WSRC.MS.

934)37

ACCEPTANCE CRYrERIA FOR

CORRODFA) CARBON STEEL PIPING
CONTAINING WELD DEFECTS (U)

G.E. Mertz, P.S. Lam and N.G. Awadalla

Westinghouse Savannah River Company

Aiken, South Carolina, 29808

A paper proposed for presentation at the

1993 ASME Pressure Vessel and Piping Conference
Denver CO
July 25 to 29, 1993

and for publication in the proceedings

This paper was prepared in connection with work done under Contract No. DE-AC09-89SR18035 with the
US Department of Energy. By acceptance of this paper, the publisher and/or recipient acknowledges the
U.S. Government's right to retain a non-exclusive, royalty-free license in and to any copyright
covering this paper, along with the right to reproduce aad to authorize others to reproduce all or part of
this copyrighted paper.

%
.OIS'rRtBUT1ONOF: lr.4:S DOCU_,,JEItTkt..;Ui\tLIh4rrEL._
i

DISCLAIMER

This report was prepared as an account of work sponsored by an agency of the United
States Government. Neither the United States Government nor any agency thereof, nor
any of their employees, makes any warranty, express or implied, or assumes any legal
liability or responsibility for the accuracy, completeness, or usefulness of any information,
apparatus, product, or process disclosed, or represents that its use would not infringe
privately owned rights. Reference herein to any specific commercial product, process, or
service by trade name, trademark, manufacturer, or otherwise does not necessarily
constitute or imply its endorsement, recommendation, or favoring by the United States
Government or any agency thereof. The views and opinions of authors expressed herein
do not necessarily state or reflect those of the United States Government or any agency
thereof.

This report has been reproduced directly from the best available copy.

Available to DOE and DOE contractors from the Office of Scientific and Technical
Information, P.O. Box 62, Oak Ridge, TN 37831; prices available from (615) 576-8401,
FTS 626-8401.

Available to the public from the National Technical Information Service, U.S. Department
of Commerce, 5285 Port Royal Rd., Springfield, VA 22161.
 ACCEPTANCE CRITERIA FOR CORRODED
CARBON STEEL PIPING CONTAINING WELD DEFECTS

Greg E. Mertz, Poh Sang Lam and Bill G. Awadalla

Westinghouse Savannah River Company
Aiken, SouthCarolina

ABSTRACT in the early 1960s consisted of checking the minimum

Acceptance criteria for corroded low temperature, low measured wall thickness against the minimum wall thickness
pressure carbon steel piping containing weld defects is required by the construction code for internal pressure. Normal
presented along with a typical application of these criteria, operating bending stresses were limited by allowable span
The acceptance criteria are intended to preclude gross rupture or guidelines, and seismic loading was not considered by the
rapidly propagating failure due to uniform wall thinning, local original code of construction.
wall thinning, pitting corrosion and weld defects.
Over the past thirty years, the ISI program has evolved,
The minimum allowable uniform wall thickness is based on increasing the number of lines inspected and the number of
the code-of-record allowable stress and fracture criteria. Weld inspection points on each line. Additionally, the use of
defects are postulated as potential sites for fracture initiation, auto:hated UT inspection equipment is increasing, allowing a
The CEGB/R6 failure assessment diagram is used as the fracture better characterization of corroded piping. Modern piping
criteria to determine the minimum allowable wall thickness, construction codes, which include limits on bending stresses
and ueismic loading, are considered. As part of the recent K
The design of a large portion of the low temperature, low Reactor safety upgrades, an as built stress analysis of the
pressure piping is dominated by axial stresses. Existing local secondary cooling water piping was performed, and the
wall thinning acceptance criteria address high pressure piping cooling water piping has been seismically qualified.
where hoop stress dominates the design. The existing criteria
is over conservative, in some cases, when used on low pressure This paper presents acceptance criteria for corroded carbon
piping. Local wall thinning criteria is developed to limit the steel piping that is consistent with (1) the current capabilities
axial stress on the locally thinned section, based on a reduced of available UT inspection equipment, and (2) the current
average thickness. Limits on pit density are also developed to piping construction codes, as represented by the analysis of
provide acceptance criteria for pitted piping, record. Additionally, the acceptance criteria consider local
buckling of a thinned section and fracture, both of which are
not usually addressed explicitly by piping construction codes.
INTRODUCTION
The K Reactor at the Savannah River Site (SRS) has been in The acceptance criteria are arranged into four tiers, as shown
operation since 1954. Carbon steel secondary cooling water in the following table. A section of piping that meets any one
systems have experienced corrosion over the reactor's life. As of these four tiers at the end of the next evaluation period is
a pipe corrodes, the stresses in the remaining ligament acceptable for continued service.
increase, reducing the structural margins of safety and
increasing the possibility of gross rupture or rapidly Screening Piping with a minimu, L wall thickness greater
propagating failure. Instantaneous rupture of a second_,ry Criteria: than 87.5% on the nominal wall thickness is
cooling water pipe has been postulated as an accident initiator, acceptable.

