As seen in the Ocotober 2009 issue of
Developing standards for
composite repair systems
Recently developed ASME PCC-2 and ISO 24817 standards
offer guidance on the design of composite repair technologies.
Dr. Chris Alexander, Stress Engineering A properly-designed repair system Applicability sections from Part 4
Services, Inc., Houston, Texas ensures that strains in the reinforced (Nonmetallic and Bonded Repairs) of the
steel and reinforcing composite material 2006 edition of ASME PCC-2.
A
significant amount of work has do not reach unacceptable levels. This 1.1 Scope. This Article provides the
transpired over the past several article provides a basic overview of the requirements for the repair of pipework
years in generating consensus- design philosophy embedded into the and pipelines using a qualified Repair
based standards that include ASME current design codes, as well as present- System. The Repair System is defined as
PCC-2 and ISO 24817 for developing ing results associated with several spe- the combination of the following ele-
composite repair systems. The intent in cific studies that were conducted to eval- ments for which qualification testing has
developing these standards has been to uate composite repair performance. been completed.
provide industry with guidelines for (a) substrate (pipe)
designing composite repair systems to Repair codes (b) surface preparation
ensure that damaged pipelines are safely In order for composite systems to repair (c) composite material (repair laminate)
and properly reinforced. damaged pipelines, it is critically important (d) filler material
With the numerous composite repair that they be designed to ensure that adequate (e) adhesive
systems currently available to pipeline reinforcement is present. The ASME PCC-2 (f) application method.
operators, the importance of evaluating the and ISO 24817 composite repair codes (here- The composite materials allowed for
capabilities of each system cannot be over- after referred to as Codes) provide the the Repair System include, but are not
stated. The fundamental design variables required details to design a system that has limited to, glass, aramid, or carbon fiber
available to manufacturers are stiffness, sufficient stiffness, strength, and thickness. reinforcement in a thermoset resin (e.g.
strength, and thickness of the composite. Provided below are the Scope and polyester, polyurethane, phenolic, vinyl
ester, or epoxy) matrix. Fibers shall be
continuous.
1.2 Applicability. This Article addresses
the repair of pipework and pipelines origi-
nally designed in accordance with ASME
B31.1/B31.3/B31.4/B31.8, and ISO 15649
and 13623.
The Applicability section goes on to
state that the Code covers situations
involving damage that include internal
and external corrosion, external damage
such as dents, gouges, and cracks, as
well as manufacturing defects. The repair
of leaks is also permitted, although for
high pressure transmission pipelines, this
repair option is unacceptable at the pres-
ent time based on the author’s opinion.
Because the focus of this article is repair-
ing high pressure gas and liquid trans-
Figure 1. Strains measured in composite reinforced corroded pipe sample. mission pipelines, there is no discussion
October 2009 www.pipelineandgastechnology.com
considering hoop stresses based on
internal pressure. Note that by including
the Ps term credit is taken for strengths
associated with the remaining steel.
(ASME PCC-2 Equation 1)
In reviewing Equation 1, is it clear that
Figure 2. Schematic diagram of composite repair pipe test sample. the relative stiffness values of the steel
(Es) and the composite (Ec) are inte-
grated to calculate the minimum required
on the repair of leaking pipes. corroded steel, although the option for thickness. The use of this equation
The function of the Codes is design not including this contribution is an assumes that the substrate (e.g. remaining
and within ASME PCC-2 there are three alternative provided by the Codes that reinforced pipe material) does not yield
basic approaches for determining the results in a greater required minimum and remains elastic throughout operation.
minimum required thickness for a par- composite thickness. The second design option that is
ticular composite material in repairing The following ASME PCC-2 nomen- available in PCC-2 is calculating the
corrosion and are listed below. clature (i.e. variable descriptions) is minimum composite thickness based on
• Section 3.4.3 − Pipe Allowable Stress used in the calculations that follow. hoop strain due to internal pressure
• Section 3.4.4 − Repair Laminate D External pipe diameter (inches) using Equation 4. Also included in PCC-
Allowable Strains Ec Tensile modulus for the compos- 2 is Equation 3 that integrates the effects
• Section 3.4.5 − Repair Laminate ite laminate in the circumferential of internal pressure in the pipe at the
Allowable Stresses Determined by direction (psi) time of the composite installation,
Performance Testing. Es Tensile modulus for the pipe steel although this equation is not presented
The contents of this article should not (psi) in this article.
be used as a substitute for actually con- f Service factor (inverse of safety
sulting and utilizing the composite factor, provided in ASME PCC-2
repair design codes (i.e. ASME PCC-2 Table 5)
and ISO 24817). These design codes P Internal design pressure (psi) (ASME PCC-2 Equation 4)
provide details that deal with specific Ps Maximum Allowable Working
of an allowable long-term strain, εc, is
issues when using composite materials Pressure (MAWP) for corroded In solving Equation 4 the designation
in repairing and reinforcing damaged pipe using B31G, etc.
