Leak Detection Research Articles

Artificial Intelligence based Leak Detection in Blended Hydrogen and Natural Gas Pipelines

In the transition towards a hydrogen-based energy system, the strategic use of pipelines is crucial for efficient hydrogen distribution. Leveraging existing natural gas pipelines to carry a mix of hydrogen and natural gas offers a cost-effective alternative to building new infrastructure. This study explores the development of a leak detection system compatible with existing pipelines, specifically tailored for the blended hydrogen and natural gas mix. Given the scarcity of leak data for blended hydrogen-natural gas pipelines, the study introduces a Real-Time Transient Model (RTTM) for blended gases, simulating leak dynamics to generate necessary data. Additionally, a leak detection system (LDS) is developed using a fusion of Convolutional Neural Network (CNN) and Explainable Artificial Intelligence (XAI) through Adaptive Neuro-Fuzzy Inference Systems (ANFIS). This LDS framework overcomes the “black box” issue common in AI-driven systems, enabling reliable detection. The integration of Explainable and traditional AI techniques holds promising implications for blended hydrogen pipelines by enhancing the safety and efficiency of hydrogen transportation, thereby mitigating economic and environmental impacts, and addressing public concerns.

Underground Hydrogen Storage: Comparison of High-pressure Hydrogen, Liquid Hydrogen, and Ammonia

One of the alternatives for effective storage of irregularly produced renewable energy is hydrogen energy. In order to realize a hydrogen-based society, not only environmentally friendly production of hydrogen but also effective storage is very important. Underground hydrogen storage technology is a technology that has evolved from the technology for storing natural gas underground, and includes waste gas fields, salt domes, aquifers, and rock cavities. When stored underground, hydrogen is converted into high-pressure gaseous hydrogen, liquefied hydrogen, and ammonia. Liquefied hydrogen requires extremely low storage temperatures, and ammonia is a toxic substance that requires separate handling, and energy loss occurs during the conversion process. To compensate for this, research on liquefied hydrogen, such as multilayer insulation technology, is being conducted. Ammonia has successfully extracted high-purity hydrogen by developing a membrane reactor. Ammonia toxicity can be prevented by strengthening leak detection and blocking facilities. Among these, ammonia was found to be the most suitable for underground storage in terms of economic feasibility, environment, and commercialization.

Improved Early Detection of Tube Leaks Faults in Pulverised Coal-fired Boiler Using Deep Feed Forward Neural Network

Boiler tube leaks significantly reduce the operational availability of power units, yet their early detection and prediction have not been fully realised in the industry. This paper introduces a novel approach employing deep feedforward neural networks for early detection of boiler tube leaks in pulverised coal-fired boilers. Early detection enhances repair planning, minimising downtime and production losses. It also improves monitoring and control of boiler tube failures, thereby optimising power plant operations and revenue. Diverse deep neural network models were developed and rigorously tested by leveraging 9 years of operational data (2012–2020). Exhaustive hyper-parameter optimisation, involving seven parameters, substantially improved predictive accuracy. By achieving training and testing accuracies of 82.8% to 99.3%, the study assessed their ability to detect boiler tube leaks over the same 9-year span, providing insights into leak detection capabilities. The models notably predicted all 12 identified tube leak events, averaging a 14-day lead time before boiler shutdown. In addition to leak prediction, a leak detection matrix was devised to analyse residual behaviour and reduce the likelihood of false alarms. However, the models’ predictive performance was observed to be limited to the following year, with satisfactory results for 2021 only. Incorporating the 2021 data into retraining significantly improved the predictions for 2022. The study concludes that while the models demonstrate robust short-term prediction capabilities, they require continuous retraining to maintain accuracy and relevance. This ongoing refinement is essential for keeping the models up-to-date and reliable in predicting future boiler tube leaks.

Miniaturized Electrochemical Gas Sensor with a Functional Nanocomposite and Thin Ionic Liquid Interface for Highly Sensitive and Rapid Detection of Hydrogen.

