Experimental Study on Effective Propping of Multi-Level Fractures Using Micro-Proppants

Abstract

In deep shale gas fracturing, the narrow width of micro fractures presents a challenge for conventional proppants (40/70 mesh, 70/140 mesh), which often fail to enter branch fractures, resulting in inadequate effective support volume. To address this, a high-efficiency propping strategy is proposed based on the hybrid use of micro-proppants and conventional proppants. Utilizing a proppant transport experiment device, the effects of proppant size ratios and injection timing on proppant distribution were investigated to determine the optimal design parameters. The results indicate that the 200/400 mesh micro-proppant can effectively enter the distal micro fractures, thereby mitigating the problem of the non-uniform distribution of the proppant within the fracture network. To ensure effective propping of secondary fractures, the optimal pumping sequence is to inject quartz sand first, followed by ceramic proppants. The recommended ratio of 70/140 mesh quartz sand to 40/70 mesh ceramic proppants is 7:3. Additionally, for blended injection, the optimal mixing ratio of 70/140 mesh quartz sand to micro-proppant is 8:2. Field trials at the L-X1 well in the LZ block demonstrate that this strategy significantly boosts post-fracturing production, with test yields increasing 2.4 to 4 times.

1. Introduction

Deep shale gas is a key component of unconventional hydrocarbon resources, and its large-scale development is of great importance for increasing China’s natural gas production, reducing reliance on imports, easing domestic energy supply–demand conflicts, and ensuring national energy security. Horizontal well multi-stage, multi-cluster hydraulic fracturing technology is the core technology for achieving the efficient development of deep shale gas. Compared with shallow and medium-depth shale, deep shale has higher in situ stress, larger stress differentials, and pronounced rock plasticity. This results in a relatively lower complexity of fractures, characterized by a more regular structure of “main fracture + branch fracture + micro fracture”. The overall fracture width is narrower, and there is a significant difference in width between the main and secondary fractures. This causes most proppants to be concentrated in the main fracture, making it difficult for them to enter secondary fractures [1,2]. Consequently, there is a severe “liquid entry without proppant” phenomenon in branch and distal micro fractures, which negatively impacts the overall stimulation effect.

Effective stimulated reservoir volume (ESRV) directly affects the stimulation effectiveness of deep shale [3]. How to construct high-ESRV fracture networks has always been a key concern in engineering. From the perspective of proppant transport, fractures expand within the reservoir and eventually form a three-dimensional fracture network. Achieving effective proppant placement in both the horizontal and vertical fracture regions is crucial for enhancing ESRV. Currently, studies on proppant placement patterns are primarily conducted using two approaches: laboratory-scale proppant transport experiments [4,5,6] and numerical simulations [7,8,9]. Numerical simulations often require simplifying assumptions during model development and are affected by factors such as mesh quality, boundary condition treatments, and equation selection. Consequently, the accuracy and reliability of numerical predictions must be validated through experimental results. Therefore, this study employs laboratory experiments to investigate the proppant placement law.

Kern et al. [10] first systematically proposed the variation mode of proppant stacking behavior within a single fracture. The proppant injected in the early stage piles up at the fracture front, while the proppant injected later bypasses the formed sandbank and moves towards the fracture depth, piling up behind the sandbank. Babcock et al. [11] expanded on Kern’s research, introducing the concepts of equilibrium velocity and proppant stacking constants. Based on the particle concentration along the fracture height, he divided it into four zones from bottom to top. Medlin et al. [12] proposed three proppant transport mechanisms in low-viscosity fluids and analyzed the proportion of each mechanism by establishing differential fluid diffusion equations and motion equations for fluidized particles at the sandbank surface, combined with experimental observations. He suggested that the turbulent transport mechanism has a limited range of action and is confined to the area near the injection port. The bed transport mechanism only occurs after the sandbank has accumulated to a certain height and has certain requirements for fluid shear rate. Only fluid viscous resistance transport is the key to proppant transport efficiency. Sahil et al. [13] conducted experiments using alternating injections of proppant slugs with different viscosity fluids and confirmed that the difference in fluid viscosity could temporarily enhance proppant transport distance.

