Optical Fiber Technology 40 (2018) 82–92

Contents lists available at ScienceDirect

Optical Fiber Technology

journal homepage: www.elsevier.com/locate/yofte

Regular Articles

Multi-interface level in oil tanks and applications of optical fiber sensors T

a b,⁎ a a
Arnaldo G. Leal-Junior , Carlos Marques , Anselmo Frizera , Maria José Pontes
a
Telecommunications Laboratory, Graduate Program on Electrical Engineering, Federal University of Espírito Santo, Fernando Ferrari Avenue, 29075-910 Vitória, ES,
Brazil
b
Instituto de Telecomunicações and Physics Department & I3N, Universidade de Aveiro, Campus Universitário de Santiago, 3810-193 Aveiro, Portugal

A R T I C L E I N F O A B S T R A C T

Keywords: On the oil production also involves the production of water, gas and suspended solids, which are separated from
Optical fiber sensors the oil on three-phase separators. However, the control strategies of an oil separator are limited due to un-
Fiber Bragg Gratings availability of suitable multi-interface level sensors. This paper presents a description of the multi-phase level
Multi-interface measurement problem on the oil industry and a review of the current technologies for multi-interface level assessment. Since
Oil production
optical fiber sensors present chemical stability, intrinsic safety, electromagnetic immunity, lightweight and
Oil density
multiplexing capabilities, it can be an alternative for multi-interface level measurement that can overcome some
of the limitations of the current technologies. For this reason, Fiber Bragg Gratings (FBGs) based optical fiber
sensor system for multi-interface level assessment is proposed, simulated and experimentally assessed. The re-
sults show that the proposed sensor system is capable of measuring interface level with a relative error of only
2.38%. Furthermore, the proposed sensor system is also capable of measuring the oil density with an error of
0.8 kg/m3.

1. Introduction measurement of interface level sensors. In addition, there may be a

Liquid level and the level of the interface between fluids are im-
portant parameters in industries such as agriculture, automobile, food
storage, chemical, medical, oil and gas [1]. In the oil and gas industry,
there is the necessity to measure level and interface level between fluids
on wells, tanks, reservoirs, processing vessels and storage vessels for
fluids with different density, corrosiveness and viscosity [2]. Further-
more, processes in oil and gas industry may have high range of pressure
and temperature [3], which increase the complexity and robustness
requirements for the instrumentation.
One of the key process on the oil and gas industry is the separation
of the oil, water and gas. The hydrocarbons obtained on producing
wells and transported are a mixture of oil, gas, the produced water and
suspended solids [4]. These different components are separated on the
oil separation unit, which generally makes the separation between oil,
gas and water through the density difference of the immiscible fluids
[4]. However, there is no laminar and well-defined layers for each one
of the fluids. Instead, there is an emulsion layer between the oil and
water, which has its composition and behavior affected by the water
and oil properties [5], as will be depicted later in this paper. Further-
more, there may be the formation of foam between the oil and gas
layers on atmospheric tanks, which also has a dynamic behavior [6].
The foam and emulsion layer discussed can introduce errors on the
formation of sludge or wax on the tank walls, which also can have in-
fluence on the sensors performance.
Besides the operational problems of the multi-interface level mea-
surement caused by the formation of the emulsion and foam layers,
there are also limitations on current technologies for multi-interface
level measurement related to the variety of fluids that can be inside of
an oil separator, the operation on harsh environments and safety issues
[2]. There is a presence of high inflammable gases in crude oil tanks
that may generate explosions. For this reason, the devices employed on
the tank instrumentation cannot surpass certain limits of voltage, cur-
rent and capacitance [7].
The lack of reliability on multi-interface level measurement systems
takes to simplified control strategies on crude oil separators [2]. This
limitation on the interface level measurement leads the companies to
employ subsequent separators to achieve better separation of each
phase of fluid, which greatly enhances the cost of the plant and the
complexity on the maintenance [4]. If the separation between oil and
water does not occur correctly, it is possible to obtain oil with high
amount of water on the refining process, which has influence on the oil
processing prices and efficiency [4]. Another scenario is the water with
some amount of oil that can contaminate the environment. The effects
of the oil contamination are broadly discussed in [8].
In order to overcome the limitations related to the interface level on

⁎
Corresponding author.
E-mail addresses: arnaldo.leal@aluno.ufes.br (A.G. Leal-Junior), cmarques@av.it.pt (C. Marques), frizera@ieee.org (A. Frizera), mjpontes@ele.ufes.br (M.J. Pontes).

https://doi.org/10.1016/j.yofte.2017.11.006
Received 9 October 2017; Received in revised form 6 November 2017; Accepted 9 November 2017
1068-5200/ © 2017 Elsevier Inc. All rights reserved.
A.G. Leal-Junior et al. Optical Fiber Technology 40 (2018) 82–92

