Filacoste

Why 3D prints fail and what it really costs you

The wasted filament is the cheap part. Here is what you actually lose with every failed print, and where failures come from in order of frequency.

Written by Álvaro Melgar Merenciano·

Close-up of a desktop 3D printer on a workbench.

When a print peels off halfway through, the first thing you look at is the tangle of plastic on the bed and you think about what you have thrown away. It is 96 cents. The expensive part is what you cannot see: two and three quarter hours of printer time and ten minutes of yours cleaning up and restarting.

This article breaks down what a failure really costs, how much each failure rate takes from you in a year, and why the "failure %" field in the calculator is not there for decoration.

The wasted filament is not the problem

Let us start by dismantling the most repeated idea. Take the site's reference part: 60 grams of PLA, five and a half hours of printing. If it breaks exactly halfway, this is what is lost:

Direct cost of a failure halfway through, with the 60 g and 5 h 30 part.
ItemCost Share of total
Material, power and machine wear0.96 €32%
Your time cleaning up and restarting (10 min)2.00 €68%
Total direct cost2.96 € —

Two thirds of the cost of a failure is your time, not the plastic. And that is before counting the 2.75 hours of printer that are gone and appear in no account at all, because you do not pay for them: you simply produced nothing during them.

That distinction changes how to attack the problem. If the enemy were filament, the answer would be buying cheaper. Since the enemy is time, the answer is that failures should happen early, or not at all.

When it fails matters more than whether it fails

A first-layer failure and a failure at 90% have nothing in common. The material and the hours consumed up to that point do not come back.

What is lost depending on when the print is interrupted.
Moment of failureMaterial PowerMachine Loss
First layer0.02 €0.00 € 0.01 €0.04 €
At 25%0.30 €0.02 € 0.15 €0.48 €
Halfway0.60 €0.05 € 0.31 €0.96 €
At 90%1.08 €0.09 € 0.56 €1.73 €
Finished but unusable1.20 €0.10 € 0.62 €1.92 €

A first-layer failure costs 4 cents. The same failure caught at 90% costs 1.73 €, more than forty times as much. That is why watching the first five minutes of a long print is the habit that saves the most money of all.

And it is why a camera, or simply a glance every half hour on eight-hour prints, pays for itself: it does not prevent the failure, but it stops the failure eating all eight hours.

Where failures come from, in order

According to the breakdowns used by manufacturers and print services, the causes cluster into four groups:

Approximate breakdown of failure causes in FDM printing.
CauseShare When it shows up
First layer and adhesion35-40% In the first few minutes
Filament condition (moisture, batch, diameter)25-30% At any point
Slicer settings20-25% Tends to repeat identically
Mechanics (belts, nozzle, hotend)10-15% Gets worse over time

Note the fortunate coincidence: the most frequent cause is also the cheapest to suffer. First-layer failures, nearly four in ten, happen when only a few cents have been spent.

The bad combination is the second row. A moisture problem gives no warning at the start: the filament begins clicking and extruding badly halfway through, exactly when the loss is real.

What each failure rate costs per year

Suppose you print around 200 parts a year of this size, which is what you get printing about three hours a day. This is what each rate takes:

Annual cost of the failure rate with 200 parts a year, assuming they fail on average halfway through.
RateParts lost MaterialYour time Printer hours
5%109.59 € 20.00 €28 h
10%2019.18 € 40.00 €55 h
15%3028.77 € 60.00 €82 h
20%4038.35 € 80.00 €110 h

Look at the last column, the one that hurts. At a 20% failure rate you lose 110 printer hours a year: nearly five full days of machine producing rubbish.

If you sell, that column is sales that do not exist. Dropping from 15% to 5% saves 59.18 € in material and time, but above all it frees up 55 hours, enough for about ten more parts. At 11.90 € profit each, that is 119 € of income that did not exist before.

You can see the effect on your own case by moving the failure percentage field in the cost calculator: you will watch the cost per part rise without touching anything else.

