What Key Performance Indicators do you monitor to manage the maintenance and reliability of your plant?

In my opinion, too many.

MTBF and MTTR are simple for maintainability and reliability performance measurement

and much more

I currently have 1 Key Performance Indicator. Planned x Reactive Maintenance.

Interesting, Planned times Reactive Maintenance. How do you measure it?

Try measuring MTBF (mean time between failure) and tell me what you find. I attached a MTBF Users guide I developed out of frustration. MTBF is the best way to measure reliability.

I would like to know the results once you begin measuring MTBF. start with your most critical piece of equipment and monitor it for a month.

MTBF_User_Guide_Rev_1_1.doc (60 Kb, 213 downloads) MTBF Users Guide

Take out all PPM workorders
Then you will have a list of maintenance workorders.
We have 4 types of priority:
1 - emergency
2 - urgent
3 - normal
4 - combined shutdown

Priority 1 and 2 we call reactive, the rest is planned. Our users have the "freedom" to give a priority, but there were abusers of the system.
We explanied the rules -
1 - emergency means maintenance will stop all other work immediately to tackle the problem (no leave, no other work), all hands on.
2 - urgent means, working until the problem is fixed
If someone wants to change a valve, and not because he wants it yesterday he will put priority 1 or 2. What about delivery time of materials of 2 or 3 months?

We evaluate the KPI over a year, thus the formula is (1 + 2)/Total workorders for reactive workorders
For planned workorders is that 1 - reactive

Indicator:
80 - 100% planned = good Green
60 - 80% planned = alert yellow
<60% planned = bad red

Put it in a nice dashboard, divided by trade group, when they open-up the CMMS, the first thing that the section supervisor sees is his maintenance mode. Works like a breeze. When I click on reactive, I see the the reactive jobs and status (They also, and the customer)

Benefits:
1) Maintenance doesn't get overloaded with unreasonable reactive work, because if it is too big, it is easy to loose focus of what is important and what not.
2) People not knowing how to prioritize are spotted and receive attention/training.
3) Tradesmen with to much open reactive work are spotted and receive special attention.

3) Believe in the workorder system is enforced.

4) Lip service is not allowed, the excuse of "I told them to fix it 20 times" is not valid without a workorder.

5) CYA politics by antidating is spotted, and immediately corrected

The B.S. of notourios abusers was pinned one or two times on the publication board, and they got the message.

Steven,

Awesome KPI. I hope others can benefit from your KPI.

Thanks,

Ricky

This looks like a very interesting KPI but forgive me, I am slow and I am not sure I fully understand what you are saying. If 1 and 2 are reactive then (1 + 2)/Total workorders for reactive workorders is always going to be 1?

Can you give us an example with some sample figures to clarify?

My maintenance department receives of 250-300 workorder p/month. Part of this is scheduled maintenance (PPM). Let us say we have a peak of non-ppm workorders. This can be resolving a leak on a flange, fixing the lightning, calibrating a transmitter or a problem on an airconditioning system. We run a plant 7x24 with a dedicated shift Electric/Instrumentation mechanic on every shift.
Also we have a list of critical equipment.
A priority 3 (normal) does not mean that the problem will not be resolved immediately, but it will not stop production or result in a dangerous situation if unattended. By the way it is an oil refinery. If I get a priority 2 in the afternoon, that will result in most cases in scheduling/planning on the fly with probably overtime.
A brake problem on a car is not an emergency or a priority 2 urgent. If the owner waits to report problems until it is too late and bangs into a wall, he is an idiot and I do not have mercy with him.

For reactive workorders I define the percentage as: (emergencies (1) + urgent (2) )/ total workorders released in the current year

The Planned workorders are 1 - above defined number

I started with this in the middle of last year, and initially there were sections in the red zone. We didn't went back in the history to correct the priorities on already closed workorders.
This year we started with a leap, all sections started with green or yellow.

This message has been edited. Last edited by: svanels,

I will give you an example of what I used to use while I was in maintenance. As a maintenance supervisor I used 8-10 KPIs and as a maintenance manager I used on 5-6. As a maintenance supervisor I was concerned with % of effective PMs based on a 30 day interval, % of planned work (you have to define what planned work is), Scheduled Compliance (again must be defined, etc. I recommend maintenance technicians use asset availability and MTBF. These two are the results of all their hard work. Which ones do you think maintenance technicians should use? Do you agree with my two we used to use?

