TL;DR — Off-Grid Solar System Sizing
Correct off-grid solar system sizing follows one sequence: calculate worst-case daily load, size battery bank for that load plus days of autonomy, size panel array for winter production at your latitude, size inverter for peak simultaneous demand, size charge controller for the panel array. Every deviation from this sequence produces an undersized system. The most common failure point is using summer data to size a winter system.
A veteran in rural Montana installed a 4kW solar system with a 10kWh battery bank in August. The system ran beautifully through September and October. By December his battery bank was cycling to 20% state of charge every night — well below the LiFePO4 rated depth of 80%. By March of the following year two of his four battery modules showed permanent capacity loss. The contractor had sized the system on July load data. Montana in July has 7 peak sun hours. Montana in December has 3. That one substitution cost him $8,400 in battery replacement.
Table of Contents
The error: sizing for average instead of worst case
Undersizing happens when designers use the wrong inputs. The two most common wrong inputs:
Wrong input 1: Annual average peak sun hours instead of winter minimum. Peak sun hours vary from 3.5 in Seattle in December to 7 in Phoenix in July. A system sized on annual averages underperforms every winter. The months when you need the most stored power are the months when panels produce the least.
Wrong input 2: Current appliance load instead of future load. A homestead that adds a chest freezer, an electric water heater, or a shop tool in year two is operating a system sized for year one. Build in 20–30% headroom.
"Residential solar PV systems in northern U.S. climates experience a seasonal production variation of 40–55%, with December generation averaging 45% below June generation for the same system in states above the 40th parallel."
— National Renewable Energy Laboratory, PVWatts Calculator Regional Analysis, 2024
A 40–55% production drop paired with higher winter loads — lighting runs longer, heating draws more power, hot water demand increases in cold weather — means the worst-case scenario occurs every December through February. Size for that window. The rest of the year takes care of itself.
Step 1: Calculate worst-case daily load
Build two load tables: summer and winter. Identify the loads specific to each season.
Summer additions: Air conditioning (if electric), attic fans, irrigation pumps, outdoor lighting.
Winter additions: Auxiliary heating (electric resistance or heat pump in extreme cold), more interior lighting hours (shorter days), higher refrigerator efficiency loss in freezing locations (paradoxically, refrigerators work harder when ambient air is very cold in unheated spaces), water heating demand.
Take the larger number — usually winter for most northern and central US locations — as your design load number.
| Appliance | Summer Wh/day | Winter Wh/day |
|---|---|---|
| Refrigerator + freezer | 1,800 | 2,100 (longer cycle in cold) |
| LED lighting | 400 | 700 (4 more dark hours) |
| Laptop + routers | 600 | 600 |
| Well pump | 375 | 375 |
| Mini-split (1.5 ton) | 3,600 | 4,800 (heating mode) |
| Washing machine | 500 | 500 |
| Misc | 600 | 600 |
| Total | 7,875 Wh | 9,675 Wh |
This example home uses 23% more power in winter. The panel array produces 40–50% less. Size for winter.
Step 2: Size the battery bank for autonomy
Formula:
Battery bank (Wh) = Daily load (Wh) × Days of autonomy ÷ Usable depth of discharge
Days of autonomy: How many days can your system run with zero solar input? Two days is the minimum for a primary residence. Three days is more appropriate for locations with regular multi-day overcast periods.
Usable depth of discharge: LiFePO4 — 80%. AGM — 50%. Flooded lead-acid — 50%.
Example with winter load of 9,675 Wh:
LiFePO4, 3 days autonomy: 9,675 × 3 ÷ 0.80 = 36,281 Wh → round to 40kWh
AGM, 2 days autonomy: 9,675 × 2 ÷ 0.50 = 38,700 Wh → round to 40kWh
Same total bank size — but the LiFePO4 system has 3 days of autonomy while AGM has 2. And LiFePO4 lasts 4,000–6,000 cycles versus AGM's 400–600. The math favors LiFePO4 significantly over a ten-year period.
The Battle Born 100Ah LiFePO4 is the battery Wattson rebuilt his own system with. For a 40kWh bank at 48V, you need approximately 833Ah of capacity — eight 100Ah batteries wired in appropriate series/parallel configuration. Check current pricing and availability on Amazon before comparing alternatives.