the oil separation processes, a multi-interface level sensor must detect

the water, oil, emulsion and foam layers [2]. In addition, the costs of its
maintenance have to be low and the time between maintenances has to
be high, in order to achieve high production rates and a more reliable
control of the oil separators [7]. Furthermore, the device must be in-
trinsically safe [7]. If a device is capable of meet these requirements, it
will provide a more effective utilization of the separator, which also
includes the design of more compact separators with lower safety
margins and higher efficiency [4].
Optical fibers sensors are technological systems that result on de-
vices that are compact, lightweight, immune to electromagnetic fields,
chemically stable and allow multiplexing [9]. In addition, they can be
considered intrinsically safe, especially when compared to conventional
electronic technologies. Optical fiber sensors have been employed in
industrial applications on the measurement of different parameters,
such as temperature [10], liquid level [11], acceleration [12], pressure
[13], acoustic [14] and refractive index [15]. The advantages of optical
fiber sensors and their broad use on industry make them an interesting
option for multi-interface level sensing.
However, in order to obtain an efficient optical fiber based multi-
interface level sensor, it is necessary a review of the oil separation
process and its issues. Moreover, a review on the current technologies
for multi-interface level assessment has to be made to evaluate the Fig. 1. Phase distribution of a tank containing oil, water and gas.
advantages and drawbacks of each technology and how the optical fiber
based technology can overcome the limitations of conventional tech-
nologies. For this reason, this paper presents a review of the process of
oil separation and the conventional technologies for interface level
measurement. After this review, an optical fiber based multi-interface
level measurement system is proposed and the principle of measuring
water and oil layers are experimentally assessed and validated.
This paper is organized as follows. Section II presents the descrip-
tion of the oil separator and its layers. Moreover, the emulsion and
foam layers are described and some of its properties are presented. A
review of multi-interface level measurement technologies is discussed
in Section III. Section IV presents an overview of optical fiber based
sensors. Also, a novel proposal of optical fiber based multi-interface
level sensor is presented on this section. Simulation analysis and ex-
perimental results of the sensor and its ability to detect water and oil Fig. 2. Schematic illustration of a three-phase gravity separator.
layers are presented in Section V. Conclusions and suggestions for fu-
ture works are discussed in Section VI. not have the sludge layer. In addition, the oil phase collected on the oil
outlet also present some amount of water, which is further separated on
2. Multi-interface level problem description the next separation stages. Each gravity separator is able to separate
about 90–95% of water on the oil [19].
An important problem on oil extraction is the combined production There are different control strategies for the multiphase separator of
of water and gas with the crude oil [16]. In order to address this issue, Fig. 2, which are simplified due to the uncertainty of the interface level
oil separators are employed. Since oil, water and gas are immiscible provided by the limitations of the current techniques for multi-interface
fluids and have different densities, the separation is achieved by settling level measurement [2]. One of these strategies is the one presented in
them on large tanks until obtain a layer of water on the bottom of the [4], which consists in a closed loop control for real-time control of the
tank, a layer of oil in the middle and gas at the top [17]. However, there level of interfaces and is presented on the block diagram of Fig. 3.
may also be an emulsion formation, which is a layer with a mixture of On the control strategy presented in Fig. 3, the three-phase se-
water and oil [16]. A foam layer may also be formed between the oil parator controller has to be able to control the sand flushing, the control
and the gas due to carbonation processes [18]. The hydrocarbon multi- valve, the water pump and the chemical injection process. All the
phase mixture also includes sand or other solids, which are present on control commands have to be done with only two input data: the sand
the well and form a sludge on the bottom of the separator [4]. The estimation and the multi-level interface measurement. Since the sand
phases that are formed inside a tank or an oil separator are presented in estimation is applied only to control the sand flushing, the interface
Fig. 1. level sensor system provides the information to control the pump,
A common approach to separate the phases abovementioned is to control the valve for oil inlet and to control the chemical injection on
employ a three-phase separator, which is based on the principle that the the oil.
gravity will separate immiscible fluids with different densities [4]. Since the separator can reach its full capacity or the water layer can
Fig. 2 shows a typical oil separator. In this device, a deflection plate surpass the bulkhead, a control valve is applied to block the flow if the
obstructs the inlet flow of the mixture and redirects the fluid for the left separator capacity is reached. Whereas, the chemical injection is made
part of the separator shown in Fig. 2. The reason for place the mixture to reduce the layers of emulsion and foam, which generally are not
on the left side is due to the positioning of the outlet valve of the water precisely measured with some of the conventional technologies of
on the left side of the bulkhead presented in Fig. 2. There also may have multi-interface level measurement [20].
a sand monitor and sand flushing nozzles on the separator or close to Besides the inaccuracies on the foam and emulsion layers
the separator stage [4]. For this reason, the presented separator does

83
A.G. Leal-Junior et al.
Fig. 3. Block diagram of the control for a three-phase oil separator.
estimations, these two undesirable layers can lead to operational pro-
blems on the oil separation process. The foam has large weight to vo-
lume ratio, which means that it occupies a large volume on the tank and
it leads to reduction on the oil production [4]. To reduce the amount of
foam, anti-foam products are injected on the fluid on chemical injection
process, which are controlled by the separator controller based on the
information provided by the multi-interface level sensor. Moreover, the
presence of the emulsion layer increases the time that the fluid has to be
inside the separator, also known as the residence time [20], which also
reduce the production. As it is done to reduce the foam layer, emulsion
breaking compounds are applied on the hydrocarbon mixture and it is
represented by the chemical injection block.
To propose a sensor that can effectively measure the emulsion and
foam layers, it is necessary to have some knowledge of these layers
behavior and formation. Foam is a metastable dispersion of gas in a
continuous oil phase and its formation is related to crude oil properties
such as density, viscosity and surface tension [21]. Furthermore, foam
is stabilized by the asphaltenes and resins present on the crude oil [22].
Since the foam presents unstable behavior, it may cause a fluctuation on
the multi-interface level sensor signal, which can limit its resolution
and precision [4].
The emulsions are colloidal systems with thin droplets of one liquid
dispersed on another liquid, where the two liquids are immiscible and
there is the presence of emulsifying agents, which are asphaltenes and
resins on the water and oil emulsions [23]. Its formation occurs when
there are turbulence or sufficient mixing to obtain a dispersion between
liquids [24]. Several properties of the oil and water can affect the
emulsion characteristics and stability, which include different water
volume fractions, temperature and shear rate [25], the salinity of the
water [26] and electric stress [27]. Furthermore, the emulsion presents
a non-Newtonian behavior and are classified as viscoelastic fluids [28].
These effects can lead to an even more unpredictable behavior of the
emulsion characteristics with the shear rate [25] and asphaltenes and
resins affect the elastic modulus of the emulsion due to its viscoelastic
nature.
On the crude oil case, there are three main types of emulsions:
water-in-oil (W/O) emulsion, oil-in-water (O/W) emulsion and multiple
emulsion. W/O emulsion is formed when the water droplets are dis-
persed on the oil continuous phase. When the opposite happens, i.e.
84
Optical Fiber Technology 40 (2018) 82–92
when the oil droplets are dispersed on the water, it is an O/W emulsion
[5]. Multiple emulsions can occur in two manners: water-in-oil-in-water
(W/O/W) or oil-in-water-in-oil (O/W/O) [24].
There are different parameters that affect the emulsion stabilization
and dynamics. For this reason, there are different classifications of the
emulsion with respect to its stability. For W/O emulsion, the four
classes of emulsions are stable, mesostable, entrained water and un-
stable [24]. This classification is related to the time that each emulsion
type remains unbroken on controlled conditions of temperature and
fluid turbulence. Stable emulsions remain unbroken for more than a
week on a laboratory. Whereas, mesostable W/O solutions are the ones
that are divided in water and oil after three days on controlled condi-
tions. Entrained water remains unbroken in less than one day in la-
boratory. Finally, on unstable emulsions, W/O solutions are broken into
water and oil after some hours [24].
Since most of the emulsions found on oil and gas industry are stable
[24] and conventional technologies for multi-interface level measure-
ment generally cannot effectively measure the emulsion layer [2], it is
necessary to break the emulsion to reduce the residence time of the oil
and increase the production rate. Thus, demulsification agents are
employed [7]. Although these agents can break the emulsion, this ap-
proach is costly and can lead to environmental problems due to the
chemicals composition [7]. The same happens with the chemical in-
jection to break the foam layer.
The problems aroused by the foam and emulsion layer, as well as
the control limitation of separators, can be reduced if reliable tech-
nologies for multi-interface level assessment are able to detect these
additional layers with high precision and resolution.
3. Conventional technologies for multi-interface level
measurement
In order to overcome the limitations and attend the necessities of
the multi-interface level application, different sensors have been pro-
posed throughout the years with varying degree of success and are they
are presented on the next sub-sections.
A.G. Leal-Junior et al. Optical Fiber Technology 40 (2018) 82–92

Fig. 4. Displacer-based multi-interface level sensor.

Fig. 5. Differential pressure-based multi-interface level sensor with multiple pressure
sensors.
3.1. Displacer-based sensors

These sensors rely on the principle that a body with its density
between the densities of the fluids will be positioned on the interface
between these two liquids [2]. For this reason, the sensor is comprised
of a displacer that has its density higher than the one of the oil and
lower than the water density (see Fig. 4). The displacer movement es-
timates the level of water and oil.
Although this sensor has the advantages of low cost and simplicity,
it can only measure sharp differences on the densities [7]. Therefore, if
the density of the interface is not constant due to the presence of an
emulsion layer, the displacer response will be unstable [4]. Further-
more, the thickness of the foam layer cannot be measured with this
approach [2].

