Factorio Forums

Once you have productivity research or productivity modules, the LDS casting recipe, available in Foundries, is significantly less efficient than the default LDS recipe in Assemblers.

The reason for this is that for the casting recipe, sure you melt ore into molten form (+50% productivity from Foundries), then use the molten ore to cast LDS (another 50% productivity), but if you go the assembler route, you must also cast the molten metal into steel and copper plates, adding another +50% productivity step onto the process.

With Epic Productivity Modules (+76% Productivity for 4 of them) and let's say +70% research, you have the following:

LDS Casting:
+50+76% Prod from Molten Metal (Foundry) (+146% total)
+50+76+70% Prod from LDS Casting Recipe (Foundry) (+196% total)
Total Productivity in chain: 343%

"Vanilla" LDS:
+50+76% Prod from Molten Metal (Foundry) (+146% total)
+50+76% Prod from Metal plate casting (Foundry) (+146% total)
+76+70% Prod from Assembler LDS (Assembler) (+146% total)
Total Productivity in chain: 438%

This comes out to a big lead for the Assembler recipe, already at Epic level productivity modules, and with additional modules, the gap will become even bigger. Is this intended? Is the LDS casting recipe only meant to be used early on Vulcanus and then be replaced with the Assembler recipe? It feels odd that you get additional productivity gains from an extra, arguably, unnecessary step in the LDS chain (the casting of plates first, rather than casting directly into LDS).

Even with the +50% Productivity bonus of Foundries, they will never catch up. In my stage of the game (from where the numbers come from), the Assembler route uses about half as many ores /s as if I were to use LDS casting.

this heavily neglects the insane crafting speed multiplier of foundries. at base, 1.25 for assembler3, and 4 for foundry. you also skip belt downtime, only limited by molten iron/copper and plastic input. if you value exclusively input to output, the extra step is yes, worth it. but if you value time at all, the foundry is way better.

Yes, I'm aware the Foundry is more UPS efficient because of less machines, less belts, less inserters, but in the way that matters for everything but megabases, the fact that the Assembler route is >50% efficient in resource drain is honestly huge.

If on Vulcanus, only the drain of calcite and coal matters though ?

----

But ignoring this, what's the breakpoint when one should switch ?

Given that you can get iron and copper from lava in the planet where you unlock foundries and in space in an infinite amount, being efficient in ressource drain is not very valuable compared to requiring less machines/footprint to do the same thing imo. And with the way research cost scale, with foundries it's easier to get to the max productivity level ( 300%) compared to assembler.

All direct casting recipes are less efficient in this way than casting plates and then doing regular processing. All of them, some just earlier than others. It’s a choice you can make: Better resource efficiency, or smaller and cleaner designs. And with all the other frankly insane new ways to decrease resource consumption, some people may settle for the latter option in this specific case. Personally, I think it’s balanced.

CyberCider wrote: ↑Sat Nov 16, 2024 2:00 pm All direct casting recipes are less efficient in this way than casting plates and then doing regular processing. All of them, some just earlier than others. It’s a choice you can make: Better resource efficiency, or smaller and cleaner designs. And with all the other frankly insane new ways to decrease resource consumption, some people may settle for the latter option in this specific case. Personally, I think it’s balanced.

Ehm, that's not correct at all.

Using the same research and same modules, smelting 120 Iron Plates using regular Furnaces, you spend 86.9 iron/s to get 120 Iron Plate/s in 20 Epic Furnaces, each covered by 2 epic beacons.

With Foundries, you spend 23.4 Iron/s (and 0.4 Calcite/s) to smelt iron ore into molten iron, which is casted into 120 Iron Plates/s, using 3.3 Furnaces smelting ore and 0.5 Furnaces casting plates, again covered by 2 epic beacons.

It's not even close.

The LDS recipe is the only direct casting recipe that loses productivity.

It's true, and the more productivity you have, the worse it gets. It does however use less plastic, which makes it better on Volcanus.

But more calcite, hmm...

Nice graphic, what did you use to make it ?

Looks like it's from Foreman.

It is from Foreman.

When is resource drain a concern for you? With Big Drills offering effectively a 2x resource multiplier already (and up to a 12.5x multiplier for legendary), and productivity further down the chain being huge either way, how many weeks if not months of 24/7 gameplay are you going to need in order to deplete a single 5M iron deposit right outside your starting area? I'll take the less infrastructure footprint option by far.

