical basis of the micro-tectonic influence on the distribution of residual oil is the influence of oil reservoir inclination and gravity differentiation of oil and water on the development of water injection, so that the speed of the waterline advance tends to be uniformly distributed in the plane, the oil and water are differentiated according to gravity, and more residual oil will be retained in the high part of the structure.

The third type is the imperfect type of well network: the well network factor is mainly manifested in the degree of control of the well network and injection system on the reserves and the degree of adaptation to the non-homogeneity of the oil reservoirs. The most important is the degree of perfection of the injection system and its configuration with the relationship between the geological factors. The existing well network has a low degree of control over the oil reservoir, resulting in imperfect injection and extraction, injection without extraction, or extraction without injection, thus forming residual oil.

The fourth type is the unreasonable injection and extraction differential pressure type: the production differential pressure is unreasonable. The production differential pressure is too small to meet the demand of supplementing formation energy and production capacity, and the production differential pressure is too large, resulting in the injection of the horizontal plane along the main stream line and vertical single layer, resulting in violent flooding of oil wells, low degree of utilization of water drive, low efficiency of oil drive, and enrichment of residual oil.

Pressure distribution across the region indicates that the pressure on the western side of the 256-fault zone is generally high, at about 60 MPa, while the formation pressure on the eastern side of the 256-fault zone is low, at about 25 MPa. The remaining oil is mainly distributed in the peripheral zones and rezones of well areas with limited oil injection and production efficiency. Oil enrichment is observed around the edges of fault corners. In areas with well-developed faults and multiple oil-water systems, wells controlled by higher structural positions near the fault are fewer in number. These wells often follow a depletion-style development with minimal water injection influence, making them important areas for the distribution of remaining oil. The masking effect of small faults contributes to the accumulation of remaining oil. In the later stages of development in complex fault-block oilfields, low-order small faults play a significant role in controlling the distribution of remaining oil. Due to the obscuring effect of these small faults, remaining oil accumulation may occur in both hanging-wall and footwall sections of the fault, as shown in Figure 5 (left).

Microstructures are minor undulations on the top (or bottom) surface of oil layers within the context of trap structures. They mainly consist of positive, negative, and inclined structures. From Figure 5, it can be inferred that the theoretical basis for the microstructure’s impact on the distribution of remaining oil lies in the influence of the oil layer’s dip angle and oil-water gravity differentiation on water injection development. This results in a more uniform advancement of the waterfront on the plane and the gravity-induced separation of oil and water. More remaining oil tends to accumulate in high-structural areas.

Well network factors primarily involve the degree of control the well network and injection-production system have over reserves, as well as their adaptability to the heterogeneity of the reservoir. Most importantly, the adequacy of the injection-production system and its configuration in relation to geological factors play a vital role. Inadequate well network control may lead to imperfect injection and production systems, with scenarios like injection without production or production without injection resulting in the formation of remaining oil.

Unreasonable production pressure differentials are mainly caused by either insufficient pressure differential, which fails to meet the requirements for replenishing reservoir energy and production capacity, or excessive pressure differential, leading to the advancement of the waterfront along the main flow path in the horizontal plane and causing single-layer advancement in the vertical direction. This can result in sudden water flooding of oil wells, low water drive efficiency, and poor oil displacement efficiency, leading to the accumulation of remaining oil.

4. Dynamic Geomechanical Model

This section discusses the process of geomechanical modeling (Figure 6, Figure 7 and Figure 8). The geomechanical modeling process before refracturing involves five stages: the isotropic in situ stress field, anisotropic in situ stress field, initial post-fracturing stress field, stress field after production and water injection, and the stress field after refracturing. The calculation of the entire stress field is a complex fluid–solid coupling problem that is time-dependent. The key to achieving 4D stress simulation is to primarily employ a static geomechanical model, as depicted in Figure 8. This is then combined with reservoir simulation and geomechanical models to achieve the coupling of fluid–solid parameters. Finally, a cross-iteration coupling method is applied to solve the model, as illustrated in Figure 6. The structural factors of rocks, such as faults, microfractures, etc., need to be taken into account in geostress prediction, especially due to the existence of a complex fracture network after the initial fracturing, and therefore the directional difference of local tectonic stress needs to be considered in geostress prediction, and therefore an anisotropic geologic model was established.

