INTRODUCTION
Corrosion is a common form of structure degradation that reduces both the static
and cyclic strength of a pipeline. There is always the chance that pipelines
could leak or rupture and a pipeline failure can cause serious human, environmental
and financial losses (Hopkins, 1995; National
Energy Board, 1996). As reported by Riemer (2000),
the transportation of petroleum products and natural gas has begun since early
of 20th century by using buried steel pipelines over long distance. There are
over 1.28 million miles of buried steel main-line pipe for the transport of
natural gas alone (American Gas Association, 1999).
It is the safest method to burry onshore and offshore transportation pipelines
under soil for mechanical protection and to provide thermal insulation. Researchers
have classified failure mechanisms associated with pipelines failures into three
categories which are third-party damage, material defects and corrosion (Alodan
and Abdulaleem, 2007; Ahammed and Melchers, 1997;
Cheuk et al., 2007). Corrosion is recognised
as one of the most dominant forms of deterioration process and has been identified
as one of the major causes of loss containment for offshore steel pipelines
(Ahammed, 1998).
Corrosion failures represent a significant proportion of the total number of
failures of natural gas pipelines. Early detection of pipe thinning owing to
corrosion in real time before failure occurs will enhance transmission pipeline
reliability (Bullard et al., 2005). For that reason,
mechanical devices and probes have been introduced to the industry so as to
continuously monitor the internal and external condition of the lines under
corrosion.
Even though periodic inspection has been regularly conducted to inspect pipeline condition both internally and externally, pipeline failure due to corrosion attack is still a major problem. Projecting the future growth of defects in order to determine the time to failure of the operational pipelines is not a straightforward task. This is due to inherent uncertainties associated with soil properties, material properties and imperfect measurement by the inspection tool. The operators are intensifying the need for a reliable empirical model especially for external condition. A pipe line damaged by external corrosion has been shown in Fig. 1.
|
| Fig. 1: |
Pipeline damaged by external corrosion |
Research issues: Pipeline corrosion is most prevalent when the breakdown of coatings, inhibitors, or cathodic protection takes place in a corrosive environment. When it comes to pipeline integrity and corrosion, the two most important questions to a pipeline operator are how severe the corrosion is, and how quickly the integrity of the pipeline is deteriorating. The problems are that these factors do not affect the pipeline equally at all locations and corrosion does not grow at the same rate throughout a pipeline. If operator can identify those corrosion defects which are active or growing, then predictions of future corrosion severity for each and every defect can be made.
The existing corrosion mitigation programs are quite effective to combat corrosion
problem. The difficulty is that for various reasons, programs can fail to provide
adequate protection in specific, isolated areas. The reasons may include soil
conditions, cathodic protection shielding and interference, or inadequate inhibitor
concentrations. Any integrity management plan which addresses when to conduct
future repairs and when to re-inspect the pipeline has a tendency to make some
assumption of corrosion rate, whether explicitly or implicitly, due to insufficient
information pertaining to metal loss mechanism. An estimation of corrosion rates
may be applied from engineering judgement based on years of experience and intricate
knowledge of a pipeline, or it may be a sophisticated scientific assessment
incorporating detailed pipeline data (Nesic, 2007).
A proper and accurate corrosion model can greatly assist engineers in making
decisions related to design, operations and control.
Available model: Melchers and Jeffrey (2008)
reported that models for corrosion loss are generally commenced with Tammann
in 1923 who solved the mathematics for the diffusion of oxygen through the tarnish
layers formed on copper, then followed by Booth (1971)
who then refined the mathematics although this usually is simplified to a corrosion
or pit depth c(t) growth law of the form as follows:
where, t is time and A and B are constants.
|
| Fig. 2: |
Corrosion loss-exposure time model showing each of the phases
and the parameters used to define them. |
This model is used in at least one very refined model for pit initiation and early
pit growth (
Englehardt and MacDonald, 1998). It also forms
the basis for much modelling of long term atmospheric corrosion loss.
Melchers
(1997) postulated that the corrosion process changes with time and could be
presented by a number of sequential phases as shown in
Fig. 2.
Melchers and Jeffrey (2008) then calibrated the model
using an extensive re-examination of literature data and field tests on the involvement
of anaerobic bacterial activity of mild steel in marine environments.