In 2009, Dayan et al. [14] first discovered that sufficient proppant accumulation in the main fracture and ensuring fluid velocity are essential for proppant entry into branch fractures. Since then, research on proppant transport mechanisms in multi-level fractures has been initiated. Sahai et al. [15] proposed two transport mechanisms for proppant entering secondary fractures from the main fracture. (1) When fluid velocity reaches a threshold, the proppant is suspended and enters the secondary fracture under fluid carrying. (2) Proppant particles roll into the secondary fracture under gravity. He also believed that fracture width and proppant particle size must match, with the secondary fracture width being 2.5–2.8 times larger than the proppant particle size for proppant entry. Wen et al. [16], building on Sahai’s research, suggested that the proppant enters secondary fractures near the injection port in a suspended manner and distal branch fractures via rolling. Zhang et al. [17] recognized that proppant entering secondary fractures undergoes a process of turning, decelerating upon wall collision, and re-accelerating. After entering the fracture, proppant particles collide with the fracture wall, losing energy and reducing velocity, which increases their tendency to settle. However, fluid can replenish energy to the particles, enabling their reactivation. Zeng et al. [5] conducted experiments using varied particle and fluid parameters in a narrow planar channel to investigate particle transport and distribution. Based on the results, two correlation models were developed using multiple linear regression, incorporating four dimensionless parameters to predict the bed equilibrium height and the coverage area of small particles in narrow fractures. Guo et al. [18] carried out proppant transport simulation experiments under different fracture geometries, sand ratios, branched fracture opening times, and injection sequences involving proppants of different particle sizes. The results showed that injecting proppants with varied particle sizes in a specific sequence can enhance proppant placement lengths in both main and branched fractures. Bahri et al. [19] explored the influence of fracture roughness and varying fracture widths on proppant transport. Their study also investigated the effects of proppant density, size, and concentration on transport behavior within rough fracture surfaces.

Previous laboratory experiments on proppant distribution in multi-level fractures [20] have shown that the filling degree of distal secondary fractures is relatively low in fracture systems. Conventional 40/70 mesh proppants mainly distribute in the main fracture and struggle to enter 1 mm width fractures. Although 70/140 mesh proppants can enter secondary fractures, they only pile up at the front of the fracture, leaving most of the rear region of the fracture unfilled. To effectively fill these extremely narrow (micron-level) micro fractures, it is necessary to design and develop micro-proppants that match the micro fractures for targeted filling. However, current research on proppant placement patterns in multi-level fractures primarily focuses on conventional proppants with particle sizes equal to or larger than 70/140 mesh. Experimental studies on finer micro-proppants remain limited, and key parameters such as injection strategies and mixing ratios with conventional proppants are yet to be clearly defined. This lack of understanding hinders the practical application of micro-proppants in the field due to insufficient theoretical support.

Based on this, this paper aims to address the uneven distribution of proppants in multi-level fractures and the insufficient effective support volume in secondary fractures. It seeks to achieve effective fracture support by employing physical experiments to explore the placement patterns of micro-proppants in multi-level fractures. This study provides scientific guidance and theoretical support for the design and optimization of proppant pumping parameters in deep shale fracturing.

2. Experiment Equipment

The experiments were conducted using a self-developed proppant transport experiment system [21], which consists of four main components: a circulation and pumping module, a control module, a fracture module, and a sand bank morphology acquisition module.

The circulation and pumping module simulates the field blending and injection process. The proppant is fed into a mixing tank at a controlled sand addition rate, where it is thoroughly blended with the fracturing fluid. The resulting slurry is then injected into the fracture using a screw pump. The control module enables centralized operation of the entire system through a computer interface. It allows the user to adjust the frequency of the variable-frequency drives on the stirrer, mixing tank, and screw pump, and it also collects real-time data from pressure and flow sensors. The sand bank morphology acquisition module includes high-resolution industrial cameras and custom image processing software. During experiments, multiple cameras capture real-time images of proppant placement. These images are stitched together, and the software uses image recognition techniques to extract the sand bank contours and calculate the placement area, enabling detailed analysis of proppant placement behavior.

The fracture module is composed of smooth flat plates that can be flexibly assembled (Figure 1). Given the complex and variable geometry of fracture networks in deep shale formations, a modular design was implemented using standard slot-type fracture connections with variable widths, multiple angles, and full transparency. These can be freely combined with the base fracture plates to construct various fracture geometries. The specific fracture structures used in this study were designed based on data from true triaxial mechanical tests [1,22], core sampling [23], and microseismic fracture monitoring [24]. Two representative fracture geometries were constructed: double-branch fractures and multi-level fractures. At the entrance of the fracture module, a simulated wellbore unit is installed to replicate field injection conditions. This unit is equipped with three perforation ports—upper, middle, and lower—which allow for the injection of slurry from different positions along the wellbore. In the experiments conducted in this study, all three ports were opened simultaneously. Each injection port has a diameter of 10 mm, closely matching the size of perforation holes created by shaped charges in actual field operations.

