Dry season in Bali is almost unfair for solar. Clear skies, five-plus hours of strong sun, panels running at 90-plus percent of rated output. Then November rolls in, the sky turns grey, and your production app shows a number that looks worryingly low. If you own a villa with solar or you're about to install one, this is the article to read before you size the system or start panicking about winter-equivalent output.
The short version: yes, Bali's rainy season cuts solar output. About 20-25% on average, and more on the worst single overcast days. But the system doesn't stop working, and if the system was sized correctly for the cloudy months, you won't notice the difference in your day-to-day comfort. Here's the full picture, with real numbers instead of marketing optimism.
Reading this in Bahasa Indonesia? Switch to: /blog/cuaca-mendung-berhari-hari-panel-surya
TL;DR
- Bali wet season (November to March) reduces average panel output by 20-25%. A 5 kWp system drops from about 22 kWh/day in August to 16-18 kWh/day in January.
- Single overcast days can fall as low as 3-5 kWh. Size for the worst sustained 30-day average, not the worst single day. That keeps battery costs reasonable.
- Hybrid systems handle cloudy weeks comfortably: battery covers night load, PLN tops up after 2-3 consecutive overcast days. Full off-grid needs 2-3 days of battery autonomy.
- Don't oversize panels hoping to compensate for rainy-season dips. Permen ESDM 2/2024 removed net metering: surplus dry-season production has nowhere to go and is simply wasted.
- Wet-season rain keeps panels cleaner. The real long-term threat is connector humidity and inverter salt-air ingress, not reduced sun hours.
What the numbers actually show
Bali's solar resource varies more by month than most buyers realize. Peak sun hours (PSH, roughly equivalent to kWh/m² of solar irradiance per day) tell the story clearly.
Approximate monthly PSH for Bali (south-facing panels, standard tilt), using Global Solar Atlas reference data:
| Month | Approx PSH | Season |
|---|---|---|
| January | 4.0 | Wet |
| February | 4.1 | Wet |
| March | 4.3 | Transitional |
| April | 4.9 | Transitional |
| May | 5.2 | Dry |
| June | 5.4 | Dry |
| July | 5.5 | Dry |
| August | 5.5 | Dry |
| September | 5.2 | Dry |
| October | 5.0 | Transitional |
| November | 4.5 | Transitional |
| December | 4.1 | Wet |
Dry-season average (May-September): about 5.4 PSH. Wet-season average (December-February): about 4.1 PSH. That's a 24% drop, which matches the 20-25% rule of thumb.
For a real system: a 5 kWp array with a typical 80% system efficiency produces roughly 5 × 5.5 × 0.8 = 22 kWh/day on a peak dry-season day, and 5 × 4.0 × 0.8 = 16 kWh/day on a wet-season average day. That's the band you're working with. Not zero, not half. About 16-22 kWh depending on the month.
The floor, not the average, is what trips people up. A fully overcast Bali day during a heavy squall can drop production to 3-5 kWh for a 5 kWp system. That's real. But those days are brief. The sustained 30-day wet-season average is 16-17 kWh/day for the same 5 kWp array, not 5. Design for the average, not the storm.
How hybrid systems handle cloudy weeks
A hybrid system (solar + battery + PLN backup) is designed specifically for this scenario. Here's what happens on a typical overcast week in January:
Day 1-2 of cloud cover: The panels still produce 50-70% of normal output. The battery charges partially during the day, covers nighttime load as usual. You might not notice anything different.
Day 3-4 of sustained cloud: The battery starts each morning at a lower state of charge. Your monitoring app shows PLN taking a larger share of the daytime load. The system is still doing its job: using all available solar first, then drawing from PLN only for the shortfall.
Day 5+: Heavily overcast week? PLN carries most of the load. Your bill for that stretch is higher than in July. This is exactly what you paid for: PLN as a reliable top-up, not a problem.
The key discipline here is battery depth of discharge. LiFePO4 batteries handle 80% DoD safely across 6,000-plus cycles. During cloudy stretches, your system may cycle deeper each night. That's fine, that's what the battery is there for. Where it becomes a problem is if the battery is undersized for your daily usage. A 10 kWh battery for a 30 kWh/day villa will struggle. A 20 kWh battery for the same villa handles a 2-day cloudy patch comfortably.
Our rule for hybrid sizing: battery kWh = one day of your actual usage, at 80% DoD. PLN handles the rest. That's the sweet spot between cost and autonomy for most Bali villas on the grid.
