cause or causes will eliminate the effect.
What is root cause analysis
Root cause analysis is a logical sequence of steps that leads the investigator through the process of isolating the facts or the contributing factor surrounding an event or failure. Once the problem has been fully defined, the analysis systematically determines the best course of action that will resolve the event and assure that it is not repeated. A contributing factor is a condition that influences the effect by increasing the probability of occurrence, hastening the effect, and increasing the seriousness of the consequences. But a contributing factor will not cause the event. For example, a lack of routine inspections prevents an operator from seeing a hydraulic line leak, which, undetected, led to a more serious failure in the hydraulic system. Lack of inspection didn’t cause the effect, but it certainly accelerated the impact.
There is a distinction between failure analysis, root cause failure analysis and root cause analisis.
Root Cause Analysis: Implies the conducting of a full‐blown analysis that identifies the Physical, Human and Latent Root Causes of HOW any undesirable event occurred. The word “Failure” has been removed to broaden the definition to include such non‐mechanical events like safety incidents, quality defects, customer complaints, administrative problems (i.e. – delayed shutdowns) and the similar events.
Failure Analysis: Stopping an analysis at the Physical Root Causes. This is typically where most people stop, what they call their “Failure Analysis”. The Physical Root is at a tangible level, usually a component level. We find that it has failed and we simply replace it. I call it a “parts changer” level because we did not learn HOW the “part failed.”
Root Cause Failure Analysis: Indicates conducting a comprehensive analysis down to all of the root causes (physical, human and latent), but connotes analysis on mechanical items only. I have found that the word “Failure” has a mechanical connotation to most people. Root Cause Analysis is applicable to much more than just mechanical situations. It is an attempt on our part to change the prevailing paradigm about Root Cause and its applicability.
RCA can be done reactively (after the failure – RCFA) or proactively (RCA). Many organizations miss opportunities to further understand when and why things go well. Was it the project team involved? The change management methodology applied during implementation? The vendor used or the equipment selected? I would argue that performing RCA on successes is just as, if not more, important for overall success than performing RCFAs on failures
The objectives for conducting a RCA are to analyze problems or events to identify:
What occurred
How it occurred
Why it occurred
Actions for averting reoccurrence that can be developed and implemented
The root cause analysis process – RCA has five identifiable steps.
1 Define the problem
2 Collect data
3 Identify possible causal factors
4 Identify the root cause
5 Recommend and implement solution
Define the problem
One of the important steps in root cause failure analysis (RCFA) is to define a problem. Effective and event descriptions are helpful to ensure the execution of appropriate root cause analyses. The first step to define the problem is by asking the four questions:
What is the problem?
When did it happen?
Where did it happen? and
How did it impact the goals?
The investigator or the RCA analyst seldom present when an incident or failure occurs. Therefore, the first information report or FIR is the initial notification that an incident or failure has taken place. In most cases, the communication will not contain a complete description of the problem. Rather, it will be a very brief description of the perceived symptoms observed by the person reporting the problem.
It involves failure reporting regarding incident which includes details of failure time, place, nature of failure, and failure impacts on organization.
Consider a problem on a centrifugal pump AC Motor. A typical problem report could state “pump ABC motor has a problem”. Even though this type of problem reporting could be worse, for example, “fan is bad” or “shrill noise from one of the pumps.” “Pump ABC Motor has a problem” it is still not a very good definition.
A better definition may be “AC Motor of pump ABC” is hot. Can we do better with some basic Root Cause Analysis steps? Sure! Let’s ask the traditional, WHAT, WHERE, WHEN, EXTENT. The problem is:
What: AC Motor of pump ABC (already answered) Where: Motor is hot close to the front (belt drive side) When: Don’t know exactly, but 7 days ago a 138 F reading was recorded (normal) Extent: Front of motor is running 210 F.
The above definition is usually enough to get a problem started. Is it ideal? Perhaps not, but it’s pretty good for a problem statement. This level of problem reporting for craftspeople and operators would be a huge improvement for most plants in improving day‐to‐day Root Cause Analysis.
Collection of data
Data collection is the second and important phase of RCA process. Acquiring, gathering, or collecting the failure data regarding the incident are a key for getting the valuable results of RCA investigation. Comprehensive and relevant failure data are crucial to identify and understand the root causes of a failure accurately. Unavailability of correct, adequate, and sufficient data can lead to undesired results of RCA.
It is important to collect data immediately after occurrence of failure for accurate information and evidence collection before the data is lost. The information that should be collected consists of personnel involved; conditions before, during, and after the event; environmental factors; and other information required for root cause analysis process.
Every effort should be made to preserve physical evidence such as failed components, ruptured gaskets, burned leads, blown fuses, spilled fluids, partially completed work orders, and procedures. Event participants and other knowledgeable individuals should be identified. All work orders and procedures must be preserved and effort should be made to preserve physical evidence such as failed components and ruptured gaskets. After the data associated with the event have been collected, the data should be verified to ensure accuracy.
Data for any failure could include the previous failure reports, maintenance, and operations data, process data, drawings, design, physical evidences, failed part of equipment and any other necessary information related to the particular failure. It is not necessary that every failure required comprehensive data but sometimes data could be missing and gathered data is not sufficient to identify actual causes of the failure. So it is necessary that collected data must be accurate and relevant. Failure can’t be investigated properly without availability of correct and related data. Usually, data collection consumes more time as compare to other steps of RCA process so data must be precise and meaningful
