N. Casari, M. Pinelli, A. Suman, M. Manganelli, Mirko Morini, K. Brun, L. Larosiliere, V. Jariwala
The operability region of a centrifugal compressor is bounded by the low-flow (or high-pressure ratio) limit, commonly referred to as surge. The exact location of the surge line on the map can vary depending on the operating condition and, as a result, a typical Surge Avoidance Line is established at 10% to 15% above the stated flow for the theoretical surge line. The current state of the art of centrifugal compressor surge control is to utilize a global recycle valve to return flow from the discharge side of a centrifugal compressor to the suction side to increase the flow through the compressor and, thus, avoid entering the surge region. This is conventionally handled by defining a compressor surge control line that conservatively assumes that all stages must be kept out of surge at all the time. In compressors with multiple stages, the amount of energy loss is disproportion-ally large since the energy that was added in each stage is lost during system level (or global) recycling. This work proposes an internal stage-wise recycling that provides a much more controlled flow recycling to affect only those stages that may be on the verge of surge. The amount of flow needed for such a scheme will be much smaller than highly conservative global recycling approach. Also, the flow does not leave the compressor casing and therefore does not cross the pressure boundary. Compared to global recycling this inherently has less loss depending upon application and specific of control design.
{"title":"Dynamic Model of Multistage Centrifugal Compressor With a Stage-by-Stage Anti-Surge Recirculating System","authors":"N. Casari, M. Pinelli, A. Suman, M. Manganelli, Mirko Morini, K. Brun, L. Larosiliere, V. Jariwala","doi":"10.1115/gt2021-04273","DOIUrl":"https://doi.org/10.1115/gt2021-04273","url":null,"abstract":"\u0000 The operability region of a centrifugal compressor is bounded by the low-flow (or high-pressure ratio) limit, commonly referred to as surge. The exact location of the surge line on the map can vary depending on the operating condition and, as a result, a typical Surge Avoidance Line is established at 10% to 15% above the stated flow for the theoretical surge line. The current state of the art of centrifugal compressor surge control is to utilize a global recycle valve to return flow from the discharge side of a centrifugal compressor to the suction side to increase the flow through the compressor and, thus, avoid entering the surge region. This is conventionally handled by defining a compressor surge control line that conservatively assumes that all stages must be kept out of surge at all the time. In compressors with multiple stages, the amount of energy loss is disproportion-ally large since the energy that was added in each stage is lost during system level (or global) recycling.\u0000 This work proposes an internal stage-wise recycling that provides a much more controlled flow recycling to affect only those stages that may be on the verge of surge. The amount of flow needed for such a scheme will be much smaller than highly conservative global recycling approach. Also, the flow does not leave the compressor casing and therefore does not cross the pressure boundary. Compared to global recycling this inherently has less loss depending upon application and specific of control design.","PeriodicalId":252904,"journal":{"name":"Volume 8: Oil and Gas Applications; Steam Turbine","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125164183","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Griffin C. Beck, N. Andrews, A. G. Berry, A. McCleney
In gas processing, boosting, and gathering applications, gas-liquid separator equipment (typically referred to as a scrubber) is placed upstream of each reciprocating compressor stage to remove water and hydrocarbon condensates. However, field experience indicates that liquids are often still present downstream of the separation equipment. When liquids are ingested into the reciprocating compressor, machinery failures, some of which are severe, can result. While it is generally understood that liquid carryover and condensation can occur, it is less clear how the multiphase fluid moves through equipment downstream of the scrubber. In this paper, mechanisms responsible for liquid formation and carryover into reciprocating compressors are explored. First, the effects of liquid ingestion on reciprocating compressors reported in the open literature are reviewed. Then, the role of heat and pressure loss along the gas flow path is investigated to determine whether liquid formation (i.e., condensation) is likely to occur for two identical compressors with different pulsation bottle configurations. For this investigation, conjugate heat transfer (CHT) models of the suction pulsation bottles are used to identify regions where liquid dropout is likely to occur. Results of these investigations are presented. Next, liquid carryover from the upstream scrubber is considered. Multiphase models are developed to determine how the multiphase fluid flows through the complex flow path within the pulsation bottle. Two liquid droplet size distributions are employed in these models. Descriptions of the modeling techniques, assumptions, and boundary conditions are provided.
