Page 1 of 1
Does barreling make any sense now?
Posted: Sat May 06, 2017 4:39 am
by Mendel
Is there a situation in which you decide to go for barreling liquids instead of using pipes and liquid wagons? I can´t seem to think of any where this would make sense now. If you have a design where this brings an advantage, please let me know.
I haven´t run any numbers or anything, just seems like barreling right now is just an unneccessary hassle. Of course I´m not obliged to use it and having that option is not harmful in any way. Just wondering if people actually use it.
Re: Does barreling make any sense now?
Posted: Sat May 06, 2017 6:23 am
by Ingolifs
My oil setup is on the other side of my base from my processing unit factory, and far away from things that use lubricant, so barrelling (as well as use of the logistic system) comes in quite handy.
Re: Does barreling make any sense now?
Posted: Sat May 06, 2017 6:26 am
by Mendel
Please be more specific. How is it better than just putting lubricant in a liquid wagon?
Re: Does barreling make any sense now?
Posted: Sat May 06, 2017 6:33 am
by Lav
When you have a large factory, weaving a belt through it is generally easier than a rail track.
Re: Does barreling make any sense now?
Posted: Sat May 06, 2017 6:52 am
by Mendel
ah, so you would barrel without actually putting the barrels in a cargo wagon. Didn´t even think of that for some reason
So I guess the throughput of belts is better than throughput of pipes? By how much?
This offsets the hassle of actually having to build barreling and unbarreling assemblers?
Re: Does barreling make any sense now?
Posted: Sat May 06, 2017 7:41 am
by Lav
Don't remember the numbers, but pipe throughput sucks at mid-to-high distances. There are relevant threads on this.
UPD. Yellow belt can provide 800 * 250 = 200K LpM (Liquid per Minute). A single offshore pump can fill a pipe fully and supply 20 boilers, and those consume 60 per second each, that's 20 * 60 * 60 = 72K LpM. So even at pipe's peak performance a yellow belt is almost three times better (disregarding the cost of all those barrels
).
Re: hassle. A single unenhanced assembler-2 fills 45 barrels per minute - that's 11,250 liquid per minute.
Now, for consumption. The most liquid-demanding recipe is the production of sulfuric acid, which consumes 7500 water per minute. Most other recipes need a lot less:
1. Express belt and splitter assemblers require 1800 lubricant per minute
2. Express underground belt assembler requires 3600 lubricant per minute.
3. Oil refinery consumes 1200 crude oil per minute.
So one barreling and one unbarreling assembler can supply quite a few factories easily. Add another barrel-making assembler next to barreling one, and only allow it to make barrels if there are no empty barrels incoming, and that's it.
Re: Does barreling make any sense now?
Posted: Sat May 06, 2017 9:13 am
by BlakeMW
Lav wrote:Don't remember the numbers, but pipe throughput sucks at mid-to-high distances. There are relevant threads on this.
The numbers is that pipe throughput sucks at extreme to ultra extreme distances, unless you use repeater pumps every 1000 tiles (500 rail pieces) or so. Train throughput also starts to suck over extreme distances unless you add more trains to the route. I believe that pipes are cheaper at any distance because repeater pumps are so much cheaper than trains and you also avoid all barreling, barrelling, station and junction infrastructure. The only real downside of pipes is that biters are more likely to nom through them and you may not get a warning that has happened (unless you setup a circuit network alarm that liquid has stopped flowing).
Mendel wrote:
So I guess the throughput of belts is better than throughput of pipes? By how much?
The throughput of belts is waaay better because each barrel holds a ridiculous amount of liquid, but you need a return route (meaning, 2 belts), and you need about 4 Assembler 3's doing barrelling (at each end) to match the throughput of a single pipe. So if you're trying to move more liquid than two pipes can you'll have a honking great array of barreling/unbarrelling assemblers at each end and I cannot imagine what you would've gained by doing that.
In my opinion barrels are best for low throughput (using logistic bots, or for example adding a single sulfuric acid + empty barrel to a uranium train), and pipes are best for high throughput. Trains are good if you already have a rail going someplace, especially if that route is not congested. I often lay pipeline next to rail to keep oil trains (+junctions+stations) off the rail network.
Re: Does barreling make any sense now?
Posted: Sat May 06, 2017 9:39 am
by alexzzzz
Since bots can carry barrels, you don't have to lay belts or pipes or railtracks from your oil processing site to the customers that don't need large amounts of liquids.