Train layouts: stations, intersections and signaling

The capacity numbers that drive station design

Before designing a station, it helps to know what a single train can deliver. The cargo wagon has 40 inventory slots, each holding one stack. The stack size depends on the item being carried, so the per-wagon capacity varies by material:

Material	Stack size	Per wagon	4-wagon train
Iron ore	50	2,000	8,000
Electronic circuits	200	8,000	32,000

A four-wagon ore train carries 8,000 ore per trip. The fluid wagon holds 50,000 units of fluid. The locomotive has a top speed of 1.2 tiles per tick, which is 72 tiles per second — though that is a cap, not an average, and real travel time depends on acceleration, fuel, train weight, and track geometry. These numbers come from data/2.1/machines.json and data/2.1/items.json. The train throughput page works through the belt equivalents and round-trip math in full detail.

Station design: loading and unloading bays

A train station is a straight section of track with a train stop, space for inserters and chests on one or both sides, and a way for the train to enter and leave without blocking the main line. The dimensions of the bay are determined by the train length.

Bay length

Each cargo wagon is 6 tiles long and a locomotive is also 6 tiles. A 1-4-1 train (one locomotive, four wagons, one locomotive) is therefore 6 × 6 = 36 tiles long, though the stop itself only needs to cover the wagon positions for inserter placement. The practical rule is to build the straight section at least two tiles longer than the train at each end, giving the train room to stop accurately and room for the signal that guards the station exit.

A 2-4-2 train adds two more locomotives, reaching 48 tiles. The longer train has better acceleration and bidirectional capability, but every station it visits must have a longer bay. If some stations on the network serve short ore trains and others serve long mixed-cargo trains, the bay must be sized for the longest consist that will ever stop there.

Inserter placement

Each wagon has two long sides, and each side has room for up to six inserters — twelve per wagon in total. In practice, stations typically use six inserters per wagon (all on one side) for a compact layout, or twelve for maximum unloading speed. The inserters place items into buffer chests (steel chests with 40 slots each, or logistic chests for robot-fed stations), and belts pull from those chests into the factory.

The choice between six and twelve inserters is a trade-off between unloading speed and footprint. Six stack inserters per wagon can empty a full ore wagon in a few seconds; twelve does it roughly twice as fast but requires chests and belts on both sides of the track. For most bases, six per side is sufficient. For high-throughput megabase stations where train turnaround time is the bottleneck, twelve per wagon is the standard.

Station entry and exit

A station should never be placed directly on the main line. A train stopped at a station blocks any train behind it, turning the main line into a parking lot. Instead, the station sits on a siding or bay that branches off the main line, with a chain signal at the entrance and a regular signal at the exit. A train entering the siding reserves the block and does not block through traffic.

The siding needs enough track between the main line divergence and the train stop for the full train to clear the main line before stopping. If the train is still partially on the main line when it halts, it blocks the junction. As a rule of thumb, the divergence point should be at least one train length before the stop.

Intersection types

An intersection is where two or more tracks cross or merge. The shape of the intersection determines how much track it uses, how many trains can pass through it simultaneously, and how resistant it is to deadlock. Three shapes cover most cases.

T-junction

A T-junction joins three directions: a through track and a branch. It is the simplest intersection and the most common in rail networks that grow organically from a main line. A well-signaled T-junction allows a train on the through track to pass while a train on the branch waits, and vice versa. It cannot handle two trains crossing conflicting paths at the same time.

The T-junction's strength is its compact footprint and simplicity. Its weakness is throughput: every train that turns crosses the path of through traffic, and a busy T-junction becomes a bottleneck as the base grows. For a main line with a handful of branches, it is the right choice. For a network with heavy traffic from multiple directions, a larger intersection type is better.

Roundabout

A roundabout is a circular loop that connects multiple incoming and outgoing tracks. Trains enter the loop, travel around it in one direction, and exit at their desired branch. A roundabout can connect four or more directions in a compact space and has no crossing paths — every movement is a merge or a diverge, which means chain signaling is straightforward and deadlocks are rare as long as the loop itself is not allowed to fill up.

