Introduction

In the enhanced oil recovery (EOR) method, it is widely known that at least 60–70% of crude oil remains trapped as oil drops in pores in the discrete phase after primary and secondary recovery because of capillary forces (Brea et al. 2016). Since the demand for energy is increasing, the petroleum industry is investigating new methods to recover trapped crude oil where nanotechnology has shown promising results. The application of nanotechnology has been successful in some cases, including reservoir characterization (Rahmani et al. 2015) and drilling (Hoelscher et al. 2012).

For the oil and gas industry, different nanotechnology applications have been proposed based on laboratory experiments (Suleimanov et al. 2011; Zhang et al. 2014; Onyekonwu et al. 2010; Cheraghian et al. 2020; Udoh et al. 2021). Most of the reported results in the literature showed the potential of nanoparticles in improving oil recovery. Nevertheless, to the best of our knowledge, few field trials have been reported. The first reported attempt to utilize nanoparticles in a reservoir occurred in 2010, when Saudi Aramco performed a push–pull test using A-Dots (carbon-based fluorescent nanoparticles) in the Arab D formation of the Ghawar field. The results showed a high recovery percentage, up to 86%, suggesting their high stability (Kanj et al. 2011). Another trial was conducted in the same field and confirmed their high stability (Kosynkin and Alaskar 2016). In the Columbia oilfield, aluminum oxide and silica NPs were used for the inhibition and remediation of formation damage. After 8 months of injecting aluminum oxide, the oil rate has increased by 300 bbl/d. Another trial using silica resulted in an oil and gas rate increase of 134 bbl/d and 1 MMSCF/d, respectively (Franco et al. 2017). Also in the Columbia oilfield, an unnamed nanofluid was used to enhancing the mobility ratio of heavy oils, and an immediate increase in the oil production rate was observed along with a reduction of 11% in the basic sediment and water production (Zabala et al. 2016) In Brazil, stabilizing a shale formation was achieved using a water-based drilling fluid containing nanoparticles. The results showed good performance in terms of shale hydration inhibition and wellbore stability. The same fluid was stored and used to drill another section in a different well after approximately 3 months and resulted in a 15% reduction in the well’s cost (Barroso et al. 2018).

Due to their sizes ranging between 1 and 100 nm, the physical and chemical properties of NPs differ from their bulk behaviors (Suryanarayana et al. 1992 and Alsaba et al. 2020), and their unique properties allow them to have multiple impacts on the recovery of oil. NPs such as aluminum, iron, titanium dioxide, and silica were found to act as nano-catalysts which are highly beneficial to catalytic reactions during steam injection into heavy oil reservoirs due to their large surface-to-volume ratio, small size, and varied shapes (Hashemi et al. 2014). These nano-catalysts can be used to conduct upgrading in heavy oil reservoirs, converting bitumen to lighter products (Yoosuk et al. 2008; Almao 2012). These catalytic reactions fall under aqua-thermolysis which also include the breaking of carbon–sulfur bonds within asphaltenes, increasing saturates and aromatics in the heavy oil, which also has an impact on oil recovery (Hyne 1986).

Another path through which NPs affect oil recovery is their adsorption, which is a surface interaction that leads to the transfer of a molecule from a fluid bulk to a solid surface. This interaction mainly takes place between nano particles and rock surfaces. The major forces that can contribute to the adsorption process include electrostatic (Coulombic) interactions, charge transfer interactions, van der Waals interactions, repulsion or steric interactions, and hydrogen bonding (Cheraghian and Hendraningrat 2016; Olayiwol and Dejam 2019; Murgich 2002; Kokal et al. 1995). The main functions of NPs adsorption are to alter the rock wettability, lower the oil–water interfacial tension (IFT), and reduce chemical adsorption on the reservoir rock surfaces (Al-Anssari et al. 2016; Bera and Belhaj 2016; Hendraningrat et al. 2012; Ju et al. 2006; Kazemzadeh et al. 2019; Nowrouzi et al. 2019; Olayiwola and Dejam 2019; Saien et al. 2017; Zaid et al. 2013).

Wettability is a property of a fluid to cover a surface in the presence of other immiscible fluids (Van and Chon 2016). When oil and water are the immiscible fluids in oil reservoirs, the NPs strongly affect the rock surfaces wettability to reduce the oil–water IFT (Ali et al. 2018), which systematically influences the capillary pressure, permeability, and flow behaviors of fluids in the rock pores (Khalil et al. 2017). During their injection into rock pore spaces, NPs have been shown to arrange themselves in the oil–water–rock system as a well-structured wedge film between the surface of the rock and the oil, exerting a disjoining pressure on the film and separating oil from the surfaces of the rock (Azizr et al. 2018; Khalilnezhad et al. 2019; Kondiparty et al. 2011).

