NDE of Composites
NDE of Composites
NDE of Composites
Authorisation Sheet
Report Title:
Customer Reference:
HOIS
Project Reference:
UC000138-01-01-12
Report Number:
HOIS GP1
Issue:
Issue 2
Distribution List:
Open Publication
Updated
Author:
Dr Martin Wall
16/4/2012
Reviewed:
Richard Lee
16/4/2012
Authorised:
Dr Stephen Burch
2 May 2012
ii
Executive Summary
This updated good practice guide covers the inspection and non-destructive evaluation
(NDE) of offshore composite components in-service and includes lessons learnt from several
inspection trials on HOIS FRP samples including the glass reinforced flow loop pipe sample
supplied by Petrobras. Specifically this guide excludes manufacturing quality control, quality
assurance or inspection, for which the reader is referred to ISO 14692.
All components that form part of a GRP piping system (e.g. pipe, branches, bends, tees,
tapers, flanges, fittings and joints) are covered. This includes piping, from 50 mm (2) to over
1 m (40) diameter, and fittings. Specific guidance is provided on composite connections
including adhesively bonded joints, laminated joints and flanged connections. Fittings include
T-joints, bends, branches and flanged connections and may typically be manually overwound
or laminated giving a wide range of wall thickness, 2 mm to 50 mm.
The guide encompasses the recommendations regarding NDE of composites in available
standards including ISO 14692, NORSOK M-622, preceding NORSOK and UKOOA studies,
and relevant API and ASTM standards. Good practice has been included from individual oil
companies design and engineering specifications, HOIS members inputs, and ESR
experience in composite inspection. Relevant published studies and initiatives in the UK
funded by the Department for Business, Innovation & Skills (formerly the Department of
Trade & Industry) and the Technology Strategy Board, TSB have been cited. In addition, a
watching brief of technologies employed in the USA on inspection and monitoring of
composites have also been assessed.
We have restricted consideration to defect types that may occur in-service or be present after
installation. It is assumed that an appropriate inspection and quality control plan has been in
place during manufacture including monitoring the state of cure in line with that in ISO 14692.
Composite vessels and tanks are not specifically included as there is limited use of these
offshore. The recommendations regarding process pipework inspection will however
generally be applicable to vessels and accessibility is often better.
Secondary structures such as ladders, walkways, gratings, and equipment housings are not
included. The major issue here is removal of the protective gel coat in marine environments,
exposing fibres. Visual inspection and painting or refurbishment of the gel coat is usually
adequate here. The inspection methods considered for pipework applications may also be
relevant to these applications if the application is safety critical.
Although this guide refers to offshore components, it is equally relevant to onshore
applications and in several cases the techniques used would be similar.
Inspection of GRP piping is generally more difficult to carry out compared to metal pipes for a
number of reasons. This includes general unfamiliarity by inspectors of inspecting
composites and the anisotropic and heterogeneous nature of these materials causing signal
noise and attenuation. Porosity and the laminated nature of the microstructure account for
some of these effects.
This guide has been updated to take account of lessons learnt from a series of inspection
trials conducted under the auspices of the FPSO and Flexible Risers Working Group on a 6
GRE spool piece (flow loop) prepared by Petrobras. The flow loop contained a number of
intentional bondline defects including paper inserts, incorrect application of adhesive and
incorrect surface preparation. The 3 metre flow loop contains two 90-degree elbows and a
iii
iv
Contents
1.0
SCOPE ........................................................................................................... 1
2.0
3.0
BACKGROUND.............................................................................................. 4
4.0
3.1
Introduction .................................................................................................4
3.2
3.3
3.4
Materials .....................................................................................................6
3.5
Applications.................................................................................................6
3.6
Fabrication methods....................................................................................7
3.7
3.8
3.9
Life prediction..............................................................................................9
3.10
Painting .......................................................................................................9
3.11
REFERENCES ............................................................................................... 9
4.1
4.2
4.3
4.4
5.0
DEFINITIONS ............................................................................................... 11
6.0
Installer requirements................................................................................ 13
6.2
7.0
EQUIPMENT ................................................................................................ 13
8.0
9.0
8.1
Inspection .................................................................................................14
8.2
Overview ...................................................................................................14
10.0
11.0
12.0
13.0
9.2
9.3
9.4
10.2
10.3
10.4
10.5
10.6
10.7
10.8
10.9
11.2
11.3
In-service ..................................................................................................33
Pipework ...................................................................................................36
12.2
Fittings ......................................................................................................38
12.3
Supports ...................................................................................................38
12.4
Joints ........................................................................................................39
12.5
12.6
12.7
12.8
Repairs .....................................................................................................45
Delaminations ...........................................................................................45
13.2
13.3
13.4
13.5
13.6
Materials degradation................................................................................ 52
13.7
vi
14.0
14.2
14.3
14.4
Ultrasonics ................................................................................................57
14.5
Radiography..............................................................................................62
14.6
14.7
Thermography...........................................................................................67
14.8
Shearography ...........................................................................................69
14.9
14.10
Acousto-Ultrasonics ..................................................................................72
14.11
14.12
15.0
16.0
17.0
18.0
19.0
ACKNOWLEDGEMENTS ............................................................................ 80
Appendices
APPENDIX A GRP CODES AND STANDARDS................................................ A-82
APPENDIX B INSPECTION GUIDANCE FROM ISO 14692 .............................. B-85
APPENDIX C DAMAGE MECHANISMS IN COMPOSITES .............................. C-97
APPENDIX D EXAMPLES OF IN-SERVICE DEGRADATION OF COMPOSITE
COMPONENTS ......................................................................... D-110
vii
1.0 Scope
This good practice guide covers the inspection of offshore composite components in-service
and after installation. The focus is on inspection of low to medium pressure process
pipework (~10-50 bar), the primary application of composites offshore. This good practice
guide specifies additional and optional requirements to ISO 14692 (all parts). All components
that form part of a GRP piping system (e.g. pipe, branches, bends, tees, flanges, and joints)
are covered. The good practice is directed towards piping systems in GRP materials used on
offshore production platforms, but may also be used for similar onshore systems dependent
on the location and fabrication method.
Composite materials in the context of this good practice guide are defined as, and limited to,
fibre reinforced thermosetting matrix (or resin) systems. These have several features that
make them attractive for use in the Oil and Gas industry, namely ease of installation, light
weight and good corrosion resistance. Glass fibres are the most common reinforcement and
epoxy resins are primarily used as the matrix material. Other types of resins can be used
depending on the application and in many cases corrosion resistant liners are used which
may affect the inspection results obtained.
Conventional inspection strategy and practice is described. In addition, this document
provides guidance on selection and how to apply non-destructive examination (NDE) and
testing (NDT) methods; an area where very limited guidance is available in existing
standards such as ISO 14692. Newer NDE methods such as microwave inspection, acoustoultrasonics, tap-testing, ultrasonic B-Scan and laser shearography are also included in
addition to those found in the main industry standards ISO 14692, NORSOK M622 and the
ASME Boiler and Pressure Vessel Code Section V. Such methods are widely used
elsewhere for inspection of composites and are starting to find application in the Oil and Gas
sector.
All components that form part of a GRP piping system (e.g. pipe, branches, bends, tees,
tapers, flanges, fittings and joints) are covered. This includes piping, from 50mm (2) to over
1m (40) diameter, and fittings. Specific guidance is provided on composite connections
including adhesively bonded joints, laminated joints and flanged connections. The guidance
is applicable both to high quality filament wound pipework, and to piping systems that have
been manufactured by manual overwinding or hand laminated giving variability in thickness
and surface quality. Wall thicknesses ranging from 2mm to 50mm are considered. Higher
wall thicknesses are generally associated with fittings or connections.
Inspection of GRP piping is generally more difficult to carry out compared to metal pipes for a
number of reasons. This includes general unfamiliarity by inspectors of inspecting
composites and the anisotropic and heterogeneous nature of these materials causing signal
noise and attenuation. Porosity and the laminated nature of the microstructure account for
some of these effects.
This document is directed towards GRP piping and components used topside on offshore
production platforms, but may also be used for similar onshore systems. The same
technologies have applicability to subsea, downhole and pipeline applications of composites,
though there will be issues in terms of access and marinisation if carried out in-situ.
The following generic component types have been considered in the development of the
NDE recommended practice:
Connections
Following installation;
The role of inspection is to assure technical integrity during operations and form a basis for
maintenance evaluation/planning. It may also serve to provide a fitness-for-purpose
evaluation and contribute to the improvement of current and future designs and inspection
strategy. Composites are becoming used in an increasing range of structural and process
applications, and at higher pressures and more severe environments. For this reason NDE
in-service is becoming more common.
3.0 Background
3.1
Introduction
Fibre reinforced plastic (FRP) composite materials are seeing increased use in the Oil and
Gas industry. Composite materials have been used in chemical, processing and refinery
applications for over 40 years primarily in containment applications. The first application in
the North Sea was in the early 1980s 1. Corrosion resistance, light weight and in some cases
design flexibility and continuous manufacture are the primary business drivers, which when
used to advantage in design, can lead to either reduced life-cycle costs or improved safety.
There are also advantages in ease of fabrication, mechanical and chemical properties and
cost.
Composites are fundamentally different to metallic materials, and provided the correct resin
and fibre types have been selected and the components are properly installed and qualified,
few issues are normally encountered in service. Particular problems may be encountered
with adhesive joints especially those made in the field. Material selection in corrosive
environments is a specialist area and end users are reliant on service experience and advice
and ratings supplied by the main composite suppliers.
3.2
HSE research Report RR039 Cost Effective use of composites offshore; Part B: Summary of the Joint industry industry programme on the cost effective use of fibre reinforced composites offshore. UK Health & Safety
Executive HSE; HSE Reports, 2003 http://www.hse.gov.uk/RESEARCH/rrhtm/rr039.htm
3.3
The main factors affecting inspection of composites offshore are access, surface finish,
material quality, complex geometry and thickness. These and other factors such as
unfamiliarity with composites may limit the applicability of traditional NDE methods such as
ultrasonics and favour composite or polymer specific methods such as microwave inspection,
thermography, acoustoultrasonics or radiography for complex geometries and thick
sections. An advantage is the fact that composites generally have good defect tolerance and
fatigue properties.
The way in which NDE methods are applied is similar to metallic materials though the types
of defects that may be present are different. Delamination and disbonding are the most
important defects compared to cracking in metallic materials. Composites are damage
tolerant and can support a significant amount of damage compared to steel where an
individual fatigue crack can become critical and lead to failure.
Composite materials by their construction and diverse methods of fabrication pose some
physical challenges and benefits to inspection as some are transparent or translucent. Their
mechanical properties are generally anisotropic (in many cases they are orthogonally
anisotropic, i.e. orthotropic) due to the methods of lay-up and the different properties of
polymer and reinforcement. Surface roughness is generally higher than would be expected
for a metallic pipe and hence there may be difficulties in probe coupling. Composite materials
are usually more highly attenuating than steel and lower frequencies are used when
inspecting using ultrasonics.
There are differences associated with individual NDE methods. Visual inspection can show
up a wider range of defects in composites. With suitable illumination it is possible to look
through some composite components to see internal defects. Ultrasound is more highly
attenuated by composites due to the many internal interfaces and porosity so it is normally
necessary to use lower frequencies. The surface finish can also pose difficulties in coupling.
Common electromagnetic methods such as eddy current and magnetic particle inspection,
MPI, are not applicable to glass-fibre reinforced epoxy (GRE) composites as the materials
are non-conducting. Composites are less absorbing of X- or gamma- rays so it is necessary
to use lower energy sources or less penetrating isotopes. Thermal diffusion is slower than in
metals which simplify thermography inspection. Long wavelength methods such as
microwaves which would cause reflection on metal surfaces are well suited to polymers and
composites because of their dielectric properties and better match of wavelengths to
microstructural differences. Similarly surface strain measurements such as laser
shearography are easier because of the lower stiffness and the ease of distortion of
composites compared to metals.
Composite components are often painted for cosmetic or protective reasons. This may be
for pipe identification and also provide some additional protection against ultra-violet rays
(UV) or from the external environment. A consequence is that it is no longer possible to
inspect the component visually using internal illumination, a standard method. In this case
detection and monitoring of service damage may be more difficult.
On FPSOs process equipment can be more closely packed than on conventional platforms
making access for NDE difficult.
3.4
Materials
The composite materials used in the offshore industry are primarily glass fibre reinforced
plastic (GRP) which can be used in tanks, vessels or piping systems. A variety of glass fibre
types and resins may be used dependent on environment and conditions.
The primary fibre used is glass, although carbon and aramid are used in limited applications.
The primary resin system (thermosetting) used is epoxy or polyester, although vinyl esters,
polyurethanes and furanes are also used to a limited extent. The method of manufacture is
predominantly filament winding implying continuous fibre composites, although some
components are pultruded, resin transfer moulded or made by hand lay-up.
3.5
Applications
piping systems;
process equipment;
access equipment (stairs, gratings);
beams;
modules and secondary structures;
tubing and casings;
tanks and vessels;
lifeboats;
risers;
pipelines;
mudmats;
protective covers.
Figure 1 Example of GRP pipework offshore. Left, water treatment and firewater mains
GRP components, Petrobras P50 FPSO (Courtesy Petrobras); right, GRP firewater mains.
3.6
Fabrication methods
Commercial grade composites in these industries are normally produced by filament winding,
with increased use of moulding processes such as resin transfer moulding (RTM). Manual
overlay, lamination or moulding is used on filament wound pipe in the tapered regions
associated with nozzles, flanges, manifolds and attachments.
3.7
Correct handling and installation is important for composite components. There is a risk of
surface or impact damage. Adequate supports are required for pipework, in accordance with
the guidance in ISO 14692. Care is needed in preparation of adhesive and laminated joints
to ensure good alignment, clean surfaces and adequate distribution and curing of adhesive.
A common issue for flanged joints is over-tightening leading to overstressing of joints and
flange cracking. The integrity and leak tightness of piping systems is usually verified by
pressure testing. Composite pipework is normally subjected to thorough inspection and
acceptance criteria at the manufacturing stage in accordance with ISO 14692. There is the
likelihood that some manufacturing flaws, usually benign, may carry through to service in the
nature of the fabrication methods used.
