Aalborg Universitet

Benefits of Risk Based Inspection Planning for Offshore Structures

Straub, D.M.; Goyet, J.; Sørensen, John Dalsgaard; Faber, Michael Havbro

Published in:
Proceedings of the 25th International Conference of Offshore Mechanics and Arctic Engineering

Publication date:
2006

Document Version
Publisher's PDF, also known as Version of record

Link to publication from Aalborg University

Citation for published version (APA):

Straub, D. M., Goyet, J., Sørensen, J. D., & Faber, M. H. (2006). Benefits of Risk Based Inspection Planning for
Offshore Structures. In Proceedings of the 25th International Conference of Offshore Mechanics and Arctic
Engineering

General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners
and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

? Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
? You may not further distribute the material or use it for any profit-making activity or commercial gain
? You may freely distribute the URL identifying the publication in the public portal ?
Take down policy
If you believe that this document breaches copyright please contact us at [email protected] providing details, and we will remove access to
the work immediately and investigate your claim.

Downloaded from vbn.aau.dk on: octubre 13, 2021

 Proceedings of OMAE’05
25th International Conference on Offshore Mechanics and Arctic Engineering
June 4-9, 2006, Hamburg, Germany

OMAE2006-92089

BENEFITS OF RISK BASED INSPECTION PLANNING FOR OFFSHORE

STRUCTURES

Daniel Straub Jean Goyet

Matrisk, Switzerland Bureau Veritas, Marine Department, Paris
[email protected] [email protected]

John D. Sørensen Michael H. Faber

Aalborg University, Denmark Swiss Federal Institute of Technology, Switzerland
[email protected] [email protected]

ABSTRACT structures have not been systematically quantified in past

The economical benefits of applying risk-based inspection publications. Furthermore, in the public domain little
planning (RBI) for offshore structures subject to fatigue are information is available on realistic estimates of costs related to
evaluated based on experiences from past industrial projects. the structural integrity management of these structures. As an
To this end, the factors influencing the cost of inspection, repair example, in Dalane et al. (1990) the resulting expected cost for
and failure of structures are discussed and realistic values of different designs and inspection strategies are compared for
these costs are presented. These are then applied to assess the example details in fixed and floating offshore structures, yet
expected costs from different inspection strategies, both risk- without presenting the underlying cost model.
based strategies as well as inspection strategies with fixed In this paper, the factors influencing the costs of
inspection intervals for all potentially critical elements. By inspections and repairs are reviewed. Based on past experience,
comparing these expected costs, the financial benefit of RBI is typical costs of different types of inspections and repair actions
assessed. are presented for fixed and floating offshore steel structures,
including FPSO’s, subject to fatigue damages. Furthermore, the
INTRODUCTION factors influencing the cost of a failure in the structure are
For offshore structures, risk and reliability based discussed. On this basis, the financial benefits of performing
inspection planning (RBI) procedures have been developed and risk based inspection planning are determined for some typical
implemented since the 1980s, mostly for fatigue deterioration cases by comparing the associated expected costs with the
of fixed jacket steel structures, Skjong (1985), Madsen et al. expected cost of an inspection strategy with fixed, predefined
(1989) and Fujita et al. (1989), but more recently also for ship inspection intervals. This assessment is based on the generic
and floating production storage and offloading systems subject approach to RBI.
to corrosion and fatigue, Lotsberg et al. (1999) and Goyet et al.
(2004). While the significant computational efforts required by INSPECTION STRATEGIES
RBI hindered the applications in the past, this restriction has The potential inspection strategies can be divided into
been resolved with the development of the generic approach to three groups, namely prescriptive strategies, qualitative
RBI, see Faber et al. (2000), Straub (2004), Faber et al. (2005) strategies and quantitative risk-based strategies.
and Straub and Faber (2006), which facilitates the highly
efficient application of RBI for portfolios of offshore structure. Prescriptive (or rule-based) inspection planning
Although it has been observed that RBI in general reduces Prescriptive inspection plans require that inspections are
the amount of inspections considerably, Moan et al. (2000), the performed for all joints in the structure at fixed intervals in
financial benefits of applying RBI strategies for offshore time. Such rule-based inspection planning is still commonly

1 Copyright © 2006 by ASME

applied in the offshore industry. Its main advantage is that it facilitate the updating of any deterioration model in the
does not require further structural and risk analysis of the presence of an inspection outcome through the application of
structure. Additionally, it may facilitate the planning of the Bayes’ rule, as demonstrated by Madsen (1987). Through the
logistic aspects of the inspection campaigns, because the assumption of no-indication at the inspections, the required
number of inspections in each campaign can easily be adjusted inspection times to comply with a given threshold on the
to the available inspection capacities. Rule-based strategies are acceptable annual failure probability can be determined, as
thus defined completely by the inspection interval ΔTInsp . illustrated in Figure 1 for thresholds 10-3 yr-1 and 10-4 yr-1. The
assumption of no-indication implies that mitigation measures
Partly risk-based inspection strategies (qualitative or (repair, monitoring, follow-up inspections) are taken in case a
semi quantitative risk-based strategies) defect is indicated at any of the inspections. The fact that
Some owners and operators of offshore structures inspection intervals increase with time reflects the increased
implement an inspection policy which corresponds to a confidence in the fatigue performance of the hot spot after the
combination of the rule-based and the RBI approach, i.e., they inspections.
apply inspection strategies which are partly risk-based. These
strategies prescribe inspection intervals separately for groups of ΔpFT = 10-3 yr-1
10-3
joints, in accordance with the considered qualitative and/or

Annual probability of failure ΔpF

quantitative risk indicators. As an example, in Pemex (2000),
inspection intervals are determined for fixed steel platforms as
a function of various indicators, including the member ΔpFT = 10-4 yr-1
-4
importance (primary, secondary or tertiary member) and the 10

calculated fatigue life.

