e ect of droplet size on the macroscopic morphology of

Effect of Droplet Size on the Macroscopic

Morphology of Methane Hydrates

Marıa Alejandra Aguirre

Olga Juliana Mesa

Department of Chemistry and Chemical Engineering

Royal Military College of Canada, Kingston

June, 2012

A thesis submitted to Universidad de Los Andes in partial fulfillment of the

requirements of the degree of Ingeniero Quımico

c©Maria A. Aguirre, Olga J. Mesa 2012

Abstract

Methane clathrate formation on water films without previous hydrate formation his-

tory was studied to assess the effect of water droplet on the macroscopic morphology

of hydrates. Two volumes of water were evaluated under a constant subcooling of 2

K, and it was found that hydrates formed from a 20 µL water droplet presented a

coarse morphology that suggests bigger hydrate grains than the ones found in the

hydrates formed from a 60 µL droplet. It was also observed growth sites counting

with several sites in the 20 µL samples, while in the other volume a single growth

site appeared on the periphery of the water droplet. A third hydrate layer growing

outside of the original water boundary was also observed in both cases, with a clear

difference in morphology between volumes. A hydrate band that separated the halo

from the interior of the hydrate was observed in both volumes, suggesting a common

mechanism that governs hydrate growth. In addition, it was established comparable

local growth rates which may indicate the same transport limitations in both cases.

Resumen

La formacion de hidratos de metano a partir de pelıculas de agua sin historia

previa fue estudiada. Dos volumenes de agua se evaluaron manteniendo constante el

∆Tsub en 2 K, encontrando que los hidratos formados a partir de las gotas de 20 µL

presentan una morfologıa rugosa lo cual sugiere granos mas grandes en relacion a los

encontrados en los hidratos de 60 µL. Adicionalmente, se observo una diferencia en

el numero de sitios de crecimiento contando con la aparicion de diversos puntos en 20

µL, mientras que en el caso de 60 µL un unico sitio de crecimiento se evidencio en la

periferia de la gota de agua. En ambos casos se observo una tercera capa de hidrato

que crecio afuera de los lımites iniciales de la gota de agua, con una clara diferencia

en la morfologıa entre los volumenes evaluados. Otra caracterıstica observada en

los dos volumenes fue una banda que separo el interior del hidrato de la capa que

se extendio por fuera del volumen inicial de agua, esta observacion sugirio que

el crecimiento del hidrato fue gobernado por el mismo mecanismo . Finalmente,

se determino una razon de crecimiento local comparable lo cual puede indicar las

mismas limitaciones en cuanto a la formacion de los hidratos.

Acknowledgements

We would like to thank our thesis advisor, Dr. Juan G. Beltran, the members of

the hydrate research group at the Royal Military College of Canada (RMCC) and

the department of Chemical Engineering at RMCC and Los Andes University. We

also want to thank our families for the company and support during the develop-

ment of our work.

Agradecemos a nuestro asesor de proyecto de grado, Dr. Juan G. Beltran, los

miembros del grupo de investigacion en hidratos en la universidad Royal Military

College of Canada (RMCC) y al departamento de ingenierıa Quımica de RMCC y

de la Universidad de los Andes. Tambien queremos agradecer a nuestras familias las

cuales sirvieron de apoyo durante el desarrollo del presente trabajo.

Contents

1 Introduction 1

2 Background 4

2.1 Clathrate Hydrates . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

2.1.1 Thermodynamics . . . . . . . . . . . . . . . . . . . . . . . . 6

2.1.2 Hydrate Kinetics . . . . . . . . . . . . . . . . . . . . . . . . 9

2.1.3 Location and impact on human activities . . . . . . . . . . . 11

2.2 Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

3 Experimental apparatus 26

3.1 Experimental Apparatus . . . . . . . . . . . . . . . . . . . . . . . . 26

3.2 Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

4 Results 29

4.1 Hydrate Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . 29

4.2 Growth Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

5 Discussion 38

5.0.1 Growth Rate . . . . . . . . . . . . . . . . . . . . . . . . . . 43

6 Conclusions, Recommendations and Future Work 46

6.1 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

6.1.1 Recommendations and Future Work . . . . . . . . . . . . . . 47

i

CONTENTS ii

Bibliography 52

List of Figures

2.1 The three common hydrate unit crystal structures. . . . . . . . . . 6

2.2 Schematic of a pressure vs. temperature diagram for the system methane

+water. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2.3 Region of interest from the partial phase diagram of the system

methane+water. Literature data as compiled by (Sloan and Koh,

2008) was used to give an estimate of temperature and pressure. . . 8

2.4 Map showing worldwide locations of known and inferred gas-hydrate

deposits around the world. . . . . . . . . . . . . . . . . . . . . . . . 11

2.5 Primary methane hydrate film formed in static (non stirred) on a free

gas-water surface.(Makogon et al., 2000). . . . . . . . . . . . . . . . 14

2.6 Methane hydrate whiskery crystals growth with seawater (Makogon

et al., 2000). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

2.7 Soft gel massive methane hydrate crystals formed in water phase

(Makogon et al., 2000). . . . . . . . . . . . . . . . . . . . . . . . . . 15

2.8 Methane hydrate covering the surface of water droplets under high

driving force, 10 minutes after nucleation. (Servio and Englezos, 2003). 16

2.9 Methane hydrate covering two water droplets under low driving force

at three different times (Servio and Englezos, 2003). . . . . . . . . . 17

2.10 Formation of a polycrystalline hydrate shell in water presaturated

with R-141b (Ohmura and Mori, 1999). . . . . . . . . . . . . . . . . 18

2.11 Sequential graphs of the growth of dentritic hydrate crystals in liquid

water presaturated with CO2. . . . . . . . . . . . . . . . . . . . . . 19

iii

LIST OF FIGURES iv

2.12 Sequencial graphs of the growth of methane-hydrate crystals into

liquid water presaturated with methane (Ohmura et al., 2005). . . . 21

2.13 Methane hydrate formation experiment from dissolved methane with-

out any stirring in the cell (Subramanian and Sloan, 2000). . . . . . 22

2.14 Water films completely covered by hydrate at low (left) and high

(right) driving force. Regions appreciated: interior of the film, pe-

riphery, and clathrate extending outside the original water boundary

(Beltran and Servio, 2010). . . . . . . . . . . . . . . . . . . . . . . . 23

2.15 Detail of the hydrate film formed at low driving force (Beltran and

Servio, 2010). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

2.16 Sequence of images of methane hydrate films formed at the surface of

suspended water droplets at 273.35 K and 4.86 and 6.65 MPa (Zhong

et al., 2011). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

3.1 Experimental apparatus. . . . . . . . . . . . . . . . . . . . . . . . . 27

4.1 Hydrate formation sequence on a water droplet (20 µL) with no pre-

vious hydrate formation history. . . . . . . . . . . . . . . . . . . . . 31

4.2 Detail of the hydrate film formed from a 20 µL water droplet with

no previous hydrate formation history. . . . . . . . . . . . . . . . . 32

4.3 Hydrate formation sequence on a water droplet (60 µL) with no pre-

vious hydrate formation history. . . . . . . . . . . . . . . . . . . . . 33

4.4 Detail of a hydrate film formed from a 60 µL water droplet with no

previous hydrate formation history. . . . . . . . . . . . . . . . . . . 34

4.5 Crystal growth progression of a hydrate formed from a 20 µL water

droplet with no previous hydrate formation history. . . . . . . . . . 36

4.6 Crystal growth progression of a hydrate formed from a 60 µL water

droplet with no previous hydrate formation history. . . . . . . . . . 37

5.1 First clathrate growth site located on the periphery of the water droplet. 39

5.2 Water films completely covered by hydrate. (a) 20 µL sample and (b)

60 µL sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

LIST OF FIGURES v

5.3 Detail of a hydrate film (interior), 20 µL (left) and 60 µL (right) . . 41

5.4 Detail of a hydrate film (boundary), 20 µL (left) and 60 µL (right). 42

5.5 Detail of a hydrate film (halo), 20 µL (left) and 60 µL (right) . . . 42

5.6 Initial growth rates of hydrates formed from 20 and 60 µL water

droplets under the range of experimental temperatures. . . . . . . . 44

5.7 Initial growth rates of hydrates formed from 20 and 60 µL water

droplets under the range experimental pressures. . . . . . . . . . . . 45

List of Tables

4.1 Experimental conditions for methane hydrate formation on 20 µL

and 60 µL water droplets without previous hydrate formation history. 29

5.1 Crystal growth rates for methane hydrate without previous hydrate

formation history. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

vi

Chapter 1

Introduction

Aqueous clathrate hydrates, also known as gas hydrates are nonstoichiometric,

crystalline compounds consisting of hydrogen bonded water molecules that trap

small molecules at high pressures and low temperatures (Sloan, 2003). There are

more than a hundred species which can combine with water and form these struc-

tures. Examples of gaseous guest molecules that can be enclathrated are methane,

carbon dioxide, ethane, propane, argon and krypton, among others (Makogon,

2010).

