Patent Application
20120283963
The present invention relates generally to monitoring operating environments and in particular to components enabled for transmitting data with respect to the condition of individual components within an operating environment such as a gas turbine engine. More specifically, the invention relates to conditioned based maintenance systems and methods used for predicting the remaining useful life of complex engine systems such as turbine engines and components thereof.
Gas combustion turbines are used for a variety of applications such as driving an electric generator in a power generating plant or propelling a ship or an aircraft. Firing temperatures in modern gas turbine engines continue to increase in response to the demand for higher efficiency engines. Superalloy materials have been developed to withstand the corrosive high temperature environment that exists within a gas turbine engine. However, even superalloy materials are not able to withstand extended exposure to the hot combustion gas of a current generation gas turbine engine without some form of cooling and/or thermal insulation.
Thermal barrier coatings are widely used for protecting various hot gas path components of a gas turbine engine. The reliability of such coatings is critical to the overall reliability of the machine. The design limits of such coatings are primarily determined by laboratory data. However, validation of thermal barrier coating behavior when subjected to the stresses and temperatures of the actual gas turbine environment is essential for a better understanding of the coating limitations. Such real world operating environment data is very difficult to obtain, particularly for components that move during the operation of the engine, such as the rotating blades of the turbine.
Despite the extreme sophistication of modern turbine engines, such as gas turbines for generating electrical power or aircraft engines for commercial and military use, designers and operators have very little information regarding the internal status of the turbine engine components during operation. This is due to the harsh operating conditions, which have prevented the use of traditional sensors for collecting reliable information of critical engine components.
Many current turbines are equipped with sensors capable of limited functions such as exhaust gas-path temperature measurements, flame detection and basic turbine operating conditions. Based on this information, turbine operators such as utility companies operate engines in a passive mode, in which maintenance is scheduled based on prior histories of similar engines. Engine rebuilds and routine maintenance are performed in the absence of a prior knowledge of the remaining or already utilized life of individual components. The lack of specific component information makes early failure detection very difficult, often with the consequence of catastrophic engine failure due to abrupt part failure. This results in inefficient utilization, unnecessary downtime and an enormous increase in operating cost.
Currently, the gas turbine industry approach is to depend on the measurement of gas path temperature, which is related back to specific component problems based on experience and history regarding a class of engines. This approach is highly subjective and only allows for determining already severe situations with an engine. It does not provide indications of impending damage or insight into the progression of events leading up to and causing engine damage due to component degradation or failure.
The instrumentation of a component such as a blade or vane within a steam turbine typically includes placing wire leads on the balance wheel, which continue on to the blade airfoil. The wire leads are typically held together by an epoxy. These wires are routed from within the component to the turbine casing. The pressure boundary of a component may be breached to introduce a sensor such as a thermocouple and a braze is back filled to hold the thermocouple in place. Each thermocouple sensor has wire leads coming out of the component that are connected back to a diagnostic unit. Instrumenting a plurality of components of a turbine in this manner results in an extensive network of wires just for monitoring the single operating condition of temperature. Instrumenting components using this technique is expensive, which is a barrier to instrumenting a large number of components within a single turbine. Further, the wire leads and data transfer is frequently poor, which can result in costly repairs and flawed data analysis.
Using thermocouples for temperature measurements in the gas path of a turbine may be disadvantageous because it only provides feedback to an operator that a temperature change has occurred in the gas path. It does not provide any indication as to why the temperature change has occurred. For diagnosing problems with blades or vanes based on a measured temperature change, there has to be a historical correlation between the measured temperature differential and the specific problem, such as a hole in a vane. This correlation is difficult and time consuming to derive to within a reasonable degree of certainty and needs to be done on an engine-by-engine basis taking into account turbine operation conditions. When a temperature differential is measured, it is difficult, if not impossible, to predict what the problem is or where it is located. Consequently, the turbine must typically be shut down and inspected to determine the scope of repair, replacement or other maintenance to be performed.
