See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/266295985

Lifetime Prediction of Fiber Optic Cable Materials for Nuclear Power

Applications: Evaluation of Failure Mechanism, End of Life Criteria, and Test
Methodology

Article · October 2010

CITATIONS READS
5 1,544

5 authors, including:

Brian Risch Shawn Fox

Prysmian Group Marlin
52 PUBLICATIONS   196 CITATIONS    1 PUBLICATION   5 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

High Density Optical Fiber Cables with Flexible Ribbon Technology View project

Fiber Optic Buffer Tube Reliability View project

All content following this page was uploaded by Brian Risch on 15 January 2015.

The user has requested enhancement of the downloaded file.

 Lifetime Prediction of Fiber Optic Cable Materials for Nuclear Power
Applications: Evaluation of Failure Mechanism, End of Life Criteria,
and Test Methodology
Brian G. Risch*, Shawn Fox*, and Richard A. van Delden**
*Draka Telecommunications Americas **Draka Comteq Telecom B.V.
P.O. Box 39 IJzerweg 2
Claremont, NC 28610-0039 9936 BM Delfzijl
1-828-459-8435 · [email protected] The Netherlands
+31 596 642 735· [email protected]

Abstract year renewals, thus extending their licensed lifetimes to 60

years.2 With this growth a variety of cable designs will be
This study examines important variables that must be needed for the construction of new nuclear power plants,
considered while performing end of life testing for fiber optic upgrades in existing nuclear power plants, and replacement of
cable materials for use in nuclear power plants. Variables cables that have exceeded their rated lifetimes.
investigated include selection of test temperature range, end of
life criteria, identification of failure mechanism, the importance Many nuclear power plants that are in service today were designed
of considering competing failure mechanisms, and data analysis with 30 to 50 year old analog instrument and control systems.
techniques that are used to determine lifetime limitations. Many of these are in operation with initial systems still in place.
Lifetime predictions are made for a variety of fiber optic cable Old control and instrumentation systems will need to be replaced
materials including PVC (traditionally used in non-nuclear and upgraded to digital systems, even in plants with decades of
applications), Polypropylene, PBT, LSZH compounds, and service left, due to a variety of reliability, safety, and efficiency
other fiber optic cable materials. Differences in failure reasons.3 Especially in new nuclear power plant construction, many
mechanisms and lifetime prediction methodologies are of the control and instrumentation cables are being upgraded to fiber
discussed for these various materials in terms of chemical optic technology.
composition and various potential failure mechanisms.
Important considerations based on materials and end use are Design and reliability requirements for fiber optic nuclear cable
noted which differentiate test methods for fiber optic nuclear materials can be far more critical and demanding than for other fiber
cables and tight buffer materials from test methods that have optic cable applications. Specific requirements have already been
been developed for more conventional cable materials. It is established for many power and copper cables for nuclear
demonstrated that the lifetime prediction techniques which application, and over 40 years of real life experience and test data can
incorporate known failure mechanisms for these materials be leveraged for cable materials testing.4 However, the materials and
predict practical service lifetimes for many combinations of end use requirements for fiber-optic cables can vary substantially
materials that exceeds 50 years under normal service compared to more traditional nuclear cable designs. To ensure
conditions. proper reliability requirements are met for fiber-optic cable materials,
the test procedures and requirements must be optimized.
Keywords: fiber optic cable, reliability, lifetime, Arrhenius,
nuclear power plant.
2. Defining Criteria for Failure
The failure mechanisms for fiber optic cables may be very
1. Introduction different from those in copper instrumentation and equipment
cables. Copper cables may be somewhat more resistant to
Currently there are over 400 operational nuclear power plants
abuse during service and installation since optical fiber can only
which produce 14% of global electricity.1 In Europe the
withstand a few percent bending or tensile strain before being
percentage of electricity produced from nuclear power plants is
permanently damaged. Copper instrumentation and equipment
about 30% and increasing. Currently there are dozens of new
cables, however, can fail due to electrical shorts when the
nuclear reactors under construction with 150 or more being
insulation breaks down. An optical fiber, may continue to
planned for the coming years. With the demand for nuclear
function even if the cable around it totally disintegrates as long
power increasing due to concerns about greenhouse gas
as the fiber is not subjected to high tensile or bending stresses.
emissions, limitations of fossil fuel resources, and increasing
Typical failures that would be expected in a fiber optic cable
global energy demands, continued growth in the nuclear power
due to environmental exposure would be due to a situation that
industry is expected. Additionally many existing nuclear power
creates a kink in a buffer tube or cable thereby resulting in an
plants, especially in the United States are reaching the end of
excessively small fiber bend radius that either results in fiber
their 40 year licensed lifetimes. To continue to meet energy
breakage or bending induced attenuation. As a result of the
demand in the US, these older nuclear power plants will need to
differences in cable failure modes, fiber optic cable reliability
be re-commissioned with new cables and equipment or replaced
testing often investigates the impact of environmental factors on
completely resulting in additional demand for cables and
buffer tube and cable resistance to failure during dynamic or
equipment. In the US 48 reactors have already received 20-