One of the most favorable NPs is zinc oxide (ZnO), which has a high surface charge and can function as a surface-active agent to replace surfactants (Soleimani et al. 2016). The small size of the NPs allows penetration of smaller pores to mobilize the capillary-trapped oil (Yahya et al. 2014). The adsorption of NPs on rock surfaces in oil reservoirs can modify the wettability condition from oil wet to water wet (Gurgel et al. 2008; Lianga et al. 2019) with the formation of an interface between the oil and water surfaces. Rezk and Allam (2019) reported an increase of about 8% of oil-recovery efficiency when both ZnO NPs and a surfactant mixture were employed when compared to a surfactant-based oil-recovery process in the case of sandstone.

However, NPs also have their limitations. For example, their release into the aquatic ecosystems through industrial wastewaters can induce pernicious effects on fish and other organisms, increasing concerns of environmental hazards. Several characteristics of ZnO NPs (e.g., size, shape, surface charge and agglomeration state) play a central role in biological effects such as genotoxic, mutagenic, or cytotoxic effects. Further, ZnO NPs may interact with the bacterial surface and/or with the bacterial core, exhibiting different bactericidal mechanisms (Jiang et al. 2009). Furthermore, economic feasibility is the major drawbacks when employing nanoparticles (NPs) in the petroleum industry (Bera and Belhaj 2016). Consequently, it has become necessary to investigate applications of nanotechnology within a laboratory environment, especially before applying NPs in the field.

Low-field nuclear magnetic resonance (LF-NMR) relaxometry techniques were developed in the laboratory to enhance and support comparable NMR logging tools that are currently used downhole. LF-NMR relaxometry has shown that discrimination of water and oil saturation in core and raw material can be easily determined. In such cases, the NMR can detect the total water weight fraction and the total oil weight fraction, the viscosity of the oil, the amount of bound or mobile water and the amount of mobile or bound oil (Mirotchnik et al. 1998; Mirotchnik and Kantzas 1999).

Additionally, LF-NMR has been applied in the crude oil industry because of its high potential to determine fluid and rock properties (Barbosa et al. 2015; Hou et al. 2020; Ok and Mal 2019) using both in situ and ex situ methods. LF-NMR has several advantages such as being non-destructive, fast, reliable, and easy-to-operate (Barbosa et al. 2013; Jiang et al. 2021). LF-NMR provides time-domain relaxation data. In a typical NMR relaxation measurement, the relaxation processes reinstate the equilibrium magnetization after excitation of the spin ensemble (Ridwan et al. 2020). Longitudinal (T1) and transverse (T2) NMR relaxation times explain magnetization vector components that are parallel and perpendicular to the external magnetic field (B0), respectively. The T1 and T2 values of molecules depend on fluctuations of the NMR interactions due to molecular motions. Hence, T1 and T2 measurements have become conventional techniques to explore molecular reorientations both in pure bulk state and in confined geometries (Abragam 1961; Gautam et al. 2017; Vogel 2010). LF-NMR measurements of fluids confined into rocks can be utilized to predict several petrophysical properties including porosity, pore size distribution, and free fluid index (Connolly et al. 2019).

In wettability measurements, nuclear magnetic resonance (NMR) interrogates the character of water molecules which changes based on whether the water molecules are in contact with rock or in the liquid phase. Borysenko et al. (2009) demonstrated that NMR determination of wettability showed a good correspondence with contact angle measurements. Odusina et al. (2011) used NMR to examine shale wettability, and Sulucarnain et al. (2012) studied shale wettability and effective surface relaxivity. NMR offers advantages such as being less expensive and faster than the USBM or Amott methods for single measurements. NMR can also monitor wettability changes, and the results can also be compared with normal geophysical logs that directly interrogate the reservoir in a continuous manner.

The goal of this study is to assess the potential of ZnO NPs in EOR processes with the aid of LF-NMR, to reduce IFT and alter wettability in the confined geometries and nanopores of sandstone rock samples, where oil and water molecules show strong deviations from their bulk behaviors. To achieve this goal, ZnO NPs were first thoroughly characterized by elemental composition analysis and surface area determination. Then, blends of formation water (FW) and crude oil were studied in bulk using various approaches, including IFT tests and water contact angle measurements. Finally, the non-destructive and reliable LF-NMR technique was applied to evaluate sandstone samples saturated with oil and nanofluid.

From a practical viewpoint, this study will be valuable in EOR for developing a new high-precision LF-NMR approach, which is faster and more reliable in measuring or estimating rock wettability through different chemical conditions present in the oil fields. This method will be an alternative to methods of Amott-Harvey, USBM (US Bureau of Mines test), (sessile drop) methods used in the laboratory and reservoir, Anderson (1986), and Abdallah et al. (2007). The novelty of the work lies in the fact that the wettability of the rock surface affects the distribution of fluids within the pore space, and the oil and water distribution can be obtained by comparing the NMR relaxation data at different saturations (Al Harbi et al. 2017). Also, accessible advance integrated petrophysical evaluation for in situ wettability to support the field development and improve the reservoir characterization.