3. Optimization of Effective Propped Parameters of Fractures

After clarifying the approach of adopting micro-proppants that match the width of micro fractures in deep shale fracturing and considering the transport and distribution patterns of proppants in multi-level fractures, the key evaluation indicators for effective fracture propping are defined as follows: ① Sufficient placement in the main fracture: As the primary conduit for hydrocarbon flow from the reservoir to the wellbore, the main fracture must maintain high conductivity. This requires reducing flow resistance by fully filling the main fracture with high-conductivity proppants—meaning the height of the proppant bank in the main fracture should not be too low. ② Ensuring proppant entry and long-distance placement in branch and micro fractures: This implies the need to use smaller-sized proppants during fracturing operations and to ensure an adequate proportion of them in the mixture.

This subsection focuses on the optimal combination ratio of multi-sized proppants and the proppant pumping strategy for deep shale fracturing in South Sichuan. Specifically, it aims to investigate the injection timing and ratio of 70/140 mesh quartz sand and 40/70 mesh ceramic proppants, explore the integration of micro-proppants with conventional proppants, and develop a suitable proppant pumping strategy for deep shale reservoirs.

3.1. Proppant Pumping Mode

The design of the proppant pumping program includes determining the pumping rate, the type of proppant, the order of proppant pumping, the viscosity of fracturing fluid at different stages and the concentration of proppant, and other related parameters, and finally forming a complete and fine pumping program. To balance the objectives of creating complex fracture networks and achieving the goal of fully filling fractures with proppants, the pumping parameters (rate, fluid viscosity, and proppant concentration) were maintained within relatively fixed ranges. The treatment employed a moderate-high viscosity pad fluid (15–30 mPa·s) with stepwise pumping rate ramping to initiate and propagate fractures. This was followed by continuous high-rate pumping (14–18 m³/min) of a low-viscosity fluid (2.5–5 mPa·s) for proppant transport. The pumping of the proppant began with a low concentration (<120 kg/m³) of 70/140 mesh quartz sand to reduce near-wellbore friction by perforation erosion. During the main proppant stage, high concentrations (200–240 kg/m³) were pumped to saturate and pack the fracture body. The final stage utilized lower concentrations (<140 kg/m³) of 40/70 mesh ceramic proppant in a tail-in sequence specifically designed to pack the near-wellbore fracture region and mitigate screen-out risks (noting that some wells historically experienced mid-stage pumping of ceramic proppant). Comprehensively analyzed, the liquid viscosity and proppant concentration are not changed, but there is a “process of low viscosity fluid movement in high viscosity fluid” after the proppant enters the fracture, and it is necessary to carry out experiments to clarify the influence of this effect on the proppant transport. At the same time, in order to clarify the specific distribution of the proppant in the fracture under different pumping sequences, it is designed to carry out transport experiments under the conditions of different ratios of 70/140 mesh quartz sand and 40/70 mesh ceramic proppants and injection timing. The experimental plan is presented in

Table 1. Experimental plan for different ratios and injection timing of conventional proppants.

Case	Pumping Rate (L/min)	Proppant Concentration (kg/m3)	Proppant Ratio	Injection Timing	Liquid Viscosity (mPa·s)
#1	10	/	/	/	30→2.5
#2					15→2.5
#3	10	70/140 mesh = 200 40/70 mesh = 120	70/140:40/70 = 8:2	Tail-in 40/70 mesh	2.5
#4			70/140:40/70 = 6:4	Tail-in 40/70 mesh
#5			70/140:40/70 = 1:0	/
#6			70/140:40/70:70/140 = 4:3:3	Mid-job 40/70 mesh

, and schematic diagrams under different pumping conditions are shown in Figure 2. For Cases #1 and #2, changes in fracturing fluid viscosity are illustrated in Figure 2a. The “Tail-in 40/70 mesh” mode includes only steps 1 and 2, as shown in Figure 2b, whereas the “Mid-job 40/70 mesh” mode includes steps 1 through 3.

