I’m Sudeep, and if you’ve read my work before, you know I have a soft spot for turning vague claims into actual numbers. “Solar panels still work on cloudy days” is one of those lines I’ve said to clients probably a hundred times over the years, usually followed by a slightly uncertain “…just less.” That “just less” always bothered me, because it’s not a real answer. So this August, I decided to actually measure it. My friend runs a residential solar setup in Virginia, USA, and he agreed to share his system’s daily generation data with me for a full 30-day stretch, cross-referenced against the local weather conditions each day. What follows isn’t a lab study or a manufacturer’s marketing brochure — it’s real production data from a real rooftop system, tracked day by day through an entire month, including the genuinely gray stretches Virginia summers can throw at you. Here’s exactly what we found.
1. The Setup: What System We Tracked
Before getting into results, let me walk you through what we were actually measuring, because the specifics matter for how you interpret the numbers. My friend’s system is a 6 kW residential rooftop array, south-facing, installed at a tilt appropriate for his latitude, using standard monocrystalline panels — the most common configuration for U.S. homes today. The system includes production monitoring through his inverter’s app, which logs daily kWh generation automatically, so the data wasn’t hand-recorded guesswork; it was pulled straight from his system’s logs each evening.
For 30 consecutive days in August 2026, he sent me two things daily: the system’s total kWh output for that day, and a note on the general sky conditions — clear, partly cloudy, overcast, or rainy — cross-checked against local weather station data for that county. I categorized each day into one of four buckets based on cloud cover, then averaged the production within each bucket to see how output actually shifted as skies changed.
I want to be upfront about the limits here too: this is one system, one location, one month. It’s not a substitute for peer-reviewed research across many climates and seasons. But it is real, unfiltered, single-household data — the kind most homeowners never actually get to see, because most published numbers come from lab conditions or aggregated regional averages rather than one specific rooftop’s day-to-day reality.
2. Clear-Sky Baseline: What “Full Output” Looked Like
To make sense of the cloudy-day numbers, we first needed a clear-sky baseline from the same system, same month, same panel angle. Across the handful of genuinely clear days in our 30-day window, the 6 kW system averaged production comfortably in line with its rated capacity for August sun hours in Virginia, generating strong output from early morning through late afternoon with a clean, predictable bell-curve production graph on the monitoring app.
This baseline matters because “how much power did cloudy days produce” is meaningless without a reference point. A cloudy day producing 2 kWh sounds unimpressive in isolation, but tells a very different story once you know the same system was producing 30+ kWh on a clear day nearby. Percentages, not raw numbers, are what actually let you compare fairly.
One thing I noticed almost immediately reviewing his logs: even “clear” days weren’t perfectly uniform. A stray patch of cloud drifting past around midday could dip production briefly before it recovered. This is normal, and it’s actually a preview of the more interesting phenomenon we saw more dramatically on the partly cloudy days, which I’ll get to shortly.
3. Heavy Overcast Days: The Real Numbers
This is the category everyone actually wants to know about — the genuinely gray, thick-cloud days where you can’t see a shadow at noon. We logged seven such days during the 30-day window, and the system’s output on these days averaged roughly 15–20% of the clear-sky baseline, translating to meaningful but clearly reduced generation throughout the day.
This lines up closely with the general industry consensus I’ve seen cited across multiple sources — most reporting that heavily overcast conditions typically produce somewhere in the 10-25% range of a panel’s rated output. Our real-world data sat comfortably inside that range, which was honestly reassuring; it meant my friend’s system wasn’t underperforming or overperforming relative to what physics and prior research would predict.
What struck me more than the average, though, was the shape of the production curve on these days. Rather than a clean bell curve, heavy overcast days produced a flatter, choppier line — smaller peaks and dips throughout the day as cloud thickness varied hour to hour, rather than one predictable midday peak. The panels were clearly still working, converting the diffuse, scattered light coming through the cloud layer into usable electricity, just at a meaningfully reduced rate compared to direct sun.
4. Partly Cloudy Days: The Most Interesting Data
If heavy overcast days confirmed what I expected, partly cloudy days genuinely surprised me. We logged eleven days that qualified as partly cloudy — a mix of sun and passing cloud cover throughout the day — and average output on these days came in far higher than I’d anticipated, landing in the 55–75% range of clear-sky baseline production.
