The Solar performance page on the Insights tab summarizes the performance for a solar system. This helps you answer: over the last month, quarter, or year, how much energy did this system lose, and to what cause.
Each solar meter has its own Solar performance page. The date range you select applies to the Expected vs. Actual chart and the Solar metrics table. The Long-term trends charts at the bottom always show the meter’s full recorded history.
Note on solar meters. The performance metrics are based on what the solar meter measures. A solar meter could be set up to measure all or just part of the solar system, like an inverter or string.
Expected vs. Actual
The waterfall chart accounts for every kilowatt hour between expectation and reality. Read it left to right. Bars that drop reduce production, and a bar that rises means the system produced more than the model expected.
- Expected: the energy output the weather should produce, based on available sunlight, outdoor temperature, panel orientation, and the system’s capacity and age.
- Soiling: energy lost to dirt on the panels, relative to a clean array. It is zero or negative; more negative means dirtier. It builds up between rains and clears after them.
- Outage: energy lost while the system was offline or otherwise down during a detected outage event. Check Outage events and Total downtime below to see whether that was a few long failures or many short ones.
- Other: the balance left after expected production, soiling, and outages are accounted for. A large negative bar usually means a partial fault, such as an underperforming array or a single failed inverter, that never dropped production far enough to register as an outage. This is the case to take seriously. A positive bar over a short range is normally local irradiance variation between the nearby weather station and your site; over a long range it usually means the model is conservative because it does not have your system’s exact specifications.
- Actual: actual energy output from the solar system that has been measured.
Helpful Hint: Run the chart at 1y first to see which category dominates over a full year of weather, then narrow to 90d or 30d to find out when the losses happened.
Solar metrics
This table reports the full solar metrics over the same period as the waterfall chart.
- Effective capacity (whole record): this is computed as the DC capacity (kWdc) multiplied by a derate factor. If this information has not been provided, then the effective capacity is estimated based on historic performance. This is a constant, not tied to the selected range.
- Expected production: the energy output the weather should produce, based on available sunlight, outdoor temperature, panel orientation, and the system’s capacity and age.
- Actual production: actual energy output from the solar system that has been measured.
- Performance index: your actual production divided by what the weather based model expected. Near 1.0 means the system is meeting expectations at its typical cleanliness, not pristine panels. Below 1.0 means it produced less than expected, and the Soiling, Outage, and Other categories break down why. Above 1.0 is valid: briefly, just after a cleaning or from short term local irradiance differences. If it stays above 1.0 over a long period, the model is likely conservative without your system’s exact specifications.
- Soiling loss: the percentage of energy lost to dirt on the panels.
- Degradation (whole record): the percentage output decline per year. This is a single lifetime estimate, not tied to the selected range.
- Outage loss: the percentage of energy lost while the system was offline or otherwise down during a detected outage event.
- Outage events: number of discrete outage periods in the range.
- Total downtime: total downtime from outages in the range.
A metric reports a dash when we do not yet have enough history to compute it.
Note on effective capacity. When Gridium has your system’s provisioned DC capacity, effective capacity is the total capacity reduced by a derate factor, which accounts for the real world losses between what the panels are rated for and what the system delivers. If this information has not been provided, we estimate capacity from the system’s observed performance on clear days. This means that a system with a long standing fault could appear healthy. Providing your system’s capacity, orientation, and tilt recalibrates the model to be more accurate.
Long-term trends
These charts cover the full history of data and ignore the date range above. They answer a question a single period cannot: is this system getting worse?
The Performance index chart plots the index over history. A well performing system is close to the reference line at 1.0. A step down that never recovers marks a fault that began on a particular date, and a slow drift downward over years points to degradation, new shading, or a gradually failing string.
The Soiling ratio chart plots the degree to which dirt is accumulating and leading to soiling loss. A value of 1.0 indicates clean panels. The sawtooth shape is normal, since soiling accumulates through dry weather and clears when it rains or is cleaned.
