Sir Godfrey Hounsfield invented the first computed tomography (CT) scanner utilizing X-ray technology at EMI Central Research Laboratories in 1967. Hounsfield explained that the study team intended to create a three-dimensional (3D) image of a box by re-examining it as a succession of slices. This spurred additional research at EMI, which resulted in the first commercially successful CT scan in 1971 (Arns et al., 2005; Gharieb, 2022). CT scanning generates cross-sectional images of the body using computers and rotating X-ray machines. They may display the soft tissues, blood vessels, and bones in several body parts, including the head, shoulder, spine, knee, and chest, among others. It is used for diagnosing diseases and assessing injuries, and it aids physicians in identifying bone fractures, muscle issues, and monitoring therapies for conditions such as cancer and heart disease (Gharieb, 2022). By examining the intricate details of one or more core samples, the image data generated by these instruments are utilized to better comprehend the nature and characteristics of light oil reserves. It is capable of revealing interior microscopic properties such as grain size and shape, pore size and network, lithology, and in situ liquid distribution (Coles et al., 1991; Simjoo et al., 2013). In addition to image analysis, CT scanners are utilized for a vast array of applications, including routine and specific core analyses, petrophysics, formation damage, petroleum geology, reservoir modeling, and digital rock typing (Saraf & Bera, 2021; Simjoo et al., 2013). The primary objective of this review article is to provide useful information to maximize the usage of CT scanning in the oil and gas industry.

The advent of nano-scale techniques such as nano-X-ray microscopy (nano-XRM), helium ion microscopy (HIM), and focused ion beam scanning electron microscopy (FIB-SEM) (Wargo et al., 2013) has allowed for resolutions of many nanometers for FIB-SEM and nano-XRM, and tens of angstroms for HIM (Hlawacek et al., 2014; Menke et al., 2022). Thus, 3D X-ray CT scans with a voxel count of up to 8 × 109 on core plugs with a diameter of 5 cm and a pixel count as small as 2 μm can be acquired. With this method, the pore space of a rock can be seen in 3D at varying scales (Sakellariou et al., 2003b). The data were examined after scanning an Estaillades limestone micro-core with X-ray microcomputed tomography (μCT), monitoring the pressure gradient during single-phase flow and flooding the rock with changed brine. The contrast between the images was used to assign porosity to each microporosity-linked voxel. The pore-space flow was determined using both Stokes–Brinkman (S–B) and Stokes-only solvers, and the differences between the observed and calculated permeabilities were evaluated (Hlawacek et al., 2014; Menke et al., 2022).

Pore network models are often used to simulate a wide range of geophysical and contaminant processes, such as non-Newtonian displacement, phase exchange, thermodynamically consistent reactive transport, and oil layers that do not flow in a Darcy-like manner (Xiong & Jivkov, 2015; Xiong et al., 2016). Using 3D image segmentation, a pore-network model of a core sample was derived. Using X-ray CT and specialized software, a thorough digital 3D representation of the core sample was produced and subsequently utilized in this experiment (Rasmusson et al., 2021). Consequently, the engineering community has recognized that pore network models are a useful tool for comprehending and predicting meso-scale networks, as they bridge the gap between single-pore processes, for which other methods are more accurate and homogenized in continuous porous media (Riepe et al., 2011; Xiong et al., 2016).

The practical importance of using pore models is related to the calculation of macroscopic characteristics such as capillary pressure and absolute and relative permeabilities, which are significant variables in almost every reservoir model (Sakellariou et al., 2003a). Images of the cross-sections of porous rocks are surprisingly informative in terms of their genesis, composition, and hydrodynamic capabilities. With advancements in imaging techniques, additional features can be retrieved from images to improve the understanding of the porous structure (Rabbani & Jamshidi, 2014). Despite the fact that 3D imaging techniques are more desirable today, two-dimensional (2D) analyses of rock pictures are still commonly used due to their simplicity and cost-effectiveness. Various hydrocyanic factors can be derived from 2D images that correlate well with experimental permeability values. Analysis of pore space imaging of rocks can yield pore size, grain size, specific surface area, porosity, coordination number, and material characteristics (Rabbani et al., 2014). Numerous techniques exist to obtain cross-sectional photographs of rocks. Micro-photography, which requires a confocal or reflecting microscope for examination of the rock surface and evaluation of petrophysical qualities, is the most accessible technique for the study of rocks (Kułynycz & Maruta, 2017; Raeini et al., 2012).

The creation of images has become a vital aspect of our daily lives, from photography using a variety of gadgets to medicinal applications. The oil and gas industry has been using this technology for a long time. Currently, CT scans are predominantly used for upstream applications, including image-based visualization of microphysics parameters. Upstream applications calculate fluid characteristics and conduct pore-scale fluid flow analysis using the obtained parameters. A detailed review article on the science underlying this technology, its application in the oil and gas industry, and future enhancement suggestions for industry professionals is required due to the diverse applications of this technology in various contexts.

This review study describes the components of micro-scale imaging techniques. Using presentation diagrams, an overview of CT scan image processing and the application of imaging techniques in reservoir categorization is provided for the overall evaluation structure. After a thorough examination of the aforementioned topics, the focus shifts to network extraction, with a focus on image processing, followed by an in-depth discussion of the review's foundation and inspiration, novelty, relevance, and significance, as well as the porous network model. The following section describes exploratory efforts of petrophysical property determination and demonstration at different sizes with petrophysical characteristics affected by core flooding. It is known that the methods of calculating and processing CT scan images as well as the most recent CT scan research in the oil and gas industry through industrial CT scans provide enough information.

2 NOVELTY OF THE REVIEW

Review articles usually provide information on fundamentals of a topic, development and progress of research, current status, and future prospects. However, there are several review articles in the literature on imaging techniques for reservoir characterization. All the existing studies focus on specific aims from their perspective and in terms of the topic and presentation. This review is unique because of an in-depth discussion and thorough application-based presentation of previous and current studies. Application of the CT-scan imaging technique is a game-changing and significant addition to the reservoir characterization method. Significant developments have been made in the last decade. However, a thorough and fundamental review is not available in the literature. Therefore, this review has been carried out to highlight the importance of CT scan in reservoir characterization, especially for newcomers in this field. Moreover, this review also highlights the challenges of the analysis of CT scan data and their use. A precise description of the prospects and incorporation of DRP in the CT-scan technique represents another novelty of the review.