3.1.1. Viscosity Variation Experiment in the Pad Stage

In the pad stage, a medium-to-high viscosity fluid is first injected to open the fracture. When a lower-viscosity fluid is subsequently introduced, the interaction between the two fluids influences the overall flow behavior. Figure 3 demonstrates the injection of 2.5 mPa·s fluid into 15 mPa·s and 30 mPa·s fluids, respectively. In the figure, T1–T7 (or T6) represent different experimental stages, and the 2.5 mPa·s low-viscosity fracturing fluid is dyed with blue ink for visualization. The yellow dashed boxes in the figure indicate the stages during which the blue low-viscosity fluid was observed. When the 2.5 mPa·s fluid is injected into the 30 mPa·s fluid, a clear “viscous fingering-in” phenomenon, even in the secondary fracture, still maintains a strong penetration and low-viscosity fluid bandwidth of about 4 mm; when the 2.5 mPa·s fluid is injected into the 15 mPa·s fluid, low viscous fluid penetration is weakened, the degree of diffusion is increased, with a low-viscosity fluid bandwidth of about 10 mm. With the increase in injection time and sand mass, the viscous fingering phenomenon finally disappears under the influence of fluid diffusion and ultimately has small impact on the final proppant placement shape. Low-viscosity slickwater has a positive effect on the opening of natural micro fracture and the entry of small-sized proppant, so it should be used to control the dosage of moderate-high viscosity slickwater in the pad stage to create the main fracture, followed by a timely switch to low-viscosity slickwater when the treatment pressure becomes stable.

3.1.2. The 40/70 Mesh and 70/140 Mesh Proppant Ratio and Pumping Sequence Experiment

Figure 4 shows the final proppant distribution in the fracture under different proppant ratios and injection timing conditions. The dimensions of the multi-level fracture are as follows: the main fracture is 3.66 m long and 6 mm wide; the first-level branch fracture is 1.16 m long and 3 mm wide; and the second-level branch fracture is 1.05 m long and 1 mm wide. The height of all fracture levels is uniformly 0.3 m.

When the single 70/140 mesh quartz sand is pumped throughout the whole process, the proppant in the main fracture and the first-level branch fracture is filled to a high degree, and the proppant in the second-level branch fracture is mainly distributed at the front end of the fracture; when the ratio of 70/140 mesh quartz sand to 40/70 mesh ceramic proppant is 8:2, the main fracture and the first-level branch fracture are fully filled, and the degree of filling in the second-level branch fracture is lower than that in the pure 70/140 mesh condition. In this case, the last injected ceramic proppants tend to push the surface layer of quartz sand backward, forming a stratified ceramic band on the sand bank surface near the injection point, within the main fracture, and in the first-level branch fracture, with very low levels of ceramic proppants found in the second-level branch fracture; under a 6:4 ratio of quartz sand to ceramsite, the distribution of ceramsite in the middle and rear part of the main fracture and the first-level branch fracture will be increased, and the distribution of ceramsite in the second-level branch fracture will be improved slightly compared with that in the 8:2 ratio; When the ratio of 4:3:3 is used to pump 40/70 mesh ceramsite in the middle stage, a stacked distribution of bottom quartz sand—middle ceramsite—top quartz sand was formed in the fracture. The statistical results of the proportion of different fracture placement areas are shown in

Table 2. Statistics of proppant placement area proportion.

Proportion	Main Fracture	First-Level Branch Fracture	Second-Level Branch Fracture	70/140 Mesh Quartz Sand and 40/70 Mesh Ceramsite Ratio
Quartz sand	0.8718	0.8697	0.5643	1:0
Ceramsite	0	0	0
Total proportion	0.8718	0.8697	0.5643
Quartz sand	0.7874	0.8351	0.4688	8:2
Ceramsite	0.1235	0.0712	0.0417
Total proportion	0.9109	0.9063	0.5105
Quartz sand	0.6176	0.5066	0.2291	6:4
Ceramsite	0.2278	0.3456	0.0972
Total proportion	0.8454	0.8522	0.3261
Quartz sand	0.6147	0.8994	0.3817	4:3:3
Ceramsite	0.2451	0.6002	0.3167
Total proportion	0.8598	0.2992	0.065

.

