Unraveling Reservoir Quality: How Mineralogy Shapes Pore Attributes in Sandstone Lithofacies

Abstract

The Cenomanian Bahariya Formation exposed at Gebel El Dist in the Western Desert of Egypt provides valuable surface analogues for evaluating the reservoir quality of subsurface Bahariya sandstones. The formation was analyzed using 27 oriented samples and 91 core plugs from quartz arenite (QA) and quartz wacke (QW) facies. Analyses included XRD, petrography, SEM, helium porosity–permeability, and capillary tests, as well as measurements of pore-throat radii (R) at 35% and 36% mercury saturation. X-ray diffraction analyses reveal a heterogeneous mineral composition dominated by quartz, feldspars, dolomite, pyrite, siderite, goethite, hematite, clay minerals, glauconite, and gypsum. QA displays higher porosity and permeability than QW, along with larger pore radii, and lower specific surface area per unit pore volume (Spv) and per unit grain volume (Sgv). Multivariate regression equations, specific to each facies, were developed to convert standardized XRD mineral percentages directly into pore-system and flow attributes (ϕ, k, r, Spv, Sgv, R35, R36), quantifying capillary-based recovery contrasts between facies. Across both facies, regressions linking mineralogy to ϕ, k, r, Spv, Sgv, R35, and R36 are strong (R² = 0.78–1.00). The established predictive equations provide a low-cost method to estimate reservoir quality from mineralogy alone, enabling rapid screening of Cenomanian Bahariya analogues and similar clastic reservoirs where core data are limited.

1. Introduction

A considerable portion of the world’s hydrocarbon resources is contained within sandstone reservoirs. Their economic value is strongly influenced by reservoir quality, particularly porosity and permeability, which reflects depositional texture (grain size, shape, sorting, and packing) and diagenetic alteration [1,2,3,4]. Reservoir quality is also affected by the pressure–temperature regime and fluid–rock interactions. For example, fluid overpressure can retard mechanical compaction and preserve porosity, whereas excessive overpressure may promote fracturing, leakage, and fluid redistribution [5,6]. Similarly, magmatic–hydrothermal heating adjacent to igneous intrusions can alter diagenetic pathways and pore systems, variably enhancing or degrading reservoir quality [7,8,9]. Accurate reservoir quality prediction is a fundamental objective for characterizing hydrocarbon and geothermal reservoirs, as well as for CO₂ sequestration research. Sandstone reservoir quality results from complex evolutionary processes that depend on both the depositional environment and diagenetic history. Mineralogy is a vital yet insufficiently quantified factor in this evolution. Different mineral phases respond differently to geochemical conditions, leading to diverse pore geometries and connectivity patterns [4,10,11].

Mineralogy is a critical factor controlling reservoir quality. The mineralogical composition of sandstone determines its porosity, permeability, wettability, and diagenetic pathways [12,13,14]. Quartz and feldspars provide strong support for intergranular porosity through their resistance to compaction and chemical alteration, while clays, iron oxyhydroxides, and carbonates weaken reservoir quality by blocking pore spaces and modifying fluid interactions [15,16,17,18]. The types of minerals present control how a sandstone reservoir responds mechanically to hydraulic fracturing and reacts chemically during enhanced oil recovery (EOR) processes. For example, carbonate minerals can improve permeability through acid treatment, but clay minerals may swell and block pore throats under the same conditions [19,20,21]. The same mineralogical factors that affect oil and gas recovery also influence fluid flow, chemical reactivity, and mechanical stability in groundwater aquifers, CO₂ storage sites, and geothermal systems [22,23]. Recent basin-scale studies further document how detrital frameworks and diagenetic cements govern porosity–permeability relationships and pore-throat distributions in sandstone reservoirs [24,25].

However, a clear practical need exists for field-ready equations that convert readily available mineralogical information (e.g., XRD percentages) into pore geometry and flow attributes such as porosity (ϕ), permeability (k), pore and throat radii (r, R35/R36), and specific surface metrics (Spv, Sgv), which control recovery and injectivity. These equations would minimize the need for expensive core plug programs and allow for early screening and facies mapping even in data-poor environments.

Quantitative measures of the impact of mineralogy on pore properties remain insufficiently defined. Evidence points to several quantitative relationships; for instance, a strong positive correlation exists between quartz content and both pore volume and average pore diameter, while a weak negative correlation is observed between carbonate content and total organic carbon (TOC). This suggests that a higher quartz content leads to larger pore volumes and diameters, whereas carbonates have the opposite effect. However, these relationships are currently expressed as correlations rather than explicit predictive equations or models, indicating a gap in fully developed quantitative measures [26]. Furthermore, few existing studies attempted to correlate the reservoir’s mineralogical makeup with its pore properties [27]. Plots of porosity versus permeability are used to explore these relationships, but significant data variability implies that mineralogy is only one of several controlling factors and that its quantitative impact is not fully quantified [28]. Some studies calculate derivative parameters, such as the Normalized Porosity Index (NPI), Reservoir Quality Index (RQI), Flow Zone Indicator (FZI), and pore throat radius (R35), from porosity and permeability data. While these parameters assess reservoir quality, a direct quantitative link to specific mineralogical compositions has not been clearly established [29]. The lack of quantitative constraints on mineralogy limits our ability to predict reservoir performance in data-poor conditions and complicates the application of mineralogical parameters in reservoir modeling.

The complexity, limited data, and uncertainty associated with characterizing a poor-quality subsurface reservoir can be investigated by studying surface analogs. Such studies help establish a geological context and provide a basis for comparison, especially in areas where subsurface data is sparse [30,31]. Furthermore, the high costs of retrieving core plugs from deep reservoirs can be mitigated by using surface analogs to characterize reservoir heterogeneity [32]. Therefore, surface analog studies and subsurface reservoir characterization methods play complementary roles.

In the oil and gas industry, the specific surface area, which can be expressed either per unit pore volume (Spv) or per unit grain volume (Sgv), is a critical parameter. It plays a major role in reservoir behavior, especially in EOR, as it quantifies the area available for interaction between pore walls and fluids (oil, gas, water, and injected CO₂). Consequently, it governs adsorption/desorption behavior, particularly for gas storage in nanopores, and directly influences capillary pressure and relative permeability, which are key to understanding fluid movement through reservoir rocks [33,34].

