Pipeline Research Council International,

Inc.

Advances in Pipeline Fitness-for-Service

A PRCI Webinar

SEIKOWAVE
2
www.prci.org

We will cover
2

 A brief review of corrosion damage assessment methods

 Determination of burst pressure – two examples
 Integration with NDT tools
 3D surface measurements
 UT
 What’s next?
 Pit gage data entry interface
 ASME B31G (2012), DNV RP-F101, Kastner
 Q&A
 www.prci.org

A BRIEF HISTORY OF CORROSION

DAMAGE ASSESSMENT METHODS FOR
PIPELINES
 www.prci.org

Barlow’s Formula
4

 Barlow's formula calculated the t

maximum internal pressure that a pipe
can withstand using the dimensions
and material properties of the pipe direction of flow  D

𝜎𝑜 2𝑡
𝑃=
𝐷
 Where
t
 P = burst pressure
 so = allowable stress
 t = pipe wall thickness D
 D = outside diameter of the pipe
 www.prci.org

Maxey’s Surface Flaw Equation

5
L

 Developed in the 1960’s to describe the impact of flaws on d t

reducing the maximum pressure of a pipe; modifies the
stress based on the surface flaw geometry

𝐴
1 −
𝐴𝑜 direction of flow  D
𝜎 = 𝜎𝑜
𝐴
1 −
𝐴𝑜 𝑀
 Where
𝐴 = 𝐿𝑑

𝑑 = 𝑑𝑒𝑝𝑡ℎ 𝑜𝑓 𝑎 𝑟𝑒𝑐𝑡𝑎𝑛𝑔𝑢𝑙𝑎𝑟 𝑑𝑒𝑓𝑒𝑐𝑡

𝐴𝑜 = 𝐿𝑡 t

0.8𝐿2
𝑀= 1+ D
𝐷𝑡
 www.prci.org

Corrosion Assessment Failure Prediction

6

 Based on concepts pioneered by Maxey and Kiefner

 Modifications to Barlow’s formula to account for surface flaws
 First ASME B31G standard in 1991
 Subsequent revisions in 2009, and 2012
 RSTRENG
 Based on a more detailed assessment of the shape of the corrosion damage
 Incorporates more accurate Folias factors
 Original development by John F. Kiefner while at Battelle Memorial Institute
 Ongoing development and advancement supported by PRCI
 www.prci.org

Original B31G Parabolic defect model for defects

7
L
d t
2𝑑
𝜎𝑓𝑙𝑜𝑤 2𝑡 1 − 3𝑡
𝑃𝑏𝑢𝑟𝑠𝑡 =
𝐷 2𝑑
1 −
3𝑡𝑀
Where direction of flow  D
2
𝐴 = 𝑑𝐿
3

𝜎𝑓𝑙𝑜𝑤 = 1.1𝑆𝑀𝑌𝑆

0.8𝐿2 t
𝑀= 1+
𝐷𝑡
L = defect length
d = maximum defect depth
D = pipe diameter
D
t = pipe wall thickness
SMYS = Specified Minimum Yield Strength
For defects defined as 𝐿 ≤ 20𝐷𝑡
 www.prci.org

Original B31G Rectangular defect model for long defects

8
L
d t
𝜎𝑓𝑙𝑜𝑤 2𝑡 𝑑
𝑃𝑏𝑢𝑟𝑠𝑡 = 1−
𝐷 𝑡
Where
𝐴 = 𝐿𝑑 direction of flow  D

𝜎𝑓𝑙𝑜𝑤 = 1.1𝑆𝑀𝑌𝑆

L = defect length
d = maximum defect depth t
D = pipe diameter
t = pipe wall thickness
SMYS = Specified Minimum Yield Strength
For defects defined as 𝐿 > 20𝐷𝑡
D
 www.prci.org

Bulging stress magnification factor

0.85dL Method (Folias factor) depends on defect length
9
L
d t
𝑑
𝜎𝑓𝑙𝑜𝑤 2𝑡 1 − 0.85 𝑡
𝑃𝑏𝑢𝑟𝑠𝑡 =
𝐷 𝑑
1 − 0.85
𝑡𝑀
Where direction of flow  D
𝐴 = 0.85𝑑𝐿

