7th International Conference on Multiphase Flow,
ICMF 2010, Tampa, FL, May 30 June 4, 2010
Insitu determination of mass flow rate and visualisation of vertical
slurry pipeline flows
B. Munir and M. Wang
Institute of Particle Science and Engineering, University of Leeds, Leeds, LS2 9JT, UK,
pre4bm@leeds.ac.uk and m. ain lcck cds .I uk
Keywords: Mass measurement, vertical slurry flow, electrical resistance tomography
Abstract
Solidliquid transport systems present considerable metering and visualisation challenges, due to the physical properties
of the slurry mixture such as its wide ranging particle size distributions, opaqueness and abrasive nature. In addition the
pipeline is required to be open bore and the measurement scheme is nonintrusive. This paper presents a measurement
technique capable of sensing essential transport variables for the determination of mass flow rate of slurry insitu, by
combining technologies electrical resistance tomography (ERT) with differential pressure measurement.
As ERT measurement is based on a relative difference between a reference frame and measurement frames, it is
susceptible to temperature changes during continuous industrial operation. The change of fluid's temperature during its
process also causes additional variations to the online ERT measurements. By using a insitu density meter based on a
differential pressure measurement scheme to update the ERT measurement it can be demonstrated that this problem can
be alleviated. The flow rate of disperse phase, i.e. sands, is obtained with the concentration and velocity distributions
obtained with ERT and crosscorrelation method and corrected with the equivalent liquid model. Experimentation has
been conducted in a solids handling flow loop using silica sand in water as medium to test the performance of the
proposed measurement scheme. Slurry density measurements determined by electrical resistance tomography technique
at various transport velocities and solids concentration are reported. A validation of the proposed measurement scheme
is made against discharge samples taken on line using a load cell weigh system.
1. Introduction
The design of slurry pipeline systems carrying settling
slurries requires prediction of the hydraulic
performance of the pipeline. Modern measurement
techniques are often required for determining flow
characteristics not only industrially where the
instruments form a more monitoring and control role,
but also academically where information of the mean
and distributive values of concentration and velocity
are important for further determining the mechanisms
behind slurry flow.
Currently there is no single device that can determine
completely the characteristics of multiphase solid
liquid flow. The single phase flow meters used
commercially are based mainly on ultrasonic Doppler
or magnetic principles. However these meters are
sensitive to the flow regime changes and signal
distortion (Hammer et al. 1997, Loh 1998, Baker 2000
and Cha et al. 2001) and often flows encountered
industrially maybe unsteady in nature. In addition,
electromagnetic flow meters (EMF) only gives the flow
rate of the liquid and not the actual flow of solids,
which is where the interest lies. The presence of
Electrical Resistance Tomography sensor (ERT) in the
proposed measurement scheme presents an opportunity
to alleviate the error that is inherent in flow meters
when gauging nonuniform multiphase flows.
Continuous process monitoring represents a challenge
to ERT systems, due to its operational nature of
measurement which uses a reference quantity as a
means of establishing relative changes in the mixture.
Over time process changes in temperature and ionic
concentrations can cause the original reference to
become unsuitable. Obviously there are difficulties in
taking renewed references in a continuous process
hence there is a need to update the reference file using
an alternative technology. In this paper it will be
shown that the velocity found from ERT can be used to
update the original reference reading.
By utilizing a cross correlation technique applied to
reconstructed conductivity maps from ERT, the
velocity of the mixture can be determined. Crucially
these signals can be of any value and are not limited to
quantitative conductivity data. When this is combined
with the measured pressure drop the concentration of
the mixture can be found, which indirectly infers an
updated conductivity value for the mixture. In addition
to this the discharge or delivered measurement
technique is used for evaluating pipeline slurry insitu
concentration results obtained using ERT measurement.
Discharge measurement of the full pipe bore represents
the only reliable method of sampling hydraulic flow as
any other flow meter technology will face the same
issues of calibration.
