Generally, the contents of brittle minerals and clay have a significant influence on the effect of reservoir fracturing (Wang et al., 2016). The reservoir with high brittle minerals content and low calcite content is with a relatively short drilling cycle and is easy to be fractured (Wang et al., 2013). Therefore, the evaluation of mineral content is particularly important for shale reservoir evaluation. At present, conventional logging combining with ECS logging is often used to evaluate the mineral composition of shale (MacDonald et al., 2012). The major elements measured by ECS logging include Si, Ca, Fe, S, Ti, Gd, etc. As the Si is closely related to quartz, Ca is closely related to calcite and dolomite, and Al is closely related to clay mineral content, the ECS logging data can be used to evaluate the mineral composition of shale by establishing the logging interpretation model.
The correlation analysis between core testing data and logging interpretation results proves the feasibility and effectiveness of the model. Figure 2 shows that the logging interpretation of quartz, carbonate and clay mineral content is in a moderate or good correlation with the core testing results. The correlation coefficients R2 are 0.53, 0.61, and 0.42, respectively. The R2 value between 0 and 0.2 is considered poor, 0.2-0.4 fair, 0.4-0.6 moderate, 0.6-0.8 good, and 0.8-1.0 almost perfect (McGinn et al., 2007). Therefore, ECS logging can be well used for mineral component evaluation. ECS interpretation data shows that the clay content at the lower part of the shale (1 850 to 1 872 m) decreases slightly, with an average of 25.3%, the calcite content decreases obviously, with an average of 17.3%, the quartz content is higher, with an average of 35.1%, the content of feldspar is low, while the pyrite increases to an average of 4.5% (Fig. 3). The calcite content at the upper part of the shale (1 786 to 1 850 m) is higher, with an average of 36.7%. The clay minerals are with an average of 27.5%. The quartz is with an average of 21.3%. The content of feldspar and pyrite is relatively low, less than 3% on average. It is showed that the content of brittle minerals (quartz and feldspar) at the lower part of the Shuijingtuo Formation is higher (Sondergeld et al., 2010). Previous studies have shown that the reservoir can have a good fracturing ability when the content of brittle minerals is more than 40% (Sondergeld et al., 2010). The logging interpretation results show that there is just the bottom 20 m segment of the Shuijingtuo Formation locally meeting the above conditions, which constitutes the most favorable reconstruction intervals.
Figure 2. Cross-plot of mineral contents of core testing and elemental capture spectroscopy (ECS) logging interpretation from the shale of Shuijingtuo Formation in Well EYY1.
As the results of CMR measurement are not affected by the rock skeleton composition, and are sensitive to the flowing of pore fluid, which can accurately provide the porosity value in the stratigraphy, it is widely used in the shale porosity testing (Westphal et al., 2005). A method for calculating the true porosity of low porosity and low permeability reservoir is presented by combining acoustic and NMR (nuclear magnetic resonance) logging data (Mao et al., 2010). It can accurately provide the interval porosity relating to the pore throat radius and pore structure of the rock. The CMR porosity model is shown in Fig. 4. As can be seen from the figure, with the relaxation time increasing from left to right, it corresponds to muddy bound water, capillary bound water and movable water in turn. Movable water exists in relatively large pore space and generally has the longest relaxation time, corresponding to a relatively large T2 spectral interval. Muddy bound water exists in very small micro-pores, and the echo signal is easily attenuated, which corresponds to the minimum relaxation time. The relaxation time of capillary bound water is between that of movable water and muddy bound water. The area enclosed by the CMR decay spectrum curve (T2 spectrum) is proportional to the number of hydrogen nucleus in the pore fluid, so the T2 spectrum of CMR log can be directly calibrated to the reservoir porosity (Li et al., 2016). The calculation formula of porosity is as follows
Φ is the porosity, %. T2 is the relaxation time, ms. S(T2) is the distributed spectral function of T2.
