of the finned tubes. In the field, joints were
wrapped with electrical resistance strip heaters
and insulation. Temperature monitoring and
control was by direct attachment of thermocouples
and multiple channel controllers and chart
recorders. The extra heat sink caused by the
tube fins and the internal venting of heat due to
chimney effects increased the difficulty of performing
PWHT. The fins increased the weight
of the 50 mm (2 in.) OD tubes to an equivalent
wall thickness of 13.7 mm (0.54 in.), more than
four times their nominal wall thickness and
effectively 70% thicker than the 8.1 mm (0.32
in.) wall thickness of the header pipe.
Feedwater Economizer Performance. One
of the harp assemblies leaked during hydrostatic
testing. The design required a 15° bend to form a
dogleg between the riser and the headers. No
heat was applied for bending. Brittle cracking
originated in the weld seams, partially because
the welds had been positioned on the outside of
the bend. The welds had a coarse-grained
microstructure that was indicative of brittleness
and an elevation in the nil ductility transition
temperature (Fig. 16). Repositioning the weld
seams to the neutral axis of the bend prevented
further problems.
A leak was found in a tube-to-lower-header
weld joint within seven months of initial operation.
However, the unit was returned to service
without repair, because the leak was small, and
PWHT of type 430 welds would be required if
repairs were made.
The unit was reinspected five months later
during a scheduled outage, and the leak had
increased in size. In addition, other tube-tolower-
header welds on this harp had pitting,
weld toe cracking, and many other liquid penetrant
test (PT) indications. A second harp on this
unit had pitting on the inlet side of its lower
header. The remaining harp had no indications.
The two corroded harps were bypassed, with
their feedwater flow redirected to the carbon
steel LP evaporator. The other unit in the plant
was inspected during a later outage after two
years of operation, and no PT indications were
found on its feedwater economizer lower headers.
The inlet halves of its lower headers had a
dark and wet deposit, which was apparently a
hygroscopic acidic condensate.
Test Results and Metallographic Examination.
Hardness testing was performed on a
tube sample in the as-received condition. The
bulk hardness obtained by the Vickers 500 g test
method averaged the coarse-grained, dualphase,
microhardness of the autogeneous seam
weld (Fig. 16). Vickers 50 g microhardness testing
differentiated between the two phases in the
weld fusion zone; the ferrite phase was 175
HV50 (88 HRB), and the martensite phase was
251 HV50 (23 HRC). The tube base metal hardness
met the 90 HRB maximum requirement of
ASME SA-268. Hardness testing required by
this standard is performed away from the weld
seam, in accordance with SA-450, paragraph
21.7 (Ref 31). Table 9 lists hardness values for
the tube-to-header welds.
A corroded tube-to-header weld sample
removed from the leaking header after seven
months of service was also hardness tested. Its
weld and HAZs, particularly the tube HAZ