Gas turbine engine with active protection against pulsations and method of operating the same

Abstract

The present invention relates to a gas turbine engine having active protection against pulsations and a method of operating the same. A control system (3) includes a pulsation sensor (12) and a processing unit (13) coupled to the pulsation sensor (12). The processing unit (13) determines an azimuthal component (#imgabs1#) of a pulsation signal (P N (t)) defined by #imgabs0#, where NC is the number of can-type burners (8) in a burner assembly (3), #imgabs2# is the azimuth of the Nth can-type burner (8), and M is the order of at least one azimuthal component (#imgabs3#) in the range [-ceil(NC / 2)+1; floor(NC / 2)]. The processing unit determines a working time series (TSW N ) of spectral parameters (A1; A1, A2; F1) from the pulsation signal (P N )(t)) and selectively activates a protection measure based on the working time series (TSW N ) and an activation criterion. In response to activation of the protection measure, the control system (3) modifies the air supply and / or fuel supply to the burner assembly (3).

Description

[0001] Cross - reference to related applications

[0002] This patent application claims priority to European Patent Application No. 19220166.3, filed on December 31, 2019, the entire disclosure of which is incorporated herein by reference. Technical field

[0003] The present invention relates to a gas turbine engine having active protection against pulsations and to a method of operating a gas turbine engine. Background art

[0004] It is well known that in the past few years, flexible control of gas turbines in power generation equipment has become the most important goal. On the one hand, in fact, the awareness of environmental risks has prompted regulations that set increasingly strict requirements. On the other hand, the organization of the modern power market and the continuously changing demands do not allow operating power generation equipment under constant load conditions. Instead, the need to meet demand fluctuations (including sudden increases or decreases) and contribute to controlling the grid frequency requires flexible operation.

[0005] However, such flexible operation increases the risk of the gas turbine suffering critical conditions because a wide operating range has to be considered. In addition, the aging of components may cause a drift of critical conditions, such that dangerous vibration states may occur during the machine's service life due to deviation from the design characteristics. In fact, in the field of gas turbines, it is well known that mechanical and thermo - acoustic vibrations are generated under different operating conditions. Such conditions depend on the response of the complex structure of the gas turbine and can vary widely according to the type and size of the gas turbine and also change over time, for example, due to the aging of components. During the design and test steps of the gas turbine, oscillating operating conditions and frequencies may become apparent, and thus protective measures such as mufflers may be provided. However, when the conditions change, such measures may become ineffective.

[0006] Therefore, poor combustion conditions can cause serious damage to the machine structure because certain parts (such as thermal insulation coatings) may be vulnerable to vibrations and thermo - acoustic pulsations, especially at certain frequencies. As a result, maintenance is required more frequently than desired. Moreover, despite regular maintenance, turbine parts may lose their effectiveness and no longer be able to fully perform their functions. For example, in the case of thermal insulation coatings, loss of effectiveness may lead to severe and even irreversible damage to the metal components that are no longer adequately protected.

[0007] Other serious events, such as excessive levels of pollutant emissions, may not pose a danger to the machine itself but should be avoided anyway to comply with laws and regulations and also because the efficiency of the machine is significantly affected.

[0008] Accordingly, there is a general desire to improve the protection of gas turbines against critical operating conditions under which dangerous vibrations and / or thermoacoustic pulsations may occur. Summary of the Invention

[0009] Accordingly, an object of the present invention is to provide a gas turbine engine and a method of operating a gas turbine that allow overcoming or at least alleviating the above limitations.

[0010] According to the present invention, there is provided a gas turbine engine comprising:

[0011] a burner assembly having a plurality of burner units; and

[0012] a control system configured to control the air supply and fuel supply to the burner assembly;

[0013] The control system comprises:

[0014] a pulsation sensor coupled to a respective burner unit and configured to provide a pulsation signal from the respective burner unit, each burner unit being provided with a respective one of the pulsation sensors; and

[0015] a processing unit coupled to the pulsation sensor to receive the pulsation signal and configured to:

[0016] determine at least one azimuthal component of the pulsation signal, the azimuthal pulsation component being defined by:

[0017]

[0018] where NC is the number of can-type burners in the burner assembly, is the azimuth of the Nth can-type burner of the burner assembly, i is the imaginary unit, and M is the order of at least one azimuthal component in the range [-ceil(NC / 2)+1; floor(NC / 2)];

[0019] determine a working time series of spectral parameters from at least one azimuthal component of the pulsation signal; and

[0020] selectively activate a protection measure based on the working time series and an activation criterion.

