Starter base: What to build and what not to build.

[zOldBulldog]

SOLUTION, based on a combination of the information discussed in this thread and my personal design:

https://factorioprints.com/view/-LBLVHe1Xk2_ghVGn8Se

Full description and usage instructions are included in the FactorioPrints page.

Enjoy.

----

It seems to me that whether we build separate ones or we tear down and upgrade in place, we typically go through 3 bases:

(1) Starter base, a throw-away little base whose main purpose is to provide power and produce the things needed to build the main base.
(2) Main base, typically built around a bus, that carries us to launching the first rocket. This base could be further broken down into:
(2.1) Before solar (and/or nuclear?), bots and electric furnaces, during which building is cumbersome and we are often resource-constrained. Things built in this stage often end up torn down and upgraded later.
(2.2) The main build-up and scaling of the base, during which things are rarely rebuilt.
(3) The Mega base. Where the focus is on either major throughput upgrades or building a separate high-volume production base.

In the very early days of playing Factorio I came across the concept of a throw-away starter base in a KoS tutorial (https://www.youtube.com/watch?v=ouWUxGDNNds around minute 12). I tried playing without it and going to building the main base as quickly as possible, but things were not pretty... much too slow and lots of running around since I like spread out and organized bases. I tried it with the starter base and loved it, usually completing all Red/Green science before even being ready for the next step. I then expanded that little starter base with a few extra things like making red track and steel, resulting in a much more comfortable and enjoyable experience. I ended up using yellow belts/inserters only for the starter base while the main base started at red belt, blue assemblers, blue/red inserters.

So, I am now toying with the idea of designing a slightly beefier starter base that can carry me up to the point where I have the assembly lines for 2.1. I'd have to verify it, but I think that requires a 24-steel furnace smelter each of iron, copper and steel (later recycled into electric smelters for "some" ore), oil processing, plastic, sulfuric acid, batteries, green/red circuits, green/red science, blue science, solar and bot production lines.

My expectation is that (once designed) the time to build such a starter base would be very short and far more than compensated by time saved building the main base to the 2.1 level plus minimize "refactoring" the smelters and production lines for higher capacity. Plus, I hope it will allow postponing building too much capacity until 2.2 where it can be done in a fairly final form quickly with bot assist.

So, I think the following are things that this "beefed up" disposable starter base would produce (none for the main base bus unless specified, most of it is low-speed/volume production and capped at the chests, and the produced items are those that are used extensively in building the main base.

- Wood and Coal-fired Steam power (prioritized to wood).
- Low volume smelting of Iron, Copper, Steel, Brick. For starter red/green science and chests.
- Gears, yellow track (underground and splitters can be made by hand), yellow inserters, Green circuits. For starter red/green science and chests to build starter base and mines.
- Red/Green science.
- Medium and Big electric poles.
- Electric mining drills.
- Lamps
- Red track, underground track, splitters.
- Blue, Red and maybe green inserters.
- Blue assemblers.
- Steel furnaces.
- Landfill
- Rail, train stop, rail signal, rail chain signal. (locomotive and wagons built by hand)
- Pipe, underground pipe, engines, pumps. (red/green wire needed by oil processing built by hand)
- Chemical plant, Storage tanks. (Oil refineries built by hand)
- Grenades (for clearing trees), Ammo (Piercing Ammo if feasible at this stage), Gun turrets, Radar.

I suspect this is probably something that speed runners do (stamp down a blueprint for something like this, build progressively, keep going to the main base).

Questions:

- What do experienced players think of this?

- I am sure there are some flaws in my current thinking, as in maybe missing some needed supply or worse not noticing that I might have a pre-requisite that I didn't take into consideration?

- Has someone already done this? Is there a blueprint or screenshot I can look at for inspiration?

