Storage and Production Aspects of Reservoir Fluids in Sedimentary Core Rocks

Abstract

Understanding the fluid storage and production mechanisms in sedimentary rocks is vital for optimising natural gas extraction and subsurface resource management. This study applies high-resolution X-ray computed tomography (≈15 μm) to digitise rock samples from onshore Cyprus, producing digital rock models from DICOM images. The workflow, including digitisation, numerical simulation of natural gas flow, and experimental validation, demonstrates strong agreement between digital and laboratory-measured porosity, confirming the methods’ reliability. Synthetic sand packs generated via particle-based modelling provide further insight into the gas storage mechanisms. A linear porosity–permeability relationship was observed, with porosity increasing from 0 to 35% and permeability from 0 to 3.34 mD. Permeability proved critical for production, as a rise from 1.5 to 3 mD nearly doubled the gas flow rate (14 to 30 fm³/s). Grain morphology also influenced gas storage. Increasing roundness enhanced porosity from 0.30 to 0.41, boosting stored gas volume by 47.6% to 42 fm³. Although based on Cyprus retrieved samples, the methodology is applicable to sedimentary formations elsewhere. The findings have implications for enhanced oil recovery, CO₂ sequestration, hydrogen storage, and groundwater extraction. This work highlights digital rock physics as a scalable technology for investigating transport behaviour in porous media and improving characterisation of complex sedimentary reservoirs.

1. Introduction

Oil and natural gas produced from oil and gas fields reside in porous and permeable reservoir rocks where these fluids have formed and accumulated throughout geological time. Geologically, source rocks are sedimentary rocks that were deposited in very slow flowing aquatic environments, usually in swamps on land, shallow tranquil marine bays, or deep submarine settings. In turn, source rocks are typically composed of fine-grained mineral particles interspersed with organic matter, primarily derived from algae and other microscopic organisms. Common types of sedimentary rocks comprise limestone, sandstone, siltstone, shale, conglomerate, and breccia [1].

The porosity and permeability of rocks are defining factors that could yield good reservoir formations. A rock that is both porous and permeable would make a decent reservoir medium, as it allows for the concealed oil and gas to propagate through the pores in the rock closer to the well where they are extracted [2]. Reservoir storage and flow capacity are first-order controls on reservoir behaviour, governing fluid distribution, sweep efficiency and ultimate recovery. Li [3] and Qingchao et al. [4] shed light on reservoir behaviour and recovery, showing that porosity and permeability govern these outcomes by linking rock properties to fluid distribution, transmissibility, and production forecasting. Hommel et al. [5], Kolenković Močilac et al. [6], Mohammadi et al. [7], and Sheykhinasab et al. [8] have investigated the relationship between porosity and permeability in natural reservoirs. Their findings highlight that these parameters are strongly interdependent and that new insights pertaining to their connection is crucial for more reliable characterisation of reservoir quality and prediction of storage and production performance of fluids. There exists a good correlation (R² = 0.95) between sandstone porosity and permeability while the relation between porosity and permeability of carbonates demonstrates a scattered distribution. Therefore, achieving an optimum balance between porosity and permeability can enhance natural gas production and storage efficiency.

Although traditional core analysis methods provide reliable porosity and permeability data, they are manual and resource demanding, often destructive, and limited in their ability to capture pore-scale heterogeneity. Rock core laboratory equipment has historically probed physical rock properties through experiments conducted on core plugs. Nowadays, digital rock physics (DRP) represent a time- and cost-effective alternative for determining rock attributes [9]. Many destructive experiments are routinely replaced by digital scanning techniques. Lee et al. [10] recommended a method for designing and 3D printing micromodels that mimic sandstone pores for porous media transport studies. The models were validated using imaging tools and simulations, demonstrating strong accord with permeability predictions. Digital rock physics enable the acquisition, management, preservation, and visualisation of images of porous materials and experiments. They also facilitate the basic analysis of associated measurements, such as porosity, capillary pressure, permeability, elastic properties, and various other properties, using electrical resistivity logs and NMR techniques. Such measurements are required to validate the modelling approaches and the construction and upscaling of larger (hydro)geological models.

A key component of digital rock physics is high-resolution digital imaging, particularly X-ray computed tomography (CT), which enables the non-destructive visualisation of a rock’s internal structure at the microscopic scale. It emphasises the importance of integrating high-resolution imaging techniques to accurately characterise complex pore structures in carbonate reservoirs [11]. Wang et al. [12] presents a multi-scale digital rock modelling method by fusing three-dimensional micro-CT images, acquired at two distinct resolutions, together with two-dimensional SEM images. The research demonstrates how combining different imaging scales enhances the accuracy of digital rock models, particularly for tight sandstone reservoirs.