International Wire & Cable Symposium 183 Proceedings of the 59th IWCS/IICIT
static bending since the optical fibers are so sensitive to 3. Methodology of Lifetime Prediction
kinking of the buffer tubes and cables in which they are Almost universally the simulation of long-term service for lifetime
contained. Figure 1 illustrates the type of failure that may be estimation of cables used in nuclear power applications is simulated
observed in a fiber optic component after environmental by the use of isothermal aging at elevated temperature and using
exposure. As a result of the material failure, the coiled buffer Arrhenius extrapolation.4,9,10,11,12 Typically, for qualification of a
tube has kinked, thereby placing the optical fibers under cable, the activation energy for degradation of the materials used in
elevated fiber strain. This type of failure can occur due to the cable must first be determined in order to establish the time and
static bending, handling of the component during maintenance, temperature required to simulate and in-service lifetime simulation.
or during a seismic event. The resulting high bending strain on Then the cable must then be aged according to this criteria followed
the fiber can result in fiber breaks or attenuation that could stop by some functional and performance testing.
operation of the communications system.
The Arrhenius equation describes a specific rate constant (K) as a
function of an activation energy (Ea), the gas constant (R), and
temperature (T).

K = A e-[Ea/RT] . (1)

To determine the time to equivalent aging at different temperatures,

equation (1) can be expressed as:

t1 = t2 e[(Ea/R)*(1/T1-1/T2 )], (2)

where t1 is the time of accelerated aging at temperature T1, and t2
is the time of simulated lifetime at temperature T2.
Similarly, equation 2 can be rewritten to determine a simulated
lifetime, t2, at temperature T2 that corresponds to environmental
exposure testing at temperature T1, and time, t2:

t2 = t1 e[(Ea/R)*(1/T2-1/T1 )], (3)

The Arrhenius approach to lifetime estimation can be a very
useful tool in screening and establishing reasonable materials
reliability criteria, but valid application of this approach depends
on many factors. The application of this methodology is
dependent on the short term degradation process occurring at high
temperature being identical to the long term degradation process
that occurs in service. This requires that the physical and
chemical mechanisms at play are not a strong function of
temperature.
Especially for thermoplastic materials there are upper limits on
Figure 1: Materials failure in fiber optic cables can result test temperature for which sample integrity will not be
in increased fiber strain or fiber breaks maintained. Flow can make the sample non-measurable above
some temperature that is characteristic of the material being
A common failure criterion for failure of nuclear power plant cable
tested. Performing an Arrhenius extrapolation through a first or
materials has been 50% absolute elongation, but there is no
second order thermodynamic transition such as a crystalline
standardized criterion. Other criteria such as a 50% reduction in
melting point or glass transition can make results difficult to
relative elongation, 50% change in tensile strength, 50% change in
interpret, at best, or possibly erroneous. Additionally, multiple
modulus, evidence of cracking, or even specific reduction in OIT
failure mechanisms may be present each with differing
(Oxidation Induction Time) values have been used.4,5,6 Multiple
activation energies. The Arrhenius approach is best applied to
studies covering multiple materials have shown that similar
characterize one failure mechanism at a time, most commonly
Arrhenius activation energies can be obtained with lifetime
thermo-oxidative degradation. Characterization of materials
estimations based on very different end-of life criteria.6,7,8 In some
before initiation of Arrhenius testing is, therefore, very
cases testing of samples from fiber-optic cable components to one
important in order to understand possible degradation
failure criteria such as elongation to break may be impractical or
mechanisms, whether or not multiple, competing degradation
impossible due to difficulty in tensile testing of a suitable length of
mechanisms are present, and to establish practical limits for test
tight buffer because the material must be stripped from the fiber. It
temperatures.
is therefore, important to establish multiple valid end of life criteria
that can be used in testing the variety of materials used in fiber optic In an ideal situation test data as near to the in-service
cables for nuclear power applications. Due to the sensitivity of the temperature as possible would be preferred. Temperatures
optical fibers it is important to consider additional end of life criteria within 25°C of the maximum service temperature for a material
such as resistance to material failure during bending tests or static are recommended11,12, but simulation of 40 to 50 years of in-
bending experiments. service use would lead to impractically long test times at the

International Wire & Cable Symposium 184 Proceedings of the 59th IWCS/IICIT
resulting temperatures when dealing with typical activation Tesnile tests on dogbone and tube samples were measured using
energies for most common failure mechanisms. At very high an Instron model 4468 mechanical tester. A crosshead speed of
temperatures mechanistic changes or factors such as oxygen 50mm/min was used for mechanical testing
diffusion limitations can make measured activation energies
temperature dependent. Temperature dependence of
degradation activation energies has been observed in many
materials;7,13 therefore, simply testing at the 3 or 4 required
temperatures suggested in many standards may not give an
activation energy or accurate predicted lifetime that would
apply for real world reliability. The best experimental approach
is to utilize as broad a range of test temperatures that is
practically possible and to incorporate alternative failure
criteria in order to identify possible non-linearity in Arrhenius
plots as well as the possibility of competing failure
mechanisms.

