International Scholarly Research Network

ISRN Corrosion
Volume 2012, Article ID 237025, 8 pages
doi:10.5402/2012/237025

Research Article
Predictive Modelling of Wellhead Corrosion due to
Operating Conditions: A Field Data Approach

Chinedu I. Ossai
Production Planning Department, Overall Forge Pty Ltd., 70 R W Henry Drive, Ettamogah near Albury, P.O. Box 5275,
Albury, NSW 2640, Australia

Correspondence should be addressed to Chinedu I. Ossai, ossaic@gmail.com

Received 3 October 2012; Accepted 19 November 2012

Academic Editors: M. Criado, M. Scendo, and R. Wang

Copyright © 2012 Chinedu I. Ossai. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The flow of crude oil, water, and gas from the reservoirs through the wellheads results in its deterioration. This deterioration which
is due to the impact of turbulence, corrosion, and erosion significantly reduces the integrity of the wellheads. Effectively managing
the wellheads, therefore, requires the knowledge of the extent to which these factors contribute to its degradation. In this paper,
the contribution of some operating parameters (temperature, CO2 partial pressure, flow rate, and pH) on the corrosion rate of
oil and gas wellheads was studied. Field data from onshore oil and gas fields were analysed with multiple linear regression model
to determine the dependency of the corrosion rate on the operating parameters. ANOVA, P value test, and multiple regression
coefficients were used in the statistical analysis of the results, while in previous experimental results, de Waard-Milliams models and
de Waard-Lotz model were used to validate the modelled wellhead corrosion rates. The study shows that the operating parameters
contribute to about 26% of the wellhead corrosion rate. The predicted corrosion models also showed a good agreement with the
field data and the de Waard-Lotz models but mixed results with the experimental results and the de Waard-Milliams models.

1. Introduction resulted in pipeline failures, failure attributable to sour and

sweet corrosion which results from activities of CO2 and H2 S
Corrosion of materials is a major challenge to maintaining has contributed to over 50% of all pipeline failures [3].
the integrity of equipment in the industry. Mobile and static Over the decades, experts have worked to combat the
mechanical equipment like pipelines, vessels, tanks, com- menace of corrosion with varying degree of success. They
pressors, turbines, and so forth have been periodically sub- have used many techniques that include use of alternative
jected to degradation and failure due to corrosion. According materials, modification of microstructures, addition of new
to reports [1], the cost of corrosion in the US will hit 6.2% elements to existing alloys of metals, use of chemical inhi-
of GDP in 2012, making it the most expensive single venture bitors, and linings; however, the act of corrosion has not been
in the economy. The impact of corrosion in the oil and gas prevented from occurring. Corrosion has just continued to
industry significantly contributes to the nonproductive time occur because materials are continually trying to return to
(NPT) of 20–30% lost from exploration to production [2]. the original energy state via lowering of their Gibb’s energy
Corrosion of pipeline accounts for about 35% of failures [6].
associated with pipeline leakage and bursting in Canada [3], Considering the overwhelming importance of pipelines
while 50% of the loss of containment hazard in Europe in the oil and gas industry, the contribution of experts in
between 1980 and 2006 was a result of the ageing plant mech- combating the corrosion problem cannot be overempha-
anism [4] triggered by corrosion-related fatigue cracking and sized. This is pertinent to the safeguarding of the environ-
erosion. Corrosion also ranks second to the highest most fre- ment, personnel, and cost associated with oil and gas pro-
quent initiating factor leading to loss of containment in UK duction. The growth in the demand of oil and gas around the
[4, 5]. While external corrosion, stress corrosion cracking, world has increased the need of exploration in unfriendly ter-
and microbiologically influenced corrosion have significantly rains like deep waters and the Antarctica. This has increased
2 ISRN Corrosion

