® Registered trademarks of T.D. Williamson, Inc. in the United States and in foreign countries. © Copyright 2012

It’s all about the DataTexas Gas Association

San Antonio - 2017

Lloyd Pirtle

Transmission Roundtable

2

Contents

• History of In-Line Inspection (ILI)

• ILI tools – they’ve come a long way

• Recent Advancements

• Case Study Reviews

• Why is the data important ?

• Integrity Verification Process

• Pipe Identification (PI)

• Pipe Joint Classification (PJC) in Practice

• Positive Material Identification (PMI – NDE method)

• Questions

3

History of ILI

1942 War Emergency Pipeline

• Requirement for pipeline cleanliness

1. Improve product throughput

2. Control Internal Corrosion Rates

3. Improve first run success rate and data quality of ILI (after 1960s)

4

History of ILI -1960s

Magnetic Flux Leakage (MFL) tool: Developed in 1965 by

Tuboscope to detect areas of metal loss

• First tool had 12 sensors (6 in the patent applied for), with

90° arc on bottom of pipe (low resolution)

• Detected internal corrosion on crude oil pipelines

• Tape Recorder used as data storage

5

History of ILI – 1960s

The KALIPER® tool first developed by TDW in late 1960s

A single potentiometer provided course measurements of dents and

out-of-roundness pipe. Single channel recorder with sensors

mounted on inside of rear cups. Data recorded on pressure sensitive

paper inside the tool

6

History of ILI – 1970s and 1980s

1970s - British Gas (BG) invests in their own high-resolution MFL inspection tool. Vetco and Tuboscope succeed in developing new MFL tools.

Instead of 12 sensors per tool, these new tools had approximately 1.27 cm (0.5 in) spacing, permanent magnets, and inner diameter/outer diameter (ID/OD) discrimination.

Inspection tools could detect and size anomalies

7

History of ILI – 1970s and 1980s

Deformation (DEF) tools improve from single channel sensors to 6-12

channel sensors Nearly every vendor develops DEF tool with varying

levels of resolution.

Most advances to High resolution Deformation tools took place in the early

2000’s through today

8

History of ILI – 1990s and 2000s

� Ultrasonic Crack Detection (UTCD): Pipeline Integrity International (PII)

develops first tool in 1994 for axial cracks.

� Circumferential MFL (CMFL) - developed for axial metal loss and crack-

like defects

� Combo-tools are introduced using multiple technologies on a single tool

– DEF + MFL

Photo: geoilandgas.com

Photo: geoilandgas.com

9

History of ILI – 1990s and 2000s

Electromagnetic Acoustic Transducer (EMAT):

• Rosen and GE-PII develop first EMAT tools - crack inspections on

gas transmission pipelines

• TDW develops first EMAT tool in late 2000s

Photo: geoilandgas.comPhoto: rosen-group.com

10

History of ILI – 1990s and 2000s

� Speed Control tools – allows pipelines to operate at full rate while ILI tool travels at a slower rate.

Mapping (XYZ) tools – provides GPS coordinates of pipeline

Robotic tools – difficult to pig pipelines

• Unbarred tees, low/no-flow, diameter changes,

mitre bends

Photo: diakont.com

11

Technology Limitations

* *

*

**

12

Introduction

Technology Limitations

13

CONTINUOUS IMPROVEMENT

14

How Far have we come ?

� Tool specifications� length, weight, drag efficiency, sensor density, energy capacity

& consumption, pipeline negotiability� Tool run performance� Technology packaging & combining data sets� Data sizing parameters & specifications � Data sizing accuracies, tolerances & quality� Feature Identification & characterization� New Data Analysis techniques - same or similar technologies

INLINE INSPECTION TOOLS - THEY’VE COME A LONG WAY

15

Recent Advancements

Multiple Dataset (MDS) Tool

• TDW started in 2008 with 16-

inch MDS phase 1 tool

SpirALL® MFL, DEF, and GMFL

Multiple Data-Set Concept

• 5 different datasets

combined on a single tool

• Comprehensive threat

detection/identification

16

Recent Advancements

17

Case Study Reviews

Mechanical Damage

Selective Seam Weld Corrosion (SSWC)

T.P. Groeneveld et al. (PRCI 1991)

18

Mechanical Damage

19

Mechanical Damage

20

Dent Prioritization

MFL

4:30

12:15

DEF

4:30

12:15

SMFL

4:30

12:15

LFM

4:30

12:15

Mechanical Damage

21

• Mechanical Damage Classifier:• Distinguish dent with corrosion from dent with gouge

