Interaction Effects of Predominantly Linear and ... - ACS Publications

Subscriber access provided by FORDHAM UNIVERSITY

Fossil Fuels

Interaction Effects of Predominantly Linear and Branched Waxes on Yield Stress and Elastic Modulus of Waxy Oils Thiago O. Marinho, Carla N. Barbato, Gizele B. Freitas, Angela C. Duncke, Marcia Cristina Khalil de Oliveira, and Márcio Nele Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.8b00761 • Publication Date (Web): 29 Jun 2018 Downloaded from http://pubs.acs.org on June 29, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

Interaction Effects of Predominantly Linear and Branched Waxes on

2

Yield Stress and Elastic Modulus of Waxy Oils

3

Thiago O. Marinho1, Carla N. Barbato2, Gizele B. Freitas1, Angela C. Duncke2,

4

Márcia C. K. de Oliveira3, Márcio Nele1,2

5 6 7 8 9

1

PEQ/COPPE - Universidade Federal do Rio de Janeiro, Cidade Universitária, RJ/Brasil 2

EQ/DEQ - Universidade Federal do Rio de Janeiro, Cidade Universitária, RJ/Brasil 3

Petrobras/CENPES - Cidade Universitária, RJ/Brasil

10

In order to assess the relationship of wax chemical structure and viscoelastic properties of

11

waxy gels, model oils composed of single and blended waxes were prepared at a fixed

12

concentration of 7.5 wt% and then gelled. The investigation encompassed four different

13

well-characterized commercial paraffin waxes, solubilized in a mineral oil matrix. Previous

14

rheological measurements pointed that these systems reproduce essential features of

15

crude oil gels (e.g. gel-like mechanical response) when gelled. Among the employed waxes,

16

two are predominantly linear whereas the others are non-linear branched molecules. The

17

waxes were molecularly characterized to aid in the investigation by means of GC-FID, 13C-

18

NMR, DSC, FT-IR, and XRD. Rheological properties were measured at a controlled-stress

19

rheometer by oscillatory shear experiments. Polarized light microscopy was adopted for

20

morphological characterization of precipitated crystals. It was found that yield stress and

21

elastic modulus at linear viscoelastic region are highly correlated (R2 = 0.94). For single wax

22

systems, the increase in chain length resulted in a yield stress increase. Although, there is a

23

competitive effect among chain length (positive effect) and branching content (negative

24

effect). The results indicated that for blended systems the small-chain linear wax was able

25

to interact favorably with the long-chain non-linear wax, possibly due to its ability to

26

accommodate within the later molecule, ensuring the highest yield stress value (630.2 Pa).

27

The wax structural arrangement of 37 carbon atoms on average, including approximately

28

three tertiary carbons, was effective for lowering the yield stress of particular blended

29

systems. The lowest viscoelastic properties were measured for a blended system composed

30

by non-linear waxes, which was also characterized by the smallest and rounded crystals

31

visible by optical microscopy.

ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

Keywords: wax, model oil, yield stress, rheology

2 3

1. Introduction

4 5

Flow assurance is a major concern to petroleum industry especially in offshore

6

platforms and arctic environments. In this regard researchers generally address the

7

efforts towards rheology, DSC, and microscopy analysis in order to unveil relevant data

8

of waxy crude flow under cooling regimes1–4. Also, a relevant effort is being devoted to

9

model the oil intricate rheological behavior under gelation conditions and yield stress

10

prediction. Cooling rate, shear rate, shear stress, solid fraction and aging time are the

11

experimental variables usually assessed5–9. It is also important to consider the influence

12

of wax chemical structure on the yield stress since it is promoted by interlocking

13

networks maintained by London-Van der Waals forces among wax crystals10,11. Due to

14

petroleum multi-component nature and variable composition, the gel formation

15

mechanism is quite complex and wax structure plays an important role in this regard11.

16

The yield stress is an essential parameter to estimate the restart pressure of

17

pipelines filled with gelled oil12–14. At reservoir conditions (70 ~ 150°C and 8,000 ~

18

15,000 psi) petroleum has Newtonian flow behavior and waxes are solubilized15.

19

Although, wax crystal precipitation occurs during production, process and

20

transportation procedures, due to heat transfer from oil bulk to cold neighborhoods,

21

resulting in temperatures lower than the WAT or Wax Appearance Temperature16,17.

22

The cooling favors the fluid gelation as wax crystallization takes place with subsequent

23

deposition. This scenario leads to a strong waxy crystal interlocking network which is

24

accompanied by a drastic change in rheological properties, such as yield stress

25

appearance1,18. The issue is enhanced when a flow shutdown occurs (e.g. emergencies

26

or maintenance situations). In this case, the fluid undergoes a quiescent cooling and

27

partial or complete blockage of pipelines, damage to process equipment and

28

complications at pumping restart operations may arise3,14.

29

Waxes are divided into two main classes: macro and microcrystalline.

30

Macrocrystalline wax is regarded as the main responsible for deposition and gelation


Page 2 of 35

Page 3 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

problems during petroleum production. It is constituted primarily by n-alkanes with

2

few or no branches, H/C ratios between 1.96 and 2.05, exhibiting linear chains with 20-

3

50 carbon atoms length and melting point ranging to ~ 40-60°C19. On the other hand,

4

microcrystalline waxes are aliphatic hydrocarbon compounds containing branches and

5

rings, exhibiting also a broader carbon number distribution (~ 30-70 carbon atoms).

6

Due to the lack of large-scale crystallinity, they present wider melting point range (~

7

60-90°C)20.

8

Elucidating the composition of waxes is of utmost importance to explain their

9

behavior in liquid oil structuring. Doan et al. (2017)21 described the major chemical

10

components of seven natural waxes (such as sunflower wax and bee wax) followed by

11

evaluation of their oil structuring properties through oscillatory rheology. Model oils

12

were prepared ranging from 1.0 to 5.0 wt% of wax. The gel strength was evaluated in

13

terms of average elastic modulus at linear viscoelastic region and yield stress. It was

14

found that the main component dictating the gel strength are hydrocarbons. Free fatty

15

alcohols and wax esters also had a positive correlation with rheological properties,

16

although less pronounced than hydrocarbons.

17

According to a study performed by Bai and Zhang (2013),

22

yield stress

18

drastically decreases with the increase of average carbon number of wax for samples

19

prepared with crude oil added to single or blended waxes (regardless shear or

20

quiescent cooling conditions). They also observed smaller crystal size and simpler wax

21

structure for systems containing increasing average carbon number, which could

22

justify the lower yield stress. Zhao et al. (2012)23 employed macro and microcrystalline

23

waxes to prepare model oils solubilized in dodecane. Comparing 5.0 wt% systems,

24

macrocrystalline wax presented a much higher yield stress, indicating that non-linear

25

alkanes reduce gel strength. Although, none evaluation was made with blended waxes

26

in this sense. Han et al. (2016)24 investigated structural effects of precipitated wax

27

crystals (ranging from C18 to C45) on the yield stress of wax-decane gels. From HTGC

28

and XRD data, the authors devised a linear relationship between the conformation

29

disorder of Cn molecules and the degree of polydispersity. It was found that the

30

polydispersity of precipitated Cn has a negative impact on yield stress measured.

31

As stated by Senra et al. (2009)25, relevant effort on an industrial and academic

32

level has been dedicated to developing proper additives to weaken or break the gel


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

deposits, however, little work has been devoted to understanding how the

2

composition of the oil affects the formation of waxy gels. Thus the authors performed

3

a comprehensive study on the gelation characteristics of long-chained n-alkanes,

4

ranging from C36 to C24, dissolved in a short-chained n-alkane solvent (dodecane). The

5

authors concluded that for systems where cocrystallization occurs, the addition of only

6

a small amount (0.5 mass %) of a more soluble n-alkane reduces the gel and pour

7

points of the system. Cocrystallization is pointed as the cause of the formation of weak

8

points in the crystal structure as a result of the longer n-alkane contorting to allow for

9

the shorter n-alkane to be incorporated into the crystal structure. As a consequence,

10

decreases in the pour point and gel point of as much as 20 °C were assessed for

11

blended systems. The work is entirely focused on the role of n-alkanes on the

12

properties of waxy gels, although systems composed by cyclic or branched waxes are

13

not addressed. In a later investigation, Senra et al. (2015)26 prepared model oils

14

composed by decane, dodecane or hexadecane added to one or two n-alkane solutes.