𝜎𝑓𝑙𝑜𝑤 = 𝑆𝑀𝑌𝑆 + 10,000𝑝𝑠𝑖

2
𝐿2 𝐿2
𝑀= 1 + 0.6275 − 0.003375 t
𝐷𝑡 𝐷𝑡
L = defect length
d = maximum defect depth
D = pipe diameter
t = pipe wall thickness D
SMYS = Specified Minimum Yield Strength
For defects defined as 𝐿 ≤ 50𝐷𝑡
 www.prci.org

Bulging stress magnification factor

0.85dL Method (Folias factor) depends on defect length
10
L
d t
𝑑
𝜎𝑓𝑙𝑜𝑤 2𝑡 1 − 0.85 𝑡
𝑃𝑏𝑢𝑟𝑠𝑡 =
𝐷 𝑑
1 − 0.85
𝑡𝑀 direction of flow  D
Where
𝐴 = 0.85𝑑𝐿

𝜎𝑓𝑙𝑜𝑤 = 𝑆𝑀𝑌𝑆 + 10,000𝑝𝑠𝑖

𝐿2 t
𝑀 = 0.032 + 3.3
𝐷𝑡
L = defect length
d = maximum defect depth
D = pipe diameter D
t = pipe wall thickness
SMYS = Specified Minimum Yield Strength
For defects defined as 𝐿 > 50𝐷𝑡
 www.prci.org

Comparison
11

Criterion Original B31G 0.85dL Method Effective Area

Flow stress 1.1SMYS SMYS + 10,000psi SMYS + 10,000psi
Defect area 2 0.85𝑑𝐿 Effective Area
𝑑𝐿 𝑜𝑟 𝑑𝐿
3
Transition length 20𝐷𝑡 50𝐷𝑡 50𝐷𝑡
Folias factors 1 2 2
Defect model 2 1 Corrosion Profile
Pressure model 2 1 1

• Improved performance achieved by adjusting the bulging stress magnification factor (Folias factor) for the length of the defect
and maintaining a single failure stress (pressure) model
• Effective area is the best method for estimating the remaining strength of the pipe (hence the term rstreng)
 www.prci.org

ASME B31G Folias Factor Comparison

12

12

10

8 0.8𝐿2
𝑀𝐵31𝐺 = 1+
𝐷𝑡
6

2
4 𝐿2 𝐿2 𝐿2
𝑀0.85𝑑𝐿 = 1 + 0.6275 − 0.003375 ≤ 50
𝐷𝑡 𝐷𝑡 𝐷𝑡
2
𝐿2 𝐿2
𝑀0.85𝑑𝐿 = 0.032 + 3.3 > 50
𝐷𝑡 𝐷𝑡
0
0 10 20 30 40 50 60 70 80 90 100 110 120

𝐿2
𝐷𝑡
 www.prci.org

Maxey’s Surface Flaw Equation

13 A second look
L

𝐴 d t
1 −
𝐴𝑜
𝜎 = 𝜎𝑜
𝐴
1 −
𝐴𝑜 𝑀 direction of flow  D
What’s the best
method to 𝐴 = 𝐴𝑟𝑒𝑎 𝑜𝑓 𝑑𝑎𝑚𝑎𝑔𝑒
estimate the area
of damage, A? 𝑑 = 𝑑𝑒𝑝𝑡ℎ 𝑜𝑓 𝑎 𝑑𝑒𝑓𝑒𝑐𝑡
t
𝐴𝑜 = 𝐿𝑡
D
𝑀 = Bulging stress magnification
factor (Folias factor)
 www.prci.org

Effective Area
14

 Better calculation of the area of damage

 Does not depend on a specific defect model
 Parabolic, rectangular, or otherwise
 Requires detailed data regarding the shape of the infrastructure damage
 One step closer to a FEA (finite element analysis) for damage assessment
 www.prci.org

15

DETERMINATION OF BURST PRESSURE

– TWO EXAMPLES
 www.prci.org

16

• Broad area
corrosion
• Data collected
using a 3D surface
measurement tool
 www.prci.org

