Recent Advances in Viscoelastic Surfactants for Improved Production from Hydrocarbon Reservoirs.pdf

8/17/2019 Recent Advances in Viscoelastic Surfactants for Improved Production from Hydrocarbon Reservoirs.pdf

1/35

SPE-173776-MS

Recent Advances in Viscoelastic Surfactants for Improved Production fromHydrocarbon Reservoirs

Katherine L. Hull, and Mohammed Sayed, Aramco Research Center—Houston; Ghaithan A. Al-Muntasheri, Aramco Research Center—Houston, Saudi Aramco

Copyright 2015, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE International Symposium on Oilfield Chemistry held in The Woodlands, Texas, USA, 13–15 April 2015.

This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents

of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect

any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written

consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may

not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract

Viscoelastic surfactants (VES) are used in upstream oil and gas applications, particularly hydraulic

fracturing and matrix acidizing. A description of surfactant types is introduced along with a theoretical

description of how they assemble into micelles, what sizes and shapes of micelles can be formed under

different conditions, and finally how specific structures can lead to bulk viscoelastic solution properties.

This theoretical discussion leads into a description of the specific VES systems that have been used over

the last twenty years or so in improved oil recovery for upstream applications.

VES-based fluids have been used most extensively for hydraulic fracturing. They are preferred over

conventional polymer-based fracturing fluid systems because they are essentially solids-free systems

which have demonstrated less damage to the reservoir rock formation. Important advancements in VES

have been made by introducing “pseudo-crosslinking agents” such as nanoparticles to enhance the

viscosity. Fracturing fluid systems based on VES have also been improved recently by developing internal

breakers to lower their viscosity in order to flow back the well. The flexibility of VES-based fluids has

been demonstrated by their application as foamed fluids as well as their incorporation with brine systems

such as produced water.

A second key area that has benefited from VES-based systems is matrix acidizing carbonated-based

reservoirs. The viscosity of these VES-based fluids is mostly controlled by pH where, at low pH (low

viscosity), the acid system flows easily and invades pore spaces in the formation. During acidizing, theacid is spent, and the pH and viscosity increase. Because the spent acid has higher viscosity, fresh acid

is diverted to low permeability un-contacted zones and penetrates the rocks to form wormholes. A number

of experimental studies and field applications to these effects have been performed and will be described

here.

In order for VES-based fluids to play a more prominent role in the field, inherent limitations such as

cost, applicable temperature range, and leak-off characteristics will need to continue to be addressed. If

we can efficiently and economically overcome these issues, VES-based fluids offer the industry an

excellent clean, non-damaging alternative to conventional polymer-based fluids.


2/35

Introduction

Surfactants have been used in a wide variety of industrial products including cleaning detergents, textiles,

cosmetics, paper production, food, mining, as well as fluids for the oil and gas industry. Their versatile

nature has allowed them to be utilized for as emulsifiers, wetting agents, and foaming agents. Surfactants

are amphiphilic organic molecules which consist of a component which, on its own, would be soluble in

a given liquid and a second component which, on its own, would not be soluble in the same liquid (Witten

and Pincus 2010). In aqueous environments, the hydrophilic head group interacts favorably with the

solvent medium while the hydrophobic tail has a more favorable free energy when away from the solvent,

concentrating at the liquid boundary. The surfactant molecules form an interface between two immiscible

liquids and larger quantities of surfactant lead to more interfacing of the two liquids until eventually they

are considered mixed. The specific chemical identity of the polar head groups and hydrocarbon tail groups

varies but surfactant molecules are typically broken into classes which include anionic, cationic, nonionic,

and zwitterionic species. Common examples of each surfactant class include carboxylate or sulfate polar

head groups (anionic), quaternary ammonium head groups (cationic), long chain alcohols (nonionic), and

betaines (zwitterionic). Table 1 shows each of the surfactant types, their charge, as well as names and

structures of specific examples.

Under certain conditions, surfactant molecules arrange into colloidal structures called micelles, where

the hydrocarbon tails of the surfactants orient towards each other while the polar head groups form an

interface with the surrounding aqueous media. Figure 1 provides a simple schematic to illustrate the

micellization process. The size and structure of these micelles is controlled by a variety of parameters and

will be described in later discussions. Surfactant micelles form spontaneously in aqueous solution when

the surfactant concentration, c, exceeds the threshold referred to as the critical micelle concentration

(cmc). Certain properties of aqueous solutions of surfactants such as osmotic pressure, turbidity, surface

tension, and electrical conductivity are observed to experience discontinuities at this critical concentration,

providing confirmation of the transition from discrete molecules to aggregates (Hargreaves 2003). For

Table 1—A list of the types of surfactant molecules is provided along with specific examples that correspond with each of the

classes. Studies of the phosphobetaine (x10, n1-10) have been described (Chevalier 1988 & 1992).

2 SPE-173776-MS


3/35

example, the cmc of various solutions of cationic surfactants was determined by plotting the electrical

conductivity against the surfactant concentration. Two linear slopes were observed for all surfactants, and

the intersection of the two lines is the cmc (Gravsholt 1976).

The effects of various factors on the cmc have been noted, including temperature, surfactant type,

hydrophobe chain length, and salts/cosolutes (Hamley 2007). The cmc is largely independent of temper-

ature except in the case of nonionic surfactants, whose hydrophilic head groups are based on oxyethylene

groups, where increasing temperature causes the cmc to decrease. In the case of ionic versus nonionic

surfactants with equal hydrophobic chain lengths, the cmc is usually lower for nonionic surfactants

because the electrostatic repulsions between ionic headgroups are more difficult to overcome during

micelle formation. Similarly, as the size of the polar head group increases for ionic surfactants, the cmc

increases. However, as the length of the hydrophobic tail of the surfactant increases up to around 16

carbon atoms, the cmc decreases. Beyond this chain length, there is little change observed in the cmc.

Finally, salt addition to ionic surfactant solutions decreases the cmc because the ions reduce head group

repulsions.

A. Thermodynamics of Micelle FormationWhen surfactant molecules are dissolved in aqueous solution, the attractive and repulsive forces between

the surfactant molecules and water cause micelle assembly, if the concentration of surfactant exceeds the

cmc. Initial work by Tanford (1974 & 1979) gave rise to a free energy model to describe the formation

and growth of these surfactant aggregates resulting from (1) the tail transfer, (2) the hydrocarbon-water

interface, and (3) the head group interactions. The free energy change associated with an aggregate of size

g (aggregation number) can be summarized with the following relationship in Equation 1 ( Nagarajan

2002):

(1)

where k is the Boltzmann constant and T is the temperature in Kelvin. The tail transfer free energy (arises from the transition of the hydrocarbon tail from the aqueous environment to the hydro-

carbon-rich environment at the center of the micelle. This change provides a favorable contribution to the

overall free energy. By contrast, both the interfacial and head group interactions are unfavorable

contributions to the overall free energy. The interfacial free energy results from the residual contact that

the hydrocarbon tail has with water, while the head group free energy corresponds with the crowding

(repulsive forces) that takes place at the surface of the micelle. Both the interfacial free energy and the

head group free energy are dependent upon the surface area per molecule of the aggregate core, a. The

standard free energy change can be rewritten as in Equation 2:

Figure 1—Schematic of a surfactant molecule depicting the polar head group and the hydrocarbon tail group, along with its

transformation into a colloidal structure when the surfactant concentration exceeds the cmc.

SPE-173776-MS 3


4/35

(2)

where is the unit area for the residual contact between hydrocarbon and water and is the headgroup

repulsion parameter. For micelles in thermodynamic equilibrium, where a ae, the standard free energy

(g/kT) is minimized (Equations 3-5). Furthermore, Table 2 describes each contribution to the critical

micelle concentration in mole fraction units, Xcmc, and how each influences either the formation of aggregates or the size of the aggregates.

(3)

(4)

(5)

The role of the surfactant hydrocarbon tail on micellization free energy has also been considered ( Nagarajan 2002). The hydrocarbon chains pack differently in the bulk as compared to the micelle where

they deform non-uniformly. Although the tail is constrained to a fixed location where it adjoins the polar

head group, the other end is free to position itself anywhere within the core of the micelle as long as the

core maintains a uniform density of hydrocarbon chains. The result is an additional contribution to the

overall free energy of micellization which accounts for the packing ( Nagarajan and Ruckenstein 1991).

(6)

The packing free energy of the micelle core per surfactant molecule is then given for spheres, cylinders,

and lamellae, respectively:

(7)

where R is the radius of the micelle, L is the characteristic segment length, and N is the number of

segments. The numerical values 3, 5, and 10 correspond to the molecular packing differences in each of

the geometries ( Nagarajan and Ruckenstein 1991).

B. Micelle Size and Shape

The size and shape of micelles is dictated by a wide range of factors including surfactant properties such

as charge, geometry, and concentration as well as solution conditions such as temperature, ionic strength,

type and concentration of salt, and shear rate. The molecular packing parameter is a classical method for

Table 2—Contributions to the overall free energy of micellization are provided along with physical descriptions of their effects

upon the aggregation behavior of the micelle.

