- Email: [email protected]
Experimental study on pipeline internal corrosion based on a new kind of electrical resistance sensor
Accepted Manuscript Title: Experimental study on pipeline internal corrosion based on a new kind of electrical resistance sensor Author: Yunze Xu Yi Huang Xiaona Wang Xuanqin Lin PII: DOI: Reference:
S0925-4005(15)30496-2 http://dx.doi.org/doi:10.1016/j.snb.2015.10.030 SNB 19174
To appear in:
Sensors and Actuators B
Received date: Revised date: Accepted date:
31-12-2014 8-6-2015 12-10-2015
Please cite this article as: Y. Xu, Y. Huang, X. Wang, X. Lin, Experimental study on pipeline internal corrosion based on a new kind of electrical resistance sensor, Sensors and Actuators B: Chemical (2015), http://dx.doi.org/10.1016/j.snb.2015.10.030 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Experimental study on pipeline internal corrosion based on a new kind of electrical resistance sensor Yunze Xua,*, Yi Huanga, Xiaona Wangb,**, Xuanqin Lina a School of Naval Architecture and Ocean Engineering, Dalian University of Technology, Dalian 116024;
* Corresponding author. Tel: +86-15898193759; fax: +86-0411-84706061. **Corresponding author. Tel: +86-0411-84706061; fax: +86-0411-84706061. E-mail address: [email protected] (Yunze Xu); [email protected] (Xiaona Wang);
ip t
b School of Physics and Optoelectronic Engineering, Dalian University of Technology, Dalian 116024
cr
Postal address: School of Naval Architecture and Ocean Engineering, Dalian University of Technology, Linggong Road 2#, Dalian 116024, Liaoning Province, China.
us
Abstract:
In this paper, a ring pair electrical resistance sensor (RPERS) has been developed for an internal-pipeline corrosion on-line monitoring system. The RPERS was divided into six segments along the circumference. The
an
corrosion depth of each segment could be measured by using three alternating excitation currents injected into the sensor from different angles. In order to simulate and monitor the pipeline internal corrosion, a corrosion monitoring system which contained RPERS, wire electrical resistance sensor (WERS), thermocouples and a
M
corrosion coupon was established. First, the corrosion processes in 3.5% sodium chloride solution with the temperature varied from 30℃ to 60℃ were studied. The temperature differences between the inner and outer pipe wall surfaces were measured by the top segment of RPERS. The test results revealed that the performance of RPERS
d
is better than that of WERS in term of metal loss measurement. Then, carbon dioxide mixed with water vapour was pumped into the system with the temperature varied from 50℃ to 80℃. In the scenarios, top of the line corrosion
Ac ce pt e
(TLC) occurred and was monitored by RPERS. The monitoring results demonstrated that the gas temperature and the temperature difference were important factors for TLC in sweet conditions. Key words: ring pair sensor, internal corrosion, on-line monitoring, temperature difference, TLC
1. Introduction
The electrical resistance (ER) sensor technique, also known as electrical coupon technique,
has become one of the most important methods for pipeline internal corrosion monitoring [1]. As a result of corrosion process, the cross-section area of the sensor is reduced, leading to an increase in electrical resistance. Based on the change of the resistance, the corrosion loss in metal thickness can be obtained and the slope of the metal loss curve directly corresponds to the corrosion rate. In contrast to conventional electrochemical techniques such as the linear polarization resistance and the electrochemical impendence spectroscopy methods, the advantage of the ER technique on metal loss measurement is that the presence of an electrolyte layer on the metal surface is not inevitable. In that sense, the method can be used in a high resistance medium such as gas and crude oil. Moreover, it can be used for corrosion monitoring of high temperature corrosion and erosion corrosion [2, 3].
1
Page 1 of 43
The traditional ER sensors are almost entirely composed by a corrosion element and a compensation element. The geometry shapes of the elements are usually present in the forms of wire, tube, flush spiral and strip. The corrosion element of the ER sensor is exposed to corrosion surroundings and the compensation element is covered with a protective coating to segregate it from the corrosive environments [4, 5]. Through the resistance ratio of the two elements, the temperature interference can be eliminated under ideal conditions. Thus, the metal loss of the
ip t
corrosion element can be calculated according to its geometric dimensioning. However, in the pipelines with high temperature and pressure, the two elements of the traditional ER sensors may be
cr
subjected to different temperature and pressure conditions due to the different locations where they are emplaced. It has been found that a minor difference of 0.25℃ between the corrosion element
us
and the compensation element will cause a significant change in the resistance ratio of 1000 ppm. Although the effect of the pressure difference on the resistance ratio is much less than that caused by the temperature difference, the resistance ratios of typical pipeline steels still undergo obvious
an
changes from 2000 ppm to 4000 ppm per 100 bar [6, 7].
In recent years, localized corrosion problems such as bottom of the line corrosion (BLC) and
M
top of the line corrosion (TLC) have become focus issues of the pipeline internal corrosion [8-10]. In general corrosion surroundings, the traditional ER sensors can be sensitive to homogeneous corrosion due to their geometric forms, but the responses of these sensors to localized corrosion are
d
limited [11]. With improvements on the resolution of the electrical resistance measurement, some
Ac ce pt e
researchers have tried to use the ER sensors for localized corrosion monitoring in pipelines by changing their forms. Marc Singer [12] has used a new kind of flush head ER sensor for TLC monitoring. However, the monitoring results are dissatisfactory due to the edge effect caused by the pipeline arc. BLC is always caused by sand and corrosion products covering the bottom of the line. The deposit on the steel surface may lead to failures of the corrosion inhibitors and cause serious localized corrosion. The practical corrosion rate of the bottom area in the pipeline is hard to be monitored due to the galvanic effect between the bare steel and the deposit-covered steel [13]. For most oil and gas pipelines, the inner fluids usually endure high temperatures above 50℃,
while the temperature of the outer surroundings is always much lower than the fluid temperature, especially in subsea surroundings [14]. In these scenarios, the temperature differences between the inner and outer pipe wall surfaces exist. Gerstmann [15] found that the temperature differences in sweet conditions may lead to the formation of condensation films on the inner surfaces of the pipelines, which is an important factor for TLC. The outer temperature of the pipe wall surface can be easily measured by arranging the temperature measurement sensor outside of the pipe wall surface. However, it is hard to measure the temperature of the inner pipe wall surface in practice without destroying the pipe wall structure. Due to the unknown temperature difference between the
2
Page 2 of 43
inner and outer pipe wall surfaces, Vitse and Zhang et al. [16, 17] established complex thermodynamic models to forecast the condensation rates, but the predicted results still have deviations compared with the experimental results. The primary purpose of this paper is to present a new kind of ring pair ER sensor (RPERS) for pipeline internal corrosion monitoring. RPERS is composed of two rings, i.e. a corrosion element and a compensation element. The rings are made with an 8 inch X65 pipeline which is usual for
ip t
subsea installations. The RPERS is divided into six segments and used to monitor the metal losses of these segments in a laboratory system. In the experiment, the temperature difference between the
cr
inner and outer pipe wall surfaces is measured by the top segment of RPERS. The metal loss
measurement results of RPERS are compared with those obtained from a traditional wire ER sensor
us
(WERS) and a corrosion coupon. In the experimentation of TLC, the metal loss, corrosion rate and temperature difference of the top segment in the pipeline are monitored by RPERS. The relationship of the gas temperature, temperature difference between the inner and outer pipe wall surfaces and
an
corrosion rate is obtained. RPERS also provides a new measurement method for thermodynamic
2. Method
M
model research within the experiment.
d
2.1. Internal-Pipeline corrosion monitoring system
Ac ce pt e
As shown in Fig. 1, the internal-pipeline corrosion monitoring system is composed of three parts: pipeline information detecting section, data acquisition system and data-managing section. The pipeline information detecting section contains RPERS, traditional WERS, K-type
thermocouples and a corrosion coupon. It is the sensitive part of the monitoring system. The sensors provide continuous signals of the internal corrosion conditions and the temperature information in the pipeline.
In the data acquisition system, the ER signals generated from RPERS can be converted to
digital signals by an RM3545 micro resistance measurement device whose ER measurement resolution is 10 nΩ. The models of the K-type thermocouples are TP-K01, and the measurement surfaces of the thermocouples are coated with a layer of teflon for insulation. The temperature measurement accuracy of the thermocouples can reach ±0.1℃ in the range of 0~80℃ after having been calibrated in a Julabo FP 51 thermostatic water bath. The temperature control accuracy of the thermostatic water bath is ±0.05℃ in the range of -20~100℃. The voltage signals generated from the thermocouples are recorded by NI PCle-6320 data acquisition card. The ER signals from WERS can be obtained by CMB 1510b. The CMB 1510b is a corrosion data acquisition and storage device which is manufactured by Instituted of Corrosion & Protection of Metals in China. All the signals 3
Page 3 of 43
are transferred to the computer after disposal by a RS232 transition port. The data-managing section is an operation platform. It is compiled by C# to record and process messages conveyed from the Data Acquisition System (DAS). All data is saved in the Structured Query Language (SQL) Server database.
ip t
2.2. The construction and measurement principle of RPERS
As shown in Fig. 2a, RPERS is a part of the pipeline which needs to be monitored. The corrosion element and compensation element have the same inner diameters with the pipeline. The
cr
two elements are embedded into the pipeline through a connecting element, which is shown in Fig.
2b. The axial widths of both elements are 10 mm and the initial sensitive wall thickness of the
us
corrosion element is 13.5 mm. The rings are coated with a layer of chromic oxide by plasma spraying except the internal surface of the corrosion element, which is shown in Fig. 2c. The two
an
elements are insulation from other pipe surfaces through the Cr2O3 coating on the contact surfaces and the insulation resistance is higher than 1 MΩ among each element at a voltage of 50 V. A better temperature and pressure compensation effect can be provided as the compensation element and
M
the corrosion element are coaxial concentric placed in the pipeline. The monitoring results of the localized corrosion, i.e. BLC and TLC, will be closer to the actual situations than traditional flush
d
head ER sensors since the arc of the corrosion element is the same as the pipeline to be monitored. RPERS can provide two monitoring functions, i.e. the monitoring of the localized corrosion in
Ac ce pt e
pipelines and the measurement of the temperature difference between the inner and outer pipe wall surfaces. The two functions are introduced as follows. For localized corrosion monitoring, each element are divided into six segments using
electrodes 1, 3, 5, 7, 9 and 11, as shown in Fig. 3. Segment 1 is between electrodes 1 and 3, segment 2 is between electrodes 3 and 5 and the other segments are dealt with the same approach. These electrodes are also used for voltage measurements. The resistance values of the segments on the corrosion element are represented by
(
), respectively. The resistance values
of the segments on the compensation element are represented by
(
),
respectively. The six segments may suffer different temperatures or corrosion environments due to the different locations and flow regimes in the pipeline. Galvanic current may generate in RPERS and cause a measuring error [18]. To eliminate the additional potential caused by the galvanic current in the ring, the square excitation current waves are injected into the ring pair for ER measurement. The amplitude of the current is 1 A and the measurement period for each signal is 241 ms. In a
4
Page 4 of 43
measurement period, the voltage signals negative current
and
caused by the positive current
are measured. Then the average voltage
and
used as the final measurement
result can be calculated by:
V=
V+ + | V− | 2
(1)
and
are successively injected into the ring pair through the
even-numbered electrodes, as shown in Fig. 3. Firstly, when
is injected as the red path, the
cr
Three excitation currents
ip t
In this way, the measurement error of the voltage caused by the galvanic current can be eliminated.
us
potential differences between electrodes 1 and 3, 3 and 5, 9 and 7, 11 and 9 are respectively measured as the voltages of segments 1, 2, 4 and 5 on the corrosion and compensation elements.