{"title":"Wet Gas Formation and Carryover in Compressor Suction Equipment","authors":"Griffin C. Beck, N. Andrews, A. G. Berry, A. McCleney","doi":"10.1115/gt2021-59353","DOIUrl":"https://doi.org/10.1115/gt2021-59353","url":null,"abstract":"\u0000 In gas processing, boosting, and gathering applications, gas-liquid separator equipment (typically referred to as a scrubber) is placed upstream of each reciprocating compressor stage to remove water and hydrocarbon condensates. However, field experience indicates that liquids are often still present downstream of the separation equipment. When liquids are ingested into the reciprocating compressor, machinery failures, some of which are severe, can result. While it is generally understood that liquid carryover and condensation can occur, it is less clear how the multiphase fluid moves through equipment downstream of the scrubber.\u0000 In this paper, mechanisms responsible for liquid formation and carryover into reciprocating compressors are explored. First, the effects of liquid ingestion on reciprocating compressors reported in the open literature are reviewed. Then, the role of heat and pressure loss along the gas flow path is investigated to determine whether liquid formation (i.e., condensation) is likely to occur for two identical compressors with different pulsation bottle configurations. For this investigation, conjugate heat transfer (CHT) models of the suction pulsation bottles are used to identify regions where liquid dropout is likely to occur. Results of these investigations are presented. Next, liquid carryover from the upstream scrubber is considered. Multiphase models are developed to determine how the multiphase fluid flows through the complex flow path within the pulsation bottle. Two liquid droplet size distributions are employed in these models. Descriptions of the modeling techniques, assumptions, and boundary conditions are provided.","PeriodicalId":252904,"journal":{"name":"Volume 8: Oil and Gas Applications; Steam Turbine","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122842943","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Rasidi Mohamed, Syafeq Moazari Sukeri, Robert Mendoza, R. Kurz
A key function for a control system in a gas turbine train is to keep the operation of all components within a range of parameters that keep the unit safe. If the operating parameters of components fall outside the desired range for safe operation, the control system will detect these and create an alarm. For critical parameters, the control system may initiate an alarm and a shutdown of the unit. In many instances, an alarm may precede the shutdown command. Frequent discussions evolve around situations that lead to a shutdown of the train, as shutdowns impact the availability of the turbomachinery equipment, but in a wider sense also the availability of the compressor station. Therefore, shutdowns impact the profitability of a system. On the other hand, shutdowns may prevent significant, costly damage to the equipment, with significant downtime, and financial implications. In this lecture, we will discuss different methodologies for shutdown requirements, in the effort to maximize availability of units. Particular emphasis will be given to aging machines as well as machines where the instrumentation, and the control algorithms may no longer be state of the art, or where unnecessary or spurious shutdowns plague an installation.
{"title":"Alarms, Shutdowns and Trip Rationalization","authors":"Rasidi Mohamed, Syafeq Moazari Sukeri, Robert Mendoza, R. Kurz","doi":"10.1115/gt2021-00646","DOIUrl":"https://doi.org/10.1115/gt2021-00646","url":null,"abstract":"\u0000 A key function for a control system in a gas turbine train is to keep the operation of all components within a range of parameters that keep the unit safe. If the operating parameters of components fall outside the desired range for safe operation, the control system will detect these and create an alarm. For critical parameters, the control system may initiate an alarm and a shutdown of the unit. In many instances, an alarm may precede the shutdown command.\u0000 Frequent discussions evolve around situations that lead to a shutdown of the train, as shutdowns impact the availability of the turbomachinery equipment, but in a wider sense also the availability of the compressor station. Therefore, shutdowns impact the profitability of a system. On the other hand, shutdowns may prevent significant, costly damage to the equipment, with significant downtime, and financial implications.\u0000 In this lecture, we will discuss different methodologies for shutdown requirements, in the effort to maximize availability of units. Particular emphasis will be given to aging machines as well as machines where the instrumentation, and the control algorithms may no longer be state of the art, or where unnecessary or spurious shutdowns plague an installation.","PeriodicalId":252904,"journal":{"name":"Volume 8: Oil and Gas Applications; Steam Turbine","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129791804","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
E. Losi, M. Venturini, L. Manservigi, G. Ceschini, G. Bechini, Giuseppe Cota, Fabrizio Riguzzi
