Batteries are temp-sensitive; the solution is designed per cabin load and layout.
1. Key points
±2C uniformity target, anti-dew, zoned airflow.
2. Redundancy
N+1, BMS/fire linkage, RS485 central monitoring.
3. Process
Load calc -> airflow -> selection -> commissioning.
| Item | Solution |
|---|---|
| Uniform | ±2C target |
| Redund | N+1 |
| Link | BMS/fire |
HCK ESS cabin cooling custom keeps battery cabins uniform and safe.
4. Why battery enclosures cool differently
An energy storage cabinet is not an electrical cabinet with a battery in it. Three things make the thermal problem its own discipline:
- The heat source is inside the battery, not in the electronics, and it scales with charge and discharge current
- Cell temperature window is narrow, and uneven temperature across the pack shortens life more than the average does
- Fire safety thinking constrains what materials and what air paths are acceptable
5. Size for the worst case that actually occurs
The design case is not "fastest possible charge in the coolest weather". It is a full rate charge or discharge in the hottest ambient the site will see. A pack that only charges overnight in a mild climate does not need the capacity that the same pack needs during a midday fast charge.
6. Uniformity beats average temperature
Temperature spread across the pack matters more than the mean. Cells at the top of a stack run hotter because hot air rises; cells at one end run hotter if the airflow does not reach them.
Two cheap measures fix most of it: baffles that force air across the full pack height, and a setpoint that targets the hottest cell rather than the average sensor reading.
7. Humidity and condensation
Battery enclosures generate their own moisture during charge, and room air carries its own. If the evaporator runs below the dew point, condensate forms — that is normal, but it needs somewhere to go.
- Provide a drain path that will not block in a dusty site
- Do not set the setpoint so low that dehumidification outruns the drain
- Keep the evaporator clean; fouled coils reduce both capacity and dehumidification
8. Where customisation actually matters
| Aspect | Why it varies by project |
|---|---|
| Capacity and voltage | Depends on pack size, C-rate and site ambient |
| Protection rating | Indoor hall versus coastal outdoor drives IP rating and coating |
| Power supply | Mains, DC bus, or a supply from the pack itself |
| Communications | Monitoring against a BMS or SCADA, usually via RS485 |
| Dimensions and mounting | Constrained by the container, the cabinet and the service access |
9. Commissioning checks
- Map cell temperatures during a full charge cycle, top, middle and bottom.
- Verify the unit holds setpoint at the site's worst expected ambient — not the datasheet one.
- Check condensate drainage under a humidity spike.
- Confirm the alarm thresholds match the cell limits, not a generic default.
10. Operating and maintenance schedule
An energy storage installation is normally served for decades, while the cooling unit inside it is expected to be replaced at some point. Designing for that replacement is cheap; discovering it was not part of the plan is not.
Three things determine how the unit ages: how often the filter is changed, how clean the condenser stays, and whether the setpoint is set to protect the cells or to protect the average reading. The first two are calendar items. The third is a design decision that gets revisited during every alarm.
| Item | Interval | What it protects against |
|---|---|---|
| Filter inspection | Monthly indoors, quarterly outdoors | Airflow loss across the evaporator |
| Condenser cleaning | Quarterly; more often in dusty or coastal sites | Quiet capacity loss of 20–30 % |
| Condensate drain check | Monthly wherever condensate forms | Water ingress into the electrical compartment |
| Cell temperature mapping | Quarterly, and after any fan replacement | Uneven ageing caused by changed airflow |
| Fan and controller check | Annually | The two items that fail silently before anything else |
11. Replacing the unit without decommissioning the pack
Plan the physical removal early. A unit mounted with its condenser coil against the ceiling of the container cannot be withdrawn upwards, and the pack below it cannot be moved. The usual outcome is a service visit that turns into a week.
- Leave a clear service path at least the depth of the unit on the removal side
- Fix the mounting so the unit can be unclipped with the fixing accessible, not sealed behind a panel
- Keep the same form factor and voltage across replacements, or accept that the control wiring will change
- Leave room for the condensate connection to be re-run
Ventilate the pack before anyone opens the enclosure for service. A battery enclosure that has tripped on a thermal alarm holds heat, and hydrogen from the cells has usually had nowhere to go.
12. Where a standard unit genuinely fits
Customisation is worth paying for when the power supply, the ambient, the protection rating or the monitoring interface genuinely differ from the catalogue. It is not worth paying for when they do not. A standard unit in a standard indoor cabinet with mains supply and no monitoring requirement is a standard job; asking for a bespoke unit there buys lead time and nothing else.
The test is simple: write down the four constraints that actually differ — supply, ambient, protection rating, and communications — and see how many of them match a catalogue item. Two or fewer is a standard job.
13. Fire safety thinking in practice
Where the pack contains lithium cells, the enclosure is designed as a fire-sensitive volume rather than as a cool room. That changes what is allowed near the cooling unit as much as what temperature it must hold.
- Air recirculation from a damaged cell into a clean compartment is not acceptable, so the cooling unit's air paths are considered in the fire strategy rather than after it.
- Materials near the unit are selected for low flame spread and low smoke density.
- Detection is usually a gas or early-stage temperature signal, ahead of anything a thermostat would see, with the cooling system's own alarm feeding the same panel.
- Where a thermal runaway is a credible scenario, the cooling unit is located and fixed so it does not become the path by which heat spreads, and so it can be isolated if needed.
None of this is a reason to avoid standard cooling practice. It is a reason to agree the air path and the detection logic early with whoever owns the fire strategy, because both are easier to incorporate at design stage than to retrofit.
14. Siting and enclosure-level decisions
The cabinet-level choices matter as much as the unit. A correctly specified unit in a badly sited enclosure will underperform and will be blamed for it.
| Siting factor | Effect on cooling | Mitigation |
|---|---|---|
| Direct sun on the cabinet wall | Adds several hundred watts of envelope gain | Shading, or move the cabinet away from the west wall |
| Cabinet in a hot room with other plant | Intake air may be well above ambient | Dedicated air path, or relocate the intake |
| Container in a desert or tropics site | Wall surface temperature far above air temperature | Rated for the surface condition, not just the air |
| Cabinet in a basement or underground vault | Heat has nowhere to go; drainage may be absent | Consider heat rejection to the room rather than to ambient |
| Level ground prone to flooding | Condensate and wash-down become a water-ingress risk | Raised mounting, sealed conduit, IP rating above the site's record |
The last item is the one most often discovered in the wrong order. Condensate is a designed drainage problem, and if the site has no drain, the unit needs a sealed condensate path to somewhere deliberate rather than into the bottom of the cabinet.