Natural hydrate formations are restricted in location because their stability de-

pends on the pressure, temperature and composition of both the gas and liquid

phases. Therefore, gas hydrates can be found in the deep-ocean and in permafrost

regions where conditions are appropiate for hydrate stabilization (Smelik and King,

1997).

For gas hydrates applications include natural gas storage and transportation,

separation of gases and water, storage material for hydrogen, a mean of cool energy

storage, and desalination of seawater (Sloan and Koh, 2008). Nevertheless, the for-

mation of hydrate blockages in the oil and gas pipelines and the key role of CO2

in the greenhouse effect, among others, are examples of potential hazards and con-

1

CHAPTER 1. INTRODUCTION 2

cerns around the gas hydrates.In addition, gas hydrates represent an engineering

challenge due to the importance they have in gas-dominated systems. There is the

need to understand the initial stages of formation and growth considering variables

such as guest molecules, size of particles, temperature, and pressure, among others.

As an engineering challenge different models have been developed to simulate the

hydrate deposition in pipelines (Jassim et al., 2010). In general, there is the need to

develop models considering the size of the particles since the formation of hydrate

plugs seems to be polydispersed since different sizes of particles are formed (Jassim

et al., 2010), and as a consequence, all models and simulations should consider the

effect of different droplet volumes.

In order to take advantage of hydrate applications numerous studies have been

performed to understand the hydrate crystallization process and which factors af-

fect the morphology. In recent years, substantive progress is seen in the study of the

morphology of hydrate crystals formed at the guest water interface, which is briefly

reviewed here.

First studies on morphology reported different geometries of hydrate crystals

such as threadlike and dendritic (Makogon, 1997). Factors like supercooling, the

operation pressure and the gaseous guest can affect the geometry of gas hydrates.

Maini and Bishnoi demonstrated that hydrates grow until the bubble was com-

pletely entrapped in a hydrate layer (Maini and Bishnoi, 1981). Hydrate crystals

with dendritic morphology grew in large numbers into the liquid-water phase from

the hydrate film when ∆Tsub ≥ 3K, whereas dendritic crystals were replaced by

skeletal or polyhedral crystal when ∆Tsub ≤ 2K (Ohmura et al., 2004). Ohmura,

Sakemoto and Tanaka observed variations on the morphology of the hydrate using

methane, ethane, and propane, concluding that the size of the individual hydrate

crystal increases with decreasing the subcooling (Tanaka et al., 2009). Servio and

Englezos working with CH4 and CO2 as gaseous guests found that the type of hy-

drate guest did not have an effect on the crystal morphology (Servio and Englezos,

CHAPTER 1. INTRODUCTION 3

2003). Beltran and Servio studied the influence of the driving force, the memory

effect, the aging and the reformation on the morphology of a methane hydrate; it

was found that higher driving forces produced smaller hydrate grains and smoother

surfaces than lower driving forces in hydrates without history (Beltran and Servio,

2010).

Although several studies on morphology have been performed, the influence of

the droplet size in the morphology of a hydrate has not been defined yet. The present

study conducted crystal morphology experiments with two different droplet sizes 20

µL and 60 µL using CH4 as the forming gas in an unstirred system, and to correlate

variations in morphology with the different volumes mentioned above trying to keep

all the other variables constant. The present work is organized as follows: Chapter 2

presents general information about hydrates and previous researches about morphol-

ogy. Chapter 3 presents the experimental procedure; Chapter 4 reveals the results

obtained for hydrate morphology and growth rate are later analyzed and discussed

in Chapter 5; Finally, Chapter 6 presents the overall conclusions with some recom-

mendations for future work.

Chapter 2

Background

2.1 Clathrate Hydrates

Gas hydrates are non-stoichiometric, inorganic crystalline substances. They are

inclusion compounds which generally consist of two molecular species that arrange

themselves in space so that one (host) physically entraps the other (guest) (Engle-

zos, 1993). Gas hydrates have crystal structure and are composed of approximately

85-mol% host molecule; therefore many of the hydrate mechanical properties re-

semble those of ice. Some exceptions to these properties are yield strength, thermal

expansivity, and thermal conductivity (Sloan and Koh, 2008).

Gas hydrates are metastable compounds that can form in gas and oil pipelines,

where temperature and pressure are favorable.There are more than 100 species

which can combine with water (Englezos, 1993), but the most widely observed

guest molecules in gas mixture are methane, ethane, propane, i-butane, n-butane,

nitrogen, carbon dioxide and hydrogen sulfide. However, among those, methane hy-

drates occurs the most naturally (Sloan and Koh, 2008).

All common natural gas hydrates belong to the three crystal structures: cubic

structure I (sI), cubic structure II (sII), and hexagonal structure (sH). The spe-

4

CHAPTER 2. BACKGROUND 5

cific equilibrium hydrate structure is mainly determined by the size of the guest.

Structures I and II are the most common hydrate crystal structures and are com-

posed mainly of light hydrocarbons (Sum et al., 2009). Structure I is formed with

guest molecules having diameters between 4.2 and 6 A, such as methane, ethane,

carbon dioxide, and hydrogen sulfide. Nitrogen and small molecules including hy-

drogen (d<4.2 A) form structure II as single guests. Larger single guest molecules

(6 A<d<7 A) such as propane or iso-butane can form sII. Even larger molecules

(typically 7A<d<9 A) such as iso-pentane or neohexane can form structure H when

accompanied by smaller molecules such as methane, hydrogen sulfide, or nitrogen

(Sloan, 2003).

Hydrate structures are composed of five polyhedral cages formed by hydrogen-

bonded water molecules. Each polyhedron can be described using Jeffrey’s nomen-

clature nimi, where ni is the number of edges in face type i, and mi is the number

of faces in with ni edges. Figure 2.4 shows the different structures for gas hydrates.

Hydrate crystal structure I is composed of pentagonal dodecahedrons 512 cavities

and 14-sided cavities with 12 pentagonal faces and 2 hexagonal faces 51262. Struc-

ture II presents guests such as propane, nitrogen and iso-butane among others, and

corresponds to the hexakaidecahedron denoted 51264 based on the four hexagonal

faces, and structure H includes the irregular dodecahedron and icosahedrons called

435663 and, 51268 respectively, presenting a mixture of small and large guests such

as methane+neohexane and methane+adamantine (Sloan and Koh, 2008).

Molecules that are trapped inside the cavities of water are unbound to them so

they are free to rotate and vibrate. Hydrate cavities are prevented from collapse by

the repulsive presence of guest molecules (the cavities are not stable by themselves),

either in the cavity itself or in a large percentage of the neighboring cavities; most

of the times hydrates are formed by introducing hydrophobic gas molecules such

as methane, so the cavity expansion is mainly maintained by the guest repulsion

instead of the attraction between hydrogen bonds in water (Sloan and Koh, 2008).


Figure 2.1: The three common hydrate unit crystal structures. The numbers insquares indicate the number of cage types (Beltran, 2009).

2.1.1 Thermodynamics

Hydrate phase equilibria are normally determined in terms of four variables:

pressure, temperature, water-free hydrocarbon phase composition and the free wa-

ter composition. However, phase equilibrium is mostly discussed in terms of pressure

and temperature as they are the commonly measured variables in a process (Sloan

and Koh, 2008).