A gas turbine trip is an unplanned shutdown, of which the most relevant consequences are business interruption and a reduction of equipment remaining useful life. Thus, understanding the underlying causes of gas turbine trip would allow predicting its occurrence in order to maximize gas turbine profitability and improve its availability. In the ever competitive Oil & Gas sector, data mining and machine learning are increasingly being employed to support a deeper insight and improved operation of gas turbines. Among the various machine learning tools, Random Forests are an ensemble learning method consisting of an aggregation of decision tree classifiers. This paper presents a novel methodology aimed at exploiting information embedded in the data and develops Random Forest models, aimed at predicting gas turbine trip based on information gathered during a timeframe of historical data acquired from multiple sensors. The novel approach exploits time series segmentation to increase the amount of training data, thus reducing overfitting. First, data are transformed according to a feature engineering methodology developed in a separate work by the same authors. Then, Random Forest models are trained and tested on unseen observations to demonstrate the benefits of the novel approach. The superiority of the novel approach is proved by considering two real-word case-studies, involving filed data taken during three years of operation of two fleets of Siemens gas turbines located in different regions. The novel methodology allows values of Precision, Recall and Accuracy in the range 75–85 %, thus demonstrating the industrial feasibility of the predictive methodology.
{"title":"Prediction of Gas Turbine Trip: a Novel Methodology Based on Random Forest Models","authors":"E. Losi, M. Venturini, L. Manservigi, G. Ceschini, G. Bechini, Giuseppe Cota, Fabrizio Riguzzi","doi":"10.1115/gt2021-58916","DOIUrl":"https://doi.org/10.1115/gt2021-58916","url":null,"abstract":"\u0000 A gas turbine trip is an unplanned shutdown, of which the most relevant consequences are business interruption and a reduction of equipment remaining useful life. Thus, understanding the underlying causes of gas turbine trip would allow predicting its occurrence in order to maximize gas turbine profitability and improve its availability.\u0000 In the ever competitive Oil & Gas sector, data mining and machine learning are increasingly being employed to support a deeper insight and improved operation of gas turbines. Among the various machine learning tools, Random Forests are an ensemble learning method consisting of an aggregation of decision tree classifiers.\u0000 This paper presents a novel methodology aimed at exploiting information embedded in the data and develops Random Forest models, aimed at predicting gas turbine trip based on information gathered during a timeframe of historical data acquired from multiple sensors. The novel approach exploits time series segmentation to increase the amount of training data, thus reducing overfitting.\u0000 First, data are transformed according to a feature engineering methodology developed in a separate work by the same authors. Then, Random Forest models are trained and tested on unseen observations to demonstrate the benefits of the novel approach. The superiority of the novel approach is proved by considering two real-word case-studies, involving filed data taken during three years of operation of two fleets of Siemens gas turbines located in different regions.\u0000 The novel methodology allows values of Precision, Recall and Accuracy in the range 75–85 %, thus demonstrating the industrial feasibility of the predictive methodology.","PeriodicalId":252904,"journal":{"name":"Volume 8: Oil and Gas Applications; Steam Turbine","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131302421","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
E. Losi, M. Venturini, L. Manservigi, G. Ceschini, G. Bechini, Giuseppe Cota, Fabrizio Riguzzi
A gas turbine trip is an unplanned shutdown, of which the consequences are business interruption and a reduction of equipment remaining useful life. Therefore, detection and identification of symptoms of trips would allow predicting its occurrence, thus avoiding damages and costs. The development of machine learning models able to predict gas turbine trip requires the definition of a set of target data and a procedure of feature engineering that improves machine learning generalization and effectiveness. This paper presents a methodology that focuses on the steps that precede the development of a machine learning model, i.e., data selection and feature engineering, which are the key for a successful predictive model. Data selection is performed by partitioning units into homogeneous groups according to different criteria, e.g., type, region of installation, and operation. A subsequent matching algorithm is applied to rotational speed data of multiple gas turbine units to identify start-ups and shutdowns so that the considered units can be partitioned according to their operation, i.e., base load or peak load. Feature engineering aims at creating features that improve machine learning model accuracy and reliability. First, the Discrete Fourier Transform is used to identify and remove from the time series the seasonal components, i.e., patterns that repeat with a given periodicity. Then, new features are created based on gas turbine domain knowledge in order to capture the complex interactions among system variables and trip occurrence. The outcomes of this paper are the definition of a set of target examples and the identification of a meaningful set of features suitable to develop a machine learning model aimed at predicting gas turbine trip.