These partly risk-based inspection strategies are not
considered explicitly in this paper. It is argued that although
strategies based on qualitative indicators are preferable to 10-5
0 10 20 30 40 50 60 70 80 90 100
purely prescriptive inspection planning strategies, they do have
Year t
similar disadvantages. This holds in particular when Inspection
times:
considering fatigue deterioration, as qualitative indicators are t
poor in describing fatigue performance. Partly risk-based
inspection strategies, which are based on quantitative indicators Figure 1. Inspection times as determined from the
(such as the calculated fatigue life) are not treated in the application of Bayes’ rule, from Straub (2004)
following, as it is argued that with the availability of these
indicators it is preferable to perform a fully quantitative RBI, The calculated probabilities can be applied to determine
as, once the indicators are available, the additional effort for inspection strategies that comply with given risk acceptance
such an analysis is small, as outlined in the following section. criteria. Additionally, it is possible to assess the expected costs
associated with a given threshold and therefore identify the
RBI strategies (quantitative risk-based strategies) optimal threshold (and thus the optimal inspection strategy), as
RBI strategies are based on the priorization of inspection illustrated in Figure 2. This optimization follows the principles
efforts in accordance with the risk reduction efficiency of the of the Bayesian preposterior decision analysis according to
different alternative inspection actions. Ideally, such strategies Raiffa and Schlaifer (1961).
should be based on the preposterior analysis of the Bayesian The probability calculations required for RBI are
decision theory. In the following, the RBI model presented in computationally very demanding, especially for fatigue
Straub (2004) and Straub and Faber (2006) is summarized and problems. In the past, this has hindered the application of the
subsequently applied for the numerical investigations; the RBI methodology in practice. The generic approach to RBI was
practical application of similar models has been reported in, developed to overcome these limitations. The core of the
e.g., Pedersen et al (1992), Moan et al. (2000), Faber et al. generic approach to RBI is the pre-fabrication of inspection
(2005) or Chakrabarti et al. (2005). plans for generic hot spots which are representative for the
The deterioration mechanisms are represented by particular hot spots 1 in the considered structures. These pre-
stochastic models of the defect size as a function of time, S (t ) . fabricated plans are termed Generic Inspection Plans. All hot
For fatigue, S (t ) is the crack depth and length as evaluated by spots that are represented by the model are fully described by
a probabilistic fracture mechanics based model. Inspection the so-called generic parameters. These are the input
qualities are commonly represented by a Probability of parameters to the model that vary from one hot spot to the next
Detection (PoD) curve and a Probability of False Indication and which are indicators of the relevant deterioration
(PFI), which describe the likelihood of an inspection outcome mechanism. For structures subjected to fatigue, typical
given the state of the inspected component. Based on the PoD
and PFI, structural reliability analysis or simulation techniques 1
Hot spots are the potential locations for fatigue failures.

2 Copyright © 2006 by ASME

examples of such generic parameters are, e.g., the calculated Threshold 10 yr
-4 -1
design fatigue life TFL (respectively the dimensionless Fatigue
40
Design Factor FDF 2 ), other loading characteristics (such as the

Inspection times [yr]

shape of the distribution describing the stress ranges at the hot etc.
30
spots), the applied SN curve (which is representative for the 2
nd
detail type and the environment) and geometrical parameters
20
such as the wall thickness at the hot spot. 1st inspection

10

7·10-3
0
Total cost Insp. cost
6·10-3 0 2 4 6 8 10
Failure cost Rep. cost
Fatigue Design Factor FDF
5·10-3
Expected cost

4·10-3 Figure 3. Inspection times as a function of the FDF for

a target reliability 10-4 yr-1, Straub (2004).
3·10-3

2·10-3

1·10-3 Threshold 10-4 yr-1

1
0
-2 -3 -4 -5 Total cost
10 10 10 10
10-1 Failure cost
Threshold ΔpFT [yr-1]

