Raman spectroscopic and kinetic analysis of hydrate shell formation on hydrogel particles containing monoethylene glycol

Raman spectroscopic and kinetic analysis of hydrate shell formation on hydrogel particles containing monoethylene glycol

Accepted Manuscript Raman spectroscopic and kinetic analysis of hydrate shell formation on hydrogel particles containing monoethylene glycol Juwoon Pa...

3MB Sizes 0 Downloads 26 Views

Accepted Manuscript Raman spectroscopic and kinetic analysis of hydrate shell formation on hydrogel particles containing monoethylene glycol Juwoon Park, Hyunho Kim, Yutaek Seo, Wendy Tian, Colin D. Wood PII:

S1875-5100(17)30115-4

DOI:

10.1016/j.jngse.2017.03.009

Reference:

JNGSE 2107

To appear in:

Journal of Natural Gas Science and Engineering

Received Date: 30 October 2016 Revised Date:

28 February 2017

Accepted Date: 1 March 2017

Please cite this article as: Park, J., Kim, H., Seo, Y., Tian, W., Wood, C.D., Raman spectroscopic and kinetic analysis of hydrate shell formation on hydrogel particles containing monoethylene glycol, Journal of Natural Gas Science & Engineering (2017), doi: 10.1016/j.jngse.2017.03.009. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Graphical Abstract

ACCEPTED MANUSCRIPT

Raman spectroscopic and kinetic analysis of hydrate shell

2

formation on hydrogel particles containing Monoethylene

3

Glycol

4

Juwoon Park1, Hyunho Kim1, Yutaek Seo1*, Wendy Tian2, and Colin D. Wood 2*

SC

1

Department of Naval Architecture and Ocean Engineering, RIMS, Collage of Engineering,

6 7

M AN U

5

RI PT

1

Seoul National University, 1, Gwanak-ro, Gwanak-gu, Seoul, 151 - 744, Korea 2

CSIRO Energy, Australian Resources Research Centre, Kensington, WA, 6151, Australia

AUTHOR EMAIL ADDRESSES: [email protected] and [email protected]

9

TELEPHONE: YS +82-2-880-7329; CDW +613-9545-8160

11

EP

FAX: YS +82-2-888-9298; CDW +618-6436-8555

AC C

10

TE D

8

1

ACCEPTED MANUSCRIPT

ABSTRACT

2

Hydrate shell growth on hydrogel particles incorporating monoethylene glycol (MEG) was

3

investigated using Raman spectroscopy and a high pressure autoclave apparatus. Surface images

4

of hydrogel particles covered by a hydrate shell are presented and the obtained Raman spectra

5

indicate MEG molecules were excluded from the formation of the hydrate shell. The Raman

6

spectra and surface imaging of the particles demonstrate that the particles remain intact after the

7

dissociation of the shell. In subsequent hydrate formation cycles, the hydrate fraction in the

8

liquid phase was determined from gas consumption measurements in a high pressure autoclave.

9

The hydrate fraction reached only 0.11 for hydrogels containing MEG while a high hydrate

10

fraction of 0.41 was observed for a control system consisting of bulk water and decane mixture.

11

The growth rate in the presence of hydrogel particles was suppressed in the early stages of

12

hydrate formation, suggesting that the increasing MEG concentration inside the hydrogel core

13

may limit the further inward hydrate growth. After hydrate formation the hydrogel particles were

14

analyzed by thermogravimetric analysis (TGA), suggesting the possibility of recovering

15

hydrogel polymer through conventional heating methodologies. These results demonstrate that

16

injecting hydrogel particles containing high concentration of MEG may be feasible to manage

17

the risk of hydrate plug formation as they will absorb the free water resulting in the dispersed

18

hydrogel particles among the hydrocarbon liquid phase.

19

KEYWORDS Flow assurance, Gas hydrate, Hydrogel, Shell formation, Anti-agglomeration,

20

Plug prevention.

AC C

EP

TE D

M AN U

SC

RI PT

1

21 22

2

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

1

3

ACCEPTED MANUSCRIPT

INTRODUCTION

2

Gas hydrates are nonstoichiometric crystalline compounds, consisting of host water and guest

3

hydrocarbon molecules such as methane, ethane, and propane (Sloan and Koh, 2008; Sloan et al.,

4

2009). They are formed through hydrogen bonding between water molecules which enclathrate

5

guest molecules at high pressure and low temperature, conditions which are found in subsea oil

6

and gas flowlines. Industry practice has been relying on the injection of vast quantities of

7

alcohol- or glycol-based thermodynamic hydrate inhibitors (THI) to prevent hydrate formation in

8

subsea flowlines. However, handling large amount of THI in remote offshore fields significantly

9

increases the capital and operational expenditure, thus adversely affecting the economics of field

10

development. As such hydrate prevention strategies are now moving toward hydrate risk

11

management by either delaying the formation with polymer-based kinetic hydrate inhibitors

12

(KHIs) or preventing the agglomeration of hydrate particles by surfactant-based anti-

13

agglomerants (AAs), which are collectively known as low dosage hydrate inhibitors (LDHIs)

14

(Kelland, 2006; Clard and Anderson, 2007; Duchateau et al., 2009; Del Villano and Kelland,

15

2011)_).

16

Although there have been numerous applications of LDHIs in offshore oil and gas fields,

17

monoethylene glycol (MEG) is still a reliable solution for avoiding hydrate plug formation for

18

offshore fields due to its robust performance, negligible loss to gas phase, and lower Health,

19

Safety, and Environmental (HSE) issues (Sanford et al, 2011). The injection rate of MEG into

20

subsea flowlines is typically based on the lowest temperature and highest pressure conditions

21

that would be encountered in extended shut-in operation of the subsea production system. As the

22

concentration of MEG in water may need to be up to 50 ~ 60 wt%, the industry is spending

23

significant operating and capital expenditure. As such a small reduction in the MEG injection

AC C

EP

TE D

M AN U

SC

RI PT

1

4

ACCEPTED MANUSCRIPT

rate can result in significant saving. Recent studies of hydrate formation kinetics have suggested

2

that the amount of hydrate formed can be controlled in the presence of MEG in water where the

3

concentration of MEG is lower than the value required to fully avoid the hydrate formation under

4

experimental conditions, a practice known as under-inhibition. Li et al. (2011) reported that

5

under inhibition can have an adverse effect where the concentration of MEG in water was in

6

range between 1.0 and 5.0 wt%. Experimental results showed that MEG enhanced hydrate

7

formation kinetics when added at the lower concentration than 5.0 wt%. It has also been

8

observed that the hydrate plug formation potential increased to a maximum at MEG

9

concentration of 10 to 15 wt%. The under-inhibition effect might be effective with MEG

M AN U

SC

RI PT

1

concentration of higher than 20 wt%.

