"Kovarex Enrichment" needs a rethink -- catalysts!

[Malor]

The Kovarex cycle is a good fundamental idea, but it's really terribly flawed in its implementation. It's designed around the code base and existing art assets, instead of what's reasonable to actually do. By using the same inputs and outputs, but in different amounts, it's extremely difficult to determine what's actually surplus, and the design I've been forced to use is messy and requires a fairly ridiculous amount of additional U-235 to ensure that I'm not dipping into my 'seed corn', as it were. (as in, hundreds of additional bars to fill up sushi belts and then an overflow chest; when the chest is sufficiently full, then start pulling the excess out.) This takes hours to get running, and generates a ridiculous amount of belt traffic. It's just a poor design, one that doesn't fit the workflow for any other system in the game.

Now, I think I understand why you did it that way. The game, as it stands, can only have inputs and outputs for a production device, and you don't want to spend the money, time, and art assets to implement yet another type, especially one so closely related to what you already have. You're dealing with uranium, uranium is purified in a centrifuge, so clearly, the existing centrifuge is where that needs to happen. Given the code as it stands, the art assets available, and the physical reality of what you're trying to model, the existing Kovarex Enrichment isn't a bad solution in theory.... but man, does it ever make a mess trying to implement in actual practice.

Rather, I think a new slot type needs to be added: "Catalyst". A catalyst should be something that's required for a reaction to take place, but which isn't consumed by that reaction. This will allow all kinds of interesting mods, as well, including much more advanced chemistry; even if you guys aren't interested in pursuing the possible designs down that path, I bet your audience would be.

Major logic changes that I know about: devices would need to check if all catalysts are there before starting a reaction (in addition to the inputs and enough room in the output.) Inserters would need to fill catalyst slots first and then input slots. On emptying, they should take only from the Output, ignoring catalyst slots altogether. And the UI would need to be redone for any reaction that requires a catalyst; I have no idea how that code works, so have no idea how hard it would be. Could be easy, could be awful. (edit to add: oh, and then anytime you're dealing with a device's inventory, you'll have to consider the catalyst slot(s). If someone changes a recipe and gets everything dumped into their pockets, for example, the catalyst needs to get dumped, too.)

With catalyst slots, Kovarex enrichment would be specified this way: 40 U-235 as catalyst, 3 U-238 as input, 1 U-235 as output. This would give the exact same actual results as the current method, but it stops the confusion between inputs and outputs, eliminates the ridiculous belt spam, and works more or less like all the rest of the production devices in the game.

I really like the idea of nuclear power, but I've spent hours just getting my fuel going, and I haven't even started on an actual reactor yet. (I'm still not certain that my sushi belt design is even going to work, as I haven't been able to fill it yet!) I don't mind the slow ramp up, that kinda mirrors reality to some degree, but the input/output confusion is a major mess, and I don't see that ever getting better. That part of my factory is just doing too much useless work, unloading and then reloading the exact same stuff into the exact same device.

Catalysts are, I believe, an elegant answer.

[quyxkh]

Factorio's a design-puzzle game at least as much as anything else. There's a pretty awesomely good solution to this one, that has none of the flaws in yours. It's simple, easy to remember, efficient in every sense. I didn't find it, I and lots of people got kinda in the territory, somebody figured out how to count better, somebody else figured out how to do the sequencing (much) better, the result is obvious . . . in hindsight. There aren't many puzzles this game sets that don't have elegant or at least simple solutions, and this is a late-game facility. It's exactly what it should be, one of their best puzzles, with any number of easy and sloppy solutions driving you to think "there's _got_ to be a better way", and a really satisfying better way.

[Ingolifs]

I'm perfectly fine with the sushi belt approach. However I do feel that Kovarex enrichment is perhaps a bit too powerful.

The player goes from obtaining a single u-235 on average every 1400 seconds per centrifuge to obtaining one u-235 every 50 seconds per centrifuge. This is about a 30 times increase in rate and basically results in solving all of your energy needs forever.

[OdinYggd]

Can't really catalyze a nuclear reaction. Kovarex process gives the impression that it is actually a fuel breeding scheme, irradiating the U238 using the U235 in order to trigger a neutron capture and decay chain that transmutes it.

Also my solution for it is very compact and effective. The output chest has a filter inserter that grabs U235 and puts it back in the intput chest until the Input chest has at least 80 U235 in it. I can then feed 2 centrifuges at once from this pair of chests easily enough.

While my solution is using logi bots for the connectivity, belting together several of these centrifuge installations is fairly trivial affair. The output would be a stack inserter from the output chest onto the belt- but wired so that it only runs when the input chest has > 80 U235. Like so the loop always stays closed.


There's nothing really wrong with the Kovarex process other than the time taken to get it started initially, the player just needs to put some thought into how they interact with it in order to keep a stable loop.

