Type something to search...
to navigateto selectESC to close

Asking Vega and investigating energy performance

Evaluate solar production and self-consumption

Assess solar output, compare performance over time, and understand self-consumption coverage.
Table of contents
  1. Prepare a trustworthy solar scope
  2. Review the built-in PV KPIs
  3. Ask Vega to explain the KPIs
  4. Compare actual and expected production over time
  5. Use the PV Calculator as a reference
  6. Calculate self-consumption only with complete site flows
  7. Investigate underperformance

Prepare a trustworthy solar scope

Before judging performance, verify the setup described in Connect a solar power plant:

  • the correct solar component and hierarchy,
  • installed peak power (P0), loss, tilt, and azimuth,
  • site coordinates,
  • actual production measurements,
  • historical expected-production and irradiance data, and
  • complete coverage for the period.

Do not total a parent component together with children that already contribute to it. Use the @ picker to select the intended plant, array, or inverter.

Review the built-in PV KPIs

  1. Open Inventory and select the site.
  2. Open Components and select a component marked as a PV installation.
  3. Select PV Performance.
  4. Under KPIs, choose a Granularity: 15 minutes, hourly, daily, weekly, monthly, or yearly.
  5. Select the corresponding Period, or use the previous and next controls.
  6. Review the values, range labels, sample counts, and any incomplete-data warnings.

The cards show:

  • Plant Energy Produced (EY) — measured energy produced in kWh.
  • Plant Energy Expected (EE) — the historical expected energy for the same period.
  • Energy Performance Index (EPI) — EY / EE × 100%.
  • Specific yield (YF) — produced energy relative to installed peak power, in kWh/kWp.
  • Reference yield (YR) — plane-of-array irradiation relative to standard irradiance.
  • Performance ratio (PR) — YF / YR × 100%.

Expected production is a historical, weather-normalized benchmark for completed periods. It should be compared with actual production for the same component and timestamps.

Do not confuse that benchmark with Production Forecast, which is Vega’s short-term operational forecast for today through D+2. Use Use and interpret PV forecasts when planning upcoming production or reviewing how a retained forecast compared with a completed day.

Ask Vega to explain the KPIs

Select Ask AI at the top of the KPI panel. Vega opens a contextual conversation beside the component and asks about the displayed component, granularity, and period.

Follow up with:

Explain EY, EE, EPI, YF, YR, and PR for this selected period. State the exact values, data coverage, missing inputs, and which indicators support an underperformance concern.

Closing the panel keeps the conversation in Chat history. You can also open it as a full-page chat.

Compare actual and expected production over time

For a trend rather than one KPI card:

  1. Start or continue a chat.
  2. Insert the exact solar component.
  3. Choose a complete historical period.
  4. Ask for actual production and expected production at the same granularity.
  5. Request a table first, then a chart if useful.

Example:

For @Main Site / Roof PV, compare actual and expected daily production during August 2026. Show kWh and EPI for each complete day, identify missing data, and rank the five lowest EPI days.

Do not interpret missing expected or actual values as zero. If Vega reports incomplete data, resolve readiness problems before ranking performance.

Use the PV Calculator as a reference

The PV Calculator on the same page shows average monthly and annual PVGIS reference production, irradiation, variability, and total loss. It also shows when the reference was fetched and whether a cached result was used.

This reference requires site coordinates, peak power, and system loss. Missing-input messages point to the site General tab or component Settings. Treat long-term PVGIS values as a reference, not as the measured output for a particular day or month.

Calculate self-consumption only with complete site flows

Self-consumption requires aligned measurements for PV production and grid export. Site-load coverage also requires grid import. When all three series describe the same electrical boundary and timestamps:

  • self-consumed solar = PV production − grid export,
  • self-consumption rate = self-consumed solar ÷ PV production, and
  • solar coverage = self-consumed solar ÷ total site load.

Ask Vega to confirm the topology before calculating:

For @Main Site, identify the components representing PV production, grid import, and grid export. Confirm that they cover the same boundary and complete period. Do not calculate self-consumption if any required series is missing or overlapping.

Then ask:

Calculate self-consumed solar, self-consumption rate, and solar coverage for August 2026. Show the component used for each term, the formula, units, and coverage.

Do not substitute grid import for total site load, and do not assume all PV production is consumed on site when export is unavailable.

Investigate underperformance

If EPI or PR is low:

  1. Confirm actual, expected, and irradiance coverage.
  2. Compare several equivalent complete periods.
  3. Check whether the issue affects the whole plant or one child component.
  4. Inspect Health and synchronization status.
  5. Verify P0, loss, tilt, azimuth, and site coordinates.
  6. Check operational evidence for outages, clipping, shading, soiling, snow, or maintenance.
  7. Ask Vega to label each explanation as observed, supported, or still hypothetical.

A low daily value alone does not identify the cause. Finish by verifying the evidence and assumptions before scheduling work or estimating losses.