There is no obvious difference in the distribution of large-size proppants in the near-well zone under different experimental ratios. With the increase in the proportion of large-size proppants, the proportion of large-size proppants in the distal end of the main fracture and the first-level branch fracture is obviously improved, but the filling degree of the second-level branch fracture decreases, which may have a positive effect on the stable production in the later stage of production. At the same time, premature accumulation of large-size proppants at the narrow secondary fracture would interfere with subsequent proppant entry. Therefore, small-size proppants should be continuously pumped in the early stage, with the subsequent backward transportation of large-size proppants along the surface of the proppant pack to make the distal fracture propped.

In deep shale formations with high closure stress, maintaining a high proportion of high-strength, large-size proppant is beneficial for improving the overall conductivity of the fracture. Based on experimental results, a proppant ratio of 70/140 mesh quartz sand to 40/70 mesh ceramsite at 7:3 is recommended. The quartz sand is pumped first, followed by the ceramic proppant, to ensure that large-size proppant fills the rear section of the main fracture, while minimizing the impact on proppant entry into secondary fractures.

3.2. Micro-Proppant Mixing Ratio

The experiments used a micro-proppant with silicate as the base material (Figure 5), which has a solid powdery appearance, true density of 2.0–2.8 g/cm³, and packing density of 1.0–1.3 g/cm³, and the fracture conductivity is 10~50 times higher than that of self-supporting fracture conductivity under the same sand concentration and closure pressure.

3.2.1. Basic Properties of Micro-Proppant

(1)

Particle size: the particle size of the micro-proppant is designed to be 200~400 mesh (38~74 μm), and the particle screening is realized through a high-precision turbine airflow classifying device and laser particle size detecting device.

(2)

Dispersibility and compatibility: Surface modification and addition of dispersant are used to avoid the formation of clumps after mixing with liquids, to realize uniform mixing with slickwater of different viscosities, and to prevent adsorption on the walls of the fracture through the introduction of negative charge by graphene in the powder. The micro-proppant shows good dispersion performance in clear water, low-viscosity, and medium-high viscosity slickwater (Figure 6). The viscosity test was carried out by mixing the test liquids, and the results are shown in

Table 3. Mixed liquid viscosity test.

Case	Viscosity of Slick Water (mPa·s)	Viscosity of Mixed Liquid (mPa·s)
#7	1	1.05
#8	2.72	2.75
#9	15	15.2

. In general, the viscosity of the mixed liquids did not change significantly after the addition of micro-proppant.

On the basis of clarifying the sequence of “70/140 mesh + 40/70 mesh proppant pumping”, how to introduce micro-proppants under the existing pumping mode in order to better utilize its advantages, we need to carry out the static settling velocity test of proppants of different particle sizes, single particle size proppant transport experiments in different structural fracture as well as the mixed pumping experiments of micro-proppants with 70/140 mesh and 40/70 mesh proppants (Figure 7) to further clarify the dynamic transport characteristics of micro-proppants.

3.2.2. Proppant Static Settlement Experiment

The static settling velocities of proppants with different particle sizes were tested in slickwater with a viscosity of 2.5 mPa·s (Figure 8). The test method involved pouring equal masses of proppants with different particle sizes into a graduated cylinder. The time taken for all proppants to fully settle to the bottom was recorded as the settling time. The settling velocity was then calculated by dividing the fluid height by the settling time. The test results showed that the settling velocity of the micro-proppant was approximately 0.8 cm/min—only 1/13 that of the 40/70 mesh proppant and 1/8 that of the 70/140 mesh proppant. This indicates that the micro-proppant exhibits a significantly lower tendency to settle, possesses excellent suspension performance, and is capable of following the carrying fluid into the deeper sections of the fracture.

3.2.3. Proppant Placement Experiment

(1)

Single particle size proppant

In this case, a single particle size proppant is injected using low-viscosity fluid (2.5 mPa·s) throughout the entire process, and the experimental program is shown in

Table 4. The experimental scheme of single particle size proppant placement.

Case	Pumping Rate (L/min)	Liquid Viscosity (mPa·s)	Proppant Particle Size (mesh)	Proppant Concentration (kg/m3)
#10	10	2.5	40/70	200
#11			70/140
#12			100/200
#13			200/400(micro-proppant)

below. Figure 9 shows the proppant distribution results within the fracture for different particle sizes. The dimensions of the double-branch fracture are as follows: the main fracture is 3.36 m long and 6 mm wide; the front branch fracture is 1.1 m long and 3 mm wide; and the rear branch fracture is 1.05 m long and 1 mm wide. All fractures have a uniform height of 0.3 m. As the particle size decreases, the proppant is able to migrate deeper into the fracture, leading to a transition in its distribution within the main fracture from being higher at the front and lower at the back to a more uniform pattern. Additionally, the difference in the proppant placement area between the front and rear branch fractures gradually decreases.