Consequently, two key gaps remain. Previous studies have commonly focused on qualitative analysis or simple bivariate correlations rather than developing multivariate predictive equations that connect the complete mineral set with ϕ, k, r, Spv, Sgv, and effective pore-throat radii (R35/R36) metrics. Second, for the Bahariya Formation, outcrop-based equations that convert mineralogy to flow properties have not yet been published, limiting the transferability of insights from analog to subsurface settings.

To address these gaps, the objectives of this study are to (1) quantify how mineral composition governs porosity and permeability across QA and QW facies; (2) derives facies-specific multivariate regression equations that relate XRD-measured mineral percentages to key pore-system attributes (ϕ, k, r, Spv, Sgv, R35, and R36); and (3) assess the predictive capability of these equations for analogous reservoirs. This study provides a quantitative, transferable workflow that converts mineralogy directly into petrophysical and capillary parameters at the facies scale. The derived equations enable the estimation of reservoir properties from XRD data alone and support the prediction of heterogeneous sandstone reservoir quality in Bahariya analogs and similar clastic systems where core data are limited.

2. Materials and Methods

2.1. The Case Study

Egypt is one of the top five oil-producing countries in Africa, producing approximately 550,000–600,000 barrels per day of crude oil and condensates. It has over 65 Tcf of proven gas reserves. Egypt has three potential basins: the Nile Delta, the Gulf of Suez, and the Western Desert. The Western Desert basin is one of the most prolific oil provinces, hosting more than 2 billion barrels of recoverable reserves and contributing substantially to Egypt’s total crude oil resources [35]. Within this basin, the Cenomanian Bahariya Formation is a significant reservoir, with recoverable reserves estimated to be between 300 and 500 million barrels [36,37]. The formation is widely distributed across the northern Western Desert, encountered in productive fields such as Razzak, Abu Gharadig, Qarun, Khalda, and the Shushan Basin [37,38,39]. Stratigraphically, it reaches a thickness of up to 170 m, consisting of a lower sandstone-rich member and an upper member composed of sandstone and variegated shale [40,41,42]. Surface exposures of the upper member at Gebel El Dist provide excellent analogues for subsurface reservoir characterization.

Our investigation is grounded in a detailed case study of the Bahariya Formation from Gebel El Dist in the Western Desert, Egypt. The studied section (Figure 1a) is located on the northern escarpment of the Bahariya plateau at this location, between 28°25′42.8″ N and 25°55′41.11″ E.

Figure 1. (a) A location map of the study area, including a geological map that shows the lithologic distribution and the precise location of Gebel El Dist. (b) The lithostratigraphic sequence includes an elevation log that displays grain density, porosity, and permeability measurements, along with the sampling locations. The figure also shows the following: Sample Number (S), Facies Association (FA), Lowstand Systems Tract (LST), Highstand Systems Tract (HST), Transgressive Systems Tract (TST), Surface of Unconformity (SU), Wave Ravinement Surface (WRS), Maximum Regression Surface (MRS), and Maximum Flooding Surface (MFS).

Based on the systems tracts defined by Catuneanu et al. [40], the lower Cenomanian Bahariya Formation corresponds to a second-order depositional sequence. This unit was deposited on a continental shelf with a relatively low rate of positive accommodation (less than 200 m) over a period of three to six million years. The rise in base-level was interrupted by three episodes of base-level fall, resulting in third-order sequence boundaries. These boundaries are represented by subaerial unconformities, which may or may not be replaced by younger transgressive wave ravinement surfaces. The Bahariya Formation is divided into four third-order depositional sequences based on lateral and vertical changes within a transitional environment that ranges from shelf-proximal facies to fluvial facies landward.

The upper member of the Bahariya Formation, outcropping in the Gebel El Dist area, is dominated by three main facies: QA, QW, and mudstone [35]. The QA facies represents the Lowstand Systems Tract (LST) and is repeated four times throughout the studied section (Figure 1b). This facies association is interpreted as having been deposited in an exposed continental shelf environment influenced by sediment influx from wind, tidal, fluvial, and/or meteoric water sources. The second facies association (quartz wacke) is interpreted as either a Transgressive Systems Tract (TST) or a Highstand Systems Tract (HST). When this facies is intercalated within a quartz arenite cyclothem, it represents a TST overlying an LST (Figure 1b). In contrast, when the quartz wacke forms a cyclothem with the mudstone facies, it is interpreted as an HST; this systems tract association is repeated twice (Figure 1b). The third facies association is mudstone. Its presence consistently indicates an HST and it is also repeated twice (Figure 1b) [40].

2.2. Sample Collection and Analyses

A total of twenty-seven representative samples of QA and QW facies were collected from the studied section in Gebel El Dist (Figure 1). These samples were oriented with respect to the bedding plane. From them, 91 one-inch core plugs were produced for petrophysical analysis.

Grain size analysis (GSA) was conducted by crushing and sieving portions (approximately 210 g each) from the 27 samples. A set of sieves covering sizes from 4 mm (pebbles) down to 0.0625 mm (very fine sand and smaller fractions) was used. The mean grain size (Mz) and sorting coefficient (σI) were determined following the method of Folk [43], which uses the Ø scale to describe grain-size distribution and sorting. For petrophysical modeling, both parameters were later converted into millimeter units. The calculation of these parameters on the Ø scale utilizes the grain diameters corresponding to the 5th, 16th, 50th, 84th, and 95th percentiles of the cumulative weight distribution.

Mineralogical and textural features of the samples were examined through thin-section petrography (using end-trim plugs from cylindrical samples of 2.5 cm³), X-ray diffraction (XRD) of powdered samples (D8 Advance, Bruker, Billerica, MA, USA), and scanning electron microscopy (SEM) (Benchtop Neoscope^TM SEM, JOEL Solution for Innovation, Peabody, MA, USA). The classification and naming followed the sandstone scheme proposed by Pettijohn et al. [44].

2.3. Petrophysical Measurements

The Porosity ɸ of the core plugs was measured using a Helium gas expansion porosimeter (HEP-P Helium Porosimeter, Vinci Technologies, France). The instrument applies Boyle’s law through isothermal helium expansion to determine the porosity and grain density. Permeability (k) was evaluated under steady-state conditions with a Gas Permeameter (KA-210, Coretest Systems, Knoxville, TN, USA), using the method described by El Sayed [45] and Tiab and Donaldson [46].

$$k={\displaystyle \frac{2\mu qL}{A\left(P_1^2-P_2^2\right)}}\times 1000$$

(1)

where μ represents the nitrogen viscosity in centipoise (0.018 at surface conditions), q is the flow rate measured in cm³/s, A is the cross-sectional area of the sample in cm², and P is the pressure difference across the core plug between the upstream pressure (P₁) and the downstream (P₂).