Hydrogen has been widely used in industrial and commercial applications as a carbon-free, efficient energy source. Due to the high flammability and explosion risk of hydrogen-air mixtures, it is vital to develop sensors featuring fast-responding and high sensitivity for hydrogen leakage detection. This paper presents a miniaturized electrochemical gas sensor by elaborately establishing a nanocomposite and thin ionic liquid interface for highly sensitive and rapid electrochemical detection of hydrogen, in which a remarkable response time and recovery time of approximately 6 s was achieved at room temperature. A screen-printed carbon electrode was modified with a reduced graphene oxide-carbon nanotube (rGO-CNT) hybrid and platinum-palladium (Pt-Pd) bimetallic nanoparticles to realize high sensitivity. To achieve miniaturization and high stability of the sensor, a thin-film room-temperature ionic liquid (RTIL) was employed as the electrolyte with a significantly decreased response time. The fast-responding hydrogen sensor demonstrates excellent performance with high sensitivity, linearity, and repeatability at concentrations below the lower explosive limit of 4 vol %. The engineered high-performance interface and gas sensor provide a promising and effective strategy for gas sensor design and rapid hazardous gas monitoring.

Leak detection and leak localization in a smart water management system using computational fluid dynamics (CFD) and deep learning (DL)

Leak detection and leak localization in a smart water management system using computational fluid dynamics (CFD) and deep learning (DL)

A Novel Optimized Vibration-based Energy Harvester for Leak Detection in Wall-mounted Water Pipelines

A Novel Optimized Vibration-based Energy Harvester for Leak Detection in Wall-mounted Water Pipelines

Segmentation of tunnel water leakage based on a lightweight DeepLabV3+ model

Abstract The accurate and efficient detection of water leakage with complex backgrounds is crucial for the safety of metro operations. A lightweight segmentation method for metro tunnel water leakage based on transfer learning is proposed. Firstly, this is based on the Deeplabv3+ model and adopts MobileNetv3-Large as the backbone feature extraction network, which significantly reduces the network parameters and improves the detection speed; secondly, it incorporates the efficient channel attention mechanism, which enables the model to adaptively adjust the weights of the channel features and capture the inter-channel relationships in the image, which significantly improves the model’s ability for feature extraction ability; furthermore, for the problem of severe imbalance between positive and negative samples in the dataset, the recognition accuracy of complex samples is increased by optimizing the loss function; finally, the training method of transfer learning is utilized to solve the problem of scarcity of water leakage dataset, and to improve the model’s accuracy and generalization ability. The results show that the model has more significant detection accuracy and segmentation speed advantages than today’s mainstream semantic segmentation model. With strong generalization ability in complex environments (e.g., low illumination and multiple obstructions), the model can be used for intelligent operation and maintenance in metro tunnel projects.

Water Leak Detection: A Comprehensive Review of Methods, Challenges, and Future Directions

This paper provides a comprehensive review of the methods and techniques developed for detecting leaks in water distribution systems, with a focus on highlighting their strengths, weaknesses, and areas for future research. Given the substantial economic, social, and environmental impacts of undetected leaks, timely detection and precise location of leaks are critical concerns for water authorities. This review categorizes existing methods into traditional approaches, such as manual sounding, and modern techniques involving smart water management and sensor technologies. A multidimensional bibliometric analysis was employed to systematically identify, select, and evaluate 600 scholarly articles on water leak detection, sourced from the Scopus database over a 23-year period (2000–2023). The paper evaluates each method based on leak sensitivity, burst detection, continuous monitoring, alarm accuracy, and implementation costs. Novel insights include an analysis of emerging smart water technologies and their integration into real-world water distribution networks, offering improved efficiency in leak detection. The paper also identifies key gaps in current research and suggests future directions for advancing the accuracy and cost-effectiveness of these technologies.

Border Gateway Protocol Route Leak Detection Technique Based on Graph Features and Machine Learning

In the Internet, ASs are interconnected using BGP. However, due to a lack of security considerations in the design of BGP, a series of security issues arise during the propagation of routing information, such as prefix hijacking, route leakage, and AS path tampering. Therefore, this paper conducts research on the detection of route leakage. By analyzing BGP routing information, we abstract the routing propagation relationship between ASs into a network topology graph, and extract graph features from the graph abstracted from routing data at certain time intervals. Based on the structural robustness features and centrality measurement features of the graph, we determine whether a route leakage has occurred during the current time period. To this end, we use machine learning methods and propose a weighted voting model. This model trains multiple single models and assigns weights to them, and through the weighted analysis of the results of multiple models, it can determine whether a route leakage has occurred. In addition, to determine the corresponding weights, we use genetic algorithms for identifying route leaks. The experimental results show that the method used in this paper has a high accuracy rate, and compared with a single model, it performs better on multiple datasets.

