Fuzzy Inference System Modeling for Corrosion Risk Management of Natural Gas Pipelines
Subject Areas : Fuzzy Optimization and Modeling Journal
Nazila Adabavazeh
1
,
Mehrdad Nikbakht
2
*
,
Atefeh Amindoust
3
,
Sayed Ali Hassanzadeh-Tabrizi
4
1 - Department of Industrial Engineering, Najafabad Branch, Islamic Azad University, Najafabad, Iran
2 - Department of Industrial Engineering, Najafabad Branch, Islamic Azad University, Najafabad, Iran
3 - Department of Industrial Engineering, Najafabad Branch, Islamic Azad University, Najafabad, Iran.
4 - Advanced Materials Research Center, Department of Materials Engineering, Najafabad Branch, Islamic Azad University, Najafabad, Iran
Keywords: Modeling, Fuzzy Inference System, Natural Gas Pipelines, Risk Management, Corrosion Risk, Corrosion Management.,
Abstract :
Due to the complexity of operating conditions, predicting corrosion in natural gas pipelines is a challenge, and therefore its use in the gas industry is problematic. So, the present study employs corrosion risk management in control systems and decision prediction with fuzzy inference systems to address the uncertain phenomena of critical factors affecting natural gas pipelines. In the study, 84 factors were derived and 14 critical factors were identified using the Lawshe approach. The Boehm risk management framework was used and the factors were analysed in pairs in a fuzzy inference system. The probability and consequences of each factor were determined according to the API 581 standard and the weight coefficient of each factor was determined using the FUCOM method. In this study, a fuzzy inference system with two fuzzy inference subsystems was developed to comprehensively address both aspects of evaluation and response. First, the system uses the API 581 standard risk matrix to identify the risk level, followed by the risk response strategy determination through the sensitivity risk priority index chart. The results show that at the first level of the system, the highest risk was classified as severe and harsh, while at the second level of the system, the maximum output behavior occurred during the wear-out phase of the pipeline. The proposed fuzzy inference system has the potential to significantly contribute to the effective management of pipeline corrosion risk and the economic productivity of the gas industry.
1. Adabavazeh, N., Nikbakht, M., Amindoust, A., & Hassanzadeh-Tabrizi ,S. A. (2024). The identification and analysis of pivotal factors influencing the corrosion of natural gas pipelines using fuzzy cognitive map. Engineering Failure Analysis, 166, 108806.
2. Agarwal, R., & Nishad, A.. (2023). A Fuzzy Mathematical Modeling for Evaluation and Selection of a Best Sustainable and Resilient Supplier by Using EDAS Technique. Process Integration and Optimization for Sustainability. 8. 1-10.
3. Ahmed, S., A., Akmar, M., A., Candra, K., J., Lashari, N., Sarwar, U., Jameel, S. M., Inayat, M., & Oladosu, T. L. (2022). A review on Bayesian modeling approach to quantify failure risk assessment of oil and gas pipelines due to corrosion. International Journal of Pressure Vessels and Piping, 200, 104841.
4. Aloqaily, A. (2018). Cross-Country Pipeline Risk Assessments and Mitigation Strategies. Gulf Professional Publishing, v, ISBN 9780128160077..
5. API Recommended Practice 581 (2019). Risk-Based Inspection Methodology, American Petroleum Institute, USA.
6. Azoor, R.M., Deo, R.N., Birbilis, N., & Kodikara, J. (2019). On the optimum soil moisture for underground corrosion in different soil types. Corrosion Science, 159, 108116.
7. Bai, Y., Wu, J., Ren, Q., Jiang, Y., Cai, J. (2023). A BN-based risk assessment model of natural gas pipelines integrating knowledge graph and DEMATEL. Process Safety and Environmental Protection, 171, 640-654.
8. Batbayar, K., Takács, M., & Kozlovszky, M. (2016).Medical device software risk assessment using FMEA and fuzzy linguistic approach: Case study, IEEE 11th International Symposium on Applied Computational Intelligence and Informatics (SACI). Timisoara, Romania. 197-202.
9. Boehm, B. W. (1991). Software risk management: principles and practices. IEEE Software, 8, 1, 32-41.
10. Cox, W. M. (2014). A Strategic Approach to Corrosion Monitoring and Corrosion Management. Procedia Engineering, 86, 567-575.
11. Fang, J., Cheng, X., Gai, H., Lin, S., & Lou, H. (2023). Development of machine learning algorithms for predicting internal corrosion of crude oil and natural gas pipelines. Computers & Chemical Engineering, 177, 108358.
12. Fink, S. (2021). Untersuchung zur Wasserstoffversprödung von Gasrohrleitungen; Technische Universität Graz, Graz, Austria.