Just doing some quick math, a big patch right outside my starting zone has about 450k iron ore in a 5x6 square of a single drill + belt line. With 8% resource drain from legendary and 2.5/s mining speed that means it will take 625 hours of pure 100% throughput mining to deplete. With normal quality drills its still 100 hours. And I dunno about you but my throughput on miners is basically never 100%, it's far below it.

Not saying this isn't an interesting fact but I don't think depleting an ore patch right next to my base in say 500 hours as opposed to 525 hours is something I'm going to lose sleep over.

To me this mostly means that I'll reduce the availability of ores (whether in settings or via mods) until it starts to matter again...
(except probably for those designed to be ~infinite like lava ? which always seems to need calcite anyway ?)

TheToblin wrote: ↑Sat Nov 16, 2024 5:20 pm

CyberCider wrote: ↑Sat Nov 16, 2024 2:00 pm All direct casting recipes are less efficient in this way than casting plates and then doing regular processing. All of them, some just earlier than others. It’s a choice you can make: Better resource efficiency, or smaller and cleaner designs. And with all the other frankly insane new ways to decrease resource consumption, some people may settle for the latter option in this specific case. Personally, I think it’s balanced.

Ehm, that's not correct at all.

Using the same research and same modules, smelting 120 Iron Plates using regular Furnaces, you spend 86.9 iron/s to get 120 Iron Plate/s in 20 Epic Furnaces, each covered by 2 epic beacons.

With Foundries, you spend 23.4 Iron/s (and 0.4 Calcite/s) to smelt iron ore into molten iron, which is casted into 120 Iron Plates/s, using 3.3 Furnaces smelting ore and 0.5 Furnaces casting plates, again covered by 2 epic beacons.

It's not even close.

The LDS recipe is the only direct casting recipe that loses productivity.

No no no that's not what I meant
I was referring specifically to recipes like casting gears, casting cable and casting steel. It always eventually becomes more resource efficient to first use the foundry to cast plates, than craft/smelt those plates into the finished item, instead of casting the item out of molten metal. For steel it requires some research, for cables it takes high quality modules, and for gears it's immediately more efficient, even with common quality tier 1 modules.

Even with only 2 of them ?

CyberCider wrote: ↑Mon Nov 18, 2024 12:32 pm No no no that's not what I meant
I was referring specifically to recipes like casting gears, casting cable and casting steel. It always eventually becomes more resource efficient to first use the foundry to cast plates, than craft/smelt those plates into the finished item, instead of casting the item out of molten metal. For steel it requires some research, for cables it takes high quality modules, and for gears it's immediately more efficient, even with common quality tier 1 modules.

Except apparently not? This is with Epic Productivity modules. (Don't mind the Beacons. It's just my default. The Iron/s is what we're looking for). Foundries are still more efficient. By a hair, in this case, but still more efficient. Same for Steel. The only case where Foundries lose, and lose BIG, is with LDS.

How did you calculate the other direct casting recipes if you got assemblers to win? I can see it happen with legendary modules, but not before that, apparently. (Nope, actually checked. With Legendary Productivity modules, they come out exactly the same at 2.4 iron/s.)

Note that if you want quality LDS, the foundries receipe is absolutely OP, you just need quality plastic and you have quality LDS, and then you can recycle the LDS and you get back the plastic, plus some extra copper and steel of quality leading to free ressources.

Low density structure casting is vastly superior to assembler recipe depending on how you use it ^^

TheToblin wrote: ↑Mon Nov 18, 2024 4:07 pm

Except apparently not? This is with Epic Productivity modules. (Don't mind the Beacons. It's just my default. The Iron/s is what we're looking for). Foundries are still more efficient. By a hair, in this case, but still more efficient. Same for Steel. The only case where Foundries lose, and lose BIG, is with LDS.

How did you calculate the other direct casting recipes if you got assemblers to win? I can see it happen with legendary modules, but not before that, apparently. (Nope, actually checked. With Legendary Productivity modules, they come out exactly the same at 2.4 iron/s.)

Oops, completely my bad! The values used to be different before release, and I sort of just kept them in my mind even after I got the game myself. In that case, your suggestion could definitely be implemented for consistency's sake. I still personally prefer casting, for reasons I already stated, but I can see your reasoning now.

"vastly" is a bit strong word for this

Stargateur wrote: ↑Mon Nov 18, 2024 5:08 pm "vastly" is a bit strong word for this

It might be a touch hyperbole, but 50% is still 50%. It's a lot.