3 DIFFERENT SCALE IMAGING TECHNIQUES

An image is the visual representation of an object or concept. Various imaging techniques have been used to better understand phenomena, store relevant records, and, in some cases, for esthetic purposes (Al-Marzouqi, 2018; Almetwally & Jabbari, 2020). Drawing is the most basic form of visualizing something in one's mind. Humans use drawing tools to create images of things both seen and unseen. However, sketches can be inaccurate due to the artist's interpretation, and they might not include all the details that are necessary for a complete scientific understanding. Photoshop, PowerPoint, and so forth have made digital drawing possible, but they still do not provide diagrams of real scientific value (Lopez et al., 2016; Ravelli & Van Leeuwen, 2018). Photography is a more refined alternative to traditional imaging methods like drawing. Here, objects are photographed with cameras set up to capture images of a specific pixel quality.

Technological advancements have made photography somewhat less difficult, with some portable devices now capable of generating high-quality photos. Most of the time, people take pictures to record events, make visual observations, and for esthetic purposes (Li et al., 2010; Zou et al., 2016). Researchers cannot obtain the high-quality images needed for a rigorous scientific study. Therefore, advancements in scientific imaging techniques have been made to facilitate the generation of higher-quality images suitable for scientific inference. SEM, CT scan, and MRI are all examples of imaging methods that are used to investigate the surface and bulk structures of materials. These methods can capture images behind barriers, like human organs, the insides of rocks, fluid saturation profiles in core samples, and so forth, each of which can be used in a wide range of scientific fields (Hajiabadi et al., 2019; Saraf & Bera, 2021; Wang et al., 2015).

3.1 X-ray imaging

X-ray (a form of electromagnetic waves) has been used extensively in medicine (Ahn et al., 2013; Simjoo, Dong, et al., 2012; Simjoo et al. 2013). The methods are based on the concept that the rays can travel unimpeded through flesh and fat and are impeded only by bones. The patient is exposed to irradiation, and the radiation is sent to a digital recorder, where it is captured as an image (Sherer et al., 2013). It is well known that X-ray imaging is used in the medical industry for the production of chest X-rays for the diagnosis of conditions related to the lungs (Ahn et al., 2013; Sherer et al., 2013). The method of medical image processing from initiation to interpretation is shown in Figure 1.

However, X-rays have been widely used in diffraction testing, also known as X-ray diffraction, in the oil and gas industry (Gharieb, 2022; Macallister et al., 1993). A specimen is placed in the path of a concentrated X-ray beam and then examined. Depending on the crystal structure of the specimen, the resulting rays will bend in a variety of ways. Since different structures have different diffraction patterns, a pattern database has been created to make it easy and quick to identify structures (Arns et al., 2005; Withjack et al., 2007). Due to the fact that X-ray diffraction (non-imaging technique) tests are non-destructive and are typically performed at room temperature, the sample retains its original characteristics after the tests are completed. X-ray diffraction testing can provide more information such as the number of crystals, strain, and crystal defects within the sample (Tang et al., 2013; Zhang et al., 2014). However, it is worth mentioning that X-ray diffraction is not included in the discussion as it is not an imaging technique for rock characterization.

3.2 CT scan imaging

Sir Godfrey Hounsfield, who invented the first commercial CT scanner, undoubtedly had no clue how widely this device would be utilized in the industry. CT scanning machines used in medicine and oil exploration share a number of remarkable similarities. The X-ray beams from the scanner are aimed at the specimen from all angles, creating cross-sectional tomography images (slices) while the scanner rotates. A computer is then able to interpret these signals (Simjoo et al., 2013; Withers et al., 2021). The CT scanner will perform this process many times to obtain multiple slices, which will then be converted into a 3D image that a doctor can examine and interpret. By using a motorized X-ray source as opposed to the fixed X-ray sources present in conventional X-ray machines, the scanner is able to produce more detailed topographic slices (Orlov et al., 2015; Simjoo et al., 2009). The medical community has expressed concerns over the increased radiation exposure from CT scans compared to standard X-rays, but these concerns do not apply to nonliving specimens used in the oil and gas industry (Gharieb, 2022; Siddiqui & Khamees, 2004). There have been significant advancements in CT scanners in recent years, with the most recent models being able to produce images with a higher resolution and, therefore, more technically useful data.

The actual oil displacement mechanism during core flooding oil displacement is difficult to characterize. Seepage characteristics and residual oil distribution in the rock core can be analyzed with CT scanning technology to identify the micro-oil displacement process based on the micro-pore structure (Gao et al., 2009). CT technology can be utilized to carry out quantitative and visual analyses on the micro-pore structure without compromising the exterior shape or internal structure of the core. It is possible to examine the displacement process and identify the oil displacement mechanism (Cheng et al., 2021). Guo et al. (2018) utilized digital core and high-resolution micro-CT scanning technologies to quantify the impact of pore structure and permeability on oil displacement efficiency during the high water cut stage.

Excessive coordination numbers increase the effectiveness of oil displacement, whereas excessive tortuosity and pore throat ratios decrease the effectiveness of oil displacement. A gray slice picture was obtained using the image preprocessing program (reconstruction program) provided by the CT scanning equipment and is illustrated in Figure 2. The microscopic structure of reservoir rocks has a significant impact on oil and gas accumulation and transportation. Studies on the pore structure and water–oil dispersion at a tiny pore scale are crucial (Li et al., 2012). Numerous researchers have undertaken a great deal of research in this field.

