Sensors and Actuators B 210 (2015) 602–617

Contents lists available at ScienceDirect

Sensors and Actuators B: Chemical

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

Non-invasive process tomography in chemical mixtures – A review

Yasmin Abdul Wahab a , Ruzairi Abdul Rahim b,∗ , Mohd Hafiz Fazalul Rahiman c ,
Suzanna Ridzuan Aw d , Fazlul Rahman Mohd Yunus b , Chiew Loon Goh b ,
Herlina Abdul Rahim b , Leow Pei Ling b
a
Department of Instrumentation & Control Engineering (ICE), Faculty of Electrical & Electronics Engineering, Universiti Malaysia Pahang, 26600 Pekan,
Pahang, Malaysia
b
Process Tomography and Instrumentation Engineering Research Group (PROTOM-i), INFOCOMM Research Alliance, Faculty of Electrical Engineering,
Universiti Teknologi Malaysia, 81310 UTM Skudai, Johor, Malaysia
c
Tomography Imaging Research Group, School of Mechatronic Engineering, Universiti Malaysia Perlis, Pauh Putra Campus, 02600 Arau, Perlis, Malaysia
d
Faculty of Electrical & Automation Engineering Technology, Terengganu Advance Technical Institute University College (TATiUC), Jalan Panchor, Telok
Kalong, 24000 Kemaman, Terengganu, Malaysia

a r t i c l e i n f o a b s t r a c t

Article history: Decades of research, visualizing and monitoring chemical mixtures in the process plant non-invasively
Received 7 September 2014 plays an important role in making sure that the system produces a high quality product without distur-
Received in revised form bing the nature of the plant. In this case, process tomography can be applied for the chemical mixtures.
21 December 2014
Thus, the objective of this paper is to review the current research on non-invasive process tomogra-
Accepted 24 December 2014
phy for detecting chemical mixtures based on hard-field and soft-field tomography. The discussion on
Available online 17 January 2015
liquid–liquid, liquid–solid and liquid–gas two-phase flow for every type of non-invasive process tomo-
graphy is explained in detail. Consequently, non-invasive process tomography has expanded strongly
Keywords:
Non-invasive
in chemical applications. Finally, potential future research on the chemical mixture based on ERT is
Process tomography addressed.
Chemical mixtures © 2015 Elsevier B.V. All rights reserved.
Hard-field tomography
Soft-field tomography

1. Introduction Tomography (IPT) or simply as Process Tomography (PT) [3].

A chemical mixture such as liquid–liquid, liquid–solid and
liquid–gas two-phase flow regime is the main concern in chem-
ical applications. Visualization at an early stage of the chemical
mixtures in a chemical application, such as in a vertical or hor-
izontal pipeline, promises a good performance and prevents any
unwanted condition in the process plant. One of the methods that
can be implemented in chemical applications is process tomo-
graphy. Originally, the term ‘tomography’ comes from the Greek
words ‘tomos’ meaning a section and ‘graphy’ meaning an image
[1]. Simply, the tool of tomography is called the ‘tomograph’ and
the reconstructed image is a ‘tomogram’. Tomography can also
be defined as a process to reconstruct the image of an object
of interest within the sensing zone [2]. Since the 1970s, tomo-
graphy has been widely employed in medical applications. After
decades of research, tomography is also implemented in indus-
trial applications, when it is commonly known as Industrial Process
∗ Corresponding author. Tel.: +60 197104000; fax: +60 75566177.
E-mail address: ruzairi@fke.utm.my (R. Abdul Rahim).
Due to its main advantage of being non-destructive to the pro-
cess or object being measured, process tomography has become
of widespread interest to researchers investigating its industrial
application.
Process tomography can be divided into three main parts; the
sensing system, interfacing and an image reconstruction algorithm
for displaying the tomogram, as shown in Fig. 1. The sensing sys-
tem includes the sensor, measurement circuits for transmitting and
receiving a signal, and a signal conditioning circuit to amplify and
filter the signal before it is used for the interfacing part. The inter-
facing part refers to the data acquisition system (DAS). Then, in the
image reconstruction algorithm, it can be divided into two parts:
a forward problem and an inverse problem. The forward problem
is solved first in order to know the theoretical value of the sen-
sor output based on the signal projection, whilst the algorithm,
such as a non-iterative algorithm and iterative algorithm, is solved
later in the inverse problem. The types of process tomography are
based on the sensors used in measuring the medium of interest. As
a result, the different types of process tomography work based on
the different principles of sensors involved and implemented in the
applications.

http://dx.doi.org/10.1016/j.snb.2014.12.103
0925-4005/© 2015 Elsevier B.V. All rights reserved.
 Y. Abdul Wahab et al. / Sensors and Actuators B 210 (2015) 602–617
Fig. 1. General system configuration of process tomography.
Thus, a review of the several types of non-invasive process
tomography for chemical mixtures developed within the last ten
years of research is the main purpose of this paper. The types
of non-invasive process tomography in the subtopics can be
divided into soft-field and hard-field tomography. Lastly, the pro-
posed improvement of current non-invasive process tomography
is briefly discussed in this paper.
2. Non-invasive process tomography
Process tomography consists of several sensing method
approaches which are invasive, non-invasive, intrusive and non-
intrusive measurements. The researchers should decide carefully
which method they are going to use based on the principles of
process tomography chosen. The term ‘invasive’ means the sen-
sor is applied to the inner surface of the wall of the pipeline, while
‘intrusive’ relates to how the sensor protrudes into the medium
of interest. Thus, the non-invasive technique means the sensor
only mounted on the pipe wall without making any hole through
the pipe. Despite, the non-intrusive technique will not necessary
placing the sensor outside the pipe wall; it also can be invasive,
but the sensor element should not protrudes into the vessel pipe,
if not the technique is called as intrusive technique [4,5]. There-
fore, the non-invasive and non-intrusive should be differentiated
clearly. These concepts are represented in detail in Fig. 2. In the past
decade, developments of the non-invasive technique promise to be
very significant for process plants and the chemical industry. The
non-invasive method is beneficial for the industry compared to the
non-intrusive method because of several advantages:
1. Minimizing the safety and accountancy difficulties with valuable
process materials.
2. Assisting installation (and even retrofitting) and looking after the
instruments even when the plant is on-stream.
3. Avoiding contamination of pure or sterile materials.
4. Decreasing the hazards of working with poisonous, radioactive,
explosive, flammable or corrosive materials.
Therefore, the non-invasive method is one of the favourite
methods applied in process plants compared to other sensing tech-
niques.
3. Non-invasive hard-field tomography
Hard-field tomography refers to the condition where the sen-
sitivity of the medium imaged is independent of the distribution
of the measured parameter in the whole volume [6]. X-ray
tomography, optical tomography, and ultrasonic tomography are
603
examples of hard-field tomography. Non-invasive hard-field tomo-
graphy is reviewed from year 2004 till the present.
3.1. X-ray computed tomography
X-ray tomography was the first type of tomography introduced
in medical applications during the 1970s. It then also expanded
into the process industry where it is now established. X-ray tomo-
graphy reconstructs the image based on the attenuation of the
X-rays. X-ray tomography is widely used in the chemical mixture
for detecting liquid–gas phase flow compared to other types of
two-phase flow. An investigation of liquid–gas in a large bubble col-
umn (approximately 32 cm in diameter) has been reported by [7,8].
Hubers et al. [7], for instance, presented an investigation of X-ray
tomography for a bubble column with diameter 32 cm and length
of almost 4 m (see Fig. 3). Tubes of different sizes and materials
were applied for the study, such as acrylic tubes (6.35 mm, 9.53 mm,
12.27 mm, and 12.6 mm), Teflon (6.35 mm) and PVC tubes (8.99 mm
and 12.7 mm). Based on the filtered back projection, by compar-
ing the real position of the tubes in the column with the image
obtained; it is observed that the system is capable of visualizing
the water–gas in a large bubble column.
Moreover, the research on X-ray tomography also focuses on
improving the quality of the reconstructed image just as in the
applied cross-correlation method [9] and algebraic reconstruc-
tion technique (ART) [10]. Moreover, X-ray tomography is also
compared with a wire mesh sensor as two high-speed imaging
modalities. The wire mesh sensor can provide a phase fraction
distribution in the cross-section of the flow. However, the wire
mesh sensor is using the intrusive and invasive technique. Thus,
modality X-ray tomography with wire mesh is not non-invasive
technique. However, dual-modality with the combination of X-
ray tomography and wire mesh sensor was found to produce a
better-reconstructed image compared to a single modality X-ray
tomography, as in [11,12]. The study concerned measuring the
water–gas flow in a vertical pipeline, as in Fig. 4.
Boden et al. [13] presented the implementation of X-ray cone
beam tomography (CBCT) in chemicals in a stirred tank reactor. The
schematic diagram of CBCT, photograph of stirred tank reactor and
the reconstructed image are illustrated in Fig. 5(a)–(c). The dimen-
sion of stirred tank reactor in the research was 0.008 m thickness
with an inner diameter of 0.08 m. The material of the stirred tank
reactor was borosilicate glass. It showed a different image when
the gas velocity increased from 1000 min−1 to 1200 min−1 .
Even though X-ray tomography is useful in producing high
resolution of the images reconstructed, the subject of hardware
development is not discussed in detail. In addition, the technique
604 Y. Abdul Wahab et al. / Sensors and Actuators B 210 (2015) 602–617

