Tuev Nord Concept Loop: Lifetime Optimisation of Pipelines

1 Copyright © 2014 by ASME

TUEV NORD CONCEPT LOOP -

LIFETIME OPTIMISATION OF PIPELINES

Ralf Trieglaff TÜV NORD SysTec GmbH & Co. KG

Hamburg, Germany Email: [email protected]

Christian Schrandt TÜV NORD SysTec GmbH & Co. KG

Hamburg, Germany Email: [email protected]

Axel Schulz

TÜV NORD SysTec GmbH & Co. KG Hamburg, Germany

Email: [email protected]

Mayk Schulz IGN GmbH & Co. KG Greifswald, Germany

Email: [email protected]

ABSTRACT

LOOP is a concept to evaluate corroded or damaged

pipelines based on detailed data from UT-pigging. The

procedure of LOOP delivers a 3D-model generated from the

data of a commercial in-line inspection tools (ultrasonic,

magnetic flux). This makes it possible to use the full

functionalities of the relevant finite element software like

evaluation of wall-thinning (LOOP 1) and fracture mechanics

analysis to evaluate cracks in the wall (LOOP 2). In this paper is

given the basic ideas of the LOOP concept, where the main

focus is directed to the LOOP 1 assessment procedure.

Based on a real example of a corroded pipeline is

demonstrated the assessment procedure, which is based on an

elastic-plastic analysis of a real inner contour of the corroded

surface transferred in the finite element geometry model. The

unique element is that the surface data of the UT-pigging is used

directly to generate the geometry model in the FE-software

ANSYS. The assessment procedure is validated by a burst

pressure test of a corroded pipeline. The result of the burst

pressure test is compared with the calculated limit load from an

elastic-plastic analysis based on measured material properties.

Additionally, the assessment procedure is compared with the

results of a limit load analysis based on DIN EN 13445-3 and

with the results of the standard assessment procedure. At the end

the assessment procedure is compared with the procedure given

in API 579-1 standard.

NOMENCLATURE

A Elongation at fracture

Da Outer Diameter

E Elastic modulus

KB Stress characteristic number

KBmax Maximum permissible stress characteristic number

KFL Fault characteristic number

Lf Length of the damage

pi Internal pressure

pmax Maximum allowable internal pressure

RMd Fictive flow stress

Rp Yield strength

Rm Tensile strength

Tf Depth of the damage

sn Nominall wall thickness

sf Minimum wall thickness

v Poisson ratio

σn Nominal stress

Proceedings of the ASME 2014 Pressure Vessels & Piping Conference PVP2014

July 20-24, 2014, Anaheim, California, USA

PVP2014-28755

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 11/25/2014 Terms of Use: http://asme.org/terms

mailto:[email protected]




Copyright © 2014 by ASME 2

INTRODUCTION

A pipeline is exposed to a number of operational and

environmental conditions during its life cycle. It can experience

various types of damage, such as corrosion on the interior and

exterior surfaces caused by transport medium or a envelope.

Further can occur geometric abnormalities like buckling due to

external influences or cracks due to manufacturing and/or cyclic

loadings (see Figure 1). To determine the state of the pipeline

UT pigging is periodically used for inspection. Then it is

necessary to analyze the data obtained and to evaluate the

findings. The time at which repairs are to be made is a central

question.

With the LOOP-concept the service life of damaged

pipelines can be calculated in advance. LOOP is based on FE-

methods and methods of fracture mechanics, and is particularly

well-suited for lines of various dimensions subject to pressure

and temperature stresses, such as those for natural gas pipelines,

oil pipelines and product pipelines.

In this report, the use of LOOP for determining the

maximum limit load of a defect is presented on the basis of a

real failure in a practical application. In addition, a comparison

of the analysis results with the results of an evaluation method

frequently used in Germany is performed, and a validation of the

method on the basis of a burst pressure test is presented.

Appropriate reserves are indicated in the evaluations and the

consistency of the proposed concept with the requirements of the

API 579-1 standard is presented.

Fig. 1: Possible damage to a pipeline in the life cycle.

