WRC 533 2010

BULLETIN 533
jllllllllllllllj Welding Researelt couneil

\!\/RC
111111111111 11 bulle~in
THE ROLE OF WELDED JOINT VULNERABILITY
AND VARIOUS DAMAGE MECHANISMS ACTIVE
IN PROCESS AND PETROCHEMICAL PLANTS:
RELIABILITY ANALYSIS WITH RISK BASED
INSPECTION (RBI} APPROACH
Giancarlo Canale
Marco De Marco
Stefano Pinca
These Bulletins contain final Reports from projects sponsored

by the Weld ing Research Council, important papers presented
before engineering societies and other reports of current
ISSN 0043-2326 interest.
WELDING RESEARCH COUNCIL, INC.
PO BOX 1942
NEW YORK, NY 10156
www .forengineers.org
INTENTIONALLY LEFT BLANK
WRC - The Welding Research Council brings together science and engineering specialists in
developing the solutions to problems in welding and pressure vessel technology. They exchange
knowledge, share perspectives, and execute R and D activities. As needed, the Council organizes
and manages cooperative programs.
MPC – A Council of the WRC, the Materials Properties Council is dedicated to providing
industry with the best technology and the best data that can be obtained on the properties of
materials to help meet today’s most advanced concepts in design and service, life assessment,
fitness-for-service, and reliability and safety.
PVRC – A Council of the WRC, the goal of the Pressure Vessel Research Council is to
encourage, promote and conduct research in the field of pressure vessels and related pressure
equipment technologies, including evaluation of materials, design, fabrication, inspection and
testing.
For more information, see www.forengineers.org
WRC Bulletins contain final reports from projects sponsored by the Welding Research Council, important
papers presented before engineering societies and other reports of current interest.
No warranty of any kind expressed or implied, respecting of data, analyses, graphs or any other
information provided in this publication is made by the Welding Research Council, and the use of any
such information is at the user’s sole risk.
All rights are reserved and no part of this publication may be reproduced, downloaded, disseminated, or
otherwise transferred in any form or by any means, including photocopying, without the express written
consent of WRC.
Copyright © 2010 The Welding Research Council.

All rights, including translations, are reserved by WRC.
Printed in the United States of America.
ISSN 0043-2326
Library of Congress Catalog Number: 85-647116
Welding Research Council

20600 Chagrin Blvd.
Suite 1200
Shaker Heights, OH 44122
www.forengineers.org
The Welding Research Council, Inc.
The Role of Welded Joint Vulnerability and

Various Damage Mechanisms Active in Process
and Petrochemical Plants: Reliability Analysis
with Risk Based Inspection (RBI) Approach
Giancarlo Canale
Marco De Marco
Stefano Pinca
WRC Bulletin 533
THE WELDING RESEARCH COUNCIL, INC.

PO Box 1942
New York, NY 10156
www.forengineers.org
WRC Bulletin 533

Weld Joint Vulnerability Assessment and Reliability Assessment Using RBI i
ISBN No. 1-58145-540-2
Library of Congress
Catalog Card Number: 85-647116
Copyright © 2010 by
Welding Research Council, Inc.
All Rights Reserved
Printed in U. S. A
ii WRC Bulletin 533

Weld Joint Vulnerability Assessment and Reliability Assessment Using RBI
FOREWORD
With growing regulatory acceptance, industry is now applying the concepts of risk-informed decision
making, especially in the area of in-service inspection i.e. risk-based inspection (RBI). It is no secret that
welds are principal areas of concern, are potentially subject to complex failure modes, and generally
inadequately understood. However, RBI must be done using defensible, systematic approaches that
yield reasonably uniform results independent of the performing party, often an outside vendor.
Often today there are wide variations in the work products. RBI requires modeling damage accumulation
and consequences. Some good work has been done. However, caveat emptor – buyer beware – with
the buyer at a serious disadvantage. It is the responsibility of the owner/operator to ensure that
underlying assumptions and technology are capable of producing reliable RBI results and predictions.
RBI must sit on a strong technical foundation of models and accompanying documented technology.
Data acquisition and analysis require rigor. Responsible engineers must do an effective job of identifying,
predicting and quantifying potential damage modes and contributing factors. NDT is in need of major
upgrading and understanding in consideration of the unique failure modes associated with welds.
Damage models, materials properties and consequence models must be justified. Computerized
assessment procedures must be continuously improved, updated and the underlying technology made
more transparent. Terminology must be uniform and fully understood.
This WRC Bulletin The Role Of Welded Joint Vulnerability And Various Damage Mechanisms Active In
Process & Petrochemical Plants: Reliability Analysis With Risk Based Inspection (RBI) Approach is an
important reference work in the process of assessing weld reliability. It provides an in-depth look at the
approach to RBI of WRC’s sister organization, Istituto Italiano Della Saldatura in Genoa, Italy with
extensive experience in the area
Martin Prager, Ph. D

Executive Director
Welding Research Council
WRC Bulletin 533

Weld Joint Vulnerability Assessment and Reliability Assessment Using RBI iii
Table of Contents
1 INTRODUCTION ................................................................................................................................. 1
2 IS RISK BASED INSPECTION METHODOLOGY REALLY UNDERSTOOD? .................................. 2
2.1 When Can RBI Be Best Applied In The Ageing Life Cycle Of Existing Equipment?............ 2
2.2 Overview Of RBI Methodology (What-Where-How-When Comprehensive Approach) ........ 2
2.3 RBI Methodology – Probability and Consequence of Failure ................................................ 3
2.4 The Aim Of RBI – The Inspection Planning Outcome ............................................................. 5
2.5 How Far Can I Go With The Risk – The Risk Target................................................................ 6
3 THE WELDED JOINTS ROLE IN INCREASING THE FAILURE FREQUENCY OF A COMPONENT6
3.1 Overload...................................................................................................................................... 7
3.2 Brittle Fracture............................................................................................................................ 7
3.3 Fatigue ........................................................................................................................................ 8
3.4 Creep ........................................................................................................................................... 9
3.5 High Temperature Hydrogen Attack (HTHA)............................................................................ 9
3.6 Corrosion & Stress Corrosion Cracking ................................................................................ 10
4 HOW CAN RBI CONTROL WELDMENT TENDENCY TO INCREASE FAILURE FREQUENCY? . 13
4.1 Time & Inspection Effectiveness - The Damage Factor........................................................ 13
4.2 How To Estimate Damage Factor – Inspection & Corrosion Engineers Main Scene......... 15
4.2.1 The Thinning Damage Factor – Can The Weld Zone Corrosion Rate Be Estimated In A
Reliable Manner?............................................................................................................................. 15
4.2.2 The SCC Damage Factor – The Escalation Factor .......................................................... 16
4.2.3 The Brittle Fracture Damage Factor – The Cold Effect ................................................... 17
4.2.4 The Mechanical Fatigue Damage Factor – The Shaking Effect...................................... 18
4.2.5 The HTHA Damage Factor – The Time/Temperature/Hydrogen Effect .......................... 19
4.2.6 Creep – Life Extension....................................................................................................... 19
4.2.7 Example Of A Damage Factor Calculation....................................................................... 19
5 CONCLUSION ................................................................................................................................... 20
6 NOMENCLATURE ............................................................................................................................ 21
7 REFERENCES .................................................................................................................................. 22
8 TABLES............................................................................................................................................. 23
9 FIGURES ........................................................................................................................................... 28
WRC Bulletin 533

Weld Joint Vulnerability Assessment and Reliability Assessment Using RBI iv
ABSTRACT
Most equipment in process plants and energy production is fabricated by welding. Welding is also
extensively used to repair, modify and overlay plant equipment during shutdowns. The statistics of failure
show weldments to be one of the most vulnerable areas. Risk Based Inspections (RBI) methodology in
the area of asset integrity and risk management focuses on prioritizing and optimizing inspection
strategies and should look at the vulnerability of welded joints taking into account mechanical behavior,
and damage mechanisms including corrosion and stress-corrosion degradation. This WRC Bulletin
provides a survey of the interaction between potentially active damage mechanisms and welded joints
and describes the RBI approach to rank the criticality of such an interaction for the purpose of inspection
and risk management.
1 INTRODUCTION
This WRC Bulletin was written with the aim to present an overview of the possible influence that the
welding process, which has historically been essential to the manufacture of pressure equipment, may
have on reliability during the life cycle of process plants.
Welding technology has played a significant role in the stormy technological development of recent years.
The improvement of welding techniques allows not only for the sudden increase in the size of plants, but
also for the development of completely new designs. On the other hand, it is not surprising that with such
development, material and manufacturing problems have been encountered, especially with the new and
high performance materials which are more sensitive to the presence of a welded joint (i.e. Cr-Mo-V low
alloy steels, duplex stainless steels, nickel alloys, etc.).
In such an industrial context, it has been seen that the recently developed, complex pressure equipment
risk management and monitoring tools have to deal with the presence of welded joints and with all the
problems they entail in terms of reliability and inspection issues. The interest in such asset integrity
management tools for pressure equipment has been mainly dictated both by the progressive and
inevitable aging of process plants, and by the growing awareness of safety, financial savings, and
environment-related issues.
The Risk Based Inspection (RBI) approach is the methodology for assessing, monitoring and mitigating
risk that most takes into account the complexity of the chemical and physical aspects of the multiple, and
often synergistic, damage mechanisms active in process plants. It also takes into account the best
inspection methods to evaluate these mechanisms in the field, while trying to move forward along these
lines as quantitatively as possible.
In the early days of Risk Based Inspection the term risk-informed inspection was sometimes used. This
was first introduced by the US Nuclear Regulatory Commission in order to emphasize the link – but not a
direct correlation – between risk and inspection. If risk based inspection is understood to be inspection
planned on the basis of information obtained about the risk, then the two terms are synonymous.
Generally speaking, Risk Based Inspection involves the planning of an inspection on the basis of the
information obtained from a risk analysis of the equipment. Within this article, the term inspection refers
to the planning, implementation and evaluation of examinations to determine the physical and
metallurgical conditions of equipment in terms of Fitness-For-Service (FFS). On the other hand, the
purpose of risk analysis is to identify potential degradation mechanisms and threats to the integrity of the
equipment, and to assess the consequences and risks of failure. The resulting inspection plan can then
target the high risk equipment and can be designed to detect potential degradation before FFS is an
issue.
WRC Bulletin 533

Weld Joint Vulnerability Assessment and Reliability Assessment Using RBI 1
2 IS RISK BASED INSPECTION METHODOLOGY REALLY