4. Experimental
OIT determinations were carried out according to ASTM D3895
at except temperatures were varied. Aluminum sample pans
were used instead of copper since copper has been shown to
catalyze some degradation reactions in polyolefins resulting in
reduced OIT values as well as variability of test results
depending on the state of oxidation of the pan’s surface.

A TA 2920 robotic dual sample DSC (Differential Scanning

Calorimeter) was used for OIT determination as well as for
determination of melting characteristics of the materials
investigated. The DSC was calibrated with indium standards.
Temperature calibration was repeatable to within 0.2°C and
specific heat calibration was repeatable to within 0.5J/g. The
cell to cell specific heat variability was approximately 2 J/g
with a temperature variability of less than 0.5°C.

Dry heat aging experiments were conducted by placing multiple

sample geometries for several different materials in ovens at
temperatures ranging from 80°C to 180°C. Not all materials
were suitable for testing at all temperatures, and not all sample
geometries were possible with each material, but several
materials could be tested and compared with equivalent test
conditions. Additionally several materials could be tested with
different sample geometries and end of life failure criteria in
order to determine the interaction of these factors on Arrhenius Figure 2: Tensile testing being performed on a jacket
activation energies. sample in order to determine elongation to break

5. Results and Discussion

Sample geometries and failure mechanisms that were tested Arrhenius data collected from multiple experiments on stabilized
include tensile testing of dogbone tensile bars to 50% residual impact polypropylene buffer tubes, dogbones, and OIT samples is
elongation, tensile testing of buffer tubes to 50% residual summarized in Figure 3. The activation energies obtained from the
elongation, mandrel wrap testing, inspection for cracks on different end-of –life criteria are equivalent if obtained from similar
coiled buffer tubes and tight buffered fibers in static coils, and temperature ranges, but a temperature dependence of activation
OIT measurements on virgin and aged materials. Mandrel energy is seen at higher temperatures. Data measured at
diameters for mandrel wrap testing and coil formation were intermediate temperatures for changes in elongation to break on
selected using 20 times the buffer diameter for tight buffers and buffer tubes and dogbones, however, correlates well to very long
loose tubes, the larger of 20 times diameter or 50mm for term studies performed on buffer tubes based on embrittlement
interconnect cables, or 150mm for general purpose, riser and (mandrel wrap testing) and spontaneous crack formation. The
plenum cables as explained in ICEA 596 Section 7.16 Low and activation energy for degradation is about 120 ± 10 KJ/mol. across
High Temperature Bend Test. all sample geometries and end of life criteria. Data obtained from
the longest term testing produced a slightly lower activation energy
and would be the most suitable for conservative lifetime estimation.

International Wire & Cable Symposium 185 Proceedings of the 59th IWCS/IICIT
The data points in the upper right hand corner represent 50 year Figure 4 illustrates the broad range of melting behaviors that can be
lifetimes at 40°C and 50°C respectively, so Arrhenius data for this observed across materials that can be utilized in nuclear power plant
material supports a service lifetime in excess of 50 years at these cables. From material to material large differences can be seen in
temperature as the Arrhenius extrapolation trend line goes to the left both crystalline melting point and crystallinity. The melting points
and above these reference points. Some EU standards today extend of the materials place practical limitations on test temperatures and
required lifetimes as far as 55°C for 60 years, a requirement which acceleration factors that are practical for Arrhenius lifetime
the results of this test data also easily exceed. prediction. The differences in crystallinity can have an influence on
how major an impact that Arrhenius extrapolation through this
1000000 thermodynamic transition will have.

100000 1000000
y = 1E-11e13292x
Time to Reach Failure Criteria (h)

R2 = 0.9916
y = 6E-14e15295x
10000 100000 R2 = 0.9846

Time to Reach Failure Criteria (h)

1000
10000
y = 9E-12e13252x
100 y = 2E-23e25573x R2 = 0.891
R2 = 0.9792 1000

10

100
1

10
0.1
0.0015 0.0017 0.0019 0.0021 0.0023 0.0025 0.0027 0.0029 0.0031
1/T(°K) 1
0.0015 0.0017 0.0019 0.0021 0.0023 0.0025 0.0027 0.0029 0.0031

Figure 3: Arrhenius Plot for Stabilized Impact 1/T(°K)