In any application, combustion turbines are routinely subject to various maintenance procedures as part of their normal operation. Diagnostic monitoring systems for gas turbines commonly include performance monitoring equipment that collects relevant trend and fault data used for diagnostic trending. In diagnostic trend analysis, certain process data (such as exhaust gas temperature, fuel flow, rotor speed and the like) that are indicative of overall gas turbine performance and/or condition are compared to a parametric baseline for the gas turbine. Any divergence of the raw trend data from the parametric baseline may be indicative of a present or future condition that requires maintenance. Such diagnostic monitoring systems can only predict or estimate specific component conditions and do not collect data from or provide any analysis with respect to the actual condition of a specific component itself.
In this respect, conventional methods of predicting component failure for gas turbines and of scheduling maintenance have not been entirely accurate or optimized. The traditional “duty cycle” used for predictive maintenance does not reflect real operational conditions, especially off-design operations. The actual life of specific components of a gas turbine depends strongly on the actual usage of that gas turbine and the specific components within the turbine.
For example, elevated temperatures and stresses within the turbine, and aggressive environmental conditions may cause excessive wear on components in the turbine beyond that predicted with the standard design duty cycle. Off-design operating conditions, which are often experienced by industrial gas turbines, are not reflected by the standard duty cycles. The actual life of components in the gas turbine may be substantially less than that predicted by the design duty cycle. Alternatively, if more favorable conditions are experienced by an actual gas turbine than are reflected in the design duty cycle, the actual component life may last substantially longer than that predicted by maintenance schedules based on the design duty cycle. In either event, the standard design duty cycle model for predicting preventive maintenance does not reliably indicate the actual wear and tear experienced by gas turbine components.
Known techniques for predicting maintenance and component replacement rely on skilled technicians to acquire or interpret data regarding the operation of a combustion turbine. Such techniques are subject to varying interpretations of that data by technicians. Technicians may manually evaluate the operational logs and/or data collected from gas turbines. Technicians, for example, may evaluate start and stop times and power settings to determine how many duty cycles had been experienced by the gas turbine, their frequency, period and other factors. In addition, if the data log of a gas turbine indicated that extraordinary conditions existed, such as excessive temperatures or stresses, the technicians may apply “maintenance factors” to quantify the severity of these off-design operational conditions.
None of these techniques provides accurate information with respect to the actual condition of a specific component or component coating, which may lead to unnecessary repair, replacement or maintenance being performed causing a significant increase in operating costs. Monitoring systems and methods now exist that incorporate sensors mounted directly on components by embedding the sensors within component coatings, such as thermal barrier coatings, for detecting component operating conditions such as wear of a component, heat flux across a component coating, spallation of a coating, strain across an area of a component or crack formation within a component substrate or coating. Such systems may have capabilities for the wireless transmission of data relative to component operating conditions, and provide more accurate real-time data relative to such operating conditions. However, such conditioned based monitoring/maintenance systems have not been incorporated into component or engine life prediction systems or methods.
The invention is explained in the following description in view of the drawings that show:
Returning to
In use, air is drawn in through compressor 12, where it is compressed and driven towards combustor 14. Combustor 14 mixes the air with fuel and ignites it thereby forming a working gas. This working gas will typically be above 1300° C. This gas expands through turbine 16, being guided across blades 18 by vanes 22. As the gas passes through turbine 16, it rotates blades 18 and shaft 20, thereby transmitting usable mechanical work through shaft 20. Combustion turbine 10 may also include a cooling system (not shown), dimensioned and configured to supply a coolant, for example, steam or compressed air, to blades 18 and vanes 22.
The environment wherein blades 18 and vanes 22 operate is subject to high operating temperatures and is particularly harsh, which may result in serious deterioration of blades 18 and vanes 22. This is especially likely if the thermal barrier coating 26 should spall or otherwise deteriorate. Embodiments of the invention are advantageous because they allow components to be configured for transmitting data indicative of a component's condition during operation of combustion turbine 10. Blades 18, 19, vanes 22, 23, and coatings 26, for example, may be configured for transmitting component specific data that may be directly monitored to determine the respective condition of each component during operation and to develop predictive maintenance schedules.
