Determination of NMR T 2 Cuto ﬀ and CT Scanning for Pore Structure Evaluation in Mixed Siliciclastic–Carbonate Rocks before and after Acidiﬁcation

: Nuclear magnetic resonance (NMR) is used widely to characterize petrophysical properties of siliciclastic and carbonate rocks but rarely to study those of mixed siliciclastic–carbonate rocks. In this study, 13 di ﬀ erent core samples and eight acidiﬁed core samples selected amongst those 13 from the Paleogene Shahejie Formation in Southern Laizhouwan Sag, Bohai Bay Basin, were tested by scanning electron microscopy (SEM), micro-nano-computed tomography (CT), and NMR. SEM and CT results revealed a complex pore structure diversity, pore distribution, and pore-throat connectivity in mixed reservoirs. Sixteen groups of NMR experiments addressed changes in these properties and permeabilities of mixed siliciclastic–carbonate rocks before and after acidiﬁcation to determine its e ﬀ ects on such reservoirs. NMR experimental results showed no “di ﬀ usion coupling” e ﬀ ect in mixed siliciclastic–carbonate rocks. Distributions of NMR T 2 cuto ﬀ values ( T 2 C ) are closely related to the pore structure and lithologic characteristics before and after acidiﬁcation. The T 2 C index separates irreducible and movable ﬂuids in porous rocks and is a key factor in permeability prediction. Centrifugation experiments showed that, before acidiﬁcation, the T 2 C of mixed siliciclastic–carbonate rocks with 60–90% siliciclastic content (MSR) ranged widely from 1.5 to 9.8 ms; the T 2 C of mixed siliciclastic–carbonate rocks with 60–90% carbonate content (MCR) ranged from 1.8 to 5.6 ms. After acidiﬁcation, the T 2 C of MSR ranged widely from 2.6 to 11.6 ms, the T 2 C of MCR ranged from 1.5 to 5.6 ms, and no signiﬁcant di ﬀ erence was observed between MCR reservoirs. Based on an analysis of the morphology of NMR T 2 spectra, we propose a new T 2 cuto ﬀ value prediction method for mixed siliciclastic–carbonate rocks based on a normal distribution function to predict various T 2 C values from morphological di ﬀ erences in NMR T 2 spectra and to calculate the irreducible water saturation ( S wir ), i.e., the ratio of irreducible total ﬂuid volume to e ﬀ ective porosity. The reliability of the proposed method is veriﬁed by comparing predicted T 2 C and S wir values with those from NMR experimental results. New experiments and modeling demonstrate the applicability of NMR for the petrophysical characterization of mixed siliciclastic–carbonate rock reservoirs. Our results have potential applications for identiﬁcation and evaluation of mixed siliciclastic–carbonate rock reservoirs using NMR logging.


Introduction
The study of mixed sedimentation of siliciclastic rocks and carbonate rocks started in the 1980s. In 1984, Mount first proposed the term "mixed sediments" to describe the mixed sedimentation of siliciclastic and carbonate rocks [1]. Scholars around the world have successively conducted research on mixed sedimentary conditions [2][3][4], the main controlling factors of mixed sedimentation [3,5], mixed sedimentary sequences [6,7], and mixed sedimentary lithofacies distribution patterns [8,9]. Due to the late start of the study of mixed deposition, siliciclastic rocks and carbonate rocks of mixed deposition have traditionally been studied as independent systems, but such studies cannot effectively guide the exploration and development of oil fields with mixed siliciclastic-carbonate rock reservoirs, and there is a paucity of research on the mechanism of oil-gas migration and the significance of mineral geology in mixed sediments. Mixed siliciclastic-carbonate rocks have important geological value for oil and gas exploration and development. The success of oil and gas exploration in the Bohai Sea area of China has led scholars to conduct basic research on the petrophysical characteristics of the mixed rocks [10], such as the effective porosity, pore size distribution, pore structure, and volume of irreducible water, which are used to estimate the hydrocarbon storage capacity of mixed siliciclastic-carbonate rock reservoirs [11]. When studying the migration law of fluids in the reservoir space, the mature seepage mechanism theory is used as a reference for both siliciclastic rocks and carbonate rocks, but there is no such theory for mixed siliciclastic-carbonate rocks. Moreover, pore structure evaluation is the basis for improving the seepage mechanism on mixed siliciclastic-carbonate rocks [12]. Considering the diversity of mixed siliciclastic-carbonate rock reservoir space types, combinations, and genetics, scholars across the globe have not established a unified evaluation method for the pore structure, which greatly limits the understanding of the seepage mechanism of mixed siliciclastic-carbonate rock reservoirs and the formulation of development strategies.
Since the introduction of nuclear magnetic resonance (NMR) technology as a relatively new, rapid, and nondestructive method into the petroleum industry, this method has provided a technical means to evaluate the fluid properties and petrophysical characteristics of siliciclastic rocks, carbonate rocks, shale, and coal [13,14]. NMR has solved many problems in traditional and unconventional reservoir engineering, for example, providing reliable reservoir physical parameters, identifying formation fluids, measuring saturation accurately, and improving logging accuracy. However, the application of NMR core analysis technology in mixed siliciclastic-carbonate rocks is rarely reported in the literature [12]. The NMR T 2 spectrum distribution can provide information about the fluid distribution, pore structure, and pore volume change to obtain important reservoir evaluation parameters, such as total porosity, effective porosity, permeability, irreducible water saturation, and movable water saturation. T 2 cutoff values (T 2C ) is a necessary parameter to obtain these evaluation parameters. T 2C is the T 2 boundary value between the movable fluid and the irreducible fluid [15]. For example, using scanning electron microscopy (SEM) observations, micro-nano-computed tomography (CT), and NMR, researchers have found that siliciclastic pores usually have a typical transverse relaxation time T 2 distribution ranging from 0.01 to 100 ms, carbonate pores usually have T 2 distribution ranges from 0.01 to 1000 ms, and shale pores usually have T 2 distribution ranges from 0.01 to 10 ms [16]. Globally, scholars have conducted extensive experimental studies on the determination of T 2 cutoff values for sandstone, carbonate, and shale, but there are relatively few studies on T 2C values for mixed siliciclastic-carbonate rocks [12,17]. If there are no NMR core experimental data, Schlumberger's T 2C for medium-and high-permeability sandstone reservoirs is 33 ms, and the Numar Corporation regards this limit as being between 25 and 30 ms [18]. Other conducted research on low-permeability sandstone reservoirs has shown that T 2C has a wide distribution range, and researchers have recommended a T 2C average value of 12.85 ms. For carbonate reservoirs, Schlumberger recommended T 2C = 92 ms [19]; for shale reservoirs, Liu et al. showed that movable fluid T 2C ranged from 3.87 to 16.68 ms and recommended that the T 2C average value should be 8.29 ms [20]. The above T 2C values are not universally applicable to all reservoirs since the paramagnetic materials (such as iron and manganese) in the reservoir can affect the transverse relaxation time characteristics of NMR logging. A variable T 2C value, rather than a single value, should Energies 2020, 13,1338 3 of 29 be used to distinguish the movable fluid and irreducible fluid [21]. For this special lithology, we use the NMR research results of predecessors on siliciclastic and carbonate reservoirs to estimate the pore T 2 distribution and determine the NMR T 2C values of mixed siliciclastic-carbonate rock reservoirs with different lithologies before and after acidification. The effects of acidification on the alteration of mixed siliciclastic-carbonate rock reservoirs are evaluated by SEM, micro-nano-CT, and NMR experiments.
In contrast to traditional siliciclastic and carbonate reservoirs, the types and mineral compositions of rocks formed by mixed sedimentation of siliciclastic and carbonate rocks are very complex and are characterized by an intricate pore structure. Previous and recent studies have mainly focused on the study of macroscopic mixed sedimentation [8,22,23], while the micropore structure characterization and fluid occurrence state of mixed siliciclastic-carbonate rocks remain open questions. In this study, we performed 100% saline water saturated, cyclic centrifugation, and cyclic heat treatment on 13 samples before acidification and on eight samples after acidification. These treatments were carried out, and 21 sets of NMR experiments were performed. Combining the results of casting thin-section observations, SEM analysis, micro-nano-CT scanning, and nuclear magnetic resonance analysis, the centrifugation and normal distribution function calculated by NMR experiments were used to predict the T 2C values of siliciclastic-carbonate rocks with different mineral compositions. The pore structure characteristics and the factors that influence structure characteristics before and after acidification were also identified. Finally, due to the lack of core data in the offshore exploration area, we hope to establish a set of detailed pore structure evaluation data of mixed siliciclastic-carbonate rocks stratum in the entire well section with the help of the above experimental methods and the final research results.

Geographical Location of Core Samples
The experimental samples in this paper were taken from the Laizhouwan Sag, Bohai Bay Basin, China ( Figure 1A). Due to the difficulty in obtaining cores from an offshore drilling platform, full-diameter large core data are lacking, and most of the data are from the borehole measurement. The Bohai Bay Basin is one of the most important oil-and gas-enriched continental basins in China and is a Mesozoic-Cenozoic composite basin developed above the Paleozoic platform area of North China. After a long geological history, the current large depression Bohai Bay Basin was developed in the Neogene due to basin subsidence, and consists of many divided Paleogene fault depressions. Thus far, nearly 100 medium and small uplifts and sags have developed in the Bohai Bay Basin ( Figure 1B), such as the Weibei uplift, the Laizhouwan sag, and the Huangkekou sag.
In the Paleogene sedimentary period of the Bohai Bay Basin, the fault depression structures developed abundantly. Paleogene lake groups developed under this tectonic setting. Most of the lake basins are dustpan-like fault depressions with inclined basements; steep slopes often have multistage fault terraces, and gentle slopes often have hills and ridges controlled by anti-dip faults [24]. Combined with magmatic activity and the formation of local igneous rocks, these features often make the shape of the lake bottom complicated. In such special basins, nearly 10,000 m of sandstone, mudstone, and carbonate rock reservoirs were deposited. Among these basins, the Laizhouwan Sag is located in the southeastern part of the Bohai Bay Basin and is a Cenozoic sag in the Jiyang depression. Its eastern and western boundaries are limited by the east and west branches of the Tanlu fault zone, respectively [25]. The southern and northern parts are close to the Weibei uplift and the Laibei low uplift, the eastern part is close to the Ludong uplift, and the western part is close to the Kendong uplift. The mixed siliciclastic-carbonate rocks in the sampling area ( Figure 1C) developed under this tectonic setting.

