Solar Panel & Battery Size Calculator
Panels, batteries and controller for your daily kWh — NASA sun data for 237 cities.
2. Sunshine
Typed in. Choose a city to use NASA POWER climatology instead.
3. Panels and losses
What the 14% PVWatts default is made of
- Soiling2%
- Shading3%
- Snow0%
- Mismatch2%
- Wiring2%
- Connections0.5%
- Light-induced degradation1.5%
- Nameplate rating1%
- Age0%
- Availability3%
Combined as 1 − Π(1 − loss) = 14.08% (NREL PVWatts Version 5 Manual). Temperature and inverter losses are counted separately.
4. Battery bank
5. Charge controller and cold-weather check
How it was calculated
Month by month with this design
| Month | Sun hours | Delivered / day | vs load |
|---|
Results are estimates from the formulas shown, not a professional design or certification. Have a qualified engineer verify anything safety-critical (structures, electrical installations, gas or pressure systems).
About the Solar Panel & Battery Size Calculator
An off-grid solar system has to make, store and deliver your daily energy. This calculator works through that chain: it adds up your appliances (or takes a daily or monthly kWh figure), allows for inverter and battery losses, and divides by the peak sun hours of the month you design for to get the panel watts and the number of panels. It then sizes the battery bank for the days of autonomy and depth of discharge you choose, lays the batteries out in series and parallel for a 12, 24 or 48 V system, and checks the charge controller: its output current, the PV input current of 1.25 × Isc, and the string voltage on the coldest morning against the controller’s limit.
Peak sun hours for 237 cities — 83 of them in India — come from NASA POWER’s 2001–2020 climatology, month by month, for flat panels, panels tilted at the latitude and latitude + 15°. Everything runs in your browser.
How to use it
- List your appliances — watts, hours per day and how many — or switch to Total per day and enter kWh per day, or the units (kWh) on a monthly electricity bill.
- Choose your city for NASA sun data, the panel tilt, and the month to design for: Worst month keeps an off-grid system running all year, Annual average suits a grid-connected (hybrid) system. Or type the peak sun hours for your site.
- Enter the panel’s datasheet values (Wp, Voc, Isc and the temperature coefficients), and check the losses: 14% system losses is the NREL PVWatts default, and the temperature loss comes from the cell temperature.
- Choose the battery type, system voltage and days of autonomy, and the voltage and Ah of the batteries you plan to buy.
- Enter the charge controller’s maximum PV voltage and the coldest temperature at the site. Read the result, the monthly balance and How it was calculated; copy or download the summary for your installer.
Examples
1,678 Wh/day (lights, fans, router, TV), 5 peak sun hours, 550 W panels (Voc 49.6 V, Isc 14 A), 50% DoD, 1 day
614 W needed → 2 × 550 W (1 in series × 2 strings); battery 3.73 kWh → 4 × 12 V 150 Ah (2 in series × 2); controller ≥ 45.8 A; PV current 35 A
300 kWh per month
300 ÷ 30.44 = 9.86 kWh per day to size for
Voc 49.6 V, −0.27 %/°C, coldest morning 0 °C
49.6 × [1 + 0.0027 × 25] = 52.95 V per panel → 105.9 V for two — too much for 100 V; use a 150 V controller or one panel per string
NASA POWER 2001–2020
January: 3.67 peak sun hours (annual average 4.90)
The method
- Energy at the battery: E_dc = E_load ÷ η_inverter (for AC loads)
- Energy the panels must deliver: E_pv = E_dc × share from solar ÷ η_battery — off-grid sizing assumes all of it passes through the battery
- Derating: f = (1 − system losses) × (1 − temperature loss) × η_MPPT, with temperature loss = −γ_Pmax × (T_cell − 25 °C)
- Array size: P = E_pv ÷ (peak sun hours × f), rounded up to whole panels
- Battery: C = E_dc × days of autonomy ÷ depth of discharge; batteries in series = system voltage ÷ battery voltage, and enough parallel strings for C
- Cold Voc: V_oc,max = V_oc × [1 + β_Voc × (T_min − 25 °C)] per panel, times the panels in series — NEC 690.7(A)(1), or the NEC Table 690.7(A) factor
- PV current: I_max = 1.25 × I_sc × strings (NEC 690.8(A)(1)); cables and fuses on that circuit need 125% of it (690.8(B))
- Controller output current: about P_array ÷ V_battery
Peak sun hours and the NASA data
Peak sun hours are the day’s solar energy on the panel in kWh/m², which equals the hours of full 1,000 W/m² sun. The city table is NASA POWER climatology for January 2001 – December 2020: irradiation on an equator-facing surface that is flat, tilted at the latitude or at latitude + 15° (POWER’s SI_EF_TILTED_SURFACE values). They are averages for the POWER grid cell around each city, about 1° across, so a valley, a coast or local haze can differ. For another place, look up your coordinates in the NASA POWER Data Access Viewer and type the value.