UNDERSTOOD?
2.1 When Can RBI Be Best Applied In The Ageing Life Cycle Of Existing Equipment?
The equipment in plants around us is progressively ageing with time and becoming more and more
seriously affected by the various mechanisms of degradation active in process plants. Ageing equipment
is equipment for which there is evidence or likelihood of significant deterioration and damage taking place
since new, or for which there is insufficient information and knowledge available to know the extent to
which this possibility exists.
By its nature, industrial equipment containing hazardous and/or pressurized fluids is exposed to stress
and environmental conditions that will ultimately degrade the material from its initial state. Damage will
accumulate until the equipment reaches a state where it is no longer judged to be fit-for-service. During
the ageing life, as damage accumulates, failure becomes increasingly likely and, if not monitored with
increasing confidence, an accidental event of some kind will eventually occur.
Typically, accumulated damage and degradation rate rise with time (see Figure 1), and hence the
probability that an individual component will fail due to this accumulated damage normally increases over
time. However, the probability of failure can be altered by appropriate inspection, maintenance, and
repair of damaged areas. The risk of failure then oscillates between the minimum and maximum
operating risk levels, with the periodicity decreasing and maintenance, inspection and repair becoming
more frequent later in life (see Figure 2).
In Figure 3 depicts the four stages that can be associated with ageing equipment life, also shown in
Figure 1 and Figure 2. For the purpose of life-cycle management, it may be helpful to consider an item of
equipment as having four stages in its life, each having specific characteristics and needing different
management, inspection and maintenance strategies.
The scope of RBI is to optimize the inspection and maintenance driven force of risk reduction during the
ageing life of equipment. It fits mainly – but not only – within Stage 2 of equipment life.
2.2 Overview Of RBI Methodology (What-Where-How-When Comprehensive Approach)

As described above, Risk Based Inspection involves the planning of an inspection on the basis of
information obtained from a risk analysis of the equipment. The purpose of risk analysis is to identify the
potential degradation mechanisms and threats to the integrity of the equipment, and to assess the
consequences and risks of failure. The inspection plan can then target the high risk equipment and be
can designed to detect potential degradation mechanisms.
RBI is not innovative in risk analysis from the point of view of consequence of failure. The innovation in
RBI is because of the introduction of two new concepts when evaluating probability of failure:
• the role of inspection in providing new information about the condition of the equipment; this may
be better or worse, or the same as previously estimated but the aim is to reduce existing
uncertainty so new information can modify the estimated probability of failure;
• the global, analytical and structured approach to the evaluation of all the various damage
mechanisms active in process and power generation plants extrapolated from many years of
industrial experience and expertise applied research.
RBI has become the first analytical and documented approach to collate the interdisciplinary of different
sciences such as process engineering, corrosion engineering, inspection and Non-Destructive
Examination (NDE) engineering, and Health Safety and Environment (HSE) study to support Asset
Integrity Management.
The aim of RBI is to define and measure the level of risk associated with an item, then evaluate safety,
WRC Bulletin 533

environmental and business interruption risks in order to reduce risks of failure by the effective use of
inspection resources. It is worth remembering that the real scope of RBI is the latter, that is the
optimization of inspection resources and that it is wrong to look at RBI as a stand-alone risk analysis or as
a self-supporting detailed inspection programming tool at plant levels. It could be dangerous for
inexperienced personnel to employ such an advanced and analytical method as a self-supporting tool.
In the qualitative and semi-quantitative approaches each plant item is evaluated and assigned a position
in a 5 x 5 risk matrix (see Figure 4).
In general, risk in an RBI assessment is calculated as the product of the likelihood or probability of failure
and the consequence of failure the damage mechanisms determined to be active, and the total risk for an
item is the sum of all the risks in the scenario:
R(t ) = Pf (t ) ⋅ C (t ) (1)
Riskitem = ∑ R(t ) (2)
Note that in general, the probability of failure, Pf (t ) , and the consequence of failure as both shown as a
function of time.
An inspection program is then developed to reduce that risk. The following needs to be considered to
develop an inspection program:
a) What type of damage to look for,
b) Where to look for damage,
c) How to look for damage, and
d) When to look for damage.
What and where is established from reviewing the design and process data and the equipment history.
How to look for the damage is decided by reviewing damage density and variability, inspection sample
validity, sample size, detection capability of method, and validity of future predictions based on past
observations. When to look for damage is related to the estimated risk increasing up to a maximum
tolerable value. The RBI approach prescribes actual methods to employ, with specific values that can be
applied to given situations and conditions.
2.3 RBI Methodology – Probability and Consequence of Failure

In a structured RBI analysis that contains an actual risk evaluation, the likelihood or probability of failure,
Pf (t ) , is determined from the sum of five weighted factors:
a) Damage mechanism.
b) Usefulness of inspection.
c) Current equipment condition.
d) Nature of process.
e) Safety design and mechanisms.
The time dependence of the probability of failure may be introduced in a RBI analysis using a damage
factor as a function of time, D f ( t ) . This methodology was introduced into the American Petroleum
Institute (API) RBI methodology which defines Probability of Failure as:
Pf ( t ) = gff ⋅ D f ( t ) ⋅ FMS (3)
WRC Bulletin 533

In this equation, the probability of failure, Pf (t ) , is determined as the product of a generic failure
frequency , gff , a damage factor, D f (t ) , and a management systems factor, FMS . The API RBI
methodology is described in detail in the API RBI RP 581 Risk Based Inspection Technology, Second
Edition, 2008. This Recommended practice provides a quantitative procedure to establish an inspection
program using risk-based methods for pressurized fixed equipment in the process industry, including
pressure vessels, piping, tankage, pressure relief devices, and heat exchanger tube bundles.
In the API RBI approach the damage factor modifies the industry generic failure frequency, gff , and
makes it specific to the component under evaluation. The Damage Factor is determined on the basis of
active damage mechanisms (local and general corrosion, cracking, high temperature stress, etc.) which
are relevant to the material of construction and the process service, the physical conditions of the
component and the inspection techniques used to quantify damage.
Damage factors do not provide a definitive Fitness-For-Service assessment of the component. Its basic
function is to statistically evaluate the amount of damage that may be present as a function of time in
service and the effectiveness of an inspection activity to quantify that damage. The damage factor has to
be determined for the following main damage mechanism categories:
a) Thinning (both general and local)
b) Component linings
c) External damage (corrosion and stress corrosion)
d) Stress Corrosion Cracking (internal and based on process fluid, operating conditions and materials
of construction)
e) High Temperature Hydrogen Attack
f) Mechanical Fatigue (piping only)
g) Brittle Fracture (including low temperature brittle fracture, temper embrittlement, 475°C embrittlement
and sigma phase embrittlement).
If more than one damage mechanism is present, then the principle of superimposition - with a special
modification for general thinning and external damage, and component linings - is used to determine the
total damage factor for a given piece of equipment (see Figure 5).
The management systems adjustment factor, FMS , accounts for the influence of the facility’s
management system on the mechanical integrity of the plant equipment. This factor accounts for the
probability that accumulating damage which results in loss of containment will be discovered in time and
is directly proportional to the quality of a facility’s mechanical integrity program. This factor is derived
from the results of an evaluation of a facilities or operating unit’s management systems that affect plant
risk.
The Consequence Of Failure, C (t ) , is divided into three factors:
a) Fire/Explosion;
b) Toxicity;
c) Financial.
Loss of containment of hazardous fluids from pressurized processing equipment may result in damage to
surrounding equipment, serious injury to personnel, production losses, and undesirable environmental
impacts (see Figure 6). In API RBI, the consequences of loss of containment are determined using well
established consequence analysis techniques and are expressed as an affected impact area or in
financial terms (see Figure 7).
Impact areas of an event outcome, such as pool fires, flash fires, fireballs, jet fires and vapor cloud
explosions are quantified based on the effect of thermal radiation and overpressure on surrounding
WRC Bulletin 533

equipment and personnel. Additionally, cloud dispersion analysis methods are used to quantify the
magnitude of flammable releases and to determine the extent and duration of personnel exposure to toxic
releases. Event trees are utilized to assess the probability of each of the various event outcomes and to
provide a mechanism for weighing the loss of containment consequences (see Figure 8).
2.4 The Aim Of RBI – The Inspection Planning Outcome

The premise of inspection planning using RBI is based on the fact that at some point in time, the risk, as
defined above, will reach a specified risk target. In order to reduce the uncertainty of knowledge about
the relevant state of the piece of equipment, when, or before, the risk target is reached, an inspection of
the equipment is recommended based on ranking of the component damage mechanisms that have the
highest calculated damage factors (see Figure 9).
In RBI, reduction in uncertainty associated to an inspection program is a function of the effectiveness of
the inspection in identifying the type and extent of damage. Some inspection techniques are better, for
example, in detecting thinning (general corrosion) damage than others.
The main issue in the RBI approach is to analytically develop the link between the effectiveness of a
particular inspection method and risk reduction, once the actual damage mechanism and failure mode
active on a piece of equipment are quantitatively defined.
Inspection programs (the combination of NDE methods such as visual, ultrasonic, etc.) vary in their
effectiveness for locating and sizing damage. Limitations in the ability of a program to improve knowledge
of the damaged state result from the inability to inspect 100% of the areas subject to damage, and from
inherent limitations of NDE methods to detect and quantify the damage. The effectiveness of an
inspection program can be quantitatively expressed as the likelihood that the observed damage state
(and thus the predicted damage rate) actually represents the true state.
By identifying credible damage mechanisms, determining the damage rate and susceptibility, and
selecting an inspection effectiveness category based on defined level of inspection, a probability of failure
and associated risk may be determined using the following equations that in the calculation of probability
take into account the role of the condition, or of the damaged state, observed by an inspection program
with variable effectiveness.
R(t , I E ) = Pf (t , I E ) ⋅ CA for Area − Based Risk (4)
R(t , I E ) = Pf (t , I E ) ⋅ FC for Financial − Based Risk (5)
Note that in Equations (4) and (5), the risk is varying with time since the probability of failure is a function
of time. Also note that in API RBI, the consequence of failure in Equation (1), C (t ) , is assumed to be
invariant with time. In Equations (4) and (5), C (t ) is replaced with CA when the consequence is
expressed as an impact area expressed in units of area and FC when the consequence is expressed in
expressed in financial terms.
It is worth noting that in a well developed RBI analysis it is important to identify the damage mechanism
before determining the actual inspection effectiveness. The role of the different personalities involved in
the analysis is therefore of primary importance. Only a qualified and experienced materials engineer has
the ability to identify and properly characterize the damage mechanisms and modes of failure, and only a
qualified and experienced NDE engineer or specialist can evaluate the best traditional and advanced
inspection methods employed in the past that can be suggested for future plans.
RBI is not for a single old style-all knowing expert or consultant but is for structured, skilled and able-to-
communicate team of knowledgeable personnel comprised of inspection, reliability, and operations
personnel.
WRC Bulletin 533

2.5 How Far Can I Go With The Risk – The Risk Target
The risk target is defined as the level of acceptable risk for inspection planning purpose. Although RBI
analysis can be used to rank equipment on the basis of risk, and to effectively inspect the top ranking
ones, it can also be applied with a risk target approach. The risk target is in terms of area for area-based
consequence analysis and in terms of financial limits for financial-based consequence analysis. A Risk
Target may also be set for the damage factor, taking into account the important and time dependant role
of such parameters.
A risk target is not absolute. Specification of a risk target is the responsibility of the Owner-User. A risk
target may be developed based on Owner-User internal guidelines. Many companies have corporate risk
criteria defining acceptable and conservative levels of safety, environmental and financial risks. These
risk criteria should be used when making risk-based inspection decisions since each company may have
different acceptable risk levels, and risk management decisions can vary among companies.
3 THE WELDED JOINTS ROLE IN INCREASING THE FAILURE