17

• Corrosion damage
analyzed to
determine depth
and extent of metal
loss
 www.prci.org

18
 www.prci.org

Fitness for Service Determination

19

ASME B31G-2012 Revision

 www.prci.org

20

Defect length = 12.389”

𝐿≤ 20𝐷𝑡 = 17.321“

Area model for ASME B31G (1991)

2
𝐴𝐵31𝐺 = 𝐿𝑑 = 1.148 𝑖𝑛2
3

Area model of 0.85dL

𝐴0.85𝑑𝐿 = 0.85𝑑𝐿 = 1.464 𝑖𝑛2

Area model Effective Area

𝐴𝐸𝐴 = 0.860 𝑖𝑛2
 www.prci.org

Area Estimate Comparison Corrosion (River bottom) Profile

21

Effective Area

Area model for ASME B31G (1991)

Defines Original ASME B31G 2
boundary for area estimation 𝐴𝐵31𝐺 = 𝐿𝑑 = 1.148 𝑖𝑛2
3
Defines 0.85dL boundary Area model of 0.85dL
for area estimation 𝐴0.85𝑑𝐿 = 0.85𝑑𝐿 = 1.464 𝑖𝑛2

Area model Effective Area

𝐴𝐸𝐴 = 0.860 𝑖𝑛2
 www.prci.org

22

• Isolated pits that

likely interact to
form a single
defect
• Data collected
using a 3D surface
measurement tool
 www.prci.org

23

• Isolated pits that

likely interact to
form a single
defect
• Data collected
using a 3D surface
measurement tool
 www.prci.org

24

Defect length = 141mm

𝐿≤ 20𝐷𝑡 = 225𝑚𝑚

Area model for ASME B31G (1991)

2
𝐴𝐵31𝐺 = 𝐿𝑑 = 408 𝑚𝑚2
3

Area model of 0.85dL

𝐴0.85𝑑𝐿 = 0.85𝑑𝐿 = 520 𝑚𝑚2

Area model Effective Area

𝐴𝐸𝐴 = 263 𝑚𝑚2
 www.prci.org

Corrosion (River bottom) Profile

25

Defines Original ASME

Defines 0.85dL boundary
B31G boundary for area
for area estimation
estimation
 www.prci.org

26

INTEGRATION WITH NDT TOOLS

27
www.prci.org

Integration with NDT tools

27

 Seikowave Tools
 3DSL Rhino
 3D Toolbox
 Ultrasound
 Olympus
 Other 3D surface measurement tools Need 3D data in
 Handyscan 700
 Coordinate Measurement Machines (e.g. Mitutoyo)
ply or stl format
 www.prci.org

28

WHAT’S NEXT?
 www.prci.org

29

• Complexity
• Difficult to acquire the
data
• Difficult to perform the
calculation
• Conservatism
• More conservative
generally means more
cost to maintain
 www.prci.org

30

ASME B31G-2012 includes

• Expanded definition of flow stress
• Applicability to metal loss in field bends,
induction bends and elbows

ASMEB31G-2012 does not include

• Preferential corrosion affecting pipe
seams or girth welds
• Metal loss in fittings other than bends
and elbows
 www.prci.org

ASME B31G (2012) Flow Stress

31

Material SMYS Temperature Flow Stress

Carbon Steel T < 250F (120C) sflow = 1.1 X SMYS
sflow < SMTS
Carbon Steel 𝑆𝑀𝑌𝑆 ≤ 70𝑘𝑝𝑠𝑖 (483MPa) T < 250F (120C) sflow = SMYS + 10kpsi (69MPa)
and low-alloy
Steel sflow < SMTS
Carbon Steel 𝑆𝑀𝑌𝑆 ≤ 80𝑘𝑝𝑠𝑖 (551MPa) sflow = (sYT + sUT)/2
and low-alloy
Steel sYT and sUT are specified at the
operating temperature (YT is
the yield strength and UT is the
ultimate strength in tension)

