1. Survey obstacles and how they change
Survey chimneys, trees, neighbouring buildings and the horizon. Request hourly and seasonal shading with assumptions about vegetation and known future work. Sunshine during one visit does not establish the same winter exposure. Compare a layout that leaves persistently shaded areas unused.
2. Separate light losses from electrical losses
Have cell layout, bypass diodes, strings and MPPT inputs described. Diodes can bypass parts of a module under certain conditions; they cannot restore missing light. PVsyst shows that interconnections and irradiance conditions affect losses. “Half of one panel means half of all generation” is not a project calculation.
3. Compare the actual proposed configurations
Compare a string inverter with suitable MPPTs, DC optimisers and microinverters on the same layout. Check compatibility, voltage, current and module count. Module-level electronics can change mismatch management and monitoring but do not remove every loss. Request estimated recovered kWh, assumptions and service requirements for each option.
4. Compare costs and plan monitoring
Compare installed extra cost, additional useful energy, warranties, access and replacement. Check which data the owner can access and who handles alerts. Keep the connection plan and generation forecast at handover. Module-level app data can aid diagnosis but cannot alone establish that extra hardware pays for itself.