Aven CH323.

tex 17/5/2007 11: 42 Page 2543 Risk, Reliability and Societal Safety – Aven & Vinnem
(eds) © 2007 Taylor & Francis Group, London, ISBN 978-0-415-44786-7 Vibration fatigue in process
pipework: A risk based assessment methodology R. Swindell Bureau Veritas, Soutampton, UK
ABSTRACT: Risk Based Inspection (RBI) has been used for some time to optimise inspection
schedules based on assessing the risk of various corrosion related degradation mechanisms.
However, one degradation mechanism that has been traditionally overlooked is that of vibration
induced fatigue. A proactive methodology for assessment of vibration induced fatigue was
published in 2000 as part of a Joint Industry Project to identify, using a risk based approach,
which systems would be at risk from vibration induced fatigue. This paper describes the use of
the overall risk based methodology and the key learning points in its application over the last 6
years. 1 INTRODUCTION Vibration induced fatigue has always been recognised as a potential
failure mode of process pipework, although the problem has usually been addressed on a
reactive, or ‘fix-as-fail’, basis – the only exception being the guidance given in API618 and
API674 (American Petroleum Institute 1995a, 1995b) for reciprocating compressors and pumps. In
the early 1990’s a sudden increase in pipework failure due to vibration induced fatigue was
noted in the North Sea, especially on new offshore installations where extensive use had been
made of duplex alloys. This resulted in thinner walled and lighter weight piping systems which
were more susceptible to vibration related fatigue problems. More recently there has also been
an increasing trend of vibration related problems on existing or mature assets (both onshore and
offshore) where process or operational changes have been made. Typical examples where
problems have been experienced include: – During commissioning of a new plant – When
changing the operating conditions on an existing plant (e.g. debottlenecking) – Transient events
over an extended period In the UK, the Health and Safety Executive has collated data on
reported hydrocarbon leaks on offshore installations. The results have been published for the
period April 2000 to March 2001 (Health and Safety Executive 2002), and the results, in terms of
‘the immediate causes of releases’ are summarised in Table 1. Table 1. Causes of failure offshore
(from Health and Safety Executive 2002a). Total reported failures Description % Degradation of
material properties∗ 28 Incorrect installation 21 Fatigue/vibration 21 Corrosion/erosion 13 ∗E.g.
loss of flexibility in flange gaskets and valve stem packing. In addition, it was found that ‘... the
majority of releases involved gas, and occurred during normal production. Pipework was the
main release site, with small bore piping a significant contributor’. (Health and Safety Executive
2002). One North Sea operator has quantified the cost of a ‘typical’ failure to be approximately
€0.1M, based on lost production and subsequent repair costs. 2 FACTORS INFLUENCING
VIBRATION INDUCED FATIGUE For a vibration problem to occur, there must be (a) some form of
excitation, (b) response of the main line to that excitation, causing either fatigue damage to the
main line itself, or (c) damage to small bore connections (SBC’s) on that main line. This ‘failure
chain’ is shown in Figure 1. 2543 Aven CH323.tex 17/5/2007 11: 42 Page 2544 Vibration of main
line Vibration of small bore connections Excitation • flow turbulence • mechanical • pulsation •
acoustic • transient • pipe diameter • wall thickness • support • length of connection • type of
connection • no. of valves • diameter • parent pipe wall thickness • location Figure 1. Vibration
induced fatigue failure chain. Table 2. Typical vibration excitation mechanisms. Steady state
excitation – Flow induced turbulence – Flow induced pulsation – Pulsation due to rotating stall on
centrifugal compressors – Pulsation due to reciprocating pumps & compressors – Mechanical
excitation Transient excitation – Valve opening (acoustic fatigue, flashing, cavitation) – Valve
closure (surge, momentum effects) 2.1 Excitation mechanisms The most common excitation
mechanisms can be subdivided into two main categories – those occurring under ‘steady state’
operation, and those which are experienced under transient conditions (e.g. plant start-up or
shutdown). The majority of these mechanisms are sensitive to process and operational changes
– for example, increasing the flow velocity in a line by 20% will increase the level of turbulent
energy by between 40% and 50%. In addition, key aspects associated with the geometry of
each main line and small bore connection influences the level of fatigue damage that will
occur for a given level of excitation; these are shown in Figures 2 and 3. 3 STRATEGIES FOR
DEALING WITH VIBRATION INDUCED FATIGUE Historically, there are three common strategies
which have been adopted in an attempt to overcome the issue of vibration induced fatigue of
process pipework. Each has advantages and disadvantages. 3.1 Bracing of all small bore
connections The principal advantage of bracing all small bore connections is that it is a ‘one off
’ exercise. However, Figure 2. Factors influencing main line response. Figure 3. Factors influencing
small bore connection response. if undertaken retrospectively (i.e. on an operational asset) then
the cost can be high, especially when access and insulation removal and reinstatement costs
are included. In addition, the long term maintenance issues associated with bolted braces need
to be taken into account. 2544 Aven CH323.tex 17/5/2007 11: 42 Page 2545 Finally, this
approach does not address main line integrity problems (i.e. fatigue damage at main girth
welds). 3.2 Visual survey A visual survey on an existing plant has the key advantage that it
captures the true ‘as built’ condition of the pipework. This is the only way that potential problems
associated with local construction issues can be identified (for example, bracing of a
connection to the neighbouring deck structure rather than the parent pipework). However, a
visual survey can be very subjective. It also cannot identify potential issues at the design stage of
a new plant. 3.3 Vibration measurement survey Another approach has been to identify
problems using some form of vibration measurement survey. Whilst this approach can yield
useful information and can identify any ‘immediate’ fatigue issues, it only provides a ‘snapshot’
