Hydraulic fracturing response characteristics and differentiated pumping parameter optimization of coal reservoirs in the Yushe-Wuxiang block, Qinshui Basin, China

1 Introduction

As low-permeability reservoirs, coal reservoirs rely on hydraulic fracturing as an indispensable stimulation measure for efficient coalbed methane (CBM) development (Palmer, 2010). The effectiveness of hydraulic fracturing is influenced not only by geological factors but also significantly by engineering factors (Liu X. et al., 2025). Therefore, analyzing the fracturing responses under different injection rates, displacements, and fracturing fluid types, and constructing an appropriate operational parameter system on this basis, is of great practical significance for improving hydraulic fracturing effectiveness.

Scholars in China and abroad have conducted extensive theoretical analyses, physical simulations, and field empirical studies on the influence of fracturing operational parameters on fracture propagation and stimulation effectiveness, forming a series of fundamental understandings (Zhang et al., 2018). Regarding injection rate control, displacement is the core engineering parameter determining fracture geometry and complexity (Liu S. et al., 2025): at low injection rates, fracturing fluids mainly extend slowly along natural fractures and bedding planes, with strong fracture controllability but a limited stimulation range; medium-to-high displacement can drive fractures toward the far end by increasing fluid kinetic energy, widening fracture aperture, and activating natural micro-fractures to form complex fracture networks, but excessively high displacement increases operational control difficulty and the risk of sand screenout. True triaxial experiments have confirmed that the initiation pressure of coal is positively correlated with injection rate (Jiang et al., 2016; Huang et al., 2025), and that this effect is jointly modulated by stress contrast and the degree of natural fracture development. Proppant and sand ratio also matter: proppant volume and sand ratio directly determine the long-term fracture conductivity and stimulation effect (Ahamed et al., 2021). At low sand ratios, proppants mainly fill the distal fractures, and fluid energy is mainly expended in driving fracture extension; medium-to-high sand ratios increase the net pressure within fractures, promote lateral fracture extension, and effectively prop the fracture walls against closure. However, coal reservoirs with well-developed cleats and low rock strength are prone to near-wellbore screenout at excessively high sand ratios, which instead constrains vertical fracture extension (Zhang et al., 2018). The fracture-forming mechanisms and applicable scenarios of fracturing fluids with different viscosities differ significantly (Barati and Liang, 2014; Yang et al., 2020). High-viscosity cross-linked fluids transmit pressure efficiently and easily form long and wide main fractures with strong proppant-carrying capacity, but their ability to activate natural fractures is weak; low-viscosity active water and slickwater easily infiltrate natural cleats and micro-fractures, promoting branch fracture development and fracture network complication, but their proppant-carrying capacity has an obvious upper limit, making it difficult to form high-conductivity main channels (Cipolla et al., 2009).

In addition to the pumping parameters discussed above, the perforation method, which serves as the engineering interface connecting the wellbore with the coal reservoir, also markedly influences fracturing effectiveness. Perforation method and parameters also markedly influence fracturing effectiveness. Oriented perforating along the maximum horizontal principal stress direction reduces the breakdown pressure and near-wellbore fracture tortuosity, while higher perforation density increases the flow area but may induce competing initiation points that weaken uniform fracture propagation (Behrmann and Elbel, 1991). In the Qinshui Basin, cased-hole perforated completion is the dominant mode owing to its good wellbore stability and adaptability to selective fracturing, and an appropriate match between perforation design and pumping parameters is a prerequisite for achieving the expected stimulation effect (Liu X. et al., 2025). CBM wells mainly adopt two completion modes: openhole (barefoot or slotted liner) completion, which provides a large flow area but offers poor wellbore stability and is unsuitable for strongly broken coal intervals, and cased-hole perforated completion, which is the most widely used mode in the Qinshui Basin owing to its good wellbore stability and its adaptability to the selective fracturing of multiple coal seams. For perforated completions, parameters such as perforation density, phase angle, hole diameter, and penetration depth determine the flow area between the wellbore and the reservoir, thereby affecting the near-wellbore friction loss and the breakdown pressure; oriented perforating along the direction of the maximum horizontal principal stress can reduce the fracture initiation pressure and the near-wellbore fracture tortuosity, and is therefore conducive to smooth fracture extension. A reasonable match between the perforation method and the pumping parameters is thus a prerequisite for achieving the expected stimulation effectiveness; however, the perforation methods applied in the Yushe-Wuxiang block have not yet been systematically compared, which is beyond the scope of this study.

