NEW INSIGHTS INTO POLYMER RETENTION IN POROUS MEDIA

Transcription

1 NEW INSIGHTS INTO POLYMER RETENTION IN POROUS MEDIA By GUOYIN ZHANG A DISSERTATION SUBMITTED TO THE DEPARTMENT OF PETROLEUM ENGINEERING IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY NEW MEXICO INSTITUTE OF MINING AND TECHNOLOGY DECEMBER, 2013

2 DEDICATION Dedicated to my Mom, Huatang Xu, and my Dad, Xingshan Zhang.

3 ABSTRACT When water-soluble, high molecular weight polymers are used to reduce water mobility and improve volumetric sweep efficiency in enhanced oil recovery technology, polymer retention occurs when propagating through the reservoir. It is widely accepted that polymer retention comprises the adsorption on the rock surface, mechanical entrapment in small pores, and flow-induced hydrodynamic retention. Retention leads to polymer loss in reservoir and delay oil bank during a polymer flooding. Therefore, proper characterization of polymer retention in porous media is critical for a polymer flooding project. In this study, we first investigated the effect of concentration on HPAM retention because the literature showed controversial results on this issue. To accomplish this, both static adsorption on disaggregated sands and dynamic retention in sandpacks and sandstone cores were measured. Different retention behaviors were observed in dilute, semidilute and concentrated regions. In both dilute and concentrated regions, polymer retention is basically concentration-independent. In contrast, in the semidilute region, polymer retention shows concentration-dependent behavior. Little re-adsorption occurs at high concentration after the adsorbent s pre-contact with low concentration polymer. The

4 results also show macromolecule polymers have high adsorption tendency on rock surface and the adsorption can be considered almost instantaneous and irreversible. Based on experimental results, a concentration-related retention mechanism is proposed which correlates the orientation of adsorbed polymer molecules on rock surface with the interaction between molecular coils in solution. Next, hydrodynamic retention caused by the increase of hydrodynamic force acting upon polymer molecules was evaluated. As flow rate rises from 3.26 ft/day (base reference) to 6.52 ft/day and ft/day, retention of 500 ppm HPAM in 1.9 Darcy sandstone core increases by 13.2% and 39.16%, respectively. Hydrodynamic retention shows strong flow dependence. In low flow region, the retention increases abruptly with increased flow rate. By comparison, in high flow region, the increase becomes much more gradual. Our results also demonstrate that this flow-induced retention is totally reversible (no incremental irreversible retention) and residual resistance factor is not affected by this reversible retention. To prove polymer rheology in porous is an intrinsic property, not caused by retention related permeability reduction, xanthan polymer was also tested. The results indicate with increase of flow rate, both HPAM and xanthan retention goes up in a 71 md sandstone core. However, the resistance factor measurements show HPAM and xanthan show distinct rheology in porous media. At low flow rate, even with retention increase, HPAM shows Newtonian fluid behavior and no resistance factor increase is observed. In contrast, xanthan polymer exhibits shear thinning behavior. Therefore, the hydrodynamic retention has limited effect on polymer rheology in porous media.

5 Negative polymer inaccessible pore volume (IAPV) is observed with increase of flow rate and decrease of permeability. This presence of negative IAPV is caused by reversible polymer retention. Adsorptive retention on rock surface proves to be almost irreversible, therefore, the mechanical entrapment and hydrodynamic retention should account for this phenomenon. Keywords: Adsorption, mechanical entrapment, hydrodynamic retention, reversibility, inaccessible pore volume, core flooding

6 ACKNOWLEDGEMENTS There are many people who helped me during my study at New Mexico Institute of Mining and Technology. This work would have never been done without them. First, I would like to thank my advisor, Dr. Randall Seright for his persistent guidance, encouragement, patience and tolerance. His extraordinary knowledge, wisdom in science and engineering and attitude toward pursuit of truth always impress and inspire me. Next, I would like to thank my committee members, Dr. Thomas Engler, Dr. Mike Kelly and Dr. Reid Grigg for reading my dissertation and providing valuable comments and suggestions. I want to thank Kathryn Wavrik for her tremendous help in the lab. She taught me how to use rheometer, TOC analyzer, prepared cores and setup equipment for me. I learned a lot by working with her. I would also thank Dr. Jianjia Yu and James Mclemore. They offered me generous help whenever I got into trouble with my experiment. Thanks also go to Dr. Robert Lee, Dr. Ning Liu, Elizabeth Bustamante, Xu Han, and other PRRC staff. Finally, special thanks go to my wife, Ms. Dongling Xu and our daughter Yifan Zhang, who have been creating a loving and supportive family environment for me. ii

7 TABLE OF CONTENTS DEDICATION ABSTRACT ACKNOWLEDGEMENTS... ii TABLE OF CONTENTS... iii LIST OF FIGURES... v LIST OF TABLES... viii CHAPTER 1. INTRODUCTION Significance of Polymer Retention during Polymer Flooding Statement of Problem Approach Batch Adsorption or Static Retention Measurement Flow Experiment or Dynamic Retention Measurement Outlines of Dissertation CHAPTER 2. LITERATURE REVIEW Brief Introduction to Polymer Flooding Overview of Polymer Retention Mechanisms in Porous Media Mechanisms of Polymer Retention in Porous Media Polymer Inaccessible Pore Volume (IAPV) Factors Influencing Polymer Retention in Porous Media Langmuir Adsorption Isotherm Concluding Remarks CHAPTER 3. METHODS AND PROCEDURES Introduction Equipment and Material iii

8 3.3 Experimental Procedures Polymer Injection at Different Concentrations Polymer Injection at Different Flow Rates CHAPTER 4. RESULTS AND DISCUSSIONS Introduction Dependence of Retention on HPAM Concentration Static Measurements Dynamic Measurements Proposed Adsorption Model Overlap Concentration (C* and C**) Measurement Effect of Flow Rate on Polymer Retention Method Established to Detect Hydrodynamic Retention Hydrodynamic Retention in 1.9 Darcy Dundee Sandstone Core Is HPAM Shear Thickening Behavior Caused by Hydrodynamic Retention? Hydrodynamic Retention in 71 md Berea Sandstone Core Polymer Inaccessible Pore Volume (IAPV) Effect of Polymer Retention on Permeability Reduction Steady-State Flow in Porous Media CHAPTER 5. CONCLUSIONS Conclusions Discussions and Future Work NOMENCLATURE REFERENCES iv

9 LIST OF FIGURES Fig Polymer bank delay factors associated with polymer retention Fig Delay in oil recovery caused by retention... 4 Fig Polymer retention mechanisms in porous media (Szabo and Corp, 1975) Fig Effect of polymer retention and IAPV on polymer propagation Fig Typical Langmuir adsorption isotherm Fig Rheology of HPAM polymer in a viscometer Fig Viscosity vs. concentration at shear rate of 7.3 s Fig Sand shaker (IKA KS 4000) Fig Schematic diagram of polymer retention determination system Fig Roller for static measurement Fig Total Organic Carbon (TOC) analyzer for concentration determination Fig Correlation between TOC and polymer concentration Fig Polymer retention and inaccessible pore volume (IAPV) determination Fig p vs. Cp when HPAM flowing through a 10 m filter Fig Kinetics of polymer adsorption on sand Fig Desorption tests for 100-, 500-, and 1,000-ppm HPAM Fig Adsorption isotherm of HPAM using static method Fig Comparsion of retention on fresh sands and used sands v

10 Fig Adsorption isotherm using dynamic method (fresh sandpacks used for each case) Fig Retention determination for 50 ppm HPAM Fig Retention determination for 500 ppm HPAM Fig Retention of 1,000 ppm in pre-treated sandpack with 500 ppm Fig Effect of concentration on polymer retention, 347 md core Fig Retention determination for 80 ppm HPAM, 347 md core Fig Retention isotherm of HPAM in 71 md core Fig Retention determination for 20 ppm HPAM Fig Retention determination for 40 ppm HPAM Fig Polymer molecule interaction at different concentrations Fig Proposed polymer adsorption mechanism on the rock surface Fig Overlap concentration (C*) determination by linearity deviation Fig Overlap concentration (C*) determination by intrinsic viscosity Fig Polymer retention in the near wellbore region, radial flow Fig Mehod to determine hydrodynamic retention Fig Effect of flow rate on KI (tracer) retention Fig Polymer retention at flow rates from 3.26 ft/day to ft/day Fig Incremental retention of HPAM vs. flux Fig Determination of irreversible retention at ft/day Fig Determination of irreversible retention at ft/day Fig Resistance factor of HPAM at different flow rates Fig Rheology of HPAM and xanthan polymers in a viscometer vi

11 Fig Hydrodynamic retention of 150 ppm xanthan in 1.9 Darcy core Fig Resistance factor of xanthan at different flow rates Fig Hydrodynamic retention and resistance factor for HPAM Fig Hydrodynamic retention and resistance factor for xanthan Fig IAPV determination for 100 ppm HPAM, 60 ml/hr, 347 md core Fig IAPV determination for 20 ppm at flow rate of 1,000 ml/hr Fig IAPV determination for 20 ppm HPAM at flow rate of 60 ml/hr Fig nd injection of 160 ppm HPAM and tracer at 60 ml/hr, 71 md core Fig nd injection of 1,000 ppm HPAM and tracer at 240 ml/hr, 71 md core Fig Effect of polymer concentration on residual resistance factor, 347 md core. 94 Fig Effect of polymer concentration on residual resistance factor, 71 md core Fig Pressure drop during polymer injection in 347 md core Fig Pressure drop during polymer injection in 71 md core vii

12 LIST OF TABLES Table 2.1-Summary of Polymer Retention Using Dynamic Measurement Table 3.1-Core Properties Table 4.1-Dynamic Retention in Sandpacks Table 4.2-Retention on Sandstone Cores Table 4.3-Summary of Adsorption vs. Polymer Concentration Table 4.4-Properties of the Dundee Sandstone Core Table 4.5-Retention Summary viii

