METHOD AND SYSTEM FOR DETECTING ABNORMAL OPERATION IN A STIRRED VESSEL

• US 20080082294A1
• Filed: 09/27/2007
• Published: 04/03/2008
• Est. Priority Date: 09/28/2006
• Status: Active Grant

• Current Assignee: Fisher-Rosemount Systems Incorporated (Emerson Electric Company)
• Original Assignee: Fisher-Rosemount Systems Incorporated (Emerson Electric Company)
• Inventors: PIHLAJA, Roger K., Miller, John P.
• Granted Patent: US 7,778,797 B2
• Time in Patent Office: Days
• US Class Current: 702/179
• CPC Class Codes: G05B 23/0221 Preprocessing measurements, e.g. data collection rate adjustment; Standardization of measurements; Time series or signal analysis, e.g. frequency analysis or wavelets; Trustworthiness of measurements; Indexes therefor; Measurements using easily measured parameters to estimate parameters difficult to measure; Virtual sensor creation; De-noising; Sensor fusion; Unconventional preprocessing inherently present in specific fault detection methods like PCA-based methods
• CPC Class Codes: G05B 23/024 Quantitative history assessment, e.g. mathematical relationships between available data; Functions therefor; Principal component analysis [PCA]; Partial least square [PLS]; Statistical classifiers, e.g. Bayesian networks, linear regression or correlation analysis; Neural networks

Abstract

A system and method for detecting abnormal situations associated with a stirred vessel in a process plant receives statistical data associated with a pressure within a stirred vessel. A pressure signal associated with the pressure in the vessel is filtered by a digital filter to isolate a frequency component corresponding to pressure changes caused by the movement of a blade of an agitator through a fluid. For example, a pressure sensor device disposed at least partially within the stirred vessel may generate the statistical data based on a pressure signal. The statistical data is analyzed to detect whether one or more abnormal situations associated with an agitator of the stirred vessel exist. For example, the statistical data may be analyzed to detect whether the agitator is broken/unbalanced, corroded, missing a blade or multiple blades, etc. If an abnormal situation is detected, an indicator of the abnormal situation may be generated.

Claims

1. A method for detecting an abnormal situation associated with a stirred vessel in a process plant, the method comprising:

collecting first data associated with pressure within a stirred vessel;

filtering the collected first data to isolate a frequency component corresponding to changes in pressure associated with blade rotation in the vessel;

generating statistical data from the filtered first data;

analyzing the statistical data to detect whether one or more abnormal situations associated with an agitator of the stirred vessel exist; and

generating an indicator of an abnormal situation if one or more of the one or more abnormal situations are detected.

2. A method as defined in claim 1, wherein filtering the collected first data to isolate a frequency component corresponding to changes in pressure associated with blade rotation in the vessel comprises filtering the collected first data to isolate a frequency component corresponding to a rate at which spikes occurs in the pressure signal generated by a pressure sensor disposed at least partially within a vessel of the agitator.

3. A method as defined in claim 1, wherein generating statistical data from the filtered first data comprises generating a standard deviation of a pressure signal generated by a pressure sensor disposed at least partially within a vessel of the agitator.

4. A method as defined in claim 1, further comprising:

collecting second data associated with a normal pressure within the stirred vessel; and

generating nominal statistical data from the collected second data, wherein the nominal statistical data represents an indication of a normal operation of the vessel;

wherein analyzing the statistical data to detect whether one or more abnormal situations associated with an agitator of the stirred vessel exist comprises comparing the statistical data to the nominal statistical data to detect whether one or more abnormal situations associated with the agitator of the stirred vessel exist.

5. A method as defined in claim 1, wherein analyzing the statistical data comprises determining if a standard deviation of the filtered first data is approaching zero or is at least approximately zero to detect whether the agitator has stopped turning.

6. A method as defined in claim 1, wherein analyzing the statistical data comprises determining if a standard deviation of the filtered first data has significantly deviated to detect corrosion of the agitator.

7. A method as defined in claim 1, wherein analyzing the statistical data comprises determining if a standard deviation of the filtered first data has deviated by a first threshold to detect a missing blade from the agitator.

8. A method as defined in claim 7, wherein analyzing the statistical data comprises determining if a standard deviation of the filtered first data has deviated by a second threshold to detect multiple missing blades from the agitator.

9. A method as defined in claim 8, wherein analyzing the statistical data comprises determining if a standard deviation of the filtered first data has deviated by a third threshold to detect massive agitator failure.