Polypropylene Tested with Varying Sample Geometries
and End of Life Criteria Figure 5: Arrhenius Plot for Hydrolytically Stabilized PBT
The four data points obtained at the highest temperatures are from Tested with Varying End of Life Criteria
OIT data collected at temperatures of 200°C to 230°C. The Arrhenius data for PBT plotted in a manner identical to the data in
activation energy obtained from OIT is nearly double that obtained Figure 3 for impact polypropylene is shown in Figure 5. The
from the physical test data. The two data points collected at the activation energy for thermo oxidative degradation for PBT
highest temperatures for change in elongation also suggest an obtained from intermediate and lower temperature testing is nearly
increase in activation energy with increasing temperature. This identical to that observed for impact polypropylene and this material
change in activation energy can be explained by the crystalline also shows an extrapolated service lifetime in excess of 50 years at
melting point of the material which is approximately 160°C. The 40°C and 50°C. A substantial difference in the data for the two
onset of melting for this material, and corresponding changes in materials, however, is evident at higher temperatures. The data for
degradation mechanism start to occur at temperatures about 150°C. PBT does not show the substantial Arrhenius slope change at test
Testing performed above the crystalline melting point of the temperatures above 150°C. The most substantial factor explaining
material shows a change in activation energy probably associated the more linear behavior of the PBT Arrhenius data at high test
with changes in degradation mechanism. temperatures is the difference in crystalline melting points between
the two materials as illustrated in Figure 6. PBT shows an onset of
melting at just over 200°C while the onset of melting for
1.4
A
1.2 polypropylene is about 150°C.
1.0 B
0.0
0.8

0.6 -0.2
A 153.72°C

C
Heat Flow (W/g)

0.4
-0.4

0.2

0.0
D -0.6
Heat Flow (W/g)

-0.2 -0.8

-0.4 -1.0

B
205.33°C
-0.6 E
-1.2

F
-0.8
-1.4
-1.0

-1.2 -1.6
50 100 150 200 250 300
Exo Up Temperature (°C) Universal V2.5H TA Instruments
-1.8

Figure 4: DSC melting behavior for a variety of materials -2.0

50 100 150 200 250 300
Exo Up Temperature (°C) Universal V2.5H TA Instruments
that can be used in nuclear power plant applications:
A: ETFE Copolymer; B: PBT; C: Flexible Polyester Figure 6: Onset of Melting for Polypropylene (A) and
Copolymer; D: Stabilized PP-impact Copolymer; E: HFFR PBT (B)
Compound 1; F: HFFR Compound 2.

International Wire & Cable Symposium 186 Proceedings of the 59th IWCS/IICIT
A substantial pitfall of performing an Arrhenius lifetime 1000000
extrapolation from too high a temperature or across a 12256x
y = 2E-10e
thermodynamic transition is overestimation of lifetime. Since 2
R = 0.9723

Arrhenius plots are on logarithmic scales for lifetime, the magnitude

Time to Jacket Failure (hours)

100000
of an error in extrapolation can be extreme. As an example, the y = 7E-11e
12001x

extrapolated lifetime of stabilized impact polypropylene from OIT R2 = 0.9786

data and data obtained from tensile testing or coil cracking tests at A
lower temperatures can differ by a factor of 50 000. Although both 10000 B
results extrapolate to passing results, the extrapolation obtained C
from the highest temperature data to in service temperatures is D
obviously in error. Use of such high temperature data on less 1000
stabilized materials could possibly result in selection of a material
with unsuitable long term stability.
100
0.00220 0.00240 0.00260 0.00280 0.00300 0.00320
Table 1. Activation energies determined for 1/T (°K)

polypropylene and PBT buffer tubes using different end

of life criteria Figure 7: Arrhenius Plot for HFFR and PVC Jackting
Compounds with Varying End of Life Criteria and Sample
End of life Ea Temp. Range
Material Geometry:
Criterion (KJ/mol)
Cracks in Static
i-PP 128 110°C-160°C A: HHFR Compound tested to 50% ultimate elongation
Coils
Cracks in with large dogbones prepared in the lab.
i-PP Mandrel Wrap 110 100°C-140°C
Test
B: HHFR Compound tested to 50% ultimate elongation
<50% Ultimate with small dogbones cut from cable jacket.
i-PP 109 100°C-140°C
Elongation
<50% Ultimate C: HHFR Compound mandrel wrap testing.
i-PP 122 100°C-150°C
Elongation D: PVC Compound mandrel wrap testing.
<50% Ultimate
i-PP 142 100°C-160°C
Elongation
i-PP OIT 215 200°C-240°C Figure 7 shows the Arrhenius plots generated from testing of
Cracks in HFFR and PVC compound samples with different sample
PBT Mandrel Wrap 110 100°C-140°C geometries and end of life criteria. The data shows essentially
Test identical Arrhenius activation energies for both materials
PBT
<50% Ultimate
127 100°C-180°C
although the HFFR compound tended to show longer lifetimes
Elongation before failure. The HFFR compound showed a slight variation
<50% Ultimate in time to failure as a function of sample geometry. The large
PBT 108 100°C-145°C
Elongation dogbones showed the longest lifetime while the small samples
for the mandrel wrap testing showed the shortest lifetime. The
Table 1 summarizes activation energies obtained for stabilized impact of sample size on time to failure may be explained by
impact polypropylene (i-PP) and PBT obtained with different end of oxygen diffusion differences as a function of sample size, but
life criteria and data from different temperature ranges. The this phenomenon did not impact the resulting Arrhenius
activation energies determined in this testing for the two materials is activation energy determined from the testing. The HFFR
nearly identical. The differences in activation energies from test-to- compound appeared to show the greatest linearity of test data
test are almost wholly due to differences in test temperatures. Due for Arrhenius testing of the materials investigated. This better
to availability of larger lengths of test sample and ease of testing, linearity to the Arrhenius equation is suspected to be partially
mandrel wrap tests and static bending tests can more easily be due to the lower crystallinity of the HFFR compounds relative
conducted for longer periods of time thereby yielding experimental to the other materials investigated. The lower volume fraction
results that can be more accurately extrapolated to in-service of crystalline material in the HFFR compounds results in a
temperatures. Mandrel wrap tests typically produce failures before smaller impact of any thermodynamic changes associated with
static bend testing, so in this dataset the mandrel wrap tests have the the crystalline phase.
lowest temperature data, and thus the lowest activation energies for
failure since failures in static bending were not observed at the
lowest test temperatures during the multi-year timeframe of the
study.