An inservice inspection (ISI) program began monitoring the Uniform Piping with a minimum wall thickness greater
thickness of critical lines in the early 1960s to detect thinned Thinning: than the minhnum uniform wall thickness required
lines that could challenge reactor safety. Acceptance criteria

1 Greg E. Mertz
 by the code-of-record, local buckling criteria and in these systems was less than the size of observed weld
fracture criteria is acceptable, defects. Thus, weld defects are conservatively postulated as
crack initiation sites.
Local Piping with an average wall thickness greater than
Thinning: the minimum required uniform wall thickness In the early 1980s, as part of the SRS L-Reactor reactivation
divided by a local thinning reduction factor is effort, an assessment of the carbon steel piping was made to
acceptable, determine their suitability for continued service. Much of this
piping had been installed during the plant erection in the early
Pitting: Piping with a base thickness greater than the 1950s and is similar in construction to the piping in the K
minimum required uniform wall thickness divided Reactor. In accordance with the codes of construction, the
by a pit density reduction factor is acceptable, carbon steel piping welds were accepted by visual inspection.
Radiographic examinations were not required.

DEGRADATION I DEFECT CHARACTERIZATION Visual, radiographic, magnetic particle and destructive

metallography were used to confirm the adequacy of the
original pipe welds for the intended service requirements [1,
2]. The inspections did not reveal any evidence of cracking.
System Description, The combination of six postulated surface defects, 1" long by
The K-Reactor cooling water piping is a moderate energy
piping system with pressures less than 100 psig and 100% deep,
bounds 99% 3"x85%,
of ali of 5"x60%, 6"x45%,
the observed weld 9"x35%
defects. and
The 10"x25%
5"x60%
temperatu_'es between 40°F and 180°F. The piping carries river
flaw postulate has the largest stress intensity of the six defects
water in the reactor building to heat exchangers and then back
and is chosen as the design flaw postulate for circumferential
out of the building, flaws.

ASTM A53 piping and pipe fabricated from ASTM A285 Since axial flaws are shop welded under controlled
plate are used in the K-Reactor cooling water piping. These
systems were designed to the 1950s B31.1 code with larger conditions
is believed asto opposed
be betterto than
field field
welds,welds,
the quality
and theof weld
shop defect
welds
diameter piping designed in accordance with the ASME Section
VIIY., Unfired Pressure Vessel code. Modifications to the data measured during the L-Reactor reactivation effort is not
piping system after 1981 are based on the B31.3 code. Recent applicable to axial welds. Two flaw postulates, consistent
with the flaw postulates used in ASME Section XI [3], were
as built piping stress analyses include pressure, deadweight, chosen for the design axial flaw postulate.
and seismic loading. The 1950 vintage piping codes did not
include seismic loading. Thermal loading on these piping The design flaw postulates are based on nominal wall
systems is insignificant, thickness. Since the majority of the flaws observed were
interior surface flaws, the relative flaw depth (flaw depth
Dearadation divided by the actual wall thickness)decreases when corrosion
- is on the inside of the pipe. Exterior corrosion removes the
A systematic assessment of degradation mechanisms was remaining ligament from an interior surface flaw increasing
conducted to determine the potential for rapidly propagating the relative flaw depth and increasing the potential for fracture.
failure or gross rupture. The degradation assessment Since exterior corrosion cannot be discounted on some lines, a
considered the effect of fluid-material compat:,bility, operating compromise between the two thinning assumptions is made.
conditions, and service history. General attack, galvanic Wall thinning is assumed to occur equally on both the inside
corrosion, pitting corrosion, and microbiologically-induced and outside surfaces. For the vast majority of lines, this
corrosion (MIC) were identified as potential degradation assumption is conservative, as thinning and surface flaws are
mechanisms. The cooling water piping is not susceptible to both on the inside of the pipe. The flaw length is assumed to
erosion corrosion, primarily because of the high oxygen remain unchanged.
content of the river water.

