We are going to a 24/7 production work week in a large manufacturing facility. We have a staff of PdM techs, Area Lube techs and an on site oil analysis lab. Do you have any suggestions to implement a meaningful PM program? We have been doing 360* walk around PM's for several years but they are not as effective as a thorough hands on, look behind covers and guarding type of PM. Since our machine down-time will be practiacally non-existent we are struggling on how to approach this task. We have very few redundant processes so this is another problem we face as well. Looking forward to reading some input from others who have tackled this obstacle successfully. TIA

It sounds as though you have some history as to what method of PM work provides value to your organization.

Assumption on my part:
1) Historically, by opening covers and looking inside the maintenance group has found failing components or parts that given more time could have gone catastrophic. – These could be considered your "Good" (value-added) inspections.

Question:
Do you have historical data that contains:
1) The specific failures found
2) When the failures were found
3) It might also be nice to know how important the findings were (what catastrophic event was prevented by doing the inspection)

Your history will hopefully support your need to perform the "Good" Inspections – which may require production outages to prevent even more severe situations.

If your history supports the need to perform the good inspections, perhaps it will also indicate the need to review your sites maintenance strategies following reliability and/or risk based engineering methods.

I.e. if your history clearly defends shutting down production to perform critical inspections, then perhaps the opportunity of gaining back some maintenance down time will gain management support to further analyze the maintenance strategies (free up some resources). Once resources exist than follow the recipes outlined in postings on this Forum – RCM/Spare parts/redesign, etc.

Who knows, there may be a failure out there waiting to happen that is not even being inspected for and this situation requires as much or more of a remedy as optimizing already existing maintenance programs.

I have only worked at 24/7 production facilities. I hope the change in production at your plant is taking into account balancing equipment maintenance needs (Theory of Constraints balanced with long-term equipment viability).

One thing I have noted at 24/7 operations is that the cost of maintenance at night is more expensive for labour, etc. If the product made at night or day is sold for the same amount then have the planned outages during regular work hours. I assume this to be a difference with facilities that run less than 24 hours as I assume most maintenance would be completed during the shutdown times (whenever that may be).

Hi, don't underestimate the importance of your walk arounds. Very often something that can seem unproductive or ineffective is actually more effective than you think.
It's often the way we are doing those things that need to be addressed.
I am a CM tech - or was - until recently and did a lot of walkarounds as part of my task.

In the process I picked up things that either operators / production personnel did not spot or if they did - did nothing about them either because:
1.They did not know they were important - lack of knowlege / understanding.
2. Did not want to rock the boat -the risk of looking foolish
3. Did not care - attitude.

Also I found getting the wider picture was useful for me in that in Vib or oil analysis we are so focused micro and can often miss the macros that tie 2 and 2 together.

Being a TPM advocate I would also suggest you make the inspections easier with a little visual control.
By visual control doing things like marking guages, making sure you have indicators such as temp sensitive tape, windows in guards, oil level indicators etc, all contribute to speeding up a walk around route - esp, on a large site.

I hope this is of some help - and I suggest if you are unfamiliar with some of the TPM methods for autonomous inspection seek out some resource - that is if you haven't already.
Shucks for all I know you may be TPM gurus.
Perhaps I should stop there!

All the best.
Mike.

Manufacturing facilities generally need set-up changes, retooling etc, so called product or production related downtimes.
Aligning PM wprk with these downtimes is one way of scheduling PMs. Secondly, if you record downtime due to breakdowns, you can justify some downtime to do PM work, if that will eliminate or minimize breakdowns.
If your failures are predominantly due to wear, fatigue or corrosion, PM is well justified. If they are of the 'no fault found' variety, it is hard to justify PM.
Overall downtime (PM +BM) is what matters, not just PM.

The answer to your question all depends on what is needed for your facility. The rather small seemingly easy question demands a rather gargantuan answer that cannot be completely answered here. We can only offer ideas. Just make sure that your program is in line with the business goals of your company. If it goes outside of the business goals, you will not get buy-in and the maintenance people will be the first to realize this.

