Dear All,

Can someone enlighten me if these two things are the same or they have some differences for communicating purposes

Predictive Maintenance and Condition-Based Maintenance

If these two are different which one is correct that PdM is just a part of CBM or CBM is a part of PdM ? Which do you think an organization should focus more PdM or CBM ?+

My Warm Regards,

PdM = CBM but predictive maintenance is subset of condition based MONITORING which includes monitoring operating performance, not just machine conditions.

Rolly,

I don't see PdM being the same as/equal to CBM. CBM is the next phase in Maintenance Strategy in a PM driven environment, to optimize PM particularly when the OEM warranty has expired on equipment. The basic application of CBM is logging/monitoring of operating parameters, to evaluate system performance. Advanced program involves Vibration, Lube oil, IR Thermography, Ultrasonics, etc. data collection and analysis; which is Predictive Maintenance. Although above techniques are in a way CBM (expensive), the objective would be to engage in prediction of degraded system condition/performance, so that an opportune window could be scheduled for remedial work; hence, it becomes PdM.
An organization will not be able to engage in PdM without first employing CBM, as part of Mtce Strategy progression.
My two bits worth.

Cheers...Rajan Muthukrishnan

Example:
PM = Replace the prefilter every 3 months.

Question:
"Replace the prefilter if pressure drop is over X value."
Is this CBM or PdM?

Eugene,

I can see where your question is aimed at and my response is...CBM, simply because the replacement call is based on condition of pressure drop/increase at a certain value. On the other hand, if the drop/increase in pressure is used to determine P-F timeframe and pinpoint failure modes, I'd call it PdM.
What say you, would that argument hold water??

Regards...Rajan Muthukrishnan

My take on it:

PdM is a series of tests taken at periodic intervals in order to estimate when the equipment/system/component is going to fail.

CBM involves a process for selecting the optimal inspections, tests or other practices, including run-to-failure, re-engineering, etc. designed to perform maintenance based upon the condition of the system. This involves the use of such strategies as RCM to select the proper maintenance strategies to support the CBM process.

In effect, PdM is an activity, CBM is a philosophy.

Howard

MotorDoc

I disagree with your take on PdM as a series of tests taken at periodic intervals in order to estimate when the equipment/system/component is going to fail.

I have never told anyone in a report I have issued when something is going to fail. I am not a member of a Predictive Failure Group and never have been.

We recommend maintenance action well in advance of that failure stage. We don't want that seal to start leaking from misalignment, or the shaft to wear because of a bad bearing.

Just my two cents.

Mick:

You are right, my terminology is incorrect.

I meant to detect a potential defect, such as a worn bearing. However, the concept is still to make the call and provide a recommendation on a repair prior to additional damage or failure.

As far as the term 'Predictive Maintenance' are we now suggesting that we can 'predict' when something is going to fail? What happens if we say that corrective action must be taken on the bearing within the next 30 to 60 days, those days pass and nothing happens?

Howard

Dear All,

This is what I know and the term predictive is used so the maintenance tasks can rhyme.

Reactive, Preventive, Predictive, Proactive, Detective (Added by John Moubray)

If we speak about Predictive Maintenance then we speak about the use of these non-destructive instruments to predict potential failures.

If we speak about Condition-Based Maintenance, we speak about the system as a whole, the Predictive Maintenance Organization, the softwares, trending results and analyzing them, monitoring schedules, cost savings generated etc. etc. which made me to conclude that PdM is only a subset of CBM. Hence, if we speak about the instruments being used then we speak about PdM but if we speak about the system as a whole then we speak about CBM.

My Warm Regards,

I think condition based maintenance is the costsaving preventive maintenance. the overhauling of a compressor or turbine is scheduled for an interval to check the bearing clearence, deposits on rotor, labrynth clearence etc. this is preventive maintenance. on the other hand if the performance of the machine and vibration trend indicate the good health condition of the machine then we can extend this intrval as well as save the money.

Predictive maitenance is the way how to find the offcoming failure. Vibration trend analysis and performance calculation are types of the predictive maintenance.

so we can say PDM people advise for CBM.

