You are on page 1of 14

Higher Energy Piping Reexaminations

Fewer Locations- Higher Confidence

Marvin J Cohn
Aptech Engineering Services, Inc.

Abstract used with metallurgical replication to justify


Many utilities select high energy piping (HEP) reexamination intervals beyond 5 years.
reexamination locations by considering the
highest sustained longitudinal stress location, Introduction
highest thermal expansion stress location, Many fossil power plants throughout the
terminal point locations, and fitting world are being considered for operation
weldments. Eighteen HEP systems were beyond their original design lives of 30 to 40
evaluated for highest priority girth weld years. To meet these long-term objectives,
examination locations. Using the above nondestructive methodologies have been
approach, the number of conventional used extensively in the past decade to
reexamination sites varied from 6 to 20 perform condition assessments, to estimate
locations per piping system. the remaining economic lives, and to provide
reexamination intervals of critical power plant
As an alternative, a high energy piping life components.
consumption (HEPLC) methodology is used
to predict maximum material damage Since it is prohibitively expensive to examine
locations. The methodology is customized to all of the piping system weldments during
prioritize expected creep damage (ECD) each inspection, the most critical sites are
considering applicable affects, such as field commonly selected using criteria such as
piping displacements, time-dependent engineering judgement and locations of
operating stresses, and multiaxial piping historical failures. Selection of examination
stresses. locations based on engineering judgment may
include high stress locations from the results of
This paper includes a case history of 18 main conventional stress analyses and welds near
steam or reheat piping systems in 12 fossil supports that are not functioning as originally
power plant units. It compares reexamination designed. Selection of examination locations
locations selected by the HEPLC based on historical experience may include
methodology to locations selected by terminal welds, fitting welds (e.g., valves,
conventional criteria. reducers, tees, wyes, and laterals), and sites of
previous nondestructive examination (NDE)
The HEPLC risk-based inspection indications unique to the unit.
methodology may be considered as a rational
approach to reduce conventionally selected High stress locations are commonly selected
reexamination locations by more than 50% in by a flexibility stress analysis developed in
most cases. This approach also revealed accordance with the American Society of
some locations of high ECD which were not Mechanical Engineers (ASME) B31.1 Power
identified by the conventional site selection Piping Code (ASME, 1995); both the
approach. The HEPLC approach can be sustained longitudinal stress, SL, and thermal

Aptech Engineering Services, Inc. TP114


1
expansion stress range, SE, results divided by HEPLC methodology considers the operating
their respective allowables (Sh and SA) are loads at the beginning of unit operation.
usually sorted and ranked in the flexibility Life Consumption
analysis programs to identify the maximum The specified scope for the Code is to
stress ratio locations. Consequently, flexibility “prescribe minimum requirements for design,
analyses programs to identify the maximum materials, fabrication, erection, test, and
stress ratio locations. Consequently, flexibility inspection of power and auxiliary service
analysis software provide a convenient piping systems.” There is no life evaluation
approach to rank high stress locations based associated with a Code analysis. Evaluation
on ASME B31.1 criteria. of the life consumption of an HEP system is
beyond the scope of the Code and should be
Why Is Effective Life Consumption performed by enhanced analysis. The
Approach More Effective? HEPLC approach uses a cumulative life
This paper suggests that the conventional fraction of the appropriate steps.
approach to select examination sites may not
be very reliable because some of the most Material Property Degradation
critical creep damage locations may not be Most main steam and hot reheat piping
identified. Furthermore, many of the system materials are subject to creep during
conventionally selected sites do not have to normal operation. Two distinct creep damage
be reexamined perpetually. The use of processes occur in low alloy ferritic steels.
HEPLC methodology provides a rational Microstructural degradation results in gradual
approach to select high ECD locations for reduction of creep strength and accelerated
inspections. It is cost effective because strain cumulation. In addition, creep
unnecessary locations selected by the cavitation primarily occurs along the grain
conventional approach can be screened out boundaries. As the creep damage proceeds,
and some high ECD locations are identified these damage processes reduce the ultimate
which are not captured by the conventional tensile strength, yield strength, and creep
criteria. The HEPLC methodology was rupture strength of material.
developed to evaluate the multifaceted life
consumption problem by improving on the An example of the time-dependent property
conventional methodology as described in the degradation is presented in Table 1-14.6D of
eight topics below. Code Cases N-47 (ASME, 1994). The
minimum stress to rupture value of
Evaluation Of Operating Loads 2-1/4Cr-1Mo at 1000°F drops from 9.4 ksi
As discussed above, the ASME B31.1 Code (65MPa) at 100,000 hours to 7.9 ksi (55MPa)
approach considers a ranking of SL/Sh and at 300,000 hours. These time-dependent
SE/SA. It does not consider any combination material property degradation effects are not
of the hoop stress with the axial stress. considered in the Code. Empirical curve fit
Although the operating load case can be rupture equations are used as part of the
evaluated in conventional software, the HEPLC approach.
ASME B31.1 high stress criteria does not
consider the results of the online piping Weldment Properties
system loads. There have been many For the purpose of this paper, a weldment is
instances where the combination of the defined as a joint consisting of unaffected
sustained and thermal expansion loads result base metal, the base metal heat affected zone
in a significantly different ranking than either (HAZ), and weld metal. The Code stress
the SL/Sh and SE/SA tables. Consequently, the allowable tables do not distinguish between