13 CHAPTER 1. INTRODUCTION 1.1 Significance of Polymer Retention during Polymer Flooding When water-soluble, high molecular weight polymers are used for enhanced oil recovery (EOR), polymer retention delays polymer propagation into the formation. The presence of the polymer is needed to provide high viscosity and low mobility levels which in turn are needed to improve oil displacement and sweep efficiency. Consequently, high polymer retention can substantially delay oil displacement and recovery. To illustrate this point, consider the range of polymer retention levels reported in the literature 9 to 700 g/g (Green and Willhite 1998) and the range of polymer concentrations used in polymer floods 500 to 3,000 ppm. Given the rock density ( rock, 2.65 g/cm 3 for quartz), porosity ( e.g., 0.3), polymer retention in g/g (R pret ), and polymer concentration in ppm (C poly ), Eq. 1.1 can be used to calculate the delay (PV ret, pore volume delay per pore volume injected). PVret rock (1 ) / Rpret / C...(1.1) poly Using this equation and the parameters mentioned above, Fig. 1.1 shows delay factors. With a very low retention level of 10 g/g and a polymer concentration of 2,000 ppm, the 1

14 delay factor is only about 3% of one pore volume (PV). In contrast, for a high retention of 500 g/g and a polymer concentration of 500 ppm, the delay factor is over 6 PV. For more typical values of 150 g/g for retention and a polymer concentration of 1,500 ppm, the delay factor is about 0.6. For this latter combination, a 20% difference in retention would mean an extra 12% PV polymer bank needed (if the retention is higher) or not needed (if the retention is lower) to accomplish a given objective. In one 40-acre 5-spot pattern with a height of 20 ft and a porosity of 0.3, 0.12 PV of 1,500-ppm HPAM (costing $1.5/lb) would represent a polymer cost of about $176,000. From another viewpoint, the mass of rock in the above 40-acre pattern is 40*43560*(12*2.54) 3 *2.65*(1-0.3)/0.3=6.10 x10 12 grams. Given the retention levels of 10, 50, 120, 150, 180, and 500 g/g, and an HPAM cost of $1.5/lb, the polymer costs required to satisfy the retention requirements of the rock would be $201,777, $1,008,883, $2,421,320, $3,026,650, $3,631,980, and $10,088,835, respectively. Of course, the delay in polymer propagation also delays oil recovery. Fig. 1.2 illustrates this point using fractional flow calculations (from For these calculations, we assumed oil viscosity of 1,000 cp, water viscosity of 1 cp, and the reservoir was initially at connate water saturation (S wr =0.3). The reservoir was then flooded with one PV of water (before continuous polymer flooding with 100-cp polymer), one homogeneous layer was present, flow was linear, and the following relative permeability curves (Eq. 1.2 and 1.3) were used. 2 krw 0.1 [( S w 0.3) / ( )]. (1.2) 2

15 Pore volume delay factor 2 kro 1 [(1 0.3 S w) / ( )]....(1.3) In Fig. 1.2, the term, IAPV, refers to inaccessible pore volume, which is defined as the fraction of the pore space being inaccessible to the large polymer molecules but accessible to the small solvent and salt molecules and ions. IAPV accelerates polymer propagation, whereas polymer retention (PV ret ) retards it. Three different levels were considered in Fig. 1.2 where retention plus IAPV (1) were perfectly balanced to cause no delay in polymer propagation (i.e., PV ret + IAPV=0), (2) caused a one PV delay (i.e., PV ret + IAPV=-1), and (3) caused a 2.5 PV delay (i.e., PV ret + IAPV=-2.5). Figure. 1.2 illustrates that the delay in the arrival of the oil bank is directly proportional to the delay in polymer propagation. Consequently, high polymer retention is economically detrimental because of increased cost for polymer and delayed oil recovery Porosity=0.3 rock=2.65 g/cm 3 IAPV= ppm 1000 ppm 1500 ppm 2000 ppm Polymer retention, µg/g Fig Polymer bank delay factors associated with polymer retention. 3

16 % of mobile oil recovered PV ret + IAPV = 0 PV ret + IAPV = PV ret + IAPV = -2.5 k rw = 0.1 [(S w -0.3)/( )] 2 k ro = 1 [(1-0.3-S w )/( )] 2 1,000 cp oil, 1 cp water, 100 cp polymer PV injected Fig Delay in oil recovery caused by retention. In this research, experimental studies are performed to yield some new insights into polymer retention in porous media. The effects of concentration, injection rate, and core permeability on polymer retention are investigated. The permeability reduction caused by polymer retention as well as retention reversibility is also analyzed based on experimental results. 1.2 Statement of Problem When enhanced oil recovery (EOR) polymers propagate through reservoir matrix, they tend to adsorb on the rock surface due to the affinity of polymer molecules for many reservoir rocks. In addition to the adsorption, polymer molecules tend to be trapped and accumulate in the small pores. The latter retention is commonly known as mechanical entrapment. Another retention, which is called hydrodynamic retention, may occur when flow rate suddenly increases. Retentions caused by different mechanisms show different 4

17 reversibility and permeability reduction behaviors. Many factors prove to influence polymer retention in porous media. Briefly, they can be divided into three categories. (1) Formation properties. These include rock permeability, mineralogy, clay content, salinity and ph of formation brine, and rock wettability, as well as reservoir temperature. Generally, polymer retention increases dramatically with decreasing permeability for less permeable rock (below several hundred md). If more clay is present, retention tends to increase because of increased surface area. Compared with permeability and clay content, other parameters may also affect polymer retention, but generally, they show minor impact (Smith 1970; Hirasaki and Pope 1974; Vela et al 1976; Espinasse and Siffert; 1979; Shah et al. 1985; Huang and Sobie 1993; Broseta et al. 1995). (2) Polymer and solvent types. Due to their molecule orientation in the solution or the distinct functional groups in the molecules, different types of polymers such as xanthan polymers, polyacrylamides (PAM), and partially hydrolyzed polyacrylamides (HPAM) show different retention behaviors under the same conditions (Chiappa et al 1999; Sydansk and Romero-Zeron 2011). Adsorption of polyacrylamide on silica sand decreases with increasing hydrolysis (Martin and Sherwood 1975). A study conducted by Meiste et al (1980) confirms that the increase of hydrolysis of polyacrylamide reduces retention on the negatively charged surface. Research from He et al (1990) indicates that smaller polymer molecules, instead of larger molecules tend to be preferentially adsorbed if a polymer mixture was injected simultaneously, resulting in a higher weightaveraged Mw for the early effluent than for the injected polymer. Studies from 5

18 both Koral et al (1957) and Stromberg et al (1959) show solvents also play an important role in polymer adsorption. About two to four times as much polymer was absorbed from poor solvent (high solubilization ability) as from the good solvent (low solubilization ability). (3) Besides these factors mentioned above, polymer flow rate should also be taken into account when investigating polymer retention in porous media. The increase of flow rate is accompanied with additional polymer retention in porous media. Extensive work has been done to describe the retention behaviors of the EOR polymers. Nevertheless, there are still some areas needed to be clarified. Among them, the effect of polymer concentration on retention is an outstanding one. Dawson and Lantz (1972) proposed that polymer retention in porous media follows the Langmuir isotherm without justification. Most of these researchers who claimed that polymer retention in porous media either fits the Langmuir isotherm or is strongly concentration-dependent arrived at their conclusions based on static adsorption measurements (Mungan 1969; Szabo and Corp 1975; Deng et al 2006). Zheng et al (1998) suggest their experimental data obtained from dynamic method using the same core fits the Langmuir isotherm. However, a careful examination of their data shows the highest retention at 1,500 ppm is less than 1.5 times higher than the lowest value at 250 ppm and no retention data was provided from concentrations below 250 ppm. Few researchers except Vela et al (1976), Shah et al (1978), and Szabo and Corp (1975) have tried to measure retention through dynamic measurement. Again, the data is very limited. Currently, the Langmuir adsorption model which is well-known for describing reversible adsorption is used in most chemical flooding simulators to describe polymer retention in porous media which 6

19 shows little reversibility (Satter et al. 1980, Vossoughi et al. 1984, Camilleri et al. 1987, Yuan et al. 2010, Dang et al. 2011). Next, we will focus on how the flow rate influences polymer retention in porous rock. Previous studies (Maerker 1973; Dominguez and Willhite 1976; Aubert and Tirrell 1980; Zaitoun and Kohler 1987; Huh et al 1990) reported that more polymer molecules would be retained with injection rate increase. However, no specific amount of retention is measured at increased flow rates. Chauveteau et al (1974) suggested that shear thickening behavior that was widely reported for HPAM solutions in porous media was caused by bridging adsorption. In our study, we will address this question: whether polymer rheology in porous media is an intrinsic property, or strongly affected by this flow-induced retention. We will also investigate the reversibility of polymer retention under various conditions and how polymer retention in porous media alters the rock permeability. Basically, the following issues will be addressed in this study: 1) Does polymer retention in porous media depend on polymer concentration? Or, does it follow the Langmuir isotherm? 2) How can quantify hydrodynamic retention be quantified for different rates? 3) Under what circumstances does polymer retention becomes more reversible? 4) Does hydrodynamic retention dominate polymer rheology in porous media? 5) How does polymer retention affect rock permeability? 1.3 Approach Two approaches were devised to measure polymer retention in porous rocks. One is called batch adsorption or static retention measurement and the other is called flow 7

20 experiment or dynamic retention measurement. They are chosen based on different scenarios Batch Adsorption or Static Retention Measurement. Batch adsorption is used to estimate polymer retention on disaggregated sand grains or in unconsolidated rocks with relatively high permeability where the adsorptive retention dominates. To determine the adsorption, sand grains with a particular size distribution are prepared by grinding sandstone cores. Next, a polymer solution with known concentration is contacted sufficiently with known mass of dry and fine sand grains. The system containing both sand grains and polymer solution will be thoroughly mixed. After the retention reaches equilibrium, the upper liquid phase is separated from the solids and sands are removed by centrifuging. Polymer concentration is determined by total organic carbon (TOC) analyzer. The amount of polymer adsorbed on the sand surface is calculated by mass balance Flow Experiment or Dynamic Retention Measurement. If consolidated porous media is used to determine polymer retention, the batch adsorption method is no longer applicable. This is because extra surface area will be generated during the fine grain preparation and the application of batch adsorption may introduce significant error. As a result, dynamic retention measurement that involves the injection of polymer solution through a porous media is used. Both polymer adsorption retention and mechanic entrapment retention can be measured this way. To date, it is still a challenging task to distinguish between these two types of retentions in porous media. 8

