Challenge: compact plastic production (no bots)

Post pictures and videos of your factories.
If possible, please post also the blueprints/maps of your creations!
For art/design etc. you can go to Fan Art.

vanatteveldt
Filter Inserter
Filter Inserter
Posts: 947
Joined: Wed Nov 25, 2015 11:44 am
Contact:

Challenge: compact plastic production (no bots)

Post by vanatteveldt »

From Blurb's reaction on my question on fluid throughput here (viewtopic.php?f=18&t=38293) I realized that my plastic production setup was pretty stupid, as I had forgotten that plastic produces 2/s.

My general goal is to use productivity everywhere and boost with as many speed beacons as possible shared by as many assemblers as possible, which often means a linear layout without gaps between assemblers 'sandwiched' between rows of beacons, allowing the beacon rows to be shared between racks (such as in my design below).

This means a plastic plant gets 2 prod3 modules and 8 sp3 beacons, for a total speed of 5.874, producing 2*5.875*1.2=14.1 plastic/s. This also means that 3 plants can fill a blue belt, but adds a challenge for a good design, since it outputs too much for a single stack inserter to put on a belt, meaning you need to place two inserters per plant. Since you also have coal and PG input, that gives very little room to spare.

This is my current solution:

Image

It outputs between 2.3 and 2.4k/min, so close to saturating the blue belt.

But I don't like how much space it takes above and below the plants. Especially the space above means that you cannot easily place a mirrored rack above this one, which would have been nice since there is plenty of coal and PG. Also, it doesn't quite produce 2.4k/min, which of course annoys my inner OCD. Finally, the design is more a result of tinkering than of deep thought, so I would think that something better is possible.

So, the challenge is to produce 2.4k/min on a single blue belt with the rack setup displayed above, minimizing room needed primarily "above" the rack and secondarily below the rack. 500 bonus points if it has space for an extra belt coming down (the plastic output of the 'mirorred' plant), but I don't think that's possible (?)
creative mode blueprint
(PS does anyone have a good blueprint setup for measuring throughput?)

NM, found siggyboy's contraption for measuring average throughput (viewtopic.php?f=8&t=29679&start=0#p190249) and combined it with xelephant's counter (viewtopic.php?f=193&t=19825&start=40#p193337. If the counter is correct my production is fine at 2390+ plastic/min.
with counter
Mehve
Filter Inserter
Filter Inserter
Posts: 318
Joined: Sat Aug 06, 2016 9:12 pm
Contact:

Re: Challenge: compact plastic production (no bots)

Post by Mehve »

How critical is that last 0.1k/min? Or getting it from only 3 plants/belt? :)

This is my current bulk plastic setup. Uses four plants/belt, outputs 4.6-4.7k/min across 2 blue belts, all within standard beacon spacing. It's lacking that last 5%, but it's pretty simple, and even has some gaps for sticking power poles for when you're not in creative mode and your adjoining sections aren't using substations :)

Image
vanatteveldt
Filter Inserter
Filter Inserter
Posts: 947
Joined: Wed Nov 25, 2015 11:44 am
Contact:

Re: Challenge: compact plastic production (no bots)

Post by vanatteveldt »

Nice setup, I guess the choice to use only one inserter per plant is reasonable; at least it's more reasonable than having only three plants in a rack, which wastes many more beacons on 'edge effects'.

Since you have quite a bit of overproduction, I would guess that you should be able to get rid of the top and bottom beacons, no?