Aptech Engineering Services, Inc. TP114


2
base metal and weldment properties by limit stops to simulate observed
(provided that the weldment has been 100% displacements. Limit stops should be used
radiographed prior to service). If the base only when necessary and with careful
metal and weld metal have similar material consideration of proper user input values.
composition, Table 1-14.10D of Code There may be instances where a piping
Case N47 (ASME, 1994) illustrates that the system may have a limit stop or a pipe roll
weldment creep properties degrade faster support, which may be active offline and may
than the base metal creep properties. Since not be loaded online or vice versa. The Code
this is primarily an in-service issue, the does not distinguish between the offline and
evaluation of this effect is outside of the scope online deadweight cases, which may be
of the Code. The HEPLC methodology nonconservative. Alternatively, the HEPLC
includes a weldment factor which results in an methodology considers creep relaxation of the
increased life consumption due to the HAZ online operating loads, which is more realistic
creep rupture strength, weld profile stress for the prediction of gradually changing
concentrations, material property mismatch stresses and cumulative material damage.
stress concentrations, and the volume of field
“tested” material. The simulation approach of thermal
expansion displacements does not account for
Multiaxial Stress material strain due to system self-springing,
The Code considers the hoop stress separate shakedown, and creep relaxation of thermal
from the longitudinal stress. Hoop stress due loads. These displacement effects are most
to pressure is considered in the Code pipe significant during the few years after 1) initial
wall thickness calculation. The Code stress operation of the unit, 2) hanger position
calculation for sustained loads specifies a indicator adjustments, and 3) hanger load
stress based on the axial stress due to pressure adjustments. Once the piping system has
and the combined moments due to torsion experienced several years of operation and a
and bending. The Code stress calculation for few outages, additional displacement changes
thermal expansion stress range specifies a due to any of the above three effects are
stress based on only the combined moments reduced to a negligible level. Beyond 5 years
due to torsion and bending. Alternatively, the of piping system operation, it is believed that
HEPLC approach determines a creep the simulation of thermal expansion
initiation stress based on the von Mises stress, displacement range provides a very good
including the hoop stress component (Cohn, estimate of the actual piping system thermal
1990). expansion stress range.

