Comparative Porosity and Pore Structure Assessment in Shales: Measurement Techniques, Influencing Factors and Implications for Reservoir Characterization
Porosity and pore size distribution (PSD) are essential petrophysical parameters controlling permeability and storage capacity in shale gas reservoirs. Various techniques to assess pore structure have been introduced; nevertheless, discrepancies and inconsistencies exist between each of them. This study compares the porosity and PSD in two different shale formations, i.e., the clay-rich Permian Carynginia Formation in the Perth Basin, Western Australia, and the clay-poor Monterey Formation in San Joaquin Basin, USA.
Yujie Yuan and Reza Rezaee
Western Australian School of Mines, Minerals, Energy and Chemical Engineering, Curtin University, Perth WA 6845, Australia.
Received: 9 May 2019; Accepted: 29 May 2019; Published: 31 May 2019
Porosity and PSD have been interpreted based on nuclear magnetic resonance (NMR), low-pressure N2 gas adsorption (LP-N2-GA), mercury intrusion capillary pressure (MICP) and helium expansion porosimetry.
The results highlight NMR with the advantage of detecting the full-scaled size of pores that are not accessible by MICP, and the ineffective/closed pores occupied by clay bound water (CBW) that are not approachable by other penetration techniques (e.g., helium expansion, low-pressure gas adsorption and MICP). The NMR porosity is largely discrepant with the helium porosity and the MICP porosity in clay-rich Carynginia shales, but a high consistency is displayed in clay-poor Monterey shales, implying the impact of clay contents on the distinction of shale pore structure interpretations between different measurements. Further, the CBW, which is calculated by subtracting the measured effective porosity from total porosity, presents a good linear correlation with the clay content (R2 = 0.76), implying that our correlated equation is adaptable to estimate the CBW in shale formations with the dominant clay type of illite.
The increasing demand of unconventional energy resources raises the significance of shale reservoir investigation [1,2]. Shales are defined as the laminated fine-grained argillaceous sedimentary rock, which are essentially constituted by minerals involving silt-sized particles (4–62.5 µm) and clays (<4 µm) in couple with organic matter (OM) [3,4,5,6].
The porosity and pore size distribution (PSD), performing as the most fundamental pore structure parameters to estimate gas storage capacity and fluid transporting behaviour in shale complex pore structures [4,7,8,9], are significantly associated with clay minerals and the promising OM that significantly varies between different shale formations [10,11,12,13,14].
The clay mineral or OM develops the micropore (i.e., pores smaller than 2 nm per International Union of Pure and Applied Chemistry (IUPAC) classification ) and mesopore (i.e., pores ranging from 2 nm to 50 nm per IUPAC classification ) system, complicating shale pore structures and resulting in the extremely low permeability, low porosity and the large distinction of PSD in shales.
To date, three types of laboratory techniques are applied for pore characterization or quantification. Microscopy techniques, e.g., transmission electron microscopy (TEM) and scanning electron microscopy (SEM), perform as the helpful petrographic-imaging approaches for porosity estimation , however, provide objective results and are not adaptable to cover the full range of PSD in shales . Radiation Scatterings, such as small angle neutron scattering (SANS) and ultra-small angle scattering (USANS) techniques, are capable to quantify the continuous PSD in tight sandstones  and coals [18,19]. However, the applications in shale systems are still under debate due to the limitation of neutron sources [2,20,21,22].
Fluid penetration methods, i.e., low-pressure (<18.4 psi) CO2 gas adsorption (LP-CO2-GA), low-pressure N2 gas adsorption (LP-N2-GA), mercury intrusion capillary pressure (MICP) and nuclear magnetic resonance (NMR), enable a wide range of pore structure detection and have been universally utilized in shale research studies [23,24,25,26,27,28,29,30,31]. However, MICP displays destructive disadvantages and is not approachable to the pore throat sizes smaller than 3.6 nm [3,11,32,33], merely inter-communicated pores are available for detection . Helium expansion is attainable to the connected pore space corresponding to effective porosity, while the acquisition of PSD is not available .
LP-CO2-GA coupled with LP-N2-GA is approachable to the pore sizes ranging from 0.35 nm to 200 nm . However, only interconnected pores are accessible , and the results are sensitive to measurement procedures and highly dependent on the sample pre-treatment such as the dewatering/ outgassing temperatures and the size of the smashed shale fragments [36,37,38]. NMR, which is acknowledged as a non-destructive technique, is adaptable for measuring the total porosity and PSD in shales [4,27,39,40,41].