15

The authors found out that solutes with shorter chains had much less impact on

16

gelation characteristics of the systems.

17

The chemical structure of waxes is of absolute importance in liquid oil structuring

18

phenomenon as the gelling behavior is governed by wax crystal interactions. A key

19

feature to evaluate the influence of wax chemical structure and crystal arrangement at

20

yield stress is the ability to assess structural and morphological properties. In this

21

regard, many authors described methods for wax physicochemical characterization.

22

Musser and Kilpatrick (1998)20 extracted and analyzed waxes from sixteen different

23

crude oils. By means of FTIR, C13NMR spectroscopy and DSC the authors were able to

24

propose average chemical structures for several waxes assessed.

25

Cazaux et al. (1998)27 investigated the gel structure of samples of waxy crude

26

using XRD, SAXS and cross-polarized optical microscope along with controlled stress

27

rheometer. According to the authors, the structure of waxy gels is determined by the

28

crystal shape and number density of wax crystals, both of which depend on the

29

temperature and cooling rate employed.

30

Zaky and Mohamed (2010)28 extracted and characterized high melting point

31

macro and microcrystalline waxes from crude residues of El-Ameria and Alexandria Oil

32

Company. The authors were able to characterize samples through multiple techniques


Page 4 of 35

Page 5 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

as GC-FID, XRD, DSC, TGA, SEM and 1H RMN verifying differences in their n-paraffin

2

contents, crystallinity, thermal characteristics, degrees of branching and crystal sizes. It

3

was found that microcrystalline wax exhibited melting point and thermal stability

4

higher than those obtained for macrocrystalline wax due to its higher boiling point

5

range and molecular weight. Also, macrocrystalline wax crystals appear in large and

6

loose needle form comparing to microcrystalline wax which crystallized in needle form

7

however as smaller crystals. The degree of crystallinity and degree of branching

8

(expressed as CH3wt%) for the waxes assessed were 83% and 14.7% (macro) and 63%

9

and 4.96% (microcrystalline).

10

Likewise, optical microscopy is generally employed to analyze and compare wax

11

crystal morphologies. From cross-polarized light technique, the sample contrast comes

12

from the rotation of polarized light through the crystalline material contained. Crystal

13

length, aspect ratio, and boundary fractal dimension can be obtained from these

14

microscopic images1,23,29–31.

15

The rheological properties of waxy gels have important technological

16

implications. Effective strategies to prevent and remediate paraffin deposition and wax

17

plugging, including the developing of more accurate deposition models, demands a

18

solid understanding of the relationship between structure and mechanical properties.

19

Thus, it is important to investigate how the gel characteristics are affected by oil

20

composition. In the present work, model oils composed by single and blended

21

commercial refined paraffin waxes of linear and branched nature were dissolved in

22

non-crystallizable mineral oil and cooled to ultimately produce a waxy gel. Rheological

23

measurements, previously performed by our group, showed that these particular

24

systems reproduce essential features of crude oil gels (e.g. exhibiting a low-

25

temperature gel-like mechanical response to an imposed low-frequency oscillatory

26

stress). The branched waxes added to the model oils are essential to obtaining a

27

system that closely mimics the gelation behavior of real crude oils. Though this is not

28

surprising, in light of findings on the composition of the real crude oils wax32, it seems

29

to be a lack of discussion on the role of branched waxes on the yielding properties of

30

waxy gels prepared from model systems. Our results indicated that for a better

31

interpretation of yield stress in terms of wax chemical structure, the chain length and

32

the degree of branching may not be taken separately, as they represent opposite


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

effects. By means of optical microscopy, it was possible to observe micrometer-sized

2

wax crystals exhibiting a marked tendency to associate into dense crystalline masses,

3

highlighting the low concentration of wax solids required for gelation. It is also shown

4

that elastic modulus is strongly correlated to yield stress for the systems assessed (as

5

both properties arise from the gelation phenomena).

6 7

2. Materials and Methods

8 9

2.1 Materials

10 11

2.1.1 Waxes. Four waxes were acquired to prepare model oils: L24 (Fluka Analytical

12

Reagents, Brazil), L29 (Sigma-Aldrich, USA), B37 (Fluka Analytical Reagents, Brazil) and

13

B53 (GM Waxes, Brazil). At this point, it is important to mention that the capital letter L

14

stands for (predominantly) linear whereas B stands for branched wax. The subscripted

15

numbers indicate the average carbon number calculated from GC-FID data.

16

2.1.2 Solvent. Spindle mineral oil, a non-crystallizable and n-alkane based oil (315°C

17

boiling point, ρ = 852 kg/m3 and µ=2.399*10-2 cP at 20°C) was kindly supplied by

18

Petrobras S.A.

19 20

2.2 Methods

21

2.2.1 Carbon Number Distribution. Gas chromatography was performed at 6890N

22

(Agilent Technologies) with on-column injection method and FID detector for waxes

23

carbon number distribution. For such, a melted silica capillary column with 100%

24

methyl silicone stationary phase was used. The experimental procedure was based on

25

the ASTM D5442 standard.

26

2.2.2

27

equipment at 75 MHz. 10 mm tubes and deuterated chloroform was employed (125

28

mg of wax to 2.5 ml of solvent). Pulse width and delay were 90° and 1 s, respectively.

29

Prior to analysis, waxes were solubilized at 120°C inside closed tubes. For L24 and L29

30

the experiment temperature was held at 40°C whilst 50°C was employed for B37 and

13

C-NMR analysis. Spectrums were obtained with Mercury 300 (Varian)


Page 6 of 35

Page 7 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

B53.

2

2.2.3 X-ray Diffraction (XRD). The X-ray diffraction patterns were recorded in the range

3

2Θ = 5 to 60° with a step size of 0.05 (2 Θ) every 2 s, in a Rigaku MiniFlex

4

diffractometer employing Cu Kα radiation with 1.54 Å wavelength and operated at 30

5

kV. Diffraction diagrams were recorded at 20°C.

6 7

2.2.4 FTIR analysis. Spectra were acquired using a Nicolet 6700-FTIR with a DTGS

8

(deuterated triglycine sulfate) KBr detector with 4.0 cm-1 resolution (16 scans) in the

9

range of 4,000 to 500 cm-1.

10 11

2.2.5 Differential Scanning Calorimetry. DSC 8500 (Perkin-Elmer) with N2 as a purge gas

12

at 50 ml/min was employed. Each sample (~ 30 mg) was placed in 50 μL pan with a

13

pierced cover which was tightly sealed. Experimental procedure comprised three

14

steps: [1] Heating ramp from 25°C to 80°C at 20°C/min followed by a 10 min

15

conditioning at 80°C; [2] Cooling ramp from 80°C to 4.0°C at 1.0°C/min followed by a

16

10 minutes conditioning at 4.0°C; [3] Heating ramp from 4.0°C to 80 at 1.0°C/min.

17

2.2.6 Model Oil Preparation. Wax solutions were prepared inside a 500 ml jacketed

18

beaker, previously heated to 80°C. The systems remained 5 minutes under constant

19

temperature (80°C) and mechanical stirring (RW20 from IKA@, 500 rpm) to ensure

20

homogeneity and solubility. The composition of 7.5 wt% was adopted for single or

21

blended wax solutions (in the later case 3.75 wt% of each wax was added). In order to

22

departure from the same thermal and shear history, the samples were freshly

23

prepared before each run. The remaining sample was used for microscopic

24

observation of wax crystals.

