DEV Community

Cover image for Sizing a Parcel Sortation System: Throughput, Chute Logic and the Data You Need Before You Buy
TrueLiSort
TrueLiSort

Posted on

Sizing a Parcel Sortation System: Throughput, Chute Logic and the Data You Need Before You Buy

Buying a sortation system on the basis of "average parcels per day" is the fastest route to an expensive bottleneck. Sorters are sized for peaks, and peaks are shaped by arrival profiles, item mix and read rates — not by the daily total. Here is the data pack we ask customers to assemble before any layout discussion.

Build a real arrival profile

You need parcel counts in 15-minute buckets for at least 4–6 weeks, including a peak season. From that, extract:

  • Peak hour volume, and the peak within the peak (often a 10-minute spike).
  • The ratio of peak hour to average hour — 2.5–4× is common in e-commerce.
  • Number of inbound docks feeding the sorter and whether they release simultaneously.
  • Cut-off times driven by carrier collections, which compress the working window.

If carrier collection is at 18:00 and 60% of orders arrive after 14:00, your effective sorting window is four hours, not eight. System capacity must be sized against that compressed window plus a recovery margin for downtime.

Item mix drives the technology choice

Parcel characteristic Implication
Weight range (e.g. 0.1–30 kg) Cross-belt vs. shoe sorter vs. tilt-tray
Min / max dimensions Induction pitch, chute width, gap control
Soft bags, polybags, irregular shapes Need belt-based transport, not rollers
Fragile / high-value items Limit drop height and transfer speed
Non-conveyable outliers Requires a manual exception lane

A sorter rated at 8,000 parcels/hour on cartons may deliver 5,000 on a mix dominated by small polybags. Ask for a rate qualified on your mix, not a brochure figure.

Read rate is a capacity multiplier

No-read parcels recirculate or go to a manual exception station. A 98% read rate at 6,000 pph means 120 parcels/hour diverted to manual handling — a full workstation, every hour. Improving barcode quality and scanner placement from 98% to 99.5% can be cheaper than adding sorter capacity.

Include in the specification: minimum barcode grade (ISO/IEC 15416), number of scan attempts, camera coverage for top/side/bottom labels, and a defined exception route that never blocks the main line.

Chute logic and destination planning

Chute count is not just "number of carriers". Consider:

  • Destinations per wave and whether chutes are dynamically re-assigned between waves.
  • Chute full handling: what happens when a chute fills mid-wave? A blocked chute that stops the sorter is a design flaw, not an operational issue.
  • Recirculation for overflow, and how many loops a parcel is allowed before exception.
  • Outbound sequencing — sorting by route, by carrier, or by delivery sequence each imposes different chute pressure.

Model the wave with a simple discrete-event simulation before committing. Even a spreadsheet-level model that assigns parcels to chutes by destination volume will expose the one chute that always fills early.

WCS/WMS interface and controls

Define the handshake explicitly: how the sort plan is downloaded, how parcel ID and destination are paired, what happens on a scanner timeout, how reconciliation is reported at end of wave, and which system owns the "parcel sorted" event. Ambiguity here is a classic source of go-live disputes.

Also specify what the operator sees: a dashboard with live throughput, chute fill levels, no-read count and jam locations beats a wall of alarms.

Acceptance testing: write it into the contract

A meaningful FAT/SAT includes sustained run at rated peak for a defined period (e.g. 2 hours at 110% of design peak), mixed-size parcels including the worst-case polybag, induced faults (full chute, no-read, emergency stop) to prove recovery, sort accuracy target (commonly ≥99.9%), and documented recovery time after each fault.

FAQ: should we buy headroom now?

Some, but not blindly. 20–30% capacity headroom is reasonable. Beyond that, modular extension (adding chutes or a second induction line) is usually better value than oversizing the whole loop on day one — provided the layout reserves the space and the controls support it.

Wrapping up

Sortation is a system-of-systems problem: arrival profile, item mix, read rate, chute logic and controls. Get the data pack right and the hardware selection becomes straightforward.

For automated parcel sortation systems, layout design and WCS integration, see TrueLiSort. More information is available at truelisort.com.

Top comments (0)