PRESENTATION OF THE TÜV NORD LOOP® CONCEPT

In order to reduce the scope of reconstruction of pipelines

and to be able to better estimate the service life, a more precise

evaluation of failure can be performed with LOOP as a

supplement to the standard procedure.

For this purpose, special procedures and calculation methods

based on FE- analysis are used. LOOP consists of two modules

(see Fig. 2), which evaluate reduction in wall thickness

according to an innovative procedure include an evaluation of

crack-like defects with Module 2. With these modules, a detailed

evaluation of nearly all defects occurring in a pipeline is made

possible in order to assure integrity and to make more accurate

service life predictions.

LOOP 1 LOOP 2

Evaluation of reductions in

wall thickness with FEM

Evaluation of cracks caused

by the manufacturing

process and operational

conditions

3D modeling

Modeling of the failure on

the basis of the measurement

data from the UT pigging

+

Plastic collapse analysis

Determination of the safety

against plastic collapse in

the area of the failure

+

Fatigue estimation

Determination of the fatigue

strength under the operating

conditions to be expected

Static verification

Fracture-mechanical

evaluation of cracks according

to API 579-1/ASME FFS-1

+

Crack growth

Calculation of crack growth

under the operating conditions

to be expected

Fig. 2: Structure of the LOOP concept

STANDARD EVALUATION OF A FAILURE

The procedure and the results from the use of the LOOP

concept (LOOP 1) are presented below on the basis of an

example from a practical application.

Large-scale material losses over greater lengths were

detected during the inspection of a product pipeline by means of

an ultrasonic test pigging. In the evaluation of the pigging data it

was determined that, on an abnormality on the interior side,

there was a maximum material loss of 3.4 mm with a measured

wall thickness of 7.2 mm. The abnormality extended on the

interior pipe wall in a range between 300° and 360° with a

maximum length of 2052 mm and a maximum width of 231

mm. This abnormality was classified as critical in the scope of

the preliminary evaluation and had to be evaluated more closely

in regard to its safety against plastic collapse.

For visualization of the described abnormality, the damage

pattern of a similar abnormality from the same pipeline section

is shown in Figure 3.



Fig. 3: Presentation of a similar failure from the same

pipeline section

Specifications for the failure are listed in Table 1. The

specifications were taken from the ultrasonic scan (UT scan)

shown in Figure 4.

Tab.1: Specifications for the failure from pigging data

Abnormality specifications

Value

Inside/outside location: Inside (material loss)

Length [mm]: 2 052

Width [mm]: 231

Depth [mm]: 3.4

Remaining wall [mm]: 3.8

Wall thickness [mm]: 7.2

Circumferential position start

[degrees]: 300 °

Circumferential position end

[degrees]: 360 °

Fig. 4: UT scan of the analyzed abnormality

Additional important specifications for the inspected

pipeline, which were partly used for the calculation, are listed in

Table 2 below.

Tab.2: Dimensions and operating parameters of the pipeline

Line data

Value

Product conveyed Liquid

Pipe Welded longitudinal seam

Outside diameter 457.0 mm

Wall thickness 7.2 mm

Material St 53.7

Minimum yield strength 360 N/mm²

Minimum tensile strength 510 N/mm²

Design pressure 46.5 bar

Max. operating pressure 42.0 bar

Pigging test UT pigging

STANDARD EVALUATION OF THE FAILURE

For the standard evaluation of critical failures in pipelines,

the evaluation process according to Mackstein and Schmidt [1]

is frequently utilized in Germany. This process is largely

empirical and is based on the results of burst and swelling

pressure tests. Although this process is very simple in use, it

provides very conservative evaluation results, as is shown

below.