Experimental details

Materials and properties

Core samples

Synthetic Berea Sandstone core plugs were purchased from Kocurek Industries Inc. (Houston, TX, USA). The core samples were cleaned using distillation–extraction Soxhlet apparatus with a 50/50 mixture of toluene/methanol and subsequently dried in a vacuum oven at 80 °C (1CE, Thermo-Fisher Scientific with Hydraulic Thermostat Controller, UK). The porosity and permeability of the core test samples were measured using Helium PHI-220 Porosimeter and KA-210 Gas Permeameter, respectively. Both instruments were supplied by Coretest Systems, Inc., USA. The porosities at pressure and temperature (400 psia, 25 °C) and for permeabilities (250 psia, 25 °C) were measured at a confinement pressure 500 psia as recommended by the manufactures. The core properties are shown in Table 1. The average element analysis using (EDXRF, Epsilon-1 Malvern analytical Ltd UK) are presented in Table 2, and X-ray diffractometer (XRD) analysis, using D8 Advance Bruker GmbH, was performed to reveal the amount of different crystals existing in the Berea Sandstone specimens as shown in Fig. 1. Formation water (FW)

Table 1 Dimensions and petrophysical properties of the core samples
Full size table
Table 2 Elemental analysis of Berea sandstone
Full size table
Fig. 1
figure 1

XRD of the dry Berea sandstone core

Full size image

FW of low salinity and low conductivity was employed in this study. The conductivity, total dissolved solids (TDS), and salinity of the FW (30,000 ppm) were measured using a VWR traceable hand-held meter (Chemicals and Laboratory Scientific Company), while the turbidity was measured using a HACH model 2100P portable turbidimeter (GmbH, Germany). The detailed physicochemical properties of the FW are summarized in Table 3. Besides, the distribution of the solid particles in the FW was determined using dynamic light scattering (DLS) by a Zetasizer Nano ZS-ZEN3600, DLS, USA. The average particle diameter was 1760 nm after filtration with sterile poly-ether sulfone (PES) syringe filters with four layers, followed by a membrane filter, resulting in average particle size of 380 nm.

Table 3 Properties of the FW
Full size table

The particle size distribution of FW averaged at 1250 nm, obtained by DLS (Fig. 2). In addition, the rock heterogeneity was qualitatively observed from the frequency graph of the pore throat diameter (Fig. 3). The fraction of pores (y-axis) was calculated as the volume of injected mercury divided by the pore volume of the core sample. The pore size distribution of Berea sandstone is normal with a relatively narrow peak less than those suggested by Gong et al. (2020). Thus, the FW is considered suitable for the present study because the particle size distribution is less than the pore size distribution. The rock heterogeneity might be also qualitatively observed from the frequency graph of the pore size distribution.

Fig. 2
figure 2

DLS results of FW

Full size image
Fig. 3
figure 3

Pore throat diameter distribution of Berea sandstone

Full size image

Crude oil

Samples of crude oil were collected from a Kuwaiti oilfield; the field produces medium to light crude oil with an API gravity of 28–36°. The samples were stored in specially designed screw-cap bottles under dry conditions in a thermostatic fume hood at 25 °C. The basic sediments and water (BS&W) were determined using the ASTM D4007-11. The density was measured at 25 °C using a precision digital Anton Paar oscillating U-tube densitometer, DMA4500, with a reproducibility of 10–2 kg m3. The dynamic viscosities were also measured as a function of temperature (20, 25, 30, and 40 °C) using an SVM 3000 Stabinger Anton Paar viscometer. SARA Analysis (saturates, aromatics, resins, and asphaltenes) was done using IATROSCAN MK-6s (Mitsubishi Chemical Medience, Japan). The physical properties of the crude oil are tabulated in Table 4.

Table 4 Physical properties of dry crude oil samples
Full size table

Zinc oxide (ZnO) nanoparticle (NP)

ZnO NP was purchased from Skyspring Nanomaterials Inc, USA, without any additional treatment. The properties of ZnO NP shown in Table 5 were investigated experimentally using an automatic absorptiometry surface area analyzer (ASAP-2010, Micrometrics USA).

Table 5 Physicochemical properties of ZnO NPs
Full size table

Methodology

Stability, dispersion, and adsorption of the ZnO NP

Particle stabilization is important for preventing particle agglomeration and formation damage. To represent real operating conditions, known masses of all the NPs of ZnO at three concentrations of (0.05, 0.075, 0.1) wt% were mixed with FW (30,000 ppm) and stirred continuously for 3 h using a digital stirring plate (Thermo Scientific, USA) at 500 rpm, with overnight storage in an oven at 30 °C. A cloudy solution was observed in all samples. To avoid high dispersion of NPs in the solution and reduce or prevent the possibility of particle agglomeration, each prepared solution was subjected to ultrasonication for 60 min using a Hielscher ultrasonic mixer (model UP200s, GmbH), as proposed by Chung et al. (2009) and Graves et al. (2019). Moreover, to evaluate the dispersion stability of nanofluids, the zeta potential (ξ-potential) was measured and calculated using the Helmholtz–Smoluchowski equation (Wilson et al. 2001; Munson et. al. 1998).