2. Theoretical Considerations
A mechanical model can be used to obtain information
such as concentration or velocity insitu pipe mixture
providing the pressure drop is known, the basis of this
model is given in this section. Solid particles in fully
developed vertical flows are considered to be uniformly
distributed across a pipe crosssection. No component
of the submerged weight of a solid particle acts against
a pipe wall. The submerged weight acts in the direction
of the flow and causes slip between solids and liquid in
the flow the extent of which is dependent on particle
characteristics. Following a force balance the basic
equation for vertical rising flow can be given by:
dp g. PLSM + 4./D (1)
dx
where dp/dx is the pressure gradient, ris the shear stress
at the pipe wall, D is the tube diameter, and SM is the
insitu relative density defined
S = 1+(Ss 1). C,, (2)
where Ss =psPL and C,, represent the relative density of
the solid and the insitu volumetric solids fraction,
respectively.
Some consideration must be given to the relative
velocity difference which is common to vertical
7th International Conference on Multiphase Flow,
ICMF 2010, Tampa, FL, May 30 June 4, 2010
transport of settling solids. Although some levels of slip
must occur in vertical flows, for the condition of these
experiments (moderate concentrations) its effect would
be small, Matousek (2002). Our tested solids showed
an effective slip with an equivalent settling velocity of
0.44m/s and a slurry settling velocity 0.24m/s,
determined using established correlations of settling
particles and mixtures (Wilson 1997). This effectively
means the influence of a slip on measurement is so
negligible that Cvd can be used in determining the in
situ density of the pipe, which in turn is measured by
our discharge measurement scheme detailed in section
3.
The frictional head loss can be described by the
equivalent liquid model (ELM) first proposed by
Durand (1951a, 1951b). According to this model the
solid particles exhibit a coring effect, and travel away
from the near wall region of the pipe. The near wall
region is then comprised mostly of water and thus ffor
the mixture can be evaluated as that of water having the
same velocity. The wall shear stress, TL can then
effectively be defined as PL fLV22 where V is the
mixture velocity, and friction factor, fL, of water for a
smooth pipe, and substituting for SM into Eq. (1) leads
to:
gpL (1+ (S,
1)C,) + pLg2 /(2gD) (3)
The ERT system was used to estimate the insitu
volumetric fraction based on the average of volumetric
fractions of individual pixels which constitute the entire
image. The simple calculation is given by
I=m
1 A
where A,, A,,t and cs,, is the area of pixel, the area of
image (the crosssectional area of pipe) and insitu
local volumetric disperse phase fraction, respectively.
3. Experimental Procedure and Setup
The flow facilities comprises PVC pipe work of 50mm
internal diameter and a vertical section of 3.65m (73D)
in which measurement is made by ERT,
electromagnetic flow meter, temperature and pressure
sensors taking place over 2m (40D), a schematic is
shown in Fig. 2. The loop consists of a main 500 litres
mixing tank, where the solids and liquid are mixed
homogeneously and introduced to the loop, and a 250
litre measuring tank placed on bending beam type load
cells, which measures the weight of its contents and as
it fills, a level sensor will measure the volume of the
liquid in the tank to a precalibrated settling. A 15kW
Warman International 2/11/2 AH heavyduty
centrifugal pump is used to transport the slurry mixture.
The installed electromagnetic flow meter in the loop is
the Aquaflux model supplied Krohne, chosen as it is
especially suited to handling abrasive slurry mixtures as
it has a resistance body lining and has typical accuracy
value are of the order of 0.5% of full deflection. The
two Danfoss flush diaphragm type pressure sensor
placed in the vertical measuring section are 1.75m
(35D) apart from each other, with 0.3% fullscale
deflection error. A Ktype thermocouple having
sensitivity of 41[tV/C, measures the temperature
variation during the experimental run.
The measurement tank is positioned on three Vishay
Tedea Huntleigh bending beam load cells type 355
7th International Conference on Multiphase Flow,
ICMF 2010, Tampa, FL, May 30 June 4, 2010
100kgC3, which have a claimed accuracy of 0.05%.
