Fault Detection for Gas Turbine Hot Components Based on a Convolutional Neural Network

Gas turbine hot component failures often cause catastrophic consequences. Fault detection can improve the availability and economy of hot components. The exhaust gas temperature (EGT) profile is usually used to monitor the performance of the hot components. The EGT profile is uniform when the hot component is healthy, whereas hot component faults lead to large temperature differences between different EGT values. The EGT profile swirl under different operating and ambient conditions also cause temperature differences. Therefore, the influence of EGT profile swirl on EGT values must be eliminated. To improve the detection sensitivity, this paper develops a fault detection method for hot components based on a convolutional neural network (CNN). This paper demonstrates that a CNN can extract the information between adjacent EGT values and consider the impact of the EGT profile swirl. This paper reveals, in principle, that a CNN is a viable solution for dealing with fault detection for hot components. Based on the distribution characteristics of EGT thermocouples, the circular padding method is developed in the CNN. The sensitivity of the developed method is verified by real-world data. Moreover, the developed method is visualized in detail. The visualization results reveal that the CNN effectively considers the influence of the EGT profile swirl.

Download Full-text

Early Fault Detection of Hot Components in Gas Turbines

Journal of Engineering for Gas Turbines and Power ◽

10.1115/1.4034153 ◽

2016 ◽

Vol 139 (2) ◽

Cited By ~ 5

Author(s):

Liu Jinfu ◽

Liu Jiao ◽

Wan Jie ◽

Wang Zhongqi ◽

Yu Daren

Keyword(s):

Fault Detection ◽

Gas Turbine ◽

Gas Turbines ◽

Working Environment ◽

Detection Sensitivity ◽

Normal Operation ◽

Large Temperature ◽

Ambient Conditions ◽

Fisher Discriminant Analysis ◽

Early Fault Detection

The working environment of hot components is the most adverse of all gas turbine components. Malfunction of hot components is often followed by catastrophic consequences. Early fault detection plays a significant role in detecting performance deterioration immediately and reducing unscheduled maintenance. In this paper, an early fault detection method is introduced to detect early fault symptoms of hot components in gas turbines. The exhaust gas temperature (EGT) is usually used to monitor the performance of the hot components. The EGT is measured by several thermocouples distributed equally at the outlet of the gas turbine. EGT profile is symmetrical when the unit is in normal operation. And the faults of hot components lead to large temperature differences between different thermocouple readings. However, interferences can potentially affect temperature differences, and sometimes, especially in the early stages of the fault, its influence can be even higher than that of the faults. To improve the detection sensitivity, the influence of interferences must be eliminated. The two main interferences investigated in this study are associated with the operating and ambient conditions, and the structure deviation of different combustion chambers caused by processing and installation errors. Based on the basic principles of gas turbines and Fisher discriminant analysis (FDA), a new detection indicator is presented that characterizes the intrinsic structure information of the hot components. Using this new indicator, the interferences involving the certainty and the uncertainty are suppressed and the sensitivity of early fault detection in gas turbine hot components is improved. The robustness and the sensitivity of the proposed method are verified by actual data from a Taurus 70 gas turbine produced by Solar Turbines.

Download Full-text

Convolutional neural network-based deep transfer learning for fault detection of gas turbine combustion chambers

Applied Energy ◽

10.1016/j.apenergy.2021.117509 ◽

2021 ◽

Vol 302 ◽

pp. 117509

Author(s):

Mingliang Bai ◽

Xusheng Yang ◽

Jinfu Liu ◽

Jiao Liu ◽

Daren Yu

Keyword(s):

Neural Network ◽

Fault Detection ◽

Convolutional Neural Network ◽

Gas Turbine ◽

Transfer Learning ◽

Combustion Chambers ◽

Gas Turbine Combustion

Download Full-text

Early Fault Detection of Gas Turbine Hot Components Based on Exhaust Gas Temperature Profile Continuous Distribution Estimation

Energies ◽

10.3390/en13225950 ◽

2020 ◽

Vol 13 (22) ◽

pp. 5950

Author(s):

Jinfu Liu ◽

Mingliang Bai ◽

Zhenhua Long ◽

Jiao Liu ◽

Yujia Ma ◽

...

Keyword(s):

Fault Detection ◽

Temperature Profile ◽

Gas Turbine ◽

Detection Method ◽

Continuous Distribution ◽

Detection Sensitivity ◽

Exhaust Gas ◽

Gas Temperature ◽

Early Fault Detection ◽

Exhaust Gas Temperature

Failures of the gas turbine hot components often cause catastrophic consequences. Early fault detection can detect the sign of fault occurrence at an early stage, improve availability and prevent serious incidents of the plant. Monitoring the variation of exhaust gas temperature (EGT) is an effective early fault detection method. Thus, a new gas turbine hot components early fault detection method is developed in this paper. By introducing a priori knowledge and quantum particle swarm optimization (QPSO), the exhaust gas temperature profile continuous distribution model is established with finite EGT measuring data. The method eliminates influences of operating and ambient condition changes and especially the gas swirl effect. The experiment reveals the presented method has higher fault detection sensitivity.

Download Full-text

Abrupt Fault Detection and Isolation for Gas Turbine Components Based on a 1D Convolutional Neural Network Using Time Series Data

AIAA Propulsion and Energy 2020 Forum ◽

10.2514/6.2020-3675 ◽

2020 ◽

Author(s):

Junjie Zhao ◽

Yi G. Li

Keyword(s):

Neural Network ◽

Time Series ◽

Fault Detection ◽

Convolutional Neural Network ◽

Gas Turbine ◽

Time Series Data ◽

Fault Detection And Isolation ◽

Series Data

Download Full-text

Numerical investigation of non-uniform flow in twin-silo combustors and impact on axial turbine stage performance

Proceedings of the Institution of Mechanical Engineers Part A Journal of Power and Energy ◽

10.1177/0957650921993948 ◽

2021 ◽

pp. 095765092199394

Author(s):

Hafiz M Hassan ◽

Adeel Javed ◽

Asif H Khoja ◽

Majid Ali ◽

Muhammad B Sajid

Keyword(s):

Gas Turbine ◽

Gas Turbines ◽

Internal Flow ◽

Large Temperature ◽

Clear Understanding ◽

Gas Temperature ◽

Turbine Stage ◽

Flow Angle ◽

Axial Turbine ◽

Stage Performance

A clear understanding of the flow characteristics in the older generation of industrial gas turbines operating with silo combustors is important for potential upgrades. Non-uniformities in the form of circumferential and radial variations in internal flow properties can have a significant impact on the gas turbine stage performance and durability. This paper presents a comprehensive study of the underlying internal flow features involved in the advent of non-uniformities from twin-silo combustors and their propagation through a single axial turbine stage of the Siemens v94.2 industrial gas turbine. Results indicate the formation of strong vortical structures alongside large temperature, pressure, velocity, and flow angle deviations that are mostly located in the top and bottom sections of the turbine stage caused by the excessive flow turning in the upstream tandem silo combustors. A favorable validation of the simulated exhaust gas temperature (EGT) profile is also achieved via comparison with the measured data. A drop in isentropic efficiency and power output equivalent to 2.28% points and 2.1 MW, respectively is observed at baseload compared to an ideal straight hot gas path reference case. Furthermore, the analysis of internal flow topography identifies the underperforming turbine blading due to the upstream non-uniformities. The findings not only have implications for the turbine aerothermodynamic design, but also the combustor layout from a repowering perspective.

Download Full-text