Tab.3: Boundary conditions for the calculation

Tensile strength Rm = 510 MPa

Outside diameter of the pipeline Da = 457 mm

Nominal wall thickness sn = 7.2 mm

Minimum wall thickness sf = 3.8 mm

Internal pressure pi = 4.2 MPa

Depth of the damage

(maximum expansion)

Tf = 3.4 mm

Length of the damage Lf = 2 052 mm

For the evaluation of the aforementioned failure, a so-called

fault characteristic number (KFL) and stress characteristic

number (KB) are calculated with this process by means of the

specified formulae. The combination of the two characteristics

provides an evaluation point in the diagram (Fig. 5) with which

a statement about the reliability of the defect can be made

immediately. In the Following the formulas to calculate the fault

characteristic number and stress characteristic number is given

below.

n

pi Da sn 2

(1)

KBn

Rm

(2)



f

pi Da sf 2

(3)

KFLf

n

1

(4)

If the calculations are now performed with the indicated

formulae (1) – (4), the results are a value of 0.26 for the stress

characteristic number KB and a value of 0.89 for the fault

characteristic number KFL. When entered into the evaluation

diagram (Fig. 5), it can be determined that this failure is close to

the burst pressure safety line 1.8 and can thus be classified as

critical.

Fig. 5: Loading capacity of pipes with longitudinally-

oriented crack-like defects with static internal pressure (3R

International 34 Mackenstein/Schmidt) [1]

For the calculation of the maximum allowable internal

pressure according to this process, a maximum allowable stress

characteristic number KBmax is taken from the diagram on the

basis of the fault characteristic number KFL. The stress

characteristic number KBmax is arrived at with a consistent fault

characteristic number KFL and on reaching the fracture safety

line 1.8.

pmax

KBmaxsn 2 Rm

Da

(5)

With (5) a maximum allowable internal pressure pmax of 46.6

bar can now be determined. In comparison to the maximum

operating pressure of 42 bar, a safety of only 4.6 bar is provided,

and therefore a decision was made to repair the damaged area of

the pipeline.

EVALUATION OF THE FAILURE WITH LOOP 1

The failure described above is evaluated in the following by

means of the LOOP 1 process. In this connection, an FE model

is created in an initial step by means of realistic fault geometry

based on the measurement data of the UT pigging. Expanding on

this FE model, an analysis of the maximum collapse loads of the

failure is performed by means of two different processes. Then a

fatigue analysis is performed for an estimation of remaining

service life.

3D-MODELLING OF THE FAILURE

In order to perform an analysis with as much detail as

possible, it is essential that the damaged area is factored in as

realistically as possible. This is achieved by means of a realistic

3D modeling of the defect geometry. However, for an exact

modeling, the pipeline must have been inspected by means of

ultrasonic measurement to determine the weakening of the wall

thickness.

The benefits of the UT process are that the wall thickness of

the entire pipeline is recorded in defined measuring grids (e.g.

1.5 mm in the axial direction and 8 mm in the radial direction)

and a highly precise measurement can be achieved. However, an

essential requirement for the modeling is that all the

measurement data can be output in one file. An appropriate

interface for the transfer of the measurement data must be

agreed upon with the pigging company in advance. The

sequence for the model generation is schematically represented

in Fig. 6.

Fig. 6: Sequence for the automatic model generation

However, before the data is processed it must first be filtered

and prepared. Implausible measurements and corresponding

outliers which would falsify the result are thereby eliminated.

Since it is a relatively large defect in this example, there is also a

correspondingly large amount of data available. Around 66,000

data sets (individual measurements) must be processed for the

conversion. For the conversion of the measurements into a

usable 3D model, special software has been developed which

saves a great deal of time and essentially provides a better model

in comparison to previous manual model creation. The generated

3D model, see Fig 7, now serves as a basis for an optimized and

realistic calculation of the maximum capacity for this damage

area, as well as a fatigue analysis.

Fig. 7: 3D model of the failure (left), as well as a detailed cut-

out of 16 mm (right)



DETERMINATION OF THE COLLAPSE LOAD

For the evaluation of the static safety of the fault in

consideration, two different approaches are addressed below.

Process I is oriented towards the processes frequently used in

Germany, wherein the fault is evaluated on the basis of the

local/global stress analysis according to German AD-Standard

S4, German standard DIN 2413 as well as scrutinizing literature

[2]. Process II is based on the loading capacity analysis

mentioned in DIN EN 13445-3 Annex B [3]. In the Following

we present at first the evaluation process I and subsequently the

evaluation process II.