25 26

2.2.7 Rheological Experiments. The rheological tests were carried out at AR-G2

27

rheometer (TA Instruments) equipped with grooved concentric cylinders (Bob

28

diameter 28.05 mm, cup diameter 30.50 mm, Bob length 50.00 mm) to avoid apparent

29

wall slip phenomenon, which may underestimate yield stress values. The initial

30

temperature was set to 80°C prior to analysis. The procedure comprised 4 steps: [1]


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

conditioning at 80°C for 5 minutes to favors the sample thermal equilibration; [2]

2

dynamic cooling from 80°C to 4.0 °C at 1.0 °C/min and shear rate of 0.8 s-1; [3] time

3

sweep at 4.0°C for 15 minutes at oscillatory stress of 0.1 Pa and frequency 0.2 Hz to

4

favor structure build-up; [4] logarithmic stress ramp ranging from 0.1 to 1,000 Pa at 0.1

5

Hz frequency. All experiments were run in triplicate. The 95% confidence interval was

6

provided by the Student’s t-test. Figure 1 exhibits the geometry employed in the tests.

7

8 9

Figure 1 - Grooved concentric cylinders used in the rheological tests in order to avoid

10

apparent wall slip phenomenon.

11

2.2.8 Optical Microscopy. Wax crystal visualizations were made at inverted Axio Imager

12

2 microscope (Carl Zeiss) equipped with MRc5 Axiocam (5.0 megapixels) and

13

temperature controller (Linkan T95-PE). The polarized light technique was adopted and

14

20x objective lens (200x magnification) was employed. Small sample volumes were

15

placed on a glass slide fitted on the preheated thermal stage of the microscope set to

16

80°C. Then, samples were cooled to 4.0°C at a 1.0°C/min rate. All images were taken

17

15 minutes after reaching final temperature (4.0°C) and analyzed with Axio Vision 4.8

18

software. The aspect ratio (defined here as the ratio of major/minor axis) and crystal

19

length were averaged for 25 random measurements.


Page 8 of 35

Page 9 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1 2 3 4

3. Results and Discussion

5 6

3.1 Waxes Characterization

7 8

Table 1 summarizes structural and physical characterizations for all waxes

9

assessed by different analytical techniques. As already mentioned (section 2.1), the

10

capital letter L stands for (predominantly) linear whereas B stands for branched wax.

11

The subscripted numbers indicate the average carbon number calculated from GC-FID

12

data.

13 14

Table 1 – Waxes structural and physical features L24

L29

B37

B53

a

42 - 50

50 - 58

56 -75

41 - 68

a

135

187

152

91

a

133

203

167

116

Melting point range (°C) Fusion enthalpy (J/g) Crystallization enthalpy (J/g)

b

CH2/CH3 molar ratio (%)

24

25

44

57

b

CH/CH2 molar ratio (%)

0.0

0.0

6.36

4.41

b

Branches per linear chain

0.0

0.0

2.79

2.51

c

Average carbon number

24

29

37

53

c

Branched wax content (wt%)

10.4

15.2

35.3

32.5

50.1

31.7

51.0

35.8

d

15

Crystalinity degree (%)

Data from: DSC(a), 13C-NMR(b), GC-FID(c), XRD experiment(d)

16 17

The waxes thermal behavior was investigated by DSC (Figure 2) and the melting

18

range along with average enthalpies of fusion and crystallization were obtained (Table

19

1). For L24 and L29 waxes, the melting temperature ranged from 42-50°C and 50-58°C,

20

respectively. For B37 and B53 waxes, it ranged from 56-75°C and 40-62°C, respectively.

21

One can observe a relatively narrow temperature range for Lx (~ 8°C) comparing to Bx

22

waxes (~ 20°C). The wax structure is likely to be responsible for such differences.


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 35

1

Indeed, broader melting ranges indicate a higher degree of polydispersity, a

2

microcrystalline wax feature. The absence of a true melting region for the Bx waxes is

3

also related to the presence of isoalkanes and/or cycloalkanes33.

4

In addition to completely amorphous or completely crystalline materials, there

5

are of course materials that are partially crystalline. In this type of materials crystallites

6

and amorphous regions coexist. In this regard, the presence of smaller peaks

7

(highlighted in Figure 2) at Lx and B53 DSC thermograms are likely to be due to

8

amorphous

9

melting/crystallization12. Regarding B37 thermogram, the smaller peak associated to

10

amorphous precipitation is not seen, which is likely due to a relatively high branch

11

content and crystallinity degree (Table 1). The average enthalpies of fusion ranged

12

from 91 to 187 J/g whereas the average crystallization enthalpies were among 116 to

13

203 J/g. The very high wax polydispersity combined with isoalkanes and/or

14

cycloalkanes presence may lead to small values of enthalpy20 as occurred to B53 wax.

15

Spindle oil did not exhibit any thermal response for the same procedure adopted at

16

waxes DSC experiment.

precipitation,

while

the

more

intense

peaks

assigned

17

18 19

Figure 2 – Thermal behavior of the waxes assessed by DSC.

20


to

Page 11 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

13

C-NMR spectra for all waxes assessed is exhibited in Figure 3. From the

2

normalized areas, a dimensionless CH2/CH3 ratio can be calculated (all NMR spectra

3

are available as supporting information). This number relates the carbons at the

4

carbon main chain to the primary carbons and provides an estimative for the chain

5

length34. As can be seen from Table 1, the results from GC-FID regarding average

6

carbon number (24, 29, 37 and 53) and the respective ones from 13C-NMR (24, 25, 44

7

and 57) are comparable. The better agreement for Lx waxes is due to their simpler 13C-

8

NMR spectra. Another ratio, the CH/CH2, provides a relative estimative for tertiary

9

carbons presented in the sample (in molar basis). In the case of B37, 6.36% of carbons

10

are due to branching. Comparing to B53, which has 4.41% of tertiary carbons, the

11

former wax is smaller and presents more branches per molecule. From CH2/CH3 and

12

CH/CH3 molar ratio it is possible to estimate the number of branch points for each

13

linear main chain. This number is about 2.79 for B37 and it is smaller for B53 (2.51). In

14

the case of predominantly linear waxes, L24 and L29 did not present signals for tertiary

15

carbons (the analysis for L24 and L29 waxes were repeated twice, leading similar

16

results). This fact does not necessarily indicate the absence of branches. Indeed, as

17

shown in Table 1, L24 and L29 waxes have non-negligible branching contents, according

18

to GC-FID data. However, the characteristic signal associated to tertiary carbons for

19

these particular waxes (~ 32.5 ppm)34 could not be detected by the NMR equipment.

20


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1 2

Figure 3 - 13C-NMR spectra of Lx and Bx waxes.

3

Figure 4 exhibits the carbon number distribution for Lx and Bx waxes (all

4

chromatograms are available as supporting information). As one can clearly notice, the

5

polydispersity is remarkable for the B53 wax distribution, whereas other samples

6

presented a fairly narrow distribution. From CG-FID data it is possible to calculate the

7

dispersity ĐM (also known as polydispersity index), defined as the ratio of the weight-

8

average molar mass and the number-average molar mass. This ratio is a measure of

9

the width of molecular weight distributions35, generally used in polymer science.

10

Values of ĐM in the range of 1.0 - 1.1 indicates a narrow distribution. For L24, L29 and B37

11

waxes the values of dispersity are 1.012, 1.022 and 1.017, respectively. Also, Table 1

12

contains the information of branched content of each wax. L24 and L29, regarded as

13

predominately linear waxes throughout the text, has 10.4 and 15.2 wt% due to

14

branched carbons, respectively. B37 and B53, on the hand, has 35.3 and 32.5 wt%

15

assigned to branched carbons, being, therefore, considered non-linear branched

16

molecules. The combined results from 13C-NMR and GC-FID suggests that B37 has more

17

tertiary carbons per linear chain and longer branches, based on the higher mass of

18

branched carbons encountered (35.3 wt%, compared to 32.5 wt% of B53).


