Introduction
The importance of subterranean petroleum hydrocarbons as a dependable energy resource has intensified on global scale owing to the ever-increasing consumption of crude oil and/or associated products in industrial, household, transportation and technological applications [1,2]. This has led to the exploration and production of oil from complicated reservoir formations, wherein problems such as low permeability, heterogeneity and less accessibility persist during extraction processes. The initial stage encompasses the application of primary and secondary recovery techniques to produce one-thirds to nearly one-half of the original oil in place (OOIP) by natural drive and water/gas injection. Enhanced oil recovery (EOR) methods have attracted widespread attention in the last few decades to attain optimized production of residual oil trapped by alteration of reservoir fluid properties after conventional recovery [3,4]. Surfactant flooding is a promising EOR technique employed by the production sector since 1970s decade [5,6]. This type of oil recovery functions by allowing surfactants or “surface-active agents” to adsorb onto the interface of oil/water, thereby decreasing interfacial free energy and increasing dimensionless capillary number [5-7]. Polymer improves the viscosity of displacing (injected) fluid to reduce the mobility ratio between water and oil [8,9]. Furthermore, polymer addition increases the viscous force perpendicular to oil-water interface and responsible for pushing the residual oil towards the production well [8,9]. When this force exceeds the capillary forces holding crude oil within rock-pores, residual oil detaches from the rock surface and mobilizes forward with increasing sweep efficiencies. Nanoparticles, in conjunction with surfactant and polymer, adsorb onto interfaces to enhance the mechanical barrier onto displaced crude oil surfaces and produce impulsive emulsions with improved crude oil attracting ability [10,11]. A pivotal aspect of chemical EOR lies in proper screening and optimization of displacing fluid, keeping in mind the effectiveness as well as cost-profitability of the method employed [12]. Hence, surfactant, polymer and/or nanoparticle in chemical fluid must be introduced to create a forward-moving oil bank within porous rock formations, which can significantly improve the oil recovery and maintain pressure gradient during chemical fluid +/ chase water injection.
Simulation studies are important to assess the flooding performance of injected chemical fluids, and predict how oil displacement will occur under specified reservoir/fluid conditions [13,14]. Prior to simulation, the technological feasibility of different EOR routes are tested by experimental investigations [14,15]. Such studies provide useful input information to allow the simulator to identify reservoir parameters, predict recovery and testing the effectiveness of different EOR projects with similar components [13-16]. The current industry is involved in the application of realistic chemical flood simulators like STARS by Computer Modelling Group (CMG), UTCHEM by the University of Texas at Austin (UT Austin), REVEAL by Petroleum Experts (Petex), and ECLIPSE by Schlumberger (SLB). The physics associated with fluid properties’ evaluation differ in each type of reservoir simulator [17-19]. UTCHEM is a compositional simulator capable of simulating different types of EOR processes owing to the provision of four different phases (gas, aqueous, oil, microemulsion) and incorporation of advanced numerical concepts [20]. REVEAL, a full field reservoir expert, is similar to UTCHEM with surfactant phase behavior and mobility control options, permeability reduction and polymer degradation parameters [21]. However, this tool is not well known among professionals and engineers in production areas and the existence of a fourth phase i.e. microemulsion may cause problems in field studies [17,20,21]. As per UTCHEM and REVEAL, the presence of a microemulsion phase is a key parameter to model displacement efficiency, in spite of the fact that microemulsion properties are not generally measured in pilot tests and field operations [18-21]. Both ECLIPSE and STARS do not consider microemulsion phase as contributor to flooding simulation and represent oil displacement behavior via analyses of relative permeability curves for experimental results [17,22,23]. However, ECLIPSE software, though common in the industry, does not encompass the technical functionalities such as salinity effects, adsorption, polymer concentration mixing, multi-component EOR, shear thickening and degradation regimes required for accurate modelling [22]. Another powerful flooding simulator is CMG, which is capable of modeling flooding results and manage complex behavior of oil-chemical-water systems in laboratory-scale and field-scale porous media [24,25]. Goudarzi and other researchers [17,19] assessed the performance of different reservoir simulators and developed an EOR benchmark to improve chemical design for field-scale as well as lab-scale operations. Pandey et al. [26] employed CMG-STARS for coreflood modelling experiments and investigated flow parameters that could be subsequently used in pilot field tests. Kazempour and others [18] investigated the validity of multi-phase component EOR systems in detail, and identified the dynamic behavior of fluid components existing within core-flood model. Tunnish et al. [27] successfully matched experimental flooding results using CMG to effectively tune relative permeability curves and predict the chemical fluid’s ability to produce in-situ crude oil. Dahbag et al. [28] reported the performance of ionic liquid/surfactant flooding during chemical oil recovery processes, and found the results can be used to predict future scenarios. CMG tool is reliable and instrumental in evaluating the potential of conventional and modern EOR methods [17-19,24-28].
In this article, a series of flooding experiments were performed to investigate the secondary and tertiary oil recoveries using surfactant, surfactant-polymer and surfactant-polymer-nanoparticle slugs. Initially, physicochemical behavior of designed fluids were evaluated by a series of experimental studies. A Cartesian grid model was developed using CMG-STARS software, and parameters such as rock-fluid properties, interfacial tension, viscosity, adsorption and injector/producer geometry were entered in the simulation model. Thereafter, the experimental results obtained in the laboratory were history-matched for specified builder and injection pattern/time was set with CMOST tool. Emphasis is put on the injected fluid composition, flow rate and flooding period. Using detailed methodical approach to identify and predict well-matched recovery data, produced recovery data were optimized with minimal error as compared to laboratory results. This model is useful to simulate surfactant/polymer/particle behavior on core-scale, and optimize brine/chemical flooding from functional viewpoint.
