METHODS OF DETECTING ONE FAILURE IN A BURNER OF A COMBUSTOR AND TURBINE SYSTEMS

Abstract

The method is used for detecting one failure in a burner of a combustor of a turbine system; the combustor comprises a plurality of burners arranged annularly; the turbine system comprises a turbine downstream of the combustor, the method comprising the steps of: A) providing a plurality of temperature sensors arranged annularly at the outlet of the turbine, B) detecting a plurality of temperatures through the plurality of temperature sensors, C) calculating a temperature spread indicator as a function of the plurality of temperatures, and D) carrying out a comparison the temperature spread indicator and a threshold; a positive result of this comparison indicates a burner failure.

Claims

1. A method of detecting one failure in a burner of a combustor of a turbine system, wherein the combustor comprises a plurality of burners arranged annularly, wherein the turbine system comprises a turbine downstream of said combustor, the method comprising the steps of: A) providing a plurality of temperature sensors arranged annularly at the outlet of said turbine, B) detecting a plurality of temperatures through said plurality of temperature sensors, C) calculating a temperature spread indicator as a function of said plurality of temperatures, and D) carrying out a comparison between said temperature spread indicator and a threshold, wherein a positive result of said comparison indicates a burner failure.

2. The method of claim 1, wherein steps A, B, C and D are cyclically repeated in time, and wherein a set of consecutive positive results of said comparison indicates a burner failure.

3. The method of claim 1, wherein step C comprises the sub-steps of: C1) determining a minimum temperature in said plurality of temperatures, C2) determining a maximum temperature in said plurality of temperatures, and C3) calculating a temperature spread indicator as a difference between said maximum temperature and said minimum temperature.

4. The method of claim 3, comprising the step of signaling an alarm at one positive result or a set of consecutive positive results of said comparison.

5. The method of claim 3, comprising the step of tripping said turbine system at one positive result or a set of consecutive positive results of said comparison.

6. The method of claim 3, wherein steps A, B, C and D are cyclically repeated in time; wherein a set of N1 consecutive positive results of a comparison between said temperature spread indicator and a first threshold indicates a first-type burner failure; wherein a set of N2 consecutive positive results of a comparison between said temperature spread indicator and a second threshold indicates a second-type burner failure; comprising the step of signaling an alarm in case of said first-type burner failure; comprising the step of tripping said turbine system in case of said second-type burner failure.

7. The method of claim 1, wherein step C comprises the sub-steps of: C4) determining a mean temperature being the average of said plurality of temperatures, and C5) calculating a temperature spread indicator as a plurality of values (Phi(i)) being functions of a corresponding plurality of differences between a temperature of said plurality of temperatures and said mean temperature; wherein at step D each of the values of said plurality of values is compared with said threshold, wherein any comparison positive result indicates a burner failure.

8. The method of claim 7, wherein any value of said plurality of values is calculated according to the formula:
Phi(i)=e.sup.(k*(T.sup.i.sup.T.sup.ave.sup.).sup.2.sup.) wherein i is an index of a temperature sensor of said plurality of temperature sensors.

9. The method of claim 7, comprising the step of signaling an alarm at any comparison positive result or a set of consecutive comparison positive results.

10. The method of claim 7, comprising the step of tripping said turbine system at any comparison positive result or a set of consecutive comparison positive results.

11. The method of claim 7, wherein steps A, B, C and D are cyclically repeated in time; wherein a set of M1 consecutive positive results of a comparison between said temperature spread indicator and a first threshold indicates a first-type burner failure; wherein a set of M2 consecutive positive results of a comparison between said temperature spread indicator and a second threshold indicates a second-type burner failure; comprising the step of signaling an alarm in case of said first-type burner failure; comprising the step of tripping said turbine system in case of said second-type burner failure.

12. A method of locating one failure in a burner of a combustor of a turbine system, wherein the turbine system comprises a turbine downstream of said combustor, wherein the combustor comprises a plurality of burners arranged annularly and a plurality of temperature sensors arranged annularly at the outlet of said turbine, comprising the detection method according to claim 1, wherein if there is one burner that has failed the plurality of temperatures are processed based on predetermined data and a burner position or a burner position range of the faulty burner is obtained.