{"title":"Data Selection and Feature Engineering for the Application of Machine Learning to the Prediction of Gas Turbine Trip","authors":"E. Losi, M. Venturini, L. Manservigi, G. Ceschini, G. Bechini, Giuseppe Cota, Fabrizio Riguzzi","doi":"10.1115/gt2021-58914","DOIUrl":"https://doi.org/10.1115/gt2021-58914","url":null,"abstract":"\u0000 A gas turbine trip is an unplanned shutdown, of which the consequences are business interruption and a reduction of equipment remaining useful life. Therefore, detection and identification of symptoms of trips would allow predicting its occurrence, thus avoiding damages and costs.\u0000 The development of machine learning models able to predict gas turbine trip requires the definition of a set of target data and a procedure of feature engineering that improves machine learning generalization and effectiveness.\u0000 This paper presents a methodology that focuses on the steps that precede the development of a machine learning model, i.e., data selection and feature engineering, which are the key for a successful predictive model.\u0000 Data selection is performed by partitioning units into homogeneous groups according to different criteria, e.g., type, region of installation, and operation. A subsequent matching algorithm is applied to rotational speed data of multiple gas turbine units to identify start-ups and shutdowns so that the considered units can be partitioned according to their operation, i.e., base load or peak load.\u0000 Feature engineering aims at creating features that improve machine learning model accuracy and reliability. First, the Discrete Fourier Transform is used to identify and remove from the time series the seasonal components, i.e., patterns that repeat with a given periodicity. Then, new features are created based on gas turbine domain knowledge in order to capture the complex interactions among system variables and trip occurrence.\u0000 The outcomes of this paper are the definition of a set of target examples and the identification of a meaningful set of features suitable to develop a machine learning model aimed at predicting gas turbine trip.","PeriodicalId":252904,"journal":{"name":"Volume 8: Oil and Gas Applications; Steam Turbine","volume":"218 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114657163","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
N. Casari, E. Fadiga, M. Pinelli, A. Suman, R. Kurz, Kevin C. Davis, Flavio Marin
Gas turbine driven centrifugal compressors often undergo a detailed performance analysis either for the acceptance by the user or to evaluate their existing condition. Often, the tests are conducted at site, where the compressor and its driver must prove their capability to fulfill the requirements of the project specifications. Typical field testing of gas turbine compressor packages requires the evaluation of head, efficiency, capacity, fuel consumption, as well as the available driver power. The process is not straightforward since the conditions at which the compressor package is tested are typically different from the originally agreed upon conditions. In particular, the ambient conditions, as well as the load conditions at the test are usually different. There are a number of standard procedures for carrying out a field test. In this work, one of them is considered for quantifying how the variation in some parameters can impact the result of the field test. Particular attention is given to the evaluation and data correction for the gas turbine. The results of the test can be affected by several factors. Some of them are related to the installation, such as the array of RTDs used. Further, the influence of the accuracy of the input data that are used for the calculations must be considered by determining the effect of test uncertainty. Other parameters that can affect the results are related to the modeling: The natural gas exhibits real gas behavior at the test conditions, and an equation of state has to be used for data conversion. The choice of an equation of state can translate into a differences in the test results. An assessment of the impact of these factors on the outcome of a field test is reported. From the results of this work, the expected error as consequence of deviation from the specification can be quantified.