Regarding your PM program, it can be developed through the principles outlined with the RCM process. The first steps with RCM is to target your equipment's needs based on a number of factors. Criticality of the system, equipment and components to the mill/plant and to its respective process is absolutely necessary in order to ensure you are headed in the right direction. In other words a PM program that will be setup to succeed.

Remember, a good PM program will:

target those pieces of equipment that require the attention that PM program requires (from the criticality assessment).

target the specific tasks that will either predict, prevent or detect equipment/system conditions that can be adjusted to be proactive in preventing failure.

target the failure modes of the equipment. In other words, how do you know the equipment is failing to provide its intended purpose. A pump is designed to provide either a specific flow or pressure. You know it is failing when it goes outside of the range that is required to keep the system reliable. So, you would test a pumps pressure or flow to determine if it is failing. Not just a bearing defect. The lack of pressure or flow would mean either the impeller is worn, damaged or you could then check the level in the tank... etc.

target the specific ways to correct the failure mode, to prevent failure. Whether a minor adjustment on the equipment or replacement of a small component requiring minimial downtime (if possible). Schedule work if the task requires downtime in excess of what is allowed.

target getting the PM Techs (inspectors, etc.) in gathering data using handheld technology for routing, and recording of data. An excellent way of gauging whether your program is successful is by using handhelds. Success in carrying out the program, and success in the effectiveness of the program. Reports can be run to indicate exceptions, and then communicated to those that need the information.

target the review of the PM itself. I like to call it Dynamic PM's. Review the effectiveness of the PM over a period of time, and the MTBF, MTBR on that piece of equipment. Is the PM extending those values, or not having any effect. If not, or not enough, review PM for proper items to focus on to extend those values.

Again, this is not a complete list of items to target with a PM program. The real trick to it is getting a focus by your maintenance/reliability group and to carry the plan out to fruition. I have seen many PM programs just being a piece of paper, and either "done" or "not-done." What does "done" mean? That is where the handhelds would help. Even thinking about it, what does "not done" mean?

Remember, ultimately the idea is to capture the work history that has been done on the piece of equipment, either on an Operator Basic Care route, PM Route, Mechanic, E&I Tech, etc... All have done work, but if we don't know what was done by what trade, we don't know how effective our program is.

James Fajcz, P.E., CMRP
Reliability Engineer

James Fajcz, P.E., CMRP
Reliability Engineer

James,
I am not sure you are addressing the original question, which is how to manage a PM program within a 24/7 scenario, not the content of a good PM program, which I think your answer fits well.

quote:

Originally posted by Vee:
Manufacturing facilities generally need set-up changes, retooling etc, so called product or production related downtimes.
Aligning PM work with these downtimes is one way of scheduling PMs.

That's the way to do it. Negotiate with production planning and production supervisors when those time windows will ocurr by each production area/line, and then bring the whole maintenance crew to execute all the PM orders for that area/line. Then move on to the next time window and area.
It will be great to do not have more than one production area/line down at the same time so to avoid maintenance crew capacity conflicts.

quote:

Originally posted by Vee:
James,
I am not sure you are addressing the original question, which is how to manage a PM program within a 24/7 scenario, not the content of a good PM program, which I think your answer fits well.

Vee:

You are correct. I didn't answer his original question. The reason I did not answer that is due to the fact that I don't know his PM program. Every PM program is different, and in each company, facility, maintenance department, mechanic there are vast differences between knowledge, worker skill, abilities, training, process knowledge, technical skills, management styles, and management's expectations of the mainteance department. Etcetera, etcetera, etcetera. Ad nauseum.

Typically what I have seen is that Management is not interested in the PM program itself. They are only interested in the results of what a PM program brings. Therefore, the PM program must fit the Management goals, objectives and the style of the organization, and must be managed accordingly by those goals. RCM is a great tool to assist in finding out those facility goals.

In a forum such as this, we can only assume what his organization is like, but there are some basics of a PM program which I did outline only a few parts of one in my earlier posting. Other than that, unless igolfat8 responds to these postings to correct our tangents we get on, we can suggest pretty much anything, so it seems.

Thanks for your input.