4 SPE-173776-MS


5/35

qualitatively predicting the micelle structure and is still widely used (Israelachvili et al. 1976; Israelachvili

1992). The parameter is defined as 0 /ael 0 where 0 is the volume of the surfactant tail, l 0 is the maximum

length of the surfactant tail, and ae is the surface area per molecule at the water-surfactant interface as

described previously. When the molecular packing parameter is 1/3, the surfactant molecules are

predicted to assemble into spherical aggregates. When the packing parameter is between 1/3 and 1/2, the

surfactant micelle is expected to adopt a rodlike or wormlike shape. Finally, vesicles and lamellar

structures are expected when the packing parameter is between 1/2 and 1 or close to 1, respectively. Asummary of the various micelle shapes and their corresponding packing parameters is provided in Figure

2.

The packing model can be used to predict how changes to the surfactant molecule or solution

conditions will influence the micelle structure. Surfactant molecules with large polar headgroups (large ae)

are expected to promote the formation of spherical micelles, while surfactants with small headgroups

should encourage lamellae formation. For example, nonionic surfactants with small ethylene oxide

headgroups (small number of carbons, m) should favor bilayer/lamellae structures. However, nonionic

surfactants with larger headgroups (e.g., m10, 12, 14, 16) yield cylindrical or wormlike micelles (Jerke

et al. 1998; Bernheim-Groswasser et al. 2000; Imanishi and Einaga 2007). Similarly, the packing model

can be used to predict that salt addition to ionic surfactant solutions will cause a transition from spherical

to cylindrical micelles. The salt interacts electrostatically with the polar headgroups, thus reducing the

headgroup repulsion parameter, , thereby decreasing ae and increasing the molecular packing parameter.

For example, the cationic surfactant hexadecyltrimethylammonium bromide, which is also referred to as

cetyltrimethylammonium bromide (CTAB), forms spherical micelles in aqueous solution. Upon addition

of sodium nitrate (NaNO3) to the surfactant solution, the spherical micelles transform into wormlike

micelles (Kuperkar et al. 2008). Another example of using the packing parameter to predict micelle

structure involves surfactants which are composed of double rather than single hydrophobic tails. These

surfactants will have molecular packing parameters which are twice as large and, given the same

headgroup, they will be much more likely to form lamellar structures.

Figure 2—Schematic diagrams of each micelle type and their respective qualitative prediction of shape based on the packing parameter

(adapted from Nagarajan 2002 and van Zanten 2011).

SPE-173776-MS 5


6/35

C. Wormlike Micelles

The micelle shape of particular interest to oilfield applications is wormlike micelles. These are elongated

rodlike micelles which are long and flexible and can entangle in solution, imparting viscoelastic properties

to the fluid. Detailed structural, chemical, and behavioral characteristics of wormlike micelles have been

recently reviewed in the chemical literature (Dreiss 2007). The review provides an extensive list of

surfactants systems that have been reported to form wormlike micelles and the corresponding methods

used to characterize these fluids such as Cryo-Transmission Electron Microscopy (TEM), rheology, staticlight scattering (SLS), and small angle neutron scattering (SANS).

The structure of wormlike micelles can be described with a series of different length scales. Figure 3,

below, provides a schematic of a wormlike micelle and the various parameters that are used to describe

the structure. The radius of gyration, R g , reflects the radius around which the micelle rotates. The total

length of the micelle, Lc, can vary from nanometers to microns and can be determined by direct imaging

via Cryo-TEM. The persistence length, l p, is the length over which the micelle is considered to be rigid.

As an example, l p was measured for 6 mM cetyl pyridinium bromide with 0.8 M sodium bromide added

(Witten and Pincus 2010; Cates and Candau 1990). At 35 °C, l p was found to be 20 nm while the

cross-section radius of the micelle, Rcs, was determined to be 2 nm.

Wormlike micelles are dynamic structures which differ from polymer solutions, in that the aggregates

are constantly breaking and reforming, versus polymer solutions where the lengths of the chains are fixed

at the point of quenching during synthesis (Cates 1990; Cates and Candau 1990). These so-called “living

polymers” are in thermal equilibrium and have molecular weight distributions which are normally quite

broad. The dynamics of breaking and reforming micelles can have an effect on the chain entanglements.

This in turn has an effect upon the diffusion properties of the chains as well as the viscoelasticity. Micelle breakage is assumed to be a unimolecular process where the lifetime of the chain before breaking into two

segments is given by the parameter break (Cates 1990; Cates and Candau 1990):

(8)

where is the average contour length of the micelles before breaking and k 1 is the rate constant. A

second stress relaxation mechanism is reptation, a diffusion phenomenon also observed in entangled

polymer solutions. From Equation 8, when break rep, micelles break and reform faster than the time

scale for reptation. In the limit when break rep is much less than break rep, the viscoelastic surfactant

Figure 3—Wormlike micelle schematic showing the fundamental structural parameters that cover a wide range of length scales

(adapted from Dreiss 2007).

6 SPE-173776-MS


7/35

solution behaves as a Maxwell fluid with a single relaxation time given by Equation 9 (Cates 1990; Cates

and Candau 1990):

(9)

The frequency, , dependent elastic (storage) modulus G = and the viscous (loss) modulus G for a

Maxwell fluid are given by the following expressions (Schubert et al. 2003):

(10)

(11)

Where G 0 is the high-frequency plateau modulus (the elastic modulus at infinite frequency or t0).

The moduli G = and G can be determined experimentally by oscillatory-shear measurements where the

sample is deformed sinusoidally and the response is measured. The inverse of the crossover frequency of

G = and G is the relaxation time R . This model has been applied to a large number of viscoelastic micelle

solutions and is generally considered a good indication that wormlike micelles are present (Dreiss 2007).

Under steady shear flow, wormlike micelles exhibit a low-shear Newtonian plateau, which is character-

ized by a zero-shear viscosity, 0 G 0 R. However, when the critical shear rate is exceeded, thewormlike micelle solution undergoes strong shear thinning. The longest structural relaxation time R can

be estimated as the inverse of the critical shear rate. Measurement of G = and G can be used to determine

whether or not a fluid is viscoelastic at a given temperature. A simple observational technique has also

been described for determining whether or not a fluid is viscoelastic (Gravsholt 1976) i.e., bubbles that

appear during swirling of a sample will recoil when the swirling stops if the solution is viscoelastic.

D. Improved Oil Recovery Applications

Viscoelastic surfactant fluids have been used in a variety of applications in the upstream oil and gas

industry including gravel packing, frac packing, fracturing fluids, and matrix acidizing. Each of these

applications will be touched on throughout the course of this paper. Substantial recent developments in

the areas of hydraulic fracturing and matrix acidizing have been seen over the last decade and will be

discussed in greater detail. In the case of fracturing fluids, a variety of nanoparticle additives have beendeveloped to “pseudo-crosslink” the surfactant micelles and enhance the viscosity of the fluids. Further-

more, methods for breaking the gelled fluids in order to flow back the well have been developed and will

be described. Viscoelastic surfactants have also been used in more specialized cases such as foamed fluids

and heavy brine solutions. A detailed description of their use in matrix acidizing will also be given,

including a variety of experimental studies and several case histories of field treatments with VES-based

fluids and hydrochloric acid.

Discussion

A. Fracturing Fluids

VES-based fluids were first reported for upstream oil and gas applications in gravel-pack completions

( Nehmer 1988) and frac-packs (Brown et al. 1996; Stewart et al. 1994) and were later also developed into

fluids that were used for hydraulic fracturing (Samuel et al. 1997, 1999, 2000). Conventional polymer-

based fracturing fluid systems incorporate a water-soluble polymer, crosslinker, and breaker among other

additives (Al-Muntasheri 2014). The fluid has viscosity which is sufficient for transporting proppant into

the fractures. Over time, the gel is broken by enzyme or oxidizer and the fluid is flowed back from the

formation to the surface. This process is operationally complex in that it requires polymer hydration and

a variety of additives such as biocides, crosslinkers, and breakers. By contrast, viscoelastic surfactants are

simpler to use in the field because there is no hydration step and fewer additives are required. A major

advantage over polymer-based systems is that VES-based fluids are essentially solids-free which means

SPE-173776-MS 7


8/35

that residue is not deposited in either the formation or the proppant pack, so these fluids are more efficient

in hydraulic fracture reservoir stimulation. This feature of VES-based fluids is significant since the

conductivity of the in-place proppant pack have a significant effect upon the overall well productivity

(Palisch et al. 2007). In addition, VES-based fluids have been observed to heal after exposure to shear,

and additives have been developed to improve the shear rehealing time (Lee et al. 2008; Chen et al. 2008).

There are some drawbacks to VES-based fluids relative to conventional polymer-based fracturing fluids.

Because they are relatively solids-free, they do not form a filtercake, so high leak-off has been observed in many cases. Also, the temperature range over which VES-based fluids can operate is also lower, with

the highest temperatures observed falling less than 300 °F. Table 4 provides a summary of the various

characteristics observed about the nature of VES-based hydraulic fracturing fluids.

