Français
Español
Typhoon Season, Two Basins, Three Warehouse Sites
A warehouse network in three cities is not one location with three addresses. Different weather systems close different gates, and that decides whether a storm is an inconvenience or a stock-out. The lever is where goods stand before the booking, not what happens after the wind arrives.
Key facts
- Three sites, two weather systems. Dongguan and Shenzhen sit inside the Pearl River delta's storm window; Suzhou sits in the Yangtze delta, a different basin with a different calendar.
- Network scale is fixed. 24,000 m² in total: Dongguan 8,000 m² for parcel stock, Shenzhen 3,000 m² for palletised and crated cargo, Suzhou 13,000 m² for Yangtze-delta consolidation.
- Closures are announced by others. A suspension is a port or municipal notice, not a forwarder's decision, hitting terminals, gates and haulage before your order.
- A few days becomes a week. Berth queues and gate restrictions rebuild slowly, so recovery outlasts the closure itself.
Two weather systems, and what each one stops
| Site | It takes | The weather that stops it | What a closure costs |
|---|---|---|---|
| Dongguan | Parcel stock and dropshipping inventory | Pearl delta storm signals and municipal work stoppages | Same-day despatch stops for the whole parcel book |
| Shenzhen | Pallets and crates for big cargo | The same system, plus terminal and gate suspension | Headhaul bookings cannot reach the terminal |
| Suzhou | Yangtze-delta supplier consolidation | East-China typhoons and inland waterway limits | Collection from delta factories pauses, not the parcel floor |
Two sites share a weather system; one does not. That asymmetry is the planning fact: a Pearl delta event stops parcel despatch and big-cargo delivery together while the Yangtze site keeps working, and the reverse is true too: the October 2026 east-China sequence closed Yangtze terminals and gates while our Pearl delta sites kept despatching, and that basin back in normal operation on 21 October.
Why a short closure turns into a long one
Operations restart before the queue does; three effects stack. Berths are occupied by vessels that waited, so the first departures after a suspension are the most contested. Gate windows tighten before they loosen, as a backlog-prone terminal restricts early deliveries. Then the cut-off cascade begins: a booking missed on Monday is not recovered on Friday, since the replacement vessel is carrying the cargo that missed Monday.
So a two-day notice costs five days of arrival, and that gap sits on top of a figure already published as an estimate. Ocean into the US West Coast is estimated at 18-25 working days, main European ports at 28-38; both assume a normal operating window at origin, and neither is a sentence to repeat to your customer.
The split that is a decision, not a hedge
Holding one line at two sites is not free insurance. It pays when the two sites feed lanes you actually use and a closure on one side leaves the other able to ship. It does not pay when both sites would serve the same storefront, because the split only doubles the count to reconcile at month-end.
Three inputs size it: how long the line can be out of stock before the storefront penalty lands; whether the destination is reachable from either site or one only; and whether the packing standard differs between a parcel line and a palletised one. Movement between sites is planned and quoted as its own line, with a second count reconciled against the first. The window that closes earliest is the booking one, not the storm: The Headhaul Booking Window.
What the closure calendar does to your dates
The Western Pacific produces late-window events that overlap the retail build, and the 2026 record carried them in east-China waters — an emergency-measures cycle at the Yangtze running into late October, and terminal suspensions the same month. Those notices land in the weeks when peak replenishment is meant to arrive, which is why the peak build starts before the curve: The Peak Build Starts Before the Curve.
Two habits survive this better than urgency. Treat the pre-peak weeks as the last reliable window rather than the earliest convenient one; a plan executable only in November has no margin. And put a latest-instructable date on every line, worked backwards from the arrival needed through the estimated band and the destination's own seasonal cut-off.
The instruction that survives a closure
A four-line note before the event is worth more than an urgent message during it: which line may ship from a fallback site; which destination it may reach from there; whether goods may be delivered into the terminal ahead of the cut-off or must wait; and what happens to orders placed while the floor is down — queued, cancelled, or moved to a channel still moving.
No provider can promise a schedule against a port notice. Stock instructable ahead of the 16:00 China-time cut-off leaves the same working day in normal conditions; under a suspension that sentence is the first thing to stop being true, and the China holiday calendar shows how routinely a date, not a distance, is the binding constraint: China's Holiday Calendar. The honest version of continuity is a documented fallback.
Four questions to settle before the next notice
- Which lines could ship from a second site without re-packing? If none, the network is not doing continuity work.
- Which lane is Guangdong-dependent? A departure from Guangdong needs cargo inside Guangdong.
- What is the latest instructable date for each arrival promised? Work it backwards, not forwards.
- Who absorbs storage if freight is already at the terminal? A commercial answer, agreed before the storm.
Send us your line list and the arrivals you have promised and we can name which of them a delta closure actually breaks: tony@sbt56.com, or the form at Contact. Standing answers are on the FAQ page.
Can you keep shipping if Shenzhen or Dongguan is under a storm suspension?
Not from the affected sites, and we would not sell you that answer. A suspension stops haulage, gates and terminal work before it stops ours, and a truck that cannot enter a terminal turns a delayed order into a detained container. What can continue is despatch from a site outside the affected basin, where the goods already stand and the destination is reachable from it. Stock placement is the decision; a line held at one site has no fallback besides waiting.
Should I split my stock across two of your three Chinese warehouses?
Where the line is genuinely at risk and both sites feed lanes you use, yes — as a priced decision rather than a default. The costs are concrete: a second count to reconcile, storage held in two places, and a movement between sites quoted as its own line. The benefit is equally concrete: one basin can close while the other still ships. Split the fast movers carrying a penalty; slow movers are usually cheaper to let wait.
Does a typhoon change my transit estimate or my transit promise?
It changes the estimate, and we never made a promise to change. Every band we publish, ocean and air alike, is an approximation that assumes a normal operating window at both ends, so an event makes the figure less accurate rather than breaking a commitment. What is worth demanding from a provider is the two dates that matter during a closure: the next departure that can still take your container, and the cut-off by which it must be inside the terminal.
Operational guidance published 1 November 2026 (2026-11-01) by FulfillDock by SBT, with 24,000 m² at three sites — Shenzhen, Dongguan, Suzhou — under the company registration Shenzhen Shibotong International Logistics Co., Ltd. Closure notices come from port and municipal authorities and are reported here as operating conditions, not forecasts. No prices or amounts appear; every transit figure is a published estimate, not a delivery promise.