Mixed Sedimentary Characteristics
The "mixed sediments" are mainly composed of carbonate and terrigenous clastic components [26]. Therefore, the mixed sedimentary environment should have provenance or geographical conditions with the simultaneous or alternating input of clastic rocks and carbonate rock minerals in the environment of sea-land transitional zone or coastal zone, continental lake, slope-basin, and so on [27,28]. We preliminarily determined that the source supply of the Paleogene Shahejie Formation in the sampling area mainly came from the Kendong uplift in the western part and the Weibei uplift in the southern part of the Laizhouwan Sag [29], and Wells 4, 5, and 6 in the sampling area were located on the platform of the central uplift at the early stage of sedimentation. In Figure 2, "deep cycles" indicate that, with frequent changes in the lake level [30], the sedimentary facies in the lower member of the Shahejie Formation are mainly mixed beaches and dams and that the strata often developed with mixed siliciclastic-carbonate rocks. Mixed sedimentation was primarily affected by external sources in the sampling area, which shows two types of mixed sedimentation [31]: (1) structural mixed deposition, which refers to the presence of 10-50% of clastics in carbonate rocks or siliciclastic rocks containing 10-50% carbonate, such as bed numbers Z8, Z10, Z17, Z18, Z19, Z23, Z26, and Z33 in Figure 2 (we use "bed number" to represent the same lithologic unit in a certain thickness stratum), and (2) interbedded mixed deposition, which refers to interbedded layers of siliciclastic rocks and carbonate rocks. The mixed siliciclastic-carbonate rocks in the sampling area are clastic dolomite/limestone, bioclastic dolomite/limestone, and carbonate-bearing sandstone, and the siliciclastic dolomite/limestone mainly includes argillaceous dolomite (Z8), silty dolomite (Z10), fine sandy dolomite (Z26 and Z33), and silty limestone (Z18). Carbonate-bearing sandstone mainly includes limy mudstone (Z17, Z19) and limy sandstone (Z23). Below, we use scanning electron microscopy (SEM), micro-nano-computed tomography (CT), and NMR results to characterize the pore structure of siliciclastic dolomite/limestone and limy sandstone reservoirs before and after acidification and to evaluate the effect of acidification on the siliciclastic-carbonate rock reservoirs.

Mixed Sedimentary Characteristics
The "mixed sediments" are mainly composed of carbonate and terrigenous clastic components [26]. Therefore, the mixed sedimentary environment should have provenance or geographical conditions with the simultaneous or alternating input of clastic rocks and carbonate rock minerals in the environment of sea-land transitional zone or coastal zone, continental lake, slope-basin, and so on [27,28]. We preliminarily determined that the source supply of the Paleogene Shahejie Formation in the sampling area mainly came from the Kendong uplift in the western part and the Weibei uplift in the southern part of the Laizhouwan Sag [29], and Wells 4, 5, and 6 in the sampling area were located on the platform of the central uplift at the early stage of sedimentation. In Figure 2, "deep cycles" indicate that, with frequent changes in the lake level [30], the sedimentary facies in the lower member of the Shahejie Formation are mainly mixed beaches and dams and that the strata often developed with mixed siliciclastic-carbonate rocks. Mixed sedimentation was primarily affected by external sources in the sampling area, which shows two types of mixed sedimentation [31]: (1) structural mixed deposition, which refers to the presence of 10-50% of clastics in carbonate rocks or siliciclastic rocks containing 10-50% carbonate, such as bed numbers Z8, Z10, Z17, Z18, Z19, Z23, Z26, and Z33 in Figure 2 (we use "bed number" to represent the same lithologic unit in a certain thickness stratum), and (2) interbedded mixed deposition, which refers to interbedded layers of siliciclastic rocks and carbonate rocks. The mixed siliciclastic-carbonate rocks in the sampling area are clastic dolomite/limestone, bioclastic dolomite/limestone, and carbonate-bearing sandstone, and the siliciclastic dolomite/limestone mainly includes argillaceous dolomite (Z8), silty dolomite (Z10), fine sandy dolomite (Z26 and Z33), and silty limestone (Z18). Carbonate-bearing sandstone mainly includes limy mudstone (Z17, Z19) and limy sandstone (Z23). Below, we use scanning electron Energies 2020, 13, 1338 5 of 29 microscopy (SEM), micro-nano-computed tomography (CT), and NMR results to characterize the pore structure of siliciclastic dolomite/limestone and limy sandstone reservoirs before and after acidification and to evaluate the effect of acidification on the siliciclastic-carbonate rock reservoirs. Among these mixed siliciclastic-carbonate rocks with different carbonate and clastic content, such as argillaceous dolomite (Z8), silty dolomite (Z10), silty limestone (Z12), Z16, Z17, Z18, Z19, Z22, and Z23, we use "MSR" to represent mixed siliciclastic-carbonate rocks with 60-90% clastic content, and "MCR" to represent mixed siliciclastic-carbonate rocks with 60-90% carbonate content in the following descriptions. According to the analysis of microresistivity imaging logging, the reservoir space of mixed siliciclastic-carbonate rocks consists of mainly pores and cracks in the sampling area. The pore throats are the channel for fluids flow, which are a key factor in measuring permeability [32]. The diameter of microcracks ranges from 1 to 10 µm in the studied samples. On the basis of the pore classification of siliciclastic rock and carbonate rock [33], the pores of the studied mixed siliciclastic-carbonate rock samples can be divided into micropores (<50 µm in diameter), mesopores (50-100 µm in diameter), and macropores (>100 µm in diameter).
Energies 2020, 13, x FOR PEER REVIEW 5 of 29 Among these mixed siliciclastic-carbonate rocks with different carbonate and clastic content, such as argillaceous dolomite (Z8), silty dolomite (Z10), silty limestone (Z12), Z16, Z17, Z18, Z19, Z22, and Z23, we use "MSR" to represent mixed siliciclastic-carbonate rocks with 60-90% clastic content, and "MCR" to represent mixed siliciclastic-carbonate rocks with 60-90% carbonate content in the following descriptions. According to the analysis of microresistivity imaging logging, the reservoir space of mixed siliciclastic-carbonate rocks consists of mainly pores and cracks in the sampling area. The pore throats are the channel for fluids flow, which are a key factor in measuring permeability [32]. The diameter of microcracks ranges from 1 to 10 μm in the studied samples. On the basis of the pore classification of siliciclastic rock and carbonate rock [33], the pores of the studied mixed siliciclastic-carbonate rock samples can be divided into micropores (<50 μm in diameter), mesopores (50-100 μm in diameter), and macropores (>100 μm in diameter).

Samples and Methodology
We obtained a total of 13 natural full-diameter core samples of mixed siliciclastic-carbonate rocks (Figures 2 and 3A) in the sampling area, numbered A, B, ..., L, and M. Their diameter is 70 mm, and their height is 85 mm ( Figure 3A). The lithology information of the studied samples can be found in Figure 2, "Sedimentary sequence and lithology." First, we used high-purity anhydrous gasoline as a solvent to displace the studied samples with a soxhlet extractor under normal temperature and pressure. When high-purity anhydrous gasoline flows through the pores of rocks, it dissolves and takes away the oil in the pores. Finally, the displaced core was soaked in pure benzene. Two sets of core plugs with the same size were obtained at the same bottom of natural full-diameter core samples A-M. The sampling position is shown in Figure 3A as two black circles with two red dotted lines. The height and diameter of the subsamples was 20 × 38 mm ( Figure 3B). The core plugs were prepared

Samples and Methodology
We obtained a total of 13 natural full-diameter core samples of mixed siliciclastic-carbonate rocks (Figures 2 and 3A) in the sampling area, numbered A, B, ..., L, and M. Their diameter is 70 mm, and their height is 85 mm ( Figure 3A). The lithology information of the studied samples can be found in Figure 2, "Sedimentary sequence and lithology." First, we used high-purity anhydrous gasoline as a solvent to displace the studied samples with a soxhlet extractor under normal temperature and pressure. When high-purity anhydrous gasoline flows through the pores of rocks, it dissolves and takes away the oil in the pores. Finally, the displaced core was soaked in pure benzene. Two sets of Energies 2020, 13, 1338 6 of 29 core plugs with the same size were obtained at the same bottom of natural full-diameter core samples A-M. The sampling position is shown in Figure 3A as two black circles with two red dotted lines. The height and diameter of the subsamples was 20 × 38 mm ( Figure 3B). The core plugs were prepared for NMR experiments and acidification numbered 1-A, 1-B, 1-C, 1-D, 2-E, 2-F, 2-G, 2-H, 3-I, 3-J, 3-K, 3-L, and 3-M. The core plugs were prepared for making rock thin sections and SEM before acidification numbered 1-A', 1-B', . . . , 1-D', 2-E', . . . , 2-H', 3-I', . . . , and 3-M'. For example, Sample 1-A was first subjected to NMR experiments, then acidification experiments, and finally NMR experiments, making acidified rock thin sections; the repeated use of Sample 1-A not only overcomes the errors caused by core heterogeneity, but also saves many precious cores. Furthermore, the geochemical parameters of Sample 1-A and Sample 1-A' are almost the same, thus, the two sets of subsamples heterogeneity can be neglected during photomicrographs and SEM image analysis in the paper.

Centrifugation
We put the selected samples into a high-speed centrifuge CSC-12(S) to rotate, and used the centrifugal force generated by the high-speed centrifuge as the external discharge displacement pressure to achieve the non-wet phase displacement of the wet phase. The fluid above the corresponding pore radius can be removed from the selected sample by adjusting the centrifugal force reached by the centrifuge, and the remaining fluids are called the irreducible fluids. The basic parameters of centrifugation used in the experiment were as follows: the test temperature was 35 °C, the centrifugal time was 60 min, and the centrifugal speed was gradually increased from 1000 rpm until the fluid signal was unchanged.