A pressure temperature diagram is used to describe the phase behavior of a sys-

tem. Figure 2.2 provides a partial phase diagram for the methane+water system,

the point Q1 represents the temperature and pressure at which ice, hydrate, water


and the vapour phases coexist. Quadruple point (Q1) is also the starting point for

three-phase lines:

1. Liquid water (LW ), hydrate(H) and vapour (V)

2. Ice (I), hydrate (H) and vapour (V)

3. Ice (I), liquid water (LW ) and hydrate

4. Ice (I), liquid water (LW ) and vapour (V)

Figure 2.2: Schematic of a pressure vs. temperature diagram for the systemmethane+water. I = ice; L = liquid rich in water; V = vapour; H = hydrate;Two-phase and three-phase regions are represented by areas, and lines respectively.Q1 represents the I-H-L-V quadruple point (Sloan and Koh, 2008).


The pressures and temperatures of the LW -H-V and the I-H-V lines mark the

limits to hydrate formation: an upper region where hydrates are stable and a lower

region where the hydrate phase does not exist. There is also a third line (I-LW -V)

that divides the two regions mentioned before as seen in Figure 2.2.

The present study is concerned in the region of the methane+water phase di-

agram above the freezing point of water (Figure 2.3). Above the equilibrium line

(H-LW -V) two phases can coexist (liquid water and hydrate).While below the equi-

librium line only liquid water and vapor phase can coexist.

320310300290280270

400

300

200

100

0

T/K

P/Mpa

Figure 2.3: Region of interest from the partial phase diagram of the systemmethane+water. Literature data as compiled by Sloan and Koh, (2008) was usedto give an estimate of temperature and pressure.

However, conditions of temperature and pressure within the stable zone do not


ensure the formation of a hydrate due to metastability (the ability of a nonequilib-

rium state to persist for a long period of time) (Sloan and Koh, 2008).

2.1.2 Hydrate Kinetics

Hydrate formation is a crystallization process and as such it can be broken down

into two steps: nucleation and crystal growth. Later, other crystallization steps may

simultaneously occur such as agglomeration (Monfort et al., 2000).

Nucleation is a stochastic process during which small clusters of water and gas

grow and disperse in an attempt to achieve critical size for continued growth (Sloan,

2003).

A delay or induction time (metastability) from the moment the system is ther-

modynamically favorable to form hydrates, to the observed crystallization time is

characteristic of clathrate formation (Sum et al., 2009). Nucleation can only be

achieved when the solution is supersaturated (when the water (solvent) contains

more dissolved gas (solute) than can be ordinarily accommodated at a tempera-

ture). Supersaturation can represent the driving force for crystallization as it indi-

cates the deviation from equilibrium mole fractions (solubility). The driving force

is the deviation from the equilibrium conditions (Sloan and Koh, 2008).

It is thought that hydrate nucleation and growth will occur within the metastable

region at the water-hydrocarbon interface (Long, 1994). This hypothesis is based

on the fact that the interface lowers the Gibbs free energy of nucleation and hydro-

carbon species concentrations are normally higher at the interface

On the molecular level the mechanism of hydrate growth is considered a combi-

nation of (1) the kinetics of crystal growth at the hydrate surface Englezos-Bishnoi

model (Sloan and Koh, 2008), (2) mass transfer of components to the growing crys-

tal surface Skovborg-Rasmussen model (Sloan and Koh, 2008), and (3) heat transfer


of the exothermic heat of hydrate formation away from the growing crystal surface

(Sloan and Koh, 2008). Nevertheless, the mechanism of growth on the water droplets

is not known, but is normally assumed that water is transferred through capillaries

within the porous hydrate layer and reacts with gas that is surrounding the droplet

(Servio and Englezos, 2003).

Makogon et al., (2000) established that kinetics can be affected because of sev-

eral factors such as (Makogon et al., 2000):

1. Supercooling.

2. Overpressurization.

3. Rate of cooling.

4. Stirring rate.

5. Previous temperature history of water available for hydrate formation.

6. Presence of the sites for hydrate nucleation, such as steel walls of the reactor

or pipeline.

7. Presaturation of water with hydrate forming gas.

8. Additives.


2.1.3 Location and impact on human activities

Deposits of gas hydrates have been discovered around the world near the conti-

nental slopes in oceans, and at depths below the permafrost, these deposits contain

a huge amount of potential energy for the 21st century (Zhong et al., 2011). There

are different locations of gas hydrates but it is still hard to quantify how much

methane, ethane, propane, i-butane, or n-butane is trapped in this sources, and it

has not been done for all the known locations (Sloan and Koh, 2008). Nevertheless,

there are approximations that state that the amount of methane present in these

deposits is equivalent to at least twice the amount of energy of all other fossil fuels

combined, which in terms of volume there are some suggestions near 200000 tril-

lion cubic feet (TCF) at STP, or at least two order of magnitude greater than the

quantity found in conventional sources (Makogon, 2010).

Figure 2.4: Map showing worldwide locations of known and inferred gas-hydratedeposits around the world. Yellow circles represent the recovered gas hydrates, whilered circles show the inferred gas hydrate locations.


The major difficulty in considering natural gas hydrates as energy sources is

the dispersed character of the locations were they are found. Hydrates are difficult

to access considering that they are located in deep oceans and permafrost regions

making harder to recover the gas from them than that from normal gas reservoirs.

Nevertheless, it is likely that in a near future mankind will need to tap that fuel

source to meet growing energy demands (Makogon, 2010).

In addition to the energy represented as gas hydrates, there are some prob-

lems associated to the formation of them in pipelines and wells. The importance of

pipeline blockage increased in the 70’s when plugging of even the largest diameter

pipelines from offshore, arctic fields or the wells from high-pressure underground

storage facilities were reported (Sloan and Koh, 2008).

Natural gas hydrates can be dangerous compounds not only during construc-

tion stages but also during operation stages of process facilities such as platforms,

pipelines and producing gas wells before the gas has been dehydrated. The pre-

vention of hydrates requires substantial investments in inhibitors up to millions

of dollars per year for oil companies (Kvenvolden, 2006). Besides the economical

impact, there are problems related to the production and transportation of gas hy-

drates. They are associated to operation safety problems when the formation of

plugs leads to the line rupture (Austvik et al., 2000).

The sizes of the droplets that may form hydrates are important in gas-dominated

flowlines. As an engineering challenge different models have been developed to sim-

ulate the hydrate deposition in pipelines (Jassim et al., 2010). Jassim et al. (2010)

proposed the concept of the particle deposition velocity as a function of the parti-

cles size; their model presents how small particles are influenced by the main fluid

velocity and how this effect diminishes for relatively large particles. Their analysis

also suggests a certain size of particles in which any growth of it has no significative

effect on the distance deposition. In general, there is the need to develop models


considering the size of the particles since the formation of hydrate plugs seems to be

polydispersed since different sizes of particles are formed (Jassim et al., 2010), and

as a consequence, all models and simulations should consider the effect of different

droplet volumes.

2.2 Morphology

Several studies have been conducted in order to provide a physical picture of the

phenomena that occur upon hydrate crystallization on the water surface (Beltran

and Servio, 2010). Servio and Englezos (2003), and later Shi et al., have observed

that unconverted water entrapped inside a hydrate shell led to a collapse of the

hydrate layer (Servio and Englezos, 2003). They studied the natural gas hydrate

formation, growth, and the variations of gas consumption (Shi et al., 2011).

According to Makogon et al., (2000), there are three types of hydrate crystals:

massive, whiskery, and gel-like crystals. Under certain conditions all three types of

hydrate crystals can form and coexist (Makogon et al., 2000).

Massive crystals grow due to the adsorption of gas and water on the crystal

surface that is being formed. Although crystals may grow in the gas phase, they

are more likely to form on the crystal surface (Figure 2.5). Porosity of the massive

crystals can be up to 80-90%, depending on growth conditions. Whiskery crystals

(Figure 2.6) can grow because of the adsorption of gas and water. These crystals

are the strongest, have the highest density, and dissociate after the increment of

temperature and once all the other crystals are dissociated. Massive or gel crystals

do not grow on the surface of a whiskery crystal. The gel-like crystals (Figure 2.7)

normally grow in the liquid water phase during a pressure or temperature drop and

once a small amount of gas is dissolved in the water. Gel crystals are very soft and

their porosity is near 95-98% (Makogon et al., 2000).


Figure 2.5: Primary methane hydrate film formed in static (non stirred) on a freegas-water surface. p= 56 bar and T= 279.4 K. (Makogon et al., 2000)

The morphology of methane and carbon dioxide hydrates formed from water

droplets, was studied by Servio and Englezos (2003) (Servio and Englezos, 2003).