Page 12 of 35

Page 13 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

It is worth mentioning that the equipment used for the analysis has a detection

2

limit at C78 although B53 wax has 23 wt% assigned to larger carbon chains. The average

3

carbon number of 53 was calculated assuming that all remaining carbon chains have

4

79 atoms, which is a conservative estimation, especially when compared to

5

results of 57 (Table 1) for B53.

13

C-NMR

6

7 8 9

Figure 4 - Carbon number distribution for Lx and Bx waxes from GC-FID data. From left to right: L24, L29, B37, and B53.

10 11

The XRD analyses are exhibited in Figure 5. The diffraction pattern of all waxes

12

are similar and the most intense peaks are located at 2Θ ~ 21.50° and 2Θ ~ 23.50°,

13

which represent materials crystallized with orthorhombic shapes24. Assuming a first-

14

order X-ray reflection (n = 1, from Bragg’s Law) the distance between atomic layers

15

according to the more intense and narrow peaks are estimated to be 0.403 nm and

16

0.378 nm. Kané et al. (2003)29 observed a thin lamellar structure, with a lamella

17

thickness between 1.5 and 3.0 nm, for all samples assessed by the authors (crude oils

18

and model oils), when crystallized by lowering the temperature.

19

All the waxes showed clear peaks with similar intensities at the low angle region

20

(2Θ ~ 12°), providing an evidence for lamellar packing21. According to Zaky and


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

Mohamed (2010)28, the peak at 2Θ = 36.07°, more intense at L24 wax, is characteristic

2

of macrocrystalline waxes. The crystallinity degree (calculated by dividing the total

3

area under Bragg peaks on the total area under Bragg peaks plus the area under hump)

4

for Lx and Bx waxes are shown in Table 1. A low degree of crystallinity indicates carbon

5

chains highly non-linear, presenting obstacles for the regular folding of chains. Also,

6

wide molecular weight distribution may contribute to this reduced crystallinity28. In

7

this sense, the relatively small enthalpy values for B53 wax (Table 1) is corroborated by

8

the low XRD crystallinity value. Although, the XRD results for L29 and B37 could not be

9

explained on this basis since both waxes have similar molecular weight distribution

10

and 13C-NMR results obtained for L29 indicated neglected branch presence, unlike B37.

11

Due to this, the

12

similar results.

13

C-NMR and XRD experiments were repeated, although yielding

13

14 15

Figure 5 - X-ray diffraction patterns for Lx and Bx waxes. From top to bottom: L24, B37,

16

B53, and L29.

17 18

The structure that develops in the gelled oils is dependent on the wax content

19

and also on the presence of other material that may reduce the crystallinity of the

20

structure. To certify for the presence of possible impurities FTIR experiments were


Page 14 of 35

Page 15 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

performed and results are exhibited in Figure 6. Similar spectra were observed for all

2

samples. The most intense peaks can be found at 3,000 to 2,840 cm-1 region which

3

indicates stretching vibrations for C-H bonds of sp3 carbons, i.e. the alkane or

4

cycloalkane chemical signature. Peaks between 1,473 cm-1 to 1,461 cm-1 and 1,378 cm-

5

1

6

bonds and symmetrical bending vibrations for C-H (sp3 carbons), respectively. The peak

7

at 720 cm-1 is associated with straight chain methylene at least four units long36,37.

8

Thus, none impurity seems to be present in waxes samples.

to 1,377 cm-1 are due to scissoring molecular vibrations at methylene groups of C-H

9

10 11

Figure 6 - FTIR spectra data of Lx and Bx waxes.

12 13 14

3.2 Measurements of WAT, Yield Stress, G’LVR and γCR

15 16

For temperatures above a threshold limit, known as Wax Appearance

17

Temperature (WAT), the crude oils behave as simple fluids without elasticity, due to

18

none (or almost none) waxy crystals presence. This scenario corresponds to the step

19

[1] of the rheological tests performed (see details in experimental section). However,

20

the viscosity of the system begins to increase as they undergo a cooling process with


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

the temperature reaching the WAT and further lower values (step [2] of the

2

rheological tests). Different techniques have been developed to measure the WAT

3

based on DSC, near-Infrared spectroscopy, nuclear magnetic resonance spectroscopy,

4

and optical microscopy4,32,38. In this work, this measurement was based on rheological

5

data. Despite existing comparative studies in the literature suggesting that rheometry

6

measurements underestimate the Wax Appearance Temperature4,39, this fact is not a

7

particular concern in this study, since we are comparing WAT values based on the

8

same metric. The WAT was defined as the onset temperature where there is a sudden

9

viscosity change evident on a linear scale plot4. This behavior is exhibited in Figure 7

10

for three different model oils. As a consequence of the continuous precipitation

11

(especially at quiescent conditions), the crystal-crystal interactions lead to a waxy-gel

12

network formation (step [3] of the rheological tests). As the cooling rate directly

13

affects the gelation process and consequently the yield stress value5, it was kept at 1.0

14

°C/min at all rheological tests performed.

15 16 17

Figure 7 - Viscosity increasing during a cooling process (80°C - 4.0°C at 1.0°C/min) for L24, B53 and L24 + B53 model oils.

18 19

The gelled oils have a viscoelastic structure that responds to the stress exhibiting


Page 16 of 35

Page 17 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

a yielding region. At oscillatory amplitude sweep tests, this region comprises the onset

2

departure of the elastic modulus (G’) from the linear viscoelastic region and the

3

crossing point of G’ and G’’ (the viscous modulus). The yielding of gelled oils can be

4

suitable captured by step [4] of the rheological tests performed and is exhibited in

5

Figure 8. In this work, the yield stress is taken as the oscillatory stress for the crossing

6

point of G’ and G’’, which provides a reasonable and fast method to probe a wide

7

experimental range40. Although, it is important to clarify that the oscillatory stress at

8

the crossing point of G′ and G′′ may not be regarded as the material yield stress, since

9

this value may vary with e.g. the applied test frequency or the stress loading rate.

10

Indeed, it is reasonable to argue that the yield stress is, in fact, an idealization since,

11

given sufficient accurate measurements (or sufficient patience), no yield stress would

12

exist. In other words, if a material flows at high stresses it will also flow, however

13

barely and slowly, at low stresses. In addition, according to Barnes (1999)41, all

14

materials can flow on long enough timescales and consequently many materials, which

15

are considered to have a true yield stress, are actually very high viscosity liquids. On

16

the other hand, the question of whether yield stress exists as a real phenomenon has

17

little relevance in engineering practice. For process timescales such as production,

18

transportation, and storage of waxy crude oil the yield stress is a practical reality and it

19

can be essentially defined as the point at which, when increasing the applied stress,

20

the solid first shows liquid-like behavior, i.e. continual deformation42. Thus, a practical

21

definition accepted and supported in this investigation.

22

Another useful property obtained from the step [4] is the elastic modulus at

23

linear viscoelastic region (G’LVR). This region can be defined by a plateau of G’ values at

24

a logarithmic plot and it extends until an obvious decreasing in G’ values43. G’LVR is

25

calculated as an average between the values of G’ at this linear viscoelastic region.

26

Finally, the critical strain (γCR) is defined in this paper as the maximum value of shear

27

strain captured before a complete material rupture, which is represented by an abrupt

28

increase of the strain located among the yielding region and G’ and G’’ crossing point

29

(Figure 8). When the oscillatory shear stress is in the low range the relationship

30

between shear stress and its corresponding strain is in the linear elastic region. When

31

the loaded stress increases above a certain value, G′ decreases but the strain increases

32

exponentially, meaning that the gelled structure is partly destroyed. The γCR is,


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

therefore, defined as the maximum shear strain in this yielding region. It is important

2

to note that our definition is slightly different from the critical elastic strain, which is

3

the value of maximum shear strain in linear elastic region43,44, thus a smaller value

4

than the critical strain.