3.8
Issues in-service
Composites do not corrode in the conventional sense but can be subject to a number of
degradation mechanisms in-service, including: physical ageing, mechanical ageing and
chemical ageing. The consequence of these can be a reduction of 20 - 40% or greater in the
strength characteristics of the polymer during the lifetime of the component and introduction
of damage including matrix cracking and delaminations. This is handled in design codes by
use of regression curves based on short term and longer term (typically 1,000h and 10,000h)
tests to determine the qualification pressure for the component and the allowed operating
pressure over the design life.
There is concern about whether such methods of life assessment are sufficiently robust,
given the increasing diversity of applications in which composites are applied. In contrast to
steel vessels or pipework, where non-destructive methods such as ultrasonics,
electromagnetics and radiography are widely applied, very little inspection other than visual
Ageing of composites in oil and gas applications, S Frost; Ageing of Composites, Ed. Rod Martin, Chapter 14 p
375-395, Woodhead publishing, ISBN 978-1-84569-352-7, 2008
3
ISO/TS 24817 Petroleum, petrochemical and natural gas industries- Composite Repairs for pipeworkQualification and design, installation, testing and inspection; 2006
3.9
Life prediction
In the onshore and offshore Oil and Gas industries a more robust approach is developing2 to
life prediction and extension, based on materials characterisation and non destructive
evaluation of the actual damage in service components. This offers to improve the accuracy
of life prediction and reduce the risk of premature failure. There would be benefits in
extending such good practice more widely in the offshore and onshore industries.
3.10
Painting
External painting is not required since GRP is not subject to atmospheric corrosion. If
painting is necessary for other reasons, the surface should be lightly blast-cleaned before the
paint system is applied and painting should be carried out after inspection and test of the
component. Painting is detrimental to inspection as it impedes visual inspection methods
particularly where internal illumination is used.
3.11
In chemical and onshore processing applications, pressure vessels and piping are often lined
with polyethylene or other resistant polymers. Most GRP pipework offshore is unlined. There
are particular issues for lined or painted GRE vessels or pipework. If a lining is used then a
compromise may be made on the resin and fibres used in the GRE vessel. These may not
be as resistant as would be used if the GRE was exposed to the environment. It is important
in this case to monitor the lining condition since lining failure could lead to degradation and
failure of the GRE vessel in a shorter timescale than might typically occur for an unlined
vessel exposed to the same environment.
4.0 References
The following codes and standards have been considered and have been referred to in this
good practice guide. The latest issue of the references shall be used unless otherwise
agreed. Other recognized standards may be used provided it can be shown that they meet or
exceed the requirements of the standards referenced below.
4.1
Standard;
DNV-OS-C501,
Composite
4.2
4.3
ASTM D2563 Standard Practice for Classifying Visual Defects in GlassReinforced Plastic Laminate Parts.
ASTM E2191 - 08 Standard Practice for Examination of Gas-Filled FilamentWound Composite Pressure Vessels Using Acoustic Emission.
ASTM D 2563 Standard Practice for Classifying Visual Defects in GlassReinforced Plastic Laminate Parts.
NDT Procedures
The following procedures have also been developed but have not reached full standards
recognition:
10
4.4
Concise good practice guides have recently been produced by HSE on FRP pipe and
composite overwrap repairs. Designed for HSE inspectors, these provide simple summaries
on what can go wrong, damage mechanisms and normal inspection practice (based on ISO
14692). Photographs are included showing damage mechanisms. The latter included advice
from members of the Association of Composite Repair Suppliers AcoRes originally formed
by ESR Technology.
HSE GRP pipe fact sheet, Health and Safety Laboratories HSL; Final draft
Revision 1, 2009.
FRP composite repair fact sheet, Health and Safety Laboratories HSL; Draft
Final, 2009.
5.0 Definitions
For the purposes of this practice document, the following terms, definitions and abbreviations
apply in addition to those in ISO 14692 Part 1.
E-glass
ECR or C-glass
Fittings
Hand lay-up
In-field hydrostatic test Short term hydrotest after installation, used as a leak test.
Defined as 1.5 times the system design pressure.
Mill hydrostatic test
Phenolic
Pipeline system
Piping components
11
Regression curve
Decay curve based on short and long term testing at the design
stage to show the predicted degradation in materials properties
and safe operating pressure during the design life of the GRP
component (See ISO 14692 for use and derivation).
R-Glass
S-Glass
Nominal diameter
DSC
DWSI
ECR
EX
FRP
GRE
GRP
GRVE
HSE
MAWP
MW
Microwave
NDE
Non-destructive evaluation
NDT
Non-destructive testing
NPD
PE
Pulse echo
PED
RT
Radiographic testing
UT
Ultrasonic testing
TT
Transient thermography
12
Installer requirements
All pipe, fittings and related items shall be installed by qualified GRP pipe fitters and
thereafter approved by a qualified GRP piping inspector. GRP pipe fitters and GRP piping
inspectors shall be qualified according to the minimum requirements detailed in Annex D of
ISO 14692 Pt 4.
As an alternative, the pipe fitters, supervisors and inspectors maybe qualified in accordance
with another internationally recognised certification scheme based on acceptance by the
company.
6.2
NDT personnel
NDT Personnel should be qualified in accordance with a recognised international code and
standard such as EN 473.
It is recommended that all personnel carrying out visual or NDE inspection of GRP
components have appropriate certification (ASTM, CSWIP etc) for the NDE method and
specific experience in the inspection of GRP components.
Routine inspections may be carried out by a Level I inspector with Level II supervision. More
sophisticated NDE methods such as transient thermography or shearography should be
undertaken by a specialist practitioner in the NDE method, preferably with at least Level II
certification.
NORSOK M622 recommends that ultrasonic inspectors shall be qualified according to EN
473, Level 2 or equivalent. Additionally, they shall also have had specific training for GRP
pipe joints in the ultrasonic test method to be used.
Within Europe, NDT inspectors of joints in piping systems falling into category III and IV
(equipment/vessels) according to PED, shall be approved by a 3rd party organisation
recognised by an European Union (EU) member state.
7.0 Equipment
The NDE equipment to be used for the inspection shall be portable and rugged enough for
the intended service. Equipment intended for laboratory use will normally not be suitable for
field use. In particular moisture is detrimental. If outdoor testing is performed, the necessary
precautions shall be taken to protect the equipment from rain, wind etc.
Most offshore platforms have EX 1 zones in which no electric equipment that can produce
sparks is allowed. The operator of the NDE equipment shall ensure that the equipment to be
used fulfils the EX requirements, or obtain special permission from the safety department on
board to execute the inspection in special zones, in shutdown periods, etc.
13
8.1
Inspection
In general, all safety precautions relevant to the NDE method being deployed should be
followed as in the test procedure. Individual NDE methods such as radiography,
shearography, thermography or microwave inspection will have their own specific safety
requirements.
8.2
Dust hazards
During machining of GRP, a dust mask and adequate work clothing should be worn in order
to prevent inhalation of, or skin irritation by, the glass-fibre dust produced. Machining should
be done in a well-ventilated room or in the open air in order to minimize contact with dust. In
the workshop a portable dust extraction unit should be used with the point of extraction as
close as possible to the work.
Overview
Composite components are subject to ageing by a number of processes which can reduce
the strength and properties of the pipework. This includes physical ageing, mechanical
ageing and chemical ageing. These are normally allowed for in design through regression
curves, encompassed in relevant design standards such as ISO 14692. A detailed overview
of damage mechanisms in composites can be found in Appendix C.
Physical ageing processes include moisture ingress, swelling and plasticization common with
other polymeric components. These processes are referred to as static fatigue. Mechanical
ageing refers to the development of defects during service including matrix cracking,
delaminations and impact damage. Chemical ageing refers to environmental processes that
change the chemical structure or bonding within the component and therefore degrade its
physical properties; this includes hydrolysis, and modification of bonding or cross linking.
These ageing processes apply equally to adhesive as well as the composite resin and result
in a change in the glass transition temperature tg and properties.
In its broadest definition ageing can be defined as the reduction in performance of a
component as a function of the applied conditions. The three primary causes of ageing for
composite components in the Oil and Gas industry are through chemical species ingress,
elevated operating temperature and length of time of load application.
Service experience in the Oil and Gas sector is that damage to the composites from ingress
of the environment (internal or external) is minimal and does not significantly affect materials
properties. Progressive damage may occur under service loadings by matrix cracking, the
normal response of composite materials to loading.
For process pipework the main service risk is weepage of the process fluid eventually
leading to failure of component. The other failure mode of concern is fibre failure. Generally,
this occurs at the ultimate load-bearing capacity of the composite component and results in
gross failure. The ageing process accelerates the failure process, be it increasing the density
of micro-cracks, affecting the glass transition temperature Tg, changing the physical
properties of the matrix, or reducing the strength of fibres.
14
9.2
Manufacturing defects
The possible causes of manufacturing flaws, and an overview of NDE techniques suitable for
detecting these defects, are summarized in the Tables in ISO 14692 and NORSOK M-622
together with acceptance criteria. These tables are included in Appendix B.
Manufacturing and handling defects are outside the scope of this guide except insofar as
they carry over into the service component and affect service life or the properties in service.
9.3
Defects that can potentially occur during handling and installation are summarised in ISO
14692 Pt. 4, Table 4, (see Appendix B). The main issues of concern to operators are impact
damage and problems with adhesive joints.
Defect types include the following:
9.4
Misaligned joints;
In-service defects
The relevant in-service defects for which NDE inspection is considered in offshore vessels,
tanks, process pipework and fittings are as follows:
Matrix cracking;
Delamination;
15
10.0
Disbonding;
Impact damage;
Fatigue;
Fibre failure;
Cavitation;
Significant cracks;
Inspection Strategy
The operator shall define an inspection strategy for the GRP systems and components to
identify system criticality and the requirements for inspection. This shall cover:
Inspection shall be linked to possible failure modes and mechanisms identified in the design
or experienced on installation or in-service. The strategy shall at least contain:
inspection frequency;
acceptance criteria;
reporting routines.
Guidance on the development of an inspection strategy for GRP components can be found in
ISO 14692 Part 2 Annex H, which groups components for inspection in terms of criticality,
probability of failure and consequences. This is referred to below and the relevant Tables
are reproduced in Appendix B. NORSOK M-622 includes additional and optional
requirements beyond that provided in ISO 14692 as well as a specific and different grouping
for pre-fabricated pipe spools and adhesive joints. The Operator should specify which
grouping method is being followed.
Specific guidance on inspection for other reasons, such as life extension, defect detection or
due to service reasons is provided later in this Section.
16
10.1
Manufacturing inspection
The possible causes of manufacturing flaws, and an overview of NDE techniques suitable for
detecting these defects, are summarized in the Tables in ISO 14692 and NORSOK M-622
together with acceptance criteria (See Appendix B).
Manufacturing processes used to produce fittings are typically more complicated and less
automated than those used to manufacture pipes. The manufacturing problems which may
occur tend, therefore, to be more prevalent in the fittings, and NDE of fittings should be
prioritised.
Manufacturing and handling defects are outside the scope of this guide except insofar as
they carry over into the service component and affect service life or the properties in service.
10.2
Documentation required
All relevant as-built drawings and records shall be available and maintained. It is
recommended that as a minimum these include the following details:
The supplier shall provide the installer with the following information, which shall include but
not be limited to:
a) Operating and Design parameters:
design pressure;
design temperature;
Tg of the resin used in component manufacture;
Tg of the adhesive used in component manufacture (if appropriate);
qualified pressure of each component and minimum qualified pressure in each
piping system;
mean and maximum velocity conditions in each piping system;
chemical resistance limitations, if applicable;
procedures to eliminate or control water hammer and cavitation, if applicable;
fire classification and location of fire-rated pipe, if applicable;
conductivity classification, location of conductive pipe, earth linkage/grounding
requirements and location of earthing points;
criticality.
17
10.3
The installation of composite structures shall be carefully planned. It shall be part of the
design analysis. Handling composite structures like metal structures may introduce severe
damage. Any aspects of handling that deviates from typically practice with metal structures
should be identified. Procedures should be in place to describe special handling
requirements for composites.
Handling of composite structures requires special care. Handling instructions should follow
each component. Point loads should be avoided. Scraping, wear and tear should be avoided.
Bending the structure into place should be avoided. Lifting shall only be done at specially
indicated spots that were designed to take such loads.
Inspection requirements after installation are covered in Sections 5.1 and 5.2 of ISO 14692.
This part of ISO 14692 assumes that the fittings and pipes have been correctly manufactured
and inspected according to the criteria given in ISO 14692-2. The handling of the GRP
components shall follow the guidelines given in Annex B of ISO 14692 Part 4 and the
requirements of the pipe manufacturer. All piping components shall be visually inspected in
accordance with Table A.1 of ISO 14692 Pt. 4 for damage that may have occurred during
storage and shipment. Rejected components shall be replaced. If doubts concerning the
extent of defects occur during inspection, a specialist approved by the operator shall perform
a second inspection of the delivered items.
Upon arrival at site the packaging shall be checked visually for possible transport damage.
Vessels should be handled and stored in the original packing for as long as possible to avoid
possible damage. The vessel shall be inspected after unpacking.
External surface cracks (e.g. caused by the hydrostatic pressure test, transport or storage)
shall not exceed Level III of ASTM D2563.
All piping components shall as far as possible be installed so that they are stress-free.
Therefore:
bending of pipes to achieve changes in direction, or forcing misaligned
flanges together by over-torquing bolts is not permitted;
the manufacturers recommendations for bolt-torquing sequence, torque
increments and maximum bolt torque shall be followed.
All installation activities shall be verified independently for high safety class components.
Whether the verification shall be done by the manufacturer himself, by the customer, or by a
third party should be decided by the project.
10.4
Traditionally, most GRP piping applications have been inspected visually and the quality
assessed by pressure testing prior to commissioning. Once commissioned no further
inspection has been performed.
18
The following routine quality assurance/control measures are intended to help in ensuring
that GRP piping systems are installed without problems. A suggested inspection strategy for
GRP piping systems after installation, which considers system criticality and availability/
accessibility, is illustrated in Figure 2 taken from NORSOK M-622 (1999). This should be
used as the basis for developing an appropriate specific strategy for a particular installation.
The limitations noted above are addressed by:
19
Piping or Tank
system to be
inspected
QC to Standard by
Supplier (1), PreFabricator (2),
Installer (2)
Is system
critical?
(3)
Are
QC findings
acceptable
(9)?
No
Yes
No
No
Yes
Is system
available/
testable?
(4)
Visual Inspection
(5)
Visual Inspection
(5)
Yes
Additional NDT
(7)
Visual Inspection
(5)
Yes
Inspection
Findings
Acceptable?
