Data processing: measuring – calibrating – or testing – Measurement system – Statistical measurement
Reexamination Certificate
2001-03-28
2004-05-11
Hoff, Marc S. (Department: 2857)
Data processing: measuring, calibrating, or testing
Measurement system
Statistical measurement
C701S029000
Reexamination Certificate
active
06735549
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates generally to process monitoring display systems and more particularly to predictive maintenance display systems that determine, with a measure of specificity, the date on which components should be serviced to avoid the likelihood that they will fail during operation.
2. Related Art
Many industries, such as the power generation industry, have experienced an increased awareness of and emphasis on the benefits and use of predictive maintenance technologies. Use of such technologies has the potential to improve the long-term availability and reliability of plant components resulting in an overall improvement to plant operability.
Predictive maintenance methodologies currently utilize a variety of techniques in order to predict subsequent equipment and system failures. With the present state of the art, the predicted failures that are based on time remaining to failure are typically depicted in terms of the future point in time in which the failure will likely occur along with a corresponding confidence interval. A typical representative prediction can be stated as:
Predicted failure point: 2000 hours (+/−500 hours with a 95% confidence interval).
The above example indicates that the system is predicted to fail within 2000 hours from now (mean-time-to-failure). It further indicates that failure will occur between 1500 and 2500 hours from now with a 95% confidence (that is, 95 out of 100 times the failure will occur between 1500 and 2500 hours from now, and 5 out of 100 times the failure will occur for times outside this range).
However, such predictions contain a number of deficiencies, for example:
1. The predicted failure range is often too wide and thus, it becomes difficult to efficiently and cost effectively schedule maintenance in advance of the predicted failure.
2. Operations and maintenance personnel are often misled by predictions of failure in which a failure interval is given (as illustrated above). An implicit assumption is often made that since equipment failure can occur anywhere within the predicted interval, that it is better to wait until the end of the interval in order to obtain additional operating life from the equipment and not replace the equipment prematurely. However, this logic is faulty because the probability of failure is based on a continuous increasing probability continuum. Thus, the probability that the equipment will fail in service before the predicted mean-time-to-failure is low for a time interval that occurs near the beginning of the predicted interval. However, the probability that the equipment will fail in service before the predicted mean-time-to-failure is high for a time interval that occurs near the end of the predicted interval.
As an illustration, considering the example noted above and referring to
FIG. 1
, assume the equipment failure prediction conforms to a Normal Distribution, and assume it is desired to refurbish/replace the equipment prior to experiencing an “in-service” failure, then:
At 1500 hours, the probability that the equipment will fail “in-service” (prior to the 1500 hour point) will be only 2.5% (another way of stating this is that there is a 97.5% probability that the equipment will fail after 1500 hours). Thus, if a decision is made to perform maintenance on the equipment at the 1500-hour point, there is only a 2.5% chance that the equipment will fail in-service, prior to the 1500-hour point. Therefore, it is highly likely that the equipment can be successfully refurbished/replaced prior to the occurrence of an “in-service” failure.
At 2500 hours, the probability that the equipment will fail “in-service” (prior to 2500 hours) will be 97.5% (another way of stating this is that there is a 2.5% probability that the equipment will fail after 2500 hours). Thus, if a decision is made to perform maintenance on the equipment at the 2500 hour point, there is a very significant 97.5% chance that the equipment will fail “in-service,” prior to the 2500 hour point. Therefore, it is highly unlikely that the equipment can be successfully refurbished/replaced prior to the occurrence of an “in-service” failure.
Such information, on the probability of equipment failing prior to a selected maintenance date, is unavailable based on current predictive maintenance display methodologies, and represents a significant shortcoming with the current methodologies.
3. Simply providing the expected interval in which the failure is likely to occur does not alone provide maintenance and operational personnel sufficiently detailed information on the probability of failure from any specific point in time within the predicted failure interval. Because failure predictions are based on a continuous increasing probability continuum, there is wide variance on the probability of failure from anywhere within the expected failure interval. In order to efficiently accommodate and plan for future equipment and system failures, the maintenance and operational personnel require the specific probability of equipment/system failure from any point within a predicted failure interval.
4. Currently, the capability for operators and maintenance personnel to perform interactive “what if” scenarios based on future points in time, does not exist. Such capability will allow personnel to explore various maintenance scheduling alternatives, by determining what the specific probability of equipment failure will be for any future point in time. By obtaining the specific probability of failure on a given date, the operations and maintenance personnel can explore “what if” scenarios to better decide when to schedule refurbishment/repair activities. Providing such capability will allow operators and maintenance personnel to fully explore the probability continuum and optimize maintenance and scheduling activities.
5. Currently, the capability for operators and maintenance personnel to perform interactive “what if” scenarios based on the desired probability of equipment failing before a calendar date, does not exist. Such capability will allow personnel to explore various maintenance scheduling alternatives, by determining what calendar date corresponds to a specified probability of equipment failing prior to the calendar date. By obtaining the calendar date for a specified probability of failure prior to the calendar date, the operations and maintenance personnel can explore “what if” scenarios to better decide when to schedule refurbishment/repair activities. Providing such capability will allow operators and maintenance personnel to fully explore the probability continuum and further optimize maintenance and scheduling activities.
SUMMARY OF THE INVENTION
Accordingly, it is an object of this invention to provide a predictive maintenance display system that will identify the specific probability of failure of a monitored component for any given date inputted by plant personnel. Additionally, it is a further object of this invention to provide such a display system that will identify the date a monitored component is not likely to fail on or before for a given probability inputted by plant personnel.
These and other objects are achieved by this invention which includes an Equipment Failure And Degradation Module that determines the remaining equipment/system life from measurements taken from plant sensors that monitor various components and subcomponents in a plant. Preferably, various methodologies are utilized by the Equipment Failure And Degradation Module to ascertain the potential of incipient equipment failures and to predict the equipment remaining life, to best fit the methodology to the component monitored. Examples of these methodologies include trend analysis, pattern recognition, correlation techniques, limits and ranges, data comparison, and statistical process analysis. The predictive maintenance algorithms utilized by the Equipment Failure And Degradation Module will employ a variety of the aforementioned techniques that best suit the equipment or system that is
Charioui Mohamed
Hoff Marc S.
Westinghouse Electric Co. LLC
LandOfFree
Predictive maintenance display system does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Predictive maintenance display system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Predictive maintenance display system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3225323