Fig. 2. Types of sensing techniques.

needs a specialist to operate the system. Also, the main disadvan-
tages of X-ray tomography are its expense and the need for major
safety precautions related with radiation issues.
3.2. Optical tomography
By contrast with X-ray tomography, optical tomography
(OT) is mostly applied in non-medical applications investigating
liquid–gas phase flow detection compared to liquid–liquid and
liquid–solid measurement. Optical tomography reconstructs the
image based on optical radiation [14], transmitting a beam of light
from one boundary and receiving the beam of light from another
boundary. Optical radiation can be obtained by using different kinds
of the optical sensor, such as laser, linear CCD, LED, etc. Schleicher
et al. [15] proposed bubble flow detection in a low gas fraction in
a bubble column (acrylic pipe with an outer diameter 280 mm and
inner diameter 50 mm). The system implements LED as the optical
transmitter and optical fibre as the receiver (see Fig. 6). The mini-
mum size of the bubble detected in this paper was 0.1 mm, and the
maximum size was 10.3 mm. It is observed that, based on the back
projection algorithm, the system is able to improve the quality of
the image of bubble flow in the bubble column.
Correspondingly, Yunos et al. [16] implemented a linear CCD
sensor as an optical sensor for measuring bubble flow in a vertical
pipeline. The linear CCD sensor applied is ILX503A which contains
2048 sensors (effective pixels). It is observed that, the ILX503A
limits the bubble velocity up to 0.04 m/s when the sampling is at
1 MHz. Besides, research on improving the reconstructed image

Fig. 3. X-ray tomography system with different tube sizes as gas phantom [7].
 Y. Abdul Wahab et al. / Sensors and Actuators B 210 (2015) 602–617 605

Fig. 4. Location of X-ray tomography and wire mesh in vertical pipeline and the reconstructed image [12].

Fig. 5. (a) The schematic X-ray cone beam computed tomography; (b) photograph of stirred tank reactor; (c) gas distribution in stirred tank reactor with increasing gas
velocity [13].
606 Y. Abdul Wahab et al. / Sensors and Actuators B 210 (2015) 602–617

Fig. 6. (a) The basic design of the OT for horizontal view (left) and sectional view (right); (b) optical sensor configuration for optical tomography. Left: tomography from
outside; right: view inside the tomography sensor from above [15].

of the liquid–gas phase flow was introduced by using a hybrid

back projection algorithm in [17]. Fig. 7 shows the optical tomo-
graphy configuration for [17]. The authors in Ref. [17] used four
dichroic halogen bulbs acting as light projectors for two planes
of optical fibre receivers. Every plane consists of 8 × 8 arrays of
orthogonal and 11 × 11 array of rectilinear projections. Here, the
minimum size of gas that could be detected by the system was
20 mm. However, in 2013, Mohd Fadzil et al. [18] proposed opti-
cal tomography for a vertical bubble column (130 mm in outer
diameter and 1.2 m in length). Here, sixteen laser pointers and six-
teen photodiode model EPD-660-5 were chosen as transmitters and
receivers, respectively. The study discussed the detail of the hard-
ware development needed to ensure that the OT system is capable
of detecting the bubble flow in the bubble column. However, the
limitation of optical tomography relates to the limitation in its opti-
cal surface; the sensors cannot be placed too close to each other so
as to avoid reflection, and can just detect the image in between the
area of the optical sensors.