Unlike the conventional rocks displaying consistent results in porosity and PSD among different fluid-penetration measurements , shales, however, tend to reveal significant discrepancies. For example, MICP porosity in Barnett shales exhibit ~25–50% lower when compared to helium porosity values . Similar porosity inconsistencies up to 50% have also been found in previous studies [32,44,45]. To fully understand the variations of shale pore structure interpretation between different measurements, the comprehensive techniques are highly required to be combined and compared in parallel.
This paper discusses the discrepant results of different measurements for the two shales in typical composition (i.e., the Carynginia shales of the Perth Basin in Western Australia and the Monterey shales of San Joaquin Basin in the U.S). Porosity is compared based on MICP, NMR and helium expansion. PSD is interpreted based on MICP, NMR, and low-pressure gas adsorption. The influencing factors are discussed for result discrepancies. Implications are provided for shale gas reservoir characterization.
Materials and Methods
Shale samples from two formations were analyzed and compared between different measuring techniques. Carynginia samples, by the name of “AC1-AC8”, were collected from Arrowsmith well in the Perth Basin, Western Australia. Monterey shales, by the name of “M1-B-M10-B” and “M1-M6”, came from well-1B and well-1, respectively, in the San Joaquin Basin, USA. Geological settings of Carynginia and Monterey shale formation were displayed in other studies [27,46].
Table 1 shows the mineralogical composition in Carynginia and Monterey shales. Carynginia shales are characterized by abundant clay minerals, constituting 31.1–50.8 wt % of the total mineral contents (e.g., the average value of Carynginia clay content is 36.6 wt %). The quartz contents occupy 35.6–53.2 wt % (e.g., quartz averages in 45.17 wt %), while the minorities are shown in K-feldspar, plagioclase and other minerals. Monterey shales present a low-clay content (e.g., the mean value of the clay content is around 9.0 wt %) but a relatively high proportion in quartz content. The clay type in both of Carynginia and Monterey shales have been identified as illite [27,47].
Table 1. XRD mineralogical composition for shales from Carynginia and Monterey formation. Some data were collected from the other studies.
Nuclear Magnetic Resonance (NMR)
Prior to NMR experiments, Carynginia shale plugs of 1.5′ diameter were cleaned with toluene/methanol mixture, and fully saturated with 30,000 ppm brine that are matched with the average formation salinity. The low-field NMR measurements were performed on saturated samples by using 2 MHz Magritek Rock Core Analyzer, which was set under 30°C with P54 probe and conducted under the constant magnetic resonance frequency. NMR T2 spectrum was acquired by using the experimental parameters, i.e., 100 µs inter-echo spacing (TE), 10,000 ms inter-experiment delay, 10,000 number of echoes and the minimum 200 signal to noise ratio (SNR), coupled with the Carr-Purcell-Meilboom-Gill sequence [48,49,50].
Applying NMR T2 spectrum to study shales pore structure is fundamentally established on the transverse relaxation dominated by surface relaxation mechanism :
where T2 is the transverse relaxation time; ρ2 is surface relaxivity, which is considered as a constant value representing the strength of surface relaxation; S V is the surface volume ratio that is closely intimated with pore sizes. Pore size distribution could be interpreted via T2 spectrum, with smaller pore sizes corresponding to shorter relaxation times.
Low-Pressure Gas Adsorption (LP-GA)
Low-pressure N2 gas adsorption (LP-N2-GA) was applied to measure the pore size distribution (PSD) and the pore volume. Prior to the measurements, shale samples were crushed into fragments of 60 mesh sizes and degassed over 8h for pore surface cleaning. LP-N2-GA was performed on Micromeritics® TriStar 3020 instrument at the bathing temperature of 77.4 K. N2 was penetrated into the degassed samples under the constant temperature for the acquisition of adsorption isotherm. PSD is obtained by using the Barrett, Joyner and Halenda (BJH) theory based on N2 adsorption isotherm . The interpretations were carried out on the embedded TriStar II 3020 standard software.
Mercury Injection Capillary Pressure (MICP)
MICP measurements were performed on Micromeritics Autopore IV 9500 V1.09 porosimeter on chip samples weighing around 10 g. Prior to the test, shale chips were evacuated under the pressure of 70 µmHg for 10 min. The non-wetting mercury, as the working probe for pore access, follows the parameters applied as: Hg density of 13.53 g/mL; adv. and rec. contact angle of 130°; Hg surface tension of 485 dynes/cm. The mercury filling pressure was performed of 0.51 psia under 10 s for equilibration, followed by injection under high pressure, ranging from 0.1 MPa (14.5 psi) to the maximum 413.7 MPa (60,000 psi), which corresponds to the pore throat size from 3.6 nm to 1100 µm.