Simulation of Field Conditions Fewer NDE Locations


The HEPLC methodology (Cohn, 1989; Microstructural creep damage is first evident
Cohn, 1990; Cohn 1994; Cohn, 1995) is in HAZ. Even in the HAZ, significant creep
based on the simulation of field conditions. damage is not expected to occur until the
This includes the effects of anomalies, such as material has reached 40 to 80% of the life
pipe interferences and topped out and exhaustion. It is common to have many girth
bottomed out hangers. The as-found stress welds at relatively low stresses and a few
analysis (Cohn, 1989) may be much more welds at significantly higher stresses. In the
accurate than the original as-designed stress following study of 18 piping systems, it was
analysis. As discussed by Cohn and May typically found that locations ranked beyond
(1996), it is not recommended that four have considerably less stress and less
improperly functioning hangers be modeled

Aptech Engineering Services, Inc. TP114


3
microstructural damage due to service crack depth for the indication in this weld is
conditions. 0.50 inch (13mm), the indication can be
reexamined in 24 months (as illustrated in the
If a weld is expected to have less than 40% t=24 months curve), with a safety factor of
life exhaustion, metallurgical replication will 3.0 on stress. If the maximum crack depth for
probably reveal creep damage no greater than the indication is 0.15 inch (3.8mm), it is
isolated cavities. Reexamination of locations justifiable to reexamine the indication within
with low ECDs is not productive because the 60 months.
results can be predicted apriori as no
reportable indications and no evidence of Higher stressed (e.g., ASTM A 335,
significant creep damage. The HEPLC Grade P22 material at 7.5 ksi (52MPa)) girth
methodology provides a rational approach to welds with no reportable UT indications may
justify that reexamination (e.g., by ultrasonic not provide reexamination intervals beyond 2
examination (UT), magnetic particle or 3 years. For example, if the above girth
examination (MT), and metallurgical weld was subject to a stress of 7.5 ksi instead
replication) of locations with low ECDs is not of 5.7 ksi, a crack depth sized at 0.06 inch has
cost effective. a reexamination interval of only 2 years, as
illustrated in the TFD presented in Figure 2.
Quantitative Justifiable Reexamination
Intervals Since the axial stress varies considerably
The HEPLC methodology provides a rational around the girth weld due to the bending
process to justify reexamination intervals until stress, metallurgical replication results are very
the next scheduled outage (18 to 60 months) sensitive to the selected circumferential
based on the results of microstructural location. It is possible that creep damage no
examinations at appropriate locations. It is greater than isolated cavities may exist 45
probable that the UT and MT would not degrees from a creep crack. Therefore, it is
reveal indications as small as 1/16 inch important to select metallurgical replica
(1.6mm) or a microcrack. For lower stressed locations based on damage from the
(e.g., ASTM A 335, Grade P22 material at cumulative creep relaxed stress distribution.
5.7 ksi (39MPa)) girth welds, high quality UT Given the metallurgical results for the highest
resulting in no reportable indications may ECD locations (including the correct
provide as much as a 5-year reexamination circumferential location and evaluation of all
interval. of the HAZ in the replica for potential
microstructural creep damage), the
This is illustrated in the Tolerable FlawSM reexamination interval for isolated cavities
diagram (TFD) presented in Figure 1. In a may be as much as 10 years. Without the
recent study, an ASTM A 335, Grade P22 proper metallurgical evaluation of the highest
main stream pipe had a 508mm (20-inch) ECD locations, the justification for the next
diameter and was 78mm (3.08 inches) thick, reexamination interval has much less
there was a significant NDE indication in a certainty.
specific girth weld subject to 5.7 ksi (39MPa)
bending stress. A fracture mechanics analysis Case History Studies
predicted that crack propagation instability Eighteen main steam or hot reheat piping
effectively occurs at the time a hypothetical systems were evaluated in twelve fossil power
crack propagates through the wall (up to 13- plants units using the HEPLC approach.
inch circumferential crack length), as Isometric piping system models were
illustrated in the t=0 curve. If the maximum developed for the stress analyses based on the