These data were obtained from the NASA Langley Research Center (LaRC) POWER Project funded through the NASA Earth Science/Applied Science Program.
Losses that matter
The 14% system losses are the default of NREL’s PVWatts Version 5 (the laboratory is now the National Laboratory of the Rockies): soiling 2%, shading 3%, snow 0%, mismatch 2%, wiring 2%, connections 0.5%, light-induced degradation 1.5%, nameplate rating 1%, age 0% and availability 3%, combined as 1 − Π(1 − loss) = 14.08%. PVWatts models temperature and the inverter separately, and so does this calculator: hot panels lose −γ_Pmax per degree above 25 °C (crystalline panels list about −0.3 to −0.4 %/°C), the MPPT controller and inverter have their own efficiencies, and a battery returns less energy than it took in — about 80% for lead-acid and 95% for LiFePO4 are used as starting values.
The cold-weather voltage check
A panel’s open-circuit voltage rises as it gets colder, and it is highest at sunrise on the coldest day. If the string voltage then exceeds the charge controller’s maximum PV input voltage, the controller can be damaged. NEC 690.7(A) asks for the lowest expected ambient temperature (ASHRAE’s extreme annual mean minimum design dry-bulb temperature is the usual source) and the module’s Voc temperature coefficient; without the coefficient, Table 690.7(A) gives factors from 1.02 (20–24 °C) to 1.25 (−36 to −40 °C) for crystalline silicon. When you pick a city, the calculator also shows the lowest hourly temperature in NASA POWER’s 2001–2020 record for that grid cell.
Sources
- A. P. Dobos (2014), PVWatts Version 5 Manual, NREL/TP-6A20-62641 — 14% default system losses
- NFPA 70, National Electrical Code: 690.7 (maximum PV system voltage, Table 690.7(A)) and 690.8 (circuit sizing and current)
- IEC 62548-1:2023, Photovoltaic (PV) arrays — Part 1: Design requirements
- NASA POWER climatology API (v2.10.0), 2001–2020, community RE: SI_EF_TILTED_SURFACE and T2M_MIN
Limitations
- Peak-sun-hour sizing uses monthly averages. It cannot show a run of cloudy days; days of autonomy and a generator or grid backup cover those.
- City values are for a NASA grid cell about 1° across and an unshaded, equator-facing panel. Shade, a different direction or a dusty site lower the output.
- The calculator assumes an MPPT charge controller. A PWM controller holds the panels at the battery voltage, so panels must be matched to the battery and deliver less.
- It does not size cables, fuses, earthing or the inverter’s surge rating — check those against local rules (for example IEC 60364-7-712 or NEC 690) and the equipment datasheets.
- Location lookup covers the 237 cities in the list; for any other place, type the peak sun hours (the NASA POWER Data Access Viewer gives them for any coordinates).
Privacy
Everything happens in your browser. What you enter or open here is not uploaded or stored by MySmartCoPilot.
Frequently asked questions
How many solar panels do I need for 5 kWh per day?
Divide by the peak sun hours and the losses. With 5 peak sun hours, 14% system losses, 7% temperature loss, a 95% MPPT controller, a 90% inverter and 95% LiFePO4 efficiency, 5,000 ÷ 0.9 ÷ 0.95 = 5,848 Wh must come from the panels, and 5,848 ÷ (5 × 0.86 × 0.93 × 0.95) = 1,539 W — three 550 W panels.
What are peak sun hours?
The daily solar energy on the panel, in kWh/m², expressed as hours of full 1,000 W/m² sun. A day with 5 kWh/m² is 5 peak sun hours, whether the sun shone weakly for 10 hours or strongly for 5.
Should I size for the worst month or the annual average?
For an off-grid system that must run all year, use the worst month — in Delhi that is January at about 3.7 peak sun hours on a latitude tilt, against 4.9 for the year. A grid-connected or generator-backed system can be sized for the annual average and import the shortfall.
How big a battery do I need?
Daily energy at the battery × days of autonomy ÷ depth of discharge. 1,864 Wh a day for one day at 50% depth of discharge needs 3,729 Wh — 155 Ah at 24 V, which is two strings of two 12 V 150 Ah batteries.
How do I size the charge controller?
Its output (battery) current must cover the array: watts ÷ battery voltage, 1,100 W ÷ 24 V = 46 A, so a 50 A controller. Its PV input must accept 1.25 × Isc of the parallel strings, and the cold-morning string Voc must stay below its maximum input voltage.
Why does the temperature matter twice?
Heat lowers the panels’ power (the temperature loss in the array size), while cold raises their voltage (the cold-Voc check). The first makes you add panels; the second limits how many you can put in series.