International Wire & Cable Symposium 187 Proceedings of the 59th IWCS/IICIT
Table 2. Activation energies for additional cable materials anticipated, the experimenter can run out of samples before
determined using different end of life criteria and test appropriate reduction in elongation is observed. Another
temperatures. consequence of the rapid reduction in properties often observed
End of life Ea Temp. Range after a long induction time is that testing becomes less sensitive to
Material the end of life criterion.
Criterion (KJ/mol)
14 Thermal
ETFE
Analysis
227 220°C-280°C In all of the testing used to determine activation energies in Tables 1
<50% & 2 materials were degraded through thermo-oxidative degradation.
ETFE15 Ultimate 88 70°C-140°C Thermo-oxidative degradation is not the only important degradation
Elongation mechanism for fiber optic cable materials, and other important
Thermal degradation mechanisms need to be considered and studied in order
FEP16 223 225°C-550°C
Analysis to ensure long term cable reliability. Other important mechanisms
<50% that can produce failures in cable materials are interactions with
FEP Ultimate 100 80°C-160°C solvents or other environmental factors, slow crack growth, or other
Elongation
chemical degradation mechanisms such as hydrolysis. For a nuclear
Cracks in
EVA Elastomer
Mandrel 104 130°C-160°C
cable designed for use inside containment structures, combined
Compound interactions of radiation, thermal decomposition, and other factors
Wrap Test
Cracks in such as exposure to high temperatures and possibly steam in a Loss
XLPE Of Coolant Accident (LOCA) event must also be considered.
Mandrel 110 110°C-130°C
Compound
Wrap Test
Cracks in PBT and other polyester copolymers are also susceptible to
PVC Compound Mandrel 106 100°C-130°C hydrolytic decomposition which can result in much more rapid
Wrap Test material failures especially in high humidity environments.
Cracks in Hydrolytic decomposition of fiber optic cable materials has been the
HFFR
Compound
Mandrel 106 120°C-160°C focus of many studies over the years.17,18,19 Materials such as
Wrap Test polyesters where other important degradation mechanisms are
<50% known to be present should also undergo testing to that results in a
Ultimate lifetime estimation based on other known failure mechanisms. A
HFFR
Elongation 104 140°C-160°C
Compound
(Large
similar methodology using Arrhenius lifetime estimation can also be
Dogbone) applied to chemical degradation mechanisms such as hydrolysis of
<50% polyesters, but partial pressure of water vapor must also be
Ultimate considered as an additional variable in the Arrhenius testing.
HFFR
Elongation 102 130°C-160°C
Compound Many polymeric compounds can fail due to slow crack growth.
(Large
Dogbone) Time to failure in slow crack growth experiments has also been
<50% shown to follow an Arrhenius temperature dependence.20 When
Ultimate using an Arrhenius dependence to model failures due to slow crack
HFFR
Compound
Elongation 101 125°C-145°C growth in cable materials, factors such as cable geometry, bend
(Small radius, and the temperature dependent modulus must also be
Dogbone) considered since slow crack growth also has a power law
dependence on the stress level at the crack tip. Additionally
Table 2 summarizes Arrhenius activation energy from the PVC, environmental factors such as the impact of solvents or
HFFR compound, and a variety of additional different materials. environmental stress cracking agents must also be considered for
The trends in the data are amazingly similar. Nearly equivalent end-of-life analysis for slow crack failures.21 It is recommended
activation energies are seen across almost all of the materials when that coil or mandrel wrap testing at temperatures at the lower range
tested at equivalent temperatures regardless of end of life criterion. of the experimental temperature range be done in addition to the
Additionally, all materials seem to show a dependence of activation traditional elongation to break testing to ensure that the presence of
energy on test temperature at high temperatures. All activation slow crack growth can be determined as a possible failure
energies obtained on materials from test data in the melt state mechanism as failures due to cracking in bending are probably the
through thermal analysis yielded activation energies in excess of most likely mode of failure for cables due to material degradation
200KJ/mol. after installation.