11

Hydrate plug formation involves growth of hydrate crystals followed by agglomeration which

12

results in a high resistance-to-flow for liquid hydrocarbon fluids. Joshi et al (2013) suggested the

13

hydrate particles would suspend homogeneously in the liquid phase during the early stages of

14

hydrate formation, however once the amount of hydrate was greater than a certain hydrate

15

volume fraction in the aqueous phase, heterogeneous particle distribution would be followed by

16

stationary bed formation before the hydrate plug formation. Akhfash et al. (2013; 2016) also

17

observed the transition of homogeneous distribution to heterogeneous segregation of hydrate

18

particles in liquid phase depending on the hydrate volume fraction. In terms of thermodynamics,

19

the presence of MEG in the liquid phase shifts the hydrate equilibrium to lower temperature and

20

higher pressure condition, thus reducing the driving force for hydrate formation. Recent studies

21

also suggest MEG induces low resistance-to-flow for hydrate slurries. Kim et al. (2014)

22

presented the amount of hydrate in liquid phase could be reduced less than 28% by adding 20 wt%

23

MEG while maintaining stable torque. Sohn et al. (2015) also showed that the resistance-to-flow

AC C

EP

TE D

10

5

ACCEPTED MANUSCRIPT

was minimized with under-inhibition of MEG for water and decane mixtures with water cut of

2

60 % and 40%.

3

In our previous works, the potential use of polyacrylamide-based hydrogel particles for

4

preventing the agglomeration of hydrate particles by locking the water phase into particles which

5

ensured that the torque was stable during the hydrate formation (Seo et al., 2014; Park et al.,

6

2016). The use of hydrogel particles was further extended by incorporating mono ethylene glycol

7

(MEG), then their performance for hydrate inhibition was tested using a high pressure autoclave

8

by measuring hydrate onset time, hydrate fraction, and torque changes during hydrate formation.

9

Although the MEG containing hydrogels showed the best performance for forming low hydrate

10

fraction and maintaining stable torque further investigations are required on the role of MEG

11

during the hydrate formation process on the surface of the hydrogel particles, especially with a

12

spectroscopic method. Raman spectroscopy was used to identify the structural characteristics of

13

the gas hydrates formed on the surface of the hydrogel particle. The hydrate structure type and

14

distribution of guest molecules on the surface of hydrate particle was investigated from Raman

15

spectroscopic analysis. The obtained Raman analysis along with the macroscopic studies on

16

hydrate formation kinetics will provide comprehensive understanding on the effect of hydrate

17

formation on the hydrogel particle. Moreover there has been no systematic study on the

18

mechanism of swelling hydrogels in subsea flowlines and separation of hydrogel polymer from

19

the absorbed liquid phase.

20

This study represents the first effort to analyze the hydrate shell formed on the surface of a

21

hydrogel particle using Raman spectroscopy. The hydrate structure formed on the surface of the

22

hydrogel particle was investigated along with the presence of MEG on the surface of the hydrate

23

shell. Then TGA analysis was carried out for the hydrogel samples to investigate the separation

AC C

EP

TE D

M AN U

SC

RI PT

1

6

ACCEPTED MANUSCRIPT

of the hydrogel polymer with conventional heating. The formation rate of gas hydrate with and

2

without MEG in hydrogel particles was also investigated by measuring the gas consumption rate

3

during hydrate formation. The obtained results suggests that MEG molecules were excluded

4

from the hydrate shell during its formation and migrated inward, suggesting the under-inhibition

5

effect was coupled with water discretization using hydrogel particles.

RI PT

1

SC

6 EXPERIMENTS

8

Synthesis of MEG-Hydrogel All chemicals for hydrogel particle synthesis were supplied by

9

Sigma-Aldrich and were used without further purification. These chemicals are as follows:

10

polyacrylamide-co-acrylic acid partial sodium salt (PAM-co-AA), Mw 520 000, Mn 150 000,

11

typical acrylamide level 80%; N-(3-(dimethylamino)propyl)-N′-ethylcarbodiimide hydrochloride

12

(EDC, commercial grade); N-hydroxysuccinimide (NHS, 98%); and 1,2-diamino ethane (EDA

13

>99%). The hydrogel synthesis has been explained in our previous work10, which involves the

14

carbodiimide-mediated cross linking (CMC) reaction of PAM-co-AA in an inverse suspension.

15

To incorporate MEG into the PAM hydrogel particles the dried PAM powder were swollen in 20

16

wt% MEG solution with final polymer concentration in the hydrogel of 13 wt%. The obtained

17

hydrogel particles were suspended in a liquid hydrocarbon such as decane where they maintained

18

their spherical shape.

19

Raman analysis of hydrate shell on MEG-hydrogel particle For the spectroscopic analysis of

20

hydrate shell on the surface of MEG-hydrogel, ethane was used to form hydrate with MEG

21

hydrogel particles. The primary objective of this Raman analysis was to observe the presence of

22

MEG in hydrate shell formed on the surface of MEG-hydrate particle. The synthesized MEG-

AC C

EP

TE D

M AN U

7

7

ACCEPTED MANUSCRIPT

hydrogel particles were exposed to ethane gas at 1oC for 3 days. After confirming the hydrate

2

formation visually, the samples are stored in liquid nitrogen to prevent the dissociation of

3

hydrate shell. Jobin-Yvon ARAMIS high resolution dispersive Raman microscope was used to

4

measure the Raman spectra by an electrically cooled (203K) CCD detector. 514.53 nm from Ar-

5

ion laser performed as the excitation source and the laser had 30mW intensity. The measurement

6

was carried out while increasing the temperature in step from 93K to 153K, 213K, and finally

7

243K using a Linkam TMS 94 temperature controller. The images of hydrate shell on the surface

8

of the MEG hydrogel particles were taken using the microscope, and the qualitative analysis of

9

hydrate shell composition was carried out from the obtained Raman spectra.