[0021] Monitoring the spectral parameters of the pulsation signal allows the identification of adverse oscillation phenomena at a very early stage and is insensitive to the aging of components. More precisely, the monitoring performed by the control system can effectively identify the early occurrence of adverse phenomena, regardless of whether such phenomena are associated with known critical conditions or are caused by unpredictable events such as aging.

[0022] The protection measures can change the operating conditions of the gas turbine by modifying the air and / or fuel supply in order to reduce the impact of the oscillation phenomenon. In some cases, very small changes may be sufficient to bring the gas turbine to safe operating conditions with a low risk of dangerous oscillations. Additionally, when a clear trend is shown by the time series, the protection measures can be triggered, thus avoiding unnecessary interventions in response to fluctuations (such as statistical errors) that do not actually reflect critical conditions. The control system can be configured to respond to the activation of the protection measures by directly modifying the air supply and / or fuel supply to the burner assembly, or alternatively by sending an alarm or warning message to prompt the operator to take action, such as changing the operating set point and thus changing the air supply and / or fuel supply.

[0023] According to one aspect of the invention, the activation criterion includes a value criterion defined by a non-permissible value range of the spectral parameter, in conjunction with a time criterion based on the persistence of the spectral parameter within the non-permissible value range.

[0024] The use of the combined criterion helps to reduce the activation of unnecessary protection measures.

[0025] According to one aspect of the invention, the time criterion includes at least one of the following:

[0026] The spectral parameter of the working time series continuously remains within the non-permissible value range for at least a first time interval;

[0027] The spectral parameter of the working time series continuously or intermittently remains within the non-permissible value range within the inspection window for at least a cumulative second time interval;

[0028] The time counter reaches a count threshold, and the counter is incremented by the processing unit when the spectral parameter of the working time series is within the non-permissible value range, and remains unchanged or is decremented by the processing unit when the spectral parameter of the working time series is outside the non-permissible value range.

[0029] Thus, appropriate time criteria can be made available, and the appropriate time criteria can be used according to design preferences in order to achieve a balance between timely activation of the protection measures and effective rejection of false positives.

[0030] According to one aspect of the invention, the spectral parameter includes at least one peak amplitude within a corresponding programmed frequency range, and the processing unit is further configured to compare the working time series of the peak amplitude with a corresponding independent amplitude threshold, and the non-permissible value range includes peak amplitude values greater than the independent amplitude threshold.

[0031] In this way, the strongest oscillations can be easily identified and distinguished from background noise that is not an indicator of potential dangerous conditions, thus allowing timely activation of the protection measures whenever critical oscillations occur at any of the monitored burner units.

[0032] According to one aspect of the invention, the spectral parameter includes a plurality of peak amplitudes in a respective programmed frequency band for at least one azimuthal component of the pulsation signal, and the processing unit is further configured to compare a combination of the working time series of the peak amplitudes with a combined amplitude threshold, and a range of non-permitted values includes values greater than the combined value of the combined amplitude threshold.

[0033] Monitoring the combination of peak amplitudes in multiple frequency bands can reveal the effects of non-linear coupling of acoustic modes that might otherwise be missed. Thus, the protection is further improved.

[0034] According to one aspect of the invention, the spectral parameter includes at least one peak frequency of the respective peak amplitude, and the processing unit is further configured to compare the working time series of the peak frequency with a lower frequency threshold and an upper frequency threshold, and a range of non-permitted values includes the frequency band bounded by the lower frequency threshold and the upper frequency threshold.

[0035] When the oscillation approaches the eigenfrequency of the structure, the structure of the burner assembly may sometimes exhibit resonance behavior even in a very narrow band. Thus, unwanted oscillations can even be triggered by relatively small amplitudes in the critical band, i.e., relatively small oscillations near the eigenfrequency may amplify and cause resonance phenomena. Monitoring the oscillation peak frequency allows preventing the gas turbine from remaining in operating conditions that may trigger resonance. In many cases, making a small adjustment to the air and / or fuel supply may change the operating conditions sufficiently to avoid resonant oscillations.

[0036] According to one aspect of the invention, the processing unit is configured to determine azimuthal components of all orders within the range [-ceil(NC / 2)+1; floor(NC / 2)].

[0037] According to one aspect of the invention, the processing unit is configured to determine a combination of azimuthal components of different orders or all orders within the range [-ceil(NC / 2)+1; floor(NC / 2)] and determine a working time series from the combination of azimuthal components.

