Reservoir Characteristics and Production Performance of Shale Oil in the MX Block, the Fengcheng Formation, Mahu Sag, China

1  Introduction

Despite the accelerating shift towards green and low-carbon energy within the global supply-demand structure against a backdrop of gradual economic recovery, fossil fuels continue to hold a significant position in the energy mix, and international demand for crude oil remains on an upward trajectory [1]. According to the oil and gas report released by China’s National Energy Administration, global crude oil production reached 4.8 billion tons in 2024, marking a year-on-year increase of 41.6 million tons, or 1%. Shale oil production accounted for 770 million tons, up 3% from the previous year, demonstrating robust growth momentum [2,3]. Thanks to two shale revolutions, the per-barrel cost of shale oil in the Permian Basin has fallen from $80 to $29 over the past decade. Its shale oil production accounts for approximately 70% of the global total, solidifying its position as the world’s top producer.

The Mahu Sag in the Junggar Basin is a world-renowned hydrocarbon-rich sag, having achieved significant breakthroughs in various types of oil and gas exploration and development. As oil and gas exploration continues to expand into unconventional domains, the hydrocarbon resources within the hydrocarbon-bearing mudstones of the Fengcheng Formation have gradually become a key research focus in the Mahu Sag [4,5]. The Fengcheng Formation of the Permian in the Mahu Sag represents a typical clay shale deposit. It constitutes the oldest alkaline lake sedimentation discovered in China to date, characterized by complex and diverse lithologies, pronounced heterogeneity, the development of distinctive alkaline minerals, and unique hydrocarbon source rocks [6,7]. To advance oil and gas exploration and development in the Fengcheng Formation of the Permian System within the Mahu Sag, China National Petroleum Corporation’s Xinjiang Oilfield strategically deployed an exploration well (MY1) in late 2018 (Fig. 1). Successful oil testing was achieved in October 2019 at the Third and Second members of Fengcheng Formation (P1f3 & P1f2), yielding a daily oil output of 20.78 t and a cumulative production of 1189.64 t over 64 days. This confirms the objective shale oil exploration and development potential of the Fengcheng Formation of the Mahu Sag. To further accelerate exploration progress in the northern slope area of the Mahu Basin and expand exploration coverage of the Fengcheng Formation, Xinjiang Oilfield deployed an exploration well (MX well) in the MX block of the Mahu Basin in May 2021. Successful oil testing in the P1f2 yielded high-yield results, confirming that the MX block’s shale oil resources possess excellent potential for commercial development. To clarify the shale oil development potential of the MX block and demonstrate its commercial viability, two high-angle wells (M1X, M2X) and one horizontal well (M3H) were successively deployed to evaluate the geological characteristics and production dynamics of the P1f2 shale reservoir within the MX block [8,9].

Figure 1: Location of the MX block

This work will comprehensively analyze the geological characteristics and production dynamics of the P1f2 shale reservoir in the MX block, based on geological exploration findings from the Permian Fengcheng Formation in the Mahu Sag. It will integrate geological static data and production dynamic data from three appraisal wells in the MX block to establish a prediction method suitable for estimating the ultimate recovery (EUR) of shale reservoirs. This study determines the EUR for the three evaluation wells targeting the P1f2 shale reservoir in the MX Block. It provides a basis for accurately assessing the reservoir’s development potential and economic viability, as well as for designing and formulating development plans.

2  Geological Setting

The Junggar Basin, located in northwest China, is rich in oil and gas resources and ranks as the second-largest petroleum basin in the region. Structurally, it lies at the junction of three tectonic domains: Kazakhstan, Siberia, and the Tarim Block. The Junggar Basin is subdivided into six primary structural units and 44 secondary structural units. The Mahu Sag, situated at the northwestern margin of the Junggar Basin, is one of its secondary structural units. Covering an area of approximately 4258 km2, it ranks among the basin’s most hydrocarbon-rich sags. The Mahu Sag underwent multiple phases of tectonic evolution, ultimately forming its present-day structural morphology characterized by a southeast-dipping monocline that tilts from the foreland into the basin.

This work investigates Block MX, which is structurally located in the northern slope zone of the Mahu Sag. This area is rich in hydrocarbon resources. Since the mid-1950s, nearly 70 years of exploration and development have uncovered oil reservoirs in the Jurassic Badaowan Formation, the Triassic Karamay Formation and Baikouquan Formation, as well as the Lower Permian Wuerhe Formation, Xiazijie Formation, and Fengcheng Formation. The fundamental topography of the northern slope zone of the Mahu Sag was formed during the late Permian Haixi period tectonic movement and finalized during the Yinzhi period. The north slope area of the Mahu Sag is primarily composed of northeast-trending faults, folds, or monoclines, manifesting as fault zones and fold belts associated with thrust fault zones. From northwest to southeast, these include the Baiwu Fault, Wunan Fault, Fengnan Fault, Xia10 Well Fault, and Xia5 Well Fault [10].