13. Hagarová, M., Cervova, J, & Jaš, F. (2015). Selected types of corrosion degradation of pipelines. Koroze a Ochrana Materialu. 59. 30-36.
14. Huang, Y., Qin, G., & Yang, M. (2023). A risk-based approach to inspection planning for pipelines considering the coupling effect of corrosion and dents. Process Safety and Environmental Protection, 180, 588-600.
15. Jamshidi, A., Abbasgholizadeh, R, Samira , Ait-Kadi, D., & Ruiz, A. (2015). A comprehensive fuzzy risk-based maintenance framework for prioritization of medical devices. Applied Soft Computing. 32, 322-334.
16. Jiang, F., & Dong, Sh. (2024). Probabilistic-based burst failure mechanism analysis and risk assessment of pipelines with random non-uniform corrosion defects, considering the interacting effects. Reliability Engineering & System Safety, 242, 10978.
17. Jianxing, Y., Chen, Y., Yang, Y. Zh. (2019). A weakest t-norm based fuzzy fault tree approach for leakage risk assessment of submarine pipeline. Journal of Loss Prevention in the Process Industries, 62, 103968.
18. Li, C, & Xi, Z. (2019). Social Stability Risk Assessment of Land Expropriation: Lessons from the Chinese Case. International Journal of Environmental Research and Public Health. 16(20),3952.
19. Li, X., Wang, J., Abbassi, R., Chen, G. (2022). A risk assessment framework considering uncertainty for corrosion-induced natural gas pipeline accidents. Journal of Loss Prevention in the Process Industries, 75, 104718.
20. Liang, X., Ma, W., Ren, J., Dang, W., Wang, K., Nie, H., Cao, J., & Yao, T. (2022). An integrated risk assessment methodology based on fuzzy TOPSIS and cloud inference for urban polyethylene gas pipelines. Journal of Cleaner Production, 376, 134332.
21. Ma, D., Mao, W., Zhang, G., Liu, Ch., Han,Y., Zhang, X., Wang, H., Cen, K., Lu, W., Li, D., & Zhang, H. (2024). Dynamic risk assessment of gas pipeline operation process by fusing visual and olfactory monitoring. Journal of Safety Science and Resilience, 5(2), 156-166.
22. Mahmood, Y., Chen, J., Yodo, N., & Huang, Y. (2024). Optimizing natural gas pipeline risk assessment using hybrid fuzzy bayesian networks and expert elicitation for effective decision-making strategies. Gas Science and Engineering, 125, 205283
23. Muhammad, H., Zhang, T. & Nasser, M. (2023). Adoption of big data analytics for energy pipeline condition assessment. International Journal of Pressure Vessels and Piping, 105061.
24. Muhlbauer, W. K. (2006). Enhanced Pipeline Risk Assessment. WKM Consultasy.
25. Pamučar, D., Stević, Ž., & Sremac, S. (2018). A New Model for Determining Weight Coefficients of Criteria in MCDM Models: Full Consistency Method (FUCOM). Symmetry, 10, 393.
26. Prasanta, K, D, (2010). Managing project risk using combined analytic hierarchy process and risk map. Applied Soft Computing, 10(4), 990-1000.
27. Qin, M., Liao, K., He, G., He, T., Leng, J., & Zhang, Sh. (2023). Quantitative risk assessment of static equipment in petroleum and natural gas processing station based on corrosion-thinning failure degree. Process Safety and Environmental Protection, 172, 144-156.
28. Raeihagh, H., Behbahaninia, A., & Aleagha, M. M. (2020). Risk assessment of sour gas inter-phase onshore pipeline using ANN and fuzzy inference system – Case study: The south pars gas field. Journal of Loss Prevention in the Process Industries, 68, 104238.
29. Speight, J. G. (2015). Subsea and Deepwater Oil and Gas Science and Technology. Gulf Professional Publishing, 213-256.
30. Taghipour, M., Lashkaripour, G.R., & Ghafoori, M. et al. (2016). Evaluating the soil corrosion of Bushehr, Iran, based on a new classification system for corrosive soils. Anti-Corros. Methods Mater. 63 (5), 347–354.
31. Velázquez, J.C., Valor, A., & Caleyo, F. (2024). Statistical study of localized internal corrosion defects in oil and gas pipelines through sampling inspection. Process Safety and Environmental Protection, 186, 566-576.
32. Wang, F. , Xu, J., Xu, Y., Jiang, L., & Ma, G. (2020). A comparative investigation on cathodic protections of three sacrificial anodes on chloride-contaminated reinforced concrete. Construction and Building Materials. 246. 118476.