As reported by Shi and Mahadevan (2000), a seven-stage
conceptual model was identified by Goswami and Hoeppner
(1995) from the onset of corrosion fatigue fracture. A three-stage probabilistic
model was proposed by Harlow and Wei (1994) to predict
the corrosion fatigue life as well as a probabilistic model for the growth of
corrosion pit induced by constituent particles. The transition model for pitting
to corrosion fatigue crack nucleation has been first proposed by Kondo
(1989) and further discussed by Chen et al. (1996).
Generally, deterministic and probabilistic models of varying degrees of accuracy
are available for various stages of the corrosion fatigue life. The model that
incorporates all the seven stages of corrosion life is proposed by Shi
and Mahadevan (2000) using a comprehensive probabilistic method. The seven
stages consists of pit nucleation, pit growth, transition from pitting to fatigue
crack nucleation, short crack growth, transition from short crack to long crack,
long crack growth and fracture.
To date, empirical model for predicting the growth of external corrosion defects
on buried pipelines is hardly available. Most of the established models such
as the De Waard et al. (1991) equation are meant
for internal growth which governed by the flown product properties and operational
condition. Moreover, the intricacy of physical corrosion models may deter pipeline
operator from using it in their maintenance program. There is an immense need
for new empirical model capable of estimating the growth rate of external defects
under influence of soil properties. For buried gas pipelines, the external corrosion
is a great concern than the internal growth due to severe exposure to various
condition of soil. The internal corrosion for buried gas pipelines is less critical
because the pipelines are used to transport processed gas or cleaned gas with
less impurity. The benefits of the research is to give better understanding
on the actual behaviour of corroding steel under exposure to various soil types
and conditions, knowing the most optimum potential for cathodic protection for
specific site, reduce the uncertainties in the estimation of corrosion growth,
assist the operators in decision making and reduce the overall operating cost.
External corrosion of buried pipelines: Corrosion may act on the pipelines
either internally or externally or both. Furthermore, it may be uniform or nearly
uniform in nature or localised in extent and severity (e.g., pitting or crevice
corrosion) (Ahammed and Melchers, 1997). External corrosion
is a major factor contributing to the deterioration of buried pipelines; it
weakens the pipe wall, which increases the risk of failure (Ahammed
and Melchers, 1997). External corrosion is a function of the interaction
between the pipelines and the soil that surrounds it (Doyle
et al., 2003; Osella et al., 1998).
The external corrosion of gas transmission pipelines is usually controlled by
the application of various polymetric coatings augmented with Cathodic Protection
(CP) (Bullard et al., 2005). Most common pipeline
corrosion protection is done by coating with special material that protects
the surface from corrosive elements, such as the type of the soil, moisture,
pH, temperature variations and other factors such as resistivity and the presence
of sulphate reducing bacteria (Alodan and Abdulaleem, 2007;
Bullard et al., 2005; Doyle
et al., 2003; Rim-Rukeh and Awatefe, 2006).
The common types of corrosion that can occur in a buried pipeline are: (a) pitting corrosion owing to material in-homogeneities, (b) chloride or sulphate induced stress corrosion cracking, (c) corrosion by concentration cells in soil arising out of differences in oxygen concentration in the soil adjacent to the pipe at different regions, (d) microbiologically induced corrosion under anaerobic conditions by sulphate-reducing bacteria (SRB) and Acid-producing Bacteria (APB), (e) tuberculation because of the build up of corrosion products on the internal pipe surfaces and (f) stray current corrosion by earth return direct currents.
Research objective: The objective of the study was to design a new technique/methodology to measure the actual metal loss of buried steel coupon on site. The metal loss information can be utilised to model the external corrosion as experienced by buried gas pipelines exposed to various soil conditions (soil types, properties and contents) or to verify corrosion potential data gained through Electrochemical Impedance Spectroscopy (EIS), for instance. It is of importance to identify the relationship between various soil conditions and the severity of corrosion rate so that the modelling of corrosion dynamic can simulate the actual mechanism as accurate as possible. The pipeline materials will be exposed to various environmental conditions and continuously monitored to determine the corrosion growth rate on site. Apart from modelling, the measured corrosion growth rate can also be utilised in the probabilistic-based assessment of pipeline integrity. Then, the outcomes of the assessment can be used for determination of inspection intervals and other integrity monitoring programmes.
The study focuses on the relationship of soil condition and degree of exposure to the external growth of corrosion defects. The available inspection data and previous maintenance record will be used to determine the exact location for soil sampling based on the severity of corrosion attack. The investigated pipelines throughout the research are made of steel from various grades and sizes.
