Several Modules Flagged: Which Investigate First Triage Decision Tree

Why this matters

A system with module level power electronics that reports multiple flagged modules on the same morning is a triage problem. Walking the roof and investigating each flag one at a time burns a full day on a system that may have one root cause behind ten symptoms. The right approach is to triage the flags into categories first, identify whether the flags share a root cause, and start the field investigation at the highest-yield diagnostic point. A disciplined triage decision tree turns a chaotic alert dashboard into a one-trip diagnosis.

Symptom presentation

The monitoring portal shows multiple modules in a flagged state on the same system. The flags may include underperformance versus peers, communication loss, zero production, or specific error codes from the module level electronics. The customer may have noticed reduced total production on the bill or may not have noticed anything if the system has many modules and the flagged share is small. The inverter may or may not be reporting a system-level fault.

The natural first instinct is to start with the most obvious flag, often the module showing zero production. That is sometimes correct and sometimes a mistake.

Quick checks

Pull the flag list with timestamps. Modules that started flagging at the same instant share a likely common cause. Modules that started flagging at different times across different days are independent failures. The timestamp pattern is the most important triage signal on the dashboard.

Map the flagged modules to physical location on the roof. Modules that are physically adjacent or on the same string or same DC home-run share more candidate root causes than modules scattered across the array. A cluster of flags in one rooftop area points to a localized cause. Scattered flags point to a fleet-level cause or independent failures.

Note the flag types. Communication loss flags often share a single gateway or RS485 bus root cause. Underperformance flags often share a string-level connection or shading cause. Zero production flags often share a DC home run or string conductor root cause.

Check the inverter and gateway. A gateway with a stale data timestamp may be reporting cached flags from hours ago rather than current condition. A gateway that lost network connectivity is reporting nothing new and the dashboard is misleading.

Decision thresholds

Use four gates.

Gate one is timestamp clustering. If multiple flags share the same start timestamp within a few minutes, treat them as one fault until proven otherwise. The first investigation is at the shared infrastructure: the string they share, the DC home run, the inverter MPPT input, the combiner if present, or the gateway communication path.

Gate two is physical clustering. If the flagged modules are physically clustered, walk that cluster first. Look for shading from new growth, debris that landed on the area, animal damage, or weather damage from a recent storm. Localized physical causes resolve quickly once on the roof.

Gate three is flag type clustering. If the flag types are all communication loss, the gateway and the powerline communication or RS485 bus are the first investigation. If the flag types are all underperformance, the string-level analysis at the inverter is the first stop. If the flag types are mixed, treat as independent failures and prioritize by impact.

Gate four is impact. Among confirmed independent failures, prioritize by production loss. A module reporting zero is roughly one-module-worth of loss. A module reporting 50 percent of peer is half that. A module reporting communication loss with normal production is no immediate production loss and is the lowest priority for the day.

Confirming diagnosis

For timestamp-clustered flags, investigate the shared infrastructure first. Pull the string IV curve at the inverter. Read DC voltage and current on the affected string. Inspect the DC home run from the inverter MPPT to the first combiner or junction. Inspect the combiner fuses if present. A failed fuse, a damaged conductor, or a corroded MC4 at the string entry to the combiner produces an entire string going dark in a single timestamp.

For physically clustered flags, walk the cluster with binoculars from the ground first to identify obvious causes without a roof access. Then walk the cluster on the roof to inspect for shading, debris, animal damage, or storm damage.

For communication-only flags with normal production, the modules are producing. The dashboard is missing data. Investigate the gateway and the communication bus. A gateway with a tripped circuit, a damaged ethernet cable, or a powerline communication coupling issue produces a cluster of communication losses without any production fault.

For mixed flags treated as independent failures, work the highest-impact flag first.

Remediation

For string-level shared root cause, repair the shared infrastructure. Replace blown fuses, repair damaged conductors, replace corroded MC4 connectors, verify torque at combiner terminations.

For physical localized cause, address the local cause. Prune vegetation, remove debris, repair animal damage, replace storm-damaged modules.

For communication-only flags, restore the gateway or the communication bus. Replace damaged ethernet, restore power to the gateway, verify powerline communication coupling, restart the gateway after the root cause is addressed.

For independent module failures, address each on its own merits. Module level electronics replacement, module replacement, or warranty action depending on the root cause of each.

Document the triage path on the work order so the next tech can replicate the decision logic. A note that reads "ten flags, all timestamp-clustered to one moment, all on string 3, root cause was a blown combiner fuse, single repair cleared all flags" is more valuable than a list of ten line items that obscure the relationship.

References

• NEC Article 690: Solar Photovoltaic (PV) Systems.
• NEC Article 705: Interconnected Electric Power Production Sources.
• IEC 62446-1: Grid Connected PV Systems - Commissioning Tests, Inspection and Documentation.
• IEC 61724: Photovoltaic System Performance Monitoring (module-level performance data interpretation).
• UL 1741: Inverters, Converters, Controllers and Interconnection System Equipment.