48-hour outlook
A little certainty for the hours after sunset.
Try a different plan
Pin this setup, then change appliances, battery or charging to compare.
What changed
Both plans & assumptions
Where the energy goes
Modeled use over the planBattery reserve
Energy by 24-hour period
kWh · solar before input clippingA year of sunlight
Same setup. A different place in the sun.
Coordinates stay in your browser. Fixed UTC offset for the entire plan and year; adjust it yourself for daylight saving time.
Daily solar harvest
After your sunlight & harvest assumptions · before clippingThe colored strip reruns your plan from its starting battery charge on every date. It is not a continuous year of battery use.
What the sunlight model does and doesn't know
Sun position follows NOAA’s daily geometric approximation. Clear-sky horizontal irradiance uses Haurwitz, then the Erbs direct/diffuse split and an isotropic tilted-panel model with ground reflectance 0.20. Your harvest-efficiency and sunlight factors apply afterward. Solar energy uses irradiance (W/m²) and insolation (kWh/m²), not lux, which measures brightness for human vision.
This is a smooth seasonal comparison, not a weather forecast or historical climate data. It has no local cloud, altitude, terrain, building, shading, snow or panel-temperature data. Sun-up times use the center of the geometric sun, without atmospheric refraction. Keep extra margin for real conditions.
Method references: NOAA solar geometry ↗ · Haurwitz ↗ · Erbs ↗ · Tilted diffuse ↗
What needs power?
Choose each power path and enter watts delivered to the appliance. Use average watts for cycling loads.
A planned top-up
Optional car, shore or other external charging.
How this estimate works
This is a planning estimate, not a forecast or an electrical safety check. Enter your own measured or manufacturer-rated specs. Example appliances are illustrative.
- 15-minute steps. The first day starts at the chosen clock time. Day 1, 2, 3 and 4 mean calendar days; an earlier Day 1 event is already past. Daily events can continue from the previous night.
- Manual mode follows a smooth 06:00–18:00 daylight curve, repeated every day. Location mode uses the selected coordinates, date, fixed UTC offset and panel orientation; each new calendar day gets its own solar curve. In manual mode, panel watts × peak sun hours × harvest efficiency × sunlight assumption gives daily solar energy, before charging-limit or full-battery clipping. Peak sun hours are equivalent full-power hours, not daylight hours.
- Appliance watts are delivered output watts. AC loads share the AC efficiency and inverter limit; DC / USB loads share a separate efficiency. Inverter overhead is battery-side watts, counted once either throughout the plan or while at least one positive-watt AC load is scheduled. Scheduled-only is a manual on/off assumption, not a hardware ECO mode. It does not model cycle-by-cycle switching for average loads. Reserve is advisory, not an automatic shutoff.
- External charging is modeled at 90% efficiency. Solar and external charging share the total input limit. Unused input is clipped when the battery is full.
- An overload means requested AC watts exceed the continuous AC inverter rating. The estimate caps AC delivery at that rating, but real hardware may trip AC output. DC / USB is excluded from this AC check. Starting surges, combined battery output, DC / USB port limits, voltage and protocol compatibility, temperature, degradation, charging taper and device-specific behavior are not modeled. Selecting a path does not confirm equipment compatibility.
- When energy is scarce, accepted AC loads, DC / USB loads and overhead share the same powered fraction of each 15-minute step. This is a time-sharing estimate, not a promise about device priority or restarts. Delivered energy, conversion losses and consumed overhead use that same fraction so energy is conserved. Requested overhead can exceed consumed overhead while the battery is empty.
- Efficiencies and fixed overhead are independent manual assumptions. Published maximum efficiency may already include some idle consumption; do not blindly combine nameplate values. Leave overhead at zero if it is already included in your measured efficiency. Example specification separating maximum efficiency and zero-load power ↗; no device settings are imported from it.
- Battery runout estimates interpolate within a 15-minute step. Real conditions can be worse. Use extra margin, check equipment compatibility, and never rely on this estimate for medical or other critical power.