The load cell setup performance was gauged by
carrying out discharge measurements using water. It
can be seen in figure 1 that the load cell setup is able to
accurately produce measurements that compared
linearly with the commercial EMF described above,
with an average deviation of 1.1%. A test to determine
water density at varying velocity was also performed,
which showed the load cells were able to produce
measurements within 0.3% for velocity below 2.5m/s
when compared with calculated values, and 1.12% for
velocity between 2.54.5m/s.
Sand slurry with median particle size range from
212pm (d15) to 355pm (d85) was used as our test
medium. The flow velocities were between 1.5 ms1
and 3.5 ms '. ERT results presented in this paper were
obtained from an ITS 2000 ERT system (Industrial
Tomography Systems Ltd, Manchester, U.K.) and a fast
system developed at Leeds University (Wang et al.
2005). ERT sensor was mounted in the working section
at a downstream distance of 2.25m (45D) from the pipe
bend, and as a study of bend effects on flow the ERT
sensor was placed 0.5m (10D) from the bend.
4
4
Pressure Drop Density
V(m/s) (Pam)Error(%)
(Pa/m) (kg/m3)
0.274
0.470
1.018
0 1 2 3
EMF Velocity m/s
1.520
2.052
2.517
I I
2.998
4 5 3.538
4.005
42.86
91.13
302.96
597.15
1013.29
1465.00
2018.58
2744.69
3461.88
Figure 1: Calibration of discharge measurement scheme with velocity of EMF (left) and determining of water
density (right)
U3
E
33
0
632
.0
r)
L1
(U
c
993.59
994.50
1000.08
1001.02
1000.27
1003.57
1010.56
1008.04
1009.14
7th International Conference on Multiphase Flow,
ICMF 2010, Tampa, FL, May 30 June 4, 2010
Flow Diversion
Switch
Vertical Section
500 Litre
Mixing Tank
Measurement
Section
T
P2)
T
P18
35D, 1.75 m
25D
Total = 73D, 3.65 m
Figure 2: Schematic of the slurry flow experimental rig along with measuring section
3.1 Concentration Measurement
The dualplane ERT Sensor with two dummy rings
was configured so that the axial separation of the
image planes was 30mm. On each plane, sixteen
stainless steel electrodes are mounted flush to the
surface of the pipe at equal intervals. The electrodes
were designed to have a length to width ratio of 3,
giving an electrode size of 18mm by 6mm. The
voltage potential differences for tomography images
were collected based on the normal adjacent protocol,
with a data collection speed of 5 frames per second
for the dual planes, at an AC current injection
frequency of 9600Hz and a current value of 15mA.
This produces 104 independent measurements for
each tomographic image.
The conductivity of a slurry mixture is dependent on
the individual conductivities of the two phases, as
well as on the relative amount of solid present in the
suspension. Maxwell (1881) found that the local
volume fraction distribution can be determined
indirectly from the conductivity distribution as shown
below
20c + 2 2,,, +
"2
m 
("1
7m 2
+ 2(c, a, )
Where a is the conductivity of the first phase, 02 is
the conductivity of the second phase, and am is the
conductivity of the mixture. Neale et al. (1973) has
shown that the validity of this equation extends into a
variety of void fractions. In our case the second solid
phase has an assumed conductivity of zero therefore
equation (5) can be simplified to
2ol 2o
Es = (6)
om, + 2ao
where (l is the conductivity of the liquid phase, and
Cm is the solidliquid mixture conductivity
determined from the pixel conductivity as reported
by the reconstruction algorithm used which in this
!
case is the Modified Back Protection (MSBP)
algorithm. The liquid phase conductivity can be
measured at the start of an experiment using a
commercial conductivity meter.
3.2 Velocity Measurement
Velocity measurement were made using fast
impendence camera system (FICA), developed by
University of Leeds and has a capability of measuring
up to 987 dual frame per second. The measurements
were taken as set of block data of 8000 frames and a
method of Cross correlation can be used to establish a
relationship between the two signals from the dual
plane ERT sensor Fig. 3.