Overall, existing studies have clarified the influence of individual operational parameters on fracturing effectiveness, but obvious deficiencies remain. On the one hand, most studies focus on the independent effects of single parameters; research on multi-parameter collaborative matching mechanisms is insufficient, and a differentiated parameter matching system based on reservoir geological classification is lacking. On the other hand, existing results mostly target conventional sandstone or high-rank coal reservoirs; systematic studies addressing the medium-rank coal, low horizontal stress contrast, and strong heterogeneity characteristics of the Yushe-Wuxiang block are scarce, leaving field operational parameter design without a targeted basis.

In view of this, taking the No. 3 coal seam of the Yushe-Wuxiang block as the main research object, this paper systematically analyzes reservoir fracturing response characteristics under different injection rates, sand ratios, and fracturing fluid types based on field fracturing data, and identifies the dominant geological and engineering factors controlling fracture propagation. Addressing the engineering characteristics of this block—soft coal reservoirs with well-developed natural fractures that cause complex fracturing fluid leakoff—reservoir type division is carried out in combination with reservoir geological characteristics, and a differentiated operational parameter system and optimization design scheme are constructed. It should be noted that, limited by data availability, direct observation of subsurface fracture geometry in the field is difficult, and the sand ratio is therefore adopted here as an indirect indicator of fracture propagation. The results provide a scientific basis for the fracturing design of CBM wells in this block and important engineering guidance for improving fracturing stimulation of low-permeability coal reservoirs and achieving efficient CBM development.

2 Regional geological setting

The Yushe-Wuxiang block is located in the central-eastern Qinshui Coalfield (Figure 1), administratively covering six counties of Jinzhong and Changzhi cities, with a total area of 1219.72 km² (Zhang et al., 2017). Structurally, the block lies on the eastern limb of the middle segment of the Qinshui synclinorium, forming as a whole a monocline striking NE and dipping NW, with strata dips generally of 3°–10° and locally up to 30° where disturbed by structures. Owing to the superposition of multiple tectonic movements from the Hercynian to the Himalayan, compressional and extensional deformations coexist in the area, with 14 synclines/anticlines, 29 faults, and a few collapse columns developed. The overall structural framework is relatively simple, dominated by gentle long-axis linear folds: over 80% of anticlinal limb dips are less than 10°, and more than 70% of structural amplitudes are less than 50 m. The faults are dominated by NE- and NNE-striking high-angle normal faults of generally small scale, with only six extending over 20 km, more numerous in the west than in the east. The gentle structural setting and the low horizontal stress contrast constitute the fundamental mechanical environment for hydraulic fracture propagation in this area.

The main coal-bearing strata in the area are the Carboniferous Taiyuan Formation and the Permian Shanxi Formation (Su et al., 2005). The Shanxi Formation is continental in origin, with an average thickness of 57.39 m, and its No. 3 coal seam is minable over most of the block. The main mined coal seams span a wide range of burial depths, with strong reservoir heterogeneity and well-developed natural cleats and fractures; these are the core geological factors controlling fracturing initiation pressure and fracture propagation in this area, and they also provide the geological basis for subsequent reservoir classification evaluation and differentiated fracturing parameter design.

3 Hydraulic fracturing response characteristics of coal reservoirs

The data used in this study were obtained from field fracturing operation records, well logging surveys, and well test interpretations of CBM wells in the Yushe-Wuxiang block, which were systematically collected and compiled by the research team during the CBM exploration and development project in this area. Based on field data collected from CBM wells in the Yushe-Wuxiang block (fracturing operation curves recording bottomhole pressure, displacement, and sand ratio; breakdown pressure statistics; in situ stress measurements; and density and acoustic logging and well test interpretation results), the hydraulic fracturing response characteristics of the coal reservoirs are systematically characterized from three dimensions—classification of fracturing curve morphologies, distribution of breakdown pressure (also termed rupture pressure), and correlation of characteristic parameters—to clarify the pressure evolution patterns under different geological and operational conditions and provide a field basis for the subsequent analysis of the dominant factors controlling fracture propagation and for fracturing parameter optimization.