NMR Experiments and Parameters Calculated
Based on SEM and micro-nano CT scans, we performed nuclear magnetic resonance analysis on 13 original samples (numbered 1-A, …, 1-D, 2-E, …, 2-H, 3-I, …, 3-M) and eight acidified samples (as shown in "3.5. Acidification") using a "Recore 3100" rock sample NMR analyzer. Their height and diameter was 20 and 38 mm ( Figure 3B). First, the core was placed in a high-pressure saturator, and after vacuuming, the prepared formation water solution was injected under confining pressure until

Photomicrographs and SEM
We made 20 casting thin sections from acidified Samples 1-A~2-H and 40 casting thin sections from Samples 1-A'~3-M' and observed them using Zeiss Imager A1m 1221 and Leica DMRX 226376 polarizing microscopes. According to the rock thin section identification method SY/T 5368-2000, we observed and described the rock types, mineral compositions, and pore structure characteristics of the cast thin sections (Figure 4). Using an EVO/MA15 15-16-11 scanning electron microscope and based on the rock sample SEM analysis method SY/T 5162-1997, we observed and described the mineral types and the filling state of minerals in intergranular/intergranular pores or throats in the selected samples before acidification. When sampling by SEM, due to the looseness after washing the oil, 35 samples were taken and 28 samples were completed; we avoided gravel when sampling and focused on the dissolution phenomenon and the type of clay minerals in observation and photography.

Acidified Samples
Comparing Figure 4A'-H' reveals that the rock components and pore types can be seen to have changed significantly compared with their original appearance after the dissolution of the rock components that are easily soluble in acid, such as calcite and iron calcite containing CO3 2− . Figure 4A'-D' shows that the acidification plays a constructive role in the MSR reservoir; the pore volume inside the rock increases, and the connectivity between the pores improves. The pore characteristics of carbonate sandstones in Figure 4A reveal the most evident change and the best acidification effect. While the amounts of dolomite and calcite in Figure 4B'-D' are higher than those in Figure 4A', the rocks are denser, and the pores between the rock particles or inside the particles are mostly filled or cemented by calcite, iron calcite, and dolomite, thus, the acidification effect is not as good as that shown in Figure 4A. Figure  4B',C' show that the filling is acid-soluble calcite and iron calcite; the pore volumes become larger after acidification, and the intergranular dissolution pores and dissolution cracks located at the edges of the particles newly appear, while the connectivity of pores improves, which is beneficial to the storage and transport of fluids. While the majority of fills in Figure 4D are dolomite, the pore volume changes after acidification are not evident, but notable dissolution cracks can be observed at the edges of the rock particles. After acidification, the volume and connectivity of pores with high amounts of calcite or iron calcite change more evidently than those of pores with a high amount of dolomite in MSR samples. In the MCR samples, Figure 4E,F have more terrigenous debris and calcite, and Figure 4E',F' show that the dissolution pores increase evidently because of the mixed distribution of clastic materials and calcite among dolomite particles; however, the change in pore connectivity is not evident. For mixed siliciclasticcarbonate rocks with low clastic content ( Figure 4G,H), the rock is dense, and the combined characteristics

Micro-Nano CT Scanning
First, full-diameter core samples (Samples B' and E') were selected for micro-nano-CT scanning ( Figure 3A). These samples were obtained from the upper part of Samples B and E, and their height and diameter were 58 and 70 mm ( Figure 3C). The lithology of Sample B' is limy sandstone, and that of Sample E' is sandy dolomite. Since Samples 1-B and 2-E were loosely deformed after acidification ( Figure 3B), acidified samples were not subjected to CT scanning. In this study, the CT testing instrument was a GE v|tome|x m 300 CT scanner that imported the scanned data into the data processing workstation and reconstructed a digital three-dimensional model. The professional software packages Volume Graphics Studio Max and Fei Avizo were used for data processing on the reconstructed three-dimensional model, the statistical analysis of sample pores, and throat and simulated permeability data.
A small feature surface could be fully displayed through a large amount of image data without destroying Samples B' and E'. Figure 3C shows that the pore structure and relative density of Samples B' and E' are positively correlated with the grey levels of the three-dimensional (3D) CT images. The resolution of Samples B' and E' scanning volume is 0.80 µm, and the highest scanning accuracy is 50 nm. From the top section ( Figure 3D), the tangent plane ( Figure 3E), and the side section ( Figure 3F) of the obtained full-diameter core column, we can visually observe the interior pore structure development characteristics of the mixed rock. From the top slice of the core section ( Figure 3D), the right slice ( Figure 3E), and the front slice ( Figure 3F) of the full-diameter core, we can directly observe the Energies 2020, 13, 1338 8 of 29 pore structure development characteristics in the mixed siliciclastic-carbonate rocks. The pores are developed, and the connectivity between the pores is good. Red pores represent large effective pore volumes, and blue pores represent large invalid pore volumes in Figure 3C-F, and the pores are cemented or filled during diagenesis.

Centrifugation
We put the selected samples into a high-speed centrifuge CSC-12(S) to rotate, and used the centrifugal force generated by the high-speed centrifuge as the external discharge displacement pressure to achieve the non-wet phase displacement of the wet phase. The fluid above the corresponding pore radius can be removed from the selected sample by adjusting the centrifugal force reached by the centrifuge, and the remaining fluids are called the irreducible fluids. The basic parameters of centrifugation used in the experiment were as follows: the test temperature was 35 • C, the centrifugal time was 60 min, and the centrifugal speed was gradually increased from 1000 rpm until the fluid signal was unchanged.

NMR Experiments and Parameters Calculated
Based on SEM and micro-nano CT scans, we performed nuclear magnetic resonance analysis on 13 original samples (numbered 1-A, . . . , 1-D, 2-E, . . . , 2-H, 3-I, . . . , 3-M) and eight acidified samples (as shown in "3.5. Acidification") using a "Recore 3100" rock sample NMR analyzer. Their height and diameter was 20 and 38 mm ( Figure 3B). First, the core was placed in a high-pressure saturator, and after vacuuming, the prepared formation water solution was injected under confining pressure until the pores of the core were completely saturated by the formation water. Core magnetic resonance experiments were carried out, and the T 2 spectrum distribution of 100% saline water saturated and centrifuged samples was measured. The basic parameters of NMR used in the experiment were as follows: the inter-echo spacing was 0.2 ms (TE), the salinity was 25,000 mg/l, the test temperature was 35 • C, the waiting time was 6 s, the echo interval was 0.6 ms, the number of scans was 128, and the number of echoes was 2048. The surface magnetization of the mixed siliciclastic-carbonate rocks was uniform, and the pores were isolated, thus, the relaxation process of each pore was independent, which shows that the relaxation time is related to the distribution of pore size. The transverse relaxation time of the samples can be expressed as [14]: where T 2 is the transverse relaxation time of the pore fluid, ms; T 2B , T 2S , and T 2D are the bulk, surface, and diffusion transverse relaxation time, respectively, and their units are ms; ρ 2 is the transverse surface relaxivity, µm/ms; S is the surface area of the pore space, cm 2 ; V is the pore space volume, cm 3 ; D is the fluid diffusion coefficient, µm 2 /ms; γ is the gyromagnetic ratio, rad/(ms·Gs); G is the magnetic gradient, Gs/m; t E is the echo spacing, ms. When the mixed siliciclastic-carbonate rocks are fully saturated with water, the G and t E values are low, T 2B and T 2D values are much larger than T 2S , and both can be neglected. Under these conditions, Equation (1) is a function of surface relaxation time only, and can be simplified as Equation (2): The T 2 distribution of NMR can be used to characterize the pore size distribution of rocks [34]. The theoretical basis is that, during the NMR measurement, hydrogen protons collide with the surface of the particles to cause energy attenuation, which is called surface relaxation; the specific surface (S/V, unit is µm −1 ) of macropores is low, the probability of collision between hydrogen protons and particle surfaces is small, the relaxation time in T 2 spectra is long, and the S/V of micropores is large; therefore, Energies 2020, 13, 1338 9 of 29 the probability of collision between hydrogen protons and particle surfaces is high, and the relaxation time of T 2 spectra is short [35,36]. T 2 reflects the size of the specific surface within the rock pores, which is proportional to the pore radius and fluid content [37]. The movable fluid and irreducible fluid saturation of the mixed siliciclastic-carbonate rocks were quantitatively calculated by the change in water volume, T 2 increment, and T 2 cumulative distribution before centrifugation (saturated water state) and after centrifugation (irreducible water state). NMR parameters such as sample weight (g) and NMR porosity (%) were used to calculate the irreducible fluid porosity (%) and BVI/S wir (%). In a typical T 2 spectrum, a boundary value for the T 2 cutoff is required, which divides the T 2 spectrum into a movable fluid portion and an irreducible fluid portion. The portion of the T 2 spectrum larger than T 2C represents the movable fluid, and the portion of the T 2 spectrum smaller than T 2C represents the irreducible or immobile fluid. If the proportion of the T 2 spectrum on the left side of T 2C is large and the proportion on the right side is small, then the mixed siliciclastic-carbonate rock sample has high irreducible water saturation and low movable water saturation [38,39]. Therefore, when the porosity value on the saturation accumulation curve is equal to the irreducible fluid porosity, the T 2 transverse relaxation time is the T 2C value.

Acidification
We obtained the X-ray diffraction samples by crushing the top of the core plugs, 1-A'~3-M', removed the mud contaminated part, and carried out an X-ray diffraction analysis of clay minerals and common non-clay minerals using a D/max-2500 diffractometer. The lithologic name was given through the composition and content of various minerals in the samples. According to the content of calcite and dolomite from low to high (approximately from 10% to 80%), eight natural core samples that have a full diameter and that are morphologically intact were selected for acidification experiment and numbered 1-A, 1-B, 1-C, 1-D, 2-E, 2-F, 2-G, and 2-H. During the acidification process, (1) the simulated reservoir temperature was 70 • C; (2) the experimental fluid included KCl brine with a salinity of 25,000 mg/L (same as the formation water salinity); (3) according to the holoclastic rock analysis on studied samples and acidification experience [12], we selected 7.5% HCl as the prepositioned acid and 6.0% HCl + 1.5% HF as the host acid. During the experiment, (1) the dry samples were weighed; (2) the KCl brine with the same salinity of the formation water was prepared first, then held for one day and filtered; (3) the studied samples were placed in the coreflood holder, and the confining pressure was maintained at 3-4 MPa; (4) formation water was injected to measure permeability, and the flow rate was 0.2 mL·min −1 ; (5) the sample was saturated under pressure and then weighed again; (6) the pre-acid 1 PV was injected to react for 0.5 h at a 70 • C reservoir temperature at a flow rate of 0.2 mL·min −1 ; (7) the host acid 1.75 PV was injected to react for 0.5 h at a 70 • C reservoir temperature; (8) 4% NH 4 Cl was injected to measure permeability; (9) the core was removed, and photographs were taken to record the core appearance after acidification. The acidified samples were used to make casting thin sections and NMR analysis.

