Hacker News new | past | comments | ask | show | jobs | submit login
New Study Details Toxic Particles Spewed by 3D Printers (gizmodo.com)
146 points by elijahparker on Nov 13, 2018 | hide | past | web | favorite | 84 comments

Anyone with a working nose cannot be really surprised about that, especially w.r.t. ABS printing. There's a reason why things that burn/melt things usually have an exhaust of some kind. I'm sure those will become mandatory at some time, and rightfully so. Just like thermal runaway protection should be mandatory, and yet many cheap 3D printers don't have it either. That should actually worry people more than nano particles. I'm still amazed how many people let their printers run while no one is at home (yes, there's OctoPrint+Webcam, but last I checked that couldn't put out a fire...).

> I'm still amazed how many people let their printers run while no one is at home

When it takes 12-24 hours (or perhaps more) to complete a larger print, are you really going to sit around and watch the thing? No, you're going to go about your life until its finished. That means sleeping, going to work, etc.

I have done 3-day long prints and have seen reports of 21+ days for a single print.

This all from hobbyists who have full-time jobs. It would be impossible to sit next to the printer for the entire print time for any big/complex enough print.

The argument is not about it being practically possible, but whether you should.

Though I also disagree with that.

Personally i never leave my printers "unattended", as in i'm at home, but i don't watch them constantly when they're printing. They're in a room that is normally closed when they're running, and is properly ventilated to the outside after. ABS is printed in the garage.

I made a "better than nothing" auto off switch using an Arduino, a 230V relay, a couple of MQ2 smoke detectors, and a couple of DS18B20 digital thermometers.

The MQ-2 is mounted about 50cm above my printers frame, one for each printer, connected to an arduino which has a relay controlling power to the printers. If smoke is detected, or the temperature rises above 45C, it cuts the power to all printers. The arduino defaults to off, requiring manual intervention to turn it on again via a push button. If the arduino loses power, so does the printers.

It probably won't prevent the house from burning down, but in case of an electrical fire it should shut that down pretty good. The problem is the filament itself. It burns like a candle, so if the flames ever make it to the filament, it should have no problems continuing to the spool, setting the house ablaze.

    > The problem is the filament itself. It burns like a candle...
I wonder if this can be addressed by keeping a slight positive pressure of nitrogen or CO2 in the enclosure of the 3D printer?

Filling your room with CO2 by accident won't be great either. That is, proper ventilation would be needed anyway.

Could do a really low volume flow of nitrogen or argon at the extruder like in welding. You'd still want some ventilation but you'd just need a tank, regulator, and tubing instead of a sealed enclosure and you'd probably use less gas. Just enough to prevent anything at the tip from burning.

Yes, ventilation here is more like a fan in a room window than a sealed box and tubes.

If you are worried about it, you can buy automatic fire extinguishers. If you hang one of those directly over the printer, I think the risk of fire spreading from that area would go close to zero.

I've made a point of buying only "quality" printers. I have a couple of Prusa i3 Mk2/Mk2.5, and an Emotion Tech Micro Delta.

While they're maybe not the highest build quality, they're made from quality parts, with safety features enabled in software. Sometimes they're a bit too sensitive, like printing PETG with an open window, which will trigger a thermal runaway 85% of the times because the heater is unable to keep a steady temperature with the draft.

The main "fire prevention" though is i never print while not at home, and i never print while sleeping. That of course limits the time i have for printing, but so be it.

I'm thinking about upgrading to a Prusa i3 Mk3, as this supports pausing/resuming prints and cutting power between it.

Is your setup documented anywhere? I'd be interested in seeing those schematics and code.

You could always mount some of those fire extinguisher balls over the printers - no need for fancy electronics. Might ruin the printer though...

I may have an old schematic in Fritzing, but i doubt it. I never got around to documenting it, but it's simple enough.

My Macbook developed a "hunchback" and is currently having it's battery replaced, so i can't check the details.

IIRC, i wired the relay to a default off, and there's button wired to the arduino that needs to be pushed before it turns on the relay. It then stays that way until one of the MQ-2 or DS18B20 sensors detect a reading that's outside normal operating parameters, or power is cycled.