13. The method of claim 12, further comprising the steps of: E) identifying a temperature sensor detecting a minimum temperature, F) determining a sensor angular position of said temperature sensor, and G) subtracting a predetermined angular shift or a predetermined angular shift range from said sensor angular position and a burner position or a burner position range of the faulty burner is obtained.

14. The method of claim 13, wherein said predetermined angular shift or predetermined angular shift range depends on the operating status of the turbine system.

15. A turbine system comprising a compressor, a burner downstream of said compressor, a turbine downstream of said combustor, and a plurality of temperature sensors arranged annularly at the outlet of said turbine, wherein said combustor comprises a plurality of burners arranged annularly; and further comprising a digital signal processing unit adapted to carry out the method according to claim 1.

Description

BRIEF DESCRIPTION OF DRAWINGS

[0014] The accompanying drawings, which are incorporated herein and constitute an integral part of the present specification, illustrate exemplary embodiments of the present invention and, together with the detailed description, explain these embodiments. In the drawings:

[0015] FIG. 1 shows a schematic diagram of a turbine system using a combustor failure detection method according to the subject matter disclosed herein,

[0016] FIG. 2 shows a schematic section view of a combustor of the turbine system of FIG. 1,

[0017] FIG. 3 shows a schematic front view of a turbine of the turbine system of FIG. 1,

[0018] FIG. 4 shows a flowchart of a first embodiment of the method according to the subject matter disclosed herein,

[0019] FIG. 5 shows a flowchart of a second embodiment of the method according to the subject matter disclosed herein, and

[0020] FIG. 6 shows a flowchart of a third embodiment of the method according to the subject matter disclosed herein.

DETAILED DESCRIPTION

[0021] The following description of exemplary embodiments refers to the accompanying drawings.

[0022] The following description does not limit the invention. Instead, the scope of the invention is defined by the appended claims.

[0023] Reference throughout the specification to one embodiment or an embodiment means that a particular feature, structure, or characteristic described in connection with an embodiment is included in at least one embodiment of the subject matter disclosed. Thus, the appearance of the phrases in one embodiment or in an embodiment in various places throughout the specification is not necessarily referring to the same embodiment. Further, the particular features, structures or characteristics may be combined in any suitable manner in one or more embodiments.

[0024] FIG. 1 shows a turbine system 1 comprising: a compressor 2, a combustor 3, a first turbine 4 (that may be called high-pressure turbine), a second turbine 5 (that may be called low-pressure turbine), and a digital signal processing unit 6. Compressor 2 has an own shaft 21; the first turbine 4 has an own shaft 41; the second turbine 5 has an own shaft 51 (mechanically disconnected from shaft 41). An outlet of compressor 2 is fluidly connected to an inlet of combustor 3; an outlet of combustor 3 is fluidly connected to an inlet of turbine 4; an outlet of turbine 4 is fluidly connected to an inlet of turbine 5. It is to be noted that combustor 3 has at least another inlet (not shown in FIG. 1), for example a fuel inlet.

[0025] Combustor 3 comprises a plurality of burners 32 arranged annularly in a single annular combustion chamber (see also FIG. 2). In FIG. 2, there are twelve burners labelled from 32-1 to 32-12; a lower number of burners is possible for example down to five; a higher number of burners is possible for example up to fifty.

[0026] Turbine system 1 comprises a plurality of temperature sensors 52 arranged annularly at the outlet of turbine 5 (see also FIG. 3). In FIG. 3, there are six sensors labelled from 52-1 to 52-6; a lower number of sensors is possible for example down to three; a higher number of sensors is possible for example up to twenty. It is to be noted that the number of sensors 52 is lower than the number of burners 32.

[0027] According to embodiments alternative to the one of FIG. 1, the turbine system may comprise one or two or three or more compressors (for example serially connected) and/or one or two or three or more turbines (for example serially connected).

[0028] According to embodiments alternative to the one of FIG. 1, the sensors may be arranged elsewhere typically annularly for example at the outlet of turbine 4, but not inside the combustion chamber of combustor 3 or at the outlet of combustor 3.

[0029] Digital signal processing unit 6 is electrically connected to sensors 52 and receives measure signals from these sensors.

