Work out your daily watt-hour consumption, divide by your location’s peak sun hours, then add 30–40% for real-world losses. That gives the panel wattage you need. Size the battery to hold one to two days of that consumption. The arithmetic is simple; the part that catches people out is that panels deliver far less than their rated wattage, every single day.
The formula, in one box
Panel watts = (daily watt-hours ÷ peak sun hours) ÷ 0.7
Battery watt-hours = daily watt-hours × days of autonomy ÷ 0.85
The 0.7 is the derate factor — the share of rated panel output you actually collect once temperature, angle, dirt, wiring and controller losses have taken their cut. The 0.85 accounts for inverter losses on the battery side. Everything below is an explanation of how to fill in the three inputs honestly.
Step 1: your daily consumption
Everything downstream depends on this number, so it is worth getting right rather than guessing. List every load, multiply watts by hours used per day, and total it.
| Load | Watts | Hours/day | Watt-hours/day |
|---|---|---|---|
| 12V compressor fridge | 45W avg | 24 | 400–600Wh |
| Laptop | 50W | 6 | 300Wh |
| Starlink Mini | 40W | 10 | 400Wh |
| LED lighting | 20W | 5 | 100Wh |
| Phone and tablet charging | 15W | 4 | 60Wh |
| Water pump | 60W | 0.5 | 30Wh |
| Roof fan | 25W | 8 | 200Wh |
| Induction hob | 1,400W | 0.3 | 420Wh |
A van with a fridge, a laptop, Starlink and lights lands around 1,200–1,800Wh per day. A weekend camping setup without a fridge is closer to 200–400Wh. A small cabin with a well pump and a chest freezer can reach 3,000Wh.
Two pieces of advice on this step. First, use average draw rather than rated draw for anything that cycles — fridges and freezers especially, which is covered in detail in our refrigerator wattage guide. Second, if the system already exists, measure rather than estimate: a plug-in energy monitor or the battery’s own app will tell you the truth in a week, and the truth is usually 20–30% above the estimate.
Step 2: peak sun hours where you live
A peak sun hour is one hour of sunlight at 1,000 watts per square meter — the standard test condition panels are rated under. It is not the same as daylight hours. A location with 14 hours of daylight in June may deliver only 6 peak sun hours, because early and late sun arrives at a shallow angle through more atmosphere.
| Region | Summer | Winter | Annual average |
|---|---|---|---|
| Southwest (AZ, NM, NV) | 7–8 | 4–5 | 5.5–6.5 |
| California, Texas, Florida | 6–7 | 3.5–4.5 | 5–5.5 |
| Mountain West | 6–7 | 3–4 | 4.5–5.5 |
| Midwest | 5.5–6.5 | 2–3 | 4–4.5 |
| Northeast | 5–6 | 1.5–2.5 | 3.5–4.5 |
| Pacific Northwest | 5–6 | 1–2 | 3.5–4 |
Use the winter column if the system has to work year-round. Sizing on the annual average means a system that works beautifully in June and leaves you short every December, which is precisely when outages and heating needs peak.
For a location-specific figure rather than a regional one, the National Renewable Energy Laboratory’s PVWatts calculator is free and takes a postal code. It is worth five minutes.
Step 3: the derate factor
Panels are rated under laboratory conditions: 1,000 W/m² of light, a cell temperature of 25°C, and perfectly perpendicular light. Your roof is not a laboratory. Losses stack up:
| Loss | Typical | Notes |
|---|---|---|
| Temperature | 10–20% | Panels lose ~0.3–0.4% per °C above 25°C |
| Angle and orientation | 5–20% | Flat mounting on a van roof is the common case |
| Dirt, dust and pollen | 2–5% | More in dry, dusty regions |
| Wiring and connections | 2–3% | Worse with undersized cable |
| Charge controller | 3–10% | MPPT at the low end, PWM at the high end |
| Panel tolerance and ageing | 2–3% | Plus ~0.5% per year thereafter |
| Shading | 0–100% | See below — this one is not linear |
Multiply it out and 0.7 is a realistic planning figure for a typical installation. A well-oriented, well-ventilated, unshaded array with MPPT might reach 0.8. A flat-mounted van roof in Arizona in August might be closer to 0.6.
Shading deserves special attention because it is not proportional. Panels wire cells in series, so shading one corner can cut output from the whole panel — sometimes by 50% or more from a shadow covering a fraction of the surface. A single vent, an antenna or an overhanging branch can cost far more than its size suggests. Bypass diodes help, and panel-level optimizers or microinverters help more, but the real fix is not putting panels where shadows fall.
Sizing the battery
Panels generate during the day; you consume around the clock. The battery bridges the gap, and it needs to cover both the overnight period and the days when the sun does not appear.
Days of autonomy is the design choice. One day means one cloudy day empties you. Two days is the common compromise. Three or more is for remote installations where being without power is a serious problem.
- 1,500Wh/day × 2 days ÷ 0.85 = ~3,500Wh of battery
Two adjustments to that figure. With LiFePO4 you can use most of the rated capacity, so the calculation above works directly. With older lead-acid batteries you must not discharge below 50%, so you would need double — one of several reasons lead-acid has effectively disappeared from new installations. Our chemistry guide covers why LiFePO4 won.
Also allow for cold. A battery in an unheated space delivers 70–80% of its rated capacity at freezing, and most cannot be charged below 0°C at all without integrated heating.
Charge controllers: MPPT vs PWM
The controller sits between the panels and the battery and decides how much of the available power actually arrives.
- PWM is a simple switch. It pulls the panel down to battery voltage, wasting the difference. Cheap, and acceptable only where panel voltage closely matches battery voltage.