7th International Conference on Multiphase Flow,
ICMF 2010, Tampa, FL, May 30 June 4, 2010
time delay, T is the time over which cross correlation
is calculated. The correlation function reaches a
maximum for time delay tmax, this time delay being
an estimate of time taken for the measured substance
to travel between the two planes.
A direct correlation method described in detail by
Beck and Plaskowski (1987) has been developed into
the AimFlow software package at the University of
Leeds and the Chinese Academy of Sciences. The
software utilises the conductivity map produced by
the linear back projection algorithm. The axial flow is
calculated using the following equation:
k
R2(n)= Z f(m)f2(m+n)
m1
Flow
Figure 3: Point to point cross correlation (adapted
from Mosorov, 2002)
The flow velocity can be calculated by
V d (7)
= (7)
t
Where V is the flow velocity, d is the distance
between sensor 1 and sensor 2 t is the time taken for
the flow signature to be recognized by both sensors.
The cross correlation function used in this case, as
reported by Mosorov (2002), is defined by Eq. (9).
The basic function is to find a time delay that can
make the difference, R, minimum.
Rsxsy(r) = lm Sx(t)S(t )dt (8)
T> T 0
Where Sx and Sy are the signals detected from planes
x and y, R is the error function which gives , the
Where k is the sample length, n is the offset number,
fl(m) and f2(m) are the mth upflow and downflow
images respectively. The volumetric flow rate of the
dispersed phase is calculated by integrating the
product of the local volume fraction and axial
velocity of the dispersed phase in the flow cross
section:
Qd= advddA
A (10)
4. Results and Discussion
In order to determine the functionality of this concept
flow metering technique, a means of validating ERT
results will be produced. In addition to this results
from ERT will be corrected using the derived cross
correlation velocity and pressure measurements using
the ELM model. Tomography system results in two
different scenarios are presented; developed section
at 45D, and undeveloped at 10D and the insitu
concentration and velocity are the primary variables.
4.1 Equivalent liquid model validity
Figure 4a and 4b shows the pressure gradients of
solidliquid two phase mixtures according to the
discharge velocity which can be taken as the
superficial mixture velocity in the pipe according to
the previous discussion in the theoretical section. It
can be seen that there is a higher pressure drop with
increasing concentration of solids from figure la. The
hydraulic frictional gradient is plotted in Figure lb
and shows that there is no additional frictional drop
due to the increased number of particles. This would
seem to suggest that the ELM is a valid
approximation of the fluid properties in this case, as
any increased pressure gradient is due an increase in
density.
* Water
*9.40%
*18.30%
0 1 2 3 4
Mean discharge velocity (m/s)
0.7
0.6
0.5
E
 0.4
E 0.3
0.2
0.1
0
* Water
* 9.40%
*18.20% /o
* l. jfi
4.0
E
3.0
0)
n
S2.0

0 1 2 3 4
Mean discharge velocity (m/s)
Figure 4: Total pressure gradient (a) and
frictional hydraulic gradient (b) as a function of
discharge velocity
4.2 Velocity from cross correlation
The mean cross correlation velocity was obtained for
developed (45D) and undeveloped nonhomogenous
(10D) sections, using a conductivity map obtained
from the FICA system and processed using the
Aimflow software. This was compared with a
discharge measurement, which is taken at the exit of
the vertical pipe section. This is shown in Fig. 5a for
a sand discharge concentration of 9.4% and 18.3% in
Fig. 2b. It is important to note this result is consistent
and repeated in the developed and undeveloped
sections near the bend. Another observation is that
7th International Conference on Multiphase Flow,
ICMF 2010, Tampa, FL, May 30 June 4, 2010
the electromagnetic flow meter is able to give
comparable values to the discharge measurement at
45D but results are much less convincing near the
bend position at 10D, and especially when a higher
insitu concentration is used. This is a scenario where
an ERT based flow meter could provide a flow
solution as principally it should not be affected by
nonhomogenous flow conditions as many
commercial based flow meters are.
