Walk into any large warehouse in Metro Manila in July, and you'll usually find the same setup: a row of split air-conditioning units bolted to the wall, condensers humming on the loading dock, and a thermostat reading 22 °C. The facility manager will tell you the climate is “fine.”
Pull out a humidity meter and you'll often read 78% RH. The product specification — for cured tobacco, milled grain, dried fish, finished cigarettes, paper substrate, anything hygroscopic — is 60% RH ± 5%. The dashboard says everything is fine. The product is silently being damaged.
1. Standard split AC has a Sensible Heat Ratio (SHR) of 0.70–0.85 — built to remove heat from human-occupied spaces.
2. A tropical warehouse load is dominated by latent moisture, with required SHR around 0.40–0.50.
3. The split AC hits temperature setpoint in minutes, compressor stops, dehumidification stops, and humidity climbs unchecked.
4. Once RH crosses 65%, mould (primarily Aspergillus) forms on hygroscopic product within 3–6 days. Across the Philippine agricultural supply chain, post-harvest loss from this exact failure routinely reaches 12–50%.
What the engineering is, plainly
Every air-conditioning machine does two things: it lowers the temperature (sensible cooling) and it removes moisture (latent cooling). The ratio of those two jobs is called the Sensible Heat Ratio. SHR = 1.0 means all the work is temperature; SHR = 0 means all the work is moisture removal.
Split AC is engineered for offices, hotels, classrooms — spaces full of warm bodies and equipment giving off sensible heat. The design SHR sits around 0.75. That's exactly right for a meeting room.
A tropical warehouse is the opposite. The shell is well-insulated, internal heat is low, but moisture pours in continuously through:
- Loading-dock door cycles — a 14×14 m bay opens for a forklift every 10–15 minutes during operating hours
- Required ventilation — ASHRAE 62.1 mandates outdoor air for worker air quality and to maintain positive pressurisation
- Product respiration — hundreds of tonnes of stored organic matter equilibrating with the surrounding air
- Envelope infiltration — the outdoor vapour pressure in Manila wet season is so much higher than indoor that moisture is actively driven through the building fabric
Run a first-principles ASHRAE Heat Balance calculation on a 200 m² warehouse zone and you'll get a sensible load around 6.4 kW and a latent load around 8.8 kW. That's SHR = 0.42. The split AC is built for SHR 0.75. It is, mathematically and physically, the wrong machine.
What happens in the room when the wrong machine is installed
The compressor starts. Cold refrigerant runs across the evaporator coil. The room — being well-insulated and having low internal heat — drops from 25 °C to 22 °C in about five minutes. The thermostat is satisfied. The compressor stops.
And then, with the compressor off, dehumidification stops too. There is no other moisture-removal mechanism. Outdoor humid air continues to enter through every door cycle, every ventilation intake, every gap in the envelope. Inside the closed warehouse the relative humidity climbs from 60% to 70% to 78% within an hour.
The thermostat eventually senses the warehouse warming back up (latent gain becomes sensible gain over time) and the compressor restarts. By then the humidity excursion has already happened. Cured tobacco can begin showing mould inside one week of sustained high RH; finished cigarettes show paper rehydration even faster.
The standard operator workaround is to drop the thermostat setpoint to 16 °C. This forces the compressor to run longer and removes more moisture as a side-effect. The warehouse becomes freezing cold (which causes its own problems with brittle product) and the electricity bill triples.
Some operators add an electric resistance reheat coil to overcool then reheat the supply air. This is explicitly forbidden under ASHRAE 90.1 §6.5.2.3 (simultaneous heating and cooling). Exception 5 only permits reheat for dehumidification when the reheat energy comes from site-recovered sources (like condenser heat) — which a standard split AC cannot do.