33. Wang, W., Shen, K., Wang, B., Dong, Ch., Khan, F., & Wang, Q. (2017). Failure probability analysis of the urban buried gas pipelines using Bayesian networks. Process Safety and Environmental Protection, 111, 678-686.
34. Woldesellasse, H., & Tesfamariam, Solomon (2023). Consequence assessment of gas pipeline failure caused by external pitting corrosion using an integrated Bayesian belief network and GIS model: Application with Alberta pipeline. Reliability Engineering & System Safety, 240, 109573.
35. Xu, W., Yuan, K., Li ,W. , & Ding, W. (2023). An Emerging Fuzzy Feature Selection Method Using Composite Entropy-Based Uncertainty Measure and Data Distribution. IEEE Transactions on Emerging Topics in Computational Intelligence, 7, 1, 76-88.
36. Yang, Y., Faisal, Kh., Thodi, P., & Abbassi, R., (2017). Corrosion induced failure analysis of subsea pipelines. Reliability Engineering & System Safety, 159, 214-222.
37. Zhang, Y., Chen, L., Qin, H. Jiang, Z., Du, Y., & Lu, M. (2024). Research on fluctuation characteristics of AC interference parameters and corrosion behavior of buried gas pipelines. International Journal of Electrochemical Science, 19(1), 100396.
38. Zhao, Lei, Qi, Gang, Dai, Yong, Ou, Hongxiang, Xing, Zh., Long, Zhao, & Yan,Y. (2023). Integrated dynamic risk assessment of buried gas pipeline leakages in urban areas. Journal of Loss Prevention in the Process Industries, Volume 83, 105049.
Fuzzy Inference System Modeling for Corrosion Risk Management of Natural Gas Pipelines
Nazila Adabavazeha,b, Mehrdad Nikbakhta,b*, Atefeh Amindousta,b, Sayed Ali Hassanzadeh-Tabrizic
a Department of Industrial Engineering, Najafabad Branch, Islamic Azad University, Najafabad, Iran.
b Modern Manufacturing Technologies Research Center, Department of Industrial Engineering, Najafabad Branch, Islamic Azad University, Najafabad, Iran.
c Advanced Materials Research Center, Department of Materials Engineering, Najafabad Branch, Islamic Azad University, Najafabad, Iran.
A R T I C L E I N F O |
| A B S T R A C T Due to the complexity of operating conditions, predicting corrosion in natural gas pipelines is a challenge, and therefore its use in the gas industry is problematic. So, the present study employs corrosion risk management in control systems and decision prediction with fuzzy inference systems to address the uncertain phenomena of critical factors affecting natural gas pipelines. In the study, 84 factors were derived and 14 critical factors were identified using the Lawshe approach. The Boehm risk management framework was used and the factors were analysed in pairs in a fuzzy inference system. The probability and consequences of each factor were determined according to the API 581 standard and the weight coefficient of each factor was determined using the FUCOM method. In this study, a fuzzy inference system with two fuzzy inference subsystems was developed to comprehensively address both aspects of evaluation and response. First, the system uses the API 581 standard risk matrix to identify the risk level, followed by the risk response strategy determination through the sensitivity risk priority index chart. The results show that at the first level of the system, the highest risk was classified as severe and harsh, while at the second level of the system, the maximum output behavior occurred during the wear-out phase of the pipeline. The proposed fuzzy inference system has the potential to significantly contribute to the effective management of pipeline corrosion risk and the economic productivity of the gas industry.
|
Article history: Received 5 August 2024 Revised 21 September 2024 Accepted 30 September 2024 Available online 12 October 2024 | ||
Keywords: Modeling Fuzzy Inference System Natural Gas Pipelines Risk Management Corrosion Risk Corrosion Management. |
1. Introduction
Natural gas, a crucial part of the energy supply chain, has seen a significant expansion of pipeline networks, which has contributed to the economic development of countries. However, this growth has also brought challenges to pipeline safety and reliability [22]. The International Energy Agency (IEA) predicts that demand for natural gas will continue to grow for several decades, leading to an increase in the total length of pipelines [31].
Corrosion is a natural phenomenon that affects the reliability of pipelines and can lead to natural gas leaks, resulting in a chain of flammable gasses, fires, and explosions. Natural gas pipeline accidents caused by corrosion can have devastating consequences for society, the economy, and the environment. Therefore, it is essential to conduct risk assessments to understand the probability and impact of natural gas pipeline corrosion accidents and develop effective emergency planning [19]. All stakeholders are concerned with assessing and analysing risks associated with pipelines. However, the lack of a comprehensive and accurate dataset often leads to unpredictable risk assessment, exacerbating the consequences of such accidents [22]. The limited availability of public datasets on pipeline corrosion further complicates corrosion prevention [11].