3.3 Scanning electron microscope (SEM)

A SEM is a microscope that produces images of sample with focus based on electrons. The electrons interact with atoms to provide information about the surface topography and composition of that sample. Different types of signals are generated by SEM such as secondary electrons (SEs), backscattered electrons (BSEs), characteristics X-rays and cathodoluminescent (CL), and absorbed current (Bansal, 1991; Inkson, 2016). SEs have a low energy of 50 eV, which interrupts their mean free path in solid matter. It collects the image of resolution below 1 nm and is highly localized at the primary electron beam (Chen et al., 2020; Coles et al., 1991). The electrons that are reflected or knocked off the near-surface region of a sample are used to produce an image by scanning the beam in a raster-like pattern using a specific set of coils that are used in SEM (Bisweswar et al., 2020). Figure 3 depicts the components of SEM.

In BSE, the electrons are reflected from the sample by elastic scattering because of much higher energies. The SEM uses electron instead of light to form an image. A beam of electrons is produced by heating a metallic filament. It travels to electromagnetic lenses, and detectors collect SEs and produce an image (Mohammed & Abdullah, 2018; Ma & Chen, 2014). It yields characteristic information such as topography, morphology, composition, and crystallography. Components of a SEM instrument are an electron gun (Filament), condenser lenses, an objective aperture, detectors, and a vacuum chamber (Akhtar et al., 2018). SEMs have a variety of applications in industrial and technological fields such as semiconductor inspection, production line of miniscule products, and in the development of research tools (Brook, 1999; Mady et al., 2020).

4 APPLICATION OF CT SCAN IN OIL AND GAS INDUSTRIES

The CT scan method uses a scanner in which a sample is placed on a horizontal stage, and this plate rotates 360°, taking X-ray images at regular time intervals. This technique yields many 2D images of the sample from a variety of perspectives. After that, a piece of software transforms the 2D images into a 3D model (Du et al., 2007; Rezaeizadeh et al., 2021). Image samples range in size from several centimeters down to the micrometer range. Other methods are used to analyze petroleum industry core samples. Some core integrity damage is always inevitable with the use of any method. There is a way to mitigate the inevitable core damage due to X-ray analysis in CT scans (Gong et al., 2020; Zhang, Jing, et al., 2019; Zhang, Zhou et al., 2019). By penetrating the material with an X-ray, an image of the core can be obtained, from which data can be extracted by image analysis. The extracted data include internal features such as grain size and shape, grain distribution, and heterogeneity (such as fracture and lamination) (Li et al., 2018). CT scan also yield information about time-varying 3D fluid saturation.

Porosity and permeability are the main petrophysical properties of rocks. Evaluation of these properties helps to identify the economic value of reservoirs. To determine the porosity through CT, the core is saturated with two different fluids and CT is performed successively. X-ray attenuation is measured in both the cases and the difference of both the images yields the porosity distribution (Vega et al., 2014; Yerramilli et al., 2015). If a reservoir has saturation of two fluids, then a single energy source performs the CT, while in the case of saturation of three fluids, a dual energy source is required to performed the CT (Fernø et al., 2015). A dual energy source was also used to determine the sample's bulk density. Core flooding can also be monitored by a CT scan that uses various injection techniques, such as water, gas, surfactant, polymer, and foam. Moreover, researchers can monitor oil production to understand phenomena like gravity segregation, channeling, and fingering (Wang et al., 2020).

High magnification of CT scan provides the pore network data, which can be used for fluid–rock interaction interpretation. Traditional CT scanners are stationary, but based on recent developments, permission is provided for the use of portable CT scanners for field measurements of rock properties and geological interpretation in drilling sites (Mady et al., 2020; Wanniarachchi et al., 2018). A detailed discussion of these features is presented in the following section.

4.1 Application of CT scan in pore modeling

Micro-CT scan is used for scanning reservoir rocks at very high resolution and microfocus CT is used to obtain 3D microstructures present in reservoir rocks. It also helps to determine or characterize the flow of multiphase fluids in porous media (Jiang et al., 2017; Rezaeizadeh et al., 2021). The full use of CT scan requires remote processing of digital image information to compute petrophysical properties and fluid saturations from the basic attenuation data. The medical CT scan does not have the ability to show the pores present in the reservoir rocks (Honarpour et al., 1985; Tovar et al., 2014). For example, a fine fracture or tiny hole within the core plug would show as a blurred line or lump, respectively. Micro-CT scans are particularly suitable for extraction of important data on little cuttings of little supply rocks that require determination close to 1 μm for clear visualization and identification between pore throats, pore bodies, and shake grains (Blunt et al., 2013; Gong et al., 2019a). The pores formed at the surface are clearly interconnected with those inside the sample, as shown in Figure 4.

Another application that also uses μCT scanning is pore network modeling. Pore network modeling is an advanced reservoir engineering tool that helps determine the petrophysical properties of a reservoir from drill cuttings (Ali et al., 2021; Saraf & Bera, 2021). An appropriate size of cutting is selected and then a μCT scan is conducted at a resolution of a few microns. The pore network model of the rock is extracted by using advanced image analysis techniques (Andrianov et al., 2012; Bai et al., 2013). Characterization of the microstructure of scaffolds is essential due to its crucial role in the scaffolds' performance. Micro-CT is an excellent instrument for characterizing the microstructure of scaffolds (Cengiz et al., 2018). The network model preserves the vital pore space properties obtained from the rock samples. Surface area and volume are determined from the porosity and pore size of a material (Figure 5) (Cengiz et al., 2018; Yang et al., 2001). Some of the most important petrophysical properties that can be predicted with flow simulations are porosity, permeability, capillary pressure, and resistivity. Understanding the 3D structure and mechanical properties of a material dependent on pore analysis is essential in different fields of application (Figure 6) (BRUKER AXS, 2021; du Plessis et al., 2020). Bone porosity is a parameter of the overall mechanical strength of bone in the medical field, while pore network connectivity is of interest to petroleum engineers for the transportation of liquids in oil and gas applications (BRUKER AXS, 2021; Gong et al., 2019b).

There are two methods to extract the pore network (Bansal, 1991; Blunt et al., 2013):

• 1.