Presumably the most important factor governing the physical behaviour of an actual rock sample is the rock microstructure, i.e., the distribution of the grains and pores. An understanding of pore-scale physics can provide valuable insights into the prevailing structure and estimation of the physical properties of lithologies. Digital rock physics provide high-resolution CT images of rock pores and mineral geometries from which various physical properties can be gleaned. Virtual rock cores can help in clarifying the relationship between porosity and permeability, the effect of lithological composition on physical properties, the geometry of grains, pores, and throats while conducting some experiments without subjecting the natural rock samples to physical alteration or destruction. Advanced rock physics is an effective tool that can be integrated with other techniques to capture the capabilities and properties of natural rock cores [9,13].

Although digital rock physics (DRP) has advanced significantly, its application as a validated alternative to conventional core analysis remains rather limited. Many studies focus on pore-scale simulations but stop short of comparing digital predictions to experimental data [14,15]. Additionally, DRP is seldom applied to simulate production-relevant metrics, such as gas flow rates or storage capacity [16]. The discovery of large accumulations of natural gas in the Eastern Mediterranean including Cyprus’ Aphrodite and Glaucus-1 fields calls for a better understanding of rock mechanics and the mechanisms of natural gas production and storage. Frontier exploration basins like onshore Cyprus, despite their geological relevance to offshore hydrocarbon systems, are largely underrepresented in digital reservoir studies [17]. This study addresses this gap by integrating high-resolution CT imaging, core plug validation, and synthetic grain pack modelling to evaluate how porosity, permeability, and grain morphology govern gas production and storage. The results demonstrate that DRP not only complements but, in many cases, outperforms traditional experimental methods in characterising complex reservoir rocks [18,19]. Accordingly, this research maps the relationship between porosity and permeability, and the volume of gas or the storage capacity of the porous media, which in turn define the natural gas transport velocity and the lithology’s gas production capabilities.

2. Geological Settings

After obtaining permission from the Cyprus Department of Geological Survey (DGS), which is the island’s competent authority for geological matters, the consortium of FabRocks research project identified several physical locations from which rock samples were collected. More specifically, rock specimens were retrieved from the Kantou area in Limassol and from Tochni, Agios Theodoros, and Xylophagou, in Larnaca, Cyprus. Figure 1 illustrates the lithological profiles of each section visited during the field campaign, along with the stratigraphic sequences of the collected samples. Physical locations of the sampling sites are depicted in Figure 2. The geological settings of Kantou, Tochni, Agios Theodoros, and Xylophagou villages were shaped by Troodos Ophiolite, the Mamonia Complex, and Messaoria Basin formations, respectively [20]. Rocks belonging to the Mamonia Complex are mainly deep-marine sediments and may serve as analogues to offshore source rocks. The lithologies found at Agios Theodoros, Tochni, and Kantou consist of sandstones and calcarenites, which can be compared to offshore clastic reservoirs. In contrast, the Xylophagou area is characterised by carbonate formations, particularly limestones, making it a good analogue for offshore carbonate reservoirs [21].

Kantou village located at the southern part of the Troodos Ophiolite is dominated by pillow lavas and sheeted dikes, which were once part of the upper oceanic crust. The geology of the village is associated with sulphide mineralisation and Cyprus-type volcanogenic massive sulphide (VMS) deposits. Of equal geological interest are the Neogene and Quaternary aged sediments, which contain sandstones, calcarenites, marls, and conglomerates. At the villages of Tochni and Agios Theodoros, located in the Larnaca district (Figure 2), one could encounter late Miocene sedimentary rocks, primarily sandstones and calcarenites. These formations were deposited in shallow-marine environments, influenced by fluvial, deltaic, or nearshore processes. However, the Mamonia Complex and the Kannaviou formation, which are much older in age, do not directly pertain to the cores studied in this research. Instead, these Neogene formations provide better analogues for offshore sandstone reservoirs. Geographically, the village of Xylophagou, situated at the boundary of the Larnaca and the Famagusta districts, is part of the Mesaoria Basin, which boasts thick sequences of Miocene to Pleistocene marine sediments, primarily limestones, calcarenites, and marls. These units constitute a shallow-marine carbonate platform, with evidence of karstification features in the limestones. The area’s relatively flat topography also includes quaternary coastal deposits. Notably, the limestone outcrops from Xylophagou, being carbonate dominant in nature, renders them valuable analogues for offshore hydrocarbons bearing carbonate reservoirs.

The selection of these formations stems from their potential as analogues to hydrocarbon systems encountered in the Eastern Mediterranean, exhibiting reservoir, seal, and source rock characteristics. Rock samples from the Kantou area contain sandstones from Neogene-Quaternary deposits, which resemble fluvial and coastal sandstone reservoirs. Rock samples from Tochni and Agios Theodoros consist of sandstones and calcarenites potentially serving as source rocks. In contrast, outcrops from Xylophagou are predominantly reef limestones, representing a shallow-marine carbonate platform like some Middle Eastern and Mediterranean carbonate reservoirs. These formations exhibit good secondary porosity emanating from karstification and dissolution, analogous to Egypt’s offshore Zohr gas field. Therefore, studying these core samples provides valuable insights into regional hydrocarbon exploration [22,23,24].