Embodiments of the present invention allow for a plurality of sensors to be embedded within the respective coatings of a plurality of components within combustion turbine 10. Alternate embodiments allow for the sensors to be surface mounted or deposited onto components, especially those contained in areas where components do not require a barrier coating such as the compressor. Exemplary embodiments of sensors may be used to provide data to system 30 with respect to physical characteristics of a component and/or properties of a component's coating as well as other component or coating specific information.
For example, exemplary sensors may be used to detect wear between two components, measure heat flux across a component's coating, detect spallation of a coating, measure strain across an area of a component or determine crack formation within a component or coating. Those skilled in the art will recognize other properties and/or characteristics of a component or component coating that may be measured and/or detected in accordance with aspects of the invention.
It will be appreciated that aspects of the invention allow for various sensor configurations to be embedded within a barrier coating such as a barrier coating 26 of blades 18 or vanes 22 of turbine 16. U.S. Pat. Nos. 6,838,157; 7,270,890; 7,368,827; and, 7,618,712, which are specifically incorporated herein by reference, describe various embodiments of methods for instrumenting gas turbine components, such as blades 18, 19 and vanes 22, 23 that may be utilized for depositing sensors in accordance with aspects of the present invention. These patents disclose various methods of forming trenches in a barrier coating, forming a sensor in the coating and depositing a backfill material in the trench over the coating. Embodiments of those methods and components may be used to form smart components as disclosed herein.
U.S. Pat. No. 6,576,861, which is specifically incorporated herein by reference, discloses a method and apparatus that may be used to deposit embodiments of sensors and sensor connectors with transmitters in accordance with aspects of the present invention. In this respect, methods and apparatuses disclosed therein may be used for the patterning of fine sensor and/or connector features of between about 100 microns and 500 microns without the need of using masks. Multilayer electrical circuits and sensors may be formed by depositing features using conductive materials, resistive materials, dielectric materials, insulating materials and other application specific materials. It will be appreciated that other methods may be used to deposit multilayer electrical circuits and sensors in accordance with aspects of the invention. For example, thermal spraying, vapor deposition, laser sintering and curing deposits of material sprayed at lower temperatures may be used as well as other suitable techniques recognized by those skilled in the art.
Embodiments of the invention allow for a plurality of sensors 50 to be deployed in numerous places within combustion turbine 10 for monitoring component-specific or coating-specific conditions as well as collecting other data with respect to the operation or performance of combustion turbine 10. For example,
The sensors 50 may be incorporated in wireless telemetry systems as that disclosed in U.S. Publication No. 2009/0121896 and U.S. application Ser. No. 13/015,765, which are incorporated herein by reference. Such telemetry systems utilize power induction systems, such as resonant energy transfer systems or induction coil systems and incorporate transceivers for the wireless transmission electronic data. The transceivers are provided in electrical communication with sensors 50 for the transmission of electronic data signals that may be indicative of an operating condition of a component.
With respect to embodiments of the present invention, the above-described sensors and wireless telemetry systems are provided for communication with the control system 30, including the antenna 32 and receiver 33 for receiving electronic data signals indicative of one or more operating conditions associated with components such as blades 18, 19 and vanes 22, 23 and/or the coatings on such components of the turbine engine 10. The database 36 may include historical data relative to past operating conditions of the engine 10 and the components of the engine. For example, such historical data may include the different loads under which the turbine engine 10 has been operating and the amount of time the turbine engine 10 has been operating under each such loads and the inlet and exhaust turbine temperatures and pressures over time, the number of cycles that the engine has operated in and the rate of each such cycle. This historical data relative to the engine 10 and its components may also include data relative to certain ambient parameters that can affect the condition of the engine 10, the components and the occurrence of failure modes. For example, turbine engines that are used in power generation plants are located at various geographic locations. Accordingly, data relative to ambient temperature, humidity and air pressure may be provided. As explained in more detail, this historical data relative to operating conditions of the engine and the component are input into algorithms to determine a remaining useful life of multiple components of the engine to more efficiently and accurately determine maintenance schedules for the engine 10.