Risk management includes hazard assessment, risk analysis, risk assessment, risk control, supervision and continuous improvement. As risk levels reach an endurable limit, the process will be established for supervision and continuous improvement to make sure that risk levels will be remained manageable. The whole cited processes included ranging from hazard analysis and risk documentation to continuous improvement [4]. Corrosion management refers to a risk-based approach for locating corrosion, choosing an up-to-the-minute precise tool which is able to figure out the beginning and spread a local attack of corrosion that provides an opportunity by inspected consequences for diminishing the intensity and the time of attack [10].
Corrosion management, which has been developed in recent years, is a crucial tool for making sure of safety and reliability in pipelines; nonetheless, developing in precautionary corrosion management is demanding owing to the limited availability of data in targeted pipelines and corrosion causes. Developing an active corrosion management approach, based on a forecast framework, will bring about facilitating in active maintenance strategies and improving asset management in the pipeline industry.
Several studies have conducted risk assessments based on probabilistic models, such as Petri nets, Markov chains, Monte Carlo simulations, fault trees, and Bowties [3]. However, these models are based on unrealistic assumptions and are static in nature. To address the uncertainty and ambiguity in pipeline risk assessment, the use of fuzzy set theory is recommended [22]. Fuzzy logic can play a useful role in accurate and correct risk management decision-making and systematic risk analysis using mathematical methods. Fuzzy inference system models adapt the process resulting from experience and observations to the performance of a particular system to establish management and engineering knowledge. The purpose of this expert system is to obtain and apply knowledge and inference methods to achieve a more advanced solution to difficult problems that require a significant effort from scientific experts to solve.
The results of several studies on the risk assessment of gas pipelines have relied on expert opinion rather than data [7]. In this regard, the current research focuses on modeling the fuzzy inference system for managing the corrosion risk of natural gas pipelines for the strategic capability of the gas industry. This model can be used to prevent corrosion and improve risk management processes to increase safety. The purpose of this study is to develop practical knowledge in the field of corrosion risk assessment and to guide its practical application.
The remainder of this paper is organized as follows: Section 2 presents the literature review, Section 3 explains the research methodology, Section 4 presents the numerical results of the research and the research model, and Section 5 concludes the study and makes suggestions for future work.
2. Literature review
The natural gas industry proactively deals with threats to pipelines, especially corrosion. Pipeline organizations need to make quick and cost-effective decisions regarding damaged pipelines and corrosion damage to prevent incidents from escalating into high-consequence areas. The aim of this study is to provide a comprehensive structure for managing the corrosion risk of natural gas transportation pipelines through a review of the literature. In the following sections, the studies conducted in this area are discussed:
Velázquez et al. [31] analysed the internal corrosion defects of oil and gas pipelines based on inspections. If the distribution of corrosion defect characteristics is known, historical damage data from corroded oil and gas pipelines can help to estimate risk and reliability using statistical models. This paper presents a field study on the characteristics of corrosion defect characteristics in Mexican oil and gas pipelines. The investigation revealed that the depth of corrosion damage exhibits a two-dimensional behavior. The fluid properties are crucial for the statistical description of the internal corrosion damage. The results showed that the Bayesian approach allows statistical estimation of the depth of internal corrosion defects using limited information from the pipelines.
Zhang et al. [37] investigated the characteristics of AC fault parameters and the corrosion behavior of buried gas pipelines. An internal simulation experiment was conducted to investigate the effects of AC current density and cathodic protection potential on AC corrosion. The results showed that with the same cathodic protection, a higher AC current density led to a higher dynamic AC corrosion rate. Finally, the study discussed the hazards of the AC interference corrosion mechanism.
Jiang and Dong [16] analysed the mechanism of pipeline burst failure due to corrosion considering mutual effects. Using nonlinear finite element analysis and Monte Carlo simulation, an integrated method was developed to describe the failure behavior of pipelines with random defects. The failure mechanism and reliability of the pipelines were analysed and the performance changes were discussed.
Mahmood et al. [22] have developed a hybrid fuzzy Bayesian network to optimize the risk assessment of natural gas pipelines. In this paper, a fuzzy framework based on historical event data is presented. The diagnosis revealed that pipeline materials and their service life pose a significant threat to the integrity of pipelines in the Midwestern United States. This study underscores the importance of a targeted mitigation strategy to minimize risks in the pipeline network.