Expected cost
Inspection cost
10-2
Figure 2. Optimization of inspection efforts, from Repair cost
Straub (2004).
10-3
Once the generic inspection plans are calculated, the
10-4
inspection plans for the specific hot spots in a structure can be
obtained by an interpolation of the generic plans, see Straub
10-5
and Faber (2006) for details on the procedure. For this task,
0 2 4 6 8 10
software tools such as iPlan, see Faber et al. (2005), can be
Fatige Design Factor FDF
developed. Because the generic parameters are obtained from
standard fatigue evaluation procedures, the RBI can, in
Figure 4. Expected cost as a function of the FDF for a
principle, be performed without specialist knowledge once the
target reliability 10-4 yr-1, Straub (2004).
generic inspection plans are available. In this way, the RBI is
easily integrated in the daily asset integrity management
procedures of the owner or operator of the structure.
CHARACTERISTICS OF FATIGUE PERFORMANCE IN
As an example consider Figure 3: The inspections required
OFFSHORE STRUCTURES
to comply with given acceptance criteria are here shown as a
Offshore structures are subject to fatigue mainly due to
function of the generic parameter FDF, i.e., for fixed values of
environmental loads (waves). In addition, parts of the structure
all other parameters, the inspection times are obtained as a
are subjected to fatigue loads from machinery or other
function of the FDF. Similarly the expected costs can be
operational loadings. Typically, fatigue performance is assessed
expressed as a function of the FDF, Figure 4. The calculations
in terms of the fatigue design life (or the FDF) as calculated
are based on marginal costs of failure C F = 1 , cost of repair
using the SN approach. The FDF is a main indicator for the
C R = 0.01 , cost of inspection C I = 0.001 and an interest rate
fatigue performance and the required inspection efforts. In the
r = 0.05yr -1 .
following, we focus entirely on the FDF when describing the
fatigue performance of offshore structures. Other generic
parameters (such as the uncertainty in the load modeling) also
have a large influence on the fatigue performance and/or the
required inspection times, however, these other parameters
often are the same for the entire structure or do not vary much
from one hot spot to the next. For the comparative study
presented later, it is sufficient to assume that these other
2
The FDF is a deterministic safety factor, defined as the ratio of the
calculated design fatigue life to the design service life.

3 Copyright © 2006 by ASME

parameters are the same for all fatigue hot spots in the FPSO. For the conversion, the presented FDF values represent
structure. hot spots in 3 cargo tanks located in the aft part, the midship
part and the fore part respectively. The number of the web
Fixed steel structures frames in these 3 tanks is 9 (3 per tank) and the total number of
As an example, Figure 5 shows the distribution of FDFs welded connections is slightly higher than 1050. The fatigue
calculated for 4 steel jacket structures built in the late 1970s in calculations underlying these FDF values are described in
the Gulf of Mexico. The platforms are all eight-leg drilling Goyet et all. (2004). Based on the results of these calculations,
platforms with an anticipated service life of 35 to 36 years. The only 96 welded connections (about 9% of the total) were
fatigue calculations were performed as part of the reassessment considered in the detailed RBI analysis (all hot spots with
study described in Chakrabarti et al. (2005). A majority of the FDF<8). Thereby, the critical welded connections in this FPSO
hot spots have a FDF larger than 10. For those, no inspections are situated as follows:
will be required according to the RBI (see Figure 3), but also - Side shell longitudinals: 71
according to standards such as NORSOK (1998) or API (2002). - Longitudinal bulkhead longitudinals: 23
The distribution of the FDFs provides an indication of the - Bottom longitudinal: 1
fatigue strength of the installation and will be decisive for the - Bracket toe weld: 1
required inspection efforts following an RBI approach. In the second FPSO, the FDF values presented in Figure 6
are located in a condensate tank and a water ballast tank. The
35
~ 600 number of the web frames is 11 and the total number of welded
connections is higher than 1500. The fatigue calculations lead
30
to a selection of 294 welded connections (about 19% of the
total) for further detailed RBI analysis, which are situated as
Number of hot spots

25

follows:
20
- Side shell longitudinals: 74
15 - Longitudinal Bulkhead (side) longitudinals: 5
10 - Bottom longitudinals: 215
It is pointed at the fact that the critical connections are
5
located at different areas in the two FPSOs (in the side shell
0 longitudinals respectively the bottom longitudinals). This
0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9 - 10 > 10
indicates the difficulty in identifying the relevant hot spots for
FDF Range
inspection without detailed fatigue calculations.
Figure 5. Distribution of FDFs observed on 2 steel
~ 950 / 1200
jacket structures. 80
FPSO 1 (Conversion)
70
It is noted that the fatigue performances of the individual FPSO 2 (Purpose-built)
60
hot spots are dependent, in particular for similar types of details
Number of hot spots

in adjacent locations in the structure. This inter-dependency 50

allows considering system effects in the planning of 40

inspections, which may reduce the amount of required 30

inspections, see Straub and Faber (2005).
20

Floating structures 10

Floating structures include FPSOs (Floating Production, 0

0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 >8
Storage and Offloading units) but also semi-submersible
FDF Range
platforms. Whereas the fatigue assessment for floating
platforms can be considered in analogy to the assessment for Figure 6. Distribution of FDFs observed in 5 different
fixed structures, the fatigue assessment for FPSOs is different tanks on 2 FPSOs.
due to the large amount of hot spots (potential locations of
fatigue failures) in the structure and due to the large
redundancy of the structure. For the same reasons, the COST OF INSPECTION, REPAIR AND FAILURE OF
consideration of inter-dependency between the fatigue OFFSHORE STRUCTURES
performances of the individual hot spots is even more relevant The cost-relevant activities and events included in the
for floating structures than for fixed structures. optimization of the inspection efforts are the inspections
Figure 6 shows exemplarily the distribution of FDFs as themselves, with associated cost C I , repair actions with costs
calculated for 2 units. The first one is a conversion, i.e. a tanker C R and failure of individual hot spots with cost C F . Inspection
converted into a FPSO, while the second is a purpose-built