In the Brunauer–Emmett–Teller (BET) method of surface area analysis (Brunauer et al. 1938), liquid nitrogen is usually used at partial vacuum conditions to cool surfaces and detect adsorption since the interaction between gaseous and solid phases is generally weak. N2 (− 195 °C) gas used was a 99.999% pure product of Kuwait Oxygen and Acetylene Company KOAC (Kuwait). The specific surface areas (SBET) were calculated from the BET equation in its linear form (Brunauer et al. 1938) with the nitrogen molecule cross-sectional area taken to be 16.2 × 10–20 m2 (linearity region between 0 and 0.35 p/p0). The total pore volume was estimated from a single point on adsorption isotherm at p/p0/0.975 (Badalyan et al. 2003 and Brundle et al. 1992). The pore size distributions were calculated in the standard manner using the Barrett–Joyner–Halenda (BJH) method (Barrett et al. 1951), and the pore size analyses done as stated by Siegbahn et al. (1967) are shown in Table 4.

IFT and contact angle measurements

The main purpose of these measurements is to evaluate the effect of ZnO NP on IFT and contact angle and to subsequently determine the optimum concentration of NP using NMR. The IFT for the oil/FW and oil/nanofluid systems was characterized, using a drop shape analyzer (DSA 100, Kruss, GmbH), based on the pendant drop method (Ayatollahi and Zerafat 2012). The contact angles between the fluids under study and the selected Berea Sandstone were determined, and the wettability was identified according to the criteria (Teklu et al. 2015).

The apparatus was calibrated according to manufacturer recommendations with the standards provided; these standards consist of glass slides with modeled drop shapes, which are accurately calculated. Glass slides were used to calibrate the apparatus using the Young–Laplace method. Shapes with contact angles of 30°, 60°, and 120° each for standard and microscope optics deviated by less than 0.1° from their nominal values. The sessile drop was used for preferential determination of wettability test used in core flooding tests for measuring the contact angle directly. The flooded cores were cut in slices and prepared following the same method proposed by Ayatollahi and Zerafat (2012), and the drop of the nanofluid was introduced at the surface. The DSA-100 apparatus equipped with a high-resolution camera and digital processing software was used to perform contact angle measurements, and the results of measurements were checked for repeatability at least three times for each experiment. Finally, the results were averaged.

Fluid displacement experiments

Fluid displacement experiments were carried out in the core flooding system shown in Fig. 4. The system was mainly used to prepare the core plugs needed to conduct the NMR experiments based on the optimum results originated from IFT and contact angle. Besides the dry core sample (S0), which was used as reference for NMR, the following six samples were prepared: (S1) 100% saturation with FW; three pore volumes of FW were injected in the clean dry rock sample followed by soaking the core in FW at a rate of 0.5 cm−3 min−1 for about 1 h to ensure complete saturation. (S2): 100% saturation with crude oil; three pore volumes of crude oil were injected in the clean dry rock sample; then the core was infused with the crude oil at the rate of 0.5 cm−3 min−1 for about 1 h to ensure complete saturation. (S3–S6): rock restoration and nanofluid injection; the remaining rock samples were restored to represent reservoir saturation profile; initially flooded with FW until reaching 100%; in the next step, oil sample was injected at a rate of 0.5 cm−3 min−1 until no more water was removed; then nanofluid was injected with a specified concentration and required pore volumes.

Fig. 4
figure 4

Flow diagram of core saturation and flooding equipment

Full size image

LF-NMR relaxometry details

The LF-NMR relaxometry data of the rock core samples were obtained on a 2.35 MHz Oxford GeoSpec2 Instrument, UK, with a 43-mm-diameter probe using the software Lithometrix 8.5.0. In flooded sandstone rock cores, the duration of the 90° pulse was 9.5 μs. To obtain T2 data, the Carr–Purcell–Meiboom–Gill (CPMG) pulse sequence was used with a recycle delay time of 1125 ms, while an inversion recovery pulse sequence with a 3000 ms recycle delay time was used for the T1 data. The number of detected points (spin echoes) in the T2-pulse sequence was varied within the range from 32,401 to 4630 depending on the sample, and 12,963 was for oil and 115,741 for water. Three-exponential fitting analysis of T1 relaxation data was performed according to the following equation:

$$y\left( x \right) = \mathop \sum \limits_{i = 1}^{3} A_{i} exp\left( {\frac{x}{{T_{1\left( i \right)} }}} \right)$$
(1)