M AN U

SC

RI PT

1

High pressure autoclave experiments Hydrate formation kinetics were investigated during

11

hydrate formation for four different systems including bulk water, 20 wt% MEG solution, PAM-

12

hydrogel, and MEG-hydrogel particles. To simulate a liquid phase composed of water and

13

hydrocarbon, distilled water, purchased from the OCI, and decane, supplied by SIGMA-

14

ALDRICH Inc, were mixed with the volumetric ratio of water phase, 60%. Synthetic natural gas

15

(CH4: 90 mol%, C2H6: 6 mol%, C3H8: 3 mol%, n-C4H10:1 mol%) and pure C2H6 were supplied

16

from the Special Gas Company (Daejeon, Korea), with a stated minimum purity of 99.995

17

mol%. The details of the high pressure autoclave is presented in our previous works. (Kim et. al.,

18

2014; Park et. al., 2016; Seo, et. al., 2014; Sohn, et. al., 2015)

19

To investigate the hydrate formation characteristics, a liquid phase (80 mL) was loaded into the

20

autoclave with total internal volume of 360 mL Decane and distilled water were used to

21

represent liquid hydrocarbon phase and water phase respectively, which is a control without

22

hydrogel polymer in aqueous phase. The volume ratio of water in liquid phase was 60% for all

23

experiments in this work whether it was bulk water or hydrogel particles, thus 48ml of water was

AC C

EP

TE D

10

8

ACCEPTED MANUSCRIPT

mixed with 32 ml of decane. The concentration of MEG to investigate the effect of MEG on the

2

hydrate formation characteristics was 20 wt%, thus 12g of MEG was mixed with 48g water. For

3

studying hydrate formation in hydrogel particles, hydrogel particles were mixed with decane

4

instead of bulk water. The reactor was pressurized to 120 bar at 24 oC without mixing, then the

5

stirring was commenced at 600 rpm for 2 hours to saturate the liquid phase with natural gas. The

6

cooling of the system was commenced from 24 oC to 4 oC with a cooling rate of 0.25oC/min,

7

which was to simulate the cooling of hydrocarbon and water mixture in subsea flowlines. The

8

cooling rate can be varied depending on overall heat transfer coefficient of the flowline surface.

9

Once the temperature reached 4 oC, it was maintained at 4 oC for 10 hours to observe the hydrate

10

formation characteristics for the studied system. A total of 32 cycles of hydrate formation and

11

dissociation were repeated for i) bulk water, ii) 20.0 wt% MEG solution, iii) hydrogel with no

12

MEG, and iv) MEG hydrogel with MEG concentration of 20.0 wt% (MEG-20 hydrogel).

13

Hydrate formation rate and the volume fraction in the liquid phase was characterized by

14

analyzing pressure decrease due to gas consumption during the hydrate formation.

15

Thermogravimetric analysis (TGA) TGA for MEG 20% aqueous solution and recovered

16

MEG-Hydrogel samples from Raman analysis were carried out using a Setsys 16/18 (France,

17

Setaram). The temperature ramp went from 25 oC to 325 oC with heating rate of 5 oC per min

18

under nitrogen injection of 30mL/min. Comparison of the TGA curves between the MEG 20 wt%

19

solution and the MEG-20 hydrogel was made to confirm the possibility of evaporating MEG

20

from the hydrogel particles using conventional heating method.

AC C

EP

TE D

M AN U

SC

RI PT

1

21

9

ACCEPTED MANUSCRIPT

RESULTS AND DISCUSSION

2

Spectroscopic analysis of hydrate shell on hydrogel particles

3

It was presumed that hydrate would form on the surface of hydrogel particles and the

4

composition change on the surface of hydrogel was expected due to exclusion of MEG during

5

the hydrate formation. The Raman spectroscopic study was carried out along with optical

6

imaging to investigate the composition of hydrates formed on the MEG-20 hydrogel particle.

7

The formation of hydrate on the MEG-20 hydrogel particle was visually confirmed (Figure 1),

8

and the Raman spectroscopic observation for the region of C-C stretching for ethane and MEG

9

provides evidence of the hydrate structure and its composition. The temperature of the Raman

10

probe was maintained at 93K initially to ensure that there were significant amounts of crystals on

11

the surface of the hydrogel particle. The melting and boiling temperatures of ethane are 90.4 K

12

and 184.6 K, respectively. After the temperature was increased to 153 K (Figure 1b) there were

13

minimal changes to the shape of the surface. The crystals on the surface of the MEG-20 hydrogel

14

were still present at 213K as confirmed in the spectra(Figure 1c), however 243 K they were

15

dissociated (Figure 1d). The obtained images in Figure 1 confirmed the formation of crystal

16

structures on the surface of MEG-20 hydrogel particle, which would be structure I hydrate

17

formed from ethane and water. After the dissociation of crystal structures, the MEG-20 hydrogel

18

particle recovered its original sphere shape without losing water to the surrounding environment,

19

confirming the performance of polymer PAM-co-AA even at cryogenic conditions. The exact

20

hydrate structure and its composition was confirmed from the analysis of Raman spectra.

AC C

EP

TE D

M AN U

SC

RI PT

1

21

10

SC

RI PT

ACCEPTED MANUSCRIPT

M AN U

1

Figure 1. MEG (20wt%)-PAM-hydrogel with ethane hydrate on the surface. When the temperature was increased, the hydrate shell started to dissociate and no hydrate was observed in the final image at 243K

6

Figure S1 shows the Raman spectra obtained in the range of 300 and 3600 cm-1 for the MEG-20

7

hydrogel particle while maintaining temperature at 93K, 153K, 213K, and 243K. The C-H

8

stretching bond of MEG molecules would appear at 2888 and 2940 cm-1, while the C-H

9

stretching bond of ethane would appear at 2888 and 2943 cm-1(Seo et al., 2009; Kim et al.,

10

2014). They were superimposed on each other as shown in Figure S1 and were difficult to

11

discriminate, although two sharp peaks seemed to disappear at 243 K. The presence of hydrate

12

structure was confirmed with the C-C stretching bond of ethane molecules in Figure 2, where the

