Root cause analysis (RCA) is a vital tool utilized daily in the realm of reliability. Properly understanding problem-solving methods and the analytical tools that function within it are key in aiding your site’s journey to world-class maintenance.
Identifying the root of the problem can lead to minimized downtime, defect elimination and an optimized workforce. The objective of this session is to distinguish, review and comprehend multiple methods to approach root cause analysis problems.
RCA is a continuous improvement tool that utilizes multiple methods or avenues to aid in the discovery of a problem or unexpected event such as inferior quality, inhibited performance or increased concern for safety through the process of scrutinizing and eliminating symptoms of a problem to identify the underlying, bottom-line issue.
While RCA often starts out reactive in nature due to knee-jerk reactions based off plant failures, over time this process, if nurtured, can move to drive results on a proactive as well as reactive state of equipment reliability at a facility.
RCA works to improve the overall quality of a product or process and as such merits a certain organized approach to be effective and drive real results. There are several iterations to this methodology, but the key steps to consider are:
State and clarify the specific problem or concern,
Collect data and information related to the problem,
Categorize likely and potential causes,
Identify what causes need to be removed to eliminate recurrence,
Detect effective, potential solutions,
Establish a plan for implementation, and
Review and document changes made to ensure the issue has been properly corrected.
It is worthwhile to note that while these actions and analyses can be completed by an individual, RCA works best when a group of peers with varying levels of experience, job duties and unique perspectives are utilized.
Now that a baseline for understanding what RCA is has been established and general processes have been outlined, an examination of common methods applied during this development can take place. It is imperative to note that each of these methods have a varying level of complexity, and the nature in which each technique is deployed can greatly depend on the active involvement in the site’s continuous improvement initiatives.
Some methods are quick processes, while others can require a long and arduous journey. Certain methods are utilized proactively during installation and while equipment is in service. Others rely greatly on the process of reviewing a noted postmortem failure or problem with the asset.
There are several names for the first method being covered. Some refer to this exercise as a fishbone diagram, Ishikawa or maybe as a cause-and-effect diagram. The idea behind this RCA method is to draw the failure or problem on a box with a long, straight line leading from it. Off this long line there will be several branches or bones that divide off the primary line.
Each line will have a header and act as a specific category failure mechanism in identifying possible reasons for the root cause problem. Common category headers are people, procedures, practices, equipment, materials, measurements, environment, and so on.
As an example, under the people category, the RCA team would identify all notable individuals or positions that would have had a possible impact on the outcome of the problem. The team would continually carry out this process for all reasonable categories and work to identify a common theme of realistic root causes based on one or more categories. This exercise is one of the most commonly used RCA methods available and is used as a driving force to identify root causes after failures or problems occur.
Barrier analysis can be a reactive as well as a proactive RCA technique used to discover why the problem occurred and to eliminate it from recurring. The idea behind this methodology is that the problem can and should be avoided by having proper barriers established. The RCA team works to recognize the target and hazard, and then tries to identify a means to create barricades between the two.
In some cases, more than one barrier is utilized. This case is referred to as layers of protection or depth in defense. These barriers, depending on the problem, can be engineering or administrative controls. Barrier analysis is most commonly used in safety-based incident evaluations.
Failure mode effects analysis (FMEA) is primarily a proactive RCA tool that is used to determine potential failures, risks and causes. The exercise has qualitative traits but is a more practical application of quantitative traits based on calculated matrices and is often considered the first step in proactive RCA.
FMEA is executed by the RCA team investigating equipment prior to problems occurring and understanding the related concerns based on a weighted ranking system. Plans and strategies for solutions deriving from the FMEA will mainly be prioritized based off the methodical results. FMEA should typically occur during any installation, design change or increase in production numbers.
Fault tree analysis is a method similar to FMEA in that is often considered to be a proactive approach to RCA. This technique establishes a waterfall or top-down logical analysis of causes shaping event rates and leading to specific root solutions. This methodology is generally used in process concerns or safety-related incidents.
Eight Disciplines is an RCA approach developed to correct recurring defects or problems. This method works to establish long-term corrective actions to eliminate the concern based on statistical analysis. This practice includes creating a team, describing the problem, developing short-term containment plans, determining and confirming the root cause, verifying long-term resolution, executing corrective actions, preventing recurrence, and recognizing success. Eight Disciplines was initially applied in the automotive industry, but now has many practical applications within RCA.
The “Five Whys” operational approach to RCA in its basic form can be a very simplistic method to identifying issues. The general process of how this approach works is to identify the concern and ask why this failure occurred five times. The number of whys is set at five as somewhat of a standard, but this process can be as simple as one why or as complex as more than 100 whys to the concern or failure occurred.
Just because the “why” process is small or large in number does not make the process more or less true. There can also be multiple failure mechanisms or concerns based on the category you are reviewing. The same failure maybe be correctly identified separately as a design issue, an operational issue, or a process or procedural issue. Further detailed why analysis can also work to identify multiple categories in the same failure mechanism.
There are a multitude of variations of why analysis, but it is up to the RCA team to decide how deep to dive in this linear evolution. The why analysis is typically used directly after a failure or problem has occurred and is often used to kick-start the RCA team on most projects after data collection has occurred and the process and team have been defined. This technique is rarely found on the proactive side of RCA.
The Is/ Is Not Method is another somewhat simplified approach to RCA. The idea is to create a list of who, what, where, when and why rows with “Is” and “Is Not” columns influencing the problem or concern based on all data and information that has been collected. This practice is often utilized after the problem has occurred and the RCA team needs to narrow the scope, establish solutions based on common themes and establish priorities.
Pareto analysis is a great visual presentation of RCA. This method uses data to track and trend root cause concerns. Pareto charts typically appear in bar graphs that display values in descending order totaling the whole charge of the problem. As Pareto is commonly referred to as the 80/20 rule, it often lives up to its name in regard to identifying failures.
Once these critical few items are addressed though Pareto analysis, the majority of cause for concern will be eliminated. This technique is most commonly applied with quality-control concerns and generally conducted while the failures are occurring.
Much like Pareto charts function well in other areas of industry, the same can be said for scatter plot diagrams. These diagrams can be a useful RCA tool when working to correlate one or a multitude of paired variables. Variables are separated on an X and Y axis of a chart to establish if the noted variables are related. This particular method is typically used in conjunction with other methods as a means of validation.
In conclusion, identifying the root of the problem, understanding the RCA process, and knowing how and when to deploy certain analysis methods can lead to minimized downtime, defect elimination and an optimized workforce. Properly deploying these practices proactively and reactively as well as driving the discoveries to fruition are key in aiding your site’s journey to world-class maintenance.