21 Several methods have been proposed to measure dynamic polymer retention in porous media (API RP , Dawson and Lantz 1972, Szabo 1975, 1979, Dominguez and Willhite 1977, Gupta and Trushenski 1978, Castagno et al. 1987, Huh et al. 1990, Mezzomo et al. 2002). Several of them advocate injection of a slug of polymer solution, followed by brine, and performance of a mass balance on the polymer (i.e., retention = polymer injected minus polymer produced). Key problems with this type of method are: (1) recovery of the polymer may require an extended period of brine injection because of the unfavorable displacement and (2) cumulative errors associated with measurements of low polymer concentrations in the produced fluid can introduce considerable uncertainty to the mass balance. We prefer the method used by Lotsch et al. (1985), Hughes et al. (1990), and Osterloh and Law (1998). In this method, two banks of polymer solution are injected which are separated by a brine slug. Polymer retention can be determined by the plot of the two effluent polymer concentration profiles versus pore volume injected. Hydrodynamic retention can also be measured this way by varying polymer injection rate. Another important parameter affecting polymer propagation in porous media is polymer inaccessible pore volume (IAPV). It is defined as the faction of pore space not contacted by polymer molecules due to a smaller inlet diameter compared to the size of polymer molecules. To estimate the inaccessible pore volume, KI tracer with concentration of 40 ppm is injected together with polymer solution and its concentration in the effluent is measured by an absorbance detector. The area between the second polymer and tracer breakout curves is used to estimate polymer inaccessible pore volume. 9

22 During dynamic retention measurement, pressure drops across the core will be recorded which are used to calculated resistance factor and residual resistance factor. Again, polymer rheology in porous and its dependence on hydrodynamic retention will also be addressed in our study. 1.4 Outlines of Dissertation Chapter 2 is the literature review. It will briefly introduce the concept of polymer flooding, polymer retention and the retention mechanisms in the porous media proposed by the researchers. Most importantly, in Chapter 2, the effect of polymer concentration, and flow rate on retention will be reviewed in detail. Chapter 3 describes the experimental equipment setup and testing procedures. As introduced earlier, polymer retention will be estimated using both static and dynamic methods. The measurement of polymer and tracer concentration in the effluent is a key part of this test. In Chapter 3, a new and convenient method to determine effluent polymer concentration is established. Chapter 4 deals with the experiment results and discussions. The effect of concentration, flow rate on HPAM retention, polymer reversibility and permeability reduction caused by retention will be included in this chapter. Besides these observations, a concentration-related retention mechanism is proposed that considers the orientation of the adsorbed polymer molecules and the interaction between molecular coils in solution. Chapter 5 summarizes this work and recommends areas needs further study. 10

23 CHAPTER 2. LITERATURE REVIEW In this chapter, a brief introduction to polymer flooding and the main role of polymer in improving sweep efficiency is provided, followed by the concept and mechanisms of polymer retention in porous media. Most importantly, the previous findings associated with the effect of concentration and flow rate on polymer retention will be reviewed. 2.1 Brief Introduction to Polymer Flooding Waterflooding is usually performed after the primary recovery during the development of a typical oil reservoir. However, due to the low viscosity of water or brine, viscous fingers form during water injection, resulting in an early breakthrough and poor sweep efficiency. To mitigate this unfavorable situation, a water-soluble, high molecular weight polymer is usually added to the water phase to increase its viscosity and thus reduce its mobility. Both biopolymers (e.g., xanthan) and synthetic polymers such as partially hydrolyzed polyacrylamide polymer (HPAM) have been tried. Currently, HPAM polymers are most widely used in polymer flooding due to their low cost, vast commercial availability, excellent viscosity-enhancing performance and resistance to microbial degradation. 11

24 Mobility ratio, M, the ratio of the displacing phase to displaced phase mobility, is the most important parameter for polymer flooding operation. D krw kro M ( ) / ( ) (2.1) d w o where, D is mobility of the displacing phase (water) and d is mobility of the displaced phase (crude oil). k rw and k ro are the relative permeability to water and oil, respectively. w and o refer to the water viscosity and oil viscosity. Based on the value of mobility ratio (M) relative to unity, the displacing process is considered to be either favorable, where M 1, or unfavorable, where M > 1. To attain a favorable mobility ratio (M) and to improve the sweep efficiency, increasing viscosity of the water phase is the most common way used. The most important mechanism of polymer flooding is its capability of improving volumetric sweep efficiency and conformance control, which can be attributed to viscosity-enhancing property of polymers. Polymer flooding is not expected to reduce residual oil saturation lower than waterflooding because the addition of polymer into the aqueous solution does not significantly change the interfacial tension between aqueous phase and oil phase. However, some researchers (Mohammad et al. 1992, Wang et al. 2001, Mojdeh, et al. 2008, Zhang et al. 2010, Urbissinova et al. 2010, Wang 2010) proposed that the viscoelasticity of polymer solution could improve the microscopic sweep efficiency after extensive pore volumes of water injection. So far, this is still a controversial subject. 12

25 Polymer flooding has been applied in the field on a substantial scale. For instance, in Daqing oilfield, China, it presently contributes to about one quarter of the annual oil production. There are 37 polymer flooding operations and about 9,000 wells involved as of 2005 and over 10 percent original oil in place (OOIP) has been recovered by conducting this technique (Liu et al. 2009). Early screening criteria indicated that polymer flooding should be applied in reservoirs with oil viscosity between 10 and 150 cp (Taber et al. 1977a, 1977b). However, with extensive use of horizontal wells and fracturing technology, polymer flooding also shows great potential for heavy oil recovery (Seright 2010, Delamaide et al. 2013). 2.2 Overview of Polymer Retention Mechanisms in Porous Media Among the factors influencing the performance of a polymer flooding, polymer retention is recognized as very important. Suppose severe retention occurs in the reservoir but it is not properly considered during the project design, it may cause polymer flooding to fail technologically and economically. In this section, three mechanisms on how polymer molecules tend to be retained in the porous media are first reviewed. When dealing with polymer retention, another parameter called inaccessible pore volume (IAPV) should not be avoided. IAPV describes the fraction of pore space that cannot be contacted by the injected polymers due to the small pore diameter relative to the polymer molecule size. The combination effect of IAPV and retention on polymer propagation in reservoir will be demonstrated. Finally, the review of effects of polymer concentration, flow rate, rock mineralogy on retention, 13

26 retention reversibility as well as the role of retained polymer molecules in altering permeability will be conducted Mechanisms of Polymer Retention in Porous Media Polymer retention primarily comprises three mechanisms. The first mechanism is caused by physical adsorption onto the pore surface. It is the result of the high affinity of polymers for many reservoir rocks, for example, due to van der Waal s and hydrogen bonding forces (Stutzmann and Sffert 1977, Pefferkorn et al. 1985, Shah et al. 1985, Sorbie 1991). This retention is believed to be basically irreversible and the amount of polymer adsorption is proportional to the surface area accessible to polymer molecules. The second retention is called mechanical entrapment, which happens when polymer molecules enter pores with smaller outlet diameter relative to the size of polymer molecule. Small molecules such as water and salt can travel through, but large polymer molecules will be trapped and accumulate in these small pores (Gogarty 1967; Szabo and Corp, 1975). The third retention is called hydrodynamic retention, which is associated with the local velocity of the polymer. After the retention reaches equilibrium, sudden increase of flow rate will cause extra polymer loss in the porous media. This flow-related hydrodynamic retention is believed to be reversible, i.e., when the flow rate is reduced or flow is completely stopped, the newly-retained polymer molecules will be released and migrate to the main flow channels (Maerker 1973, Dominguez and Willhite 1977, Huh et al. 1990). Maerker (1973) suggested that a significant pressure gradient causes polymer molecules to deform and become trapped within the core, particularly in relatively small 14

27 pores. Zitha et al (1998) and Chauveteau et al (2002) proposed a mechanism called bridging adsorption to explain hydrodynamic retention. In concept, polymer molecules may be stretched sufficiently in the elongational flow field during flow through a porous medium so that the molecules can span the distance over a pore constriction. If the ends of the molecules attach to the rock, a plugging or increased resistance to flow might develop. Chauveteau et al. (1974, 2002) suggested that the shear-thickening behavior that was widely reported for HPAM solutions in porous media was caused by bridging adsorption. Figure. 2.1 shows the polymer retention model proposed by Szabo and Corp (1975). In medium-permeability (several hundred millidarcies), high surface-area Berea cores, physical adsorption is more dominant than mechanical entrapment. By comparison, in low permeability rocks (several tens of millidarcies), mechanical entrapment is expected to increase. As shown by Fig. 2.1, adsorption dominates the retention in the main flow channel, while, mechanical entrapment occurs in the small pores with a pore throat inlet large enough for polymer molecules to enter but an outlet small enough to trap polymer molecules. In these pores, though restricted, a slow flow of brine is allowed. It also demonstrates the concept of polymer inaccessible pore volume (IAPV), i.e., pores with a small inlet that prevents the polymer penetration will be unreachable for polymer molecules. Mungan (1969) suggests that if only connate water-not oil-occupies these small and narrow channels, oil recovery by polymer flooding could be significantly improved. 15

28 Fig Polymer retention mechanisms in porous media (Szabo and Corp, 1975) Polymer Inaccessible Pore Volume (IAPV) As mentioned earlier, inaccessible pore volume (IAPV) plays a significant role in influencing polymer propagation in porous media. Results show that IAPV exists for EOR polymers (Shah et al. 1978, Vela et at. 1976, Liauh et al. 1978, Lotsch et al. 1985). The presence of IAPV theoretically accelerates the polymer propagation to be more than expected from the pore volume injected. On the other hand, polymer retention will retard polymer transportation in porous media. The combined effect of IAPV and retention of polymer propagation in the reservoir is illustrated in Fig. 2.2 (Dawson and Lantz 1972), assuming piston-like displacement. Case A: no retention and no IAPV. Polymer breaks through at precisely 1 PV; Case B: no retention, 0.25 PV IAPV. The polymer slug breaks through at 0.75 PV; PV; Case C: 0.25 PV retention and no IAPV present, polymer bank breaks through at