3. Sampling and Methodology

3.1. Experimental Measurements of Porosity and Permeability

After collecting the rock samples from the geological sites of interest, suitable size cores were obtained by drilling the rock samples and then trimming their edges. In total, 11 cores were extracted from the preceding rock specimens. The core samples were prepared at the Petrophysics Laboratory, at the University of Nicosia, which is fully equipped with Routine Core Analysis Laboratory (RCAL) devices, where essential petrophysical tests are conducted. Further to coring and trimming, all rock cores were thoroughly washed with water and oven dried at a temperature of 70 °C for a duration of 2 days to preserve any embedded carbonate material liable to the exposure of excessive heat. This setpoint is far below the temperature at which calcite (CaCO₃) decomposes (≥750 °C) and below the ~80 °C which is the dehydration threshold for gypsum.

Table 1. Physical dimensions of the cylindrical natural rock cores.

Core Name	Rock Type	Average Diameter (mm)	Average Length (mm)
K1	Sandstone	36.33	43.77
K2	Sandstone	36.04	46.24
A.T.C1	Sandstone	36.39	75.53
A.T.C2	Sandstone	36.89	66.83
A.T.S1	Sandstone	36.5	72.91
A.T.S2	Sandstone	36.84	66.15
A.T.S3	Sandstone	36.34	65.91
A.T.S4	Sandstone	36.79	59.31
T1	Sandstone	36.89	67.84
X1	Carbonate	36.83	54.63
X2	Carbonate	36.96	62.64

lists the dimensions of the core samples.

Helium porosimetry measurements were performed using a Vinci Helium porosimeter, which measures the effective porosity of cores samples by applying Boyle’s law in the context of helium gas expansion. The bulk volume of each sample was determined using a vernier calliper. Individual core samples were placed in a sealed chamber, and the porosimeter was calibrated using a reference volume billet to ensure precision. This reference volume billet is a solid stainless-steel cylinder measuring 36.8 mm in diameter and 65.9 mm in length. The billet has a volume of 70.06 cm³, at 20 °C, while measurements exhibit an uncertainty of ±0.5 mm. Once the hardware attained equilibrium at 20 °C after 30 min, the pressure sensors were reset to 0, and we ran an empty cell leak test. During this test, the cell was sealed in the absence of a rock core, pressurised to ~300 kPa (3 bar; 43.51 psi), and isolated for 60 s. The test was successful provided the pressure drop was less than 2 kPa (0.02 bar; 0.3 psi). For volume verification, we performed multi-pressure checks (100–500 kPa or 14.5–72.5 psi). Volume calibration was realised when the computed cell volume matched the billet within ≤0.1% during 3 iterations, which deviated within ≤0.05 cm³. During measurements, the reference cell was charged to 200 psi (≈1380 kPa); then, helium was allowed to expand into the sample cell, and the equilibrium pressure was recorded. The pressure sensor exhibited an accuracy of 0.1%.

Using Boyle’s law, it was possible to determine the grain volume by subtracting the void volume from the total chamber volume. This method yields accurate measurements of effective porosity [25].

Absolute permeability was determined using a Vinci permeameter in a Hassler-type core holder. Cores were loaded into the sleeve, and a confining pressure was applied to simulate overburden pressure and ensure radial sealing while the confining pressure was kept higher than the pore pressure. Distilled water was used as the test fluid. An upstream water reservoir was pressurised with nitrogen gas to provide a controlled inlet pressure. Nitrogen gas presumably did not completely permeate the core. After steady state was reached, the effluent water was collected in a container of known volume, and the time to fill that volume was recorded to calculate the flow rate. The pressure drop across the core was measured between the inlet and the outlet. Utilising Darcy’s law, permeability was determined from the measured flow rate, water viscosity, core dimensions (length and cross-sectional area), and the pressure drop. This procedure yielded reproducible absolute permeability values for the samples. Accurate permeability measurements of rock cores are instrumental for evaluating their fluid flow potential and the transmissibility characteristics of petroleum and water reservoirs. Accurate permeability measurements of rock cores are instrumental for evaluating their fluid flow potential and the transmissibility characteristics of petroleum reservoirs [24].

3.2. Micro-CT Scans of Sandstone and Carbonate Rocks

After the rock cores were prepared, the next step was to scan them and generate their 2D images. These renderings were subsequently used to create digital models of the respective porous rocks. Rock core samples were scanned at the Agios Therissos medical laboratory, in Nicosia, using a Philips 256 CT imaging devise (Philips Healthcare, Best, The Netherlands). The cores were scanned using the fine-bone ear/sinus protocol (120 kV). Data were acquired as multiple small, overlapping scans at 15 μm isotropic voxels, each reconstructed at 1024 × 1024 pixels and 256 slices. Scanned images overlapped by 10% across the core diameter and along its length and were then fused into a single 3D volume. Pixel spacing and slice thickness from the DICOM headers both measured 0.015 mm.

