electrical impedance measurements of...
TRANSCRIPT
UNIVERSIDADE TECNOLÓGICA FEDERAL DO PARANÁ
PROGRAMA DE PÓS-GRADUAÇÃO EM ENGENHARIA ELÉTRICA E INFORMÁTICA INDUSTRIAL
JEAN PAULO NAKATU LONGO
ELECTRICAL IMPEDANCE MEASUREMENTS OF CLATHRATE HYDRATES
MASTER DISSERTATION
CURITIBA
2015
JEAN PAULO NAKATU LONGO
ELECTRICAL IMPEDANCE MEASUREMENTS OF CLATHRATE HYDRATES
Master dissertation submitted to the Graduate Program in Electrical and Computer Engineering of Federal University of Technology - Paraná, as a partial requirement for the degree of Master of Science – Concentration Area: Automation and Systems Engineering.
Advisor: Prof. Marco José da Silva, Dr.
Co-advisors: Prof. Amadeu K. Sum, Dr.
Prof. Rigoberto E. M. Morales, Dr.
CURITIBA
2015
Dados Internacionais de Catalogação na Publicação
L856e Longo, Jean Paulo Nakatu 2015 Electrical impedance measurements of clathrate hydrates /
Jean Paulo Nakatu Longo.-- 2015. 84 f.: il.; 30 cm
Texto em inglês, com resumo em português. Dissertação (Mestrado) - Universidade Tecnológica Federal
do Paraná. Programa de Pós-graduação em Engenharia Elétrica e Informática Industrial, Curitiba, 2015.
Bibliografia: f. 68-76.
1. Gás - Escoamento. 2. Hidratos - Monitorização. 3. Compostos de clatratos. 4. Impedância (Eletricidade). 5. Espectroscopia de impedância. 6. Detectores. 7. Medição. 8. Métodos de simulação. 9. Engenharia elétrica - Dissertações. I. Silva, Marco José da, orient. II. Sum, Amadeu K., coorient. III. Morales, Rigoberto Eleazar Melgarejo, coorient. IV. Universidade Tecnológica Federal do Paraná - Programa de Pós-graduação em Engenharia Elétrica e Informática Industrial. V. Título.
CDD 22 -- 621.3
Biblioteca Central da UTFPR, Câmpus Curitiba
“Se cheguei onde cheguei e consegui fazer tudo o que fiz, foi porque tive a oportunidade de crescer bem, num bom ambiente familiar, de viver bem, sem problemas econômicos e de ser orientado no caminho certo nos momentos decisivos de minha vida.” (Ayrton Senna da Silva)
ACKNOWLEDGMENTS
I thank my advisor Dr. Marco José da Silva for every effort and time devoted to me and
to this work, without which it would not be such an achievement. My gratitude to co-advisors,
professors Amadeu K. Sum from Colorado School of Mines and Rigoberto E. M. Morales
from UTFPR, who actively participated in the development of this work.
I am extremely grateful to my family for all the support and encouragement, especially
my parents Julia and Paulo, and my sister Yuri.
To all professors involved in my training in masters and those involved in the project in
some way. Thanks to Celina Kakitani the attention and helpfulness.
To all my friends and laboratories involved, especially LASII.
To people closest to me, which one way or another pushed me and walked with me
these last times, my deepest gratitude: Aluísio Wrasse, Bruna Leme, Caroline Bertoncelli,
Eduardo Nunes, Frederico Aguiar, Greg dos Santos, Guilherme Weber, Heloísa Stalchmidt,
Izabel Gallera, Maína Worm, Murilo Gabardo, Pedro Murakami, Rafael Dias and Tiago
Vendruscolo.
To those who contributed to the completion of such work.
Financial support of Agência Nacional do Petróleo, Gás Natural e Biocombustíveis –
ANP, the Financier of Studies and Projects – FINEP and the Ministry of Science and
Technology – MCT through the Program of Human Resources ANP for the Oil and Gas
Sector – PRH-ANP/MCT – and Training Programme of Human Resources PETROBRAS -
PRH10-UTFPR.
ABSTRACT
LONGO, J. P. N. Electrical Impedance Measurements of Clathrate Hydrates. 2015. 84 p. Master Thesis (Graduate Program in Electrical and Computer Engineering of Federal University of Technology – Paraná.). Curitiba, 2015.
Among the challenges in the oil and gas industry for hydrocarbon production, a featured area in recent years is known as flow assurance, which involves to guarantee the continuously stream of fluids (oil or gas) through pipelines connecting wellhead to separation systems at topside. One of the main problems is related to the deposition of gas hydrates in pipelines, since these deposits may reduce the effective pipe diameter or even clog pipelines, causing considerable financial losses mainly due to production stop. Gas hydrates or clathrates are crystalline solid ice-like structures, typically formed by water and gas molecules under certain conditions of pressure and temperature. Currently there is no established measurement technique for monitoring the hydrate formation in pipelines. One candidate technique is impedance (or impedance spectrum) measurement of fluids, since it is simple, robust and low cost. With the objective of investigating hydrate formation in a controlled environment, several experiments with three different measuring systems operating three different measuring cells have been performed and evaluated. A fully commercial, a mixed, and a dedicated measuring system were applied for obtaining impedance data of hydrates formation. The experimental tests were performed with a model substance (tetrahydrofuran - THF) in mixture with water which allows the monitoring of hydrates formation at ambient pressures (i.e. no need to use a pressurized cell). The first two systems (commercial and mixed) are able to measure the impedance spectrum in the range 10 Hz to 10 MHz. The dedicated system operates at a fixed frequency (typically 5 MHz). The results show that considerable differences in impedance values are observed for the THF-water mixture in liquid conditions and with the presence of hydrates, hence being promising in hydrates formation monitoring. In this way, the developed measurement system allied to appropriated data processing routines has the potential to be applied as simple tool to monitor hydrate formation in pipes.
Keywords: hydrates formation, flow assurance, electrical impedance measurements, electrical impedance spectroscopy, impedance sensor.
RESUMO
LONGO, J. P. N. Medidas de Impedância Elétrica de Hidratos de Clatrato. 2015. 84 p. Dissertação de Mestrado (Pós-Graduação em Engenharia Elétrica e Informática Industrial da Universidade Tecnológica Federal do Paraná). Curitiba, 2015.
Dentre os desafios no processo de produção de hidrocarbonetos na indústria de óleo e gás, uma área em destaque nos últimos anos é conhecida como garantia de escoamento, a qual envolve assegurar de forma contínua que os fluidos (óleo ou gás) escoem pelas tubulações que conectam o poço aos sistemas de separação na superfície. Um dos principais problemas enfrentados está ligado à deposição de hidratos de gás em tubulações, podendo estes diminuir o diâmetro útil ou até obstruir as tubulações, gerando perdas financeiras consideráveis devido a, principalmente, parada de produção. Hidratos de gás ou clatratos são estruturas cristalinas sólidas semelhantes ao gelo, normalmente formados por moléculas de água e gás em determinadas condições de pressão e temperatura. Atualmente não existe uma técnica de medição estabelecida para monitoração da formação de hidratos em tubulações. Uma técnica candidata é a medição da impedância (ou o espectro de impedância) dos fluidos, pois se trata de uma técnica simples, robusta e de baixo custo. Com o objetivo de investigar o processo de formação de hidratos de forma controlada, foram realizados e analisados diversos experimentos com três sistemas de medição diferentes operando três células de medição distintas. Um sistema totalmente comercial, um sistema misto e um sistema dedicado foram utilizados para realizar medidas de impedância do processo de formação de hidratos. Os testes foram realizados com uma substância modelo (tetrahidrofurano – THF) em mistura com água, com a qual é possível a formação de hidratos sem a necessidade do uso de uma célula pressurizada. Os dois primeiros sistemas (comercial e misto) são capazes de medir o espectro de impedância na faixa 101 Hz até 107 Hz. Já o sistema dedicado opera em frequência fixa (tipicamente 5 MHz). Os resultados mostraram-se promissores no que diz respeito à monitoração da formação de hidratos, já que diferenças consideráveis nos valores de impedância são observadas para a mistura de THF-água em estado líquido e com a presença de hidratos. Dessa maneira, o sistema desenvolvido aliado ao processamento dos dados experimentais pode ser empregado em trabalhos futuros como ferramenta simples para monitorar a formação de hidratos em tubulações.
Palavras-chave: formação de hidratos, garantia de fluxo, medidas de impedância elétrica, espectroscopia de impedância elétrica, sensor de impedância.
LIST OF FIGURES
Figure 2.1 - Water molecules - (a) two pairs of free electrons and (b) Covalent and hydrogen bonds ......................................................................................................................................... 19
Figure 2.2 - Ice structure. Blue spheres represent oxygen and green spheres represent hydrogen. Image generated with help of VESTA 3.2.1 software............................................. 20
Figure 2.3 – Hydrate cages (unscaled) - (a) 512, (b) 51262, (c) 51264, (d) 51268 and (e) 435663. Image generated with help of VESTA 3.2.1 software .............................................................. 21
Figure 2.4 – Hydrates structures - (a) sI, (b) sII and (c) sH. Image generated with help of VESTA 3.2.1 software ............................................................................................................. 22
Figure 2.5 – Typical phase diagrams – pressure (log) versus temperature .............................. 24
Figure 2.6 – Trace for formation of methane hydrates in a closed system .............................. 25
Figure 2.7 – Hydrates formation in pipelines. Grey represents the fluid and white represents hydrates ..................................................................................................................................... 26
Figure 2.8 – Conceptual model for hydrate formation in a multiphase flow system (gas, oil and water) ................................................................................................................................. 27
Figure 2.9 - Most common techniques for inhibition and mitigation of hydrates .................... 29
Figure 2.10 - Representation of impedance and admittance in the complex plane .................. 33
Figure 2.11 - Schematic representation of different dielectric mechanism .............................. 36
Figure 2.12 - Auto-balancing bridge circuit ............................................................................. 36
Figure 2.13 - Frequency response of the auto-balancing bridge .............................................. 37
Figure 2.14 - Two-terminal configuration ................................................................................ 38
Figure 2.15 - Four-terminal configuration ................................................................................ 38
Figure 2.16 - Typical setup of a parallel-plate sensor .............................................................. 39
Figure 2.17 - Typical configuration for interdigital sensors ..................................................... 40
Figure 2.18 - Electric field lines behavior in interdigital sensor .............................................. 40
Figure 2.19 - Two-wire sensor schematic ................................................................................ 41
Figure 3.1 - Block diagram of systems and sensors ................................................................. 42
Figure 3.2 - Diagram of the apparatus ...................................................................................... 42
Figure 3.3 - Precision impedance analyzer block diagram ....................................................... 43
Figure 3.4 - Commercial sensor exploded view diagram - Agilent 16452A ............................ 44
Figure 3.5 - Pulse function arbitrary generator block diagram................................................. 45
Figure 3.6 - Whole dedicated system block diagram ............................................................... 46
Figure 3.7 - Sensor specification. Blue electrode is the receiver. ............................................ 47
Figure 3.8 - Meshes compared: (a) pre-defined and (b) manually refined ............................... 48
Figure 3.9 - Comparison between manually refined and pre-defined meshes ......................... 48
Figure 3.10 - Simulations to find the best sensor configurations: Normalized output X Penetration depth (mm) ............................................................................................................ 49
Figure 3.11 – Electric field behavior in interdigital sensor. hL(6)=4e-4 hL(76)=0.006; Multislice: Electric potential (V) .............................................................................................. 50
Figure 3.12 - Excitation and reception electrodes and ground plane of interdigital sensor ..... 51
Figure 3.13 - Electric field behavior in two-wire sensor. hL(51)=0.05; Multislice: Electric potential (V) .............................................................................................................................. 52
Figure 3.14 - Two-wire sensor - (a) Schematic (b) PCB picture .............................................. 53
Figure 3.15 - Two-wire sensor with the recipient .................................................................... 53
Figure 4.1 - Linearity test results .............................................................................................. 54
Figure 4.2 – Measured system output versus water concentration ........................................... 55
Figure 4.3 - Capacitance of selected substances....................................................................... 56
Figure 4.4 - Comparison of measured capacitance during ice formation from all sensors ...... 57
Figure 4.5 - Comparison of measured resistance during ice formation from all sensors ......... 58
Figure 4.6 - Signal obtained with mixed system –connected to interdigital sensor during ice formation .................................................................................................................................. 58
Figure 4.7 – Permittivity of THF, THF + water mixture and water ......................................... 61
Figure 4.8 - Ice/Hydrates capacitance comparison - Interdigital sensor – Commercial system .................................................................................................................................................. 62
Figure 4.9 - Ice/Hydrates resistance comparison - Interdigital sensor – Commercial system . 62
Figure 4.10 - Ice/Hydrates comparison - Interdigital sensor - Mixed system .......................... 63
Figure 4.11 - Ice/Hydrates comparison - Interdigital sensor - Dedicated system (5 MHz) ..... 64
Figure 4.12 - Ice/Hydrates comparison - Two-wire sensor - Dedicated system (5 MHz) ....... 65
Figure 4.13 - Interdigital sensor with (a) Hydrates and (b) Ice ................................................ 65
Figure A 1 - THF-water mixture without formation capacitance - Commercial sensor - Impedance analyzer .................................................................................................................. 77
Figure A 2 - THF-water mixture without formation resistance - Commercial sensor - Impedance analyzer .................................................................................................................. 77
Figure A 3 - THF-water mixture with hydrates formation capacitance - Commercial sensor - Impedance analyzer .................................................................................................................. 78
Figure A 4 - THF-water mixture with hydates formation resistance - Commercial sensor - Impedance analyzer .................................................................................................................. 79
Figure A 5 - THF-water mixture without formation capacitance - Interdigital sensor - Impedance analyzer .................................................................................................................. 79
Figure A 6 - THF-water mixture without formation resistance - Interdigital sensor - Impedance analyzer .................................................................................................................. 80
Figure A 7 - THF-water mixture with formation capacitance - Interdigital sensor - Impedance analyzer ..................................................................................................................................... 81
Figure A 8 - THF-water mixture with formation resistance - Interdigital sensor - Impedance analyzer ..................................................................................................................................... 81
Figure B 1 - Dedicated system block diagram ......................................................................... 82
Figure B 2 - Pi filter .................................................................................................................. 82
Figure B 3 - Representation of the measurement principle ...................................................... 83
Figure B 4 - Software control menu ......................................................................................... 84
LIST OF TABLES
Table 2.1 - Comparison of Ice, sI and sII hydrates properties.................................................. 23
Table 2.2 - Relative permittivity for selected substances liquid at 25 ºC ................................. 34