the pressure on the exploration equipment and consequently applied a mechanistic model to predict the corrosion dam-
the pipeline which plays a major role in the transportation age rate of distribution observed in oil wells operating
from the oil and gas reservoirs. The consequences of this within mature oil field. The work showed that CO2 can be
pressure on pipelines and other equipment are overutiliza- quantitatively determined for a uniform corrosion in pipe-
tion, reduction in life cycles, and failures [7, 8]. line that resulted in localized attack.
To reduce the problem of corrosion and enhance pipeline Many other experimental results about the impact of
integrity, corrosion experts have worked on different cor- CO2 , H2 S, temperature, O2 , acetic acid, and bacteria on cor-
rosion prediction models in a bid to identify the best rosion of carbon steel are found readily in the literature
way to determine corrosion in pipelines. Many authors [9–13, 27, 28]. Temperature increase has been shown to
have investigated the corrosion problem in oil and gas increase corrosion rate until a particular threshold when
pipelines using electrochemical, mechanistic, semiempirical, the increase stops to affect corrosion [22, 29]. The presence
empirical, hybrid, and probabilistic methods and so forth [9– of O2 enhances the corrosion process, while H2 S has been
16]. Their works have focused on knowing the contribution shown to cause localized pitting corrosion in pipelines [27].
of different operating parameters in the oil and gas such Acetic acid and other organic and inorganic acids also aid
as CO2 , H2 S, temperature, bacteria, and so forth on the in the increasing of corrosion rate and initiation of loca-
entire corrosion process. Others investigated the impact of lized corrosion and pitting [30, 31]. Other authors utilized
the mechanical process of fluid flow and associated effects mechanistic modelling to predict corrosion rate in steel sur-
of turbulence on corrosion of pipelines, well tubing, vessels, faces. This approach which involved homogenous chemical
tanks, and so forth [17–19]. Despite the enormity of the reaction, electrochemical reaction at the steel surface, and
work on corrosion, it is difficult to summarily prescribe a transportation of specie in the bulk solution was utilized
holistic solution to the corrosion process. This is because of by many authors in their prediction of CO2 corrosion rate
the complex electrochemical and physical processes involved. [11, 21, 32].
The most important parameters that can enhance the result Due to the importance of wellheads in the transportation
of corrosion prediction are those related to the steel proper- of oil and gas from the reservoir to the pipelines, it is neces-
ties, water chemistry, flow pattern, oil-versus-water wetting, sary to maintain the integrity at all times; however, due to
and operating conditions [20]. This is why the works of turbulence, corrosion, erosion, and other factors, they are
numerous authors focused on them [7, 8, 10, 11, 15, 19–22]. continuously deteriorating in operation. To be able to under-
Corrosion prediction models that were aimed at the mul- stand more about this deterioration as a result of corrosion,
tiphase flow regime, pH, H2 S, CO2 , and so forth were used the effect of certain operating parameters on the wellhead
by some authors [20–22] to establish the extent of internal corrosion rates was studied.
corrosion of pipelines at different operating conditions. This In this study, historic field data of wellhead corro-
helped to establish the point where mitigation is necessary sion rates measured with ultrasonic thickness measurement
in order to reduce the level of risk that corrosion poses (UTM) technique, temperature, CO2 partial pressure, flow
on the pipeline. Chemical inhibitors are injected in most velocity of crude oil, mixed flow velocity, and pH were used.
instances to mitigate the effect of the corrodents; while in The main objective of this research is to establish to which
extreme cases, more severe measures like replacement of the extent these operating parameters affect wellhead corrosion
pipeline with more resistant materials are considered. A lot of rate.
authors have numerous corrosion experimental results [11,
15, 16, 21, 22] that were not verified with field conditions;
this resulted in numerous laboratory results not having good
2. Research Methodology
practical relevance in the field. It is, therefore, important to Historical data from the company operating the Nigerian oil
understand the trend thatcorrosion field data follows with mining license (OML61) were obtained from the organiza-
respect to the operating condition in order to enhance the tion’s historical records. The wellhead corrosion rates were
acquisition of corrosion prediction results that will have high determined by using the ultrasonic thickness measurement
practical relevance. Though the works of the utilization of (UTM) technique, while the temperature, operating pres-
case-based reasoning (CBR) [23], and stochastic modelling sure, pH, CO2 content, and production rates of oil, gas, and
for CO2 corrosion via metal loss and precipitation process water were obtained as a routine well monitoring procedure.
[24] were excellent works, they have limited validation with The flow rates of the fluids were obtained through calculation
field data. This made to have little application in the industry. using the information from the parameters obtained. The
Many pipeline corrosion models failed to recognize the summary of the data obtained is shown in Tables 1 and 2.
contribution of H2 S, organic acids, and microorganisms in
the corrosion process; however, some authors [25] built a
2.1. Model Development. The wellhead corrosion rate can be
comprehensive integrated CO2 /H2 S corrosion multiphase
expressed as a function of the operating parameters of the
flow model. The interaction of the experimental results with
production well according to the following:
the field data made it possible for the authors to predict the
critical velocity for entraining free water by the flowing oil.
 
This result shows a good practical relevance for corrosion CR = f T, PCO2 , Vm , pH , (1a)
prediction of pipelines. In a bid to show the contribution  
of gas flow in pipeline corrosion, another researcher [26] CR = f T, PCO2 , V , (1b)
ISRN Corrosion 3

Table 1: Summary of oil, gas, and water production rates of the studied wells.