The northern slope area of the Mahu Sag primarily encountered the following strata from top to bottom: the Cretaceous System (Tugulu Group), Jurassic (Qigu Formation, Toutunhe Formation, Xishanyao Formation, Sangonghe Formation, Badaowan Formation), Triassic (Baijiantan Formation, Karamay Formation, Baikouquan Formation), and Permian (Lower Wuerhe Formation, Xiazijie Formation, Fengcheng Formation, Jiamuhe Formation). The Permian Fengcheng Formation is subdivided into three members: Fengcheng member 1 (P1f1), Fengcheng member 2 (P1f2), and Fengcheng member 3 (P1f3) (Fig. 2). The hydrocarbon source rocks of the Permian Fengcheng Formation constitute the primary reservoir crude oil source in the northern slope area of the Mahu Basin. Deposited in a typical closed saline lake environment, these formations exhibit vertically unevenly thick interbedding of silt-to fine-grained sandstone reservoirs with dark mudstones. The reservoirs are tight and exhibit extensive, integrated oil saturation. The Fengcheng Formation underwent four sedimentary evolutionary members: the initial volcanic activity and basin development phase during the P1f1, the basin expansion phase during the early deposition of P1f2, the basin contraction phase during the late deposition of P1f2, and the subsequent basin re-expansion and development phase during P1f3. The deposition of P1f2 is the peak period of alkaline lake development.

Figure 2: Sequence stratigraphy of the Permian Fengcheng formation

3  Reservoir Characteristics

3.1 Geological Structural Form

The northern slope zone, where the MX block is located, exhibits active tectonic activity, primarily characterized by northeast-trending thrust faults and the folds and monoclines formed by thrusting. This is manifested as thrust belts, cover zones, and slope zones associated with the process of thrusting. The Permian Fengcheng Formation exhibits a southwest-dipping monocline overall, with locally developed low-amplitude platforms, anticlines, or nose-shaped structures.

The MX block is a fault block bounded by six faults (Fig. 3). A relatively short northeast-trending secondary fault extends through the center of the block but does not function as a control zone. The overall structure of the block dips southwestward, exhibiting a two-convex-one-concave configuration. The structural high point is located in the easternmost part, while the lowest point is situated at the center of the western syncline. The closure area spans 143.7 km2, with a closure height of 1265 m and a maximum burial depth of 3933 m. The stratigraphic distribution of the P1f2 in the MX Block is stable laterally, with an encountered thickness ranging from 185 to 205 m, reaching its maximum near the M2X well (Fig. 4).

Figure 3: The geological structure of the MX block

Figure 4: The thickness distribution of MX block

3.2 Formation Sedimentary Characteristics

During the Early Permian, the Mahu Depression was in the developmental stage of a foreland basin, with the Fengcheng Formation exhibiting typical foreland basin sedimentary characteristics. Within the foreland sedimentary context of the Fengcheng Formation, both deep-water and shallow-water depositional environments were present. Deep-water environments primarily occurred in the basin center and lower slope areas of the P1f1 and P1f3. Shallow-water environments manifested in several scenarios: First, along the basin margins and upper slopes, where relatively shallow water deposited coarser clastic sediments. Second, around volcanic groups, where sedimentary water bodies were comparatively shallow. Third, evaporite-rich zones represented shallow-water environments, with the Fengcheng Formation Stage 2 developing extensive alkaline evaporites of varying thickness.

The coarse clastic source material in the Mahu Sag basin primarily originates from the western and northern parts of the depression. The northern slope area of Mahu is characterized by fan-delta to lacustrine facies deposits. Within the fan-delta deposits, subfacies include the plain subfacies, inner-front subfacies, and outer-front subfacies. The fan-delta plain is distributed near and north of the Wuxia Fault, while the fan-delta front is in the upper slope zone of northern Mahu Lake. The lacustrine deposits comprise the lacustrine subfacies, including the littoral-shallow lake subfacies, shallow lake subfacies, and semi-deep lake-deep lake subfacies. These deposits are primarily developed on the lower slope of northern Mahu and within the basin interior.

The Fengcheng Formation in the MX Block primarily consists of deposits from the fan delta forearm subfacies. Within this formation, the P1f1 developed deposits from the subaqueous channels and distributary bays within the fan delta forearm. The P1f2 developed the fan delta outer forearm facies belt, with its lower part mainly exhibiting deposits from distal subaqueous channels and estuarine bar microfacies (Fig. 5). The P1f3 generally developed deposits at the outer front of the fan delta. Its lower part consists of distal subaqueous channels, while the middle and upper parts comprise deposits interbedded between distal sand bars and meander pools.

Figure 5: The formation sedimentary characteristics of MX block

3.3 Reservoir Lithology

The MX Block Fengcheng Formation reservoir exhibits diverse mineral composition, primarily comprising two categories of eight terrigenous clastic and carbonate minerals: quartz, potassium feldspar, plagioclase feldspar, calcite, dolomite, siderite, clay minerals, and pyrite. Among these, clay minerals are relatively scarce, averaging 9.2% and showing a vertical gradient from lower to higher concentrations. Clay minerals are dominated by illite-montmorillonite and green montmorillonite, accounting for 34.47% and 32.93%, respectively. Felsic-carbonate minerals exhibit higher abundances (average quartz-feldspar content: 59.9%; average calcite-dolomite content: 25.2%). The grain size of felsic minerals is similar to that of muddy carbonate minerals, with a wide range of content variation. Carbonate minerals are primarily composed of muddy calcite and dolomite, featuring extremely fine grains ranging from several micrometers to tens of micrometers. Under polarized light microscopy, they exhibit distinct granularity, though crystal forms are difficult to discern. They are relatively pure in composition and well-rounded, indicating significant alteration before deposition. Vertically, as the content of felsic minerals increases, the carbonate minerals exhibit an inverse relationship.