[0038] Thus, more complete information can be obtained for the purpose of protecting against resonant oscillations not only based on the pulsations occurring at individual burner units, but also based on the correlations between the pulsations at different burner units.

[0039] According to one aspect of the invention, the processing unit is configured to determine the in-phase pulsation component as the 0th order azimuthal component of the pulsation signal and determine a working time series from the in-phase pulsation component, where the in-phase pulsation component is defined by the following formula:

[0040]

[0041] The in-phase pulsation component may cause torsional vibration of the rotor. Therefore, monitoring the spectral parameters of the in-phase pulsation component allows activation of protective measures also directed at the torsional vibration of the rotor, and thus torsional vibration of the rotor can be prevented.

[0042] According to one aspect of the invention, the processing unit is configured to determine an original time series of spectral parameters from the pulsation signal and to determine a working time series from the respective original time series by low-pass filtering.

[0043] Thus, insignificant false positives and short-lived transient phenomena can be ignored, and unnecessary activation of protective measures can be avoided.

[0044] According to one aspect of the invention, there is also provided a method of operating a gas turbine, the gas turbine comprising a burner assembly having a plurality of burner units and a control system configured to control the air supply and fuel supply to the burner assembly, each burner unit (8) being provided with a respective pulsation sensor (12);

[0045] The method comprises:

[0046] providing a pulsation signal from each of the burner units;

[0047] determining at least one azimuthal component of the pulsation signal, the azimuthal pulsation component being defined by:

[0048]

[0049] where NC is the number of cylindrical burners in the burner assembly, is the azimuth of the Nth cylindrical burner of the burner assembly, i is the imaginary unit, and M is the order of at least one azimuthal component in the range [-ceil(NC / 2)+1; floor(NC / 2)];

[0050] determining a working time series of spectral parameters from at least one azimuthal component of the pulsation signal; and

[0051] selectively activating protective measures based on the working time series and an activation criterion;

[0052] where activating the protective measures includes modifying the air supply and / or fuel supply to the burner assembly. Description of the Drawings

[0053] The present invention will now be described with reference to the drawings, which show some non-limiting embodiments of the present invention, in which:

[0054] - Figure 1 is a simplified block diagram of a gas turbine engine according to an embodiment of the present invention;

[0055] -Figure 2 is Figure 1 a schematic front view of a burner assembly of a gas turbine engine;

[0056] - Figure 3 is a flow chart of a method performed by a control system of a gas turbine according to an embodiment of the present invention; Figure 1 a gas turbine engine;

[0057] - Figure 4 is a diagram showing spectral quantities used in the method of; Figure 3 a gas turbine engine;

[0058] - Figure 5 and Figure 6 is a diagram showing a time series used in the method of; Figure 3 a gas turbine engine;

[0059] - Figures 7-9 is a diagram showing a time series and alternative activation criteria available in the method of; Figure 3 a gas turbine engine;

[0060] - Figure 10 is a diagram showing a time series used in a method performed by a control system of a gas turbine according to another embodiment of the present invention; Figure 1 a gas turbine engine;

[0061] - Figure 11 is a diagram showing activation criteria available in the method of; Figure 10 a gas turbine engine;

[0062] - Figure 12 is a diagram showing a time series used in a method performed by a control system of a gas turbine according to another embodiment of the present invention. Figure 1 a gas turbine engine; DETAILED DESCRIPTION

[0063] Referring to Figure 1 , the numeral 1 generally defines a gas turbine installation, which includes a gas turbine engine 2 and a control system 3. The gas turbine engine 2 in turn includes a compressor 4, a burner assembly 5, and an expansion section or turbine 6, all of which extend about an axis A. The compressor 4 and the turbine 6 share a common rotor 7.

[0064] The compressor 4 supplies an air flow drawn in from the outside to the burner assembly 5. The air supply to the compressor 4 can be controlled by the control system 3 by adjusting the orientation of the inlet guide vanes 4a of the compressor 4.

[0065] The burner assembly 5 (e.g., a can-type or can-annular burner) is provided with a plurality of burner units 8, which are circumferentially distributed about the axis A, as Figure 2As schematically shown. The burner unit 8 mixes air from the compressor 4 and fuel from the fuel supply line 9 to form a mixture for combustion. The fuel can be gaseous, such as natural gas or syngas, or liquid, such as diesel. The gas turbine engine 2 can be configured to use different types of fuels, both gaseous and liquid. The fuel supply can be controlled by the control system 3 via the fuel valve system 10.