3.3. Image-Based Rock Characterisation and Pore-Scale Modelling

Pore network modelling involves segmenting 2D images of rock samples into pore and solid phases, with the goal of reconstructing the pore space geometry and simulating fluid transport behaviour through the network [26]. The accuracy of such estimations depends on the fidelity of the digital pore network that replicates geometrical and topological features of the natural pore system. Our efforts initially focussed on identifying a software tool that combines various processes, such as segmentation, reconstruction, volume rendering, meshing, and eventually the computation of the rock physical properties. The Avizo materials characterisation suite was used to segment the 2D images. Many subroutines and algorithms, such as the watershed filter, 12 auto-thresholding, axis connectivity, and the object separation function were applied to distinguish between the solid matrix and the pore network of rocks (Figure 3).

Porosity was determined using the volume fraction method, which calculates the ratio of pore voxels to total sample voxels. Simply put, Avizo can compute the collective porosity of the pores and their throats. The Avizo network modelling extensions of Xlab and Xpore (Avizo (2019.1)) were utilised to determine the absolute permeability and the natural gas flow rate through the porous structure.

According to Zhang et al. [27], Avizo’s pore-network module estimates throat conductance using the Hagen–Poiseuille relation. In our study, permeability was computed from the CT-segmented voxel geometry using two independent approaches under single-phase, incompressible, low-Re conditions (Re << 1): Continuum flow on voxels (Avizo Xlab extension) and Pore-Network Modelling (Avizo Xpore Network extension).

While using Xlab extension, we solved flow in the creeping-flow (Stokes) limit of the incompressible Navier–Stokes equations using XLab’s lattice-Boltzmann method. The governing equations are:

$$\begin{array}{l}\nabla \cdot \overrightarrow{V}=0\\ \mu {\nabla }^2\overrightarrow{V}-\nabla P=0\end{array}$$

(1)

where ∇·is the divergence vector operator, ∇P denotes the vector derivative of the scalar pressure field (P), μ is the dynamic viscosity of the fluid, and ∇² is the Laplacian operator.

A pressure drop, $dP$, was prescribed between inlet and outlet faces, and no-slip was imposed at the solid surface. After steady state, the global flow rate ($Q$) is converted to absolute permeability via Darcy’s law:

$$Q={\displaystyle \frac{kA}{u}}{\displaystyle \frac{dP}{dx}}$$

(2)

where Q is the global flow rate supported by the porous medium (m³/s), A is the cross-sectional area of the sample that the fluid permeates (m²), k is the absolute permeability (m²), and μ is the dynamic viscosity of the flowing fluid (Pa∙s).

Flow rates were computed from Darcy’s law, using each core’s permeability, geometry, the applied pressure drop, and the gas viscosity at laboratory temperature.

The second approach was using Xpore Network extension, where a pore–throat network was extracted from the binary images. For each throat t between pores i and j, flow is given by Poiseuille conductance:

$$Q_t = g_t \left(p_i - p_j\right),g_t ={\displaystyle \frac{{A_t}^2}{{\alpha }_t \mu {\mathcal{l}}_t }}$$

(3)

$Q_t$ is volumetric flow through throat t (m³/s). Positive from pore i → j, $g_t$ hydraulic conductance of throat of throat t (m³·Pa⁻¹·s⁻¹), $\left(p_i - p_j\right)$ pressures at the two pore nodes (Pa) connected by throat t, A_t is the throat area, ℓ_t is the length of throat t, μ is the dynamic viscosity of the fluid (Pa·s), and ${\alpha }_t$ a shape factor (provided by Avizo/equivalent-radius model).

Nodal pressures satisfy mass conservation ${\sum }_{t\in i} Q_t =0$ The network’s total outlet flow $Q$ is converted to permeability using Darcy’s law as above.

3.4. Synthetic Sediment Pack Modelling (DEM Approch)

Many difficulties were encountered while converting the DICOM into vector images. For this reason, realistic unconsolidated sediments were built using Particula 1.3 [26]. The deposition of grains of specified shape and size for 17 sand pack (structure) distributions was simulated by means of a discrete element method (DEM). Such a robust approach considers the interactions between the individual elements and is commonly used in the context of soil mechanics and geotechnical studies [26]. Because individual grains were successively deposited under the influence of gravity, this process resembles— to a large degree—the natural sedimentation process. In this project, the DEM approach was utilised owing to the possibility of implementing irregularly shaped convex particles. Depending on the application, this may constitute a major advantage as simplified spheres or ellipsoids generally do not reflect the granulometric characteristics of natural sandstone and carbonate rocks.