[Frightning]

What your describing reminds of the concept of a BBS: A Bussed Building Store. This concept can take you all the way to Prod/High Tech science level for making most of what you need to build your base. The idea is pretty simple: Make a belt w/ iron+copper plates (1 side for each), and another belt that is gears+green circuits and runs next to plate belt. From there, everything early game can be automated with few exceptions (be sure to group inserters together, and likewise for belts, and assembly machines). Later, you can run two belts on the far side for more advanced items. I found that a belt with pipes+steel and another with batteries+red circuits got almost everything midgame that I wanted to automate. (including things like chem plants and robot frames (grouping engines, electric engines, and robot frames so you can direct feed up that recipe chain helps). I've taken to using that concept to start out in most of my games (once I have power+basic smeltery going).

Since 0.15 research changes, I've come to think that it may be a good idea to start modularizing the base as soon as you can start to build up train infrastructure. The real art there being deciding how to group related processes to keep the number of bases reasonably small while still making it easy to meet throughput demands for all involved items, since you won't have bots until later and a single red, later blue, belt of iron will only take you so far.

[Hedning1390]

If you are building a blueprint to carry you from the start to robots then you need to not only think about what to produce, but in which order to expand. If you naively design a blueprint which is great at the start of the robot stage then you may realize when you start a new game that there is lots of expensive building and routing in that blueprint, which you really can't afford.

Also if you are designing a standardized blueprint you really should not have any sort of bus idea because a bus is inherently inefficient. The benefit of the bus layout is that it is easy to keep organized, but with a blueprint to follow you obviously don't need that, rendering the bus layout useless. So, if you intend to make a standard blueprint to use many times, spend a little extra time creating a good one.

What is most important to get right is the smelting (including steel) and circuits (3 types). If you have those organised then the rest will be a piece of cake.

[zOldBulldog]

I am not 100% certain that I communicated correctly. What I am trying to come up with is a blueprint for (1), the throwaway startup base... and making it beefy enough to provide supplies to support the construction of the (2.1) level of the main base.

The idea is that by the point I reach 2.1 I will be able to use the robots to construct 48x48 solar modules, my full array of smelters (10 copper, 8 iron, 4 steel, 1 stone brick and 1 concrete factory), their supporting mines and train network. Then with that build the rest.

Of course, pretty soon after reaching (2.1) the (1) starter base becomes obsolete and gets wiped.

---

I think Frightning's approach of organizing the (1) starter base will do the trick, although I am still trying to visualize how the assemblers fit around the belts.

It might take some creative work but positioning assemblers I can visualize addressing most of the supplies with the 3 belts.
- iron/copper
- gears/green
- pipes/steel

I had certainly not conceived the possibility of including a mini-oil processing, plastics, red circuits, sulfuric acid, batteries, and robot frames in the startup base as a way to achieve early robots to build the main base. Very intriguing.

[dood]

From my experience, half-assing it won't do.

Going for 1 iron 1 copper belt which you also use to make greens only means you'll have to rip it up and rebuild it sooner rather than later and doing it sooner is slower and more painful.

Spaghetti something together and rush to red belt as soon as possible and use that instead of lingering about with yellow belt, go for 4 iron and copper belts.
That way you have enough capacity to get things done in a reasonable timeframe.
Along the way, you should also secure enough oil, copper and iron to get through science and build defenses as soon as possible because once you run low on iron without any oil science upgrades, it's probably too late and you no longer have the capacity to expand.

Next up is getting oil set up and squeezing out advanced oil processing asap so oil doesn't suck.
Then you get tons of efficiency mods, blue belts and t3 assemblers.
Slap efficiency mods into everything and the unending biter attacks will cease for some room to breathe but your steam power and smelters still suck and create a massive red pollution brick so probably not so next up is getting solar up and replacing your horrible burner smelters with electric ones so all pollution sources are minimized and you don't have to chug coal 24/7.

If you have enough power to support it, you can now expand easily with laser turrets.
Build a train network and welcome to the endgame.

[zOldBulldog]

From my experience, half-assing it won't do.

Going for 1 iron 1 copper belt which you also use to make greens only means you'll have to rip it up and rebuild it sooner rather than later and doing it sooner is slower and more painful.