The droplets were placed on a 316 stainless-steel cylinder covered with a layer of

Teflon and each experiment was performed with two droplets 5 mm and 2.5 mm in

diameter or three droplets with a diameter of 2.5 mm. It was observed that within

less than 5 s after the first growth evidence, the surface of the droplet quickly be-

came jagged and exhibited many fine needle-like crystals extruding away from the

gas-hydrate-water interface. That behavior was observed for all cases working un-

der high driving force, independent of the hydrate forming gas. Servio and Englezos

(2003) established that the thickness and length of the hydrate needles extruding

from the surface is related to the size of the droplet.

Servio and Englezos, (2003) found that independent of the driving force all the

droplets nucleated at the same time, although hydrate formation under high driv-


Figure 2.6: Methane hydrate whiskery crystals growth with seawater. p= 81 barand T= 274 K (Makogon et al., 2000)

Figure 2.7: Soft gel massive methane hydrate crystals formed in water phase. p=93 bar and T= 280.6 K (Makogon et al., 2000).

ing force was observed to evolve in three phases that are divided as follows: (1) a

hydrate layer appeared around the water droplet along with the needle-like crystals


which grew after time in size and thickness, (2) the crystal needles collapsed onto

the hydrate layer covering the water droplet, and (3) were depressions appeared in

the hydrate layer surrounding the water droplet. The authors suggested that the

collapse of the hydrate layer surrounding the droplets shows evidence that the water

is still being converted to hydrate. This can be appreciated in Figure 2.9.

(d)

(a) (b) (c)

Figure 2.8: Methane hydrate covering the surface of water droplets under high driv-ing force, 10 minutes after nucleation (Servio and Englezos, 2003).

On the other hand, for the experiments performed under low-pressure Servio and

Englezos (2003) observed the lack of any hydrate needles from the hydrate-covered


layer. The texture was smooth and shiny which is opposite to the results obtained

for high driving force, in which the surface was rough and dull. This can be related

to the rate of growth which increases with the degree of supersaturation (Smelik

and King, 1997).

(b)

(a)

(c)

Figure 2.9: Methane hydrate covering two water droplets under low driving forceat three different times. (a) Water droplets at the beginning of the experiment,(b) hydrate covered the initial water droplet 10 h after the experiment began, (c)hydrate covered the initial water droplet 25 h after the experiment began (Servioand Englezos, 2003)


Ohmura, (1999) and Uchida, (1999) also observed the appearance of crystals

growing radially around the hydrate covered surface (Uchida et al., 1999) . Ohmura

et al., (1999) described these crystals as platelike and standing upright on the outer

surface of the drop-enclosing hydrate shell formed as it is seen in Figure 2.10. These

crystals were not observed under low subcooling conditions of approximately 2 K

(Ohmura and Mori, 1999).

(a) 10 sec (b) 30 sec (c) 1 min

(d) 11 min (e) 45 min (f) 180 min

Figure 2.10: Formation of a polycrystalline hydrate shell in water presaturated withR-141b, (a)-(c), and subsequent formation of single-crystal plates growing into waterphase from the hydrate-shell surface, (d)-(f).Indicated below each picture is the lapseof time after artificial hydrate nucleation on the drop surface. (Ohmura and Mori,1999).

Visual observations of the variations in macroscopic morphology of hydrate crys-

tals growing in liquid water saturated with CO2 has been reported (Ohmura et al.,

2004). Water droplets (approximately 4 cm3) were poured into the test cell to form

a pool in the lower portion of the inner space of the test cell. After 10 hours the hy-

drates crystals were dissociated in order to ensure the presaturation of liquid water

with CO2 and hence, shorten the induction time for hydrate re-formation creating

a memory of the prior hydrate formation. The morphology of individual crystals is


presumably feather-like or dendritic shape. As the subcooling is reduced, the authors

recognized dendritic crystals growing downward into liquid water which resulted in

thicker and less densely packed dendrites at the lowest subcoolings (Ohmura et al.,

2004).

(a) 15 s (b) 22 s

(c) 30 s (d) 55 s

Figure 2.11: Sequential graphs of the growth of dentritic hydrate crystals in liquidwater presaturated with CO2. p= 3.4 MPa, T=277.6 K, and ∆Tsub= 3.6 K. Thetime lapse after the formation of a hydrate film covering the CO2-water interfaceis indicated below each graph. Some of the dendritic crystals detached from thehydrate film are falling in liquid water in figures (b) and (d). (Ohmura et al., 2004).

Ohmura et al., (2005) showed distinct variations in the morphology of hydrate

crystals that grew in liquid water depending on the pressure. At pressures between


6 and 8 MPa, hydrate crystals presented skeletal and columnar morphology, while

at the pressure of 10 MPa, hydrates showed dendritic crystals. Hydrate crystals

seem to appear first at the inner surface of the test cell in contact with liquid water

instead of the methane-water interface. Those crystals floated up to the methane-

water interface, where they became a polycrystalline hydrate film, and continued to

grow in the liquid phase (Ohmura et al., 2005).

The crystal growth of methane hydrate presented by Ohmura et al., (2005) was

believe to depend on a mechanism of mass transfer of dissolved methane to the

hydrate-crystal surfaces in contact with liquid water presaturated with methane.

They concluded that hydrate crystals first form a thin polycrystalline layer between

methane and water, and then hydrate crystals grew into the liquid-water phase from

the hydrate film.

Sugaya and Mori, (1996) found that the surface morphology of the hydrate layer

formed at the interface depends strongly on the degree of saturation of the water

with the guest component (Sugaya and Mori, 1996), while Ohmura et al., (1999) es-

tablished that as a general trend, it was observed that subcoolings ≥ 3 K produced

sword-like crystals while at smaller subcooling there was evidence of polygonal faces

and bigger size of latters (Ohmura and Mori, 1999) .

In the same way, visual observations were made by Subramanian et al., (2000)

during methane hydrate formation from dissolved methane in non stirred systems.

They show that hydrates first form as a film at the vapor-liquid interface. After

that, further hydrate growth occurred as fine needles that extended into the bulk

aqueous phase. New methane guest molecules supplied either by the vapor or the

aqueous phase, were needed in order to keep the growth into the bulk liquid. The

authors suggested it may occur by the slow diffusion of methane through open mi-

croscopic cracks in the hydrated V-L interface (Subramanian and Sloan, 2000).

There are two possible sources of the methane molecules that end up being en-


(a) 8 min (b) 10 min

(c) 37 min (d) 300 min

Figure 2.12: Sequencial graphs of the growth of methane-hydrate crystals into liquidwater presaturated with methane. T= 273.7 K and p= 8.2 MPa. The time lapseafter the hydrate nucleation at the methane-water interface is indicated below eachgraph (Ohmura et al., 2005).

clathrated in the growing needles: (1) vapor phase by diffusing through the cracks in

the film to dissolve in the aqueous phase, (2) from the aqueous phase where it is al-

ready dissolved. Nevertheless, in the aqueous phase, water molecules tend to cluster

around the methane molecule in order to maximize the hydrogen bonding around

the hydrophobic solute methane. Figure 2.13 shows the hydrate needles growing

into the bulk aqueous phase.

Tanaka et al., (2009) reported the visual observations of the formation and


0.1 cm

Figure 2.13: Methane hydrate formation experiment from dissolved methane withoutany stirring in the cell. There is evidence of the hydrated V-L interface and thehydrate needles growing into the bulk aqueous phae. The point at which the laseris focused is indicated by a star (Subramanian and Sloan, 2000).

growth of clathrate hydrate crystals on the surface of a water droplet exposed to

gaseous methane, ethane or propane (Tanaka et al., 2009). The growth first occurred

at a random point on the water droplet and then grew to form a polycrystalline

layer covering the surface. They observed the individual crystals that constitute the

polycrystalline hydrate layer and classified the morphology of the hydrate crystals

depending on the system subcooling (∆Tsub). They concluded as a general trend

that at ∆Tsub ≥ 3.0 K, the shape of the hydrate crystals is typically swordlike or

triangular, whereas at ∆Tsub from 2.0 to 3.0 K the shape changes to a polygon. It

was concluded that crystal morphology for methane, ethane or propane gas can be

classified using the ∆Tsub as the common criterion (Tanaka et al., 2009).