3.3. Ultrasonic tomography

Ultrasonic tomography (UT) works based on the interaction

between the ultrasonic wave and the medium measured. The capa-
bility of ultrasonic sensors depends on the frequency of interest and
also the beam angle of the sensors. High frequency and a large beam
angle influence the result of the reconstructed image of UT because
the acoustic impedance of the UT is increased, and the area covered
by each of the sensors is high enough to cover the projected area.
Besides, the number of sensors used also affects the reconstructed
image. If the number of sensors around the pipeline is increased, the
image will be more accurate because the projection signal covers Fig. 7. Optical tomography configuration for [17]. Four dichroic halogen bulbs act
as light projectors and thirty-eight optical fibre as receivers (two orthogonal and
most of the area inside the pipeline. In ultrasonic tomography, the two rectilinear projections).
received voltage received from the ultrasonic receiver is processed
 Y. Abdul Wahab et al. / Sensors and Actuators B 210 (2015) 602–617
Fig. 8. Cross-section view of UT system with example of oil–water flow in the pipeline [20].
to reconstruct the image of the medium of interest. Recently, UT has
also been applied in chemical application for detecting chemical
mixtures such as liquid–liquid, liquid–solid, and liquid–gas.
Abdul Rahim et al. [19] presented a principle study on trans-
mission ultrasonic tomography for oil–water phase flow using four
pairs of ultrasonic sensors in a horizontal pipeline. The initial results
of the average output voltage for each of receivers were plotted and
it is concluded that the average voltage drops when the percentage
of oil in the pipeline rose. Correspondingly, an experiment with
16 pairs of ultrasonic sensors and parallel projection with differ-
ent kinds of oil such as corn oil, soybean oil and palm oil in water
flow was discussed by Fazalul Rahiman et al. [20]. The linear back
projection (LBP) algorithm was used for the purpose of getting the
tomogram as shown in Fig. 8. The paper also explained detail the
implementation of LBP algorithm to the system; including the cal-
culation of the percentage error between the real condition and
the measurement sensor values for each of the water and oil com-
position. Analysis of the real, measurement sensor and percentage
error is shown on the right side for each of tomogram in Fig. 8.
Thus, the accuracy of the tomogram can be obtained based on that
607
percentage measurement. Moreover, Steiner et al. [21] applied UT
for detecting oil–water. Parallel projection with 16 pairs of ultra-
sonic sensors is used in the study. The study focused on time-of-
flight method applied in UT system. It is reported that, the method
able to determine the oil–water interfaces in the oil production.
Additionally, the work by Wockel et al. [22] proposed the detec-
tion of interfaces between a phase flow of solid acryl and liquid
kernel layers. The 2D ultrasonic propagation in the vessel was
modelled by using a finite element model (FEM) in COMSOL Mul-
tiphysics software. The research implemented a fast estimation
algorithm for the reconstructed image, and it was observed that
the system produced a high quality tomogram of the liquid–solid
phase flow. Moreover, the detection of water-particles in a verti-
cal PVC pipe was then investigated in [23]. Different sizes of PVC
plastic rod assumed as particles (12 mm, 26 mm, and 60 mm) were
used in the study to work out the smallest sizes of particles that can
be detected by the UT system. The result showed that the detection
of the smallest sizes of particles dropped depends on the ultra-
sonic wavelength; if the wavelength could be blocked by the object,
information on the sensing area could be determined. Muhamad
608 Y. Abdul Wahab et al. / Sensors and Actuators B 210 (2015) 602–617
Fig. 9. (a) UT experimental setup; (b) bubble column with attached baffle (half-blocked); (c) gas hold-up distribution in the bubble column with the attached baffle for
half-blocked at different heights: (a) 80 cm, (b) 55 cm and (c) 35 cm [26].
et al. [24] employed UT in investigating the water–solid flow regime
in a vertical pipeline. By using four pairs of ultrasonic sensors in fan
beam geometry, the UT system could be used to detect the solid
medium in the vessel.
In the work by Zakaria et al. [25], the competence of different
numbers of sensors in the projection for water–gas phase flow in
a horizontal pipe is discussed in detail. A transceiver type of ultra-
sonic sensor which can function as a transmitter or receiver was
used in the study. The advantage of using a transceiver in the UT
system is that fewer sensors are needed to produce the same qual-
ity as separate ultrasonic sensors. Simulations using 8, 16 and 32
transceivers at 40 kHz were compared and it was concluded that
the 32-transceiver set-up was the optimum number of sensors for
producing a good reconstructed image of the water–gas. Supar-
dan et al. [26] delineated the pattern of gas hold-up distribution
in a bubble column. In the research, the column applied was a
methyl-methacrylate material with 16 cm in diameter and 200 cm
in height. Fig. 9 shows the detail of the UT system for [26] with the
image reconstructed at different heights of the bubble column. The
research was deployed three parallel ultrasonic transducer arrays,
and each array had six pairs of ultrasonic sensors in parallel geom-
etry. Simultaneously, each array was arranged non-symmetrically,
and rotated at 20◦ each. One of the example experiments done is
by attaching the baffle with half blocked to the column as shown in
Fig. 9 (b). The gas sensed by the UT system was between 5.8 mm and
6.1 mm in mean diameter for 2 MHz frequency. The photograph of
the bubble column was captured to validate the tomogram of gas
hold-up in the column.
The reflection mode of ultrasonic tomography system applied
for liquid–gas detection in an insulating pipe with fan beam geome-
try was investigated by Steinar et al. [5,21]. Furthermore, ultrasonic
tomography based on 32 transceivers at high frequency (333 kHz,
125◦ beam angle) for detecting gas hold-up in a bubble column
progressively was studied by Fazalul Rahiman et al. [27–30]. The
 Y. Abdul Wahab et al. / Sensors and Actuators B 210 (2015) 602–617
Fig. 10. Dual plane of UT system for bubble column [33].
novelty of the adjacent criterion method in the inverse problem
was proposed in the study and it proved that the smallest size of
bubble that can be sensed by the bubble column is 2.84 mm. How-
ever, Nor Ayob et al. [31–33] improved the study on detecting the
gas size in the bubble column. The minimum size of gas detected in
the bubble column was still smaller than in the research by Faza-
lul Rahiman et al., namely 2.5 mm. Otherwise, the UT system used
333 kHz and 16 pairs of ultrasonic sensors but in a dual plane by
using the cross-correlation method. This is shown in Fig. 10.
Nonetheless, UT gives advantages in detecting chemical mix-
tures, and the sensors for ultrasonic tomography are readily
available from manufacturers. However, it will require an expen-
sive cost if the number of ultrasonic sensors used is increasing (i.e.,
32 pairs and above) for high resolution of the image reconstructed.
4. Non-invasive soft-field tomography
Soft-field tomography means that the sensitivity of the medium
imaged depends on the distribution of the measured parameter
in the whole volume [34]. Alternatively, it will cause a challenge
in solving the inverse problem of the medium of interest. Electri-
cal tomography can be categorized as a soft-field tomography and
variants include electrical capacitance tomography (ECT), electrical
resistance tomography (ERT) and electrical impedance tomogra-
phy (EIT). However, only non-invasive soft field tomography is
reviewed here, which are the ECT and non-invasive ERT methods.
4.1. Electrical capacitance tomography
Electrical capacitance tomography (ECT) is the first type of
tomography that evolved in industry from the 1980s. ECT recon-
structs the image based on the permittivity distribution of the
medium of interest [35]. In other words, ECT is also known as a
dielectric system. Essentially, ECT is applied non-invasively to the
system. The conductive electrodes are mounted on the periphery
of the pipeline or vessel, and the shielding is necessary to elim-
inate the external electrical interferences and also to protect the
electrodes from damage as illustrated in Fig. 11 [36]. The
609
Fig. 11. Example of ECT sensor design with eight electrodes [36].
Fig. 12. Image reconstructed for water/polyethylene particles with different distri-
butions [38].
permittivity of the medium can be investigated; the capacitance
value between adjacent electrodes in terms of different potential
units can be measured. Later, the image is reconstructed based on
the capacitance data and the algorithm applied.
ECT is used in chemical mixtures for detecting liquid–liquid,
liquid–solid and liquid–gas. For example, Chen and Han [37]
discussed in detail the improvement of high speed, high-image
resolution, with little measurement error for oil–water two-phase
flow in ECT. Here, the authors focused on development of the
image data capture system (IDCS) for better ECT system perfor-
mance. However, application of ECT in detecting liquid–liquid flow
in chemical mixtures was not much investigated by the researcher.
Besides, Cao et al. [38] presented the development of ECT for
liquid–solid two-phase flow. Water and polyethylene particles
were placed in different types of flow in the pipeline, such as
annular flow, core flow and laminar flow. The paper detailed the
hardware development for liquid–solid flow in a vertical insulat-
ing pipeline with a frequency of 500 kHz and implemented digital
signal processing (DSP). Based on the result, shown in Fig. 12, the
system was capable of getting a good image of the medium of inter-
est using linear back projection (LBP), the Landweber method and
Newton algorithm. The image colour from blue to red indicates that
the permittivity of the material increasing; the blue and red were
610 Y. Abdul Wahab et al. / Sensors and Actuators B 210 (2015) 602–617
Fig. 13. (a) Experiment setup for ECT in micro-channel; (b) image and (c) close-up image of the diamond-shaped groove of the cross-sectional micro-channel; (d) schematic
diagram of the cross-sectional microchannel [39].
for minimum and maximum permittivity, respectively. However,
it can be observed that, the image reconstructed by LBP is showing
closer to the real image in the pipeline compared to the iterative
algorithms such as Landweber and Newton algorithm. Besides, the
multiple trickle system with different position of different num-
ber of organic glass rod in the pipeline was also tested in Ref. [38]
for the purpose of comparing the effectiveness of the algorithms.
Thus, it can be concluded that LBP algorithm is more suitable for a
simple system while iterative algorithm is suitable for the multiple
trickle system. In 2013, the research on ECT has been improved for
measuring the capacitance of micro-particles in deionized water
for a micro-channel with different particle diameters used, such
as 1.3 mm, 1.5 mm, and 2.1 mm, as in Ref. [39]. A quartz glass
diamond shaped micro-channel in 800 ␮m (height), and 700 ␮m
(width) was used for the purpose of the study. Also, the size for
every twelve platinum electrodes applied for the system were
200 ␮m (length) × 0.25 ␮m (height). In micro-channel research,
due to the small sizes of particles, the frequency for the ECT system
is 12.5 MHz. Fig. 13(a)–(c) shows the detail of experiment set-up,
image reconstructed of the cross-sectional micro-channel, and a
schematic diagram for the ECT system in the micro-channel. It is
clearly seen that the frequency used in a large diameter pipeline
is low compared to a small diameter pipeline. Thus, the sizes of
pipeline and the medium of interest will influence the frequency
used in ECT system.
In reviewing the application of ECT for chemical mixtures,
liquid–gas two-phase flow emerges as a primary concern for
researchers compared to liquid–liquid and liquid–solid two-phase
flow. Different approaches and improvements have been proposed
by researchers for detecting water–gas flow in chemical applica-
tions, such as Yang et al. [40] and Yu and Deyun [41]. In addition,
Teixeira and Fan [42], Al-Masry et al. [43] and Banasiak et al. [44]
presented 3D ECT for detecting liquid–gas flow in the pipeline. In
particular, glycol–gas detection in horizontal and vertical pipelines
in 3D, for instance, were compared with the equivalent image from
CCD camcorder view, as shown in Fig. 14 [44]. In this Ref. [44],
due to the high density of the using mesh in 3D ECT which is
around 111 k tetrahedrons for forward problem and 80 k tetra-
hedrons for the inverse problem; the 3D images were performed
by using GPU NVidia CUDA acceleration (CUDA technology). It
is promising a useful method of detecting liquid–gas two-phase
flow in chemical uses because it produces precise and more accu-
rate conditions of the medium in the pipeline compared to a 2D
image.
Instead of water–gas detection, researchers are also concerned
with oil–gas detection using the ECT system especially for the oil
and gas industry. Gamio et al. [45] presented oil–gas detection
method for a pressurized pipeline. An insulating pipe with 12 elec-
trodes mounted on the periphery of a 50 mm diameter horizontal
pipeline was used for that purpose. Here, the gas velocity was
from 0 to 16.5 m/s and linear back projection was used as a recon-
struction algorithm. Omit to focus on the hardware development,
besides, some researchers also focus on improvement of the ECT
image based on the inverse problem, such as a hybrid algorithm
[46], simulated annealing for metallic pipes [47] and fuzzy logic
[48]. Xie et al. [48] applied the diesel fuel flow rate (0–15 m3 /h)
and air flow rate (0–30 m3 /h) for a horizontal pipe, as shown in
Fig. 15.
In 2012, Mohamad et al. [49] proposed a new design of ECT
sensor system that is segmented and portable for two-phase flow
such as water–gas, palm oil–gas and water–palm oil detection.
Fig. 16 shows the portable ECT system and examples of the image
reconstructed. The ECT system offers industry a more flexible mea-
surement for two-phase flow.
Oil-gas detection in a small diameter (26 mm) pipeline was
investigated by Zhao et al. [50] who focused on the high viscos-
ity effect on the characteristics of oil–gas with the oil velocities
(0.06 m/s–0.5 m/s) and gas velocity (0.3 m/s–12.0 m/s). Moreover,
Zhang et al. [51] investigated dual modality tomography (ECT-
gamma source) for a horizontal acrylic pipe in measuring the
oil–gas phase flow. The different size of gases tested at differ-
ent locations in the pipeline in the research was 8 mm, 12 mm,
16 mm, 22 mm, and 29 mm. It is observed that the combination of
dual modality tomography can improve the image reconstructed
for the system. However, ECT systems have a lot of advantages
when analysing the chemical mixtures when the main concern
of the study is an insulating medium rather than a conductive
medium such as oil, gas and particles. On the other hand, if the main
 Y. Abdul Wahab et al. / Sensors and Actuators B 210 (2015) 602–617
Fig. 14. Reconstructed image of 3D ECT system of propylene glycol-gas in horizontal and vertical pipeline (40 mm in diameter) with equivalent CCD camcorder view [44].
Fig. 15. The ECT system and example of the image reconstructed for homogeneous
flow (a), stratified flow (b) and annular flow (c) [48].
concern is a conductive medium such as water, another type of
process tomography should be used.
4.2. Electrical resistance tomography
Conventional Electrical Resistance Tomography (ERT) uses
an invasive but non-intrusive sensing approach. The invasive
611
technique is used to make sure that there is continuous contact
between the electrodes and main fluids so that current can be con-
ducted through the medium of interest [52]. Later, the resistance of
the medium can be determined, and the electrical conductivity dis-
tribution of the medium interest can be reconstructed. Table 1 is a
summary of recent research on conventional ERT systems [53–76]
in the last 10 years applied for chemical mixtures with different
types of chemical mixtures.
Nevertheless, the conventional ERT electrode has a direct con-
tact with a fluid. In that case, it will cause the electrochemical
erosion effect and polarization effect to the sensor and may cause
unpredictable measurement error to the system. This limitation
of conventional ERT system limits the practical application in the
industry plant. It follows that in 2010; a group from Tianjin Uni-
versity proposed a non-invasive ERT system for chemical mixtures.
It means that the new approach of ERT system does not have a
direct contact with the fluid. The main Cao et al. [77] started with
sixteen electrodes applied on the horizontal acrylic pipeline, and
the image between water–gas with a stratified flow gave a good
result (as shown in Fig. 17). Also, the sensor configuration was sim-
ilar to an ECT system. This means that the image reconstruction for
non-invasive ERT is similar to that achieved with a reconstruction
algorithm in ECT. The image reconstructed was compared between
linear back projection and Lavrentiev regularization. However, the
first non-invasive ERT just applied an impedance analyzer (Agilent
4294A) as a source for the exciting electrode and signal condition-
ing process with a frequency of 1 MHz; no detail was given on the
development of the hardware part of the system.
Later, in 2012, the group from Zhejiang University, Wang
et al. [78–80], improved the research by designing and developing
twelve segmented electrodes for non-invasive ERT. The study was
based on capacitively coupled contactless conductivity detection
(C4 D). Phase-sensitive demodulation (PSD) is a main component
in the AC circuit used for the study. A voltage signal is used as the
612 Y. Abdul Wahab et al. / Sensors and Actuators B 210 (2015) 602–617