Structurally, the irregularly shaped grains were generated by stochastically perturbing regular shapes by means of a three-dimensionally correlated Perlin noise, whereby the volume of the underlying regular particles was conserved. Considering that each particle exhibits a separate realisation of Perlin noise, all distinct grains are unique in shape and size. Moreover, the selected amplitude of the noise governs the strength of the perturbations and, thus, it offers a good approximation for natural shapes of sub- to well-rounded grains. Primarily, the basic input parameters for generating the synthetic sediments are the grain size distribution and the aspect ratio, which in turn, determine the ellipsoidal grain shape. In this respect, a Perlin noise of 0.75 was used to generate the irregularly shaped particles, as visualised in Video S1. The pack consisted of 5000 quartz grains of a mass-density of 2650 kg/m³ and with a controlled size distribution, featuring three distinct grain diameters, namely 50 μm, 35 μm, and 20 μm. All grains were deposited under the influence of gravity into a cylindrical container measuring 36 mm, in diameter, by 66 mm, in length (Video S1). This irregularity in shape, raised inter-particle friction and mechanical interlocking, which in turn enhanced the complexity of the interactions and led to a more realistic packing structure compared to ideal spherical grains. Furthermore, the model considered compaction forces, simulating the gradual densification of the pack as the grains precipitated at the bottom of the container. Key assumptions referred to the grains accumulating in a cylindrical geometry resembling the size of the rock cores under atmospheric pressure conditions, as shown in Video S1.

Thanks to the homogeneous nature of the sandstone sediments, sand packs of sandstones were relatively easy to create. Layers of carbonates, on the other hand, cannot be easily replicated owing to the high heterogeneity that carbonate rocks exhibit. Avizo was also able to deduce the porosity of the constructed packs, like the one illustrated in Figure 4.

After the generation of these samples, the Avizo suite was used to segment, reconstruct, and render the volume of the lattices, as summarised in Figure 5. In parallel, it was possible to validate the porosity and compute the absolute permeability, the gas entrained in the pores, and the gas flow rates. Equally important, the same porous material can be replicated if the same properties, such as porosity, grain distribution, and mass density are maintained. In other words, fixing some of these parameters offers control over permeability. Hence, unconsolidated sample material structures were obtained from the same porosity and permeability with minute changes to their internal structure.

4. Results

4.1. Experimentally Determined Porosities and Permeabilities

Both the porosities and the permeabilities of the eleven natural rock cores are shown in

Table 2. Experimentally obtained core porosity and permeability measurements.

Core Name	Experimental Porosity (%)	Porosity SD (%)	Experimental Permeability (mD)	Permeability SD (mD)
K1	29.10	0.25	2.044	0.03
K2	27.03	0.22	2.141	0.02
A.T.C1	18.99	0.20	<1	-
A.T.C2	22.10	0.21	<1	-
A.T.S1	27.62	0.24	<1	-
A.T.S2	22.39	0.23	<1	-
A.T.S3	25.97	0.21	<1	-
A.T.S4	22.88	0.22	<1	-
T1	32.51	0.25	3.188	0.05
X1	30.67	0.21	0.049	0.07
X2	33.10	0.24	0.031	0.06

SD = Standard deviation across triplicate measurements (n = 3).

. Sandstone cores possess a porosity that ranges between ≈19% and 32.5%. Notably, carbonate cores, such as X1 and X2, exhibit high porosities that range between 30.7% and 33.1%. The preceding values indicate that these cores are highly porous and could potentially constitute good reservoir formations. Porosity alone is not sufficient to evaluate the quality of sandstone and carbonate reservoir rocks. Although it indicates the total void fraction available for potential storage, it does not provide enough information about the connectivity of the pore space. Permeability must also be considered, as it reflects the ability of fluids to flow through the rock. Only by combining porosity and permeability can the storage capacity and producibility of a reservoir be reliably assessed.

Obtaining the permeabilities of such cores proved more challenging than expected. Owing to the tight nature of the rock samples, it was possible to measure the permeability of some of the cores, such as K1, K2, T1, X1, and X2. Permeability values ranged between 0.03 mD and 3.188 mD. For several tight rock samples, steady flow was very small, where a value below 1 mD could still be quantified reliably and reported numerically. These operations typically required extended acquisition times—from several hours up to multiple days—to accumulate sufficient effluent volume, at a stable pressure drop. Where flow remained below the method’s detection limit under our test conditions, permeability is reported as “<1 mD”. In other words, this notation reflects the instrument limitations and does not imply zero permeability of a core. Put differently, these measurements indicate that the studied cores exhibit, in general, low permeabilities and may not adequately represent productive reservoir lithologies. Instead, they indicate heterogeneity and potentially low-permeability zones of the reservoir.

A total of 11 cores were subjected to CT scanning, and 3D volumes were generated for all. However, due to the computationally intensive nature of the simulations, six representative cores were selected for detailed reconstruction and pore-scale modelling.

Table 3. The experimental and digital porosities and permeabilities of various sandstone and carbonate rock cores.

