Pore-Scale Flow Fields of the Viscosity-Lost Partially Hydrolyzed Polyacrylamide Solution Caused by Sulfide Ion

Abstract

The rheology of a partially hydrolyzed polyacrylamide (HPAM) solution plays an important role in its oil recovery during polymer flooding. However, multiple factors in brine, such as sulfide ions, cause a dramatic loss in the viscosity and oil recovery. To better understand the sulfide-induced viscosity loss and the consequent flow mechanisms in pore networks, the morphology of polymer solutions with and without sulfide ion was observed by scanning electron microscopy; and the variations of the pore scale flow fields were demonstrated by a microscopic visualization seepage experiment combined with Micro-PIV (Microscale Particle Image Velocimetry). The results showed that, with the presence of sulfide ion, the microstructure of the polymer changed from a uniform three-dimensional network structure to loose and uneven floccules, which resulted in viscosity loss (over 70% with 5-mg/L sulfide ion). Moreover, higher concentrations of sulfide ions (5 mg/L and 10 mg/L) resulted in earlier shear thinning characteristics than those with lower sulfide concentrations. Due to viscosity loss, the average flow velocity in the main stream of the microscopic seepage experiment increased more significantly than that without sulfide. However, the viscosity loss alone cannot independently explain the severe viscous fingering during the subsequent post-water flooding, which was about five times greater than that of the primary water flooding in terms of the velocity ratio between the mainstream and margin. A further pore-scale flow field analysis exhibited an eccentric and a bimodal velocity distribution in the throat along the radial and axial directions, respectively. The former distribution indicated that the adsorbed polymer on the pore wall was broken through by hydraulic shear due to the collapsed structure caused by sulfide ion. The latter suggested that another sulfide-induced impact was an earlier-occurring non-Newtonian characteristic with a low shear rate. Therefore, instead of viscosity loss, elastic loss is the dominant mechanism affecting the characteristics of the aggregate flow field under the action of sulfide. Microscopic flooding combined with Micro-PIV is a feasible and essential method to reveal pore scale flow mechanisms.

1. Introduction

Partially hydrolyzed polyacrylamides (HPAM) are widely used to further enhance the recovery of oil reservoirs after water flooding [1,2,3,4]. The EOR (enhanced oil recovery) of HPAM mainly relies on its solution’s viscoelasticity to reduce the oil/water mobility ratio and improve the sweep efficiency [5,6]. Generally, a large volume of brine from oil wells should be used to dilute the HPAM solution before injection [7], during which the viscosity of the HPAM would be negatively influenced by multiple brine properties, such as high temperature [8], pH [9], shearing [10], and salinity [11]. Especially, sulfide ion in brine was found to cause a dramatic viscosity loss of the polymer solution [12], then negatively affect the oil recovery [9].

The study of rheological properties is important in the petroleum field. There have been previous studies on rheological properties of drilling fluid to improve the drilling efficiency and speed up recovery [13]. The study of the pore scale is also crucial, such as the study of supercritical carbon dioxide migration in pores [14]. Although flooding experiments have suggested that viscoelasticity is a major reason for polymer EOR, limited experimental research discussed how the sulfide-induced rheology variation impacts the flow characteristics in pores and then the mechanisms of oil recovery [15]. This knowledge is essential to improve polymer performances in reservoirs with sulfide. Since the flow mechanisms of polymers are initially obtained from microscopic seepage experiments [16], it is of great significance to observe the flow field of HPAM and to analyze the mechanisms affecting recovery [17,18,19,20]. However, instead of accurate pore scale flow fields, a traditional microscopic seepage experiment only provided results on multiphase morphology and saturation [21]. In recent times, the Micro-PIV technique (Microscale Particle Image Velocimetry) has allowed researchers to reveal different flow patterns in pores, such as Haines jumps, shear-induced circulations caused, and elastic turbulence, which cannot be quantified or even observed in traditional microscopic experiments [22,23,24,25]. Therefore, Micro-PIV can serve as a feasible method to study the pore-scale polymer flow field.

In this study, the microscopic seepage experiment and Micro-PIV were combined to analyze the flow field of HPAM solution with or without sulfide ions. Additionally, a rheometer and scanning electron microscopy were used to investigate the sulfide-induced variations of rheological and morphological properties of the polymer. The purpose of this study is to exhibit the characteristics of the flow field in pores with sulfide and then to reveal how the sulfide-induced viscoelasticity loss impact in the porous flow, which are essential for the economic and rational development of polymers.

