There is no single daily generation number
A solar system does not produce the same number of units every day. Output changes with solar radiation, cloud cover, season, module temperature, shading, grid availability and the design of the plant itself.
That is why a statement such as “1 kW should always generate X units per day” is too simplistic for diagnosis. A useful benchmark must be tied to location, time period and the actual system.
Start with annual and monthly expectations
The best reference is usually the yield estimate prepared for the plant at the time of design, provided it was based on reasonable assumptions. It should account for local solar resource, module orientation, tilt, inverter efficiency, expected system losses and any known shading.
If that report is not available, use your own historical generation. Compare the same month across years or compare several clear days in similar seasons.
Daily comparisons are useful for troubleshooting, but annual and monthly trends are more reliable for judging whether a system is broadly on track.
Factors that change generation
Solar radiation
The amount of sunlight reaching the modules is the most important input. Haze, clouds, monsoon weather and seasonal sun angle can change energy production significantly.
Module temperature
Solar modules usually produce less power as cell temperature rises above the test conditions used for their nameplate rating. A very hot clear day may therefore not deliver the highest energy yield of the year.
Orientation and tilt
A roof that does not allow an ideal orientation can still produce useful energy, but its daily and seasonal profile may differ from a south-facing reference installation. East-west arrays often produce a broader generation curve with lower midday peak power.
Shading
Trees, parapet walls, water tanks, neighbouring structures and new rooftop objects can create partial shading. The impact depends on timing, array layout and inverter MPPT configuration.
Soiling
Dust, bird droppings and other deposits can reduce available light. The effect is site-specific, so cleaning decisions should be based on actual condition and performance rather than a universal loss percentage.
Inverter and grid availability
A grid-connected plant can lose generation if the inverter is offline, derating or repeatedly disconnecting because of grid conditions. This can be missed if the owner looks only at module cleanliness.
Use specific yield for fair comparisons
A useful metric is specific yield, calculated as:
Energy generated (kWh) ÷ installed DC capacity (kWp)
For example, if a 5 kW plant produces 20 kWh in one day, its specific yield for that day is 4 kWh/kWp.
Specific yield lets you compare plants of different sizes, but it still must be interpreted with local weather and design conditions in mind.
Compare the shape of the generation curve
If your monitoring portal shows power through the day, the curve can reveal more than the daily total.
A smooth bell-like profile on a clear day suggests continuous operation. Sudden repeated drops can indicate inverter or grid events. A persistent morning or afternoon dip may point to shading. A completely missing section can indicate downtime or a communication gap.
Do not diagnose from the graph alone, but use it to decide what to investigate next.
What is a meaningful warning sign?
Look for a sustained change under broadly comparable conditions rather than one unusual day.
Examples that deserve investigation include:
- a noticeable month-on-month shortfall with no obvious weather explanation;
- repeated inverter shutdown during strong sunlight;
- one MPPT or string behaving differently from the others;
- generation not recovering after a known maintenance issue is corrected;
- a new shading pattern or equipment alarm.
Build your own plant benchmark
Keep a simple record of monthly generation. Add notes for extended outages, major cleaning, inverter faults, grid issues or unusual weather. Over time, this becomes a much more useful reference than a generic “units per kW” rule.
If the plant has never had reliable monitoring or baseline data, a solar health assessment can help establish a starting point.
If generation is lower than expected
Do not jump directly to panel replacement or inverter replacement. Work through weather, season, monitoring accuracy, shading, soiling, inverter events, grid behaviour and then electrical testing.
Our guide on why a solar system may generate less electricity provides that diagnostic sequence.