Ma et al. [21] conducted a quantitative risk assessment of gas pipeline operations. They developed a 3D risk assessment model that integrates gas leakage with risk-related sensitivity and a separate 3D risk assessment model that combines visual risk factors with foreseeable risk. In addition, they introduced a quantitative expression mode for visual risks based on the radar map risk matrix method. Simulation scenarios based on field data showed that the risk assessment method of visual-olfactory integration more accurately reflects the dynamic risk level of the operation process compared to simple visual safety factor monitoring.
Zhao et al. [38] evaluated the integrated dynamic risk of gas pipeline leakage in urban areas. They considered the relationship between the probability of failure and the failure rate and used a mechanical analysis model to determine the corrosion growth model. The study indicates a general development process from pipeline failure to accidents. They then modeled and analysed the consequences of different types of accidents involving buried gas pipelines.
Bai et al. [7] presented a novel risk assessment model for natural gas pipelines using Bayesian networks and DEMATEL. This model includes a knowledge graph, DEMATEL, and a network. DEMATEL was implemented to determine complicated links within the causal network and convert it into a Bayesian network structure. The Bayesian analysis facilitates the creation of a possible causal model for gas pipeline accidents and enables the prediction of possible consequences and the optimization of risk mitigation strategies. The proposed model provides a more objective approach to improve safety management and reduce risks associated with natural gas pipelines and other digital-age facilities.
Qin et al. [27] assessed the equipment risk associated with oil and natural gas processing stations based on the thinness of pipeline walls. The researchers categorized the corrosion rates and failure consequences into five levels and developed a risk assessment matrix for the oil and gas processing station to avoid the risk assessment error caused by a mismatch in the failure database. The study concluded that the risk level of corrosion and thinning failure is moderate considering the current operational and management status of the station.
Fang et al. [11] developed machine-learning algorithms to predict the internal corrosion of natural gas pipelines. The authors simulated thousands of corrosion scenarios for crude oil and natural gas pipelines and developed two algorithms to predict the internal corrosion rate based on operating conditions. The results showed that the proposed algorithms effectively predicted the internal corrosion rates of pipelines.
Huang et al. [14] introduced a risk-based approach for pipeline corrosion inspection. The study used a dynamic Bayesian network, in-line inspection data, and corrosion knowledge to probabilistically model time-dependent pipeline failures. The researchers considered two failure scenarios (leakage and bursting) and estimated the probability of pipeline failure over time and the consequences of failure using the Monte Carlo simulation method. The authors also introduced a model for optimizing maintenance with the lowest total cost.
Woldesellasse et al. [34] used Bayesian belief network and a geographic information system (GIS) model to assess the consequences of gas pipeline corrosion. The authors proposed the integration of these two approaches to estimate the consequences of gas pipeline failure on society and the environment. The researchers employed pipe specifications, failure mode, and population density as inputs to calculate the losses. An event tree was used to represent all potential consequences of a gas release, and the social and environmental consequences of corrosion were estimated based on empirical equations and subjective judgment. The findings of this study suggest that the combination of these two approaches is an effective tool for estimating the severity of a pipe burst in a given area using the available information.
Liang et al. [20] investigated the integrated risk of polyethylene gas pipelines based on fuzzy TOPSIS and inference for urban areas. A fault tree was utilized to determine the risk status, and the results of the risk assessment of the polyethylene gas pipelines were presented in the form of an inference. The case study demonstrated that the pipeline risk was moderate and the management of time-related indicators and third-party damage was a priority.
Li et al. [19] developed a risk assessment framework that considers the uncertainty of corrosion incidents in natural gas pipelines. This approach was developed to uncover the uncertainties in the probability of failure and the impact of pipeline leakage. The identified parameters were described with a series of probability density functions. Finally, the Monte Carlo method was employed to solve the generated models. The sample of the study shows that the proposed framework is a valuable tool for the risk assessment of accidents due to corrosion of natural gas pipelines.
Raeihagh et al. [28] developed a pipeline risk assessment model using an artificial neural network and a fuzzy inference system. The fuzzy model was created using MATLAB software and the results were utilized to develop the ANN implemented in phases 9 and 10 of the South Pars gas field. The proposed model demonstrated a higher level of accuracy and reliability in assessing pipeline risks than previous approaches.