VES-based hydraulic fracturing fluid systems have shown better performance in field applications as

highlighted by several reports (Samuel et al. 1997, 1999, 2000). In Alberta, five wells stimulated with

VES-based fluid were compared with five offsets that were fractured with 20 pptg borate-crosslinked

guar. For wells that were fractured with the VES-based fluid, the absolute open flow increase was 9%

greater (Samuel et al. 1997). Similarly, when the same VES-based fluid was used to fracture a well in

southwest Kansas, the initial production rate from this well was approximately 27% greater than the

production rate from the parent well in the same section. The rate was also 52% greater than the average

of 12 adjacent wells hydraulically fluid fractured. In Rock Springs, Wyoming, two identical offset wells

were fractured with either VES-based fluid or with 25 pptg guar ( Samuel et al. 2000). The calculated

hydraulic fracture lengths for both the polymer and the VES-based treatments were similar. The fracture

height generated by the guar treatment was estimated to be more than double the height for the VES-based

fluid treatment due to the higher viscosity of the polymer fluid. It was determined that fractures occurred

outside the pay zone for the polymer treatment, resulting in propped fractures in nonproductive zones.

Analysis of the flowback fluid indicated that the VES-based fluid cleaned up faster than the polymer fluid.

The well stimulated with VES-based fracturing fluid had an initial production of 2.8 MMscf/D whereas

the polymer-stimulated well produced at 1.3 MMscf/D. Four wells in Armstrong County, Pennsylvania

were treated with either VES-based fluids or linear polymer gels, and the well treated with VES-based

fluids showed initial production levels that were three to four times higher than the linear polymer gel

treatments (Leitzell 2007). Likewise, of the four wells treated in Clearfield County, Pennsylvania with

Table 4 —A summary of characteristics associated with VES-based hydraulic fracturing fluids is provided along with a list of the

limitations of these fluids.

8 SPE-173776-MS


9/35

either VES-based fluids or linear gels, the treatments incorporating surfactants showed better sustained

production. A higher temperature VES fluid application was implemented in the El Tordillo field which

is located in the central portion of the San Jorge Basin ( Fontana et al. 2007). The well consisted of four

individual zones, and each was treated with a 6% VES-based fracturing fluid. Improved performance

relative to conventional treatments was noted in terms of reduced fracture height growth and, therefore,

reduced proppant volumes by 30-50% because fracture placement was concentrated within the zone of

interest.

A.1. VES-Based Fluid Chemistries

A variety of viscoelastic surfactant chemistries have been utilized for fluids used in hydraulic fracturing.

Examples of cationic, nonionic, zwitterionic, and anionic surfactants have all been utilized for hydraulic

fracturing applications primarily in aqueous systems, although some examples of gelled hydrocarbon

systems have been reported (Samuel 2009; Samuel et al. 2014). Beginning with cationic surfactants, early

studies of quaternary ammonium salts established their viscoelastic properties (Gravsholt 1976). Solutions

which incorporate water, a water soluble salt (electrolyte), and a quaternary ammonium salt have been

used for drilling fluids, completion fluids, hydraulic fracturing fluids, etc. (Teot et al. 1988). Figure 4(a)

shows the structure of one preferred surfactant system oleyl methyl bis(2-hydroxyethyl) ammonium

chloride. Quaternary ammonium salt chemistry was also utilized for developing a fluid system that is

stable up to 225 °F for hydraulic fracturing applications ( Norman et al. 1996). Erucyl bis(2-hydroxyethyl)

methylammonium chloride (EHAC), shown in Figure 4(b), was combined with ammonium chloride or

potassium chloride and, in some cases, also with an organic salt such as sodium salicylate, resulting in a

viscous fluid that can be used to fracture high permeability formations. This surfactant system is known

to form long wormlike micelles which entangle and impart viscoelastic characteristics to the fluid

(Raghavan and Kaler 2001). Further developments with the same surfactant system have also been seen,

showing insensitivity to a range of pH values as well as compatibility with seawater-based fluid systems

(Brown et al. 1999; Samuel et al. 2001). EHAC can alternatively be combined with a polymer such as guar

or modified guar in concentrations below 10-15 pptg to provide a dual system (Miller et al. 2003). The

polymer is crosslinked with boron, zirconium, or another metal to the extent that it forms a filter cake on

the formation face, thereby enhancing fluid loss control. Alternative cationic surfactants such as the“gemini surfactant” dimethylene-1,2-bis(dodecyldimethylammonium bromide) display viscoelastic be-

havior at lower concentrations (0.7-1.7%) than conventional surfactants (Yang et al. 2013). The critical

micelle concentration of these dimeric surfactants is usually ~10 times lower than the conventional

monomeric surfactants.

Cationic surfactants, however, present potential problems to the formation in that they can oil-wet

formation rocks, which can lead to increased resistance to oil flow, so surfactant fluids composed of

anionic or nonionic surfactants may be preferred (Whalen 2000). In the case of nonionic surfactants,

amido amine oxides have been described for hydraulic fracturing applications. One specific example that

Figure 4—The structures of specific quaternary ammonium salts that have been used for hydraulic fracturing applications are shown

including (a) oleyl methyl bis(2-hydroxyethyl) ammonium chloride, (b) erucyl bis(2-hydroxyethyl) methylammonium chloride (EHAC),

and (c) N,N,N, trimethyl-1-octadecammonium chloride.

SPE-173776-MS 9


10/35

has been reported is tallow amido propylamido oxide, TAPAO (McElfresh and Williams 2007). The

general form of this compound is shown below in Figure 5 along with specific components of the tallow

substituent. A fluid composed of 6 vol% TAPAO and 3% KCl was prepared, and the viscosity was

measured at a shear rate of 100 s-1. The viscosity of the fluid was maintained above 100 cp in the

temperature range of 100-200 °F and around 40 cp at 225 °F.

Hydraulic fracturing VES-based fluids that are composed of anionic surfactants such as alkyl sarco-

sinates have been seen and can also be applied to gravel packing, frac packing, and water control issues

(Di Lullo Arias et al. 2002). The fracturing fluid includes a combination of a water soluble salt such as

potassium or sodium chloride, 0.5-6 wt% alkyl sarcosinate surfactant, a buffer for adjusting the pH to6.5-10, and either 0.1-2 wt% carboxylic acid salt as an additional source of ions or, alternatively, 3-6 wt%

chloride or fluoride salts. The sarcosinate has been used for fracturing and has approximately 94% oleoyl

sarcosine as shown in Figure 6(a). Sufficient sarcosinate is present in solution to provide adequate

viscosity to the fluid in order to be able to transport proppant into the hydraulically created fractures. The

anionic surfactant methyl ester sulfonate (MES), which is shown in Figure 6(b), has also been incorpo-

rated as a treatment fluid (Welton et al. 2007b). 5 wt% MES was observed to gel (viscosity 20 cp at 511

s-1) in the presence of 5-10 wt% inorganic salt at pH4, ~7, and 10. 5 wt% MES was also observed to

gel in the presence of 5 wt% inorganic salt and 10% HCl, but not in the presence of 15% HCl. The MES

surfactant is generated by adding sulfur trioxide to the -carbon of a methyl ester and then neutralizing

with a base (Welton et al. 2007a). It is considered to be more environmentally friendly than other

surfactant gels because it is derived from renewable resources such as palm kernel oil, is biodegradable,and is less toxic. Several fracturing treatments were performed in South America with an environmentally

friendly anionic surfactant (EHAC) at 4 vol% (Di Lullo et al. 2001) and were reported to improve

production in three case studies.

Figure 5—The general structure of the nonionic surfactant tallow amido propylamine oxide (TAPAO) is shown, along with the major

components of the tallow amido substituent.

10 SPE-173776-MS


11/35

VES-based fluids composed of zwitterionic/amphoteric surfactants have also been reported, including

dihydroxyl alkyl glycinate, alkyl ampho acetate or propionate, alkyl betaine, alkyl amidopropyl betaine,

and alkylimino mono- or di-propionates derived from waxes, fats, or oils (Dahanayake et al. 2001, 2002,

2004). The surfactant is combined with either an inorganic salt or an organic compound such as phthalic

acid, salicylic acid, or their salts. Figures 7(a) and 7(b) show two specific surfactants that were reported

including disodium tallowiminodipropionate and disodium oleamidopropyl betaine. The viscosity of a

VES solution composed of 5% disodium tallowiminodipropionate and 2.25% phthalic acid was deter-

mined at variable shear rates. When the shear rate varied from 1 to 100 s-1, the viscosity ranged from

almost 100,000 cp down to 100 cp at both 25 and 50 °C. Similar results were also obtained when

1.75-2.0% phthalic acid and 4% of ammonium chloride (NH4Cl) were added to 5% surfactant instead. The

viscosity versus shear rate was also reported for 4-5% disodium oleamidopropyl betaine, 3% KCl, and

0.5% phthalic acid. As the shear rate increased from 0.01 to ~30 s -1, the viscosity decreased from

approximately 100,000 to 1000 cp. In the Gulf of Mexico, 17 fracturing treatments utilizing zwitterionic

VES-based hydraulic fracturing fluids in concentrations ranging from 3.5-6.0% were reported by Sullivan

et al. (2006). The bottom hole static temperature ranged from 135-214 °F, and the permeability of the

formation was as high as 170 md. Welton and Bryant (2011) furthermore describe another formulation

incorporating 7 wt% of the previously described oleylamidopropyl betaine, as in Figure 7(b), 0.14 wt% potassium stearate (soap), and 2 wt% polyamide (nonaqueous tackifying agent). Steady shear viscosities

obtained in the temperature range of 175-225 °F were enhanced relative to the same formulation that did

not contain the tackifying agent. Fluids incorporating a similar zwitterionic surfactant, erucylamidopropyl

betaine as shown in Figure 7(c), have been described for enhanced oil recovery (Morvan and Degre

2012).

