Cases Study For Corrosion in Heat Affected Zone of Carbon Steel
Cases Study For Corrosion in Heat Affected Zone of Carbon Steel
Cases Study For Corrosion in Heat Affected Zone of Carbon Steel
net/publication/285523481
CITATIONS READS
0 2,144
2 authors, including:
Mazen Elmorsy
Gulf of Suez Petroleum Company
1 PUBLICATION 0 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Mazen Elmorsy on 02 December 2015.
ABSTRACT: Corrosion in heat affected zone (HAZ) is considered as one of main reasons
for industrial facilities failures in petroleum oil production fields; where most of failure cases
are referred to severe reduction in wall thickness of area adjust to girth-welds of the industrial
facilities like to pipework systems and pipelines. Corrosion in HAZ of stainless steel has a
common mechanism called sensitization and it depends on depletion of chromium in granules
of stainless steel alloys; while HAZ corrosion in carbon steel can be described by preferential
weld corrosion mechanism. The recent work is an analysis for different practical cases that
have severe reduction in wall thickness of carbon steel piping and pipelines. This analysis is
done to determine the affected length of active corrosion section in carbon steel piping
because this length is more than defined HAZ length.
INTRODUCTION
Heat-affected zone (HAZ) is the adjacent zone to the weld metal. It is the portion of the base
metal that has not been completely melted; but whose microstructure or mechanical properties
have been altered by the heat of welding. Weldment consists of a transition from wrought
base metal through HAZ and into solidified weld metal; it includes five micro-structurally
distinct regions normally identified as the fusion zone, the unmixed region, the partially
melted region, the HAZ and the unaffected base metal. The microstructural features of the
welds has the main reason for activation of corrosion in weldment to be more active than
un-welded pieces that are in the same corrosive environment. Corrosion resistance is
decreased due to heterogeneities and differing characteristics of welded joints that is referred
to the following factors.
The cycle of heating and cooling in the welding process affects the microstructure and surface
composition of welds and adjacent base metal. Also, variation in composition which occurs
due to effects of segregation in micro and macro scales, precipitation of secondary phases,
formation of unmixed zones, recrystallization and grain growth in HAZ, Contamination of the
solidifying weld pool, and dilution at the weld fusion, weld interface and HAZ. From other
point of view, welding procedures, filler metal additions, weldment design and fabrication
technique have a significant effect on activation of corrosion in weldments. Foreign
components including moisture contamination, organic or inorganic chemical species, oxide
film, scale, weld slag and spatter have the same effect. Choice of welding process can
overcome some sides of foreign components effect; as example, using of processes that
provide shielding for molten and hot metal surfaces from reactive gases in the weld
environment.
The width of the heat-affected zone is a function of the heat input. The heat input varies with
different welding processes as well as with different variations of the same process. Rate of
heat input can determine the HAZ width; where slow rate increases the width of heat-affected
zones and provides coarse-grain structure that results poor mechanical properties, especially
in high-strength ferrous alloys.
Peak temperature is a temperature below which the welding heat does not affect the properties
of the base metal. It effects on the estimated value for HAZ width. Within the normal ranges
of heat input, HAZ width increases proportionally with the net heat input. Peak temperature
for the most of plain carbon and low-alloy steels is nearly 1350°F (732°C). At this
temperature, The HAZ width is 0.23in (5.9 mm). The area of activated corrosion, in practical
cases, is more than this value. The recent work is achieved to calculate the length of this
active area through analytical study for some of practical cases.
Variations in microstructure and composition can cause numerous of corrosion mechanisms;
likes to preferential weld corrosion, intergranular corrosion, galvanic corrosion, pitting, stress
corrosion, hydrogen cracking, and microbiologically influenced corrosion 1, 2.
Intergranular corrosion (IGC) is a common mechanism for corrosion in heat affected zone of
stainless steel, it occurs preferentially at or adjacent to the grain boundaries of a metal.
Usually, there will be only a slight attack on the remainder of the grains. IGC can occur
uniformly or, more likely, in localized regions such as welds. In some cases of IGC, metal
loss occurs as the metal grains fall completely out of the sample; however, in other cases, the
grain boundaries are attacked deep below the surface without dropout. In the latter case
although the material appears to be acceptable, the mechanical properties (strength and
ductility) of intergranularly attacked regions are severely degraded, which results in
unexpected failures. Intergranular corrosion can occur by either of three mechanisms that are
precipitation, sensitization and knife line attack. Weld decay is another mechanism, but it
might be considered as special case from sensitization.3,4
Preferential weld corrosion (PWC) is one of these mechanisms that can occur in carbon steel.
