Scrap doesn’t just happen

In reality, if you need 1,000 of whatever product, the manufacturing process is rarely going to yield exactly 1,000 of it, even if you feed into the first operation the exact quantities of raw materials system calculated as gross requirements. The process may produce 980 or 1,020, but is hardly ever going to be exactly 1,000. If you didn’t take scrap into account, or you took incorrect scrap into account, your actual output from the process might be much more unpredictable, with huge variances. Variances are always a problem, only they don’t have to be that big a problem in all scenarios.

If you are a make-to-stock company, manufacturing your products to meet forecast demand, you produce goods, than stash them away in the warehouse, even before you have any customer orders. Based on your previous experience and understanding your own business, you may eventually optimize make-to-stock manufacturing not only from supply chain management perspective (meeting the demand in time), but from costs perspective as well. Understanding your scrap is crucial in this process.

In make-to-stock manufacturing, fixed scrap can be minimized by running large lots as often as possible. Fixed scrap is the same regardless of volume, so running a 1,000 pieces per lot will cost you significantly less than 100 pieces per lot. Fixed scrap is real scrap, a certainty, and is also something you can’t avoid. You can do two things about it: cut it down to necessary minimum; and you can minimize its effect on the total cost of your products by running larger lots.

As opposed to fixed scrap which is certain, running scrap is not certain. It is something that may or may not happen, but statistically it is going to happen, and it is extremely important to understand how often it happens, why it happens, and how big it is when it happens. If you don’t know the answer to these questions, you must count with much higher uncertainty ratios.

For example, you notice your machine starts producing scrap. You may do several things, the worst one being just calculating the scrap to useful product ratio, then take this factor into account in planning, and increasing your prices accordingly. A better course of action would be to analyze why it happens. Sakichi Toyoda and Taiichi Ohno of Toyota made a really good job at how to analyze this with their 5 Whys technique, which later influenced more known and widespread concepts, such as Six Sigma. In any case, it is crucial to understand why scrap happens in order to take the most appropriate action.

In our example, say that you find out that the reason of scrap is a worn-out machine part. By deeper analysis you also find out that after 5,000 machine hours this part wears out a little, increases scrap factor by 5%. In order to replace the part, you must purchase the replacement for $1,000, then stop the machine, take it apart, and replace the part, which is a process which can take up to 6 hours. At this point, you don’t just go and replace the part. You must analyze which costs more, stopping your production and paying repairmen to do their job, or go on with the scrap, then do the repair at scheduled maintenance time. If manufactured product is cheap, it might be your opportunity cost to go on with increased scrap, because 5% scrap may amount to lesser cost in the end, than stopping your production for 6 hours, especially if your machine is a bottleneck, and stopping it would also mean stopping the whole plant.

In any case, in make-to-stock scenarios, your running scrap is also going to affect much larger lots, and statistically your running scrap variances will be lower, so you will be able to plan and operate within tighter margins.

On the other hand, if you are a make-to-order manufacturer, your scrap is going to cause you much more headaches. Make-to-order usually produces smaller lots, and smaller lots are affected more by all fixed costs, including fixed scrap. Then you must address the running scrap in completely different manner. If your running scrap varies a lot, and over 1,000,000 units produced it is on average 5% with variance of additional 4%, if you are producing 1,000 units, you must increase your running scrap estimate, up to 9%. Why? Because on 1,000 orders of 1,000 units, you are going to have average of 5%, but any single 1,000 unit order may produce anywhere between 1% and 9% of scrap. You are better off planning for 9%, otherwise you might come significantly short of original order.

Therefore, companies in make-to-order manufacturing have an even more pressing urge to understand their running scrap, analyze it and minimize it if eliminating it altogether is not possible. Because with make-to-order manufacturing, scrap distills down to customer satisfaction much more than with make-to-stock.

Your customer might be all-right with receiving 980 or 1.020, instead of 1.000 they ordered, and if these variances are low, most of customers (at least those who understand manufacturing process) will be all-right. But if your output is once 760, then 1.240, then 690, then 1.310, you might be scheduled to go out of business.

If you have these huge variances, you are unreliable. Not only the quantity variances are going to scare your customers off, but price variances as well. If you plan for 1.000 not knowing whether you end up at 690 or 1.310, with same gross requirements, it means that your costs are the same for any output between these two values, and that your cost variance is as high as 31%. With such huge cost variances, you must either vary your sales prices significantly, or keep constantly high prices. You are much more likely to keep your sales prices high, because you can never know your true costs before you finish the production process, and customers who get billed 31% more than quoted, aren’t too likely to return. So, you have to operate with higher average margin to compensate for high variance and avoid low quoting, which is going to put you out of business as soon as your competitors figure how to plan and execute with lower scrap rates.

So, scrap is not that innocent as it might seem at first, and definitely isn’t something that just affects the material planning. It doesn’t just happen, and one of your top-priority goals should be to take control of it, before it takes control of you. Long term effects of scrap may be detrimental to your business, or help you build your competitive edge. It’s completely up to you.


Vjeko has been writing code for living since 1995, and he has shared his knowledge and experience in presentations, articles, blogs, and elsewhere since 2002. Hopelessly curious, passionate about technology, avid language learner no matter human or computer.

Leave a Reply