Beltran and Servio, (2010) studied the morphology of methane-hydrate films

formed on a glass surface and reported the growth of a hydrate layer outside the

original water boundary (Beltran and Servio, 2010). It was shown two different

morphologies on the water films covered with hydrate, one within the film and the


other on the periphery of it. Differences in methane hydrate morphology were also

found to be dependant on driving force; The high driving force nucleation showed

a smooth surface while a striated pattern was observed at low driving force. It was

found for low driving force that the annulus part was smoother and narrower than

at high driving force, but as it was said, the contrary occurred for the inside of the

hydrate film where coarse grains were observed at low driving force and finer grains

at higher driving force (Figure 2.14). The authors identified hydrate history as a key

factor to determine the macroscopic hydrate growth and morphology in hydrates

growing on a water droplet deposited on a glass surface, and they concluded that

hydrate nucleation occurred on the periphery of a film first, followed by the nucle-

ation within the water film, and that the crystals formed had a completely different

morphology than the ones formed on the water edge.

Figure 2.14: Water films completely covered by hydrate at low (left) and high (right)driving force. Three different regions are appreciable: interior of the film, peripheryof it, and clathrate extending outside the original water boundary. (a) Nucleationand growth occurred at T= 275 K, p= 3.6 MPa. (b) T= 274 K, p= 8.2 MPa(Beltran and Servio, 2010).

Saito et al., (2010) reported detailed observations of the morphology of indi-

vidual hydrate crystals on the surface of a water droplet with methane, ethane, or


Figure 2.15: Detail of the hydrate film formed at low driving force. Light gray:hydrate formed in the periphery. Darker gray: hydrate formed within the waterfilm. Hydrate formed at T= 275 K and p= 3.6 MPa (Beltran and Servio, 2010).

propane, as guest gases (Saito et al., 2010). They observed that the nucleation of

the hydrate first occurred at a random point on the water droplet and then floated

up to the apex of it along the surface. The hydrate grew down to form a polycrys-

talline layer covering the surface of the water droplet. The authors also indicate

that the hydrate crystal morphology has a significant dependence on the system

subcooling, reporting that the size of the individual hydrate crystals decreased with

the increasing ∆Tsub irrespective of the guest substances. According to their results,

pressure difference had no significant effect on the hydrate crystal morphology. The

authors established that hydrates presented a rougher surface at smaller ∆Tsub, and

that the time required for the complete coverage of the water-droplet surface by the

hydrate layer depended significantly on ∆Tsub, indicating that the lateral growth

rate of the hydrate film propagation increased with the increasing ∆Tsub.

Finally Zhong et al., (2011) observed the morphology of hydrate films formed at

the droplet surface and showed that the growth rate of methane hydrate is signifi-

cantly increased as the supersaturation is increased and the droplet size is reduced


(Zhong et al., 2011).

a b c d

e f g h

i j k l

273.35 K,

4.86 MPa,

r0= 2.0 mm

273.35 K,

6.08 MPa,

r0= 2.0 mm

273.35 K,

6.65 MPa,

r0= 2.0 mm

Figure 2.16: Sequence of images of methane hydrate films formed at the surface ofsuspended water droplets at 273.35 K and 4.86 and 6.65 MPa. The time lapse afterthe hydrate formation is indicated below the pictures. (a) 0 s, (b) 10 s, (c) 30 s, (d)10 min, (e) 0 s, (f) 10 s, (g) 30 s, (h) 10 min, (i) 0 s, (j) 10 s, (k) 30 s, (l) 10 min(Zhong et al., 2011).

The sizes of the droplets that may form hydrates are important to simulate

the hydrate deposition in pipelines. In general, there is the need to develop models

considering the size of the particles since the formation of hydrate plugs seems to be

polydispersed since different sizes of particles are formed (Jassim et al., 2010), and

as a consequence, all models and simulations should consider the effect of different

droplet volumes. Jassim et al. (2010) proposed the concept of the particle deposition

velocity as a function of the particles size (Jassim et al., 2010).

Chapter 3

Experimental apparatus

3.1 Experimental Apparatus

Figure 3.1 shows a schematic of the experimental apparatus. Crystallization oc-

cured inside a 316 stainless steel cell, fitted with two sapphire windows on the top

and bottom. A Neslab RTE740, laboratory chiller provided the necessary cooling by

recirculating a mixture of ethylene glycol and water (50/50, V/V) through a cop-

per coil fitted around the pressure vessel’s body. The cell had several ports used as

follows: to feed gas, to purge gas out of the cell, to insert a thermocouple, and to con-

nect the interior of the vessel to a pressure transducer. Temperature was measured

with a type K mini thermocouple probe (±1 K) (Omega Engineering, QC, Canada).

Pressure was monitored with a Rosemount 3051S pressure transducer (Laurentide

Controls, QC, Canada) with an accuracy of ± 0.025% of the span.

A PCO.2000 camera (Optikon Corporation, ON, Canada) recorded high resolu-

tion images of the crystallization process through the top sapphire window. A Schott

KL 2500 (Optikon Corporation, On, Canada) cold light source fitted with an artic-

ulated light pipe was used to light the cell through the bottom sapphire window.

The video camera, the temperature signal and the pressure signal were connected

to a personal computer in order to acquire and analyze the data. Deionized water

26

CHAPTER 3. EXPERIMENTAL APPARATUS 27

Figure 3.1: Experimental apparatus

and methane gas, 99.99% purity (Air liquide, ON, Canada), were used.

3.2 Procedure

A precleaned, microscope, glass slide was cut to fit inside the high pressure cell

(Figure 3.1), and a water droplet was deposited on the slide. Experiments were per-

formed with 20 µL and 60 µL water droplets. Methane was fed to the reactor and

purged several times to minimize the presence of air inside the pressure vessel.

CHAPTER 3. EXPERIMENTAL APPARATUS 28

After purging, the pressure in the reactor was adjusted in order to maintain

a constant driving force (∆Tsub)∗ of approximately 2 K. A hydrate formation ex-

periment was terminated when it became apparent that no liquid water was left

in the microscope slide. After reducing the pressure, the remaining water droplet

was discarded, and a fresh water droplet was used for the next experiment. Three

replicates were performed for each droplet size adjusting the pressure.

∗. ∆Tsub (subcooling) is the difference between the system temperature and the equilibriumtemperature (on the hydrate phase boundary) at the system pressure. Subcooling can be consideredto represent the driving force for hydrate formation Tanaka et al. (2009).

Chapter 4

Results

4.1 Hydrate Morphology

The conditions for methane hydrate formation on 20 µL and 60 µL water droplets

are summarized in Table 4.1.

Table 4.1: Experimental conditions for methane hydrate formation on 20 µL and 60µL water droplets without previous hydrate formation history.

Experiment Droplet size/µL T/K p/MPa Tequilibrium/K a ∆Tsub/K1 20 275 3.8 277 2.02 20 275 4.0 277 2.13 20 276 5.1 279 2.14 60 280 6.6 282 2.05 60 278 5.6 280 2.36 60 280 6.0 281 1.7

a. From data compiled by Sloan and Koh, (2008).

Figure 4.1 presents a sequence of frames of hydrate formation and growth (T=

275 K, p= 4 MPa) on a 20 µL water film without previous hydrate formation his-

tory. Figure 4.1.(a) shows the intact water surface. The first growth site appeared

29

CHAPTER 4. RESULTS 30

on the periphery of the water film followed by several other growth sites (Figure

4.1.(b)). Hydrate propagated from each growth site until the newly formed crystals

covered the entire droplet surface in less than 20 s (Figure 4.1.(c,d)). Following the

complete coverage of the water surface by the hydrate, a thin hydrate film (halo)

extended beyond the original water boundary (Figure 4.1.(e,f)).

Close-up views of the hydrate formed from a 20 µL water droplet are shown in

Figure 4.2. A black strip, approximately 0.06 mm in thickness, separated the hy-

drate formed in the interior of the water droplet from the clathrate that extended

beyond the original water boundary (Figure 4.2.(b,c)). Both interior and exterior

hydrate exhibited a granular texture; however, the halo –or exterior hydrate– ap-

peared lighter and rougher (Figure 4.2.(d)) than the hydrate formed in the interior

(Figure 4.2.(a)).