In order to be able to evaluate a reduction in wall thickness

as permissible, all the following conditions must be met for the

first process:

a. The maximum comparison stress at nominal pressure is

less than the tensile strength of the material.

b. At nominal pressure, only a locally limited plastic

deformation should occur in the area of a magnitude of

the wall thickness.

c. The safety factor against plastic collapse must be at least

1.8.

FEM-code ANSYS [4] is used for the calculations and the

following procedure and boundary conditions are considered for

Process I:

FE model with real abnormality geometry

Elastic-plastic material behaviour (multi-linear material

law)

Geometrically non-linear

In addition, the following material characteristics were taken

as a basis for the creation of the material model. The minimum

specifications indicated in accordance to the material standard

were applied.

Tab. 4: Mechanical-technological specifications for St 53.7

Material St 53.7

Standard values

Yield strength [Rp]: 360 MPa

Tensile strength [Rm]: 510 MPa

Elongation at fracture [A]: 20.0 %

E-module [E]: 200 000 MPa

Poisson ratio [v]: 0.3

Fig. 8: Applied material model for the elastic-plastic analysis

according to Process I

In order to determine the plastic collapse load of the real

modeled failure, in the scope of the calculation the internal

pressure is increased until the solver of the program no longer

achieves any convergence due to the occurring plastic strain and

the calculation is discontinued. A representation of the complete

model can be seen in Fig. 9.

Fig. 9: Representation of the complete model and result of

the elastic-plastic analysis according to Process I

A maximum pressure of 117 bar was determined based on

this process. If an operating pressure of 42 bar is set, there is a

safety of 2.8 against plastic collaps. If a collaps safety factor of

1.8 would be used as a basis for this fault, a permissible

operating pressure of 65 bar would arise through conversion of

the formula (6). Therefore the criterion c) is fulfilled.

(6)

In addition, the plastic deformation must be considered for

the fault area, which should only occur to a locally limited

extent at nominal pressure within the fault area. For this

purpose, the course of the plastic expansion over the pressure is

shown in Fig. 10 for the range of 0 – 2 %. It is recognizable that

no noteworthy expansions occur within the fault at an operating

pressure of 42, in which case the criterion b) would be fulfilled.

For visualization, the 0.2% expansion limit is entered, which

represents the limit between the elastic and plastic deformation.



Fig. 10: Result of the elastic-plastic analysis according to

Procedure I, course of the plastic expansion from 0 – 2 %

above the internal pressure

In order evaluate the last criterion, a linear-elastic analysis

was performed (Fig. 11). In this figure is shown the equivalent

stress distribution. They must lie below the tensile strength of

the material at operating temperature, in which case the last

criterion is also fulfilled.

Fig. 11: Representation of comparison stresses at 42 bar

internal pressure

The analysis has shown that comparison stresses of a

maximum of 460 N/mm² occur in the area of the failure at a

nominal pressure of 42 bar. Since this is well below the

indicated tensile force of 510 N/mm², the criterion a) is also

fulfilled and the verification of the static safety of the fault

according to Procedure I is provided.

For comparative purposes, a limit load analysis in

accordance with the process given in DIN EN 13445-3 (Annex

B) was performed as Procedure II. In comparison with the

preceding analysis (Procedure I), the following boundary

conditions were considered for the limit load analysis according

to DIN EN:

Linear elastic ideal plastic material law (without

hardening)

Tresca flow criterion (main stress hypothesis)

Fictive flow stress RMd

Maximum plastic expansion limited to 5 %

Partial safety factor of 1.2 on the stress

For the limit load analysis, with a yield strength Rp of 360

MPa, a fictive flow stress RMd of 249.4 MPa arises. This lies

well below the yield strength of the material and is calculated

according to the formula (7).