Core Name	Experimental ϕ (%)	Digital ϕ (%)	Δ(ϕ) (%)	Experimental k (mD)	Absolute k (mD) via Xpore Network Modelling	Absolute k (mD) via X-Lab Extension	ΔLog10k (Xpore−Exp)	ΔLog10k (Xlab−Exp)
X2	33.1	35.2	2.10	0.031	0.049	0.047	0.179	0.177
T1	32.51	33.92	1.41	3.188	3.873	3.856	0.085	0.083
A.T.S2	22.39	24.56	2.17	-	3.569	3.554
K1	29.1	31.23	2.13	2.044	3.089	3.069	0.199	0.181
A.T.C1	18.99	20.67	1.68	-	1.568	1.537	-	-
A.T.C2	22.1	24.43	2.33	-	1.672	1.656	-	-

illustrates the porosities and the permeabilities obtained from Avizo. Digital porosity exceeded laboratory values by 1.97% on average (95% confidence interval is 1.60–2.34; paired t-test p = 3.6 × 10⁻⁵), indicating a small, consistent positive offset with strong overall agreement. For the three cores with numeric laboratory values, K1, T1, and X2, a paired t-test on the log10(k) showed that the Xpore method exhibited a mean difference of 0.154, with a 95% confidence interval (CI) from 0.002 to 0.306 and p = 0.0486. This corresponds to the digital permeability being about 1.43 times the laboratory value on average. A second paired t-test on the same cores showed that the X-lab method had a mean difference of 0.147, with a 95% CI from 0.0088 to 0.284 and p = 0.0445, corresponding to about 1.40 times the laboratory value. Hence, both digital permeability methods are in good agreement with the laboratory measurements, demonstrating the same trends across the three paired cores.

High porosity with low permeability indicates that much of the pore volume is poorly connected or linked by narrow pore throats; therefore, flow is restricted, even though total void space is considerable. For reservoir quality this implies sound storage but limited deliverability. Therefore, porosity alone can overstate quality in these intervals. Meanwhile, permeability must be considered when ranking or forecasting porous media performance.

4.2. The Porosity–Permeability Relationship in Sandstone Cores

The numerical analysis of the sandstone cores revealed a clear relationship between porosity and permeability (Figure 6). As demonstrated in Figure 6, the coefficient of determination (R²), which amounts almost to 1, captures a perfect straight-line relationship between porosity and permeability. Sandstone samples establish a clear porosity–permeability correlation, as reflected by the gradient of the straight line, which is almost 1 and intersects the origin. There seems to be a linear relationship between these two physical rock properties. This means that an increase in porosity will be accompanied by a rise in permeability. As porosity increases from 0% to 35%, permeability grows from 0 mD to 3.34 mD. This relation is valid for all scanned natural sandstone cores. For example, the A.T.S2 core possesses a porosity (ϕ) of 24.6%, while T1 exhibits a higher porosity of 33.9%. Likewise, the permeabilities for A.T.S2 amounted to 3.57 mD while T1 boasts an even higher permeability of 3.87 mD. Apparently, the homogenous nature of the sandstone cores explains their porosity–permeability correlation. However, caution should be exercised when interpreting these results as the cohort of core samples tested is rather limited. Thus, to draw more concrete conclusions the investigation needs to be expanded to consider a wider spectrum of samples possessing a range of permeabilities.

4.3. Gas Volumetric Flow Rate–Permeability Relationship in Sandstone Cores

To-date, Darcy’s law has been extensively applied in the analysis of fluid flow behaviour in porous media. Among others, currently, the law is used for rock core investigations, reservoir characterisation and fluid flow simulations [27]. As suggested by Dybbs and Edwards [28], four different flow regimes can be observed in porous media depending on the flow velocity and the properties of the porous body. These four regimes comprise: (1) pre-Darcy, (2) Darcy or laminar, (3) Forchheimer, and (4) turbulent flow. Generally, the differentiation between the flow regimes is based on the Reynolds number, with flow transitions in porous media considered smooth, albeit lacking a distinct threshold.

The Reynolds number in a porous medium can be obtained from Re = (ρUd)/μ, where ρ is the fluid mass density, U is the fluid velocity, μ is the dynamic viscosity of the fluid, and d is the average diameter of the grains in the porous medium. In the laminar flow domain, the Reynolds number is smaller than 1 [27]. In this regime, the flow is dominated by viscous forces, and the pressure gradient varies linearly with the flow rate. However, the Forchheimer flow is characterised by intense inertial effects, and the pressure gradient is a parabolic function of the flow rate.

Being smaller than 1, these Reynolds numbers indicate the presence of laminar flow in the rocks [28]. The average Reynolds number for the samples was ≈0.72, thus lending credibility to the utilisation of Darcy’s law which is applicable to laminar flow conditions. Figure 7 depicts the linear relationship between the natural gas flow rate (fm³/s) and the pressure gradient ΔP(psi) where R² ≈ 1. It must be noted that other sandstone natural cores show similar linearity with different slopes. Furthermore, Figure 7 demonstrates the application of Darcy’s law and confirms that the flow regime is creeping, so the gas is treated as an incompressible fluid.