13

Raman spectra obtained in the range of 300 and 1600 cm-1 were focused. The Raman peak at 999

14

cm-1 presents the C-C stretching bond of ethane molecules encaged in large cage of structure I

15

hydrate, which can be confirmed between 93K and 213K, but is not shown at 243K. These

16

Raman spectra were well aligned with the images shown in Figure 1, suggesting the ethane

17

hydrate was formed and stayed on the surface of the MEG-20 hydrogel particle between 93K and

AC C

EP

TE D

2 3 4 5

11

ACCEPTED MANUSCRIPT

213K, but dissociated when the temperature was increased to 243K. This is due to the fact that

2

the dissociation temperature of ethane hydrate in MEG 20 wt% solution was estimated to be

3

240.4K at atmospheric pressure by CSM Gem. The presence of structure I hydrate on the surface

4

of the MEG-20 hydrogel particle was confirmed from the images and Raman spectra in Figure 1

5

and 2. However the intensities of the peaks for MEG changed adversely, small peaks from 93K

6

to 213K but high intensity peaks at 243K, suggesting the concentration of MEG changed on the

7

surface of the MEG-20 hydrogel particle during the process of hydrate formation and

8

dissociation.

M AN U

SC

RI PT

1

AC C

EP

TE D

9

10 11 12

Figure 2. Raman spectra obtained in a focusing area at 93K, 153K, 213K, and 243K, respectively. When the temperature increased, the hydrate shell dissociate and no hydrate was

12

ACCEPTED MANUSCRIPT

observed at 243K. Note the Raman spectra range is between 300 and 1600 cm-1 to focus on C-C stretching bond of ethane and MEG molecules.

4

The concentration of ethane would be highest at the interface between the ethane gas and the

5

MEG-20 hydrogel. Hydrate forms on the surface of the hydrogel particles, during which the

6

MEG and polymer are excluded from the hydrate structure. Eventually the formation of ethane

7

hydrate ceases due to mass transfer limitations through the hydrate shell while a separate solid

8

phase covers the surface of the MEG-20 hydrogel particle. As the MEG molecules cannot

9

participate in the hydrate structure, the molecules are excluded from the hydrate crystal structure

10

during the formation process and are believed to be concentrated inside the hydrogel core. This is

11

the most likely scenario because the MEG is more soluble in water within the particle than in the

12

surrounding gas phase. Most of the solid phase observed in Figure 1 (a) to (c) is the ethane

13

hydrate while only a small amount of MEG exists as can be seen from the Raman spectra for

14

ethane and MEG in Figure 2. Ethane remains in the large cages of structure I hydrate up until

15

213 K from the Raman peak at 999 cm-1. However, at 243K the Raman peak for ethane in

16

hydrate cages disappears, suggesting the hydrate shell is dissociated and only dissolved MEG

17

remains on the surface of the hydrogel particle. It is noted that the intensity of the Raman peaks

18

associated with MEG (866, 1050-1150, 1459 cm-1) increased as the temperature was raised to

19

243K. Upon complete dissociation of the hydrate, the MEG molecules is absorbed back into the

20

surface of the MEG-20 hydrogel particle again while the hydrogel particle gains its original

21

sphere shape (Figure 1 (d)).

22

Overall the MEG-20 hydrogel shows reversible behavior in a hydrophobic environment during

23

hydrate formation and dissociation. This reversible behavior of the MEG-20 hydrogel particle

24

would be beneficial to utilize the MEG-20 hydrogel for managing the hydrate formation risks in

AC C

EP

TE D

M AN U

SC

RI PT

1 2 3

13

ACCEPTED MANUSCRIPT

subsea flowlines. However more investigations on the exact process of hydrate formation on the

2

surface of the MEG-20 hydrogel particle, including hydrate growth rate and volume fraction

3

comparing to bulk solution, are further required in mixing condition.

4

Hydrate formation kinetics as determined using a high pressure autoclave

5

The hydrate formation experiments were carried out in a high pressure autoclave with continuous

6

cooling and mixing of fluids at 600 rpm to ensure that the decane and water/hydrogel were

7

completely mixed. Figure 3 shows the pressure-temperature (PT) trace during the cooling of

8

natural gas and liquid mixture along with the hydrate equilibrium conditions. Figure S2 presents

9

the estimated hydrate equilibrium conditions in the presence of MEG in pressure range between

10

2.0 and 14.0 MPa. The experiments were carried out in isochoric condition and the pressure

11

continuously decreased with a rate that corresponds to the cooling of fluids from 281.1 K to

12

271.1 K until the hydrate equilibrium for water-decane-natural gas mixture was passed. When

13

the condition reached the point O1, hydrate formation was observed visually and the subcooling

14

temperature, i.e. temperature difference between equilibrium condition and onset condition, was

15

determined to be 5.6 K. Since the hydrate onset, fast pressure reduction was observed until the

16

point C1, then the PT trace returned to its initial cooling path due to gas consumption for hydrate

17

formation. The hydrate formation in the hydrogel and decane mixture started slightly earlier in

18

the point O2, resulting in a lower subcooling temperature of 4.5 K than bulk water and decane

19

however, the reduced pressure due to hydrate formation was much smaller than the bulk water

20

system.

21

The cooling was completed at 276.1 K and 10.0 MPa for water, decane, and natural gas mixture.

22

The standard industry practice for completely avoiding hydrate formation, is in this system

AC C

EP

TE D

M AN U

SC

RI PT

1

14

ACCEPTED MANUSCRIPT

would require a MEG concentration in the aqueous phase to exceed 43.0 wt% in order to shift

2

the hydrate equilibrium conditions to lower temperature and higher pressure region than the

3

cooling path of the fluids, as seen in Figure 3. The addition of 20.0 wt% MEG (i.e., under

4

inhibited system) in the aqueous phase would not be able to prevent the hydrate formation with

5

the cooling path and would induce the hydrate formation. However, the resulting hydrate onset

6

condition for MEG 20.0 wt% bulk solution and MEG-20 hydrogel particles were shifted to a

7

lower temperature region because the hydrate equilibrium condition was shifted in the presence

8

of MEG as well. From the calculated equilibrium condition for MEG 20.0 wt% MEG solution

9

with natural gas, the resulting subcooling temperature was 7.5 K in the hydrate onset condition

10

point O3. For the MEG-20 hydrogel particles and decane mixture, the subcooling temperature

11

from the onset condition in the point O4 was 5.8 K. The pressure reduction due to hydrate

12

formation was relatively small for both MEG 20.0 wt% bulk solution and MEG-20 hydrogel

13

particles. The addition of MEG increased the subcooling temperature from 5.6 K to 7.5 K,

14

however the MEG-20 hydrogel particles showed less subcooling temperature of 6.8 K,

15

suggesting that the hydrate formation happened faster in MEG-20 hydrogel particles than in

16

MEG 20.0 wt% bulk solution.