2. Materials and Methods

2.1. Experimental Set-Up

HPAM (average molecular weight of 25 million Daltons) was obtained from Daqing Refining & Chemical Company (Heilongjiang, China), which was dissolved in artificial brine with a concentration of 2000 mg/L and an initial viscosity of 62.2 mPa·s (at 7.34 s⁻¹, 40 °C). The artificial brine comprised NaCl (3300 mg/L); CaCl₂ (200 mg/L); MgCl₂ (100 mg/L); and different concentrations of Na₂S (0, 2, 3, 5, and 10 mg/L). Polymer solutions were dissolved with different sulfide ion concentrations (400 rpm, 2 h) by stirring with a stirrer. The viscosity of the fluids was measured by a Brookfield DV-III viscometer (Brookfield, MA, USA), with the shear rates ranging from 0.01 to 1000 s⁻¹ at 40 °C.

The polymer solution with different sulfide ion concentrations (0–10 mg/L) was incubated hermetically for 50 h. The viscosity was measured at regular intervals (2–5 h) at 7.34 s⁻¹ and 40 °C. After the polymer was sealed at 40 °C for 24 h, the microstructure of the polymer was observed by a scanning electron microscope (SEM) with an accelerating voltage of 5 kV and a magnification of ×20.0 K.

The pore scale flow field analysis system included a flooding system [26] and a Micro-PIV system. The flooding system consisted of a syringe pump, accumulators, and a model holder. An electronic manometer (0–50 kPa, Senex, Indianapolis, IN, USA) was used to measure the injection pressure. The micro model (Figure 1) was the heart of the apparatus, of which the pore structure was identical to the cross-section image of a natural sandstone core and was etched onto a flat glass plate. On the basis of our previous studies, which selected six observation points, it was suggested that the main stream and margin can be represented by the two selected observation sites [27]. At the two opposite corners, the covering plate had an inlet and outlet hole, making fluids flow through the pore network. The approximate permeability of the micromodel was 2000 mD (by fluid permeability measurement). The external size, pore volume, porosity, and pore diameter of the model were 40 × 40 mm, 27 μL, 42.5%, and 20–100 μm, respectively. In the steel model holder (a hollow cylinder), the micro model was clamped horizontally by thick glasses and filled with water to load the overpressure, which made the light travel through [27].

The Micro-PIV system (Figure 2) is composed of a 200-megajoule dual-pulsed Nd: YAG laser (peak emission wavelength of 532 nm and pulse frequency of 15 Hz), a 12-bit CCD camera (10× objective), and a microscope (ZEISS, Oberkohen, baden-Wurberg, Germany). Fluoro-max red fluorescence (0.005%) of 1 micron size was used to track the liquid flow. The velocity vector field and average velocity in the observation field were obtained by the supporting software DaVis (10.0.5), which also calculated the average velocity in the flow field for each time step by

$${\mathrm{V}}_{\mathrm{avg}}={{\displaystyle \sum }}_{\mathrm{i}=1}^{\mathrm{N}}\frac{{\mathrm{V}}_{\mathrm{i}}}{\mathrm{N}}$$

(1)

where V_avg is the average velocity, m/s; N is the number of velocity vectors in the measurement view, dimensionless; and V_i represents the velocity of each vector, m/s.

2.2. Velocity Field Measurement Procedures

First, the micromodel was saturated by injecting 2.0 pore volumes (PVs) of artificial brine under vacuum conditions. Next, 5.0 PV of brine and polymer (with the sulfide ion concentrations of 0 mg/L and 10 mg/L) were injected as water and polymer flooding, respectively. Afterwards, 5.0 PV of artificial brine was injected again at the post-water flooding stage. The injection rate and temperature of all the experiments were 4 µL/min and 40 °C, respectively. The pore scale velocity fields were measured in the selected observation ranges (Figure 1) in both the main stream and margin regions. In addition, the pore-scale velocity dynamics in the radial (Lines 1 and 3 in Figure 1) and axial directions (Lines 2 and 4 in Figure 1) were analyzed in a throat and a pore.

3. Results

3.1. Effect of Sulfide Ion on Rheology of the Polymer Solution

Viscosity measurements were carried out to understand the effect of sulfide ion concentrations on the apparent viscosity of the polymer, the results of which are shown in Figure 3. Figure 3a shows that sulfide ion has a negative effect on the polymer viscosity. With the increase of the sulfide concentration within 0–10 mg/L, the polymer viscosity loss increased. Over 80% viscosity loss and the highest decline rate of 1.438 mPa·s/h occurred with 5 mg/L sulfide during the initial 15 h. The viscosity of each solution became relatively stable after 15 h.