Table 1
Summary of conventional ERT for chemical mixtures.

Types of Authors Application

chemical
mixtures

[53] Water–oil flow in horizontal pipe

[54] Water–oil drop for wet logging
Liquid–liquid
[55] Water – 200 ml brine of concentration 10 g/l in
stirred vessel
[56] Dichloride methane–water and toluene–water
in plant reactor

[59] Particle distribution in the micro-channel

[60,61] Percentage of total solid content milk for milk
powder production
Liquid–solid [62,63] Percentage of total solid content milk for milk
powder production but in 3D
[64] Particle size glass beads (210–1500 mm) –
solid concentration tap water (5–30 wt%) in
agitated tank
[65] Liquid–solid circulating fluidized beds (LSCFB)
[glass beads particles (spherical shape,
2500 kg/m3 & irregular shape, 2210 kg/m3 )]
[66] Solid–liquid in wet particulate processing
[67] Pulverized kaolin clay and water

[68] Gas dispersion performance of an aerofoil

impeller and a standard Rushton turbine (pure
water – 8 L/min gas flow rate)
[69] Characterization of high concentration ionic
bubble column (homogeneous regime at
Liquid–gas 0.85 cm/s, heterogeneous regime at 2.5 cm/s
slug flow regime at 4.2 cm/s)
[70] Mixing in air–water flow (air-suspension flow:
3.0 m/s, 0.11 m/s, 5.0 m/s)
[71] Bubble distribution in bubble column for
volume fraction (gas flow rate: 0.04 m/s,
0.07 m/s, 0.09 m/s, 0.11 m/s, 0.13 m/s, 0.16 m/s,
0.18 m/s)
[72] Gas–liquid slug vertical pipe flow with data
fusion from electromagnetic flow meter
[73] 3D gas holdup in the flotation cell (gas flow
rate: 0.20 cm/s, 0.67 cm/s)
[74] Performance of a counter-current bubble
column (gas flow rate: 0.28 cm/s, 0.56 cm/s,
0.84 cm/s, 1.12 cm/s)
[75] Slug flow of horizontal pipe validated with CCD
[76] Flow rate measurement of air–water
[77] Bubble velocity in swirling flow (water–gas
velocity:0.63 m/s/, 0.065 m/s,0.037 m/s,
0.41 m/s/0.068 m/s)
Fig. 16. The portable ECT system and example of the image reconstructed for [78] Liquid–gas velocity in vertical upward pipe by
water–gas, palm oil–gas, and water–palm oil [49]. using cross correlation

