Voici comment vous pouvez éviter les pannes récurrentes dans vos projets d’ingénierie d’usine.

Reoccurring or chronic failure means finding or correcting root cause has not happened yet. Root cause failure analysis is not easy especially on chronic issues. The easy failures have been solved. Solving chronic failures requires a commitment of resources to solve. Many organizations just don’t have the will. If you don’t have right resources (SME), you can hire them. Experience and expertise is required to solve chronic failures.

Preventing recurring failures in plant, needs competent Maintenance & Operation teams to understand failure scenario and perform necessary corrective actions. And, they should inovolve teams like Asset integrity department or Condition Monitoring department etc., in the investigation process. All team involved, should have the knowledge, training & experience to perform Root cause failure analysis. The data or the conclusion are more validated. If the root cause analysis was done by competent team. If the company is small with limited manpower. They can hire an reputed 3rd party company to perform RCFA. And, management should also support Maintenance team in all aspects. From the above points best RCFA document can be developed.

Imho, in the context of RCA, one thing very hard to avoid future occurence is human error. Multiple factors involved in this type of event. And also false root cause found due to lack of evidence and/or misleaded logic construction. All in all, experience will taught us the hard way.

Audit needs to be the right level of audit - at a maintenance and reliability engineering level not a management KPI level. Knowing certain tasks are happening like PMs is nice but the quality, content and execution detail of PMs is where the difference is made. A 100% PM KPI completion means nothing if the task is pencil whipped, has wrong inspection detail or if techs don’t know how to execute PM.

Imho, real life equipment audit planning will just fall apart under the (false) bigger stake i.e "production cannot be stopped" as per our planning. This situation will push us to be more clever and do things out of the box depending on each specific circumstances. So, no static book standardized to tackle every aspects of equipment reliability and maintenance, but it is indeed a Life Book type, which will be adjusted here and there every now and then.

Imho, when we planning and scheduling maintenance activities for entire equipment, we need to be careful to account total manhours not to be very close to 100% utility. The danger come when we run into an uncontrollable critical equipment failure lead to plant shutdown (forced outage) and then due to wrong basis staffing policy, made us dont have enough manpower to overcome the problem. It is better to see our staff seemingly having under utilitize than we desperately looking for extra hands when it most needed

This is really key and most organizations fail in this. Training or lack of is a major latent cause of failures. In my experience I’d guess at least 80% has training as a latent root cause. Evaluating skill gaps and building a custom plan to address is critical. Must be constantly evaluating workforce skills

Imho, counting training manhour as part of overall maintenance personnel manhour effective enough to make sure our maintenance program and staff training program running smooth. And we institutionalized this feature in weekly planning and daily scheduling process. No more manhour unaccounted for

In industrial settings, I like when one of the new project Engineering steps is Project Leads have to present the improvement opportunity to frontline employees to get their "experience input". Monthly Shift Safety meetings are a good time to engage employees in new solutions to improve project success and buy in.

Yes, imho this is the only way for obsolete and no longer produced control system and instrumentation parts, no other way around

In the right places upgrades can be good but don’t fall for the trap that everything new is better because it is not. Infant mortality is real too. New stuff will bring with it new failures that may be worse than previous. Solve the problem in front of you before spending money on something you don’t understand.

If you want to solve chronic repeat failures hire a good Reliabilty engineer to focus on these improvements. The shift crew isn’t going to solve many of these chronic issues, they are busy keeping it run in on their shift.

What's most important is having an organized team that is focused on the same goal. Firm timelines need to be monitored and adhered to making sure that the project is completed within the necessary time frame. Often times, it is important to bring trusted resources into the team, trusted vendors and advisors who can help you guide your project to a successful completion. The team must have a leader who can guide the team members but who also must be open to input from the team.

I’ve never been a proponent of establishing a fixed plan and sticking to it, as inevitably they become out of date as soon as they are set. My preference is to implement a real time strategy that uses many of the processes mentioned above but applies real time predictive processes as conditions change as they always do. Recurring failures always are indicative of a plan that has not considered all the stressors and skewed the definition of the true root cause(s) or combination thereof. Often part of any predictive strategy is to perform system sensitivity testing to quantify how a system responds based on how engineering has predicted. Small changes in combination can destabilize a system just as they can stabilize a system.