  The medial axis algorithm: This method represents the shape of objects by identifying the topological skeleton, a set of curves running through the center of the specimen. This method characterizes the network by determining the medial axis within the pore space domain (Torrealba et al., 2016). In addition, the medial axis at the outermost spheres consists of a curved surface, as shown in Figure 7.

• 2.

  The maximal ball algorithm: This method involves growing spheres within the pore networks, assigning master spheres to the pore bodies and slave spheres to the pore throats. Once the network is extracted, petrophysical properties such as permeability, capillary pressure curves, and relative permeability are predicted applying a fluid flow simulation code applied to the network (Salimidelshad et al., 2019). The spectrum of radii from the distance field has a greater density than the center of the maximum balls algorithm, and grid line sampling results in a sub-voxel resolution. In this regard, the method for medial maximum balls is typically more precise than the approach for central maximum balls. Maximal Balls generated by the method for medial maximal balls are depicted in Figure 8 (Arand, 2017).

3D pore networks can be obtained through direct imaging methods like micro-CT scanning or synchrotron imaging, or through process-based methods such as diagenesis and simulation-based models, and thin section or SEM-based models. Micro-CT scanning is the most relevant as it enables direct observation and analysis of rock samples (Blunt et al., 2013; Wang et al., 2015).

4.3 Determination of Petrophysical properties

Micro-CT images of the rock fragments are essential in the process of determination of the petrophysical properties of rock. Therefore, acquisition of μCT images is considered as the first step. These raw images from μCT scanners may contain various image artifacts. After removing the artifacts, different petrophysical properties can be extracted from the images (Bai et al., 2013; Saputra et al., 2019).

4.3.1 Porosity

Porosity calculation using micro-CT involves working directly with the grayscale images and using “Segmentation”; it is different from conventional (medial) CT, where porosity is calculated using the average bulk density or image-subtracted CT number data (Andrianov et al., 2012; Farajzadeh et al., 2010). In the micro-CT method, rocks and non-rock materials are differentiated using a “threshold value” for the pixel of the image. Threshold values determine whether it is a rock or non-rock sample. In the standard CT porosity calculation, the core sample should be 100% saturated with a contrast agent such as xenon gas or brine. For cleaning purposes, sometimes, special equipment is needed. Based on the matrix CT number, the dual-energy scan porosity calculation method is proposed at some energy level. In this technique, cleaning and saturation of samples with a contrast agent are not necessary (Akin & Kovscek, 2003; Siddiqui & Khamees, 2004).

The distribution of porosity and raw CT images help to characterize the nature of porosity, that is, whether it is homogeneous or heterogeneous in nature. Figure 9 presents a set of raw CT images of a carbonate core at 1 cm spacing. Dark and white shades correspond to high- and low-density regions of the core, respectively (Akin & Kovscek, 2003; Song et al., 2019).

4.3.2 Permeability

Absolute permeability is commonly determined from the image by solving Stokes' law in a distinct domain acquired readily from the digitized image (Takahashi & Kovscek, 2010). The Lattice Boltzmann method (LBM) is commonly used with the continuity equation to solve Stokes' law (Bandara et al., 2021; Zhang et al., 2020). Another method like finite differences is also used to obtain the solution of the law. The relative permeability can be calculated using CT images. Auzerais et al. used a simulation method based on a 3D immiscible lattice–gas model. Auzerais et al. considered a two-fluid model where one fluid wets the solid boundaries, and the viscosity and density of both fluids are equal (Auzerais et al., 1996; Martys & Chen, 1996). Figure 10 shows a 3D representation of the segmented honeycomb web fabric, which is used for interpretation. The figure also includes the 2-point correlation function of the porous space (air material phase) for the sample. Additionally, it presents a comparison plot of the predicted versus computational fluid dynamics permeability. This comparison is made using the 2-point correlation and lineal direction function characteristic dimensions, with the analysis conducted by a cross-validated regression analysis that left one data point out (Nickerson et al., 2019).

4.3.3 Capillary pressure

To calculate the capillary pressure from micro-CT images, it is commonly assumed that the rock shows water-wet behavior, wherein water (the wetting phase) occupies the corners of pores, while oil (the non-wetting phase) is concentrated in the central regions (Hassanzadeh et al., 2003; Mahmoud, 2014). This assumption forms the basis for deriving the fluid phase distribution using a simple percolation algorithm. This algorithm computes fluid saturations by analyzing the volumes occupied by each fluid phase at specific capillary pressures, rather than directly calculating capillary pressure data at predetermined saturation levels. Consequently, capillary pressure curves are constructed, providing insights into the fluid behavior within the porous medium. However, it is important to note that this approach does not incorporate hysteresis effects, which can significantly influence fluid displacement processes in porous media (Hassanizadeh & Gray, 1993; Ma & Chen, 2014). Hysteresis effects arise due to the complex interactions between fluids and the porous medium, leading to differences in fluid behavior during imbibition and drainage processes. Despite this limitation, the use of micro-CT imaging and percolation algorithms offers valuable information about fluid distribution and capillary pressure characteristics, aiding in the understanding of fluid flow phenomena in porous media and enabling development of enhanced oil recovery strategies.

4.3.4 Resistivity index

It is a dynamic property that depends not only on rock properties but also on the properties of fluid flowing through it. The resistivity index is calculated from the measurement of conductivity (Hosseini-Nasab & Zitha, 2017; Walters and Wong, 1999). The resistivity index is the ration of resistivity values at various saturations with resistivity at 100% water saturation. The diffusivity equation is solved using the random walk algorithm to calculate resistivity (Coles et al., 1991; Withjack et al., 2007).

4.5 CT scan procedure for core flooding

A schematic depiction of the experimental setup utilized to conduct core flooding investigations for CO2-enhanced oil recovery, for short and long cores, is presented in Figure 11. Each system primarily included a core holder, a double-effect piston displacement pump, a gas mass flow controller, a back-pressure regulator, and a fraction collector (Saraf & Bera, 2021; Simjoo et al., 2013). Moreover, CT scans were also performed using a third-generation SOMATOM Volume Zoom Quad Slice scanner. CT scans were used to visualize the long-core studies. The CT scanner's X-ray tube required 140 kilovolts (kV) and 250 milli-amperes (mA) to function properly. Matlab was used to analyze the pictures using custom-written numerical programs, and Avizo was used to view the CT scans in 3Ds (Visualization Sciences Group) (Saraf & Bera, 2021; Simjoo et al., 2013).