Figure 4.3 shows the hydrate formed at T= 280 K and p= 6.6 MPa from a 60

µL water droplet. The intact water surface is shown in Figure 4.3.(a). A clathrate

growth site appeared on the periphery of the water film (Figure 4.3.(b)) and grew

into a polycrystalline layer that covered the water surface. After the appearance

of the first growth site, no new growth sites were observed (Figure 4.3.(c)). Figure

4.3.(d) shows the complete coverage of the water surface after 29 s. The hydrate ex-

tended outside the water boundary and covered the complete glass surface (Figure

4.3.(e,f)).


2 m

m

(a)

(b)

(c)

(d)

(e)

(f)

Fig

ure

4.1:

Hydra

tefo

rmat

ion

sequen

ceon

aw

ater

dro

ple

t(2

0µ

L)

wit

hno

pre

vio

us

hydra

tefo

rmat

ion

his

tory

.T

=27

5K

,p=

4M

Pa,

∆Tsu

b=

2K

.(a

)W

ater

film

bef

ore

hydra

tefo

rmat

ion.

(b)t=

0s,

Hydra

tefo

rmat

ion

isob

serv

edat

the

upp

ersi

de

ofth

edro

ple

t,fo

llow

edby

the

app

eare

nce

ofnew

grow

thsi

tes,

(c)t=

4s,

the

spot

ted

hydra

teco

nti

nues

togr

ow.

(d)t=

18s,

the

wat

erfilm

isco

vere

dco

mple

tely

,(e

)ap

pea

rence

ofa

hydra

tela

yer

outs

ide

ofth

eor

igin

alw

ater

bou

ndar

y(h

alo)

,an

d(f

)t=

30s,

hal

oex

tends

onth

egl

ass

surf

ace.


100 µm

(a) (b)

(c) (d)

Figure 4.2: Detail of the hydrate film formed from a 20 µL water droplet withno previous hydrate formation history and a ∆Tsub= 2 K. Hydrate formed at T=275 K, p= 4.0 MPa. (a) Hydrate in the interior of the water droplet, (b) and (c)boundary of the hydrate. (d) Halo

Figure 4.4 presents a detailed view of the different regions of the hydrate formed

from a 60 µL water droplet. A clearly distinct hydrate band separated the halo from

the hydrate in the interior of the film (Figure 4.4.(b,c)). The halo appeared smooth

and shiny (Figure 4.4.(d)) compared to the interior where dark spots were observed

(Figure 4.4.(a)).


3 m

m

(a)

(b)

(c)

(d)

(e)

(f)

Fig

ure

4.3:

Hydra

tefo

rmat

ion

sequen

ceon

aw

ater

dro

ple

t(6

0µ

L)

wit

hno

pre

vio

us

hydra

tefo

rmat

ion

his

tory

.T

=28

0K

,p

=6.

6M

Pa,

∆Tsu

b=

2K

.(a

)W

ater

film

bef

ore

hydra

tefo

rmat

ion.

(b)t=

0s,

firs

tev

iden

ceof

hydra

tefo

rmat

ion

occ

urs

atth

eupp

erp

erip

her

yof

the

dro

ple

tan

dth

ehydra

tefilm

floa

tsto

the

cente

rof

the

film

.(c

)t=

18s,

grow

thof

the

hydra

tein

the

mid

dle

ofth

efilm

.(d

)t=

29s,

the

wat

erfilm

isco

mple

tely

cove

red.

(e)t=

40s,

hydra

teex

tends

outs

ide

ofth

eor

igin

alw

ater

bou

ndar

y.(f

)t=

300

s,hydra

teou

tsid

eth

eor

igin

alw

ater

bou

ndar

yco

vers

the

glas

ssl

ide


100 µm (a) (b)

(c) (d)

Figure 4.4: Detail of a hydrate film formed from a 60 µL water droplet with noprevious hydrate formation history. Hydrate formed at T= 280 K, p= 6.6 MPa,∆Tsub= 2 K. (a) Hydrate in the interior of the water droplet, (b) and (c) boundaryof the hydrate, (d) halo

The first hydrate growth site appeared at the edge of the water film for both

the 20 µL and 60 µL samples (Figure 4.3.(b) and Figure 4.1.(b)). The 60 µL sam-

ples developed from this first growth site only (Figure 4.3); in contrast, the 20 µL

samples propagated from several new growth sites (Figure 4.1). Growth of a thin

hydrate layer outside of the original boundary was observed both for 20 µL (Figures

4.1.(e) and 4.1.(f), and Figures 4.2.(c) and 4.2.(d)) and 60 µL (Figures 4.3.(e) and


4.3.(f), and Figures 4.4.(c) and 4.4.(d)) droplets, after complete hydrate coverage

of the water surface. This process was significantly slower than the initial growth

inside the original water droplet.

After complete coverage of the water film the hydrate on the 60 µL samples

seemed to cave into the water droplet (Figure 4.3.(f)). This collapse of the hydrate

layer was not observed on the 20 µL samples (Figure 4.1.(f)).

The 20 µL samples presented coarse grains and a dark surface (Figure 4.2). The

contrary seemed to be true for the surface of the 60 µL samples which exhibited a

lighter and smoother surface (Figure 4.4). Wavy patterns characterized the halo of

the 60 µL samples (Figure 4.4.(d)), whereas the same layered structure appeared to

be masked by the rough surface of the halo in the 20 µL samples (Figure 4.2.(d)).

4.2 Growth Rate

Hydrate growth rates for 20 µL and 60 µL samples were calculated by tracking

the position of the crystal interface as a function of time. Crystal growth rates var-

ied between samples and appeared to slow down after 10 s (Figures 4.5 and 4.6).

Initial growth rates seemed to be fairly linear throughout the samples, and a least

squares fit was used to compare growth rates for the first 8 s of growth (Table 5.1).

On average, the 20 µL samples were covered with hydrate in 23 s while clathrates

covered the 60 µL samples in 29 s.


0

0.1

0.2

0.3

0.4

0.5

0.6

0 5 10 15 20 25

Gro

wth

/cm

t/s

0

0.1

0.2

0.3

0.4

0.5

0.6

0 5 10 15 20 25

Growth

/cm

t/s

0

0.1

0.2

0.3

0.4

0.5

0.6

0 5 10 15 20 25

Growth

/cm

t/s

(a)

(b)

(c)

Figure 4.5: Crystal growth progression of a hydrate formed from a 20 µL waterdroplet with no previous hydrate formation history. ∆Tsub= 2 K. (a) Experiment 1,T= 275 K, p= 3.8 MPa. (b) Experiment 2, T= 275 K, p= 4.0 MPa. (c) Experiment3, T= 276 K, p= 5.1 MPa


0

0.1

0.2

0.3

0.4

0.5

0.6

0 5 10 15 20 25

Gro

wth

/cm

t/s

0

0.1

0.2

0.3

0.4

0.5

0.6

0 5 10 15 20 25

Gro

wth

/cm

t/s

0

0.1

0.2

0.3

0.4

0.5

0.6

0 5 10 15 20 25

Gro

wth

/cm

t/s

(a)

(b)

(c)

Figure 4.6: Crystal growth progression of a hydrate formed from a 60 µL waterdroplet with no previous hydrate formation history. ∆Tsub= 2 K. (a) Experiment 4,T= 280 K, p= 6.6 MPa. (b) Experiment 5, T= 278 K, p= 5.6 MPa. (c) Experiment6, T= 280 K, p= 6.0 MPa

Table 5.1: Crystal growth rates for methane hydrate without previous hydrate for-mation history.

20 µL 60 µLExperiment p/MPa Growth rate (mm s−1) Experiment p/MPa Growth rate (mm s−1)

1 3.8 0.14 5 5.6 0.342 4.0 0.30 6 6.0 0.223 5.1 0.22 4 6.6 0.15

Chapter 5

Discussion

Hydrate growth sites first appeared on the periphery of the droplet for both

the 20 µL and 60 µL samples (Figure 5.1.(a,b)). This observation agrees with the

location described by Beltran and Servio, (2010) for the first hydrate growth site.

However, the macroscopic morphology that we observed (Figure 4.1 and Figure 4.3)

did not exhibit the dual morphology observed by Beltran and Servio, (2010), nei-

ther was our morphology characterized by dendritic behavior. Our observation of a

specific location for the initial growth site contrasts with Tanaka et al., (2009) work

where initial growth sites were observed to appear at arbitrary locations.