- MPPT tracks the panel’s maximum power point and converts excess voltage into additional current. It typically delivers 15–30% more energy from the same panels, and the advantage is largest in cold weather and low light.
For any system above about 200W, MPPT pays for itself quickly. Nearly all power stations with solar input use MPPT internally, so if you are charging a portable unit from panels this decision is already made for you. What you do need to check on a portable unit is the maximum solar input voltage and wattage — exceeding the voltage limit can damage the unit, and many accept far less wattage than buyers assume.
Three worked examples
Weekend camping, 300Wh/day, Colorado summer
300 ÷ 6 peak sun hours = 50W, divided by 0.7 = 71W of panel. Round to 100W. Battery: 300 × 1.5 days ÷ 0.85 = 530Wh, so a 500–700Wh power station. A 100W folding panel and a unit like the EcoFlow River 2 ProCheck price on Amazon covers this comfortably.
Full-time van, 1,500Wh/day, year-round travel
Using 4 peak sun hours to allow for travel through less sunny regions: 1,500 ÷ 4 = 375W, divided by 0.7 = 536W of panel. Round to 600W, which is three 200W panels or four 150W. Battery: 1,500 × 2 ÷ 0.85 = 3,530Wh. That points at a 3,000–4,000Wh system, such as an expandable EcoFlow Delta Pro 3Check price on Amazon or a comparable Bluetti setup.
Most vans are roof-area limited before they are budget limited — 600W is roughly the practical ceiling on a long wheelbase van, and it is worth measuring the roof before finalizing any of this.
Small cabin, 3,000Wh/day, Pacific Northwest
Winter is the binding constraint: 1.5 peak sun hours. 3,000 ÷ 1.5 = 2,000W, divided by 0.7 = 2,857W of panel to run through winter on solar alone. That is a large array, and it is why remote cabins in cloudy regions almost always pair solar with a generator rather than sizing for December. A more economical design is 1,200–1,500W of panel plus a generator for the worst weeks, or accepting reduced consumption in winter.
Sizing for winter, not for July
This is the mistake that produces the most disappointment, so it is worth stating on its own. In the northern US, December delivers roughly a third of June’s energy from the same array. Days are shorter, the sun sits lower, cloud cover is heavier, and snow can cover panels entirely for days.
Three ways to respond:
- Size for winter and accept large summer surplus. Expensive, and the right answer for genuinely off-grid installations where there is no fallback.
- Size for the shoulder seasons and plan a backup for deep winter — a generator, shore power, or reduced consumption.
- Tilt the array. Steeper angles favor low winter sun and shed snow. On ground mounts and cabins this is nearly free performance; on vans and flat roofs it usually is not practical.
Common sizing mistakes
- Using rated panel wattage as if you will get it. You will get about 70%, on a good day.
- Sizing on summer sun. The system then fails exactly when you need it.
- Underestimating consumption. Measured figures come in well above estimates almost every time.
- Ignoring shading. A small shadow costs a large fraction of output.
- Buying panels the controller cannot accept. Check maximum input voltage and wattage before ordering.
- Undersizing the wiring. Voltage drop on long, thin DC runs wastes real energy and is easy to avoid.
- Forgetting that panels degrade. Roughly 0.5% a year, so a 25-year-old array delivers about 85% of new.
Frequently asked questions
How many solar panels do I need?
Divide your daily watt-hour consumption by your local peak sun hours, then divide by 0.7 to allow for real-world losses. For a van using 1,500Wh per day at 4 peak sun hours, that is roughly 536W — three 200W panels. Use winter peak sun hours if the system must work year-round.
Why do solar panels never produce their rated wattage?
Panels are rated at 25°C with light hitting them perpendicular at 1,000 W/m². Real installations lose 10–20% to heat, 5–20% to imperfect angle, a few percent each to dirt, wiring and the charge controller, and potentially far more to shading. A 200W panel realistically delivers 130–160W in good conditions.
What are peak sun hours?
One peak sun hour is one hour of sunlight at 1,000 watts per square meter, the standard condition panels are rated under. It is not the same as daylight hours — a 14-hour summer day might yield only 6 peak sun hours, because early and late sunlight arrives at a shallow angle. Most of the US averages 3.5–6.5 annually, with winter figures often less than half the summer ones.
How big a battery do I need with my solar panels?
Enough to cover your consumption for the number of days you want to survive without sun, divided by about 0.85 for inverter losses. Two days of autonomy is the usual compromise: at 1,500Wh per day that is roughly 3,500Wh of LiFePO4 storage. In cold conditions, allow for the battery delivering 70–80% of its rating.
Is MPPT worth it over PWM?
For anything above about 200W, yes. MPPT converts excess panel voltage into usable current and typically delivers 15–30% more energy from the same panels, with the biggest advantage in cold and low light. Nearly all power stations use MPPT internally, so the decision only arises when building a system from components.
Can I add more panels later?
Usually, within limits. The charge controller has a maximum input voltage and wattage that cannot be exceeded, and mixing panels of different specifications in one string reduces the whole string to the weakest member. If expansion is likely, buy a controller with headroom now and keep additional panels on their own string or input.
Do solar panels work in winter and cloudy weather?
Yes, but at much reduced output — typically 10–25% of rated on a heavily overcast day, and in the northern US roughly a third of summer energy in December. Panels are actually more efficient in cold air; the problem is short days, low sun angles and cloud, not temperature. Snow cover stops output entirely until it clears or slides off.
Related reading. What size power station do I need? for the battery side in detail, plug-in solar if you are supplementing a home rather than going off-grid, and solar generator packages for bundles that pair panels and storage.