Spaghetti something together and rush to red belt as soon as possible and use that instead of lingering about with yellow belt, go for 4 iron and copper belts.
That way you have enough capacity to get things done in a reasonable timeframe.
Along the way, you should also secure enough oil, copper and iron to get through science and build defenses as soon as possible because once you run low on iron without any oil science upgrades, it's probably too late and you no longer have the capacity to expand.

Next up is getting oil set up and squeezing out advanced oil processing asap so oil doesn't suck.
Then you get tons of efficiency mods, blue belts and t3 assemblers.
Slap efficiency mods into everything and the unending biter attacks will cease for some room to breathe but your steam power and smelters still suck and create a massive red pollution brick so probably not so next up is getting solar up and replacing your horrible burner smelters with electric ones so all pollution sources are minimized and you don't have to chug coal 24/7.

If you have enough power to support it, you can now expand easily with laser turrets.
Build a train network and welcome to the endgame.

Believe it or not, in that recommendation for "not doing it" you hit one of the reasons why I want to do this. While one important reason is to have the supplies to build exactly what you say (and more) in a faster and easier way, the other reason is to find the sweet spot that minimizes pollution, teardown and rebuild.

My first couple attempts followed something very close to your approach, but it required building a pretty big coal power plant, and I ended up having to tear down and rebuild a bunch of stone smelter lines to replace them with electric. In my latest run I expanded that early minimal red/green micro-base (that also gave me iron/copper plate and green circuits) to provide red belt, steel and one or two other things. It was still slow, but I went straight to red belt and blue assemblers for my "not spaghetti" main bus and main base, was able to go on to build the normal base with oil, plastic, sulfuric acid, batteries, red/green circuits, logistics part of the mall, red/green/blue sciences, normal/electric engines, and electric furnaces... All with about 1/4 of a coal-based power plant, 1 full steel furnace smelter, red belt each of iron/copper and and a tiny portion of the steel smelter. I am now working to make the robots to build solar to support the conversion of the smelters to electric and their expansion to my preferred "saturated everything" level (but not certain if I can do it on the 1/4 of a coal power plant before I setup the first solar array).

So, you see... I already verified that there is not only less manual construction of basic supplies, but also noticeably reduced pollution level and reduced rebuilding (of stuff built before solar and electric furnaces) by using this approach.

Also, if I exclude the time spent thinking/designing, the construction of that mini starter base only took minutes to build, and will take seconds to instruct bots to remove it once no longer needed. The effort and resources are negligible and the benefits already were significant.

The goal now is to fne tune the process so that I can reach bots/solar/electric furnaces the fastest and with the least amount of coal power and coal-fired smelting as possible.

[Lubricus]

Blue science and red belts need's lot's of iron, so I think that should be done on the main factory with proper smelting columns. The bootstrap/starter base could do red+green science and yellow belts, the red+green science is cheap and can quickly be done in a little dinky factory.
Wait until the mega-factory to use electric smelters they are only efficient with beacons and modules.

So I would say the first big shortage is iron for red belts and blue science, so don't use red belts early where it's not needed.
The second big shortage will be modules, and circuits for yellow science and rocket control units

[Serenity]

One thing I did recently was build my starter green circuits, yellow belt, gears and red/green science in an area that would later become smelters.

You don't need a lot of copper or green circuits for quite a while. So I reserved an area for 4 copper smelters and built half of one for my current needs. You can then fit four green circuit assemblers (two are enough to begin with), science and some odds and ends assemblers in the horizontal space of just two smelters. Then it's out of the way of your other building efforts but doesn't need really need its own area. You can easily route metal belts to it just by splitting off belts from the other smelters and the green belt easily connects to the main bus. After that start your bus and your main base - and by the time you need to expand green circuits and copper smelting you can tear down the old stuff.

What you have been describing seems to be more extensive though. The above is really just to get you through the early game. After that I laid out the start of the bus I stamped down my Build Everything factory, green/red science, military science and then the labs. And I ripped it up before I got to robots. It's not a lot so it can be done by hand.