(10)
No
Fail
Functional Test
(8)
Pass
Functional Test
(8)
Pass
Pass
Fail
Fail, Replace,
Redesign
Fail
System
Acceptance
Fail, Replace,
Redesign
Figure 2 Inspection Strategy for manufacture and installation of GRP piping and tank
systems based on flowchart in NORSOK M622
20
10.5
In-service inspection
21
A system is considered ready and available for pressure testing if it is: physically
accessible; and not prohibitively expensive to prepare for pressure testing (i.e. blinding off
joints, blocking deluge nozzles, etc.).
In projects where more than one set of Regulatory Authorities' rules apply or several
contractors are involved, only one inspection strategy and one common inspection
programme shall apply for the GRP piping system.
Unless there are other specific reasons for carrying out NDE inspection, it is recommended
the in-service inspection strategy for the GRP components shall be developed in accordance
with the guidance given in ISO 14692 Pt. 3 Annex H and Tables H.1. and H.2. (see Appendix
C). Table H.1 gives suggested inspection programmes based on the likelihood of defects or
degradation occurring and the criticality of the system. The interactions between materials
and process conditions should be considered when selecting condition-monitoring methods.
This entails a comprehensive materials engineering evaluation that considers the most
probable failure/degradation mechanisms and defects.
Equipment shall be classified into Inspection Groups (A-D) given in Annex H of ISO 14692
Pt. 3 based on equipment classification (or criticality), probability of failure, and severity
(consequence) of failure. The selection of NDT methods and inspection intervals shall take
account of the recommendations for the Inspection Group given in Table H.2. Suggested
inspection intervals are given in Table H.2.
It is recommended that an initial selection of NDE methods for use in detecting defects which
are most likely to occur during operation of GRP piping systems is made using the Tables in
ISO 14692 Pt.4 Table 4 along with the recommended acceptance criteria. Possible causes
and recommended corrective actions are also included.
Relevant non-destructive testing (NDT) methods should be selected if possible, bearing in
mind the possibilities and limitations of each method. A combination of several methods may
be required in order to achieve safe and cost-effective utilization of the equipment/ system.
Since ISO 14692 was issued a number of new NDE methods have become established.
More detailed guidance on Selection of NDE methods encompassing these new methods is
given below in Section 11.0.
10.6
Concise good practice guides 4,5 on inspection of GRP pipe and FRP composite overwrap
repairs have been produced by HSE. The latter includes advice from members of the
Association of Composite Repair Suppliers AcoRes, originally formed by ESR
Technology.
4
5
HSE GRP pipe fact sheet, Health and Safety Laboratories HSL; Final draft Revision 1, 2009
FRP composite repair fact sheet, Health and Safety Laboratories HSL; Draft Final, 2009
22
10.7
It is also noted that in the case of inspection indications, which may be false calls or genuine
defects, a plan should be executed listing suggested actions to be taken depending on the
type of findings. The plan may be included in the inspection philosophy. Inspection
procedures shall be defined for:
Manufacturing control;
Detection of damage due to accidental loads or overloads;
Detection of damage due to unexpected high degradation of long term properties.
Inspection shall be linked to possible failure modes and mechanisms identified in the design.
10.8
The inspection programmes in Table H.2 of ISO 14692 Pt 3 include the use of destructive
testing of material samples to characterize long-term material degradation under the most
aggressive operating conditions, and as a justified means to extend GRP equipment past its
rated life. Such material samples should be representative of the equipment in-service, i.e.
by testing a pipe sample removed from service, or by testing coupons which have been
exposed to the same media and stress levels that are seen in service. If the initial materials
engineering evaluation indicates that destructive tests are required, the same test methods
as those used to pre-qualify the material should be used. More guidance on the assessment
of ageing is given in Section 13.6 below.
The Tables from ISO 14692 referred to above are reproduced in Appendix B.
23
Reason for
inspection?
Other
Inspection
Group in ISO
14692 Table
H.1.?
D
Inspect as Group D
in Table H.2 of ISO
14692 Pt/ 33
A-C
Inspect according to the
relevant Group A, B or C
in Table H.2 of ISO 14692
Pt/ 3
Visual Inspection
internal/ external at
intervals of 0.3 x
service life. First
inspection after 1-2
years
Visual inspection
internal/ external
Define next
inspection interval
Identify
degradation
mechanisms and
suitable NDE
Service life
beyond
original
estimated
service life?
NDE Inspection
No
Modify or change
inspection group.
See ISO 14692
pt. 3 Table H2?
Yes
Assess defects.
Replace, Repair
Defects
Found?
Yes
Destructive testing of
relevant components
or NDE to assess
current condition and
potential for life
extension
No
System assumed OK
for continued service
Define new
operating life and /or
safe operating
pressure
Figure 3 Inspection Strategy for in-service inspection of GRP piping and tank systems
taking account of guidance in Annex H and Tables H.1. and H.2. of ISO 14692 Pt 3.
24
Life
Extension
To satisfy
ISO 14692
Reason for
inspection?
Defect detection,
Other
Remove selected
components for
destructive testing
or carry out NDE
Estimate actual
regression in
condition
Define next
inspection interval
Accessible for
inspection?
Yes
No
Compare with
initial regression
curve (ISO 14692)
Visual inspection
internal/ external
Modify or change
inspection group.
See ISO 14692 pt.
3 Table H2?
NDE Inspection
Yes
Defects
Found?
Assess defects.
Replace, Repair
No
System assumed OK
for continued service
Figure 4 Inspection Strategy for in-service inspection of GRP piping and tank systems for
life extension, damage detection or other reasons (Non ISO 14692 Pt. 3 inspections).
25
10.9
When to Inspect
GRP piping systems are often used in systems that are not safety-critical and which may be
classified as ANSI/ASME B 31.3 Category D systems requiring no inspection. However,
these systems can be crucial in maintaining uninterrupted production. Therefore, the choice
of when to inspect is largely an economic question.
Unless there are specific reasons to inspect more frequently, it is recommended that the
guidance in ISO 14692 Pt 3, Annex H referred to above is used to define inspection
frequency. This will be dependent on the group (A-D) defined in Annex H for the component.
The probability and consequences of system failure must warrant the added cost of
inspection. For prefabricated pipe spools or adhesive joints the guidance on inspection timing
and grouping of components according to criticality in Table 3 of NORSOK M622 (2005) may
be followed as an alternative.
The specific guidance in NORSOK M622 (2005) is as follows:
The GRP piping systems shall be inspected within the first year (group 2 and group 3
systems) and within the second year (group 1) after start of service. The inspection interval
thereafter shall be 1 year to 2 years for group 2 systems and group 3 system and 3 years for
group 1 systems. The inspection intervals shall be adjusted, i.e. reduced or increased
depending on observed severe degradation or gained confidence in materials and
construction. An increase of inspection intervals can be considered after 5 years of service.
Destructive testing is required if the service life is extended beyond the originally estimated
service life.
Economic and risk considerations will not only determine whether a GRP system is inspected
at all, but also whether it should be periodically inspected while in service. A suggested,
reasonable balance between costs and benefits of inspections is that both non-critical and
critical piping systems should at least be visually inspected within 1-2 years after start of
service. Following this the frequency of inspection should be according to the developed
strategy.
To ensure satisfactory operation of GRP pipework and vessels over a period of years,
periodic inspections shall be performed to check that the GRP material is sound. This
inspection shall include connections and branches to the wall, bottom corner, supports and
the inner liner, if present. Although visual inspection is the most common method of
inspection, other non-destructive techniques such as ultrasonics and radiography are being
developed with increasingly reliable results.
During inspection, damage to surfaces should be avoided by suitably covering footwear and
ladders. Cleaning processes shall be checked to ensure that the internal or external
protective surfaces will not be damaged or destroyed by incorrect application.
11.0
The choice of NDE method which is practically applicable to a given component will depend
on a number of factors including the access, wall thickness and surface conditions. Available
inspection methods may not detect all critical failure mechanisms. However, the methods
may detect preceding failure mechanisms. A link between detectable failure mechanisms
and critical failure mechanisms shall be established. The reliability and functionality of all
inspection methods should be documented.
26
Since ISO 14692 was issued in 2002 a number of new NDE methods have become
established for inspection of composites such as microwaves, shearography and acoustoultrasonics. The guidance in Tables 4 and 5 of ISO 14692 Pt. 4 forms a basis for the initial
selection of NDE methods. However, the recommendations on NDE are general, do not
define the specific NDE method to be used (for example stating ultrasonics), and do not
encompass more recent developments.
For this reason it is recommended that Table 1 and Table 2 below are used for selection of
suitable NDE methods. In these Tables the advice in ISO 14692 Pt 4 has been updated to
take account of other potential NDE methods that are now available. For specific defect
types Table 12 of ISO 14692 Pt. 2 should also be considered (see Appendix C). This Table
refers to manufacturing inspection, but the recommendations are still relevant in-service and
following installation provided access is sufficient. More detailed guidance on the selection
of NDE methods and NDE practice for specific components and defect types is given in
Sections 12.0 and 13.0 below.
Section 14.0 gives advice on how to practically apply individual NDE methods as well as the
current status in regard to application offshore.
Relevant non-destructive testing (NDT) methods should be selected if possible, bearing in
mind the possibilities and limitations of each method. A combination of several methods may
be required in order to achieve safe and cost-effective utilisation of the equipment/ system.
27
Table 1
NDE selection and acceptance criteria for defects potentially occurring during fabrication, handling and installation
(Update of Table 4 from ISO 14692 Pt. 4, additions shaded).
Possible defects
Cause(s)
Consequence(s)
Recommended
NDT method(s)
ISO 14692
Criteria
Corrective action
Comments
Incorrect
dimensions
Incorrect
prefabrication.
Joint cannot be
sealed, leakage
GRP can be
overstressed if joint
pulled up
Measurement to
verify
documented
dimensions
In accordance with
6.8.5 of ISO
14692-2:2002
Weepage or pipe
failure
Visual inspection,
with light source
inside pipe
Transient thermography,
ultrasonics, microwave,
shearography, or acoustoultrasonics to detect
delamination damage
associated with impact.
In accordance with
Table A.1
Incorrect transport.
Incorrect handling
Outside temperature
and humidity
specifications.
Weakened joint or
leakage
In accordance
with 8.3.3 of ISO
14692-2:2002
Improper mixing.
Heating pad overlap
or controller
problems.
Differential Scanning
Calorimetry (DSC) to determine
Tg for adhesive and extent of
cure. See NORSOK M622
Table 3. DSC will require a
small off cut of material to be
removed and heated.
In accordance with
8.3.3 of ISO
14692-2:2002
Alignment to
project
specifications
Unlikely in a correctly
commissioned system. Should be
ironed out in system qualification.
Movement during
curing.
Bending
Residual stress,
resulting in less
than rated
performance
Incorrect dimensions
Visual inspection.
Ultrasonics
28
Cause(s)
Consequence(s)
Recommended
NDT method(s)
ISO 14692
Criteria
Corrective action
Comments
Defects in adhesive
bond
Weakened joint or
leakage
Ultrasonics, or
radiography
Microwave inspection,
shearography, acoustoultrasonics or tap testing
Debond area
greater than 30 %
of total bond area.
Remake joint
Movement during
curing
Axial length of
debond area
greater than 20 %
of total axial bond
length.
Improper treatment
of joint surfaces
Contaminated
surface after grinding
Weakened joint or
leakage
Visual inspection
Microwave inspection,
shearography, acoustoultrasonics or tap testing
In accordance with
adhesive supplier's
requirements
Remake joint
Excess adhesive
Restriction in pipe
to flow
Radiography
No flow
obstruction. 5 % of
inner diameter or
10 mm, whichever
is less
Visual inspection
In accordance with
Table A.1
Increased risk of
erosion damage of
pipe
Damaged threads
Teeth chipped
Damaged end faces
Joint cannot be
sealed, leakage
29
Table 2 NDE selection and acceptance criteria for defects that could potentially occur during operation
(Update of Table 5 from ISO 14692 Pt. 4, additions shaded).
Operational defects
Cause(s)
Consequence(s)
Recommended NDE
method(s) ISO 14692
Criteria
Corrective action
Comments
Visual inspection
Ultrasonics.
No leakage permitted
Radiography unlikely to be
successful unless well
aligned.
System failure
Visual inspection
Monitoring by acoustic
emission or leak detection
methods
No failure permitted
Weepage
Ultrasonics
Ultrasonics, microwave,
shearography, acoustoultrasonics or transient
thermography to detect
delamination damage
associated with ageing or
impact.
More than 20 %
reduction in original
axial modulus
Radiography
Design conditions,
loads, temperatures
exceeded.
Operational procedures
inadequate (e.g. water
hammer due to valve
opening).
Ageing
Long-term materials
degradation
30
Cause(s)
Consequence(s)
Recommended NDE
method(s) ISO 14692
Ageing
(Continued)
Impact damage
Criteria
Corrective action
Weepage
Ultrasonics.
inspection
Visual
Transient thermography,
ultrasonics, microwave,
shearography or acousto
ultrasonics to detect
delamination damage
associated with impact.
Comments
In accordance with
Table A.1 Visual
inspection
Earthing cable
damage
Some cables
susceptible to corrosion
in marine atmosphere
Earthing reduced or
eliminated
Visual inspection
Operating conditions
resulting in e.g. barium
sulphate deposits
Reduced flowrate
None
None permitted
Replace cables
Radiography.
Reduction in internal
diameter of more than
10 mm or 5%
Mega-ohm meter
Ultrasonics
Shearography potentially
good but expensive.
NDE unlikely to be
applicable
Applicability of NDE
methods will depend on the
nature of the scale
deposition and the
interface. Linear scan (BScan) preferable to point
measurements for
ultrasonics.
31
Cause(s)
Consequence(s)
Recommended NDE
method(s) ISO 14692
Criteria
Corrective action
Comments
Erosion
Particulates in flow
Reduction in wall
thickness leading to
pipe weepage
Ultrasonics
Ultrasonic B-Scan.
Reduction in original
wall thickness of less
than 20 %.
Applicability of ultrasonics
will depend on thickness
and surface quality. Linear
scanning methods (B-Scan,
TOF) preferred to point
measurement for erosion
damage.
Reduction in original
wall thickness of more
than 20%.