Further statistics on the difference in the percentage of placement area between the front and rear branch fracture (Figure 10) showed that the difference in the percentage of placement area between the front and rear branch fracture gradually decreased (42% → 21% → 16% → 2%), which further indicated that the micro-proppant is able to enter the branch fracture efficiently, and improve the phenomenon of non-uniform distribution of proppants in the fracture. However, when only a micro-proppant is used, the sand bank height in the main fracture remains relatively low, leading to insufficient fracture support and difficulty in maintaining high conductivity. Therefore, it is necessary to further investigate the placement characteristics of proppants when the micro-proppant is mixed and co-injected with a conventional size proppant.

(2)

Mixed proppant

The 40/70 mesh and 70/140 mesh are used to mix with the micro-proppant for pumping, and the experimental scheme is shown in

Table 5. The experimental scheme of mixed proppant placement under different ratios.

Case	Pumping Rate (L/min)	Liquid Viscosity (mPa·s)	Proppant Particle Size and Ratio	Proppant Concentration (kg/m3)	Experimental Fracture Shape
#14	10	2.5	70/140 mesh: 200/400 mesh = 8:2	200	Double-branch fractures/Multi-level fracture
#15			70/140 mesh: 200/400 mesh = 6:4		Double-branch fractures/Multi-level fractures
#16			70/140 mesh: 200/400 mesh = 4:6		Double-branch fractures
#17			40/70 mesh: 200/400 mesh = 6:4		Double-branch fractures/Multi-level fractures

. Figure 11 and Figure 12 show the results of proppant distribution under different ratios in the double-branch fracture, and when 70/140 mesh proppant and micro-proppant were mixed with different ratios, the proppants in the fracture are laid gently and distributed evenly, but the degree of filling is different. The overall filling degree of the fracture is the highest when the ratio is 8:2. When 40/70 mesh proppant and micro-proppant are mixed, compared with the single-size 40/70 mesh proppant, the equilibrium height of the sand bank decreases, the proppant placement in the front branch becomes more uniform, while the amount of proppant entering the rear branch increases.

Figure 13 shows the results of proppant distribution under different blending ratios within the multi-level fracture system, a horizontal fracture with an aperture of 1 mm is located in the middle to rear section of the main fracture. For clarity, it is represented by a blue box in the figure, and the enlarged top view is pointed out by a red arrow. It can be observed that when 70/140 mesh proppant is mixed with micro-proppant, a higher proportion of micro-proppant leads to a slower overall settling velocity of the mixture. The proppant distribution within the fracture becomes more uniform, and a fuller sand bank is formed in the horizontal fractures. However, the sand bank height in the main fracture decreases as the proportion of micro-proppants increases. When the ratio is 8:2, the filling degree of the second-level branch fracture reaches the highest 32%, and the micro-proppant and 70/140 mesh proppant are interleaved with the class of stripe-like distribution in the horizontal fracture. When 40/70 mesh proppant is mixed with the micro-proppant, there is no obvious increase in the filling degree of the second-level branch fracture, but the proppant placement in main fracture and the first-level branch fracture is gentler, and only the white sand surface formed by the micro-proppant is observed in the horizontal fracture. In summary, the micro-proppant is beneficial to the uniform distribution of 70/140 mesh proppant in the fracture, increases the filling degree of the horizontal fracture, and has a poor effect on the 40/70 mesh proppant.

The above experimental results further confirm that the micro-proppant has the following advantages compared with conventional proppants: (1) It can enter the distal secondary fracture and significantly improve the phenomenon of non-uniform distribution of proppants in the fracture. (2) It can enter the horizontal fracture in large quantities and form support. Combined with the micro-proppant and conventional proppant mixing and transport experimental results, it is considered that the mixed injection of micro-proppant and 70/140 mesh proppant is beneficial to both the uniform distribution and adequate fracture filling. The optimal ratio is 70/140 mesh/ micro-proppant = 8:2, and the proportion of micro-proppant can be appropriately reduced according to the construction situation in the field construction. To maximize the goal of “proppant reaches wherever fluid reaches,” it is recommended that the micro-proppant be co-injected continuously throughout the 70/140 mesh proppant stage.