Length of pipeline 900–14650 m Oil production rate 23–249 m3 /day

Pipeline external diameter 0.114 m Gas production rate 0.006–0.497 SMm3 /day
Pipeline internal diameter 0.09718 m Water production rate 0.12–1094.64 m3 /day

Table 2: Summary of the field operating parameters at the wellheads.

CR (mm/yr) T (◦ C) PCO2 (MPA) Vm (m/s) V (m/s) Pop (MPA) pH

Mean 0.545 61 0.203 0.447 0.158 8.62 7.65
Standard deviation 0.283 7.78 0.128 0.442 0.088 3.02 0.64

 
CR = f T, PCO2 , Vm , (1c) Solving for the residual sum of squares by introducing an
  error factor (e) to (2a)–(2d), a normal least equation of the
CR = f T, PCO2 , (1d) form shown in (3) is obtained as follows:

x xβ = xy  ,
where CR = wellhead corrosion rate (mm/yr), T = temper- (k+1 × k+1)(k+1 ×1)
(3)
ature (◦ C), PCO2 = CO2 partial pressure (mpa), Vm = mixed ⎡ ⎤
velocity of fluid flowing through the wellhead (m/s), and V n xi1 xi2 ... xik
= flow velocity of the crude oil (m/s). ⎢ ⎥
⎢ xi1 xi1
2
xi1 xi2 ... xi1 xik ⎥
⎢ ⎥
To model the effect of the listed operating parameters on ⎢ ⎥
x x = ⎢
⎢
xi2 xi1 xi2 xi2
2
... xi2 xik ⎥,
⎥ (4)
the corrosion rate of the wellhead, a linear regression model ⎢ ⎥
⎢ .. .. .. .. .. ⎥
is adopted. The regression equation that is used to predict ⎣ . . . . . ⎦
the impact of the operating parameters on the wellhead xik xi1 xik xi2 xik ... xik2
corrosion rate (due to the combination of these operating
parameters) is shown in (2a)–(2d): ⎡ ⎤
yi
⎢ ⎥
⎢ xi1 yi ⎥
⎢ ⎥
CR = α11 + β11 T + β12 PCO2 + β13 Vm + β14 pH, (2a) ⎢ ⎥
xy  = ⎢
⎢
xi2 yi ⎥
⎥. (5)
⎢ .. ⎥
CR = α21 + β21 T + β22 PCO2 + β23 V , (2b) ⎢ . ⎥
⎣ ⎦
xik yi
CR = α31 + β31 T + β31 PCO2 + β32 Vm , (2c)

CR = α41 + β41 T + β42 PCO2 , The solution of the normal equation (3) gives the solution
(2d)
of the independent coefficients of the operating parameters.
This is given as follows:
where αik and βik are the normal and slope coefficients due
to the operating parameters (T, pH, V , Vm , and PCO2 ) β = (x x)−1 xy  . (6)
Equations (2a)–(2d) can be written as a single matrix
equation as shown in (1). Consider Multiple regression and analysis of variance (ANOVA)
test was performed at 95% confidence level to examine the
combined effect of temperature, CO2 partial pressure, flow
y = xβ , (1) velocity, and pH on wellhead corrosion rate of the studied oil
(n+1) (n×k+1)(k+1 ×1)
fields.

where x = (T, PCO2 , Vm , V , or pH), y = (CR1 , CR2 , . . . , 2.2. Validation of the Model. To test the accuracy of these
CRn ) is a vector of measured wellhead corrosion rates at models, three out of the thirty-three field data obtained from
different levels of the operating parameters (x variables). OML61, de Waard and Milliams model (DM) [13], modified
The matrix form of the operating parameters is shown to de Waard and Milliams model (MDM) [9], de Waard and
be in the form of Lotz model (DL) [12], and experimental results of Zhou and
Jepson (1993), Jepson and Menezes (1994), and Bhongale
⎡ ⎤
1 x11 . . . x1k et al. (1996) from the NSF/IUCRC multiphase system centre
⎢ ⎥ (as reported by Zhang et al. (1997) [21]) were used.
x = ⎣1 x21 . . . x2k ⎦, (2)
1 xn1 . . . xnk
2.3. De Waard and Milliams Model. This is one of the oldest
known mechanistic models for predicting CO2 corrosion.
β = (α, β1, β2 , . . . , βk ) and contains the regression coefficients This model is based on the electrochemical studies carried
due to the operating parameters. out by de Waard and Milliams to show the correlation
4 ISRN Corrosion