(
(
) and
an
Through the measurement results, the resistance ratios ) can be expressed as:
(2)
cor ( comp ) cor ( comp ) ε cor 2( comp 2) = R4cor (comp ) / R5cor (comp ) =V9,7 / V11,9( I1 ) ( I1 )
(3)
d
M
(comp ) cor ( comp ) ε cor1( comp1) = R1cor (comp ) / R2cor (comp ) =V1,cor / V3,5( I1 ) 3( I1 )
and
represent the resistance ratios between segments 1 and 2 of the two elements,
Ac ce pt e
where
represent the resistance ratios between segments 4 and 5 of the two elements.
,
,
and
respectively represent the voltage
measurement results of segments 1, 2, 4 and 5 of the two elements when Then, when
is passing through.
is injected into the ring pair as the blue path shown in Fig. 3, the voltage values
of segments 1, 3, 4 and 6 can be measured successively using the same steps as through. The resistance ratios
(
) and
(
is passing ) can
be expressed as: cor ( comp ) cor ( comp ) ε cor 3( comp 3) = R6cor (comp ) / R1cor (comp ) =V11,1 / V1,3( I2 ) (I2 )
(4)
cor ( comp ) cor ( comp ) ε cor 4( comp 4) = R4cor (comp ) / R3cor (comp ) =V9,7 / V7,5( I2 ) (I2 )
(5)
5
Page 5 of 43
where
represent the resistance ratios between segments 6 and 1 of the two elements, represent the resistance ratios between the segments 4 and 3 of the two elements.
,
and
respectively represent the voltages of
segments 1, 3, 4 and 6 of the two elements when Finally, when
is passing through.
is injected as the purple path shown in Fig. 3, the voltage values of segments
2 and 3 can also be measured and the resistance ratios
(
cor ( comp )
ε cor 5( comp 5) = R3
=V7,5( I(3 )
cor comp )
cor ( comp ) / V5,3( I3 )
)can be
(6)
an
represent the resistance ratios between segments 3 and 2 of the two elements,
and
is passing through.
d
two elements when
respectively represent the voltages of segments 2 and 3 of the
M
and
cor ( comp )
/ R2
us
expressed as:
where
ip t
,
cr
and
Ac ce pt e
Because the two rings are in series, the current passing through the corrosion element is equal to that passing through the compensation element. Thus when
is passing through, Eq. (7) can be
written as:
cor V1,3( I1 )
R1cor
+
cor V11,9( I1 )
R5cor
=
comp V1,3( I1 )
R1comp
+
comp V11,9( I1 )
R5comp
= I1
(7)
Combining Eqs. (2) to (7), Eq. (8) can be obtained:
6
Page 6 of 43
comp i
i=2 i=3 (8)
i=4 i=5
cr
/R
i =1
ip t
ki = R
cor i
cor comp V1,3( I1 ) + V11,9( I1 )ε cor 2ε cor 1 / ε cor 4ε cor 5 comp comp V1,3( I1 ) + V11,9( I1 )ε comp 2ε comp1 / ε comp 4ε comp 5 ε comp1 k1 ε cor1 ε cor 5ε comp1 k ε comp 5ε cor1 1 = ε cor 4ε cor 5ε comp1 k 1 ε comp 4ε comp 5ε cor1 ε cor 4ε cor 5 ε comp 2ε comp1 k1 ε cor 2ε cor1 ε comp 4ε comp 5 ε cor 3 k1 ε comp 3
us
,
is the resistance ratio between the corresponding segments of the
an
where
i=6
corrosion and compensation elements, and
is the resistance ratio between
and
. As
M
is obtained, the metal loss of each segment can be calculated according to the geometric dimensioning of the ring.
d
The temperature difference between the inner and outer pipe wall surfaces of each segment can also be measured by using RPERS. As mentioned above, the compensation element is coated with a
Ac ce pt e
layer of chromic oxide coating for anticorrosion. The resistance values of the segments on the compensation element only change with the fluctuation of the temperature. When a temperature difference exists, there will be a nearly liner temperature gradient along the diameter direction [19], which is shown in Fig. 4a.
and
surfaces, respectively, and
is the difference between
Because the wall thickness
represent the temperatures of the outer and inner pipe wall
and the width
and
.
along the axial direction are much smaller
than the arc length a in a ring element, which is shown in Fig. 4a, the geometric model of a segment can be simplified as a rectangle model, as shown in Fig. 4b. The temperature gradient is shown in Fig. 4c. At the height of z along the vertical direction, pick up the thin rectangular plate with a thickness of
, the resistance of the thin plate can be described in Eq. (9):
dR = ρ 0 where
[1 + α(Tz − T20 )]a b × dz
is the electrical resistance of the thin rectangular plate,
(9) is the electrical resistivity of
7
Page 7 of 43
the steel material at 20℃,
is the temperature coefficient of the resistance,
of the thin rectangular plate, and
equals to 20℃.
Tz = Ti −
can be calculated by Eq. (10):
Ti − To z Δr
(10)
in Eq. (7) is a known constant, the electrical resistance of each
ip t
As the current value of
is the temperature
segment can be calculated from Eqs. (2) to (7). The electrical resistance of each segment on the
) can be measured in the temperature control device before
cr
compensation element at 20℃ (
being installed into the pipe. The outer pipe wall temperature of segment (
) can be measured by
us
the temperature sensor arranged on its outer pipe wall surface. Because the thin rectangular plates are connected in parallel within a segment on the compensation element, the resistance of segment ) can be calculated by solving Eq. (9). The calculated result is shown in Eq. (11):
an
(
,
is the inner pipe wall temperature of segment
and the value of
which cannot be
can be calculated by using the numerical calculation method,
Ac ce pt e
directly monitored. From Eq. (11),
(11)
d
where
1 + α (Tii − T20 ) 1 + α (Toi − T20 )
M
Ricomp = Ricomp 0 α (Tii − Toi ) / ln
can also be obtained.
As the geometric shape of the corrosion element is also a ring form like the compensation
element, the resistance of the corrosion element
can also be calculated when a temperature
difference exists. When the metal loss of segment
on the corrosion element is
, Eq. (12) can be
obtained by using the same derivation processes as the compensation element.
Ricor = Ricor 0
α∆r
∆r − xi
(Tii − Toi ) / ln
1 + α (Tii − T20 ) 1 + α (Toi − T20 )
(12)
Combining Eq. (8), Eq. (11) and Eq. (12), the metal loss can be determined:
xi = ∆r × (1 − The resistance ratio of
and
Ricor 0 ) comp ki Ri 0
(13)
in Eq. (13) can be measured before corrosion occurs and
used as a constant. From the calculative process, it can be found that whether a temperature
8
Page 8 of 43
difference exists or not, it makes no difference to the calculated metal loss in theory.
2.3 The corrosion monitoring circulation system in the laboratory
As shown in Fig. 5a, the pipeline circulation system in the laboratory is composed of six parts. The temperature control heater and the PT100 temperature sensor are installed in the water tank to
ip t
control the fluid temperature. The corrosion liquid or gas could be pumped into the monitoring system by using different kinds of pumps. The traditional WERS is directly inserted into the 8 inch
pipe with its data collection device installed out of the pipe. The main chamber which contains
cr
RPERS and two thermocouples is butt welded by two pieces of reducer pipes, and it is filled with
heat transfer oil. The two thermocouples are respectively emplaced on the inner and outer pipe wall
us
surfaces of the top segment in the pipeline. The corrosion coupon is hung nearly at the end cap of the delivery port that could be removed for weighing after the experiment. The data acquisition
an
system and power supply devices are placed in the electronics chamber.
All the corrosion sensors and the corrosion coupon are made of X65 pipeline steel of which chemical composition (wt%) is shown in Table 1.
M
As shown in Fig. 5b, the whole system was working in a simulative subsea environment during the test period. The water temperature was monitored for 12 days and was fluctuated within 1℃ in
d
the laboratory, which is shown in Fig. 6. Firstly, 3.5% sodium chloride solution by weight with a PH of nearly 6.5 was pumped into the system, and the solution was launched to circulate with oxygen
Ac ce pt e
continually providing. The flow rate of the solution was 0.3 Lpm. When the flow regime and the temperature of the solution became stable, the solution temperature was successively increased from 30℃ to 60℃. Finally, the temperature differences (
) and metal losses were recorded. After
the first experiment, the corrosion coupon was taken out and used to weigh the weight loss. After the solution was released from the system, oxygen gas in the system was eliminated by pure nitrogen. Then, carbon dioxide mixed with water vapour was pumped into the system for circulation, and the flow rate of the gas was 1.5 Lpm. The total pressure was 0.5 bar and the partial pressure of the carbon dioxide was 0.3 bar. Varying the gas temperature from 50℃ to 80℃, the values of
and
the metal loss were monitored by the top segment of RPERS. 3. Experimental Results
3.1 The temperature difference measurement results using RPERS and thermocouples
As mentioned above, the two thermocouples were respectively mounted on the inner and outer 9
Page 9 of 43
surfaces of the pipeline. One thermocouple was placed on the outer pipe wall surface of the top segment on the compensation element, and the outer surface’s temperature ( ) of the top segment could be measured. Another thermocouple was inserted into the pipe near the compensation element and it was flush with the inner pipe wall surface of the top segment. The inner surface’s
temperature difference (
ip t
temperature ( ) of the top segment could also be measured. Based on the measurement results, the ) of the top segment could be obtained.
the value of
could also be calculated from Eq. (11) in conjunction with the measurement value
which was obtained by the thermocouple emplaced on the outer pipe wall surface. Fig. 7
us
of
cr
Through the resistance measurement results of the top segment on the compensation element,
shows the temperature difference measurement results of the top segment monitored by the both
when gas was flowing past the system.