4.5.1 CT scan imaging technique

The Siemens SOMATOM CT scanner takes the image of dry and fluid saturated cores in high resolution at 140 kV and 460 mA · s (Ireland, 1986; McCollough & Morin, 1994). Each scan produces an image of 256 × 256 volume elements (voxels) that represent a volume of 0.5 mm × 0.5 mm by either 4 or 8 mm, depending on the slice that was selected for the scan. The images are stored temporarily in the CT scan hard disk and are later transferred to another computer for further examination (Ireland, 1986; Ishutov et al., 2018; McCollough and Morin, 1994).

While measuring the porosity and fluid saturation, it is important to keep in mind that the scan is conducted on the same area of the core under the same scanning conditions in which the computer-controlled positioning network is also constructed (Alsuwaidi et al., 2021; Hassan et al., 2020). As X-ray is used for CT scan, it operates by rotating around the object being scanned, so there is a chance that two images may be displaced. To avoid this error while imaging, it is important to capture the odd-numbered images because X-ray rotates in one direction, while even-numbered images rotate in other directions (Martinez Antunez, 2016; Vega et al., 2014).

After obtaining pictures from every angle, all files are merged into one file to obtain a 3D image to further analyze the porosity and fluid distribution. The porosity–saturation distribution calculation requires the CT value of various fluids (Zhang et al., 2019). The scale is calibrated in such a way that the CT value for air is found to be 500. Then, the CT value for other fluids is measured by filling the void space with that fluid. Changes in the CT value are affected by the position of the void; for example, the CT value of a void located at the edge of the sample will be different from that of a pore located in the middle of the sample (Fu et al., 2015; Sandoval-Martínez and Muñoz-Navarro, 2019). Micro-voids are smaller than meso-voids in transverse cross-section, while they are usually larger in planar view; the calculation using the graphical method is shown in Figure 12 (Chen et al., 2015; Park & Woo, 2011). The figure illustrates the competition between the viscous flow and the capillary flow; inclined arrows show the transverse impregnation of the two; and micrographs illustrate micro- and meso-voids within and between the two (Chen et al., 2015).

Image analysis extracts quantitative information from an image. Reservoir rock images are captured by thin-section optical microscopy through a 2D digital version. They are readily accessible, relatively affordable (e.g., thin slices taken with an optical microscope), and can include important information regarding the pore microstructure (Blunt et al., 2013). Image analysis of micro-CT samples can therefore provide valid information regarding the geometric complexity of reservoir pore networks (Saraf et al., 2019). This section focuses on image generation, acquisition, and processing. Figure 13 depicts the evolution of the considered information (pictures) and the processes involved, representing the best way of presenting the content.

The saturation distribution images are generated by subtracting the CT scan of a dry core from the CT scan of a saturated core (Simjoo, Dong, et al., 2012; Simjoo, Nguyen, et al., 2012; Zhang et al., 2017). In this case, the scale is changed from CT density to a fractional value. In the single-fluid system, the fractional value represents the open pore space. In two-fluid systems, fractional values represent the relative amount of oil and brine (Simjoo, Dong, et al., 2012; Simjoo, Nguyen, et al., 2012; Zhang et al., 2017). The researchers introduced an image subtraction method to extract the density change from the 2D CT image data (Sato & Ikeda, 2015; Suekane et al., 2006). However, the amount of data becomes massive in the 3D image data; thus, the histogram of grayscale in the 3D image data is utilized (Sato & Arkin, 2021; Sato & Ikeda, 2014). The fundamental idea of this method is illustrated in Figure 14.

4.5.2 Core flooding steps

The chemical used in core flooding is known as brine, as the water phase, whose composition includes NaCl, CaCl2, and MgCl2 (Asrilhant et al., 2007; Yerramilli et al., 2013). In the core flooding experiment, initially, brine with the same composition is used to saturate the core. However, brine composition can be changed as required. A case study can be considered to understand the process. In the core flooding experiment, HEPLER oil and North Burbank Unit (NBU) oil were used (AlYousef et al., 2017; Barri et al., 2016). Due to the differences in oil type, their properties are also different. Therefore, while using CT technology to visualize the oil saturation, 20% iododecane is added to solution. As a result, the viscosity of the HELPER oil was affected more than that of NBU oil (AlYousef et al., 2017; Llave & Gall, 1995).

Second, the core flood apparatus was developed and assembled in such a way as to facilitate completion of a wide scope of exploration related to core flooding in an independent coordinated framework. The following steps are considered for the core flooding experiment (Saraf & Bera, 2021; Steel et al., 2016):

• 1.

  In the core flooding experiment, first of all, a dry scan is conducted. This scan is very important as it is used in all subsequent calculations.

• 2.

  The core is evacuated and saturated with CO2 and then saturated with degassed brine, and the composition depends on the condition in which the experiment is being performed.

• 3.

  This brine-saturated core is scanned. In this step, it is recommended to perform a tracer test to define and confirm the presence of heterogeneity in the core.

• 4.

  Micro-CT monitored core flood is pre-evaluated with tagged brine (10% NaI) and then the core is flooded with tagged oil (20% iododecane) for residual water saturation and then flooded with brine for residual oil saturation.

• 5.

  A CT scan is conducted after each and every step.

• 6.

  After that, finally, a surfactant and a polymer slug are flooded into the core.

• 7.

  A CT scan is conducted after the injection of the surfactant and polymer slug with a pore volume of at least one.

4.5.3 Results of the core flood experiment

The objective of using CT scan technology in traditional core flood experiments is to gain a detailed understanding of the fluid movement and fluid saturation distribution at various stages of flooding (Javidi et al., 2019; Zhang et al., 2016). Results of application of a CT scan in a core flood experiment suggest that it can contribute significant information to enable interpretation and evaluation of polymer and surfactant EOR processes (Alfazazi et al., 2021; Gall, 1992; Liu et al., 2022).