[zOldBulldog]

@Lubricous:

Quite correct on the iron demand issues. But it is still feasible to use red belts from the beginning, as follows:
- Route the iron/copper/steel to starter red/green science 1st, starter base production (including those red belts) 2nd, GC and main busses 3rd.
- Throttle the red belt production appropriately (2-6 stacks is adequate).
- Don't attempt to build the whole bus(es) first. Lay it out, but intentionally build only what is needed to handle the part of the main base you are building.
- Stock up your toolbelt with all the red belts available at the moment you head out from the starter base to work on something (i.e. GCs, oil, whatever), and place all of your remaining belts on the bus on the way back.
- Using this technique I always had enough belts for the work I was doing, and by the time I reached Blue science I had completed the GC iron/copper bus and the iron/copper/steel/GC/RC/plastic/batteries/coal of the main bus... and it was all nicely full before I reached Blue science, allowing me to do all the critical blue researches before I drained the iron.
- No need to wait until the megabase to use electric smelters. If you are able to place your first 48x48 solar module with robots at the end of 2.1 you should have the power to place the first long-term electric iron smelter line, and from there on it is just a matter building more solar and more smelters. Electric furnaces without modules are exactly as productive as steel furnaces, and perfectly paired to red belt. For my main base I will use efficiency modules to reduce pollution (not to increase productivity and pollution), so I won't be upgrading those belts to blue. That means that this way I build most of my main base once and done.
- No worry about the blue circuit/yellow science shortage. By the time I get there I have already scrapped the started base and fully depend on my main base with its 10 lines of copper electric smelters, 8+ of iron, 4+ of steel, plus some odds and ends. In an earlier base those were far more than sufficient. And I have space reserved for further expansion should I need to add additional smelters for my first rocket production.

@Serenity:

Brilliant idea!!!

The area reserved for the smelter arrays is absolutely perfect for the starter mini-base:
- Close to the starter ore fields
- Out of the way of other construction
- Makes for shorter running distances
- After the initial small iron/copper stone-furnace smelter lines for the red/green science and micro-mall, it is easy to add the one steel-furnace smelter line of iron/copper (with the option to add more if needed) that will take you all the way to 2.1.
- Easy to dump all excess iron/copper/etc onto the main bus to stave off adding extra steel-furnace smelter line.

I will totally rip out the old starter base, design and blueprint it this way before I start assembling my full Electric smelter lines. It should make my next run through a lot more pleasant. Thank you very much for the idea.

[Lubricus]

@zOldBulldog
You would be surprised how far you can go with yellow, if you want a quick start don't use red belts where they isn't needed, definitely nowhere before blue science.

[Frightning]

Here is an image of a somewhat stripped down version of an early BBS I have made in my current base:

20180429100517_1.jpg (511.2 KiB) Viewed 63300 times

As you can see, I'm only make yellow belts+some of the types of inserters, along with assembly machines and a few other things (including Gun turrets).
I also put my Red science on the far side since it only needs gears+copper plates, but now that I've done all red science only research, it remains idle, waiting for me to setup dedicated R+G science setup somewhere else (I'd recommend dedicated science block because of throughput needs, that way you can feed that+BBS from dual sided plate belts from your smeltery, and thus won't need red belts early (1 dual-lane yellow belt=1 lane of a red belt). The real point of the BBS isn't throughput, it's automation. By having all the early game basics automated, you can simply come there and grab more as needed and the stock will be replaced while your off building (I limit them to 1 stack, but I could see yellow belts being set to 2-4 stacks instead if u want more stock so you don't have to return to your BBS for more belts as often).

Also this setup doesn't have the other belt(s) yet. I will likely run a pipes+steel belt on the east side if I want to include R+G science level recipes like chem plants and medium/big power poles to facilitate building up my modular midgame base (I will tear each of that base's 'modules' down later once I am ready for endgame beaconized setups, but not before that).