MXene anchored fabric for room-temperature operated ammonia sensing

Low-cost and fast-response fabric sensors are highly desirable for shop floor worker daily monitoring and early trace ammonia leak detection. However, the difficulty is how to establish a high bonding capacity between the active material and the fabric to meet the needs of sensing performance on practical applications. Herein, MXene anchored fabric sensors have been fabricated by spray casting technology and the impacts of the preparation process on sensor performance have been investigated. When the casting concentration of MXene is 10 mg/mL and the fabric is spandex, the ammonia sensor exhibits the best sensing performance with a response time of 0.62 s for 100 ppm ammonia and a low limit of detection of 0.4 ppm. Notably, when the sensor used for detection ammonia in wafer cleaner volatiles, the responsivity was up to 96.61 %. This research offers an efficient route for developing MXene anchored fabric for room-temperature operated ammonia sensing.

Neural network-enhanced internal leakage analysis for efficient fault detection in heavy machinery hydraulic actuator cylinders

This work presents a method for detecting internal leakage faults in hydraulic actuator cylinders using signal analysis and a supervised artificial neural network classifier. An artificial leakage is introduced into the hydraulic cylinder actuator, and a signal-based fault detection method is employed to process and transform the signals for internal leakage detection. The analysis focuses on extracting features from the pressure signal, particularly the peaks, which include information such as location, height, and width. After the neural network has undergone the training process, it is deployed for the purpose of categorizing fault levels into three distinct categories: “a healthy system,”“a system with a low fault,” and “a system with a high fault.” The proposed technique utilizes pressure difference signals from the cylinder chambers and extracts features from the peak signals to reduce dimensionality. The extracted features are then used for supervised training of an artificial neural network using a feed-forward algorithm. The trained network is capable of predicting the leakage class of unknown datasets. This method offers advantages such as reduced computational cost through feature extraction and dimensionality reduction, and it is capable of detecting multiple leakage classes. This study introduces a notably efficient approach, predicated on artificial neural networks, for the identification of internal leakage faults in hydraulic cylinders, with specific emphasis on faults arising from component wear and seal damage. The practical value of this research lies in its potential to significantly improve the reliability and efficiency of hydraulic systems used in heavy machinery. By utilizing neural networks for internal leakage detection, this research addresses a critical issue that can lead to costly downtime and maintenance in industrial operations.

Research on gas pipeline leakage identification based on SE-2DCNN with ultra-weak fiber Bragg grating distributed acoustic sensing system

Abstract A new method for identifying gas pipeline leaks using an ultra-weak fiber Bragg grating (UWFBG) distributed acoustic sensing (DAS) system is proposed. To enhance the accuracy of gas leakage detection, an improved convolutional neural network (CNN) incorporating a two-dimensional structure based on the squeeze-and-excitation (SE) attention mechanism was introduced. The principle of acoustic leakage sensing using this technology is explained in detail, and an experimental setup simulating gas pipeline leaks is constructed. During this process, 9,340 DAS data points across varying leakage volumes and pipeline pressures were collected to create ten distinct datasets. The Mel-frequency cepstral coefficients (MFCC) are employed as the feature input for the optimized SE-2DCNN model, which performs identification tasks. The results show that this optimized model achieves an average leakage identification accuracy of 95.33%, demonstrating superior performance over other methods. This approach offers a robust reference for accurately detecting gas pipeline leakages.

EU Methane Regulation and its impact on LNG imports

Abstract The EU Methane Regulation entered into force on 4 August 2024. It creates rules and obligations on measurement-based Monitoring, Reporting, and Verification of methane emissions of relevant activities, and on Leak Detection and Repair, and introduces limits on routine venting and flaring within the European Union (EU). In addition, and importantly for this article, it creates the first import-related requirements on methane emissions for imported fuels globally. This article will focus on the impact of the Methane Regulation on Liquified Natural Gas (LNG) imports into the EU. In this respect, the Methane Regulation creates far-reaching and difficult-to-implement obligations. The concerns for the LNG industry and imports into the EU are that (i) many of the central rules are not set out in the Methane Regulation and will be provided in the years to come and (ii) some of the key concepts and key provisions lack clarity and are open to various interpretations. As we will discuss in detail, the uncertainty created by the Methane Regulation will make contract negotiations more difficult and will negatively affect the security of gas supply within the EU.