5

6 7 8

Figure 8 - A profile of the logarithmic stress sweep (step [4] of the rheological protocol adopted) used to obtain τ0, G’LVR and γCR values.

9 10

3.3 The Influence of Wax Chemical Structure at Yield Stress

11 12

Model oils composed by L24, L29, B37 and B53 waxes solubilized in spindle oil were

13

employed to evaluate the influence of different chemical structures and provide

14

insight into crystal-crystal interactions. The composition was set to 7.5 wt% for each

15

system. Blended systems employed 3.75 wt% of each wax. Figure 9 and Table 2

16

summarizes the obtained results from rheological tests.

17


Page 18 of 35

Page 19 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

Figure 9 - Yield stress of model oils (singles waxes and blends).

2 3 4 5

Table 2 - Rheological properties and cohesive energy density (Ec) of waxy gels assessed. WAT (°C) 10.6

G’LVR/103 (Pa) 82.426

Critical strain (%)

L24

Yield stress (Pa) 112.5 ± 10.7

0.3795

Ec (J/m3) 5.934

L29

199.5 ± 31.2

27.3

319.89

0.2154

7.420

B37

8.43 ± 0.92

34.3

3.304

0.3236

0.173

B53

26.7 ± 5.5

49.6

14.407

0.5311

2.032

L24 + L29

223.9 ± 27.9

21.5

305.28

0.0910

1.265

B37 + B53

2.99 ± 0.54

35.1

1.548

0.1017

0.008

L24 + B37

335.1 ± 30.5

20.1

179.07

0.1876

3.152

L24 + B53

630.2± 40.4

28.2

214.64

0.1740

3.250

L29 + B37

134.4 ± 16.9

19.7

116.01

0.1290

0.965

L29 + B53

266.7 ± 16.3

29.2

164.40

0.1692

2.354

Model oil

6 7


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

3.3.1 Single wax model oils

2 3

For systems prepared with single wax, the yield stress was favored by an increase

4

in chain length for predominantly linear and branched wax compared separately

5

(Figure 9). This is possible due to the earlier precipitation of longer wax molecules

6

which favors the molecule packing and guarantees more available nuclei for

7

crystallization. In this reasoning, the longer is the saturated linear chain the higher is

8

the WAT, which in fact is observed by comparing the WAT values for L24, L29, B37 and

9

B53 (10.6°C, 27.3°C, 34.3°C and 49.6°C, respectively). Although, non-linear alkanes

10

might reduce the gel strength3,23. As the interlocking network resulting from gelation

11

process relies on crystal-crystal interactions, it is reasonable to argue that branching

12

points reduce gel strength, since crystal interactions may be hampered. Therefore, the

13

highest WAT value for the branched B53 wax does not guarantee the highest yield

14

stress, which suggests a competitive effect among carbon chain length and tertiary

15

carbons presence. Thus, it is important for further evaluations do not dissociate the

16

impact of these two structural features at rheological properties. In this basis, the very

17

low yield stress for the B37 system might be due to the number of branches (2.71 per

18

linear saturated aliphatic chain) in relatively narrows carbon chains when compared to

19

B53.

20

Despite having the same unit, yield stress and elastic modulus does not

21

represent the same quantity. The elastic modulus is related to the resistance of some

22

material to being deformed elastically whereas yield stress can be regarded as the

23

highest stress at which no flow is detectable. In the flow assurance context yield stress

24

is a key parameter. The elastic modulus, however, has not the same relevance.

25

Although, in the particular case of waxy gels, both properties arise from the same

26

phenomenon i.e., the oil gelation2. Thus it is expected a correlation among them. This

27

correlation is exposed in Figure 10, which exhibits an almost linear proportionality

28

between the yield stress and G’LVR for single and blended model oils since R2 is 0.94 and

29

the fitting equation exponent (1.0801) is very close to the unity. Considering that the

30

elastic modulus is also affected by the material internal structure, the same reasoning

31

applied to yield stress is applicable to G’LVR.


Page 20 of 35

Page 21 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1 2

Figure 10 - G’LVR and yield stress data for single wax systems and blended waxes.

3 4

An incipient idea of the relationship between elastic modulus and yield stress is

5

presented by Lopes-da-Silva and Coutinho (2007)11 in the investigation of the gel

6

structure development under quiescent conditions for three different waxy crude oils.

7

Mendes et al. (2017)8 presented quantitative correlations between G’ and yield stress

8

for a Brazilian waxy crude oil, using small amplitude oscillatory tests to obtain the

9

rheological data. The authors observed analogous behavior for G’ and gel yield stress

10

development at static cooling conditions. For mixed cooling conditions (i.e., dynamic

11

and static conditions employed during cooling) the linear relation between those two

12

properties was not so clear. Although, it could be devised a global tendency, showing

13

that the higher the storage modulus, higher the yield stress. The importance of such

14

results is well stated by the authors: “If a relation between G′ and yield stress can be

15

drawn, it can be useful as a method for reducing the number of experiments to detect

16

yield stress tendencies with respect to some cooling parameter.” Fernandes et al.

17

(2016)45 also presented a similar reasoning for samples of oil-based drilling fluid that

18

exhibits a complex rheological behavior, i.e., elasticity, viscoplasticity and thixotropy.

19

The authors proposed a correlation between the stress overshoot observed in flow

20

start-up rheometric experiments and the storage modulus, obtained during a low

21

amplitude oscillatory sweep over the aging time. The results showed an excellent

22

agreement for linear relationship of these two properties, with error bands of 10% in a


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 35

1

range of shear rates varying from 0.5 to 50 s-1. At the best authors' knowledge, there

2

are at least three complex materials (drilling fluids, waxy crude oil and model waxy oil)

3

that presented this correlation between elastic modulus and yield stress.

4

Non-linear alkanes do not pack as closely together as linear waxes. Their

5

cohesive energy is thus weaker than those observed for their linear counterparts.

6

Cohesive energy is related to the interaction energy among wax crystals molecules and

7

generally, the stronger it is, the higher is the intensity of association formed by

8

hydrophobic groups46,47. Table 2 contains the values of cohesive energy densities, Ec,

9

which is essentially the driving force of the ordering transformation from soft to hard

10

wax in the gelation process. These values were calculated from G’LVR and γcr (obtained

11

from step [4] of the rheological tests performed) based on the following equation by

12

Tadros (2010)44:

13 ଶ ‫ܧ‬௖ = 1ൗ2 ∗ ߛ௖௥ ∗ ‫ܩ‬′௅௏ோ

Eq. (1)

14 15

It is observed that the values of Ec follow the same trend of yield stress. The

16

branched wax B37 presented the lowest Ec value (0.173 J/m3, Table 2). This indicates

17

that these wax molecules would be less available to aggregate together, thus

18

significantly lowering their cohesive energy as well as their chances to nucleate wax

19

crystallization. In fact, the structural features of Lx and Bx wax affect the yield stress,

20

G’LVR, and the cohesive energy similarly, which can be observed from Table 2.

21 22

3.3.2 Blended wax model oils

23 24

Rheological properties and cohesive energy density are also available for blended

25

wax systems (Table 2). As can be noticed, the WAT values for systems containing L24

26

are approximately the average between the WAT for each wax that composes the

27

respective blend. This is clearly seen in Figure 7: the WAT for L24, B53 and L24+B53 model

28

oils are 10.6 °C, 49.6 °C, and 28.2 °C, respectively. Since the WAT is sensitive to the

29

carbon number distribution29, it is clear that occurred a shift in this distribution for

30

blends containing L24. It suggests that, in these cases, the waxes were completely

31

associated during the precipitation. Although, for the association of B37+B53, L29+B37 ACS Paragon Plus Environment

Page 23 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

and L29+B53 it is observed that the WAT corresponds to approximately the same WAT

2

value of the smaller wax in the blend (e.g., the WAT for L29, B53, and L29+B53 model oils

3

are 27.3 °C, 49.6 °C, and 29.2 °C, respectively). It indicates some lack of interaction

4

between the different wax in these blends.

