CHAPITRE 03: WELDING IN SERVICE:

INTRODUCTION:

Pipeline operators are often faced with opportunities to repair corrosion

damage on pipelines, which is second only to mechanical damage as the
primary cause of natural gas pipeline failures in the U.S.(Ref' 0 To prevent an
area of corrosion damage from causing a pipeline to rupture, the area
containing the corrosion damage must be rein forced to prevent the pipeline
from bulging. The most predominant method of rein forcing corrosion damage
in cross-country pipelines is to install a full encirclement repair sleeve. The
applications of this and other repair methods are described in detail elsewhere.
<Ref 2) there are two basic types of full-encirclement sleeves; Type A and Type
B (Figure 1). Type a sleeves are for repair of external corrosion only. While not
necessary to prevent bulging, the ends of Type A sleeves are often welded to
the pipeline to prevent further corrosion. Type B sleeves are for repair of leaking
defects or defects that will eventually leak (e.g., internal corrosion). Therefore,
the ends of Type B sleeves must be welded to the pipeline to contain the
pressure. There are often significant incentives for avoiding interruption of
pipeline service; therefore, this welding is often performed "in-service». There
are two primary concerns with welding onto in service pipelines, whether for
installing repair sleeves or installing a branch connection prior to "hot tapping".
The first is for "burn through", where the welding arc causes the pipe wall to be
penetrated. The second concern is for hydrogen cracking, sincewelds made in-
service cool at an accelerated rate as the result of the flowing contents' ability
to remove heat from the pipe wall.

An alternative method of repair for pipeline corrosion is deposited weld metal,

or weld deposition repair (Figure 2).This method of in-service repair is attractive
because it is simple and direct, and because it can be applied where the use of a
full-encirclement sleeve is impossible, such as for the repair of fittings and field
bends. In addition to the concerns for burn through and hydrogen cracking, the
rein concern for ensuring that the deposited weld metal has adequately
restored both the static and fatigue strength of the pipeline. The results of a
recent project at Edison Welding Institute (EWI) indicate that it is feasible to
safely carry out weld deposition repair on corroded areas with a remaining wall
thickness as thin as 0.125 in (3.2 mm), provided that special precautions are
taken. (Ref 3)Properly made repairs we resh own to have the ability to restore
the strength of the pipeline and be resistant to pressure cycles.

-Type B Sleeve

-Type A Sleeve
 Figure 1. Type A and Type B sleeve

Figure2. Exemple of weld deposition repair

DEFINTION:
Sleeve-repair welding is a widely used technique for in-service repair of
defective pipelines. In this technique, two half-sleeves are assembled around
the damaged pipe section with a longitudinal but weld and then a
circumferential fillet weld is used to attach the sleeve to the pipe as illustrated
in Fig. 1. While the longitudinal but weld has no effect on the pipe, the
circumferential fillet weld is expected to induce thermal stresses and high pipe-
wall temperature as a result of the welding process. There are three main
concerns associated with the in-service sleeve repair welding: (a) the possibility
of pipe-wall melt through due to localized heating during welding [1]; (b)
formation of marten site in the heat affected zone (HAZ) due to the high cooling
rates at the weld zone resulting from the high speed flowing gas [2]; (c) residual
stresses in the pipe wall resulting from fillet welds may cause crack opening and
growth as well as enhancement of hydrogen cracking [3,4]. This technique can
modify the structural response of the pipeline, increasing its stiffness and
initiating zones of high stresses at the ends of the sleeve. Hence, stress
distributions in the area of the sleeve-to-pipe fillet welds can be identified as a
critical point of the repair [5].

Over the last two decades, a number of finite element simulations of sleeve
repair welding have been conducted to investigate the transient thermal fields,
cooling rates, stress fields, and prediction Of melt-through using thermo-
mechanical elastic-plastic models [5e16]. Bang et al. [6] developed an ax
symmetric finite element (FE) model for in-service multi-pass sleeve repair
welding of gas pipes and predicted transient temperature distributions, heat
affected zone (HAZ) size, and residual stress distributions. Cisilinoet al. [7]
employed 3D FE modeling to determine the minimum

Weld able wall thickness under different operating pressures and was found to
increase with gas flow rate. Teague et al. [9] studied experimentally and
numerically the effects of adjacent repair sleeves on the structural integrity of
gas pipes using a set of ax symmetric FE models. Sabapathy et al. [10, 11]
developed a numerical procedure to predict burn-through failure during in-
service welding of gas pipe sand was able to predict bursting pressure.

Wahabetal. [12] Attempted to enhance the prediction accuracy of temperature

fields, using empirical relationships involving welding process inputs, weld bead
size and shape, and the modeling of heat source coordinates. A direct
simulation of pipe bursting during welding.
Fig. 1. Schematic illustration of sleeve repair welding of a defective pipe.

Repair models and development:

In the present work, different models have been developed to study the effects
of the repair procedure adopted during in-service sleeve repair welding, and its
influence on resulting residual stresses is monitored. The first model is
concerned with the effect of welding sequences on residual stresses. Two
welding sequences are compared. The first is the sequential welding procedure
(3DSEQ)where one side of the sleeve is welded initially and then the other side
is subsequently welded by the same welder (see Fig. 3a).The second is the
simultaneous welding procedure (3D-SIM) where both sides of the sleeve are
welded concurrently by two welders(see Fig. 3b). Both procedures were
simulated using 3D FE models. For 3D-SIM model, only half of the complete
model is required due to the symmetry in loading, boundary conditions, and
geometry around the sleeve/pipe mid length plane. In addition, a
2Daxisymmetricmodel (2D-SIM) is developed, whose results are used for
comparison with 3Dmodels. Table 2 summarizes the characteristics of the
developed FE models, including model acronyms, type of analysis, element
type, and numbers of elements and nodes. The time consumed during the in-
service sleeve repair is very important as the welding process maybe conducted
under severe weather conditions. Therefore, the influence of the number of
welders, their positions and welding directions on induced residual stresses and
corresponding distortion are of concern. Therefore, the second model concerns
the effect of welding schemes (i.e., welding direction and number of welders)
on residual stresses and is achieved by modeling four different welding
schemes, as shown in Fig. 4. A single welder is needed to weld the repair sleeve
in welding Schemes # 1 and 4, while two welders are required in Schemes #
2and 3. The 3D-SEQ model is used in the welding scheme investigation.

Loads and boundary conditions:

Both thermal loads due to welding process and structural loads due to the
internal pressure of the flowing gas are considered in the FE models. The
thermal load due to fusion is represented by a volumetric heat source at the
weld bead location, while the internal pressure is reflected in the induced hoop
and axial stresses in the pipe wall. The axial stress in the pipe Wall due to the
internal gas pressure is represented by a uniform axial tensile stress at. The end
of the pipe wall. The analysis consists of three steps to apply both structural
and thermal loads.
Fig. 4. Welding schemes considered in the present work. Fig. 3. Boundary
conditions and Welding sequence;

(a) sequential Welding (3D-SEQ) and simultaneous

(b) Welding (3D-SIM).

The first step is a pre-weld loading step simulating the application of internal
pressure along with the corresponding axial stress. The second step simulates
the deposition of the weld material and continues until cooling to room
temperature. The third step involves unloading the pipe in order to remove
stresses induced by internal pressure, to determine the remaining residual
stresses under no pressure. These steps are illustrated in Fig. 5a. The deposition
of the filler material during the welding process is modeled using element birth
technique [22]. At the beginning of the analysis, all elements representing the
welding passes are deactivated by adjusting their thermal and structural
properties to zero. During the analysis, at the beginning of each welding step,
the deactivated elements under the moving weld torch are reactivated to
represent the molten droplets, as shown in Fig. 5bed. The heat input is modeled
as a body heat flux distributed uniformly throughout the molten electrode. The
net heat input from the welding arc to the pipe is calculated as:

Table 1: Welding parameters.

Filler Voltage(V) Current(Amp) Welding Heat

metals speed(mm/min) input
(kJ/mm)
Type Dia.
(mm
)
E8010 5.0 29 195 170 1.5
Fig. 5. Steps used in the FE simulation of the Sleeve repair welding process
steps: (a) loading and boundary conditions, and (bed) moving heat source
acting on the reactivated elements at different welding steps.

Q ¼hVvoltageIcurrentðWÞ (1)

Where, Q, h, V voltage, and I current are net heat input, arcefficiency, arc
voltage and current, respectively [23]. The heat of the moltendropletsis
modeled using a volumetric body heat source and is calculated as:
q¼Q, Vpass.W=m3Σ(2)

Where q and Vpass are heat flux and volume of the weld pass, respectively
[24, 25].Natural and forced convection boundary conditions are assumed on
the outer and inner surfaces of the model, respectively. The natural
convection coefficient is taken as 28 W/m.K while the forced convection
coefficients calculated according to Ref. [6].Analyses were performed using
a thermo-mechanical coupled temperature-displacement procedure in
ABAQUS™ version 6.11e2 commercial finite element package. Temperature
dependent thermo-mechanical properties of X65 [6]are adopted in the pre-
sent work for temperatures ranging from room temperature up
to1500○C.Thematerialmodelsadoptedforpipe,sleeveandweldmetal are
assumed to gradually change from elastic-plastic with linear hardening at
room temperature to perfectly plastic at elevated temperatures. Although
the more general coupled temperature-stress analysis termisadopted here,
the numerical procedure utilized here is in effect a sequentially coupled one,
 since the stress field solutions are dependent on the temperature field while
there is no inverse dependency. Effectiveness of sequentially coupled
thermo mechanical simulation of residual stress in arc welding has been
demonstrated in FE-welding simulation litera- true [26, 27].

CONCLUSION:

Pipelines carrying gases and fluids often contain defects that can lead to their
rupture. These defects are classified into five major types: pitting and corrosion
craters, cracks, notches, indentations and so-called combined faults. The
difference in geometry and nature of the defect sledre searchers to create
specifictools for each type of defect based on limit analysis, linear and nonlinear
fracture mechanics, as well as finite element analysis and experimentation.

One of the many challenges operators face each day is keeping their
pipelines safe from the public. It is important that operators are aware of
the methods available for repairing mechanically damaged pipe and have a
plan in place for when they encounter it. Composite sleeves have been used
to constrain the dent.
Composite sleeves are proven to provide almost two orders of magnitude
improvement in fatigue life compared to grinding alone and almost three
for unrepaired pipe. The strain, as the dent tries to flex, is transferred
through the incompressible material into the composite repair.
It is also important that operator’s fully assess the composite they want to
utilize to repair their pipelines, as not all composite repairs are the same.
Any company offering a composite repair system for mechanical damage
should be able to back up its claims of repair suitability with valid and
independent testing.