At present time it only possible to speculate at the
reason for this overrepresentation of the velocity
derived from ERT. One idea is that the sheer number
of sand particles in the flowing mixture that are of
similar characteristic essentially confusing the
correlation between each sensor plane. The second
plane may recognize an exaggerated number of
correlated particles as a result of this, thus
determining a higher velocity.
*EMF 45D
*EMF10D
a ERT velocity 45D
* ERT Velocity 10D
) 2.0 3.0
Discharge Velocity m/s
* EMF 45D
*EMF 10D
* ERT velocity 45D
* ERT Velocity 10D .' f
E
>
03.0
2
0
SrL.
1.U
1.0 1.5 2.0 2.5 3.0 3.5
Discharge Velocity m/s
Figure 5: Velocity determined from cross
correlation compared to that determined from
EMF for discharge concentration of 9.4% (a); and
compared with discharge concentration of 18.3%
4.3 Validation of ERT
Tomography data from actual slurry flow was also
recorded, 300 conductivity images were continuously
collected over approximately 5 minutes for all
experiments. These images were averaged and
merged and then the volumetric fraction profile was
obtained using the Maxwell relationship. ERT data
and discharge measurements were collected in an
equivalent time frame, so that the two methods could
be more comparable. The results presented in Fig. 3
include a comparison with the discharge
measurement, the equivalent liquid model using the
discharge velocity (denoted as ELM in Fig 3.), and
show the insitu pipe concentration derived from the
ELM model but using the ERT cross correlation
velocity (as Conc. From ERT velocity) as well as
directly obtained from ERT (as ERT). The results are
also tabulated in table 1.
0.300
0.250
0.200
0
E0.150
0.100
0
0.
0.050
0.000
0.14
0.12
C
0
" 0.08
w0.06
0.04
@0.04
* ERT
* Discharge
* ELM
xConc. from ERT velocity
t A
'C
1.0 1.5 2.0 2.5 3.0 3.5 4.0
Discharge velocity m/s
*ERT
A Discharge
SELM
x Conc. from ERT velocity
NI
*
0.02
1.0 2.0 3.0 4.0
Discharge velocity m/s
Figure 6: Concentration determined from cross
correlation velocity, compared to that determined
7th International Conference on Multiphase Flow,
ICMF 2010, Tampa, FL, May 30 June 4, 2010
from EMF and ELM for 9.4% discharge
concentration (a); for 18.3% discharge (b)
It can be seen that insitu concentration determined
by ERT are generally lower than those determined by
discharge measurement and from pressure drop.
However the result is consistent with the general
trend and has a regular offset. At 9.6% loading
concentration of sand the ERT records an insitu
concentration of average 2.5% lower than the
discharge, and at 18.4% the result gives an offset of
3.6% volumetric concentration. As both profiles
show a similar trend of steeply decreasing
concentration with increasing velocity, then this
phenomena is likely to be present in the flow system
and not just erroneous data. It is unexpected as
previous experimentation, Xu et al., (2007), did not
reveal such dramatic concentration sensitivity to fluid
velocity. This effect can be explained with the
principle of mass conservation and flow continuity,
but due to the large changes observed it is more likely
it is being influenced by mixing in main tank or flow
loop arrangement. Nevertheless the ERT system has
demonstrated that with appropriate calibration it will
be able to give reliable repeatable results.
Results from the ELM model were determined using
insitu pressure measurement, and mean velocity
from the discharge measurement. The ELM model
has predicted results with some similarities to the
discharge measurements. The model has a tendency
to consistently over predict and particularly at higher
velocities beyond 3m/s. Current research has shown
that the ELM model does not predict accurately in
scenarios where fine/medium sand (d5o=120tm) is
used Matousek (2002) especially when high
velocities are involved. As coarser particles exhibited
even lower solids friction and concentration affects at
high velocities than the predicted the author argues
that a repelling effect from hydrodynamic lift is
present. It is not sufficient to simply treat the slurry
mixture as having equivalent frictional properties of
the carrier phase but also consider the Bagnold
collision force Bagnold (1954) and liquid lift force in
the near wall region to treat mixtures of varying
transport properties.