We use ADGPRS1.0 to establish a 3D geomechanical grid model and initialize the model grid with attributes from single-well mechanical profile data and attribute models, and use a reservoir differential model coupled with data conversion to adjust the attributes with a grid mapping strategy program, and initialize the stress field using a geologic finite-element model, and ultimately complete the estimation of the 3D stress field. We have continued the calculation of the initial stress field through the single well profile data and the initial 3D geological model established in part 2, combined with the test pressure data, and ensured its accuracy. After obtaining the static initial stress field, the initial fracturing is carried out, and the ground stress field after the initial fracturing is obtained by combining the fracture geometry after the initial fracturing. Reservoir production is then carried out, and the pore pressure field changes to obtain the ground stress field for depletion production. Then water injection is performed to obtain the ground stress field before refracturing. This process establishes the dynamic stress response model.

Due to variations in the dynamic stress field patterns under different development modes, this study considers two development modes: depleting development reservoir-geomechanical coupling and water injection development reservoir–geomechanical coupling numerical models. The rock mechanics and stress field models for the working area were constructed, as depicted in Figure 7. A total of 35 well-specific rock mechanics and stress calculations were carried out. Based on these, the study conducted initial fracturing and refracturing analyses.

The range of elastic modulus values in the study area falls between 19.5 to 59.2 GPa, with an average of 34.2 GPa. The Poisson’s ratio ranges from 0.20 to 0.31, averaging at 0.29. The brittleness index ranges from 22.1 to 72.5, averaging at 41.3, indicating a moderate overall brittleness.

5. Analysis of 4D Stress Field in Different Development Modes

This section will discuss the 4D stress field under different development modes. The entire time period of 44 years (1977–2021) is divided into four key stages: 1977, 1992, 2007, and 2021. The evolution of the minimum horizontal principal stress across the entire working area is illustrated in Figure 9, reflecting the overall decrease in pore pressure (stress) with significant local variations. Two specific well groups, Group 1 and Group 2, were selected to illustrate the dynamic stress field simulation results under different development conditions. Additionally, the analysis incorporates individual wells to examine the dynamic stress field changes resulting from injection and production development.

5.1. Water Injection Development Model

For the water injection development model, a diamond-shaped inverted nine-point well pattern was selected. A dynamic stress field analysis was conducted on a typical water injection well and an oil production well within the working area under various development stages. The results of the stress field changes within the well group under water injection development conditions reveal that as the development progresses, the magnitude of the stress decreases in a downward pattern, with the area of decrease gradually expanding. Moreover, a larger stress-reduction zone appeared near the oil production wells. In the context of the water injection development model, the overall stress reduction trend within the well group is influenced by the injection–production relationship, resulting in a relatively uniform reduction trend around each well. As shown in Figure 9 and Figure 10, the average decrease in the minimum horizontal principal stress across the working area is approximately 16 MPa, while the average decrease in the maximum horizontal principal stress is around 20 MPa. The overall reduction is about 18% compared to the initial decline during reservoir development.

Individual grid cells around the water injection and production wells were segmented to analyze the change in principal stress in the near-well zone. For water injection wells, continuous development led to a decreasing trend in both maximum and minimum horizontal stress magnitudes. The histogram of the minimum principal stress within the working area shifted to the left, with a decrease of approximately 2 MPa from 1977 to 1992. Similarly, for the production wells, continuous development resulted in a decreasing trend in both maximum and minimum horizontal stress magnitudes, with a decrease of about 5 MPa from 1977 to 1992.

Comparing the stress reduction patterns of water injection wells and production wells, it is evident that the average reduction in stress around water injection wells is relatively smaller than the average stress reduction around production wells. Additionally, the stress reduction trend around water injection wells is more gradual compared to the stress reduction trend around production wells, confirming that water injection has a certain restorative effect on reservoir pressure.

5.2. Fracture Depletion Development Model

For the fracturing depletion development model, numerical simulations were conducted on fracturing wells. Two adjacent fracturing wells within the working area were selected for dynamic stress field analysis under different development stages. One of the wells underwent refracturing operations. Under various development stages the two fracturing wells exhibited similar stress reduction trends, with the reduction primarily concentrated around the well vicinity. The overall decrease in the minimum horizontal principal stress is approximately 12 MPa, which is higher than the stress reduction caused by water injection development (around 6 MPa).

Due to the creation of fracturing fissures around the two fracturing wells, stiffness-induced deformations occurred, altering the local stress field. As a result, the stress reduction exhibited an elliptical pattern. By analyzing the stress difference in different production stages, it was observed that continuous production led to an increase in the horizontal stress difference. In regions with high stress differences, temporary plugging and fracturing redirection techniques should be employed to mitigate the effects of high stress differentials.