between temperature (◦ C), CO2 partial pressure, and corro- This could imply that only the corrosion rate determined
sion rate (mm/yr). The equation for the CO2 corrosion rate using temperature and CO2 partial pressure is not statis-
is shown as follows: tically significant; however, the result shown in Table 3 for
2320 adjusted square of correlation coefficient R2 shows that none
log(CRt ) = 7.96 − of the combinations is statistically significant due to the low
T + 273 (7)
 
effects of the operating parameters on the wellhead corrosion
− 5.55 × 10−3 T + 0.67 log PCO2 , rate.
The P value of the variables shows that the coefficients of
where: CRt = corrosion rate (mm/yr), T = temperature (◦ C), the operating parameters have values higher than the 0.05 for
PCO2 = CO2 partial pressure (mpa). all the variables as shown in Table 5.
De Waard and Milliams revised the constants of the This is a further proof that the test is not statistically sig-
formula in (7) based on the experimental results of Dugstad nificant. Though some of the studied operation parameters
et al. (1994) [10] and obtained the following expression: CR(T, PCO2 , pH, Vm ), CR(T, PCO2 , V ), and CR(T, PCO2 , Vm )
  were statistically significant with the result of ANOVA in
1710
log(CRt ) = 5.8 − + 0.67 log PCO2 . (8) Table 4, the P value and square of correlation coefficient (R2 )
T + 273 show statistical insignificance. This result could imply that
To further validate this model, de Waard and Lotz modelled the effect of other electrochemical, chemical, and mechanical
corrosion rate with respect to velocity in the absence of sur- activities associated with the flow of the fluid might have also
face scale with a parallel resistance model [12]. The result of affected the corrosion rate. The activities of bacteria and the
their experimental analysis is summarized in (9) the follow- organic and inorganic compounds could have also affected
ing: the corrosion rate as was pointed by some of these authors
[30, 33–35]. Another possible cause of minimal effect of the
1 1 1 operating parameters on the wellhead corrosion rate is the
= + , (9)
Vcr Vr Vm bubbles production effect of multiphase flow system due
where Vcr = corrosion rate, Vr = flow-independent contri- to turbulence in the flow regime. This is supported by the
bution denoting the reaction rate, and Vm = flow-dependent research on the effect of bubbles on mass transfer in multi-
contribution denoting the mass transfer rate as follows: phase flow [32].
The coefficients of the operating parameters in Table 5
1119   were used to develop the set of linear regression equations in
log(Vr ) = 4.93 − + 0.58 log PCO2 , (10)
T + 273 (12)–(15) for wellhead corrosion rate as follows:
where T = temperature (◦ C), and PCO2 = CO2 partial pressure CR = 3.8 × 10−3 T + 9.97 × 10−1 PCO2 + 2.379 × 10−1 Vm
(mpa). Consider
+ 5.97 × 10−2 pH − 4.508 × 10−1 ,
U 0.8 (12)
Vm = 2.45 PCO2 , (11)
Dh0.8
CR = 4.1 × 10−3 T + 9.48 × 10−1 PCO2
where U = liquid flow rate (m/s), and Dh = hydraulic diam- (13)
eter of the pipe. + 2.32 × 10−1 Vm − 9.1 × 10−4 ,

CR = 3.196 × 10−3 T + 8.196 × 10−1 PCO2

3. Results and Discussions (14)
The correlation results of the corrosion rate in the wellhead + 1.333V − 2.61 × 10−2 ,
as a dependent variable and the operating parameters (T, CR = 7.01 × 10−2 + 5.138 × 10−3 T + 7.97 × 10−1 PCO2 .
pH, Vm , V , and PCO2 ) are shown in Table 3.
(15)
The correlation coefficient R2 ranges from 0.168 to 0.336,
while the adjusted R2 ranges from 0.106 to 0.259. This result The modeled result for wellhead corrosion rate due to
implies that only 10.6% of the effect of temperature and CO2 temperature, CO2 partial pressure, pH, and mixed flow rate
partial pressure contributed to the corrosion rate on the oil (see (12)) was compared with field data, and the result is
field wellheads when the corrosion rate was determined as a shown in Figure 1.
function of only the two parameters, while 25.9% of wellhead The result shows that there are some similarities between
corrosion was caused by the combined effect of temperature, the modelled result and field data. Whereas the field data
CO2 partial pressure, oil flow rate, and pH of the associated ranges from 0.06 to 0.51 mm/yr, the modelled result ranged
formation of water in the mixture. from 0.13 to 0.38 mm/yr. This result shows that there is a
The field data showed significance at F values of reasonable degree of agreement between the field data and
0.045, 0.013, and 0.026 for wellhead corrosion rates with the modelled prediction results.
the following combination of operating parameters: CR(T, The result of wellhead corrosion rate as a function of
PCO2 , pH, Vm ), CR(T, PCO2 , V ), and CR(T, PCO2 , Vm ), while temperature, CO2 partial pressure, and mixed flow velo-
F is not significant at 0.083 for CR(T, PCO2 ) at 95% con- city of the fluid (see (13)) was compared with experi-
fidence level as shown in Table 4. mental results of Zhou and Jepson (1993), Jepson and
ISRN Corrosion 5

