Hybrid Model-Based and Data-Driven Diagnostic Algorithm for Gas Turbine Engines

Author(s):  
Amare Fentaye ◽  
Valentina Zaccaria ◽  
Moksadur Rahman ◽  
Mikael Stenfelt ◽  
Konstantinos Kyprianidis

Abstract Data-driven algorithms require large and comprehensive training samples in order to provide reliable diagnostic solutions. However, in many gas turbine applications, it is hard to find fault data due to proprietary and liability issues. Operational data samples obtained from end-users through collaboration projects do not represent fault conditions sufficiently and are not labeled either. Conversely, model-based methods have some accuracy deficiencies due to measurement uncertainty and model smearing effects when the number of gas path components to be assessed is large. The present paper integrates physics-based and data-driven approaches aiming to overcome this limitation. In the proposed method, an adaptive gas path analysis (AGPA) is used to correct measurement data against the ambient condition variations and normalize. Fault signatures drawn from the AGPA are used to assess the health status of the case engine through a Bayesian network (BN) based fault diagnostic algorithm. The performance of the proposed technique is evaluated based on five different gas path component faults of a three-shaft turbofan engine, namely intermediate-pressure compressor fouling (IPCF), high-pressure compressor fouling (HPCF), high-pressure turbine erosion (HPTE), intermediate-pressure turbine erosion (IPTE), and low-pressure turbine erosion (LPTE). Robustness of the method under measurement uncertainty has also been tested using noise-contaminated data. Moreover, the fault diagnostic effectiveness of the BN algorithm on different number and type of measurements is also examined based on three different sensor groups. The test results verify the effectiveness of the proposed method to diagnose single gas path component faults correctly even under a significant noise level and different instrumentation suites. This enables to accommodate measurement suite inconsistencies between engines of the same type. The proposed method can further be used to support the gas turbine maintenance decision-making process when coupled with overall Engine Health Management (EHM) systems.

Author(s):  
A. Peretto

The present paper evaluates the behavior, in design and part load working conditions, of a complex gas turbine cycle with multiple intercooled compression, and the optional preheating of the air at the high pressure compressor outlet by means of the gas turbine outlet hot gas. The results are then compared with those obtained by a Brayton cycle gas turbine, with or without preheating of the air at the high pressure compressor outlet. Subsequently, the performance of complex combined cycles, with intercooled gas turbine as topper and one, two or three pressure level steam cycle as bottomer, in design and part load working conditions is also evaluated. The performance of these complex combined plants is then compared with that obtained by a Brayton cycle gas turbine as topper and one, two or three pressure level steam cycle as bottomer. Part load working conditions are realized by varying either the inlet guide vane angle of the first compressor nozzles or the maximum temperature at the combustor outlet. The study shows that in part load working conditions obtained by varying IGV, the complex cycles, in the examined gas turbine or in the combined cycle power plants, give conversion efficiencies decidedly greater than those obtainable by varying combustor exit temperature. Furthermore it is found that these complex power plant efficiencies, in part load working conditions, are far greater than those obtained by the Brayton cycle gas turbine, or by combined cycle with Brayton cycle gas turbine as topper, if IGV adjustment is adopted. If power variation is obtained with combustor outlet temperature adjustment, the efficiencies of the combined power plants with complex or Brayton cycle gas turbines, are substantially the same, for the same relative power variation.


Author(s):  
Georg Kro¨ger ◽  
Christian Voß ◽  
Eberhard Nicke ◽  
Christian Cornelius

Engine operating range and efficiency are of increasing importance in modern compressor design for heavy duty gas turbines and aircraft engines. These highly challenging objectives can only be met if all components provide high aerodynamic performance and stability. The aerodynamic losses of highly loaded axial compressors are mainly influenced by the leakage flow through clearance gaps. Especially the leakage flow due to the radial clearances of rotor blades affects negatively both, the efficiency and the operating range of the engine. Recent publications showed that the clearance flow and the clearance vortex can be influenced by an additional static pressure gradient at the outer casing, which is created by an axisymmetric wavy casing shape. A notable performance increase of up to 0.4% stage efficiency at design point conditions was reported for high pressure stages with large tip clearance heights [1] as well as for a transonic stage with a relatively small radial clearance gap [2]. An analytic approach to predict the effects of axisymmetric casing contouring has been developed at DLR, Institute of Propulsion Technology, and is outlined in the first part of this work. The characteristic behavior of the clearance vortex in an adverse pressure gradient is discussed by means of an inviscid vortex model [3]. The critical vortex parameters are isolated and related to the static pressure increase due to the casing contour. The second part illustrates the application of an axisymmetric endwall contour. A three dimensional optimization of the outer casing and the corresponding blade tip airfoil section of a typical gas turbine high pressure compressor stage with a high number of free variables is presented. The optimization led to a significant increase in aerodynamic performance of about 0.8% stage efficiency and to a notable reduction of the endwall blockage at ADP conditions. Furthermore, an improved off-design performance was found and a simple design rule is given to transfer both, the casing contour and the blade tip section modification on similar high pressure compressor blades. Based on these design rules the results of the optimized stages were applied to the rear stages of a Siemens gas turbine compressor CFD model. An increase of 0.3% full compressor performance was reached at design point conditions.