excitation source and the current signal is measured on the detec-

tion electrodes. Fig. 18 illustrates the sensor configuration and the
image reconstructed in annular flow. The optimum sensor con-
figuration was modelled based on a finite element model (FEM)
in COMSOL Multiphysics software. The latest study shows that it
was implemented for detecting liquid–solid (water–plastic rod as
a solid medium) in a vertical insulating pipeline. The frequency of
100 kHz applied in the non-invasive ERT system was much lower
than in the first non-invasive ERT system. ERT is one type of pro-
cess tomography that offers advantages to the system when the
primary medium is a conductive medium rather than an insulating
medium.
Based on Refs. [77–80], since the insulating pipe was imple-
mented as a part of the sensing field for the non-invasive ERT
system; the coupling capacitance of the insulating pipe also gen-
erally considered for the sensing field. The coupling capacitance
is referred to the capacitance between the electrode mounted on
the pipe wall and the conductive liquid. Thus, it will have two cou-
pling capacitances for one pair of the measurement electrodes. It
is exactly different if compared to the conventional ERT system;
which is only considered the resistance between each pair of the
measurement electrodes. In that case, the non-invasive ERT system
will have a total impedance of resistance in series with the coupling
capacitances. Nevertheless, the non-invasive ERT system is not
exactly an electrical impedance tomography (EIT) system because
the complex impedance model of the EIT system considers an
equivalent resistor and capacitor in parallel in measuring circuit as
studied in Ref. [81]. Simultaneously, the EIT system reconstructs the
image based on both electrical conductivity and permittivity distri-
bution of the medium interest [82]. Therefore, the non-invasive ERT
system is works based on the resistance value when the phase angle
of the measured impedance could be optimized due to the some
high enough working frequency applied for the system. Finally,
based on the resistance value, the electrical conductivity distribu-
tion of the chemical mixtures for non-invasive ERT system can be
known.
5. The proposed non-invasive process tomography
Non-invasive process tomography is a promising imag-
ing technique for physiological monitoring, with potentially
 Y. Abdul Wahab et al. / Sensors and Actuators B 210 (2015) 602–617
Fig. 17. ERT system using ECT sensor [77]; (A) ECT sensor, (B) switching unit and (C) Agilent 4294A.
enormous applications for chemical mixtures analysis. However,
due to the ill-posed nature of the image reconstruction problem,
the requirements on instrumentation are demanding. Electrical
resistance tomography (ERT) is a concern here. Alternatively, as
suggested by [77–80] in the previous subtopic, there is potential
for improvement of non-invasive techniques of process tomogra-
phy in chemical mixtures, especially for ERT. Here, an improvement
of non-invasive ERT is proposed as one of the non-invasive methods
for measurement of chemical mixtures.
ERT is very useful when the primary medium of interest
is a conductive medium rather than an insulating medium. By
using non-invasive ERT, the chemical mixtures can be detected
quickly without sensors breaking the wall of the pipeline. Ref. [78]
introduced non-invasive ERT based on the series resistor–capacitor
(RC) circuit. By assuming the insulating pipeline as a coupling
capacitance and the conductive liquid as a resistor as shown in
Fig. 19, with the necessary frequency applied to the non-invasive
ERT system, the resistance of the medium can be measured by opti-
mizing the phase angle of the measured impedance. It means that,
the phase angle of the measured impedance should be removed so
that the measured impedance just consists of the resistance value.
Based on Ohm’s law, the resistance of the ERT system can be
measured by the exciting voltage (or current) and detecting cur-
rent (or voltage) via sensors mounted on the circumference of the
pipe walls [62]. A conventional invasive ERT system excites a cur-
rent signal with different current approaches, such as the adjacent
strategy, the opposite strategy, the conducting boundary strategy
or the diagonal strategy [83], and measures the voltage as an out-
put. In that case, the total independent measurements of resistance
613
are different depending on the specific method of excitation strat-
egy implemented. However, non-invasive ERT excites the voltage
and measures the current as an output. Voltage is supplied to the
excitation electrode in non-invasive ERT because the pipeline used
is an insulating pipeline. If current is applied, measurement with
non-invasive ERT is impossible. A complete cycle for non-invasive
ERT system measurement starts with the first electrode as excita-
tion electrode and all the other electrodes as detecting electrodes.
Thus, the internal resistance is measured between the first elec-
trode and the adjacent electrodes. The process is repeated for all
other electrodes until each of the electrodes has become a source
electrode, and a complete cycle is done. Hence, the total indepen-
dent resistance measurement, M is given by the following equation
N(N − 1)
M= (1)
2
where M is the total resistance measurement and N is the total
electrode sensors.
Unfortunately, there is not much current research on non-
invasive ERT systems. For instance, it is investigated for a horizontal
pipeline with stratified flow and also for a vertical pipeline in detec-
ting liquid–solid (water–plastic rod). Potential improvement of the
non-invasive ERT system could apply, for example, in detecting
bubbles in bubble columns, or for real-time monitoring. In terms
of the number of electrodes used, this can be increased to sixteen
since it has been proved that this is a typical number of electrodes
implemented in ERT for chemical mixtures, where the material of
the electrode is flexible printed circuit board (PCB). In the study
undertaken, Fig. 20 shows an overview of proposed non-invasive
614 Y. Abdul Wahab et al. / Sensors and Actuators B 210 (2015) 602–617
Fig. 18. (a and b) Segmented non-invasive ERT system; (c) example of reconstructed image for annular flow [80].
Fig. 19. Non-invasive ERT sensor and its equivalent circuit.
Fig. 20. Overview of proposed non-invasive ERT system.
Fig. 21. Block diagram of proposed non-invasive ERT system.
ERT system, which sixteen electrodes are placed on the insulating
pipe wall for detecting chemical mixtures. Accordingly, the spe-
cific dimension of the size of electrodes would be optimized by
simulating the model using COMSOL Multiphysics software before
the proposed system can be applied for the real hardware sys-
tem. The flexible PCB is recommended for electrodes since it can
be bendable to give accurate geometry to the circle surface of the
pipe. Here, an acrylic or glass pipe will be used as the insulating
pipeline. Also, Fig. 21 shows the rough idea of a block diagram for
the proposed non-invasive ERT system for detecting chemical mix-
tures. Determining the most suitable excitation frequency for the
proposed non-invasive ERT electrodes is one of the chief concerns
here.
In short, non-invasive ERT can also be applied as one type of
non-invasive process tomography for detecting a two-phase flow
regime such as liquid–liquid, liquid–solid or liquid–gas.
 Y. Abdul Wahab et al. / Sensors and Actuators B 210 (2015) 602–617
6. Conclusion
Non-invasive process tomography seems to be a useful and
powerful tool for investigating and monitoring various kinds of
chemical mixtures. In industrial process plants it probably needs
to be constructed non-invasively so that the nature of the process
in the pipeline is not disturbed. Conversely, from the literature, not
much work has been done on non-invasive ERT systems for chem-