Reject
Depth limited to
surface resin layer
Accept
NDE unlikely to be
necessary
Chalking
Exposure to UV
radiation
Minor breakdown of
outer surface
Visual inspection
None
32
11.1
Manufacturing inspection
Manufacturing NDE is outside the scope of this document except insofar as manufacturing
defects persist into the final system or component. Defects can occur in either the GRP
material or in the mechanical and/or adhesive-bonded joints that make up the piping
system. Joint defects, including defects in prefabricated pipe spools, are more likely to
occur than defects in the GRP material, provided QA procedures are followed during
manufacture, handling and delivery.
Manufacturing processes for fittings are typically more complicated and less automated
than those used to produce pipes. Therefore, manufacturing related defects tend to be
more prevalent in fittings.
NDE methods recommended for use in detecting those defects most likely to occur during
specific stages in the manufacture and operation of GRP piping, together with acceptance
criteria and recommended corrective actions are listed in Table A.1 of ISO 14692-4:2002.
(reproduced in Appendix C).
Similar NDE methods may be applicable in-service, subject to the more limited access
restrictions. The Operator shall be notified of all repairs. On agreement between the
Operator and the manufacturer, a mill hydrostatic test on all minor repaired items shall be
performed.
11.2
After installation
Inspection requirements after installation are covered in Sections 5.4 to 5.9 of ISO 14692
Part 4. It is recommended that the Operator carry out hydrotesting of representative sitefabricated joints and fittings soon after the start of installation to verify the standard of
workmanship. This is particularly important for pipe diameters above about 200 mm
(ISO14692 Part 4, Section 5.5.8). This is due to correct on-site installation, workmanship,
alignment and achieving required dimensional tolerances being more difficult for larger
diameter pipes.
11.3
In-service
Inspection requirements in-service are covered in Section 6 of ISO 14692 Part 4 and
guidance on NDE methods provided in Annex E. Recommendations on NDE methods for
the defects that may occur during operation and relevant inspection methods are given in
Table 5 of ISO 14692 Part 4 (see Appendix C). Since ISO 14692 was released in 2002 a
number of new NDE methods have become available. These have been identified in Table
1 and Table 2 above and are identified by component and defect type in Sections 12 and 13
below. It is recommended that these updated recommendations regarding NDE techniques
are followed.
12.0
This Section provides guidance on the selection of NDE methods for specific GRP piping
system components. In each case the recommendations in ISO 14692 have been updated
to take account of newer alternative NDE methods that have become available since the
last revision of ISO 14692 in 2002. The guidance from ISO 14692, encompassed in the
HSE Good Practice Guide is summarised below in Table 3 together with the potential
alternative NDE methods identified earlier in this document in Table 2.
33
Measurement to verify
documented
dimensions.
Visual inspection
Incorrect curing of
adhesive/laminated joint
Misaligned joints
Defects in adhesive bond
Visual inspection/ultrasonics
Ultrasonics/Radiography
Visual inspection
Excess adhesive
Radiography
Damaged Threads
Visual inspection
Ultrasonics, microwave,
shearography, acoustoultrasonics or transient
thermography to detect
delamination damage
associated with impact.
Acoustic emission to monitor.
Differential Scanning
Calorimetry (DSC) to
determine Tg for adhesive
and extent of cure. See
NORSOK M622 Table 3.
Radiography for intact joints
Microwave inspection,
shearography, acoustoultrasonics or tap testing
Microwave inspection,
shearography, acoustoultrasonics or tap testing to
detect disbonding
Internal visual inspection with
endoscope, Radiography.
Radiography, Leak testing
34
Visual Inspection
Visual Inspection
Visual Inspection
Ageing
Ultrasonics
Impact Damage
Visual Inspection/Ultrasonics
Visual Inspection
Visual
Inspection/Megohmmerter
Visual (reduced flow),
Radiography
Erosion
Ultrasonics
Chalking
Visual Inspection
Ultrasonics.
Radiography
Monitoring by acoustic
emission or leak detection
methods. Shearography,
ultrasonic angled shear wave
inspection, TOFD,
radiography or possibly
microwaves for cracks.
Monitoring by acoustic
emission or leak detection
methods
Destructive testing and
characterisation of condition
or NDE using Ultrasonic
velocity to measure matrix
cracking non-destructively.
Ultrasonics, microwave,
shearography, acoustoultrasonics or transient
thermography to detect
delamination damage
associated with ageing.
Transient thermography,
ultrasonics, microwave,
shearography or acoustoultrasonics to detect
delamination damage
associated with impact.
Acoustic emission to monitor.
As impact damage
None
Radiography.
Ultrasonics
Ultrasonic B-Scan.
Ultrasonic Time of flight
(TOFD).
Radiography.
Acousto-ultrasonics.
None
35
12.1
Pipework
ISO 14692 recommends the following inspection methods for GRP pipework in-service as
detailed in Table 3 above:
Visual inspection.
Ultrasonics.
Radiography.
The applicability of ultrasonics and other NDE methods will depend on the thickness, quality
and surface condition of the GRP pipework, defect type, location and the purpose of the
inspection (defect detection or assessment of ageing).
Inspection trials, conducted under the auspices of the HOIS FPSO and Flexible Risers
Working Group, confirmed the above recommendations. Data obtained using manual
ultrasonics were somewhat variable but were generally able to detect the pipe back wall
echo but not necessarily the deliberately introduced bondline defects. Automated
ultrasonics using either twin probe TOFD or phased arrays produced reasonable B-scan
images showing the layered nature of the material and in some cases reflections from the
introduced defects. Interpretation of reflected signal waveforms was more complex than
would be expected for steel components as the GRP structure is elastically anisotropic and
heterogeneous with signal attenuation due to voids/porosity and scattering nature of
laminated structure. It was sometimes difficult to get a consistent back wall echo with some
variability in response between joints. Inspection operators need to familiarise themselves
with particular glass reinforced epoxy components and choose optimum probe types.
Ultrasonic B-scan images are the preferred acquisition mode as they can give a clearer
delineation of bond line defects and back wall echo. Generally the lower frequency probes
( MHz - 2 MHz) gave better sample penetration than 5 MHz probes but with reduced
resolution. Phased array wheel probes were relatively quick to scan the pipe surfaces but
had some positional difficulties around elbows and fittings due to roller sliding.
Of the various non contact inspection methods trialled, i.e. laser shearography, microwave
inspection and radiography, the latter technique produced the best images using the
portable pulsed X-ray source used in conjunction with digital detector arrays. In this case
some details of the joints were revealed including bondline defects such as porosity
(however there was no evidence of the paper inserts) and good images of the pipe wall and
fittings.
After the inspection trials were completed the flow loop was cut open and macrophotographs were taken of joint sections to reveal the location and extent of the defects. In
many cases the joint defects were lack of adhesive and porosity. The quality of the elbow
end fittings was also assessed and showed evidence of poorly wetted out fibres which was
detected during several ultrasonic trials. A section of the centre bell and spigot joint was
prised open, as recommended by a consultant from the pipe manufacturer, to reveal the
quality of the bond.
In summary, each of the inspection trials had practical challenges and most were able to
detect some of the defects. Ultrasonic inspection was able to detect loss of back wall echo,
bondline features and end fitting irregularities. Radiographic examination using digital
detectors together with image enhancement filters was able to reveal pipe wall and end
fitting details, bondline porosity and quality of adhesive fillet but not gaps in the adhesive
bond.
36
Care shall be taken in the use of conventional methods for removal of scale and
other blockages (e.g. high-pressure water lances, mechanical and chemical
cleaning methods). The manufacturers recommendations shall be followed in all
cases.
Earthed GRP piping systems shall be periodically checked to ensure that all
earthing leads are functional and that the requirements for continuity of electrical
path and the resistance to earth are not greater than the specification given in the
operator's documentation. The testing of resistance to earth shall be in accordance
with 5.5.4.4 of ISO 14692 Part 4.
Taking account of the above information and our experience in inspection of GRP, it is
recommended that the advice in ISO 14692 and NORSOK M622 and the HSE good
practice guide 4,5 is followed for GRP pipework after installation and in service, but subject to
a number of additional considerations:
For damage mitigation pipes should not used as a step up, not be exposed to
deleterious foreign fluids and not be exposed to heat sources. Earthing cables (if
required) shall be present and in good condition.
For commercial grade GRP pipework visual inspection for impact damage and
possible signs of delaminations is recommended as the primary inspection method,
supplemented by periodic NDE inspection. The guidance on when to apply NDE
methods in Section 10.9 and ISO 14692 Pt. 4 should be followed.
The general appearance of pipework should be uniform with smooth surfaces and
none of the following, or within prescriptive limits, given in associated standard/
manufacturers guidance: cracks (including adhesive fillets between repair laminate
and substrate), delamination, wear, chips/gouges, resin loss, exposed fibres/cut
edges, leaks/weepage, dry spots, variation in resin colour, contamination/foreign
matter, heat damage, softening/ bulging/ discolouration/ burn, blisters/ pin holes/
pores, chemical attack, erosion .
The use of suitable alternative NDE methods to ultrasonics identified in Table 2 and
Table 3 such as microwave inspection, tap-testing, radiography or shearography is
acceptable provided these are qualified on the component or suitable reference
samples.
37
For more critical systems monitoring using acoustic emission or graded optical fibres
should be considered.
Guidance on NDE and destructive testing methods for the assessment of ageing is given in
Section 13.6. Ultrasonic velocity measurement to determine the matrix cracking density,
following methods developed by ESR Technology9, offers an alternative to destructive
testing.
12.2
Fittings
Manufacturing processes for fittings are typically more complicated and less automated
than those used to produce pipes. Therefore, manufacturing related defects tend to be
more prevalent in fittings. It is possible that such defects could extend in service. Providing
the fittings have been adequately designed and qualified then the main concern is likely to
be matrix cracking leading to weepage later in the life of the component.
Fittings pose specific difficulties for inspection due to the geometry, thicker section and
quality of the GRP arising from the methods of construction. Filament wound fittings will
generally be easer to inspect than fittings that are manually overlaid. Fittings are usually
tapered to accommodate the higher loading in these regions, which makes inspection more
difficult than for parallel surface pipework. The interface should be considered for fittings
that have been manually overlaid or laminated to the pipework.
It is recommended that fittings are subject to periodic visual inspection to check for impact
damage or signs of ageing. If NDE is required then the recommendations given above for
thicker section or hand laminated quality GRP pipework should be considered. Ultrasonics
will be difficult and require low frequencies (0.5 to 1 MHz). For more complex or less
accessible geometries, radiography may be the only feasible inspection option.
NDE methods applicable to flanged connections are discussed below (Section 12.7).
12.3
Supports
GRP piping systems may be supported using the same principles as those for metallic
piping systems. However, due to the proprietary nature of piping systems, standard-size
supports will not necessarily match the pipe outside diameters. The use of saddles and
elastomer pads may allow the use of standard-size supports. Guidance on support methods
is provided in Section 5.5.3 of ISO 14692 Part 4.
It is recommended that the areas around pipe supports are subject to more detailed visual
inspection as such areas can experience higher or more variable loading than the normal
pipework. Enhanced visual inspection using dye-penetrant or application of NDE methods
should be considered if defects have been encountered in service or inadequacies are
suspected in support design.
It should be established from visual inspection after installation that the guidance on pipe
supports in ISO 14692 has been followed:
38
Supports in all cases should have sufficient width to support the piping without
causing damage and should be lined with an elastomer or other suitable soft
material.
Clamping forces, if applied, should be such that crushing of the pipe does not occur.
Local crushing can result from a poor fit and all-round crushing can result from overtightening.
Supports should preferably be located on plain pipe sections rather than at fittings or
joints.
Supports shall be spaced to avoid sag (excessive displacement over time) and/or
excessive vibration for the design life of the piping system. Guidance on support
spacing is given in ISO 14692 and in the AWWA M45 Manual Chapter 8.
GRP pipe shall not be used to support other piping, unless agreed with the principal.
GRP piping should be adequately supported to ensure that the attachment of hoses
at locations such as utility or loading stations does not result in the pipe being pulled
in a manner that could overstress the material.
The anchor support shall be capable of transferring the required axial loads to the
supporting structure without causing overstress of the GRP pipe material.
Anchor clamps are recommended to be placed between two double 180 saddles,
adhesive-bonded to the outer surface of the pipe. The manufacturers standard saddles are
recommended and shall be bonded using standard procedures.
Urgent action is required if pipes are found to be inadequately supported; or pipe supports
missing, extensively damaged or corroded. Such damage is considered to be unsafe (HSE
priority level P1 urgent). Risk assessment shall be performed as a matter of urgency and
suitable mitigation strategies implemented.
Minor damage or corrosion that leaves the supports still structurally sound is acceptable but
requires caution (HSE priority level P2). In this case, a risk assessment shall be performed
as soon as reasonably practicable to identify a suitable course of action.
12.4
Joints
The joints are one of the main areas of concern for composite pipework, but generally
function reliably in-service if they are properly designed, qualified and installed. The most
likely form of failure in GRP piping systems is a leaking joint.
Visual inspection and pressure testing after installation are the primary methods of
ensuring joint integrity. Barcol Hardness may be used to establish adequate curing of
adhesive joints. Guidance on testing and inspection of joints can be found Section 5.5.6 and
Annex C of ISO 14692 Pt. 4 including the percentage of joints to be tested.
39
adhesive-bonded joints;
laminated joints;
flanged joints;
threaded joints.
Potential defect types in joints include: de-bonds (lack of adhesive), fibre breakage,
damaged or cracked threads, cracked flanges, environmental access and degradation.
Threaded connections between composite materials are not often used offshore. Metal end
connectors may be threaded.
A detailed assessment of integrity issues including NDE methods for joints has been made
in the ACLAIM programme (www.ndt.esrtechnology.com/aclaim) and followed by the UK
TSB funded Imajine programme (www.imajine.co.uk). In ACLAIM, a range of NDE methods
were assessed including tap-testing, ultrasonics, microwaves and shearography.
The ACLAIM project (Advanced Composite Life Assurance and Integrity Management), is
one of several that DTI has funded through its Technology Programme. This was aimed at
developing an integrated structural health management framework to assess the integrity of
Composite Structures. The primary objective was to provide an integrated approach that
includes detection techniques (embedded), assessment procedures and guidance
documentation to:
Specific guidance for adhesive and laminated joints is given in the following Sections. It is
recommended to visit the ACLAIM and Imajine websites if a more comprehensive
consideration of joint issues is required, including assessment of NDE methods.