Experimental and Simulation
Materials
The surfactant employed in this study is N,N′-bis(dimethyltetradecyl)-1,6-hexanediammonium bromide (abbreviated as 14-6-14 GS) with molecular weight of 726 g/mol. This gemini surfactant was synthesized and characterized in our earlier papers [29,30]. Partially hydrolysed polyacrylamide (PHPA), a water-soluble polymer was purchased from SNF Floerger, SNF SAS, ZAC de Millieux, Andrézieux, France. It has molecular weight of 2.1 × 107 g/mol with 26.4% hydrolysis. Aqueous polymer solutions were prepared in accordance with the American Petroleum Institute: Recommended practices for evaluation of polymers used in EOR operations (API RP 63). Silica (SiO2) nanopowder (5-15 nm) was obtained from Merck Industries. Sandstone core employed in flooding experiments was procured from Kalol field in Gujarat, India. Crude oil sample has total acid number (TAN) of 0.044 mg KOH/g, kinematic viscosity of 6.147×10-5m2/s and 23.55° API gravity at 303 K. It was procured from Ahmedabad oil field, ONGC Asset, India. Double distilled water was extracted from distillation apparatus in our laboratory.
Physicochemical evaluation tests
Wettability behavior of 14-6-14 GS was investigated by contact angle studies with the help of Kruss DSA25 Drop Shape Analyzer. Adsorption behavior of gemini surfactant molecules onto sand surface was conducted by UV spectrometric analyses to determine the amount of 14-6-14 GS adsorbed per weight sand (in mg/g). Interfacial tension experiments were performed by analyzing rotating crude oil drop profile in continuous surfactant/polymer/nanoparticle containing aqueous solution with the help of spinning drop SVT20 tensiometer (Dataphysics). Viscosity values of aqueous chemical fluids were measured using cup and bob geometry in Bohlin Gemini 2 Rheometer instrument at 303 K. The obtained experimental results serve as input-data during simulation studies.
Flooding procedure
The experimental flooding apparatus (Porous Material Inc.) consists of core-holder, positive displacement pump, chemical slug injectors and measuring cylinder for collecting effluent samples. Sandstone core with 8.74 cm length and 3.66 cm diameter was initially saturated with 1.0% NaCl brine for 72 h to saturate the cores; and obtain porosity values in the range 17-18%. The core sample showed permeabilities in the range 350-400 milliDarcies (mD). When placed within core-holder apparatus, a confining pressure of ~1000-1200 psi was employed to hold the core in vertical position. Crude oil was injected into the pores to displace aqueous phase, until irreducible saturation state was achieved. This was followed by an ageing period of 6 days to obtain an oil-saturated reservoir model in the laboratory. Secondary flooding was investigated by brine flooding at the rate of 10 ml/h to recover a fraction of crude oil. When water cut percentage exceeds ≥ 95%, chemical aqueous fluid containing surfactant +/ polymer +/ nanoparticle was flooded at 5-10 ml/h rate to sweep residual oil during EOR. Finally, chase water was injected at the same flow-rate to maintain pressure drop for favorable oil displacement. Effluent liquid produced during secondary and tertiary recovery studies were collected in graduated cylinders.
Simulation methodology
The STARS simulator package in CMG is widely employed compositional tool in the petroleum industry, with the capacity to develop reservoir models [24-27]. A Cartesian grid system with specified divisions along X-axis, and the developed model was simulated to match flooding history data using CMOST analysis tool. Prior to running the STARS simulation for aqueous flooding model and subsequent history-matching of experimental recovery data, the following assumptions were made to obtain accurate findings [31,32]:
  1. The reservoir initially consists of two phases, namely, crude oil and water.
  2. The amount of free gas/solvent gas in the core model is assumed as zero.
  3. A grid-based core model is considered, with uniform properties and no geological complexities/heterogeneities.
  4. Fluid flow in radial direction is negligible as compared to that in axial direction.
  5. Salinity effect on phase behavior is ignored.
  6. Chemical reactions do not occur.
  7. Oil and water flowing through porous media obeys the Darcy’s Law.
Porosity, permeability and crude oil properties were introduced as input data for reservoir characterization. In recent years, the need to develop appropriate flooding model has paved the way to make informed decisions during chemical fluid selection/optimization and field implementation [33-35]. Druetta and co-workers [35] developed a flooding simulator to investigate EOR properties of different chemical fluid compositions. Arhuoma et al. [36] found that CMG simulation model is useful to determine displacement phenomena governing flooding behavior, depending on injection fluid type. The effectiveness of numerical simulation studies on chemical flooding showed far-reaching consequences during oilfield applications, as evident from the findings of earlier papers [34,37]. This kind of grid-based model helps in understanding fluid flow behavior prior to injection; as well as achieve a sufficiently robust numerical model [33,35,36]. Table 1 presents the core and fluid properties employed in STARS model.
Table 1. Core model and fluid parameters for flooding simulations.