What this costs the customer in product
The cost isn't the electricity bill — though that's also significant. The real cost is the product:
| Product type | Spec | What goes wrong |
|---|---|---|
| Cured tobacco leaf, finished cigarettes | ISO 3402: 22°C ± 1°C / 60% RH ± 3% | Mould (Aspergillus) in 3-6 days >65% RH; paper rehydrates and unwraps; brittle leaf at <55% RH cracks during manufacturing |
| Grains (corn, rice) | Water activity (aw) < 0.7 | Aflatoxin production by Aspergillus flavus when aw reaches 0.90–0.99 — toxic, carcinogenic, regulated |
| Dried fish, processed foods | Varies; typically <65% RH | Reabsorption of moisture causes mould, rancidity in fats, loss of crispness/texture, microbial spoilage |
| Paper, packaging substrate | 50–60% RH | Curl, bleed, dimensional change disrupts downstream production lines |
PHilMech (Philippine Center for Postharvest Development and Mechanization) documents an average 12.7% storage loss on corn alone; broader agricultural loss across PH supply chains routinely reaches 50%. Inadequate warehouse storage environments are universally cited as the primary driver. Tobacco operators face the same physics — once mould begins, the affected pallet is unrecoverable.
The right architecture: AHU + iSTOR + iHEAT R290 + iVOLT
The fix is not bigger split AC. It is a different machine entirely. The Karnot warehouse architecture has four parts:
- A fully inverter-controlled AHU — Air Handling Unit with both a hot coil and a cold coil. The same machine handles both jobs from the same airstream, ducted directly into the warehouse. Runs on a humidistat, not a thermostat — so it keeps removing moisture even after the temperature is right.
- iHEAT R290 reversible heat pump — sealed monobloc, sited outside the warehouse envelope. Reverses between heating and cooling on demand to charge whichever side of the AHU needs it. R290 (propane) holds COP above 3.0 across the full Philippine ambient range; GWP of 3 versus R410A's 2,088, so no Kigali phasedown clock.
- Two iSTOR thermal batteries — phase-change material tanks, one charged at 44 °C (hot), one at 22 °C (cold). They buffer both sides of the AHU. When the heat pump reverses from heating to cooling (or vice versa), the “other” tank carries the load — the AHU sees no interruption. Karnot's iSTOR PCM formulation is patented and uses natural fluids sourced in the Philippines.
- iVOLT solar + Li-ION — roof-mounted PV array, Karnot inverter, LiFePO4 lithium-ion battery. Powers the entire system through any grid event. Eliminates the diesel generator entirely.
What the math looks like — a worked single-zone case
First-principles ASHRAE Heat Balance calculation on a 200 m² tobacco-product warehouse zone in Manila, currently cooled by 5 × 3 HP plus 1 × 2 HP split AC running 24/7 at ₱14/kWh tariff (including 3-phase service premium):
| Configuration | Annual electricity | Reduction |
|---|---|---|
| Existing splits · baseline | ₱629K | — |
| + Karnot iHEAT + iSTOR + AHU (grid-only) | ₱309K | −51% |
| + 15 kWp iVOLT solar PV (recommended) | ₱65K | −89% |
| + 16 kWh Li-ION battery (optional Phase 2) | ₱17K | −97% |
Net annual saving with the recommended (solar-included) configuration: ~₱564,000 per zone per year. CO2 avoided: ~28 tonnes per zone per year. Payback: ~2.9 years on a new-build basis (net of avoided split-AC capex), ~3.7 years on a gross retrofit basis.
The split AC system isn't just inefficient — it's the wrong tool. By matching the architecture to the load (decoupled humidity and temperature, hot-gas reheat from recovered condenser heat, R290 efficiency at tropical ambient, solar-powered electrical input) we eliminate the structural waste, not just trim around it.
How a project starts
One free site visit. Our engineers spend a day on site, datalog temperature and RH at multiple zone heights for a full operating cycle, and pull 12 months of utility data. Within 14 days you receive a one-page indicative report — block-load calculation per ASHRAE Heat Balance Method, recommended AHU + iSTOR + iHEAT sizing, indicative capex range, projected annual saving, payback period, and CO2 avoided.
No procurement triggered. No obligation. The output is yours to keep regardless of whether you proceed.
Want a worked case for your warehouse?
Book a free site visit. We model the load using the ASHRAE Heat Balance Method, size the AHU + iSTOR + iHEAT + iVOLT package, and come back with a one-page indicative proposal — no commitment.
Book a free site visit Download the brief