What actually goes wrong with solar farms
Most people talk about desvantagem da usina solar in vague terms. They say intermittent generation or high upfront cost, which is true but useless for anyone who has actually managed a project. The real problems are much more specific and tend to show up late, when you're already committed.
Know the main desvantagem da usina solar before you invest
Let me start with interconnection. This is where the biggest headaches happen. The grid connection process in Brazil alone can take anywhere from 18 to 36 months for projects above 5 MW. Not because of poor planning, but because distribution companies simply don't have the engineering staff to process requests fast enough. I had a project in GoiƔs where the utility company demanded a full power quality study, harmonic analysis, and voltage flicker simulation before even scheduling a site visit. That added eight months and roughly R$120.000 in consulting costs. The workaround was straightforward but not obvious: get a pre-study done by an independent engineer before you even file the formal request. It costs about R$40.000 upfront and usually cuts the waiting time by half because the utility sees your paperwork is already complete. The second problem that nobody warns you about is land subsidence and soil composition. Solar farms sit on vast areas of flat ground, and the ground matters more than you'd think. In the cerrado region, I encountered lateritic soils that shift significantly during the rainy season. Panel mounts designed for stable ground started tilting at angles between 2 and 4 degrees after the first heavy rains. That doesn't sound like much until you calculate the resulting power loss from soiling and tracking misalignment. We ended up replacing the pile foundations with longer steel piles driven 4 meters deep instead of the planned 2. Cost increase of about 15% on the structural side, but it prevented annual maintenance headaches.
Inverter lifespan is shorter than you assume
Most project financial models assume inverters last 25 years alongside the panels. They don't. Central inverters in tropical climates typically last 10 to 12 years before major component failures become frequent. String inverters last slightly longer but still need replacement around year 12 in high-temperature environments. I've seen projects in Minas Gerais where the inverter replacement in year 11 cost nearly R$800.000, which completely threw off their internal rate of return calculations. The financial model assumed zero opex for inverters over the project lifetime, which is wildly optimistic. Here's a detail that tripped up a client of mine recently: transformer losses. The step-up transformers in medium-voltage substations have no-load losses that run continuously whether you're generating power or not. In a project I audited, those no-load losses were consuming about 18 kW constantly, which translates to roughly 157.000 kWh per year of energy that simply disappears. That's not a small number. The fix was installing a smaller reserve transformer and keeping it offline with a disconnect switch until production ramped up to a level where the transformer efficiency justified keeping it energized. This saved about 95.000 kWh annually, which over 25 years is meaningful.
Soiling and cleaning logistics
Dust accumulation is worse than marketing materials suggest. In semi-arid regions, panels lose between 0.5% and 1.2% of output per day without cleaning. That's not theoretical. I measured it myself on a 3 MWp plant in Bahia during the dry season. The cleaning schedule matters enormously here. Some operators clean monthly and see good results. Others wait two months and accept the cumulative loss. The decision depends on water availability and labor costs, which vary by region. Water for cleaning is another hidden bottleneck. A typical 10 MWp plant uses roughly 40.000 liters per cleaning cycle. If you clean every 30 days, that's 480.000 liters per year. In drought-prone areas, sourcing that water becomes a genuine operational problem. One solution I recommended was installing a rainwater collection system connected to the substation roof, which provided about 30% of the cleaning water requirement and eliminated the need for trucked-in water during certain months.
š Clique no botĆ£o abaixo para saber mais sobre o assunto!
Metal price volatility affects your BOP costs
The balance of plant includes a massive amount of steel, aluminum, and copper. These commodity prices fluctuate independently of solar panel costs, which have been dropping steadily. In 2022 and 2023, steel prices jumped significantly, and projects that had fixed-price EPC contracts started losing money on the structural and mounting systems. I worked with a developer who locked in an EPC contract in early 2021 at R$2.800 per kW. By the time procurement started in mid-2022, the same scope would have cost closer to R$3.400 per kW due to steel and aluminum price increases. The contract had a force majeure clause but no commodity adjustment mechanism, so the developer absorbed the entire difference. It's worth negotiating price adjustment indices into your contracts if the project timeline spans more than 18 months.
The degradation curve is not linear
Panel degradation is usually quoted at 0.5% per year, which sounds straightforward. But the first two years of degradation are typically higher, around 1% to 2%, due to light-induced degradation in certain panel technologies. This means your year 1 and year 2 production will be noticeably lower than a straight 0.5% calculation would predict. I've seen developers get surprised when Year 1 yield reports came in 3% below their P50 estimates. The panels weren't defective. The model was just too optimistic on the initial degradation phase. Bifacial panels add another layer of complexity. They're marketed as producing 10% to 25% more energy than monofacial panels, which is true under ideal conditions. But that gain depends heavily on ground albedo. If you install them on dark soil or gravel, the reflective gain drops to maybe 5%. I had a project where we measured actual bifacial gain at only 7.3% because the surrounding terrain was covered in dry grass most of the year. The vendor's spreadsheet had assumed an albedo of 0.3, which applies to white concrete or sand, not the actual ground cover at the site.
Insurance and weather risks
Hail damage is a real and underpriced risk in many Brazilian regions. Standard insurance policies often have high deductibles for weather-related damage, and the claims process can take six months or more. I've seen hail storms in the south destroy entire strings of panels in minutes, with damage exceeding R$2.000.000 on a medium-sized installation. The insurance payout covered most of it, but the revenue loss during the replacement periodāabout four months for procurement and installationāwas not compensated by any standard policy. It pays to verify your coverage includes business interruption or at least calculate the revenue gap yourself and budget for it. Another risk that gets ignored is vegetation growth under panels. In humid regions, grass and weeds can grow tall enough to shade lower rows of panels if left unchecked. Mowing costs add up, and some operators use herbicides, which introduces environmental compliance issues. A hybrid approach works best: mechanical mowing every 45 days during the wet season and targeted herbicide application only in access lanes, not under the panels themselves. This keeps costs down while maintaining safe access for maintenance crews.
Grid curtailment risk
In some regions, the grid operator can and does curtail solar production when there's oversupply. This happened in the Northeast region during certain months when wind generation peaks simultaneously with solar. Projects in those areas face actual revenue risk from curtailment, not just theoretical risk. I reviewed a project in Pernambuco where curtailment events averaged 200 hours per year over a three-year period. That's about 2.3% of potential generation lost. For a 50 MWp plant, that's roughly 115.000 MWh of foregone revenue annually. Any financial model that assumes 100% grid availability in that region is fundamentally flawed. The bottom line is that solar farm projects have a long list of hidden costs and risks that only become apparent after construction begins or during the first two years of operation. Budgeting for interconnection delays, inverter replacements, commodity price swings, and realistic degradation curves is what separates a project that meets its target returns from one that barely breaks even. The technology itself works fine. The economics require careful attention to the details most people overlook.