Getting started with CBM
Condition-Based Maintenance (CBM) is an innovative, data-driven maintenance strategy designed to enhance equipment availability and reduce maintenance costs. We’ve created a simplified 6-step guide meant to help you stand up a condition-based maintenance program.
#1) Select the assets you want to monitor
As is the case with the preventive maintenance plan, you should concentrate on assets that are:
- Important to production/organizational success
- Costly to repair and replace
- Not going to be replaced any time soon
#2) Identify all known and probable failure modes
The best way to do this is by performing an RCM analysis and focus on the failure modes that can be managed using CBM strategies.
#3) Select the right CBM solutions and monitoring techniques
We had a whole section about this so we are going to skip details here. In short, you need to select the right solutions for the failure modes you identified in the previous step.
#4) Define baseline limits for chosen CBM solutions
You need to define acceptable condition limits so that the system can warn you when monitored equipment is starting to deteriorate. These limits have to be set in a way that you have enough time to perform corrective actions.
#5) Establish the CBM program
Running any maintenance program requires you to define tasks and responsibilities and assign them to your maintenance team. Collecting and recording measurements should be in the center of your plan.
#6) Analyze the data and act accordingly
Analyze the data coming in from sensors and inspections to plot a trend and schedule maintenance work accordingly.
When and how to apply CBM
Though condition monitoring can play a beneficial role in your maintenance program, it’s often implemented in a rushed manner without a full understanding of some factors that will affect its effectiveness in the overall system.
This section looks at some key preliminary points to consider for a strategic, rather than haphazard, adoption of CBM.
Preliminary questions for choosing a condition monitoring solution
With the wide range of condition monitoring solutions available, it’s important to first understand your needs, then the capabilities and limitations of these solutions, before making a final decision of which one you want to use.
Consider the following:
1) Does the solution monitor specific failure modes?
A failure mode is a specific cause of the failure or one of the possible ways in which a system can fail. The more complex equipment, the more failure modes it can have.
Understanding these failure modes and their impact will help you identify and adopt the right condition monitoring solution which is an important aspect of improving asset reliability.
Some methods analyze the causes of failures and help one understand their frequency and impact. One of these methods is the Failure Modes and Effects and Criticality Analysis (FMECA). For each asset, the failure modes and effects on the entire system are calculated and recorded before generating the full FMECA.
This step is part of the initial condition monitoring process. Thereafter, the appropriate maintenance tasks for each identified failure mode can be determined.
2) Does the technology interface with existing ERP/CMMS systems?
With the continuous advancements in the way we use data, it is increasingly necessary that all your data channels can interface effectively with minimal or zero disruptions to operations.
Such channels would normally include software like Enterprise Resource Planning (ERP) and Computerized Maintenance Management System (CMMS). Some organizations may also have incorporated tools from the Internet of Things (IoT) such as wireless technology.
Whatever the case, the data stream from the monitoring sensors should work with the other systems already in place if you want to squeeze the most value out of condition-based maintenance.
3) What is the ease of implementation?
Some condition monitoring sensors and solutions are easier to implement than others. You need to understand and completely scope the resources needed to implement a specific condition monitoring solution.
Consider the time, labor, and effort required for implementation because it can be significant depending on:
- How large your system is
- How complex are the machines you’re using
- How easy is it to install needed sensors
- How fast can your team adopt the new solution
Limble offers a Modular IoT sensor kit that streamlines the whole implementation process. Now any business can start a predictive maintenance program without breaking the bank!
4) Can alerts be generated accurately and improve over time?
Before applying a particular solution, consider how reliable the generated alerts will be.
False positives are not uncommon in condition monitoring settings and they are one reason why an organization may decide to abandon the process after implementation. If technicians are continuously called on to respond to “failures” that end up being false several times, eventually everyone will lose confidence in the system.
Therefore, check the record of the ratio of true versus false positives from your proposed solution provider. The aim is to confirm that even though these false positives may occur, they will reduce with time.
Understanding the P-F interval and the P-F curve
If maintenance personnel can find some signs or evidence that a piece of equipment is about to fail, they can take steps to prevent that failure incident. Fortunately, machinery generally exhibits warnings before they fail. This is where the P-F interval comes into play.
The P-F interval and P-F curve are a method for describing how to detect the failure.
The P-F Curve illustrates the behavior of a machine as it approaches functional failure.
As failure begins to manifest, the curve shows that the equipment gradually deteriorates to the point where it can be detected (P), i.e. the point of potential failure. However, if failure is not noticed and mitigated, the situation continues until the system fails. This point is called a functional or hard failure (F).
The time that elapses between points P and F are known as the P-F interval. The P-F interval represents the best time for detecting and arresting imminent failure.
In condition monitoring, the interval between P and F must be enough for analysis to be effective and corrective action taken as this whole process would be pointless otherwise.
Think of it this way. Finding out that something will fail a couple of days before it happens is useless information if that doesn’t give you enough time to apply corrective measures.
Ideally, the P-F interval gives the maintenance team enough leverage to plan the right maintenance task on time to prevent functional failures. However, determining the interval between inspections can be challenging. The good news is that there are a lot of different condition monitoring techniques (we described earlier) you can use to spot potential problems on time.
A general rule-of-thumb is to set the intervals for inspection somewhere midway of the P-F interval. Or at least, the monitoring time must be shorter than the span of the P-F interval. Again, using multiple monitoring points within the P-F interval is even better and more advisable.