• 1.

  Comparison of oil distribution of two good chemical systems

  A core flooding experiment was conducted on four core samples labeled as CT CF-1, CT CF-2, CT CF-3, and CT CF-4, using HELPER and NBU oil (Alfazazi et al., 2021; Gall, 1992). In both cases, the surfactant slug predominantly penetrated areas of higher porosity. Higher oil saturation was observed ahead of the surfactant slug for both oils, with CT CF-1 showing greater visibility (Llave et al., 1992; Tomutsa et al., 1993). This discrepancy may be attributed to differences in experimental procedures or a slightly higher amount of iododecane in CT CF-1. Following oil injection, saturation levels were uniformly low throughout the core in both cases. Only a small amount of oil remained near the core inlet and outlet, with the oil near the outlet showing a more heterogeneous distribution compared to that near the inlet (Gall, 1992; Llave et al., 1992). This suggests improved performance of the chemical flood, which has the potential to channel through certain layers.

• 2.

  Comparison of oil distribution for different concentrations of mobility control polymer

  In the core-flood experiment on CT CF-3, a lower concentration of mobility control polymer was used, similar to CT CF-1. Comparative analysis revealed that the average oil saturation distribution after surfactant injection in CT CF-3 mirrored that of CT CF-1 (Llave & Gall, 1995; Llave et al., 1992). However, the use of a less effective mobility control agent in CT CF-3 failed to displace additional oil from the core, with oil saturations in this area remaining unchanged from post-water flooding levels (Llave et al., 1992; Tomutsa et al., 1993). The impact of the lower concentration mobility control polymer is evident in the reduced oil production following polymer injection. Both tests demonstrated nonuniform surfactant injection, with at least one area in each core showing higher permeability to injected fluids (Boneau & Clampitt, 1977; Withjack, 1988). Significant fluid channeling occurred with the lower concentration polymer, leading to decreased overall oil production. Thus, CT images indicate that oil bypassing by injected chemicals is a major consequence of inadequate mobility control design (Gall, 1992; Tang et al., 2006).

• 3.

  Comparison of oil saturation distribution for different surfactant formulations

  The designed surfactant showed lower effectiveness in HELPER oil compared to NBU oil during core flooding. CT CF-2 was conducted to visualize the oil saturation distribution with this chemical formulation (Boneau & Clampitt, 1977; du Plessis et al., 2020). Examination of the outlet core area revealed varying oil saturation levels, ranging from 15% to 30% in some areas and from 45% to 55% in others (Gall, 1992; Skinner et al., 2015; Tovar et al., 2015). This suggests that water flooding was ineffective and the chemical failed to alter flow properties. A comparison between CT CF-2 and CT CF-4 can enable assessment of the relative effectiveness of two different surfactant systems used with NBU oil (Gall, 1992; Skinner et al., 2015). The mixed surfactant system used in CT CF-2 resulted in higher oil saturation near the core inlet compared to the petroleum sulfonate system used in CT CF-4 (Gall, 1992; Tovar et al., 2015). This difference may be attributed to the lower interfacial tension (IFT) value generated between oil and water in the petroleum sulfonate system compared to the mixed sulfonate surfactant system. The oil saturation distribution provided by CT scans at the core entrance offers insights into the relative efficiency of chemicals used in the EOR process (Tovar et al., 2014).

5 RECENT APPROACHES OF CT SCAN RESEARCH FOR THE OIL AND GAS INDUSTRY

Several issues have arisen recently that hinder oil and gas exploration and production. The use of various forms of CT scan for non-destructive testing (X-ray, magnetic resonance, and others) is currently receiving a lot of attention in oil and gas research (Dewanckele et al., 2020; Orlov et al., 2015). The scope of this kind of research, when combined with other methodologies, is limited only by the scientists’ credentials, experience, and creativity. As an alternative to destructive methods of monitoring and evaluating test results from various technologies and simulation processes carried out in the bottom-hole zones of oil- or gas-saturated reservoirs, the integrated use of 3D X-ray CT of the core sample is a good option (Orlov et al., 2015; Sarker & Siddiqui, 2009). Advanced CT scan applications in the petroleum industry are shown in Table 1. These include the determination of porosity, pore size distribution, flooding direction, and storage capacity.

5.1 Industrial CT scanner

Medical-grade CT scanners were invented for the benefit of living specimens. Therefore, the X-ray sources used have less energy, resulting in low penetration power (Gilliland & Coles, 1990; Haggerty et al., 2016). The expected resolution is 250–400 microns. To overcome difficulties in terms of resolution, use of high-resolution industrial CT scanners is the new trend in the oil and gas industry. Industrial CT scanners enable imaging of the core at 10 to 100 times better resolution than medical-grade CT scanners (Hajiabadi et al., 2020; Liu et al., 2016). The highest power setting can be helpful to look inside materials underneath a thick density material. In addition, one can observe extremely small details at lower settings too. Figure 15 shows a micro-industrial CT scanner. This scanner is four to five times more powerful than the usual medical CT scanners and can operate in two power modes. Its other benefits include the ability to handle multiple core samples together (Gilliland and Coles, 1990; Gotto, 2022).

The scanner works by capturing images while rotating the sample of the X-ray pattern formed from each angle. Using the 2D images, a 3D model is created using reconstruction software (Gotto, 2022; Grachev, 2012). This can help in obtaining proper oil recovery data, which can ultimately help in understanding oil recovery mechanisms and defining a reservoir model appropriately.

5.2 Spectral CT imaging

Spectral CT is a very recent development in this area known as a “dual-source” or “energy” CT scan. The main advantage of this technology is that it can break down X-ray photos by chemical elements, based on viewing of one part of the sample at two different kV energies with a dual-source CT scanner (Hu et al., 2013; Song et al., 2010). This technology can enable efficient observation and removal of chemical compounds only using their atomic number. Figure 16 shows the principle of spectral CT acquisition. The spectral CT problem consists of recovering the material density ρ from the energy-resolved projection images ( s), as shown in Equation ( 2). This can also create contrast and non-contrast images from a single scan (Ducros et al., 2017; Silva et al., 2011).