4 Copyright © 2006 by ASME

planning is a time-dependent problem, as inspections and (expected) cost of the failure of hot spot i is then derived from
repairs at an earlier time will prevent failures at a later time. the cost of structural collapse, CCOL , as:
Therefore, an interest rate r is included in the analysis. The
cost of the individual actions and events is highly dependent on C Fi = ⎡⎢ pCOL Fi (RIFi ) − pCOL F (RSR )⎤⎥ ⋅ CCOL (1)
⎣ ⎦
the type of structure and operation. In the following, the factors
influencing the cost of inspection, repair and failure are thus Equation (1) is based on the assumption that failures do not
discussed separately for steel platforms and FPSOs. occur in several hot spots simultaneously. This assumption does
only hold if general visual inspections are held in regular
Cost on offshore steel platforms intervals, ensuring the detection of failed members, see Straub
A large part of the cost of inspecting joints in steel and Faber (2003).
platforms is related to accessing the hot spots. This holds in
particular for the joints which are situated below sea-level and 1
in the splash zone. Below sea level, fatigue inspections require
removal of the marine growth, which is a time consuming task.

Annual collapse probability

In the splash zone, accessing the hot spots may be very difficult 10-1
or even impossible, depending on the weather conditions.
Because of the limited availability of ships and inspectors, the
inspections must be planned well in advance and cannot
10-2
account for the weather. A main cost factor is the ship which
carries both equipment and inspectors. Often a Dynamic
Positioning Ship is required, with associated cost in the order
of 10’000US$ per day. When inspecting joints below sea-level, 10-3
approximately 8 joints may be inspected per day, so that, as a
rough estimate, it can be assumed that the total cost of
inspecting a hot spot, including ship, personnel and equipment 10-4
cost, is C I = 2’000US$. 0 0.2 0.4 0.6 0.8 1
For repair actions, the cost factors are similar to those for Residual Influence Factor RIF
inspections, but, depending on the type of repair, the required
time to perform a repair can vary substantially. If a small crack Figure 7. Relation between the RIF and the annual
is found, it can be removed simply by grinding, which may be probability of collapse for a steel jacket in the Gulf of
performed directly by the inspector in very little extra time. If a Mexico with an RSR = 1.63.
larger defect is found, the necessary equipment and personnel
may not be available and must be brought to place at an extra The cost of platform collapse, CCOL , is difficult to
cost. Alternatively, an engineering assessment and follow-up estimate. Construction costs of typical offshore steel platforms
inspections may be performed. Considering these factors, it is are in the order of 20-30 106 US$. However, if the loss of
estimated that the average cost of repairing a major defect is production cannot be compensated by other installations in the
C R = 20’000US$. field, it may result in costs of an order of magnitude higher. As
In addition to the direct cost, the inspection and repair an example, the cost of the loss of Petrobras’ P-36 semi-
activities represent a significant risk for the involved personnel, submersible rig in 2001 has been estimated as 500 106 US$,
in particular when joints are inspected by divers. This risk Goldman Sachs (2004). In addition, catastrophic events may
should be taken into account when assessing compliance of an also lead to a loss of reputation, which is very difficult to
inspection/maintenance strategy with given risk acceptance quantify. On the other hand, most structures are insured and
criteria. failure costs are, therefore, compensated. However, because the
The cost of failure of a hot spot, C F , depends on the insurance premiums will depend on risk mitigation actions
importance of the associated members. The importance of implemented by the operator (at least theoretically), it is argued
member failure can be expressed by the conditional probability that compensation by insurance companies should not be
of global collapse given member failure, pCOL Fi . This included in the cost-benefit analysis. For the examples
probability can be estimated by the use of a simple indicator; presented in the latter, two cases are considered: CCOL = 30 106
the Residual Influence Factor (RIF), see Stahl et al. (2000) and US$ and CCOL = 300 106 US$. The cost of a hot spot failure,
Straub and Faber (2005a). The RIF is defined as the ratio calculated by applying Equation (1) and the relation given in
between the Reserve Strength Ratio (RSR) of the intact Figure 7, is given in Table 1 for different RIFs:
structure and the RSR of the damaged structure, which is
assessed by removing the element i in the pushover analysis.
Using a general probabilistic model it is possible to relate the
RIF for a particular hot spot to pCOL Fi , Figure 7. The

