Risk Based Inspection Planning For Jacket Structures: Where Did We Start, What Did We Do and Where To Go?

1er.
Simposio “Análisis, diseño, evaluación, inspección y

mantenimiento de plataformas marinas. Aspectos relevantes"
Risk Based Inspection Planning for Jacket

Structures
Where did we start, what did we do and
where to go?
Michael Havbro Faber
Matrisk GmbH, / ETH Zürich Switzerland, faber@matrisk.com / faber@ibk.baug.ethz.ch
1
1er. Simposio “Análisis, diseño, evaluación, inspección y
Overview of Presentation
• Who is ETH Zürich and Matrisk?
• Main interactions with IMP/Pemex
• RBI for platforms

- where did we start?
- where are we now?
- where to go?
• An outlook and some observations
2
Who is ETH Zürich and Matrisk?

• My ETH Research Group
15 researchers working in the areas of:
- Inspection and maintenence planning for offhore facilities, structures

and pressure vessels
- Probabilistic modeling of NDE methods
- Methods of systems reliability analysis
- Natural hazards risk management, earthquake, hurricanes, rockfall,
avalanches, floods
- Decision making for sustainable developments
- Fire risk modelling and assessment
- Risk criteria and safety management
3

• Matrisk GmbH – Specialist Consulting Firm
Offshore and marine engineering
- Risk Based Inspection planning for fixed steel structures

- Risk Based Inspection Planning for FPSO and FSO‘s
- Probabilistic Design Basis for FPSO‘s in the GoM
- Decision and risk assessment for decommissioning
- Development of RAC for design and operation
- Probabilistic modeling of loads
- Probabilistic modeling of deterioration
4

Civil engineering
- Hurricane modeling and loss estimation

- Probabilistic design basis fof structures
- Risk and safety project management
- Decision analysis for planning of deep thermal energy exploitation
- Risk assessment for roadway and tunnel systems
- Development of RAC for third persons
- Risk analysis of infrastructure systems
5

Client portfolio
- Det Norske Veritas, Denmark
- Bureau Veritas, France, Thailand, Brazil
- ABS, Houston, USA
- Aon Reinsurance, Japan
- Mærsk Oil and Gas, Denmark
- Instituto Mexicano del Petroleo, Mexico
- Pemex, Ciudad del Carmen, Mexico
- Zentech, Houston, USA
- Norwegian and Swiss Roadway Authorities
- COWI, Denmark
- Andermat Tourist Resort, Switzerland
6

• My own background
Positions
- PhD (1989) from Aalborg University of Denmark
- Post-doc at Technical University of Munich, with Rackwitz.
- Private company, RCP, software STRUREL and consulting for the
aeronautical and the offshore industry
- COWI, group leader risk and reliability, infrastructure and offshore
- Det Norske Veritas, lead engineer for the RBI of Jotun and Balder
fields including FPSO‘s, subsea and flexible systems
- ETH Zürich, (2000), chair of Risk and Safety,
- Leader of MAS on Natural Hazards Risk Management
- DTU (2011, Jan 1), Professor in Risk and Safety and Dean of
Civil Engineering DTU, Denmark
7
Main Interactions with IMP / Pemex

• RISC project - 1994
- Joint Industry Project with IMP (Dr. Oscar

Valle) and University College London
and Technical Software Consultants
(Prof. Bill Dover)
Together with UCL/TSC and IMP a

software was developed for RBI
for steel offshore platforms and we
tested this on Nohoch B
8

• The period 2001 - 2010
- For Pemex - in collaboration with Zentech and

UNACAR we have conducted RBI planning for
welded joints of 67 platforms in the bay of
Campeche
- For Pemex we have conducted several courses on RBI

for fixed steel jacket structures
- In collaboration with COMIMSA we have

generalized our technique and also applied it
for static process equipment on the AKAL-C
facilities
9

