Potato2024–25 recorded production · 11.57m t
52.6%
Top-five district share. #1 Rangpur · 1.48m t.
Test the servicebefore choosingcapacity.
Measure when produce arrives, how long it waits, and what buyers need before capacity is chosen.
ArrivesWaitsMoves
Before capacity
Start with what arrives.
Produce arrives unevenly and waits for handling, transport and buyers. Cooling only makes sense where enough of that flow would use it.
Aug 2025100 farmer units reported launched
Jan 2026Further units entered procurement
Jun–Jul 2026Larger network ambitions announced
Public pipeline activity shows where to look. Operating status, reach and crop fit still determine usable supply.
Crop Intelligence · 2024–25
Crop patterns shape capacity.
Potato is concentrated, vegetables are dispersed, and mango clusters differently again. A fixed cold-chain assumption can fit one crop window and miss another.
Potato
Vegetables
Mango
Vegetables2024–25 recorded production · 6.28m t
23.4%
Top-five district share. #1 Chattogram · 379k t.
Mango2024–25 recorded production · 1.81m t
60.9%
Top-five district share. #1 Chapainawabganj · 412k t.
The atlas narrows the question. It does not establish booked volume, seasonal dwell, willingness to pay, site fit or permanent capacity.
Capacity follows use
Size the service from use.
Test the assumptions. Let the consequence, not the equipment size, lead the decision.
Modeled use
5.0 tof 10 t installed capacity occupied on an average day
At 55% occupancy and 90% uptime, modeled use reaches about half of a 10-tonne design.
Occupied tonne-days / month149
Unfilled tonne-days / month122
Indicative turns / month7.4
Capacity calendar30-day occupancy strip · proportionate, not a forecast
Held assumption Average dwell · 2 days
The consequence: more design capacity can also mean more idle capacity.
Scenario only. The model does not establish demand, willingness to pay, operator capability, site fit or final cost.
First commission
What the validation delivers.
Four decision outputs before permanent equipment is specified.
- Capacity-days mapArrival, dwell, peak demand and low-season use.
- Alternatives appraisalHandling, local cooling, shared capacity or no permanent build.
- Operator economicsTariff, uptime, repair response and service ownership.
- First-cluster mandateA documented Build, Resize, Defer or No build decision.
Field check
Take the questions back to the field.
- VolumeHow much produce actually arrives during the peak?
- TimeHow long does it wait, and when does cooling matter?
- UseWho would use the service, how often, and in the low season?
- OperationWho can run it reliably, and what would users pay?
- BuyerWhat condition, timing and route does the buyer require?
Booking → settlement clock
One operating record across the service route
- BookNamed load + arrival window
- ReceiveWeight + condition recorded
- CoolTemperature + dwell
- ReleaseDispatch time + route
- AcceptBuyer condition closes
- SettlePayment closes the record
The smallest justified response may be better handling, local cooling, shared capacity or no permanent build.
Build · Resize · Defer · No build
Allocation
Where validation goes.
No permanent-asset envelope is published before validation closes. Colour identifies workstream, not spend share.
Flow + demand mappingNamed crop weeks, arrival, dwell and buyer windows
Scope firstAlternatives + service testsHandling, cooling and shared-capacity responses
TestOperator economicsTariff, uptime, repair and service responsibility
ModelFirst-cluster decisionMandate the smallest viable intervention
DecideNo percentage split is published. Cost shares are set only after the validation scope, sites and fieldwork are confirmed.
Evidence notes
Public pipeline: 2025–26 launches, tenders and announcements indicate continued national activity. They do not prove operating status, utilisation or commercial sustainability; field mapping must verify each facility.
Sustainable-cooling hierarchy: World Bank guidance prioritises demand reduction, efficiency and system-level business models before expanding supply.
Cold-chain constraints: IFC links weak business models, financing, electricity and fragmented movement. Cooling-as-a-service remains one model to test, with mixed reported scaling outcomes.
Commission validation first.
Commission a decision-ready validation package before permanent assets.