Factorio Forums

What is "Smart Upcycling"?

"Smart Upcycling" describes the following process:

You take a recipe that can be crafted with productivity. Ideally it's a recipe that can reach 300% productivity (maximum possible, with a couple exceptions). You then craft the recipe with as much productivity as possible and recycle it with as much quality as possible.

But here's the difference to standard upcycling: instead of recycling again until it's either gone or legendary, the recycled output is then used again as ingredient for the crafting recipe. When we manage to reach 300% productivity there is no loss during the upcycling process, your input is pretty much losslessly converted into legendary output.

This cannot be done with every recipe. I recommend to do this process on the following recipes:

1. processing units (blue chips): can reach 300% productivity with research lvl 13 and full legendary productivity modules
2. low density structure from the foundry: can reach 300% productivity with research lvl 15 and gives you legendary iron/copper
3. vulcanus science packs
4. supercapacitors
5. Nutrients from Spoilage: See viewtopic.php?t=132520 for more info
6. Rocket Fuel: Gives you legendary solid fuel

And as an honourable mention we can do a kind of better smart upcycling with asteroid reprocessing, but my calculations are not suitable for that.

I do not recommend smart upcycling modules directly. It eats up an enormous amount of ressources while not getting a lot of return.
For the Nutrients from spoilage recipe the calculation setup i provide here is also incorrect because it's a niche case that gives a different amount of return on recycling.
Legendary Biter Eggs and Legendary Bioflux are both very hard to come by.

What do w need calculations for?

Basically to figue out the ratios. How many buildings per recipe per quality, how many recyclers per recipe per quality, how much gain per input?

Here's the link to the calculations:
https://www.desmos.com/calculator/j9sswv5b5f

Sometimes it's more worth it to put quality modules into the building rather than productivity modules.
I created another graph, this time a 3D graph, to determine when that's the case.
Can be found here:
https://www.desmos.com/3d/ulryaowhw4

Lindor wrote: Sat Apr 04, 2026 12:13 am Sometimes it's more worth it to put quality modules into the building rather than productivity modules.
I created another graph, this time a 3D graph, to determine when that's the case.

It's more worth it to use quality modules than productivity when the expected result is quality product, there is no need to use any math for this. Sorry if that sound provocative, when reading your explanation very literally, it's not easy to understand what you tried to optimize/show with your graph. I feel players could be trying to optimize 2 differents things for 2 different reasons , 1) "return on investment on quality module" 2) UPS

In the first case, you are trying to answer the question where do i put my legendary tier 3 quality module first ? and then ? and then ? and so on. This optimization require considering the input ressource quantity and the price of buildings. You are researching high growth rate.
In the second case, you are trying to answer the question If everything is considered "legendary" already, how should my "late game" build look like. This optimization consider the input ressources secondary and discard the price of buildings. You are cutting on everything that isn't strictly necessary, you are searching for stability not growth.

In a way you can't have the cheapest factory that produce the most from the minimum amount of ressources the fastest, that's too many direction where sometimes one is contradicting the other. When you say in your graph that the line represent the break even point, and either way using one or the other module is "better" it's unclear to me what you mean by "better".

processing units (blue chips): can reach 300% productivity with research lvl 13 and full legendary productivity modules

To try and illustrate : This feels like a hybrid approach to me, as if you research more, you can swap some productivity module for quality while sill reaching the 300% cap, which then raises the question of wether or not you want to use speed modules too (imo) for late game build where i feel the fun is to explore the new design made possible with this level of research.

But you already need 13 level of research to get there so you need another method before you can use this. ( and you may find different method past this threshold).

To be fair a 3D graph isn't so common for me, and i could just be missing on things, but i feel to be smart in all situations there would need to be more dimensions, or more graphs ? can't really go wrong with more graphs.