James,
You are quite right, of course. Again I am not sure why you need to know the content of the PM program to answer a question on how to schedule it.

quote:

Typically what I have seen is that Management is not interested in the PM program itself

Should they be? They pay a Maintenance Manager to produce and execute a maintenance program that delivers business results. Whether it is RCM based or contains PMs, PdMs, CMs etc. are strictly not their concern. Are you assuming that igolfat8 is the top manager? If so, I agree completely with your comment.
If now you assume they did analyze their maintenance requirements using RCM and have the right program in place, the original question that igolfat8 raised still needs to be addressed. When should they do it if they run 24/7? I think Darth has summed it up nicely.

igolfat8, can you please sign your name, nickname or calling name as a courtesy to other readers? Thanks.

quote:

Originally posted by Vee:
James,
You are quite right, of course. Again I am not sure why you need to know the content of the PM program to answer a question on how to schedule it.

Vee, I appreciate your ultimate direction. Your question above cannot be answered in a box - or without knowing more information about his mill/plant/process. Scheduling PM's should be based on the business goals, equipment's Weibull life expectancy using PdM, history of PM's, and MTBF, MTTF, etc. Somthing not easily answered without more info.

Keep in mind that his original question was:

quote:

Originally posted by igolfat8:
Do you have any suggestions to implement a meaningful PM program?

Again, if I did not answer his question, then I would like igolfat8 to let us know. If I didn't answer yours, I apologize.

quote:

Scheduling PM's should be based on the business goals, equipment's Weibull life expectancy using PdM, history of PM's, and MTBF, MTTF, etc.

I do not exactly understand that statement. After a PM has been defined, I mean for this:
* tasks has been written down after RCM of other proper analysis,
* frequency for each PM task have been established,
* time to complete the PM order is estimated,
* spare parts, special tools, etc. are planned...

then scheduling the PM orders is reduced to obtain an agreement between the equipment owner and Maintenance on the day and time to start the job.
If an equipment has a quarterly PM, and this PM will be due on February, May, August, and November of each year; I see scheduling as obtaining the agreement (hopefully during January) with Operations on when between February 1 to February 28 Maintenance will be allowed to touch the equipment.

quote:

I see scheduling as obtaining the agreement (hopefully

Eugene we call that custody transfer, can be a hell of a job, in other words: the equipments is in Maintenance hands on a determined moment, and the owner has to wait until it is finished...

James,
At the risk of being accused of splitting hairs, let me examine your observation, as Darth and Steven have also done.

quote:

Scheduling PM's should be based on the business goals, equipment's Weibull life expectancy using PdM, history of PM's, and MTBF, MTTF, etc. Somthing not easily answered without more info.

You are absolutely right when you say "Scheduling PM's should be based on the business goals".
I dont agree with the rest of the statement though, for the following reasons.
1. Deciding what work to do, its approximate timing and the steps involved is a planning function.
2. MTBF, History, MTTF etc help in deciding the plan, i,e Pdm or PM content.
3. Since Weibull parameters, MTTF etc. are obtained by statistical analysis, they help estimate probability of failure before a defined time period, not time to failure. Using Pdm, which is based on physical degradation of a failure that has already commenced, the estimate of time to functional failure is much more robust. Still these numbers help us PLAN work, e.g., decide frequency of PdM readings or approcximate timing of PM work. Thus if a PdM reading which was planned at some frequency predicts a functional failure time, a CM can be initiated. Here again, the timing of the CM is a scheduling activity that depends on the business consequences of the failure.
4. Scheduling is an economic decision of when EXACTLY to start and to ensure logistics/resources are in place to finish work in the shortest possible time.

Darth,

quote:

scheduling the PM orders is reduced to obtain an agreement between the equipment owner and Maintenance on the day and time to start the job.

Bingo!

Thank you all for your responses thus far. I apologize for not chiming in sooner but "fire fighting" has taken precedence. As much as we don't like to admit it, fighting fires still exists.

A bit of brief background: We are a fortune 100 company and have not gotten there by a slipshod Maintenance approach. I am a reliability Engineer and not the Maintenance Manager. Our individual manufacturing plant (one of dozens in the US and many more world wide) employs 3500+ workers.

We have an active TPM program, Opex, limited RCM, in house PdM staff (with all the toys), in house WPA lab collecting & analyzing monthly oil samples on 5000+ machines, semi skilled lube techs on all three shifts, state certified skilled multi-craft Maintenance and separate Tooling departments.