3.2. Differential pressure-based sensors

These sensors employ the hydrostatic pressure difference to esti-

mate the level of the interface. Since the hydrostatic pressure depends
on the liquid density and the density varies with the temperature, a
temperature compensation is necessary when this approach is em-
ployed [7]. Although some configurations can only detect the interface
level if the fluids have high density difference, it is possible to apply
configurations with multiple pressure sensors or with a pressure sensor
with vertical movement along the tank to be able to estimate multi-
interface level when there is low variation of the density or when there
is an emulsion layer. By applying multiple pressure sensors, it is pos-
sible to further increase its precision with sensor fusion algorithms [29].
A schematic view of the multi-interface level measurement system with
multiple pressure sensors is presented in Fig. 5.
Besides the requirement of temperature compensation, another
limitation is the possibility of wax formation on the pressure trans-
mitter, which will harm its performance [2]. The temperature de-
pendency is a common limitation of some technologies for liquid level
sensing and can be overcame with a temperature sensor close to each
pressure sensor to compensate temperature effects [11].
Fig. 6. Multi-electrode capacitance sensor for multi-interface level measurement.
column with the detection electrodes [30], as presented in Fig. 6. The
second configuration relies on a single column of electrodes using the
tank metallic wall as a support for the sensors, which is presented in
[4,18]. The different configurations include the application of elec-
trodes with a transversal section [17], radial positioning of the elec-
trodes on the tank [3,19] and the combination of electrodes and coils to
obtain a capacitive and an inductive response to enhance the precision
of the level estimation [31].
The limitation of this approach is the problems caused by con-
ductive liquids. When the oil presents water content higher than 40% of
its volume, the capacitance measurement can lose sensitivity, which
lead to errors [7]. Moreover, the formation of wax on the sensor can
inhibit its operation [4] and the limitation of the electrodes current is
also an important drawback [32].

3.3. Capacitance sensors with multi-electrode array 3.4. Ultrasound-based sensors

This principle is based on the measurement of the dielectric constant
of the fluid between the capacitive plates. The charge/discharge circuit
presents low cost and the foam layer can be detected by analyzing the
fluctuation of the capacitance on the frequency domain, [6,18,19]. For
this reason, different configurations were proposed throughout the
years. Fig. 6 shows a configuration of a multi-electrode capacitance
sensor for multi-interface level assessment.
Two configurations are possible. The first one is based on two in-
dependent columns, one with the excitation electrodes and the other
An ultrasound-based sensor employs a transmitter and a receiver on
a transmission medium [2]. The different interfaces are detected by the
differences on the acoustic impedance of each layer, which is related to
the density and the speed of sound of the medium [4]. An overview of
the sensor operation principle is presented in Fig. 7.
Another approach is to place the transmitter and receiver facing
each other in vertical stands. Thus, with the vertical movement of both
transmitter and receiver alongside the tank with the multi-phase liquid,
the liquid identification can be made and each interface level can be

85
A.G. Leal-Junior et al.
Fig. 7. Ultrasound-based sensor for multi-interface level measurement.
estimated [16].
It is also possible to estimate the multi-interface liquid level with an
array of ultrasonic transducers, which can be positioned on two col-
umns, one for the emitter and the other for the receiver as presented in
[33]. Similar approach can be made, but applying a single column in-
stead of the double column configuration. In this configuration, the
ultrasonic transducers are activated in a time-multiplexed configuration
[34].
These sensors based on ultrasonic transducers usually need a tem-
perature compensation and a compensation for the density variation of
each liquid, which are achieved by the application of reference sensors
for temperature and pressure, respectively [34]. Furthermore, the
presence of bubbles, wax and sludge can harm the sensor performance
[34]. Another limitation is the difficulty to estimate the level of emul-
sion and foam layers, which can be achieved with artificial neural
networks [16] and pattern recognition algorithms [34].
3.5. Vibration-based sensors
The vibration-based sensors detect the variation of frequency re-
lated to each liquid that the vibrating device is submerged [7]. The
frequency variation is related to the variation of density of each liquid.
However, it is necessary to move these sensors along the tank and this
movement can create turbulences on the fluid, which introduce errors
on the sensor measurement. Furthermore, it is very sensitive to the
material build-up [7]. An approach to make a vibration-based sensor
without the necessity of the contact with the fluids is presented in [35].
The proximity of the water on the detection region of the resonator
changes its frequency response and Q-factor. However, this sensor does
not work in metallic tanks and the detection of emulsion and foam
layers are not addressed [35].
Different approach for a non-contact sensor is presented in [36] and
is comprised of placing copper wires around the tank with a multi-in-
terface fluid in such a manner that the wires behave as a microwave
resonator. The dielectric properties variation presented by each liquid
leads to the variation of the resonance frequency. However, the sensor
was on a 60 mm height recipient and place the wires in an oil separator
can be very difficult and may arise the safety issues of the application.
3.6. Radiation-based sensors
This approach comprises of an emitter of low energy gamma ray in a
vertical array, which is measured by Geiger detectors. The attenuation
measured by the detectors is related to the density of the medium and
86
Optical Fiber Technology 40 (2018) 82–92
which fluid is present on the medium [2]. This technology can operate
on harsh environments with high temperature, pressure and corrosive
materials, can measure all the interfaces reliably [4]. However, it pre-
sents higher costs than the other presented technologies [7]. Further-
more, there is the safety issues related to the hazardous nucleonic ra-
diation presented in these devices [2].
3.7. Thermal measurement-based sensors
These sensors are based on the differences of the thermodynamic
properties of the water and oil [4]. Since these sensors are based on the
differences of the thermal conductivity, it is necessary to obtain a re-
ference sensor with constant temperature by continuously circulating a
reference liquid, which is intrusive and can enhance the complexity of
the sensor system [2].
4. Optical fiber sensors for multi-interface level measurement
4.1. Overview of optical fiber sensor technology
As abovementioned, advantages like chemical stability, electric
isolation, immunity to electromagnetic interference, multiplexing cap-
abilities and intrinsic safe operation enable the application of optical
fiber sensors in harsh environments, such as the one in oil and gas
applications [10]. These advantages are well aligned with the re-
quirements for the multi-interface level assessment for oil and water
mixtures.
Different approaches for optical fiber sensors have been proposed
throughout the years, which include intensity variation based sensors
[37], interferometry [38], Fiber Bragg Gratings (FBG) [39], Tilted Fiber
Bragg Gratings (TFBG) [40], Surface Plasmon Resonance (SPR) based
sensors [41], Long Period Gratings (LPG) [42], distributed sensors
based on optical fiber nonlinear effects [43], among others.
An optical fiber sensor for multi-interface level application was
proposed in [44]. However, this sensor comprises only of lateral sec-
tions in different positions along the fiber, which present problems
related to the material build-up on the sensor and it was not tested on
the presence of emulsion and foam. Nevertheless, there are different
optical fiber based technologies that can assess the multi-interface level.
It is possible to apply optical fiber sensors to almost all the ap-
proaches for multi-interface level sensor presented in Section III. The
acoustic-based identification of liquid interface can be done with op-
tical fiber distributed acoustic sensors, which are widely employed for
different applications [45]. Since the optical fiber is compact and
lightweight, it is possible to coat the fiber with different materials [46].
If a material with density between the one of water and oil is coated on
an optical fiber based displacement sensor such as the one presented in
[47], it might be able to measure the interface level in a setup as the
one of displacer-based sensors. Furthermore, the dielectric constant
depends on the temperature [48]. For this reason, it is possible to apply
optical fiber temperature sensors with different approaches like dis-
tributed temperature sensing [49], FBGs [10] and interferometry [50]
to measure the temperature profile on the tank and estimate the di-
electric constant. Moreover, it can be applied the same principle as the
one presented in [2] to estimate the multi-interface level with tem-
perature sensors. Vibration-based sensors can also be applied with the
optical fiber technology, since there are many accelerometers and vi-
bration sensors based on optical fibers on the literature [12,51,52].
Another liquid interface level approach that can be applied with
optical fiber sensors is the differential pressure sensors. It is possible to
apply optical fiber pressure sensors based on interferometry [13] or
FBGs [53] to measure the multi-interface level with the same principle
presented in Section III. Since the refractive index of oil is different
from the one of the water [54], it is also possible to identify the dif-
ferent layers on an oil separator with sensors for refractive index
measurements, which includes interferometers [55], LPG-based sensors
A.G. Leal-Junior et al. Optical Fiber Technology 40 (2018) 82–92