Author(s):  
J. R. Khan ◽  
W. E. Lear ◽  
S. A. Sherif

A thermodynamic performance analysis is performed on a novel cooling and power cycle that combines a semi-closed cycle gas turbine called the High Pressure Regenerative Turbine Engine (HPRTE) with an absorption refrigeration unit. Waste heat from the recirculated combustion gas of the HPRTE is used to power the absorption refrigeration unit, which cools the high-pressure compressor inlet of the HPRTE to below ambient conditions and also produces excess refrigeration, in an amount which depends on ambient conditions. The cycle is modeled using traditional one-dimensional steady-state thermodynamics, with state-of-the-art polytropic efficiencies and pressure drops for the turbo-machinery and heat exchangers, and accurate y correlations for the properties of the LiBr-water mixture and the combustion products. Water produced as a product of combustion is intentionally condensed in the evaporator of the vapor absorption refrigeration system. The mixture properties of air account for the water removal rate. The vapor absorption refrigeration unit is designed to provide sufficient cooling for water extraction. The cycle is shown to operate with a thermal efficiency approaching 58% for a turbine inlet temperature of 1400 °C in addition to producing about 0.45 liters of water per liter of fuel consumed. Also at the above operating condition the ratio of the refrigeration effect to the net work output from the system is equal to 0.8. The ratio of mass of water extracted to the mass of fresh air inlet into the combined cycle is obtained for different values of cycle parameters, namely turbine inlet temperature, recuperator inlet temperature and the low pressure compressor ratio. The maximum value of this ratio is found to be around 0.11. It is found that it is a strong function of the recirculation ratio and it decreased by 22% as the recirculation ratio is decreased by 70%. The thermodynamic impacts of water extraction on the system performance are also discussed. Based on these results, and prior results, which showed that the HPRTE is very compact, it appears that this cycle would be ideally suited for distributed power and vehicle applications, especially ones with associated air conditioning loads.


Author(s):  
A. Boschetti ◽  
E. Y. Kawachi ◽  
M. A. S. Oliveira

This work presents preliminary results of corrosion studies for three blades, one of the low pressure compressor and two of two different stages of the high pressure compressor of a gas turbine, which has been operating for 5,000 hours. Scanning Electron Microscopy (SEM), Energy Dispersive X-ray Spectroscopy (EDS), X-ray diffraction (XRD), Electrochemical Impedance Spectroscopy (EIS) in aqueous solution containing chloride, and Atomic Absorption Spectrometry (AAS) were used to characterize the blades surfaces. The SEM and EDS results showed that the homogeneity and amount of contaminants, such as sodium, potassium, calcium, magnesium, chloride and sulphur are bigger in the high pressure compressor blade surfaces than in the low pressure compressor blade surface. The EIS results showed that the degradation process in turbine compressor blades increases with the temperature and pressure increase inside the compressors and depends of the blade composition. The low pressure compressor blade, which was made of a Ti base superalloy exhibited smaller corrosion resistance (smallest charge transfer resistance value (Rct)) than the two high pressure compressor blades, which were made of a Fe base superalloy. However, despite of its lower resistance to corrosion, after 5,000 hours of service, the low pressure compressor blade did not present pitting corrosion while the high pressure compressor blades did.


Author(s):  
A. B. Turner ◽  
C. A. Long ◽  
P. R. N. Childs ◽  
N. J. Hills ◽  
J. A. Millward

This paper reviews the current position of five major problem areas in gas turbine secondary air system design. Although the problems are of primary interest to the designer of the coolant flow paths, since they directly affect the temperature, the stresses and thus the life of the major rotating components, three of the problems interact with the main gas path and are thus also the concern of the mainstream aerodynamicist. The five problems reviewed are: prediction of the flow distribution and heat transfer in the high pressure compressor drive cone cavity from the turbine to the rim of the HP compressor running underneath the combustion chamber, the flow penetration and heat transfer in the multiple rotating cavities formed by the multiple discs of the high pressure compressor with a rotating shaft running through the bores; the control of ingestion of hot turbine mainstream gas into the rotor-stator wheelspaces through the rim-seals; the problem of compressor and turbine stator-well heating, particularly compressor stator-wells in which excessive temperatures have been occasionally measured and finally, the pre-swirl coolant system which has to take the blade cooling air across from the stationary casing to the rotating turbine disc in the most advantageous manner.