The cementation exponent (m) was determined by measuring the resistivity (R₀, in Ω·m) of fully saturated core plugs taken from representative samples of the studied section. Due to their clay content, ten of these samples required stabilization in a halite (NaCl) brine with a salinity of 300,000 ppm to prevent clay swelling. The brine used for core plug saturation had a resistivity (Rw) of 0.045 Ω·m. All measurements were conducted with the Electrical Properties System Atmospheric (EPSA, Vinci Technologies, Nanterre, France). The cementation exponent was then derived from the measured formation resistivity data using Equation (2) [47,48,49].

$$m={\displaystyle \frac{Log{R}_w-Log{R}_0}{Log\mathrm{\varnothing }}}$$

(2)

The cementation exponent (m) is an electro-geometric parameter that reflects how a water-saturated pore network conducts current; it is not a direct measure of the amount of cement. It describes how structure of the rock’s pore space, grain shape, and matrix consolidation affect electrical conductivity. A higher m value indicates greater tortuosity or more complex pore connections, which results in a higher calculated water saturation for a given resistivity. In mixed-mineral sandstones, conductive mineral surfaces (notably clays, iron oxyhydroxides, and iron-rich dolomite) provide parallel conduction pathways along their surfaces and within the rock structure, which lowers R₀ at a given porosity.

The porosity (ɸ), gas permeability (k), and cementation index results are used for the calculation of the pore-throat radius at 35% and 36% mercury injection saturation (denoted as W_R35, A_R35, and K_R36, respectively) and pore radius (r) [46,50,51,52]. The permeability values were converted in μm² for these calculations.

The Kozeny correlation, often referred to as the Carman-Kozeny equation, is a widely used empirical relationship in porous media studies, that relates measured porosity (ɸ) and permeability (k) to other important petrophysical attributes.

The Kozeny correlation [46] is used to calculate many other attributes based on the measured porosities (ɸ) and permeabilities (k). These calculations include the pore radius (r) in μm (Equation (3)), the specific surface area per unit pore (Svp) in μm⁻¹ (Equation (4)), and the specific surface area per unit grain (Svg) in μm⁻¹ (Equation (5)) [46].

$$r=\sqrt{{\displaystyle \frac{8k}{\mathrm{\varnothing }}}}$$

(3)

$$Spv={\displaystyle \frac{2}{r}}$$

(4)

$$Sgv=Spv \left({\displaystyle \frac{\mathrm{\varnothing }}{1-\mathrm{\varnothing }}}\right)$$

(5)

The pore throat radius at 35% mercury saturation (R35) was calculated using both the Winland (W_R35) and Aguilera (A_R35) methods (Equations (6) and (7)). These R35 values were then correlated with permeability measurements [45,52]. A second parameter, the pore throat radius at 36% mercury saturation (R36), was also calculated. This was done using either permeability (k_R36, Equation (8)) or porosity (ɸ_R36, Equation (9)). The resulting k_R36 and ɸ_R36 values were then compared with the R35 values [52]. This correlation helps determine which method is more effective for the rock types of interest, as R36 can be derived from either porosity or permeability [52,53].

$$W_{R35}=5.395 {\left({\displaystyle \frac{k^{0.588}}{\mathrm{\varnothing }}}\right)}^{0.864}$$

(6)

$$A_{R35}=2.665 {\left({\displaystyle \frac{k}{\mathrm{\varnothing }}}\right)}^{0.45}$$

(7)

$$k_{R36}=2277.5 k^{0.542}$$

(8)

$${\mathrm{\varnothing }}_{R36}=466.82 e^{0.193\mathrm{\varnothing }}$$

(9)

A porous plate (Vinci Technologies, Nanterre, France) was used to desaturate eight core plugs: six from the QA and two from the QW samples. The desaturation was achieved by applying an air-brine capillary pressure. Six pressure levels were used to desaturate the brine-saturated cores (with a salinity of 300,000 ppm), starting with 1 psi, followed by 4 psi, 10 psi, 30 psi, 60 psi, and finally 100 psi. At each of these six pressure steps, water saturation values (Sw1–Sw6) were calculated using Equation (10). The final saturation (Sw6) corresponds to the irreducible water saturation (Swi), which was achieved at the maximum pressure of 100 psi.

$$S_{W1-6}=S_{W100\mathrm{\%}}-{\displaystyle \frac{M_{Sat}-M_{Desat1-6}}{1.227 \left(TPV\right)}}$$

(10)

where S represents the saturation values in decimal; M is the mass of the sample, either saturated (sat) or desaturated (Desat), in grams; and TPV is the total pore volume in cubic centimeters. The constant value, 1.227 gm/cm³, refers to the brine density used to saturate the samples.

The air-brine capillary pressure method is used to estimate many petrophysical attributes that aid in the evaluation of both reservoir types in the selected section. Recovery efficiency (Re) is calculated using Equation (11) [54,55], and the mean irreducible water saturation (Swi) is determined for both rock types.

$$Re={\displaystyle \frac{(Sbmax-Sbr)}{Sbmax}}$$

(11)

where Sbmax and Sbr are the maximum and residual brine saturation in the samples, as a decimal, respectively.

2.4. Regression Analysis

The interrelationships between the mineralogical composition and measured petrophysical properties of sandstone lithofacies were analyzed using a multiple linear regression model in IBM SPSS Statistics (IBM SPSS Version 21, 2012). The objective was to determine the influence of individual mineral components on porosity (ɸ), permeability (k), pore radius (r), specific surface per unit pore (Spv), specific surface per unit grain (Sgv), and pore throat radii (R₃₅ and R₃₆). Mineralogical percentages from XRD and petrophysical properties from core plugs were standardized prior to the analysis to remove unit dependency and ensure comparability. The predictor variables included quartz (Q), feldspar (F), hematite (H), pyrite (P), siderite (S), clay (C), dolomite (D), goethite (G), gypsum (Gyp), and glauconite (Glau). The accuracy of the resulting regression equations was assessed using the coefficient of determination (R²) and the standard error of the estimate (SEE).