AC C

EP

TE D

M AN U

SC

RI PT

1

15

ACCEPTED MANUSCRIPT

12.0 Bulk water Hydrogel MEG 20wt% solution MEG-20 hydrogel Hydrate Onset conditions

11.0

O2.Onset-Hydrogel O1.Onset-Bulk water

O4.Onset-MEG Hydrogel O3.Onset-MEG solution C3.Complete-MEG solution

C2.Complete-Hydrogel

C4.Complete-MEG Hydrogel C1.Complete-Bulk water

10.0

M AN U

SC

10.5

RI PT

Pressure (MPa)

11.5

Commence cooling

MEG 20wt%

MEG 43wt% 275

280

285

290

MEG 0 wt% 295

300

Temperature (K)

Figure 3. Pressure-Temperature profile during the hydrate formation form bulk water + decane, hydrogel + decane, MEG 20wt% bulk solution + decane, and MEG-20 hydrogel + decane. The solid lines represent the hydrate equilibrium curves for bulk water (black), MEG 20 wt% solution (blue), and MEG 43 wt% solution (red).

6

The hydrate formation on the surface of hydrogel particles occurred at a faster rate than the bulk

7

water phase, possibly due to dispersion of hydrogel particles and enhance mass transfer of gas

8

molecules through the particles (Su et al., 2009). Figure 3 confirms that the hydrate formation

9

characteristics are very different when the system is under inhibited and in the presence of

10

hydrogel polymer. Further analysis on hydrate volume fraction change in liquid phase was

11

carried out from the pressure change in Figure 3 as a function of time.

12

Table 1 and Figure 4 present the hydrate formation characteristics for bulk water, hydrogel

13

particles, MEG 20.0 wt% solution, and MEG-20 hydrogel particles. The aqueous phase was

14

mixed with decane with watercut 60% for every experiments. At least eight experiments were

15

carried out for each system to determine the average values of water conversion to hydrate,

16

hydrate fraction in liquid phase, and growth rates. Figure 4 shows the hydrate fraction changes in

AC C

EP

TE D

1 2 3 4 5

16

ACCEPTED MANUSCRIPT

liquid phase as a function of time, where time zero indicates the hydrate onset point in Figure 3.

2

The consumed gas molecules due to hydrate formation was estimated from the pressure

3

difference between monitored moment and calculated pressure with the assumption of no hydrate

4

formation. This procedure has been suggested as a method for hydrate formation study in a flow

5

wheel and an autoclave systems (Ke et al., 2011; Hemmingsen, et al., 2008). The hydrate

6

fraction, Φhyd in the liquid phase at 150 min since the hydrate onset was calculated from the

7

following equation.

+

+(



M AN U

=

SC

RI PT

1

,

)

where Vw is the volume of water, Vw,conv is the volume of the water converted to hydrate, Vdecane is

9

the volume of decane, and Vhyd is the volume of hydrate calculated from the molecular weight

10

and density of hydrates calculated at a given time. The hydration number was assumed 6.5. As

11

the volume fraction of decane in the liquid phase was 0.4, the maximum hydrate fraction in

12

liquid phase can be 0.6 in case of complete conversion of water to hydrate.

13

As hydrate nuclei grew while consuming the dissolved gas molecules in bulk water phase, the

14

linear growth rate appeared in early stage of hydrate growth. Figure 4 showed the initial growth

15

rate of hydrates in bulk water was 0.0082 vol.frac./min before an inflection point at around 30

16

min. Hydrates growth continued until the hydrate fraction reached 0.4 within 60 min since the

17

hydrate onset, where the water conversion into hydrate was 60.2 %. The addition of PAM-co-AA

18

transformed the bulk water into discretized hydrogel particles mixed with decane and the

19

obtained hydrate growth for hydrogel particles was shown in Figure 4. The initial growth rate

20

was 0.0118 vol.frac./min, which was much faster than in bulk water. As the aqueous phase was

21

already dispersed in particulate format, mass transfer of gas molecules into hydrogel particles

AC C

EP

TE D

8

17

ACCEPTED MANUSCRIPT

was enhanced. Although fast formation rate was achieved in hydrogel particles, inward growth

2

of hydrate shell on the surface of hydrogel particle results in a barrier that retards the mass

3

transfer of gas molecules to the interior of the hydrogel particles. The hydrate fraction reached

4

0.1 at 10 min after the onset, then continued to 0.17 at 150 min with water conversion of 24.1 %.

5

For MEG 20.0 wt% bulk solution hydrate growth was slow for the first 60 mins initially after the

6

hydrate onset. The slow growth rate was 0.0005 vol.frac./min, an order of magnitude less than in

7

bulk water. The growth rate surged into 0.0101 vol.frac./min at 70 min after the onset, which is

8

close to the rate for bulk water. Hydrate kept growing until the hydrate fraction in the liquid

9

phase reached 0.2 at 90 min. The final hydrate fraction at 150 min was 0.22 with water

10

conversion to hydrate of 31.3 %. These results suggested that the MEG molecules in the aqueous

11

phase resulted in strong hydrogen bonding with the water molecules which retard the

12

crystallization into hydrate structures.

13 14 15

Table 1. Hydrate formation characteristics for bulk water and hydrogel systems. rslow: slow growth period due to inhibition of hydrate formation, rfast: fast growth period limited by mass transfer.