29 Case D: 0.2 PV retention, 0.25 PV IAPV. For this case, the polymer bank emerges at 0.95 PV. Undoubtedly, if both the retention and IAPV are 0.25 PV, the breakthrough of polymer solution still happens at 1 PV. C p /C PV A. No retention, no inaccessible pore volume. 1 C p /C PV B. No retention, inaccessible pore volume=

30 1 C p /C PV C PV retention, no inaccessible pore volume. 1 C p /C PV D. 0.2 PV retention, PV inaccessible pore volume=0.25. Fig Effect of polymer retention and IAPV on polymer propagation. (Modified from Dawson and Lantz, 1972). One might expect IAPV to increase with decreasing permeability. However, confusing results were reported in the literature. For instance, using Pusher 700 HPAM, Dawson and Lantz (1972) observed almost the same IAPV in 470 md Berea core (22%) as in 2090 md Bartlesville sandstone core (24%). Using Pusher 500 HPAM, Dabbous (1977) noted an IAPV value of 19% in 761 md Berea with no residual oil. In contrast, for the same polymer in Berea with a 28 35% residual oil, the permeability to water ranged from 49 to 61 md, and IAPV ranged from 17% to 37%. Osterloh and Law (1998) reported IAPV values up to 48% in sand packs with permeabilities up to 11 darcies. However, they acknowledged the experimental difficulties of accurately determining IAPV values. 18

31 2.2.3 Factors Influencing Polymer Retention in Porous Media Polymer retention in porous media has proven to be a very complicated process and many factors need to be taken into consideration when dealing with this problem. A broad range of research has been completed by earlier researchers to unveil polymer retention mechanisms. In this chapter, special efforts will be made to review the effects of polymer concentration, polymer injection rate, rock permeability on retention and retention reversibility because these are also the areas the author attempts to investigate Effect of Polymer Concentration on Retention. Numerous studies have been done to investigate the effect of polymer concentration on retention by researchers. However, most of them were completed using the static method. No systematic studies on this issue have been carried out using dynamic method. Mungan (1969) measured the co-polymer (containing approximately 25% polyacrylate and 75% polyacrylamide) retention on unconsolidated rocks using static approach. His results revealed that adsorption is higher on rocks with high specific surface area and it increases with polymer concentration. For instance, retention on Ottawa sand with BET equal to 0.50 m 2 /g is 340 g/g sand at polymer concentration of 1,000 ppm (in distilled water). For the same polymer solution, the retention increases to 680 g/g sand when Silica power with BET of 1.65 m 2 /g is used. In contrast to the retention at 1,000 ppm, retention decreases to 310 g/g sand at polymer concentration of 250 ppm for the same Silica sand. Dynamic measurement was performed for the consolidated Berea sand-packed core; the results indicate that the retention is only about 19

32 55 g/g rock at 500 ppm polymer concentration, which is much lower than the value obtained from the static method (610 g/g). Szabo and Corp (1975) studied retention of hydrolyzed polyacrylamide polymers on sand grains using the static method. They found that polymer retention showed almost linear dependency on polymer concentration, i.e., the polymer retention increases with increased concentration. But if sufficient brine or water (over 40 PV dilution) is used for sand soaking to remove the reversible adsorption, residual polymer retention depends only slightly on the initial polymer concentration. They ascribed this phenomenon to the partially reversible adsorption on the surface. Deng et al (2006) measured retention of three polyacrylamides (PAMs), cationic, nonionic and anionic polymers on clay minerals smectite, kaolinite and illite. Their results show that the adsorption isotherms of anionic PAM and nonionic PAM are L-type and could be fitted with Langmuir equations. Again, the amount of retention is obtained through static adsorption measurement. Retention isotherms from Chiappa et al (1998) show that polymer concentration plays different roles in polymer retention on pure quartzite. For cationic polyacrylamide (CAT), retention at high concentration could be several fold higher than that at low concentration. In contrast to CAT, the retention of 1% hydrolysis polyacrylamide (PAM) shows only slight concentration-dependence. Only a few papers mentioned the polymer retention results in porous media via dynamic measurement; see Table 2.1. Shah et al (1978) measured HPAM polymer retention in Berea core. The result indicates that retention increases from about 25 g/g 20

33 sand at concentration of 51 ppm, to 27 g/g sand at 500 ppm and 31.5 g/g sand at 1069 ppm. Therefore, polymer retention only increases a little bit with increased concentration. Vela et al (1976) measured the retention of 300 ppm and 600 ppm HPAM (5.5 million Mw, 20% hydrolysis) in porous media and the amounts of retention are almost independent of the concentration. As mentioned earlier, Zheng et al (1998) claimed Langmuir isotherm applied to their results; however, they only measured retention for 250, 750 and 1,500 ppm polymer solutions. In addition, the highest retention at 1,500 ppm is less than 1.5 times that from the lowest concentration (250 ppm). Szabo and Corp (1975) determined retention by injecting polymer solution into 1,200 md unconsolidated sand. Polymer retention increased from 3.50 g/g rock at 300 ppm to 6 g/g rock at 600 ppm. This concentration-dependent retention could be ascribed to the similar scenario of static measurement, since unconsolidated sand was used. Mungan (1969) only measured retention for 500 ppm copolymer flowing through porous media. Huang and Sobie (1993) found that retention of Scleroglucan in Bollotini packed columns increased from 8.21 g/g to g/g, increasing by 43% as concentration rose from 50 ppm to 200 ppm. Dawson and Lantz (1972) provided a curve to propose that Langmuir isotherm could be used to describe polymer retention in porous media, but no actual data measurement was performed. 21

34 Table 2.1-Summary of Polymer Retention Using Dynamic Measurement. References Porous Media Polymer Type Shah et al 1978 Zheng et al 1998 Vela et al 1976 Szabo and Corp 1975 Mungan 1969 Huang and Sobie 1993 Dawson and Lantz 1972 Berea core md Berea core 120 md 1,200 md sand (unconsolidated) Berea core Bollotini packed columns HPAM with 5 million Daltons Mw HPAM with 22 million Daltons Mw HPAM with 5.5 million Daltons Mw and 20% hydrolysis HPAM with 18-20% hydrolysis 25% polyacrylate+ 75% polyacrylamide, 3-10 million Mw Scleroglucan Polymer conc, ppm Polymer Retention, g/g rock Remarks Same core used Same core used Assuming =20%, g=2.65 g/cm S or = Propose Langmuir isotherm Results reported in different literature studies are comparable only if measurements were conducted under similar conditions. This is especially true when comparing results from static measurements with those from dynamic measurements. In this study, a series of tests were designed to clarify literature discrepancies concerning how polymer concentration affects retention in porous media. Several types of experiments were performed, including static measurements of polymer retention on fresh sand for each concentration case, and dynamic measurements of polymer retention in new sandpacks with similar permeability and porosity for different HPAM concentrations. We also examined polymer retention measurements where a single sand, sandpack, or sandstone core was exposed to successive solutions with increasing polymer concentration. HPAM 22

35 polymer solutions with a broad concentration range (from 10 to several thousand parts per million) were utilized Effect of Flow Rate As mentioned previously, the increase of flow rate will cause additional polymer loss in porous media by mechanical entrapment. These phenomena have been observed by many researchers (Maerker 1973, Dominguez and Willhite 1976, Aubert and Tirrell 1980, Zaitoun and Kohler 1987, Huh et al 1990). By monitoring polymer concentration profile in the effluent, they found when the retention equilibrium was reached at a low flow rate, subsequent increase in flow rate would render the effluent concentration lower than the concentration injected. Conversely, a decrease of flow rate will cause the effluent concentration to be higher than the injected concentration, which demonstrates that this flow-induced retention is reversible to some extent. An example is shown here. Using an 86 md core prepared by compressing Teflon powder, Dominguez and Willhite (1976) observed that the increase of interstitial velocity from 3.22 ft/day to 6.32 ft/day caused the effluent concentration to decrease from 391 ppm to 367 ppm. Because polymer adsorption on the Teflon powder was almost negligible, they believed the extra polymer loss in porous media due to flow rate variation should be attributed to mechanical entrapment. The results also revealed that a quantity of this flow-induced retention was reversible (reduction of the flow rate from 10.2 ft/day to 0.38 ft/day caused the normalized concentration to increase from 0.92 to 1.07). 23

36 By monitoring the mobility reduction (pressure drop during polymer injection divided by that during brine injection) or residual resistance factor instead of direct measurement of polymer retention, Zitha et al (2001) proposed a concept of pore-bridging adsorption. It could cause severe polymer loss at high flow rate if other conditions such as high polymer adsorption and low permeability are satisfied simultaneously, which leads to an unsteady-state flow (continuous build-up of injection pressure). Chauveteau et al (2002), Ogunberu and Asghari (2004) arrive at conclusions that if the shear rate (injection rate) is greater than the critical value ( > c ), the increased hydrodynamic force will push additional macromolecules into the already adsorbed polymer layer and increase both the density and thickness of the adsorbed layer. There is no doubt polymer retention is affected by flow rate, but very limited retention data is published to quantify this hydrodynamic retention for different flow rates. A correct description of the flow rate-retention relationship may be of great significance to the understanding of polymer propagation through reservoir because the polymer flow rate in reservoirs varies considerably with distance to the wellbore. Other questions related to this hydrodynamic retention are whether all or part of this retention is reversible and how does it affect polymer rheology and rock permeability? Our experiments conducted a thorough investigation on these issues Degree of Retention Reversibility Polymer reversibility is a controversial subject. For example, Szabo and Corp (1975) measured the residual polymer adsorption on the surface of sand grains and found it decreased with desorption time before it eventually reached a constant level. Therefore, 24