of the health of the process plant for the process and operational conditions experienced at the
time of the survey. Unfortunately, vibration of process pipework is very dependent on a variety of
process parameters (e.g. flow velocity, density, gas molecular weight etc) and how the plant is
being operated. 4 OVERVIEW OF A RISK BASED APPROACH Risk Based Inspection (RBI) has been
used for some time to optimise inspection schedules based on assessing the risk of various
corrosion related degradation mechanisms. For example, API581 (American Petroleum Institute
2000) provides a risk based methodology for corrosion mechanisms in refinery services. It is
interesting to note that API581 states that ‘... fatigue failures of piping systems present a very real
hazard under certain conditions...’ Following a similar logic, a proactive approach has now
been developed which identifies systems will be at risk from vibration induced fatigue. The
method can be applied at the design stage, or prior to commissioning, or when changes are
being contemplated for an operational asset (e.g. debottlenecking). The risk based approach
draws on a Joint Industry Project (Marine Technology Directorate 1999) (the ‘MTD Guidelines’)
completed in the UK in 2000 for ‘steady state’ excitation mechanisms, and also specific contract
work undertaken for the UK’s Health & Safety Executive (Health and Safety Executive 2002) for
‘transient’ excitation mechanisms. Both of these documents are in the public domain. Figure 4.
Overall schematic of the risk based approach. The risk based methodology uses a 3 stage
screening technique to identify whether any vibration related problems will occur, and where in
the process system they will be experienced (see Figure 4 for a simplified schematic). The 3 stage
approach is built on the principle that in order for a small bore connection to fail due to vibration
there must be: 1. some form of excitation (addressed in Stage 1), causing 2. vibration response of
the main line (addressed in Stage 2), causing either a main line fatigue issue, or 3. vibration of
the small bore connections on that line (addressed in Stage 3). The information required can be
obtained from Process Flow Diagrams (fluid densities, pressures and flow rates) and P&IDs (pipe
diameters and wall thicknesses). The assessment process identifies which main lines and
associated small bore connections (defined as any connection less than 2 NB) are at risk of
failure for the complete operational and process envelope, and what remedial actions (either in
terms of modification or inspection) need to be undertaken. The data required for an
assessment is very similar to the information required to implement RBI, although the output
results largely in a modification strategy rather than an inspection strategy as shown in Figure 5.
The risk based approach to vibration induced fatigue therefore identifies where modifications
need to be made (using detailed analysis methods as appropriate), and where on the plant
inspection or vibration survey effort needs to be targeted. Modifications could include changes
to main line supporting arrangement, removal or bracing of high risk small bore connections, or
changes to the process system to reduce the level of excitation. It should be noted that the
Likelihood of Failure (LOF) rating is a semi-quantitative value that represents the probability of a
certain failure occurring. To obtain a true ‘risk’ assessment, the consequence 2545 Aven
CH323.tex 17/5/2007 11: 42 Page 2546 ProProcess and Structural Information Risk Based
Inspection (corrosion degradation mechanisms) Risk Based Vibration (mechanical fatigue
degradation mechanisms) Inspection Strategy Modification Strategy Inspection Strategy
Modification Strategy Figure 5. Integration with risk based inspection. of that failure occurring
would also need to be taken into account. The MTD Guidelines methodology was originally
targeted at the design phase of new process plant, although since publication the
methodology has also been applied to assessments of existing process plant. 5 CASE STUDIES 5.1
Case Study 1: New offshore installation The risk assessment was applied at the end of the
detailed design stage, during construction. The complete hydrocarbon system was assessed, as
well as several process and safety critical systems. The results indicated that 3 additional main
line supports were required, and that 38 small bore connections (out of a total of 990) required
modification. These modifications were implemented, and the installation successfully
commenced production in early 2002, achieving its design throughput with no vibration related
issues. 5.2 Case Study 2: Existing offshore installation In this case, the risk assessment was applied
to a mature asset where a number of problems had been experienced previously. A bottom up
approach was adopted, with the small bore connection assessment being undertaken first. This
identified 261 small bore connections of poor geometry, with an average modification cost of
€4500 per connection (based on the mantime, material, insulation removal/reinstatement and
access costs). A Stage 1 and 2 assessment was subsequently undertaken to identify which of
these connections were located on lines which would have potentially sufficient excitation to
cause a fatigue problem. The results (shown in Figure 6) showed that modifications to only 29
connections were required. Figure 6. Case Study 2: Small bore connection assessment results. 5.3
Case Study 3: Existing onshore gas plant A risk assessment was undertaken for a gas plant where
debottlenecking was being considered. The scope of the study encompassed 2 modules, each
with 171 main lines which would be subjected to increased excitation due to the increased
throughput. The results of the assessment indicated that 48 small bore connections per module
required modification. The plant subsequently achieved its higher throughput with no vibration
fatigue issues. Further details of this particular example are given in Martin & Swindell et al 2002. 6
COST BENEFIT ANALYSIS A cost benefit analysis has been conducted for two projects – a refinery
and an offshore installation. See Table 3. The assumptions made in this analysis relate to the rate
of return used (3%), and that a 90% success rate is achieved through the application of the
methodology (this is based on the findings of a ‘blind’ assessment conducted for Shell in the UK).
