A large portion of photovoltaic solar power plants (PV plants) are generating significantly less than expected. Incorrect PVsyst assumptions, design errors, installation errors, commissioning errors, and defective equipment are resulting in lower-than-expected financial outcomes.
In some (not uncommon) cases, more than half of the projected return is wasted, thrown in the trash. Hidden errors, not always easy to identify, drain MWh and revenue down the drain.
Inverter shutdowns due to poor cable insulation, misaligned trackers, photovoltaic panel degradation exceeding expectations, faulty bypass diodes, poorly parameterized cabin protection, inverters downgrading power due to temperature, and even worse, the monitoring system not working.
Unfortunately, the reality described is present in several generation assets spread across Brazil, whether they are large centralized generation plants with hundreds of MW of power, distributed generation plants of up to 5 MW, and photovoltaic generators of 300 kW or more installed on the roofs of industries and large commercial buildings (C&I).
When we talk about O&M in photovoltaic plants, the "basics done right" are often overlooked. Cheap equipment that doesn't measure correctly (often out of calibration), lack of methodology in analyses, and especially a lack of critical thinking from the technical team, distract from what really matters and what needs to be resolved now, not tomorrow or in a month.
This article summarizes concrete benefits delivered by the technical team of CS Consulting and Services, at a 5 MW power plant in the CPFL Paulista concession area in São Paulo, where our work increased the plant's availability and resulted in significant generation gains during the analyzed period.
Practical result: availability and recovered energy.
Availability: from 71% to 96%
The plant's availability was increased from 70% to 97%, reducing failures that directly impacted generation.
- Generation: 22,3% more and +735 MWh in 9 months;
- Financial return (8 months): R$ 440;
- Projected annual financial return: R$ 586,7.
How were these gains generated?
Initial diagnosis and addressing bottleneck #1: monitoring
Work began in the first quarter of 2025. The first month was dedicated to diagnosing the asset, identifying problems and classifying them by criticality. After immediate actions—such as replacing communication equipment, performing corrective maintenance on the system, and making other necessary adjustments to restore monitoring—interventions ceased to be reactive and random, becoming more agile, targeted, and prioritized.
The unavailability of the monitoring system prevented the identification and quantification of the losses that were in fact the most significant, since several generation failures remained hidden without the necessary correction.
Among the problems that significantly impacted energy production, the following stand out: insulation faults in DC cables, blown AC fuses, circuit breakers with intermittent faults, and defects in the bushing of one of the transformers.
Recalibrating goals: overestimated assumptions become "unattainable goals"
Another critical gain was taking the plant out of "promise" mode and putting it into "reality" mode. In addition to the actual physical problems of the photovoltaic asset, the PVsyst generation assumptions were overestimated. Overestimated assumptions were identified and subsequently corrected, including:
- Losses in DC and AC cables that are "too good to be true";
- Lower transformer losses are inconsistent with the datasheet;
- Consumption of auxiliary equipment not considered;
- Incorrect module degradation;
- Lack of 3D modeling and poorly designed shading;
- Incorrect maximum and minimum tilt angles of the trackers.
Result: generation assumption 5,69% above actual figure.
This is more relevant than it seems: in economic terms, this deviation from the premise represents approximately R$ 350 per year. The unrealistic target generated friction between the operational team and the board of directors of the company that owns the plant. In the end, the point is that the difference between an "operating plant" and a "performing plant" lies in performance governance.
Consistent simulation and design loss assumptions, traceable measurements (calibrated monitoring and SCADA), correct indicators (PR, availability, failure rate per subsystem, etc.), and a disciplined process with agility → hypothesis generation → critical thinking → diagnosis → intervention → documentation, make all the difference. When this loop is rigorously executed, losses cease to be "noise" and begin to be treated as what they are: nonconformities with a direct impact on generation (MWh) and revenue (R$).
The opinions and information expressed are the sole responsibility of the author and do not necessarily represent the official position of the author. Canal Solar.
