Application of abnormal event detection technology to delayed coking unit

• US 7,720,641 B2
• Filed: 04/13/2007
• Issued: 05/18/2010
• Est. Priority Date: 04/21/2006
• Status: Active Grant

• Current Assignee: ExxonMobil Research & Engineering Company (Exxon Mobil Corporation)
• Original Assignee: ExxonMobil Research & Engineering Company (Exxon Mobil Corporation)
• Inventors: Worden, Kevin R., Woo, Stephen S., Emigholz, Kenneth F., Nguyen, Anh T., Alagappan, Perry
• Primary Examiner(s): Feliciano; Eliseo Ramos
• Assistant Examiner(s): HENSON, MISCHITA L
• Application Number:
  US11/786,950
• Publication Number:
  US 20070250292A1
• Time in Patent Office: 1,131 Days
• Field of Search: 702 57- 59, 702 81- 84, 702/105, 702182-184, 702190-195, 201/1, 703/3, 700108-110, 700209-211
• US Class Current: 702/183
• CPC Class Codes: C10B 55/00 Coking mineral oils, bitumen, tar, and the like or mixtures thereof with solid carbonaceous material cracking oils C10G
• 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: G05B 23/0289 Reconfiguration to prevent failure, e.g. usually as a reaction to incipient failure detection

Abstract

The present invention is a method for detecting an abnormal event for process units of a Delayed Coking Unit. The method compares the operation of the process units to statistical and engineering models. The statistical models are developed by principal components analysis of the normal operation for these units. The engineering models are based statistical and correlation analysis between variables. If the difference between the operation of a process unit and the normal model result indicates an abnormal condition, then the cause of the abnormal condition is determined and corrected.

Claims

1. A method for the early notification of an unplanned abnormal event detection (AED) for some process units of a delayed coking unit (DCU) comprising:

(a) comparing online measurements from an array of sensors for the process units to a set of models including at least two principal component analysis models for normal operation of the corresponding process units, wherein said delayed coking unit has been divided into at least two equipment groups wherein there is minimal interaction between said equipment groups and principal component analysis models correspond to equipment groups, wherein the inputs to the principal component analysis models include measurements from said array of sensors which have been divided into groups corresponding to equipment groups, wherein said equipment groups are defined by including all major material and energy integrations and quick recycles in the same equipment group and said measurements are cross-correlated with each other,

(b) determining if the current operation differs from expected normal operations from the cumulative measurements of said array of sensors that are inputs to said set of models so as to detect the presence of an unplanned abnormal event condition in a process unit which has developed on a timescale of minutes to hours, and

(c) determining the underlying cause of an abnormal condition in the DCU.

2. The method of claim 1 wherein said set of models correspond to equipment groups and operating modes, one model for each equipment group which may include one or more operating modes.

3. The method of claim 1 wherein said set of models correspond to equipment groups and process operating modes, one model for each equipment group and each operating mode.

4. The method of claim 1 wherein said delayed coking unit includes downstream towers and are decomposed into five monitors.

5. The method of claim 4 wherein each of the monitors generates a continuous signal indicating the probability of an abnormal condition in the area.

6. The method of claim 4 where a list of monitors are automatically identified, isolated, ranked and displayed for the operator.

7. The method of claim 1 wherein said process units are divided into operational sections of the DCU system.

8. The method of claim 7 wherein there are three operational sections.

9. The method of claim 8 wherein the three operational sections include the furnaces or heater, main fractionator and the gas plant.

10. The method of claim 9 wherein said set of models further identifies the consistency between measurements around a specific unit, including main fractionator, gas plant units, the wet gas compressor, flow to valve pressure drop which indicate any early breakdown in the consistency.

11. The method of claim 10 wherein said set of models further comprises suppressing model calculations to eliminate false positives on special cause operations.

12. The method of claim 1 wherein said set of models include process variables values measured by said array of sensors.

13. The method of claim 12 wherein said principal component models for different process units include some process variable values measured by a same sensor in said array of sensors.

14. The method of claim 1 wherein (a) said set of models begins with a principal components model based on questionable data, (b) gathering high quality training data and using said high quality training data in the model to modify the model and thereby improve the model, and (c) repeating step (b) of this claim to further improve the model.