M
results of
an
two methods when 3.5% NaCl solution was flowing past the system. Fig. 8 shows the measurement
3.2 The metal loss measurement results of RPERS and traditional WERS
d
Fig. 9 presents the measurement result of the metal loss monitored by WERS which was
Ac ce pt e
inserted in the pipe. Fig. 10 shows the six metal loss curves monitored by the six segments of RPERS in the 3.5% NaCl solution. In the corrosion monitoring system, RPERS was placed in the middle of the pipeline. Segments 3 of the both elements were placed at the bottom of the line and segments 6 were placed at the top of the line. Segments 1 and 5 were symmetrically placed on the upper half of the pipeline and segments 2 and 4 were symmetrically placed on the lower half of the pipeline.
It can be found that the slopes of all the metal loss curves increase with the raising of the
fluid temperature from 30℃ to 60℃. It can be seen from Fig. 9 that there were two saltus steps on the metal loss curve. The accumulative measurement error of WERS reached nearly 5 µm when the fluid temperature increased to 60℃. As shown in Fig. 10, a better compensation effect could be seen from the measurement results of RPERS. From the measurement results, it could be found that the metal losses of the six segments were not uniform. Because the vent valve was located slightly beneath the top of the pipe, some air was still remaining in the pipe after the solution was introduced into the system. The top segment did not have a thorough contact with the solution due to the remaining air which led to the metal loss of the top segment less than 2 µm in the experiment. The corrosion depth of segment 3 was 13.4 µm which was higher than that of the other segments. 10
Page 10 of 43
The corrosion depths of the symmetric segments 2 and 4 were 11.9 µm and 11.4 µm, respectively. The corrosion depths of the symmetric segments 1 and 5 were 7.8 µm and 9.2 µm, respectively. After the experiment, the corrosion coupon was taken out and the corrosion products were cleared by ultrasonic cleaning. The metal loss which was converted from the coupon weight loss was 10.2 µm. The average metal loss of segments 1 to 5 which were completely beneath the solution level was 10.7 µm. The general corrosion depths monitored by the two methods were of a good
ip t
consistency.
cr
3.3 The monitoring results of top of the line corrosion
As shown in Fig. 11, when carbon dioxide mixed with the water vapour was pumped into the
us
system, top of the line corrosion commenced and the metal loss was monitored by the top segment of RPERS. The corrosion rate was calculated from the slope of the metal loss curve, and is shown in
with the raising of the temperature difference
an
Fig. 12. It can be found that when the gas temperature was below 80℃, the corrosion rate increased which is shown in Fig. 8. However, after the gas
M
temperature reached 80℃, the corrosion rate immediately achieved to the highest value nearly 1.20 mm/y and then significantly decreased to 0.41 mm/y after 9 days. The highest corrosion rates of segments 1, 2, 4 and 5 were 0.63 mm/y，0.94 mm/y, 1.05 mm/y
d
and 0.61 mm/y, respectively. The final corrosion depths of these segments were 23.8 µm, 29.6 µm,
Ac ce pt e
31.2 µm and 22.6 µm, respectively, which were lower than the 39.4 µm of segment 6. The corrosion rate of the bottom segment 3 was higher than that of the top segment 6. The highest corrosion rate of segment 3 was 1.56 mm/y since the condensation water of the other segments accumulated at the bottom of the pipe. The finial corrosion depth of segment 3 was 46.7 µm. The corrosion rates of segments 1 to 5 also decreased to the values lower than 0.5 mm/y after the temperature reached 80℃ for 9 days. 4. Discussions
4.1 The analysis of the temperature difference measurement results
Because of some thermal insulation parts in the system, the real temperature difference between the inner and outer pipe wall surfaces is smaller than the difference between the seawater temperature and the fluid temperature. It can be seen from the measurement results that the values of
measured by the both methods are basically consistent. Though the temperature difference
measured by the thermocouples are a little higher than RPERS when the fluid temperature is above 11
Page 11 of 43
40℃,the maximum deviation of the
value is less than 0.4℃.
The measurement differences may be caused by two factors: Firstly, the thermocouples are arranged on the top of the pipe and the probe can only focus on a small area compared with the top segment of RPERS. Secondly, the temperature distribution may not be completely uniform in the top segment as the assumption above. However, compared with the results measured by the
ip t
thermocouples whose measurement accuracy is ±0.1℃, the temperature difference measured by RPERS are credible when combined using the measurement result of the thermocouple emplaced
cr
on the outer pipe wall surface.
us
4.2 The analysis of the metal loss measurement results in 3.5% NaCl solution
It can be seen from Fig. 9 and Fig. 10 that the metal loss curves become steep with the fluid increasing from 30℃ to 60℃. It indicates that the corrosion rate will increase with
an
temperature
the rising of the solution temperature. From the measurement result of WERS, a measurement error of the metal loss can be found. As shown in Fig. 8, when the solution temperature changes from 40
M
℃ to 50℃ and from 50℃ to 60℃, the temperature differences increase to 10℃ and 15.8℃, respectively. Though the carbon steel probe has an excellent thermal conductivity and the
d
compensation element of the WERS is parceled by the shrinkable tube, the temperature of the
Ac ce pt e
corrosion element is still higher than the compensation element because the corrosion element is directly inserted into the high temperature solution. The measurement error can be expressed as:
δx = r0
where
kt k0
α δT − 1 1 + 1 + α Tr
is the measurement error of the metal loss,
(14)
is the initial diameter of the wire ER sensor,
is the instantaneous monitoring result of the resistance ratio between the compensation and corrosion elements,
is the initial resistance ratio of the compensation and corrosion elements
which should be measured in the temperature control device before the corrosion occurs, temperature difference between the corrosion and compensation elements and
is the is the
temperature of the compensation element. The temperature difference leads to nearly 5 µm accumulative measurement error when the
increases to 15.8℃. The metal loss measurement
result is hard to be corrected because the real temperature of the two elements in Eq. (14) is hard to be monitored in practical application. 12
Page 12 of 43
It can be seen from Fig. 10 that the temperature compensation effect of RPERS is better than WERS. Since the corrosion rate of each segment can be monitored, the localized corrosion conditions of the pipeline can be reflected. As the flow rate in the system is 0.3 Lpm, the corrosion products may have accumulated on segments 3~5 under the low flow rate. The different coverage conditions of the corrosion products will cause the differential aeration on the internal surface among each segment and lead to galvanic corrosion [20]. The metal loss of segment 3 is higher than
ip t
those of the other segments. It may be caused by the coverage of corrosion products which lead to the forming of local anode area at the bottom of the pipe. The partial regions of segments 2 and 4
cr
bordering on the bottom segment may also be covered by the corrosion products, and it will result in
the corrosion rates of the two segments being higher than segments 1 and 5. It can be seen from Fig.
us
10 that the corrosion depths of the symmetry segments are nearly the same, which indicates that the distribution of the corrosive medium is approximately the same at the identical liquid level and there are no obvious pits corrosion patterns occurring in the symmetry segments.
an
Fig. 10 also shows that the fluctuation noise of the metal loss measurement results generated from RPERS is nearly ±0.3 µm which is higher than the ±0.14 µm of WERS. It is mainly caused
M
by two factors, i.e. the difference of the resistance values and the difference of the sensitive thicknesses between the two kinds of ER sensors. The resistance values of the both elements on WERS are near 2.2 mΩ at 25℃. However, the resistance values of the segments on RPERS are
d
only close to 0.27 mΩ at 25℃. The voltage signals generated from the RPERS are weaker than
Ac ce pt e
WERS at the same excitation current values and liable to be influenced by the surrounding electromagnetic noises. As the sensitive diameter of WERS is 1.2 mm and the sensitive wall thickness of RPERS is 13.5 mm, the resistance change in proportion to the initial total resistance of WERS is much higher than that of RPERS when the corrosion depths of the two sensors are at the same level. It causes the metal loss resolution of WERS to be better than that of RPERS. Based on these two factors, the fluctuation noise of RPERS is higher than that of WERS. Through the metal loss comparison result between RPERS and the corrosion coupon, it can
be found that the general corrosion depths monitored by the two methods have a good consistency. From the measurement results, it can be seen that RPERS can not only provide the general corrosion conditions of the pipeline, but can also reflect the localized corrosion status through the six segments.
4.3 The analysis of the TLC measurement results
The partial pressure of the water vapour is 0.2 bar during the test. As the corrosion system is arranged at the sea floor conditions with the temperature close to 20℃, a temperature difference between the inner and outer pipe wall surfaces can be found when the hot vapour flows past the 13
Page 13 of 43
system. When the partial pressure of the water vapour is higher than the saturation pressure at the temperature of the inner wall surface, the hot vapour will condense on the inner pipe wall surface and cause the carbon dioxide to dissolve in the condensation water drop or liquid film. Due to the gravity of the liquid, the condensation water will accumulate on the top and bottom segments, which results in the corrosion rates of the two segments being higher than the other segments in the test. In real oil and gas pipelines, the corrosion inhibitor in the liquid
ip t
medium will have an inhibition effect on the corrosion processes of the bottom segment. But the
inhibitor cannot reach the top segment when the flow state is of laminar flow type. Thus, the
cr
analysis only focuses on the top of the line corrosion in this study.
It can be seen from Fig. 11 that the corrosion rate of the top segment keeps increasing until
us
the gas temperature reaches 80℃. Because the temperature of the inner pipe wall surface is below 60℃ before the gas temperature rises to 80℃, the partial pressure of the water vapour is higher than the saturation pressure, which leads to continuous condensation of the vapour and makes
are presented in the following equations [21]:
Fe( s ) → Fe(2aq+ ) + 2e−
(15)
2 H 2CO3( aq ) + 2e− → 2 HCO3− ( aq ) + H 2( g )
(16)
d
M
the main corrosion product
an
more carbon dioxide dissolve in the condensation film or water drop. The generation processes of
FeCO3( s )
(17)
Ac ce pt e
Fe(2aq+ ) + CO32− ( aq )
The increase of the gas temperature also promotes the reaction kinetics of the corrosion process. The
layer is porous and less protective for the X65 steel surface when the gas temperature
is below 80℃. Thus, the corrosion rate monitored by the top segment maintains a stable status. When the gas temperature rises to 80℃, the temperature of the inner pipe wall surface reaches
about 70℃. The saturation pressure of the water vapour increases to 0.32 bar at the inner pipe wall temperature. It is higher than the partial pressure of the water vapour. The condensation rates begin to decrease, and the thickness of the condensation film starts to reduce. The corrosion product film also becomes denser and more protective after the gas temperature increases to 80℃ since the kinetics of
precipitation become faster, thus favoring the precipitation of larger
numbers of smaller crystals which pack more densely on the steel surface [22]. It can be seen from Fig. 12, due to the decrease of the condensation rate and the denser packed
corrosion
product film, the corrosion rate decreases to 0.41 mm/y after the gas temperature reaches 80℃ for 9 days.
