ML19309A125
| ML19309A125 | |
| Person / Time | |
|---|---|
| Site: | Rancho Seco |
| Issue date: | 07/31/1972 |
| From: | Keilman L, Raasch D SACRAMENTO MUNICIPAL UTILITY DISTRICT |
| To: | |
| Shared Package | |
| ML19309A119 | List: |
| References | |
| NUDOCS 8003260842 | |
| Download: ML19309A125 (15) | |
Text
- _ _
i i
.V.
(
WELD DEFECTS FOUND IN STEAM GENERATOR LATERAL SUPPORT RANCHO SECO NUCLEAR GENERATING STATION UNIT #1 JULY 1972 e
I
}Af &e Prepared By
}
LY R.'Kdilman, ' Supervising Mechanical Engineer
()
i
'?
/
(
Approved by
/ / j/ ~Z 'f 6/
D. G. Raasch, Principal Project Engineer
'8003260 [Q
. (
SIGNIFICANT DEFICIENCY DEFINITION This report on welding defects on steam generator lateral supports is submitted under the requirements of 10CFR50.55. The condition reported on herein, is considered a significant deficiency only because of the magnitude of the repairs. Normally, weld defects which are subsequently repaired are not considered significant.
This report describes the history of welding defects discovered in our Rancho Seco No. 1 steam generator lateral supports.
It provides a chronological review of events in the vendors shop and at the site as well as the program established to assure the fabrication of acceptable support structures.
INTRODUCTION k-The purpose of the steam generator lateral support is to restrain movement of the steam generator during a postulated seismic occurre.nce. The support e
is in a double web "H" beam cross section configuration and made in a 360 degree ring which fits around the top of the steam generator at the tube sheet elevation. This ring has four support legs which are bolted to an anchor plate imbedded into the concrete shield walls.
The original design of the stemn generator lateral supports specified the use of ASTM-A-36 carbon steel material.
In the evolution of design, a larger load was assumed which resulted in a change in the lateral support material to a higher strength steel, ASTM-A-514.
Since this steel requires careful attention to welding details; the manufacturer was required to'sub-mit all welding and NDT procedures to our welding, metallurgical anc NDT (c. <
specialist for approval prior to. performing any shop work.
Our specifi-cations also required 100% radiographic examination of accessible weld joints, or in lieu of this, non-radiographable joints were to be 100%
magnetic particle inspected in accordance with Section VIII of the ASMN Boiler and Pressure Vessel Code.
Periodic surveillance of the vendor's operation was performed to assure compliance with these requirements.
In spite of these special precautions, difficulties were encountered in the welding of this material which resulted in an abnormal amount of weld repair.
Additional requirements have been imposed on both the welding process and on inspection to insure the desired quality. These requirements are discussed in the history portion of' this report.
(.<
e
-l i
~. 9
~
HISTORY OF~ WELDING PROBLEM ON STEAM GENERATOR LATERAL SUPPORTS r,
La
+
s
~
The' welding effort on the north-steam generator lateral support (C-352A),
identified _ as Unit-fl by the vendor, was started in late January of 1972.
By,early February, the three segment welds on the inner ring had been completed and had passed radiographic examination. The welding of the-i stiffener plates to the inner and outer rings was performed during Febru-f ary. Each weld pass of the 2" full penetration stiffener to ring weld was magnetic particle inspected.
The first evidence of a significant problem was noted in early March when a large crack, shown on Figure 1, was discovered. This crack was ground out and dye penetrant inspected to assure complete removal of the defect.
Following the repair of the first crack, two other cracks appeared. These
,(.
are also shown on Figure 1 and were 30" and 40" in length.
The welding on the south steam generator lateral support, Unit #2, started in late February and proceeded concurrently with the north support. On March 13th, a crack approximately 42" long developed in the weld between i
the stiffener plate and the outer ring, as shown on Figure 3.
The crack 1
was ground out and the weld repair started.
During this repair, the crack re-appeared.
l Based on the cracking problems experienced on both supports, the welding
)
operation was suspended and an investigation initiated to determine the
- cause of _the cracking and to establish a program that would produce accept-(
abie su'pports. The investigating team included experts from the Bechtel-s 5 >
l Metallurgical Department in San Francisco, vendor engineers, and, by telephone communication, technical experts from the steel industry.
This investigation resulted in a number of chanzes to the welding pro-cedures. The major changes to our procedure were:
The size of 'the weld rod was reduced from 1/4" to 5/32" or a.
less.
b.
Pneumatic hammer peening would be performed after each weld pass except for the root and cover passes.
c.
The preheat temperature would be 200*F or above, d.
The maximum interpass temperature would be 400'F.