Step 3: Size the panel array for winter production
Formula:
Panel array (W) = Daily load (Wh) ÷ Winter peak sun hours ÷ System efficiency
System efficiency: Account for wire losses, charge controller efficiency, inverter efficiency. Typical combined efficiency: 80–85%. Use 0.80 for conservative sizing.
Example with winter load 9,675 Wh, winter peak sun hours 3.5 (northern location):
9,675 ÷ 3.5 ÷ 0.80 = 3,455W → size to 3,500–4,000W of panels
At 400W per monocrystalline panel: 9–10 panels minimum.
Compare to summer sizing at 7 peak sun hours: 7,875 ÷ 7 ÷ 0.80 = 1,406W → 4 panels.
The difference is extreme. A system sized for summer requires less than half the panels of one sized for winter at the same latitude. The Montana veteran's contractor used summer data — 7 hours — when he should have used winter data — 3 hours.
Use NREL's PVWatts Calculator to find precise peak sun hours for your location and month. It is free and accurate to the zip code.
Step 4: Size the inverter for peak demand
Formula:
Inverter size (W) = Peak simultaneous load (W) × 1.25 buffer
The peak simultaneous load is the worst case — every high-draw appliance running at once. The well pump startup surge (typically 2–3× running wattage), the mini-split compressor engaging, and the refrigerator cycling simultaneously.
Example:
- Well pump running: 750W; startup surge: 2,250W
- Mini-split running: 1,500W
- Refrigerator running: 150W
- Lighting and misc: 200W
- Peak simultaneous (using pump surge): 4,100W
Inverter size: 4,100 × 1.25 = 5,125W → select a 5kW or 6kW pure sine wave inverter
The AIMS Power pure sine wave inverter and the Victron MultiPlus are the field-proven standards for this range. Check current specifications and pricing on Amazon before selecting.
Note: Victron requires installation by a certified installer to maintain warranty coverage on some MultiPlus models. Verify warranty terms before purchasing for a DIY installation.
Step 5: Size the charge controller
Formula:
Charge controller (A) = Panel array (W) ÷ Battery bank voltage (V)
Example: 3,500W panel array at 48V system: 3,500 ÷ 48 = 72.9A → select an 80A or 100A MPPT charge controller
The Victron SmartSolar MPPT is available in 60A, 85A, and 100A configurations. The 100A model handles up to 4,800W of panels at 48V and includes Bluetooth monitoring. If you install additional panels in year two, the controller already has headroom.
Use MPPT only. PWM wastes 25–40% of your harvest. On a 3,500W array running six hours per day, that is 875–1,400 lost watt-hours per day. Across one winter month, it is 26–42kWh of energy you paid for in panel costs and did not capture.
The complete sizing worksheet
| Step | Input | Calculation | Example Result |
|---|---|---|---|
| 1: Worst-case daily load | Appliance list | Wattage × hours, sum all | 9,675 Wh/day |
| 2: Battery bank | Load × autonomy ÷ DoD | 9,675 × 3 ÷ 0.80 | 36,281 Wh → 40kWh |
| 3: Panel array | Load ÷ sun hours ÷ efficiency | 9,675 ÷ 3.5 ÷ 0.80 | 3,455W → 4kW |
| 4: Inverter | Peak load × 1.25 | 4,100 × 1.25 | 5,125W → 6kW |
| 5: Charge controller | Array W ÷ system V | 3,500 ÷ 48 | 73A → 80A MPPT |
Run these five steps in order. Do not skip to step three. Do not start at step five. The sequence ensures every component is sized relative to the load it must serve.
The free Solar Power Estimator runs this sequence automatically for your specific location and appliance list. It outputs a complete sizing recommendation including component specifications.
🦍 WATTSON ON WINTER SIZING: "I have rebuilt systems where the contractor used peak sun hours from May to size a Montana system that had to run through January. The contractor is back in his warm office in May. You are the one discovering that your batteries are at 15% state of charge at 5 PM on a February Tuesday. Size every system for the worst month at your location. If it runs in December, it runs in June with power to spare."
Run Your Complete System Sizing Now
The Solar Power Estimator runs all five sizing steps for your location and load. Free. Takes ten minutes. Produces a component-level specification sheet you can bring to any vendor conversation.