{"title":"Assessment of Non-Standard Procedure in Field Testing of Gas Turbine Driven Centrifugal Compressors","authors":"N. Casari, E. Fadiga, M. Pinelli, A. Suman, R. Kurz, Kevin C. Davis, Flavio Marin","doi":"10.1115/gt2021-04249","DOIUrl":"https://doi.org/10.1115/gt2021-04249","url":null,"abstract":"\u0000 Gas turbine driven centrifugal compressors often undergo a detailed performance analysis either for the acceptance by the user or to evaluate their existing condition. Often, the tests are conducted at site, where the compressor and its driver must prove their capability to fulfill the requirements of the project specifications. Typical field testing of gas turbine compressor packages requires the evaluation of head, efficiency, capacity, fuel consumption, as well as the available driver power. The process is not straightforward since the conditions at which the compressor package is tested are typically different from the originally agreed upon conditions. In particular, the ambient conditions, as well as the load conditions at the test are usually different. There are a number of standard procedures for carrying out a field test. In this work, one of them is considered for quantifying how the variation in some parameters can impact the result of the field test. Particular attention is given to the evaluation and data correction for the gas turbine. The results of the test can be affected by several factors. Some of them are related to the installation, such as the array of RTDs used. Further, the influence of the accuracy of the input data that are used for the calculations must be considered by determining the effect of test uncertainty. Other parameters that can affect the results are related to the modeling: The natural gas exhibits real gas behavior at the test conditions, and an equation of state has to be used for data conversion. The choice of an equation of state can translate into a differences in the test results. An assessment of the impact of these factors on the outcome of a field test is reported. From the results of this work, the expected error as consequence of deviation from the specification can be quantified.","PeriodicalId":252904,"journal":{"name":"Volume 8: Oil and Gas Applications; Steam Turbine","volume":"216 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114982340","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This study presents the results of measurements in a scaled model turbine test rig operated at Mitsubishi Hitachi Power Systems, Ltd. In this paper, the flow pattern obtained by traverse measurements is compared with the results of CFD. In order to investigate the flow field in the low pressure steam turbine, the tests are carried out using a test turbine (4 stages) of × 0.33 scale. The velocity and pressure fields are evaluated by traverse measurements. The corresponding CFD are performed by ANSYS CFX. Generally, shroud and stub are used in last stage rotating blades of industrial steam turbine to provide high structural stability by increasing stiffness and damping. In this study, the shroud and stub are modeled in CFD to evaluate the effect on flow pattern. Besides, in order to evaluate the effects of super cooling in blade rows, non-equilibrium condensation is modeled in CFD by ANSYS CFX. The computation model is constructed as accurate reproduction of the scaled model test steam turbine including some steam pipes, supporting sheet metal and the measurement equipment such as traverse pipes and instruments. However, the simple computation model which consists of blade rows with cavities (multi stages) and short diffuser is applied for non-equilibrium condensation calculation due to convergence problems. Comparative evaluation of the test results with the corresponding CFD results showed that the flow patterns predicted by CFD are good. In order to capture the flow pattern characteristics by CFD, it is necessary to consider both real shape modeling and non-equilibrium condensation modeling.
{"title":"Experimental and Numerical Investigations of the Effects of Real Shape Modeling and Non-Equilibrium Condensation Modeling on the Flow Pattern in Steam Turbine","authors":"S. Tabata, Y. Sasao, K. Segawa","doi":"10.1115/gt2021-01754","DOIUrl":"https://doi.org/10.1115/gt2021-01754","url":null,"abstract":"\u0000 This study presents the results of measurements in a scaled model turbine test rig operated at Mitsubishi Hitachi Power Systems, Ltd. In this paper, the flow pattern obtained by traverse measurements is compared with the results of CFD.\u0000 In order to investigate the flow field in the low pressure steam turbine, the tests are carried out using a test turbine (4 stages) of × 0.33 scale. The velocity and pressure fields are evaluated by traverse measurements.\u0000 The corresponding CFD are performed by ANSYS CFX. Generally, shroud and stub are used in last stage rotating blades of industrial steam turbine to provide high structural stability by increasing stiffness and damping. In this study, the shroud and stub are modeled in CFD to evaluate the effect on flow pattern. Besides, in order to evaluate the effects of super cooling in blade rows, non-equilibrium condensation is modeled in CFD by ANSYS CFX. The computation model is constructed as accurate reproduction of the scaled model test steam turbine including some steam pipes, supporting sheet metal and the measurement equipment such as traverse pipes and instruments. However, the simple computation model which consists of blade rows with cavities (multi stages) and short diffuser is applied for non-equilibrium condensation calculation due to convergence problems.