Selective Upgrading: How Leading Firms Reduce Failure Trends

In the past decade, the reliability engineering landscape has shifted dramatically. Companies are moving from time‑based maintenance, and even from traditional predictive maintenance, toward a focused strategy: selective upgrading.

The Role of Service Providers
Even the most advanced organizations involve their service providers early in life‑cycle cost (LCC) studies. When LCC analysis points to a higher‑quality component, these firms are quick to upgrade, even if the upfront cost is higher. This approach is rooted in a simple principle that has held true for 30+ years: partnering with vendors who provide strong application engineering support yields significant long‑term savings and fewer failures.

Networking among reliability professionals also uncovers failure trends that might otherwise go unnoticed. A process reliability engineer from the North Central United States recently identified a pattern of repeat failures that foreshadowed a catastrophic event. His insight exemplifies the value of collaborating with peers who are committed to adding measurable value and maintaining high standards of integrity.

Repeat Failures: A Critical Signal
Repeat failures are recurring problems that prevent equipment from reaching its expected lifespan. They often arise when a failure mechanism is deemed “economical to manage” rather than to be eliminated. If left unchecked, repeat failures become normalized deviations—low‑risk nuisances that carry no immediate safety or environmental consequence. Yet, this mindset can lead to extreme failures—explosions, fires, or crashes—that have devastating consequences.

Repeat failures are the intermediate step in the life cycle of an extreme failure. They also serve as reliable warning signs that can be detected before a catastrophic event occurs. By proactively addressing repeat failures, operators reclaim control over equipment reliability and reduce the likelihood of “unpredictable” incidents.

Maintenance management systems often record reactive work orders such as “bearing replaced,” which add little diagnostic value. A more insightful entry would read, “bearing failed due to oil starvation caused by a pressure‑unbalanced constant‑level lubricator.” Such detailed records highlight the root cause and inform future prevention strategies. Unfortunately, repeat‑failure work orders frequently get buried under higher‑priority items that threaten immediate production.

When a repeat failure is treated as a low‑consequence risk, the time spent on a failure analysis is diverted from addressing pressing production constraints. This reactive approach is characteristic of repair‑focused organizations and should be questioned. True reliability engineering requires systematic investigation and action.

In short, the most successful operators and service providers will add value at every maintenance event. They view each intervention as an opportunity to upgrade—strengthening the weakest link in the component chain whenever the cost is justified. This proactive stance leads to operator‑driven reliability (ODR) and protects both people and assets.

We must move beyond traditional maintenance practices. It’s time for a fundamental shift in thinking—one that eliminates unacceptable risk in uncertain times.

To summarize the key points:

• Stop conducting business as usual.
• Lip service to reliability improvements is a disservice.
• Investigation must precede remedial action; otherwise, repeat failures recur.
• Repeat failures precede extreme failures, which are catastrophic.
• Show unwavering resolve to break the destructive chain of events.

Reference
Bloch, Kenneth and Stephanie Williams, "Normalize Deviance at Your Peril," Chemical Engineering, 111, No. 5, pp. 52‑56, 2004.