Table 3: Summary of the results of regression analysis.

Operating parameter Multiple R R2 Adjusted R2 Standard error Observation

CR(T, PCO2 ) 0.4099 0.1680 0.1064 0.2677
CR(T, PCO2 , Vm ) 0.5428 0.2946 0.2132 0.2511 30
CR(T, PCO2 ,V ) 0.5793 0.3356 0.2589 0.2437
CR(T, PCO2 , pH, Vm ) 0.5589 0.3123 0.2023 0.2529

Table 4: Summary of ANOVA for wellhead corrosion and operating parameters.

Measured parameters df SS MS F Significance F

Regression 4 0.72610 0.1815 2.8385
CR(T, PCO2 , pH, Vm ) Residual 25 1.59877 0.0640 0.045
Total 29 2.32487
Regression 3 0.78025 0.2601 4.3779
CR(T, PCO2 , V ) Residual 26 1.54462 0.0594 0.012709
Total 29 2.32487
Regression 3 0.68499 0.2283 3.6201
CR(T, PCO2 , Vm ) Residual 26 1.63988 0.0631 0.026244
Total 29 2.32487
Regression 2 0.39064 0.1953 2.7265
CR(T, PCO2 ) Residual 27 1.93423 0.0716 0.083462
Total 29 2.32487

0.6 4
3.5
Corrosion rate (mm/yr)

0.5 3
2.5
0.4
2
CR (mm/yr)

1.5
0.3
1

0.2 0.5
0
20 30 40 50 60 70 80
0.1
Temperature (◦ C)

0 De Waard-Lotz model
1 2 3 CRmodel
Oil and gas fields Figure 2: Summary of modelled wellhead corrosion rate and de
CR field Waard-Lotz Model PCO2 = 0.1 mpa, V = 1 m/s.
CR model

Figure 1: Wellhead corrosion rate in the field versus the model.

relationship across the various CO2 partial pressure and oil
flow velocity.
The model reasonably predicted the corrosion rate with
Menezes (1994), Bhongale et al. (1996). This result was a small degree of variance at lower CO2 partial pressure and
reported by Zhang et al. [21]. The summary of the results higher flow velocity with modelled corrosion rate higher
comparison is shown in Table 6. than the experimental corrosion rate at PCO2 = 0.136 mpa
The result shows that there is variation between the and V = 1–1.8 m/s. The prediction was much lower than the
experimental results and the prediction model across the experimental values at lower velocity and higher CO2 partial
range of mixed velocity and CO2 partial pressure. pressure. The result of the model equation (14) also showed
The same set of experimental results reported by Zhang a good agreement with de Waard-Lotz model [12] as shown
et al. (1997) [21] was used to compare the result of the in Figure 2.
model using (14) in Table 7. The result shows that the For temperature between 30–80◦ C, 0.1 MPA, and velocity
modelled results and the experimental results have a mixed of 1 m/s, the predicted model gave a corrosion rate that
6 ISRN Corrosion

Table 5: Summary of parameter estimate.

Variable Coefficient Standard error P value t-test Lower 95% Upper 95%
β41 0.0038 0.0062 0.6098 0.5475 −0.009 0.0165
β42 0.997 0.386 2.585 0.016 0.2028 1.7927
β43 0.2379 0.1085 2.1922 0.0379 0.0144 0.4614
β44 0.0597 0.0745 0.8018 0.4302 −0.0937 0.2132
α45 −0.4508 0.6738 −0.669 0.5097 −1.8386 0.937
β21 0.0041 0.00613 0.668308 0.509827 −0.0085 0.016697
β22 0.948 0.378337 2.505766 0.01881 0.170341 1.725707
β23 0.2323 0.107545 2.160294 0.040147 0.011267 0.453389
β31 0.003196 0.005979 0.534545 0.597507 −0.00909 0.015486
β32 0.8196 0.360991 2.270454 0.031701 0.077586 1.561642
β33 1.3327 0.520421 2.560891 0.016593 0.263001 2.402483
α31 −0.0261 0.360199 −0.07247 0.942781 −0.76651 0.714296
α11 0.0701 0.393384 0.17819 0.859904 −0.73706 0.877254
β11 0.0051 0.006513 0.788925 0.437029 −0.00822 0.018501
β12 0.7973 0.396294 2.011877 0.05431 −0.01583 1.610425
α21 −0.0009 0.370577 −0.00245 0.998066 −0.76264 0.760825