3.1 Types of fracturing operation curves

Fracturing operation curves record the dynamic evolution of bottomhole pressure during injection; their morphological characteristics directly reflect the entire process of fracture initiation and propagation and constitute an important basis for judging reservoir conditions and operational status (Nolte and Smith, 1981). According to the evolution of pressure with time, combined with the synchronous changes in displacement and sand ratio, the fracturing curves in the study area can be divided into three types—normal, fluctuating, and complex—each corresponding to different reservoir conditions and formation mechanisms.

3.1.1 Normal type

The normal type is the typical fracturing curve under ideal reservoir conditions, exhibiting a clear three-stage evolution (Figure 2): the first stage is the initiation and pressurization segment, in which the bottomhole pressure rises rapidly to the breakdown pressure peak as the injection displacement increases steadily, corresponding to the first macroscopic formation breakdown; the second stage is the stable extension segment, in which the pressure briefly falls back after breakdown and then fluctuates slightly and steadily around the extension pressure while the fracture propagates continuously and uniformly toward the far end; the third stage is the pump-stop depressurization segment, in which the pressure declines rapidly after the operation ends.

Such curves show no large abnormal fluctuations, indicating that no obvious screenout, abrupt leakoff changes, or other anomalies occurred during fracturing; they mostly occur in well intervals with good reservoir homogeneity, undeveloped natural fractures, and gentle structure.

3.1.2 Fluctuating type

Fluctuating-type curves display continuous pressure undulation throughout the operation, with vigorous multi-peak, multi-trough fluctuations (Figure 3); the differences between peaks and troughs are significant, and the fluctuations do not fully correspond to the manual adjustments of displacement and sand ratio. The causes are multiple: first, natural fractures and bedding planes are repeatedly connected and opened during fracturing, causing frequent changes in fracture conductivity and abrupt pressure drops and rebounds; second, temporary screenout occurs within local fractures and is subsequently flushed open by high-pressure fluid, producing repeated pressure rises and falls. Such curves usually correspond to well intervals with strong reservoir heterogeneity and complex fracture systems (Gale et al., 2007).

3.1.3 Complex type

As shown in Figure 4, the complex-type curve is characterized by a single abrupt pressure event superimposed on an otherwise stable evolution. During the initial pressurization stage, the bottomhole pressure rises steadily without obvious fluctuation, similar to the normal-type curve; once the propagating fracture encounters a local anomaly-such as a fault, a collapse column, an abrupt lithologic change, or a large natural fracture-the pressure drops or jumps abruptly within a short time window; thereafter, the fracture rapidly adjusts to a new propagation equilibrium and the pressure re-stabilizes at a new level, remaining relatively steady until pump stop. Unlike the fluctuating-type curve, which exhibits repeated multi-peak and multi-trough undulation throughout the entire operation, the complex-type curve shows no continuous periodic fluctuation, and the single abrupt event is the only distinguishing feature. This morphological contrast is clearly visible in Figure 4, where the pressure trace is flat over most of the injection period and contains only one pronounced step-like jump.

In general, the three types of fracturing curves correspond to different reservoir development characteristics and operational statuses; curve morphology can be used to preliminarily judge the degree of natural fracture development and the strength of reservoir heterogeneity, providing an intuitive basis for the dynamic adjustment of field operations.

3.2 Distribution of fracture pressure

Fracture pressure is the most critical initiation parameter of hydraulic fracturing, defined as the minimum bottomhole pressure required to create the first macroscopic fracture in the coal around the wellbore (Yu et al., 2022). Its magnitude directly determines equipment selection, fracturing fluid system design, and operational risk control, and it is a fundamental parameter in fracturing engineering design.

According to field measurements, the burial depth of the No. 3 coal seam ranges from 1050 to 1650 m, corresponding to breakdown pressures of 20–62 MPa. A positive correlation exists between breakdown pressure and burial depth (Zhang et al., 2023): with increasing depth, the overburden load and the in situ stress level rise, and the bottomhole pressure required for tensile failure of the coal increases accordingly (Figure 5). However, the fitted correlation coefficient is only 0.4117, a moderate correlation—indicating that breakdown pressure is also jointly influenced by non-stress factors such as the mechanical properties of the coal itself and the degree of natural fracture development. Specifically, in terms of coal mechanical properties, a higher tensile strength and Young’s modulus mean that a greater tensile stress is required to induce failure, which raises the breakdown pressure, whereas soft coal intervals with low strength break down at lower pressures; in terms of natural fractures, intervals with well-developed cleats and fractures contain pre-existing planes of weakness along which fractures initiate at markedly lower pressures, while the accompanying increase in fluid leakoff raises the pressure required to sustain propagation. The strong spatial heterogeneity of these two factors within the block (see the distributions of Young’s modulus, tensile strength, and permeability in Section 4.1) is superimposed on the depth-dependent stress background, which explains the considerable scatter of breakdown pressure at the same burial depth and the only moderate overall correlation shown in Figure 5.

