Choose the right CSV, report or heatmap export and preserve its context.
Use an export that matches your question. A formatted report, a comparison table and a measurement time series contain different information.
| Starting point | Suitable result |
|---|---|
| Sites → Export | Measurements for a selected portfolio scope |
| Period performance or Analysis | The displayed analysis or measurement comparison |
| Saved dashboard | Its selected site results and period |
| Inverter heatmap | The current inverter page's CSV where indicated |
| Reports | Formatted PDF and the report type's available tabular exports |
A download of the visible page is not necessarily an export of every device in the account. Read the button label and any selected-site count. Partial source availability can also make a result incomplete.
Check the selected period, time zone and number of sites before using Download CSV. Fictional demonstration data.
Open the saved combined monthly report and wait for its preview to finish loading. Check Selected sites: 360, Combined overview, the report dates and the coverage notes before choosing CSV or Excel XML. These report exports use the current rendered report; they do not launch a new account-wide measurement query.
In the fictional example, all 360 installations remain in the combined report's site list and per-site results, while production and grid-export values are available for 354. The six missing values must remain blank with their data-quality context. Exporting the single-site report instead would give that site's result, not the complete 360-site portfolio.
Export from the report whose scope matches the analysis. A preview for one installation and a combined portfolio preview are not interchangeable.
The report's tabular export carries the report name, dates, sites, units, status and notes alongside its tables. When present, comparison tables also carry the previous value, difference, percentage and comparison dates. Preserve those fields when passing the data to another team.
A report can contain the same quantity in several forms: a portfolio total, its distribution and the per-site ranking. For an independent total, use the relevant per-site values once. Do not add the summary total to the distribution and ranking rows; that would count the same production repeatedly.
Excel XML downloads an XML spreadsheet, not an XLSX workbook. Use the format indicated by the control and keep the original exported file alongside any spreadsheet analysis.
Use the spreadsheet application's text/CSV import so you can select its delimiter, encoding and date interpretation. Inspect decimal separators and units before applying formulas. A timestamp converted into another time zone can shift a reading into another calendar day.
Do not sum power readings in kW and call the result kWh. Energy requires the applicable interval duration and aggregation method. Likewise, do not sum a cumulative counter column as though each row were interval energy.
Keep missing readings distinct from measured zeroes. Replacing every empty cell with zero can understate production and distort availability. When publishing a comparison, carry over any partial-data notes rather than presenting the export as a complete measured period.
For the example report, check that the production ranking retains 360 site entries and distinguishes 354 available values from six missing ones. A complete list of installations does not guarantee complete measurement coverage.
Compare periods using the same scope. If the available sites differ between periods, a change in the total can reflect changed coverage rather than changed production. The combined report only calculates a total difference when the two periods cover the same sites. Carry its coverage and comparison notes into any further analysis.
The current online count is a separate observation. The live example's 348 communicating installations should not be used as the denominator for the historical report's 354-of-360 coverage.