Universal Approximation Abilities of a Modular Differentiable Neural Network.

Approximation ability is one of the most important topics in the field of neural networks (NNs). Feedforward NNs, activated by rectified linear units and some of their specific smoothed versions, provide universal approximators to convex as well as continuous functions. However, most of these networks are investigated empirically, or their characteristics are analyzed based on specific operation rules. Moreover, an adequate level of interpretability of the networks is missing as well. In this work, we propose a class of new network architecture, built with reusable neural modules (functional blocks), to supply differentiable and interpretable approximators for convex and continuous target functions. Specifically, first, we introduce a concrete model construction mechanism with particular blocks based on differentiable programming and the composition essence of the max operator, extending the scope of existing activation functions. Moreover, explicit block diagrams are provided for a clear understanding of the external architecture and the internal processing mechanism. Subsequently, the approximation behavior of the proposed network to convex functions and continuous functions is rigorously proved as well, by virtue of mathematical induction. Finally, plenty of numerical experiments are conducted on a wide variety of problems, which exhibit the effectiveness and the superiority of the proposed model over some existing ones.

Development of Novel Silicon Quantum Dots and Their Potential in Improving the Enhanced Oil Recovery of HPAM.

Hydrolyzed polyacrylamide (HPAM) is commonly used in polymer flooding, however, it is prone to viscosity reduction at high temperatures and high salinities, weakening its ability to improve oil recovery. In this work, sulfonated modified silicon quantum dots (S-SiQDs) were synthesized and then added to HPAM to study the improvement of rheological properties and enhanced oil recovery performance of HPAM at high temperatures and salinities. It is found that the S-SiQDs with a concentration of only 0.1 wt % can significantly increase the viscosity of HPAM from 28.5 to 39.6 mPa·s at 60 °C and 10,000 mg/L NaCl. Meanwhile, the HPAM/S-SiQDs hybrid solution always possessed higher viscosity and viscoelastic moduli than HPAM, attributed to the hydrogen bonding between HPAM and S-SiQDs. Notably, HPAM/S-SiQDs still maintained elastic behavior at harsh conditions, indicating that they formed a strong network structure. Through oil displacement experiments, it was found that the oil recovery of HPAM/S-SiQDs was higher (28.3%), while that of HPAM was only 17.2%. Thereafter, the utilization sequence of oil during the displacement process was studied with nuclear magnetic resonance experiments. Ultimately, the oil displacement mechanism of HPAM/S-SiQDs was deeply analyzed, including viscosity thickening and wetting reversal.

Development and Gelation Mechanism of Ultra-High-Temperature-Resistant Polymer Gel.

To expand the applicability of gel fracturing fluids in ultra-high-temperature reservoirs, a temperature-resistant polymer was synthesized using the solution polymerization method. Subsequently, an ultra-high-temperature-resistant polymer gel was formulated by incorporating an organic zirconium crosslinking agent. A comprehensive investigation was carried out to systematically study and evaluate the steady shear property, dynamic viscoelasticity, and temperature and shear resistance performance, as well as the core damage characteristics of the polymer gel. The obtained results demonstrate that the viscosity remained at 147 mPa·s at a temperature of 200 °C with a shear rate of 170 s-1. Compared with the significant 30.9% average core damage rate observed in the guanidine gum fracturing fluid, the core damage attributed to the polymer gel was substantially mitigated, measuring only 16.6%. Finally, the gelation mechanism of the polymer gel was scrutinized in conjunction with microscopic morphology analysis. We expect that this study will not only contribute to the effective development of deep and ultradeep oil and gas reservoirs but also furnish a theoretical foundation for practical field applications.

Novel Anionic-Nonionic Surfactant Based on Water-Solid Interfacial Wettability Control for Residual Oil Development.