5

The yield stress for the L24+L29 blend is statistically equivalent to the single L29

6

model oil. As L24 and L29 waxes comprise mainly linear saturated aliphatic chains and

7

few obstacles are present for favored interactions (such as branching points) it is

8

comprehensible that the yield stress of this blended system matches the yield stress of

9

the longer wax. Although for B37 + B53 the yield stress is negligible. The branched

10

nature of these two waxes associated with an unfavorable combination of main chain

11

lengths is likely to be responsible for this result. It is expected that branched wax

12

molecules with ~ 37 carbon atoms experience a huge hindrance when trying to

13

approximate to wax molecules with ~ 53 carbon atoms, thus a network with many

14

voids is formed resulting in the weak gelling. Similar results were presented by Senra

15

et al. (2009)25. The authors investigated model oils prepared exclusively from linear

16

wax in dodecane and considered the effect of cocrystallization on the gelation

17

characteristics of the samples. According to the them, when two or more materials

18

crystallize together, the crystal formation will not be as perfect as a single material

19

forming a crystal structure because of their differing sizes. The different chain lengths

20

provide defects and weaknesses in the crystal structure. Although, as revealed in the

21

investigation, some compositions are prone to cocrystallize whereas others

22

associations seem to be not affected by the added alkane.

23

For L24+Bx, both blends had an increase in yield stress, comparing to the single L24

24

model oil. Combined with B53 the yield stress for this blend increased 462 % comparing

25

to the single L24 system. In the case of L24+B37, the yield stress is 199 % higher. It is

26

clear that the rheological properties of a waxy gel are influenced by the nature of

27

interactions between the microstructural components occurring at the junction zones,

28

which exist between intersecting crystal branches or entangled crystal lamellas. In this

29

regard, the reasoning is as follows: since B53 has very large carbon main chains, the

30

branches might be located apart from each other. In this case, small linear waxes are

31

able to arrange itself between the branch points, thus the crystal interactions would

32

just be slightly affected by these branches. Considering that B37 is smaller and more


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

branched than B53, this arrangement is less likely, except for very small chain waxes

2

molecules.

3

To support the above argument, one can consider the worst possibility (in terms

4

of wax-wax interactions) for the positions of tertiary carbons in a branched molecule:

5

the branches are equally spaced. In this case, the ratio of main chain length/branches

6

per linear chain provides the space available between branching points. For B37 this

7

ratio is 15.5 and for B53, 21.9 (calculated from 13C-NMR data, Table 1). Thus, it is more

8

likely that small molecules (such as L24 wax) would be able to arrange itself among B53

9

branches, thus enhancing London-Van der Waals interactions (more than at B37). It can

10

be seen by L29+B37 blend: as both waxes have carbon chains with similar size, the

11

hindrance is much more pronounced, reflecting a decrease in yield stress of 33 %

12

compared to L29 single systems.

13

In terms of critical strain (defined as the maximum value of shear strain before a

14

complete material rupture), Table 2 provides values for all assessed systems.

15

Comparing the critical strain of single and blended systems, one can conclude that

16

waxy gels formed from single wax are more deformable. Interestingly the branched

17

wax B37 and B53 have relatively high critical strain values, although low values for

18

elastic modulus. On the other hand, L24+L29 and B37+B53 blended systems have a

19

very similar critical strain (0.0910 and 0.1017, respectively) and completely different

20

elastic modulus (305.28 kPa and 1.548 kPa, respectively). Also, for the single wax

21

samples, there is no clear correlation between the critical strain values and any other

22

property featuring in Table 2. On the other hand, for blended systems, the higher

23

critical strain can account for the higher cohesive energy densities, as these two

24

properties are fairly linearly correlated (R2 = 0.79).

25

Regarding cohesive energy density, Figure 11 illustrates how this property is

26

related to yield stress for all systems assessed. Single wax and blended systems were

27

divided into two dataset since in both cases the Ec seems to be linearly correlated to

28

the yield stress, although with a different angular coefficient. The R2 value is higher for

29

single wax (0.90) when compared to blended systems (R2 = 0.75), however, the number

30

of points for single wax is smaller, thus it is not easy to state which curve is fitting best.

31

The behavior of Ec for single wax systems is somehow easy to follow: the bigger the

32

linear main chain and the fewer the branches, the higher is the cohesive energy


Page 24 of 35

Page 25 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

density and yield stress. On the other hand, blended systems were able to achieve the

2

highest yield stress values. It indicates that in terms of cohesive energy the interactions

3

among waxes of the same nature are more favorable.

4

5 6 7

Figure 11 – Cohesive energy density as a function of yield stress for single waxes and blended waxes systems.

8 9

For B37+B53 blend Ec value is very low (0.008 J/m3), representing a 99.6 %

10

decrease compared to B53 single wax. According to Jang et al. (2007)48 the cohesive

11

energy density of the crystalline wax is much smaller in the presence of molecules that

12

are able to act as an inhibitor of wax crystallization (comb-like polymers in the case

13

studied) and their presence perturbs or retards the ordering transformation of the

14

amorphous wax aggregate into an ordered phase48. In this regard, the branched B37

15

may act as a wax inhibitor for sufficient large wax molecules (e.g., L29 or B53).

16 17

3.3.3 Microscopy observations

18 19

Optical microscopy was employed to investigate morphological aspects (size and

20

shape) of waxy crystals precipitated from model oils. Although, there is an important

21

limitation that must be considered in this analysis: unlike thermal history, the shear


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 35

1

history imposed to the samples in the rheological tests (i.e., shear rate of 0.8 s-1 during

2

cooling process) cannot be reproduced by the microscope, as the equipment is not

3

able to shear samples during cooling. In this sense, the material at the end of the static

4

cooling and of the dynamic cooling are presumably not the same. Despite this

5

limitation, some interesting aspects of wax crystal morphologies can be devised. The

6

results are exhibited in Table 3 and Figure 12

7 8

9

Table 3 - Morphological features for wax crystals of Lx and Bx model oils Model oil

Average length (µm)

Aspect ratio

L24

72.3

9.09

L29

24.0

12.5

B37

12.5

7.33

B53

---*

---*

L24 + L29

25.8

9.09

B37 + B53

9.04

4.54

L24 + B37

18.9

5.55

L24 + B53

17.2

5.88

L29 + B37

14.8

7.69

L29 + B53

---*

---*

*at these images it was not possible to precisely devise isolated crystals

10

L24

L29

L24 + L29

B37

B53

B37 + B53


Page 27 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

L24 + B37

L24 + B53

L29 + B37

L29 + B53

1

Figure 12 - Microscopy images of precipitated wax crystals from fresh prepared model oils

2

cooled from 80°C to 4.0°C at 1.0°C/min (pictures were taken after 15 minutes at final

3

temperature).

4 5

The morphology and structure of the wax crystals play an important role in the

6

flow properties of waxy crude oils31. The network of interacting paraffin crystals shows

7

complex morphology and according to Singh et al. (1999)5, it is due to the flocculation

8

of orthorhombic wax crystallites, creating a highly porous and rigid structure full of

9

entrapped oil.

10

As can be seen from Figure 12 individual anisotropic particles having the shape of

11

needles and clustering pattern varying widely from system to system. Clusters of self-

12

assembled crystals might form an aggregate in different ways leading to a stronger or

13

weaker network. For single model wax, the highest yield stress (L29) is achieved in the

14

image with fewer unoccupied space, regardless the highest crystal size and aspect ratio

15

achieved by L24 wax. Although, this same trend is not completely valid for blended


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

systems. According to Zaky and Mohamed28, macrocrystalline waxes have a

2

characteristic XRD peak at 2Θ = 36.07° and appear at large and loose needle form.