12.5
Access for inspection and NDT of adhesively bonded connections is not straightforward.
Therefore, careful preparation and procedure during the adhesive bonding process and
component alignment are crucial to obtain a joint of good integrity. There is currently no
good NDE solution, particularly in the case of end connectors.
Relevant defect types include disbonding, lack of adhesive, kissing bonds, and voids and
other defects within the adhesive layer. ISO 14692 specifies rejection if the area of debond
is more than 30% of the total bond area bonds; or the axial length of any debond area is
greater than 20 % of total axial bond length.
40
Proper calibration standards should be produced using the same adhesive, jointing
method and GRP material and geometry as the pipework to be inspected.
It is recommended that at the design stage a good quality surface finish (Peel-ply or
gel coat) is procured on the joint surfaces to facilitate ultrasonic inspection.
For ultrasonic inspection, common scanning methods such as linear scanning (Bscan) or area scanning (C-scanning) are preferred over point measurements.
41
The alternative NDE methods assessed by ESR Technology as having the greatest
promise in this application are microwaves, shearography and acousto-ultrasonics.
Trials using simple tap-testing in the ACLAIM programme encountered difficulties. It
is not believed transient thermography would be particularly effective because of the
curvature and thickness. Procedures for application of these new techniques on
GRP pipe-joints have yet to be established. These latter methods need to be
investigated further to establish their limitations and potential applicability.
Tap and bond-testing methods are limited in applicability to offshore joints by the
thickness and quality of the GRP the tap and acoustic response needs to transmit
through to access the bond.
The possibility of installing graded optical fibres within the adhesive layer could be
considered to monitor the degradation and any defect growth in service. Such
technology is currently available and is being assessed in the Imajine programme.
For critical components, the use of acoustic emission should also be considered.
Detection of kissing bonds is particularly difficult and there is no good NDE solution.
Several methods have a potential capability, but the difficulty of this defect type is
well recognised by NDE providers. Laser shearography has the potential to detect
lack of adhesion and the strength or weakness of the bond, an advantage over
ultrasonics in this application and relevant to kissing bonds. However, the section
thickness in commercial joints is likely to limit the sensitivity of resolution for
shearography. Kissing bonds are a difficult defect for thermography. Ultrasonic
guided and surface waves have been evaluated for adhesive disbonds in
connectors with limited success. Shockwave methods (such as Remote acoustic
impact Doppler (RAID) developed in the aerospace industry for kissing bonds 6 may
have applicability in connectors but have so far not been applied to offshore grade
GRP or thicker section composites.
12.6
Laminated joints
For laminated joints the same NDT methods could be considered as given above for
adhesive joints, although as there is no adhesive radiography is unlikely to be so applicable.
The recommendations for delaminations In ISO 14692 Pt. 4 could be considered relevant.
The main issue is delamination or disbonding of the joint, and any degradation in joint
integrity that may occur in service. It would be normal to visually assess and proof test the
joint following installation. The following recommendations are given in the standards:
New Technologies to Improve the Quality, Speed and Safety of Production; Production; Dr. Peter H. Wu, Vice
President & Chief Scientist, Spirit Aerosystems. Manufacturers Summit 2006, Low Mass Transport Systems
Symposium 2006, Manchester 2 November 2006
42
Microwaves and shearography were also found to be the most promising methods for
disbonding in NDE tests undertaken by the offshore composite repair workgroup.
Ultrasonics was variable with some difficulty in achieving back-wall echoes. Tap-testing
performed less well than had been suspected. Taking these recommendations into account,
we would recommend visual inspection of the joints with occasional leak testing as the main
approach for assessing integrity of laminated joints. Acoustic emission monitoring could
also be considered. If NDE is required or considered to look for incorrect lamination or
degradation in-service of the joint, then the following methods that are sensitive to
delaminations or disbonds are recommended:
Note that the latter two methods are still at the development stage for offshore application.
The following alternative NDE methods are potential alternatives if first qualified on similar
joints:
Acousto-ultrasonics;
Tap testing;
Radiography.
HSE/HSL FRP composite repair fact sheet; Draft: Final, March 2009
43
12.7
Flange connections may be fixed or loose flange designs, with composite to metal or
composite to composite connections. Manufacture is by filament over-winding or manual
lay-up with moulding. Relevant defect types include flange cracking or delamination of the
flange itself or adjacent laminated material. Over-tightening, poor design or residual
stresses introduced during curing are the main causes of flange cracking.
Recommended NDT methods for flange cracking in ISO 14692 Pt. 4 include:
Visual inspection;
Thermography;
Manual UT;
Radiography.
ISO 14692 and NORSOK M-622 recommend visual inspection for flange cracking with tap
testing for general assessment of the GRP material in the taper and around the joint. For
thin components, thermography or ultrasonics are recommended; with X-radiography for
thicker components.
It is recommended that the ISO 14692 and NORSOK M622 guidance on NDE techniques is
followed in the first instance. In all cases the NDE method should be qualified on a suitable
reference sample or on the component.
If the above methods prove ineffective then the following are potential alternative methods:
Dye-Penetrant
Microwave inspection;
Laser shearography;
Acousto-ultrasonics.
These methods do not have an established record in this application and should be first
qualified on a suitable reference sample or representative component. There are a number
of NDE issues and limitations. The surface finish of the flange surround is generally less
good than pipework, particularly if laminated. Wall thickness is variable around the taper
and access may be limited.
44
12.8
Repairs
13.0
13.1
Inspection for delaminations is one of the most common uses of NDE in composites.
Delaminations can directly affect the integrity of a composite component and of adhesive
bonds. There are a variety of NDT methods that potentially can detect delaminations. The
most appropriate method will depend on thickness, surface finish and ease of coupling,
material quality, geometry of the component and access limitations.
Installation and in-service inspections are covered in ISO 14692 Part 4 as well as a
summary of defect types and NDE methods. NDE methods identified for delaminations are
visual inspection with internal illumination if appropriate, ultrasonics and radiography.
Several non contact inspection methods are available including laser shearography, pulsed
video thermography, microwave inspection as well as conventional, computed or digital
45
46
Rapid ultrasonic inspection (B-scan, C-scan) using wheel probe or phased array;
Tap testing;
Laser shearography;
For thicker sections > 25mm, more complex geometries or where surface finish or porosity
makes ultrasonic inspection difficult the following methods should be considered and can
also be applied to thin sections:
Microwaves;
Radiography (using appropriate radioactive source);
Acousto-ultrasonics (e.g. T-Scout or similar system) ;
Vibro-thermography (developmental -effective for delaminations and cracks).
13.2
Loss of wall thickness can occur by erosion or chemical attack. In lined piping there may
also be a need to assess the integrity of internal liners, since loss of the liner protection may
lead to a rapid loss of properties. Criteria and corrective action for loss of wall thickness
(LOWT) are defined in Table 5 of ISO 14692 Part 4. For erosion a reduction in original wall
thickness less than 20% is acceptable with monitoring required; reductions greater than
20% are to be rejected.
Chemical attack (liner or resin rich layer removal) is covered in Table A.1 of the above
standard. If the resin surface is absence rejection is recommended at manufacture, delivery
or installation. In operation the corrective action recommended is rejection/ major repair.
There is no acceptable wall loss if the protective resin layer or liner has been removed. Any
chemically damaged layer should be treated as having no residual strength and replaced.
Monitoring of wall thickness, liner condition and the depth of any chemical damage may
provide a short term solution based on engineering judgement, if it does not exceed
corrosion allowables, but is not recommended by ISO-14692.
ISO 14692 specifies ultrasonic thickness measurement with an averaging 1-D assessment
and a conservative minimum wall thickness approach.
The NORSOK guidelines recommend visual inspection (internal), ultrasonics, thermography
or radiography. Procedures are included in Annexes A, C, D and E. of the NORSOK
guidelines.
47
Acousto-ultrasonics (T-Scout);
Laser shearography.
The choice of technique will be determined by the same factors outlined above for
delaminations, namely: surface finish; thickness; material quality; and the complexity of the
component. NDT is difficult where material has been manually overlaid; with coupling
difficulties and significant attenuation for manual UT even at low frequency (< 2MHz).
Recent trials by HOIS in Houston suggest that tap and bond-testing methods also can
become difficult when applied to thicker section or hand laminated quality material
compared to commercial filament wound piping.
Procedures are available for ultrasonic A-scan wall thickness measurement, ultrasonic Bscan and C-Scan, shearography and thermography from the UK DTI funded CPD4D
programme8 and can be accessed from the MMS15 IKB website www.mms15.com .
13.3
Impact damage
Impact damage can occur to composite components during manufacture, transport or inservice. This is usually characterised by a conical region of damage, which may include
delaminations, matrix cracking and fibre cracking, emanating below the surface of the
component from the point of impact. Where the damage is only faintly visible to the naked
eye this is known as barely visible impact damage (BVID). The extent of damage will
depend on the force of impact.
ISO 14692 refers to use of visual inspection to detect impact damage. A maximum
diameter of 5mm is given as acceptable for impact damage. No specific NDT techniques
are recommended for impact damage.
The NORSOK guidelines do not include impact damage specifically in the summary table
on NDE methods. Methods that are given in the NORSOK and ISO tables for the
CPD4D Procedures for ultrasonics, thermography and shearography of FRP composites, ESR Technology,
DTI CPD4D Programme Final Report, 2002 (available at www.mms15.com )
48
Visual inspection;
Thermography;
Shearography;
Ultrasonic A-scan;
Trials at the NNDTC have shown thermography to be particularly suitable for impact
damage; picking up the conical region of damage, particularly if modern image processing
methods are used to give images at different depths. Simple ultrasonic pulse-echo has
potential if delaminations have occurred. More sophisticated ultrasonic imaging methods
(C-scan, B-scan, 3D dataset) can show up regions of impact damage visually. This is likely
to be too detailed for offshore application. Shearography has potential as a fast global
method and should give good sensitivity to surface strains and sub-surface damage arising
from the impact damage. Further trials are needed on offshore pipework to establish the
best methods for applying the load change required (vacuum, thermal or internal pressure)
and the thickness limitations.
Additional methods that could have potential include:
Microwaves;
Acousto-ultrasonics.
We are not aware that either has been specifically trialled on impact damage.
13.4
Matrix cracking
Frost S R and Lee R J; Integrity assessment procedures for GRP pipes, Technical Exchange Meeting, Saudi
Aramco, Dhahran, 27th May 2008
49
Acoustic emission;
Gamma-radiography.
If the GRP component is left unpainted and thin, i.e. transparent, then visual inspection with
internal illumination is a simple and effective approach. Digital X-radiography is the
accepted method in the aerospace industry but has not been applied in-field in an oil and
gas environment. Angled ultrasonic and C-scan methods have been developed in the
aerospace industry specifically for detection of matrix cracking, but again they have not
been applied to GRP components used in oil and gas applications.
Velocity measurements using ultrasonic waves for monitoring the change in axial modulus
due to matrix cracking has been trialled in the laboratory and shows potential, linking the
reduction in velocity to the increase in micro-crack density. This technique has not been
applied in the field. The axial ultrasonic velocity method is used to determine the modulus
and relate this to crack spacing. The number or density of cracks within each ply increases
as the level of load or over-stressing is increased. Modelling of the process of matrix
cracking within the wall of GRP pipes9 showed that:
Crack spacing is the parameter which can be used to characterise the amount of
damage;
The parameter which is the strongest function of crack spacing is the axial modulus;
The regression gradient relating to the long term performance of the pipe can be
estimated from short term and long term stress, strain measurements.
The most sensitive parameter to infer crack spacing is pipe axial modulus. As the crack
spacing increases the axial modulus will decrease. Hence from velocity measurements, the
remaining life of the pipe can be estimated from the inferred regression gradient and the
current level of damage (or crack spacing) within the pipe wall. The assumption here is that
the axial modulus will degrade at the same rate as the tensile properties required for
pressure containment2.
In the previous DTI DCC3 10 programme on composite materials an assessment approach
was developed and a number of NDE methods were evaluated for assessing the crack
10
Final Report from DTI/DCC3 Programme on design guidelines for Brittle-Matrix composite components, Subprogramme: Evaluation of NDT methods for transverse cracking, Eckold G C, Wall M et al, April 1997.
50
13.5
Significant cracks
This refers to the situation where cracking has extended over a number of plies through the
section and in many cases through to the component surface. This will normally include
both matrix cracking and fibre breakage and frequently link to delaminations.
The same NDE methods are applicable as for matrix cracking. Significant cracks are likely
to be revealed by visual inspection enhanced by penetrant or internal illumination.
ISO 14692 makes no specific recommendations on detection of discrete cracks but
recommends acoustic emission for general assessment, ultrasonic velocity measurement
for quantification, with a damage mechanics approach to estimate density and
consequences. The NORSOK M622 guidelines recommend visual inspection (e.g. dye
penetrant), acoustic emission (propagating cracks), ultrasonics or radiography. Procedures
are included in Annexes A, E, C and D respectively of the guidelines.
Through-thickness cracking or fatigue cracking is unlikely to be detected using conventional
0 ultrasonic probes, though associated delaminations may be detectable.
Methods focussed on detection of delaminations such as ultrasonic C-scanning or
thermography, and attenuation methods like radiography, may not reveal tight throughthickness cracks. However, the nature of crack propagation in composites is that cracks are
likely to be a mix of matrix cracking, fibre breakage and delaminations and a crack of this
type is potentially detectable by a range of NDT methods. It is recommended that the NDE
methods included in NORSOK M-622 are used in the first instance for crack detection in
GRP pipework and joints:
Ultrasonic velocity methods can be used to indirectly assess cracking based on the effect
on measured modulus.
In most cases visual inspection is likely to be adequate.
If the above NDE methods prove inadequate in the application then one of the following
techniques shall be used, subject to qualification on the component or a suitable reference
sample.
Laser shearography;
Ultrasonic pulse-echo (B-scan, angle probe);
Thermography;
Ultrasonic TOFD;
Ultrasonic phased array (conventional or roller probes);
Vibro-thermography (developmental);
Acousto-ultrasonics.
51
13.6
Materials degradation
Materials degradation is handled in ISO 14692 by use of regression curves, after testing at
the design and qualification stage. The normal approach recommended to assess the
actual condition of aged composites, in relation to life extension, is destructive testing as
detailed in ISO 14692 Pt. 4.