Bulk water

MEG 20.0 wt% solution MEG-20 hydrogel 16

TE D

rslow (vol.frac./min)

rfast (vol.frac./min)

5.6

60.2

0.41

-

0.0082

4.5

24.1

0.17

-

0.0118

7.5

31.3

0.22

0.0005

0.0101

5.8

15.4

0.11

0.0005

0.0051

AC C

Hydrogel

∆Tsub Water Hydrate volume (K) conversion (%) fraction

EP

System

M AN U

SC

RI PT

1

17

18

ACCEPTED MANUSCRIPT

RI PT

0.3

0.2

SC

Hydrate fraction in liquid phase

0.4

0.1

0.0 0

30

90

120

150

Figure 4. Hydrate volume fraction changes in liquid phase as a function of time for bulk water, hydrogel particles, MEG 20.0 wt% solution, and MEG-20 hydrogel particles.

TE D

4

60

Time (min)

1 2 3

M AN U

Bulk water + Decane Hydrogel + Decane MEG 20wt% + Decane MEG-Hydrogel + Decane

For MEG-20 hydrogel particles mixed with decane, the slow growth period was also observed

6

until 30 min after the onset and the growth rate was just same with the MEG 20.0 wt% bulk

7

solution, 0.0005 vol.frac./min. Fast hydrate growth was followed with a rate of 0.0051

8

vol.fac./min, but the rate was less than in bulk water. The hydrate growth on the surface of

9

MEG-20 hydrogel particles reached 0.10 at 70 min, then slowly reached 0.11 at 150 min with

10

water conversion into hydrate of 15.4 %. As we observed in Raman spectroscopic analysis, MEG

11

molecules on the surface of MEG-20 hydrogel particles would migrate inside the hydrogel core

12

during the inward growth of hydrate shell. As the MEG molecules migrated inside, fast hydrate

13

formation would occur. However, the hydrate formation reached steady-state early because of

AC C

EP

5

19

ACCEPTED MANUSCRIPT

the mass transfer limitation through the hydrate shell formed on the surface of the hydrogel

2

particle. The resulting hydrate fraction and water conversion into hydrate was less than other

3

aqueous system, i.e. 15% less than for bulk water. It can be concluded that the advantage of

4

employing under-inhibition was commingled with the effect of transforming the aqueous phase

5

into particulate formats, as seen in slow growth period and less hydrate fraction in liquid phase.

6

Moreover, hydrates only formed on the surface of MEG-20 hydrogel particles without sintering

7

or aggregation of the particles.

8

Unlike bulk water, there was no acceleration in hydrate growth for MEG-20 hydrogel particles,

9

as shown in Figure 4. Slow initial growth was followed by fast growth, but soon reached steady-

10

state. The mechanism of hydrate growth is different between bulk water and MEG-20 hydrogel

11

particles. Compared to fast formation in bulk water, MEG-20 hydrogels do not show any

12

evidence of hydrate plug formation. Figure 4 shows the hydrate growth on the surface of MEG-

13

20 hydrogels steadily decreases with time and the hydrate fraction in liquid phase reaches only

14

0.11. Therefore, the concentration of MEG in the hydrogel particles slightly increases to 24 wt%

15

considering the loss of water into hydrate shell.

16

Formation and regeneration of MEG hydrogel particles

17

Figure 5 shows the schematic diagram of employing MEG hydrogel particles in subsea flowline

18

to manage the risk of hydrate plug formation. The MEG hydrogel particles with MEG

19

concentration of 80 ~ 90 wt% are carried into the subsea tree to be mixed with produced

20

hydrocarbon and free water stream (Figure 5 A). MEG is a polar compound and has a strong

21

interaction with the carbon-backbone of the polymer structure. Thus MEG is absorbed

22

into the hydrogel particle in a similar way to water, which is favorable for delivering the MEG

AC C

EP

TE D

M AN U

SC

RI PT

1

20

ACCEPTED MANUSCRIPT

into a subsea flowline without loss to the surrounding environment. The swelling ratio of PAM-

2

co-AA was previously measured as 25 for water, however it was about 10.0 for MEG (Sheng et

3

al., 2014). Therefore, MEG hydrogels with desired MEG concentration can be produced with the

4

pre-determined swelling ratio for MEG. Based on the amount of free water in the subsea flowline,

5

the volume of MEG hydrogel particles can be calculated to achieve the target concentration of

6

MEG when the MEG hydrogel particles are mixed with free water (Figure 5 top B), then

7

absorbing the free water until the maximum swelling ratio is reached. The final concentration of

8

MEG in MEG hydrogel particles could be lowered to 20.0 wt% which is an under-inhibited

9

condition. Although the temperature and pressure are suitable for hydrates formation, they are

10

formed on the surface of the MEG-20 hydrogel particles with slow growth period and low

11

hydrate fraction as discussed in the previous section (Figure 5 top C).

12

The images of MEG-20 hydrogel particles dispersed in decane phase are shown in Figure 5. In

13

this work, the MEG hydrogels were produced by swelling the dry polymer with water droplets

14

suspended in a liquid hydrocarbon phase. Several MEG hydrogel particles were sized of around

15

1000 micron, however most of the hydrogel particles shows distribution with mean diameter of

16

280 micron. The contact points between the hydrogel particles in the image show the distortion

17

of the spheres without agglomeration of the particles, indicating the mechanical integrity of the

18

particles.

19

The reversible behavior observed for the MEG hydrogel particles is shown schematically in

20

Figure 5. When the MEG hydrogel particles are in contact with the gas molecules under

21

conditions where hydrate can form, a hydrate shell forms on the surface of the hydrogel particles

22

as the concentration of gas is highest on the surface. The hydrate shell grows inward with

23

increasing the thickness of the shell. The diffusion of gas molecules into the hydrogel core is

AC C

EP

TE D

M AN U

SC

RI PT

1

21

ACCEPTED MANUSCRIPT

limited by the hydrate shell. The decreasing driving force for hydrate formation due to increasing

2

concentration of MEG inside hydrogel core also prevents further growth of the hydrate shell.

3

Upon dissociation of the hydrate shell, evolved free water molecules are quickly absorbed back

4

into the PAM-co-AA polymer network and the hydrogel particle recovers its original shape.

5

MEG molecules migrate back into the surface of the hydrogel particles as witnessed by Raman

6

spectra. Theoretically when the MEG hydrogel particles arrive in the inlet facilities on the

7

offshore platform, they must be separated from the gas and liquid hydrocarbon stream. As the

8

MEG hydrogels retain the water phase inside the hydrogel core and can be suspended in liquid

9

phase without aggregation simple filtration or gravitational methods could be used to separate

10

the hydrogel particles from the bulk liquid phase. After they could be routed to a regeneration

11

process to increase the MEG concentration of the hydrogel particles.