5 Copyright © 2006 by ASME

 Table 1. Expected cost of hot spot failure as a unavailability (down time) of the corresponding tank for a
function of the RIF and the cost of collapse. period of about 10 days. Assuming that the loss of production
is equal to 10’000 bbl/day at a rate of 30US$ per bbl, the loss
RIF CCOL = 30 106 US$ CCOL = 300 106 US$ of production over 10 days amounts to: 10’000 x 10 x 30US$ =
3’000’000US$.
0.95 6’000 US$ 57’000 US$ In the above it is assumed that a component fatigue failure
is followed by an immediate repair in order to prevent any
0.9 14’000 US$ 138’000 US$ global failure scenario with much higher consequences. The
implicit assumption – which is not always realistic – is that the
0.8 43’000 US$ 433’000 US$ failure is immediately detected and repaired once it occurs. In
principle, also the influence of fatigue failures on the overall
0.5 688’000 US$ 6’879’000 US$ structural integrity should be (explicitly) accounted for. Clearly,
fatigue failures lead to an increased risk of loss of the entire
unit, as each failed connection will decrease the global
Note that fatalities are not considered here, as these must structural resistance to operational and environmental loads and
be taken care of by risk acceptance criteria, which are based on thus increase the probability of structural collapse. However,
the preferences of society, see Rackwitz (2002) and Kübler and due to the high redundancy in floating structures, it is not
Faber (2002). possible to assess the member importance with a RIF value
(RIF values would generally be very close to 1). On the other
Cost on FPSOs hand, fatigue failures will be highly inter-dependent and may
In analogy to platforms, a major contribution to inspection occur in clusters, which increases the probability of a global
and repair costs is from assessing the hot spots. The costs for collapse of the structure, see also Straub and Faber (2005a) for
the inspection of an example tank are: a discussion. It seems therefore not reasonable to quantify the
- 8000US$ for COW (Crude Oil Washing), water wash and cost of an individual failure event in a floating structure, as it
to purge/gas free the tank, activities which are required for has been done for platform structures. Instead, as a first
accessing the tank. approximation it is assumed that all fatigue failures will be
- 4000US$ for a visual inspection of the lower area. detected and immediately repaired. Because this assumption is
- 70’000US$ for visual inspection of the upper area using over-estimating the true costs (not all defects will be detected
rope access. and repair may take place during the regular inspection
- 5000US$ for Non-destructive testing (NDT) of 10 hot spots campaign), it is considered that this partly accounts for
in the tank. neglecting the effect of the failure on the overall structural
Note that these costs are not exclusively related to fatigue integrity. The cost of a failure is thus taken as C F = 3’000’000
inspections, but include inspections for other types of US$.
degradation (such as painting/coating checks and thickness For some of the hot spots, special failure scenarios become
measurements). Assuming that all other inspections are fixed, relevant, e.g. when the fatigue failure is located on the side
the cost which is associated with the inspection of one hot spot shell and may cause pollution of the environment or for the
may be estimated as C I = 500US$. case where a fatigue failure may trigger an explosion due to the
The cost of a repair is related to the cleaning and the repair emission of gas, as may be the case for the walls separating the
action itself. For an example hot spot, the related costs are: cargo tanks and the water ballast tanks. In that case, these
- 2500$ for the local cleaning before the repair. failure costs may be considered explicitly, following a
- 1500$ for a repair including a drill stop, gouging and probabilistic consequence assessment.
welding.
- 15’000$ for an insert repair (replacing a steel plate). Note BENEFITS OF RISK BASED INSPECTION PLANNING
that this cost is dependent on the size of the area which is The benefits of performing RBI as compared to
repaired. prescriptive inspection planning are assessed for examples of
For a minor repair, the associated costs are therefore fixed offshore steel structures and FPSOs.
approximated by C R = 4’000US$.
A main factor influencing the cost of failure is the Fixed steel structures
unavailability of a tank. Whereas inspections (including In the following, the expected cost of different inspection
subsequent repair) can be planned, failures may lead to an planning strategies are assessed, based on the probabilistic
unplanned, immediate shut-down of parts of the installation. deterioration model described in Faber et al. (2005) and the
The cost of such shut-downs is highly depending on the cost model presented in the previous section as it applies to
operation, but in any case these costs will be huge. As a simple platforms. It is assumed that the hot spot has a RIF=0.9. The
example consider a fatigue failure occurring between two expected cost of failure ( C F =138’000 US$), the cost of
consecutive inspection campaigns, which leads to an inspection ( C I =2’000US$) and the cost of repair

6 Copyright © 2006 by ASME

( C R =20’000US$) follow from the pervious discussion. Figure 11 and Figure 12 show the expected cost for two
Furthermore, an interest rate of r = 0.03 is taken into account. different prescriptive inspections plans, with constant
In Figure 8 and Figure 9, the expected cost over the life- inspection intervals of 10 and 4 years. It is reminded that only
time (40 years) of a hot spot with FDF=2 is presented, for a the second fulfills the acceptance criteria for hot spots with a
prescriptive and a risk-based strategy respectively. The RBI FDF=2 (but not with a FDF=1). Figure 13 presents the
strategies are given as a function of the maximum annual expected cost for an RBI plan, which fulfills the acceptance
probability of failure, in accordance with Figure 1 and Figure criteria for all FDFs. It can be observed that the RBI plans lead
2. The inspection times for ACFM inspections corresponding to to significantly lower total expected cost than the prescriptive
the RBI strategy are shown in Figure 10. plans for almost the entire range of FDFs, while at the same
Although the difference between the prescriptive and the time ensuring compliance with the acceptance criteria.
RBI strategies with respect to the expected cost appears small, Based on the expected costs presented in Figure 11 to
this is not the case if it is reminded that the inspection plans Figure 13 and the distribution of FDFs for the 4 sample
must also fulfill specified acceptance criteria. In the application platforms as presented in Figure 5, the total expected costs
presented in Faber et al. (2005), for a hot spot with RIF=0.9, E [CT ] for the different inspection planning strategies can be
risk acceptance criteria demand that the annual probability of computed. These are
failure is lower than 10-3 yr-1. For the example hot spot, when - 10 years fixed interval: E [CT ] = 5.2 106 US$
applying a prescriptive inspection plan, this would require that - 4 years fixed interval: E [CT ] = 11.7 106 US$
the inspection interval is 4 years or lower. In this case, the total - RBI plan: E [CT ] = 3.8 106 US$
expected cost is, according to Figure 8, 17’000US$. The RBI It is observed that the application of the RBI plans result in the
plan fulfilling the acceptance criteria has a total expected cost savings of 1.4 106 US$, respectively 7.9 106 US$ as compared
of 12’500US$, Figure 9. to the different prescribed inspection plans.