For destructive testing a two stage combination of measurements is recommended. The
first stage involves a simple assessment as to whether the component has suffered
degradation or not. This stage involves the following tests:
These tests can be completed on the GRE component relatively quickly; which implies that
an answer to the question has the component suffered any degradation? can be
answered in a short timeframe.
If the first stage assessment demonstrates that the component has suffered degradation
then the second stage involves further testing to establish the amount of or quantification of
degradation. This stage might typically involve the following tests:
From the results of these further tests and application of the damage mechanics model an
assessment of the remaining lifetime or remaining strength of the GRE component can be
made following the guidance in ISO 14692.The one drawback of this assessment procedure
is that GRE samples are required from the field to enable the destructive tests to be
performed.
Ultrasonic velocity measurement of modulus in combination with laminate9 potentially offers
an NDE alternative to assess the level of matrix cracking (the dominant ageing mechanism
in oil industry applications) and is recommended where removal of components for
destructive testing is difficult or not preferred.
In oil and gas industry applications the main degradation mechanism is matrix cracking and
the methods described above for matrix cracking would be applicable. In the latter stages
these lead to delaminations or larger transverse cracks arising from linkage of the matrix
cracks, which are precursors to weepage. Such severe ageing damage may be detectable
visually or with the NDE methods identified above for delaminations and cracks (See
Section 13.7 below specifically for weepage).
11
52
13.7
Environment ingress is not significant in oil industry composites except where damage has
progressed sufficiently to the surfaces to permit ingress or allow weepage through the
composites. Water or environment ingress will soften and weaken the composite.
Weepage is a more severe combination of matrix cracking and delaminations where the
cracking has extended to the surface and allows liquid to seep through the composite. This
is effectively the failure mechanism for GRP pipe as the pipe is no longer fit for purpose. If
damage has progressed this far it is likely to be evident on visual inspection and use of
NDE methods is unlikely to be applicable.
Weepage is a failure condition and it would be anticipated that damage should be detected
at an earlier stage by NDE methods that can detect delaminations. In most Oil industry
applications environment damage has been found to be cosmetic and evident on visual
inspection. The precursor damage is likely to be revealed by the NDT methods identified
above for delamination and matrix cracking.
14.0
NDE Methods
This section provides guidance on the application of NDE methods available for inspection
of GRP. Visual inspection remains the most important technique if the resin is transparent,
as large areas can be inspected quickly within the laminate. However, some of the most
important defects, such as poor bonding, may require more sophisticated NDE methods
and a high degree of operator skill. Specific guidance on the application procedures for
primary NDE methods is given in Annex E of ISO 14692 Pt. 4 with more detailed
procedures for ultrasonics and radiography in NORSOK M622. More detail on individual
NDE methods and data examples can be found in the HOIS review on NDE of
composites 12. Relevant NDE codes and standards are given in Section 4.2.
14.1
To validate an inspection method and to set the sensitivity for the inspection, appropriate
calibration samples should be used. An example of a calibration sample is shown in Figure
8. It is recommended these should be of similar thickness, material type, ply structure,
surface finish and geometry as the actual component and contain real or simulated defects.
The method used for insertion of defects should be representative of the defect and suitable
for the NDE method (or methods) being evaluated.
12
53
To simulate delaminations a double PTFE layer (50m) with the edges sealed by
heat resistant tape to leave an air gap;
For adhesive joints areas free of adhesive or use of double PTFE layers to simulate
disbonding;
Flat bottomed holes milled on the far surface to simulate erosion damage or voids;
Methods now exist which involve the insertion of individual air pockets (similar
principle to bubble wrap). These provide a better simulation of a delamination than
a PTFE film defect.
A difficulty with simulated defects is that some NDE methods may be good with simulated
PTFE flaws, but not with artificial flaws or vice versa. Ideally a range of defect sizes and
depths should be included to assess the technique sensitivity.
Examples of calibration samples are shown in the ultrasonic and radiographic procedures
included in the Annexes to NORSOK M-622, including a stepped adhesive bond sample
relevant to different thicknesses of composite joint (Figure 9).
54
Figure 9 Examples of calibration samples for ultrasonics in NORSOK M-622: left, Sketch
of calibration standard for voids and delaminations, defects achieved by machining holes
with diameter and depth selected to match defects to be detected; right, Sketch of
calibration standard for bonded pipe joints.
14.2
Visual inspection
Visual inspection is the primary inspection method used on composites at manufacture and
in-service. The advantage is the speed and simplicity and the range of surface visible
defects that can be detected. If the composite is uncoated then due to the transparency of
the material internal defects may also be visible on back illumination.
It is recommended that visual inspection is carried out in accordance with the guidance in
Section 5.7 of ISO 14692 Pt. 4, with defects classified in accordance with ASTM D 2563 13.
The main visually detectable defects are:
It is important that inspectors are familiar with inspection of GRP and the defect types that
may occur from manufacture and in-service. Suitable aids such as an illumination source,
mirrors, boroscope, and if appropriate penetrant, to maximize the extent and accuracy of
the visual inspection.
Possible defects along with acceptance criteria and corrective actions are listed in Table 4
of ISO 14692 Part 4 with further information concerning defects arising during fabrication
and installation, and corrective action, given in Table A.1. The presence of a uniform fillet of
adhesive is often an indicator that an adhesive joint has been assembled correctly.
Visual inspection, both internally (as far as physical access allows) and externally, should
be carried out of all joints and all surfaces.
13
ASTM D 2563. Standard Practice for Classifying Visual Defects in Glass-Reinforced Plastic Laminate Parts
55
14.3
Pressure testing
The hydrotest is considered by many to provide the best assurance of a GRP pipe systems
or vessels integrity, and to be more reliable than NDE techniques for assessing that the
system has been properly fabricated and installed. It is recommended that the guidance in
ISO 14692 Part 4 and NORSOK M-622 Section 5.5.4.6 is followed. It is normal to follow the
pressure test (1.5x maximum operating pressure) with a separate leak test at 1.1 times
maximum operating pressure. The hydrotest may be done in conjunction with Acoustic
Emission monitoring to detect early onset of damage and identification of damage locations.
Pressure testing in service may also be considered, particularly if defects are suspected in
adhesive joints or after long term ageing. A disadvantage is that the cost of blinding off
systems can be significant.
The main defects detectable by pressure testing are:
a) adhesive-bonded joints lacking adhesive or improperly prepared and assembled,;
inadequately cured adhesive in the bonded joints;
b) manufacturing defects in GRP materials;
c) leaking joints.
A pressure test at 1.5 times the design pressure reveals leaks and such major defects as
severe impact damage (e.g. from improper transport), improperly designed or fabricated
systems (lacking adequate strength or pressure retention), or very poor adhesive bonding.
However, adhesive-bonded joints are designed with a large margin of safety and bonded
joints having as much as 80 % unbonded area can pass a pressure test. Thus the pressure
test is a major element in ensuring that the GRP pipe system is structurally and functionally
adequate, but cannot be viewed as an absolute guarantee of performance.
For critical systems, other NDE methods (e.g. random verification of joint quality using
ultrasonics) can be used along with pressure testing and AE monitoring to determine the
presence of excessive debond or void areas that may have a detrimental effect on service
life.
The test pressure shall be raised over a period of 30 min or longer to 1.5 times the design
pressure or 0.89 times the qualified pressure, whichever is lower. The pressure-decay test
shall be conducted for a minimum of 1 hour.
Following successful conclusion of the strength test, a further leak test, at 1.1 times the
design pressure, should also be carried out for a minimum of 24 hours.
The following additional points are worth consideration:
All necessary measures shall be taken to remove air from the pipe spool or
vessel during filling;
Venting shall be carried out repeatedly at points in the test section where
air might accumulate, e.g. at ancillary piping;
56
Joint expansion and material creep, particularly when long runs of pipe are
being tested;
Water uptake properties of the GRP material, e.g. phenolic materials;
Temperature; which is more noticeable when testing is being carried out in
a hot or cold climate and if relatively large changes in temperature may
occur during the course of the test.
Additional checks shall be carried out after this time has elapsed, by visual inspection of the
complete piping system. Any leaking or weeping shall constitute a defect, and the test shall
be terminated and a repair effected.
The system shall be considered to have passed the hydrotest if there is no leaking or
weeping of water from the piping and there is no significant pressure loss that cannot be
accounted for by usual engineering considerations, e.g. thermal expansion of pipe, or other
factors previously agreed with the principal.
14.4
Ultrasonics
Ultrasonic pulse echo (PE) (where one transducer functions as both transmitter and
receiver) is the most commonly used ultrasonic test method for GRP in Oil industry
applications. In particular situations, through-transmission (using two transducers) may be
used.
The main defects detectable with ultrasonic testing are
A specific HOIS procedure on ultrasonic testing is being developed for GRP. In the interim
it is recommended that the Guidance in Annex E.4 of ISO 14692 Pt. 4 is followed together
with the procedure in Annex A of NORSOK Standard M-622.
If a more detailed or specific inspection procedure is required for ultrasonic B-scan or Cscan inspection then the following procedures may be considered:
DTI CPD4D offshore generic ultrasonic procedure and ultrasonic A-scan, B-scan
and C-scan procedure developed by ESR Technology (downloadable from
www.mms15.com )
DRA/NPL Working Draft Standard v 05, Fibre Reinforced plastics Ultrasonic Cscan inspection of composite structures: Parts 1-6, NPL, QinetiQ.
According to ISO 14692 Pt. 4, voids and areas lacking adhesive can be detected using
available ultrasonic methods to resolutions of approximately 10 mm and to depths of 100
mm; areas of poor adhesion, i.e. little or no adhesion but with joint faces in contact (also
known as kissing bonds) are not reliably detected by this method; variations in wall
57
The selection will depend on the quality of the composite and surface finish.
For filament wound offshore grade GRP piping a frequency of 2.25MHz or greater
may be used giving good resolution and better defined echoes from defects. Higher
frequencies (>5MHz) may not be beneficial because interpretation is confused by
natural signals from the ply structure.
The use of a delay-line probe is beneficial to reduce the initial reverberation in the
signal, and hence the depth near the surface that is not generally inspectable. The
depth affected is frequency and probe dependent.
A standard glycerine or gel couplant with low viscosity is preferred as this gives
better wetting of the surface. Coupling is improved if time is allowed for the couplant
to wet and seep into the surface.
It is beneficial to procure GRP joints with a good surface finish on surfaces from
which ultrasonic inspection may be conducted, to facilitate good coupling. The
increased inspectability after installation and in-service is likely to outweigh the
additional cost.
The clearest indication of laminar defects is the loss of back-wall echo, in addition to
any reflected signal that may occur.
The use of time-dependent gain can be beneficial if looking for back-wall echoes, for
example in thickness measurement.
Phased array or wheel probes may allow a more rapid inspection and provide more
visible defect indications. However, it is important to establish first that the probe
selection and UT conditions are correct using a conventional probe and A-scan.
Rapid scanning systems with wheel probes or phased arrays and good time
resolution can be used to collect 3D datasets, which can be used to reveal depth
dependent detail such as ply layers, porosity or wrinkling.
58
Time of flight diffraction (TOFD) with twin probes has been successfully used on
lined GRP vessels in South Africa.
Reflected pulses in GRP have more complex waveforms and less time separation
between the reflected pulses than is the case for steel. Therefore multiple echoes
cannot reliably be used in signal interpretation.
Two methods for increasing the time between reflected signals are: transmission
through flooded GRP pipes with the signal returning from the opposite pipe wall; and
use of a suitable (e.g. polymethyl methacrylate) stand-off.
Ultrasonic signals obtained from GRP can be of variable quality and possibly less
than would be expected for equivalent steel components as the FRP structure is
highly attenuating due to voids/porosity and signal scattering from the nature of the
laminated structure.
It was sometimes difficult to get a consistent back wall echo with some variability in
response between joints.
Ultrasonic B-scan is the preferred acquisition mode as B-scan images can give a
more clear delineation of bond line defects and back wall echo.
Generally 1 MHz probes gave a better image than 5 MHz but with reduced
resolution. 2 MHz appeared to be the optimum frequency for a filament wound
GRE pipe spool.
Some probe types produce a full width of joints on the screen whereas others were
much smaller and required multiple passes to achieve the same result.
The mixed woven Fiberbond spool was much more attenuating of ultrasound and
required a probe frequency of 1 MHz or below to get any reasonable response.
Most defects were evident by reflection within the bondline and/or loss of back wall
echo. In some cases it was difficult to hold large probes sufficiently perpendicular to
the surface and it might require a jig to be used to ensure correct alignment. A probe
manipulator or holder might be useful.
Some joints exhibited high signal attenuation with high levels of noise and coupled
with pipe elbow curvature it was sometimes difficult to get a good back-wall echo. It
was possible to obtain C-scan images by manual scanning, with appropriate gates
set for back wall and bondline but setting gates can be problematic.
Use of soft (gel) faced probes appeared to be acceptable but could be improved if
pipe surface had a smooth gel coat (e.g. apply smooth polymer peel-ply sheet to gel
coat in manufacture). There was some variability in acoustic response between
joints as some were more difficult to achieve a suitable back wall echo.
There may be some benefit in using water column probes for samples with rough
surfaces.
59
Figure 10 Typical A-scan display showing debond and back wall echo in the centre joint at
the 12 oclock position.
60
Figure 13 Advanced NDT Instruments - indications observed for loss of back wall echo
and interface signal changes
61
Figure 17 Scanning on bonded joint using 5 MHz GE Inspection phased array probe
14.5
Radiography
Radiography is used with variable success on offshore GRP components for detection of
volumetric defects and lack of adhesive in bonded joints. Of the various non contact
inspection methods trialled, i.e. laser shearography, microwave inspection and radiography,
the latter technique produced the best images. A portable pulsed X-ray system was used
with a digital detector array. In this case certain details of the joints were revealed including
bondline defects and clear images of the pipe wall section.
A specific radiography procedure for GRP is being developed within the HOIS programme.
In the interim it is recommended that the guidance in Annex E.5 of ISO 14692 Pt. 4 and the
radiography procedure in NORSOK M-622 Annex B are followed. As an alternative, inhouse or company specific procedures may be followed. Radiographic techniques are
described in the ASME Boiler and Pressure Vessel Code, Section V, Article 2.
Radiographic testing is generally not sensitive to surface roughness, but it is sensitive to the
orientation of the defect. It is relatively easy to perform onshore, while it is somewhat more
62
incorrect wall thickness (up to 20% deviation), or fit between male and female
adherends;
some voids, delaminations and lack of adhesive (up to 5 % deviation);
axial misalignment;
excess adhesive on inner wall of pipe at joint;
scale build-up on inside of pipe;
incorrect insertion of pipes in adhesive sockets.