ASME B31G-2012 Revision

 www.prci.org

Flow Stress Examples

32

SMYS + 10kpsi
SMYS SMTS 1.1 SMYS (69MPa) (SMYS+SMTS)/2
ksi Mpa ksi Mpa ksi Mpa ksi Mpa ksi Mpa
X65 65 448 77 531 71.5 493 75 517 71 490
X80 80 551 90 621 88 607 90 621 85 587
X100 100 690 110 759 110 759 110 759 105 725
2𝑑
𝜎𝑓𝑙𝑜𝑤 2𝑡 1 −
𝑃𝑏𝑢𝑟𝑠𝑡 = 3𝑡
• For X65, Pburst is higher when using SMYS + 10kpsi 𝐷 2𝑑
• For X80, SMYS + 10kpsi equals SMTS 1 − 3𝑡𝑀
• For X100, 1.1SMYS and SMYS + 10kpsi = SMTS
• Average of SMYS and SMTS is more conservative but Subject of ongoing PRCI studies
• ASME B31G-2012 does not cover pipe with SMYS > 80kpsi
 www.prci.org

Remaining Strength Assessment

33

What’s Available Now What’s Next (Q2 2016)

 Original ASME B31G  Pit Gage Data Entry
 Modified 0.85dL  Robotic Collection
 Additional Capabilities
 Effective Area  Expanded definition of flow stress
 API 579  DNV RP-F101
 Level 1 and Level 2 • Per DNV RP-F101, applicable for corrosion
in girth welds and seam welds
• Applicable for temperatures above 250F
(120C)
 Kastner (Circumferential corrosion
analysis)
• Better solution for examining extensive
circumferential corrosion
 www.prci.org

Pit Gage Data Entry

34

 Data from NDT tools is

not always available
 Sometimes a pit gage is
all you have 0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.034
0.000
0.037
0.000
0.035
0.000
0.000
0.000
0.000
0.000
0.000

 Data entry in a grid 0.000

0.034
0.000
0.034
0.036
0.039
0.050
0.052
0.058
0.060
0.056
0.059
0.038
0.048
0.000
0.000
0.000
0.000
0.037 0.037 0.040 0.052 0.054 0.221 0.049 0.042 0.041
format that matches the 0.036 0.037 0.037 0.035 0.041 0.043 0.041 0.044 0.048
0.000 0.033 0.035 0.039 0.040 0.038 0.139 0.111 0.045
grid drawn on the 0.000
0.000
0.000
0.000
0.045
0.036
0.090
0.037
0.122
0.036
0.038
0.036
0.037
0.038
0.041
0.041
0.041
0.041

pipeline 0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.035
0.037
0.039
0.038
0.038
0.041
0.039
0.041
0.039
0.039
0.034

 Integrates with existing 0.000

0.000
0.000
0.000
0.000
0.000
0.034
0.000
0.034
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

database 0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

 Integrates with existing 0.000

0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

flaw detection and

interaction rule software
 Integrates with Pipeline
FFS for fitness for service
calculations
 www.prci.org

35
36
www.prci.org

Robotic data collection

36

 Seikowave robotic systems Seikowave systems

navigating and
can move omni- measuring inside and
directionally over surfaces outside of pipes (1)
and can be operated
remotely enabling
collection of
 3D inspection data
 Other inspection data (e.g. UT,
eddy current)
 Untethered
 Able to operate as far as 300
meters from the base station
(1) Photos courtesy of Asahi and ExxonMobil
 www.prci.org

FEA – Under Development

37

𝑃𝑐𝑜𝑚𝑝𝑢𝑡𝑒𝑑
< 1 𝑡ℎ𝑒𝑛 𝑐𝑜𝑚𝑝𝑢𝑡𝑎𝑡𝑖𝑜𝑛 𝑖𝑠 𝑐𝑜𝑛𝑠𝑒𝑟𝑣𝑎𝑡𝑖𝑣𝑒
𝑃𝑏𝑢𝑟𝑠𝑡 𝑡𝑒𝑠𝑡

More work still needed but shows promise

 www.prci.org

38

February 2-4, 2016 at the Omni San Diego Hotel in San Diego, CA

COME TO THE PRCI RESEARCH

EXCHANGE TO LEARN MORE