Frequently Asked Questions
What is the most common solar system sizing mistake?
Using summer peak sun hours instead of winter minimums to size the panel array. In northern climates, this leads to a system that works well in summer and fails to keep batteries charged through winter. Size your panel array for December production at your latitude, not June.How many days of autonomy should my off-grid battery bank provide?
Minimum two days for a primary residence. Three days is the recommended target in climates with regular multi-day overcast or storm periods. In areas with reliable year-round sunshine, two days of autonomy is generally sufficient. In the Pacific Northwest or upper Midwest, plan for three days minimum.How big of a solar system do I need to run a whole house off-grid?
A modest off-grid home running 4,000–6,000 Wh per day needs a 2–4kW panel array with a 15–25kWh battery bank. A full residential home running 8,000–12,000 Wh per day needs a 4–8kW panel array with a 25–50kWh battery bank. Both figures assume a northern US location sized for winter. Use the Solar Power Estimator for your specific location and load.What happens if my solar system is undersized?
An undersized battery bank discharges deeper than its rated depth every night and on cloudy days. Deep discharge cycles damage battery chemistry — shortening LiFePO4 from 4,000 cycles to 1,500–2,000 and destroying lead-acid banks in months rather than years. You end up replacing the battery bank years before it should have needed replacement, paying twice for the most expensive system component.Can I add more panels later if I undersized initially?
Yes — if your charge controller has headroom for a larger array and your battery bank has sufficient capacity to accept the additional charge. Most MPPT controllers are available in sizes with significant overhead above the minimum. Install a controller rated 20–30% above your initial array size so future expansion requires only adding panels, not replacing the controller.Should I use peak or average sun hours for sizing?
Use local minimum monthly averages — specifically December peak sun hours for your location. Annual averages are optimistic. Peak hours on the best summer day are irrelevant for winter system design. NREL's PVWatts Calculator provides monthly production estimates for any US location. Use the December output to size your system.What is depth of discharge and why does it matter for system sizing?
Depth of discharge (DoD) is the percentage of a battery's total capacity that can be used before recharging. LiFePO4 has an 80% usable DoD — you can use 80% of the rated capacity without damage. AGM lead-acid has a 50% usable DoD. When sizing your battery bank, divide your required storage by the battery's usable DoD fraction. A 10kWh daily load needing two days of autonomy requires 25kWh of lead-acid or 25kWh of LiFePO4 capacity — but in total bank size those figure to 50kWh AGM vs 25kWh LiFePO4.How do I account for system losses in solar sizing?
System losses include: wire resistance (2–5%), charge controller efficiency loss (3–5%), inverter efficiency loss (5–10%), battery charge/discharge round-trip loss (5–10%), and temperature derating on panels in hot weather. Combined, a typical off-grid system operates at 75–85% efficiency from panel output to delivered AC power. Use 0.80 as a conservative efficiency factor in your panel array sizing calculation.Does location significantly affect how I size my solar system?
Yes — dramatically. Phoenix has 6.5 winter peak sun hours. Seattle has 2.5. A system sized to meet the same load in Seattle requires 2.6 times the panel capacity as the same system in Phoenix. Roof tilt and panel orientation also affect output by 10–20%. Use NREL's PVWatts for your exact address and tilt angle before finalizing any sizing calculation.What is the 125% rule for charge controllers?
NEC Article 690 requires that solar charge controllers be sized for at least 125% of the short-circuit current of the panel array. This means if your panel array has a combined short-circuit current (Isc) of 50A, your charge controller must be rated for at least 62.5A — select the next standard size up, typically 70A or 80A. This protects the controller from overcurrent damage during peak irradiance conditions.Size it right or pay twice
The Montana veteran paid $8,400 to fix a calculation that took the contractor thirty minutes to get wrong. The correct calculation takes you the same thirty minutes to get right. Winter peak sun hours. Worst-case daily load. Three days of autonomy. Those three inputs drive every component decision.
The system either runs on December data or fails on December nights. There is no third option.
Three inputs change everything: worst-case load, winter sun hours, and days of autonomy. Every undersized system skipped at least one of them. The Solar Power Estimator runs all three automatically for your location. Run it before you price anything. Then bring the output to every vendor conversation. If their quote deviates from your numbers, ask them to show the math.