10. A method as defined in claim 1, wherein collecting first data associated with pressure within a stirred vessel comprises collecting the first data based on a pressure signal generated by a pressure sensor disposed at least partially within the stirred vessel.

11. A method for detecting an abnormal situation associated with a stirred vessel in a process plant, the method comprising:

collecting first data associated with pressure within a stirred vessel

filtering the collected first data to isolate a frequency component corresponding to changes in pressure associated with blade rotation in the vessel;

generating first statistical data from the filtered first data;

generating second statistical data from the collected first data; and

analyzing the first statistical data and the second statistical data to detect whether one or more abnormal situations associated with an agitator of the stirred vessel exist; and

generating an indicator of an abnormal situation if one or more of the one or more abnormal situations are detected.

12. A method as defined in claim 11, wherein analyzing the first statistical data and the second statistical data to detect whether one or more abnormal situations associated with the agitator of the stirred vessel exist comprises analyzing the first statistical data and the second statistical data to determine whether both a standard deviation of the filtered first data and a standard deviation of the collected first data have significantly deviated to detect corrosion of the agitator.

13. A method as defined in claim 11, wherein analyzing the first statistical data and the second statistical data to detect whether one or more abnormal situations associated with the agitator of the stirred vessel exist comprises analyzing the first statistical data and the second statistical data determine whether a standard deviation of the filtered first data has deviated by a first threshold and whether a standard deviation of the collected first data has no significant deviation to detect a missing blade from the agitator.

14. A method as defined in claim 13, wherein analyzing the first statistical data and the second statistical data to detect whether one or more abnormal situations associated with the agitator of the stirred vessel exist comprises analyzing the first statistical data and the second statistical data determine whether a standard deviation of the filtered first data has deviated by a second threshold and whether a standard deviation of the collected first data has no significant deviation to detect multiple missing blades from the agitator.

15. A method as defined in claim 14, wherein the first statistical data and the second statistical data to detect whether one or more abnormal situations associated with the agitator of the stirred vessel exist comprises analyzing the first statistical data and the second statistical data determine whether a standard deviation of the filtered first data has deviated by a third threshold and whether a standard deviation of the collected first data has no significant deviation to detect massive agitator failure.

16. A method as defined in claim 11, further comprising:

collecting second data associated with a normal pressure within the stirred vessel;

filtering the collected second data to isolate the frequency component corresponding to changes in pressure associated with blade rotation in the vessel;

generating first nominal statistical data from the filtered data, wherein the nominal statistical data represents an indication of a normal operation of the vessel; and

generating second nominal statistical data from the collected second data;

wherein analyzing the first statistical data and the second statistical data to detect whether one or more abnormal situations associated with an agitator of the stirred vessel exist comprises respectively comparing the first statistical data to the first nominal statistical data and the second statistical data to the second nominal statistical data to detect whether one or more abnormal situations associated with the agitator of the stirred vessel exist.

17. A system for detecting an abnormal situation associated with a stirred vessel in a process plant, the system comprising:

a digital bandpass filter tuned to a frequency component corresponding to a rate at which spikes occurs in a pressure signal generated by a pressure sensor disposed at least partially within a stirred vessel;

a statistical parameter generator to generate one or more first statistical parameters based on a filtered pressure signal; and

an abnormal situation detector to detect at least one abnormal situation associated with an agitator of the stirred vessel based on the one or more first statistical parameters, and to generate one or more indicators of one or more abnormal situations situation are detected.

18. A system as defined in claim 17, wherein the digital filter comprises a finite impulse response (FIR) filter.

19. A system as defined in claim 17, wherein the digital filter comprises a digital filter designed by a Parks-McClellan filter design algorithm.

20. A system as defined in claim 17, wherein the digital filter comprises a 16^th order filter.

21. A system as defined in claim 17, further comprising a gain multiplier to multiply the filtered pressure signal by a fixed gain factor.