The terrigenous clastic grains in the P1f2 are generally of mud/silt grade, with over 80% of the clastic particle size being less than 62.5 μm (Fig. 6). Five types of bedding are developed: felsic, dolomitic-calcareous, siliceous, tuffaceous, and organic, exhibiting laminated to thinly laminated structural characteristics, making it a typical continental shale oil. Based on the characteristics of three typical components—“terrigenous felsic clastics, authigenic carbonate minerals, and authigenic silica”—the lithology is classified into four main categories: siliceous mudstone, felsic mudstone, dolomite-bearing felsic mudstone, and dolomitic mudstone (Fig. 7), with felsic mudstone and dolomite-bearing felsic mudstone being the dominant types. Among them, felsic mudstone accounts for 53.39% of the rock layer thickness, while dolomite-bearing felsic mudstone accounts for 32.56%. Additionally, felsic mudstone exhibits good oil-bearing properties with lighter oil quality; siliceous mudstone contains a higher proportion of heavy components; whereas dolomite-bearing felsic mudstone and dolomitic mudstone have a relatively high content of light components.

Figure 6: The particle size distribution of reservoir rocks

Figure 7: Typical photos of different rock types in the Fengcheng Formation

3.4 Petrophysics Properties

The primary reservoir spaces in the MX Block Fengcheng Formation consist of intergranular (intercrystalline) pores, dissolution pores, and microfractures. Matrix pores mainly develop in three types: intergranular (intercrystalline) pores, intragranular dissolution pores, and intergranular pores in clay minerals. Among these, intragranular dissolution pores and intergranular pores are predominantly developed in felsic shales, while intercrystalline pores occur in all four lithological types.

The maximum pore throat radius of reservoir rocks ranges from 0.06 to 0.82 μm, with an average of 0.14 μm; the displacement pressure ranges from 0.90 to 12.18 MPa, averaging 7.25 MPa; the median saturation pressure averages 85.78 MPa, with a median saturation radius of 0.02 μm. The reservoir rocks exhibit relatively high displacement pressure and moderate porosity sorting, displaying characteristics of small pores and fine throats.

Statistical analysis of MX block core properties reveals (Fig. 8) that reservoir porosity ranges from 1.70% to 11.80%, with a median of 4.50% and an average of 4.97%. Permeability analysis results indicate a permeability range of 0.01–13.00 mD, with a median of 0.02 mD and an average of 0.06 mD; The porosity of the reservoir with oil ranges from 3.13% to 11.80%, with a median of 5.51% and an average of 6.24%; Permeability ranged from 0.01 to 13.00 mD, with a median of 0.03 mD and an average of 0.07 mD.

Figure 8: The physical properties of MX core samples

4  Production Performance

4.1 Basic Information of Appraisal Wells

Since the second half of 2021, two directional wells (M1X and M2X) and one horizontal well (M3H) have been systematically deployed in the MX block by PetroChina Xinjiang Oilfield to evaluate the production potential and performance dynamics of different well types in the P1f2 shale reservoir. While M2X partially penetrated the underlying P1f3 formation, both M1X and M3H were well-positioned within the core zone of the P1f2 reservoir. Tables 1 and 2 show the basic information and high-pressure fluid properties of crude oil and natural gas from the P1f2 reservoir of the three appraisal wells, respectively.

To achieve sufficient oil well production capacity, all three appraisal wells underwent large-scale volumetric fracturing operations for enhanced production. The fracturing construction process employed a combination of cementing plugs and perforation for staged operations. The pumping process utilized reverse mixing and composite methods. An anti-scaling polymer fracturing fluid was selected (due to high Ca2+/Mg2+ content in formation rocks, which promotes fracturing fluid scaling). Tapered well proppants comprised 70/140 and 40/70 mesh ceramic proppants, while horizontal well proppants used quartz sand and ceramic proppants (15:85 ratio). Table 3 shows the fracturing parameters in detail.

4.2 Production Performance Characteristics of Appraisal Wells

4.2.1 Well M1X

Based on the production performance curve of Well M1X (Fig. 9), the production profile can be divided into three stages, with the first two stages being the main contributors to the well’s output.

Figure 9: Production performance curve of Well M1X

In the first production stage, in an oil well that has been hydraulically fractured, initial production is largely derived from the areas with the most fractures. This area features a well-developed artificial fracture network and good fluid flow channels, allowing rapid fluid supply to the well. As a result, the well exhibits high initial productivity with almost no fracturing fluid flowback, high oil production, and low water production. However, the production declines rapidly, with an initial monthly decline rate of 0.1527/mon. The wellhead pressure also decreases linearly and rapidly, at an initial rate of 5.1 MPa/mon.