ical mixtures. Thus, it is believed that a further exploration and
improvement of ERT systems will provide a broad alternative way
to visualize and monitor chemical mixtures, and so help to improve
the performance of the process plant.
Acknowledgements
The authors would like to thank the Ministry of Higher Edu-
cation and Universiti Malaysia Pahang for funding the study. Very
special thanks go to the Universiti Teknologi Malaysia and PROTOM
research group for their generous support and cooperation.
References
[1] R. Abdul Rahim, N. Mohd Nayan, M.H. Fazalul Rahiman, Ultrasonic tomography
system for liquid/gas flow: frame rate comparison between visual basic and
visual C++ programming, J. Teknol. 44 (2006) 131–150.
[2] S. Ridzuan Aw, H. Arshad Amari, J. Pusppanathan, M.H. Fazalul Rahiman, Y. Abd
Wahab, Initial study on ultrasonic tomography for multiphase flow application,
J. Teknol. 54 (2011) 267–288.
[3] R. Abdul Rahim, Optical Tomography; Principle, Techniques and Applications,
1st ed., UTM Press, UTM Johor Bharu, 2011.
[4] R.C. Asher, Ultrasonic sensors in the chemical and process industries, J. Phys. E
16 (1983) 959–963.
[5] G. Steiner, F. Podd, A non-invasive and non-intrusive ultrasonic transducer
array for process tomography, in: XVIII IMEKO WORLD Congr. Metrol. Sustain.
Dev., 2006, pp. 17–22.
[6] U.Z. Ijaz, J.-H. Kim, A.K. Khambampati, M.-C. Kim, S. Kim, K.-Y. Kim, Concentra-
tion distribution estimation of fluid through electrical impedance tomography
based on interacting multiple model scheme, Flow Meas. Instrum. 18 (2007)
47–56, http://dx.doi.org/10.1016/j.flowmeasinst.2006.12.005.
[7] J.L. Hubers, A.C. Striegel, T.J. Heindel, J.N. Gray, T.C. Jensen, X-ray computed
tomography in large bubble columns, Chem. Eng. Sci. 60 (2005) 6124–6133,
http://dx.doi.org/10.1016/j.ces.2005.03.038.
[8] T.J. Heindel, J.N. Gray, T.C. Jensen, An X-ray system for visualizing fluid
flows, Flow Meas. Instrum. 19 (2008) 67–78, http://dx.doi.org/10.1016/
j.flowmeasinst.2007.09.003.
[9] A. Shaikh, M. Al-dahhan, Characterization of the hydrodynamic flow regime
in bubble columns via computed tomography, Flow Meas. Instrum. 16 (2005)
91–98, http://dx.doi.org/10.1016/j.flowmeasinst.2005.02.004.
[10] C. Wu, Y. Cheng, Y. Ding, F. Wei, Y. Jin, A novel X-ray computed tomography
method for fast measurement of multiphase flow, Chem. Eng. Sci. 62 (2007)
4325–4335, http://dx.doi.org/10.1016/j.ces.2007.04.026.
[11] Z. Zhang, M. Bieberle, F. Barthel, L. Szalinski, U. Hampel, Investigation of
upward cocurrent gas–liquid pipe flow using ultrafast X-ray tomography and
wire-mesh, Flow Meas. Instrum. 32 (2013) 111–118, http://dx.doi.org/10.1016/
j.flowmeasinst.2013.04.007.
[12] H. Prasser, M. Misawa, I. Tiseanu, Comparison between wire-mesh sensor
and ultra-fast X-ray tomograph for an air–water flow in a vertical pipe, Flow
Meas. Instrum. 16 (2005) 73–83, http://dx.doi.org/10.1016/j.flowmeasinst.
2005.02.003.
[13] S. Boden, M. Bieberle, U. Hampel, Quantitative measurement of
gas hold-up distribution in a stirred chemical reactor using X-ray
cone–beam computed tomography, Chem. Eng. J. 139 (2008) 351–362,
http://dx.doi.org/10.1016/j.cej.2007.08.014.
[14] R. Abdul Rahim, J.F. Pang, K.S. Chan, Optical tomography sensor config-
uration using two orthogonal and two rectilinear projection arrays, Flow
Meas. Instrum. 16 (2005) 327–340, http://dx.doi.org/10.1016/j.flowmeasinst.
2005.06.001.
[15] E. Schleicher, M.J. Da Silva, S. Thiele, A. Li, E. Wollrab, U. Ham-
pel, Design of an optical tomograph for the investigation of single-
and two-phase pipe flows, Meas. Sci. Technol. 19 (2008) 094006,
http://dx.doi.org/10.1088/0957-0233/19/9/094006.
[16] Md.Y. Yunos, R. Abdul Rahim, R.G. Green, Initial result on measurement of gas
volumetric flow rate in gas/liquid mixtures using linear CCD, J. Teknol. 48 (2008)
1–11.
[17] S. Ibrahim, M.A.M. Yunus, R.G. Green, K. Dutton, Concentration mea-
surements of bubbles in a water column using an optical tomography
system, ISA Trans. 51 (2012) 821–826 http://www.sciencedirect.com/science/
article/pii/S0019057812000584 (accessed 02.12.13).
615
[18] N.S. Mohd Fadzil, R. Abdul Rahim, M.S. Karis, S.Z. Mohd Muji, M.F. Abdul Sahib,
M.S.B. Mansor, et al., Hardware design of laser optical tomography system for
detection of bubbles column, J. Teknol. 64 (2013) 69–73.
[19] R. Abdul Rahim, N.W. Nyap, M.H. Fazalul Rahiman, C.K. San, Determination of
water and oil flow composition using ultrasonic tomography, J. Teknol. 9 (2007)
19–23.
[20] M.H. Fazalul Rahiman, Z. Zakaria, R. Abdul Rahim, W.N. Ng, Ultrasonic tomogra-
phy imaging simulation of two-phase homogeneous flow, Sens. Rev. 29 (2009)
266–276, http://dx.doi.org/10.1108/02602280910967684.
[21] G. Steiner, C. Deinhammer, Ultrasonic time-of-flight techniques for monitor-
ing multi-component processes, Elektrotech. Informationstech. 126 (2009)
200–205, http://dx.doi.org/10.1007/s00502-009-0640-6.
[22] S. Wöckel, U. Hempel, J. Auge, Phase boundary characterization in
liquids by acoustic waves, Meas. Sci. Technol. 20 (2009) 124013,
http://dx.doi.org/10.1088/0957-0233/20/12/124013.
[23] M.H. Fazalul Rahiman, R. Abdul Rahim, Development of ultrasonic
transmission-mode tomography for water-particles flow, Sens. Transducer
117 (2010) 99–105.
[24] I.R. Muhamad, Y.A. Wahab, S. Saat, Identification of water/solid flow regime
using ultrasonic tomography, in: IEEE Int. Conf. Syst. Eng. Technol. (ICSET 2012),
2012, pp. 3–7.
[25] Z. Zakaria, M.H. Fazalul Rahiman, R. Abdul Rahim, Simulation of the two-phase
liquid–gas flow through ultrasonic transceivers application in ultrasonic tomo-
graphy, Sens. Transducer 112 (2010) 24–38.
[26] M. Supardan, Y. Masuda, S. a Maezawa, Uchida, The investigation of
gas holdup distribution in a two-phase bubble column using ultra-
sonic computed tomography, Chem. Eng. J. 130 (2007) 125–133,
http://dx.doi.org/10.1016/j.cej.2006.08.035.
[27] M.H. Fazalul Rahiman, Z. Zakaria, R. Abdul Rahim, Ultrasonic pro-
cess tomographic imaging sensor: an approach utilising transceivers
method, 2008, Int. Conf. Comput. Commun. Eng. (2008) 1147–1150,
http://dx.doi.org/10.1109/ICCCE.2008.4580786.
[28] M.H. Fazalul Rahiman, R. Abdul Rahim, H. Abdul Rahim, N.M.
Nor Ayob, Novel adjacent criterion method for improving ultra-
sonic imaging spatial resolution, IEEE Sens. J. 12 (2012) 1746–1747,
http://dx.doi.org/10.1109/JSEN.2011.2177255.
[29] M.H.F. Rahiman, R.A. Rahim, H.A. Rahim, N.M.N. Ayob, E.J. Mohamad, Z.
Zakaria, Modelling ultrasonic sensor for gas bubble profiles characteriza-
tion of chemical column, Sens. Actuators B: Chem. 184 (2013) 100–105,
http://dx.doi.org/10.1016/j.snb.2013.04.062.
[30] M.H.F. Rahiman, R.A. Rahim, H.A. Rahim, E.J. Mohamad, Z. Zakaria, S.Z.M.
Muji, An investigation on chemical bubble column using ultrasonic tomogra-
phy for imaging of gas profiles, Sens. Actuators B: Chem. 202 (2014) 46–52,
http://dx.doi.org/10.1016/j.snb.2014.05.043.
[31] N.M. Nor Ayob, S. Yaacob, Z. Zakaria, M.H. Fazalul Rahiman, R. Abdul Rahim,
M.R. Manan, Improving gas component detection of an ultrasonic tomography
system for monitoring liquid/gas flow, in: 2010 6th Int. Colloq. Signal Process.
Its Appl., IEEE, 2010, pp. 1–5, http://dx.doi.org/10.1109/CSPA.2010.5545286.
[32] N.M. Nor Ayob, M.J. Pusppanathan, R. Abdul Rahim, M.H. Fazalul Rahiman, F.R.
Mohd Yunus, S. Buyamin, et al., Design consideration for front-end system in
ultrasonic tomography, J. Teknol. 64 (2013) 53–58.
[33] N.M. Nor Ayob, M.H. Fazalul Rahiman, Z. Zakaria, S. Yaacob, R. Abdul Rahim, M.R.
Manan, Simulative study in liquid/gas two-phase flow measurement for dual-
plane ultrasonic transmission-mode tomography, J. Teknol. 54 (2011) 79–94.
[34] S.C. Murphy, T.A. York, Electrical impedance tomography with non-
stationary electrodes, Meas. Sci. Technol. 17 (2006) 3042–3052,
http://dx.doi.org/10.1088/0957-0233/17/11/025.