The real standout moments, though, were brief production spikes that occasionally pushed instantaneous output above the system’s rated capacity for a few minutes at a time. This lines up with something solar researchers call the “edge of cloud” effect — when sunlight hits the edge of a passing cloud, the cloud can act almost like a lens, briefly focusing and reflecting extra light onto the panels. My friend actually caught this happening live on his monitoring app one afternoon and sent me a screenshot, slightly baffled that his system was reporting output above its rated capacity for a few minutes before settling back down.
These super-peaks don’t last, and they don’t meaningfully change the day’s total output on their own. But they’re a great reminder that solar production isn’t a simple, static relationship with cloud cover — it’s genuinely dynamic, and partly cloudy skies can behave in ways that feel counterintuitive until you understand the physics behind it.
5. Rainy Days: The Bottom of the Range
We had five days with steady rain during the window, and unsurprisingly, these produced the lowest output of the month, averaging somewhere around 10–15% of clear-sky baseline. Heavy rain clouds are thick and block a much larger share of usable light compared to thinner overcast conditions, so this result tracked with expectations.
There was a small silver lining my friend pointed out, and one I’ve heard from other solar owners before too: rain has a cleaning effect on panels. Dust, pollen, and general grime that build up on panel surfaces over weeks of dry weather get washed away during rain, which can give a small boost to production efficiency on the sunnier days that follow. We didn’t have a clean enough dataset to isolate this “post-rain boost” precisely within our 30-day window, but the production numbers on the two clear days immediately following rain did come in slightly higher than earlier clear days in the month — consistent with, though not conclusive proof of, that cleaning effect.
The bigger takeaway from the rainy days specifically: even in the worst weather conditions of the month, the system never dropped to zero. It kept producing something, every single day, all 30 days, without exception.
6. The Full 30-Day Averages, By the Numbers
Pulling it all together, here’s how the month broke down by day type and average output relative to clear-sky baseline: clear days averaged full baseline production as expected; partly cloudy days averaged roughly 55–75%; heavy overcast days averaged roughly 15–20%; and rainy days averaged roughly 10–15%.
Across the full 30 days combined, factoring in the actual mix of weather Virginia handed us that August, the system’s total monthly output landed meaningfully below what a theoretical “30 clear days” month would have produced, but nowhere near the doom-and-gloom “solar doesn’t work without sun” narrative some people still believe. The system generated usable, grid-feeding electricity on every single day of the month, including the five rainy days and seven heavy overcast days combined — nearly 40% of the entire tracking period.
For a homeowner sizing a system or deciding whether solar makes sense in a climate that isn’t perpetually sunny, this monthly blended average is honestly the more useful number than either the clear-sky maximum or the worst-case overcast minimum in isolation. Real months are a mix, and this data shows what that realistic mix actually produces.
7. What This Means If You’re Considering Solar
If there’s one practical lesson from this experiment, it’s this: don’t evaluate solar based on your area’s cloudiest day, and don’t evaluate it based on your sunniest day either. Evaluate it based on your region’s realistic monthly and annual weather mix, the same way we tracked this Virginia system across a genuinely varied 30-day stretch.
This is exactly why a proper site assessment matters before installation — a good installer should be pulling historical weather and sun-hour data for your specific location, not just quoting generic regional averages. Two homes twenty miles apart can have meaningfully different cloud patterns depending on local geography, and that difference shows up directly in your system’s real-world output over a year.
It’s also worth remembering that oversizing your system slightly, if your roof space and budget allow, is a genuinely smart hedge against cloudy stretches. A system sized for your exact average daily need on clear days will underdeliver noticeably during overcast weeks, while a modestly oversized system smooths out that variability better across a full month like the one we tracked.
Conclusion
Thirty days of real production data from one Virginia rooftop won’t settle every debate about solar performance, but it does replace vague reassurance with actual numbers: roughly 15–20% output on heavy overcast days, 55–75% on partly cloudy days, and never a complete shutdown, even through five straight days of rain. What stood out most to me wasn’t any single statistic — it was the consistency. The panels kept producing, every single day, adjusting automatically to whatever light was actually available. If you’re weighing solar for a home in a climate that isn’t guaranteed sunshine year-round, this is the kind of ground-level, honest data that should factor into your decision far more than either the marketing brochures or the skeptics’ worst-case claims.
Disclaimer: This article reflects informal, real-world monitoring data from a single residential solar system over one month and is intended for general educational purposes only, not as scientific or engineering guidance. Actual solar output varies significantly by location, equipment, panel angle, and season. Consult a licensed solar installer for a proper site-specific production assessment before making purchase decisions.