In the second production stage, the areas with moderate production improvement effects from artificial fracturing are mainly utilized. The density of artificial fractures in this area is relatively low. At the same time, this area supplies fluid to both the oil wells and the areas with better production improvement effect. However, the supply speed has decreased compared to the initial production stage. The water production from the oil wells has stabilized, approximately at 20.8%.

In the third production stage, the original fluids within the hydraulic fracture network have almost entirely flowed into the wellbore. Production gradually begins to rely on the areas with poor production improvement effects and weakly modified original area. This area mainly supplies liquid to the artificially fractured modified area. The overall liquid supply capacity in this stage is poor, but the supply is relatively stable. Thus, the well exhibits a low decline rate and a concise production tail [11].

4.2.2 Well M2X

Based on the production performance curve of Well M2X (Fig. 10), its production profile can be divided into three stages, with the first stage being the primary contributor to total output. Although Well M2X exhibits three production stages similar to those of Well M1X, their production behaviors differ significantly.

Figure 10: Production performance curve of Well M2X

In the first stage, production mainly came from the effectively stimulated zones with a well-developed fracture network and good fluid flow capacity. Fluid supply to the well was rapid, and wellhead pressure decreased linearly at an initial rate of 2.3 MPa per month. Unlike Well M1X, however, Well M2X experienced some fracturing fluid flowback. After cleanup, the well reached a relatively stable production regime, with oil output of approximately 50.7 t/d and water cut of about 37.6%. No significant decline trend was observed during this period, indicating stable production capacity until the fluids in this zone were largely depleted.

In the third stage, similar to Well M1X, production gradually shifted to poorly stimulated and weakly modified original formations. Although the overall fluid delivery capacity remained limited, the supply was relatively stable, resulting in a low decline rate and a concise production tail.

4.2.3 Well M3H

Based on the production dynamic curve of the M3H well (Fig. 11), its production curve can be divided into three distinct stages, with the first two phases constituting the primary contributors to the well’s output.

Figure 11: Production performance curve of Well M3H

Similar to the M1X well, in the first well production stage, the main effort was made to utilize the areas with better production enhancement effects from artificial fracturing. In these areas, the artificial fracturing network was well-developed and the fluid flow channels were good. The liquid supply to the oil well was rapid, and the wellhead pressure decreased linearly and rapidly, with an initial pressure drop rate of 7.1 MPa per month. However, the oil production and water production of the oil well reached a relatively stable state (oil production was approximately 68.8 t/d, and the water content was approximately 9.1%). There was no decline trend, and the well demonstrated a stable production capacity until all the fluids in this area flowed into the wellbore.

Unlike the M1X and M2X wells, in the second production stage, due to the interference caused by the volume fracturing stimulation operation in the adjacent area on the M3H well, the fracturing fluid and crude oil from the adjoining area intruded into the artificial fracturing stimulation area for production enhancement of the M3H well, resulting in a certain degree of replenishment of the formation pressure and fluid content in this area. As a result, the pressure drop pattern in the second stage was similar to that in the first stage, while the water production volume increased sharply due to the influence of the fracturing fluid, reaching 50%.

In the third production stage, this stage utilizes the areas with poor production enhancement effect from artificial fracturing and the original reservoir with weak fracturing improvement. However, the inter-well fracturing interference still plays a role to some extent. The overall liquid supply capacity gradually decreases, but the water production volume remains at a relatively high level (with a water cut of approximately 36.5%). The well exhibits a low decline rate and a concise production tail.

4.2.4 Summary

The production performance curves of the aforementioned three appraisal wells shows that all three wells exhibit multiple decline stages. The production behavior demonstrates a significant dual-control characteristic dominated by both fracture and matrix supply. By analyzing changes in wellhead pressure along with variations in oil and water production rates, the production characteristics of the appraisal wells in the Fengcheng Formation of the MX block showed three distinct stages. The production behavior in the first stage is controlled by zones where hydraulic fracturing has achieved better stimulation results and the second stage is governed by areas with average fracturing effectiveness. The third stage is dominated by poorly stimulated zones and original regions with weak stimulation. Furthermore, the first and second stages constitute the main contribution stages to the well’s production, while the third stage determines the remaining oil production potential of the well [12]. This final stage is crucial for accurately predicting the well’s Estimated Ultimate Recovery (EUR).

Overall, the development results of the appraisal wells in the MX Block Fengcheng Formation shale reservoir are favorable, demonstrating good economic development potential. Unlike conventional shale reservoirs, these wells exhibited high-efficiency oil production from the initial production stage. Except for M2X, M1X and M3H showed virtually no fracturing fluid return (M2X had a 15-day steaming period, slightly shorter than M1X’s 20 days and M3H’s 19 days). This clearly demonstrates that during the well shut-in phase of fracturing operations, strong suction effects enabled the fracturing fluid to displace crude oil from the shale, directing it into the artificially created fracture network. Therefore, appropriately extending the well shut-in time during fracturing operations will effectively enhance the recovery rate of shale oil in this area. Comprehensive analysis of the production dynamic curves from the three appraisal wells indicates that the production characteristics of the P1f2 shale reservoir in the MX block are determined by reservoir properties-including free oil porosity, proportion of high-quality reservoirs, and reserve scale within the well control area-as well as the quality of artificial fracturing enhancement within that area. Additionally, appropriately controlling the initial production rate of oil wells helps stabilize production and reduce fluctuations. Furthermore, horizontal wells are more conducive to stable production than high-angle wells.