14
Page 14 of 43
5. Conclusions From this investigation, the following conclusions can be drawn: (1) RPERS provides a better temperature and pressure compensation effect than traditional ER sensors due to its geometric form and riding location in the pipeline. Compared to the monitoring results by using WERS and a corrosion coupon at different solution temperatures, RPERS ensures higher measurement accuracy.
ip t
(2) As RPERS has the same geometrical characteristics with the pipeline to be monitored and is divided into six segments on the circumstance direction, it can provide an accurate corrosion
cr
monitoring for the whole pipeline circumference. The temperature differences between the
inner and outer pipe wall surfaces of the pipeline can also be provided through the
us
measurement results of the electrical resistances.
(3) Due to the six segments, the localized corrosion patterns as BLC and TLC can be monitored. Through the TLC experiment, the qualitative analysis of the relationship between the
an
temperature difference and the corrosion rate has been done. When the partial pressure of the water vapour is higher than the saturation pressure at the inner pipe wall temperature and the
M
gas temperature is lower than 80℃, the increase in the corrosion rate can be observed together with an increase in the temperature difference, and the corrosion rate will retain a stable status. But when the gas temperature is higher than 80℃, the corrosion rate will decrease rapidly due to
layer.
d
the formation of a dense
Ac ce pt e
(4) Based on the performance of the laboratory test, RPERS can be used as a conventional method for localized corrosion monitoring in long distance oil and gas transportation pipelines. RPERS can be used for corrosion monitoring at some special pipeline sections such as the high and low temperature transformation areas where may face TLC risk under laminar flow conditions. It can also be used for the areas with low flow rates where may confront BLC attacks due to the accumulation of the deposit on the bottom of the pipe.
Acknowledgments
This research was sponsored by Key Projects in the National Science & Technology Pillar
Program during the Twelfth Five-year Plan Period of China: The subtask of the Oil & Gas Exploitation of South China Sea Demonstration Project: Corrosion Monitoring and Design of Cathodic Protection System for LW3-1 gas field (No.2011ZX05056). References [1] L. Yang, Techniques for corrosion monitoring, first ed., Chemical Industry Press, China, 2011. [2] M. Kouril, T. Prosek, B. Scheffel, F. Dubois, High sensitivity electrical resistance sensors for indoor corrosion monitoring, Corrosion Engineering, Science and Technology, 48(2013) 282-7. 15
Page 15 of 43
[3] T. Prosek, N. Le Bozec, D. Thierry, Application of automated corrosion sensors for monitoring the rate of corrosion during accelerated corrosion tests, Materials and Corrosion, 65(2014) 448-56. [4] P.A. Cella, S.R. Taylor, Electrical resistance changes as an alternate method for monitoring the corrosion of steel in concrete and mortar, Corrosion, 56(2000) 951-9. [5] L.R. Hilbert, Monitoring corrosion rates and localised corrosion in low conductivity water, Corrosion Science, 48(2006) 3907-23. [6] R.V. Rhoades, C.M. Finley, Corrosion measurement with multiple compensation for secondary temperature compensation, United States Patent: 4587479,1986-5-6.
ip t
[7] B. Hemblade, Electrical resistance sensor and apparatus for monitoring corrosion, United States
Patent: 6946855 B1, 2005-9-20. and acetic acid on Bottom-of-the-Line corrosion, Corrosion, 67(2011).
cr
[8] M. Singer, B. Brown, A. Camacho, S. Nesic, Combined effect of carbon dioxide, hydrogen sulfide, [9] Y. Tan, Y. Fwu, K. Bhardwaj, Electrochemical evaluation of under-deposit corrosion and its
us
inhibition using the wire beam electrode method, Corrosion Science, 53(2011) 1254-61.
[10] Y.H. Sun, T. Hong, W.P. Jepson, Pipeline corrosion under wet gas conditions, Material Performance, 40(2001) 48-52. probes, Electrochim Acta, 27 (2007) 7590-7598.
an
[11] A. Legat, Monitoring of steel corrosion in concrete by electrode arrays and electrical resistance [12] M. Singer, Study and modeling of the localized nature of top of the line corrosion, Ph.D., Ohio University, Chemical and Biomolecular Engineering, 2013.
M
[13] G.A. Zhang, N. Yu, L.Y, Yang, X.P. Guo, Galvanic corrosion behaviour of deposit-covered and uncovered carbon steel, Corrosion Scicence, 86 (2014) 202-212. pipelines, Corrosion, 66(2010).
d
[14] S. Papavinasam, A. Doiron, R.W. Revie, Model to predict internal pitting corrosion of oil and gas
Ac ce pt e
[15] G.P. Gerstmann J, Laminar film condensation on the underside of horizontal and inclined surfaces, Int J Heat Mass Transfer, 5(1967) 567-80.
[16] F. Vitse, S. Nesic, Y. Gunaltun, D.L. de Torreben, P. Duchet-Suchaux, Mechanistic model for the prediction of top-of-the-line corrosion risk, Corrosion, 59(2003) 1075-84. [17] Z. Zhang, D. Hinkson, M. Singer, H. Wang, S. Nesic, A mechanistic model of top-of-the-line corrosion, Corrosion, 63(2007) 1051-62.
[18] G. Hinds, A. Turnbull, Novel multi-electrode test method for evaluating inhibition of underdeposit corrosion-part 1: sweet conditions, Corrosion, 66(2010). [19] Y.M. Gunaltun, D. Supriyatman, J. Achmad, Top-of-line corrosion in gas lines confirmed by condensation analysis, Oil & Gas Journal, 97(1999) 64. [20] Y.-C. Chang, R. Woollam, M.E. Orazem, Mathematical models for under-deposit corrosion I. Aerated Media, Journal of the Electrochemical Society, 161(2014) C321-C9. [21] Y. Xie, L. Xu, C. Gao, W. Chang, M. Lu, Corrosion behavior of novel 3%Cr pipeline steel in CO2 Top-of-Line Corrosion environment, Materials and Design (2012) 54-57. [22] S. Nesic, Key issues related to modelling of internal corrosion of oil and gas pipelines - A review, Corrosion Science, 49(2007) 4308-38.
16
Page 16 of 43
Research Highlights
Ac
ce pt
ed
M
an
us
cr
ip t
Highlights: The structure and measurement principle of ring pair electrical resistance sensor (RPERS). RPERS can not only provide the general corrosion depth of the pipeline, but also can monitor the localized corrosion by six quadrants division. RPERS provides a better compensation than traditional wire ER sensor through the corrosion monitoring test in the laboratory. The temperature difference measurement method between the inner and outer pipe wall surface. The top of the line corrosion test has been monitored by the top quadrant of RPERS and the relationship of temperature difference, gas temperature and corrosion rates have been obtained.
Page 17 of 43
Table(s)
Table 1 Chemical component of the X65 pipeline steel: C
Si
Mn
P
S
V
Nb
Cr
Fe
Carbon steel
0.16
0.45
1.6
0.02
0.01
0.06
0.05
0.1
Bal.
Ac
ce pt
ed
M
an
us
cr
ip t
X65
Page 18 of 43
M
an
us
cr
ip t
Figure(s) illustrain
Ac
ce pt
ed
Fig. 1. Schematic illustration of the internal pipeline corrosion monitoring system.
Page 19 of 43
ip t cr us an M ed
ce pt
Fig. 2. The structure of RPERS (a) the three-dimension views of the mechanical structure of RPERS (b) the
Ac
dimensioning figure of the corrosion and compensation elements (c) the photogragh of RPERS.
Page 20 of 43
ip t cr
us
Fig. 3. The measurement principle diagram of the ring pair ER sensor with three excitation currents injected from
Ac
ce pt
ed
M
an
different directions.
Page 21 of 43
ip t
cr
Fig. 4. The geometry model and temperature distribution of a segment (a) The temperature distribution of a segment on the compensation element (b) the simplified rectangle model of a segment (c) the temperature gradient
Ac
ce pt
ed
M
an
us
distribution and the thin rectangular plate of a segment.
Page 22 of 43
ip t cr us an
Fig. 5. Internal-pipeline corrosion on-line monitoring system (a) Pipeline internal corrosion monitoring circulatory system: 1, water tank; 2, circular tube pump; 3, thermal insulation pipe; 4, wire ER sensor; 5, main chamber; 6,
Ac
ce pt
ed
M
electronics chamber (b) the simulative subsea environment.
Page 23 of 43
ip t cr us
Fig. 6. The measurement result of the temperature around the corrosion monitoring system in the simulative
Ac
ce pt
ed
M
an
seawater surrounding.
Page 24 of 43
ip t cr us
Fig. 7. The measurement results of the temperature difference by using the both two methods with 3.5% NaCl
Ac
ce pt
ed
M
an
solution flowing past.
Page 25 of 43
ip t cr us
Fig. 8. The measurement results of the temperature difference by using the both two methods with gas flowing
Ac
ce pt
ed
M
an
past.
Page 26 of 43
ip t cr us
an
Fig. 9. The results of the metal loss monitored by using WERS in 3.5% NaCl solution under different solution
Ac
ce pt
ed
M
temperatures.
Page 27 of 43
ip t cr
us
Fig. 10. The results of the metal loss monitored by using RPERS in 3.5% NaCl solution under different solution
Ac
ce pt
ed
M
an
temperatures.
Page 28 of 43
ip t cr us an
Ac
ce pt
ed
M
Fig. 11. The result of the metal loss measured by the top segment of RPERS with gas flowing past.
Page 29 of 43
ip t cr us
values.
Ac
ce pt
ed
M
an
Fig. 12. The result of the calculated corrosion rate under different
Page 30 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 1. Click here to download high resolution image
Page 31 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 2. Click here to download high resolution image
Page 32 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 3. Click here to download high resolution image
Page 33 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 4. Click here to download high resolution image
Page 34 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 5. Click here to download high resolution image
Page 35 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 6. Click here to download high resolution image
Page 36 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 7. Click here to download high resolution image
Page 37 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 8. Click here to download high resolution image
Page 38 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 9. Click here to download high resolution image
Page 39 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 10. Click here to download high resolution image
Page 40 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 11. Click here to download high resolution image
Page 41 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 12. Click here to download high resolution image
Page 42 of 43
Author Biographies
Biographies Yunze Xu was born in 1989 in Henan, China. He received his bachelor degree in marine engineering in 2012 from the Huazhong University of Science and Technology. After his diploma degree, he started to work as PhD student at School of Naval Architecture in Dalian University of Technology. He is working on the pipeline corrosion monitoring and corrosion mechanism
ip t
research. Yi Huang was born in 1964 in Dandong, China. He obtained his PhD from the Hiroshima University of Japan in September 1992. From 1992 to 2001, he worked at the Hiroshima
cr
University on studies of corrosion and anticorrosion problems for ship and ocean engineering. Then he was invited to Dalian University of Technology and became a full professor in School of
us
Naval Architecture. There he managed a group working on the corrosion and anticorrosion problems for ship and ocean engineering, monitoring and inspection technique for ship hull and
Ac
ce pt
ed
M
an
marine structures.