[AndrewIRL]

By using the same inputs and outputs, but in different amounts, it's extremely difficult to determine what's actually surplus, and the design I've been forced to use is messy and requires a fairly ridiculous amount of additional U-235 to ensure that I'm not dipping into my 'seed corn', as it were. (as in, hundreds of additional bars to fill up sushi belts and then an overflow chest; when the chest is sufficiently full, then start pulling the excess out.) This takes hours to get running, and generates a ridiculous amount of belt traffic. It's just a poor design, one that doesn't fit the workflow for any other system in the game.

Some guy managed to create a full color movie player- Darude Sandstorm. Quite a lot of timing involved.

We can probably figure out how to recycle 40 U235 without needing to keep a ton of excess on hand.

[BLuehasia]

I am not one for making super 100 ratio setups, i made something on my own without looking at other examples pretty easy.

I think this process is pretty decent took be about 40 hours into the game till i finally was able to get nuclear power up and running without worrying about fuel ever again.

if this set up ever freezes, it means im rolling in good green rock. And just add more storage chest to the system. my motto of game play its ok if a production line stops at a certain point when you got more than you can possible use for the time being

[brunzenstein]

How to use up the "bad green rock" besides creating rocket fuel? - can it be in another useful way be recycled?

[BLuehasia]

How to use up the "bad green rock" besides creating rocket fuel? - can it be in another useful way be recycled?

bad green rock can be turned into ammo

[OdinYggd]

How to use up the "bad green rock" besides creating rocket fuel? - can it be in another useful way be recycled?

Send it back to the input of the Kovarex process.

That's the neat thing about this process- the output and input are the same stuff, all it does is change the ratio between them.

You can loop the process endlessly until all of the dark green U238 is consumed- but you don't want to do this! You need some U238 to make fuel rods, bullets, and cannonballs.

I have my centrifuges set to run until U235 = U238/20. This way I always have supplies of both at a good ratio for making reactor fuel, but can always bleed off one or both components for weapons.

1 U235 + 19 U238 + I forgot how many iron plates = 10 fuel rods

[brunzenstein]

How to use up the "bad green rock" besides creating rocket fuel? - can it be in another useful way be recycled?

Send it back to the input of the Kovarex process.

That's the neat thing about this process- the output and input are the same stuff, all it does is change the ratio between them.

You can loop the process endlessly until all of the dark green U238 is consumed- but you don't want to do this! You need some U238 to make fuel rods, bullets, and cannonballs.

I have my centrifuges set to run until U235 = U238/20. This way I always have supplies of both at a good ratio for making reactor fuel, but can always bleed off one or both components for weapons.

1 U235 + 19 U238 + I forgot how many iron plates = 10 fuel rods

Thanks for the confirmation
I did not know but envisioned this.
It means a dramatic change as one can basically seize mining ore and only recuperate the U238 he has in abundance as the enrichment process provides (as far as I understood) more U238 as it consumes

[Malor]

So, to sum up so far: "I can solve this problem!'

But it shouldn't be this kind of problem. It's a bad design that comes from too narrow a focus on the existing code and existing art assets.

If the code changes just a little, it both makes this process easy, and also makes a lot of other interesting chemical reactions possible.

Edit, much later: in other words, this isn't fundamentally an interesting problem. You're just unloading and loading the exact same stuff back into the same device because that's what the engine code requires, not because that makes any sense at all, or because it leads to an interesting puzzle. It's makework to patch over a code deficiency, not a design that serves the greater purposes of the game.

It's a puzzle that shouldn't exist, an outgrowth of a UI problem, not a design decision.

[Malor]

It means a dramatic change as one can basically seize mining ore and only recuperate the U238 he has in abundance as the enrichment process provides (as far as I understood) more U238 as it consumes

Just FYI, this isn't the case. Once you get it running, the enrichment process turns 3 U-238 into 1 U-235. So there's a steady total loss of uranium ore as you make fuel rods. 1 U-235 and 19 U-238 give you 10 rods. Reprocessing them after use gives you back 6 U-238.

So, ultimately, 10 fuel rods end up costing you 3 (to make the U-235) + 19 (to make the fuel rod) - 6 (recovery), or a total of 16 U-238 to make 10 fuel. Plus, of course, quite a lot of power to process all those things, and a long lead-in time to generate the 'seed corn' of 40 U-235 to start the Kovarex process.

I haven't actually gotten a reactor up yet, so I'm not sure how long a fuel rod lasts. I don't know the total drain on ore over time. But it's quite clear that there is a net drain, that you'll eventually use up your uranium reserves.

[Ingolifs]

So, to sum up so far: "I can solve this problem!'

But it shouldn't be this kind of problem. It's a bad design that comes from too narrow a focus on the existing code and existing art assets.

