Maximum Botless SPM
TL;DR: Without using bots, the maximum possible SPM, before Lab Productivity research, is somewhere between 'slightly less than 2154.5k' and 'slightly less than 1139.8k' normal packs of each type per minute.So, this is an exercise in theorycrafting, looking at what the maximum possible SPM is if you are not allowed to use logistics bots.
First, the bottleneck: the Cargo Landing Pad. The landing pad is the only way to get items down to the surface from orbit, you can only have one per planet, and, without bots, the only way to get those items out of the pad and into the rest of the factory is inserters. This is the only fundamental bottleneck; every other potential limit can, at least in theory, be worked around by just building more.
Nauvis
Starting with Nauvis:- The landing pad is an 8x8 building, meaning there's room for 32 inserters taking items from it. However, for high throughput, you need a large number of cargo bays to allow multiple pods to land simultaneously; one of these cargo bays needs to connect to the landing pad, which will block 2 of the grids where you would otherwise put inserters. So we're limited to 30 inserters.
- The maximum container-to-container throughput of a fully-upgraded legendary stack inserter is 120 items/second. Of course, just moving items from the landing pad to a chest doesn't help: we need to get them out onto belts, and the container-to-belt throughput is much lower; the wiki doesn't have container-to-belt throughput for Space Age inserters, but by looking at the data available, I was able to determine that the container-to-blue-belt throughput of a fully-upgraded legendary stack inserter would be 64 items/second, and the container-to-green-belt throughput could be as high as 80i/s. However, what we can do is have a chain of inserters moving items from the landing pad and then through a series of chests at 120i/s, and then two inserters moving items from the last chest to belts; by making each chain one side of the landing pad a different length, we can avoid collisions.
- That lets us import a total of 3600i/s to the surface. For the moment, we'll assume all of those are science packs. There are 6 types of science pack which cannot be made on Nauvis, so that's 600/s of each, or 36k SPM.
- However, rather than importing common science packs, we can import legendary ones, which have 6 times the science value per pack.
- And we can use biolabs, which inherently produce double the research per pack, and fill each one with 4 legendary Productivity Module 3s, for an additional +100% productivity.
Vulcanus, Gleba, and Fulgora
All these worlds work the same as Nauvis, with two exceptions:- Each of these worlds produces one of the science packs Nauvis needs to import, so they only have to import 5 types of packs instead of 6
- However, they cannot use biolabs, so instead of that 4-fold multiplier, they only get the benefit of the +50% bonus from a regular lab with 2 legendary Productivity Module 3s
Aquilo
Aquilo presents some major challenges:- Because of the need to defrost the inserters with heat pipes, we can only fit 6 of them per side instead of 8. Additionally, the cargo bay blocks 3 spots: the two it occupies, plus one where, due to the presence of the bay, no heat pipe can get close enough. Thus we only have 21 stack inserters instead of 30.
- However, we can improve that a little by placing a burner inserter in the space next to the cargo bay, since burner devices are immune to freezing. The throughput of a fully-upgraded legendary burner inserter is only 8i/s, but that's better than nothing! (The burner inserter also loses a little bit of throughput due to the need to occasionally feed itself fuel, but if we use a long-burning fuel like nuclear fuel, the time loss is negligible.)
- Worse, while Aquilo can locally produce one of the science packs that Nauvis imports, the lack of local resources means that it needs to import all 6 of the basic science packs that other worlds all make locally, bringing the number of imported science pack types up to 11.
- Finally, the lack of biolabs affects here, too.
Minor Factors
There are several minor factors that slightly reduce the SPM values calculated above, but are negligible overall:- If we want to use foundries on any world but Vulcanus, we need to ship calcite down from space. However, the amount of calcite needed for even a massive base is fairly small, and if necessary this can be avoided entirely by just not using foundries for metal production on other worlds.
- If we want to use fusion power, we will need to deliver to Aquilo the holmium required to produce the fusion power cells, and then, if using them on other worlds, deliver those cells to the surface again. However, fusion cells are very efficient, so the amount of resources required is miniscule.
- The cryogenic science packs created on Aquilo require lithium plates, which in turn require lithium, which in turn require holmium. However, when taking into account the efficiencies possible in each step of this process, the ratio is 67.5 packs to 1 holmium plate delivered, so, again, negligible.
- Agricultural science packs lose their science value as they spoil, meaning more of them will have to be delivered to each other world than will be required of the other science packs. This, too, is negligible; a well-designed Gleba base will produce packs with an initial freshness of nearly 100%, and legendary agricultural science packs have a spoil time of 2.5h so they should lose very little freshness during transport.
The Problem: Promethium Science Upcycling
Producing promethium science packs requires 3 ingredients: promethium asteroid chunks, quantum processors, and biter eggs. Promethium asteroid chunks are collected in space, so they are not an issue. Quantum processors can be produced in space, so again not an issue; it does require lithium plates (which requires more holmium shipped to Aquilo) and fluoroketone (which must be barreled for transport, requiring empty barrels) but the ratios, once multiple steps with productivity bonuses are factored in, are so large as to be negligible.Biter eggs, however, present a problem. They must be produced on Nauvis, using captive biter spawners, which must be fed bioflux delivered from Gleba. Accounting for all productivity bonuses, the ratio is 90 promethium science packs per bioflux, a ratio which would be negligible... except we have so far assumed legendary promethium science packs, which require legendary biter eggs. And unlike the holmium shipped to Aquilo to make cryogenic science and lithium plates, which can be upcycled on Fulgora before being shipped, the only way to make high-quality biter eggs is to upcycle the eggs themselves (or a product made from them); high-quality bioflux does not produce higher-quality or greater numbers of eggs. Upcycling is lossy, meaning the ratio of promethium science packs per bioflux goes down, and since I don't know what the most efficient upcycling setup for biter eggs is, nor just how lossy it is, I cannot calculate the final ratio.
It might well be that it is more efficient to not use legendary promethium science packs at all, instead accepting a lower quality there in order to reduce the amount of bioflux which must be imported to Nauvis to maintain egg production. In the worst case, if we assume normal quality promethium science packs (and thus the negligible 1:90 delivery rate for bioflux), the effective SPM drops to 471.2k for Nauvis, 194.4k for each of Vulcanus, Gleba, and Fulgora, and 85.3k for Aquilo.