37 they concluded that retention adsorption on the rock surface is partially reversible. At the same time, they proposed that little or no desorption occurred in the area of mechanical entrapment in small pores. Ogunberu and Asghari (2004) found that the residual resistance factor determined after the polymer retention depended on the brine injection rate. After the injection of polymer solution at shear rate greater than 110 s -1, the residual resistance factor would decrease if the subsequent injection rate of brine was greater than 5.0 ml/min. However, at low flow rates, such as ml/min were applied, an increased residual resistance factor was encountered. Therefore, they arrived at conclusion that the polymer adsorbed on the rock surface could be weakened by brine injection. On the contrary, experimental results from Maerker (1973), Dominguez and Willhite (1976) suggested that the polymer retained in the form of mechanical entrapment proved to be reversible. Zaitoun and Kohler (1987) also showed that at higher clay content or lower permeability core, reversible polymer retention (mechanical entrapment in finest pore throats) occurred. Deng et al (2006) observed that after four consecutive washings with water, an accumulative of less than 3% of the adsorbed polymer (regardless of the charge type they have) was removed; therefore, the degree of reversibility of polymer adsorption on the rock surface was negligible. In our study, we investigated the reversibility of polymer adsorption on rock surface by determining residual adsorption after washing sands with adsorbed polymer molecules. For the dynamic case, if present, a negative IAPV could be a good indicator of retention reversibility. The variation of retention reversibility with the flow rate and rock 25

38 permeability was also studied in order to provide some ideas on the issues of which kind of retention, physical adsorption or mechanical entrapment, is more likely to be reversible Effect of Polymer Retention on Rock Permeability When the retained polymer molecules form an adsorption layer on the rock surface, the effective pore size is reduced, resulting in a decrease of rock permeability or increase of residual resistance factor. This phenomenon becomes more severe when the rock permeability decreases (Smith 1970, Vela et al. 1976, Seright 1992, Seright and Martin 1993). Hirasake and Pope (1974) proposed a model that correlated polymer adsorption on the pore surface with polymer molecular weight, water salinity, rock permeability, porosity, and flow rate. In their model, the adsorption of polymer is assumed to form a monolayer of polymer molecular coils with thickness approximately equal to the diameter of the molecular coil in that particular solvent based on previous findings (Rowland 1963, Rowland and Eirich 1966). This layer may be laterally compressed, resulting in an increase in segment density. The increase of segment density results in increased polymer loss due to adsorption, but will not affect the adsorbed layer thickness. Therefore, the permeability will not be further reduced. When polymer molecules adsorb on pore surface and form a thin layer, the effective pore size will be reduced, resulting in a reduced permeability. However, in the literature, few people have reported measuring how the resistance factor and the residual resistance factor vary with hydrodynamic retention. In our study, by measuring residual resistance factors after different concentration injection with same injection rate and also the same 26

39 polymer solution with different injection rates, we addressed the question whether the rock permeability varies dramatically when remarkable hydrodynamic retention occurs Other Factors Influencing Polymer Retention Other factors influencing polymer retention in porous media have also been studied. Smith (1970) found that the adsorption of HPAM varies from one type of mineral to another. For instance, the retention on calcium carbonate is five times higher than that on silica, which shows calcium carbonate appears to have a much greater affinity for polymer than silica. This increased retention is ascribed to the high content of calcium ions on the surface, which may provide calcium bridges to enhance polymer retention. They also found polymer adsorption increases with salt concentration. The amount of polymer retained increases from about 11 g/m 2 at 1% NaCl to 60 g/m 2 at 10% NaCl. Broseta et al found that in oil-wet porous media, polymer retention (PAM) in the presence of residual oil saturation (S or = ) will decrease considerably by factors ranging from 2 to 5 compared to the retention when the core is 100% water-saturated. But in water-wet porous media, the influence of residual oil saturation is less noticeable. For example, the retention with S or =0.2 is 7.5 g/g sand compared with the retention of 10 g/g sand at S or =0. They suggest the variation in polymer retention is due to the change of interfaces accessible to the polymer under these conditions. Chiappa et al (1998) investigated the role of electrostatic interactions in polymer adsorption. They tested polymers with different charges (cationic, anionic and weakly 27

40 anionic) on quartzite, which is negatively charged. Effects of clay content with a high specific surface area, ion strength and composition were also studied. Their results showed that polymer adsorption is dominated by electrostatic interactions between the charged groups that present at the polymer/brine and rock/brine interfaces. A correct match between the polymer and the surface charges can greatly increase adsorption. For example, adsorption on negatively-charged quartzite increased from 270 to 340 and 610 g/g sand when polymer of HPAM (anionic), PAM (weakly anionic) and CAT (cationic) were used, respectively. Because of the high surface area and predominately negative charge of the clay mineral, a small amount of clay can cause a significant increase in polymer retention. The retention of cationic polymer increased from 610 to 1.45*10 4, 1.8*10 5 g/g sand when the porous media was switched from pure quartzite to 8% clay quartzite with 8% clay and 100% clay. Again, the presence of divalent cation (as Ca 2+ ) can greatly enhance the adsorption of negatively changed polymers (HPAM and PAM) onto quartzite. For instance, HPAM retention of approximately 80, 340 and 800 g/g sand was determined on the pure quartzite in the system containing 0, 2% and 8% CaCl 2. Chiappa et al also suggest that the present of divalent calcium ions can enhance the adsorption of negatively charged polymers by forming an ion bridge. Efforts were also made to distinguish the adsorptive retention from the mechanical entrapment retention by Cohen and Christ (1986). In their study, they used HPAM with an estimated molecular weight of 5.5 million and degree of hydrolysis of 25%. Two kinds of packed silica sand beds were applied as the porous media. One was an adsorbing material and the other was a non-adsorbing material generated by the chemical modification of a siliceous surface. Their results showed that adsorption accounted for 28

41 about 35.2% of the total polymer retention and the remaining 64.8% was attributed to mechanical entrapment. 2.3 Langmuir Adsorption Isotherm Some researchers proposed that retention of EOR polymers on reservoir rocks depends on polymer concentration, or the Langmuir adsorption model applies. Therefore, it is necessary to make a brief introduction to this adsorption model. Equation 2.1 describes the Langmuir adsorption isotherm which shows the solute adsorption on the substrate surface is a function of solute concentration. a1b 1C 1 bc where, is solute adsorption. C is solute concentration in solution and a 1, b 1 are constants. 29

42 Adsorption, g/g rock a 1 =25 b 1 = Concentration, ppm Fig Typical Langmuir adsorption isotherm. For adsorption fits to the Langmuir isotherm model, constants of a 1 and b 1 can be determined graphically. Plotting 1/ versus 1/C on a linear scale ends up with a straight line. The slope of this line is 1/a 1 b 1 and it intercepts with y-axis at 1/a 1. Fig. 2.3 shows a typical Langmuir adsorption isotherm where a 1 and b 1 are assumed to be 25 and 0.02, respectively. For the Langmuir adsorption model, the adsorption depends strongly on the concentration. Especially in the low concentration range, the adsorption increases linearly with the concentration. When concentration approaches zero, the adsorption is also decreasing to zero. 30

43 2.4 Concluding Remarks The literature review shows that retention is a very complex process. Though several mechanisms are proposed to elucidate this phenomenon, many disagreements still exist. For instance, what role does polymer concentration plays in polymer retention? Besides the claim made by Dawson and Lantz (1972) without actual retention measurement, some researchers (Mungan 1969, Szabo and Corp 1975, Deng et al 2006) found that polymer retention is a function of polymer concentration; or, the retention follows the Langmuir isotherm based on the static measurement. However, a careful analysis of limited experimental results from dynamic measurement (Szabo and Corp 1975, Vela et al 1976, Shah et al 1978, Zheng et al 1998) shows that retention is almost independent of polymer concentration. A systemic study on the effect of concentration on retention using dynamic method is highly recommended. Researchers (Maerker 1973, Dominguez and Willhite 1976, Aubert and Tirrell 1980, Zaitoun and Kohler 1987, Huh et al 1990) observed flow-induced, or hydrodynamic, retention by monitoring polymer concentration in the effluent. Nevertheless, few studies found in the literature quantified retention for different rates. This may be of great importance if retention is highly velocity-dependent because the flow velocity in the reservoir varies considerably as the invasion radius changes. Regarding to the degree of retention reversibility, results from Szabo and Corp (1975) and Ogunberu and Asghari (2004) suggest polymer adsorption on the rock surface shows partially reversible behavior. On the contrary, Deng et al (2006) conclude that the reversibility of adsorption is almost negligible. Results from Dominguez and Willhite (1976) Zaitoun and Kohler (1987) indicate polymer molecules retained in the form of 31

44 mechanical entrapment proves to be reversible. In our study, we will investigate the reversibility of polymer retention on rock surface and in porous media. This may provide some ideas on the issue of which kind of retention, physical adsorption or mechanical entrapment, is more likely to be reversible. Polymer retention on the rock surface forms an adsorption layer that reduces effective pore size and causes permeability reduction. Some researchers propose that the increase of flow rate may either induce pore-bridge adsorption,which results in an unsteady-state flow (continuous injection build-up) (Zitha et al 1998 and 2001) or increases both the density and thickness of the adsorbed layer (Chauveteau et al 2002, Ogunberu and Asghari 2004). If it is true for either case mentioned above, the polymer resistance factor/residual resistance will be dramatically increased. In our research, we probed if severe pore-bridge adsorption occurred in our tests by recording the polymer injection pressure under various conditions. We also focused on the relationship of hydrodynamic retention and permeability reduction. In summary, to better understand the retention behaviors of HPAM polymers in porous media, as mentioned previously, the following issues were addressed in our study: 1) Does polymer retention in porous media depend on polymer concentration? Or, does it follow the Langmuir isotherm? 2) How should quantify hydrodynamic retention be quantified for different rates? 3) Under what circumstances, does polymer retention becomes reversible? 4) Does hydrodynamic retention dominate polymer rheology in porous media? 5) How does polymer retention affect rock permeability? 32

45 CHAPTER 3. METHODS AND PROCEDURES 3.1 Introduction In this section, the materials, experimental equipment, and procedures will be introduced. Again, both static and dynamic measurements will be used to determine polymer retention in porous media. 3.2 Equipment and Material Polymer and Brine. A partially hydrolyzed polyacrylamide (HPAM) (SNF Flopaam 3230S) and a xanthan polymer (Kelco Oil Field Group) were used in our tests. Both were provided by the manufacturer as white granular powders. HPAM is estimated to have a molecular weight of 6 8 million daltons and degree of hydrolysis of approximately 30%. HPAM solution was prepared using the magnetic stirrer vortex method. Xanthan solution was prepared using a blender. After the polymer solutions were preparation, they were filtered through a 10 m filter to remove any possible microgels and other debris present in the solution. The purpose of this filtration is to minimize the face plugging effect caused by these impurities. Studies show that during polymer injection, presence of this 33