In some cases, mixed surfactant systems have been developed. For example, quaternary amine

surfactant systems have been used in combination with anionic surfactants for fracturing applications. A

Figure 6—The chemical structure of (a) oleoyl sarcosine is shown, which constitutes around 94% of the sarcosinate product, along

with the structure of the surfactants (b) methyl ester sulfonate where R is an alkyl chain with 10-30 carbon atoms and (c) sodium xylene

sulfonate.

Figure 7—The chemical structures of (a) tallowiminodipropionate where Rtallow and (b) oleylamidopropyl betaine and (c) erucyl-

amidopropyl betaine.

SPE-173776-MS 11


12/35

solution composed of N,N,N, trimethyl-1-octadecammonium chloride, shown in Figure 4(c), and sodium

xylene sulfonate shown in Figure 6(c), for example, forms a viscoelastic gel with viscosity in the range

of 20-500 cp at a shear rate of 100 s-1 (Zhang 2002). In a follow-up study of this mixed surfactant system,

12 different ratios of the cationic/anionic (C/A) surfactants were evaluated in terms of proppant suspen-

sion ability as well as elastic and viscous properties (Gomaa et al. 2011). A number of conclusions were

drawn about the formulations examined including (1) the gel viscosity was highly dependent upon the

total surfactant concentration and the temperature and (2) good proppant suspension was only observed

when the C/A ratio was higher than 1.5 and the total surfactant concentration was greater than 30 gpt. In

addition to mixed surfactant systems, mixed surfactant/polymer systems have been formulated and tested

(Gomaa et al. 2014). A low molecular weight associative polymer, with no added crosslinker, along with

surfactant was evaluated in 33 different formulations. A polymer-to-surfactant ratio of 1.4 produced the

broadest working temperature range with the highest elastic characteristics and the best proppant

suspension.

A.2. Pseudo Crosslinkers

Recent studies in the chemical literature have established that inorganic particles can be added to

wormlike micelles to enhance the solution viscosity. For example, barium titanate (BaTiO3) pyroelectricnanoparticles increase the viscosity of solutions composed of sodium fatty acid methyl ester sulfonate

(Luo et al. 2012). Bandyopadhyay and Sood (2005) studied the rheology of semidilute solutions of the

cationic surfactant cetyl trimethylammonium tosylate (CTAT) in the presence of silica colloids with

diameters of 0.1 m. Upon addition of 1.3 wt% silica particles to 1.4 wt% CTAT solutions, the relaxation

time increases by 600%, the high frequency plateau modulus G0 increases by 37%, and the zero shear rate

viscosity 0 increases by 1600% when compared to solutions of pure CTAT. These changes in rheology

are attributed to the electrostatic interactions between the surfactant and the silica particles, namely

attractive forces between the surfactant headgroups and the surface of the silica. They describe these

interactions as resulting in the formation of bilayers where the silica particles form the center while the

surfactant headgroups form an outer layer.

Nettesheim et al. (2008) also proposed a model to describe the interaction between micelles and

nanoparticles in solution. When 30 nm diameter silica particles were added to solutions of cationic

cetyltrimethylammonium bromide (CTAB) and sodium nitrate, the relaxation time, the storage modulus,

and the zero shear rate viscosity were all observed to increase drastically. The nanoparticles are thought

to have an adsorbed layer of surfactant molecules on their surfaces with which the surfactant micelles can

interact. A micelle in solution can adsorb through its energetically unfavorable end-cap to a nanoparticle,

lowering the energy of the system by Ecap-Eads and causing the wormlike micelles to grow linearly. A

schematic of this proposed mechanism is shown in Figure 8. This interpretation is consistent with

Cryo-TEM micrographs as well as static and dynamic measurements. Helgeson et al. (2010) furthermore

described the viscoelastic behavior of solutions containing both wormlike micelles and nanoparticles as

a “double network” consisting of both micelle entanglements and particle junctions.

12 SPE-173776-MS


13/35

Oilfield researchers have capitalized on these developments over the last 5-10 years to improve the

viscosity and hence the proppant carrying capacity of VES-based hydraulic fracturing fluids. A widevariety of colloidal particles have been developed to form links between surfactant micelles in order to

improve the viscosity of VES-based fluids. Sullivan et al. (2006) describe materials including silica,

aluminum oxide, antimony oxide, tin oxide, cerium oxide, yttrium oxide, and zirconium oxide. The

particles range in diameter from 8-250 nm. Colloidal particles ranging in size form 1 nm up to 2 m have

been used to associate or “pseudo-crosslink” VES micelles (Huang 2009). These particles may be

composed of any of the following materials such as zinc oxide (ZnO), berlinite (AlPO4), lithium tantalate

(LiTaO3), gallium orthophosphate (GaPO4), BaTiO3, along with a variety of other materials. These

particles are added to the viscoelastic treating fluid in the range of 0.1 to 500 pptg. Another type of

crosslinker reported for VES micelles comprises a transition metal complex ( Reddy 2011). In particular,

the crosslinking agent used is in concentration of at least 0.15 wt%, is composed of zirconium trietha-

nolamine glycolate, zirconium triethanlamine lactate, or zirconium ammonium lactate acetate, and is

combined with at least 3 wt% VES.

Several examples of pseudo-crosslinking VES micelles with nanoparticles for hydraulic fracturing

applications have been reported. For example, two VES fluids were prepared with 4 vol% VES mixed

with 13.0-ppg CaCl2/CaBr 2, and one was loaded with 6-pptg of the 35 nm nanoparticles (Huang and

Crews 2008b). At 250 °F, the VES fluid system with nanoparticles maintained its viscosity at 200 cp while

the VES fluid without nanoparticles was reduced from 200 cp to less than 40 cp in 80 minutes. Leak-off

tests also showed that VES fluids containing nanoparticles outperform fluids without the nanoparticles.

Base fluids with 3 wt% KCl, 4% vol VES, and 2 gptg internal breaker were prepared, and 15 pptg

nanoparticles were added to one of the fluids. Both fluids were tested at 150 °F with 400-md ceramic discs

at 300 psi, and the fluid with nanoparticles showed significantly more control over fluid loss than the fluid

containing no nanoparticles.

Similarly, a fluid-loss control additive for VES-based fluids was then developed which consists of

small particles with an average diameter of less than 2 microns ( Huang and Crews 2009a). These particles

associate the micelles into a stronger network and are shown to maintain the viscosity, improve proppant

suspension, and reduce fluid loss. Proppant suspension tests were performed both with and without the

fluid loss control agent. Samples were prepared that consisted of 2% VES in 13.0 ppg brine with one

pound of 20/40-mesh ceramic proppant in one gallon of liquid. To one of the samples was added 10 pptg

of the particles. For the VES fluid without any particle addition, all of the proppant settled out within 15

minutes. For the VES fluid with particle addition, there is no noticeable proppant settling even after 90

Figure 8—Schematic of the model for the association of the end cap of a worm-like micelle with the surface of a particle (adapted from

Nettesheim et al. 2008).

SPE-173776-MS 13


14/35

minutes. The viscosities of VES fluids with and without particle addition were determined at 250 °F and

a shear rate of 100 s-1. Two fluids were prepared by combining 13.0 ppg CaCl2/CaBr 2 with 2 vol% VES,

6.0 ppg stabilizer, and 1.0 gptg internal breaker, and 10 pptg of the particles were added to one of the

fluids. The fluid with the added particles maintained its viscosity at 230 cp while the fluid without

particles only maintained a viscosity of 200 cp. Finally, leak-off tests with 400 md ceramic discs at 250

°F and 1,000 psi were performed. The fluids containing 5 gptg, 8 gptg, and 10 gptg particles showed

increasing degrees of fluid loss control over the base fluid.