Soil corrosion factors: Soil has many different meanings depending on
the field of study. To a geotechnical engineer, soil has much broader meaning
and can include not only agronomic material, but also broken-up fragments of
rock, volcanic ash, alluvium, Aeolian sand, glacial material, and any other
residual or transported product of rock weathering (Day, 2001).
Its physical and chemical characteristics are different from sea-bed sediment
soil, salina soil and tideland soil. Soil is a complex material; a porous heterogeneous
and discontinuous environment constituted by mineral or organic solid phase,
water liquid phase and airs other gas phase (Rim-Rukeh and
Awatefe, 2006) (Ferreira et al., 2007). According
to Ferreira et al. (2007), the soil is defined
as an electrolyte and this can be understood by means of electrochemical theory.
The study of the soil as a corrosive environment is necessary due to the large
number of buried pipelines, tank and other structures, as their deterioration
can represent a real economical and environmental problem throughout the years.
There are several parameters that can affect the corrosivity of a soil such
as resistivity, pH, redox potential, moisture content, type of soil, chloride
and others.
In agreement with the parameters cited above, Fitzgerald
(1993) studied shows the corrosivity of the soil influenced by oxygen content,
dissolved salts, pH, elements that form acids, concentration of chloride, sulphide
and sulphate, resistivity, total acidity, redox potential and others, depending
on specific application. The specific test for external corrosion due to soil
corrosivity will be discussed in the next section.
MATERIALS AND METHODS
The research duration is designed for 36 months period divided into several
stages as shown in Fig. 3. The project will commences with
a 6 month of initial data collection, site visits and establishment of research
method. This will be followed by another 6 months of soil sampling, experimental
set-ups, preparation of testing materials and initial experiments. Upon completion
of the activities in Stage 1, corrosion experimental studies and simulation
of predicted corrosion growth rates will be carried out in Stage 2 and followed
by Stage 3 for a period of twelve months each.
|
| Fig. 3: |
Flowchart of field work procedure |
The sequences of research activities are planned as shown in Fig.
3. An overview of test conducted in this study is shown in Fig.
4. Laboratory test may be classified conveniently as follows:
| • |
Soil properties test is to determine the type of soil, with
respect of their parameters comprises moisture content, liquid limit, plastic
limit, plasticity index, shrinkage limit and particle size distribution |
| • |
Soil corrosivity test is conducted to determine the effect
of soil towards buried gas pipelines in study area. The parameter considered
such as pH, temperature, resistivity, redox potential and chloride |
| • |
Weight loss measurement test is performed to determine the
corrosion growth of mild steel coupons that immersed in aqueous solution
of the soil samples |
| • |
Electrochemical Impedance Spectroscopy (EIS) is use to assess
the corrosion behaviour of metals and alloys in service (Bullard
et al., 2003). The most commonly used measurements are as follows: |
| |
• |
Determination of the steady state corrosion potential Ecorr |
| |
• |
Determination of the variation of Ecorr with time |
| |
• |
Determination of the E-i relationship during polarisation
at constant current density (galvanostic) the potential being the variable |
| |
• |
Determination of the E-i relationship during polarisation
at constant potential (potentiostatic) the current being the variable |
Laboratory studies have shown that electrochemistry-based corrosion rate probes
can be used to monitor the corrosion of steel in soils (Shreir,
1976). Corrosion rates were shown to have a good agreement with gravimetric
weight loss measurements and were also sensitive to changes in soil moisture
and salt content. Others have shown that electrical resistivity probes can be
used to monitor the effectiveness of cathodic protection of pipelines (Bullard
et al., 2004; Khan, 2002). According to previous
researchers there are several types of instrument, parameters and methods can
be used in corrosion testing.
Laboratory-based corrosion tests fall into the following categories comprises immersion test, simulated atmosphere tests, electrochemical tests and environmentally aggressive tests (Fig. 4). All of these are accelerated tests by design and therefore must be carefully carried out.
Soil sample collection: Soil sample for this study will be randomly
collected from different sites. At each of the selected site, samples will be
collected by digging a hole of 1 m deep.
|
| Fig. 4: |
Various test used to investigate corrosion mechanism on buried
gas pipelines |
Soil samples will be collected from each sites and kept in polyethylene bags
before sent to the laboratory immediately for further soil analysis.