46 debris and microgel may plug the core face by forming external filter cake (Seright et al. 2009). Two kinds of brine were used. One was 2% NaCl for the static measurements, dynamic retention in sandpacks, and polymer hydrodynamic retention measurements in consolidated cores. The other brine containing 2.52% TDS (2.3% NaCl and 0.22% NaHCO 3 ) was used when dynamic retentions were measured in consolidated sandstone cores. Both brines were filtered through m filters before application. The rheology of HPAM polymer was determined in an Anton Paar rheometer (Xanthan rheology will be shown in Chapter 4). As shown in Fig. 3.1, at concentration below 320 ppm, it behaves like a Newtonian fluid within the broad range of shear rates between 1 to 1,000 s -1, i.e., polymer viscosity is almost independent of shear rate. Polymer solutions with concentration of 640 and 1,000 ppm, at shear rate less than 10 s -1, they show Newtonian behavior. For shear rate greater than 10 s -1, they show slightly shear thinning. No shear thickening behavior is observed in a viscometer. The correlation of viscosity at shear rate of 7.3 s -1 with polymer concentration is shown by Fig The viscosity of 20 ppm is about 1 cp, increasing to 3.8 cp at concentration of 1,000 ppm. Tracer. 40 ppm Potassium iodide (KI) was added into the polymer solution as a tracer. Its concentration in the effluent was monitored by a Tunable Absorbance Detector (Waters 486) as an indicator of brine propagation through porous rock. 34

47 Viscosity, cp Viscosity, cp ppm 80 ppm 320 ppm 640 ppm 1,000 ppm HPAM 6-8 million MW 30% hydrolysis 2.52% TDS 25 C Shear rate, 1/s Fig Rheology of HPAM polymer in a viscometer HPAM 6-8 million MW 30% hydrolysis 2.52% TDS 25 C Polymer concentration, ppm Fig Viscosity vs. concentration at shear rate of 7.3 s

48 Sand Preparation. Sand grains with particle sizes between m were prepared as the adsorbent by crushing and sieving Berea sandstone core cuttings. To reduce the presence of very fine particles and carbons released by the sands, special processes were undertaken for the treatment of these disaggregated sands. First, sands were put into a bottle with brine and rotated at 300 rpm for 8 hours on an IKA KS 4000 shaker, (Fig. 3.3). The purpose of mixing sands with brine is to minimize the release of carbon with the sands themselves. Then, the upper mud-like phase was separated from the sand. Next, the sand was washed with distilled water to remove newly-generated fine particles and residual salt until the upper water phase was totally clear. Finally, the sand was dried at 110 C. Fig Sand shaker (IKA KS 4000). Porous Media. Disaggregated sands prepared as described above were used for static adsorption measurements. To determine dynamic and hydrodynamic polymer retention in 36

49 porous media, both consolidated sandstone cores and high permeability sandpacks were used. Four consolidated sandstone cores were used in our tests, among them, three rectangular Dundee cores and one cylindrical Berea sandstone core. Dundee sandstone cores cast in epoxy resin (Core #1, #2 and #3) have permeability of 347 md, 449 md, and 1.9 darcies respectively. The fourth core (Core #3) is a Berea sandstone core with a permeability of 71 md. It is a cylindrical core with a section area of 11.4 cm 2. This core was cast in the metal before being assembled in the Hassler-type core holder. All cores were 15 cm long except core #4, which is 12.8 cm long. Two internal pressure taps divided the Dundee cores into three sections with lengths of 2.5 cm, 10 cm, and 2.5 cm. For the Berea core, these three sections were 1.5 cm, 10 cm, and 1.5 cm in length. Sandpacks with high permeability and porosity were prepared from the same sands used for static measurements. Sandpacks were 6.35 cm long and 14.5 cm 2 in cross section. Table 3.1-Core Properties. Core No. L, cm A, cm 2 PV, ml, % k, md Note Dundee , Berea Experimental Setup. Figure. 3.4 shows the schematic diagram of the experimental unit for determining polymer retention in porous media. It can be divided into three major sections based on their functionalities: polymer injection, tracer concentration determination and effluent polymer concentration determination. All these three parts were assembled in series. 37

50 1) The first part deals with polymer injection into the core. The polymer flow rate can be accurately controlled by an ISCO syringe pump (Model 500HP). The pressure drop across the core is indicated by a Honeywell pressure transducer, which is connected to the two internal pressure taps on the core. 2) The second part is setup to determine tracer concentration in the effluent. In this part, the fluid exiting the core outlet flows through the absorbance detector (Waters 486) and the concentration of KI can be measured versus pore volume injected at light wavelength of 232 nm. Note that a 7 m Swagelok metal filter is attached in the flow line between the core and the absorbance detector to prevent any large particulates from flowing downstream. 3) The third part is used for the determination of polymer concentration in the effluent. As shown in Fig. 3.4, the effluent leaving the absorbance detector is first collected in a container and then fills the small accumulator beneath the container via the second ISCO pump every 10 or 20 minutes. Then, the fluid is forced to flow through a 10 m Millipore filter combination at a constant flow rate (controlled by another ISCO pump) and the pressure drop across the filter is recorded. For HPAM polymer concentration higher than 150 ppm, the filter combination can be connected directly to the core. 38

51 Fig Schematic diagram of polymer retention determination system. Note: 1. ISCO syringe pump #1 (Model 500HP); 2. Core; 3. Pressure transducer #1; 4. 7 m Swagelok filter; 5. Absorbance detector (Waters 486); 6. Beaker; m Millipore filter combination; 8. Pressure transducer #2; 9. Fluid container; 10. Accumulator; 11. ISCO syringe pump #2 (Model 500D). 3.3 Experimental Procedures Static Measurement. After mixing known concentration solution and known mass sands, bottles containing both sand and polymer solution were tied on a roller which rotate at a speed of 6 RPM for 1 hr to complete the adsorption process. One hour contact was considered to be sufficient because polymer adsorption on the rock surface is believed to be instantaneous (the adsorption kinetic will be discussed in Chapter 4). To reduce the effect of carbon released by sands themselves, a blank sample only containing sands and brine was prepared. 1) Polymer solutions with known concentration were prepared in 2% NaCl brine; 2) Known amounts of sand grains were added into the polymer solutions; 39

52 3) The bottle containing both polymer solution and sand grains was mounted on a roller (Fig. 3.5). The system was rotated at 6 rpm for 1 hr; 4) After the rotation, the upper polymer solution was transferred to a plastic tube to be centrifuged at speed of 3,000 rpm for about 1 hr to separate the residual polymer solution from sand particles; 5) Equilibrium polymer concentration was determined by Total Organic Carbon (TOC) Analyzer. Dilution was needed for high concentration cases; 6) Polymer adsorption for each concentration case is calculated: ( C C ) V / W 0 eq p sg where, is polymer adsorption, g/g sand, C 0 and C eq are initial and equilibrium polymer concentrations, ppm. V p is polymer volume, ml. W sg is the weight of sand grains, g. Both polymer and brine density were assumed to be 1 g/ml. Fig Roller for static measurement. 40

53 Determination of Polymer Concentration. A TOC analyzer (Shimaduz Model TOC VCSH, Fig. 3.6) was used to determine polymer concentration before and after adsorption. This is based on the excellent linearity between total organic carbon content and polymer concentration. As seen from Fig. 3.7, a very good correlation exists between these two parameters. To reduce systematic error, samples with concentration higher than 200 ppm were diluted with the same brine for polymer preparation before examination. The reason for diluting high polymer concentration solutions is because a range of 0 to 200 ppm TOC calibration curve was used. Fig Total Organic Carbon (TOC) analyzer for concentration determination. 41

54 Polymer concentration, ppm y = x R² = TOC, ppm Fig Correlation between TOC and polymer concentration. Dynamic Measurement. The detailed procedures of this method are described below and the determinations of polymer retention and IAPV are shown graphically by Fig. 3.8: 1) Core was initially saturated with degassed brine after evacuation and then the permeability to water was determined. Based on the amount of brine saturated and the core dimension, total pore volume (PV) and porosity were obtained. 2) Polymer solution was injected with KI tracer until polymer concentration C p in the effluent achieved the injected concentration C 0. Polymer concentration and tracer concentration (relative to the injected concentrations, C p /C 0 ) were plotted versus pore volumes injected. 42

55 3) Core was flushed with brine until polymer was not detectable in the effluent. In our tests, 60 to100 PV of brine were injected with intervening periods of no flow. 4) A second polymer and tracer bank were injected with the same concentration until C p achieved the injected concentration C 0. Again polymer and tracer concentrations (C p /C 0 ) were plotted versus pore volumes injected. 5) Retention is given by the area between the two plots of polymer concentration breakout curves. 6) IAPV is given by the area between the second polymer and tracer concentration breakout curves. Fig Polymer retention and inaccessible pore volume (IAPV) determination. 43

56 Note: all the tests are conducted at room temperature around 25 C. Unless otherwise specified, polymer solution is injected at a flow rate of 60 ml/hr which is about 1 meter/day or 3.3 ft/day. Attention should be paid when this method is utilized to determine polymer retention and IAPV. Firstly, any reversibly retained polymer should be flushed from the core during the extensive brine injection. The reoccurrence of this retention during the second polymer injection will make the effluent concentration profile shift closer to the first one. As a consequence, reversible retention is excluded or only irreversible retention is measured via this dynamic method. Secondly, IAPV measurement will be affected if substantial reversible retention occurs. Both of these two situations were encountered in our tests and will be discussed in the later sections. Determination of Polymer Concentration in the Effluent. How to accurately monitor polymer concentration in the effluent is very important for this test. After some experimentation, we established a rheological method based on shear-thickening behaviors of HPAM polymer flowing through porous media. HPAM polymers show Newtonian behavior at low flow rate in the porous rock, while, at moderate to high flux, they become definitely shear thickening, i.e., the resistance factor increases with increased flux. Results from Seright et al. (2011) show that shear thickening behavior of HPAM solution with a concentration even as low as 25 ppm is evident in porous media. 44