In addition, the database 36 may include historical data relative operating conditions of the components over the duration of the operating period or periods of the turbine engine 10. Such conditions may include data relative to conditions associated with the thermal barrier coating such as temperature of the component and fluid (air, exhaust, steam, etc.) pressure across the component. Other conditions may relate to a component substrate such as vibration information (frequency and amplitude data for vibrational movement), distortion (bending/twisting) of the substrate. As explained in more detail below, these operating conditions are continuously monitored over time and are associated with identified failure modes in order to estimate a remaining useful life of certain components and the engine 10.
In addition to the above-described historical data relative to components, data relative to future planned operating parameters or condition of the engine may be provided, which may be stored in database 36 or any other available memory that is accessible by the processor 34. As indicated above relative to historical engine data, the future planned operation of the engine may refer to the different loads under which the engine 10, exhaust temperature and pressure and time of operating at such conditions. The processor or CPU 34 is programmed or configured to access 1) the historical data relative to the operating conditions of the engine; 2) the future planned operating conditions of the turbine engine; and, 3) the current operating conditions of the components as provided by the above-referenced sensor and telemetry systems, to determine a remaining useful life of one or more components of the engine 10 and the remaining useful life of the engine. Based on this determination, the processor 34 is able to render a decision when to shut the,engine down for maintenance and/or service. In addition, the below-described predictive life curves may be incorporated into one or more life prediction algorithms to determine the remaining useful life of components and the engine 10.
With respect to
In reference to
In an embodiment of the present invention, coatings such as a thermal barrier coating, which may or may not include a bond coat, applied to a substrate of the blades 18, 19 and vanes 22, 23 are monitored for purposes of determining a remaining useful life of a component. Such coatings may have different failure modes such as spallation and coating depletion, which may be the result of growth of oxidation in the coating. Thus, the above-described sensors 50 deposited in connection with a component are embedded within or on a coating for monitoring an operating condition associated with the component and coating. More specifically, a component operating condition that is monitored and associated with coating depletion and spallation is TBC temperature. Accordingly, heat flux sensors including thermocouples may be embedded in the TBC for monitoring temperature of a component. In addition, strain gauges may be affixed to a substrate to monitor conditions such as static and dynamic vibrational modes, which may be associated with spallation, cracking or distortion failure modes.
In a second step 54, a predicted “failure mode rate” is generated for each component and an associated failure mode identified in the first step 52. A failure mode rate or predicted failure mode rate may be defined as the estimated or predicted occurrence of a failure mode of a component as a function of one or more operating conditions of the component. Examples of predicted failure mode rates are represented in the curves shown in
With respect to
The occurrence of a failure mode, whether the failure mode occurs in the coating or substrate, is typically a function of time and temperature. That is, as the engine operates over time at extreme temperatures, as in a turbine engine, the component substrate or coating approaches one or more failure modes. Some turbine engines are operated as “baseload” engines wherein the engine 10 is operated for an extended period of time before being brought down for maintenance. For example, some such engines may be operated for as long as three years before the engines are shut down for maintenance or service. For such engines and the components, predicting a failure mode is dependent largely on time and temperature. However, other engine operating conditions besides time and temperature may affect failure modes. More specifically, other engines referred to as “peak” engines, are operated for much shorter time durations, for example, one or two days, which are also referred to as cycles. Therefore, over a predetermined time a turbine engine 10 may be operated a set number of cycles. In addition, the rate of the cycles is also an operating condition that may be considered.
The rate of a cycle is the amount of time required to bring an engine to operate at a predetermined output or the amount of time to shut down an engine up and cooling it to ambient temperature. The number of cycles and cycle rates places stress on the interface between the TBC and bonding coat and/or substrate, which stress may create cracking at the interface. This cracking then leads to spallation. Accordingly, the predicted remaining useful life of a component due to spallation is illustrated as a function of time, temperature, number of cycles and cycle rates.