Figure 1 shows the number and geographical distribution of the relevant studies. Figure 1 has demonstrated the research gap and remind the significance of developing a risk management framework in pipeline corrosion, especially in Iran with its dry and ultra-dry climate.
|
Figure 1. Number and geographical distribution of the literature reviewed in this study |
Due to the critical nature of pipeline integrity, various studies have been conducted to identify and assess risks, which has led to a lack of consensus regarding corrosion management. This makes it challenging to provide a comprehensive corrosion prediction system. To address this issue, it is necessary to present comprehensive results that can improve the integrity of pipelines in the oil and gas industry. The literature review has shown that the complexity of the corrosion system is due to the multitude of factors at play. Furthermore, the variety and extent of factors affecting corrosion are critical issues that need to be comprehensively identified and assessed in a timely manner. Conventional modeling and mathematical calculation methods, such as classical modeling and deterministic analysis tools, are limited in their ability to address the uncertainty of human judgments, evaluations, and decision-making. Fuzzy logic, on the other hand, is a powerful tool for dealing with uncertainty, solving problems, and reducing the influence of personal judgment on decisions. The main objective of this research is to develop a method for pipeline risk management based on a fuzzy inference system. The method used in this study to predict and evaluate the corrosion of a fuzzy inference system is a two-step process. The ambiguity of data, lack of sufficient data, and statistical extrapolation are reasons for using fuzzy logic. With the aim of comprehensively identifying the corrosion factors and risk assessment, this study aims to improve the effectiveness and efficiency of the corrosion management system through proper management of these factors.
The literature review shows there are significant concepts in gas pipelines like “risk, corrosion, management”. Corrosion causes disorders in correct function of the continuity in transferring gas and it affects the safety level of pipelines. Risk management aids the reduction in risk influences and to control; also, it prevents the occurrence of disorders in gas pipelines. Moreover, the literature review describes this subject that there is an array of capabilities in pipeline corrosion management, which enhances responding ability to risks in pipelines. In the considered literature review, there are not any frameworks which provide a comprehensive attitude in detecting, assessing and responding to risks; therefore, the current research has been conducted with the aim of completing research gaps. The framework for fuzzy inference systems is plain and well-described. Data is analysed by fussy inference systems (FIS) for adapting to better situation, following the maintenance conditions and classifying the behaviour of pipelines. A FIS based system is able to detect risks, by enabling an intelligent alarm system, to offer corrective measures with the aim of reducing costs of maintaining and repairing and protecting more adapted to external environment and more efficient pipelines. The precision and the accuracy of achieved yields brings about the reduction in costs, time and human resources for protecting and maintaining pipelines and the increase in efficiency in pipelines.
3. Methodology
This study was practical in nature and descriptive in its approach to data collection. It was conducted during the winter of 2024. The target population of the study consisted of corrosion experts. The Lawshe method was used to select a sample of 30 experts with at least 10 years of experience in the fields of engineering, materials, and metallurgy. Risk management is a systematic process that involves identifying, analysing, and responding to risks with the aim of maximizing positive consequences and minimizing negative consequences [26]. Boehm [9] describes risk management as a two-step process that includes risk assessment, i.e. identifying, analysing, and prioritizing risks, and risk control, i.e. planning and monitoring. Boehm's risk management process is shown in Figure 2. Bohem considers risk management as a two-step process of “risk assessment based on recognition, analysis and prioritizing risks” and “risk control, based on providing planning and supervision” in accordance with Figure 2 [9].
Figure 2. Boehm's risk management [9]
The first step is to identify the components and indicators of the model. After identifying the main risks, the next step is to compare and analyse the criteria in pairs and assess the frequency of occurrence and consequences based on the API 581 standard [5]. Finally, the FUCOM approach is used to determine the significance of the identified critical factors. The FUCOM technique can be implemented in three steps using a single-objective nonlinear mathematical model. In the first step, the evaluation factors based on expert opinions are ranked according to their importance. In this step, the criteria are defined from the most important to the least important. If two or more criteria have the same importance coefficient, an equal sign is used between them [25].
(1) | Cj(1)>Cj(2)>…>Cj(k) |
In the second step of the process, a comparative analysis is performed to assess the relative merits of the Cj(k) criterion versus the Cj(k+1) criterion.
The final weight values for the criteria were then calculated in the third step, which were determined by two conditions.
· The first condition involves calculating the ratio of the final weight values of the criteria based on the preferences of the compared criteria in the second step.
· The second condition requires the final weight value to satisfy certain transfer conditions.
Using a nonlinear single-objective programming model, we have (Equation (2)):
Min s.t.
| (2) | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Figure 3. Flowchart of research implementation |
4. Results
4.1. Identification of the criteria for corrosion risk assessment of pipelines
Pipeline networks are always at risk due to various factors, including corrosion. Effective decision-making requires the evaluation and determination of the functional value of the phenomena under study, which implies the identification of key indicators. The assessment of the critical factors of corrosion will have a significant impact on the determination of net strategies. Effective risk management requires the identification and ranking of critical risks to mitigate their consequences. The identified risks were used to develop a response plan to improve the corrosion situation. Although various studies have investigated the critical factors of corrosion, it is impossible to consider all factors in corrosion risk management, and some factors are commonly overlooked although they are important to facilitate the design of fuzzy inference systems and avoid complexity and multiplicity. Fuzzy laws require the identification of critical factors. The corrosion process has been described in different ways in different studies. The crucial factors identified in this study are presented in Table 1 based on expert opinions.
