Hudson Giovani Zanin, Author in Canal Solar

This paradox encapsulates one of the central problems of Brazil's energy transition. What appears on the screen as a "generation interruption" is neither an isolated incident nor a spreadsheet error. It is a symptom of a system that has grown too fast in some areas and too slow in others. In just a few years, Brazil tripled its installed capacity from variable renewable sources, totaling around 30 GW to approximately 90 GW when considering wind, centralized solar, and distributed generation. The expansion was celebrated as a sign of modernization of the parent company, but it was not accompanied at the same speed by network reinforcements, model revisions, and updates to the market design. In a recent meeting with industry representatives, he laid bare the disconnect. Until August 2023, the ONS (National System Operator) worked with a "safety region" calculated from models that assumed a certain behavior of wind and solar power plants. The blackout of August 15th exposed that this margin was overly optimistic. Overnight, the area considered safe for operations shrank. The system that appeared capable of handling 33 GW of renewables and still accommodating another 29 GW already contracted revealed that part of that capacity was a statistical illusion. The reaction was typical of crisis management. Transmission limits have been reduced as an emergency measure. Models and assumptions were revised at an accelerated pace. Network procedures have been updated to restore a minimum operational comfort level. A new "playbook" was hastily constructed to try to restore the security standards of the day before the blackout, now with a more realistic view of the system's dynamics. Meanwhile, the broadcast auctions continued as normal. Billions of dollars worth of additional funding were contracted, primarily to increase the flow of renewable energy production from the Northeast to the Southeast. Export limits from the Northeast plummeted immediately after the disruption, then rebounded as new projects came online, and today they have already surpassed pre-event levels. When the dust settled, however, the diagnosis became more troubling. Even with the expansion of transmission, the main problem has shifted from electrical to energy-related. The bottleneck is no longer just the "pipe" that carries energy from one point to another, but the lack of sufficient load during the peak generation window. The duck curve, so often presented in international seminars as a conceptual curiosity, has come to describe the day-to-day operation. Zucarato presents a piece of data that summarizes the change. In just one year, with the addition of approximately 10 GW of distributed resources, the minimum midday load dropped from 39 GW to 32 GW. There is 7 GW less of net demand available for large centralized power plants to provide frequency, voltage, and inertia services. At the same time, the nature of the generation gap changed. In 2022, curtailment was mainly explained by incomplete network and reliability constraints. In 2023, after the disruption in August, the security component gained importance and the energy component began to emerge. In 2024 and 2025, the photograph reverses in several months. The problem becomes dominated by a structural excess of energy at times when the system is already saturated with cheap generation. ONS simulations for the period between 2026 and 2029 reinforce this trend. Using 2024 as the base year for hourly load, wind, and sun profiles, the operator projects the future with two supply scenarios. One scenario follows the energy trajectory of the Monthly Operating Plan. The other considers the maximum electricity supply, incorporating all signed transmission system usage contracts. The result is clear. Between 9 am and 16 pm, a growing portion of the hours requires generation shutdowns. In the scenario aligned with the PMO (Power Management Observation), almost half of the daylight hours already show some level of curtailment, with peaks that can exceed 20 GW. In the most heavily loaded electrical scenario, the situation worsens. More than 80% of daylight hours will experience generation cuts, with losses that can reach up to 40 GW. Since solar power generation is concentrated precisely in the "belly" of the curve, it tends to be proportionally more affected. Simulations indicate average annual cuts of around 10% for wind power and 20% for solar power in this most extreme scenario. Even so, there is still underestimation. Several local electrical constraints are not fully represented in the models. Even when this electrical layer is ignored, the purely energy component is already large enough to worry investors, banks, and policymakers. By including distributed generation (DG) in the hypothetical allocation of energy cuts, the Ministry of Mines and Energy (MME) found another sensitive piece of data. If the projected average 1000 MW of cuts are divided between centralized and distributed generation, micro and mini-generation could account for between 50% and 60% of the "contribution" to the solution. In other words, distributed generation can no longer be treated as an external element to the problem. It becomes a central part of the equation and will also have to be part of the solution. The analysis gains an additional layer when the perspective of economist Paulo Sehn is incorporated. He shifts the focus from individual criticism to the ONS and puts the issue back in its proper context. The Brazilian situation is not an isolated deviation, but a local manifestation of a global phenomenon. Several countries accelerated the integration of renewables without adequately coordinating grid expansion, market design, and operating rules. What changes from one case to another is how each society chooses to distribute the cost and reorganize the transition. Sehn draws attention to the immediate financial impact of curtailment. Projects were structured based on revenue curves that fail to materialize when the plants are systematically shut down at certain times of the day. This mismatch reopens contracts, worries financiers, and erodes the trust built up over years in financing generation assets, particularly in the free contracting environment. The risk ceases to be merely technical and becomes a systemic credibility risk. The subsidy trajectory adopted by the country reinforces this weakness. Incentive programs for distributed generation, discounts on grid connections, schemes like PROINFA, and other mechanisms have pushed renewable expansion at a pace and in a direction that does not align with the needs for firm power and fine-grained operational control. The cost associated with these incentives is already approaching tens of billions of reais annually. This model worked while renewable