Cost of Downtime.

Six Sigma – iSixSigma Forums Old Forums General Cost of Downtime.

Viewing 13 posts - 1 through 13 (of 13 total)
  • Author
  • #39163


    I have always had issues with this and wanted to ask how everyone else has handled this to see if I am missing something here – like the boat.
    A lot of companies will put costs on downtime such as $/minute.  How do people come up with the number?  If you have a line with extra capacity and you do have the ability to run a little faster than takt time without adding people.  So may labor costs don’t go up, I didn’t add machines so my capital costs don’t go up, and I didn’t work any overtime.  How would I really sell it that downtime costs me money? I didn’t ship at premium or anything – all I did is work a little faster with what is required.
    I have seen a lot of people try to use sales as the driver.  So every X minutes down costs me a potential lost sale.  How is this number derived?


    Jason C. Dennison

    Tough question.  I’d say get a target sales dollars for the fiscal year, and convert the $/yr to $/min from that, so effectively every minute has potential to reach the sales target but if it’s down, that’s a missed opportunity.  The other cost you may need to incur is the overhead of keeping the equipment running – I don’t know your application, but I work at a rubber extrusion company and we have certain $/hr figures on how much utility cost there is to maintain the line being on and ready to go, as we do not frequently shut down the line but keep it on during downtime periods (depending on the length of downtime- it’s a cost/benefit ratio of the peak utility needed to physically start the line compared to how much it costs just to keep it running).
    So basically your best best on selling that downtime costs money is to treat it as lost opportunity based on the fiscal year sales goal.  At least in my opinion that’s the best bet.



    Hi – my £0.02 on the cost of downtime……
    We have a project running at the moment that looks to improve the throughput of a line via OEE improvements. The value proposition is based around the fact that current efficiencies, it costs £X extra to achieve the customer demand. The extra costs come from running at the weekend, additonal shifts etc.
    Therefore the success of this project, as measured by the finance guys and the process owner, is in tracking the cost of running the line, checking that we are achieving the required volume, and measuring it against standard costs. It should all equal itself out, if there is no unexpected downtime. But its not an ideal world……
    Clearly downtime does affect the ability to meet required volume, so we are focussing the improvement activities on finding out why DT occurs, measuring it when it does occur. We expect (without jumping to the end of the project before we get started) to look to increase batch sizes (to reduce set up & change over DT’s) and to look at what has crept into the process from the time it was designed – I am thinking here of additional checks, tests etc that result as a consequence of a customer complaint. Very often these reflect the past and should be checked to see if they are really needed or just an additonal DT.
    So, as Jason says in his post, tying the cost of DT to lost opportunity is one way. My description above is another. It all comes down to what makes your senior managment sit up and take notice – Money Spent, Sales Lost, etc etc.
    Hope this helps


    Paul Gibbons

    My 1p worth,
    From my experience (limited according to Btdt, but I will give it anyway;¬), the performance element of the OEE calculation is often where people try and cheat to get a higher OEE. The cycle time used in the calculation should be either the original cycle time from when the process was new and everything was fine and dandy; or, if the process has been improved and proven as capable from a reliability and quality (the affect of reducing the cycle time has not worsened the availability and quality elements of the OEE calculation) perspective, the new cycle time should be used. When you have a performance % over a 100, you are either cheating or, you need to validate the cycle time.
    Another important factor about the OEE is that theoretically speaking, if you use standard working principles, such as used in the 5S system, your perfomance should always be at or near to 100%. I have tested this out and although my inferences matched the theory, the size of the case study (one;¬) would suggest that the external validity of the finding would not hold up in academic court; I could not generalise about the finding.
    Relating to the costs of downtime, from a maintenance engineering perspective; think about the costs of repair or replace. How much does your maintenance/TPM crew spend on maintaining the plant; labour and spares? Does the downtime due to plant failure affect the quality of output. In other words, when the plant breaks down, is it difficult to get it running well again or are there a few start-up scrap products?Although you are achieving a high OEE, is the plant performing as it was meant to? This relates back to the Performance element of the OEE, although it is not downtime, the performance target should never be set lower than the orignal specification of the plant. If it is then you have  to take that into account with regard to the repair or replace equation.
    I have more but I won’t bore you all any longer.
    Hope it helps



    Typically the cost of DT are managed within Asset Efficency and Asset Utlization KPIs. You must define DT planned and unplanned and after that it is quite simple to understand the effect of unplanned DT on Manufacturing/Supply Chain/Sales The global capacity of a facility is the results of  y = max (Machine/Manpower/Methods/Materials). If you have extra capacity of Machine and you whish use it without increasing of Manpower, you must maximize the equation changing another parameter, typically Methods, changing some processes.
    Rgs, Peppe



    This can be a tough calculation on a number of levels.  In addition to what’s already been suggested, have you looked at defect levels?  Every shutdown/startup is introducting additional variation, so intermittent downtime periods could be worsening your overall quality.
    I don’t think a sales-based # will work if you don’t have the demand.  One metric we use is “parts per labor hour”, to gauge operator productivity…obviously if the equipment is down and the operators aren’t doing some value-added activity, that metric will go to “zero” pretty quickly.