[0030] In order to detect one failure in a burner of combustor (for example combustor 3) of a turbine system, the following steps may be carried out: [0031] A) providing a plurality of temperature sensors (for example sensors 52) arranged annularly at the outlet of a turbine (for example turbine 5) of the turbine system, [0032] B) detecting a plurality of temperatures through the plurality of temperature sensors, [0033] C) calculating a temperature spread indicator as a function of the plurality of temperatures, and [0034] D) carrying out a comparison between the temperature spread indicator and a threshold,
wherein a positive result of this comparison indicates a burner failure.

[0035] During normal operation, the temperature at the outlet of a turbine of the turbine system rather uniform; for example, the temperatures detected by sensors 52-1, 52-2, 52-3, 52-4, 52-5 and 52-6 at different angular positions at a certain time are close to each other (even if never perfectly identical). Considering FIG. 1, the temperatures at the outlet of turbine 5 are more uniform than the temperatures at the outlet of turbine 4.

[0036] If, at a certain time, one burner fails, i.e. it does not contribute to combustion in the combustion chamber, the temperature spread in the temperature detected by the sensors at different angular positions (for example sensors 52) increases considerably.

[0037] For example, a nominal temperature spread during normal operation may be e.g. 30 C. or 35 C. (the average temperature being for example 400-600 C.), while the temperature spread in case of one failure is much higher and may be e.g. 55 C. or 80 C. (the average temperature being substantially unchanged). The threshold at step D may be set for example at 45 C. or 50 C. As it will be explained in the following and as it is shown FIGS. 4, 5 and 6, it may be advantageous to use more than one threshold.

[0038] In the embodiment of FIG. 1, step B is carried out by sensors 52, and steps C and D are carried out by digital signal processing unit 6.

[0039] In the flowchart of FIG. 4, there is a block 401 corresponding to a START and a block 402 corresponding to step B (temperatures detection) and step C (indicator calculation). In particular, the temperature spread indicator may correspond simply to the temperature spread; for example, considering FIGS. 1, 2 and 3, at a certain time six temperatures are detected by the six sensors 52, then the minimum temperature is determined between the six detected temperatures (step C1), then maximum temperature is determined between the six detected temperatures (step C2), and the temperature spread indicator is calculated as the difference between the maximum temperature and the minimum temperature (step C3).

[0040] The threshold at step D is typically fixed and predetermined (i.e. not variable). If the threshold is variable, it may depend on a current load state of the turbine system 1; in FIG. 4, block 403 corresponds to a pre-determination of one or more thresholds.

[0041] In the flowchart of FIG. 4, two different thresholds (i.e. threshold-1 and threshold-2) are used (see block 404 and block 406); in particular, the first threshold is lower than the second threshold (i.e. threshold-1<threshold-2); the first threshold may be e.g. 55 C. and the second threshold may be e.g. 80 C.

[0042] In the flowchart of FIG. 4, block 404 corresponds to a comparison of the calculated temperature spread indicator with a first threshold (i.e. is indicator>threshold-1?) and block 406 corresponds to a comparison of the calculated temperature spread indicator with a second threshold (i.e. is indicator>threshold-2?). A negative result N at block 404 indicates that there is no burner failure in the combustor 3 and, subsequently to step D, OK is signaled at block 405; for example, digital signal processing unit 6 may send an ok signal to a remote monitoring unit or send no signal at all. A positive result Y at block 404 indicates that there is one burner failure in the combustor 3. A negative result N at block 406 indicates there is one burner failure in the combustor 3 but the failure is not severe or, more in general, there is a first-type burner failure and, subsequently to step D, ALARM is signaled at block 407; for example, digital signal processing unit 6 may send an alarm signal to a remote monitoring unit. A positive result Y at block 406 indicates that there is one burner failure in the combustor 3 but the failure is severe or, more in general, there is a second-type burner failure and, subsequently to step D, TRIP of turbine system 1 (i.e. switching-off) is carried at block 408.

[0043] After blocks 405 and 407, the flow returns to block 402. This means that the temperature spread indicator is calculated repeatedly, in particular periodically; more precisely, at least steps A, B, C and D are cyclically repeated in time. The average repetition period may be in the range from e.g. 1 mS to e.g. 10 S.

[0044] According to the embodiment just described, as soon as a threshold is reached an action is taken.

[0045] Alternatively, it may be provided that an action is taken only after a predetermined time or after a predetermined number of consecutive positive results as for example in the flowchart of FIG. 5.

[0046] In the flowchart of FIG. 5, there is a block 501 corresponding to a START wherein: [0047] a first counter CNTR1 is set to zero, [0048] a second counter CNTR2 is set to zero, [0049] a first counter threshold CTH1 is set to N1 (for example 30), [0050] a second counter threshold CTH2 is set to N2 (for example 10), [0051] a first spread threshold STH1 is set to T1 (for example 55 C.), [0052] a second spread threshold STH2 is set to T2 (for example 80 C.);
in the following, it is assumed, as an example, that a temperature spread indicator TSI corresponds simply to the temperature spread and that the repetition period is approximately 1 S.

[0053] At block 502, counter CNTR1 is increased by 1, counter CNTR2 is increased by 1, a plurality of temperatures are detected by the plurality of temperature sensors, the temperature spread indicator TSI is calculated.

[0054] At block 503 the comparison TSI>STH1? is carried out; N corresponds to a negative comparison and Y corresponds to a negative comparison.

[0055] At block 505 the comparison TSI>STH2? is carried out; N corresponds to a negative comparison and Y corresponds to a negative comparison.

[0056] At block 504 OK is signaled and counters CNTR1 and CNTR2 are reset, i.e. set to zero, and then the flow returns to block 502.

[0057] At block 506 the comparison CNTR1>CTH1? is carried out; N corresponds to a negative comparison and Y corresponds to a negative comparison; in case of negative comparison the flow returns to block 502; in case of positive comparison the flow passes to block 507.

[0058] At block 508 the comparison CNTR2>CTH2? is carried out; N corresponds to a negative comparison and Y corresponds to a negative comparison; in case of negative comparison the flow returns to block 502; in case of positive comparison the flow passes to block 509.

[0059] At block 507 ALARM is signaled and then the flow returns to block 502.

[0060] At block 509 TRIP of turbine system (i.e. switching-off) is carried out.

[0061] This means that an alarm is signaled only if a temperature spread greater than 55 C. persists for more than 30 seconds and trip is carried out only if a temperature spread greater than 80 C. persists for more than 10 seconds.

[0062] It is to be noted that, according to a flowchart alternative to the one of FIG. 4, block 406 is not present and block 407 is located immediately before or after block 408, i.e. there is only one threshold, the turbine system is either considered OK or faulty, and when faulty an alarm is signaled and the turbine system is switched-off.

[0063] The calculation of the temperature spread indicator (step C) may be more complex than the one just described.

[0064] The comparison of the temperature spread indicator (step D) may be more complex than the one just described.

[0065] In the embodiments just described, the temperature spread indicator is a number. In the embodiment that will be described in the following with the aid of FIG. 6, the temperature spread indicator is s set of numbers.

[0066] It is to be noted that the flowchart of FIG. 6 is graphically very similar to the flowchart of FIG. 5; the black big dot at blocks 603, 605, 606 and 608 means that the corresponding test is repeated many times, in particular it is repeated for each of the sensors.

[0067] According to this embodiment, step C comprises the sub-steps of: [0068] determining a mean temperature being the average of the plurality of temperatures detected at step B (step C4),
and [0069] calculating a temperature spread indicator as a plurality of values being functions of a corresponding plurality of differences between a temperature of the plurality of temperatures and the mean temperature (step C5);
in the flowchart of FIG. 6, such step C is carried out at block 602.

[0070] An advantageous formula for calculating the temperature spread indicator, namely its values, is the following:

Phi(i)=e.sup.(k*(T.sup.i.sup.T.sup.ave.sup.).sup.2.sup.)

wherein i is an index of a temperature sensor of the plurality of temperature sensors.

[0071] If the above formula is applied to the turbine system of FIG. 1 and in the flowchart of FIG. 6, for any repetition of the loop six values from Phi(1) to Phi(6) are calculated for the six sensors from 52-1 to 52-6.

[0072] The above formula is advantageous as it increases the signal-to-noise ratio; in other words, the exponential function highlights the temperature spread variation due to one burner failure in a more clear distinct way.

[0073] At step D each of the calculated values is compared with at least one threshold; the threshold is the same for all sensors. If the above formula is used, the comparison is value<threshold?, a negative comparison N means that the value lesser than the threshold and a positive comparison Y means that the value greater than the threshold.

