Hydrocarbon Processing - May 2021 - 29

Special Focus

Maintenance and Reliability
A. GUPTA and M. AL-MUWAIL, Kuwait Oil Co., Kuwait;
and R. DONNELLY and J. NICHOLAS, The Welding Institute,
Cambridge, UK

An investigation into pressure vessel fatigue
cracking caused by a tack weld
Pressure vessels and piping constitute
the maximum percentage of static equipment in any oil and gas facility. Separators,
drums, desalters, columns, reactors and
heat exchangers are only a few examples
of the numerous types of pressure vessels
used in the hydrocarbon industry. The
design of welds is critical to any pressure
vessel fabrication process. As part of weld
joint fit-up and proper alignment prior to
welding, tack welding is done between
the pressure vessel components (such as
the shell, dish ends and nozzles) prior to
the initiation of the actual welding process. This is a common practice during
pressure vessel fabrication and repair.
During the fabrication process and any
repair during service, the significance of
tack welds to join metallic parts cannot
be ignored. However, from a practical
perspective, tack welds are not always explicitly addressed in fabrication drawings
or quality control plans. In this article, the
authors present the detrimental effect of a
tack weld in a pressure vessel fabrication,
leading to fatigue cracking and the eventual breach of containment.
Background. A pressure vessel leak
happened in an oil and gas surface production facility. The subject vessel (a
degassing boot) was constructed as per
the ASME Sec VIII Div. 1 code. The vessel separates out/disengages the gases
from liquid streams before they enter the
downstream storage tank. Separation of
gases in the degassing boot (FIG. 1) helps
to maintain a stable interface level in the
tank, along with proper separation of produced water from crude in the tank.
The operating pressure was 0.361 psig
(maximum), with an operating temperature of 15.5°C-60°C. The process fluid

was hydrocarbon gas, with crude oil and
water. During operation, crude oil seepage was observed at the reinforcement
pad of the 8-in. NB riser nozzle.
The subject vessel was isolated from
service to assess the cause of the leak and
to carry out necessary repair work. During internal inspection, a crack-approximately 70 mm in length-was observed
on the shell (10-mm-thick ASME SA516
Grade 60 carbon steel) at a peripheral distance of approximately 16 mm from the
N1B nozzle internal bore. A region measuring 1.4 m x 1 m was removed from the
shell to facilitate a failure investigation of
the cracked region.
Objectives of the failure investigation. The following were the objectives of

the failure investigation:
* Identify the mode or mechanism
of cracking
* Identify the possible causes of
cracking, and comment on their
likelihood
* Provide short-term and long-term
actions for an immediate remedial
response and for future avoidance
of reoccurrence, as appropriate.

Approach. To achieve these objectives,
the following work items were carried out:
1. Upon receipt of the sample,
the appearance of the crack
and the surrounding area were
subjected to visual examination.
Photographs were taken to record
the observations.
2. The extent of cracking was
evaluated using phased array
ultrasonic examination.
3. The cracked region was cut down
and chilled in liquid nitrogen.

FIG. 1. Typical degassing boot.

Approximately half of the crack
was broken open to reveal the
crack surfaces, and these were
examined visually before and
after appropriate cleaning. The
cleaned fracture surface was
examined using scanning electron
microscopy. Photographs were
taken to record the observations.
4. A metallographic section
through the nozzle, shell and
reinforcement pad was taken to
characterize the crack's location
and morphology. This section was
mounted in Bakelite and polished
to a 0.25-µm diamond finish,
using standard metallographic
preparation techniques. The
section was then etched in 2%
nitric acid in alcohol (nital) to
reveal the microstructure and then
examined using high-power light
microscopy. Photographs were
taken to record the observations.
5. A Vickers hardness survey was
carried out on Weld 1, Weld 2, the
sampling weld regions, the heataffected zones (HAZs) and other
regions of interest by using a 5-kg
load. A microhardness survey was
Hydrocarbon Processing | MAY 2021 29

Hydrocarbon Processing - May 2021

Table of Contents for the Digital Edition of Hydrocarbon Processing - May 2021