57 Pressure drop, psi 10 8 y = x R² = HPAM 6-8 million Daltons Mw 30% hydrolysis 2.52% TDS 40 ppm KI 300 ml/hr 10 micron filter 25 C Polymer concentration, ppm Fig p vs. Cp when HPAM flowing through a 10 m filter. Based on this finding, we established a filter combination that uses a Millipore AP10 TM filter pad upstream of a 10 m Sterlitech membrane filter to mimic porous media. When polymer solutions flow through this filter combination at constant flow rate, pressure drops given by the Honeywell pressure transducer correlate well with polymer concentrations. Fig. 3.9 is one standard curve that shows the sound linear relationship between polymer concentration and pressure drop. 3.4 Polymer Injection at Different Concentrations To investigate the effect of polymer concentration on retention, a series of polymer solutions with concentration from 20 ppm to 1,000 ppm were injected into the Dundee sandstone core. As described previously, for one specific concentration, two identical polymer solution banks with tracer were injected which were separated by a brine slug 45

58 injection. The tracer and polymer concentrations in the effluent were determined by Absorbance Detector and Millipore filter combination. By plotting C p /C 0 versus pore volumes injected for the first and second polymer slugs and second tracer slug, both polymer retention and inaccessible pore volume can be determined. Again, polymer retention is given by the area between the two polymer concentration curves and IAPV is given by the area between the second tracer and second polymer concentration curves. This process was repeated for the next concentration case. The tests were run in the sequence of low concentration to high concentration. All the tests were performed under room temperature, which is approximately 25 C. 3.5 Polymer Injection at Different Flow Rates. Hydrodynamic retention was measured by injecting polymer solutions in core #3 and #4 at various flow rates. Again, two polymer banks were injected separated by large pore volumes of brine injection, and effluent polymer concentrations were recorded. To determine whether this flow-related retention significantly affects polymer rheology in porous media, both HPAM and xanthan polymers were tested. Polymer pressure drops across the core during polymer injection and subsequent brine injection for these cases were recorded. These pressure drops were used to calculate polymer resistance factors and residual resistance factors. 46

59 CHAPTER 4. RESULTS AND DISCUSSIONS 4.1 Introduction This chapter deals with the experimental results and discussions. Mainly, two sections are included here. First, is an account of the investigation of the dependence of retention on polymer concentration. Porous media, such as disaggregated sands, high permeability sandpacks and low permeability sandstone cores were used. Based on the observed retention behaviors, a concentration-related retention model is proposed. The second part of this chapter focuses on hydrodynamic retention of both HPAM and xanthan polymers. A method is established to show, as flow rate increases, the amount of total incremental retention and how to distinguish irreversible and reversible retention. To clarify whether this flow-induced retention has strong impact on polymer rheology in porous media, the rheology and retention of both HPAM and xanthan polymer in porous media were examined. The evaluation of reversibility of retention under different flow rates and rock permeabilities and the effect of reversibility of polymer retention on IAPV is also discussed. In this section, we will also address the question of how permeability 47

60 reduction (residual resistance factor) is affected by polymer retention and under what circumstances, the increase of polymer retention will reduce rock permeability. 4.2 Dependence of Retention on HPAM Concentration Both static and dynamic measurements were used to estimate the impact of concentration on retention. Porous media employed include disaggregated sand grains, high permeability sandpacks and low permeability consolidated sandstone cores. Furthermore, retention in pre-contacted porous media was also studied before the proposition of a concentration-related adsorption model. Here, the term pre-contacted porous media refers to the porous media whose retention is previously satisfied at low concentration Static Measurements Adsorption Kinetics. The kinetics of polymer adsorption were first explored by mixing 100-ppm polymer solution with disaggregated sands. The bottle containing both polymer solution and sands was mounted to a roller and rotated at 6 RPM for 20 hrs. Liquid samples were taken periodically from the upper phase for polymer concentration determination using TOC analyzer. The adsorption calculated based on mass balance was plotted as a function of time. As shown by Fig. 4.1, adsorption reached the maximum plateau (about 40 g/g sand) within about three minutes and then leveled off. This indicates that the long chain HPAM molecule shows a very high adsorption tendency on rock surface and that the adsorption can be completed instantaneously. In our tests, the contact of polymer solution and sands lasted 1 hr to ensure adsorption equilibrium. 48

61 Adsorption, g/g sand HPAM 6-8 million Daltons Mw µm sands 100 ppm Rotated at 6 rpm 2% NaCl 25 C Time, min Fig Kinetics of polymer adsorption on sand. Desorption Test. After the completion of adsorption, desorption tests were carried out to estimate the amount of polymer that can be removed from the surface. In this test, excess polymer solution was decanted from the top of the sand after sufficient contact. Then fresh brine was added and again the bottle containing both sand and brine was rotated at 6 RPM for 1 hr. After the sands settled, the upper phase was sampled for polymer concentration determination. The residual polymer adsorption was calculated using mass balance. This procedure was repeated until no more desorbed polymer was detected. Figure. 4.2 shows the results for 100-, 500-, and 1,000-ppm HPAM. Calculations show that the percentage of the reversible adsorption for these three cases was 6.6%, 2.4%, and 2.9%, respectively. This result was similar to that from Chauvetear and Kohler (1974), Deng et al. (2006). Because EOR polymers have high molecular weights and extended chains, many polar groups along the polymer chain will attach to many different polar points on the rock surface. It is statistically very unlikely that a polymer molecule would 49

62 Residual adsorption, g/g sand release all points of attachment at the same time. Therefore, polymer adsorption on the sand surface can be treated as almost irreversible ppm 500 ppm 1000 ppm Brine/sand ratio, (weight/weight) Fig Desorption tests for 100-, 500-, and 1,000-ppm HPAM. Effect of Polymer Concentration. To investigate the effect of polymer concentration, retention from polymer solutions with concentrations from 10 ppm through 6,000 ppm was examined. The results are illustrated in Fig. 4.3, which suggests three distinct concentration-related retention behaviors. First, in the very low concentration region (from 10 ppm to about 100 ppm), polymer retention stabilized approximately at a value of 20 g/g. In the intermediate-concentration region (from 100 ppm to about 4,000 ppm), polymer retention increased from 35 to 420 g/g, increasing almost linearly with polymer concentration. In the very high concentration region (above 4,000 ppm), nearly constant retention (~ 420 g/g) was achieved. 50

63 Adsorption, g/g sand These results (especially the concentration-dependent observation) agree with the previous findings where most of the measurements were made in the intermediate concentration region (Mungan 1969; Espinasse and Siffert 1979). Our findings indicate that polymer retention does not fit the Langmuir isotherm, which is commonly used to describe the reversible adsorption of small molecules such as surfactants and gas. For EOR polymers with high molecular weights and extended chains, their adsorption on rock shows little reversibility (see Fig. 4.2). It is postulated that at very low concentration, polymer molecules continue to be adsorbed until the maximum coverage is reached. During this process, few adsorbed polymer molecules are likely to detach from the surface. Therefore, unlike the adsorption described by the Langmuir isotherm, polymer adsorption at very low concentration approaches a constant non-zero value m sand grains Rotated at 6 rpm for 1 hr 2% NaCl 25 C Polymer concentration, ppm Fig Adsorption isotherm of HPAM using static method. Re-Adsorption Test. Fresh sands were used for each case to generate the adsorption isotherm shown in Fig. 4.3, which illustrates the concentration-related adsorption 51

64 Adsorption, g/g sand behaviors. After the desorption tests described in Fig. 4.2, 1,000-ppm polymer solution was added to the sands previously contacted with 100-ppm and 500-ppm polymer solution to check if polymer re-adsorption occurred. The results (Fig. 4.4) show little additional polymer was adsorbed onto the used sands. For instance, for the 100-ppm concentration case, the retention increased from 32.4 to 35.8 g/g, increasing by 10.3%. For the 500-ppm case, retention rose from to g/g merely a 6.1% increase. Compared to the adsorption of 243 g/g at 1,000 ppm, a substantial retention difference existed between the fresh sands and pre-contacted sands whose retention was previously satisfied by low-concentration polymer. Apparently, even though adsorption was relatively small at low concentration, the surface was already fully covered by adsorbed polymer molecules, and no vacant sites were available for further attachment µm sand grains Rotated at 6 rpm for 1 hr 2% NaCl 25 C Adsorption at 100 ppm Re-adsorption at 1,000 ppm Adsorption at 500 ppm Re-adsorption at 1,000 ppm Adsorption at 1,000 ppm Fig Comparsion of retention on fresh sands and used sands. 52

65 4.2.2 Dynamic Measurements Retention in Sandpacks. Dynamic measurements were performed in sandpacks made from the same sand source that was used for static measurements. Sandpacks were constructed by tapping sands into a cylindrical Teflon-made container with both ends sealed by O-rings. A pair of stainless steel meshes with pore size of 15 microns was placed on both ends of the container to prevent any downstream clogging and in OD tubing was welded onto end caps. Procedures were followed as described in Chapter 3 for dynamic measurement. For each concentration case, a new sandpack was used. As shown in Table 4.1, these sandpacks had very similar properties. Permeability ranged from 4.69 to 5.51 darcies, and porosity ranged from 0.43 to Due to the high sandpack permeability, it was suggested that adsorptive retention dominated the retention for these cases (Szabo and Corp 1975, Huh et al. 1990). Polymer solutions with concentrations of 20, 50, 100, 500, 1,000 and 2,000 ppm were investigated at an injection rate of 120 cm 3 /hr (6.6 ft/day flux). The results were shown in Table 4.1 and Fig Retention was around 5 g/g at low concentrations from 20 ppm to 100 ppm. With the increase of concentration from 100 ppm to 2,000 ppm, retention increased from 5.71 to 27.8 g/g increasing by a factor of nearly 5. After completion of measurements for 100- and 500- ppm cases (Sandpacks #4 and #3 in Table 4.1), a 1,000-ppm solution was injected. Retention increases of 5.6% and 7.3%, respectively, were detected in two used sandpacks. This agrees with the results from the static measurements on used sands, which also confirms that the adsorbed 53