Edit: Yeah, you can shave off some beacons without harm. I also added some extra inserters where there was space after putting down power poles, which got production up to 2370/s per belt (4.7 ~ 4.8k according to prod screen):
screenshot and blueprint
Hmmm, would be possible to get this output with 7 plants as well?
Mehve
Filter Inserter
Filter Inserter
Posts: 318
Joined: Sat Aug 06, 2016 9:12 pm
Contact:

Re: Challenge: compact plastic production (no bots)

Post by Mehve »

Challenge met! 7 plants, holds 4.7k/min steady, no extension above or below the chem plants. Between the identical beacon footprint and the copious UG use, I'm not sure if there's any savings over the 8 plant version, but it was a fun little exercise. Logically, I know the output splitter at the bottom shouldn't help the overall output, but it seems to visually improve the belt product distribution a little?
Image
Blueprint
vanatteveldt
Filter Inserter
Filter Inserter
Posts: 947
Joined: Wed Nov 25, 2015 11:44 am
Contact:

Re: Challenge: compact plastic production (no bots)

Post by vanatteveldt »

Fantastic, works like a charm, output high 2370's according to the measurement. :). I had to stretch the design one tile to find place for power poles, but the beacon grid gives 1 row tolerance so that's no problem.

You can actually remove the bottom beacons, which drops output to low 2370's, which is pretty much rounding error. Also removing the top 2 beacons drops output to the 2350's, which is still 4.6-4.7 according to the production screen (Sorry, I misaligned the beacons after placing the poles).

You can actually remove the top and bottom beacons and still get output around 2380. So, you've shaved off 2 prod3 modules *and* 4 sp3 modules. Since each modules costs 3000-4000 resources in total, that's a huge saving compared to a couple of underground belts.
screenshot and blueprint with power poles and fewer beacons
User avatar
MeduSalem
Smart Inserter
Smart Inserter
Posts: 1686
Joined: Sun Jun 08, 2014 8:13 pm
Contact:

Re: Challenge: compact plastic production (no bots)

Post by MeduSalem »

Better power pole placement:
plastic.jpg
plastic.jpg (237.73 KiB) Viewed 20660 times
Yeah, I placed 8 chem plants... just for testing the layout...

But I'd do it with Bots anyways... as I'm doing pretty much everything with bots. Belts are somehow messy because the devs refuse to implement certain quality of life things like lane-swappers or smart splitters... and I can't stand looking at all the ugly contraptions anymore that are necessary to achieve that functionality.
Mehve
Filter Inserter
Filter Inserter
Posts: 318
Joined: Sat Aug 06, 2016 9:12 pm
Contact:

Re: Challenge: compact plastic production (no bots)

Post by Mehve »

MeduSalem wrote:But I'd do it with Bots anyways... as I'm doing pretty much everything with bots. Belts are somehow messy because the devs refuse to implement certain quality of life things like lane-swappers or smart splitters... and I can't stand looking at all the ugly contraptions anymore that are necessary to achieve that functionality.
I get that, but for me, bots utterly kill a lot of the fun this game offers. I almost set aside this game after my first kilo/megabase, simply because gameplay had devolved (for me) to just throwing roboports and bots at everything. But there's something immensely satisfying about seeing a massive factory physically moving copious amounts of stuff via belts, physically inputting/outputting via inserters, and sorting/organizing stuff via harnessing the behaviour of seemingly simple devices in concert with each other. It's the reason I'm approaching 600 hours on this game instead of setting it aside back before even reaching 50.

If this game was Tetris, I imagine bots would be the straight block. All bots = all straight blocks. Obviously highly capable and going to make it easier to get the highest levels, but also a little bland in certain respects. These days, I use bots for my shopping mall and border defense, but that's it.
vanatteveldt
Filter Inserter
Filter Inserter
Posts: 947
Joined: Wed Nov 25, 2015 11:44 am
Contact:

Re: Challenge: compact plastic production (no bots)

Post by vanatteveldt »

Mehve wrote:But there's something immensely satisfying about seeing a massive factory physically moving copious amounts of stuff via belts, physically inputting/outputting via inserters, and sorting/organizing stuff via harnessing the behaviour of seemingly simple devices in concert with each other. It's the reason I'm approaching 600 hours on this game instead of setting it aside back before even reaching 50.
Is this what you mean? :-D
immense satisfaction
User avatar
hansinator
Fast Inserter
Fast Inserter
Posts: 160
Joined: Sat Sep 10, 2016 10:42 pm
Contact:

Re: Challenge: compact plastic production (no bots)

Post by hansinator »

Mehve wrote:
MeduSalem wrote:But I'd do it with Bots anyways... as I'm doing pretty much everything with bots. Belts are somehow messy because the devs refuse to implement certain quality of life things like lane-swappers or smart splitters... and I can't stand looking at all the ugly contraptions anymore that are necessary to achieve that functionality.
I get that, but for me, bots utterly kill a lot of the fun this game offers. I almost set aside this game after my first kilo/megabase, simply because gameplay had devolved (for me) to just throwing roboports and bots at everything. But there's something immensely satisfying about seeing a massive factory physically moving copious amounts of stuff via belts, physically inputting/outputting via inserters, and sorting/organizing stuff via harnessing the behaviour of seemingly simple devices in concert with each other. It's the reason I'm approaching 600 hours on this game instead of setting it aside back before even reaching 50.
I decided to not use bots and do everything just with blue belts and I absolutely agree with MeduSalem about belts. However, after the initial frustration I began finding ways around the limitations. Often I end up with complex setups just to make something work at full throughput under all circumstances. In the end I found joy in building complex mechanisms and join them to build even more complex things. I started to not care about too many combinators and began to try to control all aspects of a belt based factory while maximizing resource sharing (smart furnace, train station with filters, shared belts). I find it satisfying to watch the mechanism work as a whole and see how it performs. It has a certain beauty when so many little parts work hand in hand precise down to a single tick.
Blurb
Inserter
Inserter
Posts: 29
Joined: Sun Dec 27, 2015 12:34 pm
Contact:

Re: Challenge: compact plastic production (no bots)

Post by Blurb »

I figured it'd be possible to shave off some space by keeping coal and plastic on the same belt.
With a 20% productivity bonus, 4800 plastic/min requires 2000 coal/min, and it is trivial to throttle the throughput.
Using one red belt (1600 items/min) and a half yellow belt (400 items/min) for feeding, I ensure the input never exceeds this value.

Secondly, I went with fewer beacons due to personal preferences. I initially figured fewer beacons would mean fewer modules and in turn vastly reduced build and power cost (when running, this uses 15.1 MW).
Comparing my build to previously proposed solutions, I see that they are only more expensive if the beacons aren't shared.
Even though I use fewer modules, a smaller portion of them are speed modules (which are somewhat cheaper).

Although the chemical plants have sufficient crafting speed to produce 4896 plastic/min, the actual production (going by P menu) is in the 4.6k/min area.
I suspect the last bit of production is lost on the final 4th belt lane and to poor compression, as there's often a bit of coal left over which prevents the plastic from being dropped off.
Feels like I completely missed the target goal, but a fun challenge none the less.
Blueprint
Attachments
beaconplastic.jpg
beaconplastic.jpg (228.71 KiB) Viewed 20480 times
User avatar
DaveMcW
Smart Inserter
Smart Inserter
Posts: 3717
Joined: Tue May 13, 2014 11:06 am
Contact:

Re: Challenge: compact plastic production (no bots)

Post by DaveMcW »

Blurb wrote:Comparing my build to previously proposed solutions, I see that they are only more expensive if the beacons aren't shared.
Yes, the whole point of the putting beacons on both sides is to share the beacons with other designs.
User avatar
MeduSalem
Smart Inserter
Smart Inserter
Posts: 1686
Joined: Sun Jun 08, 2014 8:13 pm
Contact:

Re: Challenge: compact plastic production (no bots)

Post by MeduSalem »

DaveMcW wrote:
Blurb wrote:Comparing my build to previously proposed solutions, I see that they are only more expensive if the beacons aren't shared.
Yes, the whole point of the putting beacons on both sides is to share the beacons with other designs.
Also having more Beacons than Assemblers is actually better energy-wise if one even cares about that with Beacons. It's weird but that's how it is.