Table 6: Summary of experimental result and modelled equation of Table 7: Summary of experimental result and modelled equation of
wellhead corrosion rate using mixed flow velocity (Vm ), CO2 partial wellhead corrosion rate using crude flow velocity (V ), CO2 partial
pressure, and temperature [T = 40◦ C]. pressure, and temperature [T = 40◦ C].

CO2 partial pressure Mixed velocity Experiment Modelled CO2 partial pressure Oil flow rate Experiment Modelled
P (MPA) Vm (m/s) CREXP CRmodel PCO2 (MPA) V (m/s)
CRexp CRpred
(mm/yr) (mm/yr)
0.27 0.28 4 1.96001
0.136 1 0.88 1.54595
0.27 0.56 4.9 2.02497
0.136 1.3 1.23 1.945772
0.27 1 5.4 2.12705
0.136 1.8 1.75 2.612143
0.45 0.28 6.25 2.13065
0.27 0.18 2.9 0.562929
0.45 0.56 7.4 2.19561
0.27 0.28 3 0.696204
0.45 1 7.9 2.29769
0.27 1 4.25 1.655778
0.79 0.28 9.2 2.45297
0.45 0.18 3.4 0.71046
0.79 0.56 10.75 2.45297
0.45 0.28 5.6 0.843734
0.79 1 11.6 2.62001
0.45 1 8.6 1.803308
0.79 0.18 5.6 0.989129
0.79 1 11.4 2.081977

is generally slightly higher than that predicted with the de

Waard-Lotz model, while at 20◦ C, the corrosion rate pre-
dicted by de Waard-Lotz model is slightly higher than that 4. Conclusion
predicted by the predicted model. The range of prediction
for the prediction model is 1.15–3.61 mm/yr, while that of de The effect of the operating parameters (temperature, CO2
Waard-Lotz model is 1.34–2.16 mm/yr. partial pressure, pH, and flow rate) on the rate of corrosion of
Figure 3 shows that as the temperature increases, the pre- oil and gas wellheads was studied using regression modelling
dicted wellhead corrosion model (see (15)) increases as both of historical field data. The study shows that about 26% of
well as de Waard-Milliams model [13] and the modified de wellhead corrosion is caused by these operating parameters.
Waard and Milliams model [12]. At temperature of 20–80◦ C, Other factors that could have an impact on wellhead cor-
the predicted wellhead corrosion rate ranges from 0.253– rosion rate are organic and inorganic acids, bacteria, tur-
0.561 mm/yr, the de Waard-Milliams model ranges from bulence, erosion, erosion-corrosion, bubbles formation, and
0.448 to 0.469 mm/yr, while the modified de Waard-Milliams condensation resulting from the flow of the oil and gas
model shows a range of 0.391–1.077 mm/yr. [18, 32, 34–37].
ISRN Corrosion 7

1.2 [11] M. Nordsveen, S. Nešić, R. Nyborg, and A. Stangeland, “A

mechanistic model for carbon dioxide corrosion of mild steel
1
Corrosion rate (mm/yr)

in the presence of protective iron carbonate films—part 1:

0.8
theory and verification,” Corrosion, vol. 59, no. 5, pp. 443–456,
2003.
0.6 [12] C. de Waard, U. Lotz, and A. Dugstad, “Influence of liquid flow
velocity on CO2 corrosion a semi-empirical model,” NACE
0.4 International Conference Sereis, Corrosion/95, paper no. 128,
Houston, Tex, USA, 1995.
0.2
[13] C. De Waard and D. E. Milliams, “Cabonic acid corrosion of
0 steel,” Corrosion, vol. 31, no. 5, pp. 177–181, 1975.
0 10 20 30 40 50 60 70 80 90 [14] R. Nyborg, “Controlling internal corrosion in oil and gas pipe-
Temperature (◦ C) line,” Business Briefing-Exploration & Production, vol. 2005, no.
2, pp. 70–74, 2005.
DM [15] W. Sun and S. Nesic, “A mechanistic model of H2S corrosion
MDM of mild steel,” NACE International Conference Sereis, COR-
CRmodel
ROSION 2007 paper 07655, 2007.
Figure 3: Summary of modelled wellhead corrosion rate, de Waard- [16] K.-L. Lee and S. Nesic, “EIS investigation on the electrochem-
Milliams model and Modified de Waard Milliams model (PCO2 = istry of CO2 /H2S corrosion,” NACE International Conference
0.1 mpa). Sereis, CORROSION 2004 paper 04728, 2004.
[17] M. E. Mohyaldinn, N. Elkhatib, and M. C. Ismail, “A compu-
tational tool for erosion/corrosion prediction in oil/gas pro-
The predicted corrosion model was validated with field duction facilities,” in Proceedings of the ICSSST2010 3rd Inter-
national Conference on Solid State Science &Technology, Kuch-
data, experimental results, and de Waard-Milliams corrosion
ing, Sarawak, Malaysia, 2010.
models. The result showed a good agreement between the
[18] S. Nesic and J. Postlethwaite, “Relationship between the
field data and mixed results of both the experimental and de structure of disturbed flow and erosion-corrosion,” Corrosion,
Waard-Milliams models. vol. 46, no. 11, pp. 874–880, 1990.
[19] S. Hassani, K. P. Roberts, S. A. Shirazi, J. R. Shadley, E. F.
References Rybicki, and C. Joia, “Flow loop study of chloride concen-
tration effect on erosion, corrosion and erosion-corrosion of
[1] “Cost of Corrosion to Exceed $1 Trillion in the United carbon steel in CO2 saturated systems,” Corrosion, vol. 68, no.
States in 2012 — G2MT Labs—The Future of Materials Con- 2, 2012.
dition Assessment,” http://www.g2mtlabs.com/2011/06/nace- [20] A. Dugstad, E. Gulbrandesen, J. Kvarekvål, R. Nyborg, and M.
cost-of-corrosion-study-update/. Seiersten, “Corrosion testing in multiphase flow, challenges
[2] R. Nicholson, J. Feblowitz, C. Madden, and R. Bigliani, and limitations,” NACE International Conference Sereis, Cor-
“The Role of Predictive Analytics in Asset Optimization for rosion/2006 paper no 06598, 2006.
the Oil and Gas Industry-White Paper,” 2010, http://www [21] R. Zhang, M. Gopal, and W. P. Jepson, “Development of a
.tessella.com/wp-content/uploads/2008/02/IDCWP31SA4Web Mechanistic model for predicting Corrosion rate in Multi-
.pdf. phase oil/water/gas flows,” NACE International Conference
[3] CAPP, “Best Management Practices: Mitigation of Internal Sereis, Corrosion /97, paper 601, 1997.
Corrosion in Oil Effluent Pipeline Systems,” 2009, http:// [22] W. P. Jepson, S. Bhongale, and M. Gopal, “Predictive model
www.capp.ca/getdoc.aspx?DocId=155641&DT=PDF. for sweet corrosion in horizontal multiphase slug flow,”
[4] Control of Major Accident Hazards, “Ageing Plant Opera- NACE International Conference Sereis, Corrosion/96: paper
tional Delivery Guide,” http://www.hse.gov.uk/comah/guid- 19, 1996.
ance/ageing-plant-core.pdf. [23] B. Khajotia, D. Sormaz, and S. Nesic, “Case-based reasoning
[5] P. Horrocks, D. Mansfield, K. Parker, J. Thomson, T. Atkin- model of CO2 corrosion based on field data,” NACE Inter-
son, and J. Worsley, “Managing Ageing Plant,” http://www national Conference Sereis, CORROSION 2007 paper 07553,
.hse.gov.uk/research/rrpdf/rr823-summary-guide.pdf. 2007.
[6] J. Kruger, “Electrochemistry of Corrosion,” 2001, http://elec-
[24] Y. Xian and S. Nesic, “A stochastic prediction model of loca-
trochem.cwru.edu/encycl/art-c02-corrosion.htm.
[7] “Review of corrosion management for offshore oil and gas lized CO2 corrosion,” NACE International Conference Sereis,
processing,” HSE Offshore Technology Report 2001/044, 2001. CORROSION 2005 paper No 05057, 2005.
[8] P.O. Gartland and R. Johnsen, “Application of internal [25] N. Srdjan, J. Cai, and K. L. John Lee, “A multiphase flow and
corrosion modelling in risk assessment of pipeline,” NACE internal corrosion prediction model for mild steel pipeline,”
International Conference Sereis, Corrosion 2003 paper no NACE International Conference Sereis, CORROSION 2005
03179, 2003. PAPER 05556, 2005.
[9] C. De-Waard and V. Lotz, “Prediction of CO2 corrosion of [26] R. Case, “Assessment of CO2 corrosion damage distribution
carbon steel,” NACE International Conference Sereis, Corro- in Oil wells by deterministic and Stochastic Modelling,”
sion /93. Paper no. 69, NACE int., Houston, Tex, USA, 1993. NACE International Conference Sereis, Corrosion/2008. Paper
[10] A. Dugstad, L. Lunde, and K. Videm, “Parametric study of 08514, 2008.
CO2 corrosion of carbon steeel,” NACE International Confer- [27] H. Fang, B. Brown, and S. Nescaronicacute, “Effects of sodium
ence Sereis, Corrosion/94. Paper no 14, NACE, Houston, Tex, chloride concentration on mild steel corrosion in slightly sour
USA, 1994. environments,” Corrosion, vol. 67, no. 1, 2011.
8 ISRN Corrosion

