MethodologyContact usLogin
By David Allan and Dr. Dr. Liudmila Ilyukhina
Dr. Liudmila Ilyukhina, CEO of Naukateks AS has broken down 12 key points about mask production in collaboration with Fastmarkets Nonwovens Markets. She has also created four face mask production cost models to demonstrate unexpected costs and other key considerations.
To download the cost models, click here.
At first you think, “It is just three pieces of material – how hard can it be?” Then you start up, and this goes wrong, then that goes wrong…every step in face mask production goes wrong. Even things you have not expected to be difficult, will be difficult.
The worst two things that can go wrong in your new project of producing face masks is that the learning curve takes longer than you expected, and you run out of money before you get certified and can sell any masks – or your masks become so expensive in production cost, that you can’t compete in the market.
Naukateks AS is both a supplier of meltblown nonwovens to other face mask producers, and a face mask producer in our own right. We discovered that our meltblown customers were as surprised as we were at how difficult it can be to produce high-quality face masks.
Ignoring advice often results in frustration – when the mask doesn’t breath well or filtration is not good, people would rather blame meltblown quality, than see if their other layers are too dense, or if they welded holes through the whole mask, destroying filtration. A few customers that did accept advice from us, and tested masks at our lab before sending them to an independent lab for qualification, had much more success in qualifying their masks on the first attempt.
Converting machines may change properties that are important for mask qualification, including breathability and filtration.
And that was a big thing for a producer of meltblown, to send test rolls to a customer and get feedback: “Your meltblown is not good enough.” The meltblown line operators say, “We did test it in the lab. It was perfect. But when the customer produced a mask with it, it suddenly stopped having good breathability.”
And it took time to realize that it’s actually things that happened in mask production that killed breathability or filtration.
The three main reasons for loss of filtration were:
For breathability there are also couple of enemies – compression and stretching. While stretching is somewhat easy to adjust, compression of material between rolls in the machine can be hard to detect.
The answer lies in how you adjust the post tension and compression.
At the dentist, you get a paper to bite and you see the places where pressure is harder. Similarly, there are ways to adjust it and then measure the result, and gradually reduce the compression until it goes wrong. You discover that when compression is not enough, then other things go wrong: it might not weld properly, not be straight, not be properly aligned, or something else.
Changes in key properties of materials can be painful for many – both meltblown producers and mask producers. Those who invested in both in meltblown and mask production, noticed the influence of conversion very fast, but those who only did masks, especially on low-cost mask machines, struggled all the way until they managed to qualify their masks or went bankrupt.
Producers of sophisticated high speed face mask machines had this knowledge from before and their customers had less or no problems in this area.
If you can’t justify installing one, establish a relationship with a vendor who operates a lab.
With an on-site lab you can get the settings on a machine right very quickly.
You make a sample, test it in your own lab, then make the necessary adjustments on the machine and see how that worked. Otherwise, if you have to send samples out to be tested, there is shipping time, and a delay getting the lab to test and send the result back to you, and meanwhile maybe you have been producing masks while you waited for the results, and maybe you get an answer: “Your mask is not good enough”. Or you kept the machine idle for that time.
It is also very helpful, if you are buying meltblown fabric for your mask operation, to buy from a meltblown supplier that has their own lab. This can eliminate delays in getting the meltblown formulation exactly right for your face mask needs.
Running your own test equipment requires high level of competence and dedication. Be prepared to talk a lot to supplier of test equipment, and be prepared for a steep learning curve here too.
Actual production of saleable masks on a face mask machine can be lower than the promised “nameplate” capacity.
One thing that can go wrong is that your throughput on your face mask line can be less than you expected. If the machine vendor promises you can make 300 masks per minute but it is only 250 saleable masks in practice, that is a problem.
If you can make 300 but you have a lot of rejects because of quality problems, the true number is not 300. And that happens very often.
The vendor should be able to guarantee actual throughput of saleable masks, and typical defect rates: how many masks you will be able to put in a box at the end of the run.