Using changes of tensile properties as an end of life criterion can

result in incomplete test data or more experimental uncertainty due 6. Application of Results
to the nature of the typical elongation vs. time behavior for many Environmental qualification of cables for nuclear power plants is
cable materials. Some materials will show a large induction time, generally achieved through a process such as that outlined
especially at low temperatures followed by a rapid reduction in below:10-12,
properties that characteristically occurs in a logarithmic function of 1.) Initial baseline testing is performed on representative cables
time. For long term tests at lower test temperatures, this means that within the family to be qualified.
the experimenter either needs to have some idea when to expect
failures, or an extremely large number of sample specimens need to 2.) Complete functional testing of the test cable is completed
be aged, sampled, and tested with regular frequency or a precise before aging.
determination of time to failure will not be obtained. Additionally,
if the induction time for reduction of elongation is longer than

International Wire & Cable Symposium 188 Proceedings of the 59th IWCS/IICIT
3.) The test cable is subjected to environmental testing with a Table 3. Calculated environmental exposure times (t1) at
thermal soak time and temperature determined to be equivalent to various test temperatures (T1) required for qualification
a specified operational lifespan at a given temperature based on for 50 year rated lifetimes at service temperatures (T2)
the Arrhenius activation energies of the materials included within Ea
the cable. If other important degradation mechanisms (such as T1 (°C) T2 (°C) (KJ/mol)
t1 (Days)
hydrolysis) are present, other environmental factors such as
100 30 110 5.0
humidity and or exposure may need to be included in the cable
aging. 100 40 110 20.3
4.) After the thermal aging of a cable at the time temperature 100 50 110 75.3
equivalent aging condition, the cable is subjected to additional 80 30 110 37.6
testing and functional characterization based on end of life design
80 40 110 151.7
based requirements.
80 50 110 561.6
This type of test sequence can be used to establish a qualified
cable lifetime at a specified service temperature. Generally it is
impractical to perform testing adequate to fully certify a cable to a Table 4. Impact of a 10% and 100% difference in
40 to 60 year lifespan without applying a large acceleration factor activation energy in equivalent environmental 50 year,
to the environmental testing in step 3 above. There are several 50°C equivalent exposure times for 80°C and 100°C.
approaches that may be used to reduce the risk associated with the
uncertainty of the lifetime extrapolation. These approaches can T1 (°C) Ea
(KJ/mol)
t1 % Change
be used independently or in combination.
The cable can be tested after environmental testing under low 100 100 124.0 39.3%
acceleration factor conditions to a shorter initial service lifetime, 100 110 75.3 0.0%
and the service lifetime can be increased after cable testing is
100 120 45.7 -64.8%
conducted for longer timeframes.
80 100 770.6 27.1%
The cable can be tested to its final rated service life under a high
acceleration factor, and follow-up testing can be completed at 80 110 561.6 0.0%
longer aging times and lower acceleration to verify that the 80 120 409.2 -37.2%
lifetime extrapolation is valid.
80 220 17.3 -3150.0%
If proper testing and analytical resources are available,
comprehensive material characterization and reliability studies
can be conducted on cable materials and components thereby pre- Table 3 illustrates the calculated environmental exposure time for
screening materials so that only the most reliable materials are 50 year environmental qualification at the given temperatures for
selected for the application. This detailed materials the typical thermo-oxidative degradation activation energy
characterization data, which includes the temperature dependent determined in this study (110KJ/mol). Table 4 illustrates the
Arrhenius activation energy, provides a much more precise variability in determination of equivalent environmental exposure
lifetime extrapolation. times, t1, associated with a 10% variation in measured activation
energy. From this data we can clearly see that relatively small
If there is good understanding of Arrhenius activation energy and changes in activation energies can result in large differences in
degradation mechanisms present an additional safety factor can be equivalent environmental exposure times. In fact, using an
applied to the lifetime extrapolation used for environmental activation energy characteristic of typical values obtained from OIT
testing while performing cable testing with a lager acceleration data results in a 3150% difference in calculated environmental
factor. exposure times. Also going to a higher acceleration factor in the
environmental exposure testing (100°C instead of 80°C) results in
An example of how Arrhenius activation energy could be used to
about a 50% greater impact of activation energy on predicted
establish environmental exposure requirements for qualification
equivalent aging times. The data in Table 4 illustrates the
for a service lifetime at a rated temperature is given below:
importance of having longer term testing at lower temperatures in
For a cable with a HFFR jacket and i-PP buffer tubes the order to limit uncertainties due to extrapolation. The data in table 4
activation energies at our lowest test temperatures are about would also suggest that safety factors of 50% or more may be
110KJ/mol. Equation 2 can be used to calculate Environmental desired in environmental exposure testing. The timeframes in Table
exposure times at various temperatures with and without safety 4, however, illustrate why there is a strong desire to get accurate
factors. Safety factors can be set as a simple function of the reliability data in using the highest practical environmental exposure
calculated thermal soak time or calculated by adjusting the conditions.
activation energy either by lowering it by some fixed factor or by
Using the Arrhenius activation energies determined from Arrhenius
adjusting it to a lower level based on confidence intervals set from
testing and Equation 3, cables or cable components can be tested in
the data.
order to determine qualified lifetimes. Examples of actual test
results on cable materials are provided in Table 6. Details on cable
construction are provided in Table 5.