1
Page 43 of 43
S0925-4005(15)30496-2 http://dx.doi.org/doi:10.1016/j.snb.2015.10.030 SNB 19174
To appear in:
Sensors and Actuators B
Received date: Revised date: Accepted date:
31-12-2014 8-6-2015 12-10-2015
Please cite this article as: Y. Xu, Y. Huang, X. Wang, X. Lin, Experimental study on pipeline internal corrosion based on a new kind of electrical resistance sensor, Sensors and Actuators B: Chemical (2015), http://dx.doi.org/10.1016/j.snb.2015.10.030 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Experimental study on pipeline internal corrosion based on a new kind of electrical resistance sensor Yunze Xua,*, Yi Huanga, Xiaona Wangb,**, Xuanqin Lina a School of Naval Architecture and Ocean Engineering, Dalian University of Technology, Dalian 116024;
* Corresponding author. Tel: +86-15898193759; fax: +86-0411-84706061. **Corresponding author. Tel: +86-0411-84706061; fax: +86-0411-84706061. E-mail address: [email protected] (Yunze Xu); [email protected] (Xiaona Wang);
ip t
b School of Physics and Optoelectronic Engineering, Dalian University of Technology, Dalian 116024
cr
Postal address: School of Naval Architecture and Ocean Engineering, Dalian University of Technology, Linggong Road 2#, Dalian 116024, Liaoning Province, China.
us
Abstract:
In this paper, a ring pair electrical resistance sensor (RPERS) has been developed for an internal-pipeline corrosion on-line monitoring system. The RPERS was divided into six segments along the circumference. The
an
corrosion depth of each segment could be measured by using three alternating excitation currents injected into the sensor from different angles. In order to simulate and monitor the pipeline internal corrosion, a corrosion monitoring system which contained RPERS, wire electrical resistance sensor (WERS), thermocouples and a
M
corrosion coupon was established. First, the corrosion processes in 3.5% sodium chloride solution with the temperature varied from 30℃ to 60℃ were studied. The temperature differences between the inner and outer pipe wall surfaces were measured by the top segment of RPERS. The test results revealed that the performance of RPERS
d
is better than that of WERS in term of metal loss measurement. Then, carbon dioxide mixed with water vapour was pumped into the system with the temperature varied from 50℃ to 80℃. In the scenarios, top of the line corrosion
Ac ce pt e
(TLC) occurred and was monitored by RPERS. The monitoring results demonstrated that the gas temperature and the temperature difference were important factors for TLC in sweet conditions. Key words: ring pair sensor, internal corrosion, on-line monitoring, temperature difference, TLC
1. Introduction
The electrical resistance (ER) sensor technique, also known as electrical coupon technique,
has become one of the most important methods for pipeline internal corrosion monitoring [1]. As a result of corrosion process, the cross-section area of the sensor is reduced, leading to an increase in electrical resistance. Based on the change of the resistance, the corrosion loss in metal thickness can be obtained and the slope of the metal loss curve directly corresponds to the corrosion rate. In contrast to conventional electrochemical techniques such as the linear polarization resistance and the electrochemical impendence spectroscopy methods, the advantage of the ER technique on metal loss measurement is that the presence of an electrolyte layer on the metal surface is not inevitable. In that sense, the method can be used in a high resistance medium such as gas and crude oil. Moreover, it can be used for corrosion monitoring of high temperature corrosion and erosion corrosion [2, 3].
1
Page 1 of 43
The traditional ER sensors are almost entirely composed by a corrosion element and a compensation element. The geometry shapes of the elements are usually present in the forms of wire, tube, flush spiral and strip. The corrosion element of the ER sensor is exposed to corrosion surroundings and the compensation element is covered with a protective coating to segregate it from the corrosive environments [4, 5]. Through the resistance ratio of the two elements, the temperature interference can be eliminated under ideal conditions. Thus, the metal loss of the
ip t
corrosion element can be calculated according to its geometric dimensioning. However, in the pipelines with high temperature and pressure, the two elements of the traditional ER sensors may be
cr
subjected to different temperature and pressure conditions due to the different locations where they are emplaced. It has been found that a minor difference of 0.25℃ between the corrosion element
us
and the compensation element will cause a significant change in the resistance ratio of 1000 ppm. Although the effect of the pressure difference on the resistance ratio is much less than that caused by the temperature difference, the resistance ratios of typical pipeline steels still undergo obvious
an
changes from 2000 ppm to 4000 ppm per 100 bar [6, 7].
In recent years, localized corrosion problems such as bottom of the line corrosion (BLC) and
M
top of the line corrosion (TLC) have become focus issues of the pipeline internal corrosion [8-10]. In general corrosion surroundings, the traditional ER sensors can be sensitive to homogeneous corrosion due to their geometric forms, but the responses of these sensors to localized corrosion are
d
limited [11]. With improvements on the resolution of the electrical resistance measurement, some
Ac ce pt e
researchers have tried to use the ER sensors for localized corrosion monitoring in pipelines by changing their forms. Marc Singer [12] has used a new kind of flush head ER sensor for TLC monitoring. However, the monitoring results are dissatisfactory due to the edge effect caused by the pipeline arc. BLC is always caused by sand and corrosion products covering the bottom of the line. The deposit on the steel surface may lead to failures of the corrosion inhibitors and cause serious localized corrosion. The practical corrosion rate of the bottom area in the pipeline is hard to be monitored due to the galvanic effect between the bare steel and the deposit-covered steel [13]. For most oil and gas pipelines, the inner fluids usually endure high temperatures above 50℃,
while the temperature of the outer surroundings is always much lower than the fluid temperature, especially in subsea surroundings [14]. In these scenarios, the temperature differences between the inner and outer pipe wall surfaces exist. Gerstmann [15] found that the temperature differences in sweet conditions may lead to the formation of condensation films on the inner surfaces of the pipelines, which is an important factor for TLC. The outer temperature of the pipe wall surface can be easily measured by arranging the temperature measurement sensor outside of the pipe wall surface. However, it is hard to measure the temperature of the inner pipe wall surface in practice without destroying the pipe wall structure. Due to the unknown temperature difference between the
2
Page 2 of 43
inner and outer pipe wall surfaces, Vitse and Zhang et al. [16, 17] established complex thermodynamic models to forecast the condensation rates, but the predicted results still have deviations compared with the experimental results. The primary purpose of this paper is to present a new kind of ring pair ER sensor (RPERS) for pipeline internal corrosion monitoring. RPERS is composed of two rings, i.e. a corrosion element and a compensation element. The rings are made with an 8 inch X65 pipeline which is usual for
ip t
subsea installations. The RPERS is divided into six segments and used to monitor the metal losses of these segments in a laboratory system. In the experiment, the temperature difference between the
cr
inner and outer pipe wall surfaces is measured by the top segment of RPERS. The metal loss
measurement results of RPERS are compared with those obtained from a traditional wire ER sensor
us
(WERS) and a corrosion coupon. In the experimentation of TLC, the metal loss, corrosion rate and temperature difference of the top segment in the pipeline are monitored by RPERS. The relationship of the gas temperature, temperature difference between the inner and outer pipe wall surfaces and
an
corrosion rate is obtained. RPERS also provides a new measurement method for thermodynamic
2. Method
M
model research within the experiment.
d
2.1. Internal-Pipeline corrosion monitoring system
Ac ce pt e
As shown in Fig. 1, the internal-pipeline corrosion monitoring system is composed of three parts: pipeline information detecting section, data acquisition system and data-managing section. The pipeline information detecting section contains RPERS, traditional WERS, K-type
thermocouples and a corrosion coupon. It is the sensitive part of the monitoring system. The sensors provide continuous signals of the internal corrosion conditions and the temperature information in the pipeline.
In the data acquisition system, the ER signals generated from RPERS can be converted to
digital signals by an RM3545 micro resistance measurement device whose ER measurement resolution is 10 nΩ. The models of the K-type thermocouples are TP-K01, and the measurement surfaces of the thermocouples are coated with a layer of teflon for insulation. The temperature measurement accuracy of the thermocouples can reach ±0.1℃ in the range of 0~80℃ after having been calibrated in a Julabo FP 51 thermostatic water bath. The temperature control accuracy of the thermostatic water bath is ±0.05℃ in the range of -20~100℃. The voltage signals generated from the thermocouples are recorded by NI PCle-6320 data acquisition card. The ER signals from WERS can be obtained by CMB 1510b. The CMB 1510b is a corrosion data acquisition and storage device which is manufactured by Instituted of Corrosion & Protection of Metals in China. All the signals 3
Page 3 of 43
are transferred to the computer after disposal by a RS232 transition port. The data-managing section is an operation platform. It is compiled by C# to record and process messages conveyed from the Data Acquisition System (DAS). All data is saved in the Structured Query Language (SQL) Server database.
ip t
2.2. The construction and measurement principle of RPERS
As shown in Fig. 2a, RPERS is a part of the pipeline which needs to be monitored. The corrosion element and compensation element have the same inner diameters with the pipeline. The
cr
two elements are embedded into the pipeline through a connecting element, which is shown in Fig.
2b. The axial widths of both elements are 10 mm and the initial sensitive wall thickness of the
us
corrosion element is 13.5 mm. The rings are coated with a layer of chromic oxide by plasma spraying except the internal surface of the corrosion element, which is shown in Fig. 2c. The two
an
elements are insulation from other pipe surfaces through the Cr2O3 coating on the contact surfaces and the insulation resistance is higher than 1 MΩ among each element at a voltage of 50 V. A better temperature and pressure compensation effect can be provided as the compensation element and
M
the corrosion element are coaxial concentric placed in the pipeline. The monitoring results of the localized corrosion, i.e. BLC and TLC, will be closer to the actual situations than traditional flush
d
head ER sensors since the arc of the corrosion element is the same as the pipeline to be monitored. RPERS can provide two monitoring functions, i.e. the monitoring of the localized corrosion in
Ac ce pt e
pipelines and the measurement of the temperature difference between the inner and outer pipe wall surfaces. The two functions are introduced as follows. For localized corrosion monitoring, each element are divided into six segments using
electrodes 1, 3, 5, 7, 9 and 11, as shown in Fig. 3. Segment 1 is between electrodes 1 and 3, segment 2 is between electrodes 3 and 5 and the other segments are dealt with the same approach. These electrodes are also used for voltage measurements. The resistance values of the segments on the corrosion element are represented by
(
), respectively. The resistance values
of the segments on the compensation element are represented by
(
),
respectively. The six segments may suffer different temperatures or corrosion environments due to the different locations and flow regimes in the pipeline. Galvanic current may generate in RPERS and cause a measuring error [18]. To eliminate the additional potential caused by the galvanic current in the ring, the square excitation current waves are injected into the ring pair for ER measurement. The amplitude of the current is 1 A and the measurement period for each signal is 241 ms. In a
4
Page 4 of 43
measurement period, the voltage signals negative current
and
caused by the positive current
are measured. Then the average voltage
and
used as the final measurement
result can be calculated by:
V=
V+ + | V− | 2
(1)
and
are successively injected into the ring pair through the
even-numbered electrodes, as shown in Fig. 3. Firstly, when
is injected as the red path, the
cr
Three excitation currents
ip t
In this way, the measurement error of the voltage caused by the galvanic current can be eliminated.
us
potential differences between electrodes 1 and 3, 3 and 5, 9 and 7, 11 and 9 are respectively measured as the voltages of segments 1, 2, 4 and 5 on the corrosion and compensation elements.