There are PI's and KPI's. Like a maintenance superintendent I am interested in planned x reactive. This will give indications where are the weak points in the organization. It is aesier to translate this to the other "players", because they also are contibuting.
A PI like MTBF means you have to define your equipment. Calculating MTBF for all registered equipment I call a waste of time, what do we with the number? If I have nothing to compare it with, probably the owner doesn't care either.
Schedule compliance is for the year report.
Another PI (not KPI) is wrench time per trades group. If one section runs above 4 hours, I know there have been a lot of overtime, and maybe I need extra manpower to equalize the load.
The technicians may use the PI's they want, but I have the experience that they are not interested in too much abstract calculus, (that is for the upper cloud)

How do I calculate MTBF for Pump Group?, If we want to know How to calculate MTBF of Failure Mode such as Mech. Seal Leak?

Example We have pump (mech. seal) 100 pc in plant

two pumps have fail cause by mech. seal leak and 98 pump normal in term of month (base on 30 days)

pump #1 downtime = 30; MTBF = 720-30 = 690
pump #2 downtime = 20; MTBF = 720-20 = 700


Option A: MTBF = 690+700 / 2 = 695 or
Option B: MTBF = ((720*98) + 690 +700 ) / 100 = 719.5

which one is correct? or incorrect bolt?

quote:

Originally posted by Bryan Weir:
(1 + 2)/Total workorders for reactive workorders is always going to be 1?

1st KPI = (1+2) / (1+2+3+4)
2nd KPI = 1 - 1st KPI, or (3+4) / (1+2+3+4)

quote:

Take out all PPM workorders ...

PPM = Predictive and Preventive Maintenance? Planned Preventive Maintenance? or Parts Per Millon?

To calculate MTBF is very simple. It is the total number of failures divided into time.

Example: 3 failures of a pump in 24 hours

24/3 = 8

Attached is a guide which will help.

MTBF_User_Guide_Rev_1_1.doc (60 Kb, 48 downloads) MTBF Users Guide

Hi AromaticsThailand,

I think you mean a duty pump and a standby pump forming a group. If so and:

. If the system is not repairable – meaning that when the duty pump fails, it is left in place as it is and the standby pump takes over until it also fails, causing the system to fail (case of a satellite, for instance) –, then, the MTTF of the system is simply given by MTTFus = MTTFdt + MTTFsb, where “us” stands for unrepairable system, “dt” for duty and “sb” for standby. Example: MTTFdt = 500 hours and MTTFsb = 500 hours. Then MTTFus = 1,000 hours.

. If the system is repairable (like in most situations), then, the solution is far more complicated and I use simulation in these cases. Considering the same individual MTTF´s above and a 50:50% regimen and after a few thousand runs, the MTTF of the system is approximately 4,960 hours (far higher than 1,000 hours as easily predictable).

Rui

AromaticsThailand,

TTF´s and TBF´s have different meanings. Please see attached an EXCEL file which shows an example of the meaning of the symbology that I use.

Rui

indicators.zip (4 Kb, 105 downloads)

PPM in the maintenance context

In the past we knew PM = Preventive Maintenance

We also knew IM = Incidental Maintenance (everything thas was not PM)

In our maintenance evolution we discovered
PPM = Periodic Preventive Maintenance

Since we also do some predictive Periodic can be interpreted as Predictive

Since all this things are Planned (known up-front), the "correct" abreviation would be

Planned Periodic Preventive Predictive Maintenance but PPPPM would be gross overkill

I visited a plant recently and they had their KPIs established from plant manager down to operators. Each management level had their own dashboard with 5 KPIs. Planners had the most KPIs with 8. All of the KPIs had others leading indicators (metrics) which helped each position manage the KPIs. For example if vendors are delivering late at 60% of the time then it will effect the % of planned work and schedule compliance and utimately MTBF and Asset Availability.
All of these KPI were aligned from top down. This company is in a down market but profits are up and cost continue to drop as they identify waste in their processes and eliminate waste in time, labor, and materials.
It is good to see someone doing things right.

We are also have a lot of KPIs in the form of 'Balance Score Card' to monitor but mostly at Plant/Department level. Looking for Section Level and below. Will be interested to look for KPI at these levels. Specially Reliability Point of view or Planning Point of view. Anyone has further details.

Raza