The more confident you can be, the faster you can run. If you have to manually check every mask when it comes off the machine, that is going to slow you down. If you have to stop the machine again and again and adjust settings because it is not welding well enough or whatever, that is a problem.
In cheaper versions of converting machines, you will have to watch quality of resulting masks constantly, because those lines are not stable enough and you are going to get a lot of bad welds and earloops falling off. Watch for that, stop the line, and fix it right there, do not wait until you have produced a lot of bad masks.
For example, you will need a lot of hand-holding during your initial trial runs. Also you want 24-7 support in case of subsequent issues, fast spare parts delivery, and hopefully a way for your vendor’s technical staff to observe your machine’s operations remotely.
I highly recommend to book installation supervision from the machine supplier at your site. And they should be there for a long enough period of time – not just 4 hours trial run, but at least full day running, and successful start next morning. Be sure the machine is running the way you want it to be while the supplier’s technician is there.
And then it’s an advantage if you have 24/7 online support from producer. If it’s a high speed machine, then maybe you can have a remote access so they can connect to your machine and see all the challenges even when they are not physically at your plant, although that may not be possible with a simple, basic machine.
For customers without deep knowledge of face masks – which can include government emergency stocks managers – be sure you understand how long your face mask will last in inventory before performance begins to suffer, and how the masks should be stored to maximize lifespan.
If you are making face masks to go into a country’s emergency storage, they will want longer storage times than a typical customer.
Each of the components has to be able to last that long. That includes not only filtration, but elasticity of ear loops, stability of nose foam, all of it.
Shelf-life testing is a specific area, where not all the things are covered by standards. Established producers usually know their materials well and it is easy to estimate shelf life, while newcomers will have to do some accelerated ageing tests, or trust their suppliers on estimated shelf lifetime of components.
Even a thing like allowing the parts and materials to become slightly mis-aligned can cause problems – for example, welds may not hold, or the nose piece might end up in the wrong place.
A mask machine is built to unwind rolls, cut the fabric, weld it, attach the nose piece and the ear loops. All of this can go wrong. If the rolls aren’t unwinding straight, you can get layers that are shifted, or the nose piece won’t fit right.
Usually this you can visually control for this on cheap machines. For good high-speed machines it’s all included in the technology. But welding is always challenging – with the two bad ends being “not enough” and “too much”. While “not enough” is easy visible, the “too much” is not – and you can weld holes in your meltblown so the filtration suffers.
Companies that make diapers or other hygiene products already had the knowledge about ultrasound welding, from the start. Big producers had a lot of collaboration from the mask machine builders, who have among other ultrasound welding experts. For newcomer companies, and machines sold without service agreement, they may have to take the whole steep learning curve alone.
Packaging can cost you a lot of money. If your machine quality is low, and you end up checking every mask before packaging, it will add a lot of labor cost.
Package design. If you don’t have a customer from before that defined packaging needs, it may take a lot of efforts to find your perfect packing way: you get customers with a variety of needs in terms of, for example, “Can I have a box with 20 pieces, not 50? Can you make a box that opens on the side? Can you package the masks first in plastic and then put them into the box? Can you print my logo on the box?”
If you get more and more orders involving cardboard printing and different designs, you can end up with a variety of different boxes in storage and maybe you cannot sell them all.
Manufacturing or packaging defects can result in unsaleable or unusable masks, generating lower sales and return expenses as well as damage to your reputation.
Quality control is a whole other subject. I can’t really give a full recipe on how to do it here, but I can maybe give some highlights on what people should look when establishing quality control routines and so on. To start, you can split it into three main groups. There are mechanical failures, such as welding, or cutting the shape so that you can visually see it is the wrong shape for the mask. There can be performance failures, where for some reason it doesn’t filter or breath properly – this can only be discovered in the laboratory. And then there are cleanness issues.