Areas of poor adhesion, i.e. little or no adhesion but with joint faces in contact, are
not reliably detected by this method.
Radiographic test (RT) parameters, i.e. source energy and exposure time, should be
selected to allow for the low density of polymers and composites.
The technique (setup) recommended for bonded joints in NORSOK M-622 is double
wall single image (DWSI), as described in ASME B&PV Code Section V Art.2
The following film and sources are recommended by NORSOK M-622 for inspection
of GRP joints: Ir 192 with focus 2 x 2mm with activity less than 25Ci (or Se 75 in lieu
of or as a supplement to Ir 192), Agfa Gevaert D3, D4, D5 (bonded joints) and D7
for mechanical couplings.
For smaller diameter pipework (<6) and thinner components Ir-192 is likely to have
too high an energy to give good contrast.
63
For X-ray inspection (may not be applicable offshore) low to medium tube voltages,
typically in the range of 10 keV to 100 keV, are likely to be required. The voltage
required will depend critically on wall thickness. Heavier duty oil and gas composites
may require a higher voltage.
As no image quality indicators (IQIs) are currently available to verify the sensitivity
on GRE bonded joints, it is very important that a fine grain film and a source with
lowest energy as practical are used. The use of 1-2mm thick GRP step wedges is
suggested as an aid to optimise contrast and sensitivity (see Figure 9).
From radiographic test results it is possible to determine wall and laminate (i.e.
repair) thicknesses, particularly if tangential radiography is used. In some cases it
has also been possible to determine the winding angle, voids or lack of adhesive by
contrast/change of density (particularly where these become filled with e.g. water).
Digital radiography could offer advantages in adjusting the contrast after the
inspection has been concluded. This needs to be established.
Figure 18 Photo and schematic diagram of 1mm and 2mm thick GRP step wedges to aid
contrast and sensitivity setting for radiography of composite pipework. Holes are same
diameter as step height.
64
Figure 19 GE Inspection Rhythm 3.0 - Digital X-radiograph (weld flash filter) pipe centre
joint with bondline defects. Stepped GRP wedge and EN10-16 IQI wires were used.
20 mm diameter steel ball used for dimensional reference.
14.6
Tap testing
Tap testing is long established as an inspection method for aerospace composites to detect
disbonds and delaminations. This can vary from a simple coin tap or small hammer to
automated tap testers, or bond testers, that may offer a number of techniques (e.g.
mechanical impedance analysis (MIA), acoustic resonance and pulse-echo). The technique
is also referred to as acoustic impact (AI). The virtue of the technique is its simplicity.
Application to thicker section GRP and in Oil industry applications is less established.
Application to hand laminated quality GRP and thicker sections is more difficult than
commercial filament wound pipework.
Equipment for manual tap testing can vary from a simple coin to the Airbus hammer, which
facilitates a more controlled impact. Suitable commercial automated equipment includes the
Olympus Bondmaster, Fokker Bondester, V-95 Bond Tester, and Woodpecker.
It is recommended that an automated tap-tester is used, preferably a system that allows a
combination of techniques to be assessed. This allows the optimum technique to be found
for a particular GRP type, quality and location. The method should be qualified in each new
application using suitable reference panels or calibration samples.
It is permissible to use manual tap testing methods if the method can be qualified on
suitable reference samples. These allow a simpler but more subjective inspection. In this
case it is recommended that a record of the area scanned is kept on a transparent sheet as
detailed below, to allow future repeat inspection.
Guidance is given in aerospace standards including DOT/FAA/AR-02/121 14 Guidelines for
Analysis, Testing, and Non-destructive Inspection of Impact-Damaged Composite Sandwich
14
DOT/FAA/AR-02/121 Office of Aviation Research Washington, D.C. 20591: Guidelines for Analysis, Testing,
and Non-destructive Inspection of Impact-Damaged Composite Sandwich Structures; U.S. Department of
Transportation, Federal Aviation Administration; Final Report, March 2003, Section 4.
65
This method is portable, but no data is recorded. The accuracy of this test depends
on the inspectors subjective interpretation of the test response; therefore, only
qualified personnel should perform this test.
Care should be taken to avoid causing impact damage to the GRP component.
Tap testing is widely used for a quick evaluation of any accessible composite
surface to detect the presence of delamination or debonding.
The basic tap testing procedure consists of lightly tapping the surface of the part
with a coin or light special hammer with a maximum of 2 ounces or any other
suitable object.
The surface should be dry and free of oil, grease, and dirt.
66
Data should be recorded if possible to allow later examination. For disbonding and
delaminations, it has been found that the signature, fingerprint or shape of the
response and decay curve varies, particularly in thicker section components.
Such variations in the signal response can provide a better delineation of defects
than the amplitude response.
There are advantages in using a system with encoder or probe location method that
allows the data to be presented as a C-scan type map.
Higher loading mechanical tapping systems developed by HOIS and others for
inspection of passive fire protection may have application in thicker section GRP
and fittings.
14.7
Thermography
Thermography is a commonly used NDE technique on composites and one of the methods
recommended for Oil industry application in ISO 14692. Thermography techniques use
infra-red sensitive camera detectors and film or video display and recording methods, to
monitor the apparent surface temperature of an area of component. The thermal response
from a component is altered in the presence of a defect.
Defects detectable by thermography include: delaminations, disbonds, voids, inclusions and
impact damage. The main defects identified for thermography in ISO 14692 are: scale
build-up, major deviations in wall thickness, and areas in the joint lacking adhesive.
ISO 14692 identifies thermography as an NDE method under development that may be
applied for inspection of GRP if reliable calibration standards are established and agreed in
advance of the inspection. It is assumed that this refers to transient thermography. To see
defects at a reasonable resolution it is necessary to provide a pulse of heat into the
component and use a high quality thermal imaging camera. This is the basis of transient
thermography, the most common thermography method applied to composite materials,
also known as pulse-video thermography or active thermography. Fully integrated systems
including flash tube, thermal camera and data processing hardware are now routinely
available. Sensitivity can be markedly improved by using synthetic processing methods.
There are a number of ways to provide heat including, flash lamps, induction heaters (not
for GRP) and hot-air guns. Simple monitoring of temperature with a thermal camera is
known as passive thermography. This is only really suitable for identifying global
differences, acting as a form of enhanced visual inspection. Passive thermography is
sometimes used to detect impact damage on GRP using natural thermal transients. Passive
thermography can in some circumstances reveal damage in GRP vessels and pipework,
particularly if the process fluids are at a higher temperature. A procedure for thermography
was included in the 1999 Version of NORSOK M-622, but guidance in ISO 14692 Pt. 4. is
limited. There is no current Oil industry standard on thermography.
If no suitable in-house procedure is available it is recommended either that the procedure in
the 1999 Version of NORSOK M-622 , the advice or procedures from the equipment
supplier, or the active thermography procedure developed by ESR technology in the DTI
CPD4D programme is used.
67
For application of thermography on Oil industry GRP components and joints the following
additional points are worth considerations:
A high speed, high sensitivity good resolution thermal imaging camera should be
used.
The camera should be capable of uninterrupted monitoring of the sample surface for
the entire duration of the acquisition. Cameras with automatic internal shuttering
mechanisms should allow the shuttering to be disabled during the data acquisition
period.
The camera should provide real-time digital output of the acquired signal (typically 8
bits at 60Hz). The camera output signal should be approximately linear over the
(post-flash) temperature range of the sample. The camera wavelength should be in
either the 2-5 micron range or the 8-14 micron range, selected such that the test
material is not IR translucent in the spectral range of the camera. A medium wave 25 micron wavelength is typical for GRP.
The optics and focal plane should be sufficient so that the projection of nine
contiguous pixels onto the sample plane is less than or equal to the minimum flaw
area that is to be detected.
For Oil industry applications flash lamp or air heater (hairdryer) are the most
practical heat sources. For GRP coatings on steel components or repairs an
induction heater could also be considered.
68
14.8
Shearography
Laser shearography is not included as an NDE method in the current version of ISO 14692
or NORSOK but accepted as a common inspection method for composites in other
industries. The method measures and images the surface strain field at high resolution
which is altered in the presence of internal and near surface defects. The method is the
preferred method for inspection for delaminations, disbonding and core damage of GRP
sandwich structures in lifeboats and marine hulls. Commercial integrated systems are
available.
The basis of all shearography techniques is that the video system captures a sheared
image of the laser speckle pattern formed of the surface at one stress state and stores this
as a reference image. While the stress state is modified, new images of the surface are
compared against the stored reference by image subtraction correlation; only the difference
is presented on the screen in real time. This process is carried out, and the screen updated
at the rate of 30 times a second until the predetermined second stress state is reached.
There is very little experience of use of shearography in Oil industry applications although it
is routinely used in the Marine industry and on GRP lifeboats. The method has been trialled
on offshore adhesive joints and overwrap repairs. Vacuum loading systems used on boat
hulls are robust. There is some concern about more delicate laser systems being
sufficiently robust to use offshore. The technique is fast and global covering several square
metres in a few hours but relatively costly.
In the absence of an offshore specific standard, it is recommended that in-house
procedures, advice or procedures from the equipment suppliers, or the guidance in the
following procedures and practice documents are followed:
DTI CPD4D
The following additional points should be noted in relation to oil industry application:
The technique relies on the images being captured in two different stress states.
69
For large diameter pipework, tanks or vessels, use of a standard vacuum loading
shearography system (vacuum hood) is recommended as used in global inspection
of lifeboat hulls.
For smaller diameter pipework the stress difference required may be achieved by
change in internal pressure or heating with an air heater (or hairdryer).
Suitable and specific image processing software is required for calculation and
display of the shearography images, phase stepping comparison etc.
Phase processing of the data can give clearer images of defects than standard
fringe pattern analysis.
The fringes in the shearography data can be used to give an indication of the
relative bond conditions (lack of, or increased adhesion).
70
Figure 21 Interference fringes were densely spaced near known bondline defects
14.9
Acoustic emission
Acoustic emission is a well established NDE monitoring method for composites and
guidance on application is provided in ISO 14692 Pt. 4 Annex E.6.
Standard procedures for performing acoustic emission inspection on GRP materials exist
e.g. ASTM E2191 and ASME V Article 11, and should be available from acoustic emission
equipment suppliers.
It is recommended that supplier guidance or in-house procedures where applicable are
followed. In addition the following standards should be consulted:
71
leakages.
Acoustic emission requires that the component be under load, up to a maximum of the
design load. Only growing or propagating defects are detected.
14.10 Acousto-Ultrasonics
Acousto-Ultrasonics (AU), or the Stress Wave Factor (SWF) technique, is a variant on
acoustic emission recently developed in the aerospace and defence industries for
inspection of thick section GRP and also applied in Oil industry applications. Commercial
equipment for AU is available but not always available for trials. An ultrasonic wave is
generated in the component and acoustic sensors used to monitor the emissions from the
component. These can be analysed to build up image maps of the component resembling
ultrasonic C-scans. Pulses of ultrasound are injected into the component and are detected
some distance away on the same side of the component.
The method seeks to characterise a component by investigating the way in which it
modifies an ultrasonic pulse that travels through it non-destructively.
Standard procedures are available from the AU supplier. The following procedure should
also be consulted:
Commercial AU systems are capable of generating C-scan images of amplitude and Time
of Flight (TOF), as well as A-scans of different areas of the composite structure using a
high-precision computer controlled scanning bridge. This has been used in onshore and
refinery applications but has yet to be established offshore.
The method is relatively new commercially but was evaluated as far back as 1988 as a
remote method for adhesive bond inspection. The name Acousto-Ultrasonics (AU) is
derived from a combination of Acoustic-emission monitoring and Ultrasonic characterization
originally developed by NASA 15.
The following points should be noted:
15
72
The method is well suited to thick section components and complex sandwich
structures.
As a remote, non-contact and high speed method it offers the potential to screen
large areas of pipes or vessels that would be impracticable and slow by
conventional ultrasonic methods.
The advantages are the speed of inspection, that it is non contact and the particular
suitability to thick section GRP components.
The disadvantage is the novelty of the method and the quality and resolution of the
data, which is less good than more conventional NDE methods such as ultrasonics
or shearography.
73
The microwave power associated with near-field systems is low (in the milliwatt
range) and therefore poses no Health and Safety concerns.
Sufficient access is required for the probe and scanning system. The data
acquisition system (PC-based) may be sited remotely.
Interpretation of the data and images arising is still quite subjective and there is
limited experience in interpreting microwave data for GRP pipework and joints.
No couplant is required.
For NDE purposes the two main operating parameters to be considered are
operating frequency and stand-off distance (the distance between the inspection
surface and the waveguide aperture).
The scanning system should be set up on the pipework or component with the probe
in contact or at a fixed stand-off (few mm) from the component surface. The region
to be scanned should be defined in the data acquisition software relative to a known
position on the specimen. The probe should be scanned along and circumferentially
on the specimen to cover the area of interest. A 3D data set through the sample is
generated.
Inspection should normally be undertaken from one side only or from both sides
looking for reflections. Through-transmission inspection may also be undertaken.
Suitable open-ended waveguide probes are generally available. Near-field the size
of the source is significant.
74
Near-field microwave methods that transmit the wave into a dielectric structure
located in the near-field of a sensor offer significant advantages: inspection systems
are compact in size and can be battery operated; spatial resolution is primarily
influenced by the dimensions of the probe rather than the operating wavelength;
spatial resolution is better compared to that achieved with plane or spherical waves
(far-field, radar); contact as well as non-contact measurements are possible;
mapping the magnitude or phase of the reflected wave can be used to create two- or
three-dimensional images of the structure under investigation; unwanted reflections
from changes in geometry associated with far-field methods are minimised.
75
Barcol or Shore hardness is an indirect method of measuring the degree of cure in vinyl
ester and polyester laminates and laminated joints.
DSC is a quantitative, accurate and relatively fast semi-non-destructive technique which is
based on the measurements of thermal changes related to phase transitions and chemical
reactions, such as the curing of thermoset resins. Small samples can be cut from the
external adhesive seams of the joints for measuring the glass transition temperature, Tg, by
DSC analysis.
It is recommended that the two methods are applied as identified in ISO 14692 Pt 4 and no
additional guidance is offered.