But it's not a bad problem, though. I had fun trying to come up with a design to breed u-235. It's a departure from the standard x ingredients ->1 product formula in the rest of the game.

[Malor]

It is stupid to unload stuff from a machine, only to immediately load it back in again, over and over and over.

That's makework to bodge past a fundamental UI problem.

[brunzenstein]

It is stupid to unload stuff from a machine, only to immediately load it back in again, over and over and over.

That's makework to bodge past a fundamental UI problem.

true enough - so how to do it the right way?

[Hannu]

In my opinion Kovarex process seems to be quite realistic breeding process. You load some stuff in special nuclear reactor and some part of U238 transforms to fissile isotopes. You can not transform all U238 but you have to take materials out after breeding and refine fissile materials away from U 238 and nuclear waste. Then you make a new batch and run again.

I do not see anything clearly wrong in Kovarex process. It is of course simplified and lack several important things, and use centrifuge instead of special breeder reactor, but it is reasonable simplification for game. But you have to make batches, breed them and refine results. It is easy to make simple combinator system to handle material flows in and out. Or copy ready solutions from internet if you do not want to learn circuits.

Catalyst would not be very interesting as suggested. It would make only building phase somewhat more complicated. Better idea would be that catalyst would get spent slowly and you should make small flow logistics for it. Nuclear power mod uses health to simulate slow usage of fuel elements.

[Taipion]

I haven't actually gotten a reactor up yet, so I'm not sure how long a fuel rod lasts. I don't know the total drain on ore over time. But it's quite clear that there is a net drain, that you'll eventually use up your uranium reserves.

(Damn it, I just wrote up everything and then this stupid forum ate my text when I was adding a picture -.- so skip the math and here's the results)

On the smallest reasonable reactor block of 2*2 yielding 480 MW, a medium starting area field of 100.000 uranium will net you ~158 hours.

A single reactor uses 18 fuel per hour, no matter if it does anything with it, therefore I highly advise using at least 6 reactor blocks or more, and turn them on only all at once and only if you need the heat.

With more than 4 reactors you will approach a better power/fuel ratio and up to 1/3 more yield per fuel, also this does not count in infinite research bonus.

And as this thread is about the Kovarex enrichment process, here is my setup:


As the forum seems to be even unable to handle images lik that, here it is as a link, just to in case.

The requestor chests only request bad uranium, the filter inserters only take out good uranium if there is more than 41 at the very least.
If you just have your first 41, you may use instead just one centrifuge with 4*3 stack inserters, 3 inserters on each side forming a "u" to take it out and hand it right back into the centrifuge, one "u" will do but 4 is a little faster, you can but don't have to place chests in between inserters (like, to take things out or get stuff in) but instead inserters can hand stuff directly to each other.

I too don't like how this process messes up your statistics, and makes it impossible to see how much you actually produced, but then again whenever my 12 reactors use 1 fuel each, it adds 11 instead to the statistics... -.-

A VERY simple solution would be, to make a centrifuge with the same model and slightly different colour and name, that takes 41 good uranium as additional building material, and then does the 3 bad => 1 good without further problems.

Also this forum suckz, seriously, I just tried to upload a screenshot for my setup in .jpg, and it says: "Image file type mismatch: expected extension bmp but extension jpg given."
Then I just saved it as .bmp and instead get the error message: "The extension bmp is not allowed."

[brunzenstein]

I found a very attractive setup - only - I don't understand how it works. And to understand, its what is all about:
https://youtu.be/UcWJH48ZpX0
Probably one of the masters will step in?

[MeduSalem]

I really don't know why people are insisting on using combinators with the Kovarex Enrichment... it's totally do-able without using any combinators at all. The setup from the youtube video above wastes like 50% of the space just for the weird combinator setup.

Here in this thread are a lot of setups that are fine doing without (though some Circuit Network conditions on inserters): viewtopic.php?f=208&t=45528


And if one is abusing a priority setup (where the order in which items are added onto a belt matters) then it's completely do-able without any circuit network stuff just by relying on the good old belt tricks.

[Taipion]

I found a very attractive setup - only - I don't understand how it works. And to understand, its what is all about:
https://youtu.be/UcWJH48ZpX0
Probably one of the masters will step in?

That does not look all too good, and quite slow with re-filling.
You can just take mine from my previous post, it's about similarily space efficient, but much faster,
you can take it as a "block" and paste it besides each other.
Simply wire all the inserters that point towards centrifuges and control them with one signal/setup to be able to quickly set it on/off or automatically on certain conditions like good uranium Vs bad uranium or plain number of good uranium.

[edit:]
Regarding the last post before this one: I can't understand why you would even need complex circuit network setups or "belt tricks" to make a fast and scale-able centrifuge blueprint.