As shown in Figure 3b, all the apparent viscosity with different sulfide concentrations declined with the shear rates. With a low sulfide ion concentration (less than 3 mg/L), the first Newtonian zone [28] (relatively constant apparent viscosity) existed within low shear rates (from 0.01 s⁻¹ to 0.1 s⁻¹), while no Newtonian zone was observed with higher sulfide concentrations (5 mg/L and 10 mg/L), during which the apparent viscosity decreased directly from initial shear rates (0.01 s⁻¹). Regarding the viscosity loss at a given shear rate, it decreased gradually with the increase of the shear rate. For instance, comparing the polymer solutions without sulfide and with 5-mg/L sulfide, the shear rate of 6 s⁻¹ resulted in a viscosity loss about 85%, while there was only a 36% loss at 297 s⁻¹.

3.2. Microscopic Morphology Analysis

Without sulfide ion, the polymer aggregates formed multilayered and three-dimension network structures with uniform sizes (Figure 4a). With sulfide ion, small, blocky substances and loose frameworks with nonuniform distribution were observed (Figure 4b).

3.3. Flow Field of Polymer Solution in Pores by Micro-PIV

In order to understand the flow field of HPAM in porous media, a set of microscopic seepage experiments of polymer solutions without and with sulfide (5 mg/L) were carried out with Micro-PIV.

Regarding the average velocity of the observed field in the main stream, the sulfide-free polymer flooding and the post water flooding were slower (7.05 μm/s and 1.16 μm/s) than water flooding (11.86 μm/s) (

Table 1. Average velocity (μm/s) of the observed regions in the microscopic flow fields.

Experiments	Stage	Regions
		Main Stream	Margin
Water	Water flooding	11.86	5.94
Polymer without sulfide	Polymer flooding	7.05	5.68
	Post water flooding	1.16	1.30
Polymer with sulfide	Polymer flooding	10.26	5.80
	Post water flooding	21.91	1.97

) (Figure 5), while, with sulfide, the velocity during polymer flooding and, especially, the post-water flooding increased significantly (10.26 μm/s and 21.91 μm/s) (Table 1).

At the margin region, both the sulfide-free (Figure 6) and sulfide-added (Figure 7) polymer flooding (5.68 μm/s and 1.30 μm/s) and post-water flooding (5.80 μm/s and 1.97 μm/s) flowed slower than water flooding (5.94 μm/s) (Table 1). As expected, during the whole process of both polymer experiments, the average velocities at the margins were much lower than those of the main stream.

Regarding the flow field in the throat in the radial direction (along Line 1 in Figure 5a) and in the axial direction (alone Line 2 in Figure 5a), post-water flooding with sulfide was the most abnormal one (Figure 8). First, its peak of the radial velocity (Figure 8a) was quite high (106.16 μm/s) and not centered, which indicated the polymer in pores was broken through by water injection. Second, its axial velocity distribution (Figure 8b) showed two peaks before and after the narrowest position (84.20 μm/s and 108.08 μm/s). The velocity distributions of five different time points during post-water flooding with sulfides in the throat (along Line 2 in Figure 5a) are shown in Figure 9, the results of which suggest all the velocity distributions obtained at these moments showed double peaks. Therefore, it can be concluded that the double-peak phenomenon is not occasional.

Regarding the flow field in the pores in the radial direction (along Line 3 in Figure 5a) and in the axial direction (alone Line 4 in Figure 5a), all the radial velocity distributions were close to the plug flow, while the post-water flooding with sulfide (maximum velocity 53.68 μm/s) was significantly higher than the others (Figure 8c). In terms of the axial velocity (Figure 8d), the post-water flooding velocity with sulfide showed a gradual decline of 19.42 μm/s within 132 μm, while the fluctuation of other cases was limited within 5.63 μm/s.

4. Discussion

4.1. Variation of Viscosity and Shear Thinning Induced by Sulfide Ion

The results of this study showed that divalent sulfide ions lead to a sharp decrease in the viscosity of the polymer solution, which is consistent with previous research [12]. In order to reveal the microscale reason of the viscosity loss, electron microscopy was used to show the destroyed microscale three-dimensional structure of the polymer with sulfide (see Figure 4). According to previous studies, sulfide ions break the molecular chain, weaken the molecular entanglement and crosslinking, and then reduce the apparent viscosity [29]. Meanwhile, due to the destroyed network structure by sulfide ions, the effect of the bond force was reduced [30]. In this way, the radius of the intermolecular hydrodynamics decreased and led to elastic loss [31]. Therefore, sulfide ions cause viscous and elastic losses by disrupting molecular structures and interactions.