Our current PM program is multi tiered consisting of: Basic operator visual PM and lube tasks [rotated weekly across three shifts], monthly skilled maintenance 360* walk around "run-time" PMs, monthly PdM routes (vibration both manual and on-line routes, ultrasonic, thermography, 23 miles of overhead chain conveyor stretch monitoring, laser / dial indicator alignment and bi-annual MCE / MCA testing. The final and most effective tier is a 10 hour, full time, PM Crew. We schedule a machine down for an 8 hour window for an in-depth, hands on PM. The first hour is spent analyzing machine history data from our CMMS for clues followed by an 8 hour [look behind covers and touch every nut and bolt] type PM gathering and recording measurements for trending. The last hour is spent generating a detailed report with pictures of items repaired and open issues that need follow up planning and scheduling.

The challenge we face is not PdM, RCM or implementation of other TPM tools but finding a window to have access for "hands on" PM. The PM Crew has proven to be the most effective tool in preventing failures because they can access areas that our PdM tools can not. PdM could but it would be a huge investment in remote accelerometers and sensors. Since machines will be running 24/7 there will not be opportunities to schedule them out of production for an 8 hour window.

We have considered brief PMs during change overs but since we use "quick die change" techniques we only have minutes of downtime which doesn't provide an opportunity to go behind the scenes for hands on PM.

Maybe I am answering my own question? If I don't have time to do hands on PM then it will require an additional investment into remote sensing hardware and an increase in PdM staff. Duh!

- I noticed most of you are proponents of RCM. I am not and here is why: it requires a cross functional team (which is difficult to dedicate for an extended period), it takes waaaaaaay to long for the process (typically 5 days) and we have difficulty with hourly buy in. I have studied it's success and failures at some of our other plants and it has not been very well received. I realize the outcome is dependant on the program leadership but we have used some of the "best" gurus in the industry to train the trainers. Yes, it has some very good points and tools but someone needs to devise a streamlined approach to the process. Just my $.02 ... end of ranting.

Do Toyota plants use remote sensing?

Every machine must come down once, or it must be build to work forever.
The satelites are an example of that, once it is up there, it is difficult to do tighten a nut on one of the shields

but would you buy a pump or press at the price of a satelite?

Even in the oil refineries, plants suffer turnarounds, and then everything goes down for mayor maintenance/inspection, but that does not mean that there is nothing done on maintenance between shutdowns.

The basis, the owner must agree to hand over the equipment, else he can do it himselve, because maintenance has enough work to do, and maintenance is not willing to work overtime on friday night and weekend because the equipment in question will suffer an emergency breakdown on the last working day, 5 minutes before closing time.
Murphies law: the emergency breakdowns are always on Friday

quote:

can be a hell of a job,

Yes, it IS, even though we do not run 24/7 ussually.

igolfat8,
You are not alone with your problems - you are probably in category where 90% of manufacturing companies live. Let me snip some of your post and then provide some diagnosis and suggestions

quote:

"fire fighting" has taken precedence

quote:

We have an active TPM program, Opex, limited RCM,

quote:

We schedule a machine down for an 8 hour window for an in-depth, hands on PM. The first hour is spent analyzing machine history data from our CMMS for clues followed by an 8 hour [look behind covers and touch every nut and bolt] type PM gathering and recording measurements for trending. The last hour is spent generating a detailed report with pictures of items repaired and open issues that need follow up planning and scheduling.

quote:

The challenge we face is not PdM, RCM or implementation of other TPM tools but finding a window to have access for "hands on" PM.

quote:

I noticed most of you are proponents of RCM. I am not....it takes waaaaaaay to long...

If your company is like the many I have worked with I will write out the brief RCA. But first some results from a company like yours:

Era.................... Annualized PM Hours / Availability (approx)
Pre debottlenecking..................416 / Mid 80%s
After courageous Management.... 208 / Mid 60%s
After PMO................................ 204 / Low 90%s
After plant mods..........................100 / Mid 90%s

Noteable achievements PM hours reduced from 416 to 100 per year
Availability up to mid 90% from a low of mid 60%
Note the name of the company has been suppressed and the name of the production area has been changed to Cell C.
Here is a brief RCA.