In addition to burial depth, coal mechanical properties and natural fracture development exert significant controls on breakdown pressure. Higher Young’s modulus and tensile strength increase the stress required for tensile failure, thereby raising the initiation pressure, whereas soft coal with low mechanical strength breaks down more readily (Pan et al., 2013). Well-developed cleats and natural fractures provide pre-existing planes of weakness that reduce the effective initiation pressure and enhance fluid leakoff, both of which modulate the breakdown pressure (Gale et al., 2007). This may reduce the correlation between burial depth and fracture pressure.

3.3 Engineering factors influencing fracture propagation

The stimulation effectiveness of hydraulic fracturing ultimately depends on fracture geometry and conductivity, while fracture length, aperture, height, and network complexity are jointly controlled by geological conditions and engineering parameters (Jiang et al., 2016). Since direct monitoring of the true fracture geometry during field operations is difficult, this study selects the operational sand ratio—closely related to the scale and aperture of fracture propagation—as an indirect indicator (Maxwell et al., 2010), and focuses on the influence of three core engineering parameters—injection rate (displacement), proppant dosage, and fracturing fluid type—on fracture propagation, clarifying the control effect of each parameter. It should be pointed out that the sand ratio adopted here to indirectly characterize fracture propagation is affected not only by fracture propagation conditions but also by artificial operational conditions, of which the main interference factors are: (1) real-time manual adjustment of the displacement and the sand-feeding program according to the pressure response during the operation, in particular the reduction of the sand ratio or the suspension of sand feeding to avoid screenout when the pressure rises abnormally; (2) the staged, stepped sand-adding procedures commonly used in field practice, which make the recorded sand ratio depend partly on the preset program rather than purely on fracture propagation; and (3) equipment capacity constraints such as the rated pump displacement and the sand-mixing capacity. Consequently, the sand ratio is essentially a semi-quantitative indicator that couples the fracture propagation state with operational decisions, and the correlations established in this section should be regarded as statistically meaningful trends rather than deterministic relationships. In future work, cross-validation with microseismic monitoring and post-fracturing production data will be required to decouple these influences and to further verify the reliability of this indicator. It should be further emphasized that, among the above interference factors, the real-time manual adjustment of the sand-feeding program is the most dominant source of uncertainty: field operators frequently reduce the sand ratio or even suspend sand feeding temporarily when the bottomhole pressure shows an abnormal rise, which directly truncates the statistical sand ratio and weakens its correlation with the actual fracture aperture. In addition, the staged sand-adding procedure means that the sand ratio at any given moment reflects not only the fracture’s proppant-carrying capacity but also the predetermined stage design, so the same fracture condition may yield different recorded sand ratios under different operational schedules. Therefore, the correlations between engineering parameters and sand ratio presented in this section should be interpreted as statistical trends under the specific operational conventions of the study area, and direct transfer of the quantitative thresholds to other blocks with different operational habits is not recommended.

3.3.1 Injection rate

Injection rate is the core engineering parameter controlling fracture propagation behavior; it influences fracture morphology by altering the bottomhole net pressure and the distribution of fluid energy (Liu S. et al., 2025). Statistics of the operations in the No. 3 coal seam show that the injection rate and the sand ratio are significantly positively correlated, with a high degree of correlation (Figure 6). The results indicate that as the injection rate increases, the sand ratio that the operation can sustain increases synchronously, corresponding to increases in fracture length, aperture, and network complexity (Jiang et al., 2016).