4. Field Application

4.1. Basic Information of Test Well

L-X1 well is located in LZ block Yanggaosi tectonic group; L-X1 well is a deep horizontal development well deployed in the lower half branch of the platform, the drilling results show that the drilling encounters are I, II type reservoirs, the horizontal section buried depth of 3637 m~3909 m, the length of the construction section is 1637 m, the stratigraphic pressure coefficient is about 2.15, the maximum horizontal principal stress is 99.6 MPa, the minimum horizontal principal stress is 87.3 MPa, horizontal stress difference of about 12.3 MPa, vertical stress of 91.2 MPa, Poisson’s ratio of 0.23, Young’s modulus of 36.2 Gpa, comprehensive brittleness index of 64.8%, the development of micro fractures, the formation of complex fracture network is moderately difficult. L-X1 well adopts long-stage multi-cluster stimulation mode 2.0, the length of the main stage is 72~104 m, the average stage length is 81.8 m, and the single stage is divided into 7~12 clusters. L-X1 well adopts differentiated perforation mode, with perforation density ranging from 6 to 20 holes per m, and part of the stage adopts directional upward perforation.

4.2. Proppant Pumping Procedure

L-X1 well is at risk of fracture hit and casing deformation, the main stage design pumping rate 16 m³/min, risk stage design pumping rate 14 m³/min, inter-cluster temporary plugging is applied once in both the main and the risk stages. In stage 17, micro-proppant was employed based on prior evaluations of dynamic proppant transport. The following pumping strategy was adopted: a moderate-to-high viscosity fluid (20–30 mPa·s) was used during the pad stage to initiate fracture creation, followed by a low-viscosity fluid (3 mPa·s) throughout the proppant-carrying stages, and finally a high-viscosity fluid (30 mPa·s) was applied during the flush stage. The ratio of 70/140 mesh quartz sand to 40/70 mesh ceramic proppant was 7:3, and micro-proppant was mixed with 70/140 mesh quartz sand in a 2:8 ratio. Pumping commenced with the micro-proppant/quartz sand mixture. A direct switch to tailing-in with ceramic proppant followed, without an intermediate flush stage. The detailed design scheme of proppant pumping procedure is provided in

Table 6. Proppant pumping procedure design.

Sequence Number	Stage	Liquid Viscosity (mPa·s)	Net Liquid Volume (m3)	Cumulative Liquid Volume (m3)	Pumping Rate (m3/min)	Sand Concentration (kg/m3)	Stage Sand Mass (t)	Cumulative Sand Mass (t)	Remarks
1	Slickwater	30	14	14	2	0	0	0	Pad Stage
2	HCl	\	14	28	2	0	0	0
3	Slickwater	3	168	196	2~14	0	0	0
4	Slickwater slurry	3	98	294	14	80	7.84	7.84	70/140 mesh quartz sand mixed with micro-proppant
5	Slickwater slurry	3	126	420	14	100	12.6	20.44
6	Slickwater slurry	3	98	518	14	120	11.76	32.2
7	Slickwater slurry	3	56	574	14	140	7.84	40.04
8	Slickwater slurry	3	56	630	14	160	8.96	49
9	Slickwater slurry	3	56	686	14	180	10.08	59.08
10	Slickwater slurry	3	56	742	14	200	11.2	70.28
11	Slickwater	3	98	840	14	0	0	70.28	40/70 mesh ceramic proppant
12	Slickwater slurry	3	70	910	14	80	5.6	75.88
13	Slickwater slurry	3	63	973	14	100	6.3	82.18
14	Slickwater slurry	3	56	1029	14	120	6.72	88.9
15	Slickwater slurry	3	49	1078	14	140	6.86	95.76
16	Linear Gel	30	30	1108	14	0	0	95.76	Flush Stage
17	Slickwater	3	70	1178	14	0	0	95.76

.