[0066] The turbine 6 receives and expands the hot gas flow from the burner assembly 5 to extract mechanical work, which is transferred to an external user, typically a generator, not shown here.

[0067] The control system 3 includes a pulsation sensor 12, a processing unit 13, and a controller 14. The pulsation sensor 12 can be, for example, a pressure sensor or an optical sensor.

[0068] The pulsation sensor 12 is connected to the corresponding burner unit 8 and is configured to provide a pulsation signal P N (t) in response to pulsations at the corresponding burner unit 8. In one embodiment, each burner unit 8 is provided with a corresponding pulsation sensor 12, thus providing pulsation signals P 1 (t), P 2 (t), …, PNC(t), where NC is the number of burner units 8 in the burner assembly 5. However, according to design preferences, other configurations with pulsation sensors 12 associated only with selected burner units 8 can be used. Any kind of pulsation sensor capable of detecting pressure oscillations inside the burner unit 8 and / or mechanical vibrations of the structure of the burner unit 8 can be used for this purpose. For example, piezoelectric pressure or vibration sensors or strain gauges can be used. The pulsation sensor 12 can be provided with detection electronics to provide the pulsation signal P N (t) in a form determined according to design preferences (e.g., as a voltage signal). For example, piezoelectric sensors typically provide a charge signal as an output. Thus, the detection electronics can include a charge amplifier to convert the charge output of the piezoelectric sensor into a voltage output.

[0069] The processing unit 13 is connected to the pulsation sensor to receive the pulsation signal P N (t). Through the Figure 3 process shown in N (t) of the pulsation signal P N to determine the working time series TSW Figures 4-6 (see NThe activation criteria selectively activate the protection measures. In one embodiment, the protection measures may include an automatic response of the control system 3. In this case, the control system is configured to modify the air supply and / or fuel supply to the burner assembly in response to the activation of the protection measures. The protection measures may also include sending an alert or warning message to prompt the operator to take measures, such as changing the operating set point, and thus changing the air supply and / or fuel supply.

[0070] Generally, a time series is a sequence of data points indexed in chronological order. The time series may be a sequence of data with equal time intervals and is thus a discrete-time data series. As understood herein, the working time series is a time series of spectral parameters used in the processing unit 13 for the purpose of implementing pulsation protection. As understood herein, the original time series is a time series directly obtained as the values of spectral parameters calculated at successive time points. In other words, the pulsation signals P 1 (t), P 2 (t), …, PNC(t) are spectrally analyzed, and the spectral parameters of interest are extracted. The successive values of the spectral parameters are extracted at each time point that defines the original time series. The working time series may be identical to the original time series or may be derived from the original time series through processing. The processing may include, for example, filtering and defining such that the resulting working time series is more suitable for implementing pulsation protection.

[0071] Specifically, the processing unit 13 iteratively converts the received analog pulsation signal P N (t) into the corresponding digital signal ( Figure 3 , block 100), and applies the fast Fourier transform (block 110) to determine the spectral sequence for each pulsation signal P N (t). At each iteration, the processing unit 13 extracts selected spectral parameters of the spectrum (block 120; also see Figure 4 , where the extracted spectral parameter is the peak amplitude A 1 , A 2 ), and updates the working time series TSW N based on the extracted spectral parameters (blocks 130 - 140). In other words, each time series TSW N is based on a sequence of values of the corresponding one of the spectral parameters as determined by the processing unit 13 in the subsequent iteration. The values of the working time series TSW N do not have to be the exact values of the spectral parameters, and processing may be added. In one embodiment, for example, the processing unit 13 updates (block 130) the original time series TSR N of the spectral parameters of the pulsation signal P N (i.e., the sequence of exact digital values of the spectral parameters, see Figure 5 for the peak amplitude A 1) and determine a working time series TSW from the corresponding original time series TSR by low-pass filtering N Determine the working time series TSW N (block 140; also see Figure 6 The filtered peak amplitude A 1 ). Thus, the working time series TSW N is smoothed to reduce false positives and avoid unnecessary activation of protection measures. However, in other embodiments, the original time series of spectral parameters can be directly used as the working time series TSW N .

[0072] In addition, at each iteration, the processing unit 13 checks the updated time series TSW with respect to the activation criteria N (block 150), and if the activation criteria are met, activates the protection measures (block 150, output yes; block 160). Then, a new iteration of the process is performed by the processing unit 13 through the steps of A / D conversion, fast Fourier transform, spectral parameter extraction, update of the working time series TSW N and checking of the activation criteria (blocks 100 - 150). If the activation criteria are not met (block 150, output no), the processing unit 13 directly starts a new iteration of the process.