15.0
Monitoring methods
The use of structural health monitoring (SHM) methods such as acoustic emission,
conventional strain gauging or optical fibre technology on composite components is well
established.
Guidance on monitoring methods for composite joints has recently been produced in the UK
TSB sponsored ACLAIM project 16. ACLAIM formed part of the Advanced Composite Life
Assessment and Integrity Management (ACLAIM) Technology Programme project, funded
by the Technology Strategy Board (TSB) of the UK Department for Business, Enterprise
and Regulatory Reform (BERR).This is recommended as the main reference if application
of monitoring methods is considered.
Relevant outputs from the SiMoNet network 17 should also be considered well as the
relevant standards noted earlier for Acoustic Emission. SiMoNet is focussed on industrial
application of monitoring and sensor technology.
Monitoring methods for GRP joints are also being considered in the current TSB funded
Imajine project, the successor to ACLAIM. This includes the use of embedded optical fibres
within the adhesive in bonded joints to monitor service degradation and properties. Similar
work on optical fibres is being undertaken by Smart Fibres Limited and at Surrey University.
By use of specially graded fibres it is possible to measure parameters including
16
ACLAIM Current practice guide: A Guide to Current Structural Health Monitoring (SHM)
Practice for Advanced Composite Applications, S Frost et al, ISBN: 978-0-946754-52-6, ESR Technology,
National Physical Laboratory (NPL), 2008.
17
SiMoNet Structural Integrity Monitoriing Network www.simonet.org.uk
76
16.0
There is a number of developing NDE methods that show promise for application offshore
but are not currently included in industry or composite standards or guidance. These
include:
Rapid scanning methods for ultrasonic C-scan and B-scan inspection of pipes and
vessels to detect delaminations, disbonds and wall thickness loss ( e.g. wheel
probes or other phased array methods);
Increased use of linear scanning (B-scan) methods including TOFD for inspection of
composite pipes and vessels in place of point measurements (A-scan or thickness).
This significantly improves the reliability of detection of erosion, wall thickness loss,
liner degradation or environmental ingress;
For matrix cracking further development of the axial ultrasonic velocity measurement
is recommended as a function of matrix cracking;
77
17.0
The evaluation and assessment of damage in composites is outside the scope of this
document. Limited information and guidance is provided below.
For GRP pipes and fittings it is not possible to develop a single all encompassing
assessment procedure. For each identified defect or damage mechanism within a
component a separate assessment procedure is required.
As an initial assessment, acceptance criteria for common defect types and locations are
given in the Tables in ISO 14692 Parts 2 and 4 reproduced in Appendix C and included
earlier in this document. This advice primarily relates to manufacture and qualification but
is worth consideration where defects are encountered.
Defect assessment for composites is normally experienced based. Information is fed back
to the original designer and a decision is made on whether to repair, replace or leave as is.
The questions an integrity assessment procedure should answer are:
The HSE Good Practice Guide4 on GRP pipework defines two priority levels in terms of
criticality. For priority level 1 (P1Urgent) a risk assessment shall be performed as a matter
of urgency and suitable mitigation strategies implemented. For priority level 2 (P2- Caution)
items, a risk assessment should be performed as soon as reasonably practicable to identify
a suitable course of action.
For damage mitigation pipes should not used as a step up, not be exposed to
deleterious foreign fluids and not be exposed to heat sources. Earthing cables (if
required) shall be present and in good condition.
The general appearance of pipework should be uniform with smooth surfaces and
none of the following or within prescriptive limits given in associated standard/
manufacturers guidance: cracks (including adhesive fillets between repair laminate
and substrate), delamination, wear, chips/gouges, resin loss, exposed fibres/cut
edges, leaks/weepage, dry spots, variation in resin colour, contamination/foreign
matter, heat damage, softening/ bulging/ discolouration/ burn, blisters/ pin holes/
pores, chemical attack, erosion .
The following damage types are considered to be unsafe (P1) for pipes and joints if
they extend deeper than the surface resin rich layer (gel coat) and risk assessment
shall be performed as a matter of urgency and suitable mitigation strategies
implemented: weepage, leak, fracture or cracking, delamination, cracking in a
flange, exposed fibres post-impact, heat damage- softening/ bulging/ discolouration/
78
The following items are also classified as unsafe (P1-Urgent): supports which
provide inadequate support or are missing, extensively damaged or corroded; or
joints which leak, are misaligned or deformed or exhibit the damage types identified
above extending below the resin rich layer.
The following damage types are defined as priority 2 (P2- Caution) for pipe and
joints provided they do not extend deeper than the surface resin rich layer: loss of
resin rich surface layer; blisters/pin holes/pores/chips/wear; heat damage discolouration, burn, foreign matter/contamination; and minor damage/corrosion of
the supports, provided they remain still structurally sound.
Techniques for integrity assessment of metallic materials are well established such as API
579. No such equivalent yet exists for composite materials. For composite materials
guidance exists in standards such as ISO 14692 but assessment methods are still under
development. Generic assessment procedures consist of the following:
18
Ageing of composites in oil and gas applications, Ed. Rod Martin, Woodhead publishing, ISBN 978-1-84569352-7, 2008
19
Measurement Good Practice Guide No. 78 Assessment and Criticality of Defects and Damage in Material
Systems;,NPL/ ESR Technology; DTI MMS13 Programme 2002-2005; National Physical Laboratory Report
ISSN 1368-6550, June 2005.
79
18.0
Failure prevention
Limit the operating temperature so that the effects of chemical attack are minimised.
If possible limit the concentration of aggressive fluids in the flow;
Limit axial and pressure loads to no greater than the long term allowable loads;
Ensure that pipe joints are installed correctly; by trained installers and that the joints
are not over-loaded.
Limit the operating temperature so that the effects of chemical attack are minimised;
In summary composite materials are used in many safety critical applications, the most
obvious examples are in aerospace structures (e.g. Airbus 350 and Boeing 787
Dreamliner).
Large structures pose particular difficulties for inspection. Examples include oil & gas (FRP
pipes and composite repairs) and Civil Infrastructure applications including new build and
bridge strengthening. Inspection tools are now available but the nature of composites still
present unique challenges compared to metals. Recent developments have included large
area, non contact inspection techniques such as microwaves, shearography and
thermography for use in inspection of large wind turbine blades used in the growing
offshore wind energy sector. Potential strategies include fast screening, global methods,
monitoring or application of risk-based inspection methods depending on how much is
known on type and location of damage and integrity requirements. Ultrasonic inspection
using phased arrays mounted inside roller probes are being used effectively for in service
inspection particularly in marine and aircraft inspection.
19.0
Acknowledgements
The inputs from several HOIS members are gratefully acknowledged, especially those who
conducted inspection trials at the ESR Technology NICE facility. The continued support of
the HOIS Project Champion Francisco Marques of Petrobras and his colleagues in Rio de
Janeiro are also acknowledged. HOIS is an established Joint Industry Programme aimed at
quantifying and improving the performance of NDT technology in the oil and gas industry.
80
Appendices
APPENDIX A
APPENDIX B
APPENDIX C
APPENDIX D
81
Appendix A
In addition to the NDE related codes and standards noted in the practice document, there
are a diversity of codes and standards relevant to design, testing and qualification of GRP
components. These include the following:
NORSOK M-622 NDE of GRP piping systems, tanks and vessels; the
Norwegian Oil industry Association (OLF); February 1999
UKOOA Guidelines OPS04 1994 Fibre Reinforced Plastics (FRP) for use
Offshore
82
DIN 16966-1, Glass fibre reinforced polyester resin (UP-GF) pipe fittings
and joint assemblies; fittings; general quality requirements and testing
DIN 16966-2, Glass fibre reinforced polyester resin (UP-GF) pipe fittings
and joints; Elbows, Dimensions
DIN 16966-4, Glass fibre reinforced polyester resin (UP-GF) pipe fittings
and joints; Tees, Nozzles, Dimensions
DIN 16966-5, Glass fibre reinforced polyester resin (UP-GF) pipe fittings
and joints; Reducers, Dimensions
DIN 16966-6, Glass fibre reinforced polyester resin (UP-GF) pipe fittings
and joint assemblies; collars, flanges, joint rings, dimensions
DIN 16966-7, Pipe joints and their elements of glass fibre reinforced
polyester resins Part 7: Bushings, flanges, flanged and butt joints; general
quality requirements and test methods
DIN 16966-8, Glass fibre reinforced polyester resin (UP-GF) pipe fittings
and joints; Laminated joints; Dimensions
83
HSE/ HSL GRP Pipe Fact sheet, Draft: Final Revision1, March 2009.
HSE/ HSL Composite Repair Fact sheet, Draft: Final, March 2009.
HSE RR039 Cost Effective use of composites offshore; Part B: Summary of the
Joint industry - industry programme on the cost effective use of fibre reinforced
composites offshore. UK Health & safety Executive HSE; HSE Research Report,
2003 http://www.hse.gov.uk/RESEARCH/rrhtm/rr039.htm
HSE-AL-FWI
Specification for full-wrapped Composite aluminium alloy gas
cylinders, Issue 1 - January 1991
HSE RR198, TEMPSC Structural Design Basis Determination; Part 1 Input Data
Capture and Review, research report 198, 2004
HSE RR059 An assessment of skin sensitisation by the use of epoxy resin in the
construction industry, HSE research report 198, 2003
84
Appendix B
Guidance tables are provided in ISO 14692 for visual inspection, inspection of process
pipework as well as recommendations on NDE techniques for specific defect types. These
are reproduced here for completeness.
85
86
87
88
89
Table 6 Defect types, acceptance criteria and recommended corrective action (Table A-1)
90
91
92
93
94
95
Table 7 Summary of recommendations for NDE method and defect type in ISO 14692 Part
4 Annex E. (Compiled by ESR from Annex E of ISO 14692 Part 4.)
NDE Method
Details
ISO 14692
Pt. 4
Annex E.
Section
Visual inspection
E.2
E.2
E.2
Pressure testing
Ultrasonic testing
Radiographic testing
Acoustic emission
testing
Differential scanning
calorimetry (DSC) and
Barcol hardness
testing
Thermographic testing
E.2
E.2
E.2
E.2
E.3
E.3
E.3
E.3
E.4
E.4
E.4
E.5
E.5
E.5
E.5
E.5
E.5
E.6
E.6
E.6
E.6
E.6
E.6
E.6
E.6
E.7
E.7
Scale build-up
Major deviations in wall thickness,
Areas in the joint lacking adhesive.
E.8
E.8
E.8
96
Appendix C
97
C.2 Overview
Composite components are subject to ageing by a number of processes which can reduce
the strength and properties of the pipework. This includes physical ageing, mechanical
ageing and chemical ageing. These are normally allowed for in design through regression
curves; encompassed in relevant design standards such as ISO 14692.
Physical ageing processes include moisture ingress, swelling and plasticization common
with other polymeric components. These processes are referred to as static fatigue.
Mechanical ageing refers to the development of defects during service including matrix
cracking, delaminations and impact damage. Chemical ageing refers to environmental
processes that change the chemical structure or bonding within the component and
therefore degrade its physical properties; this includes hydrolysis, and modification of
bonding or cross linking. These ageing processes apply equally to adhesive as well as the
composite resin and result in a change in the glass transition temperature tg and properties.
In its broadest definition ageing can be defined as the reduction in performance of a
component as a function of the applied conditions. The three primary causes of ageing for
composite components in the Oil and Gas industry are through chemical species ingress,
elevated operating temperature and length of time of load application.
Service experience in the Oil and Gas sector is that damage to the composites from ingress
of the environment (internal or external) is minimal and does not significantly affect
materials properties. Progressive damage may occur under service loadings by matrix
cracking, the normal response of composite materials to loading.
For process pipework the main service risk is weepage of the process fluid eventually
leading to failure of component. The other failure mode of concern is fibre failure. Generally,
this occurs at the ultimate load-bearing capacity of the composite component and results in
20
Ageing of composites in oil and gas applications, S Frost; Ageing of Composites, Ed. Rod Martin, Chapter 14
p 375-395, Woodhead publishing, ISBN 978-1-84569-352-7, 2008
98
Hogg, PJ and D Hull, "Micro mechanisms of Crack Growth in Composite Materials under Corrosive
Environments," Metal Science, pp. 441-449, August-September 1980.
22
Ageing of composites in oil and gas applications, S Frost; Ageing of Composites, Ed. Rod Martin, Chapter 14
p 375-395, Woodhead publishing, ISBN 978-1-84569-352-7, 2008
23
Ageing of composites in the chemical process industry,R Martin; Ageing of Composites, Ed. Rod Martin,
Chapter 17 p 375-395, Woodhead publishing, ISBN 978-1-84569-352-7, 2008
24
AG Gibson, The cost effective use of fibre reinforced composites offshore. University of Newcastle Upon
Tyne, HSE Research Report 039, 2003
25
C.M. Renaud and M.E. Greenwood, Effect of Glass Fibres and Environments on Long-Term Durability of
GFRP Composites.
99
100
Ply
Delaminations
101
Figure 26 Microstructure of GRP pipe showing matrix cracking due to the application of
continuous load.
102
Figure 27 Examples of chemical attack. Left, Breakdown of inner liner due to chemical
attack; right, SEM electron image of glass fibres from a filament wound pressure vessel,
following sulphuric acid attack. The micrograph shows considerable fibre cracking due to
stress corrosion cracking
103
Impact
damage
Weepage
failure
point
104
Figure 30 Photo of the female face of an adhesively bonded joint showing lack of adhesive
(dark areas)
105
Interface
Pipe
Figure 32 Butt and wrap joint showing the interface, the normal location for bond failure.
106
C.12 Cavitation
As with metallic components, composite piping can be subject to cavitation damage.
107
Figure 34
Failure of a GRP lined bottle showing burst failure (significant fibre breakages)
108
30 mm
diameter
through wall
defect
located at top
of bend
109
Appendix D
Examples of in-service degradation of
composite components
Figure 36 GRE Fire water system pipe damaged in service (through wall defect).
OD 60mm, WT 4mm Length 360mm
Figure 37 GRP oily waste pipe, 5 years service, weeping (from impact damage).
OD 159mm WT 5.5mm Length 975mm
110
Figure 38 Reverse osmosis RTR spools which exhibited pipe weepage during hydrotest.
1.5 and 4 HD flanges ANSI B16.5 Cl-150
Figure 39 GRV piping, one year in-service, SO2/ H2SO4 acid, wept from a manufacturing
defect. OD 65mm WT 5.6mm Length 385mm
111