(7)

In order to determine the plastic collapse load of the real

modeled failure, in the scope of the calculation the internal

pressure is increased until the solver of the program no longer

achieves any convergence due to the occurring plastic

expansions and the calculation is discontinued. In the result of

this analysis, the last convergent load step was reached at 69.5

bar. Since the plastic expansion must be limited to 5 %, a

maximum ultimate pressure of 68 bar and a allowable pressure

of 54 bar arise.

If Procedure I and II are then compared with each other, it

can be seen that a higher maximum allowable operating pressure

is possible than calculated with the standard evaluation. The

static safety of the failure geometry could therefore be verified.

For the Procedure I a considerable reserves of approx. 20 bar

was estimated and for process II approx. 10 bar with Procedure

II could be discovered in comparison with the standard

evaluation.

VALIDATION OF THE CALCULATION RESULTS

The validation of the calculation procedures was performed

by means of a comparison of the calculation results with an

experimentally determined burst pressure. For this purpose,

however, an additional elastic-plastic FE-analysis had to be

performed first. For the calculation and the comparison, it is

essential that true material behavior is taken into consideration

in the FE simulation. In order to determine the influence of the

material, the mechanical-technological characteristics of the

material used here were determined by means of tensile tests.

The tensile tests which were performed in the process provided

the material characteristics mentioned in Tab. 5 below. The

minimum characteristics according to standard are likewise

indicated for the comparison.



Tab.5: Mechanical-technological characteristics

Material St 53.7

Values determined

from tensile tests

Standard

values

Yield strength [Rp]: 381 MPa 360 MPa

Tensile strength [Rm]: 558 MPa 510 MPa

Elongation at fracture

[A]:

34.3 % 20.0 %

E-module [E]: 200 000 MPa 200 000 MPa

Poisson ratio [v]: 0.3 0.3

On the basis of the determined stress-strain curve from the

tensile test, a material-specific model was then derived for the

calculation of the fracture pressure; see Fig. 12, Measurements

(red curve). The following boundary conditions were taken into

consideration in the calculation with the FEM code ANSYS.

FEM model with real abnormality geometry

Elastic-plastic material behaviour with hardening

Large deformations are permitted

Convergence of the material curve to the material data

from tensile tests

Fig. 12: Derived material model for the analysis of the plastic

collapse load (violet line)

In the result of the elastic-plastic analysis, a fracture pressure

of 132 bar was determined (see Fig. 13).

Fig. 13: Result of the elastic-plastic analysis for

determination of the plastic collapse load, course of the

plastic strain above the internal pressure

In order to investigate the maximum pressure at which the

damaged pipeline actually fails, the removed pipeline section

with a length of approx. 8000 mm was subjected to a pressure

test with water.

The failure of the pipeline, see Fig. 14 to the right, occurred

at an internal pressure of 136 bar.

Plastic collapse load:

132 bar

Burst pressure:

136 bar

Fig. 14: Comparison of the results of the FE-analysis with

the burst pressure test

In a comparative ratio of the experimentally determined

burst pressure to the analyzed plastic collapse pressure, the

result for this validation example is a deviation of 3 %.

FATIGUE ANALYSIS

For lines which are exposed to a pulsating stress, as is

frequently the case with lines carrying liquids, a fatigue analysis

must also be performed in order to evaluate the damage from

cyclical loads. On the basis of a fatigue analysis, the service life

evaluation takes place based on the material fatigue with respect

to the crack formation. For an estimation of the service life,

however, the following boundary conditions must be taken into

consideration:

the load of the pipeline in the past as well as in the

future

growth of the fault in the past as well as in the future



The stresses also change on the basis of the diminishing

residual wall thickness due to the growth of a fault with

increasing age. For an estimation of the service life, this change

over time should be taken into consideration. This can take place

by means of further calculations, in which, for example, the

erosion due to corrosion in annual stages can be accounted for.

In this example the influences from the changing fault were not

factored (in conservatively).

The cyclical load of the pipeline takes place with a changing

internal pressure. This internal pressure fatigue stress results

from the following loading conditions in this case.

