When we do not know how to directly prevent or retard equipment failure-or it is impossible to do so — the next best thing that we can hope to do is to detect its onset and predict the point in time where failure is likely to occur in the future. We do this by measuring some parameter over time where it has been established that the parameter correlates with incipient failure conditions. When such is done, we call it a condition-directed or CD task. Thus, a CD task would pre-warn us to take action to avoid the full failure event. If the warning comes soon enough, our action can most likely be taken at some favorable timing of our choice.
The CD task, like the TD task, has a periodicity for the measurements, but actual preventive actions are not taken until the incipient failure signal is given. The CD task takes two forms: (1) we can measure a performance parameter directly (e.g., temperature, thickness) and correlate its change over time with failure onset; or (2) we can use external or ancillary means to measure equipment status for the same purpose (e.g., oil analysis or vibration monitoring). With the CD task, all such measurements are nonintrusive. The keys to classifying a task as CD are: (1) we can identify a measurable parameter that correlates with failure onset; (2) we can also specify a value of that parameter when action may be taken before full failure occurs; and (3) the task action is nonintrusive with respect to the equipment.
In large complex systems and facilities, there are almost always several equipment items-or possibly a whole subsystem or system-that could experience failure and, in the normal course of operation, no one would know that such failure has occurred. We call this situation a hidden failure. Backup systems, emergency systems, and infrequently used equipment constitute the major source of potential hidden failures. Clearly, hidden failures are an undesirable situation since they may lead to operational surprises and could then possibly initiate an accident scenario via human error responses. For example, an operator may go to activate a backup system or some dormant function only to find that it is not available and, in the pressure of the moment, fail to take the correct follow-up procedure. So, if we can, we find it most beneficial to exercise a prescheduled option to check and see if all is in proper working order. We call such an option a failure-finding task.
Let's look at a simple example-the spare tire in our automobile. If you are like us, you don't really worry about a flat spare tire because you have AAA coverage, and are never more than 10 to 15 minutes away from an ability to get emergency road service-except for that once-a-year trip with the family into "uncharted lands" (e.g., Death Valley). Again, if you are like us, you do check the spare before you leave — and that is a failure-finding task.
Notice that the only intent in such an action is to determine if the spare tire is in working order or not. We are doing nothing to prevent or retard a flat tire (a TD task) or to measure its incipient failure condition (a CD task). It is or is not in working order. And, if it is not in working order, we fix it. That is the essence of what a failure finding task is all about. (Is it OK? If not, fix it.)
0 Comments:
Post a Comment