FREQUENCY OF A COMPONENT
It is worth noting that in the pressure vessel and piping systems typically found in process plants, other
factors – the chemical, physical and mechanical influence of a welded joint – tend to increase the
probability of failure.
The RBI personnel should be well aware that weld seams are areas in welded constructions which must
be scrutinized critically. It is unquestionable that functional performance of the welded component itself
critically depends on the properties of the welded joint. Here, failures occur frequently because the
material properties change more rapidly in weldments than elsewhere in the base material. The statistics
of failure show weld cracks to be one of the most frequent causes. Despite everything, we do not have
any alternative joining methods that offer equal potential for producing metallic structures of high strength,
reliability and integrity.
The special role that sets welding apart from other production processes has to do with the simple fact
that heating, melting and re-solidification may dramatically change the properties of the materials. If weld
mechanical properties could be determined from point to point, they would be found to vary. There may
be a depletion of alloying elements from welding wire/electrode, there may be contamination from
atmospheric gases or carbon monoxide may evolve in some welding processes. Segregation of
impurities may occur, creating zones of inherent weakness; grain size may not be uniform, the structure
may be columnar rather than equiaxed and the solidified weld metal may exhibit a preferred orientation.
The effect of heating on the unfused parent metal can be equally dramatic, particularly in the Heat-
Affected-Zone, the area closest to the weld. Rapid heating and cooling will cause changes in the grain
structure. Intense hardening can also result which, with the addition of hydrogen diffusing from the weld,
may cause embrittlement and cracking tendency.
Also, corrosion resistance may be reduced by precipitation; the precipitation of chromium carbides in un-
stabilized stainless steels, for example, results in a chromium depleted band that is particularly
susceptible to corrosive attack.
In addition to these potential metallurgical changes, which must be considered in relation to the service
condition of the components – and perhaps more important – welding also produces severe plastic
deformation in both weld and parent metal after the weld solidifies. This introduces aging effects in the
material and also two associated phenomena that can frighten inspection and reliability personnel:
shrinkage and distortion, together with residual stresses generally equal to the material’s yield point (see
Figures 10, 11 and 12).
The often synergistic action of the above mentioned metallurgical and chemical effects and residual
stresses can seriously influence the reliability of welded joints and can create serious difficulties in
assessing the associated reliability (i.e. the associated damage factor) and inspection method.
From an RBI point of view (see Figure 2), the service related welded joint reliability (Stage 2 of ageing life)
is of primary importance, more so than fabrication related defects (e.g. hydrogen, solidification, relaxation
WRC Bulletin 533

and liquation cracks). Additionally, pre-existing welding defects can act as initiation sites (metallurgical
notches) for subsequent propagations of failures.
A welded component may fail in service by any of the following mechanisms:
• Overload
• Brittle fracture
• Fatigue
• Creep
• High Temperature Hydrogen Attack-HTHA
• Corrosion
• Stress Corrosion Cracking (SCC)
Failures may be initiated by a single or by a combination of mechanisms. For example, a fatigue crack
may propagate by brittle fracture; an overload failure may start with creep; fatigue may occur in a
corrosive environment, and creep may be combined with fatigue. Obviously, not all the above mentioned
failure mechanisms (and the associated risk managed with RBI) can be, by virtue of their different
propagation times, monitored by inspection: overload is a critical velocity propagation failure, while
corrosion can be very slow. The weld type of failures and their dependence on possible service
conditions are summarized in Figure 13.
3.1 Overload
Failure by simple overload is a relatively rare occurrence because mistakes in estimating the loads
imposed on a structure are seldom made and the resulting stresses from these loads are seldom
underestimated. The methods of allowable stress in design provide adequate protection against this type
of failure, so welded structures often have great strength reserves under static loading without any risk of
failure, provided brittle fracture, fatigue, buckling and particular oscillation reaction (wind for chimneys) are
treated separately.
The wrong choice of filler metal for high strength material can lead to a less resistant welded joint that
may results in a failure by overload in service (e.g. austenitic stainless steel weld metal in high strength
parent metal). Overload failure is seldom treated in RBI analysis by virtue of the propagation rate (critical
propagation rate) of the damage that cannot be controlled by inspection (fundamental of RBI approach).
3.2 Brittle Fracture

Brittle fracture seldom comes alone. It requires an initiation site and often the weldment is a focal point in
giving the starting notch a site for brittle fracture. No other area has attracted the efforts of so many
researchers and code writers in so many different countries as that of brittle fracture. Pressure vessels,
storage tanks, piping systems and other ferritic steel structures have suddenly failed in service, or even
during construction, by the development of a fast fracture which has been neither preceded nor
accompanied by significant plastic deformation, even though perfectly adequate material ductility was
indicated by tensile test results (see Figure 14). These fractures are initiated at some geometric
discontinuity which produces a relatively severe stress concentration such as weld or HAZ cracks, weld
toe fatigue cracks, accidental arc strike marks, the end of a fillet weld, and lack of penetration in a butt
weld. Residual stresses introduced by welding may play an important role, and certainly promote brittle
fractures at low applied stress in presence of a flaw.
From the above, it can be said that brittle fracture is an event-driven damage mechanism. The
combination of material fracture toughness at a given operating temperature – as measured in CVN,
CTOD, or other tests – the size, shape and stress concentration of a flaw, and the amount of residual and
applied stress on the flaw, govern the probability of the event.
WRC Bulletin 533

In most cases, brittle fracture occurs only at temperatures below the Charpy impact transition temperature
(or ductile to brittle transition temperature). Susceptibility to brittle fracture may be increased by a loss of
fracture toughness (i.e. an increase of ductile-to-brittle transition temperature) of the material involved
caused by embrittlement phenomena catalyzed by exposure of the material in some critical temperature
range such as temper embrittlement for Cr-Mo steels, 475° C embrittlement for ferritic stainless steels,
SIGMA phase embrittlement for austenitic stainless steel with delta ferrite (weld metal more prone), and
intermetallic precipitation for duplex stainless steels (weld metal and HAZ more prone).
In most process plants, equipment runs at temperatures above which brittle fracture is a concern so the
main concern is for brittle fracture during startup, shutdown or hydrotesting at ambient temperature.
Since the thickness of the component is another governing factor, thick wall equipment on any unit should
be considered (constraint to plastic deformation).
Brittle fracture can also occur during an auto-refrigeration event in units processing light hydrocarbons
such as methane, ethane/ethylene, propane/propylene, or butane. Such fluids have boiling points well
below atmospheric temperature at a point where the construction material of the component may not be
of acceptable toughness.
The goal of low temperature/low toughness fracture assessment is to rank components, evaluated as
susceptible to such a failure mode, in relation to the relative probability of their tendency to fracture. This
assessment will take into account the thickness, the material type, the PWHT, and temperature.
3.3 Fatigue
The largest number of failures in welded structures is by far due to the fatigue produced by repetitious or
fluctuating loads. Most structures are subjected to cycling loading in service, sometimes between
constant stress limits but more often between irregularly varying limits. The welded joint is a natural
collector of fatigue fracture initiation sites (edge of a weld, fillet weld fusion line, end of a weld, undercut,
incomplete penetration, centerline cracks, accidental ark strikes, excess of penetration, lack of fusion,
hydrogen cracking).
The number of stress cycles a particular type of joint made by a particular welding process can sustain
before it fails can be readily determined by experiment as long as the stress variation falls between
constant limits. The reality of random loading adds another variable to fatigue problems.
Fatigue is a purely local effect. A fatigue crack starts at the most highly stressed point, which may not be
the point where the designer attention was typically focused: the point of maximum calculated stress
design. Fatigue failure is generally caused by purely local changes in stress distribution and any
discontinuity will cause stress concentration (see Figure 15). Consequently and as far as possible, the
first basic principle in monitoring any fatigue problem in the field is to check for any kind of discontinuity.
In some cases, cracking can be induced through pressure, mechanical, or thermal stresses. Damage
may be characterized as either low cycle fatigue where the maximum cyclic stresses imposed approach
the yield strength of the material or as high cycle fatigue where the cyclic stresses imposed are
significantly lower than the yield strength of the material. In thermal fatigue failure of welds, the onset of
low cycle fatigue cracking is often directly related to the number of heat-up/cool-down cycles experienced.
Excessive vibration of rotating equipment or flow-induced vibration can cause high cycle fatigue damage.
A special category of fatigue cracking is referred to as corrosion fatigue. This occurs under the
simultaneous action of corrosion and cyclic stresses. Cracking observed in deaerator vessels is a typical
example of corrosion fatigue. As mentioned above, fatigue cracking can typically be detected first at
points of highly localized stress such as areas subject to high vibration, weld peaking and at internal or
external attachments. Locations where metals having different coefficients of thermal expansions are
joined by welding may be susceptible to thermal fatigue.
The preferred NDE methods for detecting fatigue cracking include liquid penetrant testing, magnetic
particle testing, eddy current testing, and angle beam ultrasonic examination. It should be noted that
fatigue cracking is likely to cause vessel/piping failure before detection with any NDE methods. The vast
majority of the fatigue cycles required to produce failure is required to initiate cracking and relatively few
WRC Bulletin 533

cycles are required to propagate the crack to failure. As such, design, installation and monitoring of
vibration effects to prevent fatigue cracking are important.
For piping systems that are not subjected to mechanical fatigue in as-built condition, changing conditions
such as failure of pipe support, increased vibration from out of balance machinery, chattering of relief
valves during process upsets, changes in flow and pressure cycles, or adding weight to unsupported
branch connections (pendulum effect), can render a piping system susceptible to failure.
3.4 Creep
Creep is dependent on time, temperature, and stress and is by definition time dependent strain occurring
under stress. After a period of time, creep may terminate in fracture by stress rupture (also called creep
rupture). Creep damage may be difficult to detect until significant damage has occurred. Cracking may
be accelerated by creep/fatigue interaction when operating conditions in the creep range are cyclic.
Particular attention should be given to areas of high stress concentration. If excessive temperatures are
encountered, mechanical properties and microstructural changes in metals may also take place.
NDE methods suitable for detecting creep cracking in the latter stages of damage include liquid penetrant,
magnetic particles, ultrasonic examination, radiography, and on-site metallography. Acoustic emission
can also be used to detect the presence of cracks that are activated by test pressures or stresses
generated during the test.
For welds, the situation is much more critical than it is for the base metals (see Figure 16). Although the
short term properties of welded joints can be superior to the parent metal, after long service times in high
temperature environments, a faster deterioration of the microstructure of weld metal and HAZ takes place.
The metallurgical and geometrical factors (geometrical discontinuities) give rise to a complex straining
and redistribution of stresses in welded joints.
Some welded joints may have low creep ductility; in this case crack initiation may occur with little or no
apparent deformation. Such type of failures are typical for higher tensile strength materials and welds,
more likely in the coarse-grained HAZ of some types of Cr-Mo steels (such as 1.0 Cr-0.5Mo, 1.25 Cr-
0.5Mo, and 2.25 Cr-1Mo), especially in older steels and/or joints Post Weld Heat Treated in the low end of
code-permissible range.
Certain ferritic steels operating for long periods of time in the range of 440° C to 760° C may suffer a type
of structural change called graphitization that can substantially change materials creep properties. The
steels most susceptible to graphitization are carbon and C-0.5Mo steels. In this temperature range, the
carbide may decompose into ferrite crystals and graphite nodules. Localized graphitization results in
highly concentrated graphite-nodule formation which may lead to mechanical failure. Localized
graphitization has most frequently been found in welded joints in narrow bands at the low-temperature
edge of HAZ.
3.5 High Temperature Hydrogen Attack (HTHA)