$25'000 $25'000

$20'000 Total cost $20'000 Total cost

Expected Total Cost

Expected Total Cost

Failure cost Failure cost

$15'000 $15'000
Repair cost Repair cost
$10'000 $10'000
Inspection cost Inspection cost

$5'000 $5'000

$0 $0
20 15 10 5 0 0 2 4 6 8 10
Inspection interval FDF

Figure 8. Expected costs for a hot spot with FDF=2 Figure 11. Expected cost as a function of the FDF
for different equidistant inspection strategies. when performing inspections in a constant interval of
10 years.
$25'000
$25'000
$20'000 Total cost
Expected Total Cost

$20'000 Total cost

Expected Total Cost

Failure cost
$15'000
Failure cost
Repair cost $15'000
$10'000 Repair cost
Inspection cost
$10'000
Inspection cost
$5'000
$5'000
$0
1.E-02 1.E-03 1.E-04 1.E-05 $0
Threshold on the annual probability of failure Δpf 0 2 4 6 8 10
FDF

Figure 9. Expected costs for a hot spot with FDF=2

for different RBI strategies. Figure 12. Expected cost as a function of the FDF
when performing inspections in a constant interval of
4 years.
ar
Ye

36
23

24

25

26

27

28

29

30

31

32

33

34

35
10

11

12

13

14

15

16

17

18

19

20

21

22
0

Threshold
0.01
0.001 X X X X
0.0003 X X X X X X X X
0.0001 X X X X X X X X X X X
0.00003 X X X X X X X X X X X X X X X

Figure 10. Inspection plans corresponding to

different thresholds for a hot spot with FDF=2.

7 Copyright © 2006 by ASME

 $25'000 $14'000

$12'000
$20'000 Total cost Total cost

Expected Total Cost

Expected Total Cost

$10'000
Failure cost Failure cost
$15'000 $8'000
Repair cost Repair cost
$10'000
$6'000
Inspection cost Inspection cost
$4'000
$5'000
$2'000

$0 $0
0 2 4 6 8 10 1.E-02 1.E-03 1.E-04 1.E-05
FDF Threshold on the annual probability of failure Δpf

Figure 13. Expected cost as a function of the FDF Figure 14. Expected costs for a specific connection
when applying a RBI strategy with a threshold on the on FPSO 1 (with FDF=2.41) and various thresholds.
annual probability of failure of 10-3 per year.
Due to the fact that the times for fatigue inspection for the
most critical set of components were found close to the usual
FPSO (Example 1) inspection times (years 2.5, 5, 7.5, 10, 12.5 and 15), it was
Similar calculations as performed in the previous section decided to fit the fatigue inspections to the usual inspection
for fixed steel offshore structures may also be carried out for campaigns required by the class. In addition, it was possible to
floating units. The total expected costs of an RBI plan are reduce the amount of inspections by considering systems
thereby compared with the total expected cost related to the effects (see, e.g., Straub and Faber (2005b)): As a consequence,
application of prescriptive rules. In the case of FPSOs, only 50% of the NDT inspections must be performed every 2.5
prescriptive rules are the rules issued by the classification years. Therefore, only 50% of the wing tanks are inspected
societies for maintaining class. According to Bureau Veritas every 2.5 years, and the inspection interval for those tanks is
rules (2004), a special survey has to be carried out every fifth thus increased to 5 years. Only a reference group of cargo tanks
year when the unit is younger than 15 years. For older units, will be inspected directly according to the plan coming from
the required inspection interval is reduced to 2.5 years. As may the detailed RBI (as in Table 2), to verify the assumptions made
be observed from Table 2, the required inspection times regarding system effects. For ballast tanks, system effects were
(applying Alternate Current Field Measurements, ACFM) vary not considered.
significantly when applying a RBI strategy, although a direct
comparison is not valid, because inspections must also be FPSO (Example 2)
carried out for assessing other degradation, in particular FPSO 2 is not under the class regime for in-service life.
corrosion. Because a large part of the inspection cost is related Therefore, a RBI approach has been applied for determining
to assessing the hot spots, the expected cost related to the inspection plans for the unit. Based on FDF values shown in
different inspections must be considered jointly. However, a Figure 6, the so called “equidistant RBI approach” (see Faber
RBI study can be performed including all different types of et al., 2000) was used to determine the optimal periodicity of
degradation. inspection campaigns, i.e., the inspection intervals required for
Note that the RBI calculations for the FPSOs are based on fulfilling the acceptance criteria. The distribution of the
the probabilistic models presented in Goyet et al. (2004). inspection intervals calculated for two different tanks is
presented in Figure 15. It is required to distinguish the two
Table 2. Inspection times [yr] for 7 representatives tanks under consideration:
FDF values on FPSO 1. For the condensate tank almost all of the hot spots require
FDF 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 an inspection each 5 years or each 4.3 years.
1.58 For the water ballast tank, most of the hot spots require an
2.51
3.50 inspection each 5 years or each 10 years, but a small number of
4.27 hot spots - basically the ones at the bottom shell connections -
5.54
6.45 require more frequent inspections.
7.45 The inspections times were determined for close visual
inspection, which is the usual way of inspection in maritime
transportation. In a second step, other, more accurate NDT
As an example, the expected cost for a hot spot with techniques with higher probability of detection (PoD) were
FDF=2.4 and service life time 15yr is shown in Figure 14. The used to extend the frequency of inspection of the most critical
costs as utilized in the calculations follow from the previous components. This illustrates the flexibility of RBI, which
discussion and are C F =3’000’000 US$, C I =500US$ and allows for adaptations and modifications when required.
C R =4’000US$, the interest rate is r = 0.03 .