15. The method of claim 14 further comprising the step of synchronizing pairs of measurements for two variables used in said model by time by one of the variables using a dynamic transfer function.

16. The method of claim 14 wherein said high quality training data includes historical data for the model of the processing unit.

17. The method of claim 16 wherein said set of models include transformed variables.

18. The method of claim 17 wherein said transformed variables are transformed to include pressure compensated temperature and/or pressure compensated flow measurements and/or valve positions to estimate flow.

19. The method of claim 1 wherein variables of process measurements that are affected by operating point changes in process operations are converted to deviation variables by subtracting the moving average.

20. The method of claim 1 wherein at least one of said models is corrected for noise.

21. The method of claim 20 wherein said set of models are corrected by filtering or eliminating noisy measurements of variables.

22. The method of claim 1 wherein the measurements of a variable are scaled.

23. The method of claim 22 wherein the measurements are scaled to the expected normal range of that variable.

24. The method of claim 1 wherein the step of determining the underlying cause of the abnormal condition provides diagnostic information at different levels of detail to aid in the investigation of the event.

25. The method of claim 1 wherein the number of principal components is chosen such that coefficients of the principal component become about equal in size.

26. The method of claim 1 wherein said set of models further includes engineering models.

27. The method of claim 1 wherein said principal components include process variables provided by online measurements.

28. The method of claim 27 further comprising the step of synchronizing measurements by time to one of the variables using a dynamic filter.

29. The method of claim 27 wherein the process measurement variables affected by operating point changes in the process operations are converted to deviation variables.

30. The method of claim 27 wherein the number of principal components is selected by the magnitude of total process variation represented by successive components.

31. A system for early notification of unplanned abnormal event detection (AED) for some of the process units of a DCU of a petroleum refinery comprising:

(a) on-line data acquisition of measurements from an array of sensors,

(b) a set of models including at least two principal component analysis models included in the set describing operations of said process units including automatic detection of drum switches and furnace decoking operations wherein said DCU has been divided into at least two equipment groups, wherein a single principal component model corresponding to an equipment group, and wherein the inputs to the principal component analysis models include measurements from said array of sensors, wherein said array of sensors has been divided into groups corresponding to said equipment groups, and wherein said equipment groups are defined by including all major material and energy integrations and quick recycles in the same equipment group and said measurements are cross-correlated with each other,

(c) a set of displays which indicates if the current operation differs from expected normal operations from the cumulative measurements of said array of sensors so as to indicate the presence of an abnormal condition in the process unit which has developed on a timescale of minutes to hours so as to detect an unplanned abnormal event, and

(d) a set of displays which indicates the underlying cause of an abnormal condition in the DCU.

32. The system of claim 31 wherein said DCU is partitioned into three operational sections with principal components models for selected sections.

33. The system of claim 32 wherein said principal component models include process variables provided by online measurements.

34. The system of claim 32 wherein said set of models further comprises suppressing model calculations to eliminate operator induced notifications and false positives.

35. The system of claim 31 wherein:

(a) said set of models are obtained from an initial model based upon training data, (b) use of said initial model to obtain new data and improve the model, and (c) iteratively repeating step (b) to improve the model.

36. The system of claim 35 wherein said training data set includes historical data of the processing unit for model development.

37. The system of claim 36 wherein said set of models includes transformed variables.

38. The system of claim 37 wherein said transformed variables include pressure compensated temperature and/or pressure compensated flow measurements and/or flow to valve positions to estimate flow.

39. The system of claim 36 wherein some measurement pairs are time synchronized to one of the variables using a dynamic filter.

40. The system of claim 36 wherein the process measurement variables affected by operating point changes in the process operations are converted to deviation variables.

41. The system of claim 36 wherein the measurements of a variable are scaled prior to model identification.

42. The system of claim 41 wherein the measurements are scaled by the expected normal range of that variable.

43. The system of claim 31 wherein the number of principal components is selected by the magnitude of total process variation represented by successive components.

44. The system of claim 31 wherein said set of models further includes engineering models.