• The period 2001 - 2010
- In collaboration with IMP we have established

a methodology for prioritization of repair works
(SIPRIAM)
- In collaboration with IMP we have developed

a risk based design basis for FPSO concepts
in the Gulf of Mexico
- with IMP I/ETH have an MoU concerning development

and knowledge transfer in the area of risk and reliability
assessments
10

• IMP/Pemex Risk Symposium
December 2001
LD RSRLD
- A very successful event with good

Reliability Based Inspection Planning of
discussions (Bob Bea/Bernie Stahl) Fatigue Damaged Offshore Platforms
M. H. Faber, Swiss Federal Institute of Technology

Risk Symposium
IMP/Pemex, December 3-4, 2001
- I presented my ideas on how we might Joint
Probability
use generic approaches to make RBI Density
Resistance
practically applicable Load
Safe Failure
J
- Ing. Juan de Dios Dela O Ramirez

contacted me and asked if I felt we were
ready to do this for real….
11
RBI for Platforms

• Where did we start?
We had:
The general theoretical principles for RBI

- decision theory
- reliability theory
- structural analysis (FLS/ULS)
- fracture mechanics
D d(t)
- inspection modeling dCRIT
The RBI planning was limited to experts in t
probabilistic engineering methods
Computation and data handling was involving Pfcrit
Pf(t)
We were missing practical tools
12
RBI for Platforms

Risk Based Inspection planning as an optimization problem
1.5E-02
Total expected cost
Failure Cost
Inspection cost
Expected Costs
1.0E-02 Repair Cost
5.0E-03
0.0E+00
0 2 4 6 8 10
Number of planned inspections
13
RBI for Platforms

Decision analysis and risk assessments as framework for optimization

Failure
Failure No detection
No detection Survival
Survival Detection and Failure
no repair
Survival
N
CIN (e, d )   CIN ,i (q)1  PF (Ti ) 
Detection and
Failure
1
repair
Survival i 1N (1  r )Ti
1
CREP (e, d )   CR ,i PRi
Failure
(1  r )Ti
Failure
Detection and No detection i 1
Failure Survival TL
no repair 1
CF (e, d )   CF (t )PF ,t PCOL FAT ( RSR )
Survival
Survival Detection and
(1  r )
Failure j t
no repair t 1
Survival

 
Detection and
Failure
repair
PFt  P  Bk (t  1)  S (t  1) Bk (t ) S U (t )
Survival
 k 
Detection and No detection
Failure PR (t )  P( S (t  1)  Bk (t  1)  R(t ))
Failure
repair Survival k
Survival
Detection and Failure
no repair
Survival
Detection and
Failure
repair
Survival
Inspection 1 Inspection 2
14
RBI for Platforms

Considering predictable fatigue crack growth and accounting for the

quality of inspections
Probability of Detection on tubulars, underwater
1
ACFM
MPI
0.8
0.6
POD
0.4
0.2
0
0 2 4 6 8 10
Crack depth [mm]
15
RBI for Platforms

Basis is taken in the fatigue design – the SN curves
But SN curves cannot relate performance to observable characteristics
Therefore fracture mechanic formulations are calibrated to SN approach
N  NI  NP a
t

m
dc C c  K C 
   ca 0   c0
da C A  K A 
2c
a
N a 0   N I
dN 1
 2c
da C A K A m
t
C A  1.1 CC
m
16
RBI for Platforms

The calibration is performed for different 5.0
4.5 Beta SN
design conditions/assumptions , i.e. for 4.0
3.5
Beta FM
Reliability Index (t)

different 3.0
2.5
2.0
1.5
- stress model uncertainties 1.0
0.5
- Weibull long term stress distribution 0.0

0 10 20 30
Years
40 50 60
shape parameter
- material thickness
- design fatigue life
- design SN-curve
17
RBI for Platforms

Select joint type
•
- SN curve
Where are we now? - SIF
- Critical crack depth
1,00E+00
1,00E-01
Probability of failure
1,00E-02
1,00E-03
Generic approach to RBI Determine acceptance criteria