It is equally noteworthy, however, that significant variations in sand ratio still exist under the same injection rate. For example, when the injection rate is 8 m³/min, the sand ratio varies from 1% to 6% (Figure 6), indicating that increasing the injection rate does not monotonically enhance fracture propagation. Previous studies have shown that at low injection rates, the fracturing fluid mainly undergoes leakoff and seepage along natural fractures and bedding planes; fractures extend slowly, the stimulation range is limited, but the fracture morphology is highly controllable; at medium injection rates, the fluid kinetic energy can effectively activate the natural cleat and micro-fracture system of the coal, easily forming complex fracture networks with well-developed branch fractures (Liu S. et al., 2025); at high injection rates, the high-pressure fluid drives fractures to extend rapidly toward the far end, significantly widening the fracture aperture and more readily creating new tensile main fractures (Yu et al., 2022), but the operational control difficulty and the screenout risk also rise accordingly. Therefore, parameter design must be based on the actual reservoir characteristics.

3.3.2 Proppant dosage

Proppant is the core medium maintaining the long-term fracture conductivity; its dosage not only determines the fracture filling effectiveness and the final conductivity, but also modulates fracture propagation in reverse by altering the net pressure within the fracture (Cipolla et al., 2009). The proppant volume and the sand ratio of the No. 3 coal seam show a strong positive correlation (Figure 7). Overall, as the proppant volume increases, the fracture width, extension length, and network complexity all increase synchronously; proppant dosage is the key parameter determining the stimulation intensity and the long-term production enhancement (Ahamed et al., 2021). At low proppant dosage, the propping effect on the fracture walls is weak, and the fluid energy is mainly used to drive fracture extension; at medium dosage, proppant particles occupy fracture space, increase the flow resistance, and push the fracture to extend laterally; at high dosage, the net pressure within the fracture rises significantly, promoting preferential fracture development in the width direction and effectively enhancing the conductivity of a single fracture.

3.3.3 Fracturing fluid type

The fracturing fluid performs the core functions of transmitting formation pressure and carrying proppant into the fractures, and viscosity is its core physical property controlling fracture morphology (Barati and Liang, 2014). High-viscosity fracturing fluids transmit pressure efficiently and carry proppant strongly, easily forming long and wide main fractures, but they can hardly infiltrate microscale natural fractures (Barati and Liang, 2014); low-viscosity fracturing fluids have strong leakoff, easily enter natural cleats and micro-fractures to activate branch fractures, facilitating the formation of complex fracture networks, but their proppant-carrying capacity has an obvious upper limit (Huang et al., 2019).

Fracturing operations in the study area mainly use three fracturing fluid systems—cross-linked gel, slickwater, and active water—with progressively decreasing viscosity (Huang et al., 2019). Cross-linked gel forms a three-dimensional network structure by chemically cross-linking linear polymers such as guar and hydroxypropyl guar, and it is a high-viscosity fracturing fluid; slickwater reduces pipe-string friction by adding polyacrylamide-type friction-reducing polymers, and it is a low-to-medium viscosity fracturing fluid that balances low friction with a certain proppant-carrying capacity; active water uses surfactants as the core additive, with a viscosity close to that of clean water, improving the reservoir wettability by reducing the interfacial tension, and it is suitable for high-permeability or strongly sensitive reservoirs (Yang et al., 2020).

Statistical results show no significant linear correlation between the fracturing fluid type and the sand ratio in the No. 3 coal seam, but the applicable sand ratio ranges of the three systems differ markedly (Figure 8): cross-linked gel has the strongest proppant-carrying capacity and the widest applicable sand ratio range, of 2%–16%; that of slickwater is 6%–15%; and active water, with the lowest viscosity and a limited proppant-carrying capacity, is concentrated at 1%–10%. This distribution is fully consistent with the viscosity-controlled proppant-carrying mechanism: high-viscosity systems can sustain higher sand ratios and are better suited to large-scale proppant fracturing operations (Cipolla et al., 2009).

4 Establishment of a coal seam pumping parameter optimization and evaluation technique

4.1 Reservoir geological characteristics

The geological factors influencing coal reservoir fracturing mainly include four categories: reservoir spatial distribution characteristics, in situ stress characteristics, rock mechanical characteristics, and physical property characteristics (Jiang et al., 2016).

4.1.1 Spatial distribution characteristics of the coal reservoir

The spatial distribution of coal reservoirs is mainly controlled by burial depth and thickness. Burial depth directly affects the in situ stress level, reservoir mechanical properties, and physical properties, and it is the core parameter determining fracturing difficulty (Zhang et al., 2017). The burial depth of the No. 3 coal seam in the study area ranges from 395.22 to 1580.86 m, averaging 1289.84 m, shallowing overall from northwest to southeast (Figure 9a). Burial depths differ significantly: except for the southeastern corner of the block, the reservoir burial depth exceeds 1000 m overall and is relatively great.