To further demonstrate a linear correlation of the gas flow rate as a function of pressure, the relationship between the gas flow rate and the rock permeability were considered. Figure 8 displays the gas volume flow rate relative to the porous medium’s permeability (mD).

The same diagram (Figure 8) reveals that as permeability increases, the gas flow rate rises proportionally. More specifically, the volumetric gas flow rate almost doubled from 13 fm³/s at 1.568 mD to 36 fm³/s when permeability increased to 3.873 mD. It also becomes apparent that the rock samples exhibiting high permeability and constant porosity, sustain a significantly elevated gas flow rate. As expected, enhancing the permeability of a sandstone sequence exerts a bigger impact on the gas flow rate compared to higher porosity. Finally, the gas production rate and the recovery of the natural gas are more sensitive to changes in permeability than fluctuations in porosity.

It is important to note that the flow rates derived from permeability–flow correlations (Figure 8) and those obtained from porosity–flow relationships (Figure 9) are not expected to coincide. In the first case, flow is directly related to permeability, which governs pore connectivity and fluid transport. In the second case, flow is inferred indirectly from porosity, which only reflects void space and is a weaker predictor of transmissivity. The discrepancy between the two approaches underscores the dominant role of permeability in controlling fluid flow, whereas porosity provides complementary but less definitive insights.

4.4. Gas Volumetric Flow Rate-Porosity Relationship in Sandstone Cores

To further validate the relationship between porosity and permeability of the real sandstone cores, another parameter taken into account was the gas volumetric flow rate (in femto cubic metres per second, fm³/s). Clarifying the relation between porosity and gas flow rate can improve the accuracy of reservoir performance predictions and support the optimisation of natural gas production strategies. Figure 9 shows the variation in the volume flow rate (fm³/s) of the gas as a function of porosity (%). Additionally, Figure 9 demonstrates that the increase in porosity does not exert a proportionate impact on the gas flow rate. Therefore, the enhanced porosity does not seem to play a major role in boosting natural gas production.

4.5. Storage Capacity-Porosity Relationship in Sandstone Packs

Besides the influence of porosity and permeability on the gas flow rate, the storage capacity or the volume of the gas entrained in the porous media was computed.

Table 4. The variation in gas volume as a function of the porosity and the sand pack’s permeability.

Pack Number	Porosity (%)	Permeability (mD)	Gas Volume (×10−15 m3)
1	5	0.75	3.59
2	10	0.98	7.18
3	15	1.27	10.77
4	20	1.66	13.53
5	25	1.87	17.21
6	30	2.47	21.55
7	35	3.34	25.14

lists the volume of gas entrained in various sand packs as part of a parametric investigation considering a porosity range spanning between 5% and 35%.

Increases in porosity and permeability enhance the gas holding capacity of the rocks and the stored gas volume. In other words, natural gas storage is collectively affected by the structure’s porosity and permeability. Since our sandstone samples’ permeabilities are very small compared to the permeabilities of other natural gas reservoirs, which can range between 100 mD and 800 mD, it can be inferred that, in our case, gas storage is chiefly governed only by porosity. These packs are tight with low permeabilities, so they transmit only small flow rates; the gas that can enter the connected pore network and be stored/delivered is very small.

Therefore, a higher natural gas storage capacity is possible only if the rock porosity is appreciable. Figure 10 displays the relationship between the rock porosity and the gas volume.

A coefficient of determination (R²) of almost 1 reveals a linear function (straight line fit) between the stored gas volume and the rock porosity (Figure 10). That is, the obtained linear relation facilitates the calculation of the gas volume storage capacity of the sandstone cores as a function of lithological porosity. For sandstone cores, it was possible to obtain the corresponding natural gas volume, as shown in

Table 5. The respective gas volume of natural sandstone core samples at specific porosities.

Core Name	Porosity (%)	Gas Volume (×10−15 m3)
T1	33.92	24.18
A.T.S2	24.56	17.5
K1	31.23	22.26
A.T.C1	20.67	14.73
A.T.C2	24.43	17.14

.

4.6. The Impact of Granular Roundness on Storage Capabilities

The impact of the grains’ roundness on the storage capabilities of sandstone lithologies has been examined using the Particula 1.3 software package. As mentioned earlier, several sand packs were prepared, each exhibiting distinct roundness and grain size. Individual sand packs were collected in a cylindrical container, carefully designed to replicate the dimensions of its corresponding natural rock core. Moreover, the porosity, 3D volume, and gas volume of each sandstone accumulations were determined, as illustrated in

Table 6. The variation of porosity and gas volume in relation to grain roundness.