High Temperature Hydrogen Attack (HTHA) occurs in carbon and low alloy steels exposed to a high
partial pressure of hydrogen at elevated temperatures (T> 204° C). It is the results of atomic hydrogen
diffusing through the steel and reacting with carbides in the microstructure to form methane (CH4). Such
reactions can cause both internal decarburization and fissuring in the material (see Figure 17) due to the
accumulation of methane gas at the carbide matrix interface.
Internal fissuring is more typically observed in carbon steel, C-0.5 Mo steels, and Cr-Mo steels at higher
hydrogen partial pressure. The more the steel is alloyed with Cr and Mo the less it is susceptible. The
reference document for HTHA evaluation is the API RP 941 –based on the Nelson Curves – which plots
different material susceptibility as a result of temperature and hydrogen partial pressure.
It should be noted that the secondary stress imposed, for example by a welded joint, can accumulate
HTHA damage which is critical also because of the geometrical stress concentration effect of the weld.
WRC Bulletin 533

3.6 Corrosion & Stress Corrosion Cracking

Corrosion is the prime cause of deterioration in a pressure vessel and may occur on any part of the
vessel/piping system. The severity of the deterioration is influenced by the corrosion resistance of the
construction materials. Many of the contaminants in oil and chemicals handled in process units react with
metals in such a way as to cause corrosion.
Corrosion failures of welds occur in spite of the fact that the proper base material and filler metal have
been selected, industry codes and standards have been followed, and welds have been deposited that
possess full weld penetration and have proper shape and contour. Since factors implicated in weldment
corrosion are different and synergistic, it is difficult to determine why weldments sometimes corrode
preferentially. One of the following may be influent:
• Weldment design
• Fabrication technique
• Welding practice
• Welding sequence
• Moisture contamination
• Organic or inorganic chemical species
• Oxide film or scale
• Weld slag and spatter
• Incomplete weld penetration or fusion
• Porosity
• Cracks (crevices)
• High residual stresses
• Improper choice of filler metal
• Final surface finish
The corrosion behavior of carbon steel weldments is dependent on a number of factors. Since carbon
steels undergo metallurgical transformation across the weld and HAZ microstructure and morphologies
become important, considerations must be given to the compositional effects of the base metal and
welding consumables and to the different welding processes used. A wide range of microstructures can
develop based on cooling rates and these microstructure are dependent on energy input, preheat, metal
thickness, weld bead size and reheating effects due to multipass welding. As the result of the different
chemical compositions and weld inclusions (oxides and sulfides) weld metal microstructures are usually
significantly different from those of the HAZ and base metal. Similarly, corrosion behavior can also vary.
During welding, the base metal, HAZ and underlying weld passes experience stresses due to thermal
expansion and contraction. Upon solidification, rather high levels of residual stress remain as a result of
weld shrinkage. Stress concentration effects due to geometrical discontinuities play a very important role
in SCC damage. This type of defects can also play a role in fatigue cracking and crevice corrosion.
Welded joints having excessive weld reinforcement or lack of/excessive penetration, and not subjected to
PWHT, are the most critical from this point of view.
For carbon steels, depending on service environment and type of welding, possible corrosion failure are:
• preferential HAZ corrosion (see Figure 18)
• preferential weld corrosion (see Figure 18)
• galvanic corrosion
• stress corrosion cracking
WRC Bulletin 533

There is no doubt that residual welding stresses in carbon steel weldments can cause SCC in
environments in which the material is susceptible. This is the case for failure by both active path and
hydrogen embrittlement mechanisms, and in the latter case, failure may be especially likely at low heat
input welds because of enhanced susceptibility of the hardened structure which forms. Most SCC studies
of welds in carbon steel have evaluated resistance to hydrogen induced SCC, especially under the sour
(H2S) conditions prevalent in oil and gas plants. In this case an overriding influence of hardness is
evident. The situation for active path cracking is less clear, but there are few, if any, cases in which SCC
resistance increases at higher strength levels. On this basis, it is probable that soft, transformed
microstructures around the weld are preferable, although there are mechanism that are microstructure
dependant (H2S related mechanisms) and others that are mainly stress dependant (NaOH and other
alkaline mechanisms).
In corrosion resistant materials such as stainless steel and nickel alloys, the cycle of heating and cooling
that occurs during the welding processes affects the microstructure and surface composition of welds and
adjacent base metals. Consequently the corrosion resistance of autogenous welds and welds made with
matching filler metal may be inferior to that of the properly annealed base metal because of:
• Microsegregation
• Precipitation of secondary phases
• Formation of unmixed zones
• Re-crystallization and grain growth in HAZ
• Volatilization of alloying elements from the molten weld pool
• Contamination of the solidifying weld pool
Corrosion resistance can usually be maintained in the welded condition by balancing alloy compositions
to inhibit certain precipitation reactions, by shielding molten and hot metal by removing chromium
enriched oxides and chromium depleted areas from thermally discolored (heat tinted) surfaces, and by
choosing the proper welding parameter.
In more alloyed materials, a number of important changes occur during the welding process that can
significantly affect the corrosion behavior of the weldment. Heat input and welder technique obviously
play important roles. The way in which the weld solidifies is equally important to understanding how
weldments may behave in corrosive environments.
Metallographic evaluation shows that welds solidify in various zones (see Figure 19). The composite
region, or weld nugget, consists of essentially filler metal that has been diluted with material melted from
the surrounding base metal. Next to the composite region is the unmixed zone, a zone of base metal that
melts and solidifies during welding without experiencing mechanical mixing with the filler metal. The weld
interface is the surface bounding the region within which complete melting was experienced during
welding, and is evidenced by the presence of a cast structure. Beyond the weld interface is the partially
melted zone, which is a region of the base metal within which the melted proportion ranges from 0 to
100%. Lastly, true HAZ is that portion of the base metal within which microstructure change has occurred
in the absence of melting. Although the various regions of a weldment are for a single pass weld, similar
solidification patterns and compositional differences can be expected to occur in underlying weld beads
during multipass application.
The corrosion problems commonly associated with welding of austenitic stainless steels are related to
precipitation effects and chemical segregation. These problems can be eliminated or minimized through
control of base metal metallurgy, control of welding practice and selection of the proper filler metal.
Different forms of corrosion are:
• preferential attack associated with weld metal precipitates (ferrite, σ phase, and M23C6 carbides);
• pitting corrosion;
• crevice corrosion;
WRC Bulletin 533

• decay (sensitization) caused by carbide precipitation in HAZ;

• chloride SCC (see Figure 20);
• caustic SCC; and
• microbiologically influenced corrosion – MIC (see Figure 21).
Taking into account also the most severe condition they encounter in service, the high alloyed Cr-Ni-Mo-N
austenitic stainless steels (6-moly), are more susceptible to segregation induced corrosion in weld zones
and must be welded with overmatching filler metals (Mo containing nickel alloys filler metal such as
Hastelloy). Since the diluted weld metal is more corrosion resistant, the unmixed zone pitting corrosion is
more likely to develop than in other stainless steels.
Conventional 400-series ferritic stainless steels such as AISI types 430, 434 and 446 (old generation
ferritic stainless steels) are susceptible to inter-granular corrosion and to embrittlement in the as-welded
condition. Corrosion in the weld area generally encompasses both the weld metal and HAZ.
In the wrought condition, duplex stainless steels have microstructures consisting of a fairly even balance
of austenite and ferrite. These alloys are useful because of their good resistance to chloride SCC, pitting
corrosion and inter-granular corrosion in the as-welded condition.
The distribution of austenite and ferrite in the weld and HAZ is known to affect the corrosion properties
and the mechanical properties of duplex stainless steels. In order to limit the dilution of filler metal (root
pass), and control the heat input in an interval between a lower value for ferrite/austenite balance (high
cooling rates means high ferrite and low toughness and corrosion resistance) and an upper value (low
cooling rates means high intermetallic precipitations and again low toughness and corrosion resistance),
the optimal corrosion resistance of the parent material can be maintained in weld area by means of
optimal joint design and weld parameters. Obviously, the weld zone in duplex stainless steels is the most
critical area for probability of failure (corrosion driven and embrittlement driven). An optimized weld
qualification procedure with CVN tests and pitting corrosion tests (ASTM A923) may reduce the
probability of failure of duplex stainless steel welded components.
The effects of welding on the corrosion resistance of nickel base alloys are similar to the effects on the
corrosion resistance of austenitic stainless steels described above. For example, sensitization due to
carbide precipitation in the HAZ is a potential problem that must be taken into account in reliability
analysis. However, in the case of nickel alloys, the high content of alloying elements such as chromium,
molybdenum, tungsten and niobium (e.g. Hastelloy alloys), can result in the precipitation of other
intermetallic phases such as μ, σ, and η. The influence of the weld on corrosion properties varies with
the characteristics of the various nickel-base alloys. The corrosion resistance of weldments is dictated
not only by the HAZ but also by the weld metal itself (see Figure 22). The effect of elemental segregation
on weld metal corrosion must also be taken into consideration for certain alloys. For this reason, certain
types of alloys should be welded with overmatching filler metals for overcoming the inherent lower
resistance of the weld metal in some borderline environments (see Figure 23).
The most critical corrosion and stress corrosion cracking damage mechanisms active in process plants,
and considered in most RBI analyses (e.g. API RBI), that can preferentially attack a welded joint, and/or
are mainly influenced by the presence of the latter for all the above mentioned reasons are:
• sulfuric acid corrosion;
• hydrofluoric acid corrosion;
• acidic chloride corrosion for stainless steels;
• sour water corrosion (mainly if acidic);
from the purely corrosion point of view, and:
• caustic stress corrosion cracking;
• amine stress corrosion cracking;
WRC Bulletin 533

• sulfide stress cracking (SSC);

• Hydrogen Induced Cracking (HIC)/Stress Oriented Hydrogen Induced Cracking (SOHIC)-H2S;
• carbonate stress corrosion cracking;
• polythionic acid stress corrosion cracking;
• chloride stress corrosion cracking;
• hydrogen stress cracking;
• Hydrogen Induced Cracking (HIC)/Stress Oriented Hydrogen Induced Cracking (SOHIC)-HF;
• external chloride stress corrosion cracking; and
• external under insulation stress corrosion cracking
from the Stress Corrosion Cracking point of view.
It must be emphasized that SCC problems are the most critical for welded areas, especially for
weldments not subjected to proper heat treatment after welding (PWHT).
4 HOW CAN RBI CONTROL WELDMENT TENDENCY TO

INCREASE FAILURE FREQUENCY?
4.1 Time & Inspection Effectiveness - The Damage Factor