You cant get the windows you need for maintenance.
Why?
Production does not buy in to your maintenance – they are under pressure for production but your maintenance program does not "turn them on".
Probable cause
Your maintenance is highly intrusive and production does not think that you are doing the right thing – their experience is that the machine is working fine, they give it to maintenance for 8 hours, maintenance then wants 10 or more and when production gets the machine back, it does not work as well as it used to ... at least they see that between maintenance downdays there is a lot of breakdowns and they get these regardless of whether they do your maintenance program or not. Production may be disenchanted with the whole maintenance problem – they may not be telling you but they are not playing the game you would like them to and this is usually an indicator of some of the underlying feelings.
Why
You may have the wrong maintenance program or at least your maintenance program has not been analysed to ensure it needs to be done – because you have not done RCM – you can not justify the maintenance.
Why
RCM takes too long.

Here is a case study that I hope illustrates how some companies break this cycle.
Background
Everything seems fine until company debottle necks the plant downstream.
Company has a strong master slave relationship between production and maintenance. Production says to maintenance do and maintenance says yes masser.
Cell C now becomes the bottle neck. Uptime is fair but not enough to keep the rest of the plant running at new capacity so the eight hour weekly maintenance downtime is pushed to two weekly. Little or no discussion just management decree. (We call this courageous management which is about as polite as we can be in these circumstances).
Missing the weekly maintenance window means that predictive maintenance is not done frequently enough and failures start to happen in week two. Temporary repairs become frequent.
8 hour down day starts to be filled with correctives (mostly removal of temporary repairs). Within a few months, 8 hour day is filled completely with correctives and no PM is done. Because the cell is shut down for the 8 hours as planned, the PM workorders are signed off as complete when nothing could be further from the truth. Supervisors are not game to report anything other than 100% compliance – just that every maintenance department is reporting 100% and the local supervisor is not going to admit that he is not achieving like the rest of the supervisors.
Soon, breakdown maintenance is chaotic.
Tradesmen are to blame.
Decision
Decision is taken to review the maintenance program based on RCM principles using PMO.
In three days the PM (including operator rounds) is reviewed. Attached is the report.
Results
The PM program was found to be a mess – almost 20% of it was a waste of time. 30% could be extended, almost 20 failure modes were preventable but were receiving no PM at all and 25% of the program had PF intervals less than two weeks hence the tasks needed to be done weekly. There were a number of random and sudden failures that were going to be breakdowns regardless of what PM was prescribed. Inspections were of no value and overhaul was counterproductive.
When all the facts were known, and there was a genuine understanding that the machine availability was inherent in the design and the way it was operated AND that maintenance could only achieve what was inherent AND the right maintenance could be figured out logically with cost benefit, the operations bought in.... In fact they began to be much more active in Reliability Engineering.
So why, in the first instance did operations take the courageous decision to double the maintenance interval?
I think it because they had no confidence in the maintenance that was prescribed. If this is the case, were they right? They most certainly were.
So I think golfat8, the first question you may ponder might be right there in your own back yard. Is the maintenance you are doing necessary and if you are asked about each task, would you be able to put your hand on your heart and show the justification?
Now what we are talking about is a reliability assurance program based on PMO and there is a bit more to it than a one pager. If you would like to make contact - you can get my details from Terry O I am sure. I can do a one to one PowerPoint / phone presentation (or one to many) for 45 mins. This offer extends to others as well.
Hope this helps... not sure if this is too much info or is going to be seen as sales .. Hopefully it is seen as good case study material and food for thought. There is a fine line I know and I dont want to get on the wrong side.
Regards
Steve

This message has been edited. Last edited by: Steve Turner,

CellC.ppt (176 KB, 12 downloads) Power point case study

quote:

Era.................... Annualized PM Hours / Availability (approx)
Pre debottlenecking..................416 / Mid 80%s
After courageous Management.... 208 / Mid 60%s
After PMO................................ 204 / Low 90%s
After plant mods..........................100 / Mid 90%s

Have you published the results somewhere for peer review? Do the results have impact on the company's financial performance?