[0073] The activation criteria can include a combination of a value criterion and a time criterion. The value criterion is defined by a non-permissible value range of spectral parameters, such as the time series being within a prohibited region. The time criterion is based on the persistence of the spectral parameters within the non-permissible value range.

[0074] The time criterion can include one or more of the following conditions:

[0075] - The spectral parameters of the working time series TSW N continuously remain within the non-permissible value range for at least a first time interval T 1 (in the example of Figure 7 , which refers to the working time series TSW of the peak amplitude A 1 , the non-permissible value range IRVI includes the peak amplitude value region greater than the independent amplitude threshold ATH N ); N );

[0076] - The spectral parameters of the working time series TSW N continuously or intermittently remain within the non-permissible value range IRVI within the check window W for at least an accumulated second time interval T 2 ( Figure 8 , where CT is the accumulated time in the prohibited region within the check window W); and

[0077] - The time counter TC reaches a count threshold C 0, the time counter TC is incremented by the processing unit 13 when the spectral parameter of the working time series TSW N is within the inadmissible value range IRVI, and otherwise remains unchanged ( Figure 9 , solid line) or is decremented ( Figure 5 , dashed line).

[0078] The protection measures may include changing the operating conditions of the gas turbine engine 2. To this end, the processing unit 13 may adjust the setpoint of the controller 14, whereby the control system 3 modifies the air supply and / or fuel supply to the burner assembly 5 in response to the activation of the protection measures.

[0079] In one embodiment, the spectral parameter includes at least one peak amplitude A 1 ( Figures 4-6 ). As understood herein, "peak amplitude" refers to the amplitude of the spectral peak in one or more given monitoring frequency bands, in which dangerous pulsations may be expected. The spectral peak is typically much larger than the average spectral amplitude and the local peak amplitude outside the monitoring frequency band. In Figure 4 the example of, the first peak amplitude A 1 and the second peak amplitude A 2 are extracted in the first monitoring band B 1 and the second monitoring band B 2 respectively. However, the detection of the peak amplitude need not be limited to a specific band. For example, in one embodiment, all peaks having an amplitude exceeding a threshold may be detected. The threshold may be adaptively set, for example, based on the average value and standard deviation of the spectral values. Each frequency corresponding to the respective peak and peak amplitude is defined as the peak frequency of the respective peak. The processing unit 13 is further configured to compare the working time series TSW 1 of the subsequent value of the peak amplitude A N with the corresponding independent amplitude threshold ATH N . In this embodiment, the inadmissible value range IRVI includes peak amplitude values greater than the independent amplitude threshold ATH N . The processing unit 13 may be configured to extract peak amplitudes from a plurality of programmed frequency bands. In one embodiment, each peak amplitude is independently monitored for the activation criterion by the respective working time series TSW N .

[0080] The monitoring may extend to all burner units 8, or be limited to one or more burner units 8, which may be considered more critical, for example, from a design perspective. For example, when the activation criterion for at least one burner unit 8 is met, the protection measures may be activated, thus providing effective protection for each burner unit 8.

[0081] In one embodiment, each burner unit 8 is provided with a pulsation sensor 12, which supplies a corresponding pulsation signal P N(t). The processing unit 13 is configured to also determine one or more azimuth components P of order M from the pulsation signal P N (t) (up to NC, where NC is the number of burner units 8 in the burner assembly 5), and compare the spectral parameters (such as the amplitude peak and / or peak frequency in a given band) of the azimuth components P M (t) with corresponding thresholds (an amplitude threshold and an upper frequency threshold and a lower frequency threshold, respectively). The azimuth components P of order M M (t) are defined as M where NC is the number of burner units 8 in the burner assembly 5,

[0082]

[0083] is the azimuth of the Nth burner unit 8 among the NC burner units 8, i is the imaginary unit, and the azimuth components have an integer order M in the range [-ceil(NC / 2)+1; floor(NC / 2)], where floor(NC / 2) is the largest integer less than or equal to NC / 2, and ceil(NC / 2) is the smallest integer greater than or equal to NC / 2. The azimuth of the Nth burner unit 8 (see Figure 2 ) is defined as the angle between the reference plane RP (e.g., the radial intermediate plane of the reference burner unit 8) and the corresponding plane RPN of the Nth burner unit 8 (e.g., the radial intermediate plane of the Nth burner unit 8). 1 (For example, the radial intermediate plane of the Nth burner unit 8).

