What a Healthy Solar Day Looks Like on the Curve

A healthy solar production curve usually rises through the morning, peaks near solar noon, and falls in the afternoon. Flat tops, sudden notches, or midday zeros are clues for clipping, shade, or faults.

A solar production curve plots instantaneous power in kilowatts against time of day. On a clear day, a healthy grid-tied array produces a smooth bell-shaped trace: low output at sunrise, a steady climb through the morning, a broad peak around solar noon, and a symmetric decline through the afternoon. The daily energy total in kilowatt-hours is simply the area under that curve. Learning to read the curve shape helps you spot inverter faults, string failures, and new shade long before your utility bill shows a gap.

The curve reflects how irradiance, temperature, and system hardware interact. Modules convert photons to DC current; the inverter or microinverters convert DC to AC and report power to the monitoring portal. MPPT trackers on each string or module adjust operating voltage to maximize harvest as light changes. Clouds cause rapid dips and recoveries. Heat reduces module voltage and can flatten the peak slightly on very hot afternoons. Fixed-tilt arrays in the northern hemisphere often show a slightly longer afternoon tail in summer when the sun sits higher and days are longer. None of that is a fault by itself. The diagnostic value is in comparing curve shape across similar weather days, not chasing the highest peak on a single record-breaking afternoon.

Start with a reference library of three to five clear-sky days from each season. Screenshot or export them after commissioning while the array is known healthy. Note peak power relative to nameplate DC capacity: residential systems often peak at 70 to 85 percent of DC rating because real-world conditions rarely match laboratory STC. A 10 kW DC array might top out near 7.5 to 8.5 kW AC on a mild spring day. If peak power on a sunny June afternoon suddenly drops 30 percent versus last June with no equipment change, investigate before assuming weather alone explains it.

Flat lines during hours when irradiance should be strong are the most urgent red flag. A zero plateau from mid-morning through mid-afternoon while neighbors produce normally usually means the inverter tripped, a DC disconnect opened, a blown fuse in a combiner, or an entire string lost continuity. A partial flat where power climbs then holds at a low ceiling for hours can indicate one string offline on a multi-string inverter while the other still works. Microinverter sites show a stepped-down bell: the envelope looks normal but the peak is lower because many units are offline even if the portal still shows some production.

Asymmetric curves point to shade or orientation effects. A truncated morning with normal afternoon often means east-side obstruction, tall trees on the horizon, or a string on the east roof section underperforming. The opposite pattern suggests west-side shade from a chimney, parapet, or afternoon tree line. Narrow notches at the same clock time every day are classic fixed-shade signatures: vent pipes, skylights, or HVAC equipment casting a repeating shadow across part of the array. Wide shallow bowls instead of sharp peaks usually mean thin high cloud or haze rather than equipment failure.

Step changes and jagged traces deserve context. Inverters with aggressive power limiting or export caps show a flat top at the limit value through midday even when more DC is available. That clipping looks like a plateau, not a fault. Rapid sawtooth patterns on old string inverters can reflect MPPT hunting in marginal light. Optimizer-level monitoring may show individual module steps when only some units are shaded. Compare against weather radar: widespread jagged dips across the whole curve match passing clouds; isolated steps on one string index point to hardware.

Use overlay tools when your platform offers them. Many apps let you stack today's curve on yesterday's or on the same weekday last month. Alignment of rise and fall times validates that inverter clock and timezone are correct. A curve shifted one hour early or late is often a settings issue, not a production issue. Overlaying modeled clear-sky expected power helps separate weather from faults: if actual tracks well below model all day on a cloudless day, equipment deserves a look.

Seasonal literacy prevents false alarms. Winter curves are lower, narrower, and sometimes slightly lopsided because the sun stays low and days are short. Spring and fall shoulders may show double peaks on east-west split layouts as morning and afternoon strings each hit their local optimum. Summer curves are tallest and widest but may peak slightly earlier on very hot days when temperature derating pulls voltage down before noon. Dust after a dry spell gently lowers the entire curve without changing its shape until rain restores output.

Pair curve review with a short checklist when something looks wrong. Confirm the gateway is online and timestamps are current. Check the inverter display or app for active fault codes. Walk the roof visually if safe access exists: new debris, nesting birds, or a fallen branch explain sudden shape changes. Review recent electrical work on the property that might have opened a disconnect. Document the abnormal curve with a screenshot and the fault code before resetting anything; warranty and service visits go faster with that record.

Establish a light weekly habit. On one clear day per week, open the portal and compare the last 24-hour curve to your seasonal reference. The habit takes under two minutes and catches most meaningful problems within days. Monthly, note whether peak power and curve width drift gradually downward across multiple clear days; slow soiling or connector heating sometimes appears as a gentle slope before a hard failure.

Production curves are the fastest visual diagnostic in residential monitoring. Totals and monthly bar charts smooth away the detail that reveals string loss, shade encroachment, and inverter behavior. Invest a little time learning your system's normal shapes across seasons, and abnormal days will stand out immediately without needing an engineering degree or expensive analytics.

Many portals also offer interval data at five or fifteen minute resolution. Zooming into the ramp at sunrise reveals whether all MPPT channels wake together or one lags by thirty minutes, a clue to string imbalance or a sticky relay. Exporting a week of curve CSV files into a spreadsheet lets you compute simple metrics like hours above 50 percent of peak, a compact health score you can track month to month without proprietary analytics subscriptions.