I started work in power generation in 1964, and the term "diagnostic testing" was used for all the vibration and performance analysis we did to help make major maintenance decisions.

The term "condition monitoring" I think originated in the UK. The first book of that name was in 1977 by the late Ralph Collacott, who founded the Institution of Diagnostics Engineers.

I used the term in the title of my first book from 1988, revised 2001.

The term "predictive maintenance" came along later, but to us who worked in CM, it was essentially the same thing: the effort to detect signs of degradation and try and predict when corrective action was needed.

For my second book, I took the best of both worlds! "Predictive maintenance of pumps using CM".

The recent ISO 13372 2004 Standard in this field has some good definitions.

"Condition monitoring:
detection and collection of information and data that indicate the state of a machine

Predictive maintenance
maintenance emphasizing prediction of failure and taking action based on the condition of the
equipment to prevent failure or degradation

Preventive maintenance
maintenance performed according to a fixed schedule, or according to a prescribed criterion that detects or prevents degradation of a functional structure, system or component, in order to sustain or extend its
useful life."

Ray Beebe

Dear all,

Suppose a critical pump runs with a little high vibration and analysis shows defects on bearing but it can be kept on running for more few months. on the other hand if the pump is shutoff for maintenance there is huge production loss. so it has been kept running and wating for any shutdown. after few days the whole plant has gone for a shutdown for low NG pressure. operation people have to wait for few hrs to resume the operation. the maintenance managemnet has taken the decision to replace the bearing of this critical pump in this shut down period. finally no down time has been required for this maintenance.

I want to know what types of maintenance involved in these activities?

Because you found the the job to replace the bearing using vibration, then it's a CP (corrective/preventive) type of maintenance.

Howard,

Yes, we have a fine line to walk with some people who expect us to know how long it will last. Most of us realize success in educating our customers that is not the best approach. Rather, listen to us when we say repair within 30-60 days to avoid more damage.

Have a great holiday

Mick:

Part of the challenge to 'predictive maintenance' is that work 'predictive.' The responses that I have run into relate to either decision-makers considering us like weathermen, or the literal concept of 'prediction.'

When presenting my work on electrical insulation predictive maintenace to the IEEE Electrical Insulation Conference in 1999 and 2001 based on work throughout the 1990's, I received a lot of resistance. The main reason is that my colleagues considered 'prediction' as an engineering statement of accuracy, such as if you predict something will fail in 30 days, it had better fail in 30 days. This was even though I had been using values of failure resistance and availability in association with the insulation system's chance of survival.

In 2003, I changed the name to Time to Failure Estimation of the insulation system. In this way, the application of the availability/survival of the insulation system took precedence. At this point, it seemed to make more sense and even though the material was virtually the same, it was very well received.

This worked in parallel with work on maintenance and management communication that we had been working on, at the time. We wanted to perform an RCFA on what went wrong with the communication between maintenance and management. We also worked on reviewing the translation of the maintenance and management languages.

The attached article is a draft of a larger article on the subject of Time To Failure Estimation. There are more rigorous methods that we have put together, but this provides the concept.

We also published it in a ReliabilityWeb iPresentation: http://www.rcm-1.com/forms/sbdra_reg.htm

Enjoy.

Howard

Considerations_in_Time_to_Failure_Estimation.pdf (125 Kb, 47 downloads)

Great Post Howard

Srs.
I have started the Predictive Maintenance in my company ,so I would like to know what kind of Key performane indicator I could work on Predictive Maintenance.

I really appreciate your comments.

HI ROUL,

I am working with average level plantwise as well as equipemnt wise indicator.
Avg. level of any highest value in a defined interval help me to track the equipments and gap analysis.

regards

Hi,
I don't think PdM=CBM, because I think CBM is condition based maintenance, but Pdm is condition and performance based maintenance. Pdm should consider more factors not only condition, for example maintenance history.
In fact, there are too many words, including Pdm, RBM, CBM, RCM and PLM etcs.