The No. 3 coal seam is 0.67–4.38 m thick, averaging 2.01 m. Spatially, the thickness is below 2 m over most of the area and exceeds 2 m only in parts of the southwest; it fluctuates significantly but generally shows a west-thick, east-thin trend (Figure 9b). When fracturing thin coal seams, fracture propagation is constrained by the surrounding rock and remains limited, and the influence of fractures growing out of the seam (fracture penetration) must be considered.

4.1.2 In situ stress characteristics of the coal reservoir

In situ stress controls the hydraulic fracturing effectiveness of the reservoir jointly through the stress magnitude and the horizontal stress contrast (Zhang et al., 2023). The higher the stress level, the greater the formation resistance that must be overcome for fracture propagation during fracturing, and the higher the reservoir initiation pressure, which increases the operational difficulty and constrains the fracture extension scale (Ma et al., 2026). In terms of the planar distribution of in situ stress of the No. 3 coal seam (Figure 10), the maximum horizontal principal stress ranges from 6.14 to 23.92 MPa and the minimum horizontal principal stress from 5.28 to 22.12 MPa. The stress distribution shows obvious zoning: the maximum and minimum horizontal principal stresses are generally high in the western and northwestern parts of the study area, exceeding 20 MPa, and decrease gradually toward the southeast; the southeastern margin is the low-stress zone of the whole area, where the maximum horizontal principal stress falls below 10 MPa. Overall, the pattern is high in the northwest and low in the southeast.

Horizontal stress contrast is another key parameter influencing fracture morphology (Ma et al., 2026); the horizontal stress difference of the No. 3 coal seam ranges from 0.85 to 1.79 MPa (Figure 11). Its planar distribution likewise inherits the northwest-high, southeast-low pattern: the stress difference in the northwest can exceed 1.7 MPa, decreasing to the 0.8–1.1 MPa range in the southeast. Both coal seams belong to low-horizontal-stress-contrast reservoirs. Under low stress contrast, hydraulic fractures do not readily extend along a single principal direction but more easily connect primary fractures, which favors the formation of complex fracture networks, enlarges the stimulated volume, and is quite favorable for the stimulation of CBM reservoirs (Warpinski and Teufel, 1987).

In summary, the uniformly low horizontal stress contrast (0.85–1.79 MPa) across the block constitutes a favorable stress condition for fracture network formation: it weakens the directional propagation of hydraulic fractures, promotes the activation and connection of natural cleats and micro-fractures along multiple directions, and thereby increases fracture network complexity and the stimulated reservoir volume; moreover, this favorable stress condition prevails throughout the study area, providing a block-wide geological basis for the differentiated pumping parameter optimization in Section 4.2.

4.1.3 Rock mechanical characteristics of the coal reservoir

The influence of coal reservoir mechanical properties on fracturing effectiveness is manifold; the principal factors are Young’s modulus, the key parameter of elastic deformation capacity, and tensile strength, the key fracture parameter (Pan et al., 2013). Young’s modulus characterizes the ability of a solid to resist deformation: the larger the modulus, the harder the rock, and the greater the pressure required to deform and fracture the rock during stimulation, which increases the operational and stimulation difficulty. Tensile strength is the ability of rock to resist tensile failure: the higher the strength, the greater the tensile stress required to fail the rock during fracturing.

According to elastic wave theory, combined with density logging and other data, the Young’s modulus of the coal seam can be calculated (Pan et al., 2013):where Δts is the formation shear wave slowness, μs/m; and Δtp is the formation compressional wave slowness, μs/m. This parameter of Equation 1 is obtained indirectly from logging data, and no laboratory core mechanical testing was carried out; the calculated results therefore contain a certain degree of logging interpretation error.

E = ρ Δ t s × 9.29 × 104 × 3 Δ t s 2 − 4 Δ t p 2 Δ t s 2 − Δ t p 2 (1)

Based on elastic wave theory, combined with density logging data, the rock mechanical parameters of the No. 3 coal seam in the study area were obtained. The results show that the Young’s modulus of the No. 3 coal seam is distributed between 1814.58 and 9911.38 MPa, with an average of 3757 MPa, indicating a soft coal reservoir overall (Figure 12a); the tensile strength of the coal is between 0.57 and 4.57 MPa, averaging 1.37 MPa (Figure 12b). In terms of planar distribution, the spatial patterns of Young’s modulus and tensile strength are highly similar: both are low in the north and northeast of the whole area, increase gradually toward the south, and form a clear local high anomaly near borehole 101 in the southeast. The overall pattern is low in the north, high in the south, with a local southeastern bulge (Figure 12a).