I spent quite a few hours fine tuning the MQ-2 with cigarette smoke, vape "smoke", and finally burning filament. The bad news is that the MQ-2 is rather cheap, so they need a long "burn in" period before they start giving even remotely consistent readings, and then they still vary wildly from each other. Once burn in is over though, they're rather consistent with their own readings, just don't try to match them to the neighbor :)

As for the extinguisher balls, i considered it, but they looked like they'd probably destroy more than they would save. I keep my printers on floor tiles, ~60cm away from any wall, and with 160cm to the ceiling, so any small fire would have trouble spreading. If the fire was to spread beyond that, i doubt the fire extinguisher balls could save it.

I was looking into CO2 cannisters, but couldn't find any that could be activated by an electrical circuit, and while i could probably design a mechanical part to activate a regular fire extinguisher, i'm not sure i would trust it.

What does this have to do with the topic of the article? It's about particulate emissions, not fire hazard.

As someone who has a 3d printer running about a quarter of the time less than 10 feet away from me hearing that it could be bad for me is alarming that it could be harming me. (Other than the lost sleep because its loud) When though of as a risk benefit analysis though I come to the strong conclusion that I will use my printer no matter what because it is the compiler to my hobby: no printer, much harder time making stuff.

I don't need to iterate on the greatness of them here but they have the potential to radically help teach people how to make stuff which drastically helps in industry.

My main concern with these articles is that there is no relative risk factor presented against. I have no idea if these particulates are something it would be worthy to invest in an enclosure or are there low levels which are permittable.

I know that I personally am much more likely to have health complications from almost anything else but to someone else reading the above article they may draw the conclusion that 3d could be dangerous and should be restricted. If a person shows up to a school board meeting throwing a fit that they are running 3d printers with children in the same room then it looks bad on the printers, possibly having them removed and to generalize I think that that is the last thing anyone wants.

Ninja edit: formating

You could of course try for some adequate ventilation and an enclosure instead of just ignoring the issue.

This. It's not a big deal. Soldering also produces harmful fumes and as long as you ventilate properly it's not a big deal.

Just do an image search for "solder ventilation" for ideas on how to do it.

Regardless of this I kinda do understand the motivation to ignore the issue. People are afraid that there will be a panic and it's going to be banned.

I expect that the output from a single printer won't be that bad, but it is interesting to put the "ventilate properly" comment in a wider context:

> Gosh, this process produces something that is harmful to me. But it's OK, I'll just put it outside.

Just because it is harmful to you doesn't mean that it is automatically bad for the environment. Breathing helium exclusively is bad for you, too.

It is dangerous because it's reactive. And since it's reactive it will disappear in a short while.

And some people (including myself) actually enjoy the smell of molten rosin flux.

And some people enjoy tobacco. That doesn't mean it's smart to continue breathing it in.

Yeah, the risks seem to be overblown, mostly acute lung issues if you inhale a ton of it, metal fume fever may be a possibility but seems very rare if only soldering. In a quick search I see no reports of significant long-term issues like cancer as some are suggesting.

I knew a guy that liked that smell too, he liked the cancer less.

I started this hobby with an Ender 3 a few months ago. Look into motor dampers. They honestly take away ~90% of the noise and loudness.

I bought these: www.aliexpress.com/item/Funssor-12pcs-lot-NEMA-17-Steel-Rubber-Stepper-Motor-Vibration-Dampers-Imported-genuine-42-stepper-motor/32825669027.html

Make sure you find the right dampers for your motors. Ender 3 uses Nema 17.

I actually have these on order shipping from china just takes forever.

Yeah the risk factor is really important to know. It could likely be more unsafe to stand outside because of the toxic particles from cars.

The 'research supply chain' will now start to close that gap. It typically starts with, 'Activity Y creates environmental challenge X' - and then someone else is left to determine if there is danger.

There is a big gap between aerosol science and the medical / biological perspective needed to determine the magnitude of the risk. It is great to have the likelihood - but it is frustrating to not have the whole equation.

Some figures about the UFP emitted by FDM:


From this one data point, it seems that ABS is the big emitter.

Here is another one [1] taken from [2]

Though that is just one graph out of many. With modern PLA getting better and better, it's rare people use ABS. Makes me wonder how harmful PLA is.

I want to add an HEPA filter anyway.

[1] https://www.tandfonline.com/na101/home/literatum/publisher/t... [2] https://www.tandfonline.com/doi/full/10.1080/02786826.2017.1...