4.3 Estimated Ultimate Recovery

4.3.1 Prediction Method for EUR

Production from shale reservoirs in the P1f2 typically undergoes three distinct phases, resulting in production curves that generally exhibit multiple decline characteristics. Consequently, directly applying the Arps production decline model in conjunction with the economic limit production rate to predict the EUR of a well presents challenges in effectively selecting the starting point of the decline curve, and leads to errors in evaluating the well’s EUR.

Analysis reveals that in the first two stages of production from P1f2, the wells primarily contribute to output volume. Only the third stage truly determines the well’s Ultimate Recovery Potential (EUR). Therefore, accurately identifying when a well enters this third stage is fundamental to predicting its EUR. To this end, based on the characteristics of the reservoir areas that were automatically controlled for oil production in each of the three stages of the oil well, this paper proposes a new EUR prediction method for shale oil reservoirs. The specific operational workflow is as follows:

(1) Determine the time when oil well production fully enters the third stage.

Based on the production dynamics of oil wells in MX block, once a well enters the third stage, all reserves within the well control zone have been fully utilized, and formation pressure has completely reached the well’s supply boundary. Therefore, to determine whether a well has entered the third stage of production and to establish the precise timing of this transition, it is necessary to develop a criterion for assessing whether all reserves within the well control zone have been fully utilized and whether formation pressure has completely reached the well’s supply boundary. Furthermore, since it is typically challenging to strictly maintain constant bottomhole flow pressure or constant production rates, fluctuations in both bottomhole pressure and output inevitably accompany the entire production process. Considering that the qDd-tcDd plot from the Blasingame method can be applied to both variable bottomhole pressure and variable production scenarios, and that when the reserves within the well’s controllable zone are fully utilized and formation pressure has completely reached the well’s supply boundary, the qDd-tcDd plot from the Blasingame method exhibits a distinct linear characteristic (Fig. 12). Therefore, this work will utilize the qDd-tcDd diagram of the Blasingame method to analyze and evaluate well production data from different time periods, thereby determining the timing of a well’s entry into the third production stage [13,14].

Figure 12: Blasingame type curves

(2) Establish the production decline model for the third stage

At present, the Arps decline analysis method widely applied in analyzing the production decline of shale oil reservoir wells. The Arps method encompasses three decline types: exponential decline, harmonic decline and hyperbolic decline. When the decline index n = 0, the well exhibits exponential decline; When the decline index n = 1, the well exhibits harmonic decline; when the decline index lies between 0 and 1, the well follows hyperbolic decline [15–17]. In conventional decline analysis, different decline types typically analyzed separately to assess well production patterns, with the model best matching the production decline selected as the well’s decline model. Clearly, this traditional workflow is not only cumbersome and labor-intensive but also highly susceptible to analyst subjectivity. Therefore, to rapidly obtain the third-stage production model for the MX block appraisal wells, this work employs the Levenberg-Marquardt algorithm for nonlinear fitting of well production data, based on the hyperbolic decline type (Eq. (1)) from the Arps production decline analysis method [18–20]. Furthermore, during the nonlinear fitting process, the control accuracy for the decline index n is set to 0.1. This approach significantly enhances the efficiency of oil well production decline analysis while ensuring the analytical results meet engineering accuracy requirements.

The general equation of Arps hyperbolic decline:

q(t)=qr[1+nD(t−tr)]1n(1)

where, q(t) is oil production rate, n is decline exponent, D is decline rate, t is production duration, and subscript r is reference time.

(3) Predicting the EUR of an oil well

Based on the production decline model for the third stage of oil wells, and using the economic limit daily production rate as the predictive constraint indicator for the decline model (production ceases when output falls below this threshold), the future production variation pattern for the third stage is forecasted (Eq. (2)). Subsequently, the residual recoverable oil volume for the third stage is determined based on the predictive model, and the well’s EUR can then be calculated using the current cumulative oil production (Eq. (3)).

Nres=∑t0tq(t)(2)

EUR=Np+Nres(3)

where, Np is cumulative oil production, and Nres is remaining recoverable oil in place.

4.3.2 EUR of Appraisal Wells

(1) The start time of fully entering the third stage

Based on the aforementioned method for determining when oil wells fully enter the third production stage, conducted a comparative analysis of the production dynamics data from the MX block appraisal wells at different time points. This analysis examined the overlap with the linear segment in Blasingame’s qDd-tcDd diagram to identify the exact time when the MX block appraisal wells fully transitioned into the third phase. The analysis results for the three appraisal wells are shown in Fig. 13, and the exact times when these three appraisal wells fully entered the third phase are listed in Table 4.