Despite the difference between our low magnification results (Figure 4.1 and

Figure 4.3) and those of Servio’s group (2010), our detailed views (Figure 4.2 and

38

CHAPTER 5. DISCUSSION 39

2 mm (a) (b)

Figure 5.1: First clathrate growth site located on the periphery of the water droplet.∆Tsub= 2 K. (a) 20 µL sample, T= 275 K, p= 4.0 MPa. (b) 60 µL sample, T= 280K, p= 6.6 MPa.

Figure 4.4) revealed a close resemblance to those obtained at high magnification

by Beltran and Servio, (2010). In both our work (Figure 4.2 and Figure 4.4) and

Beltran and Servio’s (2010), a hydrate band separated the interior of the original

water droplet from the hydrate that grew outside of the original water boundary

(halo). We believe that this similitude is an indication of the common mechanism

that governs hydrate growth.

Our 20 µL samples exhibited several growth sites (Figure 4.1.(b) and 4.1.(c))

similar to the numerous growth sites observed by Beltran and Servio, (2010) for 35


µL water droplets. Only one growth site was observed for our 60 µL samples (Figure

4.3.(b) and 4.3.(c)), which agrees with the observation of Tanaka et al., (2009) for

similar water droplet volumes.

Figure 5.2 compares the hydrate formed from a 20 µL water droplet to the

hydrate formed from a 60 µL droplet under the same magnification and lighting

conditions. The hydrate from the 60 µL sample appeared light gray (Figure 5.2.(b))

while the hydrate on the 20 µL sample appeared dark (Figure 5.2.(a)). Light passed

easily through the 60 µL sample, which may indicate that the hydrate layer in the

60 µL was thinner than in the 20 µL sample.

The higher light intensity transmitted through the 60 µL sample could also be

due to the rougher surface in the hydrate formed from a 20 µL water droplet (Figure

5.3). A plausible explanation for the difference in surface roughness is the number

of initial growth sites: single site in the 60 µL samples (Figure 4.3) and multiple

sites in 20 µL samples (Figure 4.1). Many crystal growth centers are known to favor

a rough surface (Servio and Englezos, 2003).

Figure 5.2.(b) shows what seemed to be a depression on the hydrate surface. Shi

et al., (2011) considered the collapse of the hydrate layer as evidence that uncon-

verted water trapped inside the hydrate shell (Shi et al., 2011).

At low magnifications we observed different hydrate morphologies in the inte-

rior of our 20 µL samples (Figure 5.2.(a)) and our 60 µL samples (Figure 5.2.(b)).

Figure 5.3 –obtained at higher magnification– revealed that this difference is due

to grain size. Furthermore, both the 60 µL and 20 µL samples presented the same

overall pattern (Figure 5.4): a hydrate band separating a granular interior from the

hydrate that grew outside of the original water boundary.

A undulating pattern was clearly observed throughout the halo of the 60 µL


2 mm (a) (b)

Figure 5.2: Water films completely covered by hydrate. ∆Tsub= 2 K. (a) 20 µLsample, T= 275 K, p= 4.0 MPa. (b) 60 µL sample, T= 280 K, p= 6.6 MPa.

80 µm (a) (b)

Figure 5.3: Detail of the interior of a hydrate film. ∆Tsub= 2 K. (a) 20 µL sample,T= 275 K, p= 4.0 MPa. (b) 60 µL sample, T= 280 K, p= 6.6 MPa.

samples (Figure 5.5.(b)) while the waves were more difficult to recognize in the 20


100 µm

(a) (b)

Figure 5.4: Detail of a hydrate film, 20 µL (left) and 60 µL (right). ∆Tsub= 2 K. (a)20 µL sample, T= 275 K, p= 4.0 MPa. (b) 60 µL sample, T= 280 K, p= 6.6 MPa.

(a) (b)

100 µm

Figure 5.5: Detail of a hydrate film, 20 µL (left) and 60 µL (right). ∆Tsub= 2 K. (a)20 µL sample, T= 275 K, p= 4.0 MPa. (b) 60 µL sample, T= 280 K, p= 6.6 MPa.


µL samples (Figure 5.5.(a)). A plausible explanation for this wavy pattern is the

movement of water underneath the hydrate layer, drawn by capillarity toward the

bare glass (Beltran and Servio, 2010). The smooth appearance of the halo from the

60 µL samples compared to the 20 µL samples could be explained by the availability

of unconverted water beneath the hydrate layer covering the 60 µL water droplet.

This hypothesis seems to be confirmed by the fact that the halo from the 60 µL

droplet extended over a much wider area than the halo of the 20 µL droplets (Figure

5.2).

5.0.1 Growth Rate

Figures 4.5 and 4.6 showed how hydrate growth slowed down as time progressed.

This is expected, as the clathrate layer reduces the water-gas interface area and in-

creases resistance for the diffusive transfer of methane (Skovborg and Rasmussen,

1994).

Clathrate initial growth rates were measured at pressures from 3.8 to 5.1 MPa at

an average temperature of 275 K for 20 µL droplets and from 5.6 to 6.6 MPa at an

average temperature of 279 K for 60 µL droplets (Table 4.1 and 5.1). Although sub-

cooling was kept constant, varying experimental pressures and temperatures could

have affected the hydrate growth rates (Makogon et al., 2000). Our results show

that, at constant subcooling, initial growth rates were comparable for both droplet

sizes, irrespective of the experimental temperature and pressure (Figures 5.7 and

5.6)

The coarse texture of the hydrate formed from a 20 µL water droplet could

suggest faster growth rates (Servio and Englezos, 2003) than the ones observed on

our 60 µL samples 5.3. However, the observed, initial growth rates for both the 20

†. The coefficient of variation (CV) quantifies scatter points. It is defined as the standarddeviation of a group of samples divided by their mean (?).


0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

274 276 278 280 282

Gro

wth

ra

te/m

ms-1

T/K

20 µL samples

60 µL samples

Figure 5.6: Initial growth rates of hydrates formed from 20 and 60 µL water dropletsunder the range of experimental temperatures. ∆Tsub= 2 K. The CV † for 20 µLsamples was 0.36 while the 60 µL samples presented a value of 0.29. Blue diamondsrepresent data for the 20 µL samples, while red squares are the data for 60 µLsamples.

µL and 60 µL samples averaged approximately 0.2 mm s-1 (Figures 5.7 and 5.6).

In other words, at constant subcooling, the droplet volume did not affect the initial

growth rate. The latter is an indication that the same transport limitations operate

on both droplet sizes.

Considering that the 60 µL droplet had three times the volume of the 20 µL

droplet and that the coverage times were comparable in both cases it could be in-

ferred that the hydrate film formed on the 60 µL samples was thinner than the

one observed on 20 µL samples. This conjecture is supported by the images which

showed that light traversed the 60 µL hydrate with less difficulty than it went

through the 20 µL clathrate (Figures 5.2 to 5.4).


0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0 2 4 6 8

Gro

wth

ra

te/m

m s

-1

p/MPa

20 µL samples

60 µL samples

Figure 5.7: Initial growth rates of hydrates formed from 20 and 60 µL water dropletsunder the range experimental pressures. ∆Tsub= 2 K. The CV for 20 µL sampleswas 0.36 while the 60 µL samples presented a value of 0.29. Blue diamonds representdata for the 20 µL samples, while red squares are the data for 60 µL samples.

Chapter 6

Conclusions, Recommendations

and Future Work

6.1 Conclusions

We performed a set of experiments for methane hydrate formation and growth

to determine the effect of the water volume on the macroscopic morphology of

methane hydrates. In most of the runs for both volumes, the hydrate growth oc-

curred at a specific point on the periphery of the water film. For methane hydrates

formed from 20 µL water droplets without previous hydrate formation history, sev-

eral growth sites appeared on the droplet surface seconds after the appearance of

the first growth site. Those hydrates also presented an apparent thicker hydrate

layer with a rougher surface that suggested coarse crystal grains.