Sizing for wet season if you're full off-grid
Full off-grid is a different conversation. With no PLN fallback, the battery has to cover every gap in solar production, including multi-day stretches of poor weather.
The rule we use: size panels to the cloudy month, not the sunny month. If you size for August's 5.5 PSH, your system is fine in dry season and insufficient in January. If you size for January's 4.0 PSH, your system is slightly over-spec'd in dry season but covers you year-round. That extra capacity doesn't cost you much (panels are the cheaper part of the system), but it makes the difference between a comfortable January and a stressful one.
For battery autonomy, we recommend:
- Two days minimum for most south Bali full off-grid villas (Uluwatu, parts of Canggu, East Bali coast).
- Three days for interior highland areas: Munduk, Bedugul, Kintamani, parts of Sidemen. These areas have longer cloudy periods during wet season, and the cloud cover is thicker at altitude.
The math: if your villa uses 30 kWh/day and you want 2 days of autonomy at 80% DoD, you need 30 × 2 / 0.8 = 75 kWh of battery. Round up to the next module multiple, so probably 80 kWh (for example, 16 modules at 5.12 kWh each, or 6 PowerGem Plus 14.3 kWh units).
More than 3 days of autonomy is rarely worth it. The cost of that fourth and fifth day of battery is significant, and the use case for it (4-5 consecutive fully overcast days with zero sun) is extremely rare in south and central Bali. A small backup generator covers that scenario for far less money than extra battery capacity.
The zero-export trap: don't oversize to compensate
Here's the mistake we see most often when owners learn about rainy-season output drops: they try to compensate by oversizing the panel array. "If I install 8 kWp instead of 5 kWp, the extra panels will cover the cloudy months."
That logic has a fatal flaw. Under Permen ESDM 2/2024, Indonesia's current residential solar regulation, there's no net metering for residential systems. Excess production during the day doesn't earn credits from PLN. It goes nowhere. In a hybrid or grid-tied system, once the battery is full and your home load is met, surplus panel output is curtailed or wasted.
What this means for sizing: an 8 kWp system in August (peak dry season) might produce 35 kWh on a clear day, but if your villa only uses 22 kWh of it, 13 kWh is wasted. You paid for 8 kWp of panels to only use 5 kWp worth of actual output. The rainy-season average for that 8 kWp system is around 26 kWh/day, which is still more than your 22 kWh/day need in most cases.
The right sizing logic: size for your actual daily consumption, adjusted for wet-season PSH, not for dry-season excess. For a 22 kWh/day villa, a 5-6 kWp panel array sized to the wet-season PSH of 4.0-4.3 covers you year-round without wasted overproduction in the dry months.
Cross-linked: if you want the full net metering picture, we've covered it in Net Metering in Indonesia: 2026 Reality.
When this doesn't fit your home
You're full off-grid in a highland interior area and haven't sized for wet season. Munduk and Bedugul villas with 1-day battery autonomy will struggle in December-February. If your installer sized to August sun hours for a full off-grid system, you need to revisit the battery spec before wet season arrives.
Your villa uses AC all night and the battery isn't big enough. A villa pulling 30-40 kWh/day with only 10 kWh of battery will cycle into deep discharge nightly during cloudy periods. This shortens battery life and leaves you grid-dependent. Size the battery for your actual night load, not the smallest module that fits the budget.
Your roof is in dense shade from November to March. Some Ubud and interior valley villas face canopy shadow that's worse in wet season when deciduous tropical trees have full foliage. Combined with cloud cover, shading can cut output an additional 20-40% on top of the seasonal PSH drop. If you haven't had a wet-season shading survey done, ask for one before finalizing the design.
In all three cases, we'd rather surface the problem in the sizing conversation than have you discover it in January. That's the whole point of an honest sizing call.
Ready to size your system?
The fastest way to get a number that works for your specific villa, including wet-season adjustment, is a 10-minute WhatsApp chat. Tell us the villa size, your current PLN bill or monthly kWh use, whether you're on or off-grid, and your location in Bali. We'll come back with a system spec that holds up in both July and January.
Frequently asked questions
On average, panel output falls 20-25% from dry-season peak to wet-season trough. A 5 kWp system producing around 22 kWh/day in August typically produces 16-18 kWh/day in January. The worst single overcast days can drop to 3-5 kWh total, but those are brief. The sustained wet-season average is what matters for sizing, not the single-day floor.