The pore throat size distribution of tested samples is obtained using Washburn equation assuming cylindrical pores (Equation (2)) :
where ri is the pore throat radius calculated under mercury pressure of P c (psi), µm; σ is the mercury surface tension (485 dynes/cm applied in the test); θ is mercury contact angle (130° applied in the test); p c is the injection pressure ranging from 14.5 psi to the maximum 60,000 psi.
The porosity values obtained from three measuring techniques (i.e., MICP, Helium, and NMR) are shown for two different shale formations (i.e., Carynginia and Monterey) (Figure 1). An obvious porosity distinction is displayed in NMR between the clay-rich samples (i.e., Carynginia) and the clay-poor samples (i.e., Monterey). An overall higher NMR porosities are exhibited in Carynginia compared to Monterey. In addition, the porosity discrepancies are apparently exhibited between NMR and the other two measurements in Carynginia samples.
Carynginia presents the highest porosity value in NMR, which is more than two times as MICP porosity, and about three times as helium porosity (i.e., the porosity measured by NMR, helium, and MICP ranges in 8.02–12.87%, 3.03–3.78%, and 1.93–4.15%, respectively). However, the Monterey exhibits a high porosity consistency in NMR, helium and MICP (Figure 1).
Figure 1. The porosity values obtained from mercury intrusion capillary pressure (MICP), Helium, and nuclear magnetic resonance (NMR) for two different shale formations.
The porosity measured from MICP and helium demonstrates high consistencies in both Carynginia and Monterey. As shown in Figure 2, the cross-plot of helium porosity versus MICP porosity generates a very good positive linear relationship, with the correlation coefficient (R2) of 0.93. The porosity values for each sample are shown in Appendix A (Table A1).
Figure 2. The cross-plot of Helium porosity (%) versus MICP porosity (%) for the studied samples.
The Pore Size Distribution from NMR
Figure 3 presents the NMR T2 spectrum in Carynginia shales, with the majority of pores identified in small pore sizes. The peak values of T2 curves correspond to the T2 relaxation time around 0.3–1 ms. The samples of higher clay contents, e.g., AC1 and AC2 (i.e., 50.8% and 43.2%), exhibit larger amplitude and narrower spectrum with the peak value locating in smaller pore sizes. The samples of relatively lower clay contents, e.g., AC8 and AC4 (i.e., 32.3% and 31.1%), display smaller amplitude and wider distributions, presenting a general larger pore sizes. A uniform pore size distribution is commonly indicated in Carynginia shales.
Figure 3. NMR T2 spectrum for Carynginia shales (i.e., AC1-AC8).
NMR T2 spectra of Monterey shales, i.e., M1-M6 and M1-B-M10-B, are displayed in Figure 4 and Figure 5, respectively. Significant variations are demonstrated in Monterey shales compared to Carynginia shales. As shown in Figure 4, Monterey shales from Well 1 (except for M5) exhibit the major pores in larger pore size ranges.
Figure 4. NMR T2 spectrum for Monterey shales (i.e., M1–M6) collected from Well 1. Modified from Rivera .
The peak locations of the spectrum correspond to T2 relaxation time ~1–100 ms, coupled with an overall wider spectrum range, indicating a general uneven pore size distributions. The peak values (except for M5) correspond to the incremental porosity between 0.1% and 0.13%, displaying overall lower values than that in Carynginia shales. M5 shows the trimodal spectrum associated with three typical pore types. The spectrum peak of M5 occurs at T2 relaxation time in 300–500 ms, representing the majority of larger pores or fractures.
Figure 5. NMR T2 spectrum for Monterey shales (i.e., M1-B–M6-B) collected from Well 1B. Modified from Rivera .
Figure 5 shows the T2 spectrum of Monterey shales from Well 1B with multiple modal types. M4-B, M10-B, M7-B, M9-B exhibit a majority of small pore sizes, corresponding to the T2 relaxation times of ~0.3–1 ms. M1-B and M2-B with semi-modal distributions present the PSD peaks at larger T2 relaxation time (i.e., 14 ms and 40 ms respectively) and higher incremental porosity (i.e., 0.27% and 0.36% respectively). M3-B exhibits trimodal spectrum with the main pore size locates at ~20 ms. When compared to T2 spectrum in clay-rich Carynginia, the pore sizes in Monterey are rather unevenly distributed and universally locating in larger pore sizes. Moreover, unlike Carynginia, no obvious correlations are observed between the clay contents and the NMR PSD amplitudes in Monterey shales.