International Wire & Cable Symposium 189 Proceedings of the 59th IWCS/IICIT
Table 5. Results of 90 day lifetime qualification testing for was shown to have little or no impact on the Arrhenius activation
several cable designs at specified temperatures. energy as long as test temperature ranges for Arrhenius testing were
kept equivalent.
Cable Cable Buffer When considering the potential failure modes of both fiber optic
Jacket Material
Design Material cables and materials, it is often more practical and in many cases
2f, 900μm preferred to perform mandrel wrap and/or static bend type testing in
A PVDF PVDF
Tight Buffer order to determine end-of-life in accelerated aging studies. It has
4f, 900μm Flexible been shown that equivalent Arrhenius activation energies are
HFFR obtained when testing within equivalent temperature ranges using
B Tight Buffer Polyester
Compound either traditional elongation to break testing or bending tests.
Breakout Copolymer
4f, Tight Chlorinated PE Flexible Combined with the equivalency in determined activation energy, the
C Buffer Jacketing Polyester additional ease of testing for some sample types, the ability to test
Breakout Compound Copolymer cable and buffer samples as they are produced, and the ability to
8f, 600μm identify slow crack growth failures are all advantages of
HFFR Urethane incorporating bending tests into fiber optic lifetime prediction
D Tight Buffer
Duct Cable
Compound Acrylate studies.
60f I/O HFFR
E
Loose Tube Compound
i-PP
Finally, it should be noted that lifetime prediction based on
Arrhenius lifetime extrapolation combined with cable
environmental exposure testing is only one part of a comprehensive
Table 6. Results of 50°C environmental qualification cable and system reliability approach. The reliability of the system
testing for materials in several cable designs at specified is dependent on each fiber, cable, and component. Optical fibers,
cable designs, and materials must be properly selected to meet the
temperatures (no failures observed).
reliability requirements for normal service conditions as well as
Qualified extreme situations that may be encountered during an accident or
T1 (°C) Cable t1 Lifetime natural disaster even at the end of the rated lifetime.
(Years)
110 Cable A 90 Days >84 years
110 Cable B 90 Days >150 years 8. Acknowledgments
110 Cable C 90 days >150 years I would like to acknowledge Boyce Lookadoo for his help in
materials testing and monitoring samples during testing. I would
110 Cable D 90 days >150 years also like to thank Jan Pirrong for editorial input.
100 Cable E 90 days >60 years

9. References
The different test conditions shown in table 6 are a result of thermal 1
International Atomic Energy Agency Nuclear Technology
stability of the materials and the requirements set forth in other Review; International Atomic Energy Agency, Vienna (2009).
design requirements for the cable. In many cases cable materials 2
were tested at higher temperatures to longer equivalent lifetimes, but A. Mc Donald, “Nuclear Power Global Status”, IAEA Bulletin,
the lowest temperature test data to certify a lifetime requirement is 49 (2), June 2010.
3
always preferred since there is less error in the lifetime U. Chandra, et. al., Managing modernization of nuclear power
extrapolation. In most situations cable testing will be done to fulfill plant instrumentation and control systems, IAEA-TECDOC-
some lifetime requirement rather than to failure since time to failure 1389; International Atomic Energy Agency, Vienna (2004).
for typical test temperatures and properly stabilized materials is 4
S.K. Aggarwal, et. al., Assessment and management of ageing of
extremely long. major nuclear power plant components important to safety: In-
containment instrumentation and control cables (Volumes 1 &2),
IAEA-TECDOC-1188; International Atomic Energy Agency,
7. Conclusions Vienna (2000).
A methodology for lifetime prediction of nuclear fiber optic cable 5
materials based on Arrhenius modeling has been presented along B. Bartonicek, V. Hnat, and V. Placek, “Aging Monitoring of
with lifetime modeling data for several cable materials that supports Plastics In Nuclear Power Plants By DSC”, Journal of Thermal
cable lifetimes well in excess of the normal 50 year requirements for Analysis and Calorimetry,64 (2001) 571-576.
nuclear power applications. Data on many cable materials showed a 6
V. Placek, “Asessment of Parameters for Simulation of Thermal
temperature dependence of Arrhenius activation energy especially Ageing of Materials in Nuclear Power Plants Using DSC”,
for testing done at extremely high temperatures. It is, therefore, Journal of Thermal Analysis and Calorimetry,80 (2005) 525-528.
important to conduct environmental exposure testing of cables at 7
D.A. Simoff, C.A. Pryde, and C. Gieniewski, “Long Term
temperatures as close as possible to the cable in-service temperature
Hydrolytic Aging Study of a Block Copoly(ether ester) Undersea
while still producing test results within a reasonable timeframe.
Cable Buffer”, 47th International Wire and Cable
While the Arrhenius activation energy showed high sensitivity to
Proceedings,(1998) 507-511.
test temperature in some cases, the end of life criterion for testing