For the purposes of these acceptance criteria, damage caused SCREENING CRITERION
by corrosion, regardless of the mechanism, is classified as The screening criterion used in ASME Code Case N-480 is
either uniform thinning, local thinning, or pitting, adopted, and piping with a minimum wall thickness greater
than 87.5% of the nominal wall thickness is acceptable. This
screening criteria is consistent with the piping fabrication
Weld Defect Characterization tolerances.
An initial crack or defect is required to initiate fracture.
Cracks in the carbon steel piping systems have not been
observed during the long operating history of the SRS UNIFORM WALL THINNING CRITERIA
reactors. Two potential fracture initiators are pits and weld The minimum required uniform wall thickness is the
defects. Pits and weld defects subject to fatigue may develop minimum thickness that satisfies (1) the code-of-record
cracks which could initiate fracture. An assessment of thermal allowable stresses, (2) local buckling criteria, and (3) fracture
and mechanical fatigue in the piping concluded that fatigue criteria.
c_'ackil_g was unlikely. Additionally, the largest pit observed

2 (;reg E. Mertz
 The piping forces for SRS low pressure, low temperature where SE is the product of the allowable stress and joint
piping are typically dominated by seismic loading, which are a efficiency.
function of the pipe's natural frequency. As a pipe thins, the
stiffness and natural frequency also decrease, possibly causing
an increase in the magnitude of piping forces. One possible, Local Buckling
but cumbersome, solution would be to reanalyze each piping Local buckling is not a limiting failure mechanism in
system using the actual measured wall thicknesses to obtain common pipe sizes with R/t ratios ranging between 5 and 20.
the thinned piping forces. Another, more tractable solution, Contrarily, large diameter low pressure piping may have R/t
is to adopt the as-built reconciliation methodology [4], which ratios near 50, and severe thinning of this piping could raise
accepts frequency shifts less than 10% and allows reanalysis the R/t ratio to 150. Reference 9 summarizes test data which
using simple structural models to bound the increase in pipin_ demonstrates that local buckling of straight pipe section can
forces when the frequency shift exceeds 10%. The later reduce the ultimate load capacity of A53 piping with an R/t
approach is used in this analysis to conservatively increase ratio above 120. Thus, the minimum thickness is limited to
the thinned piping forces, as required, preclude local shell buckling (compressive wrinkling) of
thinned pipes by limiting the longitudinal compressive stress
to the allowable stress for cylinders subject to axial
Analysis 9f Record compression.
The minimum uniform thickness that meets the code-of-
record allowable axial stress is the thickness that satisfies the The maximum compressive stress is conservatively assumed
following three equations for dead, dead + seismic and thermal to be constant over the entire cross section. Bending stresses
loading, are not intensified at tees and elbows because the intensified
stresses are peak stresses developed to predict fatigue life and
P DO 0.75i Ma ") act over a small area. ASME Section III, NC-3133.6 is used to

compression.

P 4---_
Do + 0.75i Z (Ma <+ SAllowable Normal _> determine the allowable compressive stress for pipes in axial
Mb)
4-----_
+ Z < SAllowable Upsetf (1) Fracture Criteria!

i Mc
Z < SAllowable Thermal J For the SRS cooling
transition temperature
approximately
water system, the nii ductility
of archival A53 and A285 pipe is
equal to the minimum operating temperature.
Previous fracture assessments have demonstrated that brittle
where P is the internal pressure, D O is the outer diaJneter, t fracture is not credible zt the minimum operating temperatures,
is the minimum uniform thickness, i is a stress intensification based on the nominal pipe wall thickness [2].
factor, Z is the section modulus, M a, M b and Mc are the dead
load, seismic and thermal moments, and SAllowabl e is the The applied stresses in the remaining ligament of a pipe
allowable stress for a given loading, as defined by the code-of- increase with wall thinning. A weld defect located in a region
record, with increased stresses could result in brittle fracture, ductile
tearing, or yielding through the remaining ligament,
An alternate form of the pressure term in Equation 1, depending on the stress, flaw size and material properties. In
p di 2 this analysis, the CEGB/R6 failure assessment diagram (FAD)
D° 2 - di 2 , where d i is the inside diameter, may also be used. is
in used
failureto through
determinea weld
the minimum
defect [5, thickness
6]. that will not result
The aJternate pressure term is more accurate and provides lower
stresses for thick wall piping. The stress strain curves for both A53 and A285 piping are
ela:_tic-plastic below 1.5% strain, as shown in Figure 1. A
Both the section modulus and stress intensification factor material specific FAD is used in this analysis, as shown in
vary with the minimum uniform thickness. At tees, the Figure 2 and is given by
amount of wall loss in the header is assumed to be equal to the

amount of wall loss in the branch line when calculating the 1 __

stress intensification factor. The left hand side of Equation 1 Kr_ for Lr < of (3)
is
theapproximately linear that
whensatisfies
plotted these
on a inequalities
log-log scale,can and
minimum thickness be ._/ Eeref + Lr 3 Oy Oy
determined in several iterations. _ Lr Oy 2E eref

The minimum uniform thickness that meets the code-of- where eref is the true strain corresponding to the true stress
record allowable hoop stress is Lr x Oy, Oy is the 0.2% yield stress, _3f is the flow stress, and E
p Do is Young's modulus. Also shown in Figure 2 is an evaluation
t=2 (SE + 0.4 P) (2) point (Kr, Lr) where

3 Greg E. Mertz
 !

80 I- _ _",,,,,,,,_ Unsafe
70. 0.8 .: _ -
Safe • (Kr,l_.r) 1
Flow Stress 0.6
. ,...4

_. 50 Flow Stress
0.4-
40
0.2-
30 " "1
l
A53 0- . ..

10 .-tb-. A285 Lr

0 0.2 0.4 0.6 0.8 1 1.2 .4

0 T .... I .... I .... I ''_'l .... i .... FIGURE 2 FAILURE ASSESSMENT DIAGRAM
0 0.025 0.05 0.075 0.1 0.125 0.15

True Strain (in/in) burst tests of corroded pipelines performed by Battelle in the
early 1960s [7, 8]. This criteria has been adopted in the ASME
FIGURE 1 STRESS STRAIN N-480 code case with an additional limitation on the amount of
R ELATIO N SHIP transverse wall thinning to preclude bending induced failure.
The limitation on transverse thinning imposed in the N-480
code case is not adopted in the current local wall thinning
Kr = Applied Stress Intensity "_ criteria, because the current criteria explicitly checks axial

Applied Stress
Lr = Limit Load KIC l (4) stresses.
Stress /B,xial Stress Criteria
Axial bending and pressure stresses are checked using the
The applied stress intensity includes both applied and actual thinned cross sectional geometry and the code-of-record
residual stresses. For the part throughwall design weld defects allowable stress
used in this analysis, a local limit load stress that causes
plasticity across the remaining ligament is used for the limit
load stress in the definition of Lr. Applied stresses in Equation P rc R 2 + Moment + P r_ R 2 Y < Sallowabl e (5)
4 are multiplied by the ASME Section XI Appendix tl factors of Athinned Zthinned
safety.
where Y is the distance between the neutral axis and the
Evaluation points corresponding to progressively thinner
pipe walls are determined. The last evaluation point below the centroid, and the thinned cross sectional area is equal to
FAD failure surface is the minimum wall thickness that will not rc taverag e R2. A parametric study of piping with a sinusoidal
result in failure through a weld defect, variation in wall thickness [Appendix A] shows that the
section modulus of a thinned pipe is conservatively given by

LOCAL WALL THINNING CRITERIA Zthinned = r¢taverage RLT A R2 (6)

Existing criteria for local wall thinning are primarily based
on pressure loading. Low pressure piping is dominated by where RLT A is given in Figure 3 as a function of the amount
axial stresses as opposed to high pressure piping which is of thinning and the length of region below the average
dominated by hoop stresses. For low pressure piping, the thickness. Assuming that the nominal pipe thickness is equal
existing local wall thinning criteria are used to preclude hoop to tma x, then the amount of thinning, b, is typically between
stress failure and subsequent leakage. Additional criteria, 0.125 and 0.7, which correspond to minimum wall thickness
based on the moment capacity of a thinned pipe, are used to of 87.5% tnomina I and 0.3 tnomina 1, respectively. The curve
preclude axial stress failure. A pipe with locally thinned areas for b=l is included in Figure 3 for academic interest.
is acceptable if it meets both the hoop and axial stress criteria.
Combining the pressure terms in Equation 5 and

Hoop Stress Criteria simplifying,

ANSI/ASME B31G contains criteria to preclude rupture of
piping with locally thinned areas and is based on pressure

4 Greg E. Mertz
 l OTO O O

ib=0.75 _ Pits Failure plane

.......................... 000<)0-,0--4
_[t_ 0.6

11 ii ____ tmax "tmin O-- O O O

_
-- 0.5 .............;............! ..... b= tmax -- FIGURE 4 PIT ARRAY

-"
-
0.4 .................................................
,' , ,i wn
1LA• • , •
il w
acts over tens of cycles, and that this local thinning
acceptance criteria is appropriate for the K-Reactor cooling
0.3 [b=l i,,,, water piping. A detailed fatigue analysis would be indicated fox'
piping dominated by thermal expansion.
0 0.1 0.2 0.3 0.4 0.5

Length of Region Thinner than taverage PITTING CRITERIA

Pipe Circumference Pits are conservatively assumed to be throughwall holes
resulting in a reduction of load capacity due to yielding of the

FIGURE 3 LOCAL THINNING remaining ligament. Using the idealized rectangular pits
REDUCTION FACTOR, RLT A array, as shown in Figure 4, the pit density, p, is given by

p (111
P R (1__ Y /R'_ Moment -4 g2
t,-verage_.2 + _)+ rc taverag e RLT A R 2 <; Sall°wable(7) The plastic capacity oi the pitted section is Tpi t = N 2(g-

As shown in Appendix B, Equation 7 can be approximated rp)Oy, while the plastic capacity of an unpitted section is Tpit
by = N 2g Oy, where N is the number of pits on the section, 2g is
the pit spacing, rp is the pit radius and Oy is the yield stress.
P R + Moment Define Rpit as the ratio of the plastic capacity of the pitted to
2 taverag e RLTA _ taverage RLTA R2 < Sall°wable (8) unpitted section and substituting in Equation 11, yields

The term taverag e RLT A is an effective average thickness.

The allowable
fracture criteria axial stress criteria,
are satisfied local thinned
on a locally bucklingarea
criteria,
if and Rpit = 1 _ ,_]2__ (12)

treq uniform axial < taverage RLTA (9) A pitted section is acceptable if the product of the unpitted
base thickness and Rpit is greater than the minimum required
or uniform wall thickness.

taverag e > tre.q uniform

RLTA axial (10) Triangular array,
rectangular arraysshown
were in investigated in addition
Figure 4, resulting to the
in values of
Rpit greater than er equal to the rectangular array. Pits are
where treq uniform axial is the minimum uniform thickness randomly distributed in the field, and the actual reduction in
that meets the uniform wall thinning criteria for axial stresses, load capacity is probably less than estimated by Equation 12.

The local stresses on the rema:'ning ligament of a locally For field applications, the pit density, p, can be determined
thinned area will be elevated, and a fatigue assessment of these by direct measurement of the pitted piping, or by
elevated stresses should be performed. A fatigue assessment of conservatively estimating the average pit radius and counting
the K-Reactor cooling water piping demonstrated that thermal the number of pits in a given sample area,
and pressure cycling are minor, that the seismic loading only

$ Greg E. Mertz
 0
4

50- tmax = 100% tnomina I _ \

taverage= 75% tnominal _ \
40-

10

0
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Nominal Thickness £,._o'_ __

FIGURE 5 DISTRIBUTION OF MIHIMUM
REQUIRED UNIFORM WALL THICKNESSES
Minimum Allowable Uniform Thick.hess ..... _ i

2 Number of Pits (I3) FIGURE 6 HYPOTHETICAL LOCALLY

p = r_ rpi t - Sample Area TH!NNED AREA

IMPLEMENTATION ON SRS COOLING WATER 100%trlominal-50% tnomina 1 0.5

PIPING b= =
This acceptance criteria is used to determine the minimum 100% tnominal
uniform wall thicknesses of the K-Reactor secondary cooling
water piping and to disposition locally thinned areas. Figure 5 yields RLT A = 0.78. From Equation 10,
shows a histogram of the minimum required wall thickness in a
typical cooling water line. Since only a very small portion of taverage = 75% tnominal >
the line is highly stressed, the majority of the line has treq uniform axial= 55% tnominal= 0.70 tnomina 1
generous allowances for wall thinning. This information can RLT A 0.78
be used to help prioritize future inspections and focus on areas
with the smaller corrosion allowance, the local wall thinning criteria are met, and the section is
acceptable.
A hypothetical locally thinned area in a cooling water pipe
is shown in Figure 6. Since the minimum thickness is less
than 87.5% of the nominal thickness, this section does not CONCLUSION
meet the screening criteria and further evaluation is required. Acceptance criteria for corroded low temperature, low
pressure carbon steel piping containing weld defects are
The minimum allowable uniform wall thickness meeting the presented. The acceptance criteria protect against gross
hoop stress criteria and fracture criteria for axial flaws is 20% rupture or rapidly propagating failure due to uniform wall
of the nominal wall thickness. Since the required thickness for thinning, local wall thinning, pitting corrosion, and weld
hoop stresses is less than the minimum thickness, the uniform defects.
wall thinning criteria is satisfied for hoop stresses.

The minimum allowable thickness meeting the axial stress ACKNOWLEDGMENTS

criteria, local buckling criteria, and circumferential fracture The information contained in this article was developed
criteria is 55% of the nominal wall thickness. Note that the during the course of work done under Contract No. DE-AC09-
minimum wall thickness is less than the required uniform 89SR18035 with the US Department of Energy.
thickness, and the uniform thinning criteria is not met for
axial stresses. Thus, an evaluation of a locally thinned area is The authors would like to acknowledge K.J. Stoner, G.R.
indicated. Caskey, R.S. Ondrejcin and D.M. Barnes for their work in
defining potential degradation mechanisms in the SRS piping;
Let the transverse width of the locally thinned area, Lt in G.S. Dsouza, R.E. Sprayberry and L.R. McCabe for their work
Figure 6, be half of the pipe's circumference. Using Figure 3, in characterizing weld defects; and G.A. Antaki who led the
with the length of the region thinner than taverag e equal to 0.5 seismic qualification effort for providing as-built piping stress
x pipe circumference and analyses.

6 Greg E. Mertz
I

, , 9

REFERENCES tmax

1 N.G. Awadalla,
Probabilistic D. A. of
Assessment Crowley, and W. in
Weld Quality F. Steel
Yau, taverage _
Piping Under Seismic Conditions, International tmin \

Journal of Pressure Vessels and Piping, Vol. 17, 1984. \

2 G.E. Mertz, K. J. Stoner, G. R. Caskey, and J. A.
Begley, Fracture Assessment of Savannah River c=l
Reactor Carbon Steel Piping, PVP-Vol. 288, 1992.

3 R.R. Maccary, Nondestructive Examination

Acceptance Standards Technical Basis and
Development of Boiler and Pressure Vessel Code,
ASME Section XI, Division 1, EPRI NP-1406-SR,
1980.

4 R.W. Mikitka, and R. F. Reedy, Guidelines for Piping

System Reconciliation NCIG-05 Rev. 1, EPRI NP- FIGURE A.1 VARIATION IN WALL
5639, 1988. THICKNESS

5 I. Miline, R. A. Ainsworth, A. R. Dowling, and A. T.

Stewart, Assessment of the Integrity of Structures lt

Vessels and Defects,

Containing Piping, Vol. 32, 1988.Journal of Pressure
International I = 2 I R t(0) (R Cos0) 2 dO - A "_'2 (A.3)
6
6 I. Miline, R. A. Ainsworth, A. R. Dowling, and A. T.
Stewart, Background to and Validation of CEGB Report Z= I (A.4)
R/H/R6-Revision 3, International Journal of Pressure R + I-y" I
Vessels and Piping, Vol. 32, 1988. I I

7 ANSI/ASME B31G, Manual for Detennining the where Y is the distance from the center of the section to
Remaining Strength of Corroded Pipelines. 1984.
the neutral axis, R is the nominal radius, taverag e is the
8 T.L. Gerber, P. C. Riccardella, A. Y. Kuo, and D.R. average thickness, and t(0) is the angular variation in wall
Pitcairn, Acceptance Criteria for Structural Evaluation thickness. A parametric study is performed which varies the
of Erosion-Corrosion Thinning in Carbon Steel thickness as a sinusq)idal function:
Piping, EPRI NP-591 lM.