PAC (RSR )
j
1,00E-04
1,00E-05
1,00E-06
This formed the basic Calculate probability of failure and

probability of repairs for different
0 5 10 15
Year in service
20 25
FDF 1, FDF 2, FDF i,..

building 2,00E-02
1,80E-02
stone for efficient and 1,60E-02
Expected total costs

Determine inspection plans for 1,40E-02
1,20E-02
different FDF 1, FDF 2, FDFi,.. and
practical RBI at larger scale different target levels PfT
1,00E-02
8,00E-03
6,00E-03
4,00E-03
2,00E-03
0,00E+00
1,0E-04 1,0E-03 1,0E-02
Determine costs for inspection plans
- using indicators for the joint
Acceptance criteria
FDFi
for different FDF 1, FDF2, FDFi,..
Time in service
performance
1,00E-02
0 5 10 15 20 25
Probability of failure
- constant threshold of annual Choose cost optimal inspection plan
fulfilling the acceptance criteria
1,00E-03
failure probability 1,00E-04
- prefabrication of optimal 1,00E-05
inspection plans
18
RBI for Platforms

• Where are we now? - material codes
Input
- pressure
- process - costs
- temperature - ………
Inspection plans are performed as

function of Indicators of future
joint performance such as
- Design fatigue life

- Thickness of the elements
- Inspection methods
- Design SN curves
- Thresholds of annual
probability of failure
- Stress model uncertainty
- Long term stress characteristics Generic parameter
19
RBI for Platforms

• Where are we now?
Development of the tool iPlan (MS Excel based)
Input iPlan Output
Fatigue Life
Inspection
Threshold method Years
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
10-5 MPI X X X X
10-5 ACFM X X X X
3·10-5 MPI X X X
3·10-5
Thickness
ACFM X X X
10-4 MPI X X
10-4 ACFM X X
3·10-4 MPI X
3·10-4 ACFM X
10-3 MPI
10-3 ACFM
10-2 MPI
COV
10-2 ACFM
Inspection
Threshold method Years
31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
10-5 MPI X X
Environment
10-5 ACFM X X
3·10-5 MPI X
3·10-5 ACFM X
10-4 MPI X
10-4 ACFM X
3·10-4 MPI X
3·10-4 ACFM X
Model uncertainties
10-3 MPI
10-3 ACFM
10-2 MPI
10-2 ACFM
20
RBI for Platforms

Development of the tool iPlan (MS Excel based)
21
RBI for Platforms

The procedure for generic RBI

FDF from design
calculations
Inspection Plan as a Perform first inspection at time t = t 1

function of detail type Crack
and FDF: detected
No crack detected
• Inspection times t 1 – tN
• Inspection method
Perform second inspection at time t = t 2
Crack
detected
No crack detected
Perform last inspection at time t = t N

Crack
detected
No crack detected
End of service life
Perform repair, replacement,

Calculate new FDF
monitoring, no action
22
RBI for Platforms

The RBI tool iPLAN – Input Sheet
23
RBI for Platforms

The RBI tool iPLAN – Optimal Threshold Sheet
24
RBI for Platforms

The RBI tool iPLAN – Final Inspection Plans Sheet
25
RBI for Platforms

• Where to go?
Benefit of RBI ?
4
RBI Approach
3.5
Relative cost entire platform
20 Years Interval
3 4 Years Interval
2.5
1.5
0.5
0
1 2 3
YES ! Different cost models
26
RBI for Platforms

• Where to go?
4000000
Failure
Tool for Asset Integrity Management? 3500000
3000000
Repair
Inspection
2500000
Cost
2000000
4000000
10000000 1500000
Failure 3500000
Failure
Repair
3000000 Inspection
1000000
9000000 Repair 2500000
Cost
2000000
Inspection
500000
4000000
1500000
8000000 1000000
0 Failure 500000
Platform 1 3500000
Platform 2 Platform 3 Platform 4
Repair
7000000 0
Platform 1 Platform 2
3000000 Inspection
6000000 25000004000000
Failure
Cost
3500000
Cost
Repair
Inspection
5000000 2000000
3000000
2500000
1500000
Cost
2000000
4000000 1500000
1000000
1000000
3000000 500000
500000
0
Platform 1 Platform 2
2000000 0
Platform 1 Platform 2 Platform 3 Platform 4
1000000
0
Total
YES!
27
An Outlook and Some Observations