Irreversible colloidal asphaltene adsorption layers are formed on formation rock surfaces due to long-term contact with crude oil, and large amounts of crude oil adhere to these oil-wet layers to form residual oil films. This oil film is difficult to peel off due to the strong oil-solid interface effect, which seriously restricts further improvement in oil recovery. In this paper, the novel anionic-nonionic surfactant sodium laurate ethanolamide sulfonate (HLDEA) exhibiting strong wetting control was synthesized by introducing sulfonic acid groups into the nonionic surfactant laurate diethanolamide (LDEA) molecule through the Williamson etherification reaction. The introduction of the sulfonic acid groups greatly improved the salt tolerance and the absolute value of the zeta potential of the sand particles. The experimental results showed that HLDEA altered the wettability of the rock surface from oleophilic to strongly hydrophilic, and the underwater contact angle increased substantially from 54.7 to 155.9°. In addition, compared with LDEA, HLDEA exhibited excellent salt tolerance and enhanced oil recovery performance (the oil recovery was improved by 19.24% at 2.6 × 104 mg/L salinity). Based on nanomechanical experimental results, HLDEA was efficiently adsorbed on the core surfaces and regulated microwetting. Moreover, HLDEA effectively reduced the adhesion force between the alkane chains and the core surface, which facilitated residual oil stripping and oil displacement. This new anionic-nonionic surfactant affording great oil-solid interface wetting control has practical significance for the efficient development of residual oil.

Probing the Effect of Young's Modulus on the Reservoir Regulation Abilities of Dispersed Particle Gels.

The mechanical strength of dispersed particle gels (DPGs), which can be directly characterized by Young's modulus, is an important parameter affecting reservoir regulation performance. However, the effect of reservoir conditions on the mechanical strength of DPGs, as well as the desired range of mechanical strength for optimum reservoir regulation performance, have not been systematically studied. In this paper, DPG particles with different Young's moduli were prepared and their corresponding migration performances, profile control capacities and enhanced oil recovery abilities were studied by simulated core experiments. The results showed that with increase in Young's modulus, the DPG particles exhibited improved performance in profile control as well as enhanced oil recovery. However, only the DPG particles with a modulus range of 0.19-0.762 kPa could achieve both adequate blockage in large pore throats and migration to deep reservoirs through deformation. Considering the material costs, applying DPG particles with moduli within the range of 0.19-0.297 kPa (polymer concentration: 0.25-0.4%; cross-linker concentration: 0.7-0.9%) would ensure optimum reservoir control performance. Direct evidence for the temperature and salt resistance of DPG particles was also obtained. When aged in reservoir conditions below 100 °C and at a salinity of 10 × 104 mg·L-1, the Young's modulus values of the DPG particle systems increased moderately with temperature or salinity, indicating a favorable impact of reservoir conditions on the reservoir regulation abilities of DPG particles. The studies in this paper indicated that the practical reservoir regulation performances of DPGs can be improved by adjusting the mechanical strength, providing basic theoretical guidance for the application of DPGs in efficient oilfield development.

Recent advances in enhanced polymer gels for profile control and water shutoff: A review.

Polymer gels have been effectively employed as a water management material for profile control and water shutoff treatments in low-middle temperature and low-middle salinity reservoirs. However, most polymer gel systems have limitations under high temperature and salinity reservoir conditions, such as short gelation time, poor strength, and long-term instability. Therefore, several researchers have developed enhanced polymer gels to satisfy the water control requirements in high temperature and salinity reservoirs. This work reviews the five main types of enhanced polymer gels that have been developed so far: nano silica-enhanced gel systems, cellulose-enhanced gel systems, graphite-enhanced gel systems, oily sludge-enhanced gel systems, and foam-enhanced polymer gel systems. Further, this article investigates the fundamental properties, strengthening and crosslinking mechanisms, reservoir application conditions, and field applications of several enhanced polymer systems. In this paper, it is found that the addition of strengthening materials can increase the bound water content in the gel network and significantly improve the temperature and salt resistance of polymer gel, so as to cope with the application of profile control and water plugging in high temperature and high salt reservoirs. Moreover, it also offers references and future research directions for enhanced polymer gel systems.

A Supramolecular Reinforced Gel Fracturing Fluid with Low Permeability Damage Applied in Deep Reservoir Hydraulic Fracturing.