As mentioned in paragraph 2, the aim of RBI is to define and measure the level of risk associated with an
item, then evaluate safety, environmental and business interruption risks in order to reduce the risk of
failure by the effective use of inspection resources.
selecting an inspection effectiveness category based on defined levels of inspection, a probability of
failure and associated risk may be determined using the following equations that in the calculation of
probability takes into account the role of condition, or damage state, observed by an inspection program
R ( t , I E ) = Pf ( t , I E ) ⋅ CA (6)
R ( t , I E ) = Pf ( t , I E ) ⋅ FC (7)
Note that in Equations (6) and (7), the risk is a function of both the time and inspection effectiveness since
the probability of failure is a function of these parameters. The time dependence of the Probability of
Failure (and risk) can be associated in RBI analysis to the so called the damage factor, D f , introduced by
The American Petroleum Institute (API) RBI methodology that defines the Probability of Failure as:
Pf ( t , I E ) = gff ⋅ D f ( t , I E ) ⋅ FMS (8)
The API RBI methods for determining the probability of failure is based on failure rates for generic
equipment classes modified by equipment specific factors. For certain damage mechanisms the
tabulated or estimated generic failure frequency, gff , of a component is modified by a damage factor that
is strongly influenced by welded joints, especially in fatigue, brittle fracture and SCC events.
Damage factors are intended to support the API RBI methodology by providing a screening tool to
determine inspection priorities and to optimize inspection efforts. The damage factors ranges defining the
Probability of Failure Category for risk matrix, together with the ranges of the consequence of failure are
WRC Bulletin 533

shown in Table 1.
The basic function of the damage factor is to statistically evaluate the amount of damage that may be
present as a function of time in service and the effectiveness of an inspection activity (t, I E ) . Damage
factors are not intended to reflect the actual probability of failure for the purpose of reliability analysis.
Damage factors reflect a relative level of concern about the component, based on the stated assumptions
for each of the damage mechanisms.
For future risk evaluation, where D f is the only time dependant variable in risk calculation (see above), it
is possible to project the calculation of D f , and consequently of the risk, for future planned dates (e.g.
future maintenance turnaround-MTA) while introducing a set number of inspections with optimal
effectiveness to maintain D f and risk below acceptable levels.
For Risk Based future inspections prioritization, it is obvious that the time-dependant D f parameter is a
very, if not the most, important one. The D f parameter includes all the factors that guide the probability
of damage and the role of the corrosion and inspection experts.
In API RBI, damage factor estimates are currently provided for the following damage mechanisms:
a) Thinning – D thin
f
b) Component Lining – D clin

f
c) External Damage – D extd

f
d) Stress Corrosion Cracking – D scc

f
e) HTHA – D htha
f
f) Mechanical Fatigue (piping only) – D mfat

f
g) Brittle Fracture – D brit

f
When assessing the probability of failure, it is important to take into consideration the future deterioration
rate (e.g. damage factor) for all potential mechanisms. The rate of degradation may increase with time as
a result of the interaction between mechanisms (e.g. corrosion and fatigue). If more than one damage
mechanism is present, the following rules are used to combine the damage factor (for local thinning
typical for welded joints corrosion):
D f −total = max ⎡⎣ D thin ⎤

f − gov , D f − gov ⎦ + D f − gov + D f − gov + D f − gov + D f − gov
extd scc htha brit mfat
(9)
As shown in Equation (9), the total damage factor, D f −total , is computed from the evaluation of the
different damage factors for the same category of damage; for example, for the various mechanisms
illustrated for SCC, the governing damage factor is the maximum one.
As previously discussed, the damage factor is a function of both time and inspection effectiveness. The
five inspection effectiveness categories used in API RBI are shown in Table 2. The actual effectiveness
of any inspection technique depends on many factors such as the skill and training of the inspectors and
the level of expertise used in selecting inspection locations. Inspections are ranked according to their
expected effectiveness at detecting damage and correctly predicting the rate of damage. The actual
effectiveness of a given inspection technique depends on the characteristics of the damage mechanism.
The effectiveness of each inspection performed within the designated time period is characterized for
each of the above described damage mechanism. The number of highest effectiveness inspections will
WRC Bulletin 533

be used to determine the damage factor. If multiple inspections of lower effectiveness have been
conducted during the designated time period, they can be approximated to an equivalent higher
effectiveness inspection in accordance with the following relationships:
a) 2 usually effective (B) inspection = 1 highly effective (A) inspection
b) 2 fairly effective (C) inspection = 1 usually effective (B) inspection
c) 2 poorly effective (D) inspection = 1 fairly effective (C) inspection
A strategy used to achieve high performance and reliability is to apply the principles of diversity and
redundancy when selecting the right NDE program and determining inspection procedures. The use of a
number of complementary NDE techniques can significantly reduce the likelihood of missing flaws.
The NDE method and technique selection should be based on the ability to detect and assess the
deterioration types anticipated in the component of interest (see Table 3, and Figures 24 and 25). For
established techniques, satisfactory evidence may be available in published literature. Additional
confidence is provided by inspection procedures that are created in accordance with recognized codes
and standards. For newer or more specialized techniques, where the only available evidence may be
capability data provided by the equipment supplier, an independent assessment of the abilities and
limitations may be necessary.
4.2 How To Estimate Damage Factor – Inspection & Corrosion Engineers Main Scene
In API RBI analysis the calculation of DF is conducted by the screening criteria shown in Figure 26. For
each damage mechanism, the analyst:
a) has to start with the screening criteria evaluation (is the component subject to such a damage?);
b) has to collect the data needed (what information do I need to know to carry out the analysis?); and
c) taking into account the basic assumptions (what are the basis of the calculation?), calculate the
damage factors (the only time dependant variable in risk estimation).
4.2.1 The Thinning Damage Factor – Can The Weld Zone Corrosion Rate Be Estimated In A
Reliable Manner?
For thinning mechanisms, the damage factor calculation takes into account that the active damage has
resulted in an average rate of thinning over the time period defined in the basic data that is fairly constant.
The probability of failure is estimated by examining the possibility that the rate of thinning is greater than
what expected. The probability of these higher rates is determined by the amount of inspection and on-
line monitoring that has been performed. The more thorough the inspection, and the greater the number
of inspections and continued use of on-line monitoring, the less likely is than the chance that the rate of
thinning is greater than anticipated. In the damage factor calculation it is assumed that thinning damage
would eventually result in failure by plastic collapse.
The thinning damage factor can be calculated from Table 4. The Art parameter can be calculated using
the basic required data: the last thickness inspection reading, trd ; the age of equipment, age ; the
corrosion rate, Cr ,bm , of the base material; the minimum required thickness, tmin , in accordance with the
applicable construction code; and the corrosion allowance, CA .
⎡⎛ t − Cr ,bm ⋅ age ⎞ ⎤
Art = max ⎢⎜1 − rd ⎟ , 0⎥ (10)
⎣ ⎝ t min + CA ⎠ ⎦
The calculated damage factor using the Art parameter, D thin

fb , can be adjusted on the basis of certain
factors.
WRC Bulletin 533

⎛ FIP ⋅ FDL ⋅ FWD ⋅ FAM ⋅ FSM ⎞

D thin
f = D thin
fb ⎜ ⎟ (11)
⎝ FOM ⎠
From the above discussion, we can understand how relevant the skill of the corrosion expert is in defining
the corrosion rate, and how relevant is the inspection expert in defining the best methods that were
applied in the past history and that will be applied in the future.
Once the likely influence of welded joints on corrosion phenomena, as described in the previous
paragraph, is defined and understood, the right susceptibility (as corrosion rate) and inspection methods
(as Inspection Effectiveness) can be selected (e.g. ultrasonic testing 0° probe, profile radiography,
TOFD).
It is seldom easy to predict a localized corrosion rate, as in the case of welded joint corrosion, since the
damage may only occur under very specific operating conditions (temperatures, chemical species, flow
velocity) and is more of an on/off situation and usually does not occur at a steady and constant pace.
Since localized corrosion rates are extremely sensitive to minor variations in process conditions/materials
it is difficult to find applicable reference sources of corrosion data and it is often useful to rely on past
inspection results.
For the different thinning damage mechanisms considered in API RBI methodology, the ones that have
the greatest influence on welded joint area selective corrosion are:
• Sulfuric Acid (H2SO4) Corrosion: when high flow velocity can create turbulence in correspondence
with weld excessive penetration, which can induce iron sulfate protective film removal and
consequent localized corrosion ( see Figure 27);
• Hydrofluoric Acid (HF) Corrosion: induced by galvanic effects in such an acid environment and in
some ferritic material (CS, low alloy steels) between weld metal and parent material;
• Acidic Sour Water Corrosion: especially in tank bottoms where high conductive acidic sour water
can enhance galvanic differences between weld metal and HAZ (see Figure 28).
4.2.2 The SCC Damage Factor – The Escalation Factor

Most critical for welded joint reliability are Stress Corrosion Cracking induced failures. The monitoring of
risk associated with such damage is not so easily performed because of the localized nature of the
damage and of the propagation rate that can vary significantly for each situation.
In this case, the role of RBI methodology has been limited to the collection of SCC industrial and
experimental expertise concerning driving factors, material & environment parameters and the most valid
inspection methods.
In an SCC event, the passivation, re-passivation and metal dissolution that occur locally at the crack tip
are altered such that when the crack tip stress intensity factor exceeds a critical threshold value SCC will
initiate and grow for the specified condition. Active SCC usually accelerates initially until it attains an
approximately uniform velocity that is independent of the stress intensity factor but may be dependent on
duration (time), material, temperature and specific environmental factors.
For Risk Based Inspection purposes the time dependence of the damage factor is simply incorporated in
the escalation factor that is the escalation in the damage factor for a particular damage based on the time
in service since last inspection. In this approach, it is assumed that the probability for cracking will
increase with time since last inspection as a result of increased exposure to critical conditions.
D f = D fb ⋅ ( age )
1.1
(12)
The main assumption in determining the damage factor for SCC is that the damage can be characterized
by a susceptibility parameter that is designated as high, medium, or low based on process environment,
material of construction and component fabrication variables (e.g. PWHT). Susceptibility can be
estimated by expert opinion, measured/observed or eventually calculated by proposed methodology (the
WRC Bulletin 533

so-called Damage Supplements in API RBI).