Original Samples
The information from the original sample rock thin sections shows that the pores of MSR are well developed ( Figure 4A) and unevenly distributed ( Figure 4B), with good connectivity, particle line contact ( Figure 4C), and particle support ( Figure 4D). The rock composition is mainly quartz and potassium feldspar, with small amounts of quartzite, muscovite fragments, and plagioclase, where the muscovite has a long axis orientation distribution ( Figure 4A); the interstitial material is mainly mud, calcite, and dolomite, with a small amount of iron calcite. The mud is distributed around the particles ( Figure 4A) or locally enriched ( Figure 4B), the calcite is evenly distributed between the particles ( Figure 4B), and the iron calcite fills in the cracks of the particles and the intergranular dissolution pores ( Figure 4C). The phenomenon of metasomatic siliciclastic particles shows that dolomite is enriched and distributed in the zonal phase ( Figure 4B,D). The pore types of the MSR in the sampling area are mainly intergranular pores ( Figure 4A), with a small number of secondary intergranular pores and dissolution pores in feldspar grains ( Figure 4A,D). The secondary intergranular pores were formed by dissolution of authigenic cement filled between intergranular pores. Through statistical analysis of samples, we found that, when the carbonate content among rock components was 26-40%, the pores were not well developed ( Figure 4B-D), and only a small number of intragranular dissolution pores were observed.
MCR mainly includes terrigenous clastic micrite dolomite rocks ( Figure 4E,G), terrigenous clastic micrite limestone dolomite rocks ( Figure 4F) and terrigenous clastic micrite limestone ( Figure 4H). MCR composition mainly consists of terrigenous clastic material and micrite dolomite ( Figure 4E,G), with small amounts of internal debris, calcite, and mud, or mainly dolomite and calcite, with a small amount of muddy and terrigenous debris ( Figure 4F). The terrigenous debris is mainly potassium feldspar ( Figure 4F), quartz, and debris ( Figure 4E,G). The terrigenous clastic materials are mostly distributed in dolomite or calcite ( Figure 4E-G), and the argillaceous mixture is distributed in the micrite dolomite or enriched in the micrite calcite layer ( Figure 4H). Figure 4E,F show that calcite in the terrigenous clastic micrite dolomite is the filling material after dolomite crushing and that the calcite grains are star-shaped. Some terrigenous clastic micrite limestone with a laminar texture is present in the sampling area, and its structure is special: micrite and calcite are interbedded, and micrite calcite is oriented with irregular strips of enrichment ( Figure 4H). The texture of MCR is dense, and pores are not developed. Except for the microcracks, pores, and holes shown in Figure 4E,H, no effective pores are observed ( Figure 4F,G), which differ from the pores, holes, and cracks developed in the carbonate rocks. Because the carbonate rocks in this paper are affected by mixed sedimentation, the structure changes, and the pores, holes, and cracks are usually filled or cemented by calcite and clastic materials.

Acidified Samples
Comparing Figure 4A'-H' reveals that the rock components and pore types can be seen to have changed significantly compared with their original appearance after the dissolution of the rock components that are easily soluble in acid, such as calcite and iron calcite containing CO 3 2− . Figure 4A'-D' shows that the acidification plays a constructive role in the MSR reservoir; the pore volume inside the rock increases, and the connectivity between the pores improves. The pore characteristics of carbonate sandstones in Figure 4A reveal the most evident change and the best acidification effect. While the amounts of dolomite and calcite in Figure 4B'-D' are higher than those in Figure 4A', the rocks are denser, and the pores between the rock particles or inside the particles are mostly filled or cemented by calcite, iron calcite, and dolomite, thus, the acidification effect is not as good as that shown in Figure 4A. Figure 4B',C' show that the filling is acid-soluble calcite and iron calcite; the pore volumes become larger after acidification, and the intergranular dissolution pores and dissolution cracks located at the edges of the particles newly appear, while the connectivity of pores improves, which is beneficial to the storage and transport of fluids. While the majority of fills in Figure 4D are dolomite, the pore volume changes after acidification are not evident, but notable dissolution cracks can be observed at the edges of the rock particles. After acidification, the volume and connectivity of pores with high amounts of calcite or iron calcite change more evidently than those of pores with a high amount of dolomite in MSR samples. In the MCR samples, Figure 4E,F have more terrigenous debris and calcite, and Figure 4E',F' show that the dissolution pores increase evidently because of the mixed distribution of clastic materials and calcite among dolomite particles; however, the change in pore connectivity is not evident. For mixed siliciclastic-carbonate rocks with low clastic content ( Figure 4G,H), the rock is dense, and the combined characteristics of rock particles are evidently changed after acidification ( Figure 4G',H'). The pores are still not developed, and the effect of acidification on pore transformation is poor.

Filling state of minerals in pores or throats
The unpolished SEM images show that intergranular pores (Por) and intragranular dissolution pores (Idp) were developed in the MSR; thus, I/S ( Figure

Filling state of minerals in pores or throats
The unpolished SEM images show that intergranular pores (Por) and intragranular dissolution pores (Idp) were developed in the MSR; thus, I/S ( Figure   The MCR mainly develop micropores and a small number of dissolution pores (Dp) and microcracks. The connectivity between pores is evidently weaker than that of MSR. The main reason is that mineral particles such as I/S ( Figure 5F,G), authigenic quartz ( Figure 5H), and potassium feldspar ( Figure 5I) are filled in the throat or pore. The pores of MSR contain many clay minerals ( Figure 5A-C, I/S, Kln; Figure 5D, chl; Figure 5E, I11) with much smaller particle sizes and other mineral particles ( Figure 5A, Py), which are attached to the surface of rock particles without external force [29]. Under certain external force, these particles may be separated from the wall of pores and transported with the fluid in the pores [40,41]. When clay minerals move to the throat position, some particles are trapped and deposited, causing blockage of the throat. The water-sensitive clay minerals in MSR and MCR, such as I/S, illite, chlorite, and kaolinite, are not high, nor is the degree of water sensitivity damage. Therefore, the fluid migration between pores and throats in MSR is still smooth.

Pore and Throat Size Distribution Characteristics from CT
We intercepted a 45 × 45 × 45 mm cube ( Figure 6) in the 3D stereogram of pore distribution in Figure 3C to analyze the distribution characteristics of pores and throats. When the resolution of the obtained images reaches 1 μm, the porosity of the 3D digital core ( Figure 3C) is still smaller than the laboratory measurement, indicating that there are micropores in the samples that are smaller than The MCR mainly develop micropores and a small number of dissolution pores (Dp) and microcracks. The connectivity between pores is evidently weaker than that of MSR. The main reason is that mineral particles such as I/S ( Figure 5F,G), authigenic quartz ( Figure 5H), and potassium feldspar ( Figure 5I) are filled in the throat or pore. The pores of MSR contain many clay minerals ( Figure 5A-C, I/S, Kln; Figure 5D, chl; Figure 5E, I11) with much smaller particle sizes and other mineral particles ( Figure 5A, Py), which are attached to the surface of rock particles without external force [29]. Under certain external force, these particles may be separated from the wall of pores and transported with the fluid in the pores [40,41]. When clay minerals move to the throat position, some particles are trapped and deposited, causing blockage of the throat. The water-sensitive clay minerals in MSR and MCR, such as I/S, illite, chlorite, and kaolinite, are not high, nor is the degree of water sensitivity damage. Therefore, the fluid migration between pores and throats in MSR is still smooth.

Pore and Throat Size Distribution Characteristics from CT
We intercepted a 45 × 45 × 45 mm cube ( Figure 6) in the 3D stereogram of pore distribution in Figure 3C to analyze the distribution characteristics of pores and throats. When the resolution of the obtained images reaches 1 µm, the porosity of the 3D digital core ( Figure 3C) is still smaller than the laboratory measurement, indicating that there are micropores in the samples that are smaller than the CT scan resolution [42]. The porosity of the 3D digital Sample B' is 2.51%, and that of Sample E' is 2.64%. The laboratory measurement porosity of Sample B' is 2.43%, and that of Sample E' is 2.60%. The above results indicate that microporosity was not developed and would not affect the network connectivity of digital cores.
Energies 2020, 13, x FOR PEER REVIEW 12 of 29 the CT scan resolution [42]. The porosity of the 3D digital Sample B' is 2.51%, and that of Sample E' is 2.64%. The laboratory measurement porosity of Sample B' is 2.43%, and that of Sample E' is 2.60%. The above results indicate that microporosity was not developed and would not affect the network connectivity of digital cores.   Figure 6C,C1 and Figure 6D,D1 are the CT images of the pores and throats of the Sample E' sandy dolomite rock. The pores in Sample B' are well developed, and "water marks" are often distributed around the pores ( Figure 6A). The pores are mostly concentrated in various parts of the rock body ( Figure 6A1). The throats in Sample B' are well developed ( Figure  6B,B1); the volume of a single throat is large, but the number is small. Sample E' is relatively dense ( Figure 6C), and the pores are distributed irregularly in various parts of the rock body. Most of the pores are micropores, and their number is large ( Figure 6C1). The throats in sandy dolomite rocks are scattered across almost the entire cube ( Figure 6D1).

NMR T2 Distribution of Different Samples
Under the conditions of no overlying pressure, the NMR results (Table 1) of 16 core samples before and after acidification under 100% saturated water (Sw) and irreducible water conditions (Swir) are shown in Figures 7 and 8. For example, Sample 1-A and Sample 1-A' were obtained from the same part of Sample A. The overall T2 spectral distribution is shown as a unimodal distribution; some mixed siliciclastic-carbonate rocks with high carbonate or clastic content have bimodal distributions, which are different from the T2 spectral morphologies of siliciclastic rocks and carbonate rocks. For all samples, the T2 spectra at 100% saturated water (Sw) are shown as black lines, while the T2 spectra at irreducible water (Sir) are shown as red dotted lines ( Figure 7A). The curve of 100% irreducible water cumulative porosity represents the accumulation of all porosity values during centrifugation [43]; these values could be used to calculate the T2 cutoff value and the movable fluid's porosity. According to the relationship between T2 relaxation times, larger T2 (T2 > T2C) values usually represent larger pores and pore radii, which store the movable fluids [4].  Figure 6A,B,A1,B1 are the CT images of the pores and throats of Sample B' limestone sandstone rock, respectively. Figure 6C,C1,D,D1 are the CT images of the pores and throats of the Sample E' sandy dolomite rock. The pores in Sample B' are well developed, and "water marks" are often distributed around the pores ( Figure 6A). The pores are mostly concentrated in various parts of the rock body ( Figure 6A1). The throats in Sample B' are well developed ( Figure 6B,B1); the volume of a single throat is large, but the number is small. Sample E' is relatively dense (Figure 6C), and the pores are distributed irregularly in various parts of the rock body. Most of the pores are micropores, and their number is large ( Figure 6C1). The throats in sandy dolomite rocks are scattered across almost the entire cube ( Figure 6D1).