Figure 13: Blasingame type curve analysis of three wells

Based on the qDd-tcDd chart comparison analysis using the Blasingame method, all three wells exhibited stable supply linear characteristics within the well control zone. Specifically, M1X well demonstrated this pattern three times, M2X well twice, and M3H well once. This indicates that, except for M3H, which only achieved stable supply upon entering the third stage, M1X and M2X also demonstrated stable supply capabilities within the artificially fractured volume enhancement zone. Specifically, due to the lower production rate control at the wellhead of M2X, the transition period from the artificially fractured volume enhancement zone to the original reservoir zone was brief. This allowed the original reservoir zone to promptly compensate for the diminished fluid supply capacity after the decline in the artificially fractured zone, thereby enhancing the stability of M2X’s production. Additionally, M2X was the fastest to fully enter the third phase, indicating that appropriate control of the initial production rate of oil wells is conducive to maintaining the stable production of the wells over a long period, thereby improving the development efficiency.

(2) Production decline analysis and EUR prediction

Based on the timing when the three appraisal wells in the MX block fully entered the third production stage, a decline analysis was conducted using Arps’ hyperbolic decline model and the aforementioned well decline model determination method. The results are shown in Fig. 14, with the decline model presented in Table 5. The decline analysis reveals a good fit between the decline curve and actual production data. All three appraisal wells in MX Block exhibit hyperbolic decline patterns during the third production phase, with relatively high decline indices. Notably, Well M2X shows an index of 0.9, approaching 1. This indicates that the decline patterns of the three wells are approaching harmonic decline characteristics. That is, the decline rates of all three appraisal wells are relatively slow, suggesting they will have extended lifespans and high remaining recoverable oil reserves.

Figure 14: Decline curve analysis of three wells

Based on the current economic limit, the daily oil production rate for shale oil wells in Xinjiang oilfields (2.38 t/d), the decline model for wells in the third production stage from Table 5 was applied to predict the future production trend of the evaluation wells during this stage. The remaining recoverable oil volume was calculated to determine the EUR for the three evaluation wells in the MX block. The calculation results are shown in Table 6. Based on the EUR analysis, Well M1X exhibits the highest EUR at 58,328 t, while the other two wells in the MX block also surpass 40,000 t. Considering current crude oil sales prices, a well becomes profitable when its EUR exceeds 35,000 t. Furthermore, once large-scale development of the P1f2 shale reservoir in the MX block commences, it will inevitably reduce per-well investment and production costs for reservoir development. This indicates that the P1f2 shale reservoir in the MX block possesses favorable economic benefits, enabling accelerated advancement of large-scale reservoir development.

5  Discussion

The comprehensive analysis of reservoir characteristics and production performance of the shale oil reservoir in the Second Member of the Fengcheng Formation, MX Block, reveals several key insights with implications for efficient development and accurate reserve prediction.

(1) The Dual-Control flow mechanism: fractures and matrix

The consistent multi-stage decline observed across all three appraisal wells is a direct manifestation of a complex flow mechanism governed by both hydraulic fractures and the matrix system. The initial high production rates are fueled by the rapidly depleting, highly conductive fracture network created through volume fracturing. Subsequently, the sustained, albeit lower, production is supported by the slower supply from the tight matrix. This “fracture + matrix” dualism underscores that successful development cannot rely solely on creating complex fracture networks but must also efficiently mobilize the hydrocarbons stored in the nano-pore matrix. This understanding is fundamental for optimizing completion strategies and long-term production management.

(2) Implications of the three-stage production model

The division of the production lifecycle into three stages provides a powerful conceptual and practical framework. Our findings demonstrate that the first two stages, while contributing the bulk of early cumulative production, are essentially a “depletion” phase of the stimulated reservoir volume (SRV). The third stage, controlled by the poorly stimulated or unstimulated matrix, is not merely a tail but the decisive phase that determines the well’s EUR. This challenges the conventional practice of relying solely on early-time data for decline curve analysis. Accurately identifying the transition to this boundary-dominated flow regime is, therefore, the cornerstone of reliable reserves estimation, as formalized in our proposed workflow. The decline exponents (n near 0.8–0.9) obtained for the third production stage are notably higher than the typical values (n near 0 or 0.5) expected for boundary-dominated flow in conventional liquid reservoirs. This near-harmonic decline behavior (n near 1) is interpreted as a result of the persistent fracture-matrix interaction and the multi-scale heterogeneity of the stimulated reservoir volume. Even after the onset of boundary-influenced flow, the well-connected fracture network continues to provide conductive pathways, while the tight matrix slowly replenishes the fractures. This leads to a prolonged transitional flow regime that mimics harmonic decline. Additionally, inter-well fracture interference (evident in Well M3H) and the natural fracture system further sustain a slower pressure depletion, explaining the high n-values. Thus, the third stage in these wells represents a fracture-mediated, boundary-influenced flow regime rather than a classical homogeneous boundary-dominated flow.