\u0000 Comparative evaluation of the test results with the corresponding CFD results showed that the flow patterns predicted by CFD are good. In order to capture the flow pattern characteristics by CFD, it is necessary to consider both real shape modeling and non-equilibrium condensation modeling.","PeriodicalId":252904,"journal":{"name":"Volume 8: Oil and Gas Applications; Steam Turbine","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125504903","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The steam turbine rotor is still the main power generation equipment. Affected by the impact of new energy on the power grid, the steam turbine needs to participate in peak load regulation, which will make turbine rotor components more prone to failure. The rotor is an important equipment of a steam turbine. Unbalance and misalignment are the normal state of rotor failure. In recent years, more and more attention has been paid to the fault detection method based on deep learning, which takes rotating machinery as the object. However, there is a lack of research on actual steam turbine rotors. In this paper, a method of rotor unbalance and parallel misalignment fault detection based on residual network is proposed, which realizes the end-to-end fault detection of rotor. Meanwhile, the method is evaluated with numerical simulation data, and the multi task detection of rotor unbalance, parallel misalignment, unbalanced parallel misalignment coupling faults (coupling fault called in this paper) is realized. The influence of signal-to-noise ratio and the number of training samples on the detection performance of neural network is discussed. The detection accuracy of unbalanced position is 93.5%, that of parallel misalignment is 99.1%. The detection accuracy for unbalance and parallel misalignment is 89.1% and 99.1%, respectively. The method can realize the direct mapping between the unbalanced, parallel misalignment, coupling fault vibration signals and the fault detection results. The method has the ability to automatically extract fault features. It overcomes the shortcoming of traditional methods that rely on signal processing experience, and has the characteristics of high precision and strong robustness.
{"title":"Research on Fault Diagnosis of Steam Turbine Rotor Unbalance and Parallel Misalignment Based on Numerical Simulation and Convolutional Neural Network","authors":"Chongyu Wang, Di Zhang, Yonghui Xie","doi":"10.1115/gt2021-60247","DOIUrl":"https://doi.org/10.1115/gt2021-60247","url":null,"abstract":"\u0000 The steam turbine rotor is still the main power generation equipment. Affected by the impact of new energy on the power grid, the steam turbine needs to participate in peak load regulation, which will make turbine rotor components more prone to failure. The rotor is an important equipment of a steam turbine. Unbalance and misalignment are the normal state of rotor failure. In recent years, more and more attention has been paid to the fault detection method based on deep learning, which takes rotating machinery as the object. However, there is a lack of research on actual steam turbine rotors. In this paper, a method of rotor unbalance and parallel misalignment fault detection based on residual network is proposed, which realizes the end-to-end fault detection of rotor. Meanwhile, the method is evaluated with numerical simulation data, and the multi task detection of rotor unbalance, parallel misalignment, unbalanced parallel misalignment coupling faults (coupling fault called in this paper) is realized. The influence of signal-to-noise ratio and the number of training samples on the detection performance of neural network is discussed. The detection accuracy of unbalanced position is 93.5%, that of parallel misalignment is 99.1%. The detection accuracy for unbalance and parallel misalignment is 89.1% and 99.1%, respectively. The method can realize the direct mapping between the unbalanced, parallel misalignment, coupling fault vibration signals and the fault detection results. The method has the ability to automatically extract fault features. It overcomes the shortcoming of traditional methods that rely on signal processing experience, and has the characteristics of high precision and strong robustness.","PeriodicalId":252904,"journal":{"name":"Volume 8: Oil and Gas Applications; Steam Turbine","volume":"59 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121823681","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
G. Girezzi, D. Checcacci, L. Cosi, A. Maggi, Alessandro Sani, A. Achilli