4.3. On-Site Implementation

Stage 17 of L-X1 well was successfully completed. The initial treatment pressure was 48.5 MPa, with a peak pressure of 80.3 MPa and a typical operating pressure ranging between 72 and 78 MPa. The maximum pumping rate was 14 m³/min. The cumulative injected fluid volume was 1058.79 m³, with the low-viscosity slickwater of 695.79 m³ (accounting for 65.7% of the total volume), and the medium-viscosity slickwater of 363 m³. The cumulative proppant mass reached 110.13 t, consisting of 61.24 t of 70/140 mesh quartz sand, 38.89 t of 40/70 mesh ceramic proppant, and 10 t of micro-proppant. Due to the obvious upward trend in pressure during the early pumping period, the slug sand addition mode was implemented (a1~a3), and the reason for the upward trend of pressure was analyzed to be that the micro-proppant entered into the micro fracture and was effectively propped, which increased the net pressure in the fracture. During the middle stage, a conventional single-mesh sand addition mode was used to complete the stage, with the highest sand concentration of 70/140 mesh quartz sand mixed with micro-proppant being 160 kg/m³, the highest sand concentration of single 70/140 mesh quartz sand being 240 kg/m³, and the highest sand concentration of 40/70 mesh ceramic proppant being 160 kg/m³. The construction curves for stage 17 of L-X1 well are shown in Figure 14.

4.4. Evaluation of Micro-Proppant Effect

After the completion of the whole well fracturing construction of L-X1 well, artificial fracture parameters for each stage were calculated using shut-in pressure decline curve and the G-function analysis. The surface areas of the main and secondary fracture for each stage were compared, and their relationships with individual stage gas production under different production regimes were evaluated (Figure 15).

Production test results show that the main fracture surface areas for stages 15, 16, and 17 were 22,808.5 m², 32,536.7 m², and 24,689.6 m², respectively. The corresponding secondary fracture surface areas were 52,537.2 m², 71,873.7 m², and 16,404.8 m². Compared with stages 15 and 16, which share similar geological conditions, stage 17 exhibited a comparable main fracture surface area but significantly smaller secondary fracture surface area—only 30% and 23% of those in stages 15 and 16, respectively.

Under all three production regimes, stage 17 exhibited higher gas production than stages 15 and 16. Under the 5 mm nozzle regime, gas production from stage 17 reached 4.72 × 10³ m³/d, comparable to stage 16 at 4.45 × 10³ m³/d, while stage 15 produced only 1.17 × 10³ m³/d. The production from stage 17 was four times that of stage 15. Under the 6 mm nozzle regime, gas production from stages 15, 16, and 17 was 4.04 × 10³ m³/d, 5.25 × 10³ m³/d, and 9.71 × 10³ m³/d, respectively. Stage 17 output was 2.4 times that of stage 15 and 1.8 times that of stage 16. Under the 6.5 mm nozzle regime, the gas production of stages 15 and 16 was approximately equal, 6.77 × 10³ and 6.71 × 10³ m³/d, respectively, while stage 17 reached 10.7 × 10³ m³/d, which was 1.58 and 1.59 times that of stages 15 and 16, respectively.

This production disparity is attributed to differences in proppant placement within the secondary fractures. Although stage 17 had a smaller volume of secondary fractures, the use of micro-proppant in combination with 70/140 mesh proppant enabled effective entry into micro fractures, significantly increasing the effective propped fracture volume and thus gas production. In contrast, stages 15 and 16, which did not utilize micro-proppant, developed more secondary fractures but suffered from the “fluid entry without proppant” phenomenon, resulting in lower production.

5. Conclusions

This study integrates the transport and distribution characteristics of proppants in multi-level fractures to clarify the theoretical basis for achieving extensive fracture propping in deep shale formations. The critical elements required for effective support are identified, and a technical strategy based on the combined injection of micro-proppants and conventional proppants is established.

Standard performance tests were conducted on micro-proppants to evaluate their physical properties. Using a proppant transport experimental platform, targeted tests were carried out under various proppant mixing ratios and injection sequences. Optimal combinations and timing for 40/70 mesh, 70/140 mesh, and 200/400 mesh micro-proppants were determined. The results show that conventional proppants mainly support the main fracture and first-level branch fractures. In contrast, micro-proppants that match the width of micro fractures are essential for achieving extensive and uniform proppant placement throughout complex fracture networks in deep shale. The 200/400 mesh micro-proppants exhibit a sedimentation velocity of only 0.8 cm/min—approximately 1/13 and 1/8 that of 40/70 and 70/140 mesh proppants, respectively. When co-injected with 70/140 mesh proppant, the micro-proppant significantly increases the proppant entry ratio and transport distance in both vertical and horizontal branch fractures, while markedly improving distribution uniformity.

Using the L-X1 well as a case study, a detailed proppant pumping schedule was designed and implemented in the field. Field application and post-fracturing production data confirmed that this approach effectively overcame the challenge of poor proppant entry into micro fractures, resulting in significantly enhanced stimulation performance.