Failure demand analysis technique for optimizing design reliability

Reliability Engineering is a complex discipline in and of itself. The numeric implications are many times difficult to model in real world terms. The problem is compounded by the combination of internal and external influences such that it sometimes becomes easy to become distanced from the full implications of the analytical activities at hand. Such was the case of what is referred to as EQPT. This equipment system had a single system reliability requirement of 0.92 against an associated reliability model and block diagram. At some point into the design effort the mathematical assessment of the system´s reliability was 0.937529, well above the requirement. As will be shown in the paper, looks can be deceiving, and as Reliability Engineers we owe it to the spirit of our craft to periodically come full circle back to an external objective method of evaluating the resulting goodness of our efforts. Perhaps the challenge might even be a little broader to include understanding all the implications our reliability can and will mean long after we´re finished conducting our models and analyses. Concisely put, we don´t engineer reliability for the mere sake of going through the steps of achieving a requirement. We engineer reliability for an end purpose. It behooves us to know and understand our reliability efforts within the realm of those end purposes. The math in this paper is considered elementary to the art. However, it is not always the math that needs attention for the art to advance. Sometimes it is the application of the math and the discovery of blind spots. And as will be shown later on in the paper the reward can be quite valuable, in this case, to help the Reliability Engineer add another tool to his/her toolbox in ascertaining answers to questions such as, “How much reliability is enough? How much reliability is too much? And, what does my reliability actually mean in real world operational terms?”.