The fouling phenomenon addressed in this paper is related to the deposition within steam turbines of steam impurities and to the presence of solid debris, coming from upstream plant sections, that can create solid build-ups in stationary and moving parts inside the turbine. As a consequence, fouling causes unit efficiency decline but, in severe cases, it may also lead to sticking of moving components, such as valves, that may be critical in machine control and/or safety. Despite well-studied and well-considered in design and operation of large power utility plants, where steam quality is of primary importance for boilers, super-heaters, turbines and condensers, this subject is often overlooked in small power generation or industrial applications, where efficiency may be less critical but turbine availability is of paramount importance for plant operation (e.g. LNG plants). The steam fouling is a subject that, despite widely studied in the past, has been quite neglected in more recent years. This paper, with the aim of underlining the importance of fouling in the operation of turbines for industrial applications, starts with examples of field evidences of severe fouling. Then the design of a test bench for the experimental characterization of fouling rates and validation of turbine components, exposed to fouling conditions, is presented along with the description of the deposition models that were developed on the basis of the physical phenomena involved in the fouling process. This study addresses the main deposition physical principles and their implications in the thermodynamic design of the test bench, on the basis of the specific physical properties of the impurities of interest. To better match plant real cases, the contaminants tested included those which have been usually identified within the units during maintenance activities and for which specific limits are prescribed by OEMs. In the following section, details relevant to the main deposition mechanisms due to different geometries and flow-fields are discussed. The results obtained are qualitatively in line with literature and internal practices, yet, through the test activities, it has been possible to establish a quantitative relationship between the concentrations of each contaminant at inlet section and the different thermodynamic conditions along the test bench, so capturing the impact of solubility changes along with the steam expansion.
{"title":"Test Bench for Characterization and Design Against Steam Turbine Fouling","authors":"G. Girezzi, D. Checcacci, L. Cosi, A. Maggi, Alessandro Sani, A. Achilli","doi":"10.1115/gt2021-59685","DOIUrl":"https://doi.org/10.1115/gt2021-59685","url":null,"abstract":"\u0000 The fouling phenomenon addressed in this paper is related to the deposition within steam turbines of steam impurities and to the presence of solid debris, coming from upstream plant sections, that can create solid build-ups in stationary and moving parts inside the turbine. As a consequence, fouling causes unit efficiency decline but, in severe cases, it may also lead to sticking of moving components, such as valves, that may be critical in machine control and/or safety.\u0000 Despite well-studied and well-considered in design and operation of large power utility plants, where steam quality is of primary importance for boilers, super-heaters, turbines and condensers, this subject is often overlooked in small power generation or industrial applications, where efficiency may be less critical but turbine availability is of paramount importance for plant operation (e.g. LNG plants).\u0000 The steam fouling is a subject that, despite widely studied in the past, has been quite neglected in more recent years.\u0000 This paper, with the aim of underlining the importance of fouling in the operation of turbines for industrial applications, starts with examples of field evidences of severe fouling. Then the design of a test bench for the experimental characterization of fouling rates and validation of turbine components, exposed to fouling conditions, is presented along with the description of the deposition models that were developed on the basis of the physical phenomena involved in the fouling process.\u0000 This study addresses the main deposition physical principles and their implications in the thermodynamic design of the test bench, on the basis of the specific physical properties of the impurities of interest. To better match plant real cases, the contaminants tested included those which have been usually identified within the units during maintenance activities and for which specific limits are prescribed by OEMs.\u0000 In the following section, details relevant to the main deposition mechanisms due to different geometries and flow-fields are discussed.\u0000 The results obtained are qualitatively in line with literature and internal practices, yet, through the test activities, it has been possible to establish a quantitative relationship between the concentrations of each contaminant at inlet section and the different thermodynamic conditions along the test bench, so capturing the impact of solubility changes along with the steam expansion.","PeriodicalId":252904,"journal":{"name":"Volume 8: Oil and Gas Applications; Steam Turbine","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120953099","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Bertold Lübbe, Jens Aschenbruck, O. Pütz, Mira Theidel