Table 3.1 - Absolute values of simulated capacitances ............................................................ 50
Table 4.1 – Reference permittivity (from Table 2.2) and permittivity of tested substances at high-frequency plateau (5 MHz) .............................................................................................. 56
Table 4.2 - Permittivity of THF, THF + water mixture and water ........................................... 61
CONTENTS
1 INTRODUCTION ............................................................................................................. 16
1.1 MOTIVATION ........................................................................................................................................ 16 1.2 OBJECTIVES .......................................................................................................................................... 17
2 THEORETICAL BACKGROUND .................................................................................. 19
2.1 STRUCTURE AND TYPES OF HYDRATES ................................................................................................. 19 2.1.1 Growth process and plugs ............................................................................................................... 23 2.1.2 Inhibition and mitigation of hydrates .............................................................................................. 27 2.1.3 Tetrahydrofuran .............................................................................................................................. 29
2.2 MONITORING TECHNIQUES OF SOLID DEPOSITS ................................................................................. 30 2.3 IMPEDANCE SPECTROSCOPY ................................................................................................................ 31
2.3.1 Definitions ........................................................................................................................................ 32 2.3.2 Typical permittivity values of common substances ......................................................................... 34 2.3.3 Dielectric relaxation ......................................................................................................................... 35
2.4 INSTRUMENTATION FOR IMPEDANCE MEASUREMENT ....................................................................... 36 2.4.1 Auto-balancing bridge ..................................................................................................................... 36 2.4.2 Terminal configurations ................................................................................................................... 37 2.4.3 Measuring cells ................................................................................................................................ 39
3 MEASURING SYSTEMS AND SENSOR DESIGN ...................................................... 42
3.1 OVERVIEW ............................................................................................................................................ 42 3.2 MEASURING SYSTEMS .......................................................................................................................... 43
3.2.1 Commercial System ......................................................................................................................... 43 3.2.2 Mixed System ................................................................................................................................... 44 3.2.3 Dedicated System ............................................................................................................................ 45
3.3 IMPEDANCE SENSORS DESIGN ............................................................................................................. 46 3.3.1 Interdigital Sensor ............................................................................................................................ 46 3.3.2 Two-Wire Sensor .............................................................................................................................. 51
4 TESTS AND RESULTS ................................................................................................... 54
4.1 DEDICATED SYSTEM PERFORMANCE ................................................................................................... 54 4.1.1 Linearity ........................................................................................................................................... 54 4.1.2 Water-Isopropanol Mixture ............................................................................................................. 55
4.2 PERMITTIVITY MEASUREMENT ............................................................................................................ 55 4.3 ICE FORMATION ................................................................................................................................... 56 4.4 HYDRATE FORMATION ......................................................................................................................... 59
4.4.1 Methodology ................................................................................................................................... 59 4.4.2 Results.............................................................................................................................................. 60
4.5 COMPARISON OF ICE AND HYDRATES FORMATION ............................................................................. 61
5 CONCLUSIONS ............................................................................................................... 66
REFERENCES ......................................................................................................................... 68
APPENDIX A .......................................................................................................................... 77
APPENDIX B ........................................................................................................................... 82
B.1 EXCITATION CIRCUIT .................................................................................................................................. 82 B.2 MEASUREMENT CIRCUIT ............................................................................................................................ 83 B.3 MICROCONTROLLER AND FIRMWARE ............................................................................................................ 83 B.4 SOFTWARE .............................................................................................................................................. 84
NOMENCLATURE
Roman symbols
A Area m²
C Capacitance F
D Dielectric displacement C/m²
d Distance m
E Electrical field V/m
f Frequency Hz
G Conductance S
I Electrical current A
j Imaginary Unit 1− -
R Resistance Ω
T Temperature K (oC)
V Voltage V
X Reactance Ω
Y Admittance S
Z Impedance Ω
Greek symbols
Γ Boundary
Electric permittivity F/m
λ Penetration depth m
µ Magnetic permeability -
θ Angle rad, o
ρ Electric charge density C/m³
τ Time constant s
ω Angular frequency rad/s
Subscripts
A Air
H High, Hydrates
HC Hydrocarbon
L Low
s Stray
W Water
x Unknown
Abbreviations
AC Alternating current
DAQ Data acquisition
DC Direct current
DDS Direct digital synthesizer
FEM Finite element method
LASCA Laboratório de automação e sistemas de controle avançado
LASII Laboratório de sensores e instrumentação industrial
LDHI Low-dosage hydrate inhibitor
MEG Monoethylene glycol
MRI Magnetic resonance imaging
NUEM Núcleo de escoamentos multifásicos
PCB Printed-circuit board
PID Proportional Integrator-Differentiator
sH Structure H
sI Structure I
sII Structure II
THF Tetrahydrofuran
UTFPR Federal University of Technology - Paraná
16
1 INTRODUCTION
1.1 MOTIVATION
In today's global economic scenario, petroleum and its derived products are
important commodities being traded worldwide. The extracted oily solution, known as
crude oil, is characterized as a complex mixture consisting essentially of naphtha, oil,
wax, asphaltenes and some inorganic compounds and metals. The oil and derived
products are highly valued in society and because of such importance, all processes
involved in the exploration and production of oil have been thoroughly studied.
A very important concept in oil production is known as ‘flow assurance’, which
means the field within petroleum engineering which aims at sustaining the flow of
produced fluids in the pipelines from well to processing units. Flow assurance issues in
design and operation phases of an exploration field are listed as the major technical
problem in offshore development (Welling and Associates, 2000). In addition, the
creation of a new academic discipline called flow assurance is discussed in forums
around the world.
During the exploration and production of crude oil, several substances can form
plugs and disrupt the flow, for instance asphaltenes, wax and hydrates. Once hydrate is
one of the major problems that can disrupt or interrupt the oil flow, this work will focus
on their study.
Hydrate is a crystalline structure and its formation naturally occurred in the
geosphere million years before its artificial formation in laboratories. Such hydrates are
formed under specific conditions of temperature and pressure when there is a mixture of
water and a specific type of gas (guest), such as methane. These gases are abundant in
some geological conditions such as near the sea floor or in the Arctic permafrost, where
water is abundant and favorable low temperature and high pressures are present
(BARNES, 2014). According to Pfankuch & Rose (1982), any sedimentary basin in a
water depth greater than 180 m has the potential to form hydrates. The hydrate
formation becomes even more critical with higher depths, because high pressures and
low temperatures favor the formation of hydrates.
Hydrates can also be source of problems to the industry. Its formation and
agglomeration are one of the major causes of pipelines blockages
(HAMMERSCHMIDT, 1934; SLOAN and KOH, 2008). In the other hand, because of
17
gas concentration in hydrates (the methane solubility in water is approximately 1:4000,
and its solubility in hydrate is approximately 1:6) and because of less than 15% of the
energy used in the formation is required for dissociation, hydrates could also be
considered as a form of energy in the future (ENGLEZOS, 1993), and a lot of research
is being conducted on that basis (MAKOGON, 1997).
The accumulation of hydrates at the inner pipe walls may considerably decrease
the useful diameter of the pipeline thus reducing produced fluids throughput. Further,
due to such accumulation, the instrumentation and sensors involved can be negatively
affected and compromise some control systems. In such cases, it may be necessary to
remove the deposits from the inner walls of the pipelines.
All these mentioned factors can result in unscheduled stoppages in production –
for pipe cleaning, for instance – and risky operating conditions, which could require
extensive work, instigate production losses and induce irreparable damage, demanding
the abandonment or equipment replacement. Remediation procedures for pipelines
blocks, generated by the hydrate formation, can be extremely costly. Once a hydrate
plug is formed, it can take up to hours, days or (in severe cases) weeks or months to
fully dissociate. It is thus very important to appropriately design and operate an offshore
pipeline to avoid or minimize hydrate risks (GUO et al., 2014).
The reliable detection of hydrate formation in pipelines is important as online
monitoring tool to be applied in production systems. Further, monitoring the presence
and formation of hydrates in experimental pipeline flow loops is of great interest in
order to support the development of accurate models and also for understanding the
phenomena involved. Hence, in both mentioned cases, specially developed sensors and
instrumentation are required.
1.2 OBJECTIVES
Most of the studies about hydrates are performed using PVT cells, where the
pressure, volume and temperature of mixtures are controlled, and several techniques
have been used to study and detect hydrates formation, ranging from simple pressure
and temperature measurements to sophisticated gamma ray or ultrasound measuring
techniques. Nevertheless, currently there is no established measurement technique for
monitoring the hydrate formation in pipelines.
18
A candidate technique is impedance (or impedance spectrum) measurement of
fluids, since it is simple, robust and low cost. Further, impedance measurements have
been used in several industries as reliable tool for monitoring different physical and
chemical parameters (DA SILVA, 2008).
Current work aims to serve as first step into understanding the electrical response
of hydrates formation and to develop a system able to detect hydrate formation based on
impedance measurements. Thus, in this work, a commercial system is used to acquire
the substance’s impedance spectra (complex permittivity) from 40 Hz to 5 MHz, a
mixed system is used to acquire the amplitude spectra from 100 Hz to 10 MHz and a
dedicated system is applied to monitor changes in impedance of mixtures using a single
selectable frequency.
Due to some similarities in properties of the hydrate and the ice and due to the fact
that the physics of ice are well understood, firstly some experiments with ice formation
are performed to validate the involved measuring techniques, as ice data is available in
literature.
In order to develop experiments in atmospheric pressure conditions, the THF
(tetrahydrofuran – C4H8O) is chosen to be the guest molecule for hydrate formation, due
to ease of handling and because of the structure of hydrates formed by THF is the same
as the ones formed by other guest molecules, also found in oil production.
The main objective is to measure hydrates formation, enabling the development of
new studies from the collected data and help gaining better understanding about
hydrates formation.
19
2 THEORETICAL BACKGROUND
2.1 STRUCTURE AND TYPES OF HYDRATES
Natural hydrates are composed of water (at least 82% water) and another
component (most frequently a gas such as methane) which is enclosed in a cage-like
fashion. Gas hydrates or clathrate hydrates are crystalline non-stoichiometric solids. The
guest molecules are trapped in water cavities that are composed by hydrogen-bonded
water molecules.
The hydrogen bond was proposed by Latimer and Rodebush (1920) using a
simplified electrostatic point charge model of the water molecule and Kollman (1977)
confirmed the proposed model using methods such as Morokuma component analysis
and electrostatic potentials. Note that water molecules are composed of one atom of
oxygen and two of hydrogen and it has two positive and two negative poles (Figure
2.1a). The attraction of the positive pole on one molecule to a negative pole on other
molecule causes the water hydrogen bond. Through this, each water molecule is
attached to four other, whereas two hydrogen bonds are donated and two are accepted
(Figure 2.1b).
Figure 2.1 - Water molecules - (a) two pairs of free electrons and (b) Covalent and hydrogen bonds
When water is cooled, the molecules begin to reorganize and, to a certain
temperature and pressure, ice is formed, as seen in Figure 2.2, where normal hexagonal
crystalline ice (known as iH structure) is shown.
20
Figure 2.2 - Ice structure. Blue spheres represent oxygen and green spheres represent hydrogen. Image
generated with help of VESTA 3.2.1 software
The electronic structure of ice is determined by the interaction forces between its
molecules and each one is connected to other molecules by hydrogen bonds, forming a
crystal structure. According to Fletcher (1970), the structure of ice can be regarded as a
completely hydrogen-bonded solid, each molecule forming four hydrogen bonds to its
four nearest neighbors.
Under suitable conditions (generally high pressure and low temperature) and a
suitable (right-sized) guest molecule is present in the water, other structures can be
formed. This is called hydrate cages, and these cavities are not found in liquid water,
only formed when hydrates are formed. Once trapped, the guest molecule cannot get
out, just rotate and move slightly (that is the reason it is called “cages”). There are five
common types of hydrates cages that can be observed in Figure 2.3, where the blue
spheres represents the oxygen molecules.
The ‘512’ representation is the pentagonal dodecahedron. This cage is formed by
12 faces with 5 edges each and it has 20 water molecules; the tetrakaidecahedron is
51262 (12 faces with 5 edges each and 2 faces with 6 edges each) and has 24 water
molecules; the hexakaidecahedron is 51264 and has 28 water molecules; the icosahedron
is 51268 and has 36 water molecules and the irregular dodecahedron is 435663 and has 20
water molecules.
As known, for any convex polyhedron, the number of Vertices plus the number of
Faces equals the number of Edges plus 2, as given by Euler’s Theorem of Geometry:
21
2V F E+ = + . (2.1)
Figure 2.3 – Hydrate cages (unscaled) - (a) 512, (b) 51262, (c) 51264, (d) 51268 and (e) 435663. Image
generated with help of VESTA 3.2.1 software
For instance, methane, ethane and cyclopropane form the so-called structure I (sI).