M AN U

SC

RI PT

1

AC C

EP

TE D

12

13

22

ACCEPTED MANUSCRIPT

1 2 3

Figure 5. (Top) Schematic of injecting MEG hydrogel into subsea flowline to manage the risk of hydrate plug formation. (Bottom-left) MEG-20 hydrogel particles suspended in decane phase. (Bottom-right) Hydrate shell formation on the surface of a MEG-PAM Hydrogel particle.

RI PT

4 There can be many methods available to regenerate the MEG hydrogel particles including

6

heating, vacuum distillation, and chemical treatment. Thermogravimetric analysis (TGA) was

7

performed on the recovered MEG hydrogel particles to find out the possibility of regenerating

8

MEG hydrogel particles using a conventional heating method. The TGA for MEG 20.0 wt% bulk

9

solution was also carried out. Figure S3 shows the analysis for (a) the 20.0 wt% MEG bulk

10

solution and (b) MEG-20 hydrogel particles. The TGA pattern for 20.0 wt% MEG solution was

11

quite different with that of MEG-PAM hydrogel as the water was evaporated completely at

12

around 100 oC and remaining MEG was removed at 150 oC. It clearly represents the evaporation

13

of water from MEG by heating. The conventional regeneration method in offshore platforms is

14

the distillation of water from MEG solution using the difference of boiling temperature between

15

water and MEG. For MEG hydrogel particles, as the polymer structure holds water in their

16

structure, 43% of water was evaporated at 100 oC and remaining water was completely

17

evaporated at around 150 oC together with MEG. The solid PAM-co-AA polymer was obtained

18

after completing temperature cycle and the initial amount of PAM-co-AA was recovered, which

19

was approximately 13 wt%. The unique evaporation pattern in the presence of PAM-co-AA

20

needs to be investigated further, however the TGA analysis strongly suggested the conventional

21

heating method can be used to regenerate the MEG hydrogel particles. By controlling the

22

temperature cycle, the concentration of MEG hydrogel might be controlled. More comprehensive

23

works will be carried out to find the optimized regeneration method for the MEG hydrogel

24

particles.

AC C

EP

TE D

M AN U

SC

5

23

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

1 2

24

ACCEPTED MANUSCRIPT

1

IV. Conclusions To investigate the structure and composition of hydrate on the surface of MEG hydrogel particles,

3

Raman spectroscopy and microscopy were used to obtain the Raman spectra and the images.

4

Raman spectroscopic analysis clearly indicate that hydrate shell is formed on the surface with

5

ethane peak at 999 cm-1, indicating the formation of structure I hydrate. However this peak

6

disappeared at 243 K upon dissociation of hydrates, instead the Raman peaks associated with

7

MEG increased, suggesting the migration of MEG molecules back into the surface upon the

8

dissociation of hydrate shell. The hydrate growth rate on the surface of MEG hydrogel particles

9

were investigated using a high pressure autoclave. The results revealed that due to the hydrate

10

shell formation the mass transfer of gas molecules inside the hydrogel core is retarded and low

11

hydrate volume fraction of 0.11 is obtained, which is lower than bulk water, 0.41, MEG 20.0

12

wt% bulk solution, 0.17. Moreover, the presence of MEG in MEG hydrogel particles induces

13

slow growth period for 30 min since hydrate onset, during which the growth rate was only

14

0.0005 vol.frac./min, but the rate picked up to 0.0051 vol.frac./min before reaching the hydrate

15

volume fraction of 0.11. The growth rate for MEG hydrogel particles were less than bulk water,

16

MEG 20.0 wt% solution, and even hydrogel particles without MEG. It is presumed that the

17

advantage of under-inhibition concept was commingled with the benefit of using hydrogel

18

particles. The TGA analysis also indicate the conventional heating method can be modified to

19

regenerate the MEG hydrogel particles. The obtained results in this work provide a better

20

understanding of the hydrate formation characteristics on particles, and present an insight to

21

develop an advanced hydrate management strategy using the MEG-PAM hydrogel particles.

AC C

EP

TE D

M AN U

SC

RI PT

2

22

25

1

AUTHOR INFORMATION

2

Corresponding authors: Yutaek Seo and Colin D. Wood

3

Notes

4

The authors declare no competing financial interests.

5

RI PT

ACCEPTED MANUSCRIPT

ACKNOWLEDGMENT

7

This work was supported by the Technology Innovation Program (10060099) funded by the

8

Ministry of Trade industry & Energy (MI, Korea) and also partially supported by the Industrial

9

Strategic Technology Development Program (No. 10045068) of the Korea Evaluation Institute of

M AN U

10

SC

6

Industrial Technology (KEIT) funded by MOTIE.

14

EP

13

AC C

12

TE D

11

26

ACCEPTED MANUSCRIPT

REFERENCES

2

Akhfash, M., Boxall, J. A., Aman, Z. M., Johns, M. L., May, E. F., 2013. Hydrate Formation and

3

Particle Distributions in Gas-Water Systems, Chem. Eng. Sci. 104, 177– 188.

4

Akhfash, M., Aman, Z. M., Ahn, S. Y., Johns, M. L., May, E. F., 2016. Gas hydrate plug

5

formation in partially-dispersed water-oil systems, Chem. Eng. Sci. 140, 337-347.

6

Clark, L. W., Anderson. J., 2007. Low dosage hydrate inhibitors (LDHI): Further advances and

7

developments in flow assurance technology and applications concerning oil and gas production

8

systems. International Petroleum Technology Conference, Dubai, U.A.E.

9

Duchateau, C., Peytavy, J., Clenat, P., Pou, T.-E., Hidalgo, M., Dicharry, C., 2009. Laboratory

10

evaluation of kinetic hydrate inhibitors: A procedure for enhancing the repeatability of test

11

results, Energy Fuels. 23, 962– 966.

12

Del Villano, L., Kelland, M. A., 2011. An investigation into the laboratory method for the

13

evaluation of the performance of kinetic inhibitors using superheated gas hydrates, Chem. Eng.

14

Sci., 66, 1973– 1985.