NMR T 2 Distribution of Different Samples
Under the conditions of no overlying pressure, the NMR results (Table 1) of 16 core samples before and after acidification under 100% saturated water (S w ) and irreducible water conditions (S wir ) are shown in Figures 7 and 8. For example, Sample 1-A and Sample 1-A' were obtained from the same part of Sample A. The overall T 2 spectral distribution is shown as a unimodal distribution; some mixed siliciclastic-carbonate rocks with high carbonate or clastic content have bimodal distributions, which are different from the T 2 spectral morphologies of siliciclastic rocks and carbonate rocks. For all samples, the T 2 spectra at 100% saturated water (S w ) are shown as black lines, while the T 2 spectra at irreducible water (S ir ) are shown as red dotted lines ( Figure 7A). The curve of 100% irreducible water cumulative porosity represents the accumulation of all porosity values during centrifugation [43]; these values could be used to calculate the T 2 cutoff value and the movable fluid's porosity. According to the relationship between T 2 relaxation times, larger T 2 (T 2 > T 2C ) values usually represent larger pores and pore radii, which store the movable fluids [4].  Figure 7 shows the T2 spectral distribution of MSR. Before centrifugation, the T2 distributions of Figure 7A,B are unimodal, and the T2 distributions of Figure 7C,D are bimodal, where the right peak is connected with the left peak and the amplitude of the right peak NMR signal is much smaller than that of the left peak. Figure 8 shows the T2 spectral distribution of MCR. The T2 distributions of Figure  8A,B are unimodal before acidification; importantly, the T2 spectral distributions are bimodal after acidification ( Figure 8A'-8D'), which is usually caused by the dissolution of calcite or iron calcite in rocks to form macropores or microcracks. Nuclear magnetic resonance T 2 spectrum distribution of the selected mixed carbonate-siliciclastic rocks. (A-D): Table 1. spectrum distribution before acidification; (A'-D'): T 2 spectrum distribution after acidification. Figure 7 shows the T 2 spectral distribution of MSR. Before centrifugation, the T 2 distributions of Figure 7A,B are unimodal, and the T 2 distributions of Figure 7C,D are bimodal, where the right peak is connected with the left peak and the amplitude of the right peak NMR signal is much smaller than that of the left peak. Figure 8 shows the T 2 spectral distribution of MCR. The T 2 distributions of Figure 8A,B are unimodal before acidification; importantly, the T 2 spectral distributions are bimodal after acidification ( Figure 8A'-D'), which is usually caused by the dissolution of calcite or iron calcite in rocks to form macropores or microcracks.

Acidification Experiment Evaluation
The lithology of Samples 1-A, 1-B, 1-C, and 1-D is MSR. The lithology of Samples 2-E, 2-F, 2-G, and 2-H is MCR. The basic physical parameters and the acidification experimental results of the selected samples are shown in Table 2. The displacement breakthrough pressures represent the pressures required for the samples to displace the liquid at different stages of the experiment. The breakthrough pressure of formation water displacement before acid injection is generally higher than that after acid injection. The permeability of formation water before acid injection is generally lower than that after acid injection. After displacement by pre-acid and host acid, the displacement pressure is gradually reduced. After Sample 2-H was displaced by the pre-acid and the host acid, the core as a whole became loose and deformed, and the dissolution pores may have been filled again; thus, the liquid permeability was not measured [44]. Therefore, after the mixed siliciclastic-carbonate rock samples were treated with the pre-acid and host acid, the apparent morphology of the core appeared loose or even slag-like, the displacement pressure was gradually reduced, and the liquid permeability of the core improved. For example, the displacement breakthrough pressures of MCR are generally higher than those of MSR (carbonate content 10-40%), and the acidification effect of MCR (more than 75% carbonate content) is not evident.

Acidification Experiment Evaluation
The lithology of Samples 1-A, 1-B, 1-C, and 1-D is MSR. The lithology of Samples 2-E, 2-F, 2-G, and 2-H is MCR. The basic physical parameters and the acidification experimental results of the selected samples are shown in Table 2. The displacement breakthrough pressures represent the pressures required for the samples to displace the liquid at different stages of the experiment. The breakthrough pressure of formation water displacement before acid injection is generally higher than that after acid injection. The permeability of formation water before acid injection is generally lower than that after acid injection. After displacement by pre-acid and host acid, the displacement pressure is gradually reduced. After Sample 2-H was displaced by the pre-acid and the host acid, the core as a whole became loose and deformed, and the dissolution pores may have been filled again; thus, the liquid permeability was not measured [44]. Therefore, after the mixed siliciclastic-carbonate rock samples were treated with the pre-acid and host acid, the apparent morphology of the core appeared loose or even slag-like, the displacement pressure was gradually reduced, and the liquid permeability of the core improved. For example, the displacement breakthrough pressures of MCR are generally higher than those of MSR (carbonate content 10-40%), and the acidification effect of MCR (more than 75% carbonate content) is not evident.

"Diffusion Coupling" Effect
Unlike the ideal distribution model of the NMR T 2 spectrum ( Figure 9A), a "diffusion coupling" effect usually occurs in carbonate rocks [45,46], i.e., fluids between pore throats with different radii are interconnected by diffusion, which makes distinguishing the T 2 distribution of large pores and micropores' throats in the same carbonate rock [47] sample difficult, as shown in the non-ideal type model of the T 2 spectrum distribution ( Figure 9B, black dotted line). Is there a "diffusion coupling" effect in the mixed siliciclastic-carbonate rocks? Can the T 2 spectral distribution reflect the pore size distribution of the siliciclastic-carbonate rocks? We discuss these issues below.

"Diffusion Coupling" Effect
Unlike the ideal distribution model of the NMR T2 spectrum ( Figure 9A), a "diffusion coupling" effect usually occurs in carbonate rocks [45,46], i.e., fluids between pore throats with different radii are interconnected by diffusion, which makes distinguishing the T2 distribution of large pores and micropores' throats in the same carbonate rock [47] sample difficult, as shown in the non-ideal type model of the T2 spectrum distribution ( Figure 9B, black dotted line). Is there a "diffusion coupling" effect in the mixed siliciclastic-carbonate rocks? Can the T2 spectral distribution reflect the pore size distribution of the siliciclastic-carbonate rocks? We discuss these issues below. When comparing the CT scan results ( Figure 6) and NMR T2 spectra distribution characteristics ( Figures 7B and 8A) of Samples 1-B (mixed siliciclastic-carbonate rock with more clastic content) and 2-E (with more carbonate content), we do not find that the T2 spectra of different pore throats are completely merged into one peak. The pore radius range of Sample 1-B indicates that micropores and mesopores are dominant, whereas macropores are scarce; the pore radius range of Sample 2-E reveals that micropores are dominant, and no macropores are found. The NMR T2 distribution of Sample 1-B shows that the T2 distribution ranges from 0.1 to 200 ms, and the sample contains macropores ( Figure 7B); the NMR T2 distribution of Sample 1-E ranges from 0.1 to 15 ms, and the sample is dominated by mesopores ( Figure 8A). Figure 6B1 and Figure 6D1 show that the throat distributions between the macropores and the mesopores in the sample are distributed irregularly and that the connectivity is weak. The throat radii are small, mostly ranging from 0 to 5 μm, and fluid diffuses When comparing the CT scan results ( Figure 6) and NMR T 2 spectra distribution characteristics ( Figures 7B and 8A) of Samples 1-B (mixed siliciclastic-carbonate rock with more clastic content) and 2-E (with more carbonate content), we do not find that the T 2 spectra of different pore throats are completely merged into one peak. The pore radius range of Sample 1-B indicates that micropores and mesopores are dominant, whereas macropores are scarce; the pore radius range of Sample 2-E reveals that micropores are dominant, and no macropores are found. The NMR T 2 distribution of Sample 1-B shows that the T 2 distribution ranges from 0.1 to 200 ms, and the sample contains macropores ( Figure 7B); the NMR T 2 distribution of Sample 1-E ranges from 0.1 to 15 ms, and the sample is dominated by mesopores ( Figure 8A). Figure 6B1,D1 show that the throat distributions between the macropores and the mesopores in the sample are distributed irregularly and that the connectivity is weak. The throat radii are small, mostly ranging from 0 to 5 µm, and fluid diffuses with difficulty between macropores and micropores. However, if the surface relaxation is weak, the fluid in the macropores and mesopores of the mixed siliciclastic-carbonate rocks can easily diffuse, and the "diffusion coupling" effect in the rock is evident. Thus, as Ramakrishnan suggests, according to the Monte Carlo random walk model simulation [48], when determining whether the relaxation time can reflect the pore size distribution, the transverse surface relaxation rate ρ 2 is crucial. In this paper, an NMR experiment proved that the ρ 2 of the mixed siliciclastic-carbonate rocks generally ranges from 0 to 10 µm/s, while the NMR surface relaxation of the saturated water rock is strong (Figures 7 and 8). Therefore, considering the particularity of the pore structure (mainly mesopores) of the samples, we find that the mixed siliciclastic-carbonate rocks do not exhibit a "diffusion coupling" effect in the sampling area.