Soil classification test: The major classification in identifying the
type of soil is the grain size distribution test as determined according to
ASTM D422-98 (American Standard of Material and Testing,
2007) and BS 1377-2:1990 (BSI, 1990a). Both wet sieving
and sedimentation analysis are used to get full particle distribution analysis
of the sample under investigation. The parameters analysed comprises moisture
content, liquid limit, plastic limit, plasticity index, shrinkage limit and
particle size distribution.
Soil corrosivity test: Adopting the BS 1377-3 (BSI,
1990b) and AWWA C105 standards (ANSI/AWWA C105/A21.5-99,
2000) for soil corrosivity testing, the following parameters are analysed
based on the collected soil samples which are pH, temperature, soil type, moisture
content, resistivity, redox potential and chloride (Cl–). Soil
type is evaluated using the following characteristics such as the ability of
water to penetrate soil, the ease with which soil was washed off the equipment
and the consistency of the soil when manipulated in one’s hand.
Moisture content of the soil samples can be ascertained using the weight loss
technique according to BS 1377-2:1990 (BSI, 1990a). An
amount of 30 g of each sample is dried in a drying oven at 105°C for 24
h. The weight difference between the sample before and after evaporation is
regarded as the moisture content. The pH and temperature are normally determined
in-situ using a multi-parameter quality monitor according to ASTM G51-95 (American
Standard of Material and Testing, 2005). Resistivity of the soil samples
is measured using the Wenner Four-Electrode Method adopted in ASTM G57-06 (American
Standard of Material and Testing, 2000b). The soil samples are saturated
with distilled water and placed in different rectangular boxes with an open
top. The boxes are filled to the top with soil. The value of resistivity can
be evaluated using Eq. 2.
Where:
| R |
= |
Resistance (Ω) |
| A |
= |
Cross sectional area of the container perpendicular to the current flow
(cm2 ) |
| a |
= |
Inner electrode spacing (cm) |
| ρ |
= |
Resistivity (Ω.cm) |
Temperature is an important parameter in the investigation of soil corrosivity
because it can modify the interactions between the metal and the soil conditions
(Rim-Rukeh and Awatefe, 2006). Soil resistivity is a
measure of the ability of a soil to conduct an electrical current (Rim-Rukeh
and Awatefe, 2006) (American Standard of Material and
Testing, 2000b). The resistivity of the soil is determined by moisture content
and the concentrations of the different ions and their mobility. The lower the
resistivity of a soil, the better the soil’s electrolytic properties and
the higher is the rate at which the corrosion can proceed (Rim-Rukeh
and Awatefe, 2006).
Reduction-oxidation (Redox) potential refers to the relative potential of an
electrochemical reaction under equilibrium conditions. It is affinity of a medium
for electrons and its electro-negativity compared with hydrogen. A low potential
indicates that the oxygen content of the soil is capable of sustaining Sulphate-reducing
Bacteria (SRB) (Booth, 1971). Redox potential (Eh) of the
soil samples can be measured according to BS 1377-3:1990 (BSI,
1990b). This test is carried out immediately after sampling to avoid the
changes of its characteristics. The platinum electrodes are embedded directly
in soil sample saturated with distilled water and the value of the potential
is recorded after the reading is stable. Chloride (Cl–) content
of the soil samples can also be determined according to BS 1377-3:1990 (BSI,
1990b). The contribution of chloride (Cl–) content to soil
corrosivity is very significant in corrosion process. They not only promote
corrosion because they are conductive by nature, but also inhibit passivity
of the metal surface which protects the metal from corrosion (Uhlig
and Revie, 1985).
Weight loss measurement: Based on the standard, the common weight loss
measurements are performed with mild steel coupons from the same materials that
used in the study area. A minimum of 40 cm3 of soil volume are used
for each 1 cm2 of exposed metal surface area. The test specimen is
then buried in the soil in the corrosion-resistant container for a month or
until the steady state conditions have been reached. The measurements of the
specimens are taken during exposure and after removal according to ASTM G162-99
(American Standard of Material and Testing, 2000a). Table
1 shows an overview of parameter and their testing based on the available
testing standard.
Field work set-up: The main goal of field work test of metal loss is to validate the model developed empirically in the laboratory. Since environmental parameters cannot be manipulated on site, the measurement of weight loss of buried coupon can give an indication of corrosion dynamic rate influenced by intrinsic value of environmental parameters on site. This relationship can be used to validate the level of error of corrosion model developed through series of laboratory testing.