(
(
) and
an
Through the measurement results, the resistance ratios ) can be expressed as:
(2)
cor ( comp ) cor ( comp ) ε cor 2( comp 2) = R4cor (comp ) / R5cor (comp ) =V9,7 / V11,9( I1 ) ( I1 )
(3)
d
M
(comp ) cor ( comp ) ε cor1( comp1) = R1cor (comp ) / R2cor (comp ) =V1,cor / V3,5( I1 ) 3( I1 )
and
represent the resistance ratios between segments 1 and 2 of the two elements,
Ac ce pt e
where
represent the resistance ratios between segments 4 and 5 of the two elements.
,
,
and
respectively represent the voltage
measurement results of segments 1, 2, 4 and 5 of the two elements when Then, when
is passing through.
is injected into the ring pair as the blue path shown in Fig. 3, the voltage values
of segments 1, 3, 4 and 6 can be measured successively using the same steps as through. The resistance ratios
(
) and
(
is passing ) can
be expressed as: cor ( comp ) cor ( comp ) ε cor 3( comp 3) = R6cor (comp ) / R1cor (comp ) =V11,1 / V1,3( I2 ) (I2 )
(4)
cor ( comp ) cor ( comp ) ε cor 4( comp 4) = R4cor (comp ) / R3cor (comp ) =V9,7 / V7,5( I2 ) (I2 )
(5)
5
Page 5 of 43
where
represent the resistance ratios between segments 6 and 1 of the two elements, represent the resistance ratios between the segments 4 and 3 of the two elements.
,
and
respectively represent the voltages of
segments 1, 3, 4 and 6 of the two elements when Finally, when
is passing through.
is injected as the purple path shown in Fig. 3, the voltage values of segments
2 and 3 can also be measured and the resistance ratios
(
cor ( comp )
ε cor 5( comp 5) = R3
=V7,5( I(3 )
cor comp )
cor ( comp ) / V5,3( I3 )
)can be
(6)
an
represent the resistance ratios between segments 3 and 2 of the two elements,
and
is passing through.
d
two elements when
respectively represent the voltages of segments 2 and 3 of the
M
and
cor ( comp )
/ R2
us
expressed as:
where
ip t
,
cr
and
Ac ce pt e
Because the two rings are in series, the current passing through the corrosion element is equal to that passing through the compensation element. Thus when
is passing through, Eq. (7) can be
written as:
cor V1,3( I1 )
R1cor
+
cor V11,9( I1 )
R5cor
=
comp V1,3( I1 )
R1comp
+
comp V11,9( I1 )
R5comp
= I1
(7)
Combining Eqs. (2) to (7), Eq. (8) can be obtained:
6
Page 6 of 43
comp i
i=2 i=3 (8)
i=4 i=5
cr
/R
i =1
ip t
ki = R
cor i
cor comp V1,3( I1 ) + V11,9( I1 )ε cor 2ε cor 1 / ε cor 4ε cor 5 comp comp V1,3( I1 ) + V11,9( I1 )ε comp 2ε comp1 / ε comp 4ε comp 5 ε comp1 k1 ε cor1 ε cor 5ε comp1 k ε comp 5ε cor1 1 = ε cor 4ε cor 5ε comp1 k 1 ε comp 4ε comp 5ε cor1 ε cor 4ε cor 5 ε comp 2ε comp1 k1 ε cor 2ε cor1 ε comp 4ε comp 5 ε cor 3 k1 ε comp 3
us
,
is the resistance ratio between the corresponding segments of the
an
where
i=6
corrosion and compensation elements, and
is the resistance ratio between
and
. As
M
is obtained, the metal loss of each segment can be calculated according to the geometric dimensioning of the ring.
d
The temperature difference between the inner and outer pipe wall surfaces of each segment can also be measured by using RPERS. As mentioned above, the compensation element is coated with a
Ac ce pt e
layer of chromic oxide coating for anticorrosion. The resistance values of the segments on the compensation element only change with the fluctuation of the temperature. When a temperature difference exists, there will be a nearly liner temperature gradient along the diameter direction [19], which is shown in Fig. 4a.
and
surfaces, respectively, and
is the difference between
Because the wall thickness
represent the temperatures of the outer and inner pipe wall
and the width
and
.
along the axial direction are much smaller
than the arc length a in a ring element, which is shown in Fig. 4a, the geometric model of a segment can be simplified as a rectangle model, as shown in Fig. 4b. The temperature gradient is shown in Fig. 4c. At the height of z along the vertical direction, pick up the thin rectangular plate with a thickness of
, the resistance of the thin plate can be described in Eq. (9):
dR = ρ 0 where
[1 + α(Tz − T20 )]a b × dz
is the electrical resistance of the thin rectangular plate,
(9) is the electrical resistivity of
7
Page 7 of 43
the steel material at 20℃,
is the temperature coefficient of the resistance,
of the thin rectangular plate, and
equals to 20℃.
Tz = Ti −
can be calculated by Eq. (10):
Ti − To z Δr
(10)
in Eq. (7) is a known constant, the electrical resistance of each
ip t
As the current value of
is the temperature
segment can be calculated from Eqs. (2) to (7). The electrical resistance of each segment on the
) can be measured in the temperature control device before
cr
compensation element at 20℃ (
being installed into the pipe. The outer pipe wall temperature of segment (
) can be measured by
us
the temperature sensor arranged on its outer pipe wall surface. Because the thin rectangular plates are connected in parallel within a segment on the compensation element, the resistance of segment ) can be calculated by solving Eq. (9). The calculated result is shown in Eq. (11):
an
(
,
is the inner pipe wall temperature of segment
and the value of
which cannot be
can be calculated by using the numerical calculation method,
Ac ce pt e
directly monitored. From Eq. (11),
(11)
d
where
1 + α (Tii − T20 ) 1 + α (Toi − T20 )
M
Ricomp = Ricomp 0 α (Tii − Toi ) / ln
can also be obtained.
As the geometric shape of the corrosion element is also a ring form like the compensation
element, the resistance of the corrosion element
can also be calculated when a temperature
difference exists. When the metal loss of segment
on the corrosion element is
, Eq. (12) can be
obtained by using the same derivation processes as the compensation element.
Ricor = Ricor 0
α∆r
∆r − xi
(Tii − Toi ) / ln
1 + α (Tii − T20 ) 1 + α (Toi − T20 )
(12)
Combining Eq. (8), Eq. (11) and Eq. (12), the metal loss can be determined:
xi = ∆r × (1 − The resistance ratio of
and
Ricor 0 ) comp ki Ri 0
(13)
in Eq. (13) can be measured before corrosion occurs and
used as a constant. From the calculative process, it can be found that whether a temperature
8
Page 8 of 43
difference exists or not, it makes no difference to the calculated metal loss in theory.
2.3 The corrosion monitoring circulation system in the laboratory
As shown in Fig. 5a, the pipeline circulation system in the laboratory is composed of six parts. The temperature control heater and the PT100 temperature sensor are installed in the water tank to
ip t
control the fluid temperature. The corrosion liquid or gas could be pumped into the monitoring system by using different kinds of pumps. The traditional WERS is directly inserted into the 8 inch
pipe with its data collection device installed out of the pipe. The main chamber which contains
cr
RPERS and two thermocouples is butt welded by two pieces of reducer pipes, and it is filled with
heat transfer oil. The two thermocouples are respectively emplaced on the inner and outer pipe wall
us
surfaces of the top segment in the pipeline. The corrosion coupon is hung nearly at the end cap of the delivery port that could be removed for weighing after the experiment. The data acquisition
an
system and power supply devices are placed in the electronics chamber.
All the corrosion sensors and the corrosion coupon are made of X65 pipeline steel of which chemical composition (wt%) is shown in Table 1.
M
As shown in Fig. 5b, the whole system was working in a simulative subsea environment during the test period. The water temperature was monitored for 12 days and was fluctuated within 1℃ in
d
the laboratory, which is shown in Fig. 6. Firstly, 3.5% sodium chloride solution by weight with a PH of nearly 6.5 was pumped into the system, and the solution was launched to circulate with oxygen
Ac ce pt e
continually providing. The flow rate of the solution was 0.3 Lpm. When the flow regime and the temperature of the solution became stable, the solution temperature was successively increased from 30℃ to 60℃. Finally, the temperature differences (
) and metal losses were recorded. After
the first experiment, the corrosion coupon was taken out and used to weigh the weight loss. After the solution was released from the system, oxygen gas in the system was eliminated by pure nitrogen. Then, carbon dioxide mixed with water vapour was pumped into the system for circulation, and the flow rate of the gas was 1.5 Lpm. The total pressure was 0.5 bar and the partial pressure of the carbon dioxide was 0.3 bar. Varying the gas temperature from 50℃ to 80℃, the values of
and
the metal loss were monitored by the top segment of RPERS. 3. Experimental Results
3.1 The temperature difference measurement results using RPERS and thermocouples
As mentioned above, the two thermocouples were respectively mounted on the inner and outer 9
Page 9 of 43
surfaces of the pipeline. One thermocouple was placed on the outer pipe wall surface of the top segment on the compensation element, and the outer surface’s temperature ( ) of the top segment could be measured. Another thermocouple was inserted into the pipe near the compensation element and it was flush with the inner pipe wall surface of the top segment. The inner surface’s
temperature difference (
ip t
temperature ( ) of the top segment could also be measured. Based on the measurement results, the ) of the top segment could be obtained.
the value of
could also be calculated from Eq. (11) in conjunction with the measurement value
which was obtained by the thermocouple emplaced on the outer pipe wall surface. Fig. 7
us
of
cr
Through the resistance measurement results of the top segment on the compensation element,
shows the temperature difference measurement results of the top segment monitored by the both
when gas was flowing past the system.
M
results of
an
two methods when 3.5% NaCl solution was flowing past the system. Fig. 8 shows the measurement
3.2 The metal loss measurement results of RPERS and traditional WERS
d
Fig. 9 presents the measurement result of the metal loss monitored by WERS which was
Ac ce pt e
inserted in the pipe. Fig. 10 shows the six metal loss curves monitored by the six segments of RPERS in the 3.5% NaCl solution. In the corrosion monitoring system, RPERS was placed in the middle of the pipeline. Segments 3 of the both elements were placed at the bottom of the line and segments 6 were placed at the top of the line. Segments 1 and 5 were symmetrically placed on the upper half of the pipeline and segments 2 and 4 were symmetrically placed on the lower half of the pipeline.