It is defined as active corrosion mechanism of C-Mn weldments steels in corrosive
environments; PWC is more active than corrosion of parent material in same environment
conditions. It occurs in weld itself; but it affects also on HAZ. Factors that affect the PWC are
as following:
The weld metal compositions: the change in it generates galvanic cell between weldment and
parent steel. That makes potential difference in certain environments. So, galvanic cell is
formed, leading to corrosion.
The microstructure and hardness: the change in it could decline the corrosion resistant of
weldment. Hardness change refers to microstructure; lower hardness decrease the ability for
PWC in specified cases like to carbon steel in carbon dioxide environment.
The treatment of the base metal: as example, hardened steel in acidic environments is more
active than fully tempered steel.
The composition of filler: minor additions of noble metals (e.g. Ni, Cr, Mo, Cu) to the weld
consumable can convert the weld metal to be cathode with respect to base metal and HAZ. In
addition, decreasing the content of some elements like to silicon can reduce the ability for
PWC. Corrosive environment should be considered because some environments, as sweet
ones, could cause severe preferential attack. Also, Over-alloying should be considered, where
it can cause attack for HAZ.
Welding conditions: as example, post weld heat treatment can reduce the ability for PWC
than as-welded structures.
De-oxidation products: it has complicated effect on PWC, these product are dependent on the
type of used flux. Basic flux causes higher corrosion rates than rutile flux.
Alteration in corrosive medium: addition of some chemicals, like to Biocide, might alter the
environment characteristics that could have effect on PWC nature as this alteration can
increase corrosion rate of PWC.
Operation conditions: temperature, flow rate and the formation of scales have their effect on
PWC. Using of corrosion inhibitors can prevent in PWC, but inhibitor should be tested for
protection of PWC specifically before using. 5, 6
Cases study
Six cases study are analyzed for different pipework systems and one case study is analyzed
for pipeline system. The mentioned seven cases studies are fabricated from carbon steel grade
B alloy; the metal composition of the cited alloy is mentioned in table (1). Operating
conditions for each case are described hereafter.
The thickness measurement results for some of corroded areas in the case study (7) are found
in the following table (table 7). One hundred and sixteen (116) corroded areas are found in the
cited case study; while table (table 7) includes only fifteen (15) areas as example for total
corroded area. Analysis is done for the total corroded areas
Table (8): Arbitrary Classes & sub-classes for distances from weld
Distance from Weld for
Class # Class Name Sub-Class
Sub-class
Adjust to weld (zero
A1
distance)
Closed to A2 0 to ≤ 50 mm
A
Weld
A3 51 to ≤ 100 mm
A4 101 to ≤ 350 mm
Near to B1 351 to ≤ 500 mm
B
Weld B2 501 to ≤ 1000mm
C1 1001 to ≤ 1500mm
Far than
C C2 1501 to ≤ 2000 mm
Weld
C3 More than 2001mm
The objective of the analysis, based on the cited arbitrary sub-classes, is the determination of
the percentage value of number of corroded areas that are located closed to weld. By
calculation of this value, the majority for corrosion in HAZ can be determined. Therefore, the
percentage of number of corroded areas in class “A” will be determined for cases study of
pipework (cases 1 – 6) and pipeline (case 7) individually; because the pipeline case study has
many of corroded areas more than pipework cases study due to comparable length of pipeline
with respect to each pipework case study.
Class “A” is the concerned class in the calculations of majority of corrosion in HAZ because
the corrosion at distance more than 350mm may refer to other reasons that are not related to
welding. The actual length for HAZ is not exactly known because it is depended on many
factors; and it needs to other researches to be determined. While, the recent analysis is
restricted to determine the percentage of corroded areas in the arbitrary class “A”; with
considering that the limit of this class (350mm) is assumed based on practical experience of
corrosion behavior in the oil filed which includes the concerned seven (7) cases study.
0 0 A1 6
5 20 A2 7
A 19
50 80 A3 2
150 300 A4 4
420 500 B1 4
B 6
550 600 B2 2
1210 1380 C1 2
1940 1960 C2 4 C 11
2200 3700 C3 5
The following figure (figure # 1) shows the percentage for number of corroded areas in each
class; here-after figure (figure # 2) shows statistics drawing for sub-classes.