T 2 Cutoff Value Determination and Pore Structure
In the description in the introduction, we mention the ranges of T 2C values for sandstone and carbonate rocks without NMR experiments, but variable T 2C values rather than single values are often used to distinguish between movable and irreducible fluids [17,19]. T 2C is the relaxation time threshold that divides the T 2 spectrum into two parts: movable water and irreducible water. In this section, we use three conventional methods (the centrifugal method, the T 2 spectral morphology method, and the T 2 parameter method) to calculate T 2C values of the mixed siliciclastic-carbonate rocks and discuss the advantages and limitations of these methods. We also propose a T 2C prediction method based on the normal distribution function, which can predict various T 2C values based on the morphological differences in the NMR T 2 spectra. The centrifugation method is obtained by comparing and analyzing the T 2 integral curve of saturated water samples before and after centrifugation. The T 2 spectrum method involves observing the relationship between the T 2 distribution and the location of T 2C , and its main morphological characteristics are the number of wave peaks, the location of wave peaks in the time domain, the area ratios on both sides of wave peaks, the ratios of peak values between several wave peaks, and the locations of the troughs; a standard for determining the positions of the T 2C values is thus established. The T 2 parameter method establishes an average value of the T 2C based on the NMR core experimental data. Figure 10 shows the process of determining the T 2C value for the T 2 spectra of saturated water and irreducible water after centrifugation, when the echo interval time is 0.2 ms. The sample is first saturated with 100% water, and the T 2 distribution of saturated water is measured by NMR. The effective porosity value (MPHI) and cumulative porosity curve are obtained. Then, under the given pressure conditions, the free water is removed from the rock sample by dewatering, and only the irreducible water is left in the pore space. The T 2 distribution of irreducible water is measured by NMR, and the bound water pore volume value (MBVI) and the cumulative porosity curve are obtained. According to these observations, a parallel line is drawn (Figure 10

T2 spectral Morphology Method
In this study, the T2C values of 16 NMR samples were determined by the morphological method. The T2 spectra of the study area can be divided into unimodal and bimodal distributions, and each type has universality. Without centrifugation of the selected mixed siliciclastic-carbonate rocks, we

T 2 spectral Morphology Method
In this study, the T 2C values of 16 NMR samples were determined by the morphological method. The T 2 spectra of the study area can be divided into unimodal and bimodal distributions, and each type has universality. Without centrifugation of the selected mixed siliciclastic-carbonate rocks, we established a criteria for determining the T 2C value of the samples with similar T 2 distribution curves (at 100% water saturation), lithology, and physical characteristics in terms of the relationship between the crest value (F) and the T 2C values calculated by the method described in Section 5.2.1.

Unimodal Morphological
Before acidification, the T 2 spectra of eight samples in the saturated water state show single wave peak distributions, accounting for 61.5% of all samples, of which three rock samples were MCR (Figures 7 and 8). According to the patterns, the unimodal distributions can be classified into three types ( Figure 11).
Energies 2020, 13, x FOR PEER REVIEW 19 of 29 are both one order of magnitude smaller than those of siliciclastic rocks and carbonate rocks, and micropores and mesopores develop in the mixed siliciclastic-carbonate rocks. Among these rocks, the micropores in MCR are especially well developed, which is different from the macropores that are well developed in carbonate rocks. Figure 11. T2 spectra of unimodal morphological at 100% water saturation.

Bimodal morphology
Before acidification, the T2 spectra of five samples in the saturated water state show bimodal distributions, accounting for 38.5% of all samples, and the NMR signal intensity of the left peak is significantly higher than that of the right peak. According to the distribution pattern, the bimodal distributions can be classified into two types. The first type has no evident boundary between the left peak and right peak, as shown in Figure 12, Curve A and Curve B. Only the shape of the left peak is prominent; the right peak is low and flat, and the trough is not evident. However, there is an evident inflection point between the left peak and right peak, and the T2C value point is in the middle position between this inflection point and the crest value point of the left peak. Curve A in Figure 12 represents the T2 spectral curve of the mixed siliciclastic-carbonate rocks with the lowest carbonate content ( Figure 12). The crest value of the left peak is approximately 2.68 ms in the center, and the right peak is not evident. Curve B in Figure 12 represents the T2 spectral curve of MCR; the main peaks are single peaks, double peaks rarely occur, and the crest value of the left peak is approximately 1.55 ms in the center.
In the second type, the left peak and right peak are clearly separated, and each of them is a normal symmetric distribution with an evident trough between the two peaks. This class can be subdivided into two types according to the size of the crest value. (1) Curve C in Figure 12 represents the T2 spectral curve of MSR. The crest value of the left peak is approximately 1.55 ms in the center, the crest value of the right peak is approximately 41.6 ms in the center, and the NMR signal intensities of the left and right peaks are not widely different. When the pore component of the left peak divided by the pore component of the right peak is less than 2.5, the T2C value is at the position of 1.20F on the right side of the left crest value. (2) Curve D in Figure 12 also represents the T2 spectral curve of MSR. (1) The first type is a completely symmetric normal distribution (Figure 11, Curve A and Curve B). Curve A represents the T 2 spectral curve of MSR, where the crest value (F) of MSR is approximately 13.89 ms in the center, and the T 2C value is at the position of 0.83F on the left side of the crest value; 0.83F is obtained by 11.53 ms (T 2C )/13.89 ms (F); Curve B represents the T 2 spectral curve of MCR, where the crest value of MCR is approximately 1.29 ms in the center and the T 2C value is at the position of 1.44F on the right side of the crest value. (2) The second type is the left drag asymmetric distribution ( Figure 11, Curve C), where a dragging phenomenon exists on the left side of the T 2 spectral curve. The crest value of Curve C in Figure 10 is approximately 3.22 ms in the center, and the T 2C value is at the position of 1.72F on the right side of the crest value. (3) The third type is the right drag asymmetric distribution ( Figure 11, Curve D), where a dragging phenomenon exists on the right side of the T 2 spectral curve. The crest value of Curve D in Figure 10 is approximately 1.86 ms in the center, and the T 2C value is at the position of 2.08F on the right side of the crest value. Through curve characteristic analysis, we can see that the amounts of clastic materials in the unimodal morphological distribution of MCR are ordered as follows: left drag asymmetric ( Figure 11, Curve C) > right drag asymmetric ( Figure 11, Curve D) > symmetric normal distribution (Figure 11, Curve B). The T 2 spectrum of MSR is a completely symmetric normal distribution (Figure 11, Curve A). The crest value of the T 2 spectrum with a unimodal distribution and the range of T 2C values are both one order of magnitude smaller than those of siliciclastic rocks and carbonate rocks, and micropores and mesopores develop in the mixed siliciclastic-carbonate rocks. Among these rocks, the micropores in MCR are especially well developed, which is different from the macropores that are well developed in carbonate rocks.

Bimodal Morphology
Before acidification, the T 2 spectra of five samples in the saturated water state show bimodal distributions, accounting for 38.5% of all samples, and the NMR signal intensity of the left peak is significantly higher than that of the right peak. According to the distribution pattern, the bimodal distributions can be classified into two types. The first type has no evident boundary between the left peak and right peak, as shown in Figure 12, Curve A and Curve B. Only the shape of the left peak is prominent; the right peak is low and flat, and the trough is not evident. However, there is an evident inflection point between the left peak and right peak, and the T 2C value point is in the middle position between this inflection point and the crest value point of the left peak. Curve A in Figure 12 represents the T 2 spectral curve of the mixed siliciclastic-carbonate rocks with the lowest carbonate content (Figure 12). The crest value of the left peak is approximately 2.68 ms in the center, and the right peak is not evident. Curve B in Figure 12 represents the T 2 spectral curve of MCR; the main peaks are single peaks, double peaks rarely occur, and the crest value of the left peak is approximately 1.55 ms in the center.
Energies 2020, 13, x FOR PEER REVIEW 20 of 29 ( Figure 12, Curve B), the ratio of the left peak is larger than that of the right peak in the T2 spectrum, which indicates that the mixed siliciclastic-carbonate rock samples have high irreducible water saturation and low movable fluid saturation. The crest value of the T2 spectrum with a unimodal distribution and the range of T2C values are also both one order of magnitude smaller than those of siliciclastic rocks and carbonate rocks; macropores or cracks are not very well developed, but they often develop in the MSR with T2 values ranging from 27.5 to 258.6 ms.  Table 1 shows the weight parameters of samples after drying (WD), water saturation (WS), centrifugation (WC), and water (WIW). The formula Swir = (WC − WD) / (WS − WD) is used to calculate the irreducible water saturation, and Swir is then used to reverse the T2C value. The T2C value calculated using this method has the highest accuracy, but the cost is high, a large amount of core NMR experimental data support is required, and this method is not widely available. In this study, we use the parameters of the NMR effective porosity, the comprehensive index of porosity and permeability, the T2 geometric mean, etc., to fit the normal distribution function of the T2C value. For data with large errors in irreducible water saturation, the singular value data are first removed, and the remainder is then fitted. Before the singular value data are eliminated, the fitting results in Table 3 show that the T2C values of different types of mixed siliciclastic-carbonate rocks are closely related to the comprehensive physical property index, such as √(K × φ), √(K/φ), T2 GM, and LG (T2 GM). Their correlation coefficient R 2 values are large, and the correlations are good. With increasing reservoir permeability, the comprehensive index of pore structure increases, and T2C values also increase. After the singular value data are eliminated, T2 GMV or LG (T2 GMV) has a slight difference in fitting correlation; the differences in NMR porosity (φ), permeability (K), √(K × φ), and √(K/φ) are evidently larger; their correlation coefficient R 2 values are significantly smaller; and the fitting formula of the T2C value is not highly reliable and can be excluded. Therefore, using the fitting formula with the T2 GM value and the T2C value, the correlation coefficient is the highest, and the T2C value of mixed In the second type, the left peak and right peak are clearly separated, and each of them is a normal symmetric distribution with an evident trough between the two peaks. This class can be subdivided into two types according to the size of the crest value. (1) Curve C in Figure 12 represents the T 2 spectral curve of MSR. The crest value of the left peak is approximately 1.55 ms in the center, the crest value of the right peak is approximately 41.6 ms in the center, and the NMR signal intensities of the left and right peaks are not widely different. When the pore component of the left peak divided by the pore component of the right peak is less than 2.5, the T 2C value is at the position of 1.20F on the right side of the left crest value. (2) Curve D in Figure 12 also represents the T 2 spectral curve of MSR. The crest value of the left peak is approximately 1.29 ms in the center, the crest value of the right peak is approximately 28.9 ms in the center, and the NMR signal intensity of the right peak is significantly smaller than that of the left peak. When the pore component of the left peak divided by the pore component of the right peak is more than 2.5, the T 2C value point is in the middle position between the trough point and the crest value point of the left peak. Among these curves, we can see that the amounts of carbonate in the unimodal morphological distributions of MSR are as follows: Curve C > Curve D > Curve A. While the unimodal morphology is rarely found in the T 2 spectrum of MCR (Figure 12, Curve B), the ratio of the left peak is larger than that of the right peak in the T 2 spectrum, which indicates that the mixed siliciclastic-carbonate rock samples have high irreducible water saturation and low movable fluid saturation. The crest value of the T 2 spectrum with a unimodal distribution and the range of T 2C values are also both one order of magnitude smaller than those of siliciclastic rocks and carbonate rocks; macropores or cracks are not very well developed, but they often develop in the MSR with T 2 values ranging from 27.5 to 258.6 ms. Table 1 shows the weight parameters of samples after drying (W D ), water saturation (W S ), centrifugation (W C ), and water (W IW ). The formula S wir = (W C − W D )/(W S − W D ) is used to calculate the irreducible water saturation, and S wir is then used to reverse the T 2C value. The T 2C value calculated using this method has the highest accuracy, but the cost is high, a large amount of core NMR experimental data support is required, and this method is not widely available. In this study, we use the parameters of the NMR effective porosity, the comprehensive index of porosity and permeability, the T 2 geometric mean, etc., to fit the normal distribution function of the T 2C value. For data with large errors in irreducible water saturation, the singular value data are first removed, and the remainder is then fitted. Before the singular value data are eliminated, the fitting results in Table 3 show that the T 2C values of different types of mixed siliciclastic-carbonate rocks are closely related to the comprehensive physical property index, such as