Gel fracturing fluid is the optimum fracturing fluid for proppant suspension, which is commonly applied in deep reservoir hydraulic fracturing. The content of polymers and crosslinkers in gel fracturing fluid is usually high to meet the needs of high-temperature resistance, leading to high costs and reservoir permeability damage caused by incomplete gel-breaking. In this paper, a supramolecular reinforced gel (SRG) fracturing fluid was constructed by strengthening the supramolecular force between polymers. Compared with single network gel (SNG) fracturing fluid, SRG fracturing fluid could possess high elasticity modulus (G' = 12.20 Pa) at lower polymer (0.4 wt%) and crosslinker (0.1 wt%) concentrations. The final viscosity of SRG fracturing fluid was 72.35 mPa·s, meeting the temperature resistance requirement of gel fracturing fluid at 200 °C. The gel-breaking time could be extended to 90-120 min using an encapsulated gel breaker. Gel particles are formed after the gel fracturing fluid is broken. The median particle size of gel particles in the SRG-breaking solution was 126 nm, which was much smaller than that in the industrial gel (IDG) breaking fluid (587 nm). The damage of the SRG-breaking solution to the core permeability was much less than the IDG-breaking solution. The permeability damage of cores caused by the SRG-breaking solutions was only about half that of IDG-breaking solutions at 1 mD.

Development of the Gemini Gel-Forming Surfactant with Ultra-High Temperature Resistance to 200 °C.

In order to broaden the application of clean fracturing fluid in ultra-high temperature reservoirs, a surfactant gel for high-temperature-resistant clean fracturing fluid was developed with a gemini cationic surfactant as the main agent in this work. As the fracturing fluid, the rheological property, temperature resistance, gel-breaking property, filtration property, shear recovery performance and core damage property of surfactant gel were systematically studied and evaluated. Results showed the viscosity of the system remained at 25.2 mPa·s for 60 min under a shear rate of 170 s-1 at 200 °C. The observed core permeability damage rate was only 6.23%, indicating low formation damage after fracturing. Due to micelle self-assembly properties in surfactant gel, the fluid has remarkable shear self-repairability. The filtration and core damage experimental results meet the national industry standard for fracturing fluids. The gel system had simple formulation and excellent properties, which was expected to enrich the application of clean fracturing fluid in ultra-high temperature reservoirs.

Gelling Behavior of PAM/Phenolic Crosslinked Gel and Its Profile Control in a Low-Temperature and High-Salinity Reservoir.

Gel conformance control technology is widely used in moderate and high temperature reservoirs. However, there are few studies on shallow low-temperature and high-salinity reservoirs. The difficulties are that it is difficult to crosslink at low temperatures and with poor stability at high salt concentrations. Therefore, the PHRO gel was developed, which was composed of gelatinizing agent (polyacrylamide), crosslinking agents (hexamethylenetetramine and resorcinol) and crosslinking promoting agent (oxalic acid). The PHRO could form high-strength gels in both deionized water and high-concentration salinity solutions (NaCl, KCl, CaCl2 and MgCl2). The observation of the microstructure of PHRO gel shows that a strong "stem-leaf"-shaped three-dimensional network structure is formed in deionized water, and the network structure is still intact in high-concentration salt solution. The results show that PHRO has good salt resistance properties and is suitable for conformance control of low-temperature and high-salinity reservoirs.

Mussel-inspired superhydrophilic membrane constructed on a hydrophilic polymer network for highly efficient oil/water separation.

HYPOTHESIS: Superhydrophilic/underwater superoleophobic membrane constructed by hydrophilic polymers possesses great advantage in the separation of oily waste water, due to its intrinsic oil-repellent property. The formation of hydration layer to repel and block oil is considered as the mechanism of underwater superoleophobicity and subsequent oil/water separation. Constructing a stable hydrophilic polymer network on the substrate surface would significantly improve the robustness of hydration layer. EXPERIMENTS: In this work, a feasible and universal mussel-inspired dip-coating method was developed for constructing stable hydrophilic polymer network onto target substrate surface, via successively immersing substrate membranes into aqueous solutions of polydopamine (PDA) and catechol-functionalized hydrophilic polymer (CFHP). After pre-wetting with water, the polymer network would swell with water to form a thin and stable water film layer, serving as a barrier against oil penetration. FINDINGS: The as-prepared CFHP/PDA modified membranes exhibit outstanding performance in separating various oil/water mixtures and oil-in-water emulsions stabilized by surfactants, with separation flux up to 5641.1 L·m-2·h-1 and separation efficiency achieving 99.98%. The surface modification method developed in this work can be easily extended to various materials and membrane systems, for achieving a variety of practical applications such as industrial wastewater treatment.