3

Meanwhile, the microcrystalline waxes crystallize in needle form although smaller in

4

size. Paso et al. (2005)49 obtained crystal lengths in the range of 10-20 µm for model

5

fluids consisting of n-paraffin waxes dissolved in mineral oil which is in agreement with

6

the lengths obtained in this study.

7

Among blended systems, Lx + Bx presented needle-like crystals amassed into clusters

8

of varying size. L24+ L29 model oil presented the biggest crystals (25.8 µm) whilst B37+B53 the

9

smallest (9.04 µm). Aspect ratio was also the smallest for B37+B53, indicating a more

10

rounded shape. Indeed, is clear from Figure 12 the presence of very tiny crystals for this

11

system.

12

Despite the efforts to devise useful patterns from microscopic images, quantitative

13

characterization of their morphology and structure is difficult, due the high complexity and

14

irregularity of wax crystal microstructures.

15

Conclusions

16 17 18

Model oils consisting of predominantly linear and non-linear branched waxes

19

dissolved in non-crystallizable mineral oil were investigated. Rheological measurements

20

showed that these systems reproduce essential features of crude oil gels (e.g. exhibiting a

21

low-temperature gel-like mechanical response to an imposed low-frequency oscillatory

22

stress). Therefore, they were employed to verify and further probe the processes involved

23

in the formation of wax crystal aggregates. The branched waxes added to the model oils

24

are essential to obtaining a system that closely mimics the gelation behavior of real crude

25

oils.

26

By means of optical microscopy, it was possible to observe micrometer-sized wax

27

crystals formed under temperatures below WAT. Wax physicochemical characterization

28

was performed to aid in the structure-rheological properties investigation. Features

29

compatible with linear macrocrystalline wax were obtained for L24 and L29, whereas non-

30

linear (branched) microcrystalline were assigned to B37 and B53 waxes. The CH2/CH3 ratio,

31

which allows estimating the number of branching points for each linear main chain, was

32

6.36% for B37 wax and 4.41% for B53 (in molar basis). Thus, it is clear that the former wax


Page 28 of 35

Page 29 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

has shorter molecules and exhibits more branching points.

2

The presence of branch points at the linear saturated aliphatic chains is likely to be

3

responsible for a reduced crystal-crystal interaction, consequently lowering the yield stress.

4

On the other hand, an increase in the chain length showed an opposite effect. Therefore, it

5

is important for further evaluations do not dissociate the impact of these two structural

6

features at rheological properties.

7

The rheological results indicated that B37 wax may act as a wax inhibitor for sufficient

8

large wax molecules (e.g., L29 or B53), but increases the yield stress of gels of a shorter wax

9

molecule (L24). Its presence may perturb or retard the ordering transformation of the

10

amorphous wax aggregate into an ordered phase for large molecules such as L29 or B53,

11

which are not able to arrange itself linearly among the branch points of B37. The large

12

branched wax, B53 increased the yield stress of the two predominantly linear waxes, but its

13

interaction with another branched wax B37 did not show any appreciable yield stress

14

enhancement.

15

Optical microscopy showed anisotropic particles with needle shape and varying

16

clustering pattern for different model oils composition. The general characteristics

17

resemble those of crude oils: the solid phase appears to be composed of ill-formed, needle-

18

like crystals amassed into clusters of ~ 10 – 20 µm size, which further aggregate to form a

19

space-filling network. The crystals formed by the predominantly linear waxes in the

20

absence of the iso-components are much larger, suggesting that the branched waxes

21

interfered with the regular growth of the wax crystals (this was also evident in DSC

22

experiments, in which the solution of predominantly linear waxes displayed a much

23

sharper crystallization transition). Although, the image analysis is hampered by the high

24

complexity and irregularity of wax crystal making it impossible to directly correlate

25

morphological aspects to rheological properties.

26

Acknowledgments

27

The authors wish to thank Petrobras, ANP (Agência Nacional do Petróleo, Gás Natural

28

e Biocombustíveis) and CNPq (Conselho Nacional de Pesquisa e Desenvolvimento) for

29

supporting this work. Also we would like to thank Bruno Charles Couto from Petrobras

30

for providing the GC-FID analysis.


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

References

1

2

1.

Venkatesan, R., Nagarajan, N. R., Paso, K., Yi, B., Sastry, A. M., Fogler, H. S. The

3

strength of paraffin gels formed under static and flow conditions. Chem. Eng.

4

Sci. 60, 3587–3598 (2005).

5

2.

crude oil emulsion gels. J. Nonnewton. Fluid Mech. 149, 34–39 (2008).

6

7

3.

Aiyejina, A., Chakrabarti, D. P., Pilgrim, A. & Sastry, M. K. S. Wax formation in oil pipelines: A critical review. Int. J. Multiph. Flow 37, 671–694 (2011).

8

9

Visintin, R. F. G., Lockhart, T. P., Lapasin, R. & D’Antona, P. Structure of waxy

4.

Japper-Jaafar, A., Bhaskoro, P. T. & Mior, Z. S. A new perspective on the

10

measurements of wax appearance temperature: Comparison between DSC,

11

thermomicroscopy and rheometry and the cooling rate effects. J. Pet. Sci. Eng.

12

147, 672–681 (2016).

13

5.

Singh, P., Fogler, H. S. & Nagarajan, N. Prediction of the wax content of the

14

incipient wax-oil gel in a pipeline: An application of the controlled-stress

15

rheometer. J. Rheol. (N. Y. N. Y). 43, 1437 (1999).

16

6.

emulsions. Colloid Polym. Sci. 277, 583–588 (1999).

17

18

7.

Huang, Z., Lee, H. S., Senra, M. & Fogler, H. S. Droplet microfluidics on a planar surface. Am. Inst. Chem. Eng. 57, 2955–2964 (2011).

19

20

Pal, R. Yield stress and viscoelastic properties of high internal phase ratio

8.

Mendes, R., Vinay, G. & Coussot, P. Yield stress and minimum pressure for

21

simulating the flow restart of a waxy crude oil pipeline. Energy and Fuels 31,

22

395–407 (2017).

23 24

9.

Van Der Geest, C., Guersoni, V. C. B., Merino-Garcia, D. & Bannwart, A. C. A modified elasto-viscoplastic thixotropic model for two commercial gelled waxy ACS Paragon Plus Environment

Page 30 of 35

Page 31 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

crude oils. Rheol. Acta 54, 545–561 (2015).

1

2

10.

oils. J. Pet. Sci. Eng. 7, 177–213 (1992).

3

4

Petter Rønningsen, H. Rheological behaviour of gelled, waxy North Sea crude

11.

Lopes-da-Silva, J. A. & Coutinho, J. A. P. Analysis of the isothermal structure

5

development in waxy crude oils under quiescent conditions. Energy and Fuels

6

21, 3612–3617 (2007).

7

12.

Dirand, M. et al. Normal alkanes, multialkane synthetic model mixtures, and real

8

petroleum waxes: Crystallographic structures, thermodynamic properties, and

9

crystallization. J. Chem. Eng. Data 47, 115–143 (2002).

10

13.

Lin, M., Li, C., Yang, F. & Ma, Y. Isothermal structure development of Qinghai

11

waxy crude oil after static and dynamic cooling. J. Pet. Sci. Eng. 77, 351–358

12

(2011).

13

14.

of slurry flows in a waxy oil pipeline. Ind. Eng. Chem. Res. 54, 4455–4459 (2015).

14

15

15.

Xu, J. et al. Effect of polar/nonpolar groups in comb-type copolymers on cold flowability and paraffin crystallization of waxy oils. Fuel 103, 600–605 (2013).

16

17

Peerapornlerd, S. et al. Effect of the flow shutdown temperature on the gelation

16.

Paso, K., Silset, A., Sørland, G., Gonc, M. D. a L. & Sjo, J. Characterization of the

18

Formation , Flowability , and Resolution of Brazilian Crude Oil Emulsions

19

Characterization of the Formation , Flowability , and Resolution of Brazilian

20

Crude Oil Emulsions. Energy & Fuels 23, 471–480 (2009).