The roundabout's weakness is that trains must travel the long way around for some turns, adding travel distance. It also has a fixed capacity: once the loop is full of trains, no more can enter, and a train that stops inside the loop can block everything. Roundabouts work well for moderate traffic with four roughly equal branches. They become problematic at very high train counts where the loop saturates.

Cross-junction (four-way)

A four-way cross-junction allows two tracks to intersect and permits turns in every direction, including across the opposing track. This is the most flexible intersection shape — any train can go any direction — but it has the most conflicting paths and requires the most careful signaling. Chain signals at every entrance and regular signals at every exit are mandatory; a single missing signal can cause a deadlock.

A well-designed cross-junction can allow two non-conflicting trains to pass simultaneously (for example, one going north-south and another going east-west if the geometry keeps their paths separate). However, most compact cross-junctions allow only one train through at a time, making them throughput-limited despite their flexibility. For high-traffic networks, the standard approach is to build the intersection large enough that crossing paths do not overlap, which increases the footprint significantly.

Shape	Directions	Footprint	Throughput	Signaling difficulty
T-junction	3	Small	Low to medium	Simple
Roundabout	4+	Medium	Medium	Moderate
Cross-junction	4	Medium to large	Medium to high	Complex

Signal spacing and block design

Signals divide track into blocks. Only one train may occupy a block at a time. The regular rail signal recipe costs 1 electronic circuit, 5 iron plate, and the chain signal costs 1 electronic circuit, 5 iron plate — both are cheap, which means the limiting factor on signal density is never material cost but placement care.

Where to place regular signals

On straight, open track, place a regular signal roughly every train length. This allows the next train to follow one block behind, maximising track capacity without risking a train stopping inside an intersection. Signals closer together than one train length do not increase throughput — they just create blocks too short to hold a full train, which can cause trains to stop with their rearmost wagon blocking a junction behind them.

Regular signals must always be placed after an intersection, where a train can safely wait without blocking the junction itself. They should never be placed inside an intersection, because a train stopped at that signal would occupy the crossing and block all conflicting paths.

Where to place chain signals

Chain signals go before every intersection entrance. A chain signal only shows green if the train can proceed through the chain signal and clear the next regular signal immediately. This prevents a train from entering an intersection unless it can also exit — the primary defense against deadlock.

The standard signaling pattern for any intersection is: chain signals on the way in, regular signals on the way out. If a train cannot fit entirely in the block beyond the exit signal, the chain signal holds it outside the intersection. This is why the exit block must be at least one train length long before the next signal.

Double-track signaling

A double-track line uses two parallel tracks, one for each direction. Each track is signaled independently with regular signals spaced one train length apart. No chain signals are needed on the open track because trains on each track all travel the same direction and never face each other. Chain signals only appear where the two tracks connect at crossovers or stations.

Double track is the standard for mature train networks because it eliminates head-on collisions and roughly doubles throughput compared to single track. The cost is rail: two rails per tile instead of one, plus signals and occasional crossovers. Each rail costs 1 stone, 1 iron stick, 1 steel plate and covers two tiles, so the material cost of double track is modest even over long distances.

Train configurations: 1-4-1 vs 2-4-2

The notation L-W-L describes a train with L locomotives at the front, W cargo wagons, and L locomotives at the rear. The two most common configurations for general-purpose rail are 1-4-1 and 2-4-2.

1-4-1

A 1-4-1 train has one locomotive at each end and four wagons in between, for a total length of 36 tiles. It can travel in either direction without turning around, because the rear locomotive faces the opposite way and can pull the train backward. This makes station design simple: no balloon loop or roundabout is needed to turn the train at its destination.

The 1-4-1 has moderate acceleration. With four loaded ore wagons, a single locomotive pulling from one end takes time to reach top speed. The rear locomotive helps when traveling in reverse but does not add power when both face the same direction. For ore, plates, and other high-volume, low-time-sensitivity routes, 1-4-1 is the standard workhorse.

2-4-2

A 2-4-2 train has two locomotives at each end and four wagons, for 48 tiles of total length. It has twice the accelerating power of a 1-4-1, which matters on long routes with frequent stops or on hilly terrain. It is also bidirectional, with the same station benefits as 1-4-1.