Lignosulfonate/diblock copolymer polyion complexes with aggregation-enhanced and pH-switchable fluorescence for information storage and encryption.

Employing natural polymers as building blocks is favorable to construct sustainable functional materials and realize biomass utilization. Here we developed a series of supramolecular polyion complexes (PICs) composed of lignosulfonate and double-hydrophilic diblock copolymer poly(ethylene oxide-b-N, N-dimethylaminoethyl methacrylate) (PEO114-b-PDMAEMA24), which performed a blue emission with quenching-enhancing tendency and copolymer distribution inversion-involved assembling transformation at pH 5.6 and monotonic greenish-blue fluorescence promotion at pH 9.5 by increasing PEO114-b-PDMAEMA24 contents. Electrostatic interactions and multiple hydrogen bonds were revealed controlled the assembling behavior and affected the emission via altering the restriction of molecular motion, through-space conjunction, and non-luminous complexation. The multiple interacting sites and special topology of diblock copolymer contributed to the efficient fluorescence regulation. Information writing-erasing and encryption-decryption systems were established by utilizing emission intensity regulation and pH-responsive emission chromism. This work paved a new way to enhance lignin fluorescence and broadened potential applications of lignin composites in realms of sensing, imaging, monitoring, and anticounterfeiting.

Self-growing Hydrogel Particles with Applications for Reservoir Control: Growth Behaviors and Influencing Factors.

Chemical profile control agents are the key for conducting effective reservoir control to enhance crude oil recovery. Self-growing hydrogel particles have emerged as highly competitive profile control agents as they can grow for control after migrating to deep fractures, exhibiting great potential in long-term adaptive reservoir control. In this work, self-growing hydrogel particles were prepared by mechanical shearing of self-repairing bulk gels constructed by catechol-functionalized partially hydrolyzed polyacrylamide p[AM-AANa-DOPA] and phenolic resin cross-linking agents. After aging for 15 days under the reservoir conditions, the median size of hydrogel particles increased from ∼3.5 to ∼18.0 µm, demonstrating apparent self-growing property and significantly enhanced resistant coefficient in waterflooding. Different factors affecting growth behaviors of hydrogel particles including cross-linking density, chemical re-cross-linking, hydrolysis degree, and molecular weight of the copolymer were studied. The results showed that the cross-linking density affected the strength and toughness of the bulk hydrogel, with appropriate polymer chain mobility facilitating the intermolecular interactions. Quantitative NMR results of the gelation process indicated that chemical re-cross-linking contributed little to the growth of hydrogel particles. Based on the rheological and nanomechanical results, bulk gels prepared by polymers with a lower hydrolysis degree and smaller molecular weight possessed a higher elastic modulus recovery rate, while the corresponding hydrogel particles exhibited stronger adhesion among each other. This work provides new insights into the growth behavior of hydrogel particles, which may help better understand and select a suitable hydrogel system and preparation technology and further promote efficient reservoir control.

Assembly of Ultralight Dual Network Graphene Aerogel with Applications for Selective Oil Absorption.

High-performance graphene aerogels with well-developed internal structures are generally obtained by means of introducing additive materials such as carbon nanotubes, cellulose, and lignin into the aerogel network, which not only enhances the cost but also complicates the preparation process. Therefore, tailoring the internal structure of pristine graphene aerogel in a feasible way to achieve high performance is of great significance to the practical applications. Herein, a novel cysteamine/l-ascorbic acid graphene aerogel (CLGA) was fabricated by a simple one-step hydrothermal method followed by freeze-drying. Through the creative combination of the reducing agent l-ascorbic acid and cross-linking agent cysteamine, a dual-network structure was constructed by both layered physical stacking and vertical chemical cross-linking. The addition of cysteamine not only enhanced the reduction degree but also assisted the formation of more vertical connections between graphene nanosheets, resulting in more abundant pores with smaller sizes compared with graphene aerogels prepared by the traditional hydrothermal reduction method. CLGA possessed an ultra-low density of 4.2 mg/cm3 and a high specific surface area of 397.9 m2/g. As expected, this dual-network structure effectively improved the absorption capacity toward a variety of oil and organic solvents, with an outstanding oil absorption capacity up to 310 g/g. Furthermore, CLGA possessed good mechanical properties and oil/water selectivity. The absorbed oil could be recovered by both continuous absorption-removal process and mechanical squeezing, making the as-prepared aerogel superior absorbent material for a variety of applications, such as selective oil absorption and water treatment.