Aptech Engineering Services, Inc. TP114


4
available design documents. The baseline necessary to account for field geometry
evaluations (Phase 1) were the as-designed anomalies, hanger load modifications, hanger
stress analysis models. An example main anomalies, and piping interferences. The
steam piping system isometric is shown in process of simulating observed thermal
Figure 3. It includes all major piping from the displacements in the as-found stress analysis is
superheat outlet header to the six turbine lead discussed by Cohn (1989). The most
connections. This isometric model is common modifications for all of the as-found
sufficiently represented with about 100 nodes. piping system models were adjusting the
An as-designed stress analysis is used to thermal anchor displacements at the header
evaluate design compliance with the and turbine connections. However, there
ASME B31.1 Power Piping Code sustained were many instances where a hanger did not
load and thermal expansion range stress displace as expected in the as-designed stress
allowables. Dead weight and thermal analysis and appropriate model attributes
expansion hanger loads from the as-designed were adjusted to closely match the observed
stress analysis are commonly used to size piping displacements. As discussed by Cohn
hangers. Thermal expansion displacements and May (1996), limit stops were used as little
from the as-designed stress analysis are as necessary to minimize nonlinear stress
commonly used to predict pipe movements at analysis errors.
hanger locations to size hanger travel
requirements. An example of observed versus predicted as-
designed stress analysis displacements is
In each piping system evaluation, hot and illustrated in Figure 4. The piping
cold walkdowns (Phase 2) were performed to displacement profile diagram for this as-
characterize and evaluate the field conditions. designed model shows that the expected
These walkdowns included the accessible displacement at Location 7 is 4.2 inches
piping supports from the outlet headers to the (107mm) more than the actual displacement.
turbine connections. Observations were The simulated thermal displacement for the
documented regarding pipe sagging, major same piping system, without the use of limit
insulation damage, structural steel damage, stops, is illustrated in Figure 5. The maximum
piping interferences, deformed or broken difference between the observed and the as-
support components, hydraulic pipe restraint found thermal displacement values is about
fluid levels, and hanger travel anomalies. As 0.6 inch (15mm). In this case, the as-found
part of the walkdowns, thermal displacements, piping analysis (including the SL and SE
were measured or estimated at selected values) based on similar thermal
locations, such as at the supports, hangers, displacements are judged to be more accurate
guide clearances, penthouse penetrations, and for current conditions than the values from the
near the turbine connections. In many corresponding as-designed stress analysis.
instances, hanger position indicator readings Resulting loads from the as-found stress
were documented with photographs. analyses were evaluated to estimate
remaining lives (at specific nodal locations),
The Phase 3 portion of each project was the recommend reexamination intervals and
as-found stress analysis. These analyses were select NDE locations.
performed to evaluate the revised stresses
from significant piping system thermal An HEP life management program should
displacement anomalies as compared to the include a one time 100% examination of all
expected as-designed values. The as- weldments to establish a baseline of NDE
designed models were modified where data. This baseline evaluation, which

Aptech Engineering Services, Inc. TP114


5
captures all fabrication and service-related The number of conventional examination
defects and indications detectable by the locations (i.e., terminal locations, fitting welds,
selected NDE procedures, could be phased and locations for maximum SL/Sh and SE/SA
over several years. Lead-the-fleet defects values) is listed for each of the units in
propagating over time can be predicted, Table 1. For these 18 piping systems, the
detected, and reexamined over time using the number of locations varied from 6 to 20.
stress analysis and NDE process as discussed
below. The number of conventional locations
captured in the top ranked four, five, or six
Phase 4 of the project was the prioritization of HEPLC locations is also presented in this
girth weldments. The program named table. The fewest captured locations is for
GIRTHLF was used to determine the Unit AM, where none of the top four HEPLC
multiaxial stresses from when the unit initially ranked locations was captured by the
went online to the present. Creep life conventional approach (i.e., none of the nine
consumption was estimated for each time step conventionally selected locations for Unit AM
to determine the cumulative life consumption matched any of the top four ranked HEPLC
versus time at each selected node. Thus, locations). In almost every case, at least one
expected damage locations were prioritized of the top three ranked locations was not
based on the cumulative life consumption of captured by the conventional approach. The
the material over time, considering the exception is for Unit JH, where the 18
multiaxial stresses from dead weight, external conventionally selected locations included all
loads, internal pressure, and thermal of the top four ranked HEPLC locations.
expansion loads. Prioritization results were
used to select the combined axial and Discussion of Results
circumferential sites for local microstructural The HEPLC methodology is intended to be
evaluation using metallurgical replications. used for ranking purposes. Material
properties are conservatively assumed to have
Results values representative of a significantly
In addition to obtaining the as-found deteriorated condition. Field attributes, such
sustained and thermal expansion stress ratios as greater wall thickness and nominal material
for compliance to the Code, a ranking of creep rupture properties, will result in much
highest life consumption values was less life consumption than the preliminary
determined for each piping system. Minimum prediction values. The author’s experience
life expectancy (MLE) Rank 1 life has shown that the actual service damage is
expectancy/MLE (i.e., Rank 1/MLE), less than predicted by the HEPLC approach
Rank 2/MLE, Rank 3/MLE, Rank 4/MLE, and and selection of lead-the-fleet locations is a
Rank 5/MLE for each piping system is robust methodology. For example, revised
tabulated in Table 1. In twelve cases, the information may significantly change the
fourth ranked location had more than 2.0 quantitative values of life consumption, but
times the remaining life of the first ranked ranking order may not change because a
location. Results of this table suggest that change in operational temperature or higher
lead-the-fleet expected damage sites can be wall thickness readings similarly influence all
sufficiently selected for reexaminations with of the weldments.
four to six locations. In only one case
(Unit HH) was the Rank 5 location remaining None of the conventional locations for
life less than 20% beyond the Rank 1 Unit AM matched the HEPLC top ranked
location. locations. In this case, some bottomed out