PLAs lower melting point is the issue. I only print PLA, mainly because it's less toxic, but I dare not print a part that might end up in a hot car, near a heating source or outside.

Same. 95% of my prints are PLA, but I've had PLA prints deform in my trunk on a hot day. For more durable prints, though, I prefer PETG over ABS for its lower shrinkage, better plate adhesion, and lower fumes. ABS is just generally not a good thing to print with.

I am shocked by how many people use 3d printed parts in the kitchen, printed PLA is not a food safe matieral.

And if you print a toy for your kids it's gonna end up in their mouth at some point...

See here [1]. No FDM printed item is going to be food safe because the layered properties allow the accumulation of bacteria, but PLA is generally non-toxic and will only leach a few "safe in small quantities" chemicals into things it comes in contact with.

[1] http://www.sciencedirect.com/science/article/pii/02786915940...

Could they mean that the material is suitable per 21CFR, but as you state, it would not meet the cleaning requirements. Benign material that could be single use, but a user should not be expected to be able to maintain a clean surface?

I think that's it. Ok for single use or for things that will only come in contact with food that will be cooked (like cookie cutters). Not ok for plates, cups, spatulas, etc. unless you seal them post-print with a food safe coating. I've seen people print both categories of things, though...

As for toys, I don't see a big problem. Even if you eat PLA it won't really do much harm. It's safer than the ABS most toys are made from.

That is true. I only 3d print trinkets for games, so PLA is great. Never have used ABS but the stories about the small has kept me away.

This is hardly surprising though, and it was inevitable that studies started to emerge.

I wrote a blog post about one of the earlier studies in 2013 [see below].

HEPA filters can still be effective - the study doesn't talk about the charge of the UFPs - and the 0.3 micron DoP particles used in HEPA testing are designed to be uncharged. It is common designing filter media to use charged media to attract small particles.

If you were running a 3D printer and concerned about this risk, I would suggest: 1/ Encapsulate the unit. 2/ Wear a mask when the enclosure is open. 3/ Ensure a steady flow of process air when running to prevent accumulation. You could also run at a negative pressure to keep USPs in the enclosure. 4/ Have a filter bank on the air outlet that you change out regularly. Pair a HEPA element with a highly charged media.

Link to 2013 blog post: https://fredlybrand.com/2013/10/19/filtration-in-the-popular...

Is anyone honestly surprised that melting plastic produces toxic particulates/fumes? It should go without saying that one would want these machines enclosed in a ventilated space separated from people.

The most common filament used (PLA) is labeled as non toxic and is made of mostly corn. So yes it would be surprising if that was found to be toxic.

Olive oil is non-toxic, yet if you heat it up too much, you'll get a good dose of acrolein. Just because something is non-toxic does not mean it stays that way if you heat it up.

It depends on how much you heat it up, though. What people are expecting from 3D printers is enough heat to melt the filament but not too much so the filament releases fumes or burns. I'm not saying that's how it works in practice, but I think that's how people see it and why they're surprised.

I don't find this surprising, knowing that fried or baked starchy foods contain at least one carcinogenic element, called acrylamide. Its name sounds nasty and it's provably quite bad, but bread is still one of the corner stones civilization.

We need to quantify these risks, educate builders and make 3D printers as safe as possible, but TFA did give me a slight sense of alarmism.

You should probably review the MSDS for the specific filament you're working with. The first one I looked up [1] contains 1,4-Dioxane [2]. In my experience any time you heat up plastic you're risking exposure to dangerous VOCs. It doesn't matter if most of the molecules were derived from corn.

There are plenty of things labeled "non-toxic" which, when heated, produce toxic fumes. Pretty much every children's toy in existence is toxic to breathe when melted.

[1] https://images.makerbot.com/support/production/SDS-000002ENA...

[2] https://en.wikipedia.org/wiki/1,4-Dioxane

It would seem there's an obvious startup opportunity for filaments that are sold based on cleanliness of product rather than the current marketplace of pigment and cost.

It would seem trivial to distill out the dioxane to an unreasonably low impurity level.

As with pharma, people may not like the cost of synthetic pure bespoke molecules, but some will sell.

The linked MSDS says "dioxane-dione-dimethyl-polymer", which is better known as PLA. It contains dioxane in the same way table salt contains sodium.