Roundness	Porosity(%)	Gas Volume (×10−15 m3)
0	30	22
0.1	33	26
0.2	35	29
0.3	37	30
0.4	38	31
0.5	38.3	32
0.6	38.7	32
0.7	39	35
0.8	39.4	37
0.9	39.8	40
1	41	42

.

Close inspection of the results in Table 6 reveal that when the grains assumed a more spherical shape, as reflected from an increase in their roundness indicator from 0 (no roundness) to 1 (perfectly spherical shape), an enhancement in rock porosity is induced from 30% to 41%, which translates into a 26.8% boost in porosity. In parallel, the stored gas volume grew by 47.6% to 42 fm³ (femto cubic metres).

5. Concluding Remarks

Overall, the analysis of experimental porosities and permeabilities in sandstone and carbonate cores residing onshore Cyprus offers valuable insights into their suitability as reservoir rocks and could potentially serve as analogues for correlation studies. Sandstone cores displayed a wide range of porosities, indicating substantial void spaces, while carbonate samples exhibited notably higher porosities, suggesting favourable reservoir properties. Challenges in permeability measurements prompted the adoption of high-resolution computed tomography imaging and the utilisation of the Avizo software for digital analysis. This approach revealed a strong correlation between porosity and permeability in sandstone cores, although the limited permeability range necessitates further investigation across several diverse samples. This relationship is likely attributed to the homogeneity of the samples, which facilitates uniform pore connectivity.

Additionally, exploring the relationship between porosity, permeability, and the gas flow rate underscored the intricate interplay of these factors in reservoir performance. A strong relationship between permeability and gas flow rate reinforces the validity of Darcy’s law under laminar flow conditions. Notably, computational simulations confirmed that enhancing permeability significantly boosts natural gas production, with flow rates increasing from 14 to 30 fm³/s, as permeability doubles from 1.5 mD to 3 mD. This highlights permeability as a critical factor in optimising natural gas extraction from reservoir rocks. At the same time, porosity, which expanded from 0 to 35%, was shown to influence gas storage efficiency. As a result, the entrained gas volume increased from 4.7 to 25 fm³, demonstrating that porosity plays a dominant role in gas retention, especially in low-permeability rocks. It is important to note, however, that permeability influences storage indirectly by governing pore connectivity and accessibility. While porosity defines the maximum storage volume, permeability determines how effectively this pore volume can be accessed and utilised under reservoir conditions. Our findings are consistent with Heap et al. [29], who demonstrated that X-ray computed tomography imaging can effectively capture internal heterogeneity and permeability variability, especially in volcanic and sedimentary formations. This supports the reliability of CT-derived models used in our study to assess gas flow potential.

Grain morphology, particularly roundness, plays a significant role in influencing porosity. An increase in roundness, modelled using synthetic grain packs from Particula 1.3, led to a 26.8% rise in porosity and a corresponding 47.6% increase in stored gas volume. The influence of grain shape on a rock’s storage capabilities underscores the importance of comprehensive microstructural analysis for accurate reservoir characterisation.

Compared to previous studies that focused predominantly on digital simulations, such as those by Blunt, Bijeljic, Dong, Gharbi, Iglauer, Mostaghimi, Paluszny, and Pentland [16] and Alqahtani, Alzubaidi, Armstrong, Swietojanski, and Mostaghimi [15], our work is distinctive in its integration of high-resolution imaging with experimental core plug validation and production-relevant modelling. Prior studies often lacked experimental benchmarks and did not yield entirely realistic flow and storage behaviour. In contrast, this study establishes validated relationships between digital predictions and empirical measurements, bridging the gap between pore-scale analysis and reservoir-scale performance.

In sedimentary reservoirs, the volume of natural gas trapped within the rock pores is fundamentally controlled by heterogeneity in porosity and pore structure and shaped by depositional textures and diagenetic processes. In carbonates, diagenetic alterations caused by dolomitisation and dissolution can significantly enhance porosity, while tight sandstone reservoirs may develop secondary porosity through grain dissolution, contributing positively to reservoir quality [30,31,32,33]. These observations align with our findings, reinforcing the critical role of both the original rock texture and diagenetic history.

The implications of our findings are especially significant for sedimentary formations in Cyprus and the broader Eastern Mediterranean. These regions often feature complex and heterogeneous reservoirs where traditional characterisation methods may fall short. By applying a digital workflow that incorporates both realistic simulations and experimental validation, this study offers a practical approach for reservoir assessment in frontier exploration areas with limited core availability.

Looking ahead, future research directions could expand on this foundation by examining how microstructural properties, such as grain roundness or connectivity, evolve under overburden pressure. Additionally, extending digital analysis to properties such as tortuosity and wettability may further improve predictive models and enhance reservoir management strategies. Ultimately, this study contributes to a more complete understanding of natural gas production and storage mechanisms and supports the development of optimised workflows for detailing, characterising, and managing complex sedimentary reservoirs.