To meet today’s and future market needs, large end-stage blades are obliged to fulfill high flexibility regarding the operational range and high efficiency goals while being prepared for daily start-stop cycles. The end-stage total efficiency can be maximized by enlarging the steam turbine exhaust area and thereby reducing the exhaust losses. Therefore, a new Low Pressure (LP) backend featuring an increased freestanding 41″ steel blade has been developed and is presented here, which is optimized for maximum efficiency over a wide range of operation conditions. To allow for such a large steel-blade to operate at 60Hz rotational speed and to meet the daily cycling demand, various aspects of the blade design were optimized. A new high strength blade steel was developed (Teuber [1]), which gives the designer freedom for aerodynamical optimizations, while keeping the mechanical utilization within the predefined, allowable limits. To maximize the cycling capability, a new fir tree root was developed which minimizes the static as well as the dynamic loading. To verify the success of the new fir-tree root design and to verify the natural frequencies for the relevant modes, an extensive validation measurement campaign was setup with a full-scale blade row in a spin-pit. Here, the airfoil, root and steeple of the end-stage blade were equipped with strain gauges. Additionally, the blade row was monitored using tip-timing sensors. The results of this validation measurement campaign are presented in this paper. They show a close agreement between the design calculations and the measured static strains and vibration responses in terms of natural frequencies as well as displacement and strain amplitudes. Additionally, a test turbine has been set-up featuring a direct scaling of the new LP backend with the new high strength steel and a pre-stage to simulate realistic operation conditions over the complete operation range. The blade performance was tested up to high mass-flows, condenser pressures of up to 300 mbar and at varying load points covering all potential load points from extreme part load to full load with minimal and maximal condenser pressure. Strain gauges as well as tip-timing are used to measure the vibration response of the end-stage blade during the measurement campaign. The results presented here show, that throughout the complete measurement campaign the blade experienced minimal excitation which led to vibration levels that allowed unrestricted operation in the complete, tested operation range. In summary this paper shows the main design features of a large full-speed freestanding end-stage blade and the validation measures that were performed to ensure that the design targets and the market requirements are fully met.
{"title":"Design and Validation of a Large Steam Turbine End-Stage Blade to Meet Current and Future Market Demands","authors":"Bertold Lübbe, Jens Aschenbruck, O. Pütz, Mira Theidel","doi":"10.1115/gt2021-59315","DOIUrl":"https://doi.org/10.1115/gt2021-59315","url":null,"abstract":"\u0000 To meet today’s and future market needs, large end-stage blades are obliged to fulfill high flexibility regarding the operational range and high efficiency goals while being prepared for daily start-stop cycles.\u0000 The end-stage total efficiency can be maximized by enlarging the steam turbine exhaust area and thereby reducing the exhaust losses. Therefore, a new Low Pressure (LP) backend featuring an increased freestanding 41″ steel blade has been developed and is presented here, which is optimized for maximum efficiency over a wide range of operation conditions.\u0000 To allow for such a large steel-blade to operate at 60Hz rotational speed and to meet the daily cycling demand, various aspects of the blade design were optimized. A new high strength blade steel was developed (Teuber [1]), which gives the designer freedom for aerodynamical optimizations, while keeping the mechanical utilization within the predefined, allowable limits. To maximize the cycling capability, a new fir tree root was developed which minimizes the static as well as the dynamic loading. To verify the success of the new fir-tree root design and to verify the natural frequencies for the relevant modes, an extensive validation measurement campaign was setup with a full-scale blade row in a spin-pit. Here, the airfoil, root and steeple of the end-stage blade were equipped with strain gauges. Additionally, the blade row was monitored using tip-timing sensors. The results of this validation measurement campaign are presented in this paper. They show a close agreement between the design calculations and the measured static strains and vibration responses in terms of natural frequencies as well as displacement and strain amplitudes.\u0000 Additionally, a test turbine has been set-up featuring a direct scaling of the new LP backend with the new high strength steel and a pre-stage to simulate realistic operation conditions over the complete operation range. The blade performance was tested up to high mass-flows, condenser pressures of up to 300 mbar and at varying load points covering all potential load points from extreme part load to full load with minimal and maximal condenser pressure. Strain gauges as well as tip-timing are used to measure the vibration response of the end-stage blade during the measurement campaign. The results presented here show, that throughout the complete measurement campaign the blade experienced minimal excitation which led to vibration levels that allowed unrestricted operation in the complete, tested operation range. In summary this paper shows the main design features of a large full-speed freestanding end-stage blade and the validation measures that were performed to ensure that the design targets and the market requirements are fully met.","PeriodicalId":252904,"journal":{"name":"Volume 8: Oil and Gas Applications; Steam Turbine","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131218243","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：