So wherever your assembly line ends... it's better if Assemblers are enclosed by Beacons rather than having Beacons enclosed by Assemblers.
Blurb
Inserter
Inserter
Posts: 29
Joined: Sun Dec 27, 2015 12:34 pm
Contact:

Re: Challenge: compact plastic production (no bots)

Post by Blurb »

MeduSalem wrote:
DaveMcW wrote:
Blurb wrote:Comparing my build to previously proposed solutions, I see that they are only more expensive if the beacons aren't shared.
Yes, the whole point of the putting beacons on both sides is to share the beacons with other designs.
Also having more Beacons than Assemblers is actually better energy-wise if one even cares about that with Beacons. It's weird but that's how it is.

So wherever your assembly line ends... it's better if Assemblers are enclosed by Beacons rather than having Beacons enclosed by Assemblers.
In that case, then Mehve's build should consume less power - which it doesn't.
Going by the electric network information, my setup uses 15.1 MW vs Mehve's 19.7 MW.
As is, this is of course because none of the beacons are shared - but it requires immense expansion left/right to achieve the same power consumption for a given output.
If one were to build 5 of them side by side (totalling 35 chemical plants and 60 beacons, of which 40 are shared), you'd find the total power consumption to be 80 MW, producing 10 blue belts worth of plastic.
So why is it that an allegedly power-efficient setup uses more power (8 MW for each blue belt worth of plastic) compared to my setup (7.5 MW for each blue belt worth of plastic)?

I'm aware that such 'shared beacon' setups become more efficient as they grow larger, but you'd need a setup of 280 chemical plants and 410 beacons before Mehve's design becomes preferable (in terms of power per output).
Am I missing something?
User avatar
DaveMcW
Smart Inserter
Smart Inserter
Posts: 3717
Joined: Tue May 13, 2014 11:06 am
Contact:

Re: Challenge: compact plastic production (no bots)

Post by DaveMcW »

You are not limited to plastic plants, any production line with beacons on both sides can be shared.

You end up with a giant rectangle where every beacon boosts 8 assembling machines, and every assembling machine is boosted by 8 beacons. Except for the edges... but you can reduce that problem by making the rectangle bigger!
Last edited by DaveMcW on Tue Dec 13, 2016 11:28 pm, edited 1 time in total.
User avatar
MeduSalem
Smart Inserter
Smart Inserter
Posts: 1686
Joined: Sun Jun 08, 2014 8:13 pm
Contact:

Re: Challenge: compact plastic production (no bots)

Post by MeduSalem »

Blurb wrote:I'm aware that such 'shared beacon' setups become more efficient as they grow larger, but you'd need a setup of 280 chemical plants and 410 beacons before Mehve's design becomes preferable (in terms of power per output).
Am I missing something?
Yeah there is a break-even point somewhere... of course with a low number of assemblers/chemplants/furnaces (or whatever you put between the Beacons) the efficiency will be off and can't really be compared effectively.

I have done some basic maths on that a year or two ago for regular furnaces and there the point is actually reached quite soon depending on the layout.
DaveMcW wrote:You end up with a giant rectangle where every beacon boosts 8 assembling machines, and every assembling machine is boosted by 8 beacons. Except for the edges... but you can reduce that problem by making the rectangle bigger!
Wouldn't be the most efficient actually be a square because it has the shortest circumfence (meaning less loss through edges) in comparison to the area covered? (excluding the space needed for the Assemblers/Furnaces/Chemplants, only talking about beacons now).