Based on susceptibility, a Severity Index (see Table 5) can be determined that is a measure of the
susceptibility of the component to cracking (or the probability of initiating cracks) and the probability of a
crack resulting in a leak.
If cracking is detected in a component during inspection, the susceptibility is designated as high and this
will result in the maximum value in the severity Index. Cracks or arrays of cracks that are found during an
inspection should be evaluated using FFS methods. In the SCC evaluation the screening criteria (is the
component subject to such a damage?), and the required data (what information do I know to carry out
the analysis?) approaches remain valid (see Figure 28).
Inspections are ranked according to their expected effectiveness at detecting the particular SCC damage
and correctly predicting the rate of damage. The effectiveness of each inspection performed within
designated time periods must be characterized by an experienced inspection engineer. As mentioned for
thinning DF, the combination of inspections must be used in case of multiple inspections.
API RBI describes the guidelines for damage factor calculation (Damage Supplements) for the following
SCC active mechanisms typical for process plants. Recognized practical guidelines are given to calculate
the susceptibility parameter, the severity index, the inspection effectiveness and the damage factor for the
following damage mechanisms:
• caustic stress corrosion cracking;
• amine stress corrosion cracking;
• sulfide stress cracking (SSC);
• Hydrogen Induced Cracking (HIC)/Stress Oriented Hydrogen Induced Cracking (SOHIC)-H2S;
• carbonate stress corrosion cracking;
• polythionic acid stress corrosion cracking;
• chloride stress corrosion cracking;
• hydrogen stress cracking;
• Hydrogen Induced Cracking (HIC)/Stress Oriented Hydrogen Induced Cracking (SOHIC)-HF;
• external chloride stress corrosion cracking; and
• external under insulation stress corrosion cracking.
4.2.3 The Brittle Fracture Damage Factor – The Cold Effect

The same approach (screening criteria-required data-basic assumptions) can be applied to brittle fracture
damage. If the following are true, then a particular component should be considered susceptible to brittle
fracture damages:
• the material is carbon steel or low alloy steel; and
• the Minimum Design Metal Temperature (MDMT) is unknown or the MDMT is known but the
specific component may operate at below the MDMT under normal or upset condition.
The concept of brittle fracture damage factor calculation is founded on the comparison between a
minimum temperature, Tmin , evaluated on the basis of design temperature, upset temperature, and the
boiling point of transported liquid at atmospheric pressure, and a reference temperature, Tref , evaluated
on the basis of the stated Minimum Design Metal Temperature (MDMT), impact test temperature, impact
exemption temperature (API 579-1/ASME FFS-1, Part 3), that also takes into account the thickness of a
component.
Brittle fracture requires the coincident presence of a sufficient size defect, application of sufficient stress,
WRC Bulletin 533

and a susceptible material. The susceptibility to failure by brittle fracture can change due to in service
conditions (e.g. auto-refrigeration of fluids and process upsets).
Particular weight can be given in shut down-start up procedures to prevent pressurizing of a component
below certain temperatures; such a temperature can be used as Tmin . An adjustment factor for service
experience, based on the grandfather concept permitted in API 579-1/ASME FFS-1, takes into account if
the component has been exposed for many years to the lowest expected temperature and if the
component is not in fatigue or SCC service.
Theoretically, an inspection to locate and remove any pre-existing defects would reduce the probability of
failure. However the initiating defects can be very small, and need to be exposed to the surface where it
can be found. For this reason inspection for such defects is generally not considered to be an effective
method for prevention of brittle fracture. For this damage mechanism, credit is not given for inspection.
However the results of testing carried out can be used to update the inputs ( Tref as impact test
temperature) to the damage factor calculation that may result in a change of value. The damage factor
for Brittle fracture is neither time nor inspection effectiveness dependant so it should be monitored with
RBI only for risk screening purposes of different components and rarely for risk mitigation.
The various embrittlement phenomena described in paragraph 3, such as Temper Embrittlement, 475°
Embrittlement, Sigma Phase Embrittlement (or other, such as strain age embrittlement in carbon) must be
considered in brittle fracture screening procedures for the relevant susceptible materials (low alloy Cr-Mo
steels, ferritic stainless steels, austenitic stainless steels and duplex stainless steels).
4.2.4 The Mechanical Fatigue Damage Factor – The Shaking Effect

The RBI assessment to evaluate the mechanical fatigue is limited to piping components that are, no
doubt, the most susceptible components to such damage, especially when subjected to particular service
conditions. For this reason the RBI screening criteria are limited to the following:
• the component is a pipe;
• there have been past fatigue failures in a piping system;
• there is visible/audible shaking in the piping system; shaking and source of shaking can be
continuous or intermittent; transient conditions often cause intermittent vibration; and
• there is a source of cyclic vibration within approximately 15 m and connected to the piping
(indirectly or indirectly via structure), such as reciprocating machinery, chattering PRV, valves
with high pressure drop.
The presence of any of all of the above indicators determines the base susceptibility, which can then be
modified by various adjustment factors (corrective actions, time of shaking, piping complexity, condition of
supports, joint type or branch design, branch diameter).
In API RBI fatigue cracking evaluation credit is not given for inspection and the analysis can be conducted
only for risk screening purposes. This is because most of the time, failure in piping fatigue is in the
initiation phase (crack too small to be detectable) and by the time a crack has reached a detectable size,
the crack growth rate is high, and failure will likely occur in shorter than typical process plant inspection
intervals.
For this reason, the inspection for mechanical fatigue in piping systems should focus mostly on welded
connections and depend more heavily on detection and correction of the conditions that lead to the
susceptibility, than on the detection of the actual cracking damage. Such techniques include:
• visual examinations (VT) of pipe supports;
• VT of any cyclic motion of the pipe;
• VT of the fillet welded supports and attachments;
• VT for all small branch connections with unsupported valves;
WRC Bulletin 533

• surface inspection methods (PT, MT) can be effective in a focused and frequent inspection plans;
• manually feeling of the pipe to detect vibration;
• measuring of piping vibration using special monitoring equipment; and
• VT of a unit during transient condition.
4.2.5 The HTHA Damage Factor – The Time/Temperature/Hydrogen Effect

If a component is made of carbon steel, C-0.5 Mo, Cr-Mo low alloy steels, operated at a temperature
above 204° C with a hydrogen partial pressure greater than 0.55 MPa the component should be
evaluated also for HTHA. The assessment of susceptibility to HTHA is based on the time the component
has been exposed to high pressure hydrogen at elevated temperatures. A single parameter, Pv , has
been developed to relate time at temperature and hydrogen partial pressure.
⎡ P ⎤
PV = log ⎢ H 2 ⎥ + 3.09(10 ) ⋅ (T + 273)(log[age] + 14 ))
−4
(13)
⎣ 0.0979 ⎦
Once the susceptibility is known, the analyst can calculate the DF based on number and effectiveness of
inspection. For such damage, high sensitivity and advanced UT techniques should be used by
experienced personnel; such techniques include AUBT (Advanced Ultrasonic Backscattered Technique)
and attenuation method.
4.2.6 Creep – Life Extension

It is difficult to rank and evaluate plant components in creep regime for Risk Based Inspection purpose
due to the difficulty and the complexity of a damage factor definition and calculation for such a particular
service condition. However, in prioritizing inspections and defining inspection intervals one can (and often
must!) apply the methodology used in the industry for the design and life extension calculation of
components in creep regime such as API 579-1/ASME FFS-1, ISO 13704 (ex API 530), BS-R5, or other
applicable international codes.
4.2.7 Example Of A Damage Factor Calculation

The equipment and operating conditions are defined below.
• Debutanizer Overhead accumulator
• Fabrication date: 01/01/1985
• Design pressure: 11.38 barg
• Design Temperature: 232°C
• Operating pressure: 6.96 barg
• Operating Temperature: 49 °C
• Material of construction: ASTM A 516 Gr 60 1983
• PWHT: yes
• Fluid: Propane/butane
• 0.11 % wt H2S in water phase (1100 ppm)
• No cyanides present
• pH of water phase 8.0
• No History inspection for SCC.
WRC Bulletin 533

From the Table 6 and Table 7 it is possible to estimate a HIC/SOHIC-H2S MEDIUM susceptibility. From
Table 5 the resultant Severity Index, SVI , is 10. Taking into account that there were no SCC inspections
in past history, the base damage factor, is determined from Table 8 to be DFbase = 10 . Combining the
DFbase with the escalation factor for a 2019 projection we obtain:
D f = D fbase ( age ) = 10 ( 34 ) = 484

1.1 1.1
(14)
In accordance with Table 2, a damage factor of this magnitude results in a Category 4 of probability, and
combined with a high (E) consequence of failure (the fluid is light hydrocarbon with toxic species H2S)
corresponds to a HIGH Risk item in a semi-quantitative analysis (see Figure 4).
By setting an intermediate inspection (e.g. in 2015) MTA with magnetic particle testing of 50% of
weldment from inside, and specialized combined UT with 0° probe and angle probe from outside in weld
sample areas and base material sample areas (Usually Effectiveness-grade B), from the same
calculation results in:
D f = D fbase ( age ) = 1( 4 ) = 5
1.1 1.1
(15)
that is in Category 1 of probability and corresponds to a MEDIUM Risk Category in a semi-quantitative

analysis.
5 CONCLUSION
In this article we have tried to present an overview on the possible influence that weldments may have on
reliability during the entire life cycle of process plant equipment and in Asset Integrity Management
methodology. The most Damage Mechanisms oriented methodology in Asset Integrity Management and
Risk Management is probably the Risk Based Inspections (RBI) analysis which is focused on prioritizing
and optimizing inspection strategies on the basis of risk ranking and future projections of risk value.
Risk Based Inspection assessment is considered to be an innovative risk analysis mainly because of the
introduction of two new concepts in the evaluation:
• the role of inspection in providing new information about the condition of the equipment; this may
be better or worse, or the same, as previously estimated but the effect aimed at is to reduce the prior
uncertainty so new information can change the estimated probability of failure.
• the global, analytical and structured approach to the evaluation of all the various damage
mechanisms active in process and power generation plants extrapolated from many years of
industrial experience and expertise applied research.
The aim of RBI is to define and measure the level of risk associated with an item, then evaluate safety,
environmental and business interruption risks in order to reduce risk of failure given by Equation (16) by
the effective use of inspection resources.
Risk s = Cs Fs or (COFs ⋅ LOFs ) (16)
Any reliability evaluation for pressure vessels and piping systems that takes into account all the possible
damage mechanisms active in process plants must deal with welded joints vulnerability and take into
account likely mechanical, corrosion and stress-corrosion related influences. Actually, a welded
component may fail in service by any of the following mechanisms.
• Overload
• Brittle fracture
• Fatigue
WRC Bulletin 533

• Creep
• High Temperature Hydrogen Attack
• Corrosion
• Stress Corrosion Cracking (SCC)
selecting an inspection effectiveness category based on defined level of inspection, a probability of failure
and associated risk may be determined using the following equations that in the calculation of probability
take into account the role of the condition, or of the damaged state, observed by an inspection program
Since the damage factor in the risk calculation is the only time dependant variable, it is possible to project
the calculation of this factor, and consequently the risk, for future planned dates (e.g. future Maintenance
Turn Around-MTA) while introducing a set number of inspections with adequate effectiveness to keep the
damage factor and risk below acceptable level.
It is obvious that the time-dependant damage factor variable is very - maybe the most - important one for
Risk Based future inspections prioritization. The damage factor includes all the factors that influence the
probability of damage and the role of the material and inspection experts.
It is worth noting that in a well developed RBI analysis it is important to identify the damage mechanism.
Here, the role of the different personalities involved in the analysis is of primary importance: only a
qualified and experienced materials engineer has the ability to identify and properly characterize the
damage mechanisms and modes of failure, and only a qualified and experienced NDT engineer can
evaluate the best traditional and advanced inspection methods employed in the past and those to be
suggested for future plan.
RBI is not for a single old style-all knowing expert or consultant but is for structured, skilled and able to
communicate TEAMS!
6 NOMENCLATURE
age is the time since the last thickness reading
Art is the damage factor parameter
CA Consequence impact area or the corrosion allowance, as applicable
Cr ,bm Corrosion rate of the base material
C (t ) is the consequence of failure as a function of time
Df Damage factor, also shown as D f (t ) and D f (t , I E )
D fbase Base Damage factor
D f (t ) Damage factor as an explicit function of time
D f (t , I E ) Damage factor as an explicit function of time and inspection effectiveness
FC Financial consequence
FAM Factor adjustment for tank maintenance per API 653 (for storage tanks)
FDL Factor adjustment for dead legs
WRC Bulletin 533