[42], or TFBG sensors [40].

Although the discussed optical fiber technologies can suffer with
some of the issues of the conventional technologies for interface mea-
surement, their compactness and multiplexing capabilities enable the
application of higher number of sensors to increase resolution and
precision. Furthermore, it is possible to combine different optical fiber
sensors on the same fiber to extract the best of each approach for the
multi-interface level measurement. The combination of higher number
of optical fiber sensors or different sensors for obtaining a higher pre-
cision on the measurement of a parameter, or to obtain compensation
techniques to enhance the sensor performance or even for simultaneous
measurement of different parameters are widely employed as presented
in [51,55–57].

4.2. Proposal of a FBG-embedded multi-diaphragm sensor system for

interface level assessment
The advantages of optical fiber sensors and the possibility of ap-
plication of different approaches for multi-interface level assessment
enable the proposal of a FBG-based sensor system for this application.
The sensor system employs the principle of differential pressure mea-
surement for multi-interface level assessment in a configuration similar
as the one presented in Fig. 5.
FBGs are periodic perturbation on the optical fiber refractive index
created with an incident Ultra-Violet laser beam [58]. The period of the
perturbation is determined by an interference pattern with phase mask,
which modifies the transmitted and reflected spectrum [39]. The cen-
tral wavelength of the reflected spectrum is the Bragg wavelength [58],
which is defined in Eq. (1):
λB = 2neff Λ (1)
where neff is the effective refractive index and the Bragg grating period
is represented by Λ . The wavelength variation happens due to the ef-
fects of temperature (ΔT ) and strain (ε ) represented by the elasto-op-
tical (Pe ) and thermo-optical (ζ ) coefficients. Furthermore, the fiber
thermal expansion (α ) also cause variation on the Bragg wavelength.
Eq. (2) shows the Bragg wavelength variation (ΔλB ) caused by tem-
perature and strain effects on the FBG.
ΔλB = [(1−Pe ) ε + (α + ζ )ΔT ] λB (2)
Although FBG are sensitive to strain and temperature, it can be
coated with different materials [46] or can be embedded in different
structures [39] to measure different parameters such as pressure [53],
vibration [51], liquid level [11], humidity [59] and concentration of
some chemical compounds [46].
Diaphragm-based optical sensors are employed on pressure [60],
vibration [51] and liquid level sensing [61]. For pressure and liquid
level sensing, the sensor is generally a FBG embedded on the center of a
circular diaphragm and the diaphragm deformation caused by the
pressure is directly transmitted to the FBG [11,53,61].
The operation principle of these sensors is the shift on the Bragg
wavelength due to the fiber strain, which is caused by the pressure on
the diaphragm. Since the FBG are sensitive to strain and temperature,
this principle can be applied only for applications with constant tem-
perature. If the diaphragm-based FBG pressure sensor is applied on
different temperatures, it is necessary a compensation technique to
isolate the effects of the temperature over the pressure measurement
[53,61].
Since there may be a variation of temperature on oil separators, the
application of diaphragm-based FBG pressure sensors for multi-inter-
face level measurements require a reference temperature sensor. In this
work, this temperature sensor is also based on FBG. As an additional
feature, the temperature sensor can also be applied also for an addi-
tional estimation of the liquid interface by means of the differences on
the temperature profile caused by the difference on the thermodynamic
properties of the oil and water. Furthermore, the compact design of the
Fig. 8. FBG-based multi-interface level sensor proposal.
sensor and the multiplexing capabilities of the FBG enable the appli-
cation of several sensors along the tank. The number of the temperature
and pressure sensors will affect the system resolution and precision and
the number of sensors can be estimated based on the requirements of
each application. The sensor diaphragm design has direct influence on
the pressure sensor response, especially on its sensitivity. For this
reason, the diaphragm design can be made with respect to the re-
quirements of the application and on the number of sensors that will be
applied. Fig. 8 shows the schematic illustration of the proposed FBG-
based sensor system for multi-interface level measurements.
The foam thickness can be estimated with the sensor system pro-
posed by analyzing each sensor response on the frequency domain as
presented in [6] and [18]. Similarly, it can be possible to estimate the
upper and lower positions of the layer by analyzing both pressure and
temperature signals on the time and frequency domain with sensor
fusion or identification algorithms. The block diagram of the analysis
with the proposed FBG-based system is presented in Fig. 9.
5. Proof-of-concept of FBG-embedded multi-diaphragm sensor for
interface level estimation
5.1. Simulation analysis
In order to show the feasibility of the proposed sensor system, si-
mulations are made with different configurations of water, oil and
emulsion. For the simulations presented, the temperature effects are
neglected. For this reason, the right part of the sum shown in Eq. (2) is
negligible and Eq. (2) can be rewritten as Eq. (3):
ΔλB = λB (1−Pe ) ε. (3)
However, ε is the strain on the fiber, which is embedded on the
diaphragm. Therefore, some of the diaphragm strain is transmitted to
the fiber. The strain of circular diaphragm within the elastic region can
be calculated as Eq. (4) [62]:
3 Δpa2
εdiap = − (1−υ) ,
8 Ediap t 2 (4)
where εdiap is the strain on the diaphragm, υ and Ediap are the Poison’s
coefficient and Young modulus of the diaphragm material, respectively.
Furthermore, a is the radius of the diaphragm that is in contact with the
liquid (diaphragm cavity radius) and Δp is the hydrostatic pressure on
the diaphragm, which is defined as Eq. (5):
Δp = ρgh, (5)
where ρ is the fluid density, g is the acceleration of the gravity and h is
the liquid level.
The Young modulus of the rubber employed on the simulation is

87
A.G. Leal-Junior et al. Optical Fiber Technology 40 (2018) 82–92

Fig. 9. Block diagram of the proposed multi-interface level measurement system based on FBG sensors.

