Gas turbine predictive emissions modeling, reporting, and model management via a remote framework

Patent metadata

Field	Value
Publication number	US-12455075-B2
Application number	US-202217703726-A
Country	US
Kind code	B2
Filing date	Mar 24, 2022
Priority date	Mar 24, 2022
Publication date	Oct 28, 2025
Grant date	Oct 28, 2025

Abstract

It is important to accurately measure the emissions of a turbomachine for a variety of reasons. However, continuous emission monitoring systems (CEMS) can be expensive to install and maintain. Accordingly, a digital platform is disclosed that hosts physics-based and/or statistical models that can be tailored to specific turbomachines and calibrated over the life of the turbomachine. The model for a turbomachine can be applied to data collected from the turbomachine to predict the emissions of the turbomachine. This enables monitoring of emissions, remotely and without the need of a CEMS. In addition, the platform may utilize the predicted emissions for alerts, compliance monitoring, health monitoring, control of the turbomachine, reporting, and/or the like.

First claim

What is claimed is:

1. A method of monitoring emissions of a turbomachine with at least one hardware processor, the method comprising:
with the at least one hardware processor:
storing a model configured to predict emissions of a remote turbomachine based on one or more parameters of the turbomachine;
over a plurality of iterations,
receiving data from the remote turbomachine over at least one network, wherein the data include values of the one or more parameters,
applying the model to the values of the one or more parameters to output predicted emissions of the turbomachine,
determining an offset based on a difference between the output predicted emissions determined by the model, and a measured emission of the turbomachine, and adding the predicted emissions adjusted by the offset to an emissions history for the remote turbomachine which is stored in a memory; and

monitoring the emissions history for compliance with at least one emissions requirement;
wherein the model comprises a physics-based model that determines one or more airflow characteristics of the remote turbomachine based on at least one the values of the one or more parameters;
wherein the physics-based model further determines a primary zone temperature of a combustor of the remote turbomachine based on the determined one or more airflow characteristics; and
wherein the one or more parameters comprise a differential pressure of the combustor, and wherein the physics-based model determines the primary zone temperature further based on the differential pressure.

2. The method of claim 1 , wherein the one or more airflow characteristics comprise an airflow rate through a combustor of the remote turbomachine.

3. The method of claim 1 , wherein the one or more airflow characteristics comprise a percentage of total airflow rate that is input to a combustor of the remote turbomachine.

4. The method of claim 1 , wherein the one or more parameters comprise one or more properties of fuel used in the combustor, and wherein the physics-based model determines the primary zone temperature further based on the one or more properties of fuel.

5. The method of claim 1 , wherein the one or more parameters comprise a discharge temperature of a compressor of the remote turbomachine, and wherein the physics-based model determines the primary zone temperature further based on the discharge temperature.

6. The method of claim 1 , wherein the one or more parameters comprise a percentage of pilot fuel in fuel used by the remote turbomachine, and wherein the physics-based model further determines the predicted emissions based on a ratio between the primary zone temperature and the percentage of pilot fuel.

7. The method of claim 1 , wherein the one or more parameters comprise an inlet temperature of a turbine of the remote turbomachine, and wherein the model includes a physics-based model that determines the predicted emissions based on the inlet temperature.

8. The method of claim 1 , wherein the model comprises a statistical model that includes a machine-learning model that has been trained to output the predicted emissions based on the values of the one or more parameters.

9. The method of claim 8 , wherein the machine-learning model comprises a regression.

10. The method of claim 8 , wherein the one or more parameters comprise one or more of an engine load of the remote turbomachine, a gas generator speed of the remote turbomachine, a percentage of pilot fuel in fuel used by the remote turbomachine, an ambient temperature in the remote turbomachine, or an inlet temperature of a turbine of the remote turbomachine.

11. The method of claim 1 , wherein the one or more parameters comprise a percentage of pilot fuel in fuel used by the remote turbomachine, and wherein applying the model to the values of the one or more parameters comprises adjusting the percentage of pilot fuel based on an offset or translation function.

12. The method of claim 1 , wherein the predicted emissions comprise a volume of nitrogen oxides (NOX).

13. The method of claim 1 , further comprising using the at least one hardware processor to, in response to the predicted emissions satisfying one or more criteria, send an alert to at least one recipient.

14. The method of claim 1 , further comprising using the at least one hardware processor to, in response to the predicted emissions satisfying one or more criteria, send a control command to the remote turbomachine, wherein the control command initiates a transition of the remote turbomachine from a first state to a second state that is different than the first state.

15. The method of claim 1 , wherein the model is deployed as a microservice, and wherein applying the model comprises calling the model via a microservice application programming interface.

16. A system comprising:
at least one hardware processor; and
software configured to, when executed by the at least one hardware processor,
store a model configured to predict emissions of a remote turbomachine based on one or more parameters of the remote turbomachine,
over a plurality of iterations,
receive data from the remote turbomachine over at least one network, wherein the data include values of the one or more parameters,
apply the model to the values of the one or more parameters to output predicted emissions of the remote turbomachine, and
add the predicted emissions to an emissions history for the remote turbomachine in a memory, and

monitor the emissions history for compliance with at least one emissions requirement,
wherein the one or more parameters comprise a percentage of pilot fuel in fuel used by the remote turbomachine,
wherein the model determines a primary zone temperature of a combustor of the remote turbomachine based on at least one the values of the one or more parameters,
wherein the model further determines the predicted emissions based on a ratio between the primary zone temperature and the percentage of pilot fuel.

17. A method of monitoring emissions of a turbomachine with at least one hardware processor, the method comprising:
with the at least one hardware processor:
storing a model configured to predict emissions of a remote turbomachine based on one or more parameters of the turbomachine;
over a plurality of iterations,
receiving data from the remote turbomachine over at least one network, wherein the data include values of the one or more parameters,
applying the model to the values of the one or more parameters to output predicted emissions of the turbomachine,
determining an offset based on a difference between the output predicted emissions determined by the model, and a measured emission of the turbomachine, and

adding the predicted emissions adjusted by the offset to an emissions history for the remote turbomachine which is stored in a memory; and
monitoring the emissions history for compliance with at least one emissions requirement;
wherein the model comprises a physics-based model that determines one or more airflow characteristics of the remote turbomachine based on at least one the values of the one or more parameters;
wherein the physics-based model further determines a primary zone temperature of a combustor of the remote turbomachine based on the determined one or more airflow characteristics; and
wherein the one or more parameters comprise a percentage of pilot fuel in fuel used by the remote turbomachine, and wherein the physics-based model further determines the predicted emissions based on a ratio between the primary zone temperature and the percentage of pilot fuel.

Assignees

• Solar Turbines Inc

Inventors

• Sepulveda Henry P
• Stein Brandon W
• Sanders Jamaal A
• Theron Stephen E
• Kirksey Nathan J
• Carlos Ivan R
• Di Norscia Christian
• Nadadur Karunakaran Thirumangalam
• Vijay Kumar Vishwas Banglore
• Torres John Charles

Classifications

• F23N5/022 Primary

  using electronic means (F23N5/04 - F23N5/14 take precedence) · CPC title

• G06N20/00

  Machine learning · CPC title

• F23N2005/181

  using detectors sensitive to rate of flow of air · CPC title

• F23N5/184 Primary

  using electronic means · CPC title

• G01M15/14

  Testing gas-turbine engines or jet-propulsion engines · CPC title