Core-Shell Nanohydrogels with Programmable Swelling for Conformance Control in Porous Media.

Conformance control during waterflooding in an oil reservoir is utilized to redistribute water and increase the sweep efficiency and hence oil production. Using preformed gel particles can effectively redirect the flow by blocking the high-permeability zones and forcing water into low-permeability zones where the oil is trapped. However, the size of such gel particles can limit their applications deeper within the reservoir and can result in shear-induced degradation near the well bore. Here, we fabricate core-shell nanohydrogels with delayed swelling behavior; their volume increases by a factor of 200 after about 30 days in brine under reservoir conditions. We study their effect on the flow behavior in a three-dimensional porous medium micromodel consisting of randomly packed glass beads. Using confocal microscopy, we directly visualize the spatial variations of flow in the micromodel before and after nanohydrogel injection and swelling. The swollen nanohydrogels block some pores reducing the permeability of the micromodel and diverting the water into low-permeability regions. A core flood experiment further confirms that the nanohydrogels can significantly reduce the permeability of a reservoir sample and divert the fluid flow. Our results demonstrate that these core-shell nanohydrogels might be useful for flow control in porous media and can be used as a conformance control agent.

Self-Lubricating Supramolecular Hydrogel for In-Depth Profile Control in Fractured Reservoirs.

In-depth profile control is a great challenge for high-efficiency oil displacement by water flooding. In this work, a shear-responsive self-lubricating hydrogel FPP-0.5, by combining the thixotropic FT (N-fluorenylmethoxycarbonyl-l-tryptophan) supramolecular network with high-strength PAM-PAANa (PAM: polyacrylamide, PAANa: sodium polyacrylate) polymer network, was synthesized and applied for in-depth profile control in water flooding. The disassembly of the FT supramolecular network induced by shear force, accompanied by the formation of a lubricating layer on the gel surface, gives FPP-0.5 gel self-lubricating function. Meanwhile, the PAM-PAANa polymer network, as a protective scaffold for the FT network, endows FPP-0.5 with high strength, making it difficult to be broken by water flow. Moreover, the shear responsiveness enables FPP-0.5 to adjust its own strength and self-lubricating performance according to the different stages of the profile control process, so as to realize dynamic profile control. From oil displacement results, in comparison to the PAM-PAANa gel, the plugging rate, water flooding volume sweep efficiency, and oil recovery of FPP-0.5 were improved by 83.1, 155.4, and 34.0%, respectively, up to 86.6, 83.5, and 71.3%, indicating its better in-depth profile control ability.

Study on the Reducing Injection Pressure Regulation of Hydrophobic Carbon Nanoparticles.

The interest in the application of nanofluid in reducing injection pressure has been increasing especially for tight reservoirs. In this work, a new type of hydrophobic carbon nanofluid was prepared and the pressure-reducing performance was investigated. The results of particle size distribution, zeta potential, and transmission electron microscopy image showed that the dispersion of nanofluid was uniform and stable. In addition, the hydrophobic carbon nanofluid showed excellent antitemperature and antisalinity property. The contact angle of oil-wet glass slide can range from 45 to 89° after it adsorbs hydrophobic carbon nanoparticles (HCNPs). The atomic force microscope tests showed that the core surface roughness was reduced about 16.67%. The core flooding tests showed that the pressure-reducing rate of 0.15 wt % HCNP nanofluid can reach 17.00%. HCNPs show good performance in reducing pressure and have a broad application prospect in oil field development.

Giant surfactant-stabilized N2-foam for enhanced oil recovery after water flooding.

