We've just launched our map. Add yourself by clicking here!


Soft-plastic recycling – research and feasibility

This topic contains 49 replies, has 6 voices, and was last updated by  jamyb 1 year ago.

Marc Vruggink marcvdv

Soft-plastic recycling – research and feasibility

16/05/2019 at 11:08

Dear Precious Plastic community,

I am intern in Swiss research institute working on waste management. A waste audit in one of our nepalese case studies highlighted the need for treatment solutions for soft plastics (mainly food wrappers, approximately 10kg/day). For the next two month we will  be conducting a research in Nepal on Small scale soft plastic recycling.

Among my tasks, I’ll be working with a local NGO to achieve and document the construction of a compression machine.

I’ll keep you updated on our progress through this post and I’m happy to hear about your suggestion, experience, comments, etcetera.

Suscribe to the post to stay updated!
All the best,

49 replies
9 subscribers
6 saved
sort on date
24/09/2019 at 17:24

Hey guys, sorry for the delay : life is crazy (in a good way). I felt like I owned this community a feedback from my experience. Make yourself comfortable, this is going to be a long post summarizing our experience.


As part of my studies in environmental engineering, I did an internship in a Swiss research institute (EAWAG; department Sandec). The municipal solid waste management group of this institution previously did a waste audit in a Nepalese boarding school and found out 10kg of plastic were openly burnt daily. This is not only an issue for this boarding school but for the whole Nepal.

Let’s look a little bit closer at the waste generated. Half of has a recycling market (PET bottles, Hard plastics, milk and oil pouches). There were little plastic bags within this boarding school. The majority of the remaining plastics were “soft-plastics”, mainly food (biscuit, instant noodle) packages. How could we deal with these plastics at the community level?

This waste fraction is currently not recycled in Nepal for three main reasons:

– Quality: The amount of printing decreases the commercial value the plastic. Often, the plastics are dirty
Quantity: These wrappers are disposed by diffuse source, all over Nepal. Their weight is little and one has to gather a big volume before a recycler is interested.
Diversity: Of polymers, mixtures (multi-layer) of polymers, mixture of materials (aluminium foil, …) making recycling difficult or impossible.

Machine building

We did a fieldwork of 2 month. This was my first working experience in a developing country. We collaborated with a local NGO (Clean Up Nepal) and, in particular, I worked as a tandem with a local mechanical engineer. We were both new to the field of plastic recycling.

Several solutions are possible for handling such waste (incineration, pyrolysis, use as construction material (ecobricks, roads, …), handicraft) but Precious Plastics was the one we decided to go for, as their option was:
-Scalable to the community
-Generating revenue (multiple objects)
-Maintaining the plastic lifecycle (further recycling) possible

Among the 4 Precious plastic machines, we choose compression as the injection and extrusion required solid fragments, and as the shredder is not designed to shred films. The compression machine allowed us to melt films directly, without preliminary shedding. We built the compression machine for 170 CHF (~170 USD, ~155 Euros; excl. masks, small equipment and workspace rental).

We followed the precious plastics blueprints. Electronic components were bought in Switzerland before leaving. We had to solve SSR-relay over-heating issues that were not anticipated by reading the Precious Plastics documentation. We found a pretty big oven and we decided to not set it on top of the compression frame. We built a higher compression frame (1m20) to not break our backs during operations.

Material characterization

In a first phase, we had to experiment with the plastics. Understanding what worked, what didn’t and why. We found out that most of the wrappers were multilayer films, composed of different plastics and materials. Moreover, most of these items were heavily printed. We developed a procedure allowing to asses the dominating polymer by melting successively at different temperatures.

Few biscuits were PET-based (can’t guarantee this was pure PET). These were not recyclable using compression, as the product was brittle (breaking like a cookie). Other wrappers (biscuits, rice, …) were made of distinct PP and PET layers (clear incomplete melting at 215°C), usually not separable by hand. A notable exception were the instant noddle packages, present all over Asia. These were made of an outer, metallized and heavily printed outer PET-based layer and of an inner, transparent, PP layer. These layers could be separated by hand (time intensive, but possible).  The inner layer melted completely at 215°C and produced a nice and solid product. As no solution was found for the outer layer and as it was time intensive to gather enough inner layers, no further work was done on the noodle packages. Rising awareness about the impossibility to recycle such packages could wake-up policy makers and manufacturers… but I don’t expect a rapid radical change.