• Pemex and IMP are in the forefront !
Significant investments have been directed into the

development of methodologies and tools
But availability of tools is not enough !
- tools must be distributed

- tools must be integrated into operational and strategic
routines
- tools must be learned
- tools must be updated and kept alive
28
An Outlook and Some Observations

• Which should be the next developments
Tools are presently available for

- RBI for joints in steel jacket structures – iPlan (jacket)
- RBI for static process equipment – iPlan (Process)
- Inspection and repair prioritization - SIPRIAM
We now need tools facilitating the integration of these basic

tools with the daily routines of the assets, some of the next
logical steps would be to establish links with
- graphical databases
- inspection and repair contracting
- condition reporting for updating
and to fix procedures on how to utilize the available tools in
operational and strategic routines
29
Gracias por su atencción
30

Risk Based Inspection Planning For Jacket Structures: Where Did We Start, What Did We Do and Where To Go?

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Risk Based Inspection Planning For Jacket Structures: Where Did We Start, What Did We Do and Where To Go?

Uploaded by

Copyright:

Available Formats

1er.

Simposio “Análisis, diseño, evaluación, inspección y

Risk Based Inspection Planning for Jacket

• Main interactions with IMP/Pemex

• RBI for platforms

• An outlook and some observations

Who is ETH Zürich and Matrisk?

15 researchers working in the areas of:

- Inspection and maintenence planning for offhore facilities, structures

Who is ETH Zürich and Matrisk?

Offshore and marine engineering

- Risk Based Inspection planning for fixed steel structures

Who is ETH Zürich and Matrisk?

- Hurricane modeling and loss estimation

Who is ETH Zürich and Matrisk?

Who is ETH Zürich and Matrisk?

Main Interactions with IMP / Pemex

- Joint Industry Project with IMP (Dr. Oscar

Together with UCL/TSC and IMP a

Main Interactions with IMP / Pemex

- For Pemex - in collaboration with Zentech and

- For Pemex we have conducted several courses on RBI

- In collaboration with COMIMSA we have

Main Interactions with IMP / Pemex

- In collaboration with IMP we have established

- In collaboration with IMP we have developed

- with IMP I/ETH have an MoU concerning development

Main Interactions with IMP / Pemex

- A very successful event with good

M. H. Faber, Swiss Federal Institute of Technology

use generic approaches to make RBI Density

- Ing. Juan de Dios Dela O Ramirez

RBI for Platforms

The general theoretical principles for RBI

RBI for Platforms

Risk Based Inspection planning as an optimization problem

1.0E-02 Repair Cost

RBI for Platforms

Decision analysis and risk assessments as framework for optimization

RBI for Platforms

Considering predictable fatigue crack growth and accounting for the

Probability of Detection on tubulars, underwater

RBI for Platforms

Basis is taken in the fatigue design – the SN curves

But SN curves cannot relate performance to observable characteristics

Therefore fracture mechanic formulations are calibrated to SN approach

RBI for Platforms

The calibration is performed for different 5.0

design conditions/assumptions , i.e. for 4.0

Reliability Index (t)

- stress model uncertainties 1.0

- Weibull long term stress distribution 0.0

RBI for Platforms

Generic approach to RBI Determine acceptance criteria

This formed the basic Calculate probability of failure and

FDF 1, FDF 2, FDF i,..

stone for efficient and 1,60E-02

Expected total costs

failure probability 1,00E-04

- prefabrication of optimal 1,00E-05

RBI for Platforms

Inspection plans are performed as