|
Is
Analysis An Engineering Function?
|
Written by Charles J.
Latino
President & Founder
of Reliability Center, Inc.
|
|
When failures are obviously costly, engineers are usually assigned to find
their causes. Yet, many engineers are not trained in the discipline of
failure analysis. Many are too close to the problem to exercise unbiased
critical judgments. And many do not have the skills to facilitate group
analysis. While this is not an indictment of engineers, it is an indictment
of a paradigm that says that only engineers close to a problem can effectively
conduct a failure analysis. |
|
Existing paradigms in manufacturing facilities
which stereotype people often shut out those who can be most effective
in collecting failure information and finding causes |
|
When performing root cause failure analysis of those failures that represent
80% of the facility losses, it is important, if not imperative, that the
failure analysis team leader has the following characteristics: |
Ability to remain unbiased and reject conventional wisdom.
|
Ability to facilitate a group of people toward a common objective.
|
Training in logic tree approaches to failure analysis.
|
Affinity for listening and questioning for understanding.
|
Patience and perseverance.
|
These characteristics may or may not be available within the in-house engineering
ranks.
|
|
If the failure is of a mechanical nature, this writer believes that a mechanical
engineer should be part of the analysis team. Likewise, an electrical failure
should include an electrical engineer and a process failure should include
a process or chemical engineer as part of the analysis team. These engineering
participants need not head up the analysis. |
|
Another factor to consider is that people close to a failure are supposed
to know the processes involved, have knowledge of the machines and their
nuances as well as an in-depth knowledge of procedures and the management
systems surrounding the failure. For most people it is difficult, under
these circumstances, to ask the probing questions. Psychologically, when
a question is asked it indicates that the person asking the question does
not know the answer. To many people this is too much exposure so they shy
away from the penetrating questions. |
|
It is generally best to provide a trained analyst that is not close to
the process, machine or administrative delay being analyzed. This writer
once assigned an aerospace engineer to analyze a chemical boiler failure,
that had recurred annually over a ten year period, because the engineer
could provide his failure analysis training as well as his ignorance to
the analysis of the problem. The aerospace engineer solved this costly
problem while, many experts in boiler and corrosion technology had studied
the problem but never offered a permanent fix. The point is that ignorance
opens many doors to information because probing questions can be asked
without the fear of tarnishing the questioners self image. |
|
When routine failures occur, mechanics and operators generally make judgments
about causes. It is a natural tendency for people to want to know why
something happened. The problem is that most of these people are
not trained to find causes. When they are trained, they are seldom given
the time to practice their new skill. When given the time to do failure
analysis, they have a really hard time having their analysis and recommendations
taken seriously. |
|
DIAMONDS
IN THE ROUGH
|
|
It is commonly thought that mechanics, particularly, do not want to solve
problems because solutions may endanger their job security. This writer
has not found this to be generally true. Most people have a hard time resisting
the temptation to solve problems particularly if their recommendations
are taken seriously. |
|
Managers fail to utilize the enormous resource
represented by their field people in solving problems by not providing
them with tools to do failure analyses |
|
We intrinsically know that field people know what works and does not work.
Where they see barriers, they use their creativity to skirt them. They
also know what repetitive jobs prevent them from achieving their best work.
These jobs may not be recorded in the maintenance management software systems.
Oftentimes jobs of 1 hour or less are performed on running work orders.
In some plants, most shift work is recorded on routine work orders that
stand for the year recording only time and materials, and not the individual
jobs, for accounting purposes. |
|
The specific work performed and their frequency of occurrence, on both
routine or standing work orders, can in many cases only be identified by
field people. Additionally, field people provide insights into why people
do the strange things we encounter. For instance, they know that shutting
down some equipment when they are in danger of failure is not done because
of the perceived chagrin of their supervisors for losing production. Analyst
should know these perceptions exist. They should know that since specific
field people are held in high esteem by their peers others will often defer
to them when off standard conditions call for action. Sometimes the course
chosen is wrong. The point is that failure analysts need to be aware that
this type of peer deference exists. |
|
The field people, in this writer's experience, desperately want to chip
in and feel the recognition and fulfillment of their contributions. This
desire usually overwhelms their long term desire for job security. Management
must use these contributions to grow their facilities and enhance job security.