Author(s):  
Chao Zhang ◽  
Aldo Abate ◽  
John Crockett ◽  
Eric Ho

High pressure compressor (HPC) stator vanes of small gas turbine engines frequently have high circumferential variation of vibratory stress. This is very important for vibratory stress measurement by strain gauge tests and structural high cyclic fatigue characterization. The current paper presents experimental results of studying the effects of HPC stator angular positions in gas turbine engines on the circumferential distribution pattern of vibratory stress. Strain gauge tests were done on a stator with cantilevered vanes. Each vane had a strain gauge deployed at the same location. The stator was installed in gas turbine engines at two different angular positions during strain gauge tests. The experimental results show that more than one resonant peak occurred for a given vibratory mode and engine order resonance. The frequencies of resonant peaks were close to one another. The circumferential distribution of maximum vibratory stress (i.e., the maximum magnitude of these resonant peaks) with respect to the stator itself has a similar pattern at the two different angular positions. This clearly indicates the distribution pattern does not follow the gas-path aerodynamic pressure, but follows the stator angular positions. The frequency of the maximum vibratory stress was found to vary from sector to sector instead of from vane to vane; the vanes in each sector have a same frequency. Mistuning analysis was performed on the HPC stator to illustrate a number of resonant peaks and the sector-to-sector frequency variation of the maximum vibratory stress. The approach of “subset of nominal system modes” (SNM) [1, 2] was employed for mistuning analysis and the frequency distributions of stator vanes obtained by bench frequency response tests were used as input data. At the end, one might conclude that the high circumferential variation of vibratory stress be related to mistuning effects due to small variations in vane properties.


Author(s):  
J. R. Khan ◽  
W. E. Lear ◽  
S. A. Sherif ◽  
John F. Crittenden

A thermodynamic design-point performance analysis is performed on a novel cooling and power cycle that combines a semiclosed cycle gas turbine called the high-pressure regenerative turbine engine (HPRTE) with a vapor absorption refrigeration system (VARS). Waste heat from the recirculated combustion gas of the HPRTE is used to power the VARS. Water produced as a product of combustion is intentionally condensed and harvested. A part of the VARS cooling is used to chill the gas entering the high-pressure compressor, allowing water condensation and extraction as well as large efficiency gains. The remaining cooling capacity is provided to an external refrigeration load. The cycle is modeled using zero-dimensional steady-state thermodynamics, considering conservative values of polytropic efficiencies, a conservative model for turbine blade cooling, conservative values of pressure drops for the turbomachinery, including heat exchangers, and accurate correlations for the properties of the LiBr–H2O mixture and the combustion products. The cycle is shown to operate with a thermal efficiency greater than 40% for parameters appropriate to medium sized engines, while producing about 1.5kg of water per kilogram of fuel (propane) consumed. This thermal efficiency is in addition to the large cooling effect generated in the evaporator of VARS, equivalent to 3–4% increased efficiency. The efficiency would be greater than 51% without turbine cooling bleed. The refrigeration ratio, defined as the ratio of external refrigeration load to the net work output, is found to be 0.38 for the base case. The water extracted is found to be a strong function of the recirculation ratio and low pressure compressor ratio PRc1. Based on these and prior results, which showed that the HPRTE is very compact and has inherently low emissions, it appears that this cycle would be well suited for distributed power and some vehicle applications, especially ones with associated air conditioning loads.


Author(s):  
Uyioghosa Igie ◽  
Orlando Minervino

Inlet filtration systems are designed to protect industrial gas turbines from air borne particles and foreign objects, thereby improving the quality of air for combustion and reducing component fouling. Filtration systems are of varying grades and capture efficiencies, with the higher efficiency systems filters providing better protection but higher pressure losses. For the first time, two gas turbine engine models of different configurations and capacities have been investigated for two modes of operation (constant turbine entry temperature (TET) and load/power) for a two- and three-stage filter system. The main purpose of this is to present an account on factors that could decide the selection of filtration systems by gas turbine operators, solely based on performance. The result demonstrates that the two-spool engine is only slightly more sensitive to intake pressure loss relative to the single-spool. This is attributed to higher pressure ratio of the two-spool as well as the deceleration of the high pressure compressor (HPC)/high pressure turbine (HPT) shaft rotational speed in a constant TET operation. The compressor of the single-spool engine and the low pressure compressor (LPC) of the two-spool shows similar behavior: slight increase in pressure ratio and reduced surge margin at their constant rotational speed operation. Loss in shaft power is observed for both engines, about 2.5% at 1000 Pa loss. For constant power operation there is an increase in fuel flow and TET, and as a result the creep life was estimated. The result obtained indicates earlier operating hours to failure for the three-stage system over the two-stage by only a few thousand hours. However, this excludes any degradation due to fouling that is expected to be more significant in the two-stage system.


Sign in / Sign up

Export Citation Format

Share Document