In general, the literature and expert opinions show that the criteria "long pipeline service life, in-line inspection [23], inspection frequency, soil resistivity, soil pH, soil moisture, redox reduction-oxidation potential [19], cathodic protection, pipeline coating, hydrogen sulfide gas, carbon dioxide gas, lack of injection of corrosion inhibitors [17,33,36], soil salt [19,33], ambient temperature [33] has special attention compared to other criteria identified as input for the fuzzy inference system. Since sweet natural gas has no potential for internal corrosion and appropriate measures are taken, the factors related to internal corrosion are removed from the identified factors. There is a huge number of elements in various studies which has been described as effective components in gas pipelines corrosion that critical factors, in accordance with elite’s notion, were identified in according to Table 1.
Table 1. Critical factors of corrosion [1]
Factor symbol | Critical factor for corrosion of natural gas pipelines | Sour natural gas | Sweet natural gas |
C1 | Service life of pipelines | n | n |
C2 | Frequency of inspection | n | n |
C3 | In-line inspection | n | n |
C4 | Cathodic protection | n | n |
C5 | Pipeline coating | n | n |
C6 | Soil resistivity | n | n |
C7 | Soil salt | n | n |
C8 | Soil moisture | n | n |
C9 | Redox reduction-oxidation potential | n | n |
C10 | pH value of the soil | n | n |
C11 | Operating temperature | n | n |
C12 | Hydrogen sulfide gas | n | r |
C13 | Carbon dioxide gas | n | r |
C14 | Injection of a corrosion inhibitor | n | r |
4.2. Analysis of the criteria for corrosion risk assessment of pipelines
Determining the threshold values for these indicators is a complex process that often requires the solution of optimization or laboratory problems. In this study, the thresholds for the risk assessment indicators were established through the research and laboratory results. Some assessment indicators such as "in-line inspection, inspection frequency, corrosion inhibitor injection, coating, cathodic protection, and acid gas separation" were established without prior on-site investigations to determine their thresholds. These limits were set by research experts. The qualitative forms of acid gasses, including CO2 and H2S, were determined under the general title of "acid gas". The assignment of factors in the fuzzy inference system model in this study is based on Table 2, which is presented below.
For all factors, for the verbal variable "very harsh" class 5, for the verbal phrase "harsh" class 4, for the verbal phrase "relatively harsh" class 3, for the verbal phrase "mild harsh" class 2 and for the verbal phrase "very mild" Class 1 is defined. The variables have a range of 5 or 3 degrees from 1 to 9 so the verbal variable "very harsh" has a numerical value of 9 and "very mild" has a numerical value of 1. The numerical values of these verbal expressions are listed in Table 3.
Table 2. Factors Affecting Corrosion
Verbal variable | Soil resistivity (m2) | Soil moisture (%) | Soil salt (%) | Redox potential (mV) | |||||
Very harsh (VH) | <10 | 60-85 | >0.75 | <100 | |||||
Harsh (H) | 10-20 | 50-60 | 0.15-0.75 | 100-200 | |||||
Relatively harsh (M) | 20-50 | 25-50 | 0.05-0.15 | 200-400 | |||||
Mild harsh (II) | 50-100 | 1-25 | <0.05 | >400 | |||||
Very mild harsh (I) | >100 | - | - | - | |||||
| [13] | [30] | [19] | [30] | |||||
Verbal variable | Inhibitor | Cathodic protection | Inspection frequency | Coating | |||||
Very harsh (VH) | No inhibitor used | No effect | Unplanned | No/improper coating | |||||
Harsh (H) | No inhibitor used | Low effect | Rarely | No/improper coating | |||||
Relatively harsh (M) | No inhibitor used | Relatively low effect | Delayed | Coating needs to be replaced | |||||
Mild harsh (II) | Inhibitor used | Effective | Timely | Proper coating | |||||
Very mild harsh (I) | Inhibitor used | Proper effect | Proactive | Proper coating | |||||
| Expert opinions | Expert opinions | Expert opinions | Expert opinions | |||||
Verbal variable | Acid gas | Pipeline inspection | Temperature | pH | Lifetime | ||||
Very harsh (VH) | No separation | No known inspection procedure | High/very low | <4.5 | 30~50 Years | ||||
Harsh (H) | Incomplete separation | No inspection procedure | High/very low | 4.5-5.5 | 30~50 Years | ||||
Relatively harsh (M) | Incomplete separation | Occasionally | Other | 5.5-7 | 15~30 Years | ||||
Mild harsh (II) | Complete separation | Full inspection | Other | 7-8.5 | 0~15 Years | ||||
Very mild harsh (I) | Complete separation | Intelligent inspection | Other | >8.5 | - | ||||
| Expert opinions | [15] | Expert opinions | [19] | [14] | ||||
In addition, high and very low temperatures increase the corrosion rate, which is included in the fuzzy model according to Table 3.