Because XRD mineral percentages are compositional data and therefore not fully independent, we screened for multicollinearity before fitting the models. Predictors were z-standardized, and the Pearson correlation matrix and SPSS Collinearity Diagnostics (Variance Inflation Factor, tolerance, and condition index) were examined to identify problematic collinearity. Regressions were run separately for QA and QW to further limit cross-facies dependence. For the final equations reported, all retained predictors satisfied these criteria; when a term exceeded a threshold in preliminary runs, the model was refit after removing the most collinear variable and diagnostics were rechecked.

Figure 2 presents a flowchart illustrating the data flow through the main steps of the research.

Figure 2. A flowchart representing the research’s main steps.

3. Results

3.1. Samples Characterization and Petrophysical Investigation

The grain size analysis (GSA) yielded the following results: For QA, the sorting coefficient (σI) ranges from 0.8 to 1.4 Ø, with a mean of 1.14 Ø, while the mean grain size (Mz) ranges from 0.06 to 2.7 Ø, with an average of 1.69 Ø. For QW, the sorting coefficient (σI) varies from 0.9 to 1.8 Ø, with a mean of 1.26 Ø, and the mean grain size (Mz) ranges from 0.6 to 1.9 Ø, averaging 1.2 Ø.

The X-ray diffraction (XRD) analysis revealed a diverse mineralogical composition dominated by quartz, feldspars, dolomite, pyrite, siderite, goethite, hematite, clay minerals, glauconite, and gypsum (Table 1). Quartz was identified as the principal framework mineral, occurring as grains, overgrowths, and matrix material. Feldspars, primarily plagioclase and microcline, were present in minor amounts. Dolomite appeared mainly as a cementing phase and occasionally formed iron-rich varieties due to interaction with iron oxyhydroxides [35]. Hematite and goethite were confirmed by XRD patterns (Figure S1) and were observed to occur as patches or cement, indicating diagenetic and weathering influences. Pyrite was detected with characteristic framboidal textures, while siderite occurred as an accessory cementing mineral formed under telogenetic conditions. Glauconite appeared as coatings or pellets, reflecting early diagenetic replacement of calcite, and gypsum was observed locally as vein fillings [35].

Table 1. XRD mineral percentages (%) for different lithofacies of Bahariya Formation.

Rock Unit	Lithofacies	Goethite	Quartz	Feldspars	Hematite	Pyrite	Siderite	Clays	Dolomite	Gypsum	Glauconite
Unit VI	GFQW	22.47	39.00	4.00	0.00	4.00	5.80	10.00	10.00	0.00	5.20
	QA	16.00	59.90	3.00	0.00	1.80	2.60	7.50	7.50	0.00	1.70
	FQA	16.00	50.00	0.00	0.00	0.00	0.00	9.00	8.00	0.00	17.00
	QA	12.80	64.00	1.00	1.00	2.50	2.50	7.60	7.60	0.00	1.00
	FQW	22.00	47.90	1.00	1.00	2.00	4.00	11.00	11.00	0.00	0.00
	FQA	15.50	52.95	4.10	0.00	3.20	4.25	10.00	10.00	0.00	0.00
	QA	18.00	57.90	4.00	8.60	1.50	2.00	4.00	4.00	0.00	0.00
Unit V	FGQW	25.00	32.00	5.00	3.50	3.50	7.50	9.00	9.30	1.70	3.50
	SFGQW	18.00	42.60	4.10	2.30	4.60	6.50	10.00	11.00	0.00	1.00
	GFQW	40.00	14.00	20.00	10.00	1.00	1.50	8.00	4.00	0.50	1.00
	FGQW	28.00	39.00	0.00	10.00	2.00	3.00	7.00	10.00	0.00	1.00
	LGFQW	27.00	10.10	6.00	6.00	3.40	9.00	13.00	13.70	5.00	6.80
	FGQW	18.00	54.00	0.00	4.00	0.00	4.00	4.00	12.00	2.00	2.00
	FQW	37.00	15.00	25.00	10.00	2.00	4.00	0.00	0.00	6.00	1.00
Unit IV	FQW	25.00	7.50	7.50	15.00	0.00	10.00	0.00	0.00	35.00	0.00
	LFQA	15.00	51.70	0.00	1.00	2.00	3.90	15.00	7.60	0.00	3.80
Unit III	FQA	38.00	19.00	2.00	6.00	6.00	6.00	12.00	11.00	0.00	0.00
	FQA	19.00	38.00	5.00	0.00	6.00	6.00	13.00	13.00	0.00	0.00
	GSQA	16.00	48.90	4.80	1.00	4.80	6.50	9.00	9.00	0.00	0.00
	FGQW	22.50	37.40	1.00	1.50	7.60	5.00	10.00	15.00	0.00	0.00
Unit II	FQW	16.00	51.00	10.00	0.00	0.00	0.00	1.00	22.00	0.00	0.00
	GFQW	24.00	21.20	1.80	0.90	3.00	7.50	15.00	10.60	11.00	5.00
	GQW	16.00	51.00	10.00	0.00	0.00	0.00	1.00	22.00	0.00	0.00
	FQA	22.00	35.00	3.00	2.00	4.00	6.00	9.00	9.00	9.00	1.00
Unit I	LFQA	22.00	32.00	4.00	2.00	4.00	7.00	9.00	10.00	7.00	3.00
	GFQA	23.10	49.56	0.84	1.50	4.00	7.00	7.00	7.00	0.00	0.00
	LFGQA	22.00	40.50	4.50	2.00	5.00	8.00	9.00	9.00	0.00	0.00

A = Arenite, Q = Quartz, L = Laminated, G = Glauconitic, F = Ferruginous, S = Fossiliferous.

Helium porosity measurements revealed that the QA facies exhibits the highest porosity (0.18–0.43%, mean = 0.33%) and moderate grain density (2.48–2.97 g/cm³, mean = 2.71 g/cm³) (Table 2). In contrast, QW shows lower porosity (0.06–0.41%, mean = 0.19%) with a slightly lower mean density (2.61 g/cm³). The average permeability is 21.46 mD for QA and 5.65 mD for QW, reflecting the better sorting and superior pore connectivity of the arenitic facies [24,25]. The lower permeability of QW is attributed to its finer grains, poorer sorting, and increased iron oxide and clay cementation [28]. Finally, the cementation exponent (m) was higher for QA (2.96–6.52) than for QW (2.57–3.46).