    Maybe i’m missing the point here. 
    I think you are saying that when the line goes down, it doesn’t result in extra “costs” because your TAKT time is big enough to hide the downtime.  That is, you can over-produce in the allotted time or your in-process work at each station is large enough to cover the lost work during down periods.   
    To me, the issue isn’t downtime…it’s your TAKT time.  I’m assuming your downtime is significant since you think that is a big cost.  If your TAKT time is large enough to cover this, you might want to re-look the line and lower your time.  Assuming, of course, decreasing TAKT time doesn’t increase defects…but you should be able to resolve that.
    As for caculating the cost of downtime, your finance dept. should be able to tell you the operating cost per minute at any station in the plant (machine rate or whatever).  Cost is cost.
    I wouldn’t focus on sales figures; they’re too arbitrary.  



    This is my suggest on how to calculate  downtime cost. During my work sometimes I have not to  take all the factors but some only.
    All is total dollar value:
    Equipment replacement costs as spare parts, packages, consumers etc.,  cost of lost production or real net profit not achieved during all  downtime including the possibles losses to the followings lines of production for interrupting the process.  Emergency response costs as maintenance people fees or of outside contractors, anyother cost generated by  downtime like workers compensation for doing nothing during downtime, indemnity payments, environmental cleanup costs
    Good Luck
    Seidel Muriel
    Reliability Engineer
    Cali, Colombia, South America.



    Here what I been doing on my downtime.
    Design a spreadsheet with the following information.
    Design blank column chart per hour for each shift
    Categorize downtime under each hour:
    Feeder changes
    Engineering / Programming
    Line Setup
    Machine Breakdown
    First Article
    Shift Reduction – Meetings – Safety Drills – Company extra Lunch
    No Materials / Part Shortage
    Track production rate and downtime inside the column chart by hours. Track the assembly until you can determine the real TAKT time and use it as the TAKT time. Use the in-process production to determine your downtime cost for that assembly.
    Maintain these data in a database and keep building up your database. Design another spreadsheet to calculate your downtime cost by linking your database.
    Hope this help.



    Sales figures, particularly lost sales can be very arbitrary. One area that we do look at is definitions of LD (liquidated damages) or SLAs (service level agreements). These are contractual obligations and can contain very real and expensive penalties for failing to deliver on time.
    Sales and finance will keep records of past failures. The Risk management group may track regulatory obligations for SLAs; this is common in the telecom business.



    The way $ DT is calculated depend on the cost system the company is using, if the company is using standard cost system the cost per hour or per piece could be calculated based on the operation expenses depending on the driver (machine or labor). What cost system is using yor company? definitely the cost per hour/piece should be provided by your financial department.
    If the company is working with extra capacity and with higher cicle, all the time, seems like the product cost is higher than it should be, due to the fixed expenses such a machinary depreciation, electrical energy, etc., and for the labor cost, is someone from Six Sigma/ Lean working already on this aspect? if not could you sell the idea as product cost reduction instead of DT?


    Kelly Thrasher

    Not much to go on here but I’ll throw some suggestions at you.
    1.  Unit cost.  additional time required to produce a unit will be added if there is down time that you include in your TAKT.  Less down time, less cost per unit.
    2.  Labor costs.  depending on how you track pay; .1 hr of a crew standing still and not producing is an increase in labor costs to produce.  Same with the time required to find repair parts, etc.  Any wait time increases your labor time for that job.  We track direct/indirect wage as part of unit cost.
    3.  Lost revenue.  Luckily (?) we are okay right now but will be outsourcing work in our heavy months.  If I track the downtime as lost production time, I can tie it to lost revenue.
    Tack your lost time/down time to what your senior leadership is tracking.  Get their attention by illustrating how much downtime costs them in one of their KPIs.
    To all of the UK participants, it was truly a joy to work with your folks in Basrah, etc when I was in Iraq.  True professionals and some great “kids”.



    I propose to consider TOC(Theory of constraint).
    If downtime occurs in a process which is bottleneck, cost is directly sale price of final product. 1 unit process output = 1 unit final pruduct.

Viewing 13 posts - 1 through 13 (of 13 total)

The forum ‘General’ is closed to new topics and replies.