[0074] In the flowchart of FIG. 6, two different indicator thresholds are used (see block 603 and block 605); in particular, the first threshold (block 603) is higher than the second threshold (block 605).

[0075] If a single indicator threshold is used, its value may be e.g. 0.2.

[0076] If two indicator thresholds are used, the value of the first threshold may be e.g. 0.25 and the value of the second threshold may be e.g. 0.15.

[0077] Any comparison positive result indicates a burner failure.

[0078] Considering now the flowchart of FIG. 6, the flow passes from block 603 to block 605 after that all comparisons are carried out and if at least one comparison is positive, the flow passes from block 606 to block 607 (alarm signaling) after that all comparisons are carried out and if at least one comparison is positive, the flow passes from block 608 to block 609 (trip performing) after that all comparisons are carried out and if at least one comparison is positive.

[0079] This means that alarm signaling and trip performing are carried out if measures from any (not all) of the sensors justify such actions.

[0080] As in the flowchart of FIG. 5, also in the flowchart of FIG. 6 actions are taken only in case of persistence and not immediately; this is the reason of block 605. The persistence thresholds may be identical to the flowchart of FIG. 5, i.e. 30 seconds and 10 seconds.

[0081] In order to take into account that the indicator corresponds to a plurality of values and that an action requires a plurality of repetitions, a set of counters are used for each of the thresholds used.

[0082] For example, if there are six temperature sensors (see FIG. 3) and two spread thresholds (see FIG. 6), twelve counters are used. In fact, for each sensor a first counter indicates how far alarm signaling is and a second counter indicates how far trip performing is.

[0083] As for the flowchart of FIG. 5, also for the flowchart of FIG. 6 variations are possible.

[0084] Once one burner failure is detected, it is advantageous to locate the burner of the combustor that has failed or, even better, identify exactly which burner of the combustor has failed.

[0085] In order to do that, if there is one burner that has failed the plurality of detected temperatures may be processed based on predetermined data. The result of such processing may be a burner position or a burner position range of the faulty burner.

[0086] The above mentioned predetermined data may be data obtained during a calibration phase of the turbine system carried out at the manufacturing site, or at the installation site, or partially at the manufacturing site and partially at the installation site

[0087] Such processing may comprise the steps of: [0088] E) identifying a temperature sensor detecting a minimum temperature, [0089] F) determining a sensor angular position of this temperature sensor, and [0090] G) subtracting a predetermined angular shift or a predetermined angular shift range from the determined sensor angular position and so a burner position or a burner position range of the faulty burner is obtained.

[0091] Considering FIG. 2 and FIG. 3, if, for example, the minimum temperature is detected by sensor 52-2, according to a first embodiment of the above processing, there might be an angular shift of 150 (anti-clockwise), the faulty burner is burner 32-10; if the minimum temperature is detected by sensor 52-6, the faulty burner is burner 32-6.

[0092] Considering FIG. 2 and FIG. 3, if, for example, the minimum temperature is detected by sensor 52-2, according to a second embodiment of the above processing, there might be an angular shift range of 120-180 (anti-clockwise), the faulty burner is any of burners 32-9, 32-10, 32-11; if the minimum temperature is detected by sensor 52-6, the faulty burner is any of burners 32-5, 32-6, 32-7. Even in this case, wherein the position of the faulty burner is not precisely identified, maintenance is highly facilitated.

[0093] It is to be considered that the above-mentioned predetermined angular shift or predetermined angular shift range may depend on the operating status of the turbine system (for example, full-load, half-load).

[0094] The burner failure detection methods according to the subject matter disclosed herein may be used in a turbine system like the one of FIG. 1 or in a system similar thereto.

[0095] The turbine system should comprise a digital signal processing unit adapted to carry out such methods; such unit typically comprises a software program for this purpose. In the embodiment of FIG. 1, such unit is labelled 6.

[0096] In order to carry out a burner failure detection method, at least one plurality of temperature sensors (with reference to FIG. 1, sensors 52) arranged annularly at the outlet of a turbine of the turbine system are necessary.

[0097] These sensors may allow not only to detect one burner failure but also to locate the faulty burner; location may correspond even to identification of the faulty burner.