The cost is two extra locomotives and longer station bays. Each locomotive costs 20 engine unit, 10 electronic circuit, 30 steel plate, so doubling up is not free. The 2-4-2 is worth it when the route is long enough that acceleration time is a significant fraction of total round-trip time, or when the train carries heavy cargo and struggles to maintain speed. For short routes between nearby outposts, the extra locomotives rarely pay for themselves.

Configuration	Length (tiles)	Wagons	Ore capacity	Best for
1-4-1	36	4	8,000	General-purpose ore and material routes
2-4-2	48	4	8,000	Long routes, heavy cargo, frequent stops

The cargo capacity is identical between the two configurations because they have the same number of wagons. The difference is speed and acceleration, not volume. A 2-4-2 delivers the same 8,000 ore per trip as a 1-4-1, but it may complete the round trip faster. Whether that speedup justifies the extra locomotives depends on the route distance and the value of reduced train wait time at stations.

Schedules: vanilla vs LTN

Factorio's vanilla train system uses schedules: a list of stations with wait conditions attached to each. A train travels to the first station on its schedule, waits until its conditions are met (cargo full, cargo empty, time elapsed, or circuit network signal), then proceeds to the next station. Schedules are simple and reliable for fixed routes: ore from outpost A to smelter B, repeat.

The limitation of vanilla schedules is that a train is bound to its named stations. If outpost A runs dry, the train keeps going there and waiting. If you add a new outpost C, you must manually update every train's schedule to include it. For a network with a handful of fixed-route trains, this is fine. For a network with dozens of outposts producing the same resource, it becomes tedious.

LTN (Logistic Train Network) is a mod that turns trains into a on-demand logistics system. Stations advertise what they provide or request via circuit network signals, and LTN dispatches the nearest available train to fulfil the request. It is analogous to logistic robots but for trains: you do not assign a train to a route, you define supply and demand thresholds and the network matches them. LTN excels at large bases with many producers and consumers of the same resources, where manually managing schedules would be impractical.

The trade-off is complexity. LTN requires circuit network wiring at every station, additional modded entities (depots, providers, requesters), and a thorough understanding of its threshold and priority settings. It is powerful but has a learning curve. For a base with fewer than twenty trains, vanilla schedules with well-named stations and sensible wait conditions are usually sufficient and easier to debug.

Common layout mistakes

• Stations on the main line. A stopped train blocks everything behind it. Always use a siding.
• Missing chain signals at intersection entrances. This is the number one cause of train deadlocks. If a train can enter an intersection without being able to exit, it will eventually stop across the crossing and block all paths.
• Blocks shorter than the longest train. A signal placed too close to an intersection can leave a train stopped across the junction. Every block after an exit signal must be at least one train length long.
• Crossing paths without a chain signal. If two tracks cross at grade, every approach must have a chain signal. A single regular signal at a crossing allows two trains to enter simultaneously and collide.
• Oversized intersections for low traffic. A sprawling four-way junction on a line that sees two trains per minute wastes space and materials. Match the intersection shape to the traffic volume.

Applies when…

• Cargo wagon capacity uses inventory_size = 40 from data/2.1/machines.json, multiplied by each item's stack_size from data/2.1/items.json.
• Fluid wagon capacity is 50000 from machines.json.
• Locomotive top speed is 1.2 tiles/tick (72 tiles/s). This is a cap, not an average. Actual travel time depends on acceleration, fuel, train weight and track geometry, none of which are in the extracted data.
• Train length calculations assume 6 tiles per wagon and 6 tiles per locomotive, which are game constants.
• Recipe costs for signals, train stops, rails and locomotives are read from data/2.1/recipes.json.
• Intersection type comparisons are design guidance, not computed values. Throughput ratings reflect general gameplay principles rather than simulation data.
• All figures are for game version 2.1.12. LTN is a third-party mod and its behaviour is not represented in the game data.

Questions or corrections? Contact [email protected].

Related

• Train throughput — wagon capacity, belt equivalents, and round-trip math
• Belt balancer designs — splitter counts and throughput for the belts that feed stations
• Main bus design — how to organise the factory that a train network feeds
• Smelting ratios — sizing furnaces for the ore a train delivers
• Planet-specific production — interplanetary logistics that extend the rail network