support

This document is a catalog of all case families defined for solar sites in Performance+. Each family groups one or more related event types into a single trackable issue, with configured timing and impact thresholds that drive when cases are opened, escalated, and closed.

For a detailed explanation of how clusters, states, impact bands, and case actions work, see Case Workflow. For the full list of events that feed into these families, see Solar Event Detection.

The search window used when collecting events for solar cases is 30 days. Each event type is identified by a numeric event code; these are shown in parentheses next to the event name throughout this document.

---

Impact Threshold Conventions

Impact thresholds are expressed as fractions of capacity and scaled automatically to each site:

• Cluster-level thresholds (per inverter) scale with the median inverter capacity of the site.
• Aggregated thresholds (fleet-wide) scale with the total site capacity.

The table below uses the following abbreviations:

• Inv cap - median inverter DC capacity (or AC where noted)
• Site cap - total site DC capacity (or AC where noted)

---

Case Family Catalog

1. Inverter Down

Events covered: Inverter Down (51002)

The inverter has ceased AC production entirely while irradiance is available. This is the most severe inverter-level loss condition and uses the tightest thresholds.

---

2. Grid Curtailment

Events covered: Grid Curtailment (51003)

The TSO or DSO has instructed the site to reduce injected active power. The inverter is healthy but externally constrained. Cases are opened after 3 days to allow for short-lived grid balancing orders to resolve without generating noise.

---

3. Power Limit

Events covered: Inverter Power Limit (51006)

An active-power setpoint issued by a plant power controller or SCADA constrains inverter output below its natural operating level. The command originates at site level rather than from the grid operator.

---

4. Inverter Electrical Losses

Events covered: Reactive Power Loss (51007), Max kVA Limit (51008)

These two events are grouped into a single family because they share the same physical origin: the inverter's apparent-power capacity is being consumed by reactive power obligations, leaving less headroom for active power. Thresholds are tighter than for down or curtailment events because these losses are typically a small fraction of capacity. AC capacity is used as the reference rather than DC.

---

5. Inverter Thermal Derating

Events covered: Inverter Thermal Derating (51009)

The inverter's internal protection reduces output power due to overheating of the power electronics. Thermal events occur during hot daytime hours and can be recurrent at high-temperature sites.

---

6. Tracker Issue

Events covered: Tracker Wind Stow (51013), Tracker Non-Wind Deviation (51014)

Both tracker event types are grouped into a single family. Wind stow is expected behaviour during high-wind periods, but persistent or unusually frequent stowing may indicate a misconfigured wind threshold. Non-wind deviation is always maintenance-actionable. Combining the two allows a single case to reflect any tracker-related yield loss, regardless of cause.

---

7. Snow

Events covered: Snow (51015)

Snow accumulation on module surfaces blocks irradiance and reduces output to near-zero. At sites prone to snowfall, persistent snow cover warrants an operational response (e.g. manual cleaning or monitoring of melt progress).

---

8. Soiling

Events covered: Soiling Level High (51019), Soiling Rate High (51020)

Both soiling event types are grouped together. The level-based event triggers when the cumulative soiling loss exceeds 15%; the rate-based event provides an earlier warning when the accumulation rate is high but the level has not yet reached that threshold. Combining them into one family ensures that a single case tracks the soiling condition from early-warning through to high-impact, without splitting the issue across two records.

Because soiling is a slow-accumulating, continuous degradation rather than an instantaneous fault, thresholds are set at lower absolute fractions of capacity compared to outage-type families.

---

Notes

• Clipping events (Site Clipping, Inverter Clipping) are intentionally excluded from the case workflow. Clipping reflects the site operating at its design ceiling and is not an actionable O&M issue.
• Site communication and weather station sensor health events are not yet covered by the case workflow and will be added in a future iteration.