Other food packages were mainly PP. They melted completely at 215°C and samples produced nice and solid plastic chips when compressed. Melting was achieved with transparent packages (often wrapping veggies in local markets), toast bread packages, (imported) pasta packages, (imported) korean ramen packages and biscuit packages (I don’t want to blame Nepal, but often imported from India). Melting was achieved with both metallized and non-metallized wrappers.

Recycling process 1.0

Based on the knowledge acquired, the second phase was dedicated to make value out of non-recycled waste. Our attempts were mainly done using toast-breads packages (that could be recycled but have little value due to printing), that were collected in significant amount (7 kg in 2 hours, by 2 persons) thanks to a local waste management company (Doko recyclers).

The process is described in the Figure below. We melted toast bead packages in an antiadhesive baking plate on which we added a tin frame (Second picture, sorry for the quality). Plastics were added by batches of 50-100g (10 to 20 previously cleaned and dried wrappers). Melting temperature was set at 218°C. I recommend carefully checking if the plastic is totally melted (use a wooden spatula) before adding more plastic : sometimes, a kind of air-bubble shielded wrappers from melting, in the middle of this “dough”. After melting sufficient amount, this sticky dough was transferred in a a pre-heated mold before compression.

Our best achievement was a clip board (see third picture below). As it was our first experience, we learned by doing:
– The mold was a bit convex. Once I pressed too hard and the product broke when releasing the pressure.
– Soldering was not continuous, and plastic infiltrated in the jointure.
-Air bubbles were in the product. Maybe due to a “too rapid” pressing.
-Plastic did not spread in the mold corners (Maybe due to too rapid pressing too).

The project was reconducted for a second phase. Three new interns started last week (@Gem.Wrong is one of them)! Their mission is to remediate to the previous issues, to think about other end-products, to produce nicely finished end-products and to optimize the process. I think they are doing well (see 4th picture).

As for me, I’m out. Starting my master thesis on another passionating topic (Organic waste conversion through Black Soldier Fly conversion). In a few months, I’ll almost be able to set-up a complete waste recycling facility :D.

I’ll do my best to answer further questions and comments, but can’t guarantee this.

Thank’s again to all of you for your support!


25/06/2019 at 17:39

The PET sounds challenging. I guess it depends on what the overall problem that is trying to be solved. If the issue is to avoid having these pouches littering the countryside, then the brittle chunks are an improvement. Perhaps the brittleness can be used to crush it and use it as an aggregate in bricks, pavers, retaining blocks, etc. made from other plastics. Kind of like the sand is being used in making plastic roof tiles and bricks.

Interesting that you are slowed in your work by insufficient plastic material.

25/06/2019 at 16:09

@frogfall Apparently this website is blocked from Nepal (For telling the truth?). I could access it through my VPN.

Weekly update # 4 (2/2)
Following @donald ‘s links (thank’s, I should improve my research skills) I found out how difficult it was to recycle PET compared to PP. Moisture makes PET recycling challenging and now it’s monsoon time in Kathmandu. An additional step of drying would therefore be needed to limit PET’s breakability (more info on this nice link). @microtransactions built a vaccum drying system for extruding PET, I don’t know what’s the conclusion of this (see these links: link1 link2

Given the limited time we have, considering drying add a degree of complexity to a process that I have to present as simple and easy to replicate. So I think we won’t focus on recycling these PET wrappers on this project.

Noodle packages
Noodle packages in Nepal consist in fact of two layers. An inner transparent PP layer and an outer PET metallized layer. The two layers can relatively easily be separated by hand (it’s like unsticking a sticker from plastic matterial; depending on the wrapper’s condition).

The inner layer melts nicely at 210°C and produces a solid and resistant chip. The outter one can be melted around 270°C but leads to a brittle result (as with the PET biscuits).

I’m now considering to iron these outer layer together (not sure it would hold, though) to product some kind of canvas sheet. Otherwise, if none of you guys have an easy suggestion to cope with PET brittleness, we’ll focus on other wrappers. Future research could also focus on how to separate these two layers on a bigger scale (out of the scope of our project, but if you have suggestions, I’ll document it in my report).