8 Copyright © 2006 by ASME

 thought, it would also be highly beneficial to perform a RBI
70%
study already during the design of a new structure. Considering
Tank 1 (Condensate tank) the examples presented in this paper, it would, e.g., be possible
60%
Tank 2 (Water ballast tank) to identify a cost optimal FDF values for the hot spots, when
50% the construction cost is included in the analysis.
Fraction of hot spots

For most new-built fixed offshore structures, fatigue

40% calculations are performed at the design stage for all hot spots.
For existing structures, however, such calculations are often not
30% available and are thus carried out as part of the RBI study. For
FPSOs, fatigue calculations are generally available only for a
20%
selection of hot spots which are considered critical, although
there is a tendency towards demanding more extensive fatigue
10%
assessments. Independently of whether or not fatigue
0%
calculations are available for all hot spots, RBI procedures are
1.0 2.0 2.5 3.3 4.3 5.0 10.0 based on a model of the structure which is not perfect. This is
Required inspection interval [yr] reflected by the fact that fatigue cracks occur at hot spot areas
where they were not expected. It is thus of utmost importance
Figure 15. Distribution of the inspection (close visual) that besides the detailed inspections planned according to RBI,
periodicity as determined for 2 tanks on FPSO 2. general (visual) inspections are performed to ensure the validity
of the assumptions made in the RBI analysis. The inspection
planning procedures must then ensure that fatigue calculations
DISCUSSION and consequently the RBI analyses are revised when such
In the paper, the benefits of RBI are presented for general inspections reveal defects which were not anticipated.
examples of fixed steel structures as well as floating structures It is noted that RBI is increasingly required by owners and
(FPSOs). For fixed structures it has been demonstrated that the operators of offshore structures. For owners and operators, the
application of RBI plans may lead to significant economical primary objective of RBI is to maximize safety, to minimize
benefits, which are in the range of one to several million US$ costs and to gain a technical understanding of the behavior and
for the four example platforms, depending on the alternatively performance of the facilities. It can be observed from the calls
applied prescribed inspection plans. In addition, the RBI plans for tenders that operators are moving from a reactive to a
ensure that the acceptance criteria with respect to risk to proactive vision through the introduction of risk-based,
personnel and the environment are fulfilled, which is not the optimized inspection programs. In the case of FPSOs, risk-
case when prescriptive inspection planning is performed. For based schemes are implemented as an alternative to or in
FPSOs, it has been outlined on two examples that the parallel with class inspection rules and flag state requirements.
application of RBI plans facilitates the targeted application of This is also true for fixed steel offshore structures, where RBI
inspections. Only the detailed fatigue calculations, which are and certification are related integrity management tools.
performed as part of the RBI study, allow the identification of Furthermore, many call for tenders nowadays explicitly require
the critical details (hot spots) in the structure. These have been the demonstration of compliance with risk acceptance criteria.
found to be different ones for the two considered FPSOs (the These prescribe that risks to personnel arising from
side shell longitudinal connections in the conversion FPSO and consequences of loss of pressure containment or structural
the bottom longitudinal connections in the purpose-built failure have to be retained below some limits as specified by
FPSO). On this basis, a significant reduction of inspection national authorities or operators and owners. Such acceptance
efforts on the non-critical elements has been achieved, while criteria may be expressed for example in terms of average FAR,
ensuring sufficient inspection coverage of the critical hot spots. location-specific FAR or individual risk, Straub and Faber
The examples underline the fact RBI allows to fit inspection (2005a). It is clear that only inspection plans based on RBI are
efforts to the requirements in terms of acceptance criteria for able to demonstrate compliance with the acceptance criteria.
each component individually and thus leads to a significant
(economical) benefit. CONCLUSIONS
In addition to a direct reduction of the total expected cost, By consideration of examples of fixed offshore steel
RBI enhances the understanding of the structural integrity. structures and FPSOs subject to fatigue deterioration, the
Because RBI requires a detailed analysis of the structure, the benefits of RBI are presented and discussed. It is demonstrated
deterioration processes as well as the inspection performances, that the financial benefits in terms of the expected total life-
it helps to identify the “weak points” of the structure. For some cycle cost are huge. Furthermore, it is observed that only RBI
structures, the RBI study may thus result in a recommendation plans ensure compliance of the structures with risk acceptance
for additional mitigation measures, which are more efficient criteria and its documentation. Finally, the analyses required by
than an increased inspection effort. Following the same line-of- RBI enhance the understanding of the relevant degradation