pipelines that should not be ignored. s Specified Minimum Yield Strength required. Table 4 from ASME PCC-2
(SMYS) for pipe (psi) specifies that for continuous (sustained)
Determining repair thickness slt 95% lower confidence limit of the loading conditions the allowable long-
The sections that follow provide specific long-term composite strength via term strain for the repair laminate is lim-
details on the above referenced ASME testing (psi) ited to 0.25%, while for rarely occurring
PCC-2 sections and their unique design t Nominal wall thickness of pipe loads it is 0.40%.
approaches. An example problem is also (inches) The minimum required thickness using
provided to demonstrate the level of tmin Minimum repair thickness of the ASME PCC-2 Section 3.4.5 method is
conservatism associated with each calcu- composite (inches) based on performance testing of the com-
lation method and the benefits in ts Minimum remaining wall thick- posite material itself and is the third ASME
εc Allowable circumferential strain
designing a performance-based system as ness of pipe (inches) PCC-2 design option. This approach
detailed in Section 3.4.5, even though requires additional testing of the composite
additional efforts and costs are associ- The first design option that is pro- material beyond what is required for the
ated in qualifying a given composite sys- vided in ASME PCC-2 is the most con- other calculation methods, such as the
tem to this level. Due to limited space in servative of the three options presented 1,000 hour survival test as presented in
this article, all calculations assume in this article. Equation 1 is used to cal- Section V-2.1 in Appendix V of ASME
structural contribution of the remaining culate the minimum required thickness PCC-2 based on ASTM D 1598. In this
www.pipelineandgastechnology.com October 2009
Composite Repair Focus Series
particular test internal pressure is applied
to a test sample having a minimum diame-
ter of 4 inches and a minimum thickness of
0.120 inches. The sample’s internal pres-
sure and composite laminate thickness are
selected to maximize the long-term com-
posite stress, slt, using the equation below.
(ASME PCC-2 Equation V-1) Figure 3. Schematic showing location of strain gages of photo of machined region.
In this qualification test, three identi-
cal test samples must be repaired and reviewing the calculated results provided in 50% corrosion that was reinforced using a
survive 1,000 hours of testing with no Table 1 is the reduction in the minimum carbon-epoxy system. During testing strain
deterioration of the laminate in the form required laminate repair thickness associ- gages monitored strain in the reinforced
of cracking, delamination, or leaking. ated with the three calculation options. It is steel region and were used to demonstrate
Once the long-term composite design clear from this presentation that with the the level of reinforcement provided by the
strength is established based on the 95% inclusion of the long-term data as required composite material. The second case study
lower confidence limit, the minimum com- for using Equation (9), a less conservative discusses results associated with a testing
posite repair thickness is calculated using composite thickness results due to the program to evaluate the capacity for a car-
Equation 9 from ASME PCC-2. In review- greater effort undertaken in determining bon-epoxy system to reinforce 75% corro-
ing this equation, the use of a service fac- the actual long-term strength. sion in a 12-in. nominal diameter pipeline
tor, f, is required. The service factors are subjected to cyclic pressures.
basically the reciprocals of safety factors Case studies Case study No. 1. In 2006, a program
and are listed in Table 5 of ASME PCC-2. If One of the consistent elements associated was conducted for the U.S. Minerals
one opts to establish long-term laminate with the development and qualification of Management Service to evaluate the use
strength using the 1,000 hour data, the composite repair systems has been experi- of composite materials in repairing off-
service factor is 0.5 (i.e. safety factor of 2.0 mental evaluation. This evaluation has shore risers. Part of this study involved
for the composite material’s strength). involved assessments at both the coupon repairing a burst test sample having 50%
and full-scale levels. Evaluating material corrosion using a 0.60-in. thick carbon-
performance at the coupon level is an epoxy system that included two pre-
effective means for determining the cured half-shells. Strain gages were
(ASME PCC-2 Equation 9) strength of the composite, while at the installed in the corroded region of the
same time being less expensive than full- 8.625-in. x 0.406-in., Grade X46 pipe
Calculating repair thickness scale testing. The primary emphasis in the sample and monitored during pressuriza-
Table 1 provides calculations associated Codes up to this point in time has been in tion to failure. Results from this test are
with the reinforcement of a 12.75-in. x designing composite repair systems to provided in Figure 1. Included in this
0.375-in., Grade X42 pipeline having reinforce corrosion; however, there is also plot are a few annotations that designate
50% corrosion where the MAOP is de- an abundance of data demonstrating that the lower bound collapse load (5,975 psi)
rated from 1,778 psi to a MAWP of 1,000 composite materials can be used to rein- from which the design pressure (2,988
psi due to the corrosion. Presented are force wrinkle bends, elbows, field branch psi) is calculated. This design pressure
results for all three calculation methods connections, dents, and others anomalies. exceeds the maximum allowable operat-
discussed previously (i.e. pipe allowable Results from several prior studies have ing pressure of 2,887 psi of a non-cor-
stress, repair laminate allowable strains, been presented in the previous articles roded pipe. The results of this program
and repair laminate allowable stresses associated with this series. demonstrated that the carbon repair was
determined by performance testing). It Two case studies are presented below effective in reinforcing the corroded pipe
should be noted that the contribution of that deal specifically with the reinforce- and ensured that strains in the reinforced
the remaining steel is considered in all ment of corrosion using composite materi- steel did not reach an unacceptable level.