Other wrappers
Today, we started melting different wrappers to see how they behave at 210°C. Many of them (metallized or not) melted nicely and allow us to press small chips (see third picture below). We are listing the PP wrappers (melting at 210°C) as well as the other/mixed wrappers to ease future replication.

For the time we have left (approximately 2 weeks) we will focus on gathering enough of these PP wrappers and to mold an usefull object (most probably small sheets to make paper boards), as a proof of concepts for the locals. We hope to end our fieldwork with a workshop/demo/discussion day.

As usual, any question, feedback, comment is welcome!

25/06/2019 at 07:55

Weekly update #4

Upgrading the oven
We decided to add the insulation we found and it just works well. While re-insulating, we understood why we were struggling to achieve a higher temperature: the upper side of the oven was in really bad condition (see photo). As a temporary solution, we removed some of the rust and wrapped the top with aluminium foil before covering it with insulation material.

@donald Thank you for the suggestion. If I don’t have time to try it, it will appear as a suggestion in my report. @frogfall : your link doesn’t work, do you mind re-sending it?

@s2019 Thank you also. The second switch idea sounds to be a good alternative. We added a fan on the other side of the heat sink and there’s no longer overheating (see picture 2; the casing is temporary). Concerning the healthy margin, is there a way to estimate the amperage of an oven in advance (I mean, without using a multimeter). By the size of the heating elements, maybe?

We don’t really have a strategy if the lid gets tuck. We have plastic supply issues and didn’t gather enough plastic yet to fill this mold.

Other nice upgrades would be to have a window anda light to follow the melting process without opening the oven.

Melting tests
As a reminder, we are conducting trials on two different packages, which each have three model/brand variations : Transparent biscuit packages and metallized noodle packages.

Biscuit packages
We were expecting PP, but the biscuit packages seem to be made of PET: It melts around 260°C and leads to a brittle product that sinks in water.

The three different packages melt at the same temperature and bind well together, which may allow us to play, to some extent with the different colors. You’ll see hereafter that the small batch we produces isn’t the most aesthetic. However, focussing on a waste managment perspective, I am convinced we can have an impact by finding the appropriate end-product.

Has anyone an experience on recycling PET wrapper? What would be the minimal thickness for a resistant end-product? We have produced small pieces up to 3-4 mm only, they are all pretty easy to break.

(To be continued)

13/06/2019 at 10:41

Well always best to err on the side of caution.

You could of course try to heat it (outside) and see what happens.

But why don’t you use wool?
Could also be the solution for anyone else who has no access to industrial isolation materials (or their labels).

13/06/2019 at 07:52

Picture of the insulation material mentionned above.

31/05/2019 at 08:50

Or place the oven outside before you use it.

Especially when still testing…

My outdoors is part of my workspace, just for this reason…

31/05/2019 at 05:26

Having several temperature measurement locations will help you understand the oven, the temperature gradients, and what the plastic is seeing. They are not that expensive online, but I’m not sure what is easy to get in your location. You could even use a grilling/cooking thermometer if it has a remote probe sensor.

If available a fan in the oven would only help avoid burning the plastic because of uneven temperature.

For the fumes, keeping the oven near the door and a floor fan if available should help.

31/05/2019 at 04:29

Thank you all for your feedback, comments and enthousiasm. We decided to start working only with the lower heating element to limit the risk of burning the top layer of films.

@s2019: Yes, we are adding a thermocouple connected to a PID controller (temporary/testing setting on the picture down there, it seems to work pretty well!). You mentionned “several” controller. We just have one. Would you still recommend adding a fan in the oven?

@donald True, I didn’t plan the timer. I’ll look for one or I’ll use my phone.

Sure, I know the banger of smoke. We’ll experiment gradual temperature increase, work with open doors and windows and use the good masks I have with me. Moreover, as we’re in a co-working space (next neighbors are maybe 10m from the oven, we’ll be really precautious and communicate a lot on our activities.

30/05/2019 at 19:39

The fumes are not always obvious smoke. That container is a small volume.