NMR Parameter Method
√ (K/ϕ), T 2 GM, and LG (T 2 GM). Their correlation coefficient R 2 values are large, and the correlations are good. With increasing reservoir permeability, the comprehensive index of pore structure increases, and T 2C values also increase. After the singular value data are eliminated, T 2 GMV or LG (T 2 GMV) has a slight difference in fitting correlation; the differences in NMR porosity (ϕ), permeability (K),

√
(K × ϕ), and √ (K/ϕ) are evidently larger; their correlation coefficient R 2 values are significantly smaller; and the fitting formula of the T 2C value is not highly reliable and can be excluded. Therefore, using the fitting formula with the T 2 GM value and the T 2C value, the correlation coefficient is the highest, and the T 2C value of mixed siliciclastic-carbonate rocks can be obtained intuitively and conveniently. With a large number of statistical analyses, a cubic polynomial fitting relationship between the parameters T 2C and T 2 GM is found to be the best ( Figure 13). Unlike carbonate rocks and sandstones, mixed siliciclastic-carbonate rocks can be subjected to a power function and exponential function to fit the T 2C values, but the power function and exponential function of the mixed siliciclastic-carbonate rocks with poor fitting correlations cannot be used to calculate the T 2C value. In this paper, we used a cubic polynomial to calculate the T 2C value: Due to the number of NMR experimental samples, the data points that can be used for T 2 GM values are limited. To improve the prediction accuracy of T 2C values, the method described in Section 5.2.2 is necessary to predict the T 2C value. In this way, determining a T 2C value that varies with the T 2 GM index across the entire well is possible such that the T 2C values of the mixed siliciclastic-carbonate rocks can be predicted in the study area, and the movable fluid saturation and irreducible fluid saturation can be determined. mixed siliciclastic-carbonate rocks can be subjected to a power function and exponential function to fit the T2C values, but the power function and exponential function of the mixed siliciclastic-carbonate rocks with poor fitting correlations cannot be used to calculate the T2C value. In this paper, we used a cubic polynomial to calculate the T2C value: Due to the number of NMR experimental samples, the data points that can be used for T2 GM values are limited. To improve the prediction accuracy of T2C values, the method described in Section 5.2.2 is necessary to predict the T2C value. In this way, determining a T2C value that varies with the T2 GM index across the entire well is possible such that the T2C values of the mixed siliciclastic-carbonate rocks can be predicted in the study area, and the movable fluid saturation and irreducible fluid saturation can be determined.

Micro-nano-computed Tomography (CT)
For Sample B', most of the pore radii are shorter than 400 μm, 98% of the pore radii range from 1 to 100 μm ( Figure 14A), and mesopores are often observed in Figure 5A; the throat radii range from 10 to 125 μm, and 80% of the throat radii are distributed from 10 to 50 μm, showing a normal distribution ( Figure 14B).
For Sample E', most of the pore radii are shorter than 300 μm, and 90% of the pore radii range from 10 to 50 μm ( Figure 14C). The throat radii are very short; most of the throat radii are less than 25 μm, and 90% of the throat radii range from 1 to 5 μm ( Figure 14D). The experimental results show that the sizes of the pores and throat radii of the limy sandstone rocks are larger than those of the sandy dolomite rocks. (1) The pore radii of the MSR range from 1 to 100 μm, the throat radii range from 10 to 50 μm, and the distribution of pores and throats is concentrated in the rock body ( Figure 6A1,B1). (2) The pore radii of the MCR range from 20 to 50 μm, the throat radii range from 1 to 5 μm, and the distribution of pores and throats is irregularly distributed in the rock body ( Figure 6C1,D1).

Micro-nano-computed Tomography (CT)
For Sample B', most of the pore radii are shorter than 400 µm, 98% of the pore radii range from 1 to 100 µm ( Figure 14A), and mesopores are often observed in Figure 5A; the throat radii range from 10 to 125 µm, and 80% of the throat radii are distributed from 10 to 50 µm, showing a normal distribution ( Figure 14B).  For Sample E', most of the pore radii are shorter than 300 µm, and 90% of the pore radii range from 10 to 50 µm ( Figure 14C). The throat radii are very short; most of the throat radii are less than 25 µm, and 90% of the throat radii range from 1 to 5 µm ( Figure 14D). The experimental results show that the sizes of the pores and throat radii of the limy sandstone rocks are larger than those of the sandy dolomite rocks. (1) The pore radii of the MSR range from 1 to 100 µm, the throat radii range from 10 to 50 µm, and the distribution of pores and throats is concentrated in the rock body ( Figure 6A1,B1). (2) The pore radii of the MCR range from 20 to 50 µm, the throat radii range from 1 to 5 µm, and the distribution of pores and throats is irregularly distributed in the rock body ( Figure 6C1,D1).
The throat of MSR was well developed; the throats were connected with each other in a reticular morphology ( Figure 15A), and the pores were effectively connected in series. Thus, the pore-throat connectivity of MSR will evidently improve, and it is beneficial in that it will increase the efficiency of fluid migration between pores. The permeability of MSR was good. Although the throats of MCR were well developed, the throat size was lower than that of MSR; the throats of MCR are distributed independently, endowing the pores with weak connectivity with each other ( Figure 15B). Therefore, the throats of MCR could not connect the pores effectively in series. The permeability of MCR was poor.  The throat of MSR was well developed; the throats were connected with each other in a reticular morphology ( Figure 15A), and the pores were effectively connected in series. Thus, the pore-throat connectivity of MSR will evidently improve, and it is beneficial in that it will increase the efficiency of fluid migration between pores. The permeability of MSR was good. Although the throats of MCR were well developed, the throat size was lower than that of MSR; the throats of MCR are distributed  Figure 8A-D). The above results show that the pore sizes of MSR are mainly micropores and mesopores, with a few macropores and microcracks; the pore size of MCR is mainly micropores. The greater the carbonate content in mixed siliciclastic-carbonate rocks, the smaller the pore size.  Figure 8A'-D'). Before centrifugation, compared to the T 2 distribution of non-acidified samples, the range of MSR T 2 distributions had not changed greatly, while the right peak NMR signal of MSR evidently increased ( Figure 7C',D'), with better throat connectivity. In contrast, the range of MCR T 2 distributions became larger, and a few mesopores appeared in MCR, but the right peak or the left peak NMR signal of MCR became weaker ( Figure 7G',H'), with less movable fluid.

Parameter Comparison of Non-acidified and Acidified Samples
The changes in parameters representing the characteristics of the pore structure can be observed in Table 1, such as NMR porosity, permeability, and T 2 cutoff value. By comparing the numerical changes in different parameters before and after acidification, we found that most data values increased after acidification. However, the data in MCR were not evident, and certain parameters were even smaller ( Figure 16, in the blue dotted line). This phenomenon indicates that acidification has various effects on the pore size, throat size, pore connectivity, and pore hydrophilicity of mixed siliciclastic-carbonate rocks. Figure 16 shows that the NMR porosity value ( Figure 15A) and permeability value ( Figure 16B) of MSR were significantly higher than those of MCR. That is, the higher the content of clastic material in the mixed siliciclastic-carbonate rocks, the higher the NMR parameter values were, and the more favorable the pore structure was for the storage and transport of fluids.
the pore sizes of MSR are mainly micropores and mesopores, with a few macropores and microcracks; the pore size of MCR is mainly micropores. The greater the carbonate content in mixed siliciclasticcarbonate rocks, the smaller the pore size.
NMR T2 Distribution of Acidified Samples After acidification, the T2 distributions of Samples 1-A', 1-B', 1-C', and 1-D' range from 0.1 to 1000 ms, and all the maximum values of T2 are more than 80 ms ( Figure 7A'-D'). The T2 distributions of Samples 2-E', 2-F', 2-G', and 2-H' range from 0.1 to 80 ms, and most of the maximum values of T2 are less than 20 ms ( Figure 8A'-D'). Before centrifugation, compared to the T2 distribution of nonacidified samples, the range of MSR T2 distributions had not changed greatly, while the right peak NMR signal of MSR evidently increased ( Figure 7C',D'), with better throat connectivity. In contrast, the range of MCR T2 distributions became larger, and a few mesopores appeared in MCR, but the right peak or the left peak NMR signal of MCR became weaker ( Figure 7G',H'), with less movable fluid.

Parameter Comparison of Non-acidified and Acidified Samples
The changes in parameters representing the characteristics of the pore structure can be observed in Table 1, such as NMR porosity, permeability, and T2 cutoff value. By comparing the numerical changes in different parameters before and after acidification, we found that most data values increased after acidification. However, the data in MCR were not evident, and certain parameters were even smaller ( Figure 16, in the blue dotted line). This phenomenon indicates that acidification has various effects on the pore size, throat size, pore connectivity, and pore hydrophilicity of mixed siliciclastic-carbonate rocks. Figure 16 shows that the NMR porosity value ( Figure 15A) and permeability value ( Figure 16B) of MSR were significantly higher than those of MCR. That is, the higher the content of clastic material in the mixed siliciclastic-carbonate rocks, the higher the NMR parameter values were, and the more favorable the pore structure was for the storage and transport of fluids.