Pseudo-crosslinked VES systems which incorporate pyroelectric (PE) nanoparticles such as ZnO were

developed and have been shown to delay crosslinking of VES micelles ( Crews and Huang 2008; Huang

2009). The particles are described as developing charges on the faces of the crystal as the particles are

heated, so more association between the particles and the micelles will occur as the VES fluid is heated

by the reservoir. Studies of pyroelectric nanoparticles have been performed where base fluids were

prepared with 13.0 ppg CaCl2/CaBr 2 and 2 vol% VES. Nanoparticles were added to two of the

fluids—one with 15 pptg regular nanoparticles and one with 10 pptg pyroelectric nanoparticles. A fourth

fluid was prepared with 30 pptg borate-crosslinked guar in 2% KCl. The leak-off performance of these

four fluids was tested on 400 md ¼ inch thick ceramic discs at 150 °F and 300 psi. The classical VES

system showed very high leak off over time because it was not wall-building. The other three systemswere wall-building and showed controlled fluid leak-off, with the best leak-off control coming from the

borate-crosslinked guar and the poorest leak-off control from the regular nanoparticles. The leak-off

control from the pyroelectric nanoparticle pseudo-crosslinked VES gel was comparable to the borate-

crosslinked guar gel. The viscosity and proppant suspension characteristics of the conventional VES fluid

were compared with the VES fluid containing pyroelectric particles, and the latter was found to perform

better in both cases.

Further improvement of VES-based fluids was accomplished with the development of a system that is

thermally stable up to 275 °F (Gurluk et al. 2013). Base fluids were prepared with either 2 or 4 vol%

amidoamine oxide surfactant in 14.2 ppg CaBr 2 brine. Either 30 nm MgO particles or 30 nm ZnO particles

were added to these solutions. The viscosity of the 4 vol% VES solutions containing either ZnO or MgO

particles at 275 °F and 10 s-1 were comparable at 100 cp, while the viscosity of the 4 vol% VES solution

without nanoparticles quickly decreased to zero cp. Solutions containing MgO and either 2 or 4 vol% VES

were also compared, the viscosity was found to increase when increasing from 2 to 4 vol% VES. An

additional observation was made of the difference between the effects of ZnO nanoparticles on 2% VES

solutions versus the effects of MgO nanoparticles on the same solution, namely that the viscosity of the

MgO solutions was higher than those containing ZnO. A summary of studies performed on pseudo-

crosslinked VES-based fracturing fluids is given below in Table 5.

14 SPE-173776-MS


15/35

Table 5—Summary of experiments performed with pseudo-crosslinkers. (R“regular” and PE”pyroelectric” nanoparticles)

SPE-173776-MS 15


16/35

VES technology developments suggest that polymer-based fracturing fluids may be replaced in some

regions (Crews et al. 2008a). Advances such as internal breakers and pseudo crosslinkers coupled with the

low molecular weight of VES have led to fluids that can simultaneously control leak-off, maintain high

viscosity, and controllably break and reduce viscosity in order to flow back the fluid. In a further

development, fluid loss in pseudo-crosslinked systems can be further reduced by adding 0.2 to 10 vol%

mineral oil (Huang and Crews 2009b; Huang et al. 2010a). The oil droplets collected in the filter cake over

time along with the nanoparticles and micelles, and the fluid loss continued to improve. The rheological properties of the fluid were unchanged, which was unexpected since it was previously thought that

hydrocarbons disrupt micelles and reduce the viscosity of VES-based solutions. Pseudo-crosslinked VES

fluids have also been used to remove residual polymer from hydraulic fractures ( Crews and Huang 2010).

Polymer breakers such as oxidizers or enzymes are combined with the nanoparticle pseudo-crosslinked

VES matrix during the initial fluid mixing and are believed to react more slowly than in the absence of

a VES network. The breakers are carried deeper into the fractures with the VES gel and can make better

contact with the polymer residue.

Nanoparticles have not only been used to enhance the viscosity of VES-based fluids but also dual

systems which incorporate both surfactants and polymers. For example, Fakoya and Shah (2003) studied

the following four systems to which 20 nm SiO2 nanoparticles were added: (1) 5% surfactant in 4% KCl,

(2) 33 lb/Mgal guar in 4% KCl, (3) 75 vol% surfactant and 25 vol% guar polymer and (4) 25 vol%surfactant and 75 vol% guar polymer. Rheological data in the form of viscometry and frequency sweep

testing was collected in the temperature range of 75-175 °F. The rheological properties of systems (1) and

(2) were both enhanced by particles added in concentrations of 0.24 wt% and 0.4 wt%. Dual systems (3)

and (4) were enhanced with nanoparticle concentrations up to 0.058 wt% and 0.24 wt%, respectively.

Dual purpose pseudo-crosslinking particles which improve the performance of the hydraulic fracturing

fluid but also control formation fines migration have been described (Huang and Clark 2013). Nanopar-

ticles are added to the VES-based fluid along with an internal breaker. As the fluid system is pumped into

the formation to create fractures, a filtercake is developed on the fracture face by the pseudo-crosslinked

micelle networks. As the internal breakers are released, the worm-like micelles are collapsed into spherical

micelles. The nanoparticles are released and precipitate, attaching to nearby proppants and thereby

capturing formation fines. A successful field treatment was performed in the Gulf of Mexico utilizing

similar nanoparticle technology for fines control (Huang et al. 2010b).

As previously described, particles composed of ZnO, MgO, TiO2, or Al2O3 enhance the viscosity of

VES-based fluids. It has been shown, however, that particles composed of certain other materials have the

opposite effect upon these gels. For example, inorganic semiconductor particles such as cupric oxide,

cuprous oxide, silicon, silicon carbide, germanium, gallium arsenide, indium antimonide, and gallium

nitride are found to reduce the viscosity of the gelled aqueous fluid (Huang 2014). The reaction is

described as being either transition-metal-catalyzed or transition-metal-mediated, although the exact

mechanism is not reported. Organic semiconductors such as pentacene, anthracene, rubrene, etc. are also

observed to have similar effects upon VES gels as the inorganic semiconductors. Similarly, metal ions

present in the aqueous treating fluid are found to break, reduce, and/or digest the VES within the aqueous

treating fluid (Crews and Huang 2013). Addition of nanoparticles composed of materials that pseudo-

crosslink VES micelles along with a complexing agent such as ethylenediaminetetraacetic acid (EDTA)

was reported to prevent redox reactions between VES molecules and the metals ions in the fluid.

A.3. Internal Breakers

Hydraulic fracturing fluids rely on a breaking mechanism in order to reduce the viscosity of the carrier

fluid and flow back the well. Polymer-based systems utilize enzymes, oxidizers, acids or, more recently,

decrosslinking agents to break the fluid and reduce the viscosity. Initial VES-based fluids took advantage

of two different reservoir conditions in order to break the fluid ( Huang and Crews 2008b). The first

16 SPE-173776-MS


17/35

condition is a change in brine concentration. VES-based fluids are seen to be stable over a particular

concentration of salts and by diluting this concentration with produced fluids, the gel can be broken. The

second condition that will break VES-based fluids is their contact with reservoir hydrocarbons. When

hydrophobic substances such as oil or gas dissolve in the hydrocarbon core of the micelle the structure

swells and breaks into smaller spherical micelles. This loss of larger worm-like micelles results in the

viscosity of the solution being reduced. Since hydrocarbons from the formation have this effect upon the

micelles, an internal breaker was not utilized to reduce the viscosity in order to flow back the fluid.However, depending upon these reservoir conditions is unreliable and, in some cases, additional treat-

ments had to be performed in order to clean up the formation. Data indicates that 20% of the treatments

performed by these VES-based fluids in the 1990s required remedial actions (Crews et al. 2008;

Al-Muntasheri 2014). Hence, the need was established to develop internal breakers that would administer

a controlled break of the VES fluids in a manner that is comparable to conventional polymer-based

systems. However, in this case the breaking would not deposit residue into the formation and proppant

pack because of the solids-free characteristic of the VES.

A number of methods for controlling the viscosity of VES-based fluids have been observed. First, a

method for controlling or delaying the onset of gelation in VES-based fluids after the fluid has been mixed

has been seen (Hughes et al. 1999). Three different mechanisms can be used to control the fluid viscosity:

delayed release of a specific counter-ion, change in hydrogen bonding, or a modification of the solution’sionic composition. Another method describes an intervention at the surface which can reversibly break the

viscosity of VES solutions that are used in a drilling application (Rose et al. 1988). When the fluid is

pumped into the well, it is viscous enough to carry cuttings to the surface, and the techniques developed

for breaking the fluid can be employed at the surface for solids removal. Some of these methods include

changing the temperature of the fluid, contacting the fluid with hydrocarbon, and adjusting the pH. These

treatments are reversible, and viscosity can be restored to the fluid. A third example of a system that can

be used to control the viscosity of the VES-based fluid involves the controlled addition of components that

decrease the viscosity ( Nelson et al. 2005). The breaker can be either internal or external such as

precursors which release a component such as alcohol, salt or organic acid by one of the following

mechanisms: slow dissolution, melting, reacting with a compound that is present in the fluid or added to

the fluid, or breaking a coating.