Hydrogen, nitrogen, oxygen, propane, isobutane, cyclopentane and tetrahydrofuran,
form structure II (sII) and methylcyclohexane, cycloheptane and cyclooctane combined
with a small molecule form structure H (sH). Therefore, the formed hydrates may have
three different structures: structure I (formed by 512 and 51262 cages), II (formed by 512
and 51264 cages) and H (formed by 512, 435663 and 51268 cages) (SABIL et al., 2009).
In the 1950s, von Stackelberg and co-workers summarized a lot of experimental
data about x-ray diffraction and the interpretation of this early experiments - von
Stackelberg (1949, 1954, 1956), von Stackelberg and Müller (1951a,b), Claussen
(1951a,b), and Pauling and Marsh (1952) - led to the determination of two structures of
hydrate crystal: sI and sII. Whereas the cubic structures (sI and sII) consist of two types
of cavities, small and large cages (SLOAN, 1998). At next decade, a number of
crystallographic studies were performed in those structures (MCMULLAN and
22
JEFFREY, 1965 and MAK and MCMULLAN, 1965). These studies showed hydrates to
be members of the class of compounds, labeled “clathrates” (POWELL, 1948), that
means “to encage”, and according to Sloan and Koh (2008), the clathrate hydrates are
ice-like inclusion compounds that are formed when small guests are enveloped by
hydrogen-bonded water cages in a crystal lattice.
The third hydrate structure (sH) was discovered only in 1987 (RIPMEESTER et
al., 1987). This structure requires both a small molecule (such as methane) and larger
molecules (such as condensate or an oil fraction). Structure H can only be formed in
laboratories, i.e. it does not appear in natural hydrate formations.
The three structures can be observed in Figure 2.4. The blue spheres represent the
oxygen molecule of H2O. Hydrogen molecules were hidden to enable a better
visualization of molecular organization on each structure.
Figure 2.4 – Hydrates structures - (a) sI, (b) sII and (c) sH. Image generated with help of VESTA 3.2.1
software
The first work focused on an industrial application of the hydrates was carried by
Hammerschmidt (1934). His paper presented a study on gas pipes clogs during winter
months, demonstrating that this problem was not caused by ice formation, as previously
thought, but rather by the hydrate formation.
More significant studies are useful in delineating details of the hydrate structure,
cage occupancies, formation/dissociation mechanisms, molecular dynamics and hydrate
composition, as can be seen in Sum et al. (1997), Sloan (2004) and Seo et al. (2002).
As mentioned before, hydrates and ice are similar. Table 2.1 allows the
observation of differences between ice, sI and sII hydrates properties.
23
Table 2.1 - Comparison of Ice, sI and sII hydrates properties Property Ice (iH) sI sII
Number of H2O molecules
4 46 136
Dielectric constant at 273 K
94 58 58
H2O reorientation time at 273 K (µs)
21 10 10
Linear thermal expansion at 200 K
(K-1)
56 x 10-6 77 x 10-6 52 x 10-6
Thermal conductivity (Wm-1 K-1) at 263 K
2.23 0.49 ± 0.02 0.51 ± 0.02
Density (g/cm3) 0.91 0.94 1.291 Adapted from Davidson (1983), Davidson et al. (1986) and Ripmeester et al. (1994)
There are associated phenomena with hydrates formation and an important one is
the hydrate nucleation, which is a microscopic process, involving tens of thousands of
molecules (MULLIN, 1993). It occurs when small clusters of water and gas grow to
reach the critical size triggering continuous growth. Kashchiev and Firoozabdi
(2002a,b) have good reviews on the nucleation process.
Another phenomenon that occurs is the called memory effect. Hydrates store a
memory of their structures when melted at moderate temperatures. Thus, the secondary
formations (when the mixture reaches the proper temperature and pressure) are
facilitated and occur quicker than when the phenomenon first occured. Once formed, if
the removal of water (which possess the structures of hydrates or residual dissolved gas)
is not complete, the hydrate will have a fast reconstruction. Consequently, this effect
has significant implications industry wide (SLOAN, 2008).
2.1.1 Growth process and plugs
Phase diagrams are commonly used to illustrate the hydrate formation curve.
These diagrams relate generally phases with temperature and pressure. Figure 2.5
illustrates a phase diagram for some natural gas components that form hydrates.
24
Figure 2.5 – Typical phase diagrams – pressure (log) versus temperature
According to Sloan and Koh (2008), in the phase diagrams, V is used to denote
vapor, I for ice, H for hydrates and LW and LHC for water and hydrocarbon liquid
phases, respectively. The hydrate region is to the left of the three phase lines (I-H-V)
and (LW-H-V). To the right, water occurs in the liquid state or ice and the guest
component as vapor or liquid.
There is a quadruple point Q1 (I-LW-H-V). The Q1 marks the transition from LW to
I, with decreasing temperature and is defined as the intersection of curve I-H-V with the
melting temperature of the water.
An alternative experiment for methane hydrate was made by Ouar, et al. (1992),
using an isochoric cell and changing the temperature. The cell was pressurized with
methane gas to the upper rightmost point (Figure 2.6). As the cell temperature was
lowered, the pressure decreases. A schematic diagram with the result for this experiment
is shown in Figure 2.6. Neither gas nor water is added to the system during the
experiment.
25
Figure 2.6 – Trace for formation of methane hydrates in a closed system
Ouar, et al. (1992)
The observation starts in A. The system is cooled to a temperature that causes
hydrate formation, represented by point B. The pressure drop which occurs between
points B and C is due to the formation and growth of hydrates, as the occlusion occurs
at this stage of the gas molecules in the crystal structures of hydrates. Dissociation
begins when the system is heated, until it is fully dissociated in D.
There are four basic processes of hydrate growth:
a) Growth of single crystal;
b) Growth of a hydrate film on the water – hydrocarbon interface;
c) Growth of multiple crystals in an agitated system and;
d) Growth of metastable phases.
In pipelines, for condensate gas flow, a representation on how hydrates plugs are
formed is shown in Figure 2.7. What happens in each of the five steps is detailed as
follows:
I- Hydrates start nucleation at inner surface of the pipelines and starts
deposition. It normally happens when water concentration is high in the
flow (>7 ppm), which results in a deposition scattered along the pipe;
II- After nucleation in inner walls, hydrates rapidly grow and occupy the
entire circumference of the pipe;
III- As the deposit grows, the useful diameter decreases;
26
IV- The great deposit grows until it is disturbed by some kind of phenomenon,
such as harmonic resonances, density difference, slugs, among others. The
deposit is not mechanically stable due to the disturbance, so it detaches
from the walls and flows through the duct as hydrate particles;
V- At this point, hydrate particles end up sintering and preventing the normal
flow of crude oil in pipelines.
Figure 2.7 – Hydrates formation in pipelines. Grey represents the fluid and white represents hydrates
Adapted from Sloan (2010)
To examine the hydrate nucleation, computational studies have used molecular
simulation (WALSH et al., 2009, MATSUMOTO, 2010 and SUM and YASUOKA,
2011). This kinds of simulations are also used to understand the growth process
(LIANG and KUSALIK, 2010, TUNG et al., 2010 and LIANG et al., 2011). These
simulations provides a better understanding about how hydrates are formed.
Another way to explain hydrates formation in pipelines is a model proposed by
Austvik (1992) (Figure 2.8). In a multiphase flow system containing water, oil and gas
(entrainment stage), hydrates can be formed at the interface of water droplets entrained
into oil phase and from gas bubbles into water phase (hydrate growth stage). These
particles start to agglomerate (agglomeration stage) and when a plug occurs, frequently
as little as 4 vol% of water is formed into hydrate (AUSTVIK, 1992). Agglomeration of
these hydrate particles with hydrate shells (ZERPA et al., 2011 and JOSHI et al., 2013)
finally can result in plugging conditions in the pipes (plugging stage).
27
Figure 2.8 – Conceptual model for hydrate formation in a multiphase flow system (gas, oil and water)
Zerpa (2011)
According to Chen et al. (2015), before hydrate formation, the oil phase was
present as droplets wrapped by a water film (caused by the flow conditions). This water
film gradually changed into a hydrate film network. With this hydrate formation, some
of the oil was squeezed out of the network and the oil-in-hydrate network system
shrank, precipitated from the oil phase, and deposited onto the pipe wall, finally
forming hydrate plug.
To a better understanding of hydrates plugs, research on the flow characteristics
and morphology of hydrates is needed. These studies, so far were usually performed
using only visual observation method (BELTRÁN and SERVIO, 2010 and CHEN et al.,
2013).
2.1.2 Inhibition and mitigation of hydrates
The most effective way to attenuate hydrate formation is to remove water from
whole system. Therefore, without water, there is no risk of hydrate formation. However,
this is not possible in most cases.
An alternative to maintenance of flow assurance in such conditions is the use of
inhibitors. Salts and alcohols are chemical components that, when solubilized, reduce
the amount of available water to form hydrates, hindering the formation. The study
performed by M. Karamoddin and F. Varaminian (2014) shows that salts with higher
electrical charge and less size are stronger inhibitors. In order to prevent hydrates in
pipelines is possible to use these thermodynamic inhibitors. Those elements change the
pressure and the temperature of formation, shifting out the operating condition of the
stable hydrate region. These inhibitors avoid hydrate formation by shifting hydrate
stability curve to higher pressure or lower temperature. Commonly used thermodynamic
28
inhibitors are methanol (MeOH) and monoethylene glycol (MEG). More details can be
found in literature (SLOAN, 1998).
As higher the water flow becomes, more effective inhibitors becomes necessary.
The low-dosage hydrate inhibitor (LDHI) is used in those situations. It is a chemical
that can effectively inhibit hydrate at low dose rates (ARJMANDI et al., 2005;
KELLAND, 2006 and LI, X. S. et al., 2006). The most common LDHI in industry are
kinetic hydrate inhibitor and anti-agglomerate (KELLAND et al., 1995; BALSON et al.,
2002; MEHTA et al., 2003; LACHANCE et al. 2009 and CHEN et al., 2010).
Knowing that temperature is a key factor of the hydrate formation, it is possible to
regulate the pipelines temperature and maintain it above the formation temperature. For
pipeline in deep water, temperature is normally very low (4 ºC). In order to minimize
thermal losses, a thermal insulation layer is added to the pipes. But no matter how much
insulation is added, after a long production shutdown, the fluid temperature will fall,
and eventually cool down to the seawater temperature. This way, insulation is not very
effective in mitigating the hydrate risk (GUO et al., 2014), only effective for fast
intervention operations.
Another strategy is the electric heating (LERVIK et al., 1997). This technique is
expensive, but can also be used as an intervention strategy. Once a hydrate plug is
formed, electric heating can be activated to melt the hydrate, allowing the plug to melt
faster than it would by using pipeline depressurization (with this last technique it could
take up to weeks or months to completely melt a long hydrate plug).
Another technique consists in depressurizing the pipeline to melt the hydrate plug,
but according to Sloan et al. (2010) this potentially produces a high differential pressure
across the plug. It can produce a high driving force that could result in launching the
plug at very high velocities, resulting in line ruptures, equipment damage and other
fatalities.
Briefly, Figure 2.9 illustrates the three most common techniques for inhibition and
mitigation of hydrates.
29
Figure 2.9 - Most common techniques for inhibition and mitigation of hydrates
The red point represents a position in the pressure versus temperature diagram
where hydrates may be formed. There are three most common techniques to inhibit or
mitigate hydrates. The system depressurization moves the operation point below the
hydrates formation curve; the thermodynamic changes the operation point out of the
hydrate domain and chemical inhibitors shift the hydrate stability curve.
2.1.3 Tetrahydrofuran
Most hydrates reviews are about gas hydrates, because it is more common in
nature. For the studies conducted in this work, the substance Tetrahydrofuran (THF)
was chosen to serve as a model fluid for generating hydrates under controlled
conditions. The advantages of using THF are twofold: (i) ease of handling and
formation of hydrates at atmospheric pressure conditions (i.e. no need to use a
pressurized cell), and (ii) THF presents high solubility in water (miscible at any molar
ratio), low freezing point and low cost. According to Liu et al. (2010), THF can replace
natural gas hydrate, due to formation of same structure of others guests molecules.
Tetrahydrofuran forms the structure II (ideal composition THF-17H2O) with THF
molecules occupying large cages and weakly interacting with the host water molecules
(PRASAD et al., 2007).
Several researchers performed hydrate experiments with THF-water mixture to
investigate the behavior of hydrate formation in liquid phase conditions
30
(DEVARAKONDA et al., 1999; BOLLAVARAM et al., 2000; IIDA et al., 2001;
NAGASHIMA et al., 2003; LIU et al., 2010 and WILSON et al., 2010).
The mixture of THF and water can form hydrates under atmospheric pressure and
temperatures below 4.4 ºC, considering a molar ratio of 1:17 (composition of THF to
water) (IIDA et al., 2001 and NAGASHIMA et al., 2003).
2.2 MONITORING TECHNIQUES OF SOLID DEPOSITS
Locating and quantifying solid formations is a very important step to prevent
unwanted stops in oil production. The detection of these deposits into flows can be
possible using some techniques, such as pressure pulses (GUDMUNDSSON et al.,
2001). This method detects the wax deposit, which is another problem of flow
assurance, by calculating the pressure variation, in time function, caused by pressures
pulses that travel through the tube.
Experiments in single phase flows for solid deposits detection were developed for
the cold finger method (DOS SANTOS et al., 2004) and radiograph technique
(EDALATI et al., 2006), both for wax monitoring. The first one consists in heating oil
samples and submitting them to different temperatures in contact with a cold surface.
