Abstract
Petrochemical plant feedstock stability is an important factor in ensuring consistent
production volumes and minimal maintenance intervention. Thus, detailed design and
in-line inspections are critical to reduce the risk of failure due to pre-existing or newly
developed cracks. In this work, a pipe specimen is designed according to ASME B31.4
and sound engineering practices. The pipe specimen final dimensions were a Nominal
Pipe Size (NPS) 8 (outer diameter: 219.1mm) pipe, with a wall thickness of 2.77mm
(schedule 5 pipe) and a length of 1000mm. Based on these dimensions, a numerical
model was developed in Abaqus and validated with published data. The original ASME
B31G and modified ASME B31G equations were used to determine the estimated
failure pressure and were then compared to the FEA model with varying crack length
and crack depth. The original ASME B31G calculations were found to be more
conservative than the modified equations. For a crack length of 40mm, there is a
15.37% difference between the estimated failure pressures from the original equations
and the numerical model whereas with the modified equations there is a 3.74%
difference. With the increase in crack length to 100mm, the percentage difference is
16.69% and 8.12% respectively. For a crack depth of 1mm, there is a 15.91%
difference between the estimated failure pressure from the original equations and the
numerical model where with the modified equations there is a 5.26% difference. With
the increase in crack depth to 2mm, the percentage difference is 12.55% and 13.13%
respectively. This shows that the original code equations are conservative however as
there is in increase in length or depth, the modified equations become more
conservative. The pipe specimen highlighted that there is room to be less
conservative. However, it depends on the user’s risk and the potential repercussions
of failure. The use of the AMSE B31G shows to be less attractive for elongated metal
losses since repairing the pipeline is required. It is also apparent that the major benefit
of the modified code equations is realised for small crack lengths. The pipe specimen
was then subjected to three different burial depths. The burial depths of 900mm and
1200mm are based on recommendations from ASME B31.4 and the burial depth of
1500mm was selected to determine the effect of deeper burial depths of pipelines. The
soil loading provides an interesting stress level pattern; without soil loading, a balloon
shape stress distribution profile forms around the crack but with soil loading, the stress
iii
levels form an umbrella shape especially in varied crack lengths. The comparison
between the overall stress distribution for the varying crack lengths and depths with
soil loading reveal that the crack depths have significant impact on the stress patterns.
The stress patterns for the effect on crack depth only changed slightly compared to
that of varying crack length. The deeper the soil depth, the higher the internal pressure
the pipe specimen can withstand, this factor does depend on crack orientation in
relation to the soil loading as well as crack geometry in relation to the original wall
thickness. These results clearly show that burying a line pipe has the potential to
lessen the impact of the internal pressure on the crack stability. Applying adequate
compaction force during construction, as opposed to deeper trench, could be a cost
effective to achieve such stability. In conclusion, a numerical model was successfully
developed. The study showed that the effect of crack depth is more significant than
crack length. Soil loading increases the load on the buried pipe which assists in crack
arrest, but crack orientation may result in the opposite effect. Soil loading has a
minimal effect on deeper cracks. Recommendations include conducting a physical
experiment and the comparison of crack orientations to provide more detail on the
effect of soil loading. More site realistic soil distributions may assist in getting a more
accurate result. The effect of material selection is a limited approach as there is a
minimal number of pipeline materials, but the use of high-density polyethylene (HDPE)
can be investigated. The effect of traffic loading and seismic activity may also be
explored to determine the effect of external factors on flawed pipelines.