Divisional intuitionistic fuzzy least squares twin SVM for pipeline leakage detection

Improving the accuracy of pipeline leak detection in noisy environments is significant for maintaining pipeline safety. Currently, the weighted support vector machine (SVM) is extensively employed. Among them, Intuitionistic fuzzy twin SVM (IFTSVM) has good robustness due to the integration of intuitionistic fuzzy set theory. However, IFTSVM does not fully consider the position information of samples in the feature space, cannot reduce the impact of heterogeneous clustering noise. Moreover, it requires solving two quadratic programming problems (QPPs), which is inefficient. To address the aforementioned issues, this paper proposes a divisional intuitionistic fuzzy least squares twin SVM (DIFL-TSVM) for pipeline leakage detection. This approach proposes a novel divisional intuitionistic fuzzy membership calculation strategy to reduce the impact of heterogeneous clustering noise and simplify the calculation of sample membership. In addition, DIFL-TSVM transforms the original QPPs into two linear equations using the least squares method to improve the efficiency of model. This paper verifies that the DIFL-TSVM model not only has the ability to eliminate the effects of discrete noise and heterogeneous clustering noise but also more efficiency on water supply pipeline leakage diagnosis experiments and UCI datasets. The experimental results show that DIFL-TSVM has higher leak identification with a detection accuracy of 95.6 %.

From Reactive to Proactive Infrastructure Maintenance: Remote Sensing Data and Practical Resilience in the Management of Leaky Pipes

The introduction of remote sensing technologies, AI and big data analytics in the utility sector is warranted by the need to provide critical services with the least disruption to customers, but also to enable preventive maintenance, extend the life cycle of infrastructure components and reduce grid loss—or overall, to exhibit ‘durability’ and ‘resilience’ when faced with the certainty of breakage and decay. In this paper, we first explore the concept of ‘resilience’ and the nature of practice from a performativist perspective in order to set the scene for discussing the impact of ‘datafication’ on maintenance practices and infrastructure durability. We then describe an instance of introducing remote sensing technologies in district heating network surveillance and leak detection: drone-operated thermographic cameras and underground wire sensors. Based on insights from this case study, we discuss the specificity of data-driven infrastructure maintenance practices, and what it means to exhibit practical resilience in relation to how such practices unfold, interrelate and evolve over time. We reflect on how the use of remote sensing technologies and data analytics (1) potentially changes district heating workers’ epistemic worlds (i.e., how knowledge emerges, is negotiated and ordered in practice) and (2) provides opportunities for ‘messy’ pipe repair work to tacitly adopt proactive and preventive logics to meet continuously evolving organizational and societal needs.

Model of oil pipeline tiny defects detection based on DDPM gated parallel convolutional swin transformer

Abstract To address the challenges of difficult detection of minute magnetic flux leakage (MFL) defects, insufficient inspection data, and low detection accuracy, the denoising diffusion probabilistic model (DDPM) gate dilated parallel convolution swin transformer (DGPST) is proposed. First, we introduce a DDPM-based data generation model, successfully generating a large quantity of diverse and rich MFL defect samples. Second, a gated parallel convolution layer is introduced into the backbone network. This strategy uses the characteristics of dilated convolution to broaden the receptive field of the model, thus enhancing the integration ability of global information. The addition of gating mechanism enables the model to adjust the calculation of attention weight based on broader context information in advance, which not only complicates the shortcomings of window self-attention in global dependence understanding, but also effectively suppress irrelevant calculation. Finally, the loss function of H Intersection over Union is introduced to improve the mean average precision. Following these enhancements, DGPST attains a satisfactory outcome in detecting tiny defects within the MFL problem. Experimental data indicates the accuracy of the algorithm reaches 95.6% and the delay is reduced to 7.6 ms.

A Novel Defect Quantification Method Utilizing Multi-Sensor Magnetic Flux Leakage Signal Fusion

In the assessment of pipeline integrity using magnetic flux leakage (MFL) detection, it is crucial to quantify defects accurately and efficiently using MFL signals. However, in complex detection environments, traditional defect inversion methods exhibit low quantification accuracy and efficiency due to the complexity of their algorithms or excessive reliance on a priori knowledge and expert experience. To address these issues, this study presents a novel defect quantification method based on multi-sensor signal fusion (MSSF). The method employs a multi-sensor probe to fuse the MFL signals under multiple lift-off values, enhancing the diversity of defect information. This enables defect-opening profile recognition using the characteristic approximation approach (CAA). Subsequently, the MSSF method is based on a 3D magnetic dipole model and integrates the structural features of multi-sensor probes to develop an algorithm. This algorithm iteratively determines the defect depth at multiple data acquisition points within the defect region to obtain the maximum defect depth. The feasibility of the MSSF quantification method is validated through finite element simulation and physical experiments. The results demonstrate that the proposed method achieves accurate defect quantification while enhancing efficiency, with the number of iterations for each defect depth calculation point consistently requiring fewer than 15 iterations. For rectangular metal loss, perforation, and conical defects, quantification errors are less than 10%, meeting practical inspection requirements.

