Question about base

[daisyshah]

Hey guys,

I now played 120h on factorio and I'm very happy with it.

I played the "wild" or "spagetti" strategy, I tried the "main Hub" (and ending now in the "wild one") and now I want to ask, what a Base have to consist of.

I think, there are some main points:

1. Several Mining Areas with train stations

• A "shop" Area where all your expansion stuff will be produced (all production and infrastructure) incl. enough storage inside logistic network

• Military area (walls, artillerie ammo, mines, uran ammo, defence towers, light oil für flametowers, atomic bomb,...) and artillerie-trains

• Defence perimeter/Walls itself

• normal Science Area(s) (where the SPU will be "produced") incl. space science

• elektrik power

So, am I right if I design and build area-types 2, 5 and 6 independend und just feed everything with rare material (Oil, iron ore, copper ore, coal, uran, water, stone) by train would be suficient and well scaleable?
Did I forget anything?

[Hannu]

That is one way to make a very well working base.

I personally prefer a base where I have centralized production for raw materials and intermediates and they are transported to science, military stuff of workshop factories. But it is just my aesthetic opinion. There is no technical reasons to do so. It just feel more realistic to have iron smelteries, electronic factories, oil refineries, chemical factories, rocket launchpad, research laboratory complex etc. instead transport raw materials and have whole refining chain in every factory.

[mmmPI]

The division in different blocks you described seems to me logical and functionnal.

If you choose a dedicated location for eletricity, i would advise placing near water if you don't use solar pannel. The quantity of water needed for power increases a lot over the course of the game and if you ever once have a blackout it could be very annoying to restart the whole machine if you planned onto shipping the water with trains since the loading/unloading pump would stop functionning and not restart by themselves.


Also you don't mention furnaces at all. But you wants to move raw ore , so i assume your plan is to have some furnaces in every area that requires them. As Hannu mentionned an alternative with pros and cons would be to have a centralized smelting area. One benefit from gameplay is that if your science is stopped and you need a lot of material in your military area, then the production of iron plate is redirected thanks to the train from science to military, all the furnaces could potentially still function, whereas if some furnaces where split where needed, some wouldn't be functionning, those connected only to science for example. (a bit like a hub that can dispatch more to one area when demand is lower in another). The difficulty to my eyes would then be to have an efficient enough train network for delivery. (avoid jams at the input, and functionning delivery schedule). It also has the effect of centralizing pollution, which you can do in geographic choke point for easier defense but maybe want to avoid near a desert that extend to biters territory.



Depending on the size of the base , size of ore patches , a third alternative could be to have furnaces located near the ore patch, and even why not a green circuit factory if you have rich ore patch of copper and iron next to each other. Then shipping plates, steel, green circuits, potentially gears. It only make sense if the patches of ore are big enough so that you won't have to move furnaces during the game. It makes it so that you only load refined material on trains, a wagon of steel plate or green circuit are a lot more material than a wagon of raw ore. With such setup, you can reduce by a lot the size of trains and potential congestion of train network; you create more pollution at each outpost instead of a central/several key location which spread the need for defenses, but also allows you to take advantage of more surface and trees to dump pollution.


you would be right (in my opinion) to pursue your plan, but it's not the only way , you can adapt to ressources location, ennemy, terrain, busy train network with some experiment along the way

[astroshak]

There are many plans that you can come up with for designing your factory. The question is not “is this good” because they’re all good. The question is “can I get this to work as I want it to” ...

[starlinvf]

There are many plans that you can come up with for designing your factory. The question is not “is this good” because they’re all good. The question is “can I get this to work as I want it to” ...

https://www.youtube.com/watch?v=w5r26fXTbA8


This video sums up the game's concept in way that also explains Factory design, and why its stupidly simple. The short version is "Space, Speed, Modularity, Efficiency", and "anything you can craft, you can automate".

Different design focuses lead to radically different factory designs, and the logistics that feed it. Like how building in bulk allows excellent production scalablity, but drastically increases the complexity of the belt networks to move things around. Localized production has more complicated set up, but you can drastically simplify raw material transport and fine tune production ratios to avoid feed shortages.

Even material processing deals with these same questions. Do you create massive smelting arrays by carting in raw ore on a trains? Or do you run smelters near remote ore deposits, and just cart in the plates?


When I first stated playing, I just kept building Steam engines to add more power over time. But then I became aware of pollution and its correlation to biter attacks.... and got really paranoid about it. Since then I've always done Green Runs, trying to get to Nuclear power on less then 10MW Steam. The way I have to build is radically different then the typical mass spd/prod beacon arrays in most factories. And the nice thing about it, is how the efficiency trade offs are way more tangible. I can't just upscale a factory to solve a problem, as it ends up diverting a bunch of resources to defense. And while the runs are slower (typically 60 hours because I still suck at production balance), I have to squeeze way more overall efficiency out of every design, while still being able to keep up with growing resource demand. I've explored nearly every efficiency question, every AM and Module combination, and many approaches to factory layout... and still not been able to get a "best" answer that isn't just "Hoggish Greedly Mega Factory". I'm making it way too hard on myself, and I'm loving every second of it.