Factorio Smelting Arrays Guide: Smelting Ratios, Furnace Counts & Blueprint Setups

Iron plate and copper plate smelting both use the same furnace timing, which is why most Factorio smelting ratio charts treat them as one planning problem. A steel furnace or electric furnace outputs 0.625 plates per second for iron or copper. That means the familiar yellow-belt target of 15 items per second is exactly 24 furnaces.

Stone furnaces produce half that throughput, so all the counts double if you are still in the earliest burner phase. Most players switch into steel or electric furnaces before they finalize an array blueprint because the 24-furnace module is easier to reason about than a 48-stone-furnace stopgap.

Steel is different because the recipe is not just another plate. It consumes 5 iron plates per steel and takes much longer to craft, so the output-per-furnace number drops sharply. That is why steel always looks furnace-hungry compared to iron and copper, even before you account for the upstream plate input it also requires.

The easiest way to plan a Factorio smelting array blueprint is to pick a belt target first, then size the furnace row to that exact output. For steel or electric furnaces, the practical benchmark table looks like this:

The standard early iron smelting setup is therefore a mirrored pair of 24-furnace plate arrays: one for iron, one for copper. That gives you one fully compressed yellow belt of each plate type without inventing custom math for every new ore patch.

Example metrics help because the most useful smelting blueprint is not the theoretical maximum. It is the module that matches the rest of your factory’s actual demand. These are the throughput numbers many players anchor around:

Those numbers also show why steel should be isolated from your basic plate arrays. A block that outputs 7.5 steel per second already needs the iron equivalent of two and a half full yellow belts. If steel shares the same iron line as gears, circuits, and science, the base starts starving in several places at once and the bottleneck becomes harder to spot.

Good smelting array blueprints are boring on purpose. Every furnace sees the same ore access, every plate lane exits the block the same way, and expansion is just another copy of a proven module. The most useful design rules are simple:

If you want one reliable answer for a Factorio iron smelting setup, use a repeatable 24-furnace module for each yellow-belt plate lane, leave walking and power space between mirrored halves, and move steel into its own dedicated furnace block instead of tacking it onto the main bus when demand rises.

AI builders perform better when every plate lane has a known ceiling. A deterministic smelting module tells the agent how much ore the block needs, how many plates it should output, and where that lane should reconnect to the bus or the train station. There is no need to infer hidden throughput from partial chest buffers or uneven furnace rows.

That repeatability matters for humans too. Once the furnace count and items-per-second target are locked, plate shortages become traceable again. You can ask whether ore delivery failed, whether power sagged, or whether downstream consumers simply outgrew the module, instead of rebuilding the entire array from scratch every time science demand climbs.