A lower Young’s modulus means that the coal undergoes plastic deformation more easily, while a low tensile strength means that a smaller pressure is required for tensile failure of the reservoir rock (Huang et al., 2025). The No. 3 coal seam in the study area is characterized overall by low Young’s modulus and low tensile strength, indicating a weak resistance to crushing of the coal reservoir rock: under the external force of hydraulic fracturing it initiates and breaks easily, more readily activates the primary fractures within the coal body, favors the induction of complex fracture networks, and enlarges the stimulated volume (Pan et al., 2022).

4.1.4 Permeability and reservoir brittleness

Among the reservoir physical properties, the brittleness index and permeability are the two key indicators for evaluating the fracability of coal reservoirs, and they significantly affect the initiation, propagation, and stimulation effectiveness of hydraulic fracturing fractures (Pan and Connell, 2012). The reservoir brittleness is characterized by the relative brittleness index, which can be obtained by two calculation methods based on mechanical properties or mineral composition; this study uses the calculation method based on rock mechanical properties, i.e., the brittleness index is obtained from the elastic modulus and Poisson’s ratio (Rickman et al., 2008):

B I = B I E + B I μ 2 (2)

B I E = E − E min E max − E min (3)

B I = μ − μ min μ max − μ min (4)

This parameter of Equations 2–4 is also obtained indirectly from logging data. The brittleness index of the No. 3 coal seam in the study area ranges from 0.24 to 0.64, with an average of 0.46 (Figure 13a). The planar distribution shows obvious spatial differentiation of the brittleness index: it is generally high in the northern-central part of the study area, with high-value centers concentrated in the vicinity of ZK07-2 and ZK0701 in the central part, and decreases gradually toward the southwest; the overall pattern is high in the north, low in the south, with local enrichment in the center.

Permeability directly reflects the fluid flow capacity of the reservoir and indirectly affects the fracturing fluid leakoff and fracture propagation behavior (Pan and Connell, 2012). The permeability data are mainly derived from well test data of the study area. The permeability of the No. 3 coal seam is 0.02–1.53 mD, averaging 0.81 mD. The planar permeability distribution corresponds well with the brittleness index: the permeability in the northern area generally exceeds 1.0 mD (Figure 13b), and it decays rapidly from north to south, with large areas of the south in the low-permeability range below 0.2 mD.

4.2 Pumping parameter optimization scheme

Based on the above analysis of reservoir geological characteristics, the selection and optimization of pumping parameters in fracturing operations must comprehensively consider the coal reservoir geological conditions, physical properties, and mechanical characteristics to achieve the optimal stimulation effectiveness (Liu X. et al., 2025). This study selects four parameters—burial depth, coal thickness, reservoir brittleness, and permeability—for zoned discussion, mainly on the following grounds:

• The maximum and minimum horizontal principal stresses of the No. 3 coal seam show positive correlation trends with burial depth, and the overall stress level can indirectly reflect the operational pressure conditions through depth; burial depth is therefore selected to characterize the operational difficulty imposed by the stress. Although the horizontal stress contrast in this area varies spatially from 0.85 to 1.79 MPa, it is entirely within the low-contrast range, and the whole area is favorable for the formation of complex fracture networks; its spatial difference has limited power to discriminate reservoir fracability, and it is therefore not included in the evaluation indicators of this zoning.
• Coal thickness directly controls the vertical fracture propagation behavior, and thin coal seams are prone to fractures penetrating out of the seam during fracturing (Cao et al., 2020); the brittleness index reflects the potential of the coal to undergo brittle failure and form complex fracture networks (Rickman et al., 2008); and permeability determines the fracturing fluid leakoff characteristics—the three jointly control the fracturing effectiveness (Pan and Connell, 2012). The zoning thresholds (burial depth 1000 m and 1500 m; coal thickness 3 m; brittleness index 0.4; permeability 0.5 mD) were determined with comprehensive reference to the common empirical thresholds used in the CBM fracturing industry (Figure 14).