Aptech Engineering Services, Inc. TP114


6
hangers modified the stress distribution locations was not captured by the
sufficiently to create maximum ECDs at the conventional approach. In one case, none of
top and bottom of the riser, with three of the the 15 conventionally selected locations
four locations distant from the bottomed out matched any of the top three ranked HEPLC
hangers. locations. This leads to the concern that the
high ranked HEPLC locations not identified
One of the most useful aspects of the HEPLC by the conventional approach may have
approach allows the evaluation of remaining service-related flaws leading to pipe failures as
life for different materials. As an example, the weldments accumulate more than 100,000
creep relaxation curves for two high stress operating hours.
nodes of the Unit IH piping system are
illustrated in Figure 6. Node 20B has SA 335 Many conventionally selected locations can be
P22 material and Node 1325 has SA 335 P11 reexamined less frequently. In two cases, the
material. Throughout creep relaxation, the conventional approach resulted in 20
latter node always has less stress than the reexamination sites, while the HEPLC
former node. However since creep rupture approach resulted in four to six high priority
strength of P11 material is considerably less sites. The HEPLC methodology can provide
than P22 material, the life consumption considerable cost savings to utilities because
curves for the two nodes reveal that the P11 of less scaffolding, insulation removal, and
node has significantly more life consumption examination costs. In addition, the
than the P22 node, as shown in Figure 7. examination of fewer locations with more
allocated time per location will probably result
Another outcome of the HEPLC approach is in higher quality examination results.
that TFDs can be calculated with higher
accuracy because the simulation evaluation The study illustrates that the HEPLC
provides higher confidence in the actual field approach can account for different pipe and
bending stresses. As illustrated in Figures 1 weld materials, pipe geometries, and field
and 2, a difference of less than 2.0 ksi conditions. No more than six locations per
(14MPa) can have a significant effect on TFD piping system were selected for
results. reexaminations, with high confidence that the
remaining locations will not have significant
Conclusions service damage as compared to the selected
This study of 18 HEP systems compares girth locations.
weld reexamination locations selected by the
conventional approach to locations selected Since the HEPLC approach has high certainty
by the HEPLC methodology. The in a qualitative sense, few metallurgical
conventional criteria was used to select high replicas can be performed. With the proper
damage locations by including those sites metallurgical evaluation of the highest ECD
which are the highest SL/Sh, highest SE/SA, locations, there is higher confidence in the
terminal points, and fitting weldments. justification of reexamination intervals. The
Alternatively, the HEPLC methodology was combination of the HEPLC process and
used to predict the top four to six ranked metallurgical replication can be used to
material damage locations based on the quantitatively substantiate reexamination
evaluation of material life consumption. intervals beyond 5 years. This is a significant
improvement over the conventional process,
It was determined that in almost every case, at which may not provide justifiable
least one of the top three ranked HEPLC

Aptech Engineering Services, Inc. TP114


7
reexamination intervals beyond 2 or 3 years Journal of Pressure Vessel Technology,
for higher stressed girth welds. Vol. 112 (August 1990).