The final NPV calculation takes account of the costs associated with the risk assessment and the
subsequent modification costs. 7 KEY LEARNING POINTS Since publication the methodology
given in the MTD Guidelines has been applied to over 60 process plants and has played a
significant role in helping to reduce pipework failures: – 40 offshore installations (United Kingdom
Continental Shelf) – 8 offshore installations (non-UKCS) – 8 downstream plants (gas plants and
refineries) – 6 petrochemical/chemical plants 2546 Aven CH323.tex 17/5/2007 11: 42 Page 2547
Table 3. Cost benefit examples. Average Estimated saving based on 90% Cost of risk NPV cost of
1 No of success rate assessment Cost of (over 3 failure failures Cost project modifications years)
Project € /y €/y Year 1 Year 2 Year 3 € € € Refinery 55,000 3 165,000 148,500 148,500 148,500
100,000 153,000 167,049 Offshore installn. 100,000 2 200,000 180,000 180,000 180,000 110,000
145,000 254,150 It was always the intention that the original MTD Guidelines would be reviewed
after a period of usage and update as necessary. During the last 6 years, very useful experience
has been gained in the practical application of the guidelines and associated methodologies.
Constructive comments have also been received regarding aspects for improvement,
principally: – Procedural aspects – improving the overall usability of the guidelines to provide a
clear holistic approach to the ‘through life’ management of vibration induced fatigue in process
pipework. – Technical aspects – inclusion of excitation mechanisms not covered previously and
extending the scope of the assessment methods. In addition, it was identified that the Guidelines
could be overly conservative in certain situations, and possibly not conservative in others. In 2005
a new Joint Industry Project was therefore initiated by the Energy Institute (which had acquired
the copyright to all MTD documents) with a number of sponsors including Shell, BP, Total,
ChevronTexaco, ConocoPhillips and the UK’s Health & Safety Executive. The revised Guidelines,
due to be published in March 2007, include a number of excitation mechanisms not previously
addressed including transient events and intrusive elements (e.g. thermowells). In addition, the
new Guidelines include a new qualitative (‘simplified’) screening assessment that can be used
by a non-specialist, as well as the updated quantitative methods for more detailed assessments,
thus mirroring the approach adopted in API581. The Guidelines now explicitly address the issue of
how to apply the various techniques in different scenarios (i.e. a new design, for plant change
and for troubleshooting existing vibration issues). 8 CONCLUSIONS Vibration induced fatigue of
process piping systems has been shown to be a tangible threat, accounting for approximately
21% of failures of pipework offshore. A risk based approach to addressing the problem of
vibration induced fatigue has been described, including its integration with other risk based
approaches to integrity management (e.g. RBI). The key benefits of adopting the risk based
approach are: – The risk of failure is determined for the complete operational and process
envelope of the plant. – The effect of plant changes can be established well in advance. – By
identifying the high risk areas, modification and inspection effort is targeted – resulting in
economic benefits. – A full audit trail is established. To date, the risk based approach has been
used on over 60 different assets, ranging from offshore installations to refineries and onshore gas
plants. Significant cost savings (in the region of €0.8M per asset) have been achieved as the
result of using the approach to rationalise modifications and target inspection effort. The
updated methodology will be available from the Energy Institute from March 2007. REFERENCES
American Petroleum Institute 1995a, Reciprocating Compressors for Petroleum, Chemical and
Gas Industry Services, API Standard 618 4th Edition American Petroleum Institute 1995b, Positive
Displacement Pumps – Reciprocating, API Standard 674 2nd Edition American Petroleum
Institute 2000, Risk-Based Inspection Base Resource Document, API Publication 581, 1st Edition
Health & Safety Executive 2002a, Report on the Hydrocarbon Release Incident Investigation
Project 1/4/2001 to 31/3/2002 Health & Safety Executive 2002b,TransientVibration Guidelines for
Fast Acting Valves – Screening Assessment, HSE Offshore Technology Report OTO 2002/028
Marine Technology Directorate 1999, Guidelines for the Avoidance of Vibration Induced Fatigue
in Process Pipework, MTD Publication 99/100, ISBN 1 870553 37 3 Martin J.T, Swindell R.J. et al,
2002,Reduce Risk ofVibrationInduced Pipework Failure, Hydrocarbon Processing, July 2002 2547
Aven CH323.tex 17/5/2007 11: 42 Page 2548