Table 2. Petrophysical investigation of QA and QW facies of Bahariya Formation.

Parameter	QA				QW
	Min	Max	Mean	St. Dev	Min	Max	Mean	St. Dev
Porosity (ɸ)	0.18	0.43	0.33	0.08	0.06	0.41	0.19	0.10
Grain Density (ρg) in gm/cc	2.48	2.97	2.71	0.14	2.38	3.07	2.61	0.14
Permeability (k) in mD	0.27	165.73	21.46	28.11	0.11	14.64	5.65	5.57
Permeability (k) in μm2	3 × 10−4	0.16	0.02	0.03	1 × 10−4	0.014	0.01	0.01
Cementation (m)	2.96	6.52	3.84	1.33	2.57	3.46	3.09	0.42

Although the QA samples show higher measured cementation exponent than the QW samples (Table 2), QA also exhibits higher porosity and permeability. This is not contradictory: in the clay- and Fe-oxide/dolomite-rich QW, surface/structural conduction lowers the resistivity (R₀) and thus depresses the apparent cementation exponent derived from Equation (2). This occurs even as pore throats are narrowed and permeability is reduced. In the quartz-supported QA, current is confined largely to the bulk electrolyte within larger, better-sorted pores. This configuration yields higher formation factors and a higher cementation exponent from electrical measurements while hydraulically, the coarser network gives higher permeability.

3.2. Specific Surface Areas of Grains (Sgv) and Pores (Spv)

The shape and size of grains are essential factors for controlling fluid flow within a reservoir [56,57]. These geological attributes are particularly important in the oil industry. The specific surface area of both pores and grains is a key petrophysical attribute that mathematically reflects the influence of grain shape and size on reservoir characteristics [33,34,46]. The QA facies has mean values for (r), (Spv), and (Sgv) of 0.59, 6.17, and 1.38, respectively (Figure 3). In contrast, the QW facies has corresponding values of 0.4, 7.98, and 1.55 (Figure 2). The larger pore radius (r) in the QA facies is attributed to lower amounts of formation fines, iron oxyhydroxides, and iron-rich dolomite cement (Figure 4a,d,f). Furthermore, the QA facies generally exhibits fewer formation fines and lower cementation (m) values than the QW facies.

Figure 3. Statistical distributions of the pore radius (r) and the specific surface area per unit grain (Sgv), and per unit pore (Spv) are shown for (a) the QA facies; (b) the QW facies.

Figure 4. Photomicrographs and SEM images of QA and QW facies. (a) Laminated QA, showing sub-rounded quartz grains (ɸ = 39%, k = 37.93 mD). (b) Glauconitic QW, featuring sub-angular quartz grains (ɸ = 28%, k = 11.24 mD). (c) Ferruginous QW, showing sub-angular grains with extensive cementation by iron oxyhydroxides (ɸ = 24%, k = 8.92 mD). (d) SEM image of QA, highlighting well-connected pore spaces (ɸ = 39%, k = 37.93 mD). (e) SEM image of QW, demonstrating lower porosity compared to the QA in (d) (ɸ = 28%, k = 11.24 mD). (f) QW showing a sub-angular quartz grain embedded in a matrix of rusty brown, iron-rich dolomite cement (ɸ = 28%, k = 11.24 mD). Images a, b, c, d, and f were taken under cross-polarized light (crossed Nicols).

The correlation between Sgv and pore radius (r) yields an R² value of 0.91 for QA and 0.72 for QW (Figure 5a). The higher correlation coefficient for QA is attributed to its low concentration of formation fines (Figure 4a,d), as well as the effects of better sorting (σI) and mean grain size (Mz). QA grains are relatively better sorted and more subrounded, whereas QW grains are poorly sorted with a higher proportion of subangular grains (Figure 4). A second correlation between Sgv and Spv shows R² values of 0.91 and 0.65 for QA and QW, respectively (Figure 5b). The higher R² value for QA can be explained by the same factors noted for the previous correlation. Furthermore, the depositional environment plays a role; QA is characterized by a Lowstand Systems Tract (LST), while QW is found in either LST or Highstand Systems Tract (HST) settings.

Figure 5. (a) A correlation between the Sgv and pore radius (r) for the QA and QW facies; (b) a correlation between both Sgv and Spv for the QA, and QW facies. The QA facies exhibits systematically lower Sgv at a given r, as well as a tighter Sgv-r and Sgv-Spv coupling.

3.3. Effective Pore Throat Radii (Apex)

The Apex refers to the highest point on the hyperbolic curve in a log-log plot of capillary pressure versus mercury saturation. This point is crucial because it represents the pore apertures that interconnect to form an effective pore system, which dominates fluid flow in the reservoir. Graphically, the Apex is determined by plotting the ratio of mercury saturation to capillary pressure against mercury saturation and identifying the peak of this curve [28,36,58]. R35 represents the pore throat radius at 35% mercury saturation. Calculating R35 requires both porosity and permeability data, whereas R36 can be determined using only a single parameter. This work involves calculating both R35 and R36 for the two reservoir rock types encountered in the Bahariya Formation to determine which parameter is most suitable for each facies. For R35, the QA facies has mean values of 1.65 µm and 0.64 µm using the Winland and Aguilera equations, respectively (Figure 6). The QW facies shows corresponding mean values of 1.32 µm and 0.45 µm. For the R36 apex, the QA facies shows mean values of 1.65 µm and 0.64 µm using the porosity-dependent (ϕ_R36) and permeability-dependent (k_R36) methods, respectively. Similarly, the QW facies has mean values of 1.32 µm and 0.45 µm for the two methods.

Figure 6. The statistical distribution of the pore throat radii (W_R35, A_R35, ɸ_R36, and k_R36) for (a) QA facies; (b) QW facies.