In an embodiment, laboratory testing and/or mathematical modeling is performed considering the operating conditions of a component and engine that most influence the above-identified failure modes. In addition, historical data relative to the operating conditions of a turbine engine and its components may be used to develop these predicted failure modes. This may be obtained by mining data relative to turbine fleet information for turbines of similar design and performance. As described above these failure mode rates may be utilized to display trends in coating degradation in turbine engine components, or other failure modes such as cracking or distortion of the substrate of a component. Moreover, the curves provide data relative to an estimated rate at which a component approaches a failure mode or data relative to a time when the failure mode is predicted to occur.
While the above-described failure mode rates are represented in the form of predictive curves, data related to the predicted failure mode rates may be represented in other forms or formats. For example, the failure mode rate data may be simply presented in the form of providing one or more tables identifying a component (Row One Blade) and an estimated time (8,000 hours) at which a failure mode (spallation) will occur.
As provided in step 56, using the above-described failure mode data with respect to
Once the appropriate sensors 50 and the locations of sensors 50 on the components are selected, the sensors 50 are calibrated as provided in step 58. In an embodiment, sensor calibration data as represented in the curve shown in
Fabrication of sensors and wireless telemetry system (data acquisition system) is then performed at step 60. As indicated above, the sensors 50 and wireless telemetry components are fabricated in accordance with the methods and materials disclosed in U.S. Pat. Nos. 6,838,157; 7,270,890; 7,368,827; and 7,618,712; and U.S. Publication No. 2009/012189; all of which have been specifically incorporated herein by reference.
With the engine 10 prepared for operation, including having the sensors 50 linked with the wireless telemetry system and control system 30, a planned operation of the engine is input into the control system 30. This planned operation includes data relative to a desired power output (i.e., load) for the turbine engine 10 and the amount of time over which the turbine engine 10 will operate at the load. This data may also include an estimated operating temperature of the engine including one or more estimated operating temperature for different stages of the turbine engine 10. In addition, ambient conditions such as ambient temperature and air pressure, which may vary according to geographic location of the engine 10, may be considered in developing a planned operation. For example, the desired power output for engines, such as the above-referenced “baseload” engines that operate in a power plant environment over extended periods of time, may vary over the course of a year. During certain times of the year the power output demand may be greater than other times of the year.
Again in reference to
Accordingly, at step 64 information (historical data) regarding past engine operating conditions is input into life prediction algorithms and the above-described life predictive models (
The database 36 may also include data relative to a threshold value of a remaining useful life associated with each failure mode for each identified life-limiting component. The processor 34 may be configured such that when the determined or calculated remaining useful life falls below this threshold value, or is within a predetermined range above the threshold value, to transmit an audible and/or visual notification of service or maintenance for the component and engine. This threshold value of the remaining useful life may be taken from the predicted trend of a failure mode as shown in
However, as this threshold value may be predictive in nature and not based on actual operating conditions of the engine, because it is based on models developed using data such as turbine engine fleet data, other steps may be taken to provide a more accurate remaining useful life. Accordingly, in step 66 data relative to presently planned and future planned operating conditions is input or provided. That is, the above-mentioned planned operation developed for the engine 10 at any time during the operation of the engine includes the present operating conditions of the engine including power output and temperature and the operating conditions of engine that are programmed for the future. In an embodiment of the invention, the processor 34 is programmed to input the data relating to the present and future planned engine operating conditions into life prediction algorithms to calculate a remaining useful life, as set forth in steps 66 and 68.
With respect to
where v is the velocity at which the critical operating conditions responsible for the specified failure mode such as spallation and coating depletion.
In an embodiment of the invention, the processor 34 is configured to determine the remaining useful life for each identified failure mode for each identified life-limiting component. This information may be displayed in any available format. For example, the display 38 may list all life-limiting components and the remaining useful life associated with each failure mode. Alternatively, the display 38 may show only a single number (hours) representative of the remaining useful life of a single component. This number would represent the shortest remaining useful life calculated for a failure mode.
The graph shown in
An alternative display is shown in
While various embodiments of the present invention have been shown and described herein, it will be obvious that such embodiments are provided by way of example only. Numerous variations, changes and substitutions may be made without departing from the invention herein. Accordingly, it is intended that the invention be limited only by the spirit and scope of the appended claims.