9 Copyright © 2006 by ASME

processes and may thus lead to an improved structural integrity NORSOK N-004 (1998). Design of Steel Structures. Standard
management beyond the optimization of inspection activities. N-004, revision 1, available online: www.nts.no/norsok/
Pedersen C., Nielsen J.A., Riber J.P., et al. (1992). Reliability
Based Inspection Planning for the Tyra Field. Proc. 11th
REFERENCES OMAE, Vol. 2, pp. 255-263.
API (2002). Recommended Practice for Planning, Designing, Pemex (2000). Inspección de Plataformas Marinas Fijas de
Constructing Fixed Offshore Platforms – Working Stress Acero. Petróleos Mexicanos, Technical Specification
Design. American Petroleum Institute, Recommended P.6.0131.01.
Practice 2A-WSD, 21st edition. Rackwitz R. (2002). A Rational Risk Acceptability Criterion
Bureau Veritas (2004). Rules for the Classification of Steel Based on the Life Quality Index. Proc. 21st Offshore
Ships, Part A: Classification and Surveys. Document NR Mechanics and Arctic Engineering Conference, Oslo,
467, A1 DTM R04 E, Bureau Veritas, November 2004. paper S&R-28430.
Chakrabarti P., Mukkamala A., Abu-Odeh I., Majumdar B., Raiffa H., Schlaifer R. (1961). Applied Statistical Decision
Faber M.H., Straub D., De la O J. (2005). An Overview of Theory. Cambridge University Press, Cambridge, Mass.
the Reassessment Studies of Fixed Offshore Platforms in Skjong R. (1985). Reliability Based Optimization of Inspection
the Bay of Campeche, Mexico. Proc. 24th OMAE, Greece. Strategies. Proc. ICOSSAR 85, Vol. III, Kobe, Japan, pp.
Dalane J.I., Skjong R., Lotsberg I. (1990). Optimal Fatigue 614-618.
Design of Offshore Structures. Proc. 9th OMAE, Houston, Stahl B., Aune S., Gebara J.M., Cornell C.A. (2000).
Vol. 2, pp. 73-79. Acceptance Criteria for Offshore Platforms. Journal of
Faber M.H., Engelund S., Sørensen J.D., Bloch A. (2000). Offshore Mechanics and Arctic Engineering, 122(3), pp.
Simplified and Generic Risk Based Inspection Planning. 153-156.
Proc. 19th OMAE, New Orleans. Straub D. (2004). Generic Approaches to Risk Based
Faber M.H., Sørensen J.D., Tychsen J., Straub D. (2005). Field Inspection Planning of Steel Structures. PhD thesis, Swiss
Implementation of RBI for Jacket Structures. Journal of Federal Institute of Technology, ETH Zurich.
Offshore Mechanics and Arctic Engineering, Trans. Straub D., Faber M.H. (2005a). Risk Based Acceptance
ASME, 127(3), pp. 220-226. Criteria for Joints Subject to Fatigue Deterioration.
Fujita M., Schall G., Rackwitz R. (1989). Adaptive Reliability- Journal of Offshore Mechanics and Arctic Engineering,
Based Inspection Strategies for Structures Subject to Trans. ASME, 127(2), pp. 150-157.
Fatigue. Proc. 5th ICOSSAR, Vol. 2, San Francisco, pp. Straub D., Faber M.H. (2005b). Risk Based Inspection
1619-1626 Planning for Structural Systems. Structural Safety, 27(4),
Goldman Sachs (2004). Global Energy. Energy Environmental pp. 335-355.
and Social Report, Goldman Sachs Global Investment Straub D., Faber M.H. (2006). Computational Aspects of Risk-
Research, February 2004. Based Inspection Planning. Computer-Aided Civil and
Goyet J., Rouhan A., Faber M.H. (2004). Industrial Infrastructure Engineering, 21(3), pp. 179-192.
Implementation of Risk Based Inspection Planning
Methods – Lessons Learnt from Experience: The Case of
FPSO’s. Proc. 23th OMAE, Vancouver.
Kübler O., Faber M.H. (2002). Optimality and Acceptance
Criteria in Offshore Design. Proc. 21st Offshore
Mechanics and Arctic Engineering Conference, Oslo,
paper S&R-28427.
Lotsberg I., Sigurdsson G., Wold P.T. (1999). Probabilistic
Inspection Planning of the Åsgard A FPSO Hull Structure
with Respect to Fatigue. Proc. 18th OMAE, New
Foundland, Canada, pp. 259-266.
Madsen H.O. (1987). Model Updating in Reliability Theory.
Proc. ICASP5, Vancouver, Canada, pp. 565-577.
Madsen, H.O., Sørensen J.D., Olesen R. (1989). Optimal
Inspection Planning for Fatigue Damage of Offshore
Structures. Proc. 5th ICOSSAR, San Francisco, pp. 2099-
2106.
Moan T., Vårdal O.T., Johannesen J.M. (2000). Probabilistic
inspection planning of fixed offshore structures. Proc.
ICASP8, Sydney, pp. 191-200.

10 Copyright © 2006 by ASME