The study demonstrates that for both QA and QW, the reservoir apex can be calculated using either the Aguilera equation (A_R35) or the permeability-dependent equation (k_R36), as evidenced by their strong correlation with W_R35 and A_R35 (Figure 7a,b). This correlation yielded high R² values of 0.97 and 0.82 for QA and QW, respectively. In contrast, the correlation involving the porosity-derived R₃₆ apex (ɸ_R36) produced a significantly lower R², indicating it is not a suitable substitute for the Winland or Aguilera method in this context (Figure 7c,d). Furthermore, a comparison of the Winland and Aguilera apex calculations against permeability data showed that the Aguilera equation performed better, with an R² of 0.98 for QA compared to 0.94 for QW (Figure 8b). Finally, the pore-throat radii (R₃₅) from both methods showed a strong and applicable relationship with permeability for the Gebel El Dist reservoir rocks (Figure 8a,b).

Figure 7. Comparison between apexes for the QA and the QW facies: (a) K_R36 vs. W_R35; (b) K_R36 vs. A_R35; (c) W_R35 vs. ɸ_R36; (d) A_R35 vs. ɸ_R36.

Figure 8. R₃₅ Apex-permeability correlation using (a) Winland and (b) Aguilera for QA and QW.

3.4. Capillary Pressure (Pc) and Microscopic Recovery Efficiency (Re)

The pressure generated when two immiscible fluids (wetting and non-wetting) flow simultaneously in a narrow, tortuous channel or tube is known as capillary pressure (Pc) [46,59]. This property is important for understanding multiphase flow regimes in a reservoir or an aquifer. Capillary forces in reservoirs arise from the combined effects of surface and interfacial tensions, pore size, and the wetting affinity of each fluid for the rock. When measured, these properties help determine important factors such as the irreducible wetting phase saturation (Swi), the residual non-wetting phase saturation (Sr), and the microscopic recovery efficiency (Re) for each reservoir rock type. To investigate the Pc for both facies, six QA and two QW core plugs were saturated. The maximum saturation of the QA facies reached 80 to 90% maximum saturation (Sbmax). The QW facies reaches 100% saturation as it has lower cementation than the QA facies. Figure 9 shows that at the maximum capillary pressure (Pc) applied, no more brine could be expelled from the core samples; the remaining water inside the sample is the irreducible water (Swi or Sbr), which cannot be expelled even after increasing the capillary pressure to more than 100 psi. The figure also shows that the QA facies have a mean irreducible water saturation (Swi) of 12.5% (Figure 9a), while the QW due to its clay content retains more brine which increases its mean Swi to 40.5% at the same (Pc) (Figure 9b). Consequently, due to the difference in the Swi between both facies, the mean microscopic recovery efficiency (Re) is higher in the QA facies than in the QW facies. The mean Re was calculated to be equal to 83.1% for QA and 59.5% for QW (Table S1), reflecting a better-quality reservoir for the QA facies.

Figure 9. Capillary pressures curve for (a) QA and (b) QW samples.

3.5. Petrophysical Modeling

Diagenesis significantly alters the mineralogy of lithofacies, affecting both the types of minerals present and the pore structure of the rock. Some minerals within lithofacies resist geological processes after deposition, while others are newly formed during diagenesis. These newly formed minerals, such as iron-rich dolomite, iron oxyhydroxides, and clay minerals, often act as cements that modify the pore structure and reduce porosity and permeability [60,61,62]. The present work uses regression analysis to relate the proportions of these minerals to all measured attributes in order to identify the extent of any meaningful and/or applicable relationships.

The Bahariya Formation in the Gebel El Dist section is primarily composed of two reservoir rock types: QA and QW. For the studied QA facies, several regression equations, with correlation coefficients (R²) ranging from 0.85 to 1 (Table 3), demonstrate the impact of mineralogical constituents on its pore network attributes. Similarly, for the QW facies, the correlation coefficients (R²) range from 0.78 to 1 (Table 4).

Table 3. Compilation of the regression analysis equations for the QA facies.

No.	Property	Regression Equation	R2	SEE
1	Porosity	ɸ = r (0.61) − Spv (0.03) + Sgv (0.17) − 0.007	0.85	0.47
2	Pore radius	r = − Q (4.44) − F (0.27) − H (0.71) − (P + S) (0.13) + C (1.02) + D (0.64) − Glau (2.80) + 0.603	1	0
3	Spv	Spv = − Q (2.61) − F (10) + H (12.36) − P (22.408) − S (12.41) − C (56.67) − D (32.41) − Glau (12.11) + 13.72	1	0
4	Sgv	Sgv = − Q (0.83) − F (6.47) + H (10.35) + P (19.22) + S (5.64) − C (18.44) − D (0.83) − Glau (17.23) + 2.21	1	0
5	WR35	WR35 = Q (0.82) + F (6.02) + H (2.67) − P (15.15) − S (4.67) + C (15) + D (0.76) + Glau (5) + 0.28	1	0
6	AR35	AR35 = − Q (0.12) + F (0.5) − H (0.80) + P (0.5) − S (0.6) + C (1.43) + D (0.84) − Glau (1.5) + 0.44	1	0
7	ɸR36	ɸR36 = − Q (61) − F (168.2) + H (41.13) + P 450.47) + S (99.32) − C (204.37) + D (80) − Glau (628.227) + 525.6	1	0
8	kR36	kR36 = − Q (261.4) − F (538.57) − H (265.5) + P (1892.73) + S (193.73) − C (573) + D (460.26) − Glau (2651) + 288.46	1	0
9	Sorting coefficient	σI = F (1.73) − H (0.55) − P (2.52) − (S + C − Q) − D (0.75) + Glau (8.52) − 0.8	1	0
10	Sorting coefficient	σI = − r (1.43) − Spv (0.006) + Sgv (0.17) + 1.2	0.93	0.25

Table 4. Compilation of the regression analysis equations for the QW facies.

