Cooling Math for Jeju Humidity Peaks

Operators often schedule spring installs because ferry weather is calmer and electricians have slack. The mistake is sizing cooling from March intake temperatures. On Jeju, relative humidity in July pushes effective heat rejection down even when dry-bulb readings look similar.

We logged a Hallim shed last year: 28°C intake in April with 62% RH, versus 30°C intake in July with 88% RH. Same stock fans, same rack layout — exhaust delta dropped 4°C in summer, and two boards throttled before the operator noticed hash dip.

What to Measure Instead

Record wet-bulb or RH alongside intake temp at 14:00–16:00. Compare against manufacturer derating curves if published; if not, assume conservative throttling above 85% RH. For air-cooled rows, add 15% BTU removal margin when planning a 0.05 TH/s addition destined for summer energizing.

Duct Checks Before Ordering

Verify filter loading — clogged panels mimic humidity stress. We photograph filter faces on every visit because operators forget replacement intervals when hash is stable in spring.

Immersion Notes

Coolant loop rejection still depends on ambient air for most Jeju immersion installs. Loop inlet targets should use summer design, not annual average. Pressure test fittings before adding hashboards; slow leaks show up under thermal cycling, not cold start.

Practical Gate

Do not energize a second equipment batch until you have one humid-week log after the first batch. Incremental capacity means proving the room accepted load under peak moisture, not peak dry heat.