[Hannu]

It seems to me that whether we build separate ones or we tear down and upgrade in place, we typically go through 3 bases:

(1) Starter base, a throw-away little base whose main purpose is to provide power and produce the things needed to build the main base.
(2) Main base, typically built around a bus, that carries us to launching the first rocket. This base could be further broken down into:
(2.1) Before solar (and/or nuclear?), bots and electric furnaces, during which building is cumbersome and we are often resource-constrained. Things built in this stage often end up torn down and upgraded later.
(2.2) The main build-up and scaling of the base, during which things are rarely rebuilt.
(3) The Mega base. Where the focus is on either major throughput upgrades or building a separate high-volume production base.

In my style there is no clear difference between 1 and 2. I may have about half dozen temporary assemblers for red science and belts but then I begin to build bus and whet you call main base. My main base is fast build around simple bus and have lots of hacks I would never make in phase 3. Capacity is increased up to one red belt of iron and copper, or maybe 2 if my initial ore depletes and I have to make first outposts to feed phase 2 base. I run phase 2 base until I get full bot logistics and begin then to build a new level 3 base, which is in most games railworld with different production units separated by hundreds of meters.

- What do experienced players think of this?

Probably it works well like so many styles. In my opinion question is what fits to your personal opinions, if you are not competitive speedrunner. I like that my bases grow and develop somewhat like real industrial plants. Especially if I play with complex mods which gives much side products. In history factories have dumped side products but during developing process have learned to utilize them to get more value or prevent environmental problems (which is unfortunately not modeled in the game).

- I am sure there are some flaws in my current thinking, as in maybe missing some needed supply or worse not noticing that I might have a pre-requisite that I didn't take into consideration?

Isn't it most interesting to build such a base and find it out? It is very difficult to remember all details without actual playing. I did not notice any severe flaws. I would not use red belts except main iron or copper lines which actually need them. However, it is simple to make enough production if you do not want to hassle with different belt or inserter tiers. You do not have to care about pollution at normal biter settings. Simple turrets and walls are enough until Tier 3 base begin to run. Then I enclose the whole area with bot maintained defense line. Lasers are easy but I prefer gun turrets in vanilla games because vanilla logistics are too simple in my opinion and ammunition supply is nice addition to that.

[zOldBulldog]

I implemented the Starter Base using much of what was discussed here plus my personal experience and placed the blueprint along with screenshot, instructions and associated blueprints in a book at factorioprints (https://factorioprints.com/view/-LBLVHe1Xk2_ghVGn8Se).

[Serenity]

Not sure why you are so focused on getting electric furnaces so soon though. They are only useful with modules and beacons. And that layout is different from one using no modules/beacons. If you don't use modules, you might as well use steel furnaces. They have the same speed. And coal isn't an issue if your main power comes from solar and nuclear anyways

Yeah for the megabase go electric. Sure. But the first main base can be coal powered steel furnaces all the way

[zOldBulldog]

Not sure why you are so focused on getting electric furnaces so soon though. They are only useful with modules and beacons. And that layout is different from one using no modules/beacons. If you don't use modules, you might as well use steel furnaces. They have the same speed. And coal isn't an issue if your main power comes from solar and nuclear anyways

Yeah for the megabase go electric. Sure. But the first main base can be coal powered steel furnaces all the way

Pollution. In part because of the bugs, but probably more because I just don't like it.

And yes, I know that without modules/beacons the productivity of electrical is exactly the same as steel furnaces. I also plan on using efficiency modules on the furnaces (again, pollution) not productivity, and my megabase setup will be elsewhere, completely redesigned, and completely independent from this main base. Sooo... once I build with red belt/electric from the start for almost all of main base smelters - exception being the couple steel ones I use early - I won't upgrade them at all.

That is why I want to build them electric/red belt to begin with... build them once, then forget them. I know the logic is a bit unusual, but can you think of anything truly wrong with it? So far most of the recommendations from everyone to *NOT* go electric are based on "there is no short-term advantage" and "that is how we always did things", but as long as I have the resources available... those aren't strong enough reasons to avoid electric, right? Or is there some critical concept that I'm missing that would tilt the scales for the choice?

[Frightning]

Not sure why you are so focused on getting electric furnaces so soon though. They are only useful with modules and beacons. And that layout is different from one using no modules/beacons. If you don't use modules, you might as well use steel furnaces. They have the same speed. And coal isn't an issue if your main power comes from solar and nuclear anyways

Yeah for the megabase go electric. Sure. But the first main base can be coal powered steel furnaces all the way

Pollution. In part because of the bugs, but probably more because I just don't like it.

And yes, I know that without modules/beacons the productivity of electrical is exactly the same as steel furnaces. I also plan on using efficiency modules on the furnaces (again, pollution) not productivity, and my megabase setup will be elsewhere, completely redesigned, and completely independent from this main base. Sooo... once I build with red belt/electric from the start for almost all of main base smelters - exception being the couple steel ones I use early - I won't upgrade them at all.

That is why I want to build them electric/red belt to begin with... build them once, then forget them. I know the logic is a bit unusual, but can you think of anything truly wrong with it? So far most of the recommendations from everyone to *NOT* go electric are based on "there is no short-term advantage" and "that is how we always did things", but as long as I have the resources available... those aren't strong enough reasons to avoid electric, right? Or is there some critical concept that I'm missing that would tilt the scales for the choice?

Regarding Steel v Electric, one thing a LOT of people miss is that Steel furnaces are 100% efficient with fuel usage, whereas Boilers (for steam) are only 50% efficient. This means that using steam to power Electric furnaces actually doubles fuel consumption due to smeltery, and that more or less offsets the pollution reduction of electric v steel. However, if you use even tier 1 efficiency modules in the electric furnaces, then the reduced energy consumption will actually result in a fuel savings (worth if you don't mind the fixed cost of adding modules; which is most easily justified with the relative inexpense of tier 1 modules). Using tier 1 modules gives a 20% fuel saving, using tier 2 raises that to 60% savings, but of course, the tier 2 modules are MUCH more expensive (not to mention, higher tech).

[DerGraue]

Pollution. [...]
Or is there some critical concept that I'm missing that would tilt the scales for the choice?

If you don't like pollution then electric furnaces powered by steam and without efficiency modules are slightly worse then stone and steel furnaces.

Some math: A boiler with 2 steam engines produces 1800 kW and ~27.7 pollution. The electric furnace produces 0.9 pollution and uses 180 kW.
-> so 1 electric furnace produces 2.77 + 0.9 = 3.67 pollution and it also consumes twice the amount of coal compared to the steel furnace.
The steel furnace only produces 3.6 pollution with the same crafting time. The stone furnace produces 1.8 with half the crafting time which equates to the same amount of pollution per plate produced.

It is only slightly worse but it is still worse and therefore not worth the effort.

[zOldBulldog]

@Frightning and DerGrauer:

Correct on all counts, electric smelting is worse if you use it with steam. That is why I use this mini-base to help me on the quickest path to solar (with bots since I hate doing solar by hand), and electric furnaces before building all of my smelter kines?.

The key is to build small but final (reserve the space and build only a small capacity portion of the production assembly lines at this stage) so that 2-3 steel furnace smelter lines carry you up to solar before the electric furnaces. Also, once you have bots it is easy to replace the few steel furnaces with electric.

[bobucles]

It is only slightly worse but it is still worse and therefore not worth the effort.

Two efficiency modules will fix that right up. The value of an efficiency 1 on a busy smelter is about 54kW compared to the 42kW solar average. The cost of Eff1 is very competitive, if not cheaper than solar panels. In addition to saving real estate it also reduces the number of accumulators you need, saving precious petroleum and oil.

Eff1 bases are a bit of a meme but they are definitely a different and easy going style to try. Every single ounce of efficiency helps reduce energy cost, reduces pollution and ultimately hinders the biter threat.

[dood]

If you don't like pollution then electric furnaces powered by steam and without efficiency modules are slightly worse then stone and steel furnaces.

If someone doesn't like pollution, why would they stick with steam and not use efficiency 1 modules?