Brazil occupies a rare position in the global green hydrogen landscape: a combination of high solar irradiance, consistent wind patterns in the Northeast, and a relatively clean electricity grid. This comparative advantage is real and should not be minimized. The mistake lies in treating it as a sufficient condition.
Reuters' coverage from March 31, 2026, documented something the electricity sector had already felt for months: renewable energy producers like Atlas Renewable Energy, Newave Energia, and Voltalia implemented operational cuts and reduced staff due to generation constraints, revenue losses, and a lack of regulatory clarity on compensation. Atlas reported revenue losses of approximately 25% in some units. Voltalia conditioned new investments on greater long-term return stability.
Curtailment: from exception to permanent reality
The National Electric System Operator (ONS) no longer treats curtailment, the forced shutdown of renewable generation, as an operational anomaly. It sees it as an intrinsic characteristic of systems with high penetration of variables. The data support this interpretation:
- In 2022, the average annual restriction corresponded to 0.5% of the renewable potential, with approximately 46 MW on average cut.
- In 2023, the index rose to 3.6%, equivalent to an average of 389 MW.
- In 2024, it reached 9.3%, with an average of 1,447 MW restricted.
The trajectory is unequivocal. The system has incorporated more renewable capacity, but has proportionally lost the ability to transform this potential into effectively utilized energy. The second half of 2024, especially August, September, and October, registered persistently high levels of restriction, ruling out the hypothesis of punctual seasonality.
The Ministry of Mines and Energy (MME) was even more direct in material finalized in February 2026: curtailment for energy reasons tends to become structural and predominant by 2029. When a ministry classifies a problem as structural, it is no longer possible to treat it as a short-term issue to be absorbed by the market.
Why does this make green hydrogen more expensive?
The economic logic is straightforward. Documents from the Energy Research Company (EPE) estimate that 70% to 80% of the cost of electrolytic hydrogen is determined by the price of electricity delivered. IRENA reinforces this: cost reductions in electrolyzers do not compensate for high energy prices.
The most frequent misconception in public debate is confusing the cost of renewable generation with the cost of electricity actually delivered to the electrolyzer. A wind farm may have a competitive LCOE (Liquid Cost of Energy) and still burden the hydrogen project due to a combination of factors:
- Curtailment that forces the generator to purchase supplementary energy on the market;
- Inappropriate time profile for continuous operation of the electrolyzer;
- Oversizing generation capacity to compensate for intermittency;
- Risk premium embedded in PPAs due to regulatory uncertainty regarding loss compensation.
The electrolyzer needs high utilization to dilute CAPEX. Short, unpredictable, or legally uncertain power supply erodes the competitiveness equation even before the project becomes operational.
The paradox of restricted abundance
EPE mapped, in a 2024 technical note, 11 registered projects totaling 45.4 GW of installed capacity for hydrogen production in the Northeast region of Brazil by 2038. However, the estimated aggregate prospective margin for connecting large-scale loads in that region ranged between 3.9 GW and 8.35 GW, with the explicit caveat that dynamic constraints would need to be evaluated on a case-by-case basis.
Therefore, there is a significant gap between the narrative of abundance and the physical capacity of the system to absorb new energy-intensive loads. What seems like a solution—using the renewable surplus to produce hydrogen—may simultaneously represent a new vector of pressure on a grid already operating with less margin.
What public planning already recognizes
The Brazilian government has identified the problem and has been acting on multiple fronts. In December 2025, the Ministry of Mines and Energy (MME) opened a public consultation to discuss financial compensation for generators affected by power cuts. In April 2026, the National Electric Energy Agency (ANEEL) put out for consultation the bidding rules for Transmission Auction No. 4/2026, with a forecast of R$ 11.3 billion in investments, 2,069 km of new lines and 13,564 MVA in transformation capacity.
In a 2025 study, EPE recommended a 600 kV, 3,000 MW HVDC-VSC system to increase the Northeast's export capacity to approximately 24 GW and allow for the integration of up to 60 GW of renewables by 2033.
These are moves in the right direction. But transmission projects have a maturation period of three to five years. Industrial projects don't wait indefinitely.
Six conditions for competitive hydrogen.
Based on documents from ONS, EPE, MME, and ANEEL, it is possible to summarize what needs to be improved:
- Network expansion prioritizing integration between the Northeast and Southeast regions and connecting energy-intensive loads.
- Mature regulation for storage: In April 2026, ANEEL authorized the first storage unit linked to a power plant, an important milestone, albeit symbolic given the scale of the challenge.
- Predictability regarding loss compensation, reducing the risk premium in contracts and financing.
- Locational perspective for hydrogen: the point of connection and proximity to grid reinforcements are as relevant as for solar or wind resources.
- A clear distinction exists between demonstration and industrial projects; the latter require contractual and electrical robustness that pilot projects do not need to demonstrate.
- Analytical honesty in sectoral discourse: natural abundance is not the same as delivered infrastructure.
Conclusion
Brazil's renewable energy advantage remains real. What is no longer sustainable is presenting it as an automatic equivalent of cheap hydrogen.
The competitive cost of the molecule will be defined by the quality of the integration between generation, transmission, system operation, and storage regulation, not by the nominal LCOE of a solar park in Piauí. As long as curtailment grows, the grid does not gain robustness, and operational flexibility does not mature, part of Brazil's competitive advantage will continue to be wasted potential.
Cheap hydrogen doesn't come from nominally cheap megawatts. It comes from megawatts delivered, predictable, and usable at the intensity the electrolyzer requires.
