edge of the attachment weld is not less than one-half application involving non-circumferential local heat-

the PWHT temperature, (where R =inside principal ing of a non-spherical surface must be evaluated on
radius of the sphere or end and t = the thickness). its own merit.
Further restrictions are imposed on the proximity of 6.3 Recommended Practice for Spherical Heads
the half temperature location to the junction with or Vessels
the shell (for heads) or other attachments. Welds in hemispherical heads or spherical vessels
API 510 (para. 5.25) explicitly allows non-circum- may undergo PWHT using the practices described in
ferential heating providing certain precautions and Figure 12. Note that the edge ofthe soak band must
requirements are applied. In general, these precau- be a minimum of 2 feet or 1.5 \/Rt, whichever is
tions and requirements include: greater, fron1 the head to shell junction or other
1) review and development of a procedure by a structural discontinuity. This practice has also been
pressure vessel engineer; applied to the larger radius section of double curva-
2) consideration of various factors such as ex- ture heads such that the edge of the soak band must
pected strains/distortion, material properties, be a minimum of 2 feet or 1.5 ~Rt, whichever is
thermal gradients, etc.; greater, from the knuckle transition region between
3) a minimun1 preheat temperature of 300°F the larger and smaller radius sections. Multiple
(149°C); nozzle or attachment welds may undergo PWHT
4) monitoring by a suitable number of thermo- using this practice, provided that the soak band is
couples; and increased in size to properly incorporate all of the
5) supplementary heating of intersected branch welds.
connections or attachments. 6.4 Analysis to Justify Use in Non-Spherical
Components
Although not explicitly stated in NBIC, alterna-
The acceptability of non-circumferential local
tive methods of postweld heat treatment per section
PWHT to non-spherical components must be deter-
RC-1103 (which are assumed to include non-circum- mined on a case-by-case basis. The approach out-
ferential heating) may be used if acceptable to the
lined in API 510 provides a framework for perform-
Authorized Inspector.
ing an evaluation. It is recommended that such an
6.2 Bases for Current Practices evaluation be documented in writing and include
Early published work regarding non-circumferen- consideration of all of the precautions and require-
tial PWHT appears to be limited to spherical vessels ments of API 510.
or heads [19]. Published and unpublished work The analysis should include the effect of all signifi-
generally utilize a circular heated region with sizing cant or major structural discontinuities (such as
and temperature gradient control based upon a nozzles, attachments, and head to shell junctions)
function of \/Rt as described above for AS 1210. and any mechanical loads which may be present
Recent work [44, 49] performed on a case by case during PWHT within the gradient control band.
basis showed that, by considering detailed thermal If an elastic analysis is performed, it has been
stress interactions with weld residual stresses, both suggested that thermal stress be treated as a pri-
the temperature level and heated area size for spot mary stress. It has further been suggested to con-
PWHT may be significantly reduced from typical sider only the contribution to membrane stress. AB a
codes requirements to achieve desired weld residual result, a possible acceptance criteria for such an elas-
stress reduction. From such analyses, it was shown tic analysis would be that the calculated membrane
that the reduction of weld residual stresses was stress at any point be limited to the yield strength of
mainly obtained by thermal compression effects ("hot the material for the temperature of that point.
compression") within the weld caused by spot PWHT. Of course, elastic-plastic analyses can be per-
If such effects can be generalized in the new PVRC formed to estimate strain and residual stress levels
joint industry project [45], significant economic ben- induced by spot PWHT. If such an analysis is to be
efits for using spot PWHT procedures can be realized performed, it is recommended that the weld residual
in terms of reduced spot heating area, hold time, and stress state should also be considered [e.g., 43, 44,
temperature level. In doing so, the likelihood of 48], since additional benefits can be demonstrated if
generating high thermal stresses at nearby geomet- the reduction of weld residual stresses is one of the
ric discontinuities should also be reduced. Any deci- objectives in applying spot PWHT. Acceptance crite-
sion to reduce temperature must be made with a ria for such analyses would have to be based upon
clear understanding of the objectives for PWHT. For specific knowledge of the component and its service
example, if tempering is an objective, a lower tem- envi ron men t.
perature, even with adequate reduced in residual
stress, may not be desirable.
7. Measurement of Temperature
At present, the application of generalized rules for [Refer to Section 8 and Annex C of AWS
non-circumferential local heating of non-spherical DlO.lO/DlO.lOM: 1999 for additional informa-
surfaces has not been recommended . Instead, each tion and recommendations.]

20 WRC Bulletin 452