Based on the above principle of CMR test technology, this study proposes the CMR porosity to characterize the reservoir measurement technology and can obtain more accurate information of reservoir pore parameters. As can be seen from the correlation diagram of nuclear magnetic porosity and helium porosity tested in the laboratory, the correlation coefficient R2 of the them is 0.97, indicating that nuclear magnetic porosity can well characterize the changes of shale porosity (Fig. 5). In addition, according to the distribution histogram of nuclear magnetic pores (Fig. 6), it can be seen that the ratio of medium to large pores of the lower section is higher than that in the middle section, showing that the ratio of medium to large pores gradually decreases from bottom to top. This indicates that the pore structure and effective porosity in the shale of Shuijingtuo Formation is gradually worse from the bottom to up, which is consistent with the results of core analysis. Previous studies have shown porosity characteristics. In Yichang area, the Schlumberger's CMR-plus is adopted. It adopts the modern pulse-echo that the pores of the shale in the study area are mainly composed of organic pores and inorganic pores, and the contribution of organic matter pores to porosity is relatively greater when the TOC is greater than 2%. Therefore, high abundance of TOC is conducive to the development of pore so that the high TOC shale shows high porosity (Li et al., 2019). According to the results of the nuclear magnetic porosity testing, the porosity of the shale of Shuijingtuo Formation decreases from bottom to top (Fig. 6). The porosity of the lower section is generally greater than 3%, while the porosity of the upper section is generally less than 3%, indicating that the lower section is a favorable shale gas storage segment.
Figure 5. Cross-plot of helium porosity and combinable nuclear magnetic resonance (CMR) porosity of the shale of Shuijingtuo Formation in Well EYY1.
The gas bearing content refers to the volume of natural gas contained in each ton of shale under standard conditions. According to the occurrence state, the gas bearing content of shale is composed of adsorbed gas, free gas and dissolved gas (Pan and Connell, 2015). The gas bearing content testing is an important experiment to evaluate the quality of shale reservoir and optimize favorable zone. The testing methods include desorbed method, isothermal adsorption method, and the logging interpretation method (Diamond and Schatzel, 1998). The desorbed method is a direct method to test the gas bearing content, and also a most commonly used method. Isothermal adsorption and logging interpretation method are indirect methods for the gas bearing content. The accuracy of desorbed measurement of gas content mainly depends on two factors. One is try to reduce the loss of gas, and the other one is to simulate the stratigraphy conditions, especially the geothermal condition (Wei et al., 2007). As desorbed method is the most direct method to test shale gas content, the testing results are mainly used in this study.
In this study, the logging evaluation of gas bearing content is carried out by combining conventional logging data with desorbed method testing data. As it is shown on Fig. 7, the acoustic logging data is with a good linear correlation with the total gas volume and desorbed gas volume data. The correlation coefficients (R2) are 0.607 3 and 0.611 9, respectively. Therefore, gas bearing content data can be fitted with acoustic logging data. The established logging interpretation model of gas bearing content is as follows.
Figure 7. Cross-plot of AC (acoustic) log and gas bearing data of the shale of Shuijingtuo Formation in Well EYY1.
TG is the total gas volume, m3·t-1. AC is the acoustic logging data, μs·m-1. DG is the desorbed gas volume, m3·t-1.
The gas bearing content of Shuijingtuo Formation is calculated by using the above logging interpretation models. The results show that the calculated gas bearing content of the model is in good agreement with the gas bearing content data, indicating the high accuracy of the model (Fig. 8). It shows that the total gas bearing content gradually decreases from the bottom to up, and the lower 20 m segment is generally greater than 2 m3·t-1, indicating that the lower segment is a shale gas enrichment layer.
Organic geochemical parameters are important parameters to evaluate the oil and gas abundance in shale (Lu et al., 2012). Organic matter has special physical properties, such as its poor conductivity, strong natural radioactivity, density closing to the density of water, belongs to the light component, acoustic closing to 550 μs·m-1, and hydrogen index closing to 67% (Murphy et al., 2008). The conventional logging cannot directly measure the TOC in shale reservoir, so it is necessary to establish the TOC logging calculation model of conventional logging data combining core testing data. Predecessors' research results show that the TOC can be evaluated by natural gamma ray, natural gamma ray spectrometry, resistivity, density and acoustic logging data according to the corresponding reflections on different well logging curves (Kamali and Mirshady, 2004). In this study, a TOC calculation model was established by combining well logging data with measured core TOC data, and the shale quality was evaluated. As it can be seen from Fig. 9, the correlation relationship between TOC and GR is excellent, and the correlation coefficient (R2) is 0.81. Therefore, the relation of TOC logging interpretation model is established. The formula is as follows
Figure 9. Cross-plot of GR (gamma ray) log and total organic carbon content (TOC) data of the shale of Shuijingtuo Formation in Well EYY1.