1.6 MPa [11]. Whereas, the Young modulus of the silica fiber is 70 GPa
[63]. This high difference between the Young modulus of silica and
silicone rubber leads to a high difference in elasticity of these two
materials. For this reason, the fiber will restrain the strain of the dia-
phragm, which causes a lower strain on the fiber than the one calcu-
lated in Eq. (4). The difference between the calculated strain on the
diaphragm without considering the fiber and the strain that is actually
transmitted to the fiber can be approximated with the constant k . A
possible estimation for this constant is the relation of the diaphragm
Young modulus and diameter (Ediap and ddiap , respectively) with the
fiber diameter and Young Modulus E and d, respectively), as presented
in Eq. (6):

E d
k= ,
Ediap ddiap (6)

where the relation between k and the strains of the diaphragm (εdiap )
and the one transmitted to the fiber ε is presented in Eq. (7):
εdiap
ε= .
k (7)

Table 1 presents the values of each variable employed on the si-

mulations for the diaphragm. The diaphragm response is simulated by
means of substituting Eq. (5) and (6) in Eq. (4), which is further sub-
stituted in Eq. (7) for the estimation of the strain on the fiber. Finally,
Eq. (7) is substituted in Eq. (3) for the wavelength shift estimation.
Applying the parameters presented in Table 1 for the wavelength
shift of the diaphragm under water (density of 1000 kg/m3) and for a
liquid level of 1 m, a sensitivity of 2.89 pm/cm is obtained.
The simulations presented in this paper are made with four dia-
phragms with the dimensions presented in Table 1. Fig. 10 shows the
positions of the four diaphragms employed.
Table 1
Parameters employed on the simulations.
Symbol Parameter
a Diaphragm cavity radius
ddiap Diaphragm diameter
t Diaphragm thickness
Pe Effective photoelastic constant
E Silica fiber Young modulus
υ Rubber poisson ratio
Ediap Diaphragm Young modulus
d Optical fiber diameter
g Gravity acceleration
λB Bragg wavelength
Fig. 10. Positions of the four diaphragms employed on the simulation with a water level
of 1 m.
The first simulation is made with the setup presented in Fig. 10,
where there is only water in a total level of 1 m. The wavelength shift of
each diaphragm is presented in Fig. 11.
Since there is only water, the linear regression of the wavelength
shift with respect to the diaphragm position presents a constant slope,
which is equal to the one of the diaphragm response when the liquid is
Value
water. If the liquid employed is an oil with a density of 800 kg/m3, the
7.5 mm slope of the linear regression will be −2.31 pm/cm.
20 mm For the second simulation, the water level is reduced to 0.5 m
1 mm (50 cm) and a 50 cm layer of oil with density of 800 kg/m3 is added.
0.22
Therefore, there is a layer of oil with 50 cm on top of the layer of water
70 GPa
0.49 with 50 cm. Fig. 12 shows the wavelength shift for each diaphragm
3.5 MPa when the water layer is 50 cm and the oil layer is 50 cm.
125 μm The simulation of the wavelength shift is made by applying the
9.8 m/s2 density of each liquid. The slope of the curves presented is related to the
1537 nm
weighted sum of each liquid level and density. If the level is known, it is

88
A.G. Leal-Junior et al.
Fig. 11. Wavelength shift for each of the four diaphragms employed when there is a
water level of 1 m.
Fig. 12. Wavelength shift for each of the four diaphragms employed when there is a
water layer of 50 cm and an oil layer of 50 cm.
Fig. 13. Density measurement for each diaphragm with respect to its positions.
89
Optical Fiber Technology 40 (2018) 82–92
possible to estimate an equivalent density, which is related to the total
level, layer height and liquids density, for the diaphragms response.
Fig. 13 presents the equivalent density presented by each diaphragm.
Since it presents water and oil above it, Diaphragm 1 presents an
equivalent density lower than the one of the water and higher than the
one of the oil. Although the same happens with Diaphragm 2, its
equivalent density is lower than the one presented by Diaphragm 1. The
reason for this behavior is the higher height of the water layer for
Diaphragm 1, which leads the weighted sum that represents the
equivalent density to higher values. Whereas, Diaphragms 3 and 4 are
submerged to oil only. For this reason, its equivalent density is the same
of the oil (800 kg/m3). Therefore, if the total level is known, for the case
of oil and water layers, only two diaphragms are needed to determine
the interface level. Furthermore, two diaphragms are needed to esti-
mate the density of each liquid. Hence, it is possible that a system with
four diaphragms can be able to measure the interface level and the
liquids density simultaneously.
The final simulation is with a layer of water with 40 cm height, a
layer of oil with 50 cm height and a layer between the water and oil
with 10 cm height to represent the emulsion that can be an issue of
some oil separators. The density employed for the emulsion layer is
900 kg/m3. The wavelength shift for the four diaphragms in this case
presents a difference on the slope between the Diaphragms 2 and 3,
which in this case is 2.66 pm/cm. This difference can lead to errors on
the interface estimation. Fig. 14 shows the wavelength shift for the case
where the emulsion is also considered. For this reason, if the emulsion is
also considered, the system for simultaneous estimation of the multi-
interface level and liquids density will need an additional diaphragm.
5.2. Experimental setup
Since the densities of the two liquids employed are known and the
level will be also monitored. The proof of concept of the proposed
multi-diaphragm FBG sensor is presented for the case where two dia-
phragms are employed. As presented on the simulations for the case
with water and oil layers, only two diaphragms are needed to measure
the interface level, if the densities of the fluids are known. Furthermore,
if the level is known, only two diaphragms are needed to estimate the
density of both fluids as well. Therefore, in this case, there will be only
the identification of the water and oil layers. The presence of only water
and oil layers is a common situation in oil separators that have chemical
injection to eliminate the foam and emulsion layers. Fig. 15 shows the
experimental setup employed on the tests, which presents an interface
level of 90 cm. In addition, Fig. 15 shows the sm125 (Micron Optics,
Fig. 14. Wavelength shift for each of the four diaphragms employed when there is a
water layer of 40 cm, an oil layer of 50 cm and an emulsion layer of 10 cm.
A.G. Leal-Junior et al. Optical Fiber Technology 40 (2018) 82–92

Fig. 16. Water and diesel characterization for both diaphragms.

