A speed chain line is paced by the carrier, but the output is set by the slowest station. Most line problems that look mechanical — carriers backing up, operators waiting, fixtures returning empty — are balance problems. The conveyor only makes them visible.

Key takeaways

  • Takt time is available time divided by required output. It is a target, not a guarantee.
  • Station work content should sit at 85–90% of takt to leave room for variation and walking.
  • Buffers cover variation between stations; they do not fix a station that is permanently over takt.
  • A line that runs at 100% balance on paper usually runs at 85–90% in production.

Start with available time, not with the conveyor

Take the shift length, subtract breaks, planned maintenance and changeover, and you have available time. Divide by the required good units, and you have takt. A single shift of 480 minutes with 30 minutes of breaks and 20 minutes of changeover gives 430 minutes, or 25,800 seconds. For 2,150 good units, takt is 12 seconds.

The word “good” matters. If the line runs at 97% first-pass yield, you need to produce 2,216 units to ship 2,150. That changes takt to 11.6 seconds. Small differences in yield assumptions move the line speed, the station count and the price.

Divide work by content, then check the walk

List every operation and its time. Group operations into stations so that no station exceeds takt. Then add the walk time between the operator's positions and the carrier. A station with 9 seconds of work and 3 seconds of walking is a 12-second station, not a 9-second one.

Our rule of thumb is to plan station content at 85–90% of takt. The remaining 10–15% absorbs variation in product, operator pace and minor stoppages. A line designed at 100% has no recovery margin; the first slow cycle propagates down the line and the buffer empties.

Station content12 s taktEffect
100%12.0 sNo recovery margin; queues and starvation appear quickly
90%10.8 sWorkable for stable, repeatable operations
85%10.2 sComfortable for manual assembly with variation
75%9.0 sOver-staffed unless the process is unstable

Buffer sizing: cover the stoppage you actually have

A buffer is the number of carrier positions between two stations. Its job is to keep the downstream station working while the upstream station recovers. The useful formula is simple: buffer positions equals the longest expected stoppage divided by takt, multiplied by a confidence factor.

If a station stops for 90 seconds once an hour and takt is 12 seconds, the downstream station needs 7.5 positions to ride through the stop. Add two positions for variation and round up to ten. That is a practical buffer. A buffer of 30 positions costs floor space and increases the line's work-in-progress without improving output.

Field noteOn one connector line, increasing the buffer from 6 to 9 carriers lifted output by 7% without changing any station. The bottleneck was not the station time; it was the starvation caused by a 40-second test cycle that occurred every fifth carrier.

Where the conveyor constrains the balance

The carrier pitch sets a minimum distance between stations. If the pitch is 600 mm and an operator needs 900 mm of working width, the line cannot place two stations at 600 mm; one station must be skipped. That wasted position is a real cost. On dense manual lines, a smaller carrier with a shorter pitch often produces a shorter and cheaper line than a large carrier with generous spacing.

Carrier speed also affects the operator. At 20 m/min, a carrier crosses a 1.2 m station in 3.6 seconds. If the operator needs to work while the carrier is moving, that is the available time. Most manual assembly is done with the carrier stopped, which is why accumulation matters more than speed on manual lines.

Test the balance before fabrication

A spreadsheet model is enough for most lines. Simulate 30 cycles with the actual station times, add a random variation of ±10%, and watch the queue lengths. If a buffer empties more than twice in 30 cycles, it is too small. If a buffer never drops below half, it is larger than necessary.

For lines with more than 20 stations or mixed-model production, a discrete-event simulation is worth the effort. It costs less than rebuilding a line that cannot reach rate.

Frequently asked questions

Should the line run continuously or index?

Continuous chain with accumulating carriers is the most flexible and is standard for speed chain lines. Indexing is better for very short cycles or when every station must operate simultaneously on the same carrier.

How many stations can one operator cover?

Two adjacent stations are common if the combined content is below 85% of takt and the walk is short. Three stations usually require the carriers to stop long enough for the operator to move, which reduces the benefit.

Does a faster chain increase output?

No. Output is set by takt and station balance. A faster chain reduces transfer time between stations, which matters only when transfer time is a significant part of the cycle.

Send us your operation list and target output and we will return a station balance and a buffer recommendation with the line layout. The assembly line section explains how fixtures and material feed fit into the same calculation.