Based on the above cited figures; the following analysis results are gained:
53% of corroded area are located closed to girth weld that prove the majority of
corrosion activation in class A
The largest numbers for corroded areas is in sub-class A1 &A2 that gives indication
for more activation for corrosion in areas that are nearer to girth-weld.
With referring to table (6); the reduction in wall thickness for corroded areas in class
“B” and class “C” is less than reduction for areas in class “A”, where the maximum
reduction in wall thickness of class “A” is 82% while, maximum reduction in classes
“B & C” is 69% and 63% respectively. That gives indication for more of activation for
corrosion in areas which are closer to welds. This observation is also proved by fact
that the average reduction in wall thickness for classes “A, B & C” are 65%, 58% and
57% respectively. The following table (table 10) shows the individual values of
reduction in wall thickness of different corroded areas in each class with ignoring of
other details like to size of defected area and number of case study for each value.
Table (10): Values for reduction in wall thickness percentage of different corroded area
in each class
82 81 76 71 71 70 70 70 65 65 63 61 61 59 58 58 54 54 46
A
% % % % % % % % % % % % % % % % % % %
69 67 57 56 50 48
B
% % % % % %
63 59 59 56 56 56 56 56 56 56 52
C
% % % % % % % % % % %
Figure (1): Percentages for number of corroded areas of each class in cases study (1 – 6)
Figure (2): Statistic drawing for count of each sub-class in cases study (1 – 6)
0 0 A1 0
5 50 A2 13
A 31
87 87 A3 2
102 345 A4 16
360 450 B1 8
B 28
504 951 B2 20
1016 1393 C1 17
1508 1965 C2 16 C 57
2037 4019 C3 24
The following figure (figure # 3) shows the percentage for number of corroded areas in each
class.
Figure (3): Percentages for number of corroded areas of each class in case study (7)
Based on the above shown figure; the following analysis results are gained:
In case of pipeline; the corroded areas that are closed to weld (in class A) has only
27% of total corroded areas. That refers to very long length of the pipeline and passing
of the pipeline through different earth levels, which might cause existing of some
portions with low levels that allow accumulate of associated water and increase of
corrosion rate in these areas.
Although that only 27% of corroded areas in class “A”; the most of the highest
reductions in wall thickness are found in the corroded areas of the cited class. By
referring to table (7); three areas with the highest reduction (70%) in wall thickness
are in this class and these three areas are out of the existing four highest reduction
corroded areas. That gives indication for the phenomena that the most active corrosion
rate areas mostly exist in class A.
By more analysis for the corroded areas in class A; the thirty-one (31) corroded areas in the
cited class can be divided into groups, each group has region for distance from the nearest
girth weld equal to twenty (20) mm. According to this dividing criterion, the following table
(table 12) for normalized frequency of corroded areas in each distance region can be obtained;
with considering that the regions with zero frequency are excluded, and these distances are for
different welds meaning that these distances are not from one weld. Based on the cited table;
the here-after statistic drawing (figure 4) is obtained.
Table (12): Normalized frequency for corroded areas of class A in 20mm distance
regions
Distance
Region 20 40 60 100 120 140 200 220 280 300 320 340 350
(mm)
Frequency
13% 23% 6% 6% 10% 6% 3% 6% 10% 3% 6% 3% 3%
%
Based on (figure 4); the highest normalized frequency is nearer to weld and by move away
from the weld, the frequency is decreased considering the cited drawing is not for one weld;
but corroded areas might be at different or same weld. That is clarified by the slope of the
dashed trend-line on the cited drawing. This analyzed result proves that the areas that are
adjusted to weld have more activation for corrosion and this activation is decreased by move
away from the weld. This result is also proved by another point of view; where, reduction in
wall thickness percentage is also decreased by move away from the girth weld. That is shown
in the following statistic drawing (figure 5) for reduction in wall thickness percentage against
distance from weld for corroded areas in class “A”.
15%
10%
5%
0%
0 50 100 150 200 250 300 350
Figure (4): Statistic Drawing for Normalized Frequency of corroded areas in class A
Figure (5): Statistic Drawing for reduction in wall thickness of corroded areas in class A
against distance
CONCLUSIONS
The length of active corrosion zone is found to be in range less than 350mm; more active
length is less than 50mm. For long pipelines, the majority of corroded areas in the cited
length are decreasing because of existing of other factors to activate corrosion as existing of
low levels areas that allows accumulation of associated water and increase corrosion rate.
Although, the corroded areas that are far than weld are less active than area of length 50mm.
REFERENCES