International Wire & Cable Symposium 190 Proceedings of the 59th IWCS/IICIT
 8
K.T. Gillen, R.L. Clough, and J. Wise, “Prediction of Elastomer Brian G. Risch
Lifetimes from Accelerated Thermal-Aging Experiments”, in Draka Communications
Polymer durability;Degradation Stabilization, and Lifetime
Prediction, American Chemical Society, Washington, D.C. Fiber Optic
(1996) 557-575. CableDevelopment
9
P. Holzmann and G. Slitter, Eds., Nuclear Power Plant 2512 Penny Rd.
Equipment Qualification Reference Manual, EPRI, TR 100516 P.O. Box 39
(1992).
Claremont, NC 28610
10
IEE std. 323-2003: IEEE Standard for Qualifying Class 1E
Equipment for Nuclear Power Generating Stations. Brian G. Risch is the Materials Technology Manager at Draka
11
IEC 60216 : Guide for the Determination of Thermal Communications in Claremont, NC. He holds a B.A. degree in
Endurance Properties of Electrical Insulating Materials. Ageing physics from Carleton College and a Ph.D. in Materials Science
Procedures and Evaluation of Test Results. and Engineering from Virginia Polytechnic Institute and State
12
University. Brian has studied structure property relationships in
IEEE P1682 : Standard for Qualifying Fiber Optic Cables, polymeric materials and materials reliability for the last 20 years.
Connections, and Optical Fiber Splices for Use in Safety Systems Prior to working for Draka, Brian worked for Alcatel for 7 years
of Nuclear Power Generating Stations. at Alcatel’s Optical Fiber Cable R&D center specializing in cable
13
S. Ding, M.T.K. Ling, A, Khare and L. Woo, “Activation materials and fundamental material reliability and then for 6 years
Energies of Polymer Degradation”, in Weathering of Plastics: at Hewlett Packard as a failure analysis engineer.
testing to Mirror Real Life Performance”, Society of Plastics
Engineers, Norwich, NY, (1999) 169-184. Shawn Fox
14
J.P Elders, Thermal oxidative degradation of ethylene Draka Communications
tetrafluoroethylene copolymer systems,Ph.D. Dissertation, Fiber Optic
Washington University (2002). CableDevelopment
15
K.T. Gillen and R.L. Clough, "Predictive Aging Results for 2512 Penny Rd.
Cable Materials in Nuclear Power Plants," Sandia National
P.O. Box 39
Laboratories, SAND90-2009, November 1990.
16 Claremont, NC 28610
A. V. Prasad and R. P. Singh, “Thermal decomposition kinetics
of a commercial fluoropolymer”, Materials Research Bulletin, 30,
11, (1995), 1407-1412. Shawn Fox is a Senior Materials Engineer at Draka
17 Communications in Claremont, NC. He holds a B.S. degree in
D. Parris and B. Warner, “Testing of Fiber Optic Materials Chemistry from Appalachian State University. Shawn has over
After Accelerated Heat and Humidity Aging”, International Wire 10 years experience working with thermoplastic materials and
and Cable Symposium, 39th IWCS Proceedings, (1990) 237-42. UV-curable coatings. Prior to working at Draka, Shawn worked
18
O. Gebizlioglu and I. Plitz, “Monitoring Accelerated Aging of as Plant Chemist at Marlin Chemical Company and served 4 years
Polyester Buffer Tubes in Fiber Optic Cable”, International Wire in the United States Air Force as AWACS Crew Chief.
and Cable Symposium, 42nd IWCS Proceedings (1993) 509-515.
Richard A. VAN
19
B.G. Risch and T. Auvray, “Effects of Moisture Content, CEC, DELDEN
and Process Conditions on Mechanical Properties and Long-Term Draka Comteq Telecom
Reliability of PBT Fiber-Optic Buffer Tubes”, 58th Annual B.V.
ANTEC Proceedings, May 2000, 302-306.
20 IJzerweg 2
M. Parsons, E. V. Stepanov, A. Hiltner and E. Baer,
“Correlation of fatigue and creep slow crack growth in a medium 9936 BM Delfzijl
density polyethylene pipe material”, J.Of Materials Science, 35, The Netherlands
(2000) 2659-2674.
21
J. Yu, Y. Hu , B. Risch , A. Hiltner, E. Baer, “Effect of an
Environmental Stress Cracking Agent on the Slow Crack Richard A. van Delden is the Cable Materials Manager for Draka
Propagation of Polyethylene”, 61st Annual ANTEC Proceedings, Communications, Europe. He holds a bachelor as well as a PhD
(2003), 2947-2951. degree in Organic Chemistry from the University of Groningen,
The Netherlands. Richard has studied synthetic nanomaterials and
their function in solution, liquid crystal and polymer environment
at an academic level. After working for two years as a
postdoctoral researcher he joined Draka in 2005. He started on the
PVC and HFFR compound development for Draka Energy &
Infrastructure Division and joined the Draka Communications
division in his present function in 2008.

International Wire & Cable Symposium 191 Proceedings of the 59th IWCS/IICIT

View publication stats