NMR and after Centrifugation
In the NMR experiment with the mixed siliciclastic-carbonate rocks, the T 2C values calculated by centrifugation for the 16 samples ranged from 1.55 to 9.64 ms ( Table 1). These values are significantly lower than those of sandstone rocks and carbonate rocks [12] and similar to those of coal seams and shale. The difference in the amounts of clastic and carbonate materials in the mixed siliciclastic-carbonate rocks leads to great differences in T 2C values. Before acidification, the T 2C values of MSR ranged from 1.5 to 9.8 ms, and the T 2C values of MCR ranged from 1.8 to 5.6 ms ( Figure 17). After acidification, the T 2C values of MSR ranged from 2.6 to 11.6 ms, and the T 2C values of MCR ranged from 1.5 to 5.6 ms ( Figure 17). The T 2C values of MSR and MCR were of the same order of magnitude before and after acidification, and the MSR T 2C values increased, whereas the MCR T 2C values did not change greatly. Comparing all the T 2C values reveals that the T 2C values of the same type of mixed siliciclastic-carbonate rock also differed due to differences in pore characteristics ( Figure 7C,D). For MCR, the acidification process did not cause the rock to form macropores or cracks, and pore connectivity also deteriorated. Among these samples, the acidification effects on Sample 2-G sandy dolomite ( Figure 4G) and Sample 2-H sandy limestone ( Figure 4H) with low siliciclastic content were the worst. After comparing the T 2C values calculated in Figures 7 and 8 with the pore characteristics identified in Figure 4 thin sections, we found that for MSR ( Figure 4A-D) and MCR ( Figure 4E-H) with micropores or mesopores or dissolved pores, T 2C values were larger than those for rocks with micropores or dense rocks. For example, fluids in the pores of sandy limestone/dolomite were more resistant to centrifugation than those in the pores of other types of rocks. magnitude before and after acidification, and the MSR T2C values increased, whereas the MCR T2C values did not change greatly. Comparing all the T2C values reveals that the T2C values of the same type of mixed siliciclastic-carbonate rock also differed due to differences in pore characteristics ( Figure 7C,D). For MCR, the acidification process did not cause the rock to form macropores or cracks, and pore connectivity also deteriorated. Among these samples, the acidification effects on Sample 2-G sandy dolomite ( Figure 4G) and Sample 2-H sandy limestone ( Figure 4H) with low siliciclastic content were the worst. After comparing the T2C values calculated in Figure 7 and Figure 8 with the pore characteristics identified in Figure 4 thin sections, we found that for MSR ( Figure 4A-D) and MCR ( Figure 4E-H) with micropores or mesopores or dissolved pores, T2C values were larger than those for rocks with micropores or dense rocks. For example, fluids in the pores of sandy limestone/dolomite were more resistant to centrifugation than those in the pores of other types of rocks. After centrifugation, the T2 spectrum was compared with the T2 spectrum of Sw (Figures 7 and  8). The first peak and the second peak of Samples 1-A, 1-B, 1-C, 1-D, 1-A′, 1-B′, 1-C′, and 1-D′ were weakened or even disappeared before and after acidification (Figure 7). The amounts of weakening shown in Figure 7A,A',D are most evident. The change in the second peak usually represents the presence of movable fluids in macropores or cracks [49]. All the peaks of Samples 2-E, 2-F, 2-G, 2-H, 2-E′, 2-F′, 2-G′, and 2-H′ were weakened after acidification (Figure 8). The weakening shown in Figure  8A is the most evident. The higher the amounts of clastic materials and calcite are, the more evident the decreases in the peaks are. The above results show that the storage spaces in MSR are mainly mesopores, with a few macropores and microcracks; the storage space in MCR is mainly micropores, and the pore radius scale is lower than that of MSR, with few macropores and microcracks developed in the rock after acidification. Figure 16B shows that the permeability of MSR increased after acidification, but the permeability of MCR could not be measured, and the T2C value of MCR decreased or remained unchanged. In general, carbonate rocks are more sensitive to acid than siliciclastic rocks. The acidification treatment After centrifugation, the T 2 spectrum was compared with the T 2 spectrum of S w (Figures 7 and 8). The first peak and the second peak of Samples 1-A, 1-B, 1-C, 1-D, 1-A , 1-B , 1-C , and 1-D were weakened or even disappeared before and after acidification (Figure 7). The amounts of weakening shown in Figure 7A,A',D are most evident. The change in the second peak usually represents the presence of movable fluids in macropores or cracks [49]. All the peaks of Samples 2-E, 2-F, 2-G, 2-H, 2-E , 2-F , 2-G , and 2-H were weakened after acidification (Figure 8). The weakening shown in Figure 8A is the most evident. The higher the amounts of clastic materials and calcite are, the more evident the decreases in the peaks are. The above results show that the storage spaces in MSR are mainly mesopores, with a few macropores and microcracks; the storage space in MCR is mainly micropores, and the pore radius scale is lower than that of MSR, with few macropores and microcracks developed in the rock after acidification. Figure 16B shows that the permeability of MSR increased after acidification, but the permeability of MCR could not be measured, and the T 2C value of MCR decreased or remained unchanged. In general, carbonate rocks are more sensitive to acid than siliciclastic rocks. The acidification treatment dissolved the calcite or iron calcite particles ( Figure 4A-D and Figure 5D) in MSR, and the dolomite particles were not dissolved ( Figure 4B',D' and Figure 17C). There was no evident change in the combination relationship of particle distribution ( Figure 4A'-D'), and the dissolution pore space was evidently enlarged. As shown in Figure 17A, kaolinite exists in book-like form, but the soil-like morphology of kaolinite easily absorbs water [50]. Thus, in book-like form, kaolinite does not easily absorb water [51]. Therefore, the lower kaolinite content in MSR had little effect on permeability. MCR is denser ( Figure 4E-H), and the fills in the intergranular pore are not easily acidified ( Figure 4E',F'). However, after acidification, the combined relationship of the primary dolomite and calcite particles changed evidently, and the layered MCR was loose and deformed ( Figure 4G',H'). Although the pore space volume experienced little change, the throats were filled with acidified material or the intergranular extrusion deformation disappeared, resulting in significant decreases in permeability and T 2C value, and the volume of movable fluid decreased. The acidified pore structure constrained fluid migration and formed "bound pores."

Acidification is Useful or Useless for Mixed Siliciclastic-Carbonate Rocks Reservoir
According to the experimental results and data analysis in the above sections, there are evident differences in the petrophysical characteristics of different types of mixed siliciclastic-carbonate rocks. Among them, the main T 2 value distribution ranges from 0.1 to 200 ms, the T 2C value ranges from 0.1 to 12 ms, and micropores are developed in mixed siliciclastic-carbonate rocks. For example, the T 2C value of MSR was higher than that of MCR (Figures 6 and 7); the pore radii in MSR were mainly in the range from 1 to 100 ms, and the throat radii were mainly in the range of 10-50 microns; the pore radii in MCR were mainly in the range from 20 to 50 ms, and the throat radii were mainly in the range from 1 to 5 ms (Figures 4-6). The effect of acidification on pore structure also depended on the type of mixed siliciclastic-carbonate rock. After acidification, the physical properties and seepage of MSR reservoirs were good, and the T 2C and S wir values predicted by the normal distribution function showed that acidification plays a constructive role in the petrophysical characteristics of MSR reservoirs. However, the physical properties of MCR reservoirs did not change greatly, although the grain structure inside the rock changed evidently, and the T 2C value decreased, which restricted fluid migration. Therefore, we suggest that the reservoir sections developed in MSR can be acidified to improve the pore structure and the fluid migration efficiency; the reservoir sections developed in MCR are not recommended for acidification. The scale of experimental samples was small in this study. Next, we will study the influence of the combined relationship of different types of siliciclastic-carbonate rocks in the strata on the seepage behavior, which has potential application value in the exploration and development of mixed siliciclastic-carbonate rock reservoirs.

Conclusions
A series of displacement, SEM, CT, and NMR experiments were carried out on 21 lacustrine mixed siliciclastic-carbonate rock samples, and this study provides a new method for determining the T 2C value and evaluating the pore structure of mixed siliciclastic-carbonate rocks. The main results are as follows: (1) After acidification, the displacement breakthrough pressure of mixed siliciclastic-carbonate rock samples gradually decreased, the liquid permeability of the core increased, and the displacement breakthrough pressure of mixed siliciclastic-carbonate rocks with 60-90% siliciclastic content (MSR) was generally lower than that of mixed siliciclastic-carbonate rocks with 60-90% carbonate content (MCR). The acidification effect on MCR was not evident.
(2) In this paper, the results of cast thin-section image observation, SEM image analysis, and CT scans accurately revealed the pore and throat grades of mixed siliciclastic-carbonate rocks. Among these rocks, the pore and throat grade of MSR were generally higher than those of MCR, while the throat connectivity of MCR was better than that of MSR.
(3) The T 2 spectral distribution characteristics of mixed siliciclastic-carbonate rocks depended on the amounts of clastic materials and carbonates. At 100% water saturation, the NMR T 2 spectrum distribution of MSR is usually bimodal and the surface relaxation time ranges from 0.1 to 1000 ms. The NMR T 2 spectrum distribution of MCR is usually unimodal, and the surface relaxation time ranges from 0.1 to 15 ms.
(4) In this study, the centrifugal method, the T 2 spectral morphological method, and the T 2 parameter method were used to estimate the T 2C values of mixed siliciclastic-carbonate rocks, which can be used along with the T 2 spectrum to distinguish irreducible fluids and movable fluids. The T 2C values determined by these methods agree well. Among these distributions, the T 2C obtained from the T 2 parameter method had a significant positive correlation with T 2 GM. For the selected mixed siliciclastic-carbonate rock samples, before acidification, the T 2C of MSR ranged from 1.5 to 9.8 ms, and the T 2C of MCR ranged from 1.8 to 5.6 ms. After acidification, the T 2C of MSR ranged from 2.6 to 11.6 ms, and the T 2C of MCR ranged from 1.5 to 5.6 ms.
(5) Acidification plays a constructive role in the petrophysical properties of MSR reservoirs. The pore volume increased, the throat connectivity improved, and the T 2C value became larger.