TOC is the total organic carboncontent, %. GR is the gamma ray logging data, gAPI.
The TOC of the shale of Shuijingtuo Formation is calculated by using the established logging interpretation model. The results show that the TOC gradually decreases from the bottom to the top, with a variation range of 0.86% to 11.83% (Fig. 8). The strata with TOC more than 2% are mainly located at the bottom of the Shuijingtuo Formation, which is the layer of organic matter enrichment.
Rock can be regarded as an elastic body under certain initial load, so its fracturing ability is mainly evaluated on the basis of brittle characteristics such as rock mechanical parameters (Maleki and Pouya, 2010). The mechanical parameters of rock mainly include Young modulus (E) and Poisson ratio (µ), while the deformation and stress of rock are usually linear, so the E and µ can be determined by the time difference between P-wave and S-wave (Saenger et al., 2011).
∆Tp is the offset time of P-wave, μs·m-1. ∆Ts is the offset time of S-wave, μs·m-1. E is the Young modulus, GPa. µ is Poisson ratio, dimensionless. ρb is rock density, g·cm-3.
Based on the characterization of the above rock mechanical parameters, the rock brittleness was evaluated according to the relationship between the parameters and rock brittleness. Generally, the brittleness of rock includes two meanings, the difficulty of rock fracture under the action of external force and the difficulty of rock fracture remaining open after fracturing. They are evaluated by the index of Poisson ratio and Young modulus respectively. The smaller of the Poissonratio, the more likely the rock is to break under the action of external forces. The larger of the Young modulus, the more favorable it is for the rock to keep the fractures opening after producing cracks. The brittleness of rock is usually indicated by the brittleness index (Diao, 2013). The brittleness index of rock is calculated by Young modulus and Poisson ratio respectively, and the average value of them is taken as the comprehensive brittleness index of the rock (Yang et al., 2015). The maximum and minimum values of Poisson ratio and Young modulus are determined by actual data.
Emin is the minimum value of Young modulus, GPa. Emax is the maximum Young modulus, GPa. µmin is the minimum Poisson ratio. µmax is the maximum Poisson ratio. BIE is the brittleness index of rock calculated by Young modulus, %. BIµ is the brittleness index of the rock calculated by Poisson ratio, %. BI is brittleness index of rock, %.
Figure 8 is the evaluation result of the brittleness of the shale. The results show that the main Young modulus of the shale ranges from 26.11 to 81.07 GPa, with an average of 46.08 GPa, the Poisson ratio mainly ranges from 0.15 to 0.34, with an average of 0.26 and the brittleness index of shale mainly ranges from 15.18% to 66.74%, with an average of 40.73%. Previous studies have shown that the reservoir is relatively most conducive to fracturing when Young modulus is greater than 20 GPa, Poisson ratio is less than 0.25, and rock brittleness index is greater than 40% (Jiang et al., 2010). The continuous 15 m shale at the lower part of the Shuijingtuo Formation in the study area is with the above characteristics, and the maximum value of BI is in this section (Fig. 8; Fig. 10). Therefore, this section is with the optimal brittleness condition.