You know these labels on cigarettes that state the smoke contains at least 253 toxic chemicals? The smoke from other plants of course contains at least about 200 of the same chemicals.

People think the formaldehyde and all these other chemicals with spooky names were added by the bad cigarette companies..but not so!! Whenever any hydrocarbon is burned, a characteristic distribution of alkanes, alkenes, and alkynes is formed -- with the nastiest and most "radical" chemicals (like the alkynes), appearing with the lowest frequency.

That's why smoking has almost no risk unless repeated. With enough exposure, those minimal alkynes add up. And they destroy DNA and wreak general havok due to their bond energy.

I know I just reiterated your point. Thank you for making the point - smoke is what is a probabilistic distribution of energetic hydrocarbons -- not the plant!!

7,000 chemicals when smoked, 250 of which are harmful, about 70 of which are carcinogens. 599 additives.

That's one of the things that amuses me about these new studies coming out showing smoking marijuana is healthier than nicotine. Yeah that may be so, but that doesn't mean you have to smoke it over like putting it in a tea or something

>> So yes it would be surprising if that was found to be toxic.

Tiny particulates are extremely bad for your lungs when inhaled and it doesn't matter much what they are made of. PLA is not "made of mostly corn" just because the material originates there. If you printed a corn-cob pipe with PLA it would still be plastic.

Once I got my homemade printer working, the first thing I did was build a 100% air sealed enclosure. Melted ABS stinks.

I'd say "stinks" is a bit of an overstatement, unless you have a very sensitive nose.

I've only tried printing with ABS a couple of times, but to me it smells almost exactly like an old photocopier... kind of nostalgic.

All the FDM printers I've been around when they print with ABS end up making my eyes water and give me a headache if I stay near them for more than a minute or two. Possibly the types of ABS used matters for this but I cannot stand near an ABS printer without really noticing it affect me.

PLA has a definite smell but doesn't seem to affect me physically.

What other precautions did you take? Do you turn the air over inside the enclosure or wear a mask?

That's the only precaution. I thought about adding an air filter inside the enclosure but it's overkill. If it's 100% air sealed you don't smell anything while it's printing, and if the printer has finished printing it only takes a few minutes for the smell inside to dissipate enough. I'm sure I'm still getting some exposure but it is a more reasonable amount.

So many man-designed processes produce nano-particles. What natural processes produce them, too? I expect many would. How does nature cope with this?

Soil erosion.

A dry atmosphere is full of nano-particles (and yes they are harmful). But natural particles tend to be larger and more benign than most human made ones.

Nature copes with them the same way it copes with ever other long term problem: it affects individuals, and individuals are all dead at the long term anyway.

Many allergies are due to microparticles - which have individual features that are in the nano realm.

Well, fire is the really obvious one. I'd be very surprised if these put out more than a charcoal grill, much less a campfire. But of course everybody understands the need for proper ventilation when using one of those.

Volcanic ash or, more topically, wildfires are some more good examples.

I'm not sure what you mean by "cope". There's so many things that are nature, and many are entirely unaffected while others are irreparably affected immediately. If you kill off a whole bunch of prey animals all at once you might see a decline in predators, a simultaneous explosion of scavengers, and then a correction as the predators switch to the scavengers as their primary prey. Or you might see a complete ecosystem collapse. Natural systems are so many things from the water cycle to the flora and fauna to the geological systems.

"Cope" here is "how we are not dying from them". Either our bodies have adapted to some nanoparticles (mineral, but not plastic yet), or we're far enough away that the natural particles get absorbed by e.g. ocean waters or soil and don't readily get into our lungs.

Another case, which I totally overlooked, is pollen and dust, which do result in serious negative reactions in humans, quite regularly. This may be not outright deadly but debilitating nevertheless.

Oh, well that's totally different.

Short answer: we don't. You breath enough smoke, it kills you, you die in part because these nano-particles (soot) are clogging up your lungs (Along with the CO2/lack of O2). Carbon particles in your food are carcinogenic and can accumulate in your body. Everything around you is constantly killing you by inches. Your body is pretty good at just ignoring anything that isn't causing immediate system shocks and allowing you to function.

Which is to say, the fact that toxic particles are being spewed by your 3D printer might not be any worse than working next to a grill. That's not to say that either is good, and you should probably minimize contact to prevent accumulation. We haven't "adapted" to natural chemicals, it's more that different chemicals are different degrees of problematic to the body.

