Room for one more roof?
Screening traffic light for rooftop solar, per transformer.
DateBACKTEST
Services
Billed net-meter services in 2026-10; +68.2% on a year earlier (from data)59,550REALBilled load
billed kW of the kW categories; the meaning of the units column is open (from data)422.9MWREALForecast 9 Nov
Error 7.6 W/kWp; BACKTEST; versus a physics model, not versus measurement. The whole-area forecast uses one representative point, hyderabad: its circles hold 68.6% of the billed kW (2026-10). It is a stated choice, not an area-weighted mean; the other 12 points are on the forecast page. (from data)4.99kWh/kWpError 7.6 W/kWpMODELNet export 9 Nov
Estimate; installed = billed kW proxy (unverified); load scenario assumed (from data)1,747MWhEstimateEXAMPLEStreet P5
Town street, 100 kVA, 400 m, ACSR Rabbit, Indian model; cap 50.0 kW, unit assumed; below the cap; Passes screening tests (from data)8.9kWpTown streetEXAMPLEForecast
MODELEXAMPLEOpenW/kWp, 9 Nov
MW, 9 Novsolarnet export
Areas
REALGrowth
REALMODELOpenMonths to full
Fastest circle 239.8%REALReverse flow 23.4%EXAMPLE
Evidence
OpenREALNet-meterWeatherDay-beforeMNREPM SuryaPVGISLive feedRules
Not heldTransformersRooftop outputGround sunlightHourly net-meterTGNPDCL
- Network values are assumed examples until the discom supplies its own data. Screening only, never approved capacity.
- Forecast: accuracy not yet measured against real output (no metered rooftop output is held).
- Weather actuals are satellite/reanalysis-based sunlight, not metered solar output.
- 50% cap: unit assumed: kW on the DTR kVA rating; TGSPDCL to confirm.