Fig. 15. Experimental setup for the validation of the multi-diaphragm interface sensor. In
this case, the interface level is 90 cm.
USA) FBG interrogator employed on the tests.
The oil employed on the tests is diesel due to its wider availability
and it forms a layer with lower emulsion than other oils. Although the
tank employed on the test has 1.2 m height, the level of the test is
limited to 1.15 m due to operational limitations. The tank also has a
ruler attached to provide a reference of the liquid level height. The test
is divided in three steps: the first step is the characterization of the
diaphragms with only water. The second one is with diaphragms sub-
merged in oil, which in this case is diesel. Finally, a test with one layer
of water and another one with diesel is made. Since the density of water
is 1000 kg/m3 and the density of the diesel is 840 kg/m3 at 25 °C [64], it
is expected that the layer of diesel will be above the layer of water. The
test with water and diesel layers are made with constant total level i.e.
the height of the diesel and water layers are changed, but the sum of the
water and diesel heights is constant. Furthermore, there are only two
layers: water and diesel. For this reason, it is possible to employ only
two diaphragms for the interface level estimation. All the tests pre-
sented were made at constant temperature of about 25 °C.
Both diaphragms employed present the same thickness (0.7 mm)
and the same diameter (15 mm). Two single mode optical fibers with
FBG inscribed are embedded on each diaphragm. However, in-
accuracies on the diaphragm manufacturing and on the fiber posi-
tioning on the embedment lead to variation of the sensor sensitivity. For
this reason, each diaphragm is characterized with water and diesel.
Diaphragm 2 presents an error of 2.11 pm and 1.73 pm for water and
diesel, respectively.
For the interface level tests, the water level is reduced and the diesel
level is increased, which leads to a variation of the interface level (see
Fig. 15). However, the total level i.e. the sum of the levels of water and
diesel is constant and equal to 115 cm. The test begins with a 10 cm
layer of diesel and 105 cm of water level. The water level is reduced in
10 cm and the diesel level is increased in 10 cm until the water level
reaches 65 cm.
Each diaphragm measures the hydrostatic pressure of the fluid
column above them. Since the hydrostatic pressure depends on the
gravity acceleration, height of fluid column and fluid density, it is ex-
pected a hydrostatic pressure reduction with the increase of diesel
column and reduction of water column, since the diesel density is lower
than water density. As the interface level changes, the wavelength shift
of the diaphragms responses will be related to a weighted sum of water
and oil layers with their densities. Therefore, it is possible to analyze
the slope of the responses of each diaphragm with the interface level
variation as presented in Fig. 17. The RMSE between the reference and
measured values is 2.51 cm, which means a relative error of 2.38%. The
diaphragms employed on the test have a variation on the thickness of
about 0.2 mm, which can increase its measurement errors. Therefore,
the RMSE can be further reduced by applying lower tolerances on the

5.3. Results and discussion

Water characterization tests are made on the range from 40 cm to

100 cm with steps of 10 cm. On the other hand, the characterization
with diesel is performed on the range of 35 cm and 75 cm with steps of
10 cm. The range of the diesel characterization is lower than the one of
water due to oil limitations, safety and environmental issues. The dia-
phragm’s characterization with water and diesel is presented in Fig. 16.
Diaphragm 1 presents a sensitivity higher than the one of the Dia-
phragm 2, which can be related to both inaccuracies on the manu-
facturing process of the diaphragm and differences on the FBG em-
bedment. However, both sensitivities are lower than the obtained on
the simulation due to small differences on the diaphragm thickness,
diameter and material Young modulus created at the manufacturing
process.
The root mean squared error (RMSE) between the linear regression
and the measured wavelength shift for Diaphragm 1 is 1.90 pm for the Fig. 17. Measured and reference interface level. The RMSE between the measured and
reference interface level is 2.51 cm.
characterization with water and 1.72 pm for diesel. Whereas,