The second one developed a method to measure wall thickness, corrosion inspection
and the degree of wax deposition. This way, is possible to evaluate the degree of loss in
the flow.
A non-intrusive instrumentation technique for wax deposition, which can give an
immediate inspection of the inner surface of the tube, is the ultrasonic technique
(ZAMAN et al., 2004). This technique is simpler in comparison to radiography, but
despite its high cost and radiography presents a better precision and image
reconstruction. Each industrial application has a more appropriate technique to detect
solid formations inside the pipelines (LONGO et al., 2014).
Pressure pulses can provide accurate information regarding wax deposits, such as
dimensions of the deposit along the tube. However, it presents limited range detection,
because it cannot monitor distances higher than 12 m (GUDMUNDSSON et al., 2001),
because pressure pulses cannot travel very long distances in some flow conditions.
Another cumber is the number of quick-acting valves needed on pipelines, which can
disrupt the flow.
31
More specifically to the detection of hydrates deposition, there are specific
techniques, such as ultrasonic detection (MAEDA et al., 2008), to describe the
formation process and the measurement of pressure and temperature (ZHOU et al., 2007
and SONG et al., 2008). Another technique that can be used for hydrates detection is the
gamma ray densitometry, by which some process information can be monitored, such as
flow regime and physical distribution of the hydrocarbon/hydrates production along the
pipe cross section (HJERTAKER, 2005).
Ultrasonic waves (PARK et al., 2013), MRI technique (LIU et al., 2010) and
electrical resistance measurement (ZHOU et al., 2007) have been used for hydrates
detection. A novel procedure was introduced by Karamoddin and Varaminian (2014),
showing that the electrical conductivity measurement of the solution can be used as a
powerful and instrumental method for investigating hydrate formation.
Capacitive sensors are fairly applied to void fraction detection in liquid-gas
mixtures and can be used to predict the concentration ratio of two-phase mixtures,
assuming that they have different electrical permittivity (CROWE, 2006). Earlier works
within our group at UTFPR have already applied the capacitive technique for
multiphase flow investigation (SANTOS, 2011; VENDRUSCOLO, 2012; WRASSE,
2015). The electrodes act as capacitance detectors, being proportional to the permittivity
changes of the mixture. Thus, the capacitance between the electrodes (emitter and
receiver) varies according to the ratio of the substances involved and their distribution
within the duct. This way, this capacitive technique may be an option for hydrates
detection.
2.3 IMPEDANCE SPECTROSCOPY
The general approach of impedance spectroscopy is to apply an electrical stimulus
(a known voltage or current) to the electrodes and observe the response (the resulting
current or voltage). The impedance spectroscopy aims at detaining the response of a
substance’s impedance at various frequencies.
The technique of impedance spectroscopy is a popular analytical tool in materials
research and development. It involves a relatively simple electrical measurement and its
results may often be correlated with many complex materials variables: from mass
transport, rates of chemical reactions, corrosion and dielectric properties, to defects,
microstructure, and compositional influences on the conductance of solids
32
(BARSOUKOV, 2005). However, according to Macdonald (1992), the impedance
spectroscopy measurements are simple in theory, and often complicated in practice,
since the resistive and capacitive components of this technique have large response
ranges. The frequency range may extend over 12 orders of magnitude or more (from
10 µHz to 10 MHz or higher).
2.3.1 Definitions
Impedance measurement is a commonly used method to characterize circuits,
components, materials and solutions (DA SILVA, 2008). The definition of impedance Z
is given by Ohm’s law in phasor notation, in which sinusoidal signals at angular
frequency ω are considered
( ) ( )( )ω
ωω
=V
ZI
(2.2)
where V denotes the electric voltage and I electric current. Inverse of impedance is
known as admittance Y.
Figure 2.10 illustrates a representation of impedance and admittance in the
complex plane and it is possible to demonstrate that
² ²Z R X= + (2.3)
and
² ²Y G B= + , (2.4)
where the real parts are the resistance R and the conductance G, they indicate the losses
within the circuit. The imaginary parts, which are termed reactance X and susceptance
B, respectively, are a measure of the reactive energy stored in the circuit during one
period and all these quantities are, in general, frequency dependent.
33
Figure 2.10 - Representation of impedance and admittance in the complex plane
By definition, impedance is a complex quantity and is only real when θ = 0 and
thus Z = R, that is, for purely resistive behavior.
In material science, the parameter complex permittivity is more commonly used
instead of impedance or admittance. In this work, the complex relative permittivity will
be referred to as complex permittivity, denoted by . As its name indicates, it is a
complex quantity
' "jε ε− , (2.5)
comprising a real part ε’ (permittivity) and imaginary part ε” (loss factor). According to
0ε=D , (2.6)
the complex permittivity relates the dielectric displacement D to the electric field
strength E, and the vacuum permittivity εo. Briefly, the real part of the complex
permittivity describes the ability of a material to support an electrical field. The losses
include dielectric loss due to relaxation or resonant effects in the materials as well as
loss by ionic conduction.
Each material has its own electrical permittivity, called absolute permittivity a.
However, a simplest way to represent it is usually by the named relative permittivity or
dielectric constant; all off the dielectric constant (or relative permittivity) is related to
the vacuum permittivity εo. The relative permittivity εr is given by:
0
rεεε
= (2.7)
In order to convert impedance or admittance values (normally the measured
quantity) into complex permittivity values (quantity of interest to characterize
34
substances) it is necessary to know the geometry factor kg of the measuring cell applied
to evaluate the substance. Hence,
g 0j kω ε=Y . (2.8)
For simple cell as exemplarily shown in Figure 2.16 (page 39), kg can be
approximate to
gAkd
= , (2.9)
where A the area of the square plates and d the distance separating them.
2.3.2 Typical permittivity values of common substances
Table 2.2 lists some substances and their permittivity values (for MHz range).
Table 2.2 - Relative permittivity for selected substances liquid at 25 ºC Substance Relative permittivity (εr)
Air 1.00
Crude oil 2.19
THF 7.58
2-propanol 20.00
Ethanol 24.00
Ethylene glycol 37.00
THF + water mixture* 73.10
Deionized Water 78.00
Adapted from Buckley and Maryott (1958), Csikós et al. (1982), Kumbharkhane et al. (1996), Chen et al. (1997) and Folgerø (1998).
In Table 2.2, the composition (80% deionized water and 20% THF) was chosen
from the table because is the closest to the used proportion in the hydrates experiments
(details are listed in section 4.4.1).
In open literature only a few works deal with the measurement and analysis of the
electrical properties of gas hydrates (according to Galashev (2006) εr ~ 2.5 for methane
hydrates). Haukalid et al. (2015) concluded that dielectric measurements have been
* Water 80% and THF 20%
35
demonstrated to be useful to study the impact of hydrate formation on water-in-crude oil
emulsion properties. Also working with emulsions, Jakobsen et al. (1996) studied the
kinetics of gas hydrate formation in emulsion models and Jakobsen and Folgerø (1997)
conclude that gas hydrate formation in water-in-oil emulsions can be monitored by
permittivity measurements with open-ended probes.
2.3.3 Dielectric relaxation
Impedance spectroscopy is the method of choice for characterizing the electrical
behavior of systems in which the overall system behavior is determined by a number of
strongly coupled processes, each proceeding at a different rate (BARSOUKOV and
MACDONALD, 2005). Many fundamental microscopic processes take place
throughout the substance when it receives an electrical stimulus leading to the overall
electrical response.
However, there are more complex concepts involving these microscopic
processes, as dielectric relaxation. With increasing frequency, the complex permittivity
of fluids, though not always, shows dielectric relaxation in which ε' decreases. A fluid
may present several dielectric mechanisms or polarization effects that contribute to its
permittivity as a whole.
According to Da Silva (2008), dipole orientation and ionic conduction interact
strongly at radio waves and microwave frequencies. Each dielectric mechanism has a
characteristic frequency. As frequency increases, the slow mechanisms drop out in turn,
leaving the faster ones to contribute to the permittivity. The nonrelaxation-type ionic
conduction losses are readily perceptible over the low-frequency spectrum (up to kHz
range) and decrease monotonically with frequency (Figure 2.11).
In Macdonald (1987), it was verified that a single relaxation process characterize
the dielectric mechanism behavior for many fluids from low frequency up to the
microwave frequency region.
36
Figure 2.11 - Schematic representation of different dielectric mechanism
Adapted from Agilent (2006a)
2.4 INSTRUMENTATION FOR IMPEDANCE MEASUREMENT
2.4.1 Auto-balancing bridge
The simplest way to measure impedance is the I-V method, in other words, the
conversion of the current into voltage. The measure can be enhanced using an opamp
(operational amplifier) with high input and low output resistance. The practical circuit
for measuring capacitive impedances (represented as parallel R-C network) is shown in
Figure 2.12, where Vi represents the input voltage and Cf and Rf are the feedback
network; Zx is the unknown impedance and Cs1 and Cs2 represent the stray capacitances
to ground, normally caused by cables used to connect the circuit with the sensor.
Figure 2.12 - Auto-balancing bridge circuit
Da Silva (2008)
To calculate the impedance, the voltage measurement at the opamp is used.
Assuming that the opamp is ideal, the output voltage V0 is determined by
37
0 f x
i x f
V Z Y= - =V Z Y
. (2.10)
The impedances Zx and Zf are formed by the parallel circuit of a capacitor and a
resistor. It can be assumed that stray capacitances have no influence in the circuit since
Cs1 is directly driven by the source voltage and Cs2 is virtually grounded by the opamp.
The frequency response of the auto-balancing bridge for typical expected
components can be seen in Figure 2.13.
Figure 2.13 - Frequency response of the auto-balancing bridge
Da Silva (2008)
The used values were Cf = 10 pF, Gf = 10 µS (100 kΩ), Cx and Gx (Rx) are
indicated in the figure.
2.4.2 Terminal configurations
Two types of sensors will be developed for this work. Both can be used in two-
terminal configuration or four-terminal configuration, depending on which measuring
system will be used.
According to Agilent (2006b), the two-terminal configuration is the simplest
method of connecting the sensors. Lead inductances (LL), lead resistances (RL), and
stray capacitance (Co) between two leads are added to the measurement result. Contact
resistances (R) between the test fixtures electrodes and the sensors are also added to
measured impedance. Without performing any compensation, the typical measurement
range is limited from 100 Ω to 10 kΩ. Figure 2.14 illustrate this method of connecting
38
such two-terminal to four-terminal measurement systems such as the commercial
instrument to be applied in this work.
Figure 2.14 - Two-terminal configuration
Agilent (2006b)
The terminals are marked as Hc (Hight current), Hp (High potential), Lp (Low
potential) and Lc (Low current). In this configuration, the terminals H and L are shorted.
Another possible configuration used is the four-terminal configuration, which
reduce the effects of lead impedances and contact resistances, because the signal current
path and voltage sensing leads are independent. Measurement errors due to the lead
impedances and contact resistances are thereby eliminated. This configuration is also
called Kelvin connection configuration and its accuracy is down to 10 mΩ (AGILENT,
2006b).
Figure 2.15 shows the four-terminal configuration.
Figure 2.15 - Four-terminal configuration
Agilent (2006b)
39
2.4.3 Measuring cells
Three sensor designs are employed in this work: a commercial and two developed
sensors. A short review is given below.
Parallel-plate sensors consist of two parallel plates with a dielectric between its
electrodes (emitter and receiver). When a voltage is applied in one of the plates, an
electric field is created.
Figure 2.16 illustrates a typical setup of a parallel-plate sensor. The plates can be
rectangular or other geometric shape, such as circular as in the case of the commercial
sensor from Agilent (will be explained in section 3.2.1).
Figure 2.16 - Typical setup of a parallel-plate sensor
Interdigital sensors are simple, low-cost and allow to investigate materials or
substances from measurements with a single-side surface contact. For instance, the
sensor can be mounted onto the wall of pipes and thus have a minimal influence on the
flow (DA SILVA, 2008).
A typical configuration of interdigital sensors can be seen in Figure 2.17. This
sensor has two electrodes, being one the transmitter (emitter) and the other the receiver.
40
Figure 2.17 - Typical configuration for interdigital sensors
The operating principle of an interdigital sensor follows the rule of a parallel plate
capacitor. The excitation electrode generates electric field lines, which penetrate into the
material under test (analyte) and interact with it. Figure 2.18 shows the simulation of the
electric field distribution of a conventional interdigital sensor.
Figure 2.18 - Electric field lines behavior in interdigital sensor
The spatial wavelength of an interdigital structure is defined as the distance
between the centers of “fingers” (MAMISHEV et al., 2004). The theoretical penetration
depth of the electric field generated by an interdigital sensor into analyte is proportional
to the spatial wavelength. According to Li, X. B. et al. (2006), in general, the
penetration depth can be defined as one third of λ (4 mm in this sensor). The penetration
depth basically describes the volume being sampled by the sensor or in another words
how far the sensor can measure.
The two-wire sensors can be applied in two-phase flows, as used by Waltrich et
al. (2013) to investigate the development of liquid-gas flows in a vertical flow loop,
where the conductivity of the phases are different and the environment is not an electric
insulator. This sensor has a low intrusion level and is normally used to mapping
concentration distributions. As the interdigital sensors, the two-wire sensors are excited
by an AC signal in the emitter electrode while the receiver electrode picks up the signal
response.
Figure 2.19 shows a schematic of two-wire sensor.
41
Figure 2.19 - Two-wire sensor schematic
This sensor can be mounted with the wires normal to the pipe wall. This type of
sensor was first used by Swanson (1966).