Intelligent recognition and automatic localization of pipeline welds based on multi-vision system

Abstract Currently, the leakage detection of spacecraft pipeline welds relies on manual point-by-point inspection using a detection gun, which is inefficient and inadequate for the automation needs of spacecraft production. However, the accurate recognition and precise localization of widely distributed and small pipeline welds are crucial for automated detection. Therefore, this paper proposes a multi-vision detection and localization system that integrates global and local information, considering both comprehensive global 3D search and high-precision local 3D measurement. The improved YOLOv8 model is employed for pipeline weld recognition, which improves the recognition rate of welds. Based on the deep learning recognized and segmented welds, this paper proposes stereo matching and segmentation extraction methods for 3D localization and pipeline orientation determination. Additionally, the system integrates a robot to perform automated point-by-point inspection of welds within the area without collisions. The experimental results demonstrate the effectiveness of the improved YOLOv8 and the proposed methods for 3D weld localization and pipeline orientation determination. The maximum deviation of the spatial distance of fine weld positioning is 0.20 mm, and the repeatability of the 3D coordinates is around 0.1 mm. The system can perform precise localization and detection, meeting the requirements for automatic weld recognition and localization.

Pipeline Leak Identification and Prediction of Urban Water Supply Network System with Deep Learning Artificial Neural Network

Pipeline leakage, which leads to water wastage, financial losses, and contamination, is a significant challenge in urban water supply networks. Leak detection and prediction is urgent to secure the safety of the water supply system. Relaying on deep learning artificial neural networks and a specific optimization algorithm, an intelligential detection approach in identifying the pipeline leaks is proposed. A hydraulic model is initially constructed on the simplified Net2 benchmark pipe network. The District Metering Area (DMA) algorithm and the Cuckoo Search (CS) algorithm are integrated as the DMA-CS algorithm, which is employed for the hydraulic model optimization. Attributing to the suspected leak area identification and the exact leak location, the DMA-CS algorithm possess higher accuracy for pipeline leakage (97.43%) than that of the DMA algorithm (92.67%). The identification pattern of leakage nodes is correlated to the maximum number of leakage points set with the participation of the DMA-CS algorithm, which provide a more accurate pathway for identifying and predicting the specific pipeline leaks.

Ultra-sensitive liquefied petroleum gas (LPG) sensor based on monometallic Ag nanospheres synthesized via microwave-assisted facile approach

We present the synthesis, analysis, and utility of polyvinylpyrrolidone (PVP) stabilized monometallic silver nanospheres for liquefied petroleum gas sensing at room temperature. We have synthesised silver nanospheres (Ag-NSs) in the diameter range of 13–21 nm using a microwave assisted method, which is both easy and fast, because microwave irradiation only takes 30 s. A solution of 0.1 M silver nitrate (AgNO3) in an ethanolic medium was microwave-irradiated to produce Ag-NSs. PVP was used as a stabilising agent throughout the process. Spin coating method was used to produce thin films of synthesised monometallic Ag-NSs. X-ray diffraction (XRD), scanning electron microscopy (SEM), acoustic particle sizer (APS), UV–visible absorption (UV–visible) and Fourier Transform Infrared (FT-IR) spectroscopy were used for characterising Ag-NSs. Based on the acoustic attenuation of the prepared sample, the acoustic particle sizer estimated the diameter of Ag-NSs to be around 17 nm. The deposited films were examined for LPG leakage detection at ambient temperature. This is the first report on monometallic Ag-NSs for leakage detection of LPG at room temperature operation below its lower explosive limit (LEL). The significant finding of the developed sensor is that it prompts the quick response (∼16 s) and recovery (∼64 s) as Ag enhances the reaction rate between the sensing film surface and adsorbed LPG, thereby, augments the sensitivity of sensor even below LEL. Moreover, the fabricated LPG sensor shows noteworthy reproducibility (measured up to six cycles) and stability up to 06 months after fabrication.