Laboratory studies have shown that VES-based fluids can be broken by internal phase breaker

technology. The aqueous breaker solutions have shown controlled viscosity reduction in the temperature

range of 80 °F to 225 °F (Crews 2005). The break times observed for these fluids were comparable to

crosslinked-polymer hydraulic fracturing fluid systems, including times as short as 15 minutes. The

breakers were also functional over a wide range of salinities such as 3 wt% KCl, 12 wt% KCl, and ASTM

synthetic seawater, with increasing amounts of breaker required. A second study was performed which

compared the laboratory results of VES fluids with and without internal breaker (Crews and Huang 2007).

Berea core cleanup tests showed that little pressure or time is needed to initiate cleanup when the internal

breaker is used. By contrast, the unbroken VES fluid requires five times higher pressure than just the

partially broken VES fluid. Likewise, nitrogen gas core cleanup tests were performed where cores were

shut-in (left static) for 24 hours then displaced with nitrogen gas over a 48 hour period. The results showed

that the VES fluid that was internally broken was readily producible with gas, but the unbroken

VES-based fluid was very difficult to produce. Another laboratory study incorporating an internal breaker

was performed (Crews et al. 2008b). By varying the concentration of the internal breaker, complete

viscosity reduction can be achieved at 250 °F. Core-regain permeability tests were performed where

VES-based fluid without any internal breaker showed just 2% regain permeability after 48 hours of low

pressure gas flow. The VES fluid with internal breaker, however, had a regain permeability of 141%.

The specific mechanisms that have been utilized for breaking VES-based gels have been described. A

microemulsion system has been developed which performs a variety of functions in addition to the

SPE-173776-MS 17


18/35

viscosity reduction of VES-based fluids, namely solubilization of by-products generated from breaking

and desorption/water-wetting to prevent VES residue from depositing in the formation (Crews 2010a). An

example of one such microemulsion system includes unsaturated fatty acid oil, glycol, a sorbitan

ester/ethoxylated sorbitan ester mixture, and an alkyl sulfonate. VES-based fluids can also undergo

viscosity reduction by a metal ion source and, optionally, either a second metal ion source, reducing agent,

or a chelating agent (Crews 2009). The reason that the fluid’s viscosity is reduced is described as one of

the following mechanisms: rearrangement of the micelle structure, deaggregating the VES micelle

structure, chemically altering the VES molecules, or a combination of these mechanisms. These potential

breaking mechanisms are also at play in systems that are broken by unsaturated fatty acids such as

monoenoic acid and/or polyenoic acid (Crews 2010b). For example, an amine oxide surfactant can be

broken with an oil that contains a large amount of unsaturated fatty acids such as soybean oil, fish oil, or

flax oil. The unsaturated fatty acids are described as auto-oxidizing into ketones, aldehydes, and saturated

fatty acids that break the VES fluid. Similarly, addition of specific types of bacteria such as Enterobacter

cloacae, Pseudomonas fluorescens, Pseudomonas aeruginosa have been observed to reduce the viscosity

of amine oxide surfactants such as TAPAO (Crews 2006). The mechanism for breaking may follow one

of two pathways: the VES micellar structure is directly rearranged or disaggregated or, alternatively, other

materials in the viscosified fluid may be degraded to form by-products that reduce the viscosity of the gel.

A.4. Foamed & Emulsified VES

Due to the desire to reduce the amount of water used in fracturing fluids, systems which incorporate both

viscoelastic surfactant and carbon dioxide to form a fracturing fluid have been developed in recent years.

These VES-CO2 fluid systems combine the benefits of VES fluids such as good proppant transport, low

formation damage, and low friction pressures with the enhanced cleanup and better hydrostatic pressure

of carbon dioxide. Historically, aqueous/CO2 mixtures have been characterized as foams. However, CO2is pumped as a liquid at surface conditions, so it may be more appropriate to refer to these systems as

emulsions (Chen et al. 2005). The VES fluid-CO2 mixture can be thought of as two liquids with limited

miscibility but dispersed stably in each other (an emulsion). Since the two components remain separate

phases, the CO2 will not disrupt the worm-like micelles and hence the viscoelasticity. As shown by Chenet al., foaming the VES with 70% CO2 roughly doubles the viscosity over that of 4% VES straight fluid.

The use of this system for field applications has been reported (Chen et al. 2005; Arias et al. 2008;

Al-Muntasheri 2014). Zwitterionic surfactants such as the betaines shown in Figure 7(b) and 7(c) and

carbon dioxide in a separate phase along with a cosurfactant such as C12 alkyl dimethyl benzyl ammonium

chloride have been described (Chen et al. 2007 and 2010).

There are several published case histories of wells that were fracture stimulated with VES-based

foamed fluids. VES foam is a significant unconventional fracturing fluid for tight gas reservoirs ( Gupta

2009). Combination of an anionic surfactant such as sodium xylene sulfonate with a cationic surfactant

such as N,N,N-trimethyl-1-octadecammonium chloride in ratios of 1:4 to 4:1 will form a viscoelastic gel

that is combined with 53% to 96% or more of carbon dioxide by volume (Zhang et al. 2002). Foamed

surfactant gel treatments of this nature were used in 75 producing formations (3,100 treatments) between 1998 and 2005 in the Western Canadian Sedimentary Basin (Gupta et al. 2005). Both CO2 and

N2 were used as the internal phase for the treatments, with N2 being used about 90% of the time. Proppant

concentrations over 800 kg/m3 (6.7 ppg) were used in about 35% of the wells (1,100 fracture stimula-

tions), where some jobs placed up to 1,200 kg/m3 (10 ppg). In about 10% of the work, proppant

concentrations lower than 600 kg/m3 (5.0 ppg) were used because of either lower surfactant concentra-

tions or the specific reservoir characteristics. Less than 20% of the CO2 foamed surfactant gel treatments

result in screen-outs, which is similar to non-energized surfactant gelled fluids or non-energized linear

guar gelled fluids. By comparison, surfactant gels foamed with nitrogen screen out 15% of the time. These

18 SPE-173776-MS


19/35

numbers are comparable to foamed conventional cross-linked gelled treatments which screen out 18% of

the time (Gupta et al. 2005).

Additional field cases describing the application of VES-CO2 based fluids have been reported (Table

6). VES-CO2 systems were used to improve the Olmos production in the Caterina SW field in Texas

(Semmelbeck et al. 2006). One test well was pumped with VES-CO2 then ten additional wells were

stimulated in this area. The VES-CO2 (70% quality) was pumped at 6 to 12 bbl/min, typically with 25,000

lbm of proppant (1 to 5 lbm proppant added per gallon of fluid). The propped fracture half-lengths

averaged 425 ft, with minimal height growth. The outcome from 11 treatments resulted in an average

production of 430 Mscf/D and 82 BOPD in zones that were previously bypassed. In another case, a

VES-CO2 based fluid was applied in the Frontier Formation, Big Horn Basin, Wyoming where two wells

have been reported utilizing fracturing fluids with high foam quality ( Bustos et al. 2007). The fracturing

treatment in each well included the following: 2,000 gallons of a 10% surfactant solution, followed by a

PAD of 3% VES-CO2 at 70% foam quality and then 2.5% VES-CO2 as carrier fluid for all 20/40 Jordan

sand stages from 1 to 5 PPA, flushed with 1% VES-CO2. The total amount of proppant in the two

treatments was 61,000 lbs and 63,000 lbs, with an injection rate of 30 BPM. P3D (Pseudo three-

dimensional) fracture simulation demonstrated a fracture half-length of about 338 ft with an average

conductivity of 587 md-ft for the first well and a fracture half-length of about 511 ft with an averageconductivity of 553 md-ft for the second well. Furthermore, wells in Waltman field in Wyoming provide

another example of VES-CO2 fluid systems (Arias et al. 2008). Four of the wells that were treated with

VES-CO2 fluid (70% quality) were compared with nearby wells treated with linear hydroxypropyl guar

(LHG) polymer system with gel loading of 40 lbm/mgal. Initial production from the wells treated with

VES-CO2 was observed in the range of 5 to 7 MMcf/D, which is higher than the gas rates of 2 MMcf/D

seen from wells treated with the polymer-based fluid. Estimated fracture lengths for the two systems are

estimated to be similar because of the similarities in the viscosities and hence the proppant transport

characteristics. The difference in production is attributed to the clean nature of the VES-CO2 and its

minimal proppant pack damage versus the LHG polymer system. A final example of the application of

VES-CO2 in the field was in the Morrow Sands in Southeast New Mexico ( Pandey et al. 2007). Three

Morrow completions covering eight stages were successfully pumped with VES-CO2. The fluid system

consisted of 4.5 vol% VES (for the pad; less for later stages) along with 70% liquid CO2 as the fracturing

fluid. In the first of the wells, 20/40 US Mesh size ceramic proppant was used in the treatments at 27,000

lbs (1 to 2.5 ppa). The pseudo 3D fracture simulator determined an average fracture length of 605 ft for

the four stages along with a propped fracture width of 0.05 inch and conductivity of 1385 md-ft.