[35] W. Wang, H. Yang, Imaging a rectangular fluidised bed by electrical capac-
itance tomography: preliminary results, Sens. Rev. 31 (2011) 315–320,
http://dx.doi.org/10.1108/02602281111169721.
[36] H. Wang, Y. Li, G. Qiu, G. Song, W. Yang, Measurement of gas–solids flow in loop
seal and external heat exchanger in a circulating fluidized bed, Powder Technol.
266 (2014) 249–261, http://dx.doi.org/10.1016/j.powtec.2014.06.046.
[37] D. Chen, Y. Han, An image data capture system for electrical capacitance tomo-
graphy of oil/water two-phase flow, in: IEEE Int. Conf. Inf. Acquis. (ICIA 2006),
2006, pp. 722–726.
[38] H.Y. Cao, X.M. Duan, H.X. Wang, Design of electrical capacitance tomo-
graphy hardware system, Adv. Mater. Res. 753-755 (2013) 2311–2315,
http://dx.doi.org/10.4028/www.scientific.net/AMR.753-755.2311.
[39] N.T. Ali Othman, H. Obara, M. Takei, Cross-sectional capacitance
measurement of particle concentration in a microchannel with
multi-layered electrodes, Flow Meas. Instrum. 31 (2013) 47–54,
http://dx.doi.org/10.1016/j.flowmeasinst.2012.12.003.
[40] W.Q. Yang, A. Chondronasios, S. Nattrass, V.T. Nguyen, M. Betting, I. Ismail,
et al., Adaptive calibration of a capacitance tomography system for imag-
ing water droplet distribution, Flow Meas. Instrum. 15 (2004) 249–258,
http://dx.doi.org/10.1016/j.flowmeasinst.2004.07.001.
[41] C. Yu, C. Deyun, Two-phase flow parameters measurement and Gauss–Newton
image reconstruction algorithm for electrical capacitance tomography, in: IEEE
Int. Conf. Ind. Mechatronics Autom. (ICIMA 2009), IEEE, 2009, pp. 192–195,
http://dx.doi.org/10.1109/ICIMA.2009.5156593.
[42] F. Teixeira, L. Fan, Advances in Electrical Capacitance Tomography, The Ohio
State University, 2006, http://dx.doi.org/3220969 UMI.
[43] W.A. Al-Masry, E.M. Ali, S.A. Alshebeili, F.M. Mousa, Non-invasive imaging
of shallow bubble columns using electrical capacitance tomography, J. Saudi
Chem. Soc. 14 (2010) 269–280, http://dx.doi.org/10.1016/j.jscs.2010.02.022.
616 Y. Abdul Wahab et al. / Sensors and Actuators B 210 (2015) 602–617
[44] R. Banasiak, R. Wajman, T. Jaworski, P. Fiderek, H. Fidos, J.
Nowakowski, et al., Study on two-phase flow regime visualiza-
tion and identification using 3D electrical capacitance tomography
and fuzzy-logic classification, Int. J. Multiph. Flow. 58 (2014) 1–14,
http://dx.doi.org/10.1016/j.ijmultiphaseflow.2013.07.003.
[45] J.C. Gamio, J. Castro, L. Rivera, J. Alamilla, F. Garcia-Nocetti, L. Aguilar,
Visualisation of gas–oil two-phase flows in pressurised pipes using elec-
trical capacitance tomography, Flow Meas. Instrum. 16 (2005) 129–134,
http://dx.doi.org/10.1016/j.flowmeasinst.2005.02.011.
[46] HuangF Z., D. Xie, H. Zhang, H. Li, Gas–oil two-phase flow mea-
surement using an electrical capacitance tomography system
and a Venturi meter, Flow Meas. Instrum. 16 (2005) 177–182,
http://dx.doi.org/10.1016/j.flowmeasinst.2005.02.007.
[47] C. Ortiz-Alemán, R. Martin, Inversion of electrical capacitance tomo-
graphy data by simulated annealing: application to real two-phase
gas–oil flow imaging, Flow Meas. Instrum. 16 (2005) 157–162,
http://dx.doi.org/10.1016/j.flowmeasinst.2005.02.014.
[48] D. Xie, Z. Huang, H. Ji, H. Li, An online flow pattern identifica-
tion system for gas–oil two-phase flow using electrical capacitance
tomography, IEEE Trans. Instrum. Meas. 55 (2006) 1833–1838,
http://dx.doi.org/10.1109/TIM.2006.881558.
[49] E.J. Mohamad, R. Abdul Rahim, L.P. Ling, M.H. Fazalul Rahiman, O.M.F.
Marwah, N.M. Nor Ayob, Segmented capacitance tomography electrodes:
a design and experimental verifications, IEEE Sens. J. 12 (2012) 1589–
1598.
[50] Y. Zhao, H. Yeung, E.E. Zorgani, A.E. Archibong, L. Lao, High viscosity effects on
characteristics of oil and gas two-phase flow in horizontal pipes, Chem. Eng.
Sci. 95 (2013) 343–352, http://dx.doi.org/10.1016/j.ces.2013.03.004.
[51] R. Zhang, Q. Wang, H. Wang, M. Zhang, H. Li, Data fusion in dual-mode tomogra-
phy for imaging oil–gas two-phase flow, Flow Meas. Instrum. 37 (2014) 1–11,
http://dx.doi.org/10.1016/j.flowmeasinst.2014.03.003.
[52] M. Sharifi, B. Young, Electrical resistance tomography (ERT) applications
to chemical engineering, Chem. Eng. Res. Des. 91 (2013) 1625–1645,
http://dx.doi.org/10.1016/j.cherd.2013.05.
[53] F. Dong, C. Xu, Z. Zhang, S. Ren, Design of parallel electrical resistance tomo-
graphy system for measuring multiphase flow, Chin. J. Chem. Eng. 20 (2012)
368–379, http://dx.doi.org/10.1016/S1004-9541(12)60400-5.
[54] H. Zhou, L. Xu, Z. Cao, J. Hu, X. Liu, Image reconstruction for invasive
ERT in vertical oil well logging, Chin. J. Chem. Eng. 20 (2012) 319–328,
http://dx.doi.org/10.1016/S1004-9541(12)60394-2.
[55] S. Kim, A.N. Nkaya, T. Dyakowski, Measurement of mixing of two
miscible liquids in a stirred vessel with electrical resistance tomo-
graphy, Int. Commun. Heat Mass Transf. 33 (2006) 1088–1095,
http://dx.doi.org/10.1016/j.icheatmasstransfer.2006.06.010.
[56] F. Ricard, C. Brechtelsbauer, X.Y. Xu, C.J. Lawrence, Monitoring of multiphase
pharmaceutical processes using electrical resistance tomography, Chem. Eng.
Res. Des. 83 (2005) 794–805, http://dx.doi.org/10.1205/cherd.04324.
[57] Y.H. Liu, N. Wang, H.Y. Sun, W. Chen, Solid-liquid two-phase flow image
reconstruction in microchannel based on tikhonov regularization method,
Appl. Mech. Mater. 333–335 (2013) 1013–1019, http://dx.doi.org/10.4028/
www.scientific.net/AMM.333-335.1013.
[58] M. Sharifi, B. Young, Towards an online milk concentration sensor using ERT:
correlation of conductivity, temperature and composition, J. Food Eng. 116
(2013) 86–96, http://dx.doi.org/10.1016/j.jfoodeng.2012.11.010.
[59] M. Sharifi, B. Young, Qualitative visualization and quantitative analysis of milk
flow using electrical resistance tomography, J. Food Eng. 112 (2012) 227–242,
http://dx.doi.org/10.1016/j.jfoodeng.2012.04.008.
[60] M. Sharifi, B. Young, 3-Dimensional spatial monitoring of tanks for the milk
processing industry using electrical resistance tomography, J. Food Eng. 105
(2011) 312–319, http://dx.doi.org/10.1016/j.jfoodeng.2011.02.041.
[61] M. Sharifi, B. Young, The potential utilisation of electrical resistance tomogra-
phy (ERT) in milk powder processing for monitoring and control, in: 6th World
Congr. Ind. Process Tomogr., 2005, pp. 1–14.
[62] S. Hosseini, D. Patel, F. Ein-Mozaffari, M. Mehrvar, Study of solid–liquid mixing
in agitated tanks through electrical resistance tomography, Chem. Eng. Sci. 65
(2010) 1374–1384, http://dx.doi.org/10.1016/j.ces.2009.10.007.
[63] S.A. Razzak, S. Barghi, J.-X. Zhu, Application of electrical resistance tomography
on liquid–solid two-phase flow characterization in an LSCFB riser, Chem. Eng.
Sci. 64 (2009) 2851–2858, http://dx.doi.org/10.1016/j.ces.2009.02.049.
[64] S.J. Stanley, G.T. Bolton, A review of recent electrical resistance tomography
(ERT) applications for wet particulate processing, Part. Part. Syst. Charact. 25
(2008) 207–215, http://dx.doi.org/10.1002/ppsc.200701126.
[65] A. Madupu, A. Mazumdar, J. Zhang, D. Roelant, R. Srivastava, Electrical
resistance tomography for real-time mapping of the solid-liquid inter-
face in tanks containing optically opaque fluids, in: C.A. Bouman, E.L.
Miller (Eds.), Proc. SPIE-IS&T Electron. Imaging, SPIE, 2005, pp. 36–46,
http://dx.doi.org/10.1117/12.587959.
[66] M. Sobri, A. Ahmad, M. Irwan, S. Jantan, ERT visualization of gas dispersion
performance of aerofoil and radial impellers in an agitated vessel, J. Teknol. 64
(2013) 75–78.
[67] A.D. Okonkwo, M. Wang, B. Azzopardi, Characterisation of a high concen-
tration ionic bubble column using electrical resistance tomography, Flow
Meas. Instrum. 31 (2013) 69–76, http://dx.doi.org/10.1016/j.flowmeasinst.
2012.10.005.
[68] W. Yenjaichon, J.R. Grace, C. Jim Lim, C.P.J. Bennington, Characteri-
sation of gas mixing in water and pulp-suspension flow based on
