If you need a new process and don't install it, you end up paying for it without getting it." — Ken Stork
When Ken Stork talked about a new process he didn't mean only a brand new production line. In extrusion that need can show up several ways — fixing a chronic equipment problem that has been absorbed into the daily routine, replacing a worn component the line has quietly outgrown, adding a technology that should have been there from the start, or in some cases replacing an end-of-life line that costs more to run than it would to replace.
The common thread is the same every time. The need is known. The cost of not acting is real. And it keeps accumulating while the decision waits.
Here are three situations from three different plants that I've witnessed, where that played out. Each one is different. The pattern is identical.
The Pipe Line — Measuring By Hand
On a pipe line we were checking OD, wall thickness, wall uniformity and ovality by hand ...at the start and end of a reel. That is not quality control, that is a spot check at best. What was happening between those checks was largely unknown. To stay out of trouble on minimum wall we ran heavy. It was the safe call and everyone knew it. It was also expensive and nobody was adding it up.
The downstream operations that used this pipe needed consistent wall uniformity and ovality... these were critical dimensions for their process. What we were shipping them was in spec but not as consistent as it should have been, and the hand measurement method meant we had no real visibility into what was happening during the run.
We installed a combined ultrasonic and dual axis laser measurement system... continuous real time measurement of OD, wall thickness, wall uniformity and ovality. The capital cost was eighty thousand dollars.
The quality improvement was immediate and visible. Downstream operations noticed it. Wall uniformity and ovality tightened considerably and the feedback reflected it. On the material side a two and a half percent reduction in giveaway at one dollar twenty per pound running thirty-eight hundred pounds per hour twenty-four hours a day five days a week came to fifty-nine thousand dollars a month in material saving.
The request had sat in the system for twenty-two months before approval. At fifty-nine thousand dollars a month that delay cost one million three hundred thousand dollars in material alone... before the quality benefits are even counted.
System cost: $80,000. Payback: just over 1 month. Material giveaway during 22 months of delay: $1,304,000 — before the quality cost to downstream operations is counted.
The Sheet Line — Leaking Deckles
The deckles on a sheet die wept polymer from the die ends. The line was losing ten hours every week to deckle-related downtime ...stoppages, cleanouts, startup scrap every time the line came back up. We knew what it needed. New deckle assemblies, a die face inspection. Quoted at fifty thousand dollars installed.
The request sat in the system for fourteen months.
On lost opportunity we want to be straightforward ...whether downtime translates to lost revenue depends on whether you have orders to fill and no other line to fill them. In our case we did. But even setting lost sales aside entirely and using machine time rate only — the fully loaded cost of running that line including labor and overhead... the number is hard to argue with.
At five hundred dollars per hour machine time rate, ten hours of downtime per week, the monthly cost of those deckles not being fixed was twenty-one thousand seven hundred dollars. Over fourteen months that came to three hundred and three thousand dollars.
Fix cost: $50,000. Payback: just over 2 months. Cost of waiting 14 months: $303,000.
The Coated Profile Line — Faulty Payoff Unwind
On a coated profile line a faulty payoff unwind caused constant substrate breaks. Every break meant stopping, rethreading, purging and restarting. The crew knew the signs and had the rethreading routine down to a system. It had been going on for two years.
Twenty-five hours of downtime a month. Eight thousand dollars a month in wasted substrate and coating material. A contractor was brought in to find the root cause in a faulty E stop circuit and properly repair the unwind mechanism ...labor only, no major parts. Nine thousand dollars.
The breaks stopped. At three hundred dollars per hour machine time rate the monthly downtime cost alone was seventy-five hundred dollars. Add the eight thousand in material and the total monthly loss had been fifteen thousand five hundred dollars for two years.
Fix cost: $9,000. Payback: under 1 month. Cost of waiting 24 months: $372,000.
The Common Thread
Three operations. Three different problems. Fix costs ranging from nine thousand to eighty thousand dollars. And in every case the cost of waiting was a multiple of the fix ... sometimes several times over, sometimes dramatically more.
These are not unusual results. They are what happens when a known problem goes unaddressed long enough for the losses to compound. The numbers will be different on every floor. The pattern will not be.
What makes these situations so costly is that nobody treats them as an emergency because on any given day they don't look like one. The line is running. The crew is managing. The losses are spread thinly enough across shift reports and variance lines that no single number demands attention. It is only when someone stops and adds it up...machine time rate, material loss, months elapse... that the full cost becomes visible.
By that point you've paid for the new process many times over.
Why The Decision Keeps Stalling
The cost of doing nothing is invisible. The fix sits on one line of a capital request and looks large. The loss is scattered across shift reports, material variance and rework logs and looks like nothing. Nobody sees the total because nobody adds it up.
The approval process is not built for smaller capital items. A fifty thousand dollar deckle repair going through the same sign-off chain as a major line upgrade loses momentum fast. It falls in the gap between too small to escalate and too large to approve locally. And it dies there.
The line is still running. A line producing badly is still producing. Taking it down to fix it properly creates visible, immediate lost production. The ongoing daily loss is quiet and distributed. We are consistently better at seeing the cost of action than the cost of inaction.
Political risk makes deferring the safer choice. The manager who signs off the spend owns the outcome. If the payback doesn't land as projected they are exposed. Deferring carries no personal risk. Approving does.
Nobody fully owns the problem. Maintenance thinks it is a process issue. Process thinks it is a maintenance issue. The request has no champion and sits in the gap between departments going nowhere.
There is always a reason this quarter is the wrong quarter. Budget freeze, customer audit, key person away, waiting on a study. Difficult timing has a way of becoming permanent.
The team gets too good at the workaround. When the crew manages a bad process well enough the problem stops looking urgent. The case loses momentum and the request gets buried.
The same internal voice stops being heard. A supplier engineer or industry peer saying what has been said internally for six months lands differently. It shouldn't work that way. But it does.
What You Can Do About It
Build a one page case in the language of finance. Fix cost, installation downtime, monthly loss from inaction, payback period, and the cost to date since the problem was first raised. That last number is the one that changes the room. Real numbers from your own operation are far harder to argue with than projections.
Use machine time rate not just material cost. It is the fully loaded cost of that line standing still — labor, overhead, fixed cost contribution. It is a number the finance team understands and cannot easily dismiss.
Stop letting the cost hide. Put the running loss somewhere people actually look. A weekly number on the production board. A line in the morning meeting. Update the capital request every quarter with what the delay has cost since last submission. Make it harder to keep deferring than to approve.
Find a champion above the approval line. A request with no senior sponsor dies quietly. Find whoever feels this problem most and get them alongside the case before it goes up the chain.
Attach the fix to something already moving. A standalone ROI case is easy to defer. The same case under a live customer complaint, a capacity target or a quality initiative suddenly carries company priority.
Resubmit every budget cycle and update the numbers every time. Each resubmission should show the accumulated cost since the last one. Some of these are won by persistence not persuasion.
The fix is almost always right there. The equipment exists, the cost is clear, the payback makes sense.
What is missing is the decision.
And while that decision waits, the meter is running. Pull the numbers together. Put them on one page. Show what the delay has already cost.
That conversation moves fast when someone finally does it.
Recognize one of these on your floor right now? Drop it in the comments. This industry learns best from people who have lived it.