SPE-173776-MS 19


20/35

A nitrogen-foamed viscoelastic surfactant-based system for use in unconventional natural gas wells has

been developed that is a coal/carboniferous shale-compatible solids-free (CCSF) fluid (Fredd et al. 2004).

The CCSF fluid showed greater than 70% retained permeability and cleanup factors less than 100

psi-min-K with formation coal samples from seven different basins in North America. By comparison,

conventional slickwater and polymer-based fracturing fluids exhibited significant coal pack damage with

retained permeabilities as low as 46% and 18%, respectively. Furthermore, the CCSF fluid exhibited a59% one-year cumulative production increase relative to conventional polymer-based fluids in the

Devonian Shale formation.

A.5. Brines & Produced Water

VES fluids have been observed to be compatible with high-density brines which make them attractive

candidates for a variety of applications. As seen in Table 5, VES fluid formulations commonly contain

high salt content. An example of a high density fluid that is suitable for a broad range of applications

contains erucylamidopropyl betaine or oleylamidopropyl betaine, an alcohol such as methanol, and at least

12.5 ppg of a salt or mixture of salts of divalent cations such as chlorides or bromides of calcium and zinc

(Fu et al. 2006). The structures of these zwitterionic surfactants are shown in Figure 7(b) and 7(c). The

same surfactant in high-density brine fluid has been further applied as a perforation fluid (Samuel et al.

2011). Similar chemistry is described in the development of a fluid loss or lost-circulation-control pill

(Samuel et al. 2007). These particular surfactants, erucylamidopropyl betaine and oleylamidopropyl

betaine, have furthermore been combined with arylalkylsulfonate cosurfactant, and a polar solvent such

as water, alcohol, or glycol (Berger and Berger 2006). Viscosities are reported from 400-900 cp over the

temperature range of 20-100 °C for surfactant solutions containing 30 wt% CaCl2. These fluids are

described as having applications in fracturing, acidizing, gravel packing, and other similar operations. In

a second example by Berger and Berger (2008), an amphoteric alkyl amido betaine surfactant is used to

viscosify an injection brine and reduce the interfacial tension between water and oil. The aqueous fluid

is pumped into an injection well, displacing hydrocarbons into the production well.

Table 6—A summary of the published case histories of wells that were fracture stimulated with VES-based foam fluids. The foam

quality in each treatment is 70%.

20 SPE-173776-MS


21/35

The development and first field application of these VES fluids for lost circulation applications have

been described (Samuel et al. 2003). Typical fluid-loss control systems are composed of high concen-

trations of polymers which are viscous and form a filter cake on the face of the rock. However, the

difficulty in cleaning up these systems led to the development of a solids-free fluid-loss pill. This

VES-based system is compatible with more completion brines that are used in well completions and was

demonstrated in the laboratory to be stable up to 375 °F. Around 10-20% of zwitterionic viscoelastic

surfactant is mixed with heavy brine such as 12.5 ppg CaBr 2. At high temperatures, around 5% methanolis also added in order to stabilize the system. A second application of brine-based VES fluids includes

nonionic surfactants for frac packing (McElfresh et al. 2003). In the Adriatic Sea and in the Gulf of

Mexico, non-ionic VES gels based on brines up to 10.5 ppg CaCl2 were used in over 30 treatments.

Newer methods have also been developed to stabilize VES fluids in high-density brines. For example,

low molecular weight surfactant polymers can be used to control the curvature of surfactant micelles (van

Zanten 2011). Tethered polymers, which are nonionic surfactants, can be added to cationic/anionic

VES-based systems in order to stabilize the brine solutions and maintain the viscosity. Three different

CmEn alkyl poly(ethoxylate) nonionic surfactants (E10, 20, and 100) were examined for their effects on

five different cationic/anionic VES systems. The tethered polymer was observed to increase the viscosity

of the cationic/anionic VES systems in a variety of brine concentrations as well as prevent the precipi-

tation of surfactants at high brine densities. Further experimental work on similar tethered polymer systems was performed on cationic alkyl quaternary ammonium salt and cationic/anionic alkyl amine/

alkyl sulfate salt (van Zanten and Ezzat 2011). For the latter system, maintaining a viscosity of 1000 cp

in high-density brine had a temperature limit of 225 °F.

Viscoelastic surfactants such as N,N,N, trimethyl-1-octadecammonium chloride and sodium xylene

sulfonate can also be used in combination with produced water (Gupta and Tudor 2005). A Case History

has been published from the Western Canadian Sedimentary Basin where 50 individual wells were treated

with VES-flowback water (Gupta and Hlidek 2010). It is preferred to use recycled water in which VES

was used in the first treatment. The concentration of chemicals needed for the second treatment is lower

than the first treatment, with the loading reduced from 210 to 120L for a typical job. It has been observed

that the anionic component of the VES fluid often remains in the flowback water while the cationic

component is believed to be adsorbed on the clay components of the formation. A cost savings of 12%

was realized from the project, and the well production performance was unaffected.

B. Matrix Acidizing

Matrix acidizing is extensively used to enhance the productivity and injectivity of wells drilled in

carbonate reservoirs. In matrix acidizing, acids are injected at pressures less than the fracturing pressure

of the formation in order to dissolve part of the rock, remove the damage and create open flow paths for

the reservoir fluids to flow through. Hydrochloric acid (HCl) is the main fluid used in matrix acidizing

both because of its inexpensive nature and because the reaction products are soluble in water and cause

no subsequent damage to the formation. The main issues with HCl are its incompatibility with some crude

oils and its fast reaction with rocks at temperatures above 200°F. These short reaction times will result in

acid spending, face dissolution and failure of the treatment. Another issue with HCl is its inability to

stimulate heterogeneous zones with large changes in rock permeability. HCl will invade the high

permeability zone and little acid will be diverted to the low permeability zones and zones with high

damage. This problem is more pronounced in acidizing long horizontal wells. Acid diversion techniques

are necessary in such cases to guarantee the proper distribution of acid and to obtain better acidizing

results.

Several approaches can be employed to enhance the acid diversion to assure better acid distribution.

Injection rates can be increased during the treatment to increase the injection pressure and help to divert

acid in the low permeability regions (Alleman et al 2003). Acid diversion can also be accomplished using

SPE-173776-MS 21


22/35

mechanical diverters such as coiled tubing, ball sealers, rock salts and acid flakes, and using plugs and

packers for isolating different zones (Economides and Nolte 1989). The main disadvantage of these

techniques is their inability to work in large permeability contrast formations (Lynn and Nasr-El-Din

2001). Chemical diverting agents are also available for acid diversion. Use of foamed fluids like foamed

KCl solutions, foamed ammonium chloride solutions and gelled pills were reported by Zerhbouh 1993;

Zeilinger et al. 1995; Taylor and Nasr-El-Din 2001. The main limitation of these foamed and gelled

systems is their instability at temperatures above 200°F and their ineffectiveness in plugging formationswith permeabilities above 500 md (Alleman et al. 2003). Gelled acids and cross-linked gelled acids are

used as chemical diverters to guarantee the proper distribution of acid in heterogeneous reservoirs (Gomaa

et al. 2010).

Viscoelastic surfactants have also been developed and applied for acid diversion (Chang et al. 2001a;

2001b). As mentioned before, viscoelastic surfactant based fluids develop their viscous nature through the

formation of micellar structures. Also, the pH value will control the rate of building the viscosity of

VES-based acid systems. At low pH values, the viscosity of VES-based acids is very low allowing the acid

system to flow and penetrate into the formation. Upon the reaction of acid with carbonates, the pH

increases and the concentration of divalent cations in solution increases and therefore the VES-based acid

system starts to build viscosity. At pH around 4, the viscosity of the spent acid will be high enough to

divert the fresh acid to low permeability un-contacted zones and fresh acid starts to penetrate and reactwith these rocks to form wormholes ( Nasr-El-Din et al. 2006b; Crews and Huang 2007; Crews et al. 2008;

Huang et al. 2008b; Yu et al. 2011). The viscosity of VES fluids can be reduced upon mixing with

hydrocarbons. This may make it unfavorable when used with dry gas wells. Surfactant based acids were

introduced in the petroleum industry by Chang et al. (2002) and Qu et al. (2002).

B.1. Experimental Studies

Alleman et al. (2003) developed a VES diverting agent with a vesicle structure type. The VES was stable

at temperatures up to 250°F and also up to 350°F by adding a material referred to as an intensifier to

interact with VES molecules and enhance the charge on the micelles. Adding this intensifier in quantities

of 0.2 to 0.3 wt% made the structure of the VES larger, stronger and more stable at temperatures up to

350°F. At a temperature of 250°F, the viscosity of VES diverting agents with 0.2 polyquat at 100 s -1shear

rate was 130 cp, while the viscosity increased to 330 cp when the polyquat concentration was increased

to 0.4 wt%.