42
3 MEASURING SYSTEMS AND SENSOR DESIGN
3.1 OVERVIEW
In order to evaluate the electrical properties of substances and in special of
hydrates, three different measuring systems operating three different sensors were
applied. The diagram in Figure 3.1 represents how each measuring system may be
operated with each sensor.
Figure 3.1 - Block diagram of systems and sensors
The commercial system can operate all sensors. The mixed and dedicated systems
can only operate the interdigital and the two-wire sensors. Details are given below.
Figure 3.2 illustrates a simple diagram of the utilized experimental setup.
Figure 3.2 - Diagram of the apparatus
43
All measurement systems communicate with a computer for controlling and data
recording and a thermostatic bath was used to cool the sensors and control the
temperature of experiments.
The temperature control of the experiments was made by a thermostatic bath. The
chosen bath was a Julabo F12EH. This equipment is a PID controlled thermostatic bath
and can reach temperatures from -20 ºC to 150 ºC, depending on the utilized fluid. This
project used propylene glycol and water mixture and the working temperatures were
−10 ºC for ice formation experiments and 0.5 ºC for hydrates formation experiments.
3.2 MEASURING SYSTEMS
3.2.1 Commercial System
The first system is a precision impedance analyzer (Agilent 4294A), which is a
commercial system. With it is possible to measure the spectrum of the substances,
making a logarithmic frequency sweep, starting in 40 Hz and ending in 5 MHz. Higher
values would imply in operating problems on the analyzer. A USB/GPIB interface was
used to communicate the impedance analyzer with the computer, allowing data transfer.
A full scan (from 40 Hz to 5 MHz) lasts about 9 seconds and scans were repeated every
minute until the end of the analyzed phenomena (defined by the operator). After
performing the experiments, data were processed and analyzed.
A block diagram of this configuration is shown in Figure 3.3.
Figure 3.3 - Precision impedance analyzer block diagram
The system has its own capacitive sensor (Agilent 16452A), which was used as a
reference sensor for some measurements. It has two circumferential and parallel
44
electrodes with 3.7 mm of diameter. The sensor has some spacers to increase or
decrease its internal volume, and within this work, the lower spacer was used (1.3 mm),
leaving the sensor with 3.4 ml. The commercial sensor exploded diagram is shown in
Figure 3.4.
Figure 3.4 - Commercial sensor exploded view diagram - Agilent 16452A
Operation and service manual
3.2.2 Mixed System
In the same way as for the precision impedance analyzer, the second system
considered in this work, also is able generate a logarithmic frequency sweep using an
arbitrary signal generator (Agilent 81150A), however, the frequencies vary from 100 Hz
to 10 MHz and the system is able to measure only amplitude values of impedance. A
LabVIEW software developed by Leitzke (2014) was used and controls the system via
USB.
The system output is a voltage signal, since the used electronics comprises a
logarithmic detector and an auto-balance bridge, converting a current signal into a
voltage signal. This voltage signal passes through the DAQ card (data acquisition – NI
USB-6009) and goes to the software.
In this system, a full scan lasts about 18 seconds, and scans were performed every
1 minute too. A block diagram of the whole system is shown in Figure 3.5.
45
Figure 3.5 - Pulse function arbitrary generator block diagram
3.2.3 Dedicated System
Looking for a simple way to measure the hydrate deposition rate with an easy to
use and low-cost system, a technique based on the use of impedance probe was
developed, using a single selectable frequency to measure the signal amplitude. This
dedicated system is the last one used in this work and is responsible to generate and
measure the sinusoidal wave signals required to determine the impedance of fluids.
Figure 3.6 depicts the block diagram of the system. A direct digital synthesizer
integrated circuit (DDS – AD9835) generates the excitation sinusoidal wave voltage.
The alternating excitation voltage is supplied to the transmitter electrode by means of a
coaxial cable. A reference direct digital synthesizer was used as amplitude and phase
reference signal, which is fed into a subtractor amplifier (OPA830). Hence, it is possible
to detect small variations of impedance, as expected for hydrates formation. The
electrical currents flowing from the transmitter electrode are converted into voltage by
an auto-balancing bridge (OPA656). A linear power detector (LTC5507) was used to
convert the sinusoidal wave into DC proportional voltages. Signal conditioning
amplifiers are used to remove the DC offset introduced by the detector and scale it for
the full analog-to-digital conversion in the microcontroller (PIC18F2320).
The voltage values measured by the sensor are proportional to the electrical
current flowing from the transmitter electrode to the receiver electrode, as well as the
unknown permittivity of the material.
46
Briefly, a developed electronic circuit generates an excitation signal conducted to
the transmission electrode of the impedance sensor. This signal then has its amplitude
and phase changed, according to the electrical characteristics of the substance present
on the sensor. The response signal for the present fluid properties on the sensor is driven
by a measuring electronic circuit to the microcontroller circuit, managing and
controlling the entire system in a closed loop via a firmware code. The microcontroller
circuit communicates, via a serial protocol RS-232, to the software which enables the
graphical interface between the user and the system as a whole.
Figure 3.6 - Whole dedicated system block diagram
More details about the operation of the system can be found in APPENDIX B.
3.3 IMPEDANCE SENSORS DESIGN
3.3.1 Interdigital Sensor
The interdigital sensor consists of a double–side printed circuit board, where the
top layer contains two interdigital electrodes and the bottom layer contains a ground
plane, creating an electromagnetic shielding for the electrodes. Information about width
and spacing can be found in Figure 3.7. The recipient used to store the analyte is a
transparent type of PVC, with dimensions 54 x 54 x 54 mm and total volume of
approximately 157 ml.
47
Figure 3.7 - Sensor specification. Blue electrode is the receiver.
According to this configuration, λ = 12 mm (as mentioned in section 2.4.3),
because the distance between the center of two finger from the same electrode is 12
mm.
Using the commercial FEM simulation software COMSOL, a few geometry
simulations were run to find the best setting for the sensor, and the value found had a
slight difference from the one theoretically given. The central idea of FEM is to
discretize the domain, representing it, even though approximately, by a finite number of
elements; therefore, this method does not solve the original problem, but one that is
associated with it (GIACCHINI, 2012). In order to qualitatively evaluate the field
distribution of the sensor, a perturbation approach was employed (similar as described
by Da Silva, 2008). A perturbation is placed within the simulation volume and the
impact on the capacitance is analyzed. The simulated capacitance of a point Csim
between the transmitter electrode and receiver electrode is calculated by integrating the
induced surface charge density ρ over the electrode boundaries Γ
1 ( )simC dV
ρΓ
= Γ ⋅ Γ∫ , (3.1)
where the integral is implemented as post-processing option in the COMSOL software.
In order to find a most appropriate mesh for the simulations, balancing simulation
time and accuracy of results, the mesh was manually refined. The mesh refinement was
done only at receiver electrode, which is the site of interest.
48
Figure 3.8 allows the visualization of the differences between the pre-defined and
the refined meshes, respectively represented. The number of elements was increased
from 433643 to 976563.
Figure 3.8 - Meshes compared: (a) pre-defined and (b) manually refined
The mesh utilized to perform the simulations was the pre-defined. It has a
maximum element size of 3.5 mm. In order to obtain more accurate results, the mesh
was manually refined to a maximum element size of 0.2 mm.
The simulations were performed considering firstly the recipient empty (with air)
and then adding water films from 0 to 6 ml in increments of 0.08 mm.
The comparison between the simple mesh simulation (pre-defined by software)
and the manually refined mesh is represented by the Figure 3.9. Nevertheless, the
difference is negligible to the point of interest (Y = 0.97).
Figure 3.9 - Comparison between manually refined and pre-defined meshes
49
Knowing that difference in the value of interest can be disregarded for this work,
the pre-defined mesh may be used.
This way, several simulations were performed in order to define the better
geometry to the sensor (based on higher penetration depth). In Figure 3.10 it is possible
to observe the results of 10 simulations (with normalized values of capacitance versus
height of water in the recipient) and conclude that the best configuration is the one with
three fingers, 2.0 mm of spacing between two adjacent electrodes and 4.0 mm of
thickness for each electrode. The red dot on this curve (illustrated near the superior right
corner) represents the coordinate (X,Y) where value of X equals 5.2 and Y equal 0.97,
which means that the penetration depth is slightly beyond 5 mm.
Figure 3.10 - Simulations to find the best sensor configurations: Normalized output X Penetration depth (mm)
The peaks at the beginning of the curves are due to mesh refinement used in the
simulations (pre-defined).
Table 3.1 lists the absolute values of capacitance for all performed simulations. It
is possible to see the values with the sensor empty (with air) and full of water.
50
Table 3.1 - Absolute values of simulated capacitances Configuration Empty - air Full - water (Number of
Fingers/Spacing(mm)/Width(mm) (pF) (pF)
3/0.5/0.5 1.4471 6.0330 3/0.5/1.5 2.3829 11.1756 3/1.0/1.5 1.8209 8.2291 3/2.0/2.0 1.4197 6.9074 3/2.0/3.0 1.8281 9.1311 3/3.0/3.0 1.7705 7.8448 3/2.0/4.0 2.0743 10.7897 5/0.5/0.5 2.1214 10.0232 5/0.5/1.5 4.0491 18.8382 5/1.0/1.0 1.9346 9.7833
Figure 3.11 provides a better understand of the electric potential behavior within
the sensor, detailing two different views of the same electric potential. The color red
indicates the location of stronger electric field and blue represents the weakest.
Figure 3.11 – Electric field behavior in interdigital sensor. hL(6)=4e-4 hL(76)=0.006; Multislice: Electric
potential (V)
The input and output sensor connections were manufactured using coaxial cables
with the purpose of immunize the electromagnetic interference from the electromagnetic
environment in which the sensor is inserted.
The electronics is responsible for applying a sinusoidal voltage to the transmitter
electrode. This excitation causes the electric potential to concentrated along the active
51
electrode, allowing the measurement of the displacement current flowing to the receiver
electrode.
Figure 3.12 brings a picture of the interdigital sensor. The ground plane created on
the top layer is disposed on the printed circuit board to induce the convergence of the
electric field in the region between the excitation and the reception electrodes.
Furthermore, the ground plane on the bottom layer creates an electromagnetic shielding,
making the circuit less susceptible to electromagnetic interference from external sources
and ensuring a more stable measurement.
Figure 3.12 - Excitation and reception electrodes and ground plane of interdigital sensor
This sensor has four terminals and can operate in two-terminals or four-terminals
configuration, as explained in section 2.4.2.
3.3.2 Two-Wire Sensor
The operating principle of the two-wire sensor is very similar to the interdigital
sensor. There are two electrodes, an emitter and a receiver, but now each electrode is a
0.15 mm thickness stainless steel wire.
Again, the sensor consists of a double–side printed circuit board, where the top
layer contains the four electric trails of the electrodes and the bottom layer contains a
ground plane, creating an electromagnetic shielding for the electrodes.
Similar to interdigital sensor, the two-wire sensor can operate in two-terminal
configuration or four-terminal configuration.
52
The sensor geometry was based on earlier studies (PARRA, 2013) and therefore
no geometry optimization was performed in this work. In order to illustrate the electric
field behavior within the sensor, a COMSOL simulation was carried out (Figure 3.13).
It can be seen that the sensor interrogating region is concentrated in between the wires
but spreads (with less intensity tough) over the rectangular recipient.
Figure 3.13 - Electric field behavior in two-wire sensor. hL(51)=0.05; Multislice: Electric potential (V)
The emitter wire is parallel to the receiver wire. The distance between them is 2.5
mm and both wires are 26 mm long. Figure 3.14a shows a schematic of the two-wire
sensor and Figure 3.14b shows the PCB (printed-circuit board).
53
Figure 3.14 - Two-wire sensor - (a) Schematic (b) PCB picture
The sensor was mounted in a recipient similar to the recipient used for the
interdigital sensor. Here it was 54 x 54 x 80 mm, with a volume of approximately 233
ml. The PCB was placed in the recipient shown in Figure 3.15.
Figure 3.15 - Two-wire sensor with the recipient
54
4 TESTS AND RESULTS
4.1 DEDICATED SYSTEM PERFORMANCE
4.1.1 Linearity
The concept of linearity is connected to the constant of proportionality between
the input and the output values of a system. To check the linearity of the developed
system, ten commercial capacitors were selected to simulate the impedance sensor. The
range chosen represent expected impedance values in fluid measurement. The real
values of capacitance were previously measured by a precision RLC meter (Agilent
E4980A). For the linearity test, 10 repeated experiments were performed in order to
take an average of the results. Capacitors were connected directly to the developed
system by coaxial cables.
The developed system can measure impedance variations, being necessary to
zero the output to a given reference. For this reason, the commercial capacitors were
divided into two groups. First group was from 1 pF to 10 pF and the second one from
15 pF to 47 pF. In both groups the lowest capacitance was used to zero the system.
The relationship between the capacitance and the circuit response to the voltage
value obtained from the linearity test is shown in Figure 4.1. The system response
shows linear behavior for both groups.
Figure 4.1 - Linearity test results
55
4.1.2 Water-Isopropanol Mixture
In this test, the interdigital sensor was connected to the dedicated system. In
sensor’s recipient a mixture of isopropanol and water at defined concentrations was
poured and investigated. After the system calibration with isopropanol, analysis of the
mixture of isopropanol and water was performed.
Measurements were made by changing the water concentration from 0 to 100%.
Ten repeated tests were carried out, recording the response of the measuring circuit.
Average values as well as the standard deviation (error bar) are shown in Figure 4.2. As
expected (see Da Silva 2008) the permittivity of isopropanol-water mixture presents a
linear behavior with concentration change.