(3) The influence of geological characteristics on production performance

In this study, the mineral composition, lithology distribution, and physical properties of the second member of the Fengcheng Formation in the MX block exert a significant control on production performance. The high felsic mineral content (averaging 59.9%) combined with low clay content (averaging 9.2%) results in high reservoir brittleness, which favors the development of complex fracture networks during hydraulic fracturing. This, in turn, supports the rapid flow characteristics observed in the early high-production stages (Stages 1 and 2). Furthermore, the presence of microfractures and dissolution pores within the reservoir provides additional flow pathways, further enhancing the conductivity of the fracture system. For instance, the M2X well is located in an area with the thickest reservoir layer (approximately 205 m), and it has a higher content of felsic minerals, which makes it more brittle. After fracturing, the fracture network becomes more developed, and fluid supply is more rapid. Therefore, it entered the third stage (boundary-dominated flow stage) earlier. In contrast, the M1X well has a slightly thinner reservoir layer, and the proportion of carbonate minerals in the mineral composition is higher, resulting in a relatively simpler fracture network and a slower transition to the flow stage.

(4) The influence of completion and production strategies

The minimal flowback water in Wells M1X and M3H strongly suggests the occurrence of counter-current imbibition during the extended shut-in period, where fracturing fluid displaces oil from the matrix into the fractures. This phenomenon enhances initial oil mobility and recovery, advocating for an optimized, extended soaking time as a key component of the completion design. The superior performance stability of the horizontal well (M3H) compared to the highly deviated wells highlights the advantage of maximizing reservoir contact. Furthermore, the case of Well M2X illustrates that controlling the initial production rate can mitigate rapid pressure depletion, leading to a more stable decline and a smoother transition between production stages, ultimately improving recovery efficiency.

(5) A robust workflow for EUR prediction in complex systems

The proposed EUR prediction methodology directly addresses the limitation of traditional Arps decline analysis in multi-stage shale oil production. By leveraging Blasingame type curves to objectively identify the start of the third stage, the method removes the subjectivity associated with choosing a decline curve start point. Coupling this with a non-linear regression for the hyperbolic decline model provides a more rigorous and data-driven forecast. The high decline exponents (n near 1) obtained for all wells indicate a trend towards harmonic decline, which implies a slower decline rate and a longer productive life, further affirming the substantial remaining potential in the third stage.

6  Conclusions

This study presents a detailed evaluation of the shale oil reservoir in the Second Member of the Permian Fengcheng Formation, MX Block, Mahu Sag, leading to the following key conclusions:

(1)   The reservoir is characterized as a typical lacustrine shale with complex mineralogy, dominated by felsic and carbonate minerals, and exhibits poor petrophysical properties with low porosity and ultra-low permeability, necessitating volume fracturing for economic production.

(2)   The production dynamics of appraisal wells consistently exhibit a three-stage decline pattern, reflecting a dual-control production mechanism where initial flow is dominated by the fracture network, followed by a long-term, stable supply from the matrix.

(3)   A practical workflow for EUR prediction is applied. This method utilizes Blasingame type-curve analysis to objectively identify the commencement of the boundary-dominated third stage and applies a hyperbolic decline model fitted with the Levenberg-Marquardt algorithm to forecast long-term production.

(4)   The EUR of the three appraisal wells is predicted to range between 43,500 and 58,300 t, significantly exceeding the economic limit and confirming the commercial development potential of the MX Block.

(5)   Operational strategies, including the use of horizontal wells, appropriate extension of fracturing fluid soaking time, and controlled initial production rates, are identified as key factors for stabilizing production and enhancing ultimate recovery.

(6)   These findings provide a solid scientific basis and technical framework for the efficient development of the Fengcheng Formation shale oil reservoir in the MX Block and offer valuable insights for the development of similar unconventional resources.

Acknowledgement: We would like to express our gratitude to PetroChina Xinjiang Oilfield Company for providing the precious research data in this work.

Funding Statement: None.

Author Contributions: The authors confirm contributions to the paper as follows: study conception and design: Lin Chen, Anqi Zhao, Yunpei Zhang, Sai-Mi-La XiaFuKaiTi; analysis and interpretation of results: Yao Qin, Chuan Wang, Gang Chen, Jiqiang Li; draft manuscript preparation: Anqi Zhao; review: Shilai Hu. All authors reviewed the results and approved the final version of the manuscript.

Availability of Data and Materials: The data that support the findings of this study are available from the corresponding author upon reasonable request.

Ethics Approval: Not applicable.

Conflicts of Interest: The authors declare no conflicts of interest to report regarding the present study.

References

1. Moghani AM, Loni R. Review on energy governance and demand security in oil-rich countries. Energy Strategy Rev. 2025;57:101625. doi:10.1016/j.esr.2024.101625. [Google Scholar] [CrossRef]

2. Luo J, Wang H, Li H, Zheng B. Structural shifts in China’s oil and gas CH4 emissions with implications for mitigation efforts. Nat Commun. 2025;16(1):2926. doi:10.1038/s41467-025-58237-z. [Google Scholar] [PubMed] [CrossRef]

3. Wang Z, Fan Z, Chen X, Li Y, Fan Z, Wei Q, et al. Global oil and gas development situation, trends and enlightenment in 2023. Petrol Explor Dev. 2024;51(6):1536–55. doi:10.1016/s1876-3804(25)60558-1. [Google Scholar] [CrossRef]

4. Polat O, Cunado J, Cepni O, Gupta R. Oil price shocks and the connectedness of US state-level financial markets. Energy Econ. 2025;141:108128. doi:10.1016/j.eneco.2024.108128. [Google Scholar] [CrossRef]