4.2. Effects of Viscosity Loss Induced by Sulfide Ion on Flow Field

Compared with water flooding, the polymer without sulfide greatly reduced the overall flow velocity, as expected (Table 1), due to the high viscosity. The ratio of average velocity between the main stream and margin decreased from 2.00 to 0.89, indicating an improved sweep efficiency by the polymer [32].

For the polymer with sulfide, though its viscosity declined from about 60 mPas to 12 mPas, it was still more viscous than water. If only viscosity is considered, the polymer with sulfide should also result in a slower flow than water flooding. However, the actual velocity greatly increased with sulfide, leading to a quite higher ratio of the average velocity between the main stream and margin (11.2) than water flooding (2.0). This severe viscous fingering phenomenon indicated an even worse sweep efficiency than water flooding [9], although there was still residual viscosity with sulfide. Therefore, the viscosity loss alone could not explain this fingering phenomenon independently.

4.3. Effects of Elastic Loss Is the Key of the Special Flow Field by Sulfide Ion

In order to explain the fingering, the flow field was further analyzed in the throat and pores. In an observed throat during post-water flooding without sulfide, the overall flow rate was relatively low (see Figure 8a), which suggested that the pore networks were plugged by the polymers attached in the pores. However, the radial velocity peak with sulfide during post-water flooding was not centered (see Figure 8a), indicating that the polymer attached in the pore was partially broken through by injected water. Consistently, the shear thinning experiment (see Figure 3b) showed a weaker shear resistance of the polymer with 5 mg/L of sulfide due to the earlier declined (within low shear rates) viscosity than that without sulfide. Once the attached weaker polymer was broken through, the reduced effective flow radius led to a high flow rate under a constant injection rate. Therefore, the sulfide-induced fingering was the result of polymer adsorption and low shear resistance because of the destructed molecular conformation. In the study of polymer flooding at high salinity, the flow field also presented a finger-like heterogeneity [33], however, which failed to discuss the causes on a pore scale. Therefore, this study is the first report that discuss the flow mechanism of the sulfide-induced fingering mechanism.

Another abnormal phenomenon is the two velocity peaks before and after the throat during the post-water flooding with sulfide (see Figure 8b). In terms of Newtonian fluid (water flooding), the velocity should increase until the throat (the narrowest point), which results in only a peak at the throat. Regarding non-Newtonian fluids, experimental studies have reported that the velocity abnormally increased just before entering the narrowest channels, because partial kinetic energy is transformed into elastic energy [34]. In this study, the two peaks before and after the throat indicated twice energy conversions at the inflow and outflow, while the two-peaked phenomenon was not obvious in the polymer experiments without sulfide, which could be consistent with the first Newtonian zone within the low shear rates (see Figure 3b). The bimodal phenomenon of velocity distribution in this flow field has not been described in previous studies, and this paper is the first case.

Therefore, the influence of elasticity on the characteristics of the flow field is essential. Micro-PIV is feasible and necessary to understanding the pore-scale flow mechanisms.

5. Conclusions

In order to understand the sulfide-induced variations of polymer flooding mechanisms, multiple pore-scale flow fields were observed with Micro-PIV. The results showed that a severe viscous fingering occurred along the main stream of the micromodel. The ratio of the average velocity between the main stream and margin of post-water flooding with sulfide is five times higher than water flooding, which indicates severe fingering with sulfide. A further analysis at the pore scale combined with a shear-thinning test revealed the mechanism that is the shear resistance of the polymer attached in pores severely diminished by sulfide. For the polymer with 5-mg/L sulfide, the experiment showed that its viscosity is about 85% lower than that of the polymer without sulfide at a shear rate of 6 s⁻¹. Due to the weakened shear resistance, a fingering flow was observed. Additionally, a two-peaked velocity distribution appeared before and after the throat along the flow direction, which indicates elastic energy transformation at the abrupt radius changes. Therefore, in terms of the flow field in pore networks, elastic loss is more dominant than viscosity loss. This study proved that Micro-PIV is an essential and novel research method to understanding the characteristics of the flow field at the pore scale. Moreover, understanding sulfide induced variations of flow is of great significance to developing sulfide-tolerant agents.