energy penetration was proportionally low and the system had hydraulic slack to accommodate frequency fluctuations. At this new level, it tends to produce an increasingly pronounced duck curve and a price structure that clearly signals neither scarcity nor abundance. The Brazilian scenario takes on additional dimensions when compared to other jurisdictions. In 2024, electrical systems with high solar and wind penetration recorded technical and economic losses associated with curtailment of between 5% and 20% of total renewable generation in certain regions. In Texas, operated by ERCOT, the annual renewable energy cut already exceeds several terawatt-hours, driven by grid congestion and recurring negative daytime prices. In China, western and northern provinces have recorded cuts exceeding 10% in wind farms in some years, despite the aggressive expansion of ultra-high voltage power lines. In Spain and South Australia, the incidence of zero or negative prices has become a structural component of the market since 2021. This body of evidence reinforces one conclusion. Curtailment is not about "excess energy" in the simplistic sense. It's about the misalignment between generation profile, network infrastructure, and consumption patterns. This misalignment first appears in the hourly price. When the spot price approaches zero, the system indicates that the additional megawatt-hour has a marginal value of zero at that instant. When the price turns negative, the message is even more forceful. That megawatt-hour has a negative marginal value and needs to be avoided to preserve the stability of the system. In nodal markets such as ERCOT, CAISO, and PJM, prices between minus 10 and minus 150 dollars per megawatt-hour are already a recurring reality during peak solar hours. These negative prices are at the economic heart of the dilemma between battery storage systems and bitcoin mining. BESS comes in as a technology that buys very cheap or negative energy in order to resell it during peak hours. Bitcoin mining, in turn, converts cheap energy into processing power and immediate financial revenue. A crucial difference is that an investment leaves a systemic infrastructure for the electrical system. The other generates only speculative financial flow tied to a volatile digital asset. From a cost perspective, utility-scale BESS systems have shown global average values between 140 and 180 dollars per kilowatt-hour for projects lasting four hours, with even more aggressive figures in markets such as China. Full-cycle efficiencies between 85% and 92%, lifespans of thousands of cycles, and levelized storage costs ranging from $60 to $150 per megawatt-hour allow batteries to replace thermal power plants in peak-demand and operational support services in markets such as California, Australia, and the United Kingdom. A BESS (Business Support System) model combines arbitrage between cheap and expensive hours, ancillary services, and remuneration based on firm capacity, which enables contracts of 10 to 20 years and reduces uncertainty for investors. Bitcoin mining operates on the opposite logic. Its global energy consumption already represents a significant fraction of the world's electricity, and the total cost of operation is dominated by the price of energy, which can account for up to 70% of recurring expenses. To survive after successive halvings, miners need very cheap energy, in the range of $0,04 to $0,06 per kilowatt-hour, or less. Curtailment energy, priced at zero or with a negative value, fits perfectly into this equation. This is why miners concentrate on regions with a renewable energy surplus and transmission bottlenecks, such as West Texas, remote areas of China, parts of Kazakhstan, Canada, and, more recently, the Brazilian Northeast. In this context, the relevant question ceases to be whether or not there is a private interest in exploiting this surplus. The question then becomes what kind of system architecture the country wants to build. Storage alters the structural dynamics and simultaneously reduces energy losses, indirect costs associated with thermal power plants running idle, and risks of instability. Mining only absorbs the surplus at the point of failure, without strengthening the long-term electrical infrastructure. In Brazil, the PLD (Price of Settlement of Differences) has already begun to signal significant distortions during peak renewable generation and low load periods, especially in the Northeast region. Given the rapid growth of solar and wind power, the trend is for this situation to worsen. Without proper storage and market design, the system accumulates energy waste and tariff burden. With storage and arrangements that reward flexibility, there is room to transform part of the problem into an opportunity for efficiency. The decision, however, is not merely a technical one. It involves regulatory choices, subsidy review, repositioning of agents, and confronting simplistic narratives that have gained traction in the last decade. Curtailment is neither the villain nor a statistical detail. This is a strong indication that the system has grown faster than the governance intelligence that organizes it. Ignoring this signal means pushing the bill forward, with an increased risk of instability and higher costs for the consumer. Facing it head-on means admitting that the energy transition requires more than slogans, vague incentives, and photos next to solar panels. It requires market redesign, coordination between institutions, and political courage to correct course while there is still time to transform the current discomfort into learning, and not into a permanent pattern. You might also be interested in! https://canalsolar.com.br/regulacao-do-armazenamento-energia-brasil/

Even with "infinite transmission," there isn't enough load to absorb so much renewable energy during the day.

Even with "infinite transmission," there isn't enough load to absorb so much renewable energy during the day.

How to size photovoltaic systems for Group A?

Tax on solar panels threatens business

Effects of reverse flow and its impacts on photovoltaic systems

See the ranking of the main module suppliers until the 3rd quarter of 2023

AFCI requirement sparks race among manufacturers to adapt inverters

Even with "infinite transmission," there isn't enough load to absorb so much renewable energy during the day.

Even with "infinite transmission," there isn't enough load to absorb so much renewable energy during the day.

How to size photovoltaic systems for Group A?

Tax on solar panels threatens business

Effects of reverse flow and its impacts on photovoltaic systems

See the ranking of the main module suppliers until the 3rd quarter of 2023

AFCI requirement sparks race among manufacturers to adapt inverters

Subscribe to our weekly newsletter