The potential is available, in most cases, for quantum leaps in productivity
and profits that can be used for growth. |
|
To fully utilize field people, training in failure analysis is needed.
The training will provide them with skills to do their own analyses. It
will enable them to remove barriers to performing with excellence and to
contribute to the analyses of others. The field people represent an untapped
army for eliminating obstacles to world class performance. |
|
INCOMPLETE
FAILURE ANALYSIS
|
|
What most plant management people want to know is why did we lose the compressor?
Why did our qualities deteriorate? When presented with the conclusions,
they show that they are pleased with the outcome. For example, if a rotor
failed prematurely due to fatigue, we can redesign it to provide longer
life in the future or we can have a spare available for the next occurrence.
If an operator did not follow procedures and failure occurred, we can discipline
the operator and send him or her for specific training to avoid ignoring
or misinterpreting procedural steps in the future. |
|
Accepting physical causes and improper human
interventions as a complete analysis does not reveal the root causes of
failures |
|
It is no wonder that failure rates do not go down in many plants even though
the plants feel they are doing a credible job of failure analysis. As analysis
proceeds from the top of the logic tree down through the physical and human
causes, we are asking "how can" the previous condition occur. This allows
us to deductively pursue the specific mechanical cause and the specific
human error, mistake or slip that occurred. If we utilize this information,
we should be able to prevent that event from recurring. BUT THIS IS NOT
ENOUGH. |
|
At the point of poor human intervention we need to change our question
and ask "why" did the error, mistake or slip occur. The answer to this
question is almost always inductive as it provides information that not
only applies to the failure being studied but to a multiplicity of potential
failures. For example, if we find that an operator did not follow a procedure,
instead of concluding the analysis, we have to ask "why" did he or she
not follow the procedure. The answer may be that the procedure was too
complicated or that the procedure was not understandable. Examination of
procedures usually reveals that the problem is universal and that correcting
the procedure fault avoids a myriad of potential failures. |
|
THE
TEST OF GOOD MANAGEMENT
|
|
Picture the scenario that often occurs when true root cause failure analyses
are being performed. Upon hearing that a management system is absent or
inadequate, the manager will become angry and sometimes even abusive. Often
times this behavior is a result of interpreting the management system deficit
as a personal reflection on their ability to manage. The result is that
the heart of what is wrong in a facility is not uncovered. |
|
Root cause failure analysis should not be
discouraged because of political considerations |
|
When the analyst goes beyond the point of wrongful human intervention,
he or she is getting into areas of inappropriate, flawed or absent management
systems. It is sometimes painful for managers to realize that they may
not have adequately covered the needs of their organizations. If managers
can get past their own internal security needs, the benefits of the process
will far outweigh the pain. Both the performance of their plants and the
personal recognition they will receive for that performance should provide
adequate reasons to continue. |
|
We are on the threshold of a new era of manufacturing performance. Indeed,
it is being forged as this paper is being written. The new era is certainly
a lot more competitive than before but, it is also more exciting. We have
in our grasp the knowledge and the foresight to be one of the prosperous
survivors. |
RCI Offers the full
range of Reliability Consulting Services and Training Programs for Industry.
We conduct facilitations, reliability assessments, FMEA & Root Cause
Failure Analysis Training - Public & On-Site.
For more information
contact:
Reliability Center, Inc.
P.O. Box 1421
Hopewell, Virginia 23860
Phone: (804) 458-0645
Fax: (804) 452-2119
Website: http://www.reliability.com
|
|
Return
to Failure Analysis Reference Library Index
|