electrical resistance tomography, Chem. Eng. J. 214 (2013) 285–297,
http://dx.doi.org/10.1016/j.cej.2012.10.057.
[69] C. Yang, H. Wang, Z. Cui, Application of electrical resistance tomography in
bubble columns for volume fraction measurement, in: IEEE Int. Conf. Instrum.
Meas. Technol. (I2MTC 2012), 2012, pp. 1199–1203.
[70] X. Deng, G. Li, Z. Wei, Z. Yan, W. Yang, Theoretical study of vertical slug
flow measurement by data fusion from electromagnetic flowmeter and
electrical resistance tomography, Flow Meas. Instrum. 22 (2011) 272–278,
http://dx.doi.org/10.1016/j.flowmeasinst.2011.03.007.
[71] J. Kourunen, T. Niitti, L.M. Heikkinen, Application of three-dimensional
electrical resistance tomography to characterize gas holdup distribu-
tion in laboratory flotation cell, Miner. Eng. 24 (2011) 1677–1686,
http://dx.doi.org/10.1016/j.mineng.2011.09.004.
[72] H. Jin, S. Yang, G. He, M. Wang, R.A. Williams, The effect of gas–liquid
counter-current operation on gas hold-up in bubble columns using electrical
resistance tomography, J. Chem. Technol. Biotechnol. 85 (2010) 1278–1283,
http://dx.doi.org/10.1002/jctb.2428.
[73] Y. Xu, H. Wang, Z. Cui, F. Dong, Application of electrical resistance tomography
for slug flow measurement in gas–liquid flow of horizontal pipe, in: IEEE Int.
Work. Imaging Syst. Technol. (IST 2009), 2009, pp. 319–323.
[74] Z. Meng, Z. Huang, B. Wang, H. Ji, H. Li, Flowrate measurement of air–water
two-phase flow using an electrical resistance tomography sensor and a Venturi
meter, in: IEEE Int. Conf. Instrum. Meas. Technol. (I2MTC 2009), IEEE, 2009, pp.
118–121, http://dx.doi.org/10.1109/IMTC.2009.5168427.
[75] M. Wang, G. Lucas, Y. Dai, N. Panayotopoulos, R.A. Williams, Visualisa-
tion of bubbly velocity distribution in a swirling flow using electrical
resistance tomography, Part. Part. Syst. Charact. 23 (2006) 321–329,
http://dx.doi.org/10.1002/ppsc.200601063.
[76] F. Dong, Y. Xu, L. Hua, H. Wang, Two methods for measurement of gas–liquid
flows in vertical upward pipe using ERT system, IEEE Trans. Instrum. Meas. 55
(2006) 1576–1586.
[77] Z. Cao, L. Xu, C. Xu, H. Wang, Electrical resistance tomography (ERT) by using
an ECT sensor, in: IEEE Int. Conf. Imaging Syst. Technol. (IST 2010), 2010, pp.
63–66.
[78] B. Wang, Y. Hu, H. Ji, Z. Huang, H. Li, A novel electrical resis-
tance tomography system based on C4D technique, in: IEEE Int. Conf.
Instrum. Meas. Technol. (I2MTC 2012), IEEE, 2012, pp. 1929–1932,
http://dx.doi.org/10.1109/I2MTC.2012.6229290.
[79] B. Wang, Y. Hu, H. Ji, Z. Huang, H. Li, A novel electrical resistance tomogra-
phy system based on C 4D technique, IEEE Trans. Instrum. Meas. 62 (2013)
1017–1024.
[80] B. Wang, W. Zhang, Z. Huang, H. Ji, H. Li, Modeling and optimal design of sensor
for capacitively coupled electrical resistance tomography system, Flow Meas.
Instrum. 31 (2013) 3–9, http://dx.doi.org/10.1016/j.flowmeasinst.2012.10.009.
[81] R. Thorn, G.A. Johansen, B.T. Hjertaker, Three-phase flow measure-
ment in the petroleum industry, Meas. Sci. Technol. 24 (2013) 012003,
http://dx.doi.org/10.1088/0957-0233/24/1/012003.
[82] M. Cheney, D. Isaacson, Issues in electrical impedance imaging, IEEE Comput.
Sci. Eng. 2 (1995) 53–62, http://dx.doi.org/10.1109/99.476369.
[83] F. Dickin, M. Wang, Electrical resistance tomography for process appli-
cations, Meas. Sci. Technol. 7 (1996) 247–260, http://dx.doi.org/10.1088/
0957-0233/7/3/005.
Biographies
Yasmin Abdul Wahab received her a B. Eng. degree (Honours) in Electrical Engineer-
ing (Instrumentation and Control) and her M. Eng. degree in Electrical Engineering
(Mechatronics and Automatic Control) from Universiti Teknologi Malaysia (UTM),
Skudai, Malaysia, in 2008 and 2010, respectively. In 2010, she joined Universiti
Malaysia Pahang (UMP), Pahang, Malaysia, as a Lecturer. Currently, she is work-
ing towards a Ph.D. in Process Tomography at UTM. Her current research interests
are Electrical Process Tomography, Sensors and Instrumentation.
Ruzairi Abdul Rahim received a B. Eng. degree with Honours in Electronic System
and Control Engineering in 1992 from Sheffield City Polytechnic, UK. He received his
Ph.D. in Instrumentation & Electronics Engineering from Sheffield Hallam Univer-
sity, UK in 1996. At present he is a Professor and a Director of Research Management
Centre, Universiti Teknologi Malaysia. His current research interests are Process
Tomography and Sensor Technology.
Mohd Hafiz Fazalul Rahiman received a B. Eng. (Hons) degree in Electrical Engi-
neering (Control and Instrumentation), M. Eng. and Ph.D. degrees in Electrical
Engineering from Universiti Teknologi Malaysia (UTM), Johor, Malaysia, in 2003,
2005, and 2013 respectively. In 2006, he joined Universiti Malaysia Perlis (UniMAP),
Perlis, Malaysia, as a teaching staff member and at present he holds the position of
Senior Lecturer. His research interests include Process Tomography, Sensors and
Instrumentation.
Suzanna Ridzuan Aw received her B. Eng. degree (Honours) in Electrical Engineer-
ing (Instrumentation and Control) and her M. Eng. degree in Electrical Engineering
(Mechatronics and Automatic Control) fromUniversiti TeknologiMalaysia (UTM),
Skudai, Malaysia, in 2009 and 2011, respectively. Currently, she is pursuing her Ph.D.
degree at UTM in Process Tomography. Her current research interest is in Electrical
Resistance Tomography.
 Y. Abdul Wahab et al. / Sensors and Actuators B 210 (2015) 602–617
Fazlul Rahman Mohd Yunus received a B. Eng.degree (Honours) in Electrical
Engineering (Instrumentation and Control) and an M. Eng. degree in Electrical Engi-
neering (Mechatronics and Automatic Control) from Universiti Teknologi Malaysia,
Skudai, Malaysia, in1999 and 2009, respectively. In 1999, he joined STmicroelectron-
ics as Test Engineer for two years before being called by the government of Malaysia
to serve in the Japan–Malaysia Technical Institute (JMTI), Penang (2001–2009) and
the Industrial Training Institute, Ledang (2009–2012) of the Manpower Department,
Ministry of Human Resources Malaysia, as a vocational training officer. Currently he
is working towards a Ph.D. in Process Tomography. His current research interest is
in Dual-modality Process Tomography.
Chiew Loon Goh received her M.Sc. Master in Electrical and Electronic Engineering
from the University Technology Malaysia, Malaysia, in 2006. After several years of
working in the field of R&D engineering and software programming, she joined the
process tomography & instrumentation research group (PROTOM-i) at the Universiti
Teknologi Malaysia (UTM) as researcher in 2012. Her research interests include the
design of electronic circuits, embedded systems, wireless system and application
programming of image processing for tomography systems.
617
Herlina Abdul Rahim received a B. Eng. in Electrical Engineering (Control and
Instrumentation) and M. Eng. in Electrical Engineering from Universiti Teknologi
Malaysia in 1998 and 2000, respectively. She received her Ph.D. in Electri-
cal Engineering from Universiti Teknologi MARA, in 2009. At present, she is
an Associate Professor at Faculty of Electrical Engineering, Universiti Teknologi
Malaysia. Her current research interests are sensor technology and artificial intel-
ligent system. She is a senior member of the IEEE, a Professional Engineer,
Board of Engineers Malaysia (BEM), a corporate member of the Institute of Engi-
neers Malaysia (IEM) and Chartered Engineer of the Institute of Engineering and
Technology.
Leow Pei Ling received the B. Eng. and M. Eng. degrees in Electrical Engineering from
the Universiti Teknologi Malaysia, Johor, Malaysia, in 2003 and 2005, respectively,
and the Ph.D. degree in bioengineering from Imperial College London, London, U.K.,
in 2009. She is currently Lecturing at the Universiti Teknologi Malaysia. Her research
interests are microfluidic devices, sensors and instrumentation.