where x stands for the signal detection time and T1(i) is the longitudinal relaxation time of the i-th component with respective amplitude Ai. A three-exponential fitting analysis of the T2 relaxation data was then performed according to the following equation:

$$y\left( x \right) = \mathop \sum \limits_{i = 1}^{3} A_{i} \exp \left( { - \frac{x}{{T_{2\left( i \right)} }}} \right)$$
(2)

where x stands for the signal detection time and T2(i) is the transverse relaxation time of the i-th component with respective amplitude Ai (Aursand et al. 2008; D’Agostino et al. 2012). The three-component fitting gives the freedom of determination of differing fractions, either heavy or light, of the crude oil in bulk and assigns the confined fluids inside such pores with different porosities.

Considering bulk crude oil, we suggest that the shortest T2 value (T2(1)) is attributed to the heaviest components, asphaltenes and resins, while the longest T2 value (T2(3)) to saturated aliphatic chains, where T2(2) values belong to aromatics; referring to A1, A2, and A3 values, it is possible to suggest the percentages of crude oil fractions. In the confined fluids, three-component fitting is preferred to describe how the two fluids are distributed in the pores with various sizes.

The continuous distributions of T1 were obtained from T1-inversion recovery relaxation using the CONTIN algorithm (Provencher 1982). The permeability and porosity were determined using the following equation for T1 distributions (Alvarado et al. 2003; Kenyon and Kolleeny 1995; Aghda et al. 2018):

$$k = \phi^{4} *T_{1}^{2} \left( {k = \frac{\phi }{C}} \right)^{4} *\left( {\frac{{{\text{FFI}}}}{{{\text{BVI}}}}} \right)^{2}$$
(3)

where k is the permeability (millidarcy: md), T1 is the longitudinal relaxation time (s), ϕ is the fractional porosity (m3 m−3), C is an empirical coefficient (md–0.25), BVI is the bound water volume index, and FFI is the free fluid index.

For processing the T1 distribution data of sandstone cores using Eq. 3, the default value of 10 was employed for C (Coates coefficient: p.u./md1/4) (Trevizan et al. 2015). T1 distributions were utilized in the present study to estimate both the volume of producible fluid (Straley et al. 1991) and to show potential wettability alterations in sandstone rocks with certain porosities because T1 distributions are reflections of porous media saturated with different fluids: low viscous FW versus highly viscous crude oil (Giraldo et al. 2013; Dang et al. 2013). The NMR estimate of a product is most often considered as the free fluid index (Straley et al. 1994).

Results and discussion

Stability, dispersion, and adsorption of ZnO NP

It was concluded that the stability of nanofluid depends on the pH, NP size, NP type (hydrophilic, hydrophobic, and amphiphilic), dispersion fluid, and ultrasonication time. The stability of dispersions is their long-term integrity and ability to remain in their initially formulated state by remaining as close as possible to their initial physical state. Because complex formulations are unstable by nature, the apparent dispersion stability can only be evaluated when the dispersed phase remains suspended, as stated by Tso et al (2010). The particle size distribution of ZnO nanofluid is shown in Fig. 5.

Fig. 5
figure 5

Average particle size distribution of ZnO in FW after sonication

Full size image

The zeta potential is a key indicator of the stability of a colloidal dispersion. The measured ξ-potential values of the ZnO nanofluid of concentration (0.05, 0.075, and 0.1) wt% were − 28.8, − 29.8, and − 29.1, respectively, with an average value of − 29.23, indicating a certain degree of electrostatic repulsion between adjacent similarly charged particles. Colloids with high zeta potentials (negative or positive) are electrically stabilized. A value of ± 25 mV can thus be taken as the arbitrary threshold that differentiates low-charge surfaces from highly charged surfaces (Dukhin and Goetz 2010).

The dispersion of ZnO NP in FW was measured using DLS based on the non-negative least-square algorithm method (Lawson and Hanson 1974). This suggested that using the ultrasonic processor during nanofluid preparation might break down the agglomeration of NPs and improve their dispersion. Figure 6 shows the normal distributions using the algorithm method to reconstruct particle size distribution (PSD) from DLS data at 60 and 1000 min indicating the stability of the dispersion with time.