So like a line of 30 Beacons in one direction and 30 rows of beacons in the other direction? (and machines in between each row of course) ... In theory these 900 beacons in a "square-like" arrangement should be more energy efficient because there are only 116 edge beacons than placing 90 beacons in 10 rows which would have 196 edge Beacons.
Last edited by MeduSalem on Tue Dec 13, 2016 11:43 pm, edited 1 time in total.
Blurb
Inserter
Inserter
Posts: 29
Joined: Sun Dec 27, 2015 12:34 pm
Contact:

Re: Challenge: compact plastic production (no bots)

Post by Blurb »

DaveMcW wrote:You are not limited to plastic plants, you can share beacon strips for every product in your factory.
Oh wow, I never would have guessed. Is it necessary to insert rolling eyes, or is my sarcasm sufficiently obvious?

Instead of tracking down various "accepted" builds and piecing them together into a factory, I simply scaled up the design at hand.
Scaling up makes it easier for everyone to understand what's going on, and makes it easier for me to crunch the relevant numbers.
It also makes for a more accurate comparison - ceteris paribus and what not.
I also imagined it would leave fewer openings for pointless nitpicking.
MeduSalem wrote:
Blurb wrote:I'm aware that such 'shared beacon' setups become more efficient as they grow larger, but you'd need a setup of 280 chemical plants and 410 beacons before Mehve's design becomes preferable (in terms of power per output).
Am I missing something?
Yeah there is a break-even point somewhere... of course with a low number of assemblers/chemplants/furnaces (or whatever you put between the Beacons) the efficiency will be off and can't really be compared effectively.

I have done some basic maths on that a year or two ago for regular furnaces and there the point is actually reached quite soon depending on the layout.
Is it possible you can find (and post) the designs and said math?
You're expressing yourself in very vague terms, which doesn't go very far to convince me.
User avatar
MeduSalem
Smart Inserter
Smart Inserter
Posts: 1686
Joined: Sun Jun 08, 2014 8:13 pm
Contact:

Re: Challenge: compact plastic production (no bots)

Post by MeduSalem »

Here... but it is not really documented because I only intended it for my own use so you would have to find your way around that excel sheet... especially if you want to edit something or add a different layout for comparison. :roll:

It considers the edge beacons and edge furnaces and whatnot... or at least I tried to do that... but you have to enter everything manually and to know what you are doing, so nothing is "automated".
Item Productivity Test.xlsx
(66.13 KiB) Downloaded 155 times
And here's one of the layouts I think I calculated and compared in the excel-sheet above. I think it's the second one from the bottom labeled with Furnaces/Beaoncs 4x14 + 5x12 on the second sheet of the file.
Furnaces
The setup is not really that efficient though because it's not that big, but I made it that way to fit it between the Roboports.

The throughput of 8525 Units/min seems to be right because that's what used to be displayed by the game production statistics as well (give or take a little margin) if I remember right, so I know that the fundamental maths of the configuration should be correct.

Well at least I haven't tried to update the excel sheet in a while or improved its usability instead of having the random mess that it looks like... because honestly I stopped giving a damn about doing maths quite a while ago... it's a game and doing everything with spreadsheets really sucks the fun out of it quite fast because you realize that you have efficiency issues here and there and whatnot and that robots are better than belts for high throughput and blah blah blah. It gets frustrating to say at least.

Maybe someone else will write a better calculator in jscript for the web or something, one with graphical representation of what you are doing... but that's not going to be me.

If you don't dare to open the Excel file (which is understandible) I'll try and post some of the maths in text form here. Just let me know in that case.

Also sorry that it is a little bit off-topic though.
vanatteveldt
Filter Inserter
Filter Inserter
Posts: 947
Joined: Wed Nov 25, 2015 11:44 am
Contact:

Re: Challenge: compact plastic production (no bots)

Post by vanatteveldt »