• Train model to recognize these types based upon ILI signal:

• Low field and High field MFL amplitude

• No. of metal loss signatures

• Location of metal loss signatures (apex, shoulder, both)

• Estimated metal loss depth

• 88 dent samples available from combination of actual ILI runs and pull tests through manufactured dents

Mechanical Damage

22

Selective Seam Weld Corrosion

SMFLMFL

23

MFL SMFL

Selective Seam Weld Corrosion

24

MFL SMFL

TDW Confidential

Selective Seam Weld Corrosion

25

Dent PrioritizationSelective Seam Weld Corrosion

• Selective Seam Weld Corrosion Classifier• Key attributes of MFL and SMFL signatures are measured in analysis software• SSWC detection algorithm:

• Detects SSWC anomalies while dismissing many corrosion anomalies near the long seam

• SSWC identification model mirrors evaluation process used by SMEs• Identification models accepts these measurements and produces a

likelihood that a candidate features is SSWC

26

PITS & PINHOLE - IMPROVED SIZING

Pinhole Analysis has been validated. It increases defect-sizing certainty, and reduces the corresponding errors

associated with comparing repeat ILI runs, which improves growth-rate determination.

* Small sample size

27

The Value of Data

• Establish pipeline condition

• Material Properties Confirmed for a pipeline segment

� Not limited to % of segment

� Known materials validated

� Unknown materials established

� No missing data

• Calculate safe test & operating pressures

• Test to targeted MAOP’s

• Minimize short and long term risks

• Prevent Failures

28

INTEGRITY VERIFICATION

29

Integrity Verification Process

30

PIPE IDENTIFICATION

31

Pipe Identification (PI)

PJC + PMI = PI

• Pipe Joint Classification (PJC) – When using multiple datasets

(MDS), the shared characteristics of pipe, from the same

manufacture, can be identified and used to assign each joint into

associated groups or bins

• Positive Material Identification (PMI) – Non destructive in-situ

acquisition of mechanical properties and chemical composition

correlated to API 5L for grade determination

• Pipe Identification (PI) – PJC to group pipe joints based on their

characteristics; PMI to identify material properties where they do

not exist; align PMI to PJC to close records gap for segments of

pipe or entire pipeline

• Also identifies “rogue” joints where records are not up to

date

32

PIPE MANUFACTURING PROCESS

33

Pipe Manufacturing Process

34

MULTIPLE DATA SETS (MDS)

Pipe Joint Classification with

35

Multiple Data Set Platform

Drive Section

with Odometers

SpirALL® MFL

Technology

High Field

Axial MFL

Low Field

Axial MFL

High Res

Deformation

• DEF – is a measurement of the changes of the inner bore and is sensitive to the

rolling in forming process.

• LFM – is the changes in magnetic flux at low field strengths and is sensitive to

chemical and metallurgical properties.

• MFL – less sensitive to local material differences a measure of the bulk magnetic

properties of the steel, especially circumferentially oriented markings.

• SMFL – similar to MFL but applied at an oblique direction so it detects

longitudinal aspects of material changes and long seam characteristics.

• IDOD – Designed to detect metal loss on the ID but is sensitive to ID surface

permeability changes in the radial direction.

What Each Data Set Measures

36

Multiple Data Sets

37

MULTIPLE DATA SETS

Pipe Joint Classification with

38

PJC IN PRACTICE

39

PJC in Practice: Low Field MFL (LFM)

Bin 1

Bin 2

Bin 3

Bin 4

Bin 5

40

PJC in Practice: High Field MFL (MFL)

Bin 1

Bin 2

Bin 3

Bin 4

Bin 5

41

PJC in Practice: SpirALL® MFL (SpirALL® MFL)

Bin 1

Bin 2

Bin 3

Bin 4

Bin 5

42

PJC in Practice – The End Result po id type dist begin lat long pipe ys pipe sf pipe matl pipe manuf