Using fuzzy logic, the evaluation indicators were considered as inputs to the fuzzy control system for pairwise comparisons, and a corresponding fuzzy membership function was defined for each. A total of 78 pairwise comparisons between corrosion evaluation indices based on secretions were investigated, and owing to the limitations in presentation, only crucial observations have been described.
As for soil corrosion factors, important comparisons between soil parameters such as "soil resistivity, soil moisture, soil salinity, pH, and redox potential" are shown in Figures 4-8. It has been observed that pipelines exposed to harsh soils for longer than their lifetime show an increasing corrosion trend, as shown by Azoor et al. [6].
This emphasizes the importance of developing a reliable approach for estimating smart management of pipeline corrosion. In addition, the degree of soil corrosiveness is determined by the maximum correlation between the parameters "redox potential, soil resistivity, pH, moisture, and soil salinity" based on five levels, as shown in Table 2.
Table 3. Evaluation of factors based on verbal variables [8]
Verbal variable | Corresponding fuzzy numbers |
VH | (8.5,10,10) |
H | (6,7.5,9) |
M | (3.5,5,6.5) |
L | (1,2.5,4) |
VL | (0,0,1.5) |
|
|
Figure 4. Corrosion risk based on soil resistivity and pipeline lifetime | Figure 5. Corrosion risk based on soil moisture and pipeline lifetime |
|
|
Figure 6. Corrosion risk based on soil salt and pipeline lifetime | Figure 7. Corrosion risk based on redox potential and pipeline lifetime |
|
|
Figure 8. Corrosion risk based on soil pH and pipeline lifetime
| Figure 9. Corrosion risk based on cathodic protection and pipeline lifetime
|
Cathodic disbondment is widely recognized as the main cause of coating damage on coated pipelines. Effective monitoring of cathodic protection can significantly reduce the occurrence of cathodic disbondment and maintain the long-term integrity of the pipeline, as shown in Figures 9-11 and supported by the research of [32].11 As shown in Figure 11, inspection plays a critical role in pipeline service life. Corrosion monitoring is a valuable strategy in the fight against corrosion and offers significant economic benefits to the gas industry.
The synergistic effect of corrosive acid gasses such as H2S and CO2 significantly increases the corrosion rate and often leads to the corrosion of pipelines. The use of inhibitors helps to prevent the expansion of the area of acid gas corrosion, as shown in Figure 12. However, corrosion inhibitors can be ineffective. To increase their effectiveness, it is recommended to propose deterrents that do not consider environmental aspects or use prohibited substances, which is consistent with the findings of Fink et al. [12] and Qin et al. [27].
|
|
Figure 10. Corrosion risk based on cathodic protection and coating | Figure 11. Corrosion risk based on cathodic protection and inspection |
|
|
Figure 12. Corrosion risk based on acidic and inhibitory gasses | Figure 13. Corrosion risk based on inhibitors and pipeline lifetime |
Based on the results shown in Figures 4-13, it is clear that the probability of corrosion can only be reduced if the soil corrosion factors, inhibitor, coating, and cathodic protection are optimal.
4.2.1. Determining the Importance and Weight of the criteria for corrosion risk assessment of pipelines
4.2.1.1. Sour natural gas
A key aspect of analysing the risk assessment factors is evaluating the weight coefficient of each factor. The ranking of the criteria based on expert opinion and their importance is shown in Table 4. A 9-point scale was used to prioritize the criteria and the experts' preferences were used to establish comparative priorities.
Table 4. Comparative prioritization of critical factors of natural gas pipeline Corrosion by experts
C10 | C7> | C11> | C8> | C6= | C9> | C12> | C13> | C14> | C4> | C5= | C3> | C2= | C1> |
1 | 2 | 3 | 4 | 4 | 5 | 5 | 6 | 6 | 7 | 7 | 8 | 8 | 9 |
Following the prioritization of the criteria by the experts, weight coefficients wk were calculated based on the comparative priorities.
|
|
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|
|
|
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|
(3) | Min c s.t.
Table 5. Weight of the natural gas pipelines corrosion assessment factors
4.2.1.2. Sweet natural gas Based on the experts' priorities for the criteria, the weight coefficients wk were calculated according to the comparative priorities.
|
