TPM Metrics & Financial Impact

During a recent kaizen event we reviewed the TPM program for one of our clients. The Operations Manager raised a good question regarding TPM metrics and how to show savings through TPM. During a plant budget meeting, the Controller had questions about the increased expenses and hours spent for maintenance as a result of TPM. These were costs to support TPM, including cleaning supplies for 5S and materials for equipment kaizen. The controller wanted to know how they were saving the company money.
The Operations Manager knew that there had been less downtime, and that repair costs had been lower, but just telling stories was not good enough for the financial Controller. We reviewed the TPM metrics, what the goal of each one was, and how the benefits can be stated in terms of financial imput. The following is a summary of our meeting:
The first metric for TPM is MTBF (Mean Time Between Failures). This is measured by machine, and for this metric, the larger number the better. This means that breakdowns are less frequent. This means that repair costs are lower. That can be measured as a direct cost savings when compared on an annual or quarterly basis. Indirectly, less downtime could mean more production output.
The second metric is Percent Reactive Maintenance (% Reactive). The smaller the better for this number. World class is 20% or less reactive and 80% preventive, improvement, or scheduled maintenance. The best way to show direct cost savings for this metric is when repairs that were previously outsourced can be done in-house because of less reactive activity, and because of time spent improving maintenance technicians’ skills. Indirect benefits include using the same maintenance hours to speed up machines or improve quality.
Mean Time to Repair (MTTR) is the third metric. For this metric, the smaller the better. As TPM progresses, repairs are less serious and are quicker. Tracking repair hours and showing an overall reduction is a direct cost savings. Another cost savings is in avoiding expedite charged and added downtime by creating a spare parts list and identify sources of critical components.
The fourth metric is Overall Equipment Effectiveness (OEE). This is measured by machine or by process. The higher the better for this number. World class is 80% or better. Direct financial impact can be shown as machines run faster with better quality more reliably. To improve OEE, you need to use SMED to reduce set up time, reduce scrap and improve yield, and eliminate minor stops and brining machines up to their optimal speeds. By achieving higher outputs of good product in fewer hours, cost savings can be realized.
If you manage to avoid getting lost in the alphabet soup of acronyms (MTBF, % Reactive, MTTR, OEE) then you will be able to effectively link numbers that are meaningful to both the factory and the financial managers.

5 Comments

  1. Erik

    December 4, 2007 - 7:49 am

    Jon,
    What do you mean by “effectiveness” in OEE? Do any of these metrics relate to operational availability? I recent check of one of our lines showed that in the past 96 days, one line had a availability of 76%. (At best. I suspect the group leaders aren’t reporting what they view to be “minor” losses of 5-10 min) I knew that we were bleeding here but now having done the math I’m downright mortified. Do you have any statistics on what constitutes world class availability?
    Thank you.

  2. Jon Miller

    December 4, 2007 - 8:22 am

    Hi Eric.
    The word “effectiveness” is used instead of “efficiency” or “availability” because OEE is more holistic measurement.
    The formula for OEE is:
    Availability x Performance x Quality
    The first category of losses include changeovers and equipment breakdowns, the second category includes minor stops losses to speed, and the third includes scrap and yield loss.
    Here is a more detailed explanation.

  3. s k murthy

    May 18, 2008 - 8:14 am

    How is OEE calculated for a process industry where it is an integrated system wherein each equipment’s OEE may not represent the OEE of the process as a whole?

  4. Jon Miller

    May 19, 2008 - 11:16 pm

    One approach is to measure the OEE of the constraining process. If the OEE was highly variable, or more than one process acted as a constraint, you could measure several spots. At some point as you solve problems to reduce losses and improve OEE of the constraining process, the constraint will move to another process and you will need to measure the other processes also or instead.

  5. B. Huey

    January 20, 2009 - 6:42 pm

    What is considered a world class ratio for direct labor to indirect & salary? In oother words total indirect & salary divided by total labor (direct+indirect+salary). I heard .75 to .85 is world class.