Based on the analysis of above key parameters of shale reservoir, the comprehensive classification and evaluation of the shale of Shuijingtuo Formation were carried out by taking into account the mineral composition, porosity, gas bearing content, TOC and mechanical property. High shale gas production areas, especially in North America, often evaluated the reservoir quality by one or several reservoir parameters due to the complexity of geological conditions, and lack of a systematic classification standard of reservoir evaluation parameters. It is showed that the basic criteria for high-quality shale reservoirs include high brittle mineral content (more than 40%), high organic matter abundance (TOC > 2%), porosity more than 3%, gas bearing content more than 2 m3·t-1, Poisson ratio less than 0.25, and rock brittleness index more than 40% (He et al., 2015; Xiao et al., 2013). Among them, the gas bearing property of the reservoir is the most important reference factor. When the gas bearing property of the reservoir is less than 1 m3·t-1, the quality of reservoir is relatively poor (Bi et al., 2014). Therefore, combining with the basis of reservoir classification and evaluation standards from the domestic and overseas, the shale of Shuijingtuo Formation is mainly classified into three types of reservoirs in this study, and the classification criteria are shown in Table 1.
Reservoir type Criteria for the classification Type Ⅰ Gas bearing content more than 2 m3·t-1, the content of brittle minerals more than 40%, TOC more than 2%, porosity more than 3%, Poisson ratio lower than 0.25, and brittleness index more than 40% Type Ⅱ Gas bearing content more than 1 m3·t-1, the content of brittle minerals more than 40%, TOC more than 2% or porosity more than 3% or Poisson ratio lower than 0.25 or brittleness index more than 40% Type Ⅲ Gas bearing content lower than 1 m3·t-1, the content of brittle minerals lower than 40%, TOC lower than 2% or porosity lower than 3% or Poisson ratio more than 0.25 or brittleness index lower than 40%
Table 1. Criteria for the reservoirs classification of the shale of Shuijingtuo Formation
Based on the above criteria and the evaluation results of reservoir parameter, the reservoir parameters are comprehensively characterized (Fig. 11). The Type Ⅰ reservoir is mainly located at the bottom, with a total thickness of 8.5 m. Type Ⅱ is mainly located at the middle section, with a total thickness of 44.38 m. The other reservoirs are classified as Type Ⅲ, mostly located at the upper section, with a total thickness of 31.83 m. Among them, the Type Ⅰ reservoir is optimal, the Type Ⅱ is general, and the Type Ⅲ is the worst.
Figure 11. Comprehensive reservoir classification and evaluation of the shale of Shuijingtuo Formation in Well EYY1.
The Type Ⅰ is chiefly made of carbonaceous shale and the gas bearing content is about 2 to 4.37 m3·t-1, with an average of 3.16 m3·t-1. The content of brittle minerals is about 40% to 60%, 49.9% in average. TOC ranges from 2% to 5.3%, 3.6% in average. The porosity ranges from 3% to 5.24%, 4.08% in average. µ is about 0.19 to 0.25, 0.21 in average. The BI is about 40% to 60.9%, with an average of 47%. On the whole, this type reservoir is with the best quality and meets all the criteria of high quality reservoir.
The lithology of Type Ⅱ reservoir is mainly carbonaceous shale and grey shale. The gas bearing content is about 1 to 4.46 m3·t-1, with an average of 2.28 m3·t-1. The content of brittle minerals is about 11% to 56.78%, 30.9% in average. TOC ranges from 1.43% to 7.88%, 2.39% in average. The porosity is about 1.7% to 5.13%, 3.67% in average. µ is about 0.19 to 0.33, 0.25 in average. The BI is about 16.12% to 53.78%, 38.71% in average. Generally speaking, this type reservoir is next to Type Ⅰ, with the gas bearing content more than 1 m3·t-1 and at least one of the other reservoir parameters meets the above criteria.
The lithology of Type Ⅲ is primary ash shale and calcareous shale. The gas bearing content is about 0 to 1 m3·t-1, with an average of 0.55 m3·t-1. The content of brittle minerals is about 3.35% to 36.44%, 17.73% in average. TOC is about 1% to 1.98%, 1.47% in average. The porosity ranges from 0.49% to 4.22%, 2.19% in average. µ is about 0.21 to 0.34, 0.28 in average. The BI ranges from 19% to 58.63%, 42.88% in average. In general, the quality of this type reservoir is the worst. The gas bearing content is less than 1 m3·t-1, and at least one of the other reservoir parameters does not meet the above standard of high quality reservoir.