90
A.G. Leal-Junior et al.
diaphragm fabrication. In addition, by increasing the number of sen-
sors, the error also will be reduced. By increasing the number of sen-
sors, it is possible to gather more information about the interface level
in different points of the tank, which leads to lower errors and higher
resolution of the sensor system.
However, with the two diaphragms setup, the interface level is es-
timated by knowing the density of each fluid. Nevertheless, it is possible
to employ an extra diaphragm to estimate the density that will be ap-
plied on the interface level estimations. In order to demonstrate this
possibility, the diaphragms are applied to estimate the diesel density as
the interface level variation occurs. The density estimation is made by
comparing the sensitivity of the sensors when only water is applied with
the one when there is a variation of the interface. In this case, the in-
terface level is known and the sensors are applied to estimate the
density, which leads to a mean density estimation of 839.2 kg/m3 on
the tests with variation of the interface level. Moreover, the standard
deviation of the density estimation is 46.22 kg/m3. Since the density of
the employed diesel is about 840 kg/m3, the deviation between the
reference and measured density is only 0.8 kg/m3. Therefore, by a first
characterization with water, the sensor system is capable of measuring
the fluid density as well.
The sensors capability to measure both interface level and oil den-
sity points towards the application of a higher number of diaphragms to
measure simultaneously the interface level, total level and oil density.
Furthermore, it is also possible to employ an even higher number of
FBG-embedded diaphragms to measure the emulsion and foam layers.
In addition, such increase on the number of diaphragms can also reduce
the interface level measurement errors.
6. Conclusions and future works
Oil and gas industries face the problem of control of the oil se-
parators. These problems are due to the issues of current technologies
for multi-interface level measurement, which can present high cost,
safety issues, lack of precision on each layer measurement and on the
estimation of foam and emulsion layers. The issues of current tech-
nologies for multi-interface level assessment inhibit the application of
optimized control of the oil separation process and the design of opti-
mized oil separators. If the interface level is measured with reliable
sensors, it is possible to reduce the costs with oil production, increase
the production rates and even reduce the environmental impact of oil
companies.
For these reasons, this paper presented a review of the multi-in-
terface level measurement application and its issues. Furthermore,
conventional techniques employed to measure interface level proposed
throughout the years are presented. In order to overcome the limitation
of conventional techniques, a brief review of techniques based on op-
tical fiber sensors that can be applied on this application is presented
and a proposal of a FBG-based multi-interface level measurement
system is discussed and experimentally demonstrated.
Future works include the simulations of the proposed FBG-based
multi-interface level sensor system to obtain the optimum number of
diaphragms and its design parameters, the evaluation of coating ma-
terials to eliminate the problem of material build-up and the con-
struction of a prototype to assess experimentally the functionality of the
sensor system proposed on an oil separator. Since it can lead to a more
sensitive sensor system [11,56], the application of FBG in polymer
optical fibers for interface level assessment will be investigated.
Acknowledgments
This research is financed by CAPES-PGPTA (88887.123920/2014-
00 and 88887.095626/2015-01), FAPES (72982608), CNPq (304192/
2016-3). C. Marques acknowledges the financial support from FCT
through the fellowship SFRH/BPD/109458/2015. This work was also
funded by FCT/MEC through national funds and when applicable co-
91
Optical Fiber Technology 40 (2018) 82–92
funded by FEDER-PT2020 partnership agreement under the project
UID/EEA/50008/2013.
References
[1] A. Tondon, M. Singh, B.S. Sandhu, B. Singh, A compton scattering technique for
concentration and fluid-fluid interface measurements using NaI(Tl) detector, Nucl.
Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms 403
(2017) 21–27.
[2] S.F.A. Bukhari, W. Yang, Multi-interface level sensors and new development in
monitoring and control of oil separators, Sensors 6 (4) (2006) 380–389.
[3] G. Holler, T. Thurner, H. Zangl, and G. Brasseur, “A novel capacitance sensor
principle applicable for spatially resolving downhole measurements”, in: IMTC/
2002. Proceedings of the 19th IEEE Instrument. Measurement Technology
Conference (IEEE Cat. No.00CH37276), vol. 2, no. May, pp. 1157–1160, 2002.
[4] B.T. Hjertaker, G. A. Johansen, P. Jackson, I. C. I. Synetix, P. O. Box, and C. TS,
“Recent developments in hydrocarbon separator interface imaging,” vol. 4188, pp.
81–92, 200.
[5] D. Langevin, S. Poteau, I. Hénaut, J.F. Argillier, Crude oil emulsion properties and
their application to heavy oil transportation”, Oil Gas Sci. Technol. Rev. IFP 59 (5)
(2004) 511–521.
[6] W.Q. Yang, M.R. Brant, M.S. Beck, A multi-interface level measurement system
using a segmented capacitance sensor for oil separators, Meas. Sci. Technol. 5 (9)
(1999) 1177–1180.
[7] M. Meribout, A. Al Naamany, and K. Al Busaidi, “Interface layers detection in oil
field tanks: a critical review”, in: Expert Systems for Human, Materials and
Automation, InTech, 2011.
[8] N. Edema, Effects of Crude Oil Contaminated Water on the Environment, Prof.
Manar El-Sayed Abdul-Raouf (Ed.), InTech, 2012.
[9] K. Peters, Polymer optical fiber sensors – a review, Smart Mater. Struct. 20 (1)
(2011) 13002.
[10] R. da Silva Marques, A.R. Prado, P.F. da Costa Antunes, P.S. de Brito André,
M.R.N. Ribeiro, A. Frizera-Neto, M.J. Pontes, Corrosion resistant FBG-based quasi-
distributed sensor for crude oil tank dynamic temperature profile monitoring,
Sensors (Switzerland) 15 (12) (2015) 30693–30703.
[11] C.A.F. Marques, G.-D. Peng, D.J. Webb, Highly sensitive liquid level monitoring
system utilizing polymer fiber Bragg gratings, Opt. Express 23 (5) (2015) 6058.
[12] T. Li, Y. Tan, P. Xia, Z. Zhou, Paralleled structure-based string-type fiber bragg
grating acceleration sensor, IEEE Sens. J. 17 (5) (2017) 1325–1332.
[13] J. Zhu, M. Wang, L. Chen, X. Ni, H. Ni, An optical fiber fabry-perot pressure sensor
using corrugated diaphragm and angle polished fiber, Opt. Fiber Technol. 34 (1)
(2017) 42–46.
[14] M. Vidakovic, C. Mccague, I. Armakolas, T. Sun, J. S. Carlton, and K. T. V Grattan,
“Fibre Bragg Grating-based cascaded acoustic sensors for potential marine struc-
tural condition monitoring”, vol. 34, no. 19, pp. 4473–4478, 2016.
[15] M.F. Domingues, P. Antunes, N. Alberto, R. Frias, R.A.S. Ferreira, P. André, Cost
effective refractive index sensor based on optical fiber micro cavities produced by
the catastrophic fuse effect, Meas. J. Int. Meas. Confed. 77 (2016) 265–268.
[16] A.M. Al-Naamany, M. Meribout, K. Al Busaidi, Design and implementation of a new
nonradioactive-based machine for detecting oil-water interfaces in oil tanks, IEEE
Trans. Instrum. Meas. 56 (5) (2007) 1532–1536.
[17] G. Lu, H. Hu, S. Duan, and H. Wang, “A simple and valuable method for detecting
levels of interface of oil -water and oil layer”, pp. 1836–1839, 2009.
[18] T.M. Shi, C.G. Xie, S.M. Huang, R.A. Williams, M.S. Beck, Capacitance-based in-
strumentation for multi-interface level measurement, Meas. Sci. Technol. 2 (10)
(1991) 923–933.
[19] O. Isaksen, S. Dico, E. Hammer, A capacitance-based tomography system for in-
terface measurement in separation vessels, Meas. Sci. Technol. 5 (10) (1999)