[0084] The azimuth components P of order M M (t) themselves are pulsation signals and provide further information about the pulsation state of the burner units 8. Thus, the azimuth components P of order M M (t) can be used to activate protection measures.

[0085] Then the process described in the reference Figure 3 can be applied. In fact, such a process is independent of the specific nature of the pulsation signal used and the specific spectral parameters selected for testing. In other words, the process does not depend on how (e.g., through pressure or optical or other sensors) a specific pulsation signal is generated, and does not depend on whether the pulsation signal represents the pulsation at a single corresponding burner unit 8 (e.g., the independent pulsation signal P N (t)), or is provided as a combination of pulsation signals P N (t) and thus represents the pulsation at different burner units 8 (e.g., the azimuth components P of order M M(t)). In addition, the spectral parameter can be the peak amplitude or peak frequency as defined above, and relevant amplitude and frequency thresholds can be used as appropriate to define the range of unacceptable values. Figures 4-10 and Figure 12 The spectral parameter of an example can be the azimuth component P of order M M (t) of the spectral parameter.

[0086] Therefore, after applying the fast Fourier transform, the azimuth component P of order M can be extracted M (t) of the spectral parameter, such as the peak amplitude A 1 、A 2 or the peak frequency F 1 、F 2 , and the original time series TSR of the spectral parameter can be established and iteratively updated N . The spectral parameter can include the peak amplitude and / or peak frequency of the spectral peaks in one or more or all of one or more monitoring bands for the azimuth component P M (t) in any combination, or throughout the spectrum. The original time series TSR can be smoothed, for example, by low-pass filtering N , to produce the working time series TSW N , which is then tested to meet the activation criteria defined above. The compliance test includes comparing the values of the working time series TSW M (t) of the azimuth component P N with the (azimuth) amplitude and / or frequency thresholds, as well as the matching value criteria and / or time criteria. In one embodiment, the processing unit 13 determines all azimuth components P of order M within the range [-ceil(NC / 2)+1; floor(NC / 2)] M (t). In one embodiment, the processing unit 13 determines the combination of different orders M or all orders M of the azimuth component P within the range [-ceil(NC / 2)+1; floor(NC / 2)] M (t), and determines the working time series TSW M (t) from the combination of the azimuth components P N .

[0087] In one embodiment, the processing unit 13 determines the azimuth component of order 0, that is, the in-phase pulsation component P IP (t) defined by the following formula:

[0088]

[0089] The in-phase pulsation component P IP (t) is in turn a pulsation signal. The spectral parameter includes the in-phase pulsation component P IPThe processing unit 13 is configured to update and compare the in-phase pulsation component P IP (t) The working time series TSW of the in-phase peak amplitude N The same as the pulsation threshold (in-phase pulsation threshold). The unacceptable value range includes in-phase pulsation values ​​greater than the pulsation threshold. In-phase pulsation component P IP (t) may cause torsional vibration of the rotor 7. Therefore, the in-phase pulsation component P is monitored. IP The spectral parameters of (t) allow activation of protective measures also against torsional vibrations of the rotor, thus preventing torsional vibrations of the rotor.

[0090] refer to Figure 4 and Figure 10 In one embodiment, the spectrum parameters include for each pulsatile signal P N (t) in the corresponding programming frequency band of the multiple peak amplitudes A K , and the processing unit 13 is configured to convert the peak amplitude A K Working time series TSW N The combination of and the combined amplitude threshold A 0 As a non-limiting example, the combination may be for each pulsating signal P N The peak amplitude A of (t) K The linear combination of the values ​​of 1 A 1 +…+k NI A NI , where NI is the pulse signal P N (t) is the number of significant peak amplitudes in the spectrum, and K 1 , ..., K NI is the coefficient of the linear combination. Therefore, for the pulsating signal P N The impermissible value range IRVI of (t) includes values ​​greater than the combined amplitude threshold A 0 The combined values ​​of:

[0091] IRVI={k 1 A 1 +…+k NI A NI >A 0}.

[0092] exist Figures 10-11 In the example, the two peak amplitudes A 1 , A 2 are considered as spectral parameters. Figure 10 The solid line in FIG. 1 shows an exemplary working time sequence TSW. N . Figure 11 An activation signal ACT for activating the protective measure supplied by the processing unit 13 is shown.N Remain within the non - allowed value range IRVI for a time interval exceeding a first time interval T 1 At the time interval, the activation signal ACT switches to an activation value.