7th International Conference on Multiphase Flow,
ICMF 2010, Tampa, FL, May 30 June 4, 2010
Conc.
CERT onc.m D e EM Discharge ERT ERT EMF EMF Discharge ERT Mass
ConE ERT Conc. Con.e velocity velocity velocity 45D 10D Mass Flow Flow Rate
ET Com/s 45D m/s 10Dm/s m/s m/s Rate kg/s kg/s
velocity
0.071 0.099 0.096 0.108 1.45 1.72 1.82 1.45 1.50 3.29 3.77
0.067 0.089 0.093 0.099 1.99 2.44 2.4 2.05 1.93 4.49 5.31
0.065 0.085 0.085 0.098 2.56 2.95 3.03 2.56 2.48 5.72 6.40
0.061 0.080 0.085 0.097 2.98 3.43 3.45 3.00 2.72 6.67 7.40
0.052 0.080 0.073 0.095 3.52 4.26 4.04 3.49 3.25 7.73 9.06
0.186 0.211 0.220 0.227 1.41 1.9 1.93 1.42 1.58 3.76 4.87
0.174 0.197 0.216 0.212 2.08 2.4 2.47 2.10 2.23 5.53 6.05
0.164 0.191 0.194 0.208 2.49 2.86 2.95 2.49 2.48 6.45 7.12
0.148 0.181 0.187 0.203 3.05 3.4 3.45 3.09 2.91 7.81 8.30
0.134 0.172 0.183 0.202 3.58 4 3.92 3.53 3.40 9.15 9.58
Table 1: Discharge values for concentration and velocity compared with those derived through cross
correlation using ERT
An important aspect of this study is to determine a
means of correcting ERT concentration reading in a
continuous process when the possibility of taking a
new reference measurement is not possible. This is
done by using a differential pressure measurement in
the pipe and using the ELM model, but substituting
the mean superficial velocity measured by the
discharge for that of the ERT derived cross
correlation velocity. This result is also shown in
figure 3, and typically gives a value higher than the
ERT measured concentration and very close to the
discharge value. It is important to note though the
ERT determined mean velocity is over predicted in
the first place and a much closer representation to the
ELM model prediction would be likely if this wasn't
the case.
The mass flow rate of the mixture can be determined
by combining measurements for concentration and
velocity, the result of which is shown in Fig 7. The
individual readings for ERT concentration and
velocity are under and over predicted respectively,
and this dual effect helps the ERT measurement to
realise values close to that determined from discharge
measurements, albeit still showing an offset of 11%
on average. For the sole purpose of determining mass
flow rate in this case the over determination from the
ERT velocity has had a bigger contribution overall in
the final offset.
10
9 A
8
E 5
g 4
3
3 4 5 6 8 9 10
Discharge mass flow rate kg/s
Figure 7: Insitu mass flow rate determined from
ERT derived concentration and velocity,
compared with mass flow rate from discharge
measurements
0.25
0.2
0.15 __  
0
o 0.1
0.05 "'2 "
0
0 0.2 0.4 R 0.6
r/R
0.8 1
0.25
0.2
0.15
o
C)
0
0 0.1
0.05
0 0.2 0.4 0.6
r/R
9.4%
 = 1.45m/s
 V = 1.99m/s
V = 2.56m/s
V = 2.98m/s
V = 3.52m/s
0.8 1
18.3%
SV=1.41m/s
 V = 2.08m/s
 V = 2.49m/s
V = 3.05m/s
V = 3.58m/s
Figure 8: Insitu slurry concentration profiles
from ERT at 45D (a); and 10D (b) from bend, for
a range of discharge concentrations and velocities
Fig. 8 (a) and 8 (b) shows velocity profiles of slurry
mixtures at varying velocities when ERT is
positioned 45D (fully developed flow) from the bend
and 10D (undeveloped flow) from the bend
respectively. A striking feature of the flow in the
undeveloped section is the major concentration of
flowing mass is off centre but away from the wall on
the far side of the bend exit. This is typically
attributed to lift force in the wall region in
horizontally flowing slurries, but similar phenomena
can be seen here. Subjectively the result is similar to
that of Huber and Sommerfeld (1994) who conveyed
particles pneumatically and observed similar profiles
near bend regions. The nonuniformity gradually
formed a developed section downstream as
accumulation near outer wall will disintegrate due to
the secondary flow induced by the bend and due to
flow turbulence. It is clear that if ERT can produce
profiles able to see changing features then in fully
developed section it should be able to also capture the
coring effect that is described previously for larger
particles. Following this it is hoped further
7th International Conference on Multiphase Flow,
ICMF 2010, Tampa, FL, May 30 June 4, 2010
experimentation using ERT will be able to visually
represent this phenomena.