Al-Ghamdi et al. (2004) studied the effect of different acid additives on the rheology of VES-based

acid systems. The acid system was 15 wt% HCl with 6.0 vol% VES. Additives like corrosion inhibitor,

iron, nonionic surfactant, anti-sludge agents and hydrogen sulfide scavengers were tested. They found

that, at 100°C and 87 s-1 shear rate, the viscosity of spent VES based acid was increased from almost zero

to 130 cp by increasing the VES concentration from 1 to 6 vol%. Also, they noticed that addition of a

mutual solvent and iron (III) with concentrations higher than 1000 ppm caused reduction in the apparent

viscosity of the VES based acid systems. This finding is important since mutual solvent as well as

hydrocarbon fluids can be used to break down the remaining VES gel in the formation for better cleanup

and better performance. Care should be taken to prevent contact of the VES-based system with mutual

solvent during injection of the treatment in the well. Also, corrosion product addition needs to be

controlled to prevent effects upon the rheology of the VES-based acid systems which in turn will affect

the performance of the diversion system.

Nasr-El-Din et al. (2006a) studied the rheology and diversion ability of VES based acid systems. Table

7 shows the relationship between the VES concentration and the apparent viscosity of the formulation

measured at 40 s-1 and at 200°F and 250°F temperature. It is clear that the apparent viscosity increased

when increasing the VES concentration. A parallel core flow test was performed using 711 permeability

ration contrast core samples to study the acid diversion in such core samples at 200°F. They noticed that

22 SPE-173776-MS


23/35

acid breakthrough occurred first in the low permeability core sample where one main wormhole was noted

by using the CT scanner. This occurred as a direct result of the effective diversion of acid that was

achieved by the VES formulation.

Lungwitz et al. (2007) experimentally studied the use of VES based fluids as diverting systems in

acidizing treatments. They performed coreflood experiments using carbonate core samples with perme-

ability in the range of 0.1 to 50 md at a temperature of 200 to 240°F using 15 wt% HCl acid systems. Fluid

loss was examined using limestone samples of permeability in the range of 1 to 3 md at a temperature of 150°F using 2 wt% KCl brines. It was found that both cross-linked based fluids and VES based fluids

showed similar leak off properties as well as the same initial viscosity (30 cp at 170 s-1 and 70°F) Using

VES based acid systems, acid breakthrough was noticed in limestone and dolomite core samples after

injection of about 1 to 1.6 pore volumes of acid, while using cross-linked polymer acid systems did not

achieve breakthrough when tested under similar conditions. Also, VES based acid systems were compared

to 15 wt% plain HCl acids using the conductivity cells. The VES based HCl system created etched

surfaces which enhanced the conductivity of the fractures, while HCl alone caused face dissolution and

fewer enhancements in fracture conductivity.

Huang et al. (2008a) examined the use of VES-organic acid systems for acid treatments in carbonate

formations. The viscosity of the fresh mixture of 2.0 vol% VES – 10 wt% organic acid – 0.2 wt% internal

breaker at 100 s-1 shear rate and pH of 3 was almost zero. Upon acid spending, the pH increased to around

6, and the viscosity of the spent acid increased to 210 cp at 100 s-1. Also, they found that the use of

internal breaker is very important to break down the remaining VES structure to better clean up the

formation after the acid treatment.

Yu et al. (2011) studied the retention of VES surfactant in the porous media through using coreflood

experiments. The VES-based acid system used in their study consisted of 15 wt% HCl, 7 vol% VES and

0.3 vol% corrosion inhibitor. The acid injection rate was found to affect on the volume of acid

breakthrough and the wt% of VES retained inside the core sample. Figure 9 shows a relationship between

volume to breakthrough and the retained surfactant (wt%) as a function of the acid injection rate. As the

acid injection rate increased both the volume to breakthrough and the retained surfactant in core sample

decreased. They furthermore described that VES can be removed by the hydrocarbon fluids produced after the treatment such that there is no need for a cleanup treatment to remove the remaining gel. In the event

that the cleanup after the treatment is not complete, they recommended the use of an internal breaker

system or a mutual solvent as a post flush to break down the remaining VES.

Table 7—Rheology data measured by Nasr-El-Din et al. (2006a)

VES CONC. (vol%) VISCOSITY Water, 200 °F (cp) VISCOSITY Water, 250 °F (cp) VISCOSITY CaCl2

,250 °F (cp)

0.5 138 113 118

1.0 152 - 141

1.5 - 152 160

3.0 433 312 -

SPE-173776-MS 23


24/35

Wang et al. (2012) studied the rheology of VES based system designed for high temperature

applications. They found that the viscosity of the spent acid (at a pH value around 4.5) was around 200

cp at 325°F for 4 vol% VES-based fluid at 10 s -1 shear rate. This will help the acid plug the high

permeability zones upon acid spending and divert the fresh acid to the low permeability zones.

Yu et al. (2012) studied the effect of hydrolysis on the rheology of VES-based acids systems. Under

high temperature and in acidic environment, the CO-NH bond (peptide bond) is broken and the system

is hydrolyzed. This bond breakage resulted in the formation of smaller molecular weight molecules which

were responsible for reducing the viscosity of the VES-based acids systems. At temperatures of 190°F, an

increase in the apparent viscosity of the VES based acid system was observed for hydrolysis times less

than 2 hours. For hydrolysis times greater than 3 hours, the acid breaking rate is very high and phase

separation was noticed. This resulted in a large decrease in the acid viscosity. Yu et al. (2012) shows the

VES-based acid phase separation as a function of hydrolysis time at 190°F for 4, 6, and 8 vol% VES

concentration (Figures 3-5 of the manuscript). Similar behavior was noted at each surfactant concentra-

tion. He et al. (2013) studied the effect of the hydrolysis of VES-based acids on the rheology and the

performance of such systems at 250°F. As the degree of hydrolysis increases, more breaking takes place

and less damage occurred in the high permeability core samples. Figure 10 shows the effect of hydrolysis

on the apparent viscosity of a VES solution prepared using 4 vol% VES and hydrolyzed at 190°F. The

results of these studies highlight the importance of proper design of VES-based acid systems in order to

prevent breaking the acid system before acid treatment is complete.

Figure 9—Change of volume of acid to breakthrough and retained surfactant (wt%) as a function of acid injection rate.

24 SPE-173776-MS


25/35

B.2. Case Histories

VES-based diverter was applied in several field trials, and a summary of the reported trials is provided

in Table 8. One of these cases was described by Alleman et al. (2003). The temperature of the well was

198 °F, and the well depth was 14500 ft including an oil bearing formation that was 53 ft. The VES

diverter pill was mixed with 10 wt% HCl and 5 wt% acetic acid, and the treatment took place in stages.

The injection pressure increased from around 300 psi to 1,400 psi upon injection of the diverting VES pill

as a direct result of its high viscosity and its ability to plug the high permeability zones. In the final stage

of injection, the pressure increased by 200 psi upon the injection of the VES diverter pill. A second case

study was also reported by Alleman et al. (2003) with similar results.

Figure 10—Effect of hydrolysis time and shear rate on the VES-based acid systems formulated using 4 vol% VES at 190°F (He et al.

2013).

Table 8 —Summary of field trials performed with VES-based acid systems. *The bbl/day rate for this well is the injection rate.

LOCATION WELL INFO VES (vol%) HCl (wt%)

PRODUCTION

(bbl/day) REFERENCE

Gulf of Mexico 198 °F, 14,500 ft - 10 5 Aceti c “Above expec ted

rates”

Alleman et al. 2003

Gulf of Mexico 150 °F, 9,600 ft - 10 5 Aceti c “Above opera tor

expectations”

Alleman et al. 2003

- Mainly calcite 6 20 6,000 Mohammed et al.

2005

Saudi Arabia Water Injector*

Horizontal

3 20 75,000 (450% inc.) Nasr-El-Din et al.

2006aSaudi Arabia Tubing: 6,270 ft

Vertical

6 20 (150% inc.) Nasr-El-Din et al.

2006b

Saudi Arabia - 6 20 3,700 Nasr-El-Din et al.

2006b

Saudi Arabia 200 °F, 9,310 ft

3,270 psi

4 20 - Nasr-El-Din et al.

2006b

Niagaran Reef (MI) Mainly dolomite - 15-20 80 (400% inc.) 2X

acid alone

Kreh 2009

Bombay High

Oilfield, India

5-500 md 3 20 800 (150% inc.) Rawat et al. 2014

SPE-173776-MS 25


26/35

Another successful treatment using VES-based acids was described by Mohammed et al. (2005). The

treatment was performed on a vertical oil-producing well where coiled tubing was used to inject the acid

to increase the efficiency of diversion in the wellbore section. The treatment included several steps: (1)

a preflush of 6% mutual solvent in diesel, (2) 20 wt% HCl with 0.4 vol% corrosion inhibitor and 0.4 vol%

H2S scavenger, (3) nitrified 20 wt% HCl with the same amount of corrosion inhibitor and H 2S scavenger

with nitrogen at a ratio of 400 Scf/bbl, (4) 6 vol% VES, 20 wt% HCl, and 0.4 wt% corrosion inhibitor with

nitrogen at a ratio of

Recent Advances in Viscoelastic Surfactants for Improved Production from Hydrocarbon Reservoirs.pdf

Documents