Figure 4.2 – Measured system output versus water concentration
4.2 PERMITTIVITY MEASUREMENT
To confirm the theoretical permittivity values, a preliminary investigation using
the commercial system and the three sensors in the measurements of electrical
permittivity was carried out and the results are shown below.
Four substances (water, glycol, 2-propanol and air) were investigated and its
permittivity was calculated for the commercial sensor and the water permittivity value
(ε ~ 75.72 in this case, because the experiments were conduct at temperatures of ~
30ºC) was used to fit the other permittivity values for the interdigital and the two-wire
sensors (Table 4.1). The good agreement between the expected linear dependence and
the obtained regression lines is confirmed by the good coefficients of about 0.99. This
way, the commercial system can be used as a reference system for the investigation of
56
the electrical properties of fluids. Figure 4.3 shows the spectra of capacitance for
measured substances.
Figure 4.3 - Capacitance of selected substances
Due to stray capacitances at low frequency, the plateau of highest frequency is
indicated to measurements; therefore, the electrical permittivity of substances was
measured for 5 MHz, as shown in Table 4.1.
Table 4.1 – Reference permittivity (from Table 2.2) and permittivity of tested substances at high-
frequency plateau (5 MHz) Substance εref εmes(interdigital) Deviation
(%) εmes(two-wire) Deviation
(%) εmes(commercial) Deviation
(%)
Air 1 1 0 1 0 1 0
2-propanol 20.1 18.20 9.45 21.80 3.48 19.49 3.03
Glycol 41.1 38.56 6.18 42.93 4.45 40.62 1.16
Deionized water
78.0 75.72 2.92 75.72 2.92 75.72 2.92
4.3 ICE FORMATION
Knowing the similarity of the ice with hydrates, experiments freezing deionized
water were performed. First, the commercial sensor (presented in section 3.2.1) was
57
used in the impedance analyzer. Then the commercial sensor was replaced by
interdigital and two-wire sensors, which were used in the three available systems.
The commercial sensor was calibrated, as the indications of operating manual, and
was filled with deionized water and placed in the thermostatic bath freezing the water.
This cooling process was monitored until all volume was fully frozen.
Each experiment was performed at least 3 times, and the results are summarized
as follow.
Figure 4.4 shows the spectrum of liquid water (when the experiments started) and
ice (when the experiments ended) for capacitance from 40 Hz to 5 MHz. The impedance
analyzer was used with the three sensors.
Figure 4.4 - Comparison of measured capacitance during ice formation from all sensors
It is possible to see that the start spectrum of the experiments is different of the
end spectrum and the capacitance decreases while the water freezes. Between 2 kHz and
60 kHz, the dielectric relaxation (already explained in section 2.3.3) is observed.
According to Fletcher (1970), the frequency at which Cp starts to decrease can be
associated with the inverse relaxation time for reorientation of molecular dipoles and
depends strongly on temperature and specimen purity.
The two-wire sensor is more susceptible to noise, so at low frequencies it presents
noisy behavior.
58
The resistance spectrum of the same experiment can be observed in Figure 4.5.
Figure 4.5 - Comparison of measured resistance during ice formation from all sensors
For resistance, the relaxation is visible at frequency between 200 Hz and 5 kHz
and contrary to the capacitance values, resistance increases when the water is frozen.
The results of ice formation measured with the mixed system are shown in Figure
4.6. A frequency range of 100 Hz to 10 MHz was used.
Figure 4.6 - Signal obtained with mixed system –connected to interdigital sensor during ice formation
59
All experiments in both capacitance and resistance and output voltage
(proportional to amplitude of admittance), allow the observation of a significant change
in the spectrum when water freezes and passes from the liquid to the solid state.
Finally, to complete this part of the experiments, dedicated electronics were used
to make the same measurement performed in previous experiments, after appropriate
calibration, and using deionized water as zero. The frequency used was 5 MHz and the
result is shown in section 4.5, when the ice formation is compared to hydrate formation.
4.4 HYDRATE FORMATION
To validate the technique, experiments were performed using THF for hydrate
formation in the measuring cell. As explained in the section 2.1, the molecular structure
formed by THF (sII) is similar to some other common gases.
4.4.1 Methodology
The samples were prepared using a precision balance. A mixture of deionized
water and THF was prepared, and the concentration, by weight, was 81% water and
19% THF. This is the exact balance in hydrate formation so that there is no free water or
THF.
In order to successfully perform and monitor the formation of hydrates inside the
sensor, some procedures were developed. An auxiliary recipient was used in the
process.
• The mixture of THF with deionized water in suitable proportions was cooled to
0.5 ºC;
• The sensor was cooled to 0.5 ºC;
• Agitation of the recipient containing the mixture was necessary in the
experimental conditions, with the objective of initiating the formation of the first
hydrate crystals;
• After complete solidification of the mixture, the auxiliary recipient was removed
from the bath in order to initiate the dissociation;
• When half of the mixture was dissociated, the liquid was poured into the sensor,
which was sealed (since THF is volatile), and the monitoring of the hydrates
formation was initiated. The first training was carried out of the sensor because
60
hydrate formation have the memory effect, which facilitates the subsequent
formations;
• Sometimes, in the interval between acquisitions, the sensor was stirred to induce
formation;
• The other half of the mixture was maintained in the recipient, but out of the bath,
in order to completely dissociate, to be used in the second part of the experiment:
measurements without formation of hydrates;
• After the complete formation within the sensor, the acquisition was finished and
the sensor was cleaned for the second part of the experiment;
• The sensor and the recipient with the mixture were cooled again to 0.5 ºC;
• The mixture was poured into the sensor, which was then sealed and the second
part of the experiment was initiated. At this stage there was no agitation to avoid
hydrates formation;
• The mixture was monitored for the same period of time considered in the first
stage.
4.4.2 Results
The experiments followed the procedures described above and were performed in
pairs (with and without hydrates formation). Firstly, the commercial sensor was used to
perform measurements with and without hydrates formation, then the interdigital sensor
was used. The separated results are presented in APPENDIX A.
It was observed that the spectra of interdigital sensor follow the same trend of the
commercial sensor spectrum, validating the fabricated sensor for this work and enabling
its use in future developments.
Finally, to confirm the values found in the literature, the commercial sensor was
used and the permittivity of THF, THF mixture and water were obtained using the
commercial system with commercial sensor. The calculated values can be observed in
Figure 4.7.
61
Figure 4.7 – Permittivity of THF, THF + water mixture and water
The measured values for the plateau (frequencies over 40 kHz) are very close to
values found in the literature. Table 4.2 summarized the results.
Table 4.2 - Permittivity of THF, THF + water mixture and water
THF THF + water Water
εref 7.58 73.10 78.00
ε 7.95 66.35 75.72
Deviation (%) 0.13 9.23 2.92
The error observed to THF + water mixture due to lack of data in the literature.
There was not found value that utilizes the same proportion of THF and water that was
used in this work, and again the temperature conditions of the experiments were
different from those of the reference measures.
4.5 COMPARISON OF ICE AND HYDRATES FORMATION
As mentioned in chapter 2, hydrates and ice present some similarities in their
molecular structure; however, as can be seen in Figure 4.8 and Figure 4.9, it is possible
62
to differentiate these two substances with spectroscopic method (capacitance and
resistance).
The Figure 4.8 and Figure 4.9 compare three experiments: ice formation, THF and
water mixture with and without hydrate formation (capacitance and resistance measured
with commercial system and interdigital sensor).
Figure 4.8 - Ice/Hydrates capacitance comparison - Interdigital sensor – Commercial system
Figure 4.9 - Ice/Hydrates resistance comparison - Interdigital sensor – Commercial system
63
With these results, it can be concluded that the impedance spectroscopy technique
for the deposition of hydrates analysis is promising, as the spectrum of the mixture is
different in liquid and solid phases due to changes of electrical characteristics.
It can be seen that the initial spectra of the three experiments are similar.
However, the final values are different when ice or hydrate formation is completed,
allowing the identification of whether there has been formation on the sensor and which
substance was formed (ice or hydrate) by inspecting the shape of spectra. It is also
noticed that the hydrates experiment without formation has overlapping lines (start and
end).
Using the mixed system, with the interdigital sensor, it is also possible to
distinguish the spectrum of the ice and hydrate, as shown in Figure 4.10.
Figure 4.10 - Ice/Hydrates comparison - Interdigital sensor - Mixed system
It is possible to reach the same conclusion, that the ice deposition can be
differentiated from hydrate observing the graph of the experiments performed with the
dedicated system and interdigital sensor, which can realize a difference between these
substances, as shown in Figure 4.11.
64
Figure 4.11 - Ice/Hydrates comparison - Interdigital sensor - Dedicated system (5 MHz)
Due to the penetration depth, it is possible to observe a sharp increase in the value
of the output voltage for ice formation, indicating that ice is formed on the surface of
the sensor.
A similar tendency can be seen in Figure 4.12, whereas hydrates formation is a
slower process than ice formation. This experiment used the dedicated system with the
two-wire sensor, but as the available volume of recipient of this sensor is considerably
larger than the one of the interdigital sensor, and the sensor unit is at the center and not
on the edge, more time was required to complete the experiments.
65
Figure 4.12 - Ice/Hydrates comparison - Two-wire sensor - Dedicated system (5 MHz)
The ice formation process occurs faster than the hydrate formation, since ice
formation needs only lower temperature, and the hydrate has other known factors which
influence the formation.
Figure 4.13 shows two pictures that illustrates the differences between hydrates
and ice in the interdigital sensor.
Figure 4.13 - Interdigital sensor with (a) Hydrates and (b) Ice
Beyond the differences that can be measured, it is possible to see the difference
between these solid substances. Hydrate is more opaque and the ice is transparent.
66
5 CONCLUSIONS
In oil and gas production, the occurrence of gas hydrates is often related to
financial losses due to production decrease or stoppage. Currently there is no
established measuring technique to online monitor hydrate formation or deposition in
pipelines which could help in the process of decision making or could deliver
information on system behavior.
In this work, the behavior of electric impedance (and its spectrum) during
hydrates formation was investigated. Impedance measurements were chosen due to the
fact of being simple, low-cost and non-intrusive nature (or minimal intrusive, in the case
of the two-wire sensor). Three different measuring systems were used to compare and
validate the obtained data. The first two were based on measurement of impedance
spectra (impedance over frequency from 40 Hz to 10 MHz). A third measuring system
was developed and tested in this work which operates at a single frequency and may be
later applied in systematic studies of hydrates formation. The chosen technique for
measuring impedance in fluids with the dedicated system was the auto-balancing
bridge, in which the amplitude of the output signal of the fluid is proportional to the
analyzed impedance. The developed circuit is able to measure variations in the
capacitance in pF range. Thus, this circuit is capable of monitoring the impedance
change between a reference substance and a test substance. Characterization tests of the
system were performed to validate the functionality proposed in the project, showing
appropriate results.
Three measuring cells were applied in conjunction with the systems in order to
investigate the impedance of fluids and deposits, being one commercial cell and two
developed ones based on interdigital and two parallel wires structures.
Since ice presents many similarities to hydrates, as initial tests in order to verify
the developed sensors, water freezing process was monitored with all measuring
systems and sensors, showing appropriate sensibility to detect ice formation.
Subsequently, hydrates formation from water-THF solution was monitored. For this
purpose a procedure was proposed in order to generate hydrates at controlled
conditions. The impedance spectra and measurements obtained show that there is a clear
change in the values for liquids and their solids formations.
67
The main contribution of this work is a systematic study of electrical impedance
of hydrates and its formation. In this way, the developed sensors and measuring system
may be further developed and enhanced, enabling the online monitoring of hydrates
formation and/or deposition in pipelines. Hence, the present work may be regarded as
the first step in the development of a tool for hydrate monitoring.
Future work shall involve obtaining impedance spectra of hydrates from other
hydrocarbons such as methane, which however requires the construction of special
pressurized cells with electrical access to its interior. Further direction of research is the
detailed analysis of impedance (complex permittivity) spectra including the description
of relaxation process and derived parameters.
68
REFERENCES
AGILENT. Basics of Measuring the Dielectric Properties of Materials. Application Note, Agilent Technologies, 2006a.
AGILENT. The Impedance Measurement Handbook. Application Note, Agilent Technologies, 2006b.
ALEXANDER, C. K; SADIKU, M. N. O. Fundamentos de Circuitos Elétricos. Trad: Gustavo Guimarães Parma. – Porto Alegre: Bookman, 2003.
ARJMANDI, M.; TOHIDI, B.; DANESH, A.; TODD, A. C. Is subcooling the right driving force for testing low-dosage hydrate inhibitors? In: Chemical Engioneering Science, v. 60, p. 1313–1321, 2005.
AUSTVIK, T. Hydrate formation and behavior in pipes. 1992. Ph.D. Thesis - Norwegian University of Science and Technology, Trondheim, 1992.
BALSON, T.; CRADDOCK, H. A.; DUNLOP, J.; FRAMPTON, G.; PAYNE, G.; REID, P. The development of advanced kinetic hydrate inhibitors. In: Royal Society of Chemistry: Chemistry in the Oil Induntry. VII. RSC Publishing, Cambridge, UK, 2002.
BARNES, B. C.; BECKHAM, G. T.; WU, T. D.; SUM, A. K. Two-Component Order Parameter for Quantifying Clathrate Hydrate Nucleation and Growth. In: Journal of Chemical Physics, v. 140 (16), p. 164506-6, 2014.
BARSOUKOV, E.; MACDONALD, J. R. Impedance Spectroscopy: Theory, Experiment, and Applications, Eds., Wiley-Interscience, 2005.
BELTRÁN, J. G.; SERVIO, P. Morphological investigations of methane hydrate films formed on a glass surface. In: Crystal Growth and Design, v. 10, p. 4339–4347, 2010.