The filament dryer: the honest sums

Here is a calculation you probably do not expect. A filament dryer costs around 55 €. If moisture causes roughly 27% of failures, and you have a 10% rate over 200 parts, you are losing about 5.27 € a year to damp filament.

At that rate, the dryer takes ten years to pay for itself counting material alone.

Does that mean do not buy one? No, it means do not buy it for the material. It pays for itself another way: every moisture failure you avoid gives you back 2.75 printer hours and ten minutes of your own. Counting time, the sum changes completely.

And there is one case where it is not up for debate: if you print materials that genuinely absorb moisture — nylon, PETG in summer, anything with fibre — the dryer stops being optional. PLA is fairly forgiving; nylon forgives nothing.

For drying PLA, manufacturer guides agree on 40-45 °C for 4 to 6 hours, without going over 50 °C because the spool starts to deform. And a detail that gets forgotten: do not leave the spool out in the open after drying it, because it reabsorbs moisture within days.

How to tell your filament is damp

You do not need a meter. The symptoms are clear enough once you know what to listen for:

Clicking or popping at the extruder. That is water flashing to steam inside the hotend at 200 °C. If you hear a pop every few seconds, there it is.

Rough surface where it used to come out smooth. Same file, same printer, and suddenly the part has a pitted texture.

Parts that snap by hand. Moisture degrades layer adhesion, so the part holds far less than it should.

Visible bubbles in the strand as it leaves the nozzle.

The definitive test is cheap: dry a suspect spool and repeat the same print. If the problem disappears, you know which side it was on.

One clue that helps rule things out: if the problem appears with every spool at once, it is not moisture, it is the printer. Moisture goes spool by spool, depending how long each one has been open. When everything fails equally, check the nozzle and belts before spending on a dryer.

The most expensive failure: the one the customer finds

Everything above is failures you see yourself, at home, before anything leaves the house. There is another category that costs an order of magnitude more: the part that ships defective and the buyer finds out.

A normal sale of that part at 20 € plus 4 € shipping leaves 11.90 € of profit. If it has to be remade and reshipped:

What happens to the profit when a sold part turns out wrong.
ItemAmount
Profit on the sale11.90 €
Second production run−5.11 €
Second shipment, on you−4.00 €
Messages and handling (15 min)−3.00 €
Final profit−0.21 €

The sale goes from paying you 11.90 € to costing you 21 cents. And that is the good case, the one where the customer agrees to wait for the replacement.

If they ask for their money back it is worse than it looks: you refund the full 24 €, but the marketplace fees are not refunded. You lose the 5.11 € of production plus the 2.99 € of fees: 8.10 € in total. It takes almost a whole good sale just to plug the hole left by a bad one.

Hence the practical conclusion: checking the part before you pack it is not optional. Two minutes looking at the surface and testing the fit are worth far more than the two minutes they cost.

What percentage to put in the calculator

The practical question. The failure field exists so that the cost of the parts you bin is paid for by the ones you sell, and putting zero is lying to yourself.

My rule of thumb, for what it is worth:

5% if you always print the same thing, in PLA, with profiles you have run in and a printer you know. That is a rate for somebody with production well dialled in.

10% is the calculator's default and the most reasonable for most people: varied jobs, the occasional different material, new parts you have not printed before.

15-20% if you do custom commissions, with geometries you are seeing for the first time, technical materials or large parts where a peel-off at six hours is a real possibility.

If you want to see it in numbers before deciding, open the calculator with a 5% failure rate and change that field to 20%: it is the same part, and what moves at the bottom is exactly what each point of the rate costs you.

And one rule worth more than any estimate: log your failures for a month. With thirty parts you already have a number of your own, which will always beat mine.

What actually lowers the rate

In order of impact according to the breakdown above:

1. Level properly and clean the bed. It attacks 35-40% of failures, it is free and takes five minutes. Isopropyl alcohol, and re-level when you change the surface or notice the first layer is not squishing the same.