22. A pressure transmitter for detecting an abnormal situation associated with a stirred vessel in a process plant, the pressure transmitter comprising:

a processor;

a memory;

a routine stored in the memory and adapted to be executed by the processor to receive data associated with a pressure signal generated by a pressure sensor disposed at least partially within a stirred vessel;

a routine stored in the memory and adapted to be executed by the processor to filter the received data to isolate a frequency component corresponding to a rate at which spikes occur in the pressure signal;

a routine stored in the memory and adapted to be executed by the processor to generate statistical data from the filtered data, wherein the statistical data comprises an indication of a standard deviation of the pressure signal at the frequency component; and

a routine stored in the memory and adapted to be executed by the processor to analyze the statistical data to detect whether one or more abnormal situations associated with an agitator of the stirred vessel exist based on changes in the standard deviation of the pressure signal at the frequency component.

23. A pressure transmitter as defined in claim 22, further comprising:

a routine stored in the memory and adapted to be executed by the processor to determine if the agitator of the stirred vessel is rotating;

a routine stored in the memory and adapted to be executed by the processor to;

generate an indication that the agitator has stopped rotating if the routine to analyze the statistical data determines the standard deviation of the filtered data is approaching zero or is at least approximately zero and if the routine to determine if the agitator is rotating determines the agitator is not running,

generate an indication that the agitator has fallen off or broken if the routine to analyze the statistical data determines the standard deviation of the filtered data is approaching zero or is at least approximately zero and if the routine to determine if the agitator is rotating determines the agitator is running,

generate an indication that the agitator is corroded if the routine to analyze the statistical data determines the standard deviation of the filtered data has significantly deviated,

generate an indication of a missing blade from the agitator if the routine to analyze the statistical data determines the standard deviation of the filtered data has deviated by a first threshold,

generate an indication of multiple missing blades from the agitator if the routine to analyze the statistical data determines the standard deviation of the filtered data has deviated by a second threshold greater than the first threshold, and

generate an indication of massive agitator failure if the routine to analyze the statistical data determines the standard deviation of the filtered data has deviated by a third threshold greater than the second threshold.

24. A pressure transmitter for detecting an abnormal situation associated with a stirred vessel in a process plant, the pressure transmitter comprising:

a processor;

a memory;

a routine stored in the memory and adapted to be executed by the processor to receive data associated with a pressure signal generated by a pressure sensor disposed at least partially within a stirred vessel;

a routine stored in the memory and adapted to be executed by the processor to filter the received data to isolate a frequency component corresponding to a rate at which spikes occur in the pressure signal;

a routine stored in the memory and adapted to be executed by the processor to generate first statistical data from the filtered data, wherein the first statistical data comprises an indication of a standard deviation of the pressure signal at the frequency component;

a routine stored in the memory and adapted to be executed by the processor to generate second statistical data from the unfiltered received data, wherein the second statistical data comprises an indication of a standard deviation of the unfiltered pressure signal; and

a routine stored in the memory and adapted to be executed by the processor to analyze the first statistical data and the second statistical data to detect whether one or more abnormal situations associated with an agitator of the stirred vessel exist based on changes in one or more of the group consisting of;

the standard deviation of the pressure signal at the frequency component and the standard deviation of the unfiltered pressure signal.

25. A pressure transmitter as defined in claim 24, further comprising:

a routine stored in the memory and adapted to be executed by the processor to determine if the agitator of the stirred vessel is rotating;

a routine stored in the memory and adapted to be executed by the processor to;

generate an indication that the agitator has stopped rotating if the routine to analyze the first and second statistical data determines the standard deviation of the filtered data is approaching zero or is at least approximately zero and if the routine to determine if the agitator is rotating determines the agitator is not running,

generate an indication that the agitator has fallen off or broken if the routine to analyze the first and second statistical data determines the standard deviation of the filtered data is approaching zero or is at least approximately zero and if the routine to determine if the agitator is rotating determines the agitator is running,

generate an indication that the agitator is corroded if the routine to analyze the first and second statistical data determines the standard deviation of the filtered data has significantly deviated and the standard deviation of the unfiltered pressure signal has significantly deviated,

generate an indication of a missing blade from the agitator if the routine to analyze the first and second statistical data determines the standard deviation of the filtered data has deviated by a first threshold and the standard deviation of the unfiltered pressure signal has no significant deviation,

generate an indication of multiple missing blades from the agitator if the routine to first and second analyze the statistical data determines the standard deviation of the filtered data has deviated by a second threshold greater than the first threshold and the standard deviation of the unfiltered pressure signal has no significant deviation, and

generate an indication of massive agitator failure if the routine to first and second analyze the statistical data determines the standard deviation of the filtered data has deviated by a third threshold greater than the second threshold and the standard deviation of the unfiltered pressure signal has no significant deviation.