21

17.

of Suspended Crystals. Energy and Fuels 30, 1–11 (2016).

22

23 24

Cabanillas, J. P., Leiroz, A. T. & Azevedo, L. F. A. Wax Deposition in the Presence

18.

Marchesini, F. H., Alicke, A. A., De Souza Mendes, P. R. & Ziglio, C. M. Rheological characterization of waxy crude oils: Sample preparation. Energy and ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Fuels 26, 2566–2577 (2012).

1

2

19.

Highly Paraffinic ‘ Waxy ’ Mixtures. Can. J. Chem. Eng. 82, 162–174 (2004).

3

4

20.

21.

22.

Bai, C. & Zhang, J. Effect of carbon number distribution of wax on the yield stress of waxy oil gels. Ind. Eng. Chem. Res. 52, 2732–2739 (2013).

9

10

Doan, C. D. et al. Chemical profiling of the major components in natural waxes to elucidate their role in liquid oil structuring. Food Chem. 214, 717–725 (2017).

7

8

Musser, B. J. & Kilpatrick, P. K. Molecular characterization of wax isolated from a variety of crude oils. Fuel Energy Abstr. 40, 715–726 (1998).

5

6

Tiwary, D. & Mehrotra, A. K. Phase Transformation and Rheological Behaviour of

23.

Zhao, Y., Kumar, L., Paso, K., Safieva, J. & Sariman, M. Z. B. Gelation Behavior of

11

Model Wax−Oil and Crude Oil Systems and Yield Stress Model Development.

12

Energy and Fuels 26, 6323–6331 (2012).

13

24.

Han, S., Zhang, J. & Zhu, L. Effect of the amount and the composition of

14

precipitated n-alkanes on the yield stress of wax-decane gels. J. Pet. Sci. Eng.

15

147, 228–236 (2016).

16

25.

Senra, M., Scholand, T., Maxey, C. & Fogler, H. S. Role of polydispersity and

17

cocrystallization on the gelation of long-chained n-alkanes in solution. Energy

18

and Fuels 23, 5947–5957 (2009).

19

26.

Senra, M., Grewal, M. G. & Jarbore, J. H. The Role of n‑Alkane Solvent Carbon

20

Number on the Gelation of Long-Chainedn‑Alkanes in Solution. Ind. Eng. Chem.

21

Res. 54, 4505–4511 (2015).

22 23

27.

Cazaux, G., Barre, L. & Brucy, F. Waxy Crude Cold Start: Assessment Through Gel Structural Properties. SPE Annu. Tech. Conf. Exhib. 729–739 (1998).


Page 32 of 35

Page 33 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

28.

Zaky, M. T. & Mohamed, N. H. Comparative study on separation and

2

characterization of high melting point macro- and micro-crystalline waxes. J.

3

Taiwan Inst. Chem. Eng. 41, 360–366 (2010).

4

29.

Kané, M., Djabourov, M., Volle, J.-L., Lechaire, J.-P. & Frebourg, G. Morphology

5

of paraffin crystals in waxy crude oils cooled in quiescent conditions and under

6

flow. Fuel 82, 127–135 (2003).

7

30.

Gao, P., Zhang, J. & Ma, G. Direct image-based fractal characterization of

8

morphologies and structures of wax crystals in waxy crude oils. J. Phys. Condens.

9

Matter 18, 11487–11506 (2006).

10

31.

Jiang, C., Zhao, K., Fu, C. & Xiao, L. Characterization of Morphology and Structure

11

of Wax Crystals in Waxy Crude Oils by Terahertz Time-Domain Spectroscopy.

12

Energy and Fuels 31, 1416–1421 (2017).

13

32.

Hansen, A. B., Larsen, E., Pedersen, W. B., Nielsen, A. B. & R∅nningsen, H. P.

14

Wax Precipitation from North Sea Crude Oils. 3. Precipitation and Dissolution of

15

Wax Studied by Differential Scanning Calorimetry. Energy and Fuels 5, 914–923

16

(1991).

17

33.

Luyt, A. S. & Krupa, I. Thermal behaviour of low and high molecular weight

18

paraffin waxes used for designing phase change materials. Thermochim. Acta

19

467, 117–120 (2007).

20

34.

CRC Press (2000).

21

22

35.

Rogošić, M., Mencer, H. J. & Gomzi, Z. Polydispersity index and molecular weight distributions of polymers. Eur. Polym. J. 32, 1337–1344 (1996).

23

24

Brandolini, Anita J., D. D. H. NMR Spectra of Polymers and Polymer Additives.

36.

Silverstein, R. M., Webster, F. X. & Kiemle, D. J. Spectrometric Identifications of


Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

organic compounds. Wiley (2005).

1

2

37.

organic compounds. Quim. Nova 27, 670–673 (2004).

3

4

Lopes, W. A. & Fascio, M. Flow chart for infrared spectra interpretation of

38.

Zhao, Y. et al. Utilization of DSC, NIR, and NMR for wax appearance temperature

5

and chemical additive performance characterization. J. Therm. Anal. Calorim.

6

120, 1427–1433 (2015).

7

39.

Coutinho, J. A. P. & Daridon, J. L. The limitations of the cloud point

8

measurement techniques and the influence of the oil composition on its

9

detection. Pet. Sci. Technol. 23, 1113–1128 (2005).

10

40.

Tinsley, J. F., Jahnke, J. P., Dettman, H. D. & Prud’home, R. K. Waxy gels with

11

asphaltenes 1: Characterization of precipitation, gelation, Yield stress, and

12

morphology. Energy and Fuels 23, 2056–2064 (2009).

13

41.

Non-Newtonian Fluid Mech 81, 133–178 (1999).

14

15

42.

43.

44.

Tadros, T. F. Rheology of Dispersions: Principles and Applications. Rheology of Dispersions: Principles and Applications. Wiley VCH (2010).

20

21

Yang, F., Li, C. & Wang, D. Studies on the structural characteristics of gelled waxy crude oils based on scaling model. Energy and Fuels 27, 1307–1313 (2013).

18

19

Chhabra, R. P. & Richardson, J. F. Non-Newtonian Flow and Applied Rheology. Butterworth-Heinemann, (2008).

16

17

Barnes, H. a. The yield stress — a review or ‘ panta roi ’— everything flows ? J.

45.

Fernandes, R. R., Andrade, D. E. V., Franco, A. T. & Negrão, C. O. R. Correlation

22

between the gel-liquid transition stress and the storage modulus of an oil-based

23

drilling fluid. J. Nonnewton. Fluid Mech. 231, 6–10 (2016).


Page 34 of 35

Page 35 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

46.

Claudy, P., Letoffe, J.-M., Bonardi, B., Vassilakis, D. & Damin, B. Interactions

2

between filter plugging point in a diesel fuel . A thermodynamic and cloud

3

depressants study. Fuel 72, 821–827 (1993).

4

47.

Ji, Y., Kang, W., Liu, S., Yang, R. & Fan, H. The relationships between rheological

5

rules and cohesive energy of amphiphilic polymers with different hydrophobic

6

groups. J. Polym. Res. 22, 1–7 (2015).

7

48.

Jang, Y. H., Blanco, M., Creek, J., Tang, Y. & Goddard, W. A. Wax inhibition by

8

comb-like polymers: Support of the incorporation- perturbation mechanism

9

from molecular dynamics simulations. J. Phys. Chem. B 111, 13173–13179 (2007).

10

11 12

49.

Paso, K., Senra, M., Yi, Y., Sastry, A. M. & Fogler, H. S. Paraffin polydispersity facilitates mechanical gelation. Ind. Eng. Chem. Res. 44, 7242–7254 (2005).

13


Interaction Effects of Predominantly Linear and ... - ACS Publications

Recommend Documents