Fig. 6
figure 6

Non-negative least square method shows the dispersion of ZnO NP in formation water at 60 and 1000 min

Full size image

The ability of ZnO for adsorption built on BET isotherm model was investigated by plotting the amount of gas adsorbed as a function of the relative pressure (Gomez-Serrano et al. 2001; Maa et al. 2019). The adsorption type is (Type II). This is most frequently found when adsorption occurs on nonporous powders or powders with diameters exceeding micropores and the Inflection point occurs near the completion of the first adsorbed monolayer. Figure 7 shows the relationship between relative pressure (P/Po) and ZnO adsorption capacity measured by quantity adsorbed (Q) = 62.79 cm3 g−1 STP. Indeed, the conventional methods such as BET and BJH models cannot distinguish between different pore structure morphologies to account for the effects of microporosity and predict the pore sizes that could be independently determined using XRD and transmission electron microscopy (TEM) with the precision unavailable earlier. Density functional theory (DFT) methods have been first suggested by Lastoskie et al (1993) for calculating the pore size distribution of carbons from nitrogen adsorption data. The main advantages of the DFT methods are related to its rigorous theoretical basis that covers the whole region of micro- and mesopores and provides an opportunity of customization to different adsorbates (nitrogen, argon, and carbon dioxide), materials (silicas and carbons), and pore morphologies (slit-like, cylindrical, and spherical); the hybrid models that include different groups of pores were designed for hierarchical materials. Furthermore, the computational quantum mechanical modeling method, used in materials science, aids the investigation of adsorption structures and mechanisms of water adsorption on a high-index polar surface of ZnO. It provides explanations of not only the water adsorption behaviors of high index polar surfaces of ZnO but also guidance to all the adsorption behaviors of nanomaterial surfaces.

Fig. 7
figure 7

Relation between relative pressure (P/Po) and measured quantity adsorbed (Q) calculated by BET methods for ZnO nanoparticles

Full size image

Effect of ZnO NP on IFT and contact angle

Different nanofluids were prepared by mixing ZnO NP at various concentrations (0.05, 0.075, 0.1) wt% with FW of different salinities. Figure 8 shows the IFT between the crude oil sample and the prepared nanofluids. It was found that as the concentration of NP increased the IFT of crude oil, and that of nanofluid decreased, where the minimum IFT value was obtained when using the nanofluid with the highest ZnO NP (0.1 wt%); these results agree with other reported results (Hendraningrat et al. 2013, 2012).

Fig. 8
figure 8

Measured IFT between crude oil samples and different concentrations of ZnO NPs

Full size image

On the contrary, wettability measurements were obtained before and after surface modifications with different concentrations of NPs (0.05, 0.075, and 0.1 wt% ZnO) in FW (30,000 ppm). The calculated values were obtained by analyzing the complete shape of the oil droplet using a precise video system and analysis software. Figure 9 shows that the surface immersed plates with and without nanoparticles. It could be seen that the value of dynamic contact angles for FW with no NP decrease at range (70.46–56.90°), with the drop value near to 19% of the original value, but for FW with NP at concentration range (0.05, 0.075, and 0.1) %, were (59.26–21.00°), (51.64–15.39°), and (47.92–12.3) respectively, indicating altered wettability in the water-wet condition. A stronger shift in wettability was achieved by increasing the NP concentration in the FW up to 0.075 wt% but small effect reaches the concentration 0.1 wt% (Fig. 9).

Fig. 9
figure 9

Contact angles of rock-FW (30,000 ppm) and NP-oil systems

Full size image

LF-NMR relaxometry

As stated earlier, the LF-NMR experiments were designed based on the results obtained through the IFT and contact angle measurements. Accordingly, 0.1 wt% ZnO nanofluid and 30,000 ppm of FW were used. Table 6 summarizes the details of the LF-NMR experiments conducted.

Table 6 Mass of rock core samples before and after LF-NMR measurements
Full size table

Dynamics of confined fluids

Figure 10 shows a representative comparison of the T1 relaxation curves of bulk crude oil and confined crude oil (S2). As expected, the T1 relaxation time of the confined crude oil (S2) shifted to lower values compared with that of the bulk crude oil. A similar, even stronger deviation in the T1 relaxation curve toward shorter times was observed in the case of FW (S1), as seen from Table 7. The T1 values were determined using Eq. 1, based on three-component exponential growth fitting.

Fig. 10
figure 10

Comparison of T1 relaxation curves of bulk and confined crude oil (S2)

Full size image
Table 7 T1 values of fluids in bulk and confined states
Full size table

Figure 11 compares the T2 relaxation curves of the bulk and confined crude oil (S2). As demonstrated in Table 7, the T2 relaxation values of the confined crude oil (S2) are lower than those of the bulk crude oil. Additionally, compared with that of bulk FW, a stronger deviation in the T2 relaxation curve with a shift toward shorter relaxation times was observed in confined FW (S1). This stronger deviation in FW is explained by the strong interaction between the water molecules and pore walls. The former preferably fill the smallest pores, and the surface relaxivity controls and reduces the T2 value. Thus, the T2 values of confined fluids are even lower than the corresponding T1 values. Similarly, significant reductions in the relaxation times of the crude oil confined in rock samples compared with the bulk relaxation times offer evidence of surface relaxation (Freedman et al. 2003). A possible explanation is that when oil fills the larger pores, the ZnO NPs form an interphase (Myint et al. 2013; Soleimani et al. 2016), forming different types of environments: water in the smallest pores with the shortest relaxation, oil molecules in the larger pores, and mixed wettability, with water molecules wetting the pore walls, ZnO NPs between oil and water, and oil molecules not interacting directly with the pore walls. Table 8 summarizes the T2 results, confirming that the deviations in the dynamic behaviors of the confined fluids from those of the bulk fluids are independent of NP treatment. The T2 values are determined using Eq. 2, according to a three-component exponential decay fitting.