[28] M. Singer, B. Brown, A. Camacho, and S. Nešić, “Combined

effect of carbon dioxide, hydrogen sulfide, and acetic acid on
bottom-of-the-line corrosion,” Corrosion, vol. 67, no. 1, 2011.
[29] Y. Song, A. Palencsár, G. Svenningsen, J. Kvarekvål, and T.
Hemmingsen, “Effect of O2 and temperature on sour corro-
sion,” Corrosion, vol. 68, no. 7, pp. 662–671, 2012.
[30] V. Fajardo, C. Canto, B. Brown, and S. Nesic, “Effect of organic
acids in CO2 corrosion,” NACE Internatinal Conference &
Expo. CORROSION 2007 paper 07319, 2007.
[31] C. Mendex, M. Singer, A. Camacho, S. Hernndez, and S. Nesic,
“Effect of acetic acid pH and MEG on CO2 top of the line
corrosion,” NACE International Conference Sereis, Corrosion
2005, paper 05278, 2005.
[32] H. Wang, W. Paul Jepson, J. Y. Cai, and M. Gopal, “Effect of
bubbles on mass transfer in multiphase flow,” NACE Inter-
national Conference Sereis, CORROSION 2000 paper 00050,
2005.
[33] X. Hu, A. Neville, J. Wells, and V. De Souza, “Prediction
of erosion-corrosion in oil and gas—a systematic approach,”
NACE International Conference Sereis, CORROSION 2008
paper 08540, 2008.
[34] A. A. Sami and A. A. Mohammed, “Study synergy effect on
erosion-corrosion in oil and gas pipelines,” Engineering &
Technology , vol. 26, no. 9, 2008.
[35] J. Wen, T. Gu, and S. Nesic, “Investigation of the effects of fluid
flow on srb biofilm,” NACE International Corrosion Con-
ference Series, Corrosion/2007 Paper no 07516, 2007.
[36] A. Keating and S. Nesic, “Prediction of two-phase erosion-cor-
rosion in bends,” in Proceedings of the 2nd International Con-
ference on CFD in Minerals and Processes Industries CSIRO,
Melbourne, Australia, 1999.
[37] S. Hassani, K. P. Roberts, S. A. Shirazi, J. R. Shadley, E. F.
Rybicki, and C. Joia, “Flow loop study of chloride concen-
tration effect on erosion, corrosion and erosion-corrosion of
carbon steel in CO2 saturated systems,” Corrosion, vol. 68, no.
2, 2012.
Journal of International Journal of International Journal of Smart Materials Journal of
Nanotechnology
Hindawi Publishing Corporation
Corrosion
Hindawi Publishing Corporation
Polymer Science
Hindawi Publishing Corporation
Research
Hindawi Publishing Corporation
Composites
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of
Metallurgy

BioMed
Research International
Hindawi Publishing Corporation Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Nanomaterials
Submit your manuscripts at
http://www.hindawi.com

Journal of Journal of
Materials
Hindawi Publishing Corporation
Nanoparticles
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Nanomaterials
Journal of

Advances in The Scientific International Journal of

Materials Science and Engineering
Hindawi Publishing Corporation
Scientifica
Hindawi Publishing Corporation Hindawi Publishing Corporation
World Journal
Hindawi Publishing Corporation
Biomaterials
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of Journal of Journal of Journal of Journal of

Nanoscience
Hindawi Publishing Corporation
Coatings
Hindawi Publishing Corporation
Crystallography
Hindawi Publishing Corporation
Ceramics
Hindawi Publishing Corporation
Textiles
Hindawi Publishing Corporation
http://www.hindawi.com
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Volume 2014