It can be found that the slopes of all the metal loss curves increase with the raising of the
fluid temperature from 30℃ to 60℃. It can be seen from Fig. 9 that there were two saltus steps on the metal loss curve. The accumulative measurement error of WERS reached nearly 5 µm when the fluid temperature increased to 60℃. As shown in Fig. 10, a better compensation effect could be seen from the measurement results of RPERS. From the measurement results, it could be found that the metal losses of the six segments were not uniform. Because the vent valve was located slightly beneath the top of the pipe, some air was still remaining in the pipe after the solution was introduced into the system. The top segment did not have a thorough contact with the solution due to the remaining air which led to the metal loss of the top segment less than 2 µm in the experiment. The corrosion depth of segment 3 was 13.4 µm which was higher than that of the other segments. 10
Page 10 of 43
The corrosion depths of the symmetric segments 2 and 4 were 11.9 µm and 11.4 µm, respectively. The corrosion depths of the symmetric segments 1 and 5 were 7.8 µm and 9.2 µm, respectively. After the experiment, the corrosion coupon was taken out and the corrosion products were cleared by ultrasonic cleaning. The metal loss which was converted from the coupon weight loss was 10.2 µm. The average metal loss of segments 1 to 5 which were completely beneath the solution level was 10.7 µm. The general corrosion depths monitored by the two methods were of a good
ip t
consistency.
cr
3.3 The monitoring results of top of the line corrosion
As shown in Fig. 11, when carbon dioxide mixed with the water vapour was pumped into the
us
system, top of the line corrosion commenced and the metal loss was monitored by the top segment of RPERS. The corrosion rate was calculated from the slope of the metal loss curve, and is shown in
with the raising of the temperature difference
an
Fig. 12. It can be found that when the gas temperature was below 80℃, the corrosion rate increased which is shown in Fig. 8. However, after the gas
M
temperature reached 80℃, the corrosion rate immediately achieved to the highest value nearly 1.20 mm/y and then significantly decreased to 0.41 mm/y after 9 days. The highest corrosion rates of segments 1, 2, 4 and 5 were 0.63 mm/y，0.94 mm/y, 1.05 mm/y
d
and 0.61 mm/y, respectively. The final corrosion depths of these segments were 23.8 µm, 29.6 µm,
Ac ce pt e
31.2 µm and 22.6 µm, respectively, which were lower than the 39.4 µm of segment 6. The corrosion rate of the bottom segment 3 was higher than that of the top segment 6. The highest corrosion rate of segment 3 was 1.56 mm/y since the condensation water of the other segments accumulated at the bottom of the pipe. The finial corrosion depth of segment 3 was 46.7 µm. The corrosion rates of segments 1 to 5 also decreased to the values lower than 0.5 mm/y after the temperature reached 80℃ for 9 days. 4. Discussions
4.1 The analysis of the temperature difference measurement results
Because of some thermal insulation parts in the system, the real temperature difference between the inner and outer pipe wall surfaces is smaller than the difference between the seawater temperature and the fluid temperature. It can be seen from the measurement results that the values of
measured by the both methods are basically consistent. Though the temperature difference
measured by the thermocouples are a little higher than RPERS when the fluid temperature is above 11
Page 11 of 43
40℃,the maximum deviation of the
value is less than 0.4℃.
The measurement differences may be caused by two factors: Firstly, the thermocouples are arranged on the top of the pipe and the probe can only focus on a small area compared with the top segment of RPERS. Secondly, the temperature distribution may not be completely uniform in the top segment as the assumption above. However, compared with the results measured by the
ip t
thermocouples whose measurement accuracy is ±0.1℃, the temperature difference measured by RPERS are credible when combined using the measurement result of the thermocouple emplaced
cr
on the outer pipe wall surface.
us
4.2 The analysis of the metal loss measurement results in 3.5% NaCl solution
It can be seen from Fig. 9 and Fig. 10 that the metal loss curves become steep with the fluid increasing from 30℃ to 60℃. It indicates that the corrosion rate will increase with
an
temperature
the rising of the solution temperature. From the measurement result of WERS, a measurement error of the metal loss can be found. As shown in Fig. 8, when the solution temperature changes from 40
M
℃ to 50℃ and from 50℃ to 60℃, the temperature differences increase to 10℃ and 15.8℃, respectively. Though the carbon steel probe has an excellent thermal conductivity and the
d
compensation element of the WERS is parceled by the shrinkable tube, the temperature of the
Ac ce pt e
corrosion element is still higher than the compensation element because the corrosion element is directly inserted into the high temperature solution. The measurement error can be expressed as:
δx = r0
where
kt k0
α δT − 1 1 + 1 + α Tr
is the measurement error of the metal loss,
(14)
is the initial diameter of the wire ER sensor,
is the instantaneous monitoring result of the resistance ratio between the compensation and corrosion elements,
is the initial resistance ratio of the compensation and corrosion elements
which should be measured in the temperature control device before the corrosion occurs, temperature difference between the corrosion and compensation elements and
is the is the
temperature of the compensation element. The temperature difference leads to nearly 5 µm accumulative measurement error when the
increases to 15.8℃. The metal loss measurement
result is hard to be corrected because the real temperature of the two elements in Eq. (14) is hard to be monitored in practical application. 12
Page 12 of 43
It can be seen from Fig. 10 that the temperature compensation effect of RPERS is better than WERS. Since the corrosion rate of each segment can be monitored, the localized corrosion conditions of the pipeline can be reflected. As the flow rate in the system is 0.3 Lpm, the corrosion products may have accumulated on segments 3~5 under the low flow rate. The different coverage conditions of the corrosion products will cause the differential aeration on the internal surface among each segment and lead to galvanic corrosion [20]. The metal loss of segment 3 is higher than
ip t
those of the other segments. It may be caused by the coverage of corrosion products which lead to the forming of local anode area at the bottom of the pipe. The partial regions of segments 2 and 4
cr
bordering on the bottom segment may also be covered by the corrosion products, and it will result in
the corrosion rates of the two segments being higher than segments 1 and 5. It can be seen from Fig.
us
10 that the corrosion depths of the symmetry segments are nearly the same, which indicates that the distribution of the corrosive medium is approximately the same at the identical liquid level and there are no obvious pits corrosion patterns occurring in the symmetry segments.
an
Fig. 10 also shows that the fluctuation noise of the metal loss measurement results generated from RPERS is nearly ±0.3 µm which is higher than the ±0.14 µm of WERS. It is mainly caused
M
by two factors, i.e. the difference of the resistance values and the difference of the sensitive thicknesses between the two kinds of ER sensors. The resistance values of the both elements on WERS are near 2.2 mΩ at 25℃. However, the resistance values of the segments on RPERS are
d
only close to 0.27 mΩ at 25℃. The voltage signals generated from the RPERS are weaker than
Ac ce pt e
WERS at the same excitation current values and liable to be influenced by the surrounding electromagnetic noises. As the sensitive diameter of WERS is 1.2 mm and the sensitive wall thickness of RPERS is 13.5 mm, the resistance change in proportion to the initial total resistance of WERS is much higher than that of RPERS when the corrosion depths of the two sensors are at the same level. It causes the metal loss resolution of WERS to be better than that of RPERS. Based on these two factors, the fluctuation noise of RPERS is higher than that of WERS. Through the metal loss comparison result between RPERS and the corrosion coupon, it can
be found that the general corrosion depths monitored by the two methods have a good consistency. From the measurement results, it can be seen that RPERS can not only provide the general corrosion conditions of the pipeline, but can also reflect the localized corrosion status through the six segments.
4.3 The analysis of the TLC measurement results
The partial pressure of the water vapour is 0.2 bar during the test. As the corrosion system is arranged at the sea floor conditions with the temperature close to 20℃, a temperature difference between the inner and outer pipe wall surfaces can be found when the hot vapour flows past the 13
Page 13 of 43
system. When the partial pressure of the water vapour is higher than the saturation pressure at the temperature of the inner wall surface, the hot vapour will condense on the inner pipe wall surface and cause the carbon dioxide to dissolve in the condensation water drop or liquid film. Due to the gravity of the liquid, the condensation water will accumulate on the top and bottom segments, which results in the corrosion rates of the two segments being higher than the other segments in the test. In real oil and gas pipelines, the corrosion inhibitor in the liquid
ip t
medium will have an inhibition effect on the corrosion processes of the bottom segment. But the
inhibitor cannot reach the top segment when the flow state is of laminar flow type. Thus, the
cr
analysis only focuses on the top of the line corrosion in this study.
It can be seen from Fig. 11 that the corrosion rate of the top segment keeps increasing until
us
the gas temperature reaches 80℃. Because the temperature of the inner pipe wall surface is below 60℃ before the gas temperature rises to 80℃, the partial pressure of the water vapour is higher than the saturation pressure, which leads to continuous condensation of the vapour and makes
are presented in the following equations [21]:
Fe( s ) → Fe(2aq+ ) + 2e−
(15)
2 H 2CO3( aq ) + 2e− → 2 HCO3− ( aq ) + H 2( g )
(16)
d
M
the main corrosion product
an
more carbon dioxide dissolve in the condensation film or water drop. The generation processes of
FeCO3( s )
(17)
Ac ce pt e
Fe(2aq+ ) + CO32− ( aq )
The increase of the gas temperature also promotes the reaction kinetics of the corrosion process. The
layer is porous and less protective for the X65 steel surface when the gas temperature
is below 80℃. Thus, the corrosion rate monitored by the top segment maintains a stable status. When the gas temperature rises to 80℃, the temperature of the inner pipe wall surface reaches
about 70℃. The saturation pressure of the water vapour increases to 0.32 bar at the inner pipe wall temperature. It is higher than the partial pressure of the water vapour. The condensation rates begin to decrease, and the thickness of the condensation film starts to reduce. The corrosion product film also becomes denser and more protective after the gas temperature increases to 80℃ since the kinetics of
precipitation become faster, thus favoring the precipitation of larger
numbers of smaller crystals which pack more densely on the steel surface [22]. It can be seen from Fig. 12, due to the decrease of the condensation rate and the denser packed
corrosion
product film, the corrosion rate decreases to 0.41 mm/y after the gas temperature reaches 80℃ for 9 days.