On the other hand, the hydrates formed from 60 µL water droplets grew only

from one growth site and presented a smooth and shiny surface. Other regions of

the hydrate where also analyzed and compared, such as the halo and the boundary

(division between the halo and the interior of the hydrate). A wavy pattern was

observed on the halo surface for both volumes. However, the halo observed in the

hydrate formed from the 20 µL droplet was rougher and thicker. Both the 60 µL

46

CHAPTER 6. CONCLUSIONS, RECOMMENDATIONS AND FUTUREWORK47

and 20 µL samples presented the same overall pattern: a hydrate band separating a

granular interior from the hydrate that grew outside of the original water boundary.

A depression on the surface of the 60 µL sample was observed and explained as an

effect of the water migration theory. No visible collapse or transformation of the

surface of the hydrate formed from a 20 µL water droplet was appreciated.

In general, four main characteristics were observed on the methane hydrate mor-

phology for the two volumes of water evaluated. (1) A difference in the number of

growth sites, (2) a smoother surface was observed in the 60 µL hydrates which sug-

gest smaller grains, (3) a band that separated the interior of the hydrate from the

halo was present in 20 µL and 60 µL samples, and (4) the presence of a layered halo

was observed in both cases, although the roughness in the 20 µL samples masked

it, making harder to recognize the different layers on it.

In addition, it was established that both volumes share the same transport limi-

tations due to equal values of local growth rates. Based on the difference in coverage

times, it could be inferred that the hydrate film formed from a 60 µL water droplet

was thinner than the one observed on 20 µL samples.

Through the study of the initial stages of formation and growth considering

variables such as guest molecules, size os particles, temperature, and pressure, our

results provide a better understanding of the crystallization process, which may help

to develop models that would describe accurately hydrate behavior in pipelines and

wells, providing solutions to one of the most important problems in the oil industry.

6.1.1 Recommendations and Future Work

Based on the analysis of the data showed in the present work, we formulated

the following recommendations or future work:

1. Experiments were performed with water films with no previous hydrate for-

CHAPTER 6. CONCLUSIONS, RECOMMENDATIONS AND FUTUREWORK48

mation history. However, hydrate history is a key factor when determining the

morphology of a gas hydrate. Our recommendation is to carry experiments with

water films with previous hydrate formation history.

2. Perform more experiments with different water volumes to track the behavior

of the system and obtain a better idea of trends in the morphology and growth rate

of the system.

3. Study the dissociation mechanism to observe the main characteristics in the

decomposition of hydrates formed from different water volumes.

Bibliography

Austvik, T.; Li, X. and Gjertsen, L., 2000. Hydrate Plug Properties: Formation and

Removal of Plugs. Annals of the New York Academy of Sciences, 912: 294–303.

Beltran, J., 2009. Equilibrium and Morphology Studies of Clathrate Hydrates. Ph.D.

thesis, McGill University.

Beltran, J. and Servio, P., 2010. Morphological Investigations of Methane-Hydrate

Films Formed on a Glass Surface. Crystal Growth & Design, 10: 4339–4347.

Englezos, P., 1993. Clathrate Hydrates. Ind. Eng. Chem. Res., 32: 1251–1274.

Jassim, E.; Abedinzadegan, M. and Muzychka, Y., 2010. A new approach to in-

vestigate hydrate deposition in gas-dominated flowlines. Journal of Natural Gas

Science and Engineering, 2: 163177.

Kvenvolden, K., 2006. Gas Hydrate and Humans. Annals of the New York Academy

of Sciences, 912: 17–22.

Long, J., 1994. Gas Hydrate Formation Mechanism and Its Kinetic Inhibition. Ph.D.

thesis, Colorado School of Mines, Golden, CO.

Maini, B.B. and Bishnoi, P.R., 1981. Experimental Investigation of Hydrate Forma-

tion Behaviour of a Natural Gas Bubble in a Simulated Deep Sea Environment.

Chemical Engineering Science, 36: 183–189.

Makogon, Y., 1997. Hydrates of Hydrocarbons. PennWell Books.

49

BIBLIOGRAPHY 50

Makogon, Y.; Makogon, T. and Holditch, S., 2000. Kinetics and Mechanisms of Gas

Hydrate Formation and Dissociation with Inhibitors. Annals of the New York

Academy of Sciences, 912: 777–796.

Makogon, Y., 2010. Natural Gas Hydrates- A promising source of energy. Journal

of Natural Gas Science and Engineering, 2: 49–59.

Monfort, J.; Jussaume, L.; El Hafaia, T. and Canselier, J., 2000. Kinetics of Gas

Hydrates Formation and Tests of Efficiency of Kinetic Inhibitors: Experimental

and Theoretical Approaches. Annals of the New York Academy of Sciences, 912:

753–765.

Ohmura, R.; Matsuda, S.; Uchida, T.; Ebinuma, T. and Narita, H., 2005. Clathrate

Hydrate Crystal Growth in Liquid Water Saturated with a Guest Substance:

Observations in a Methane+Water System. Crystal Growth & Design, 3: 953–

957.

Ohmura, R. and Mori, Y.H., 1999. Formation, growth and dissociation of clathrate

hydrate crystals in liquid water in contact with a hydrophobic hydrate-forming

liquid. Journal of Crystal Growth, 196: 164–173.

Ohmura, R.; Shimada, W.; Uchida, T.; Mori, Y.; Takeya, S.; Nagao, J.; Minagawa,

H.; Ebinuma, T. and Narita, H., 2004. Clathrate hydrate crystal growth in liquid

water saturated with a hydrate-forming substance: variations in crystal morphol-

ogy. Philosophical Magazine, 84: 1–16.

Saito, K.; Kishimoto, M.; Tanaka, R. and Ohmura, R., 2010. Crystal Growth of

Clathrate Hydrate at the Interface between Hydrocarbon Gas Mixture and Liquid

Water. Crystal Growth & Design, 11: 295–301.

Servio, P. and Englezos, P., 2003. Morphology of Methane and Carbon Dioxide

Hydrates Formed from Water Droplets. AICHE Journal, 49: 269–275.

Shi, B.H.; Sunb, C.Y.; Zhaoa, J.K.; Dinga, Y. and Chenb, G.J., 2011. An Inward

and Outward Natural Gas Hydrates Growth Shell Model Considering Intrinsic

BIBLIOGRAPHY 51

Kinetics, Mass and Heat Transfer. Chemical Engineering Journal, 172: 1308–

1316.

Skovborg, P. and Rasmussen, P., 1994. A Mass Transport Limited Model for the

Growth of Methane and Ethane Gas Hydrates. Chemical Engineering Science,

49: 1131–1143.

Sloan, E., 2003. Fundamental principles and applications of natural gas hydrates.

Nature, 426: 353–359.

Sloan, E. and Koh, C., 2008. Clathrate Hydrates of Natural Gas. CRC Press.

Smelik, E. and King, H.J., 1997. Crystal-growth studies of natural gas clathrate

hydrates using a pressurized optical cell. American Mi, 82: 88–98.

Subramanian, S. and Sloan, E.J., 2000. Microscopic Measurements and Modeling

of Hydrate Formation Kinetics. Annals of the New York Academy of Sciences,

912: 583–592.

Sugaya, M. and Mori, Y.H., 1996. Behavior of Clathrate Hydrate Formation at the

Boundary of Liquid Water and a Fluorocarbon in Liquid of Vapor State. Chemical

Engineering Science, 51: 3505–3517.

Sum, A.; Koh, C. and Sloan, E., 2009. Clathrate Hydrates: From Laboratory Science

to Engineering Practice. Ind. Eng. Chem. Res., 48: 7457–7465.

Tanaka, R.; Sakemoto, R. and Ohmura, R., 2009. Crystal Growth of Clathrate

Hydrates Formed at the Interface of Liquid Water and Gaseous Methane, Ethane,

or Propane: Variations in Crystal Morphology. Crystal Growth & Design, 9: 871–

879.

Uchida, T.; Ebinuma, T.; Kawabata, J. and Narita, H., 1999. Microscopic obser-

vations of formation processes of clathrate-hydrate films at an interface between

water and carbon dioxide. Journal of Crystal Growth, 204: 348–356.

BIBLIOGRAPHY 52

Zhong, D.; Yang, C.; Liu, D. and Wu, Z., 2011. Experimental investigation of

methane hydrate formation on suspended water droplets. Journal of Crystal

Growth, 327: 237–244.

e ect of droplet size on the macroscopic morphology of

Documents