A novel giant surfactant, APOSS-PS50, possessing good surface activity, and viscosifying and reinforcing ability as a foam stabilizer, was synthesized successfully to enhance the physical properties of foaming solutions and foam. APOSS-PS50 was widely distributed at the foam gas-liquid interface and adjacent liquid layers through diffusion and adsorption, obviously decreasing the surface tension and improving the foamability and stability of the foam. Furthermore, the aggregation of APOSS-PS50 in the foam films resulted in the formation of a self-assembled nano-sized network through supramolecular interactions (such as hydrogen bonding, π-π stacking, and van der Waals attraction), thus increasing the foam viscoelasticity, including its interfacial viscoelastic modulus and apparent viscosity. Meanwhile, from the sandpack flooding experiments, compared with HPAM/AOS (HPAM: partially hydrolyzed acrylamide and AOS: alpha olefin sulfonate), the differential pressure and final oil recovery after APOSS-PS50/AOS foam flooding increased by 23.5% and 23.2%, up to 2.68 MPa and 81.7%, respectively. In general, APOSS-PS50 significantly promoted the plugging, profile control and oil displacement performance of foam.

Alternating electric field-induced ion current rectification and electroosmotic pump in ultranarrow charged carbon nanocones.

Pumping fluid in ultranarrow (sub-2 nm) synthetic channels, analogous to protein channels, has widespread applications in nanofluidic devices, molecular separation, and related fields. In this work, molecular dynamics simulations were performed to study a symmetrical sinusoidal electric field-induced electroosmotic pump in ultranarrow charged carbon nanocone (CNC) channels. The results show that the CNC channels could rectify the ion current because of the different ion flow rates in the positive and negative half circles of the sinusoidal electric field. Electroosmotic flow (EOF) rectification yielded by the ion current rectification is also revealed, and net water flow from the base to the tip of the CNC channels is observed. The simulations also show that the preferential ion current conduction direction in the ultranarrow CNC channels (from base to tip) is opposite to that in conical nanochannels with tip diameters larger than 5 nm (from tip to base). However, the preferential EOF direction is the same as that of large conical nanochannels (from base to tip). We also investigated the influences of ion concentration and the amplitudes and periods of the sinusoidal electric field on the EOF pump. The results show that high ion concentration, large amplitudes, and long periods are desired for high EOF pumping efficiency. Finally, through comparison with a constant electric field and a pressure-induced water pump, we prove that the EOF pump under an alternating electric field has a higher pump efficiency. The approach outlined in this work provides a general scheme for pumping fluid in ultranarrow charged conical nanochannels.

Investigation of Active-Inactive Material Interdigitated Aggregates Formed by Wormlike Micelles and Cellulose Nanofiber.

In this work, a novel active-inactive material interdigitated aggregates (AIMIAs) structure was constructed by self-assembled wormlike micelles (WLMs) and one-dimensional cellulose nanofiber (CNF). The rheological behaviors and microstructures of the AIMIA systems with different CNF concentrations were investigated by rheometer, cryogenic transmission electron microscopy, and environmental scanning electron microscope. Some key parameters, including zero-shear viscosity (η0), relaxing time (τR), and contour length ( L), were calculated to analyze the changes in the properties of different systems. Meanwhile, a proper mechanism describing the interaction between CNF and WLMs was proposed. Through this work, we expect to deepen the understanding of the AIMIAs structure and widen its application.

A Study on Preparation and Stabilizing Mechanism of Hydrophobic Silica Nanofluids.

Nanofluids have increasingly drawn interest in recent years with their various applications in a number of fields. The method for the preparation of stable nanofluids is a key concern for extending the application of nanofluids. This study focuses on the effect of pH, dosage of surfactant (TX-100), and nanofluid concentration on the stability of a silica nanofluid. Particle size and zeta potential are two important factors to consider in evaluating the stability of the silica nanofluid. Results indicate that the stability of the silica nanofluid highly depends on pH, dosage of surfactant (TX-100), and nanofluid concentration. On the basis of these experiments, the best conditions for the preparation of a silica nanofluid are 0.1 wt. % for the concentration of silica nanoparticles and TX-100 and 10 for pH. A transparent and stable silica nanofluid can thus be obtained.