On March 28th, follo, wing the institution of these new procedures, we sent
,.(,
a metallurgist from the Bechtel San Francisco staff into the vendor's
-)
shop to provide any assistance that might be required to assure the proper application of these procedures.
He witnessed the welding on the stiffener to ring and reported that the work was proceeding without problems.
l At this point, welding and weld repairs on both supports, using the improved procedures, were started again.
By mid June, the top side of the north support was complete and the unit turned over.
Several cracks, as shown 4
on Figure 2, were found on the bottom side and the repair of these is in progress.
The welding of the stiffeners to the rings on the south support was completed on April 20 and a preliminary shop magnetic particle inspection performed l
i
(
l
{ +
{}'
. prior to machining. There were some minor indications noted at that.
time,(porosity, pinholes, etc.).
The indications were: marked but since-the ring was cold, the vendor proceeded to do the machining. before exploring these indications. After machining was completed, work' pro-ceeded with the welding of the support arma to the rings without too much difficulty.
Some indications from both the Xray and magnetic particle inspection were found as noted on the attached weld map, Figure 3.
The support was complete by May 26th and was allowed to set until May 31st before it was magnetic particle. inspected. The minor indications that were noted at this time were removed and repaired.
In early June, the unit's final inspection was witnessed and accepted by our source inspector.
SMUD and Bechtel engineers met at the vendor;s shop on June 6th to review t;e final inspection of the south support and the progress of the weld repairs on the north support.
The south unit was shipped to the Rancho Seco site in mid June where it 4
was sand blasted, given a visual inspection of all weld joints, and accepted. The unit was then spray painted and as the paint dried, it 4
was noted that in three places it appeared that the paint surfaces had separated. The weld joint in this area of the support was sand blasted and magnetic particle and dye penetrant inspected.- It appeared that there were several minor cracks (see Figure #4) in the fil.let welds of the support legs.
It was decided that the entire support would be sand blasted and re-examined.
Several additional minor defects (as shown on Figure #4) were discovered.
-n 4
e - r
~(*
On June 21st, a meeting was held at the site with the Beci.tel San
~
Francisco metallurgists to discuss the cracks discovered a' the site.
Since the origin or time reference of these cracks could not.be adequately. defined, it was decided that these latest minor defects would be repaired and that a more extensive acceptance program be developed by the District.
(See the following section of.this report.)
A meeting with the vendor was held at the site on June 29th to discuss the method of handling the repairs on these latest defects.
It was agreed that this support would be returned to the vendor's shop for repair and f!nal Fench mark inspection. Repair work on this support is proceeding satisiertorily at dhis time with a scheduled completion date in mid August.
p(-
e 1
(
l I
i,
DISCUSSION
'h-It is our opinion that the large cracks experienced initially were caused by too much heat input'which resulted 'in high thermalstret es during subsequent cooling. The corrective action of reducing welding rod size and interpass peening has eliminated these large cracks and none have occurred since making these changes. The small cracks, or lack of fusion defects as shown on Figures 3 and 4, are felt to be a result of improper attention to the following specific welding details:
1.
NinL n preheat temperature of 200'F has not always been maintained.
2.
Some of the fillet welds were made with a full weave technique instead of the stringer bead technique.
3.
Butter welds in the toe area of fillet welds have not always
{
been made.
The manufacturer of A-314 steel recommends that close attention be given to these specific details and our source' inspector is now closely monitoring these to correct these causes of weld joint cracking.
Prior to selection of A-514 steel for this application considerable thought was given to the design. The plant and NSS System configuration have space restrictions khich' limit the choice of supporting systems.
Considerable study was conducted before selecting the final supporting system. This study involved the investigation of various configurations including the use of suppressors, cables and a number of different steel systems. We did not want a system.which would inhibit the normal growth of the system in any way.
Because of vent area requirements, piping, electrical and o
.y.
s j
other structural supports and access platforms, an attempt was made to minimize the size. of this support.
Whenlthese different configurations ~ were.being considered various materials were also considered. As a general policy,-we always attempt.
to design items using materials.which are commonly used for a particular application.
In the case of welded -structures A-36 steel is used where ever feasible. Under certain conditions, there are limitations such as dead weig t, space or available material thickness which dictate the use of materials other than A-36.
The use of A-514 steel is normally limited to application where dead weight and/or physical size are a prime considera-tion.. The latter was the case with the support rings.
i I
(,
We were well aware of the potential welding problems involved with the use of A-514 steel.
On the other hand, many successful structures have been constructed using this steel such as bridges and penstocks. 'These structures, in many cases, have been functioning successfully under high atress conditions for years and also have been subjected to vibrating and impact loadings.
