Blue circuits requires a lot of green... GAH, MATHS!

[BitterSwede]

So, I've been considering doing what I do for everything else and use maths to figure out how many circuits I need. I've sorta skipped that part until now, for some undescribable reason. But I have had a serious drain on blues and I wanted to math it out.

But I just looked at the numbers, and I'm not sure if I'm counting right.

1 blue takes 2 red and 20 green and it takes 10 seconds to make. So if I want 1 blue per second, I'll need to feed the line with 20 red and 200 green. That's easy enough.
1 green takes half a second to make, so that's 10 factories for greens on top of the 10 for the blues.
Then we get to reds, which requires 2 green each and takes 6 seconds. Since we need 20 red per blue, that would be 60% of the blue factories, which is 6 for red. To feed those we need an additional 12 greens every 6 seconds, so that's 6 more factories for the greens.
So in total we need 16 green factories, 6 reds and 10 blues to get 1 blue circuit per second, with no spare greens or blues, did I count that right?

So to feed that, we need wires. 4 for each red and 3 for each green. Green and wire has a perfect ratio of 2:3, so we need 50% more wire factories than greens. 24 more factories there.
Red to wire ratio is 6:1, so we only need 1 factory for wires there.

Now we're up to 57 factories to get 1 blue circuit per second.

But wait, we need acid and plastic too?

Reds need 2 plastic, one chem plant gives 2 plastic per second, so that's a 6:1 ratio as well.

Blue needs 5 acid and one chem plant gives 50 acid per second, so that's a 10:1 ratio, but acid is a two-step process, so you need 2 plants.

So in total, I would need 57 factories and 3 chemical plants to make 1 blue circuit per second. And that leaves no spares of either greens or reds, nor any acid or plastic for other processes.

And the worst part is that I was trying to get to 1 High-Tech science pack per second, but if my calculations are correct, I would need 3 blue circuits per second for that. That's a whooping 171 factories and 9 chem plants for the blue circuits alone, then an additional 5 reds and 5 greens for the 15 speed module factories. That adds an exponential number of factories for circuits so big I am having trouble calculating it in my head right now.

And for an even bigger headache, I tried to calculate the raw resources needed as well. I get it to be 72.3 iron plates per second needed. That's 138 miners and 127 furnaces dedicated to iron for blue circuits alone. Copper is even worse, that's 120 plates per second needed, so 229 miners and 210 furnaces dedicated to the copper for the blue circuits alone.

My math must be off somewhere. Blues are needed... a LOT. No other resource - that is required in the quantities blue circuits are needed - take up that much space and that many factories, so where did I count wrong?

And please, don't say I need modules... That just adds even more of every single circuit needed per second, and is already one of the things I need more blues for. Besides, modules kindof even out the numbers a bit when it comes to blues. The more speed mods the blue factories get, the more greens will be needed. The more productivity in the blues, the more blue factories are needed.

[Dprince]

My head hurts

[Radioactive Panda]

Solution. Just build too much.

[BitterSwede]

Originally posted by Radioactive Panda:

Solution. Just build too much.

That's kindof the problem... Until now, I've built too much on either side of the circuit production.
My only option, short of starting over completely again (which I probably will do anyway), is to either move the rest of the factory down the main bus or move circuit production to the end of the bus and have the blues go the opposite direction.

And looking at those numbers, I think option 1 is the easiest. I'm not sure I have that many factories for the entire rest of the base at this point.

[brian_va]

Originally posted by BitterSwede:

My only option, short of starting over completely again (which I probably will do anyway), is to either move the rest of the factory down the main bus or move circuit production to the end of the bus and have the blues go the opposite direction.

consider setting up a smelting depot for iron and copper, shipping them to the greens factory that makes an obscene amount of greens. kick them over to the reds factory next door along with the excess greens and send whatevers left to the blues. finally send the blues to a train station and ship back to your bus. if you cant support the reds, add a few more columns of greens. once that is sorted send to the blues and add however many more columns of greens you need to make them run right. beacons and modules are kinda needed to keep the production/space ratio in check.

[BitterSwede]

Originally posted by brian_va:

Originally posted by BitterSwede:

My only option, short of starting over completely again (which I probably will do anyway), is to either move the rest of the factory down the main bus or move circuit production to the end of the bus and have the blues go the opposite direction.

consider setting up a smelting depot for iron and copper, shipping them to the greens factory that makes an obscene amount of greens. kick them over to the reds factory next door along with the excess greens and send whatevers left to the blues. finally send the blues to a train station and ship back to your bus. if you cant support the reds, add a few more columns of greens. once that is sorted send to the blues and add however many more columns of greens you need to make them run right. beacons and modules are kinda needed to keep the production/space ratio in check.

Yeah... At the current state I don't have access to enough modules for that to be viable, so we're right back at having to put out hundreds of assemblers to make that a reality. And lacking the circuits already, those assemblers surely ain't gonna be Mk3's.

So I will either have to stop research right now to get the beacon/assembler3/module production going to get enough circuits to make it a reality, or I'll have to slowly let science continue and build the hundreds of assemblers needed to do it without modules.

Either option does not sound nice.

So, if I do restart from scratch to do it right from the start, how should I go about it? Just allocate enough space by the circuit factory for hundreds of factories, or skip High-Tech until I have literally hundreds of Mk3 modules of both types ready to go?

[brian_va]

restarting will just delay you further.
leave what you have running. set up an area to make assembler 3's, speed and productivity modules. send half your current modules, or all of them if continued science isn't really necessary for your short term goals. in the same area (or ship them in), build rails, signals, fast and stack inserters, steel chests, power poles, blue belts/undergrounds/splitters (could do bot based instead of belt based, your call) furnaces, beacons, miners, locomotives and cargo wagons. add or subtract depening on what you currenlty make and feel you need and what I have forgotten.

set it up and set it running with high limits on output chests.. go build 2+ iron and 2+ copper depots; make sure they are suitably large deposits; 20 million+ should be a decent target. setup a beaconed smelting station for both iron and copper, account for 8-16 lanes of furnaces each depending on what you think you will want. use a blueprint from the factorio blueprints site if needed. connect them both to your rail network. you should have a decent amount of building materials by now.

then go setup the green/red/blues factory, sending plates from the new arrays and adding in petro gas and coal to make sulfer acid and plastic. as your production capacity increases, your output will increase meaning those modules come quicker. add the modules in as they become available; furnaces should probably be first, then the circuits roughly evenly would be my course of action. once you have the modules you want, turn those assemblers off and dedicate 100% of the blues where you want them.

[Jethu]

Yep, if biters aren't beating down your door, and you don't have anything you need to urgently research, stop your labs and concentrate on making an obscene amount of assemblers/beacons/modules. I do this all the time, as my module setup eats almost all my red and blue circuits, which is quite a bit at my current point in the game. While those are churning away, you can be out getting some outposts setup and by the time you come back you'll have some stuff to work with...rinse...repeat...profit.

Edit: forgot to mention INSERTERS. I've learned to keep at least 200 fast/stack inserters in my mall. Found out the hard way when building a massive steel smelting outpost.

[BitterSwede]

Originally posted by brian_va:

restarting will just delay you further.
leave what you have running. set up an area to make assembler 3's, speed and productivity modules.

[...]

then go setup the green/red/blues factory

I need blues for the modules.

Also, when I say restart, I was planning on saving a lot of time by just leaving what I have, find a nice cluster of ore patches and just build a new base elsewhere on the map.

I tend to use bot-augmented belt-based designs, as I could never figure out how to get bots to deliver stuff quickly enough. I've tried several different methods, and none of them have been able to deliver materials quick enough to a single fully beaconed green circuit factory, let alone an entire base. Not even when there's like 10 tiles between the provider chests and the request chests and more than 200 bots idle in a network consisting of only that factory.

So I only use bots for productions that take more than 1 second (post-beaconing), so that the robots can keep up and do belts for those that are quicker.

I have most of the stuff on that list already being produced, but I still can't see how I would reasonably get away with less than ~250 assemblers for the blue circuits to the high-tech science, which is a gigantic step up from the 20-25 assemblers needed for all pre-blue science AND production.
I'm just really stumped about that very steep increase. The only other thing I know that would require a massive factory like that would be the rocket parts, all of them combined might get to that level. But blues are needed for so many things, and in such high quantity. It feels like a serious design flaw in the game to have it grind to a halt all of a sudden and require you to stop everything you're doing and dedicate 2-3 hours to just getting that one component going.

[BitterSwede]

Originally posted by Jethu:

Edit: forgot to mention INSERTERS. I've learned to keep at least 200 fast/stack inserters in my mall. Found out the hard way when building a massive steel smelting outpost.

Yeah, I keep my mall stocked at all times. 200 is a rather low number for me. But even with that, the entire mall pre-blue takes up about 20 assemblers for the circuits. It's a massive step from that to over 250 assemblers, just a single tier up.

[SeaBee]

Call in the robot army and let the rebuilding begin. Find yourself a clear area away from your current setup and use it to design the production setup the way you want, then blueprint it and go back to your base and get your supplies then start rebuilding. You can let the production of your current setup provide you with materials as you rebuild.

[brian_va]

you dont' have to have all the assemblers and modules at the start. put down what you have and let it crank away making the stuff you need to generate the modules and other stuff. sure, it is less effective at first, but put in a few modules and they start adding up.

[BitterSwede]

Originally posted by SeaBee:

Call in the robot army and let the rebuilding begin. Find yourself a clear area away from your current setup and use it to design the production setup the way you want, then blueprint it and go back to your base and get your supplies then start rebuilding. You can let the production of your current setup provide you with materials as you rebuild.

Probably gonna lay track as I run and set up a supply train to pick it up for me at the old base and concentrate on tapping deposits :)

Thanks for all the tips, I'll try it out and see if I can wrap my head around it.

Good thing I ran into this now, my streaming save is just about to get into the blues, so now I have a few days to test before the next stream.

[BitterSwede]

Oh boy, one thing I did not think about was power. Ran some calculations on a production line of productivity modules, 2 factories for the mk3 need 5 of mk2 and 10 of mk1 and will give one module every 6 seconds or so when fully beaconed.
The entire chain, from miners to finished product, will draw about 420MW when fully moduled.
Sure, it will take some time to get to that stage, but that is still over 10 000 solar panels' worth.
Or just over 70 steam turbines.

For just module production!

It's so odd, because I've been watching 5 or 6 people go from scratch to rockets unmodded, and I've never seen them having to ramp up production that high, so I must be missing something. All of them at least claim to have 1 per second production of all science.

[ZomBeGone]

I tend to take a differant approach. I work hard, fast, and sloppy to get to the requester chest level. Then I rip out pretty much everything from mining to all production, and replace it with a bot driven system using blueprint stamps of small manufacturing units.

I have no buss. Everything is in a sort of grid/matrix of assembers and furnaces. I monitor the longistics menu, and production areas to find the weak links, then add more production as needed. Even the mining uses all bots. Its much more efficient and faster to set up. If I am starved for green boards, I stamp out a bunch of assembers for them as needed. Use trains to deliver the copper and iron plates and plastic in large quantities and let the bots do all the work

By late midgame, I have almost no belts. I never make the red or blue belts as they will be obsoleted by the bots.

There should be no problem with the bots keeping up production if you research all the bot techs. In a coop game we did, we got to well over 1 rocket a minute sustained and almost zero belts

Bot delivery should NOT be a problem.

[Jethu]

Originally posted by BitterSwede:

Oh boy, one thing I did not think about was power. Ran some calculations on a production line of productivity modules, 2 factories for the mk3 need 5 of mk2 and 10 of mk1 and will give one module every 6 seconds or so when fully beaconed.
The entire chain, from miners to finished product, will draw about 420MW when fully moduled.
Sure, it will take some time to get to that stage, but that is still over 10 000 solar panels' worth.
Or just over 70 steam turbines.

For just module production!

It's so odd, because I've been watching 5 or 6 people go from scratch to rockets unmodded, and I've never seen them having to ramp up production that high, so I must be missing something. All of them at least claim to have 1 per second production of all science.

Nuclear power for the win! Expensive to build a setup, but man it can be really powerful. Even OP in my opinion. I'm currently running my entire factory on just 4 nuclear reactors...but I haven't started using electric furnaces and high production bot based builds. In due time I guess...this is my first mega base attempt.