MeduSalem wrote:Well at least I haven't tried to update the excel sheet in a while or improved its usability instead of having the random mess that it looks like... because honestly I stopped giving a damn about doing maths quite a while ago... it's a game and doing everything with spreadsheets really sucks the fun out of it quite fast because you realize that you have efficiency issues here and there and whatnot and that robots are better than belts for high throughput and blah blah blah. It gets frustrating to say at least.
I recently started using the helmod base planning mod, and it's really fantastic. You choose the item to produce, how much of it to produce per minute, and it tells you the input and energy required. You can add beacons and modules, i.e. you can say I need 10k/min green circuits with 4xprod3 and 8 sp3 beacons per ass3, it will tell you how many ass3 you need and how much energy will be consumed in total, and how many materials it needs (cable+iron). You can then click on e.g. the cable and create a production chain: you specify how many modules and beacons per cable plant, and it will tell you how many plants you need to produce the cable required by the circuit plant, how much energy is consumed in total, etc.

https://mods.factorio.com/mods/Helfima/helmod

I'm using it now to plan my 1 rocket per minute base, I was doing everything on a spreadsheet but this is just so much easier...
screenshots
MeduSalem wrote:Also sorry that it is a little bit off-topic though.
:). Well, the original topic was more or less done, I think, unless someone is going to propose a setup that achieves two belts with fewer modules in total. Note that I don't care about energy use, for me what I want to optimize is total # of modules, as these are really the only thing that has significant cost (and a prod3 is something like 20-25% more expensive than sp3). Of course, modules go in beacons and plants, so # modules is similar to total energy consumption, but I would think a setup with more assemblers filled with eff3 modules would have lower total energy per unit produced than a prod3+sp3 setup, right?
User avatar
MeduSalem
Smart Inserter
Smart Inserter
Posts: 1686
Joined: Sun Jun 08, 2014 8:13 pm
Contact:

Re: Challenge: compact plastic production (no bots)

Post by MeduSalem »

vanatteveldt wrote:Of course, modules go in beacons and plants, so # modules is similar to total energy consumption, but I would think a setup with more assemblers filled with eff3 modules would have lower total energy per unit produced than a prod3+sp3 setup, right?
Yeah, filling everything with EM2s is in the end outclassing everything else in energy consumption by magnitudes, especially P3+S3 beaconized setups. But that said if you only put EM2s in everything the layout will explode in size if you want to keep up with the throughput.

So in the end it's a question if you want to sacrifice more space or more energy. Hypothetically both is infinite... so it's more like a matter of taste.

In the excel sheet from above I did some basic comparison for the furnaces:

My PM3+SM3 beaconized setup from the picture above:
  • Units/Minute = 8525
  • Total Power = 97 MW
  • Furnaces = 56
  • Beacons = 60
  • Area = ~1890 tiles
Comparable setup using only EM2s:
  • Units/Minute = 8503
  • Total Power = 9 MW
  • Furnaces = 248
  • Beacons = 0
  • Area = ~5208 tiles
So you see... the PM3+SM3 setup draws more than 10 times the power and only requires 83.3% the amount of Ore resources... but the EM2 setup requires more than 2.5 times the layout space and 5 times the Furnaces.

Approximately... depends on how efficient you are with the layout and how much of the infrastructure you can share between the Furnaces etc.


Actually after playing around for 2-3 years with modules I might even be convinced that a Setup using PM3s in the machines and a very, very specific combination/arrangement of SM3s and EM3s in the Beacons might actually be better than pure SM3s in the Beacons... but it would require so much fine tuning and testing that I never really went down that route. The problem with EMs in Beacons is... that the damn beacons aren't affected by the EMs themselves, otherwise it might turn out in really interesting ways.
vanatteveldt
Filter Inserter
Filter Inserter
Posts: 947
Joined: Wed Nov 25, 2015 11:44 am
Contact:

Re: Challenge: compact plastic production (no bots)

Post by vanatteveldt »

Interesting idea about having EM beacons consume less power themselves.

For me the reduction in materials in the whole chain is the main reason for using modules, and if you have prod modules you simply need speed modules as well... Also, the beacons add an interesting design challenge because all of a sudden you have reason to be space efficient in your designs, and you can't just plunk down assemblers with two rows of belts on both sides anymore :)
Post Reply

Return to “Show your Creations”