Why this page exists
A calculator is only useful if you can see how it reached its answer. This page explains the data, formulas and default assumptions behind the Solar Metrix tools. It is a summary, not a full engineering specification. Each calculator also has a guide with a worked example and a longer limitations section.
Data sources
The sizing and yield tools use historical weather and temperature records for the coordinates you enter. The model uses five complete calendar years of daily radiation and temperature statistics, then runs a one-year hourly simulation and scales each month to the five-year monthly average. This captures both the shape of a typical day and the year-to-year variation that drives the P90 estimate.
Location lookup uses a geocoding service to turn a city or address into coordinates. If you enter coordinates manually, no geocoding request is needed. Where weather data is unavailable, the calculator may fall back to a latitude-based planning yield and will label the result as a fallback rather than a site-specific simulation.
Common assumptions
- System losses: 8% for cabling, inverter conversion, mismatch and availability, editable.
- Soiling: about 2% per year for a moderate environment, editable.
- Temperature coefficient: −0.35% per °C above 25 °C, editable.
- NOCT: 42–45 °C depending on the module class, editable.
- P90 uncertainty: combined inter-annual and 5% system uncertainty, multiplied by 1.282.
- Tilt rule: fixed tilt ≈ 0.87 × absolute latitude; seasonal ±10°.
- Inverter DC headroom: up to 1.35 × the AC rating for an existing inverter.
- Fallback yield: 4 kWh/kWp/day for generic sizing, 1,440 kWh/kWp/year for the payback model, 130 kWh/kWp/month for expansion.
These are planning values, not datasheet values. Replace them with your own equipment specifications, meter readings or written quotes wherever you have them.
Calculator by calculator
System size
The calculator sizes the array to cover the monthly kWh you enter. With a location, it uses the hourly weather simulation to find the specific yield, then divides annual consumption by that yield. Without a location, it uses a generic 4 sun-hours-per-day planning yield. The result is rounded to a practical panel count at your chosen wattage. Peak load is reported separately for inverter and wiring checks; it does not increase the array size. The recommended inverter is selected around a healthy DC/AC ratio, not around peak load alone.
Energy yield
Hourly plane-of-array irradiance is adjusted for tilt and azimuth, then converted to cell temperature using ambient temperature and NOCT. A temperature multiplier is applied, followed by fixed losses and soiling. P50 is the central annual total. P90 subtracts 1.282 combined standard deviations from P50 to produce a conservative planning figure.
Roof space
Required kWp is converted to a panel count and a panel area. The tool then packs real panel dimensions into your roof rectangle in portrait and landscape orientations, applies edge setbacks and access allowances, and removes panels that conflict with obstacles or winter shadows. It is a layout study, not a structural or waterproofing approval.
Tilt and site
Azimuth is measured from true north. The fixed tilt follows the 0.87 × latitude rule, with summer and winter alternatives 10° flatter and steeper. The tool also reports climate zone, temperature range, wind and insolation. It does not model nearby shading or roof-specific constraints.
String sizing
Every valid series-parallel layout is enumerated. Cold-weather open-circuit voltage is calculated with a 0.30%/°C temperature coefficient from the record low, and compared with the inverter’s maximum DC voltage. Operating voltage must exceed the MPPT start voltage. Current per MPPT is the number of parallel strings multiplied by Imp and must stay within the tracker limit. The DC/AC ratio is checked against an editable 1.15–1.35 band. Layouts outside the band are shown with warnings, not hidden.
Battery backup
Backup energy is sized from what must run and for how long, not from total daily consumption. Each appliance contributes kW × its own outage hours. The energy chain is: AC energy → DC energy after inverter efficiency → charged energy after battery round-trip efficiency → solar input after charging efficiency. Nominal capacity is DC energy divided by depth of discharge and a temperature factor, then multiplied by a reserve margin. The inverter must cover 1.1 × continuous load and the motor-start surge. Recharge is reported as kWh/day because peak sun hours vary too much to infer a panel count safely.
Battery ROI
The model runs solar with and without a battery, day by day, with state of charge carried between days. It values avoided imports at the tariff and subtracts the export credit given up, then applies degradation, maintenance, replacements and tariff escalation. Outputs include simple and discounted payback, 10-year net benefit, LCOS, NPV and IRR. Backup value is optional and reported separately from electricity savings.
Grid vs battery
A 365-day simulation splits solar into direct daytime use, battery charging and export. The battery discharges into night load after round-trip losses, subject to usable capacity and power limits. The comparison holds the array and load constant, so the difference is the battery’s effect on imports and exports. Outage value is not included in this tool.
Solar EMI
The loan payment uses the standard equal-payment amortisation formula. If you provide a total repayable amount or a monthly instalment instead of a rate, the tool solves for the implied monthly rate. It reports the nominal annual rate, the effective annual rate after monthly compounding, total interest and a 25-year cash-flow view. It is not a bank offer and does not include fees, insurance or early-settlement charges.
Expansion
The extra array is sized from the kWh you still import, divided by your own measured yield per kWp where available. The result is rounded to 0.5 kWp and then to whole panels. The inverter check compares the future total DC capacity with 1.35 × the existing inverter rating; if that is exceeded, the smallest standard inverter that keeps the DC/AC ratio in band is selected. Appliance lists and grid units are cross-checked rather than mixed.
EV cost
Daily EV energy is daily distance divided by range, multiplied by battery capacity, then divided by one minus charging loss. Cost per km uses a blended home/work/public rate normalised to 100%. Fuel cost per km is fuel price divided by km per litre. Payback is the price premium divided by annual energy and maintenance savings. The solar array is sized from daily EV energy divided by peak sun hours.
What the models do not do
Unless a tool explicitly says otherwise, the models do not include roof-specific shading, module mismatch beyond the loss allowance, inverter downtime, snow, demand charges, time-of-use windows, export caps, financing fees, insurance, resale value or changes in government policy. They are planning tools, not bankable studies or engineering approvals.
Corrections and review
We review formulas and assumptions when a calculator changes or a user reports a credible error. If a correction changes a published result, the page is updated and the date at the top of this page is revised. To report an issue, email privacy@solarmetrix.app with the page URL and the inputs you used.