BOLLAVARAM, P.; DEVARAKONDA, S.; SELIM, M. S.; SLOAN, E. D. Growth Kinetics of Single Crystal sII Hydrates: Elimination of Mass and Heat Transfer Effects. In: Annals of the New York Academy of Sciences, v. 912, p. 533, 2000.
BOMBA, J. G. Offshore Pipeline Transport of Waxy Crude Oils, Paper SPE14622, 1986.
BUCKLEY, F. and MARYOTT, A. A. Tables of Dielectric Dispersion Data for Pure Liquids and Dilute Solutions. U.S. Dept. of Commerce, National Bureau of Standards, USA. Circular 589, 1958.
CHEN, J.; SUN, C. Y.; PENG, B. Z.; LIU, B.; SI, S.; JIA, M. L.; MU, L.; YAN, K. L.; CHEN, G. J. Screening and compounding of gas hydrate anti-agglomerants from commercial additives through morphology observation. In: Energy Fuels, v. 27, p. 2488–2496, 2013.
69
CHEN, J.; YAN, K. L.; CHEN, G. J.; SUN, C. Y.; LIU, B.; REN, N.; SHEN, D. J.; NIU, M.; LV, Y. N; LI, N; SUM, A. K. Insights into the Formation Mechanism of Hydrate Plugging in Pipelines. In: Chemical Engineering Science, v. 122, p. 284-290, 2015.
CHEN, L. T.; SUN, C. Y.; CHEN, G. J.; ZUO, J. Y.; NG, H. J. Assessment of hydrate kinetic inhibitors with visual observations. In: Fluid Phase Equilibria, v. 298, p. 143–149, 2010.
CHEN, Z; HOJO, M. Relationship between triple ion formation constants and the salt concentration of the minimum in the conductometric curves in low-permittivity solvents. In: The Journal of Physical Chemistry B, v. 101, n. 50, p. 10896-10902, 1997.
CLAUSSEN, W. F. A Second Water Structure for Inert Gas Hydrates. In: Journal of Chemical Physics, v. 19, p. 1425, 1951b.
CLAUSSEN, W. F. Suggested Structures of Water in Inert Gas Hydrates. In: Journal of Chemical Physics, v. 19, p. 662, 1951a.
CROWE, C. T. Multiphase Flow Handbook. Boca Raton, FL: Taylor & Francis, 2006.
CSIKÓS, R.; FREUND, M.; KESZTHELYI, S.; MÓZES, G. Y. Paraffin products, properties, technologies and applications. Budapeste, Hungria: Elsevier, 1982.
DA SILVA, M. J. Impedance Sensors for Fast Multiphase Flow Measurement and Imaging. 2008. 170p. Tese (Doutorado) - Dresden: Fakultät Elektrotechnik und Informationstechnik, Technische Universität Dresden, 2008.
DAVIDSON, D. W. Natural Gas Hydrates (Cox, J.L., ed.) Butterworths, Boston, 1, 1983.
DAVIDSON, D. W.; HANDA, Y. P.; RIPMEESTER, J. A. Xenon-129 NMR and the thermodynamic parameters of xenon hydrate. In: Journal of Physical Chemistry, v. 90, p. 6549-6552, 1986.
DEVARAKONDA, S.; GROYSMAN, A.; MYERSON, A. S. THF–water hydrate crystallization: an experimental investigation. In: Journal of Crystal Growth, v. 204, p. 525-538, 1999.
DOS SANTOS, Eduardo Nunes. Técnicas para extração de parâmetros de escoamentos bifásicos em regime intermitente utilizando o sensor wire-mesh. 2011. Dissertação - UTFPR, Curitiba, 2011.
DOS SANTOS, J. DA S. T.; FERNANDES, A. C.; GIULIETTI, M. Study of paraffin deposit formation using the cold finger methodology for Brazilian crude oils. In: Journal of Petroleum Science Engineering, 2004.
70
EDALATI, K.; RASTKHAH, N.; KARMANI, A.; SEIEDI, M.; MOVAFEGHI, A. The use of radiography for thickness measurement and corrosion monitoring in pipes. In: International Journal Press. Vessels Pip., 2006.
ENGLEZOS, P. Clathrate hydrates. In: Industrial and Engineering Chemistry Research, v. 32, p. 1251–1274, 1993.
FLETCHER, N. H. The Chemical Physics of Ice. Cambridge University Press, USA, New York, 1970.
FOLGERØ, K. Broad-band dielectric spectroscopy of low-permittivity fluids using one measurement cell. In: IEEE Transactions on Instrumentation and Measurement, v. 47, p. 881-885, 1998.
GALASHEV, A. E.; CHUKANOV, V. N.; NOVRUZOV, A. N.; NOVRUZOVA, O. A. Molecular-dynamic calculation of spectral characteristics of absorption of infrared radiation by (H2O)(j) and (CH4)(i)(H2O)(n) clusters. In: High Temperature, v. 44, p. 364–372, 2006.
GAMIO, J. C.; YANG W. Q.; STOTT A. L. Analysis of non-ideal characteristics of an ac-based capacitance transducer for tomography. In: Measurement Science and Technology, v. 12, p. 1076-1082, 2001.
GIACCHINI, B. L. Uma breve introdução ao Método dos Elementos Finitos. Departamento de Matemática, Instituto de Ciências Exatas, Universidade Federal de Minas Gerais, 2012.
GUDMUNDSSON, J. S.; DURGUT, I.; CELIUS, H. K.; KORSAN, K. Detection and monitoring of deposits in multiphase flow pipelines using pressure pulse technology. In: 12th International Oil Field Chemistry Symposium, Geilo, Norway, 2001.
GUO, B.; SONG, S.; GHALAMBOR, A.; LIN, T. R. Offshore Pipelines. Design, Installation, and Maintence. 2nd Edition, GPP Elsevier, MA, 2014.
HAMMERSCHMIDT, E.G. Formation of gas hydrates in natural gas transmission lines. In: Industrial and Engineering Chemistry, v. 26, p. 851–855, 1934.
HAMOUDA, A. A.; DAVIDSEN, S. An Approach for Simulation of Paraffin Deposition in Pipelines as a Function of Flow Characteristics With a Reference to Teesside Oil Pipeline. In Proceedings of SPE International Symposium on Oilfield Chemistry. Society of Petroleum Engineers, p. 14-17, 1995.
HAUKALID, K.; ASKVIK, K. M.; FOLGERØ, K.; Thomas, P. J. Impact of Gas Hydrate Formation/Dissociation on Water-in-Crude Oil Emulsion Properties Studied by Dielectric Measurements. In: Energy & Fuels, v. 29 (1), p. 43-51, 2015.
HENLEY, H.; EDWARD THOMAS, E.; ANGELO LUCIA, A. Density and phase equilibrium for ice and structure I hydrates using the Gibbs–Helmholtz constrained equation of state. In: Chemical Engineering Research and Design, v. 92, p. 2977-2991, 2014.
71
HJERTAKER, B. T.; TJUGUM, S. A.; HAMMER, E. A.; JOHANSEN, G. A. Multimodality tomography for multiphase hydrocarbon flow measurements. In: IEEE Sensors Journal, v. 5, p. 153-160, 2005.
IIDA, T.; MORI, H.; MOCHIZUKI, T.; MORI, Y. Formation and dissociation of clathrate hydrate in stoichiometric tetrahydrofuran–water mixture subjected to one-dimensional cooling or heating. In: Chemical Engineering Science, v. 56, p. 4747-4758, 2001.
JACCARD, C. Mechanism of the electrical conductivity in ice. In: Annals of the New York Academy of Sciences, v. 125, n. 2, p. 390-400, 1965.
JAKOBSEN, T.; FOLGERØ, K. Dielectric measurements of gas hydrate formation in water-in-oil emulsions using open-ended coaxial probes. In: Measurement Science and Technology, v. 112, p. 73–84, 1996.
JAKOBSEN, T.; SJÖBLOM, J.; RUOFF, P. Kinetics of gas hydrate formation in emulsion. The model system trichlorofluoromethane/water/non-ionic surfactant studied by means of dielectric spectroscopy. In: Colloids and Surfaces, v. 112, p. 73–84, 1996.
JOSHI, S. V.; GRASSO, G. A.; LAFOND, P. G.; RAO, I.; WEBB, E.; ZERPA, L. E.; SLOAN, E. D.; KOH, C. A.; SUM, A. K.; Experimental flowloop investigations of gas hydrate formation in high water cut systems. In: Chemical Engineering Science, v. 8, p. 1006-1015, 1997.
KARAMODDIN, M.; VARAMINIAN, F. Performance of hydrate inhibitors in tetrahydrofuran hydrate formation by using measurement of electrical conductivity. In: Journal of Industrial and Engineering Chemistry, v. 20, p. 3815–3820, 2014.
KASHCHIEV, D.; FIROOZABADI, A. Driving force for crystallization of gas hydrates. In: Journal of Crystal Growth, v. 241, 220-230, 2002a.
KASHCHIEV, D.; FIROOZABADI, A. Nucleation of gas hydrates. In: Journal of Crystal Growth, v. 243, p. 476-489, 2002b.
KELLAND, M. A. History of the development of low dosage hydrate inhibitors. In: Energy Fuels, v. 20, p. 825–847, 2006.
KELLAND, M. A.; SVARTAAS, T. M.; DYBVIK, L. A. Studies on new gas hydrate inhibitors. In: Presented at Offshore Europe. Aberdeen, United Kingdom, SPE Paper 30420, p. 5-8, 1995.
KOLLMAN, P. A general analysis of noncovalent intermolecular interactions. In: Journal of the American Chemical Society, v. 99, p. 4875-4894, 1977.
KUMBHARKHANE, A. C.; HELAMBE, S. N.; LOKHANDE, M. P.; DORAISWAMY, S.; MEHROTRA, S. C. Structural study of aqueous solutions of tetrahydrofuran and acetone mixtures using dielectric relaxation technique. In: PRAMANA Journal of Physics, v. 46, n. 2, 1996.
72
KUMBHARKHANE, A. C.; PURANIK, S. M.; MEHROTRA, S. C. Dielectric relaxation studies of aqueous N, N-dimethylformamide using a picosecond time domain technique. In: Journal of Solution Chemistry, v. 22, p. 219, 1993.
LACHANCE, J. W.; SLOAN, E. D.; KOH, C. A. Determining gas hydrate kinetic inhibitor effectiveness using emulsions. In: Chemical Engineering Science, v. 64, p. 180–184, 2009.
LATIMER,W.M.; RODEBUSH,W.H. Polarity and ionization from the standpoint of the theory of valence. In: Journal of the American Chemical Society, v. 42, p. 1419-1433, 1920.
LEITZKE, JULIANA P. Desenvolvimento de um sistema de monitoração de fluidos baseado em espectroscopia de impedância. 2014. 71 f. Monografia (Mestrado em Engenharia Elétrica) – Programa de Pós-Graduação em Engenharia Elétrica e Informática Industrial, Universidade Tecnológica Federal do Paraná. Curitiba, 2014.
LERVIK, J. K.; KULBOTTEN, H.; KLEVJER, G. Prevention of hydrate formation in pipelines by electrical methods. In: Proceedings of the Seventh International Offshore and Polar Engineering Conference. p. 25-30, 1997.
LI, X. B.; LARSON, S. D.; ZYUZIN, A. S.; MAMISHEV, A. V. Design Principles for Multichannel Fringing Electric Field Sensors. In: IEEE Sensors, v. 6, n. 2, p. 434–440, 2006.
LI, X. S.; WU, H. J.; ENGLEZOS, P. Prediction of gas hydrate formation conditions in the presence of methanol, glycerol, ethyleneglycol, andtriethyleneglycol with the statistical associating fluid theory equation of state. In: Industrial Engineering Chemistry Research, v. 45, p. 2131–2137, 2006.
LIANG, S.; KUSALIK, P. G. Explorations of gas hydrate crystal growth by molecular simulations. In: Chemical Physics Letters, v. 494 (4-6), p. 123-133, 2010.
LIANG, S.; ROZMANOV, D.; KUSALIK, P. G. Crystal growth simulations of methane hydrates in the presence of silica surfaces. In: Physical Chemistry Chemical Physics, v. 13, p. 19856-19864, 2011.
LIU, Y.; SONG, Y.; CHEN, Y.; YAO, L.; LI, Q.; J. Nat. Gas Chem, v. 19, p. 224, 2010.
LONGO, J. P. N.; WRASSE, A. N.; DOS SANTOS, E. N.; DA SILVA, M. J.; MORALES, R. E. M. Sensor Capacitivo Interdigital para Detecção de Parafina em Oleodutos. In: Anais do XX Congresso Brasileiro de Automática, p. 391-395, 2014.
MACDONALD, J. R. Impedance Spectroscopy. In Annals of Biomedical Engineering, v. 20, p. 289-305, 1992.
MACDONALD, J. R. Impedance Spectroscopy. New York: John Wiley & Sons. 1987.
73
MAEDA, K.; KATSURA, Y.; ASAKUMA, Y.; FUKUI, K. Concentration of sodium chloride in aqueous solution by chlorodifluoromethane gas hydrate. In: Chemical Engineering and Processing: Process Intensification, v. 47, p. 2281-2286, 2008.
MAK, C. W.; MCMULLAN, R. K. J. Chem. Phys., v. 42, p. 2732, 1965.
MAKOGON, Y.F. Hydrates of Hydrocarbons. PennWell Books, Tulsa, OK. 1997
MAMISHEV, A. V.; SUNDARA-RAJAN, K.; YANG, R.; DU, Y.; ZAHN, M. In: Proceedings of the IEEE; IEEE: Piscataway, NJ, v. 92, n. 5, p. 809, 2004.