No.	Property	Regression Equation	R2	SEE
1	Porosity	ɸ = − r (0.06) − Spv (103.25) + Sgv (0.121) + 0.451	0.78	0.32
2	Pore Radius	r = G (83.56) − Q (103.25) − F (51.15) − H (140.36) − P (11) − S (161.4) − C (272.8) +D (0.87) − Gyp (100.13) − Glau (103.14) + 104	1	0
3	Spv	Spv = G (625.4) + Q (613.5) + F (506.73) +H (541.46) + P (71.8) +S (834.73) + C (1584.53) − D (26.58) + Gyp (619.38) + Glau (459.16) − 627.97	1	0
4	Sgv	Sgv = G (162.84) + Q (190.6) + H (239.24) + P (21.84) + S (283.75) + C (513) − D (3) + Gyp (186.62) + Glau (176.45) − 192.22	1	0
5	WR35	WR35 = Q (3.25) − F (0.66) + H (7.6) − P (1.64) + S (57) + D (31.23) − Gyp (9) + Glau (146) − 7.16	1	0
6	AR35	AR35 = Q (3.25) − F (0.661) + H (7.6) − P (0.5) + S (24.1) + D (12.3) − Gyp (1.1) + Glau (53.4) − 3.54	1	0
7	ɸR36	ɸR36 = − G (959.05) − Q (337.62) − F (1053.51) + H (1259.4) − P (101.84) + S (62.45) − C (357) + D (86.23) − Gyp (551.9) + Glau (448) + 900.05	1	0
8	kR36	kR36 = Q (768.85) − F (1993.7) − H (506.2) − P (220.33) + S (7899.22) + D (4022.34) − Gyp (1101.06) + Glau (19163.9) − 888.35	1	0
9	Sorting coefficient	σI = − G (10.07) − Q (3.44) − F (9.30) +H (14.83) − P (0.44) + S 1.77) − C (4.35) + D (1.14) − Gyp (5.85) + Glau (3.35) + 4.48	1	0
10	Sorting coefficient	σI = − r (0.06) − Spv (103.25) + Sgv (0.121) + 0.451	0.78	0.32

4. Discussion

The results of this study demonstrate that mineralogy is a key controlling factor in reservoir quality in the Bahariya Formation sandstones. High correlation coefficients between the mineralogical and petrophysical data indicate that variations in content of quartz, dolomite, clay minerals, and iron oxide govern the sandstones’ textural and diagenetic patterns.

4.1. Influence of Mineralogy on Porosity and Permeability

The QA facies exhibits better porosity and permeability than the QW facies. This pattern is commonly attributed to QA’s quartz-rich composition and lower degree of cementation by clay and iron oxyhydroxide minerals. Quartz is a hard, stable mineral that tends to form relatively larger grains in sedimentary rocks. This granular arrangement creates larger, well-connected pore spaces, which help maintain both porosity and permeability even under compaction. Furthermore, quartz grains are less likely to deform under pressure than more ductile minerals like clay and mica. This preserves intergranular pores at higher compaction levels, whereas ductile minerals can deform and close pore spaces under load [63,64]. Consequently, quartz-rich rocks are characterized by high porosity and permeability, as quartz is resistant to alteration and does not readily form or contribute to cements that fill pore spaces. This property makes quartz a key factor in maintaining pore connectivity and fluid flow within the rock mass [63,64,65].

Finer-grained facies often contain abundant clay minerals, dolomite, and iron oxides. These minerals tend to fill pore spaces and throats, reducing both porosity and pore connectivity. Clay minerals have a fine-grained, flaky nature that allows them to pack tightly into pore spaces, leading to decreased porosity and more tortuous, less connected pore networks [63,64,66]. Clay distributed within pore spaces can significantly reduce the effective pore radius and divide pores, which results in a substantial drop in permeability. Excessive clay content also decreases the hardness of the rock and increases its susceptibility to deformation under pressure [63]. Dolomite and iron oxides, as cement, further contribute to pore filling, restricting fluid flow and reducing the overall connectivity of the pore network [64].

4.2. Role of Diagenetic Processes

Diagenetic processes significantly alter primary depositional textures and pore systems [67,68]. The co-occurrence of iron-bearing dolomite/siderite with hematite–goethite in the same beds reflects a two-stage diagenetic overprint and spatially heterogeneous redox conditions, not simultaneous equilibrium. During early burial, organic-rich pore waters consumed oxidants and favored reducing fluids that precipitated ferroan dolomite and pyrite. During later uplift/meteoric ingress, especially along exposure surfaces and microfractures, oxidizing fluids partially dissolved and oxidized Fe-carbonates/sulfides, precipitating goethite and hematite as rims and patches. These diagenetic minerals can significantly reduce pore throat radii and increase tortuosity, thereby reducing permeability [66,69,70,71]. The QA samples exhibit a higher cementation exponent (m) than the QW samples, indicating stronger intergranular cement bonding. However, pore connectivity in the QA samples remains superior due to a more prevalent rigid quartz framework. These results align with studies of similar sandstone reservoirs [72,73,74,75], which have highlighted diagenetic cementation as a key factor in porosity reduction.

4.3. Pore Structure and Specific Surface Area Relationships

Similar trends are observed for the specific surface areas per unit pore volume (Spv) and per unit grain volume (Sgv), further supporting the influence of microstructure on reservoir behavior. QA has significantly larger pore radii (mean = 0.59 µm) and a lower specific surface area than QW (mean = 0.40 µm). This signifies an inverse correlation between pore radius and specific surface area, which is a signature of coarser grains and a relative lack of fine-grained matrix material in QA. Furthermore, pore radius is strongly correlated with Spv (R² = 0.7273 and 0.9124 for QA and QW, respectively). This underlines how textural maturity controls the surface area available for fluid-rock interactions, which is the fundamental basis for capillary behavior and oil recovery [76].

4.4. Capillary Behavior and Recovery Efficiency

Capillary pressure (Pc) results in Figure 9 reveal marked contrasts between the two facies. QA exhibits low irreducible water saturation (Swi = 12.5%) and high microscopic recovery efficiency (Re = 84%), suggesting efficient displacement of wetting phases and, consequently, higher hydrocarbon recoverability. In contrast, QW shows a high Swi (40.5%) and a lower Re (59%). These values align with expectations, as QW’s smaller pore-throat sizes and stronger capillary forces result in higher entry pressures, impeding fluid movement, and directly reducing drainage efficiency [77,78]. Furthermore, heterogeneity in mineral composition and pore structure affects both static properties and dynamic fluid behavior. For instance, the abundance of clay minerals and smaller pores in tight sandstones like QW amplifies adverse effects from water-based fluids, further reducing fluid mobility and drainage efficiency [79,80,81,82]. Finally, lithologic heterogeneity, common in fluvial and point-bar reservoirs, causes significant lateral and vertical variations in porosity and permeability. This variability creates barriers and baffles to fluid flow, complicating drainage and making it difficult to predict reservoir behavior [83].

