Method and apparatus for detecting faults in steam generator system components and other continuous processes

• US 7,113,890 B2
• Filed: 10/07/2004
• Issued: 09/26/2006
• Est. Priority Date: 10/16/2003
• Status: Expired due to Fees

• Current Assignee: ABB Incorporated (ABB Limited)
• Original Assignee: ABB Incorporated (ABB Limited)
• Inventors: Frerichs, Donald Karl, Toth, Frank Marvin
• Primary Examiner(s): BARNES-BULLOCK, CRYSTAL JOY
• Application Number:
  US10/962,150
• Publication Number:
  US 20050096757A1
• Time in Patent Office: 719 Days
• Field of Search: 700 28- 30, 700/32, 700/33, 700/44, 700/45, 700/47, 700/48, 700/21, 700/51, 700/79, 700/121, 700174-178, 702/179, 702181-185, 706 14- 16, 706 17- 20, 706 45- 48, 714/1, 714/2, 714/15, 714/21, 714/32, 714 25- 27, 714/37, 714/40, 714/44, 714/47, 714/48, 382155-160
• US Class Current: 702/185
• CPC Class Codes: F22B 35/18 Applications of computers to steam boiler control
• CPC Class Codes: F28F 2265/16 for preventing leakage
• CPC Class Codes: G05B 17/02 electric
• CPC Class Codes: G05B 23/0254 based on a quantitative model, e.g. mathematical relationships between inputs and outputs; functions: observer, Kalman filter, residual calculation, Neural Networks
• CPC Class Codes: Y02P 90/02 Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]

Abstract

A method and apparatus and for detecting faults in components of a continuous process such as a steam generator. A model of the process is developed using a modeling technique such as advanced pattern recognition and the model is used to generate predicted values for a predetermined number of the operating parameters of the process. Statistical process control methods are used to determine if the difference between the predicted and actual measured values for one or more of the parameters exceeds a configured statistical limit. A rule set is used to indicate an actual or probable fault in a component of the continuous process.

Claims

1. A method for detecting a fault in a component of a continuous process, comprising:

developing a model of said continuous process;

generating predicted values for a predetermined number of operating parameters of said continuous process using said model;

comparing the value predicted by said model for each of said predetermined number of operating parameters to a corresponding actual measured value for said operating parameter; and

determining whether differences between said predicted and actual measured values for one or more of said predetermined number of operating parameters exceeds a configured statistical limit using Statistical Process Control (SPC) methods;

wherein said predetermined number of operating parameters of said continuous process depends on said continuous process; and

wherein said continuous process is the water/steam side of a boiler/turbine power cycle and said predetermined number of parameters include make-up flow, feedwater flow and condensate flow.

2. The method of claim 1 further comprising determining whether differences between said predicted and actual measured values for one or more of said predetermined number of operating parameters meets a predetermined SPC pattern test.

3. The method of claim 1 wherein said model is selected from an Advanced Pattern Recognition (APR) empirical model, a first principles model or a neural network empirical model.

4. The method of claim 3 further comprising calculating in said APR empirical model a Model Fit parameter from said differences between said predicted and actual measured values for one or more of said predetermined number of operating parameters.

5. The method of claim 1 wherein said predetermined number of operating parameters of said continuous process is between about 50and about 100.

6. The method of claim 1 further comprising determining for each of said predetermined number of operating parameters a difference between said predicted values for said operating parameter and said actual measured value for said operating parameter.

7. The method of claim 6 wherein said difference for selected ones of said predetermined operating parameters are each compared to an associated three sigma limit.

8. The method of claim 6 further comprising determining a deviation for each of said differences for said make-up flow, feedwater flow and condensate flow operating parameters to a three sigma limit associated with each of said operating parameters.

9. The method of claim 8 further comprising indicating:

a large leak in a tube of a boiler in said continuous process when said deviation for any two of said make-up flow, feedwater flow and condensate flow operating parameters are positive and statistically large for a predetermined period of time; and

a small leak in a tube of said boiler when said deviation for any two of said make-up flow, feedwater flow and condensate flow operating parameters are slightly positive for a predetermined period of time.

10. The method of claim 8 further comprising:

determining from all of said differences a parameter indicative of how good said model fits said continuous process;

analyzing said deviation for each of said make-up flow, feedwater flow and condensate flow operating parameters in accordance with an associated predetermined pattern; and

indicating that a leak may be present in a tube of said boiler when any one of said deviations matches said associated predetermined pattern and said parameter indicative of said model fit is less than a predetermined value for a predetermined period of time.

11. A process plant comprising:

a computing device for detecting a fault in a component of a continuous process operating in said plant, said computing device for;

developing a model of said continuous process;

generating predicted values for a predetermined number of operating parameters of said continuous process using said model;

comparing the value predicted by said model for each of said predetermined number of operating parameters to a corresponding actual measured value for said operating parameter; and

determining whether differences between said predicted and actual measured values for one or more of said predetermined number of operating parameters exceeds a configured statistical limit using Statistical Process Control (SPC) methods;

wherein said computing device is also for determining for each of said predetermined number of operating parameters a difference between said predicted values for said operating parameter and said actual measured value for said operating parameter;

wherein said computing device is also for comparing said difference for selected ones of said predetermined operating parameters to an associated three sigma limit; and

wherein said continuous process is the water/steam side of a boiler/turbine power cycle and said predetermined number of operating parameters include make-up flow, feedwater flow and condensate flow and said computing device is also for determining a deviation for each of said differences for said make-up flow, feedwater flow and condensate flow operating parameters to a three sigma limit associated with each of said operating parameters.

12. The process plant of claim 11 wherein said computing device is also for determining whether differences between said predicted and actual measured values for one or more of said predetermined number of operating parameters meets a predetermined SPC pattern test.

13. The process plant of claim 11, wherein said computing device is operable to indicate:

a large leak in a tube of a boiler in said continuous process when said deviation for any two of said make-up flow, feedwater flow and condensate flow operating parameters are positive and statistically large for a predetermined period of time; and

a small leak in a tube of said boiler when said deviation for any two of said make-up flow, feedwater flow and condensate flow operating parameters are slightly positive for a predetermined period of time.

14. The process plant of claim 11, wherein said computing device is operable to:

determine from all of said differences a parameter indicative of how good said model fits said continuous process;

analyze said deviation for each of said make-up flow, feedwater flow and condensate flow operating parameters in accordance with an associated predetermined pattern; and

indicate that a leak may be present in a tube of said boiler when any one of said deviations matches said associated predetermined pattern and said parameter indicative of said model fit is less than a predetermined value for a predetermined period of time.