14
Page 14 of 43
5. Conclusions From this investigation, the following conclusions can be drawn: (1) RPERS provides a better temperature and pressure compensation effect than traditional ER sensors due to its geometric form and riding location in the pipeline. Compared to the monitoring results by using WERS and a corrosion coupon at different solution temperatures, RPERS ensures higher measurement accuracy.
ip t
(2) As RPERS has the same geometrical characteristics with the pipeline to be monitored and is divided into six segments on the circumstance direction, it can provide an accurate corrosion
cr
monitoring for the whole pipeline circumference. The temperature differences between the
inner and outer pipe wall surfaces of the pipeline can also be provided through the
us
measurement results of the electrical resistances.
(3) Due to the six segments, the localized corrosion patterns as BLC and TLC can be monitored. Through the TLC experiment, the qualitative analysis of the relationship between the
an
temperature difference and the corrosion rate has been done. When the partial pressure of the water vapour is higher than the saturation pressure at the inner pipe wall temperature and the
M
gas temperature is lower than 80℃, the increase in the corrosion rate can be observed together with an increase in the temperature difference, and the corrosion rate will retain a stable status. But when the gas temperature is higher than 80℃, the corrosion rate will decrease rapidly due to
layer.
d
the formation of a dense
Ac ce pt e
(4) Based on the performance of the laboratory test, RPERS can be used as a conventional method for localized corrosion monitoring in long distance oil and gas transportation pipelines. RPERS can be used for corrosion monitoring at some special pipeline sections such as the high and low temperature transformation areas where may face TLC risk under laminar flow conditions. It can also be used for the areas with low flow rates where may confront BLC attacks due to the accumulation of the deposit on the bottom of the pipe.
Acknowledgments
This research was sponsored by Key Projects in the National Science & Technology Pillar
Program during the Twelfth Five-year Plan Period of China: The subtask of the Oil & Gas Exploitation of South China Sea Demonstration Project: Corrosion Monitoring and Design of Cathodic Protection System for LW3-1 gas field (No.2011ZX05056). References [1] L. Yang, Techniques for corrosion monitoring, first ed., Chemical Industry Press, China, 2011. [2] M. Kouril, T. Prosek, B. Scheffel, F. Dubois, High sensitivity electrical resistance sensors for indoor corrosion monitoring, Corrosion Engineering, Science and Technology, 48(2013) 282-7. 15
Page 15 of 43
[3] T. Prosek, N. Le Bozec, D. Thierry, Application of automated corrosion sensors for monitoring the rate of corrosion during accelerated corrosion tests, Materials and Corrosion, 65(2014) 448-56. [4] P.A. Cella, S.R. Taylor, Electrical resistance changes as an alternate method for monitoring the corrosion of steel in concrete and mortar, Corrosion, 56(2000) 951-9. [5] L.R. Hilbert, Monitoring corrosion rates and localised corrosion in low conductivity water, Corrosion Science, 48(2006) 3907-23. [6] R.V. Rhoades, C.M. Finley, Corrosion measurement with multiple compensation for secondary temperature compensation, United States Patent: 4587479,1986-5-6.
ip t
[7] B. Hemblade, Electrical resistance sensor and apparatus for monitoring corrosion, United States
Patent: 6946855 B1, 2005-9-20. and acetic acid on Bottom-of-the-Line corrosion, Corrosion, 67(2011).
cr
[8] M. Singer, B. Brown, A. Camacho, S. Nesic, Combined effect of carbon dioxide, hydrogen sulfide, [9] Y. Tan, Y. Fwu, K. Bhardwaj, Electrochemical evaluation of under-deposit corrosion and its
us
inhibition using the wire beam electrode method, Corrosion Science, 53(2011) 1254-61.
[10] Y.H. Sun, T. Hong, W.P. Jepson, Pipeline corrosion under wet gas conditions, Material Performance, 40(2001) 48-52. probes, Electrochim Acta, 27 (2007) 7590-7598.
an
[11] A. Legat, Monitoring of steel corrosion in concrete by electrode arrays and electrical resistance [12] M. Singer, Study and modeling of the localized nature of top of the line corrosion, Ph.D., Ohio University, Chemical and Biomolecular Engineering, 2013.
M
[13] G.A. Zhang, N. Yu, L.Y, Yang, X.P. Guo, Galvanic corrosion behaviour of deposit-covered and uncovered carbon steel, Corrosion Scicence, 86 (2014) 202-212. pipelines, Corrosion, 66(2010).
d
[14] S. Papavinasam, A. Doiron, R.W. Revie, Model to predict internal pitting corrosion of oil and gas
Ac ce pt e
[15] G.P. Gerstmann J, Laminar film condensation on the underside of horizontal and inclined surfaces, Int J Heat Mass Transfer, 5(1967) 567-80.
[16] F. Vitse, S. Nesic, Y. Gunaltun, D.L. de Torreben, P. Duchet-Suchaux, Mechanistic model for the prediction of top-of-the-line corrosion risk, Corrosion, 59(2003) 1075-84. [17] Z. Zhang, D. Hinkson, M. Singer, H. Wang, S. Nesic, A mechanistic model of top-of-the-line corrosion, Corrosion, 63(2007) 1051-62.
[18] G. Hinds, A. Turnbull, Novel multi-electrode test method for evaluating inhibition of underdeposit corrosion-part 1: sweet conditions, Corrosion, 66(2010). [19] Y.M. Gunaltun, D. Supriyatman, J. Achmad, Top-of-line corrosion in gas lines confirmed by condensation analysis, Oil & Gas Journal, 97(1999) 64. [20] Y.-C. Chang, R. Woollam, M.E. Orazem, Mathematical models for under-deposit corrosion I. Aerated Media, Journal of the Electrochemical Society, 161(2014) C321-C9. [21] Y. Xie, L. Xu, C. Gao, W. Chang, M. Lu, Corrosion behavior of novel 3%Cr pipeline steel in CO2 Top-of-Line Corrosion environment, Materials and Design (2012) 54-57. [22] S. Nesic, Key issues related to modelling of internal corrosion of oil and gas pipelines - A review, Corrosion Science, 49(2007) 4308-38.
16
Page 16 of 43
Research Highlights
Ac
ce pt
ed
M
an
us
cr
ip t
Highlights: The structure and measurement principle of ring pair electrical resistance sensor (RPERS). RPERS can not only provide the general corrosion depth of the pipeline, but also can monitor the localized corrosion by six quadrants division. RPERS provides a better compensation than traditional wire ER sensor through the corrosion monitoring test in the laboratory. The temperature difference measurement method between the inner and outer pipe wall surface. The top of the line corrosion test has been monitored by the top quadrant of RPERS and the relationship of temperature difference, gas temperature and corrosion rates have been obtained.
Page 17 of 43
Table(s)
Table 1 Chemical component of the X65 pipeline steel: C
Si
Mn
P
S
V
Nb
Cr
Fe
Carbon steel
0.16
0.45
1.6
0.02
0.01
0.06
0.05
0.1
Bal.
Ac
ce pt
ed
M
an
us
cr
ip t
X65
Page 18 of 43
M
an
us
cr
ip t
Figure(s) illustrain
Ac
ce pt
ed
Fig. 1. Schematic illustration of the internal pipeline corrosion monitoring system.
Page 19 of 43
ip t cr us an M ed
ce pt
Fig. 2. The structure of RPERS (a) the three-dimension views of the mechanical structure of RPERS (b) the
Ac
dimensioning figure of the corrosion and compensation elements (c) the photogragh of RPERS.
Page 20 of 43
ip t cr
us
Fig. 3. The measurement principle diagram of the ring pair ER sensor with three excitation currents injected from
Ac
ce pt
ed
M
an
different directions.
Page 21 of 43
ip t
cr
Fig. 4. The geometry model and temperature distribution of a segment (a) The temperature distribution of a segment on the compensation element (b) the simplified rectangle model of a segment (c) the temperature gradient
Ac
ce pt
ed
M
an
us
distribution and the thin rectangular plate of a segment.
Page 22 of 43
ip t cr us an
Fig. 5. Internal-pipeline corrosion on-line monitoring system (a) Pipeline internal corrosion monitoring circulatory system: 1, water tank; 2, circular tube pump; 3, thermal insulation pipe; 4, wire ER sensor; 5, main chamber; 6,
Ac
ce pt
ed
M
electronics chamber (b) the simulative subsea environment.
Page 23 of 43
ip t cr us
Fig. 6. The measurement result of the temperature around the corrosion monitoring system in the simulative
Ac
ce pt
ed
M
an
seawater surrounding.
Page 24 of 43
ip t cr us
Fig. 7. The measurement results of the temperature difference by using the both two methods with 3.5% NaCl
Ac
ce pt
ed
M
an
solution flowing past.
Page 25 of 43
ip t cr us
Fig. 8. The measurement results of the temperature difference by using the both two methods with gas flowing
Ac
ce pt
ed
M
an
past.
Page 26 of 43
ip t cr us
an
Fig. 9. The results of the metal loss monitored by using WERS in 3.5% NaCl solution under different solution
Ac
ce pt
ed
M
temperatures.
Page 27 of 43
ip t cr
us
Fig. 10. The results of the metal loss monitored by using RPERS in 3.5% NaCl solution under different solution
Ac
ce pt
ed
M
an
temperatures.
Page 28 of 43
ip t cr us an
Ac
ce pt
ed
M
Fig. 11. The result of the metal loss measured by the top segment of RPERS with gas flowing past.
Page 29 of 43
ip t cr us
values.
Ac
ce pt
ed
M
an
Fig. 12. The result of the calculated corrosion rate under different
Page 30 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 1. Click here to download high resolution image
Page 31 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 2. Click here to download high resolution image
Page 32 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 3. Click here to download high resolution image
Page 33 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 4. Click here to download high resolution image
Page 34 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 5. Click here to download high resolution image
Page 35 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 6. Click here to download high resolution image
Page 36 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 7. Click here to download high resolution image
Page 37 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 8. Click here to download high resolution image
Page 38 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 9. Click here to download high resolution image
Page 39 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 10. Click here to download high resolution image
Page 40 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 11. Click here to download high resolution image
Page 41 of 43
Ac
ce
pt
ed
M
an
us
cr
i
Fig. 12. Click here to download high resolution image
Page 42 of 43
Author Biographies
Biographies Yunze Xu was born in 1989 in Henan, China. He received his bachelor degree in marine engineering in 2012 from the Huazhong University of Science and Technology. After his diploma degree, he started to work as PhD student at School of Naval Architecture in Dalian University of Technology. He is working on the pipeline corrosion monitoring and corrosion mechanism
ip t
research. Yi Huang was born in 1964 in Dandong, China. He obtained his PhD from the Hiroshima University of Japan in September 1992. From 1992 to 2001, he worked at the Hiroshima
cr
University on studies of corrosion and anticorrosion problems for ship and ocean engineering. Then he was invited to Dalian University of Technology and became a full professor in School of
us
Naval Architecture. There he managed a group working on the corrosion and anticorrosion problems for ship and ocean engineering, monitoring and inspection technique for ship hull and
Ac
ce pt
ed
M
an
marine structures.
1
Page 43 of 43