Fig. 11
figure 11

Comparison of T2 relaxation curves of bulk and confined crude oil (S2)

Full size image
Table 8 T2 values of fluids in bulk and confined states
Full size table

The first critical issue concerns the confinement effect on the dynamics of the fluids. Confinement in this case refers to “molecular confinement” where the crude oil and/or water molecules have spatial restrictions and different local chemical environment in the rock sample pores compared to that of bulk. In the confined geometry and environment, different properties such as dynamics of the molecules could differ from those in bulk. Independent of the chemistry, when the fluids (water or crude oil) were confined to sandstone, both T1 and T2 decreased. This shows that the confinement effect influences the wettability of the fluids with the grain walls. The second critical issue is the percentage of fluids in the bulk and confined states. Bulk FW has only a single value for both T1 and T2. However, the SARA (saturates, aromatics, resins, and asphaltenes) components of crude oil are considered with three different T1 and T2 values: longer, medium, and shorter. For the bulk crude oil, the lower T2(3) value of 7.1 ms constitutes 19.4% (A2(3)), which corresponds to the heavy components (asphaltene + resins) of the crude oil. The larger T2(1) value of 160.5 ms constitutes 44.1% (A2(1)) and is attributed to the lighter fractions (saturates) of the crude oil, as is also suggested in the literature (Volkov et al. 2021). The lighter fractions of the crude oil should thus have greater T1 and T2 values. Following the same approach, it is observed that in the confined state, crude oil has similar percentages of the different fractions under the different NP treatments.

T 1/T 2 ratios: affinity of fluids toward pore walls

To determine the affinity of the fluids toward the grain walls, the T1/T2 ratios were evaluated (Table 9) because the interactions between fluids and rock pores are reflected in the T1/T2 ratio. Moreover, being independent of pore geometry, these interactions are mainly influenced by the variations in surface relaxivity (Katika et al. 2017). Thus, the T1/T2 ratio is reliable for measuring the surface and bulk relaxations of fluids inside pores when the relaxations due to diffusion are negligible.

Table 9 T1/T2 values of fluids in bulk and confined states
Full size table

The most important result was observed in treatment with 0.1% ZnO NPs, which yielded a higher water wetting in the larger pores (Tinker 1983). Oil fractions interact weakly with the pore walls; hence, the oil-recovery efficiency is enhanced. For molecules such as simple liquids with fast anisotropic motion, T1 and T2 are equal, and hence, the ratio T1/T2 = 1 (Valori and Nicot 2018). For a molecule with slow dynamics, T1 and T2 differ and keep diverging further as the motion is hindered more. Second, the T1/T2 ratio for the oil may deviate from unity because of intrinsic bulk oil properties such as viscosity. Interpreting this deviation as the result of wettability might result in incorrect conclusions. A more reliable interpretation would be obtained by assessing the deviation from the unity of the water phase, which is expected to always have T1/T2 = 1 for the ideal non-wetting condition (Valori and Nicot 2018). As shown in Table 8, the T1(3)/T2(3) values are similar in only FW-saturated and S4 (1.0 PV) samples. This shows that the treatment in S4 facilitates the pushing of the water molecules to the pore walls. In S3 (0.5 PV), the distribution of the fluids in the pores is mixed wetting. In other words, both FW and crude oil molecules could be found within the same porous regions.

Wettability alteration

In this study, only T1 distributions were used to analyze the wettability alteration potential by determining the porosity and permeability based on Eq. 3. Table 10 shows the NMR porosity, clay-bound water, and effective porosity (free fluid index) determined by analyzing the T1 distributions obtained via inverse Laplace transform (ILT) of the T1 relaxation data. Figure 12 shows the representative T1 distributions of crude oil in the bulk and confined states for Berea sandstone (S2).

Table 10 Parameters determined by processing T1 distribution data
Full size table
Fig. 12
figure 12

Representative T1 distributions of crude oil in bulk and confined states in Berea sandstone (S2)

Full size image

The total NMR porosity was 12.5% based on T1, considering only the oil-flooded sample (S2). This value ensures suitable liquid characteristics before any NP treatment. The smaller pores resemble the clay-bound water porosity, which is approximately 0.20% based on the T1 distribution analysis. The first saturation of the rock cores was with water, then with crude oil only, and finally with oil and nanofluid. The first saturation shows how water molecules behave in the smallest pores, while the second saturation (oil only) exhibits which pores the oil molecules wet. Based on these, we assign the shortest T1, to water wet regions in the third wetting case. We use Eq. 3 to process the T1 distribution data, and this helped obtain clay-bound water vs. free fluid index. Equation 3 does not require the use of T1 cutoff value. In addition, processing T1 distribution of S1 (only water wet sample) yields 0.2% for the smallest pores.