Achieving the cleanest level is hard. Sometimes it is easier to sterilize the masks.
Just put them in plastic and put them into a gamma ray machine and then they are cleaned and sealed.
In our production it’s a clean room. You have hair covers, face masks, lab coats. Clean pants, not street pants like jeans, whatever, that you just bicycled in with. We use clean room shoes, not shoe covers but clean room shoes and we put them on before entering the production area. We have a changing room before you get into the machine room. It does not have to be an accredited clean room, with top-level air filtration and so on, but it has to be very clean, with air clean enough to secure the necessary cleanness level. You may also need to install UV-lamps to disinfect surfaces before production start/ at the beginning of each shift.
Certification is important. It is seldom customers will buy any masks before seeing certificates.
Different countries use different standards. In the US you have ASTM and FDA requirements; the EU operates with EN and ISO standards. Many other countries have their own solution, the most famous being the Chinese KN95 type of mask. You need to understand which standard you are going to fulfill before choosing raw materials and setting up production.
And then it depends also a lot if you go for industrial customers, government customers, medical customers and so on. They may have different requirements. Not all PPEs have to be medical devices – if a mask is for an environment with dust, that is different from medical use.
In the US there are different classifications of mask. The N in N95 means it can be used if no oil-based particulates are present. An R means it resists oil-based particulates for up to eight hours, and a P means it is oil proof. In Europe you have the FFP 2 and FFP 3 standards for respirators.
At the beginning of the pandemic the customers weren’t sure which standards they needed to use, so the requirements could change over time as they learned what they needed.
One of the interesting things we learned is that N95 masks are so much easier to make than FFP 2 masks, because the N95 is tested only with salt aerosols, while an FFP2 is tested with salt and oil both.
Then in the UK they want a fit test, to test how it fits against the face.
Our biggest surprise was that Oil Load test from ISO EN 149 for FFP2 and FFP3 mask. It is hidden so well inside the standard, that you don’t really expect your mask to be frozen, tinned up, and then sprayed with oil for an hour – all this before they test filtration levels, and it still has to be better than 1% for FFP3 masks (6% for FFP2 masks). You only realize there is something wrong when you get results about a failed attempt from an accredited laboratory.
Early on, a big challenge was that customers often had no idea which kind of mask they wanted, and why. Did they want square three-ply masks or masks shaped to the face, meaning surgical masks or respirators? Only a few knew the difference between “filter” and “breath-out-vent”.
We want to accent two types of masks. One will protect the wearer – a respirator. And another is designed to protect others – a surgical or three-ply square mask. Respirators sit tight on your face, so leakage on sides is minimal – all the air you breath in comes through the filter layer. Surgical masks are not even meant to protect the wearer – their task is to collect your droplets when you breath out and this way protect others. Of course, surgical masks protect you somehow, but nobody really measures how exactly.
Another important thing: Respirators may have a vent to make breathing out easier and thus improving comfort. The vent is only working in one direction – it will let your breath to go freely out without filtration, and it is blocked when you breath in. In the pandemic, when both doctor and patient can be infected, you need a respirator without a vent; this way it protects both you and others.
You can still see at many Covid-19 test stations that health workers wear surgical masks instead of respirators. Those workers are testing hundreds and hundreds potentially sick people every day, and they should first of all protect themselves. We see improvement two years after the pandemic started, but at the start when we tried to talk about it, people people were not listening.
We didn’t really have competence on mask production in Europe at first, since most face masks sold here were made in Asia. All the brave newcomers had to take a steep, expensive learning curve. Now, we have knowledge on mask production in almost every country. We hope that end users, large companies and Governments will appreciate the efforts that people made in establishing mask production in their countries, and support them with orders. This will allow producers to move to a new stage: optimized, economically efficient production, competitive with Asian price levels. Placing the orders outside your country will let all the expensive learning disappear from your country; the interpreters who put their soul into the process will get demotivated, and hardly be willing to help out the next time new challenges arrive.