FIP Factor adjustment for injection points
FMS Management System Evaluation Factor
FOM Factor adjustment for on-line monitoring
Fs Failure Frequency of scenario
FSM Factor adjustment for settlement (for storage tanks)
FWD Factor adjustment for welded construction (for storage tanks)
gff Generic failure frequency of the equipment type
IE Inspection effectiveness
Pf ( t ) is the probability of failure as a function of time
Pf ( t , I E ) is the probability of failure as a function of time and inspection effectiveness
R (t ) is the risk as a function of time
R (t, I E ) is the risk as a function of time and inspection effectiveness
Risks Risk of the scenario
S Scenario
tmin Minimum required thickness per the applicable construction code
trd Last thickness inspection reading
7 REFERENCES
1. A. Dhooge, Failure Analysis Of Welded Structures, Reveu De La Soudre N°2/3, 1992.
2. API RP 571: Damage Mechanisms in the Refinery and Petrochemical Industry, American Petroleum
Institute, Washington, D.C., 2003.
3. API RP 572: Inspection of Pressure Vessels (Towers, Drums, Reactors, Heat Exchangers And
Condensers), American Petroleum Institute, Washington, D.C., 2001.
4. API RP 577: Welding Inspection and Metallurgy, American Petroleum Institute, Washington, D.C.,
2004.
5. API 579-1/ASME FFS-1 2007 Fitness-For-Service, American Petroleum Institute, Washington, D.C.,
2007.
6. API RP 581: Risk Based Inspection Technology, American Petroleum Institute, Washington, D.C.,
2008.
7. Osage, D.A. and Henry, P.A. “Recent Developments and Technology Improvements in API Risk-
Based Inspection Planning Technology”, ESOPE Conference, Paris, France, 2007.
8. P.F. Timmins, Predictive Corrosion and Failure Control in Process Operations, ASM International,
Materials Park, Ohio, 1996.
9. Wintle, K. and Amphlett, S., Best Practice For Risk Based Inspection As A Part Of Plant Integrity
Management, HSE Contract Research Report 363/2001, Copyright Unit, Her Majesty’s Stationery
Office, St Clements House, 2-16 Colegate, Norwich NR3 1BQ.
WRC Bulletin 533

8 TABLES
Table 1 – Damage Factor Numerical Values And Area Based Consequence Associated To
Probability Categories In API RBI (from API 581)
Probability Category (1) Consequence Category (2)
2
Category Range Category Range (m )
1 D f −total ≤ 2 A CA ≤ 9.29
2 2 < D f −total ≤ 20 B 9.29 < CA ≤ 92.9
3 20 < D f −total ≤ 100 C 92.9 < CA ≤ 279
4 100 < D f −total ≤ 1000 D 279 < CA ≤ 929
5 D f −total > 1000 E CA > 929
Table 2 – General Description Of Inspection Effectiveness (from API 581)

Qualitative Inspection
Effectiveness Description
Category
The inspection methods will correctly identify the true damage state in nearly
Highly Effective
every case (or 80–100% confidence).
The inspection methods will correctly identify the true damage state most of
Usually Effective
the time (or 60–80% confidence).
The inspection methods will correctly identify the true damage state about half
Fairly Effective
of the time (or 40–60% confidence).
The inspection methods will provide little information to correctly identify the
Poorly Effective
true damage state (or 20–40% confidence).
The inspection method will provide no or almost no information that will

Ineffective correctly identify the true damage state and are considered ineffective for
detecting the specific damage mechanism (less than 20% confidence).
WRC Bulletin 533

Table 3 – Capability of Applicable Inspection Method for Weld Type Joints (from API 577)
Inspection Methods
Joint
RT UT PT MT VT ET LT
Butt A A A A A A A
Corner O A A A A O A
Tee O A A A A O A
Lap O O A A A O A
Edge O O A A A O A
Legend:
RT – Radiographic Examination
UT – Ultrasonic Examination
PT – Penetrant Examination inlcuding both DPT (dye penetrant testing) and FPT (fluorescent penetrant testing)
MT – Magnetic Particle Examination
VT – Visual Testing
ET – Electromechnical Examination
A – NDE method is applicable
O – NDE method is marginally applicable depedning on factors such as thickness, discontinutiy size, orinetation, and
location
WRC Bulletin 533

Table 4 – Thinning Damage Factors (from API 581)

Inspection Effectiveness
Art 1 Inspection 2 Inspections 3 Inspections
E
D C B A D C B A D C B A
0.02 1 1 1 1 1 1 1 1 1 1 1 1 1
0.04 1 1 1 1 1 1 1 1 1 1 1 1 1
0.06 1 1 1 1 1 1 1 1 1 1 1 1 1
0.08 1 1 1 1 1 1 1 1 1 1 1 1 1
0.10 2 2 1 1 1 1 1 1 1 1 1 1 1
0.12 6 5 3 2 1 4 2 1 1 3 1 1 1
0.14 20 17 10 6 1 13 6 1 1 10 3 1 1
0.16 90 70 50 20 3 50 20 4 1 40 10 1 1
0.18 250 200 130 70 7 170 70 10 1 130 35 3 1
0.20 400 300 210 110 15 290 120 20 1 260 60 5 1
0.25 520 450 290 150 20 350 170 30 2 240 80 6 1
0.30 650 550 400 200 30 400 200 40 4 320 110 9 2
0.35 750 650 550 300 80 600 300 80 10 540 150 20 5
0.40 900 800 700 400 130 700 400 120 30 600 200 50 10
0.45 1050 900 810 500 200 800 500 160 40 700 270 60 20
0.50 1200 1100 970 600 270 1000 600 200 60 900 360 80 40
0.55 1350 1200 1130 700 350 1100 750 300 100 1000 500 130 90
0.60 1500 1400 1250 850 500 1300 900 400 230 1200 620 250 210
0.65 1900 1700 1400 1000 700 1600 1105 670 530 1300 880 550 500
Art 4 Inspections 5 Inspections 6 Inspections
E
0.02 1 1 1 1 1 1 1 1 1 1 1 1 1
0.04 1 1 1 1 1 1 1 1 1 1 1 1 1
0.06 1 1 1 1 1 1 1 1 1 1 1 1 1
0.08 1 1 1 1 1 1 1 1 1 1 1 1 1
0.10 2 1 1 1 1 1 1 1 1 1 1 1 1
0.12 6 2 1 1 1 2 1 1 1 1 1 1 1
0.14 20 7 2 1 1 5 1 1 1 4 1 1 1
0.16 90 30 5 1 1 20 2 1 1 14 1 1 1
0.18 250 100 15 1 1 70 7 1 1 50 3 1 1
0.20 400 180 20 2 1 120 10 1 1 100 6 1 1
0.25 520 200 30 2 1 150 15 2 1 120 7 1 1
0.30 650 240 50 4 2 180 25 3 2 150 10 2 2
0.35 750 440 90 10 4 350 70 6 4 280 40 5 4
0.40 900 500 140 20 8 400 110 10 8 350 90 9 8
0.45 1050 600 200 30 15 500 160 20 15 400 130 20 15
0.50 1200 800 270 50 40 700 210 40 40 600 180 40 40
0.55 1350 900 350 100 90 800 260 90 90 700 240 90 90
0.60 1500 1000 450 220 210 900 360 210 210 800 300 210 210
0.65 1900 1200 700 530 500 1100 640 500 500 1000 600 500 500
Notes: Determine the row based on the calculated Art parameter. Then determine the thinning damage
factor based on the number and category of highest effective inspection. Interpolation may be used for
intermediate values.
WRC Bulletin 533

Table 5 – Determination of SCC Severity Index for Different Mechanisms (from API 581)
Susceptability Severity Index
from Damage
Mechnism Caustic Amine Carbonate SSC & HIC & CI SCC PTA Other
HSC-HF SOHIC
High 5000 1000 1000 100 100 5000 5000 1000
Medium 500 100 100 10 10 500 500 100
Low 50 10 10 1 1 50 50 10
None 1 1 1 1 1 1 1 1
Table 6 – Environmental Severity – HIC/SOHIC-H2S Cracking (from API 581)

Environmental Severity as a Function of H2S content of Water
pH of Water 50 to 1,000 1,000 to 10,000
< 50 ppm > 10,000 ppm
ppm ppm
< 5.5 Low Moderate High High

5.5 to 7.5 Low Low Low Moderate
7.6 to 8.3 Low Moderate Moderate Moderate
8.4 to 8.9 Low Moderate Moderate* High*
> 9.0 Low Moderate High* High*
Note: If cyanides are present, increase the susceptibility to HIC/SOHIC-H2S one category for pH > 8.3
and H2S concentrations greater than 1,000 ppm.
Table 7 – Susceptibility to Cracking – HIC/SOHIC-H2S Cracking (from API 581)
Susceptibility to Cracking as a Function of Steel Sulfur Content
Environmental High Sulfur Steel (1) Low Sulfur Steel (2) Ultra Low Sulfur (3)
Severity
> 0.01% S 0.002 to 0.01% S < 0.002% S
As-Welded PWHT As-Welded PWHT As-Welded PWHT

High High High High Medium Medium Low
Moderate High Medium Medium Low Low Not
Low Medium Low Low Low Not Not
Notes:
1. Typically includes A 70, A 201, A 212, A 285, A 515, and most A 516 before about 1990.
2. Typically includes early generations of HIC-resistant A 516 in 1980’s, with Ca additions.
3. Typically includes later generations of HIC-resistant A 516 in 1990’s.
WRC Bulletin 533

Table 8 – SCC Base Damage Factor as a Function Of The Severity Index, SVI (from API 581)
SVI 1 Inspection 2 Inspections 3 Inspections
E
1 1 1 1 1 1 1 1 1 1 1 1 1 1
10 10 8 3 1 1 6 2 1 1 4 1 1 1
50 50 40 17 5 3 30 10 2 1 20 5 1 1
100 100 80 33 10 5 60 20 4 1 40 10 2 1
500 500 400 170 50 25 300 100 20 5 200 50 8 1
1000 1000 800 330 100 50 600 200 40 10 400 100 16 2
5000 5000 4000 1670 500 250 3000 1000 250 50 2000 500 80 10
SVI 4 Inspections 5 Inspections 6 Inspections
E
1 1 1 1 1 1 1 1 1 1 1 1 1 1
10 10 2 1 1 1 1 1 1 1 1 1 1 1
50 50 10 2 1 1 5 1 1 1 1 1 1 1
100 100 20 5 1 1 10 2 1 1 5 1 1 1
500 500 100 25 2 1 50 10 1 1 25 5 1 1
1000 1000 200 50 5 1 100 25 2 1 50 10 1 1
5000 5000 1000 250 25 2 500 125 5 1 250 50 2 1
Notes: SVI is the Maximum Severity Index determined for each specific SCC Cracking Mechanism
WRC Bulletin 533

9 FIGURES
Figure 1 – Accumulated Equipment Damage During Different Stages Of Life
Risk of
Maximum Tolerable Risk
Failure
Maximum
Operating Risk
Minimum
Operating Risk
Life
Stage 1 Stage 2 Stage 3 Stage 4
Design Requirements
& Fabrication RBI Best Fits
FFS Best Fits
FAD
C
Kr Safe B Unsafe
A
Lr
Failure Assessment
Diagram (FAD)
Figure 2 – Different Stages of Life, Risk of Failure, Effect Of Periodic Maintenance and Inspection
WRC Bulletin 533

ITEM AGEING LIFE
Stage 2 Stage 3 Stage 4

Stage 1
Post
Risk-Based Deterministic Monitored
commisioning
MATURITY AGEING TERMINAL
INITIAL
Accelerating and
Design limits
accumulating
approaching
Operation well within damage
Evidence of active
design limits Beyond design limits
deterioration
Retained corporate and known operating
Repair, refits
knowledge of design/ experience
modifications
manufacture Approach safe
Design and Changes in process/
Ageing damage not operating limits
manufactoring faults use
yet significant Advanced inspection
Installation issues Lack of full history/
Routine mainenance and FFS required to
Early life operating corporate memory
Extended operating detrmine residual life
faults Changes in ownershp/
periods Decreasing intervals
Identification of second hand plant
Selected inpection by between inspections
potential ageing sites Quantitative NDT
risk to confirm Monitoring
First thorough inspection to measure
expectation of slow Major repairs and
examination extent and rate of
degradation refit repalcement
damage accumulation
Updated risk analysis needed
FFS assessment
from experience End of life based on
Degradation rate
Rate of damege costs of repairs or
increased-less
low&known replacement and
predictable
wider economic
factors
Figure 3 – Description of the Four Stages of Progressive Aging
WRC Bulletin 533

RISK
4
High
Probability Medium High

3
Medium
Low
2
A B C D E
Consequence
Figure 4 – Risk Matrix from API 581
Figure 5 – Superimposition Principle for the Calculation of Risk (from API 581)
WRC Bulletin 533

FINANCIAL
Figure 6 – Different RBI Consequences of Failure
WRC Bulletin 533

Figure 7 – Consequence Analysis Procedure in API 581
WRC Bulletin 533

Figure 8 – Overall Event Tree Used in API RBI
WRC Bulletin 533

Figure 9 – Inspection Planning when the Risk Target is Exceeded Between the RBI Evaluation
Date and Future Plan Date, and the Impact of Inspections at Various Level of Effectiveness (from
API 581)
Figure 10 – A Weld as a Collection of Possible Defects
WRC Bulletin 533

Figure 11 – Example of Modeling of Weld Residual Stresses
Figure 12 – Residual Stress Distribution Perpendicular and Parallel To Weld (from API 579-1/ASME
FFS-1)
WRC Bulletin 533

Figure 13 – Overview of Weldment Types of Failure
WRC Bulletin 533

Figure 14 – Brittle Fracture of a Pressure Vessel
Figure 15 – Stress Concentration Effects and Fatigue Failure of a Welded Component
WRC Bulletin 533

Figure 16 – Example of a Localized Creep Failure in a Weld
Figure 17 – HTHA in Correspondence with a Welded Joint Characterized by Internal Microfissuring

Adjacent to the Main Rupture (Isomerization Plant)
WRC Bulletin 533

Figure 18 – Different Morphologies of Weld Preferential Corrosion (HAZ and Weld Metal)
Figure 19 – Different Micro Structural Zones of a Fusion Welded Joint
WRC Bulletin 533

Figure 20 – Chloride SCC in an Austenitic Stainless Steel Weld
Figure 21 – Microbiologically Induced Corrosion in an Austenitic Stainless Steel Weld
WRC Bulletin 533

Figure 22 – Weld Metal Preferential Corrosion in Alloy 400 Piping in Hydrofluoric Acid Service
Figure 23 – Weld Metal Preferential Corrosion in a Hastelloy C-22 Alloy Welded with Matching
Filler Metal, and no Corrosion with Overmatching Filler Metal in “Green Death Solution” Test @
75°C
WRC Bulletin 533

Good detection capability but requires internal access.

Corrosion/Erosion (Internal) Visual Inspection (Vessel Only) - Internal
Limited sizing capability (depth, remaining wall thickness)
General good detection and sizing capability (can be poor

Manual Ultrasonic Testing/0° Probe-External
for isolated corrosion)
Very good detection and sizing capability (application
Automated Ultrasonic Testing/0° Probe-
limited to pipe/vessel walls where simple manipulation
Mapping External
can be facilitated). Slow to apply
Good detection and sizing capability (at specific

Continuous UT monitoring
monitoring locations)
Profile Radiography Very good detection and sizing capability Slow to apply
Surface replication can be used to examine

Creep Damage Surface testing microstructure. Best if assisted by preliminary stress
analysis
Ultrasonic Testing
Methods developed for detection of early stages have not
- Attenuation/loss of back wall echo
proven in the field, Standard UT are suitable at later
- Backscatter
stages
Velocity measurement
Very good detection and sizing capability but access to

Weld root corrosion/erosion TOFD - External
both sides of weld cap required
Manual/Automated Ultrasonic Testing/0° Good detection and sizing capability but requires
Probe-External extensive surface preparation i.e. removal of weld cap
Manual/Automated Ultrasonic Testing/angle

Detection and sizing capability but can be unreliable
probe- External
Detection capability/base material but can give false

Ultrasonic testing – External 0° Probe/High
HTHA indications. Use of mapping system facilitates monitoring.
sensitivity
For welds, removal of cap is required.
Detection capability/welds but cannot detect microscopic

Ultrasonic testing – External Angle Probe/
stages of HTHA. Use of automated system facilitates
medium sensitivity
monitoring of macro-cracking.
Detection capability/welds although discrimination

Ultrasonic testing – External TOFD between macro-micro- cracking and other weld defects
can be a problem
Ultrasonic Testing
- Attenuation/loss of back wall echo
Methods developed for detection of early stages,
- Backscatter
Velocity measurement
HIC H2S Ultrasonic testing – External 0° Probe Good detection at later stages
Ultrasonic testing – External 45° Probe Good detection at later stages
Good detection capability but requires access to fatigue

Fatigue cracking (External/ Internal) Magnetic particle testing crack surface. Good length sizing capability. Some
surface preparation usually required.
Penetrant testing/Eddy Current As Above for non-magnetic materials
Good detection and sizing (length and heigth), enanched

by use of automated system – TOFD gives very accurate
flaw heigth measurement and allows in-service crack
UT/Angle Probes
growth monitoring capabilitiy but requires access to
fatigue crack surface. Good length sizing capability. Some
surface preparation usually required.
Good detection capability but requires access to fatigue

crack surface. Length and some depth sizing capability.
Alternatig Current Field Measurement Unlike MT does not usually require surface preparation
and can be used through coatings. Better for inspection of
welds than Eddy Current
PT/MT/ET techniques – Good detection capability but

Stress Corrosion Cracking/SCC Surface testing
access required to crack surface
Fair detection capability. Can be used on-line. Specialist

UT-External technique have some capability to detrmine crack
features.
On-line detection of growing SCC in large components

system too complex to be inspected by other
Acoustic Emission-External
techniques.Extraneous system noise can produce false
indications.
Figure 24 – NDT Methods for Detecting Different Damage Mechanisms
WRC Bulletin 533

Figure 25 – Field Application of NDT Inspection
Figure 26 – General Procedure for Determination of the Damage Factor (from API 581)
WRC Bulletin 533

Figure 27 – Weld Area Corrosion of Carbon Steel Pipe in Sulfuric Acid Service
Figure 28 – Selective HAZ Corrosion in Crude Oil Tank Bottoms by Acidic Sour Water
Stratification
WRC Bulletin 533

Figure 29 – SCC DF Calculation (from API 581)
WRC Bulletin 533


WRC 533 2010

Uploaded by

Copyright:

Available Formats

WRC 533 2010

Uploaded by

Document Information

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

WRC 533 2010

Uploaded by

Copyright:

Available Formats

BULLETIN 533

jllllllllllllllj Welding Researelt couneil

These Bulletins contain final Reports from projects sponsored

For more information, see www.forengineers.org

Copyright © 2010 The Welding Research Council.

Welding Research Council

The Role of Welded Joint Vulnerability and

WRC Bulletin 533

THE WELDING RESEARCH COUNCIL, INC.

WRC Bulletin 533

ISBN No. 1-58145-540-2

ii WRC Bulletin 533

Martin Prager, Ph. D

WRC Bulletin 533

WRC Bulletin 533

WRC Bulletin 533

2 IS RISK BASED INSPECTION METHODOLOGY REALLY

2.2 Overview Of RBI Methodology (What-Where-How-When Comprehensive Approach)

WRC Bulletin 533

Riskitem = ∑ R(t ) (2)

2.3 RBI Methodology – Probability and Consequence of Failure

Pf ( t ) = gff ⋅ D f ( t ) ⋅ FMS (3)

WRC Bulletin 533

WRC Bulletin 533

2.4 The Aim Of RBI – The Inspection Planning Outcome

R(t , I E ) = Pf (t , I E ) ⋅ CA for Area − Based Risk (4)

R(t , I E ) = Pf (t , I E ) ⋅ FC for Financial − Based Risk (5)

WRC Bulletin 533

3 THE WELDED JOINTS ROLE IN INCREASING THE FAILURE

WRC Bulletin 533

3.2 Brittle Fracture

WRC Bulletin 533

WRC Bulletin 533

3.5 High Temperature Hydrogen Attack (HTHA)

WRC Bulletin 533

3.6 Corrosion & Stress Corrosion Cracking

WRC Bulletin 533

WRC Bulletin 533

• decay (sensitization) caused by carbide precipitation in HAZ;

WRC Bulletin 533

• sulfide stress cracking (SSC);

4 HOW CAN RBI CONTROL WELDMENT TENDENCY TO

4.1 Time & Inspection Effectiveness - The Damage Factor

Pf ( t , I E ) = gff ⋅ D f ( t , I E ) ⋅ FMS (8)

WRC Bulletin 533

b) Component Lining – D clin

c) External Damage – D extd

d) Stress Corrosion Cracking – D scc

f) Mechanical Fatigue (piping only) – D mfat

g) Brittle Fracture – D brit

D f −total = max ⎡⎣ D thin ⎤

WRC Bulletin 533

The calculated damage factor using the Art parameter, D thin

WRC Bulletin 533

⎛ FIP ⋅ FDL ⋅ FWD ⋅ FAM ⋅ FSM ⎞

4.2.2 The SCC Damage Factor – The Escalation Factor

WRC Bulletin 533

so-called Damage Supplements in API RBI).

4.2.3 The Brittle Fracture Damage Factor – The Cold Effect

WRC Bulletin 533