Flowering plants and fungi.

All plants and animals, right?

Everything that expires an O2 or a CO2 molecule is producing nanoparticles.

Gasses are not particles.

Isn't something like, a HEPA available in most home air filters, enough to remove most particles?

Should be better than nothing in any case. A guideline mentioning that there should be extraction and proper ventilation near 3D printers - just like in laser printers, soldering stations, etc - would probably be wise.

Yes, let's not forget what laser printers used to spew out. Not sure if that got any better with time?

not the 0.1 size particles they talk about - all but the highest end filters cant remove that small. The best bet is to treat it like a gas and exhaust it.

A 0.3 micron filter will still remove some of the 0.1 micron material. Filtration works in lots of ways. Also, it is likely that the 0.1 micron particles here have a charge. The 0.3 micron test particle used for HEPA evaluation specifically does not have a charge. Using charged filter media is a common method to capture particles that are too small for the holes / pores in the filter element.

How safe are the printed objects?

As safe as any other plastic object made from the same material.

Well for example - printed objects are generally not foodsafe. They are fine to touch and whatnot, but they have too many little niches for food that I would not 3d print myself a spoon..

In case this feels like dejavu for the graybeards


TL;DR: "We were able to measure that there are some toxic particles emanated, but failed to produce any meaningful results about quantity of said particles, or potential danger to consumers, in any way."

To me that says, the quantities they did measure were too low to publish for this alarmist article.

From the article: >all tests indicated at least some level of toxic response (though the toxic response varied by filament type). >The sheer variety of the toxic substances produced by these printers was alarming. No less than 200 different volatile organic compounds (VOCs)

uh huh. The "sheer variety is alarming" line was about where my eyes started to roll, finishing their 135° journey after confirming there are no meaningful stats in this article.

The corrolary of an alarmist report is a dismissive viewpoint. Theres evidence here that FDM printers emit toxic fumes. If you want to believe it's really not so bad, fine, close your made up mind. If there is a chance that filament fumes are harmful, people need to be made aware. They're setting these things up in grade school STEM/STEAM programs; there is potential for real harm here and not pursuing more research is irresponsible.

Agreed 100%. Them blowing this off because they don't feel like it has enough data or teeth I think is a big mistake. There's a distinct possibility that there will be a 3d printer in every classroom/office/home at some point in the future (just like regular printers eventually graduated to be everywhere). Discussing that there might actually be some toxic issues with these printers now can head off a potential future where tons of people including children are getting sick due to something that we didn't test enough or look at with enough skepticism before it gained mass acceptance. It's win/win really... they become safer if they are found to be emiting toxic fumes and people can feel more comfortable having a safer 3d printer around them. I just don't see the downside...

Much like how there are wall paints that have low or no VOCs I imagine we'd want these printers to head in the same direction. Why would anyone say... oh yeah these paints have a variety of VOCs they emit. Sounds alarmist and I don't see enough evidence so we shouldn't explore further and come up with ones that don't potentially harm people.

I'm worried it will impact the cost of filament and printers. Unless there is a cheap way to test and certify filaments. Volumes are still quite low. Regulation risk concentrating filament production into monopolies. Then it will be hard to trust that those certification anyway.

Super valid concern. I imagine it will raise the price considerably and there will be 'safe!' 3d printers that cost a ton because they've been certified and have filaments and suction/vacuum/air pushing systems that have been tested and show no ill effects towards humans. It seems inevitable, but I'd imagine price will come down in a while as things scale.

It also does open the door for someone to come in and make a vent/hood system that works for many types of 3d printers so I guess there's that...

The study specifically targeted ABS which is known to emit fumes but the banner headline is overly alarmist by attacking all of 3d printing even though few use ABS. Working in a Harbor Freight store is a far greater hazard for VOC exposure.

I also tend to be a bit numb to these "x found to be (somewhat) toxic" news items because, well, there are so many of them. That being said it's not very rational to dismiss them entirely either.

Especially in this case, unless you're working in a 3D printing farm it shouldn't be too difficult to work around these problems. If you have a 3D printer on your desk consider moving it to a more isolated, well ventilated location. Maybe buy a couple more fans and an enclosure.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact