Factorio Forums

Maximum Efficiency, x1 factory, Wasteless Blue Science packs research[Science Pack 3].

1 SP3 [Science Pack 3, Blue] Requires the following:
- 1 Rocket
- 1 Advanced Circuit
- 1 Smart Inserter
- 1 Steel

Times of production are as follows:

SP3 - 12 seconds
Rocket - 8 seconds
Smart Inserter - 0.5 seconds
Advanced Circuit - 3 seconds

With base 'time unit' of 12 seconds, we can make 1 and a half rocket; therefore for safety [for time being] we will double that time unit to 24 seconds, effectively allowing for the following to be produced in that time per assembler:
SP3 - 2
Rocket - 3
Advanced Circuit - 8
Smart Inserter - 48 - 2 units per second, 24 seconds.

All As this forced out cycle time to 48 seconds we will optimize everything to 48 second cycles for time being to check scalability of project.

In essence - to allow for ALL blue Assembler working at optimal efficiency, and not producing any additional goods we get the 'core' of 48 Assemblers [already a lot, even though we're at stage 1].

48 Blue Assembler
16 Rocket Assembler
6 Advanced Circuit Assemblers.
1 Smart Inserter Assembler

Fine, but neither rockets nor smart inserters are 'valid' raw materials, Let us 'disassemble' a bit Rocket Factory Module [Rocket FM]

---------------------------------------------------------------------------------------
1 Rocket needs the following:
3x Iron Gear
2x Electronic Circuits
5x Iron Plates //good, raw material

As iron gears and electronic circuits operate at 0.5 second cycles these are not of a problem, however :
1 rocket FM needs
3/8th of iron gear per second
1/4th of electronic circuit per second

with 16 assembers we will need:
6 Iron Gears
4 Electronic Circuits

As iron gears will be produced in speed of 0.5, just as electronic circuits we will need:
3 Iron gear assemblers
2 Electronic Circuits assemblers

---------------------------------------------------------------------------------------
Advanced Circuits needs the following:
4x Electronic circuits
4x Copper wires

Again - we will need :
4/3 of Electronic Circuit
4/3 of Copper Wires

with total of 6 assemblers we will need:
8 Electronic Circuits
8 Copper wires
As each operates with 0.5 cycle time we get the following factory numbers
4 Electronic Circuit factories
2 Copper Wire factories [as they are also made in packages of 2]

---------------------------------------------------------------------------------------
Smart Inserters needs the following :
4x Electronic Circuit
and 1x Fast Inserter

as all inserters work at 0.5 speed we will just build them 1:1

Fast inserter needs
2x Electronic Circuit
1x Inserter

Inserters finally need
1x Iron Gear
1x Electronic Circuit
1x Iron Plate

In essence - semi-raw materials are the following needed as all of these are in cycles of 0.5 second we will need double that amount per second i.e.

2 Iron Gear
14 Electronic Circuits
6 Iron Plate
Per second, in essence - we will need
1 Iron Gear assembler
7 Electronic circuit Assemblers

---------------------------------------------------------------------------------------
Iron Gears are quite straightforward
3 Iron Plates used per gear
2 Iron gears produced per second [0.5 cycle time]

These are already calculated in all other systems, with producers listed in specific parts.

---------------------------------------------------------------------------------------
Final Two pieces of this puzzle are Electronic Circuits and Copper Wires; as we will use copper wires in Electronics let us analyze first electronics:
Electronic circuit need the following
3x Copper Wire
1x Iron Plate

We should mind however cycle time is 0.5 second, in essence we will need
6x Copper Wires
2x Iron Plates
per second of operation

now, how many electronic circuit assemblies we need ?
- 7 - Smart Inserters
- 2 - Rockets
- 4 - Advanced Circuits
totalling to 13
and that is a problem.
As Copper Wires produce 2 units per cycle,w ith 0.5 cycle time they produce 8 units per second. In essence - 3 Copper wire factories [that are used exclusively in whole project to electronics production] can supply 2 Electronic Circuit factories; we have 13 electronic circuit facilities needed therefore either we will have not enough copper wires OR too much electronic circuits
Easy way to get rid of this issue is to simply DOUBLE all the rest of project.

In essence - to make 100% operational factory complex without any wasted time or cycles whole project must be doubled except for Copper wires. Totalling to 26 electronic factories, needing 39 copper wire factories.

All together it totals to :
96 Science Pack 3 Assemblers
39 copper wire Factories
32 Rocket Assemblers
26 Electronic Circuit Factories [13 x2]
12 Advanced Circuit Factories
8 Iron gear Factories [4x2]
2 Inserter Factories
2 Fast Inserter Factories
2 Smart Inserter Factories

In essence : 219 assemblers of one predefined level.

Limiting factor for scalability - copper wire factories; all others can be scaled down cutting by half, or introduced 'empty cycles' cutting them even further down, and rounding up to 1 at any given time for your preferable size. In essence -while certainly possible to create operation in such scale it is massive overkill producing 96 science pack in time unit [only 24 seconds], to speed of 4 blue science packs per second; It is useless to say at this point you'd need more likely more than a hundred research lab to fully exploit it.

Problem is - this is only package viable for 'simple' production of one blue pack only; While Green Packs are synchronous with blue packs [2 greens take same time as 1 blue] horror comes with red packs
as SMALLEST amount of blue factories to synchronize with red ones is 5 times this amount [both 96 and 5 are relative primes], putting it to whooping 480 factories needed for blue science packs or if you prefer - 20 science packs per second to allow for equal amount of red and blue packs to get out at any given time space.

If there are any mistakes in this small research feel free to correct me and of course - discuss ;]

I have never encountered to need that much sciece packs. I have 4 assembling machines for science packs 3 (stuffed with productivity modules 3), the whole pre production chain needs less than half a screen of space. and that was enough to feed 10 science labs.

[edit] use productivity modules 3 in your science labs, this sppeeds up your research without needing more packs. Don't see need to exactely syncronize production, why would you want to do that?

I think this is not necessarily the most practical approach to producing blue science packs(since it overproduces), but it is something some players want to achieve: reach the most efficient way of producing something. If that is your goal, this is an excellent analysis. I've been wondering what was needed to reach maximum efficiency to produce science packs.

Gammro wrote:I think this is not necessarily the most practical approach to producing blue science packs(since it overproduces), but it is something some players want to achieve: reach the most efficient way of producing something. If that is your goal, this is an excellent analysis. I've been wondering what was needed to reach maximum efficiency to produce science packs.

Then the calculation would be completely different, because you would use productivity modules in every assembler, or speedmodules in selected assemblers, thus saving a lot of assemblers, I can post you the chain out of my excel-sheet later if you want.

All right, this might be interesting for some of you, one more time about the insane power of productivity modules.
I took the 10 "science packs 3" the OP wanted per second to achieve, even if thats a very, very high number, but nevertheless you will see the differences.

First the number of machines in total and the raw material needed (per second) for producing 10 science packs 3 per second. decimal numbers of assemblers have to be rounded up of course, they will slightly overproduce then, but then stop after their buffer is full. The bottleneck are the last assemblers, calculated for exact output number of the desired end product.

From my sheet I don't get the exact same numbers as you do: (I am pretty sure my numbers are correct, have you checked your calculations?)

I get without productivity modules 3 and with assembling machines 3:

96 Science Pack 3 assemblers (the same)
86 copper wire Factories
64 Rocket Assemblers
52 Electronic Circuit Factories
24 Advanced Circuit Factories
16 Iron gear Factories
4 Inserter Factories
4 Fast Inserter Factories
4 Smart Inserter Factories
and in addition of course:
85 furnaces for Steel
365 Furnaces for copper
578 furnaces for Iron (it's getting ridiculous) ^^
340 iron ore per second
215 copper ore per second

Enhanced with producitvity modules 3 and with assembling machines 3 this breaks down to:

60 Science Pack 3 assemblers (instead of 96)
12,61 copper wire Factories (instead of 86)
25 Rocket Assemblers (instead of 64)
11,37 Electronic Circuit Factories (instead of 52)
9,38 Advanced Circuit Factories (instead of 24)
3,31 Iron gear Factories (instead of 16)
0,61 Inserter Factories (instead of 4)
0,98 Fast Inserter Factories (instead of 4)
1,56 Smart Inserter Factories (instead of 4)
and in addition of course:
40,87 furnaces for Steel (instead of 85)
41,23 Furnaces for copper (instead of 365)
124,17 furnaces for Iron (instead of 578)
73,04 iron ore per second (instead of 340)
24,25 copper ore per second (instead of 215)

Divide this whole chain by 10 and you'll get 1 blue science pack per second achieved with a very sane amount of assemblers, furnaces and space.

Sorry guys, this is useless, unless you didn't validate your calculation with the reality. In this case with factorio.

Before you didn't do that, it is like if mathematicans try to make physics without measuring the reality.

Please be advised - there is no place where I said it is the most efficient way of doing it in general; I said it is the mathematical way of optimizing waste/production [i.e. no 'stale' cycles on any point in production chain].
With your calculations for 10 BSP - [note - standard efficiency I claim is only 4; efficiency of 20 is needed to synchronize it with red science packs]. Point of this 'research' was to calculate exact proportions of assemblers [without additional modifiers and without taking transport into account] to maintain 100% productivity efficiency of all assemblers in every chain. Additionally I did mention it is terribly cost inefficient and terrible overkill that apparently may have been missed

As per validity - if you want I can create small scale 'subchains' scaled down as proof of concept, mind you - while mathematicians DO 'make physics' without 'basis in reality' many of these calculations prove later to be valid 'jump pads' into further research and experiments, that quite often prove said 'mathematical, useless theories'.

P.S. This Productivity enchancers sheets may be of use actually Thanks for sending me notifier of these numbers

P.P.S. This is LEAST practical approach that is still efficient as far as time goes once you set it up; please mind you - research costs would have to go to thousands packages of blue science packs if not hundreds of thousands for it to be anywhere near viable method.

ssilk wrote:Sorry guys, this is useless, unless you didn't validate your calculation with the reality. In this case with factorio.

Before you didn't do that, it is like if mathematicans try to make physics without measuring the reality.

Hm, what do you mean? If I provide the materials per second mentionend, in the calculated assemblers (rounded up) than the desired endproduct will come out by the calculated amount per second. Of course you'll have to look, that the inserters etc are'nt the bottleneck, but where do you see that these calculations are not representing the realitity in factorio?

I did validate these calculations for my production chains ingame for the science packs completely. An assembler speed of 1 produces nearly exactly 2 products per second with 0,5 crafting time. I measured the time for smelting one ore in one electric furnace (smelting speed one) with 1,7 seconds, and the smelting for 1 steel bar with 8,5 seconds.

@BurnHard:
Where? Eh? 578 iron furnaces? Theory and reality? Do I need to say more?

I did validate these calculations ...

If this is so easy, why is it so hard to send some Christmas greetings? Answer: because in reality you have chaotic effects, which can influence your calculations much more, that your might think. I never came from the theoretic output to the real, just because you need to transport the stuff, have hidden storages on the belts/inserters etc.

@Both:

Useless theory.

Ok, more useful would be to answer this question:

What is the best ratio of miners, furnaces, assemblies to gain the production for 1, 2, 3, 4 ... Blue-potion assembly with the most minimal afford of buildings?

Or in other words: how many iron-gear assemblies, miners, ... do I need to build a theoretical (!) constant feed for a blue assembly?

or: to feed 1, 2, 3, 4... blue potion assembly constantly, how many miners, furnaces, assemblies do I need, provided, that the transport is immediately? How many items are used to build the factory and what are the minimums for minimal items or minimal down times?



PS: additional question: how many and in which entity did I have to put modules to get a optimal solution with 2 minimums?

ssilk wrote:@BurnHard:
Where? Eh? 578 iron furnaces? Theory and reality? Do I need to say more?

I did validate these calculations ...

If this is so easy, why is it so hard to send some Christmas greetings? Answer: because in reality you have chaotic effects, which can influence your calculations much more, that your might think. I never came from the theoretic output to the real, just because you need to transport the stuff, have hidden storages on the belts/inserters etc.

@Both:

Useless theory.

Ok, more useful would be to answer this question:

What is the best ratio of miners, furnaces, assemblies to gain the production for 1, 2, 3, 4 ... Blue-potion assembly with the most minimal afford of buildings?

Or in other words: how many iron-gear assemblies, miners, ... do I need to build a theoretical (!) constant feed for a blue assembly?

or: to feed 1, 2, 3, 4... blue potion assembly constantly, how many miners, furnaces, assemblies do I need, provided, that the transport is immediately? How many items are used to build the factory and what are the minimums for minimal items or minimal down times?

Just divide my calculations by 60 (That would be for 1 blue assembler then for all things stuffed with producticity 3) and the other appropiately. eg inserters will wildy overproduce of course (bzw stand still most of the time, but you need at least one assembler of course) and be surprised.

It'l be easier on german:

Das Verhältnis der Montagemaschinen untereinander ist ja fix: zb eine Rakete pro Paket, soundsoviele kabel pro Schaltkreis etc. Selbst wenn die getesteten Geschwindigkeiten nicht 100% (Sie tun es zu 95% mit sicherheit) stimmen, so tut es das Verhältis der montagegeschwindkeiten zu den Produzierten Materialen auf jeden Fall. Alles was entscheidet ist das verhältnis der Maschinen zueinander und das stimmt so 100%. Diese Rieseanzahl von Schmelzen ist ja eben absurd, weil man ja NIE so viele blaue Pakete pro sekunde produziert. Eben 60 Montagemaschinen für paket blau auslasten will. Bau es nach für 6 blaue assembler und die Schmelzenanzahl reduziert sich schon auf 5+5+13 Schmelzen (Gerade ungünstiger Breakpoint, module nicht vergessen) Kupferkabel und Iron gears am besten mit mehreren inserter in kisten zwischenlagern, eine 3ermaschine mit modul hat nen irren durchsatz)

Wie gesagt bau es nach mit einem beliebigen Multiplikator, die Maschinen die du auf ne volle Zahl aufrundest um das Komma wegzubekommen produzieren antürlich nen kleinen Überschuss, aber den verwende ich per Logistiknetz zb dann halt woanders.

Wenn du willst stelle ich dir die exceldatei zur Verfügung per PN, damit du siehst ich stütze mich auf sinnvolle Fakten die auch so der Realität entsprechen. Reichfuchsen musst dich halt selber, das ist nicht wirklich dokumentiert ^^ Hab da einiges an Zeit (und Hirnschmalz) reingesteckt. Die Produktionszeiten sind vom Spiel festgelegt, die Zutaten der Rezepte, die Montagegeschwindigkeiten der Maschinen... Warum sollte man das also nicht auf nackte Zahlen runterbrechen können?