Systems and methods for determining risk of operating a turbomachine

1. A method for operating a turbomachine, the method comprising:

receiving, by a processor, from a repository, historical failure data associated with a fleet of two or more turbomachines, wherein the historical failure data comprises turbomachine failure data and ambient conditions at the time of the turbomachine failure data;

receiving, by a processor, ambient conditions of an environment in which a first turbomachine is to be operated;

developing, by a processor, by clustering risk events, at least one risk threshold for at least one known operating profile based at least in part on the historical failure data in view of the ambient conditions of the environment in which the first turbomachine is to be operated;

determining, by comparing some or all of the ambient conditions to the at least one risk threshold and applying pattern recognition to historical failure data, that the at least one risk threshold for the at least one known operating profile is reached; and

based at least in part on a determination the at least one risk threshold is reached, facilitating a control action to change an operation of the first turbomachine, wherein the control action includes at least one of the following: applying bleed heat to one or more inlet guide vanes (IGV), increasing a ramp-up speed, decreasing the ramp-up speed, or controlling the bleed heat during a ramp-up cycle.

2. The method of claim 1 , further comprising mitigating an anomaly event associated with the risk.

3. The method of claim 1 , wherein the developing the at least one risk threshold includes clustering risk events using statistical techniques.

4. The method of claim 1 , wherein the determining that the at least one risk threshold is reached includes matching signatures of the at least one known operating profile to signatures of the historical failure data.

5. The method of claim 4 , wherein the determining that the at least one risk threshold is reached includes using a pattern recognition algorithm parsed through a subset of the historical failure data.

6. The method of claim 5 , wherein the pattern recognition algorithm includes at least one of the following: a dynamic time warping, a similarity based modeling, and a Gaussian mix.

7. The method of claim 1 , wherein the determining that the at least one risk threshold is reached is indicative of at least one of the following events: icing on one or more inlet guide vanes (IGV) of the turbomachine, degradation caused by atmospheric dust, damage caused by corrosive contaminants, a lean blow out, high combustion dynamics, or excessive emissions.

8. The method of claim 1 , wherein the at least one known operating profile of the turbomachine includes at least one of the following: transient operations, a load of the turbomachine, an IGV angle, an inlet temperature, an inlet humidity, or an inlet dew point.

9. The method of claim 1 , wherein the ambient conditions includes at least one of the following: atmospheric data, air composition, air particulates, a temperature, a humidity, or a pressure.

10. The method of claim 1 , wherein the historical failure data is collected at a power plant level and stored in the repository.

11. A system for operating a turbomachine, the system comprising:

a repository configured to store historical failure data associated with a fleet of turbomachines, wherein the historical failure data comprises turbomachine failure data and ambient conditions at the time of the turbomachine failure data;

a processor in communication with the repository and configured to:

receive the historical failure data;

receive ambient conditions of an environment in which a first turbomachine is to be operated;

develop, by clustering risk events, at least one risk threshold for at least one known operating profile based at least in part on the historical failure data in view of the ambient conditions of the environment in which the first turbomachine is to be operated;

determine, by comparing some or all of the ambient conditions to the at least one risk threshold and applying pattern recognition to historical failure data, that the at least one risk threshold for the at least one known operating profile is reached; and

based on a determination the at least one risk threshold is reached, causing an action to change operation of the first turbomachine, wherein the action includes at least one of the following: applying bleed heat to one or more inlet guide vanes (IGV), increasing a ramp-up speed, decreasing the ramp-up speed, or controlling the bleed heat during a ramp-up cycle.

12. The system of claim 11 , further comprising an on-site monitor configured to monitor operational data of the first turbomachine associated with the at least one known operating profile and to send the operational data to the processor.

13. The system of claim 11 , wherein the determining that the at least one risk threshold is reached includes matching signatures of the at least one known operating profile to signatures of the historical failure data using a pattern recognition algorithm parsed through a subset of the historical failure data, the pattern recognition algorithm including at least one of the following: a dynamic time warping, a similarity based modeling, and a Gaussian mix.

14. The system of claim 11 , wherein the at least one risk threshold is associated with at least one of the following: icing on one or more IGV of the first turbomachine, degradation caused by atmospheric dust, a damage caused by corrosive contaminants, a lean blow out, high combustion dynamics, or excessive emissions.

15. A method for operating a turbomachine, the method comprising:

receiving, by a processor, from a repository, historical failure data associated with a fleet of turbomachines, wherein the historical failure data includes turbomachine failure data and ambient conditions at the time of the turbomachine failure data;

receiving, by the processor, ambient conditions of an environment in which a first turbomachine is to be operated, wherein the ambient conditions include at least one of the following: atmospheric data, a temperature, a humidity, or a pressure;

developing, by the processor, by clustering risk events using statistical techniques, at least one risk threshold for the at least one known operating profile based at least in part on the historical failure profile data in view of the ambient conditions of the environment in which the first turbomachine is to be operated;

determining, by the processor, by comparing some or all of the ambient conditions to the at least one risk threshold and applying pattern recognition to historical failure data, that the at least one risk threshold for the at least one known operating profile is reached;

based on a determination the at least one risk threshold is reached, causing an action to control the first turbomachine, wherein the action includes at least one of the following: applying bleed heat to one or more inlet guide vanes (IGV), increasing a ramp-up speed, decreasing the ramp-up speed, or controlling the bleed heat during a ramp-up cycle.