2. Watch the first few minutes. It does not lower the failure rate, but it turns 1.73 € failures into 4-cent ones. It is the best effort-to-saving ratio on the whole list.

3. Store filament sealed. Airtight bag with silica gel, and change the gel now and then. It attacks a good share of that 25-30% from moisture without spending on a dryer.

4. Do not touch a profile that works. Slicer failures usually turn up after "improving" something. If a profile works for you, save it and duplicate it before experimenting.

5. Maintenance before it breaks. Nozzle, belts and fans degrade slowly, and by the time they show themselves they have been spoiling parts for weeks.

What to take away

The wasted material is small change. At a 10% failure rate it is 19 € a year. What hurts is the 55 printer hours and your own.

When it fails matters more than whether it fails. Forty times more expensive at 90% than on the first layer. Watch the start.

Never put 0% in the failure field. If one part in ten gets binned, that tenth is paid for by the nine that sell, and that has to be in the price.

The failure that reaches the customer plays in another league. Remaking and reshipping turns 11.90 € of profit into 21 cents of loss; a refund takes 8.10 €. Checking the part before packing is the best-paid inspection in the workshop.

A warning about the cause percentages above: they are approximate breakdowns that circulate among manufacturers and print services, useful for knowing where to start, but your breakdown will differ. Somebody printing only PLA in a dry room will have far fewer moisture failures than somebody working with nylon in a basement. Use those numbers to prioritise, not as if they were yours.

Once you have your real number, take it to the selling-price calculator to see how much it changes what you should charge. And if you want to see where every cent of the cost comes from, it is broken down in the guide to what 3D printing costs.

Want the exact number for your own case? 3D printing cost calculator does the full sum with your filament, your electricity tariff and your printer, and leaves the quote ready to copy.

Frequently asked questions

How much does a failed print cost?

With a 60 g, 5 h 30 part that fails halfway, the direct cost is 2.96 €: 0.96 € of material, power and machine wear, and 2.00 € of your time cleaning up and restarting.

Only 32% is material. On top of that, 2.75 printer hours are lost that appear in no account because you do not pay for them, you simply do not produce during them.

What is the most common cause of failure?

First layer and adhesion problems, which account for between 35% and 40% of failures.

They are followed by filament condition (25-30%), slicer settings (20-25%) and printer mechanics (10-15%).

What failure percentage should I put in the calculator?

5% if you always print the same thing in PLA with run-in profiles; 10% for varied work, which is the default; 15-20% if you do custom commissions with new geometries or technical materials.

The best approach is logging your failures for a month: with thirty parts you have your own number, better than any estimate.

Is a filament dryer worth buying?

On material saved, no: with a 10% rate over 200 parts you lose about 5.27 € a year to moisture, so a 55 € dryer would take ten years to pay for itself.

It pays off on time: every failure avoided gives back 2.75 printer hours. And with nylon, PETG in summer or fibre-filled materials it stops being optional.

How do I know if my filament is damp?

The symptoms are clicking or popping at the extruder, a rough surface where it used to be smooth, parts that snap by hand, and bubbles in the strand leaving the nozzle.

The definitive test is to dry a suspect spool and repeat the same print: if the problem disappears, it was moisture.

At what temperature and for how long do you dry PLA?

Between 40 and 45 °C for 4 to 6 hours, without going over 50 °C because the spool starts to deform.

Afterwards do not leave the spool out in the open: it reabsorbs moisture within days. Store it in an airtight bag with silica gel.

What does a 20% failure rate cost me per year?

With 200 parts a year: 38.35 € of material, 80 € of your time and 110 printer hours, which is nearly five full days of machine producing rubbish.

If you sell, those 110 hours are sales that do not exist.

What do I gain dropping the failure rate from 15% to 5%?

You save 59.18 € between material and time, but above all you free up 55 printer hours a year.

Those 55 hours are enough for about ten more parts. At 11.90 € profit each, that is 119 € of income that did not exist before.

More tools and guides

Tools Guides