초록 ▼

A method for monitoring performance of a gas turbine system comprises providing a plurality of combustor cans; placing a plurality of temperature sensors circumferentially around an exhaust plane of the plurality of combustor cans; operating the plurality of combustor cans while varying a plurality

A method for monitoring performance of a gas turbine system comprises providing a plurality of combustor cans; placing a plurality of temperature sensors circumferentially around an exhaust plane of the plurality of combustor cans; operating the plurality of combustor cans while varying a plurality of gas turbine operating parameters, where exhaust gas issues from each combustor can of the plurality of combustor cans during operation; measuring temperature of the exhaust gas in the exhaust plane using the plurality of temperature sensors to obtain a plurality of individual temperature measurements; determining a correlation of the individual temperature measurements of exhaust gas temperature with corresponding individual combustor cans of the plurality of combustor cans issuing the exhaust gas; and developing a swirl model, where the model uses the correlation to predict swirl values in the exhaust plane as a function of the operating parameters.

대표청구항 ▼

What is claimed is: 1. A method for monitoring performance of a gas turbine system, said method comprising: providing a plurality of combustor cans; placing a plurality of temperature sensors circumferentially around an exhaust plane of said plurality of combustor cans; operating said plurality of

What is claimed is: 1. A method for monitoring performance of a gas turbine system, said method comprising: providing a plurality of combustor cans; placing a plurality of temperature sensors circumferentially around an exhaust plane of said plurality of combustor cans; operating said plurality of combustor cans while varying a plurality of gas turbine operating parameters, wherein exhaust gas issues from each combustor can of said plurality of combustor cans during operation; measuring temperature of said exhaust gas in said exhaust plane using said plurality of temperature sensors to obtain a plurality of individual temperature measurements; determining a correlation of said individual temperature measurements of exhaust gas temperature with corresponding individual combustor cans of said plurality of combustor cans issuing said exhaust gas; developing a swirl model, wherein said model uses said correlation to predict swirl values in said exhaust plane as a function of said gas turbine operating parameters; setting said plurality of combustor cans to operate at a selected set of predetermined operating conditions; measuring real-time exhaust gas temperature data during operation of said plurality of combustor cans at said selected set of operating conditions using said plurality of temperature sensors; calculating at least one actual swirl value based upon said real-time exhaust gas temperature data; calculating at least one predicted swirl value for said selected set of operating conditions using said model; comparing said actual swirl value and said predicted swirl value and calculating at least one failure test statistic; comparing said failure statistic to a standard; and generating at least one failure flag signal when said failure test statistic exceeds said standard, wherein said at least one failure flag signal indicates a presence of a failed combustor can in said plurality of combustor cans. 2. The method of claim 1, wherein said failure test statistic is a difference between said actual swirl value and predicted swirl value. 3. The method of claim 1, wherein comparing said failure test statistic to a standard comprises comparing failure test statistic values for a respective can over a period of time and wherein a selected characteristic of the distribution of said failure test statistic values over said period of time is compared to said standard corresponding to said selected characteristic. 4. The method of claim 1 further comprising generating at least one failure classification signal, wherein said failure classification signal identifies a particular failed combustor can from among said plurality of combustor cans. 5. The method of claim 1, wherein operating said plurality of combustor cans while varying a plurality of gas turbine operating parameters further comprises varying each of said gas turbine operating parameters within a respective value range, said respective value range ranging from a respective value used at a start-up condition for said plurality of combustor cans to a respective value used at a base load condition for said plurality of combustor cans. 6. The method of claim 5, wherein said gas turbine operating parameters comprise at least one of inlet guide vane angle, compressor discharge pressure, fuel flow rate, compressor discharge temperature, corrected mass flow rate, output power. 7. The method of claim 1, wherein developing said swirl model comprises developing a linear model. 8. The method of claim 1, wherein developing said swirl model comprises developing a non-linear model. 9. The method of claim 1, wherein measuring the temperature of said exhaust gas in said exhaust plane comprises creating an exhaust gas temperature profile within said exhaust plane, said profile comprising a plurality of local maximum points, wherein each local maximum point corresponds to one of said combustor cans of said plurality of combustor cans, and wherein determining said correlation comprises correlating each local maximum point of said plurality of local maximum points with its corresponding combustor can. 10. The method of claim 1, wherein said plurality of temperature sensors are components of a temperature-monitoring portion of an online control system of the turbine. 11. A method for monitoring performance of a gas turbine system, said method comprising: providing a plurality of combustor cans; placing a plurality of temperature sensors circumferentially around an exhaust plane of said plurality of combustor cans; operating said plurality of combustor cans while varying a plurality of gas turbine operating parameters, wherein exhaust gas issues from each combustor can of said plurality of combustor cans during operation, wherein varying a plurality of gas turbine operating parameters comprises varying each of said gas turbine operating parameters within a respective value range, said respective value range ranging from a respective value used at a start-up condition for said plurality of combustor cans to a respective value used at a base load condition for said plurality of combustor cans, and wherein said gas turbine operating parameters comprise at least one of inlet guide vane angle, compressor discharge pressure, fuel flow rate, compressor discharge temperature, corrected mass flow rate, output power; measuring the temperature of said exhaust gas in said exhaust plane using said plurality of temperature sensors, wherein measuring the temperature of said exhaust gas in said exhaust plane comprises creating an exhaust gas temperature profile within said exhaust plane, said profile comprising a plurality of local maximum points, wherein each local maximum point corresponds to one of said combustor cans of said plurality of combustor cans; determining a correlation of individual temperature measurements of exhaust gas temperature with corresponding individual combustor cans issuing said exhaust gas, wherein determining said correlation comprises correlating each local maximum point of said plurality of local maximum points with its corresponding combustor can; developing a swirl model, wherein said model uses said correlation to determine a swirl value in said exhaust plane as a function of said gas turbine operating parameters; setting said plurality of combustor cans to operate at a selected set of predetermined operating conditions; obtaining real-time exhaust gas temperature data during operation of said plurality of combustor cans at said selected set of operating conditions; calculating an actual swirl value based upon said real-time exhaust gas temperature data; calculating a predicted swirl value for said selected set of operating conditions using said model; comparing said actual swirl value and said predicted swirl value and generating at least one failure test statistic indicating a presence of faulty combustor can; wherein said at least one failure test statistic is a difference between said actual swirl value and said predicted swirl value; comparing said difference to a standard; generating at least one failure flag signal when said difference exceeds said standard, wherein said at least one failure flag signal indicates a presence of a failed combustor can in said plurality of combustor cans; and; generating at least one failure classification signal, wherein said failure classification signal identifies a particular failed combustor can from among said plurality of combustor cans. 12. A method for identifying potentially faulty combustor cans in a gas turbine system, said method comprising: setting a plurality of combustor cans in said gas turbine system to operate at a selected set of predetermined operating conditions; placing a plurality of temperature sensors circumferentially around an exhaust plane of said plurality of combustor cans; measuring real-time exhaust gas temperature data during operation of said plurality of combustor cans at said selected set of operating conditions using said plurality of temperature sensors; developing a swirl model, wherein said swirl model predicts swirl values of exhaust gas issuing from said combustor cans as a function of said gas turbine operating parameters; calculating an actual swirl value based upon said real-time exhaust gas temperature data; calculating a predicted swirl value for said selected set of operating conditions using said model; comparing said actual swirl value and said predicted swirl value and calculating at least one failure test statistic; comparing said failure test statistic to a standard; and generating at least one failure flag signal when said failure test statistic exceeds said standard, wherein said at least one failure flag signal indicates a presence of a failed combustor can in said plurality of combustor cans. 13. The method of claim 12, further comprising: generating at least one failure classification signal, wherein said failure classification signal identifies a particular failed combustor can from among said plurality of combustor cans. 14. The method of claim 12, wherein the failure test statistic is a difference between said actual swirl value and predicted swirl value.