Chapter 10
Prediction of Residual Life of Oil
and Gas Pipeline Corrosion Based
on Deep Learning
Wang Xinying, Zhang Huiran, Zhang Ruicheng, Zhao Bin, Huang Xuan,
and Zhang Ying
Abstract Corrosion of buried oil and gas pipelines will lead to perforation, leakage
and even ruptures of oil and gas pipelines, causing huge property losses. In order to
reduce gas oil leakage accidents caused by pipeline corrosion, a deep learning method
for predicting residual life oil and gas corroded pipeline is proposed. For this, the
author simulated pipeline corrosion in the laboratory, using acoustic emission detection method to detect the corrosion state of the pipeline, and with the deep learning
model, studied the corrosion rate changes of the pipeline. Finally, the corrosion rate
of the pipeline was predicted to further obtain the remaining life of the pipeline.
The final result proved that the method can accurately predict the remaining life of
corroded pipeline.
10.1 Introduction
Generally, China mainly uses pipelines to transport natural gas. Because of the
noise generated during transportation, there is little waste and almost no pollution.
However, since gas pipelines are distributed underground, it is difficult to be noticed
when gas pipelines are corroded or even perforated. It is easy to cause a large leakage
accident and cause great consequences. Therefore, it is particularly important to find
the corrosion status of the pipeline in time to predict the remaining life of the pipeline
[1]. At present, there are many prediction methods for the residual life of pipeline
corrosion, such as the residual life prediction model based on the finite element analysis method to predict the residual strength of the pipeline and predict the remaining
life of the pipeline, relying on a large number of corrosion detection data. Reliability probabilistic residual life prediction model for secondary corrosion detection
data and pipeline life prediction model based on artificial neural network theory that
requires measurement of various corrosion factors and corrosion results. In recent
years, the deep learning boom has emerged. The deep learning network structure
is a network with a high level of nonlinear computing [2]. The obvious difference
W. Xinying (B) · Z. Huiran · Z. Ruicheng · Z. Bin · H. Xuan · Z. Ying
School of Environmental and Safety Engineering, Changzhou University, Changzhou 213164,
China
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021
G. Shen et al. (eds.), Advances in Acoustic Emission Technology, Springer Proceedings
in Physics 259, https://doi.org/10.1007/978-981-15-9837-1_10
105
Prediction of Residual Life of Oil
and Gas Pipeline Corrosion Based
on Deep Learning
Wang Xinying, Zhang Huiran, Zhang Ruicheng, Zhao Bin, Huang Xuan,
and Zhang Ying
Abstract Corrosion of buried oil and gas pipelines will lead to perforation, leakage
and even ruptures of oil and gas pipelines, causing huge property losses. In order to
reduce gas oil leakage accidents caused by pipeline corrosion, a deep learning method
for predicting residual life oil and gas corroded pipeline is proposed. For this, the
author simulated pipeline corrosion in the laboratory, using acoustic emission detection method to detect the corrosion state of the pipeline, and with the deep learning
model, studied the corrosion rate changes of the pipeline. Finally, the corrosion rate
of the pipeline was predicted to further obtain the remaining life of the pipeline.
The final result proved that the method can accurately predict the remaining life of
corroded pipeline.
10.1 Introduction
Generally, China mainly uses pipelines to transport natural gas. Because of the
noise generated during transportation, there is little waste and almost no pollution.
However, since gas pipelines are distributed underground, it is difficult to be noticed
when gas pipelines are corroded or even perforated. It is easy to cause a large leakage
accident and cause great consequences. Therefore, it is particularly important to find
the corrosion status of the pipeline in time to predict the remaining life of the pipeline
[1]. At present, there are many prediction methods for the residual life of pipeline
corrosion, such as the residual life prediction model based on the finite element analysis method to predict the residual strength of the pipeline and predict the remaining
life of the pipeline, relying on a large number of corrosion detection data. Reliability probabilistic residual life prediction model for secondary corrosion detection
data and pipeline life prediction model based on artificial neural network theory that
requires measurement of various corrosion factors and corrosion results. In recent
years, the deep learning boom has emerged. The deep learning network structure
is a network with a high level of nonlinear computing [2]. The obvious difference
W. Xinying (B) · Z. Huiran · Z. Ruicheng · Z. Bin · H. Xuan · Z. Ying
School of Environmental and Safety Engineering, Changzhou University, Changzhou 213164,
China
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021
G. Shen et al. (eds.), Advances in Acoustic Emission Technology, Springer Proceedings
in Physics 259, https://doi.org/10.1007/978-981-15-9837-1_10
105