1262–1271.
[20] R. P. Lees, “Increase oil production and reduce chemical usage through separator
level measurement by density profiling”, Xx, vol. xx, pp. 1–8.
[21] M.K. Poindexter, N.N. Zaki, P.K. Kilpatrick, S.C. Marsh, D.H. Emmons, Factors
contributing to petroleum foaming. 1. Crude oil systems, Energy Fuels 16 (3) (2002)
700–710.
[22] C. Blázquez, E. Emond, S. Schneider, C. Dalmazzone, V. Bergeron, Non-aqueous and
crude oil foams, Oil Gas Sci. Technol. Rev. 69 (3) (2014) 467–479 d’IFP Energies
Nouv..
[23] M. El-Sayed, A. Raouf, Factors affecting the stability of crude oil emulsions, in:
M. El-Sayed, A. Raouf (Eds.), Crude Oil Emulsions- Composition Stability and
Characterization, InTech, 2012.
[24] S.F. Wong, J.S. Lim, S.S. Dol, Crude oil emulsion: a review on formation, classifi-
cation and stability of water-in-oil emulsions, J. Pet. Sci. Eng. 135 (2015) 498–504.
[25] T.S.T. Ariffin, E. Yahya, H. Husin, The rheology of light crude oil and water-in-oil-
emulsion, Procedia Eng. 148 (2016) 1149–1155.
[26] R. J. Issa and E. M. Hunt, “Rheology of water-in-oil emulsions for a medium crude
oil”, pp. 2–4, 2015.
[27] S. Bandaru, A. Tiwari, A. Bohori, and T. Asokan, “Interface stability of oil-water
system under electric stress”, in: IEEE International Conference Dielectr. Liq. 2005.
ICDL 2005, pp. 36–39.
[28] R. Lakes, Viscoelastic Materials, first ed., Cambridge University Press, New York,
USA, 2009 ch. 1, sec.1.1, pp. 1-3.
[29] N. Skeie, S. Mylvaganam, and B. Lie, “Using multi sensor data fusion for level es-
timation in a separator”, pp. 1383–1388, 2006.
[30] F. Reverter, X. Li, G.C.M. Meijer, Liquid-level measurement system based on a
A.G. Leal-Junior et al.
remote grounded capacitive sensor, Sens. Actuators A Phys. 138 (1) (2007) 1–8.
[31] G. Lu, H. Hu, B. He, and S. Chen, “A new-type sensor for monitoring oil-water
interface level and Oil level”, pp. 981–983, 2009.
[32] R. Casanella, Ó. ́ Casas, and R. Pallàs-Areny, “Oil-water interface level sensor based
on an electrode array”, in: Conference Rec. - IEEE Instrum. Meas. Technol. Conf.,
no. April, pp. 710–713, 2006.
[33] M. Meribout, M. Habli, A. Al-naamany, A New ultrasonic-based device for accurate
measurement, lnsllumenlalion Meas. Technol. Conhence (2004) 1942–1947.
[34] M. Meribout, A. Al Naamany, K. Al Busaidi, An acoustic system for providing the
two-phase liquid profile in oil field storage tanks, IEEE Trans. Ultrason. Ferroelectr.
Freq. Control 56 (10) (2009) 2241–2250.
[35] M.H. Zarifi, M. Rahimi, M. Daneshmand, T. Thundat, Microwave ring resonator-
based non-contact interface sensor for oil sands applications, Sens. Actuators B
Chem. 224 (2015) 632–639.
[36] H. Moghadas, M. Daneshmand, P. Mousavi, A passive non-contact microwave loop
resonance sensor for liquid interface, Sensors and Actuators B 241 (2017) 96–98.
[37] M. Lomer, J. Arrue, C. Jauregui, P. Aiestaran, J. Zubia, J.M. López-Higuera, Lateral
polishing of bends in plastic optical fibres applied to a multipoint liquid-level
measurement sensor, Sens. Actuators A Phys. 137 (1) (2007) 68–73.
[38] C. Li, T. Ning, C. Zhang, X. Wen, J. Li, C. Zhang, Liquid level and temperature sensor
based on an asymmetrical fiber Mach-Zehnder interferometer combined with a fiber
Bragg grating, Opt. Commun. 372 (2016) 196–200.
[39] Z. Yazdizadeh, H. Marzouk, M.A. Hadianfard, Monitoring of concrete shrinkage and
creep using fiber bragg grating sensors, Constr. Build. Mater. 137 (2017) 505–512.
[40] T. Guo, F. Liu, B. Guan, J. Albert, Tilted fiber grating mechanical and biochemical
sensors, Opt. Laser Technol. 78 (2016) 19–33.
[41] R. Slavik, J. Homola, J. Ctyroky, Single-mode optical fiber surface plasmon re-
sonance sensor, Sens. Actuators B Chem. 54 (1) (1999) 74–79.
[42] V. Mishra, N. Singh, U. Tiwari, P. Kapur, Fiber grating sensors in medicine: current
and emerging applications, Sens. Actuators A Phys. 167 (2) (2011) 279–290.
[43] B. Culshaw, “Optical fibre sensors for industrial applications in safety and security”,
Optics InfoBase Conference Paper, pp. 1–3, 2013.
[44] G. Betta, A. Pietrosanto, A. Scaglione, Microcontroller-based performance en-
hancement of an optical fiber level transducer, IEEE Trans. Instrum. Meas. 47 (2)
(1998) 489–493.
[45] A. Owen, G. Duckworth, and J. Worsley, “OptaSense: fibre optic distributed
acoustic sensing for border monitoring”, in: Proc. - 2012 Eur. Intell. Secur.
Informatics Conf. EISIC 2012, pp. 362–364, 2012.
[46] Y. Zhang, H. Peng, X. Qian, Y. Zhang, G. An, Y. Zhao, Recent advancements in
optical fiber hydrogen sensors, Sen. Actuators B Chem. 244 (2017) 393–416.
[47] A. Vallan, M.L. Casalicchio, M. Olivero, G. Perrone, Assessment of a dual-wave-
length compensation technique for displacement sensors using plastic optical fibers,
92
Optical Fiber Technology 40 (2018) 82–92
Instrum. Meas. IEEE Trans. 61 (5) (2012) 1377–1383.
[48] E.E. Havinga, The temperature dependence of dielectric constants, J. Phys. Chem.
Solids Pergamon Press 18 (213) (1961) 253–255.
[49] A. Ukil, H. Braendle, P. Krippner, Distributed temperature sensing: review of
technology and applications, IEEE Sens. J. 12 (5) (2012) 885–892.
[50] T. Zhu, T. Ke, Y. Rao, K.S. Chiang, Fabry-perot optical fiber tip sensor for high
temperature measurement, Opt. Commun. 283 (19) (2010) 3683–3685.
[51] T. Li, C. Shi, Y. Tan, R. Li, Z. Zhou, and H. Ren, “A diaphragm type fiber bragg
grating vibration sensor based on transverse property of optical fiber with tem-
perature compensation”, vol. 17, no. 4, pp. 1021–1029, 2017.
[52] P.F.C. Antunes, H. Varum, P.S. Andre, Intensity-encoded polymer optical fiber ac-
celerometer, IEEE Sens. J. 13 (5) (2013) 1716–1720.
[53] V. Neeharika, P.K. Pattnaik, Optical MEMS pressure sensors incorporating dual
waveguide bragg gratings on diaphragms, IEEE Sens. J. 16 (3) (2016) 681–687.
[54] H. El Ghandoor, E. Hegazi, I. Nasser, G. Behery, Measuring the refractive index of
crude oil using a capillary tube interferometer, Opt. Laser Technol. 35 (5) (2003)
361–367.
[55] C. E. S. Castellani, H. C. B. Ximenes, R. L. Silva, A. Frizera-neto, M. R. N. Ribeiro,
and M. J. Pontes, “Multi-parameter interferometric sensor based on a reduced
diameter core axial offseted fiber”, vol. 29, no. 2, pp. 239–242, 2017.
[56] C.A.F. Marques, A. Pospori, D. Saez-Rodriguez, K. Nielsen, O. Bang, D.J. Webb,
Aviation fuel gauging sensor utilizing multiple diaphragm sensors incorporating
polymer optical fiber bragg gratings, IEEE Sens. J. 16 (15) (2016) 6122–6129.
[57] C.W. Lai, Y.L. Lo, J.P. Yur, C.H. Chuang, Application of fiber bragg grating level
sensor and Fabry-Perot pressure sensor to simultaneous measurement of liquid level
and specific gravity, IEEE Sens. J. 12 (4) (2012) 827–831.
[58] K.O. Hill, G. Meltz, Fiber bragg grating technology fundamentals and overview,
IEEE J. Light. Technol. 15 (8) (1997) 1263–1276.
[59] L. Alwis, T. Sun, K.T.V. Grattan, [INVITED] Developments in optical fibre sensors
for industrial applications, Opt. Laser Technol. 78 (2016) 62–66.
[60] E.A. Al-Fakih, N.A. Abu Osman, F.R. Mahamd Adikan, A. Eshraghi, P. Jahanshahi,
Development and validation of fiber bragg grating sensing pad for interface pres-
sure measurements within prosthetic sockets, IEEE Sens. J. 16 (4) (2016) 965–974.
[61] O.F. Ameen, M.H. Younus, M.S. Aziz, A.I. Azmi, R.K. Raja Ibrahim, S.K. Ghoshal,
Graphene diaphragm integrated FBG sensors for simultaneous measurement of
water level and temperature, Sensors Actuators A Phys. 252 (2016) 225–232.
[62] M.F. Ashby, Materials Selection in Mechanical Design, Design (2005) 624.
[63] P. Antunes, F. Domingues, M. Granada, P. André, Mechanical Properties of Optical
Fibers, INTECH Open Access Publ., 2012 pp. 1–15.
[64] C. Schaschke, I. Fletcher, N. Glen, Density and viscosity measurement of diesel fuels
at combined high pressure and elevated temperature, Processes 1 (2) (2013) 30–48.