T13000370 Girth Weld 9231.704 29.970838 240.647984 42000 1 ERW X42

T13000380 Girth Weld 9241.471 29.970843 240.64784 42000 1 ERW X42

T13000390 Girth Weld 9251.202 29.970848 240.647696 42000 1 ERW X42

T13000400 Girth Weld 9260.939 29.970853 240.647554 42000 1 ERW X42

T13000410 Girth Weld 9270.682 29.970849 240.647411 42000 1 ERW X42

T13000420 Girth Weld 9280.428 29.970834 240.647268 42000 1 ERW X42

T13000430 Girth Weld 9290.102 29.970833 240.647253 42000 1 ERW X42

T13000440 Girth Weld 9291.157 29.970831 240.647232

T13000450 Girth Weld 9292.56 29.970828 240.647206

T13000460 Girth Weld 9294.397 29.970826 240.647191 35900 1 ERW B

T13000470 Girth Weld 9295.416 29.970824 240.64717 35900 1 ERW B

T13000480 Girth Weld 9296.843 29.970819 240.647029 35900 1 ERW B

T13000490 Girth Weld 9306.384 29.97082 240.647002 35900 1 ERW B

T13000500 Girth Weld 9308.231 29.97082 240.646985

T13000510 Girth Weld 9309.378 29.970822 240.646936

T13000520 Girth Weld 9312.657 29.970825 240.646797

T13000530 Girth Weld 9322.011 29.970827 240.646694

T13000540 Girth Weld 9328.989 29.97083 240.64659 42000 1 ERW X42

T13000550 Girth Weld 9336.112 29.970832 240.646556 42000 1 ERW X42

T13000560 Girth Weld 9338.512 29.970839 240.646383 42000 1 ERW X42

T13000570 Girth Weld 9350.682 29.970848 240.64621 42000 1 ERW X42

T13000580 Girth Weld 9362.893 29.970857 240.646041 42000 1 ERW X42

Sections of

unidentified

pipe can be

matched to

sections with

records

Groups of

unmatched

joints can be

tested using

PMI

43

POSITIVE MATERIAL

IDENTIFICATION (PMI)

44

Optical Emissions Spectrometry

45

Optical Emissions Spectrometry (OES)

The OES technology provides positive material property test

results for both Chemical Analysis (CA) and Carbon

Equivalency (CE).

� Spark creates non-destructive burn which creates a plasma burn

� Light is emitted and captured by the equipment to be analyzed

� Existence & concentrations of chemical components are measured

� CA & CE of that data point are determined

� Results compared with API-5L-Table 4

46

Mechanical Properties Assessment (MPA)

The MPA technology provides positive material propertytest results for both EYS and ETS by applying a loadand measuring the materials response to that load.

47

Mechanical Properties Assessment (MPA)

The MPA technology provides positive material property

test results for both EYS and ETS by applying a load and

measuring the materials response to that load.

� Indenter applies a load multiple times at a single location

� MPA indenter automatically measures and adjusts the load as

necessary to achieve a predetermined depth throughout the

sequential load/depth measurement processes

� final maximum depth is 6 mils (0.006”/0.152mm)

� stress/strain data is analyzed

� determine the EYS & ETS of that data point

� Results compared with API-5L-Table 6

48

Positive Material Identification (PMI)

49

Specifications for TDW PMI NDE Services

This PMI method has been fully vetted and validated with hundreds of

data sets and thousands of data points. The documented test data

includes internal and external validation. It can be shared in more detail

discussed further in an appropriate forum

50

MATERIAL IDENTIFICATION

51

PMI Applies to Fittings Identification (FI)

• Fitting Identification (FI) – The same Internal & External Validation

processes are being applied to statistically support the established

performance specifications

• Chemistry – The same OES technology and techniques are used to

establish fitting chemistry

• Yield & Tensile Strength – The same MPA technology and techniques

are used to establish fitting Yield & Tensile Strength

• Tensile is calculated from the Yield values

• Fitting Surfaces - Due to the shape of many fitting in the industry,

unique ‘fixtures’ may be required to install the MPA equipment

52

Pipe Identification (PI)

PJC + PMI = PI

• Pipe Joint Classification (PJC) – When using multiple datasets (MDS)

the shared characteristics of pipe, from the same manufacture, can

be identified and used to assign each joint into associated groups or

bins

• Positive Material Identification (PMI) – Non destructive in-situ

acquisition of mechanical properties and chemical composition

correlated to API 5L for grade determination

• Pipe Identification (PI) – PJC to group pipe joints based on their

characteristics; PMI to identify material properties where they do

not exist; align PMI to PJC to close records gap for segments of pipe

or entire pipeline

• Also identifies “rogue” joints where records are not up to date

® Registered trademarks of T.D. Williamson, Inc. in the United States and in foreign countries. © Copyright 2012

Thank You

Questions ?