30/05/2019 at 11:42


I think you answered your own question 🙂

Having the heating on the side would give a more even heating when looked at top to bottom.

i believe most top ‘heaters’ are actually grills, meant to put a brown colour to the top of your ‘turkey’, and as such not very usefull for even heating.

Could differ per over though (but I know mine does this).


I guess the best guess would be to experiment.

If you get even heating (using only the bottom heater) in the upside down oven, great. If it goes sideways, go sideways!


Also scored some noodle packages that might fit your description to do some testing on (like washing it in a blender). If anything usefull comes from this, I’ll let you know.



30/05/2019 at 11:21

Weekly update #2

Our second week is going well!We finally got this big oven; so we won’t have to build it ourself (precious time is saved!). We just removed the circuits we didn’t need (the four electric plates) and are now experimenting with the electronics. We also got the metal and car jack for the compression frame.

I saw in V2 compression video that the oven was flipped by 90°, having the heating elements on the side. Is that better? We planned to use the oven racks to nicely melt our plastic in molds, on grids.

Also, if we don’t rotate the oven, I considered only using the lower heating element to avoid foil burning on the top, as it occured in previous experiments of this community. Any inputs?

We are looking forward to build this oven soon, and start crazy-melting stuff!

27/05/2019 at 07:04


Yes, some of these cheap shredders already overheat just showing it a piece of paper!


Re: dvds – as the plastic is not flexable, it almost needs to be grinded to dust for simple (non-extruder) uses. Most office shredders I know are merely meant to destroy the disc, so it can no longer be used, not to shred them into small pieces, like the paper, but I would love to stand corrected on this.

I’ll try to find a second-hand blender to test cds/dvds also according to the tetrapak setup. Would be easier if the printed foil did not have to be removed…


For starters I’m going to make pave part of my garden with plastic bricks (no, not gonna pave paradise, just the pathways 😉 ).
Why is this important to mention?
Well experiments are fun, but you should always aim for a product you can actually use, instead of crap you’re just gonna melt again !


P.S. mostly in research mode this week.

26/05/2019 at 18:01

If you can find/make a metal fan blade and connect it with a motor on the outside, convection should help quite a bit.

26/05/2019 at 17:17

@marcvdv @donald – Ok – swapping threads…

The major producer of noodles produces 2100 packets a minute; so the 1000 a resource is definitely applicable.Before leaving, I had access to an infrared spectrometer and ran tests on 8 random noodle packaging (export grade) and they all matched with polyproylene on the library.Concerning shredding, we’ll start with scissors. Then, I have different creative ideas on the scale-up phase.

If the base film is polypropylene, then at least you should have fewer problems trying to heat-mould packaging than if the material was was PET-based.

26/05/2019 at 06:08

Weekly update #1

The goal for this week was to meet stakeholders and to find a workspace. We are happy to share that we’ll work at Nepal communitere; where we can use an container and an equipped makerspace.

We aim on gathering all the material to build the compression this week. We’re now facing a challenge: “european sized” ovens are rare in Nepal; so you can’t really find one in scrap (at least we didn’t find one yet). However, what’s common are there smaller ovens (up to 30 litter).

For the first trials, we are considering using a smaller oven. The quantity of plastic to be treated will thus be limited; but I think that’s OK. However, I’m worried about heat uniformity in these kind of ovens; some films might burn. Has anyone experienced working with these small ovens? Would it be difficult to upgrade a small oven to a bigger one (by reusing the heating element and making a new insulated box)?

06/01/2020 at 13:09

By the way, my successors did a promotional video for our project. An educational video will follow. Enjoy.


04/01/2020 at 17:29

@marcvdv yeah you are right, Parle-G is a big brand here for biscuits, but more or less all biscuits are packaged in the same thin foil of plastic kinda way. All those packet wreck much havoc on the environment.

@sowyourseeds nice to see someone working in North East India on plastic recycling. Let me know if you need any help with the machines. I am a machine builder

03/01/2020 at 06:04

Hey Nick,
Happy to share if it can help! In India, I believe one big biscuit brand is parle. We experimented with these wrappers and they melted pretty good (but some emitted fumes, though).