9 T.V. Galambos, Guide to Stability Design Criteria for t(0)=tma x ( 1 - b 1 + Cos(c

2 0)) forO<0<it (A.5)
Metal Structures, 4th Edition, 1988.
tma x - tmi n
where b= , and c is the number of equal spaced
tmax
APPENDIX A DEVELOPMENT OF LOCAL WALL locations around the circumference with a thickness below the
THINNING REDUCTION FACTOR, RLTA average thickness. The angular variatioii of thickness is
The elastic section properties of a thinned pipe, based on shown in Figure A.I for two cases of wall thinning. The
thin shell theory, can be calculated from angular variation of thickness is assumed to be symmetric
about the vertical axis.

A= _dA=2 Rt(0) d0=2ittaverage R (A.I) region with a thickness less than the average thickness is
Of Since the circumference of a pipe is ltD, the length of a
nD
n Lt<taverage = "_-c (A.6)
AY = IydA = 2
j R t(0) R Coso d0 (A.2) section modulus,
compact form as
Equation A.4, can be rewritten in

Z = RLT A It R2 taverag e (A.7)

7 Greg E. Mertz
 q _ 0

° 9

I.I tmax

taverage ._

1 ..... _ " = .... "._j¢ train _\

0.9
1. 5
o.s J /I

FIGURE A.2 RLTA, b=0.50 .I

where RLT A is a reduction factor for locally thinned areas,

and is given by FIGURE A.3 VARIATION IN WALL
THICKNESS, c=1.3
I
RLTA = (A.8) 1.2

[R + {"y-]]_R2taverage /b=l

Equation A.8 is shown graphically in Figure A.2 for b=0.50,

Note that for integer values of c greater than 2, RLT A is equal
to 1. The minimum value of RLT A is at c=1.3, which
el_t_
_

+,e
"
\
1
b=0.50

/
......
2! 4 6 8 10
corresponds to the thinned geometry in Figure A.3. Note that [.2
.1 0.9 I c
as the size of the thinned area decreases (increasing c), the _ \
effective average thickness, RLT A taverag e, approaches the
average thickness, taverag e. A conservative, lower bound 0.8 \ 1
envelope of the thin shell analytical solutions for RLT A is
given in Figure 3.

. Thick shell solutions, with double integration of Equations FIGURE B.1 PRESSURE MULTIPLIER
A.1 to A.3, for 24" Schedule 20 (R/t=48) and Schedule 160
(R/t=5) pipe yield RLT A within 3% of Figure A.2 for c=1.3 and
b=0.5. This limited study suggests that the reduction factor for
local wall thinning may also be used for thick wall piping. At elbows and tees, the intensified pressure term is

APPENDIX B SIMPLIFICATION OF PRESSURE P R (1 + 0.75i Y lR / (B.3)

INDUCED AXIAL STRESS EQUATION taverage_. _ RLTA )
The pressure induced axial stresses are given by
For intensification factors in the 2-3 range and moderate

P R
/R) RL'I'A+_ (B.1) accurate to about 10%. For larger intensification factors the
(1
taverage_,'2 --Y
+ RL'I'_'A) = 2 RLTAp tR
a verage (2_-) wall thinning,
accuracy or this bapproximation
< 0.5, thendegrades.
the approximation
The accuracy will be
of this
approximation is a secondary issue for low pressure piping
The term RLTA+ 2 Y lR is shown in Figure B.I for b=0.5 systems, where the pressure term represents less than 10% of
the allowable stress. Equation B.3 can be used as the pressure
and
value b=l. This when
term isc=1.3
typically
and less
b=l.0.than Recall
1.10, with
that a b=l.0
peak term for high pressure systems. Note, regardless of the
of 1.13
pressure, as the size of the locally thinned area becomes
corresponds to tmi n = 0, which is only of academic interest, smaller (large c), then the distance to the neutral axis
Accepting up to 10% error in the pressure term, allows approaches zero and the approximation is valid.
Equation B.1 to be approximated by

e R (! v/R) PR
taverage\ 2 + RUI,Aj=2 RLTA taverag e (B.2)

8 Greg E. Mertz