Is Your Solar Plant Underperforming? Start With the Evidence

A lower electricity total can raise an urgent question: is the solar plant developing a technical problem? Sometimes it is. Sometimes the difference comes from weather, a grid restriction, a changed reporting boundary or missing data. Replacing equipment before separating these possibilities can consume time without resolving the real issue.

A useful investigation starts with a clear comparison and follows the evidence. Identify what is known, quantify defensible estimates and leave unsupported conclusions open.

1. Define the shortfall and the baseline

Write down the period, measured energy, expected energy and measurement point. Clarify whether the reported number is inverter output, plant generation or net export after internal consumption and other losses. Comparisons must use compatible boundaries.

Next, identify the baseline. A budget based on typical weather is useful for planning, but it does not represent the exact weather in a particular month. NREL’s PVWatts guidance distinguishes long-term modeled expectations from production in an individual period. [1]

Ask whether the model reflects the installed capacity, orientation, equipment and operating restrictions. Retain the original baseline and document any corrections separately. Repeated adjustments can hide the original difference.

2. Check the data before diagnosing the equipment

Build a working dataset using available meter readings, inverter records, irradiance measurements, temperature measurements, alarms and operator logs. Check timestamps, time zones, units, missing intervals, duplicated entries and abrupt changes following sensor or communications work.

A zero in a monitoring portal may represent lost communications rather than zero generation. Compare independent records where possible. If the export meter and inverter totals disagree, first investigate their boundaries and recording intervals instead of choosing whichever number appears more favorable.

IEA-PVPS’s analytical monitoring guidance explains the importance of measurement arrangements and methods for interpreting plant behavior. [2] In practice, the investigation report should show which data passed quality checks, which were excluded and how the exclusions affect confidence.

3. Separate weather, availability and operating restrictions

Compare production with the sunlight received during the same period. Where suitable data exist, evaluate temperature effects using an appropriate plant model rather than treating every hot-day reduction as a fault. NREL explains that an uncorrected performance ratio is influenced by weather as well as plant behavior. [3]

Then build an event timeline. Identify equipment outages, planned maintenance, grid interruptions, export limits and curtailment instructions. Record the source of each event classification.

Availability and energy loss answer different questions. A short interruption during strong sunlight may matter more to production than a longer interruption at low irradiance. State the availability definition and estimate lost energy separately, with the assumptions visible. Avoid counting the same lost interval under both an equipment outage and a grid restriction.

4. Test likely causes rather than choosing a convenient explanation

After separating known events, investigate patterns. Is the shortfall plant-wide or concentrated in one inverter group? Did it begin suddenly or develop gradually? Does it coincide with cleaning, vegetation growth, changed settings or equipment replacement?

Soiling is worth examining, particularly where dust accumulation is significant, but a dirty appearance does not quantify the loss. IEA-PVPS notes that soiling can vary across modules and across a plant. Representative measurements matter when evaluating cleaning decisions.

[4]

For a proposed cleaning trial, agree the comparison method in advance and account for changing weather. For suspected temperature-related behavior, compare the records with the equipment specifications and operating conditions. For suspected degradation, examine a sufficiently long, quality-controlled history; a weak month alone does not establish a long-term degradation rate.

Inspection or electrical testing should be planned and performed by qualified personnel under site safety procedures. Remote analysis is not a reason to bypass safe access or electrical controls.

5. Turn findings into a prioritized action plan

For each issue, state the evidence, confidence level, likely consequence, next action and responsible party. Separate confirmed faults from hypotheses requiring inspection. If energy loss cannot be isolated reliably, provide a qualified estimate or explain why no defensible estimate is available.

Prioritize safety-related findings first, then consider expected recoverable energy, cost, downtime, warranty requirements and available resources. Set a verification method before corrective work starts. After the action, compare suitable operating periods and confirm whether the relevant pattern changed.

An effective report can conclude that more evidence is needed. That is more useful than presenting a precise loss figure unsupported by the measurements.

6. Give the owner a clear next decision

SPP’s proposed support can be scoped around data-quality review, baseline checks, event reconciliation and a prioritized investigation report. Any site testing or specialist analysis should be separately defined, with responsibilities and limitations clear.

If your plant is missing expectations, contact SPP to discuss the available data and a focused review. The first objective is a defensible explanation and a practical next step, not a promised recovery figure.