[0093] Reference Figure 12 , in one embodiment, the spectral parameter includes a corresponding peak amplitude A 1 At least one peak frequency F 1 , and the processing unit is further configured to compare the working time series of the peak frequency with a lower frequency threshold FTHL and an upper frequency threshold FTHU. In this example, the non - allowed value range IRVI includes the frequency band bounded by the lower frequency threshold FTHL and the upper frequency threshold FTHU.

[0094] Finally, it is clear that changes and variations can be made to the described and illustrated gas turbine and method without departing from the scope of protection of the appended claims.

[0095] Specifically, it is clear that any combination of the described spectral parameters can be advantageously used and monitored according to design preferences for the purpose of pulsation protection.

Claims

1. A gas turbine engine, comprising: a burner assembly (5) having a plurality of burner units (8); and a control system (3) configured to control the air supply and fuel supply to the burner assembly (5); the control system (3) includes: A pulsation sensor (12) connected to a respective burner unit (8) and configured to provide a pulsation signal (P N (t)) from the respective burner unit (8), each burner unit (8) being provided with a respective one of the pulsation sensors (12); and A processing unit (13) connected to the pulsation sensor (12) to receive the pulsation signal (P N (t)), and configured to: Determine at least one azimuth component (P N (t)) of the pulsation signal (P M (t)), the azimuth pulsation component being defined by the following formula: where NC is the number of burner units (8) in the burner assembly (5), is the azimuth angle of the Nth burner unit (8) of the burner assembly (5), i is the imaginary unit, and M is at least one azimuth component (P M (t)) in the range [-ceil(NC / 2)+1; floor(NC / 2)]; Determine the spectral parameters (A N (t)) from at least one azimuth component (P M (t)) of the pulsation signal (P 1 ; A 1 , A 2 ; F 1 ) of the working time series (TSW N ); and Based on the working time series (TSW N ) and the activation criteria, selectively activate the protection measures.

2. The gas turbine engine according to claim 1, wherein, The activation criteria include a value criterion defined by a non - allowable value range (IRVI) of the spectral parameters (A 1 ; A 1 , A 2 ; F 1 ), in coordination with a time criterion based on the persistence of the spectral parameters within the non - allowable value range (IRVI).

3. The gas turbine engine according to claim 2, wherein, the time criterion includes at least one of the following: The spectral parameters (A N ; A 1 ; A 1 ; F 2 ; F 1 ) of the working time series (TSW 1 ) continuously remain within the said inadmissible value range (IRVI) for at least a first time interval (T 1 ); The spectral parameters (A N ) of the working time series (TSW 1 ; A 1 , A 2 ; F 1 ) are continuously or discontinuously maintained within the inadmissible value range (IRVI) within the inspection window (W) for at least a cumulative second time interval (T 2 ); The time counter (TC) reaches the count threshold (C 0 ), and the time counter (TC) is incremented by the processing unit (13) when the spectral parameters (A N ; A 1 ; A 1 ; F 2 ; F 1 ) of the working time series (TSW N ) are within the non-permissible value range (IRVI), and remains unchanged or is decremented by the processing unit (13) when the spectral parameters of the working time series (TSW N ) are outside the non-permissible value range (IRVI).

4. The gas turbine engine according to claim 3, wherein, The spectral parameters include at least one peak amplitude (A 1 ; A 1 , A 2 ), and the processing unit (13) is further configured to compare a working time series (TSW 1 ; A 1 , A 2 ) of the peak amplitude (A N ) with a corresponding independent amplitude threshold (ATH N ), and the non-permitted value range (IRVI) includes peak amplitude values greater than the independent amplitude threshold (ATH N ).

5. The gas turbine engine according to claim 4, wherein, The spectrum parameters include the following: N (t)) of at least one azimuthal component (P M (t)) of the corresponding programming frequency band (B 1 , B 2 ) in the multiple peak amplitudes (A 1 , A 2 ), and the processing unit (13) is further configured to convert the peak amplitude (A 1 , A 2 )Working Time Series (TSW N ) and the combined amplitude threshold (A 0 ), the impermissible value range (IRVI) includes values ​​greater than the combined amplitude threshold (A 0 )’s combined value.

