Failure reduction - Targeting Analysis

I became disillusioned with Root Cause Analysis many years ago when I first bumped into Mr Bob latino from the Reliability Centre in Virginia. (Real class character and all round nice guy)

Since then I have worked a lot in this area and my continued work has sharpened my appreciation of the problems in this field.

Initiatives end up being long lists of redesigns waiting for capital/approval/time/labour and all the while the problem persists. Or we dedicate so much of our time working on things that are going to take weeks to analyse correctly. Or worse still, we end up blaming the fallible human and sending them off for unneeded retraining.

This post is the first in a series of posts on ReflexRCA. A method of to embed root cause analysis across the organization using straightforward principles and methods, and based on tools you probably have at your disposal today. 

The series will include:

• Targeting Analysis
• Causal Analysis (Depth of analysis)
• Fixing failure (Resolving problems)
• Case Studies (From a range of different industries)

We won't be talking here about issues such as implementation or justification of these programs. I am pretty sure that most companies see this as self evident. 

Targeting Analysis

Duran told us the difference between sporadic and chronic failures many years ago, yet we still tend to get a little confused. 

Sporadic failures are large scale failures that demand attention. They are often rare, and have a very large impact on the operation of the plant. Root Cause Analysis is rightly required on these issues to ensure that they never happen again. 

However, most of the effort expended is to restore the function back to normal, and as the event was rare, the benefits could be calculate but are unlikely to occur again in any case. 

Sporadic failures - Large scale, demand attention, require high levels of effort. 

Due to their large scale impact these types of failures are generally where we find ourselves spending most of our analysis time. 

However, there are other types of failures also that often provide far greater impact, and require less effort to resolve. 

These are chronic failures. Chronic refers to something that occurs continually, or a recurring failure in the terms of physical asset management context. 

They occur regularly, they have small impacts, and they are often accepted as part of the costs of doing business. 

Chronic Failures - Small scale, continuous or recurrent, accepted as part of doing business

The difference between the two failure types is that chronic failures will cause a step change in asset, team or financial performance. 

As mentioned, chronic failures tend to be accepted as acceptable losses. The costs of doing business that are too small and insignificant to deal with.

However, the impact of regular 10 minute outages, or frequent $200 belts, or regular trips requiring reset have a large impact when they are summed together. 

Finding chronic failures

Example of chronic failure analysis in asset industries

The above graph represents a good example for highlighting Chronic failures. We regularly help clients to apply this sort of analysis across mobile fleet, process plants, utilities and the graph above is for a Longwall installation.

The beauty of these types of illustrations is that they represent three different areas. Frequency of occurrence, average duration and total downtime. This helps people to see graphically which failures are frequent, or chronic, and which are sporadic, or large and infrequent.

The next step is to determine which of the chronic failures will provide the most value to the organisation. 

Value is not defined by the total return, but understanding the return and the level of effort required to resolve the situation. 

Effort versus Impact of chronic failures

The above graph shows a small example of an effort versus impact matrix for you to grade chronic failures against. This would require some estimation on your behalf, and would also need to be modified to suit your own requirements. 

However, it will ultimately provide you with a list of projects you can start with in your RCA program. 

The next post on this issue will talk about causal analysis and some of the thinking that needs to go into detailing failure modes at the right level of causality.