MATSUMOTO, M. Four-Body Cooperativity in Hydrophonic Association of Methane. In: Journal of Physical Chemistry Letters, v. 1 (10), p. 1552–1556, 2010.
MCMULLAN, R. K.; JEFFREY, G. A. Polyhedral Clathrate Hydrates. IX. Structure of Ethylene Oxide Hydrate. In: Journal of Chemical Physics, v. 42, p. 2725, 1965.
MEHTA, A. P.; HEBERT, P. B.; CADENA, E. R.; WEATHERMAN, J. P. Fulfilling the promise of low-dosage hydrate inhibitors; journey from academic curiosity to successful field implementation. In: SPE Products Facilities, v. 18 (1), p. 73-79, 2003.
MULLIN, J. W. Crystallization, 3rd Edition, Butterworth-Heinmann, Oxford, U.K, 1993.
NAGASHIMA, K.; YAMAMOTO, Y.; TAKAHASHI, M.; KOMAI, T. Interferometric observation of mass transport processes adjacent to tetrahydrofuran clathrate hydrates under nonequilibrium conditions. In: Fluid Phase Equilibria, v. 214, p. 11-24, 2003.
OUAR, H.; CHA, S.B.; WILDEMAN, T.R.; SLOAN, E.D. The Formation of Natural-Gas Hydrates in Water-based Drilling-Fluids. In: Transactions of the Institution of Chemistry E, v. 70-A, p. 48, 1992.
PARK, S. S.; KIM, N. J. Study on methane hydrate formation using ultrasonic waves. In: Journal of Industrial and Engineering Chemestry, v. 19(5), p. 1668–1672, 2013.
PARRA, V. E. L. Escoamento bifásico líquido-gás em golfadas com leve mudança de direção, Dissertação (Mestrado em Engenharia) - Programa de Pós-graduação em Engenharia Mecânica e de Materiais, Universidade Tecnológica Federal do Paraná, Curitiba, p. 137, 2013.
PAULING, L.; MARSH, R. E. The Structure of Chlorine Hydrate. In: Proceedings of Natl Academy of Science, v. 38, p. 112, 1952.
POWELL, H. M. The structure of molecular compounds. Part IV. Clathrate compounds. In: Journal of the Chemical Society, p. 61-73, 1948.
PRASAD, P. S. R.; PRASAD, K. S.; THAKUR, N. K. Laser Raman spectroscopy of THF clathrate hydrate in the temperature range 90–300 K. In: Spectrochimica Acta Part A, v. 68, p. 1096–1100, 2007.
74
RIPMEESTER, J. A.; RATCLIFFE, C. I.; KLUG, D. D.; TSE, J. S.; Molecular Perspectives on Structure and Dynamics in Clathrate Hydrates. In: Annals of the New York Academy of Sciences, v. 715, p. 161-176, 1994.
RIPMEESTER, J. A.; TSE, J. S.; RATCLIFFE, C. I.; POWELL, B. M.; Nature, v. 325, p. 135, 1987.
ROSE, W. & PFANKUCH, H., Society of Petroleum Engineers, SPE Paper 11106, 1982
SABIL, K. M.; WITKAMP, G. J.; PETERS, C. J. Phase equilibria of mixed carbon dioxide and tetrahydrofuran hydrates in sodium chloride aqueous solutions. In: Fluid Phase Equilibria, v. 284, p. 38-43, 2009.
SARACENO, P. Estudo do fenômeno de parafinação a partir de um óleo cru. 2007, XIV. Dissertação - Universidade Federal do Rio de Janeiro, COPPE, 2007.
SEO, Y.; LEE, H.; RYU, B. Hydration number and two-phase equilibria of CH4 hydrate in the deep ocean sediments, In: Geophysical Research Letters, v. 29, p. 85, 2002.
SLOAN, E. D. Clathrate Hydrates of Natural Gases. 2nd Edition, Marcel Dekker, Inc., New York, NY, 1998.
SLOAN, E. D. Introductory overview: Hydrate knowledge development. In: American Mineralogist, v. 89, p. 1155-1161, 2004.
SLOAN, E. D.; KOH, C. A. Clathrate Hydrates of Natural Gases, 3rd Edition, CRC Press, NewYork, 2008.
SLOAN, E. D.; KOH, C.; SUM, A. K. Natural Gas Hydrates in Flow Assurance, GPP Elsevier, MA, 2010.
SONG, Y. C.; YANG, M. J.; LIU, Y. Nat. Gas Ind. (Tianranqi Gongye), v. 28 (8), p. 111, 2008.
STATOIL – WaxControl. Disponível em: http://www.ipt.ntnu.no/~jsg/undervisning/prosessering/gjester/LysarkAske2011.pdf Acesso em: Março/2015.
SUM, A. K.; BURRUSS, E. C.; SLOAN, E. D. Measurement of clathrate hydrates via Raman spectroscopy. In: Journal of Physical Chemistry B, v. 101, p. 7371-7377, 1997.
SUM, A. K.; WU, D. T.; YASUOKA, K. Energy science of clathrate hydrates: Simulation-based advances. In: MRS Bulletin, v. 36, p. 205-210, 2011.
SWANSON, R. W. Characteristics of the gas-liquid interface in two phase annular flow. Ph.D. Thesis, Univ. of Delaware, Delaware, 1966.
75
TUNG, Y. T.; CHEN, L. J.; CHEN, Y. P.; LIN, S. T. The Growth of Structure I Methane Hydrate from Molecular Dynamics Simulations. In: Journal of Physical Chemistry B, v. 114 (33), p. 10804-10813, 2010.
VENDRUSCOLO, Tiago Piovesan. Sensor Wire-Mesh de Impedância para Investigação de Escoamentos Multifásicos. 2012. Dissertação - UTFPR, Curitiba, 2012.
VON STACKELBERG, M. Solid Gas Hydrates. In: Die Naturwissenschaften, v. 3612, p. 359-362, 1949.
VON STACKELBERG, M., MÜLLER, H.R., Die Naturwissenschaften, v. 38, p. 456, 1951a.
VON STACKELBERG, M., MÜLLER, H.R., Journal of Chemical Physics, v. 19, p. 1319, 1951b.
VON STACKELBERG, M., Rec. Trav. Chim. Pays-Bas, v. 75, p. 902, 1956.
VON STACKELBERG, M., Zeitschrift Elektrochem., v. 58, p. 104, 1954.
WALSH, M. R.; KOH, C. A.; SLOAN, E. D.; SUM, A. K.; WU, D. T. Microsecond Simulations of Spontaneous Methane Hydrate Nucleation and Growth. In: Science 20, v. 326 n. 5956 p. 1095-1098, 2009.
WALTRICH, P. J.; FALCONE, G.; BARBOSA JR., J. R. Axial development of annular, churn and slug flows in a long vertical tube. In: International Journal of Multiphase Flow, v. 57 p. 38-48, 2013.
WELLING AND ASSOCIATES. Survey cited by Macinosh, Flow Assurance Still Leading Concern Among Producers. In: Offshore, v. 60(10), 2000.
WILSON, P. W.; HAYMET, A. D. J. Hydrate formation and re-formation in nucleating THF/water mixtures show no evidence to support a “memory” effect. In: Chemical Engineering Journal, v. 161, p. 146-150, 2010.
WRASSE, Aluísio do Nascimento. Sensor Capacitivo para Imageamento Direto de Escoamentos Bifásicos. 81 f. Dissertação - Programa de Pós-Graduação Em Engenharia Elétrica e Informática Industrial, Universidade Tecnológica Federal do Paraná. Curitiba, 2015.
ZAMAN, M.; BJORNDALEN, N.; ISLAM, M. R. Detection of precipitation in pipelines. In: Petroleum Science Technology, 2004.
ZERPA, L. E.; SALAGER, J. L.; KOH, C. A.; SLOAN, E. D.; SUM, A. K. Surface chemistry and gas hydrates in flow assurance. In: Industrial Engineering Chemical Research, v. 50, p. 188–197, 2011.
ZERPA, L. E.; SLOAN, E. D.; KOH, C.; SUM, A. K. Hydrate Risk Assessment and Restart-Procedure Optimization of an Offshore Well Using a Transient Hydrate
76
Prediction Model. In: Society of Petroleum Engineers: Oil and Gas Facilities, v. 1, p. 49-56, 2012.
ZHOU, X.; FAN, S.; LIANG, D.; WANG, D.; HUANG, N. Use of Electrical Resistance to Detect the Formation and Decomposition of Methane Hydrate. In: Journal of Natural Gas Chemistry, v. 16, p. 399-403, 2007.
77
APPENDIX A
In this appendix is possible to visualize the individual results of experiments
mentioned in section 4.4.2. Using the commercial sensor with the impedance analyzer,
some experiments were performed and the results are shown in Figure A 1 and Figure A
2 (Capacitance and resistance without hydrate formation).
Figure A 1 - THF-water mixture without formation capacitance - Commercial sensor - Impedance
analyzer
Figure A 2 - THF-water mixture without formation resistance - Commercial sensor - Impedance analyzer
78
As can be seen, when there is no hydrate formation, the capacitance and resistance
values do not vary over the course of the experiment. This fact no longer occurs when
there is the formation, as Figure A 3 and Figure A 4 shows (capacitance and resistance
with hydrates formation respectively).
Figure A 3 - THF-water mixture with hydrates formation capacitance - Commercial sensor - Impedance
analyzer
79
Figure A 4 - THF-water mixture with hydates formation resistance - Commercial sensor - Impedance
analyzer
Using the interdigital sensor, more experiments were performed, as follows. It is
possible to observe the response of interdigital sensor using the commercial system,
firstly without hydrates formation (in Figure A 5 and Figure A 6) and after with
hydrates formation.
Figure A 5 - THF-water mixture without formation capacitance - Interdigital sensor - Impedance analyzer
80
Figure A 6 - THF-water mixture without formation resistance - Interdigital sensor - Impedance analyzer
Again, no change in the spectrum of the mixture when no hydrate formation, and
when there is the formation, the spectrum is modified as shown in Figure A 7 and
Figure A 8.
81
Figure A 7 - THF-water mixture with formation capacitance - Interdigital sensor - Impedance analyzer
Figure A 8 - THF-water mixture with formation resistance - Interdigital sensor - Impedance analyzer
The most conclusive results were observed in section 4.5.
82
APPENDIX B
In this section details on the developed system are given. The block diagram is shown
again in Figure B 1. The constituting parts are described in details below.
Figure B 1 - Dedicated system block diagram
B.1 Excitation circuit
The excitation circuit is the part of hardware responsible for generating and
conditioning the signal that will be injected into the excitation electrode of the sensor.
The DDS (Direct Digital Synthesizer) is responsible for generating a sine wave
voltage with known amplitude and frequency. The integrated circuit AD9835, from
Analog Devices was used. Through the microcontroller circuit it is possible to set
frequency and phase of the DDS output signal. To eliminate harmonics present in the
digitalized sine wave, a third order low-pass filter is placed in the DDS output. The
filter is dimensioned to have a cutoff frequency of 5 MHz (Figure B 2).
Figure B 2 - Pi filter
83
After being filtered with a low impedance output, the sinusoidal voltage signal
passes through a voltage amplifier circuit and is finally directed to the excitation
electrode of the sensor.
B.2 Measurement Circuit
The measuring circuit is characterized for amplifying the signals and then passing
it to the microcontroller. It comprises two sub-circuits: subtractor/reference and
demodulator. Impedance measurement is based on an auto-balancing bridge, that is, the
amplifier output voltage is increased proportionally to the current flowing through the
non-inverting input (Figure B 3). Thus, there is an auto-balancing bridge output voltage
signal proportional to the current signal flowing through the sensor, which is
characterized as the response signal to the substance present on the sensor. OPA656,
from Texas Instruments, was chosen as the auto-balancing bridge in this work.
Figure B 3 - Representation of the measurement principle
The output signal of the subtractor circuit is driven by a voltage amplifier to
ensure an acceptable voltage level at the demodulator input. The demodulator circuit
detects the envelope of the signal, making it into a DC signal, which is sent to the
microcontroller after voltage gain stages to ensure the reading range. The demodulator
component used is the LTC5507, manufactured by Linear Technology.
B.3 Microcontroller and Firmware
The microcontroller configures both DDS and the two digital potentiometers, and
is responsible for signal acquisition and transmission between the interface and the
84
software. It sets the DDS to generate the sinusoidal signals of excitement and reference.
For the voltage amplifiers that require a software control, digital potentiometers are
disposed. They are connected to the SPI interface of the microcontroller.
Communication with the computer was developed through an RS-232 interface,
configured to operate in asynchronous mode, eight data bits and no parity bit.
By only allowing the interface between the different modules, the processing
power was not considered a limiting factor of choice, since the microcontroller is not
processing function of information to be read.
The firmware consists of a set of operational instructions programmed directly on
the hardware. The development language utilized was C. The reason for this choice is
linked to the popularity of this programming language, and for possessing a compatible
architecture with many compilers. The program was implemented through the
Microship MPLAB environment with the help of CCS compiler.
PIC’s USART was used to perform the microcontroller communication with the
computer, with an operating configuration in asynchronous mode. The operation rate
was 9600 baud.
B.4 Software
Visual Studio from Microsoft was chosen to develop the operating software.
Basically, the software is the interface that connects the user with the system. It is
possible to simply send commands to the microcontroller according to an internal
protocol that was created. In that matter some letters or codes were determinate in
hexadecimal do avoid communication problems. The program performs tasks only if it
is controlled by the user, it is therefore, an event-oriented software. Figure B 4 gives an
overview of the developed software.
Figure B 4 - Software control menu