Yes, I am currently in Cameroon and we are planning on feeding chicken with larvae. That’s pretty out of the topic, for more info you can DM me 🙂


02/01/2020 at 12:30

Hi Mark,
Thanks for sharing your process.
Your results & realisations will come in real handy for my recycling project in North East India, where there is similar kind of issues regarding plastic waste & mixed plastic types.
How is the masterthesis going?
Soldier flies converting organic waste, i guess that could be combined with system that feeds chickens a protein-rich diet…

24/09/2019 at 22:09

@marcvdv , thank you for the great summary. Great work, especially given the limited resources you had. Hopefully the new team will connect with the forum and share their experience.

Thanks again and good luck in your studies.

24/09/2019 at 17:28

(Last picture, Product trial made by the new team)

26/06/2019 at 06:54

@s2019 Maybe a few more words on this “plastic supply” issue could be intersting for you. Don’t take me wrong: these wrappers are laying all over the place. We partnered with a local organization to collect this material for us but it took more time than expected (someone threw away the wrappers that were saved for us).

Indeed, it could be interesting to assess the construction material opportunities. Pilot portions of “plastic roads” have been built recently (600kg waste for a 200m street). But if PET is brittle, could we still make durable construction material, without other plastic/crude oil?

25/06/2019 at 21:44


2 weeks is a very short time, so focusing on a working PoC would be key for your research, but I also agree with Stan @s2019 that ‘teasing’ the PET as a potential building material (bricks/tiles, mixed with sand) “if only you had more time…” would also keep the momentum going.


PM me for a more ‘tactical’ approach or if you want me to mail you the @frogfall webpage (I can confirm it works) 😉

25/06/2019 at 13:19

@marcvdv – that’s odd. the link still works for me. It goes to:

25/06/2019 at 08:33

Hi Marc @marcvdv
Good to hear you got the oven working!

These topics have some more technical info about PET:
PET General / Common bottles melting help!
Crystal clear PET plastic ?

14/06/2019 at 13:39

According to this link, there is also a lot of acrylic “wool” used in Nepal – so it would be wise to do a quick “burn test” before using the material as insulation 😉

13/06/2019 at 23:06

The general recommendation on some of the ebay type SSR is to have a healthy margin on amperage. Otherwise some heatsink compound and lots of airflow. If it is overheating mostly during the long initial warmup, you could add a switch so the warmup just goes through the original oven thermostat and once it gets warm you manually switch to the SSR for PID control.

Do you have a strategy on how to separate your double pot mold if it sticks? Maybe weld some more reinforcement around the rims and maybe use the nuts you are using as spacers to also work as jack screw locations. Also make sure the outside of the inner pot is smooth.

If electricity is at a premium, for smaller batches, you may want to build a smaller enclosure that uses the heating elements from the top of the stove and some sheetmetal, insulation and your PID/SSR as a mini oven.

13/06/2019 at 07:51

Weekly update #3 (On week 4)

I didn’t update you last week. I was busy solving the overheating issues of our SSR. I think we managed it now : we added an heat sink and will have an additional fan on our electronic box.

We didn’t melt any plastic yet, but we made slow but steady progress: The compression frame is almost built, our first mold is ready to use (a bit a dodgy setting of two stainless steel kitchen pots that should allow us to make bowls), our supply in plastic should be insured thanks to a local women association of waste pickers and the cleaning procedure is established.

We decided to work on these “thin arrowroot” biscuits and these “waiwai” noodle packagings.

Our oven has poor insulation and is degraded on top (rust and hole); so it’s not really energy efficient. I needed 25min to go from 27°C to 200°C; and the SSR heat sink was really hot after 45 min. Moreover, we never reached a temperature higher than 214°C and I saw several case studies of melting PP at 220°C.

Oven insulation is hard to find in Nepal; but I found this box containing some foam at home. I don’t really know what material it’s made of: Will it be temperature resistant? Can it start burning if I use an inappropriate one? Otherwise, do you guys have a solution for “DIY” insulation of an oven?

30/05/2019 at 19:21


True, I assumed this is the case as they are “now experimenting with the electronics.”, but indeed so is installing a timer….


Once it fumes: experiment failed.

But this is no news to you, right @marcvdv ?

Viewing 30 replies - 1 through 30 (of 48 total)

You must be logged in to reply to this topic.

Support our projects on Patreon so we can keep developing 💪