Tab. 6: Loading condition

Loading condition Range

[bar]

Cycle

[/year]

Start and shut down 0 - 42 2

In service fluctuation 42 – 26 – 42 730

With a starting process, the pressure fluctuates in the positive

range between 0 and 42 bar, and/or 0 and y bar with an output

adjustment (pressure fatigue stress, 0<R<1).

For these load conditions, the significant stresses

(circumferential stresses) are calculated in the scope of a linear-

elastic analysis.

Fig. 15: Result of the elastic analysis at 42 bar internal

pressure, representation of the comparison stresses

On the basis of the determined stresses, we have evaluated

this in accordance with the notch stress concept of DIN EN

13445 [3] and determined a permissible number of load cycles

of 300 000. When this number is compared with the actually

occurring and predicted cycles, consideration of the effect of the

changing abnormality becomes superfluous in this case. In

principle, the result of the fatigue analysis can still be assured by

a prediction of the progression of corrosion to be expected. In

the present calculation example, we have dispensed with this.

CLASSIFICATION OF LOOP 1 IN THE EVALUATION PROCESS OF API 579-1 STANDARD

According to API 579 / ASME FFs-1 [5], there is a different

level for the evaluation of abnormalities. In this connection,

Level I represents the lowest evaluation level at which a very

conservative initial evaluation is based on very simplified

assumptions.

Level II, on the other hand, is a detailed evaluation with less

conservativeness and improved results in comparison to level I.

Level III, in turn, is the highest evaluation level. In API 579-

1, 4.4.4 [5], it is shown that an evaluation according to Level III

is usually based of finite element analyses, wherein non-linear

analyses are given priority for determining the ultimate bearing

capacity in accordance with Annex B1. In addition, explicit

reference is made to a direct transfer of the measured wall

thickness profiles to the FE geometry. Therefore, the process

presented here consistently implements the Level III evaluation

procedure. The procedures presented in Annex B1 of API 579-

1[5] for the evaluation of the safety in regard to plastic collapse

in the form of a limit load analysis method and an elastic-plastic

stress analysis method correspond to the procedures I and II

presented in chapter 5.2 for determining the maximum capacity.

Only an adjustment of the safety factors to be used in regard to

Process I has to be considered.

CONCLUSIONS AND SUMMARY

The following was presented in this article:

The Loop concept of the TÜV NORD Group is a very

innovative procedure for the evaluation of wall-

thinning in pipelines and power plants. The concept is

based on a Level 3 verification corresponding to the

requirements of the American regulation API 579 /

ASMIE FFS-1.

In LOOP, an accurately-detailed CAD/FEM model is

generated automatically and affordably from the

measurement data of an ultrasonic test pigging using a

special converter software. The use of the very realistic

FEM model enables the use of the full potential of

inelastic FEM analyses (ultimate bearing capacity) for

the evaluation of wall-thinning. As a result, significant

reserves are revealed in comparison with the simplified

evaluation process according to Level I and II.

The process according to the LOOP concept was

demonstrated using a practical example. In the process, a

computer-permitted operating pressure increase of +35% greater

than Level I/II verification was achieved.

A comparison of the fracture pressure calculations (ultimate

bearing capacity analysis) according to LOOP with the results of

a fracture pressure test revealed a good match of the test and

calculation (deviation: 3%).

Preparation of the LOOP concept for large-scale use is

currently under way.



REFERENCES

1. Mackenstein, P.; Schmidt, W.: „Beurteilung der Festigkeit

von fehlerhaften Pipelinerohren – Verfahren und

Bewertungskriterien.“, Journal 3R International 34 (1995)

Page 667-673

2. Engbert, F.; Engel, A.; Steiner, M.: „Erfahrungen bei der

Bewertung von Wanddickenminderungen mit der Methode

der Finiten Elemente“, Journal 3R International 40 (2001)

page 642-644

3. DIN EN 13445-3; „Unbefeuerte Druckbehälter – Teil 3:

Konstruktion“; German edition EN 13445-3:2009

4. ANSYS FEM-Software Version 14.5

5. API 579-1/ASME FFS-1, Fitness-For-Service, June 5, 2007


Tuev Nord Concept Loop: Lifetime Optimisation of Pipelines

Documents