References Cohn, M.J., “Life Management of High


American Society of Mechanical Engineers, Energy Piping Girth Welds,” Plant
“Nuclear Case N-47-32, Class 1 Components Systems/Components Aging Management,
in Elevated Temperature Service, Section III, ASME PVP Vol. 283 (June 1994).
Division I,” Cases of ASME Boiler and
Pressure Vessel Code (August 1994). Cohn, M.J., “Prediction of Creep Life
Exhaustion in High Energy Piping System
American Society of Mechanical Engineers Girth Welds,” Fitness-for-Service and
Code for Pressure Piping, B31.1, 1995 Decisions for Petroleum and Chemical
Edition, Power Piping, New York. Equipment, ASME/JSME PVP Vol. 315 (July
1995).
Cohn, M.J., “Mathematical Simulation of
Thermal Displacements for a Steam Piping Cohn, M.J. and May, G.H., “Evaluation of
System in the Material Creep Regime,” Limit Stop User Input Errors in Piping
Design and Analysis of Piping Components, Flexibility Analyses,” Integrity of Structures
Pressure Vessels and Piping Conference, and Fluid Systems, Hazardous Release
ASME PVP Vol. 169 (July 1989). Protection, Piping and Pipe Supports, and
Pump and Valves, ASME PVP Vol. 333 (July
Cohn, M.J., “Remaining Life of High-Energy 1996).
Piping Systems Using Equivalent Stress,”

Aptech Engineering Services, Inc. TP114


8
Table 1
HEPLC STUDIES

MLE Conventional
Unit (1000 hrs) R1/MLE R2/MLE R3/MLE R4/MLE R5/MLE Locations Captured
AM 321 1.00 1.01 1.07 >4.00 --- 9 0/4
AH 447 1.00 1.06 1.11 1.27 1.31 11 1/5
BH 252 1.00 1.46 1.71 >4.00 --- 9 2/4
CH 148 1.00 2.64 3.16 3.80 3.87 6 2/5
DH 182 1.00 1.16 1.53 1.68 2.46 12 3/5
EH 388 1.00 1.03 1.04 2.00 3.12 10 2/5
FH 319 1.00 1.01 1.42 2.69 3.48 18 2/3
GM 283 1.00 1.01 1.06 1.75 1.81 20 3/6
GH 1000 1.00 1.02 1.04 2.59 --- 12 1/4
HM 617 1.00 1.40 1.65 1.73 --- 19 2/4
HH 682 1.00 1.05 1.08 1.11 1.18 13 2/5
IM 190 1.00 2.39 2.45 3.19 --- 7 2/4
IH 225 1.00 1.62 3.05 >4.00 --- 10 3/4
JM 201 1.00 1.26 2.07 3.94 --- 20 2/4
JH 93 1.00 2.80 >4.00 >4.00 >4.00 18 4/5
KM 152 1.00 1.26 >4.00 >4.00 --- 15 1/4
KH 215 1.00 1.00 1.60 2.10 --- 7 2/4
LM 239 1.00 2.20 2.25 >4.00 >4.00 8 2/5

Aptech Engineering Services, Inc. TP114


9
Figure 1 — TFD for a Lower Stressed Main Steam Pipe Girth Weld.

Figure 2 — TFD for a Higher Stressed Main Steam Pipe Girth Weld.

Aptech Engineering Services, Inc. TP114


10
Figure 3 — Isometric Model for the Unit JM Piping System.

Aptech Engineering Services, Inc. TP114


11
Figure 4 — Observed Versus As-Designed Piping Displacements.

Figure 5 — Simulation of Observed Piping Displacements.

Aptech Engineering Services, Inc. TP114


12
Figure 6 — Example Creep Relaxation Curves for Two Materials at Specified Model Nodes.

Figure 7 — Example of Consumption Curves for Two Materials at Specified Model Nodes.

Aptech Engineering Services, Inc. TP114


13

You might also like