5.3. Reservoir Layer Evaluation Method for Potential

Based on the remaining oil in reservoirs and the evolution distribution pattern of stress fields, an assessment is conducted on the potential for rejuvenating old fractures through re-fracturing and the potential for creating new fractures through re-pressuring. This assessment aims to select reservoir properties, initial completion, production dynamics, and other parameters that influence the effectiveness of re-fracturing. A parameter system for evaluating the re-fracturing potential layers is established, incorporating comprehensive evaluation theories to form a method for assessing the re-fracturing potential layers.

Numerical simulations of re-fracturing and production forecasting are carried out to identify the critical turning points and thresholds for re-fracturing transformations. A quantitative system for constructing indicators is illustrated in Figure 10 below.

• Class I Re-pressuring Potential Wells: >0.6 (uniform distribution of remaining oil)
• Class II Re-pressuring Potential Wells: 0.45~0.6 (local distribution of remaining oil)
• Class III Re-pressuring Potential Wells: <0.45 (sparse distribution of remaining oil)

5.4. Model Reliability Validation

This section focuses on validating the reliability of the models established using the methods described above. By interpreting pressure and shut-in pressure curves at different time points from well tests, the minimum horizontal principal stress at different development moments was inversely determined. As can be seen from

Table 2. Comparison of predicted and measured geostress values in the field.

Well Number	Stop Pump Pressure (MPa)	Minimum Horizontal Principal Stress Calculated Value (MPa)	Errors (%)
TD104XX	42.4	38.2	9.91
T740XX	44	36.8	16.36
BJ94XX	53	45.55	14.06
T867XX	59	52.7	10.68
T875XX	62	55.8	10.00
T870XX	66	57.1	13.48
T853XX	60	51.5	14.17
T875XX	62	57.2	7.74
T867XX	67	62.5	6.72
T322XX	40	36.4	9.00
T870XX	59	62.3	5.59
8D60XX	57	60.7	6.49
Average error: 11.24%

, the numerical simulation results of the reservoir pressure and stress field were then compared with field measurements to analyze the error, which was found to be below 15%. This validation confirms the reliability of the model.

Furthermore, based on the dynamic stress field results obtained from geomechanical calculations, numerical simulations of refracturing were conducted, and production forecasts were made. The results are shown in

Table 3. Comparison of forecast and actual production values.

Well Number	Repeat Fracture Type	Daily Oil Production Forecast/t	Daily Oil Production in Actual Terms/t	Error/%
T862XX	Fracture single shift in direction in refracturing	27.58	29.09	5.18
T870XX	Fracture single shift in direction in refracturing	22.51	20.72	8.61
T866XX	Fracture secondary shift direction in refracturing	22.1	19.55	13.04
T850XX	Fracture secondary shift direction in refracturing	20.20	18.16	11.23

, the simulation results were compared with field measurements, demonstrating a match. This further validates the reliability of the model.

6. Conclusions

Based on the analysis of reservoir characteristics in the relevant block, this study has established a reservoir flow model to uncover the influence of different development methods on the distribution pattern of remaining oil in the reservoir. Special emphasis was given to the remaining oil distribution under the conditions of water injection development and depletion development. A physical model coupling reservoir development methods with reservoir geomechanics was formulated. This model helped reveal the four-dimensional stress field characteristics in the context of the reservoir–geomechanics interactions for both water injection and depletion development models. In conclusion, the following findings were obtained:

(1)

Combining the results of the remaining oil distribution, the expansion of the near-wellbore area was achieved through initial fracturing and subsequent refracturing. This strategy aimed to redirect the fractures and control the fracturing parameters to ensure that the fractures extended over a broader vertical and horizontal range or increased their length to reach the remaining oil zones. The ultimate goal of this approach was to enhance the efficiency of remaining oil recovery.

(2)

Through constructing a static stress model, it was observed that the stress field exhibited a distribution pattern where the vertical stress was greater than the horizontal principal stresses. The variation range of the maximum horizontal principal stress was 35 to 70 MPa, while the variation range of the minimum horizontal principal stress was 30 to 60 MPa. The difference between the two horizontal principal stresses was in the range of 5 to 12 MPa.

(3)

In the water injection development model, the decrease in stress was smaller compared to the stress decrease in the fracturing development model. Additionally, the decrease in stress was more evenly distributed in the water injection development model. In the fracturing development model, the stress distribution exhibited a distinct elliptical shape, which was correlated with the extent and trajectory of the fractures’ propagation.

(4)

Due to the extended development period, significant stress differences emerged in localized regions. Only through temporary plugging during refracturing could the redirection of fractures be achieved, allowing for a sustained and stable production increase in the mature reservoir areas.