provided calculations. als. The first case study involves the repair This study is classified as one based on
An extremely important observation in of an 8-in. nominal diameter pipeline with strain-based design limits.
October 2009 www.pipelineandgastechnology.com
Table 1. ASME PCC-2 Calculated Thickness Values
ASME PCC-2 ASME PCC-2 Calculated Values
(see Note below for
Equation Number Equation variable values)
(1) 0.787 inches
(4) 0.306 inches
(9) 0.138 inches
Notes (input variables used in above equations) Ps 1,000 psi (de-rated operating pressure due to slt 50,000 psi (long-term composite strength based
Es 30 x 106 psi (steel pipe modulus) presence of corrosion) on ASME PCC-2 Appendix V directives)
εc
Ec 4.5 x 106 psi (composite laminate modulus) t 0.375 inches (pipe nominal wall thickness) ts 0.188 inches (remaining pipe wall thickness due
s 42,000 psi (pipe Minimum Specified Yield 0.25% (allowable long-term composite strain to corrosion)
Strength, or SMYS) from ASME PCC-2 Table 4)
P 1,778 psi (MAOP) f 0.5 (Service Factor from ASME PCC-2 Table 5)
Case Study No. 2. Most of the experi- sion in a 12.75-in. x 0.375-in., Grade X42 cycles to failure
mental research associated with the com- pipeline. Figure 2 is a schematic showing • Carbon system (Pipe #4): 532,776
posite repair of corroded pipelines has the geometry of the test sample used in cycles to failure.
focused on burst tests. The general philos- this study, while Figure 3 shows the posi- Minimal information is provided
ophy has been that in the absence of tioning of strain gages beneath each repair with the above data (e.g. no informa-
cyclic pressures during actual operation, in the corroded region. The test samples tion provided on thickness, composite
there are few reasons to be concerned were pressure cycled at a pressure range of modulus, filler materials, fiber orienta-
with qualifying composite repairs for 36% SMYS (i.e. 894 psi for this pipe). tion, etc.). However, one can definitely
cyclic conditions. One could argue that Tests were performed on five different conclude that all composite repair sys-
only liquid transmission pipelines need to composite systems that included the fol- tems are not equal. The study on the
be concerned about cyclic pressures. lowing cycles to failure: carbon composite system having four
However, recent studies have indicated • E-glass system: 19,411 cycles to failure different pipe samples was specifically
that for severe corrosion levels (on the • E-glass system: 32,848 cycles to failure conducted by a manufacturer to deter-
order of 75%) there is a need to take a • E-glass system: 140,164 cycles to failure mine the optimum design conditions
closer look at the ability of the composite • E-glass system: 165,127 cycles to failure for reinforcing the severely corroded
to provide reinforcement. The case study • Carbon system (Pipe #1): 212,888 pipe. Figure 4 shows the strains
presented herein was actually preceded by cycles to failure recorded in the four carbon-reinforced
a series of tests using E-glass materials • Carbon system (Pipe #2): 256,344 test samples. What is noted in this plot
that evaluated the number of pressure cycles to failure is that the lowest recorded mean strains
cycles to failure in reinforcing 75% corro- • Carbon system (Pipe #3): 202,903 occur in Pipe #4, which also corre-
sponds to the test sample that had the
largest number of cycles to failure.
Conclusion
Composite materials continue to play an
important role in repairing damaged
pipelines. When properly designed and
installed, they are able to restore the
integrity of damaged pipelines back to
their original integrity. The relatively
recent development and application of
composite repair standards such as ASME
PCC-2 and ISO 24817 are contributing
significantly to the proper design of the
composite repair technologies. These
standards will continue to develop as the
pipeline industry requires that composite
materials provide repair solutions for
pipeline anomalies as part of their
Figure 4. Measured strain range in 75% corroded test sample. integrity management programs. ■
Copyright, Hart Energy Publishing, 1616 S. Voss, Ste. 1000, Houston, TX 77057 USA (713)260-6400, Fax (713) 840-8585