    (2)

where denotes the (nonlinear) forward model that takes into account the acquisition geometry and the detector response function. This area still remains underexplored in the field of oil and gas, but it has been proven to be very useful theoretically based on advanced research.

5.3 Nano-CT scan

Nano-CT scan is a kind of non-destructive testing method based on micro-CT. By using specific detectors and analysis techniques, higher spatial resolution can be achieved. Nano-CT images can be produced from samples with very less damage to the sample (Li et al., 2013; Yonebayashi et al., 2018). As many samples can be scanned, the preparation time is decreased, which ultimately results in cost savings and also the samples can be reused. Nano-CT has been used for detailed analysis of the microscopic pore structure of tight sandstone formations (Gao and Li, 2016; Yan et al., 2020). The Nano-CT equipment achieves its resolution using a laboratory X-ray source (rotating copper anode) that emits X-ray with a photon energy level of 8 kV. As the schematic below shows (Figure 17), the X-ray beam passes through a mono-capillary condenser lens that uses grazing incidence reflection of X-rays to efficiently focus on the sample. This efficient condenser is key to using a laboratory X-ray source rather than a synchrotron beam (Gao and Li, 2016; X-ray Computed Tomography Facility, 2020). Nano-CT has also enabled identification of fossil fuel reservoirs based on pores, their characteristics, and interconnections. This method is the best in this field since it enables perfect analysis of the 3D pore throat distribution and interconnectivity, even better than a field emission scanning electron microscope (FE-SEM). Nano-CT, unsurprisingly, has revealed a more porous structure than micro-CT, as shown in Figure 18 (Haugen et al., 2020).

6 ADVANTAGES OF THE CT SCAN IMAGING TECHNIQUE

Most of the other imaging techniques convert a 3D object into a 2D image, as structures lying on top of each other get projected onto a single image (Hassan & Al-Hashim, 2017; Khalil et al., 2019). The advantage of a CT scan is the improvement in image contrast in the picture using a 2D image to show an almost 2D section of the object without the effect of overlapping structures. The CT scan provides us with a cross-sectional view of the object rather than an X-ray shadow of a beam passing through it (Kwong et al., 2003). The CT scan generates a trans-axial image oriented in the anatomic plane of the transverse dimension of the anatomy. The CT image is produced by the process of reconstruction: digitally combining information from X-ray projections through the object from many different angles to produce a cross-sectional image (Almetwally & Jabbari, 2020). Because the image is digital, it is made up of a group of pixels.

It should also be noted that only CT scan can be relied upon for the above-listed purposes. The ability of CT scan to analyze the object without destroying it reduces the analysis time significantly. CT images allow us to identify the shape, size, density, and texture, which can be used to analyze any object in detail (Cnudde and Boone, 2013; Deng et al., 2018). It can also help with the detection of any abnormality in the object. Figure 19 shows the CT workflow. Another advantage is that it can rapidly acquire images and can take images of a small portion or the full object at the same time, which cannot be done with any other imaging techniques (Szczykutowicz, 2021).

7 CHALLENGES OF THE CT SCAN IMAGING TECHNIQUE

There are several challenges associated with use of CT scan techniques. It is very important to handle the instrument for safety purpose. The equipment cost for a CT scan of a rock sample is expensive in addition to its large space with magnetic free, electric free, and vibration free conditions required (Belyadi et al., 2019; Bera & Shah, 2021; Ma & Holditch, 2015). The operator requires special training for rock sample operation and preparation. To capture the image of the particular zone of interest, experience in the use of CT scan equipment is a must. The modeling and data generation for characterization of the reservoir rock sample are main components that require utilization of the extensive capabilities of the laboratory (Barree et al., 2014; Wijaya & Sheng, 2020). Since CT scan techniques are expensive, it should be considered whether a project can afford the cost involved in sample characterization using modern imaging techniques. The proper resolution of the images should be set at a particular range so that clear images with the targeted zone can be captured. X-ray CT scans, while invaluable in various fields, often have several common issues that can hinder their effectiveness. First, the resolution can be too low, which impedes the ability to discern fine details. Additionally, limitations in scanning the entire sample can result in incomplete data. Sometimes, the CT images are either too dark or too bright, making it difficult to interpret the results accurately. A lack of density contrast can further obscure important distinctions within the sample. The time required for scans can also be excessively long, causing delays. Moreover, the resulting files are often too large, posing storage and handling challenges. Finally, the analysis of these scans can be overly complex, requiring specialized knowledge and tools. To avoid these problems, it is crucial to use advanced, high-resolution scanners, ensure proper calibration, optimize scanning protocols, utilize efficient data storage solutions, and use streamlined analysis software.

8 INCORPORATION OF DIGITAL ROCK PHYSICS (DRP) INTO CT SCAN STUDY

DRP combines microscopic images with advanced numerical stimulations of effective material properties. Mineral matrix and pore structure are digitalized in DRP, which is a key concept (Andrä et al., 2013). DRP uses a high-resolution 3D imaging technique to place mineral phases and pore structure in the digital object for numerical stimulation to determine the permeability, conductivity, and elastic modulus. This technique is a game changer for computation of electrical and transport properties as high-resolution 3D imaging techniques are available (Elwegaa & Emadi, 2019; Hu et al., 2020; Li et al., 2017). DRP aims to provide deeper understanding of pore geometries and petrophysical properties on the basis of the “principle of image and compute.” The DRP workflow process mainly consists of five steps for processing images, as shown in Figure 20 (Al-Marzouqi, 2018).