Expcrience with this type of steel shows that welding stress cracks occur within thici to four days af ter completion of the welding.. Since the configuration of our lateral supports are very similar to stiffener rings for hydroelectric generating unit penstocks, we are confident that once these supports are fabricated free of defects they will satisfy all our i
functional 1 requirements.
l
]:..
~
(.
In addition to previously_ mentioned welding parameters, we should make note that a low hydrogen E-9018 D1 weld rod has been used throughout the fabrication of these supports. Warming ovens located close to the work are being used and are routinely checked by our -source inspector.
In conclusion, we feel that we have 1) defined the problem as being related to the welding process, 2) isolated the cause of the cracks in the welding, and 3) imposed adequate corrective action to eliminate the problem.
e O
((.
C
f ACCEPTANCE PROGRAM-To assure a good end product, phase one of our acceptance program is to conduct our own 1007. magnetic particle and dye penetrant inspection of both lateral supports prior to shipment from the vendor's shop. Weld surfaces will be ground smooth prior to testing. To assure compliance i
with approved weld procedures, we have assigned a full time Bechtel Source Inspector to the vendor's shop for the repairs and the remaining support fabrication work.
Phase two of our program will be to instd1 the lateral supports leaving all weld areas unpainted. After the supports have hung, stressed by its own dead weight for ten days, we will reinspect all critical weld joints with the dye penetrant method.
L.
The final phase of our program is to dye penetrant. test all critical weld joints following cooldown from hot functional testing. This functional test will subject the supports to a complete thermal cycle of normal operating stresses.
Any defects found in the acceptance program will be subject to a thorough engineering evaluation.
'1 l0
/06
- 7 88 pRcn I>*
CE4(E
-lOdTER RlWG 2"hS/Ae i
- l, i
--S STIFFEN.ER
./
2 IHWER rid
~
-.g I-
\\
{.
j, 40'CA W
~
SkMOK
/
@y!f85
/
h1hRCH ll B
]
30"
~
crancK A PPPsow.
(oursies.) 2. 80p N A Rc H H,72 FIGURE #1 - NORTH STEAN GENERATOR LATERAL SUPPORT - C 352-A UNIT #1 - WELD DEFECT MAP TOP SIDE
~
VGAA hD$
C..
j
.,3
- r l*240d UTSID h/
l
)
~
W
\\/
Q s
/
\\\\
j CRACE
/
/f" C/?ncE ourSibe 007.S82 0
(
FIGURE d2 - NORTH STEAM GENERATOR LATERAL SUPPORT - C 352-A UNIT #1 - WELD DEFECT MAP BOTTOM SIDE a
SOUTH SUP T " B"
.m N UGs$g
~
N 2.
83 6" X Ray Ind.
Lack of Fusion.
Seam Weld Cracked during In-process R paired X Ray & Mag OK MagInd.1"Longx3/16 welding.
6-1 Deep.
(Visually Detected) 4 Repaired, Mag OK, 6-1-72 Approx.42"long.3/8" deep.
Repaired, X Bay &
ag. OK.
p Slag Inclusions detected at mag.
I_
testing.
Removed, repaired, & mag. OK 6-1-72 Note; Other Slag
- Inclusions
.I g
entirely removed by grinding k
\\
. tas yarious Mag j
r
. (Non Metalic Inclue L Shaped Mag. Ind.
ons) upper edge of Crack 4" Long 4" high nter ring around dia.
1/k" deep.
l l-72 Repaired. Mag & X Ray OK 6-3-72
\\
)
i 84 B,
X-Ray, Mag. Indications
&' Weld Repair Map. -
FIGURE #3 - SOUTH STEAM GENERATOR LATERAL SUPPORT - C-352-B UNIT #2 - WELD DEFECT MAP
-J
C 3L
, ; g.,
2riz 1
33 LENGTH i" DEEP T-7-72 p
2ia X 1" DEEP 4"
SURFA.
fif "
30-~TO V s...
VEw i.. X.g Sg SURF.
DEEP l"SUR 4'LO, FULL i
'. WEL esS si'/r SU2FAf DEEP
,,, b 6"i 7" 2" SU2F.
Z" SURF. i DEEP-LEMGTH -
4" LONG f
f(" DEEP FULL WELC l d'x { '
1"9 7-7-72
\\ 'OEEP A
B
\\
}
NOTE: FOE 52 $ 53, ENTIEE FLANGE PLATES BEIAI6 R.E-MOVED PRIOR 7 0 R E - W E L D.
FIGURE 4 SOUTH STE AM GENER ATOR LATERAL SUPPORT l
C-352A - UNIT #2 - WELD bEFECT MAP l
m