The pumping parameters are selected following these principles: as burial depth increases, the fracturing difficulty and the closure pressure rise, and higher pumping pressure and sand ratio are required to prop the fracture; thin coal seams require low displacement to control the fracture height and avoid penetration; high brittleness favors fracture network complication and allows a moderately lower sand ratio; and high permeability increases the fracturing fluid leakoff, so high-viscosity fracturing fluid should be used to reduce the leakoff (Ahamed et al., 2021; Cipolla et al., 2009). In addition, the pumping parameters are synergistic: a high sand ratio requires a matching high-viscosity fracturing fluid, and a high-viscosity fracturing fluid in turn requires a matching high pumping pressure. Parameter selection in this study prioritizes the fracturing success rate, followed by the inter-parameter synergy, and finally the fracture network complexity (Table 1).

TABLE 1. Characteristics of the No. 3 coal seam zones and recommended pumping parameters.

Constrained by the available data, the zonation evaluation does not include the degree of natural fracture development as an evaluation indicator, although natural fractures significantly alter the fracturing leakoff and fracture propagation behavior. The pumping parameters given in Table 1 are all qualitative empirical recommendations; owing to the lack of sufficient post-fracturing production data from corresponding wells, quantitative effectiveness verification has not yet been carried out. In actual field operations, the operational pumping parameters should be dynamically adjusted on the basis of this recommended scheme, combined with the results of drilling and logging and log-based fracture identification. In addition, the applicability of the proposed scheme merits discussion. The zoning framework and parameter-matching principles are methodologically transferable: the four evaluation parameters can be obtained from routine drilling, logging, and well test data in most CBM blocks, and the zoning thresholds can be recalibrated against local reservoir statistics. However, the specific parameter combinations recommended in Table 1 are conditioned on the medium-rank coal, low horizontal stress contrast, and strong heterogeneity of the Yushe-Wuxiang block, and direct application to reservoirs with markedly different stress regimes, coal ranks, or permeability levels is not recommended; for such reservoirs, re-analysis of local fracturing data is required. With the accumulation of post-fracturing production and microseismic monitoring data, quantitative verification and optimization of the recommended parameters will be carried out in future work. Furthermore, the limitations of the proposed optimization scheme should be explicitly noted. First, the zonation is based on four static geological parameters and does not incorporate dynamic operational feedback such as real-time pressure response, so the recommended parameters serve as a design baseline rather than a fixed operational recipe. Second, the scheme is calibrated against the medium-rank coal (Ro,max approximately 2.0%–2.5%), low horizontal stress contrast (0.85–1.79 MPa), and strong heterogeneity characteristics of the Yushe-Wuxiang block; for reservoirs with significantly higher stress contrast (>3 MPa), higher coal rank (anthracite), or markedly different permeability levels, the zonation thresholds and parameter combinations should be recalibrated using local field data. Third, the recommended pumping parameters are qualitative empirical guidelines, and their quantitative effectiveness—including the actual fracture geometry, stimulated reservoir volume, and post-fracturing gas production enhancement—has not yet been verified due to the lack of systematic post-fracturing production and microseismic monitoring data. Future work will focus on accumulating such data to quantitatively validate and iteratively optimize the proposed scheme.

5 Conclusion

Through statistical analysis of CBM well fracturing data and reservoir geological conditions in the Yushe-Wuxiang block, the hydraulic fracturing response characteristics were revealed and a differentiated pumping parameter optimization scheme was established. The main conclusions are as follows:

• Fracturing curves are classified into normal, fluctuating, and complex types, corresponding to homogeneous reservoirs, strongly heterogeneous reservoirs with developed natural fractures, and geological-engineering anomalies, respectively; curve morphology provides a basis for dynamic operational adjustment.
• Breakdown pressure of the No. 3 coal seam ranges 20–62 MPa and increases moderately with burial depth, jointly controlled by coal mechanical properties and natural fractures. Injection rate, proppant dosage, and fracturing fluid type are the core engineering parameters controlling fracture propagation.
• The No. 3 coal seam is divided into nine reservoir types based on burial depth, coal thickness, brittleness index, and permeability, with differentiated pumping parameter recommendations. This empirical scheme provides a reference for CBM fracturing design in the block and awaits further verification by production data.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Yunteng Wang, Universität für Bodenkultur Wien, Austria

Reviewed by: Chenliang Hou, Anhui University of Science and Technology, China

Dan Li, Xi’an University of Science and Technology, China