Fast Discharge Ore Factory

[rlerner]

Totally expandable to as many furnaces as you want. This one has 16 total. It makes use of the inserter bonus to quickly transfer any queued stacks out of the furnace and into a chest, which further buffers the end product. Three fast inserters pull out of the chest, per furnace, to allow less furnaces to have higher throughput for load spikes. If constructed along the factory's main iron/copper supply, this also has no corner penalty. The basic inserters are more than enough for the chests, but one may desire to use fast inserters here too if demand spikes too often.

At the end of the belt is two splitters to help funnel the lanes into respective lanes to cope with parts where the balance on one side of the belt is uneven.

The use of filter inserters on the ore-input portion may also be of use to keep out any accidental inclusions (stone hiding amidst an iron ore field), etc.

Seems the earlier this is built, the better it can fill up before usage exceeds its capacity. This is not an ideal ore factory for those with a "lean manufacturing" mindset, making the least amount of product needed to run a factory.

[elkar]

Buffering like this works only up to a point where you have so many furnaces that they will flood your express belt to max. capacity.

[rlerner]

That's the point of this.

[Zourin]

Buffering like this works only up to a point where you have so many furnaces that they will flood your express belt to max. capacity.

I would beg to correct the flaw in logic. Output buffering like this only would only work until demand exceeds the production of the furnaces. Furnaces, while 'scaleable' are still significantly space consuming.

This is a (while admittedly very clean) just a glorious way of buffering metal. My only gripe is the increased volume compared to a fairly standard dual-central layout that would have a 5 block narrower profile. Smelting facilities are already pretty bulky to begin with, so I'm not fond of exacerbating the problem.

using a 2x7 inline buffering system (splitter, gap, inserter, chest, inserter, belt, splitter, with a parallel straight belt) may be more efficient, especially if the primary lines are split multiple times or have more than three inserters feeding off them. Using an inline buffer past half of the material draw is better, as then later factories along the resource belt line aren't as likely to starve and stall. You then have 18 minutes either for demand to subside or for you to increase production rates. it's 'not max', but it gets the job done and is more practical than simply assuaging the compulsion to have perpetually backed up material lines.

[ssilk]

I don't see the sense of this. If you want to make a storage, then build a storage.

This layout has the same problem as without chests: if your belts are full, the items are stored in the furnace output stack. Now, when the production continues, the left furnaces empty first. The right can empty only, if the left are cleared. This means: in the time, you need the most material, you produce the least. The buffer gives you a wrong impression of the real need.

I use a storage for that: a splitter cascade divides the output, the overflow runs into the storage, from where it is reinserted after the cascade, with a splitter inserting from the right (left priority ). You see then very clearly, how much goes to the storage, how much comes back. The storage could be anything, anywhere, the distance is unimportant (within radar range). It's a black box.

I prefer using a logistic storage for that case. I can put anything into the storage and have for every type of item one output belt. The bots do the sorting and because it's a own logistic network I always see, how much is stored.

[Zourin]

I have to aggree with ssilk on this one, although I try not to abuse my logistic drones with high volumes of iron and copper, save for loading/unloading trains. Belts do have their strengths when a lot of one or two things need to go to certain places continuously. Logistic Storage is good for items other than iron and copper (except in small quantities for things like AP ammo) and ore. For those things, just look at the belts and decide if there's enough.

Buffers, however, are best placed between the smelting plant and the beginning of the factory chain, and again about halfway through the factory chain, so that late-chain factories don't get starved out completely if the factories upstream start working at capacity and swiping most of the material.

[ssilk]

No, it's important to make that storage an own logistic network!!
Normally 60 drones are enough to handle one full input/output belt.

[Zourin]

No, it's important to make that storage an own logistic network!!
Normally 60 drones are enough to handle one full input/output belt.

Interesting, although I'd need to find space (and spare drones) for that.. and excess supply. As soon as I switch over to advanced chips my system ends up in a resource deficit

[ssilk]

Well in THAT state of game you don't need such a storage. Some chests at the end of the production line, where you store in the overflow are enough.

I talk about after the first module-phase. In my opinion that is the moment, when you can begin to relax, and expand and suddenly need lots (10,000) amount of iron or copper for creating a new outpost-base.

[rlerner]

This is, more or less, the point of buffering it this way.

There is the output storage spot on the furnaces, however I tend to have large spikes (expanding the factory, reworking the belts), or large deficits on supply (reworking how that is transported in, trains, belts, different mines, etc).

I used to always use the chest buffering methodology where it is a discrete system apart from the furnaces, but I found that it is difficult to get around the corner penalties by using this method. The one I put above takes care of any corner penalty.

As for the space it consumes -- I used to go for the hyper-compact factories, which is a fun challenge, but now I enjoy targeting throughput even more... It's easier with a large factory.

ssilk - Furnaces cannot compare to this technique, with the stack size bonus, this clears them out quickly and allows them to get to work sooner, and has three outputs. You can do that with an electric furnace, but I like this method more.

And you're right, it does overproduce iron plates. That's why, when looking at production metrics, I don't look at my ore consumption. I know I will use it eventually. I look at the plate consumption and go from there.

I use logistics bots for intermediate products usually, and end products I want delivered to me.

[Marqee]

nice but i prefer another kind of setup.

[Zourin]

That layout practically begs for splitters to merge the streams together, although I see the wisdom in not bothering. That's still a lot of lateral real estate. Get pegged between a lake and a treeline and things get gnarly for space in short order pre-bots.

[Sir Nick]

OP: That desing is extremely NOT scalable. Even an express belt has a limited throughput, and your factory will sooner or later reach it. Doing it parallel instead of consecutive might be better, but then you run into splitting ussues.

In lategame I use a bot-powered setup: four furnaces fed by a single Requestor chest set to 10 copper ore and 10 iron ore and outputting into two provider chests placed diagonally. With two regular power poles in the middle, this setup is VERY expandable. Its main advantage is versatility: when a train full of ore comes, it is instantly distributed among the furnaces. The main disadvantage is that if you at some point have an overwhelming surplus of one type of ore, than it practically chokes smelting of the other one. It may be solved, but usually I just bump up the number of furnaces.

This setup gladly shares a roboport with train unloading, BTW.

[rk84]

Here my 2 belts in and out -desing. Smart inserters are limited to input (max) 20k plates to LogNet and top row of request chests are connected to ore field with green cable to limit ore supply.

[Zourin]

Here my 2 belts in and out -desing. Smart inserters are limited to input (max) 20k plates to LogNet and top row of request chests are connected to ore field with green cable to limit ore supply.

Wild, and dependent on a lot of bots moving ore around. What I don't get are the fast inserters pushing items in the Provider chests into storage. Wouldn't it just be better to add a LogNet condition to the smart inserters to stop once you have around 30k iron in the system, and not mess with storage chests or blindly inserting into them?

[Deathmage]

Storage chests and simply use so much iron/copper that you never have it buffering onto the belt