The field work set-up requires hole of about one metre deep to accommodate
two specimens of steel coupon placed at 1 and 0.5 m below the ground surface
as displayed in Fig. 5. The example of layout of holes on
site is represented by Fig. 6 show the digging process took
place on site. The arrangement of coupons in hole is specifically designed to
determine the influence of soil depth upon metal loss rate as well as to increase
the number of specimens per hole.
| Table 1: |
Test for corrosion parameters and soil characteristic based
on the available standard/code |
 |
|
|
| Fig. 5: |
Location of buried coupons |
|
| Fig. 6: |
Locations of coupons |
| Table 2: |
Various types of NDT inspection techniques |
 |
|
Since the experiment is conducted under worst-case scenario, therefore it is
of importance to keep a minimum ground distance of 5 m between the pipeline
route and the hole. This is to deter unnecessary protection upon buried coupon
by stray current sourced from Cathodic Protection system (CP) on the operated
line.
The samples afterwards will be retrieved off the site once every three month for the whole period of two years. The retrieved samples must not be restored since the sample has to be taken back to laboratory for cleaning and metal loss weighing, hence disturbs the original condition of the sample. Therefore, the coupon samples must be prepared under strict quality control to assure the uniformity of the samples in terms of dimension and weight. By doing so, time dependent-metal loss data which is not recorded from the same sample every time weighing process is done can be used to develop a relationship of corrosion dynamic against time with minimum error.
The steel coupons are originated from actual segment of steel pipe with grade ranged from X42-X70. The oxy-cutter is used to cut the pipe segments into small coupons. The edge of pipe segment which severely affected by excessive amount of heat during cutting process is then removed using cold-cutting technique leaving unscathed area for coupon preparation. Coupon sized of 80 cm x 60 cm can then be produced using cold-cutter device. The existing coating must be removed using grinder machine so that the coupon can be tested under worst condition of corrosion attack.
Data sources: Real inspection data of corrosion defects can be acquired from on site periodical inspection activities. The required data can be classified as a secondary data, collected by professional inspection vendor. Table 2 displays the types of NDT tools used to inspect corroding structures and the collected data from these NDT tools may be useful for this research. The inspection data and previous record maintenance kept by the pipeline operators are vital to give an indication on the historical record and current condition of the investigated pipelines. The selection of area from which the soil samples will be collected depends on the maintenance record and inspection data.
On-site measurement: On site measurement is a requirement to systematically record in situ soil parameters during field work and later compared with the secondary results from laboratory soil testing for verification. Mobile apparatuses with high accuracy and minimum limitation are needed to measure water content, salinity, pH, conductivity and resistivity of fresh soil. The measurement of water content on site is crucial. This is due to the fact that the fresh soil used to measure water content must not be exposed to environment more than 24 h prior to the testing. Failure to comply means, the measurement of water content via lab testing is considered not valid. If the soil sample is not feasible to reach laboratory within 24 h, therefore a mobile moisture content apparatus is needed to record the reading. The recorded parameters can then be used to develop empirical relationship between metal loss and soil content.
Expected results: The proposed new techniques in soil-corrosion study for underground pipelines is designed to assist researchers in collecting real-time metal loss data under worst corrosion condition using steel coupon similar to the steel grade of operated line. The data, if properly analysed and combined with laboratory data, can be fully transformed into empirical information for the following benefits;
| • |
Corrosion rates of corroding pipelines for specified materials,
locations and conditions |
| • |
Simplified external corrosion growth prediction model to be
used by industries |
| • |
An overall improved interpretation of inspection data and
prediction methodology in the integrity assessment of corroded pipelines |
| • |
Assist the operators in making accurate decisions on what,
when and where of future inspection, repair and maintenance resources to
be deployed |
| • |
Reduce the overall operating cost of the line |
CONCLUSIONS
A new technique of conducting on site experimental work to measure the real-time metal loss of buried coupon under influence of soil-corrosion factors has been presented. The potential model of external corrosion growth rate with the inclusion of soil factors based on this technique; specifically developed for buried steel pipeline; can be used to evaluate the best corrosion protection scheme to be deployed on the pipelines prior to the installation according to specific site/location. Hence, improving the reliability of the line at a minimal cost and be able to avoid unnecessary inspections in the future. A database of corrosion growth potential for pipeline installation sites can be developed for specific site and soil condition for future reference.
ACKNOWLEDGMENT
The first author is pleased to acknowledge the Ministry of Science, Technology and Innovation, Malaysia (MOSTI) and the Ministry of Higher Education (MOHE) for the support by providing the research funds and scholarship (NSF).
Subscribe Today
