V-9901&V-9902 - Equity Cracks Report Dated Oct 06 PDF
V-9901&V-9902 - Equity Cracks Report Dated Oct 06 PDF
V-9901&V-9902 - Equity Cracks Report Dated Oct 06 PDF
VALE018-0-06
During a scheduled shutdown of the Valero Aruba Refinery Coker Unit, an initial
inspection of two coke drums V9001 and V9002, the lower cone section of both
vessels were found t contain significant weld seam cracking on the middle (upper to
lower cone section) girth seam. The decision was made to have CBI, the on site
welding contractor, remove the cracked weld clad backing, repair the seam, and
complete the joint with a full circumference Post Weld Heat Treatment (PWHT.)
The Equity Engineering Group, Inc. (E2G) was contracted to assist in examining the
remaining weld seams and recommending a repair plan for the vessel internals that
would fit within Valeros scheduled timeframe for the turnaround.
It must be noted, that these weld repair recommendations are only temporary and
will not eliminate the cracking seen in drums V9001 and V9002. Until the E309
stainless steel weld metal is removed and replaced with a Nickel base material such
as ENiCrFe-2 or ENiCrFe-3, cracking will progress and lead to eventual through-wall
leak of one of the lower girth seams in V9001 or V9002.
Discussion:
Vessels V9001 and V9002 are typical examples of standard coke drum vessel
fabrication. They are of 1.25Cr-0.5Mo steel construction with an internal bonded
cladding of type 410S stainless steel for corrosion resistance. The weld seams of
the 1.25Cr-0.5Mo plate are covered on the inside of the vessel with a high alloy
weld material chosen to match the corrosion resistance of the 410S clad. Two
weld metal alloy types have been used for almost all weld metal cladding of coke
drum internal seams. These two are E309 type stainless steel and a Nickel base
type electrode, either ENiCrFe-2 (INCO-A) or ENiCrFe-3 (INCO 182.) Earlier
designs used the less expensive, and at that time more readily available, E309
stainless steel. Problems arose, however, after the coke drums experienced some
period of cyclic service. Due to the number of cycles and significant difference in
thermal expansion between the E309 type stainless steel weld metal and the
1.25Cr-0.5Mo base material, thermal fatigue cracking began in these weld joint
areas.
Cracking seen most often in the vertical weld seams of a drum would be transverse
type cracks running across the weld, usually from weld toe to weld toe. These
cracks would typically occur at approximately every 1 to 2 inches along the vertical
weld seam and almost always penetrated through the E309 weld metal cladding,
stopping just at the weld metal/ base metal interface. Cracking in the horizontal
girth seams of the drums would be seen at the weld toes, between the weld beads,
and some times as transverse cracking across the weld. Repair of this type of
thermal fatigue cracking by rewelding with E309 weld metal would only delay, not
eliminate future cracks. Elimination of cracking due to thermal fatigue stresses
requires a complete change in weld filler metal. Changing from E309 stainless to
Both of Arubas coke drum vessels, V9001 and V9002, were back welded with E309
stainless steel. They have seen approximately 10 years of cyclic service and as
expected, have begun to show signs of cracking. The areas of the T joints
appear to have the most significant cracking, probably due to the fact that they are
at a location where the residual stresses are pulling in two different directions.
Several of these T joints were blend ground to the point where the cracks had
been eliminated. The areas were then tested with copper sulfate to determine if
the grinding had penetrated the high alloy weld and entered the 1.25Cr-0.5Mo base
plate. There were indications that the cracks had penetrated greater than 1/8 of an
inch into the base metal For that reason, the T joints were listed as the number 1
repair priority. Repair welding was to be done using CBIs Temper Bead weld
procedure. To maintain the constant 350 to 400F weld preheat listed on the
procedure during welding, CBI planned to use local heating by electric resistance
coils. All alloy weld repairs would be done using ENiCrFe-3 (INCO 182) electrodes.
The horizontal girth seams, from C-7 down to C-3, had indications of longitudinal
toe cracks. Several spots on selected seams were blend ground to determine the
crack depth. These toe cracks measured only 1/16 to less than 1/8 inch deep into
the high alloy weld. The recommendation for these joints therefore, was to blend
grind out the existing cracks so that the stress level at that point would be lower.
The priority list of weld seams to complete was based on the fact that the lower
welds tend to have a higher stress level than those above. Seam C-7, however,
was near the top of the list because it was a field joint with an apparent weld
quality much less than that of the shop welded seams C-5 or C-6.
The scope of E2Gs work was limited to the on-site assistance and the final
disposition of recommended repairs is not part of this report. At the time these
recommendations were made, the coker vessels V9001 and V9002, still had
approximately two weeks of turnaround time remaining.
This report is the final version of the work product and E2G is not responsible for
any changes as a result of electronic transmission or any changes that the client
makes to the work product. E2G will maintain a hardcopy or a permanent
electronic copy (CD) of the work product in a client file and that copy will be
considered the final and complete document.
This completes the E2G assistance and if you have any questions, please call me on
216-658-4756.
Sincerely,
Robert R. Young
Principal Welding Engineer