5. Conclusion
It is seen that typically ERT is able to reliably predict
the insitu concentration of the pipeline but with an
offset from the discharge values. An investigation
into the effects of solids on ERT measurement can be
introduced as a means of calibration, and applying
correction to measurements when in pipeline.
Concentration derived from mechanistic models
agreed somewhat with discharge measurement but
tended to be slightly over predicted and especially at
higher velocities. A more robust model is required to
validate using this model particularly in the treatment
shear stress in the near wall region; potentially this is
a good method of lending credence to discharge
measurements if good agreement is found between
the two measurement schemes.
ERT is also able to determine the velocity of the
dispersed phase, and it was able to provide repeatable
data, but which was consistently over predicted.
Certainly it is encouraging that cross correlation
technique is able to produce results especially in
homogenous flow conditions, and encouraging
enough to warrant further investigation to rectify
problems with resolution brought about inadequate
sensor spacing.
The velocity derived from ERT can also be used as a
means of correcting the ERT concentration
measurement using a differential pressure
measurement in an allied scheme. This provides a
basis of providing an updated concentration
measurement in a continuous process that is
independent of temperature and ionic solution effects,
in effect it eliminates the need for an updated
reference that ERT measurement is conventionally
dependent on.
a~Eff~ha;c~_g
7th International Conference on Multiphase Flow,
ICMF 2010, Tampa, FL, May 30 June 4, 2010
References
Bagnold, R.A., (1954), Experiments on a gravityfree
dispersion of large solids spheres in a Newtonian
fluid under shear, Proc. Roy. Soc. A 225 pp. 403.
Baker, R.C., (2000), Flow Measurement handbook,
Industrial Design Operating Principles, Performance,
and Applications, Chapter 17, Cambridge, UKL
Cambridge Univ. Press
Cha, J., Ahn, Y., Kim, M., (2002), Flow
measurement with an electromagnetic flowmeter in
two phase bubbly and slug flow regimes in Flow
measurement and instrumentation, Vol 12, issues 56,
pp 329339
Durand, R. (1951a), Transport hydraulique des
materiaux solides en conduite, etudes experimentales
pour les cendres de la central Arrighi, Houille
Blanche, Vol.6, No. 3, pp. 384393.
Hammer E A, Johansen G A, 'Measurement
Principles in Multiphase Metering Their Benefits
and Limitations', The Future of Multiphase Metering
Conf, 2627 Mar 1998.
Loh W W, 'Realtime Monitoring of Drilling
Cuttings Transport using Electrical Resistance
Tomography', PhD Thesis, UMIST, Sep 1998.
Mosorov, V., Sankowski, L., Mazurkiewicz., L.,
Dyakowski, T., (2002), "The 'bestcorrelated pixels'
method for solid mass flow measurements using
electrical capacitance tomography"; Meas. Sci.
Technol. 13, 18101814.
Neale G H and Nader W K, 'Prediction of transport
process within porous media: diffusion flow
processes within an homogenous swarm of spherical
particles', AICHE Jour 19, pg112119., 1973.
Wang M., Ma Y., Holliday N., Dai Y., Williams R.A.
and Lucas G.P., (2005), A High performance EIT
System, IEEE Sens. J., Vol. 5, pp. 289
