Hi All,

I am trying to come up with a list of criteria that, when met, will kick-off a formal RCA team to investigate a failure. I need your opinion/ideas on when do you think that should happen. Thanks in advance.
Original Post
Josh & MWP,
Investigating every failure maybe a bit overwhelming. We want to investigate safety incidents and machinery failures that have adverse production, environmental and maintenance cost consequences. MTBF maybe useful for machinery/component failures but what about safety incident investigations?
If you have your maintenance KPI's being tracked and also the Production related KPI's then i guess a brain storming of them will help you come out with the most important ones. Once decided you can always have targets inline with your business requirements and threshold limits can be agreed in the group for triggering an RCA.

My 2cents, IMHO
We do investigate near misses but not necessarily as a team. It is mostly an individual function. My question is when is it necessary to put a formal team together? If you come across a problem, how would you decide whether to form a team or assign an individual to solve the problem.
Sohail,
You are quite right in seeking a filtering mechanism, as it is not economically viable to invest in full RCA team studies for all failure events.

The starting point is a register of actual failures as well as potential ones. This is constantly updated with additions and deletions.
A ranking system to evalute the risks posed, typically with a Risk Matrix helps sort them out by risk levels.
High risk events, usually involving safety or environmental hazards, and those causing very high production losses are clearly worth the effort.
Those events falling in the medium risk group may initially be analysed with a two person team using a simple method e.g., 5-why in a day or less. Some of these may be solved while in others, more complications may evolve during the simple analysis. These may then be tackled with a full team using a more advanced method (Kepnor Tregoe©®, Proact©®, Apollo©®, TapRoot©®, Tripod©® etc,).
As analysis gets completed the register gets depleted; as new events occur, they get added on.
Low risk events can be analysed at site by suitably trained technicians themselves.
Josh,
The titles I quoted are mainly software solutions with different flavors.
RCA is NOT software. I like to see it as 'thought-ware'. One can follow the RCA process without any software , but using them can help us with the process. I conduct RCA training courses (next one in KL in Nov. through petroEdge Singapore) w/o the use of software, so people know the process first. I do mention the various packages they could use to speed up and to organise rheir work.
In a public forum, I do not feel it is appropriate to discuss these opinions. In general. all of them add value – speed, rigor, presentation. But do not use them till you understand the RCA process itself well. That is independent of the softeare package
quote:
Originally posted by Sohail:
Josh & MWP,
Investigating every failure... We want to investigate safety incidents and machinery failures...MTBF maybe useful for machinery/component failures but what about safety incident investigations?


Mr. Sohail
In our site each area are responsible for your RCA. The security department uses the 5-Why and department maintenance / reliability analysis using a spreadsheet based on Reliability Centered Maintenance.
All depends on how your system is organized.

Add Reply

Likes (0)
Post

HACKATHON in association with IMC-2019

×
×
×
×