66 Adsorption, g/g sand molecules occupied almost all the vacant sites on the sand surface and prevented further attachment. Table 4.1-Dynamic Retention in Sandpacks SP No. L, cm A, cm 2 W sd, g PV, cm 3 k, D C p, ppm R pret, g/g sand , , HPAM 6-8 million Daltons Mw 30% hydrolysis 2% NaCl, 120 ml/hr 25 C Polymer concentration, ppm Fig Adsorption isotherm using dynamic method (fresh sandpacks used for each case). Among this series of tests, two examples are given here. Figure. 4.6 shows the first and second effluent polymer concentration profiles for 50 ppm HPAM injection. The area between these two breakout curves is an output of polymer retention in porous media. The same test was carried out for 500-ppm polymer solution, and with the result shown in Fig Retentions calculated for 50 ppm and 500 ppm are 4.85 g/g sand and 10.2 g/g 54

67 Concentration ratio (C p /C 0 ) Concentration ratio (C p /C 0 ) sand, respectively. We can see that when polymer concentration rises from 50 ppm to 500 ppm, retention increases by about 110% Darcies sandpack 50 ppm HPAM st polymer injection 2nd polymer injection Pore volumes injected, pv Fig Retention determination for 50 ppm HPAM Darcies sandpack 500 ppm HPAM st polymer injection 2nd polymer injection Pore volumes injected, pv Fig Retention determination for 500 ppm HPAM. 55

68 Concentration ratio (C p /C 0 ) Static measurements show that no significant re-adsorption occurred in used sands (Fig. 4.4). To check whether this is also true for dynamic measurement, 1,000 ppm polymer solution was injected through the sandpack whose retention was originally satisfied at 500 ppm (SP No. 4 in Table 4.1). Approximately 0.75 g/g sand (Fig 4.8) or 7.3% additional retention was determined in this sandpack. Again, the re-occurrence of retention is negligible for pre-contacted porous media, even though the injected solution had much higher concentration than the previous one st polymer injection 2nd polymer injection 4.88 Darcies Sandpack Pre-treated with 500 ppm HPAM 1,000 ppm HPAM Pore volumes injected, pv Fig Retention of 1,000 ppm in pre-treated sandpack with 500 ppm. Retention in Sandstone Cores. To date, polymer retention in high permeability sandpacks has been evaluated where adsorptive retention is believed to dominate. However, besides adsorption on rock surface, mechanical entrapment occurs simultaneously in pore throat constrictions and dead-end spaces when consolidated cores are used (Huh et al. 1990, Ranjbar et al. 1991). To investigate the dependence of polymer 56

69 retention on concentration in less permeable porous media, retention of HPAM with different concentrations was measured in several consolidated sandstone cores. The first measurement was performed using 347 md Dundee sandstone core (see the description of this core in Chapter 3). In this experiment, a series of polymer solutions were injected sequentially through this core at a rate of 60 ml/hr. Concentrations of 20, 40, 80, 100, 200, 400, 700 and 1,000 ppm were considered. A higher flow rate was used when dealing with the hydrodynamic effect on retention; this will be discussed in the next section. For this series of tests, the same sandstone core was repeatedly used and retention measurement followed the concentration sequence from low to high. For instance, the retention at 40 ppm was measured only after the retention at 20 ppm (the lowest concentration) was completed. Accordingly, the retention for the highest concentration case (1,000 ppm for this series of tests) was carried out in the final run. The results are shown in Fig From 20 ppm through 1,000 ppm, all the retention values fall into the range of approximately g/g sand with a relative error of 1.1%. When polymer concentration was greater than 100 ppm, the retention reached a plateau, with the maximum retention of approximately 16 g/g rock. Therefore, if one core was repeatedly used, polymer concentration showed almost no influence on polymer retention. The result was consistent with other studies where the same core was used repeatedly (Shah et al. 1978, Zheng et al. 1998). Fig shows two effluent concentration profiles from the following 80 ppm polymer solution injection. The small area between these two curves represents the increased retention, which is only 0.7 g/g sand. 57

70 Concentration ratio (C p /C 0 ) Polymer retention, g/g sand Dundee sandstone core 347 md Porosity: 23% Polymer concentration, ppm Fig Effect of concentration on polymer retention, 347 md core Dundee sandstone core Porosity: 23% 347 md st polymer injection 2nd polymer injection Pore volumes injected, PV Fig Retention determination for 80 ppm HPAM, 347 md core. 58

71 Polymer retention, g/g sand The above results were obtained from relatively high permeability core (Dundee sandstone core with permeability of 347 md). For the retention in low permeability porous rock, a 71 md Berea sandstone core was used. The same strategy was followed to determine the polymer retention at different polymer concentrations. As shown in Fig. 4.11, the retention increases a little bit with the increase of polymer concentration, but compared with the concentration increase from 20 ppm to 1,000 ppm which increases by 50 times, the increase of polymer retention is very small, only about 1.6 times. Figure demonstrates the effluent concentration profile for the first and second polymer slug injections. For the 20 ppm case, around 16 PV of polymer were injected to satisfy the polymer retention. Nevertheless, for the following 40 ppm case, shown by Fig. 4.13, fewer than 4 PV of polymer are needed to reach the maximum retention Berea sandstone core 71 md Porosity: 18.5% Polymer concentration, ppm Fig Retention isotherm of HPAM in 71 md core. 59

72 Concentration ratio (C p /C 0 ) Concentration ratio (C p /C 0 ) st polymer injection 2nd polymer injection Pore volumes injected, PV Fig Retention determination for 20 ppm HPAM st polymer injection 2nd polymer injection Pore volumes injected, PV Fig Retention determination for 40 ppm HPAM. 60

73 The two tests conducted on both high and low permeability sandstone cores indicate that polymer concentration shows very little effect on retention in porous media if the same core is repeatedly used. These results are consistent with our static measurements and also agree with findings by other researchers (Shah et al. 1978, Zheng et al. 1998). To make a comparison, a similar Dundee sandstone core with 23.4% porosity and 449 md permeability was used. Polymer solution with concentration of 1,000 ppm was injected and 56.5 g/g rock retention was measured. Again, for higher concentration, more retention occurs if fresh core used. In addition, polymer retention shows strongly concentration-dependent behavior. Core No. L, cm A, cm 2 PV, ml Table 4.2-Retention on Sandstone Cores, % k, md Polymer Conc, ppm Retention, g/g , , , When static and dynamic retentions were compared, we found that at the same polymer concentration, static measurement showed much higher retention. For instance, at concentration of 2,000 ppm, static retention of 335 g/g was detected on disaggregated sand grains (Fig. 4.3). Nevertheless, dynamic retention was only about 27.8 g/g in 5.51 Darcy sandpack (Table 4.1). The usually high retention from static measurement is attributed to the large surface area exhibited by disaggregated sand grains, which provide more vacant sites for polymer molecule attachment. The other dynamic retention by 449 md fresh sandstone core is about 56.5 g/g at concentration of 1,000 ppm (the second core in Table 4.2). The difference between these two dynamic retentions can be explained 61

74 by the effect of rock permeability. Research shows that polymer retention strongly depends on the permeability of porous media. Polymer retention usually increases with decrease of rock permeability (Vela et al. 1976) Proposed Adsorption Model Based on the experimental results, a polymer concentration-related retention model is proposed, which accounts for the observed retention behaviors. It is well known that polymer molecules may interact with each other in solution and the degree of interaction depends greatly on polymer concentration. Three concentration regimes were proposed (de Gennes 1979; Ying and Chu 1987) as dilute (C < C*), semidilute (C* < C < C**), and concentrated (C** < C), where C* is the overlap concentration crossover from dilute to semidilute regimes and C** is the overlap concentration crossover from semidilute to concentrated regimes. More specifically, as shown in Fig. 4.14, in the dilute regime, polymer molecules exist in solution as free coils where little interaction occurs. In the semidilute regime where polymer concentration is greater than the overlap concentration, C*, macromolecules start to contact each other and intermolecular interactions occur. With further increase in concentration (especially when the concentration is above C**), the intermolecular entanglements dominate the interaction, resulting in the formation of a network structure (Ferry 1948). For our HPAM, brine, and temperature, we measured C* to be 300 ppm and C** to be 3000 ppm (see the following section). When dealing with polymer retention on sand surfaces, this concentration-based interaction among polymer molecules in solution may be used to explain the adsorption mechanism. 62

75 In the dilute regime, polymer molecules exist in solution as free coils but tend to take a flat orientation when they adsorb onto the rock surface. In this configuration, most, if not all of the molecular segments are in contact with the surface. It was called twodimensional adsorption (Peterson and Kwei 1961). In this regime, two-dimensional adsorption dominates retention, and polymer molecules continue to be adsorbed until the maximum coverage is reached. As shown by Region A in Fig. 4.15, adsorption is independent of polymer concentration. For practical purposes, the retention in the dilute regime indicates the minimum amount of polymer needed to occupy the available vacant sites. In field applications of polymer and chemical floods, reduced polymer retention may be achieved by first injecting a low-concentration polymer bank. In the semidilute regime, the intermolecular interaction in solution will result in a mixed adsorption, i.e., some molecules will be adsorbed with all the segments in contact with the surface, while others will be adsorbed with only partial segments in contact with the surface. The latter orientation will be labeled as three-dimensional adsorption. Increasing the polymer concentration will increase the three-dimensional adsorption as well as the total adsorption, as shown by Region B in Fig Polymer retention is concentration-dependent in the semidilute regime. In the concentrated regime, the molecular entanglement in solution causes the threedimensional adsorption to dominate, i.e., most polymer molecules are adsorbed with segments partially attached to rock surface. Put another way, only one end of the polymer molecule is attached to the surface, while the majority of the molecule dangles free in the solution. In this case, almost no additional polymer molecules can be adsorbed with increasing concentration because all sites have already been occupied. As shown by 63

76 Region C in Fig. 4.15, the adsorption is concentration-independent. This concentrationrelated adsorption model is also summarized in Table Fig Polymer molecule interaction at different concentrations. Fig Proposed polymer adsorption mechanism on the rock surface. 64