6. The gas turbine engine according to any one of claims 2 to 5, wherein, The spectral parameters include at least one peak frequency (F 1 ) corresponding to a respective peak amplitude (A 1 ), and the processing unit (13) is further configured to compare a working time series (TSW 1 ) of the peak frequency (F N ) with a lower frequency threshold (FTHL) and an upper frequency threshold (FTHU), and the non-permissible value range (IRVI) includes a frequency band defined by the lower frequency threshold (FTHL) and the upper frequency threshold (FTHU).

7. The gas turbine engine according to any one of claims 2 to 5, wherein, The processing unit (13) is configured to determine the azimuth components (P M (t)) of all orders (M) within the range [-ceil(NC / 2)+1; floor(NC / 2)].

8. The gas turbine engine according to any one of claims 2 to 5, wherein, The processing unit (13) is configured to determine combinations of azimuth components (P M (t)) of different orders (M) or all orders (M) within the range [-ceil(NC / 2)+1; floor(NC / 2)], and to determine the working time series (TSW M (t)) from the combinations of azimuth components (P N ).

9. The gas turbine engine according to any one of claims 2 to 5, wherein, The processing unit (13) is configured to determine the in-phase pulsation component (P IP (t)) as the zero-order azimuth component (P N (t)) of the pulsation signal (P M (t)), and determine the working time series (TSW IP (t)) from the in-phase pulsation component (P N ), the in-phase pulsation component (P IP (t)) being defined by the following formula:

10. A method of operating a gas turbine, the gas turbine including a burner assembly (5) having a plurality of burner units (8) and a control system (3), the control system being configured to control the air supply and fuel supply to the burner assembly (5), and each burner unit (8) being provided with a corresponding pulsation sensor (12); the method comprises: Providing a pulsating signal (P IN (t)) from each of the burner units (8); Determine at least one azimuth component (P N (t)) of the pulsation signal (P M (t)), the azimuth pulsation component being defined by the following formula: where NC is the number of burner units (8) in the burner assembly (5), is the azimuth angle of the Nth burner unit (8) of the burner assembly (5), i is the imaginary unit, and M is at least one azimuth component (P M (t)) in the range [-ceil(NC / 2)+1; floor(NC / 2)]; Determine the spectral parameters (A N ; A M , A 1 ; F 1 , A 2 ; F 1 ) of the working time series (TSW N ) from at least one azimuth component (P N (t)) of the pulsation signal (P M (t)); and Based on the working time series (TSW N ), selectively activate protection measures according to the activation criteria; wherein activating the protection measure includes modifying the air supply and / or fuel supply to the burner assembly (5).

11. The method according to claim 10, wherein, The activation criteria include a value criterion defined by a non - allowable value range (IRVI) of the spectral parameters (A 1 ; A 1 , A 2 ; F 1 ), in coordination with a time criterion based on the persistence of the spectral parameters within the non - allowable value range (IRVI).

12. The method according to claim 11, wherein, the time criterion includes at least one of the following: The spectral parameters (A N ) of the working time series (TSW 1 ; A 1 , A 2 ; F 1 ) continuously remain within the said inadmissible value range (IRVI) for at least a first time interval (T 1 ); The spectral parameters (A N ; A 1 ; A 1 ; F 2 ) of the working time series (TSW 1 ) are continuously or discontinuously maintained within the inadmissible value range (IRVI) within the inspection window (W) for at least a cumulative second time interval (T 2 ); The time counter (TC) reaches the count threshold (C 0 ), and the time counter (TC) increments when the spectral parameters (A N ; A 1 ; A 1 ; F 2 ; F 1 ) of the working time series (TSW N ) are within the non - allowed value range (IRVI), and remains unchanged or decrements when the spectral parameters of the working time series (TSW N ) are outside the non - allowed value range (IRVI).

13. The method according to any one of claims 10 to 12, wherein, including determining azimuth components (P M (t)) of all orders (M) in the range [-ceil(NC / 2)+1; floor(NC / 2)].

14. The method according to any one of claims 10 to 12, wherein, including determining combinations of azimuth components (P M (t)) of different orders (M) or all orders (M) within the range [-ceil(NC / 2)+1; floor(NC / 2)], and determining the operating time series (TSW M (t)) from the combinations of azimuth components (P N ).

15. The method according to any one of claims 10 to 12, wherein, Including determining the in-phase pulsation component (P IP (t)) as the 0th azimuth component (P N (t)) of the pulsation signal (P M (t)), and determining the working time series (TSW IP (t)) from the in-phase pulsation component (P N ), the in-phase pulsation component (P IP (t)) is defined by the following formula:

Images