15

Hemmingsen, P. V., Li, X., Kinnari, K., 2008. Hydrate plugging potential in underinhibited

16

system s. Proceedings of the 6th International Conference on Gas Hydrates (ICGH 2008),

17

Vancouver, British Columbia, Canada.

18

Joshi, S. V., Grasso, G. A., Lafond, P. G., Rao, I., Webb, E., Zerpa, L. E., Sloan, E. D., Koh, C.

19

A., Sum. A. K., 2013. Experimental flowloop investigations of gas hydrate formation in high

20

water cut systems, Chem. Eng. Sci. 97, 198-209.

21

Kelland, M. A., 2006. History of the development of low dosage hydrate inhibitors, Energy

22

Fuels. 20, 825– 847.

23

Ke, W., Svartaas, T. M., Abay, H. K.. 2011. An experimental study on sI hydrate formation in

24

presence of methanol, PVP and PVCap in an isochoric cell. Proceedings of the 7th International

25

Conference on Gas Hydrates, Edinburgh, Scotland, U.K.

AC C

EP

TE D

M AN U

SC

RI PT

1

27

ACCEPTED MANUSCRIPT

Kim, J., Shin, K., Seo, Y., Cho, S. J., Lee, J. D., 2014, Synergistic Hydrate Inhibition of

2

Monoethylene Glycol with Poly(vinylcaprolactam) in Thermodynamically Underinhibited

3

System, J. Phys. Chem. B. 118, 9065-9075.

4

Li, X., Hemmingsen, P.V., Kinnari, K., 2011, Use of under-inhibition in hydrate control

5

strategies, Proceedings of the 7th International confernce on gas hydrates, Edinburgh, Scotland,

6

UK, July 17-21.

7

Park, J., Lee, H., Seo, Y., Tian, W., Wood, C. D., 2016. Performance of Polymer Hydrogels

8

Incorporating Thermodynamic and Kinetic Hydrate Inhibitors, Energy Fuels. 30, 2741–2750.

9

Sanford, E., Alapati, R., 2011. A new, field-proven, cost-effective solution for MEG

M AN U

SC

RI PT

1

10

regeneration unit issues in offshore Australia gas production, APPEA J. 51,193– 199.

11

Seo, Y., Shin, K., Kim, H., Wood, C. D., Tian, W., Kozielski, K. A., 2014. Preventing Gas

12

Hydrate Agglomeration with Polymer Hydrogels, Energy Fuels. 28, 4409-4420.

13

Seo, Y., Kang, S.-P., Jang, W., 2009. Structure and composition analysis of natural gas hydrates:

14

13

15

113, 9641– 9649.

16

Sheng, Q.; Tian, W.; Lapierre, F.; Gao, S.; Mulder, R. J.; Zhu, Y.; Kozielski, K. A.; Wood, C. D.

17

Arrays of polyacrylamide hydrogels using a carbodiimide-mediated crosslinking reaction, J.

18

Appl. Polym., 2014, 131, 40416-40426.

19

Sloan, E. D., Koh, C. A., 2008. Clathrate Hydrates of Natural Gases, 3rd ed., CRC Press (Taylor

20

and Francis Group): Boca Raton, FL.

21

Sloan, E. D., Koh, C. A., Sum, A. K., Ballard, A. L., Shoup, G. J., McMullen, N., Creek, J. L.,

22

Palermo, T., 2009. Hydrates: state-of-the-art inside and outside flowlines, J. Pet. Technol. 61,

23

89–94.

24

Sohn, Y., Kim, J., Shin, K., Chang, D., Seo, Y., Aman, Z. M., May, E. F., 2015. Hydrate plug

25

formation risk with varying watercut and inhibitor concentrations, Chem. Eng. Sci. 126, 711–

26

718.

AC C

EP

TE D

C NMR spectroscopic and gas uptake measurements of mixed gas hydrates J. Phys. Chem. A.

28

ACCEPTED MANUSCRIPT

Su, F., Bray, C. L., Carter, B. O., Overend, G., Cooper, C., Iggo, J. A., Khimyak, Y. Z., Fogg, A.

2

M., Cooper, A. I., 2009. Reversible hydrogen storage in hydrogel Clathrate hydrates, Adv.

3

Mater., 21, 2382–2386.

AC C

EP

TE D

M AN U

SC

RI PT

1

29

ACCEPTED MANUSCRIPT

Supporting information

2

TE D

M AN U

SC

RI PT

1

Figure S1. Entire range (300 to 3600 cm-1) of Raman spectra of MEG 20 hydrogel + ethane

4

hydrate from 93 K, 153K, 213 K, and 243 K. The blue arrows indicate the C-C stretching bonds

5

of ethane and MEG as seen in Figure 2.

AC C

EP

3

30

ACCEPTED MANUSCRIPT

14 Commence cooling

RI PT

10

8

SC

Pressure (MPa)

12

6

M AN U

4

2 275

280

285

290

295

300

Temperature (K)

Figure S2. Pressure and temperature changes during cooling of bulk water, 20.0 wt% solution,

3

hydrogel particles, and MEG-20 hydrogel particles. Solid lines indicate the hydrate equilibrium

4

condition for bulk water (black), MEG 10.0 wt% solution (pink), MEG 20.0 wt% solution (blue),

5

MEG 30.0 wt% solution (green), and MEG 43.0 wt% solution (red).

EP AC C

6

TE D

1 2

31

SC

RI PT

ACCEPTED MANUSCRIPT

1

Figure S3. Thermogravimetric analysis (TGA) of (a) MEG 20.0 wt% bulk solution and (b)

3

MEG-20 hydrogel particle.

M AN U

2

AC C

EP

TE D

4

32

ACCEPTED MANUSCRIPT Highlights for Manuscript entitled “Raman spectroscopic and kinetic analysis of hydrate shell formation on hydrogel particles containing Monoethylene Glycol”

1. This work investigate the formation of hydrate shell on the surface of hydrogels.

RI PT

2. Raman spectra indicates MEG molecules are excluded during hydrate shell formation. 3. The presence of MEG in hydrogel particles reduce the hydrate fraction less than 0.11.

AC C

EP

TE D

M AN U

SC

4. Reversible behaviour of MEG hydrogels can manage the risk of hydrate plug formation.