Life cycle of control systems
Any piece of electronics follows a life cycle curve roughly looking as shown on the right.
The initial high failure rate (Y-axis) is commonly referred to as ‘infant mortality rate’ and occurs during the first half year. Through burn-in periods at the manufacturer, during in-house tests and through warranty clauses, this period is well covered.
After the system has proven itself there is a window of 10 to 15 years with only a few failures; thereafter the mortality rate goes up again.
At roughly the same time spare parts, due to obsolescence and support availability decreases. And the old technology is no longer thoroughly understood by the new generation of engineers not trained on these systems.
This typically means that control systems start to fail when support and repair become harder to accomplish and often at a high cost. Even if the system is still functional, it has reached end of economical life. This is further accelerated by the fact that new technology offers technical and cost-saving operational advantages.
The consequence is that one must make a choice: either invest heavily in spare parts and know-how on site –or– ensure that systems are being modernised before the end of the life cycle is reached.
Note that the ‘amplitude’ of the curve depends several factors such as mechanical stress, thermal stress, corrosion, type of component and complexity of the system.