4.5. Predictive Modeling and Reservoir Implications

The strong linear regressions obtained for the Bahariya Formation sandstones (R² = 0.85–1.0) between the quartz, dolomite, and clay contents and the porosity, permeability, pore throat radius, Spv, and Sgv parameters indicate a direct and quantifiable effect of mineralogy on pore structure. In the QA facies, the positive influence of quartz and dolomite on pore throat radius and permeability is the strongest, demonstrating good preservation of intergranular porosity and the absence of significant cementation in these units. Moreover, the higher Fe oxyhydroxide and clay contents are associated with a smaller pore throat size and lower pore connectivity in the QW facies, resulting in lower permeability and higher specific surface area.

The good agreement between measured and regression-fitted data in this study also validates using the established equations for evaluating similar facies of the Bahariya Formation at Gebel El Dist. For instance, the QA permeability from the established model is similar to the measured mean permeability (21 mD) for this facies, while the QW model reproduces the lower average permeability (6 mD). In addition, the pore-throat radii predicted from the Winland and Aguilera equations are close to the measured values for the R₃₅ and R₃₆ radii, respectively, further verifying the ability of these equations to reproduce the true petrophysical response of the two sandstone facies.

These strong regressions and validated quantitative models provide robust evidence of a mineralogical effect on reservoir heterogeneity at Gebel El Dist. The study implies that changes in quartz, dolomite, and clay mineral content can be used to predict changes in reservoir quality across different stratigraphic units without the need for extensive laboratory petrophysical measurements. In the sampled section, the proposed models were able to distinguish between quartz-rich, high-permeability arenites and finer-grained, low-permeability wackes, thus serving as a useful tool for evaluating sandstone facies at the Gebel El Dist outcrop. Although these equations are currently tailored to this particular dataset, their excellent performance indicates that they can be used with a high degree of confidence to characterize other Bahariya Formation outcrops with a similar mineralogical composition and depositional setting.

4.6. Broader Geological Significance and Practical Applications

The findings from the Bahariya Formation provide a broader scope for the study of mineralogical effects on reservoir quality in clastic systems worldwide. From a geological perspective, the regression-based methodology presented in this study establishes a valuable link between petrographic observations and petrophysical analysis. By enabling the translation of mineralogical data into quantitative estimates of pore geometry parameters, the developed approach offers a framework that can be applied to a variety of siliciclastic reservoirs, regardless of their specific depositional setting (fluvial, shallow-marine, deltaic, etc.). Indeed, its relevance is especially pronounced in the early stages of hydrocarbon exploration, where subsurface data are often limited, and conventional reservoir characterization methods may be constrained. Outcrop-based analog studies, such as the one conducted in this research, can provide realistic input parameters for reservoir models, leading to improved predictions of reservoir heterogeneity and flow behavior even before drilling commences. This highlights the significance of outcrop studies, not only as a means of understanding past sedimentary and diagenetic processes but also as a source of practical guidance for the hydrocarbon industry.

The derived mineralogical–petrophysical equations can be used to quickly identify potential reservoir intervals from cuttings or outcrop samples, reducing the need for costly core retrieval and laboratory measurements. Predicting porosity and permeability directly from XRD data can helps in estimating reservoir quality and prioritize drilling targets. Furthermore, since mineralogical affects capillary pressure and wettability, it can determine the choice of EOR strategies, such as chemical or gas injection.

The same petrophysical relationships also govern the injectivity and long-term storage integrity of CO₂ reservoirs. The regression framework could be used to predict storage capacity and cap-rock sealing potential from mineralogical data, aiding in safer site selection and risk assessment. These predictive relationships between mineralogy and pore structure can further be applied to improve the assessment of aquifer productivity and geothermal reservoir performance, where permeability and fluid–rock interaction surfaces control the efficiency of heat and mass transfer.

The quantitative equations presented in this study can serve as empirical inputs for digital rock physics and geostatistical simulations, allowing geoscientists to upscale mineralogical heterogeneity into 3D reservoir models, improving the predictive capability of models for both conventional and unconventional systems.

4.7. Limitations

This investigation provides a quantitative mineralogical–petrophysical framework for a carbonate–clay–quartz sandstone system, but the study has some limitations. The dataset was generated from a small number of outcrop samples from the Bahariya Formation at Gebel El Dist. While these samples are representative of the reservoir facies, they might not capture the full lateral and vertical heterogeneity of the system. Moreover, outcrop analogs are inherently different from subsurface reservoirs in terms of burial history, stress state, and fluid evolution, which can influence cementation and pore geometry. The regression models are based on bulk XRD mineral data and do not differentiate between specific clay minerals or carbonate minerals that may have distinct effects on petrophysical behavior. However, these limitations do not invalidate the study’s conclusions but rather highlight research directions for refining predictive mineralogical–petrophysical relationships for diverse clastic systems.

5. Conclusions

This study provides an integrated mineralogical and petrophysical characterization of the Bahariya Formation at Gebel El Dist, in the Western Desert, Egypt, using its surface outcrops as analogues for subsurface reservoir units. The main findings are:

• Two sandstone facies, quartz arenite (QA) and quartz wacke (QW), display distinct mineralogical and petrophysical signatures that govern reservoir quality.
• QA shows higher ϕ and k, larger pore/throat radii, and lower specific surface area than QW, reflecting cleaner, better connected pores driven by mineralogy and diagenesis. QA is enriched in quartz and dolomite, which preserve intergranular porosity and promote better pore connectivity, while QW contains higher proportions of clays, iron oxyhydroxides, and dolomitic cements that occlude pores and restrict fluid flow. Consistently, QA has lower irreducible water saturation and higher recovery.
• Multivariate regressions (R² = 0.78–1.0) translate XRD-derived mineral percentages into ϕ, k, r, Spv, Sgv, and apex radii (R35/R36), yielding predictive equations on a facies scale. This model enables the estimation of reservoir properties directly from mineralogical data, supporting early screening and heterogeneity mapping in Bahariya analogues and similar clastic reservoirs where core data are limited.
• For both facies, the Aguilera R35 and permeability derived R36 parameters reliably reproduce the effective pore throat radius and outperform the porosity-derived R36 alternative.
• The methodology and predictive framework presented here can be readily applied to other clastic systems of similar composition and depositional setting, supporting more reliable reservoir characterization, performance forecasting, and exploration risk assessment in both conventional and unconventional sandstone reservoirs.