Turbine blade (bucket) health monitoring and prognosis using neural network based diagnostic techniques in conjunction with pyrometer signals

• US 6,786,635 B2
• Filed: 11/06/2002
• Issued: 09/07/2004
• Est. Priority Date: 11/06/2002
• Status: Active Grant

• Current Assignee: General Electric Company
• Original Assignee: General Electric Company
• Inventors: Choi, Sukhwan
• Primary Examiner(s): VERBITSKY, GAIL KAPLAN
• Application Number:
  US10/288,296
• Publication Number:
  US 20040086024A1
• Time in Patent Office: 671 Days
• Field of Search: 374/45, 374/57, 374/102, 374/104, 374/103, 374/137, 374/141, 374/144, 374/152-153, 374/120-121, 374/112
• US Class Current: 374/153
• CPC Class Codes: F01D 21/003 Arrangements for testing or measuring for measuring vibrations G01H
• CPC Class Codes: F01D 21/10 responsive to unwanted deposits on blades, in working-fluid conduits or the like
• CPC Class Codes: F01D 21/12 responsive to temperature
• CPC Class Codes: F01D 21/14 responsive to other specific conditions
• CPC Class Codes: G01J 5/0022 for sensing the radiation of moving bodies
• CPC Class Codes: G05B 23/0283 Predictive maintenance, e.g. involving the monitoring of a system and, based on the monitoring results, taking decisions on the maintenance schedule of the monitored system; Estimating remaining useful life [RUL] preventive maintenance, i.e. planning maintenance according to the available resources without monitoring the system G06Q10/06

Abstract

Oxidation of turbine buckets may cause unexpected and expensive turbine failures. Turbine bucket oxidation condition may be estimated to predict remaining useful bucket life during operation of a turbine by processing time-varying temperature distributions measured with a pyrometer of at least one rotating turbine bucket.

Claims

1. A method of estimating turbine bucket oxidation condition comprising:

(a) measuring, with a pyrometer, time-varying temperature distributions of at least one rotating turbine bucket; and

(b) determining a condition index based on the measured time-varying temperature distributions including a difference between a peak temperature value and a mean temperature value of the at least one rotating turbine bucket, the condition index reflecting at least one of an overall condition of a bucket set or a specific condition of a single bucket.

2. A method according to claim 1, wherein step (b) is practiced by applying a diagnostic algorithm to selected data from the pyrometer.

3. A method according to claim 1, wherein step (b) is practiced according to CI_k=SDD_k−

SDD_k,ref, where k is a bucket number, SDD_k is an output from smoothing, and SDD_k,ref is a smoothing reference output.

4. A method according to claim 3, wherein SDD_k is determining by applying a smoothing algorithm to DD_k, which is determined according to DD_k=D_k−

Mean_Dk, where D_k is a relative amount of difference between a peak temperature and mean of two stationary points, and Mean_Dk is a mean of D_k'"'"'s of all normal blades.

5. A method according to claim 4, wherein D_k is determined according to D_k=T_k,2−

(T_k,3+T_k,4)/2, where T_k,2 is a peak temperature value for blade k, and T_k,3 and T_k,4 are stationary points on a temperature distribution curve of blade k.

6. A method according to claim 1, further comprising (c) determining a bucket oxidation amount according to a comparison with known inspection data.

7. A method according to claim 6, wherein step (c) is practiced by establishing a feed-forward neural network and defining an oxidation level of a failed turbine bucket.

8. A method according to claim 7, further comprising determining remaining bucket life via the feed-forward neural network according to the bucket oxidation amount.

9. A method of estimating turbine bucket oxidation condition comprising:

(a) measuring, with a pyrometer, time-varying temperature distributions of a plurality of rotating turbine buckets; and

(b) determining a condition index based on the measured time-varying temperature distributions, the condition index reflecting at least one of an overall condition of a bucket set or a specific condition of a single bucket by calculating a ratio of a peak temperature value within each passage of one of the plurality of rotating turbine buckets to a mean temperature value of the time-varying temperature distributions of all of the plurality of rotating turbine buckets measured by the pyrometer for one revolution of the turbine.

10. A method of estimating turbine bucket oxidation condition and predicting remaining useful bucket life during operation of a turbine by processing time-varying temperature distributions measured with a pyrometer of at least one rotating turbine bucket, the time-varying temperature distributions including a difference between a peak temperature value and a mean temperature value of the at least one rotating turbine bucket.

11. A method according to claim 10, wherein the step of estimating turbine bucket oxidation is practiced by applying a diagnostic algorithm to selected data from the pyrometer.

12. A method according to claim 11, wherein the diagnostic algorithm is expressed as to CI_kSDD_k−

SDD_k,ref where k is a bucket number, CI is a condition index, SDD_k is an output from smoothing, and SDD_k,ref is a smoothing reference output.

13. A method according to claim 12, wherein SDD_k is determining by applying a smoothing algorithm to DD_k, which is determined according to DD_k=D_k−

Mean_Dk, where D_k is a relative amount of difference between a peak temperature and mean of two stationary points, and Mean_Dk is a mean of D_k'"'"'s of all normal blades.

14. A method according to claim 13, wherein D_k is determined according to D_k=T_k,2−

(T_k,3+T_k,4)/2, where T_k,2 is a peak temperature value for blade k, and T_k,3 and T_k,4 are stationary points on a temperature distribution curve of blade k.

15. A method according to claim 10, wherein the step of estimating turbine bucket oxidation is practiced according to a comparison with known inspection data.

16. A method of estimating turbine bucket oxidation condition and predicting remaining useful bucket life during operation of a turbine by processing time-varying temperature distributions measured with a pyrometer of a plurality of rotating turbine buckets, wherein the step of estimating turbine bucket oxidation is practiced by calculating a ratio of a peak temperature value within each passage of one of the plurality of rotating turbine buckets to a mean temperature value of the time-varying temperature distributions of all of the plurality of rotating turbine buckets measured by the pyrometer for one revolution of the turbine.

17. A system for estimating turbine bucket oxidation condition comprising:

a pyrometer that measures time-varying temperature distributions of at least one rotating turbine bucket; and

a processor receiving output from the pyrometer, the processor determining a condition index based on the measured time-varying temperature distributions including a difference between a peak temperature value and a mean temperature value of the at least one rotating turbine bucket, wherein the condition index reflects at least one of an overall condition of a bucket set or a specific condition of a single bucket.