5. Guan Q, Chen C, Pu X, Wan Y, Xu J, Zeng H, et al. Production performance analysis of a continental shale oil reservoir in Bohai Bay basin. Petroleum. 2024;10(2):294–305. doi:10.1016/j.petlm.2023.11.002. [Google Scholar] [CrossRef]

6. Dang W, Gao G, You X, Wu J, Liu S, Yan Q, et al. Genesis and distribution of oils in Mahu Sag, Junggar basin, NW China. Petrol Explor Dev. 2023;50(4):840–50. doi:10.1016/s1876-3804(23)60432-x. [Google Scholar] [CrossRef]

7. Huang Y, Wang G, Zhang Y, Xi J, Huang L, Wang S, et al. Logging evaluation of pore structure and reservoir quality in shale oil reservoir: the Fengcheng Formation in Mahu Sag, Junggar Basin. China Mar Petrol Geol. 2023;156:106454. doi:10.1016/j.marpetgeo.2023.106454. [Google Scholar] [CrossRef]

8. Du X, Jin Z, Zeng L, Liu G, He W, Ostadhassan M, et al. Formation of natural fractures and their impact on shale oil accumulation in the Mahu Sag, Junggar Basin, NW China. Int J Coal Geol. 2023;279:104385. doi:10.1016/j.coal.2023.104385. [Google Scholar] [CrossRef]

9. Wang S, Wang G, Huang L, Song L, Zhang Y, Li D, et al. Logging evaluation of lamina structure and reservoir quality in shale oil reservoir of Fengcheng Formation in Mahu Sag. China Mar Pet Geol. 2021;133:105299. doi:10.1016/j.marpetgeo.2021.105299. [Google Scholar] [CrossRef]

10. Hu ZM, Hu MY, Yu Y, Jiao SJ, Wei T, Meng LF, et al. Fault-fracture body identification and reservoir development model: taking Baikouquan Formation in Madong area of Mahu Sag, Junggar Basin as an example. Fault-Block Oil Gas Field. 2024;31(2):283–8. (In Chinese). [Google Scholar]

11. Li ZY, Chen JB, Zuo HL, Wang Q, Ouyang W, Lin T, et al. Interaction laws of hydraulic fractures and natural fractures in shale reservoirs. Fault-Block Oil Gas Field. 2024;31(2):232–40. (In Chinese). [Google Scholar]

12. Meng XB. Numerical simulation of fracture propagation in the whole well stage of multi-cluster fracturing of horizontal wells in shale oil reservoirs: a case study of Shahejie Formation in Jiyang Depres-sion. Fault-Block Oil Gas Field. 2024;31(3):459–66. (In Chinese). doi:10.1007/978-981-97-0260-2_43. [Google Scholar] [CrossRef]

13. Dang WL, Gao G, You XC, Fan KT, Wu J, Lei DW, et al. Geochemical identification of a source rock affected by migrated hydrocarbons and its geological significance: fengcheng Formation, southern Mahu Sag, Junggar Basin, NW China. Petrol Sci. 2024;21(1):100–14. doi:10.1016/j.petsci.2023.10.022. [Google Scholar] [CrossRef]

14. Du P, Wang Q, Zhang J, Wang Y, Wang D, Han Y, et al. The Blasingame method considering the effect of inter-well interference in a circular homogeneous closed reservoir. Front Energy Res. 2024;12:1376848. doi:10.3389/fenrg.2024.1376848. [Google Scholar] [CrossRef]

15. Wei M, Duan Y, Dong M, Fang Q. Blasingame decline type curves with material balance pseudo-time modified for multi-fractured horizontal wells in shale gas reservoirs. J Nat Gas Sci Eng. 2016;31:340–50. doi:10.1016/j.jngse.2016.03.033. [Google Scholar] [CrossRef]

16. Maurenza F, Yasutra A, Tungkup IL. Production forecasting using arps decline curve model with the effect of artificial lift installation. Sci Contrib Oil Gas. 2023;46(1):9–18. doi:10.29017/scog.46.1.1310. [Google Scholar] [CrossRef]

17. Gupta I, Rai C, Sondergeld C, Devegowda D. Variable exponential decline: modified ArPs to characterize unconventional-shale production performance. SPE Reserv Eval Eng. 2018;21(4):1045–57. doi:10.2118/194005-pa. [Google Scholar] [CrossRef]

18. Arps JJ. Analysis of decline curves. Trans AIME. 1945;160(1):228–47. doi:10.2118/945228-g. [Google Scholar] [CrossRef]

19. Moré JJ. The Levenberg-Marquardt algorithm: implementation and theory. In: Numerical analysis. Berlin/Heidelberg, Germany: Springer; 2006. p. 105–16. doi:10.1007/BFb0067700. [Google Scholar] [CrossRef]

20. Matinkia M, Hashami R, Mehrad M, Hajsaeedi MR, Velayati A. Prediction of permeability from well logs using a new hybrid machine learning algorithm. Petroleum. 2023;9(1):108–23. doi:10.1016/j.petlm.2022.03.003. [Google Scholar] [CrossRef]