A modern method that is mainly used to acquire images of pore geometries is micro-scale X-ray CT. A traditional CT scan machine generates low-resolution rock images that cannot be used in DRP. An alternative method for acquisition of the 3D microstructure of rock is stochastic methods and is sometimes referred to as 2D- to -3D reconstruction (Keehm, 2003; Liang et al., 2000). The main concept involves estimation of statistical properties of a rock sample from a 2D image such as by scanning electron microscopy. Since it produces a large number of digital samples, each of which has only very slight variations in its micro-structural properties (Sadeghnejad et al., 2021), this method suffers from a serious flaw in that it cannot capture the complexity of the pore structure of rocks. When performing a micro-CT scan, a diameter of 1 mm is left free. For CT scan, X-ray images of the target object are taken from different angles and then put together using tomography reconstruction to generate a 3D image (Wildenschild et al., 2002).

Hence, segmentation is one of the main aspects in DRP. The purpose of this step is the removal of artifacts of imaging, such as concentric shadows in CT images, and to delineate pores and channels. Use of a simple algorithm threshold can be avoided because it results in inaccurate segmentation results for complex behavior of the rock (Buades et al., 2005; Peng et al., 2011). Different types of algorithms used for segmentation and their quality are assessed by comparing the physical rock properties such as porosity and permeability obtained through laboratory experiments. Otsu's method and watershed algorithms are traditional algorithms that have their own limitations. Modern algorithms such as 3D hierarchical segmentation algorithms can be used to obtain better results than traditional algorithms (Deepa & Devi, 2011; Peng et al., 2011). Numerical simulation is mainly used to compute the porosity, permeability, elastic modulus, and electrical conductivity for segmented volumes. The LBM is the most widely used method for computing permeability in digital rock physics. Shear and bulk modulus are calculated using the well-established finite element method (khalil et al., 2020, 2021; Tiu and Advincula, 2015).

9 FUTURE PROSPECTS OF CT SCAN STUDY IN OIL AND GAS FIELDS

CT scan is widely used in laboratory research to gain a good understanding of different processes like pipe flow, separators, mixers, and so forth. Furthermore, CT scan can enable study of the multiphase flow in porous rock samples (Johansen, 2013). This study can include core analysis, viscous fingering, and fluid mobility, which in turn facilitate reservoir modeling and simulation with the ultimate goal of enhanced oil recovery (EOR). CT scan can also be applied for midstream and downstream purposes like fluidized catalytic cracking, separation of gas and oil, multiphase pipe flow, and so forth. With the advancement in new sensor technologies and compact versatile signal recovery electronics, the limits of what can be measured have decreased and accuracy has increased for CT imaging (Bai et al., 2013; Bartko et al., 1995). The new sophisticated algorithms can be used for in situ on-line measurement systems if proper research is carried out. Figure 21 shows an outline of a typical CT system from a sensor acquiring data through to image reconstruction and providing cross-sectional images (Johansen, 2013). The future prospects of CT scan studies in oil and gas fields are highly promising, driven by advancements in imaging technology and data analysis. CT scanning, which provides detailed cross-sectional images, is increasingly being applied to study rock samples and reservoir characteristics. This non-destructive method allows for precise analysis of pore structures, fluid distributions, and mineral compositions, enhancing the understanding of reservoir properties. As technology evolves, the integration of AI and machine learning with CT scan data is expected to further revolutionize the field, enabling more accurate predictions of reservoir behavior and optimizing extraction processes. Additionally, improved CT scan resolutions and faster processing times will facilitate real-time decision-making, reducing operational risks and costs (Johansen, 2013). Overall, the ongoing innovations in CT scan technology are set to significantly enhance the efficiency and effectiveness of exploration and production activities in the oil and gas industry.

10 CONCLUSIONS

The main purpose of this review was to provide detailed information about the application of CT scan in the petroleum industry and the current status. In this article, the authors have provided the fundamentals of utilizing digital rock samples with CT scanning. The foundational procedures are laid out in considerable detail. Since DRP provides a more accurate analysis of rock properties, advances in this field are likely to have far-reaching effects on how rocks are characterized. While there have been exciting developments in CT scan imaging techniques in recent years, much ground remains to be covered. If the research is carried out right, CT scan imaging techniques could become widely used in the downstream and midstream sectors of the petroleum industry.

CT scan imaging has undoubtedly enhanced the accuracy of analysis across the board in the oil and gas industry. It has progressed rapidly over the past two decades for application in the oil and gas industry. CT scan imaging has its own benefits that set it apart from other imaging techniques, which are not negligible. The reservoir rocks are scanned with a micro-CT scan and a microfocus CT to obtain a very high resolution and to draw conclusions about the 3D microstructures. Micro-CT scans are also used in pore network modeling, which is used to estimate reservoir petrophysical properties from drilled cuttings. Filtering raw images from scanned rock fragments is the task of the network extraction process. Micro-CT scanning, grayscale images, and segmentation are used to determine petrophysical properties like porosity, permeability, capillary pressure, and the resistivity index. Raw CT images and porosity distribution data can be used to determine whether a porous medium is homogeneous or heterogeneous, and the absolute permeability can be calculated using a digital image and Stocks’ law. The micro-CT image method can also be used to determine relative permeability. However, petrophysical properties can be determined using micro-CT scan technology precisely.

The CT scan method can be used in the chemical EOR process using surfactants, polymers, and alkalis in the core flooding process to evaluate the performance of the chemicals. In order to perform a CT scan, chemicals are injected into an oil-saturated core sample. The entire procedure begins with taking of 360° photos, which are then processed to generate a 3D model. The efficiency of the chemical can be evaluated using a 3D model. Researchers in the oil and gas sector are increasingly showing interest in CT scan imaging thanks to developments like industrial CT, nano-CT, and micro-CT. This new technology has the potential to greatly aid professionals in the oil industry in reservoir analysis and modeling. CT scan imaging stands out because of the unique insights that it can provide into material properties, non-destructive testing, failure analysis, and so forth.

The Department of Petroleum Engineering, School of Energy Technology, Pandit Deendayal Energy University, Gandhinagar, Gujarat, India, is gratefully acknowledged by the authors for providing the facility to conduct the research. Individuals are also acknowledged for their indirect contributions to the project work.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.