The completely water-wet pore size is best described as approximately 2.3% (0.29%/12.5%) of the total pore network. The effective porosity decreases when both FW and crude oil are used for flooding. Therefore, the effective porosity interphase should be a mixture of the oil + water region (between the clay-bound water and the free-oil region) and the free-oil zone. The effective porosity region is divided into two types: (1) bulk porosity-irreducible and (2) free fluid index. The first region is water dominant and has a transitional mixed zone of oil and water, while the second zone contains free crude oil. Water molecules wet the pore walls and simultaneously form an interphase with crude oil molecules. These observations pertain to two different wetting regimes: pore wall wetting and water–oil interphase wetting. In NP treated samples, the percentage of clay-bound H2O/free fluid index was maximum in S4 with 1.0 PV NP treatment. This treatment should allow maximal crude oil recovery because the inner region of the sandstone with FW + NPs has the highest fraction and the affinity of crude oil molecules to interact with the pore walls of the sandstone was the lowest. Figure 13 demonstrates both T1 and T2 distribution comparison of the samples from S2 to S6. As expected, the time scale for T1 distribution is longer than T2 distribution. There was a shift, in both T1 and T2 distributions, to shorter times when ZnO NPs were utilized. The longer times in the distribution curves arise from larger pores with crude oil. T1 and T2 distribution curves of both S5 (1.5 PV) and S6 (2.0 PV) overlap each other, and longer times in their distributions resemble S2, oil only sample. In S3 (0.5 PV) and S4 (1.0 PV), the distribution curves differ from S2 (oil only), S5, and S6. Among the samples, S4 seem to be in between completely oil wet and completely mixed wetting. Therefore, S4 treatment with 1.0 PV, as also discussed above, will provide better wetting alteration and hence higher crude oil recovery.

Fig. 13
figure 13

T2 (a) and T1 (b) distributions of crude oils in confined states in Berea sandstone with and without utilization of ZnO NPs

Full size image

ZnO NPs demonstrate a hydrophobic/hydrophilic switching ability depending on the treatment (Myint et al. 2013) as well as super-oleophilic properties (Jianlin et al. 2018). These NPs can be useful for establishing the interphase between water and crude oil molecules. Hence, crude oil molecules are segregated and do not wet the pore walls. The possible decrease in the surface energy of the water/ZnO NPs upon confinement in sandstone pores and upon interactions with the pore walls could lead to segregation of the crude oils.

Conclusion

This study was conducted to exhibit the potential of hydrophilic ZnO nanoparticles in enhance oil recovery by improving fluids–rock interaction properties such as wettability alteration. The nanoparticles are active on wettability alteration due to their ability to adsorption on or interact with the surface of the rock and altering the wettability from oil wet toward water wet. Interfacial tension measurements showed that ZnO NPs were able to reduce the FW-oil IFT from 18 to 9.52 mN/m depending on the ZnO NP concentration and FW salinity. The minimum IFT values were obtained at an optimal salinity of 30,000 ppm of FW with 0.1 wt% nanofluid. The effects of various concentrations of ZnO NP on wettability alterations, determined via contact angle measurements, indicated that depending on their concentration, the ZnO NPs were able to alter the wettability by changing the water–oil contact angle from 71° to 13°. When FW or crude oil was confined in the rock, deviations from their corresponding bulk dynamic behaviors were seen. This is attributed to the well-known confinement effect and is important because the NMR relaxation curves presented herein were acquired under confinement, where molecules have restricted motion. The percentage of clay-bound H2O/free fluid index was the highest in S4 with 1.0 PV NP treatment (for treated samples). The S4 treatment thus has a higher potential for wettability alteration under the confined state in porous rocks. A comparison of the T1/T2 ratio, an important parameter quantifying the affinity between the minerals and wetting fluid, also proved that the S4 treatment enables enhanced oil recovery because the T1(3)/T2(3) values are similar in only FW-saturated and S4 (1.0 PV) NP-treated samples. Based on these results, it is possible to alter wettability using ZnO NPs under a confined geometry in rock samples for EOR studies. Furthermore, the LF-NMR method would be optimal for assessing the conditions for NP treatment in EOR. The potential of LF-NMR relaxometry was also demonstrated in water–oil interaction investigations with NP treatment at the subsurface of the rock matrix under a confined geometry. LF-NMR is therefore a promising technique for rock core analysis, petroleum chemistry, and behavioral studies of inorganic NPs in rock cores. Finally, the theories of porous-media relaxation support the applications of LF-NMR technologies.