Overview

PSR remains actively engaged in the global research community through publications, conference presentations, and collaborative projects. This article highlights a selection of recent and upcoming academic activities that reflect the company’s ongoing work in energy analytics, optimization, and decision-support methodologies.

Last year, Mario Veiga Pereira, PSR’s founder and Chief Innovation Officer, was included for the fifth consecutive year in the Stanford-Elsevier Top 2% Scientists ranking, which identifies researchers with the highest global scientific impact based on publication and citation metrics. More information is available at https://topscinet.com/.

Upcoming Academic Engagements

Upcoming presentations further illustrate the breadth of PSR’s current research agenda, spanning uncertainty modeling, decomposition methods, time-series representation, and expansion planning.

XXIV Power Systems Computation Conference (PSCC) 2026

PSR had two papers accepted for PSCC 2026 (June 08-12), in Limassol, Cyprus. The PSR team is set to deliver the following presentations:

• “Integrated Investment and Operational Planning for Sugarcane-Based Biofuels and Bioelectricity under Market Uncertainty”, by Carolina Monteiro, Bruno Fanzeres, Rafael Kelman, Raphael Sampaio, Luana Gaspar, Lucas Bacellar, Joaquim Dias Garcia, Bruno Santos.
• “An Efficient Hybrid Heuristic for the Transmission Expansion Planning under Uncertainties”, by Joaquim Dias Garcia, Yure Rocha, Teobaldo Bulhões, Anand Subramanian.

2026 SIAM Conference on Optimization (OP26)

PSR had one paper accepted to the 2026 SIAM Conference on Optimization (June 2-5), in Edinburgh, United Kingdom:

• “General Time Series Approximation in SDDP Using Local Markov Switching Autoregressive Models”, by Raphael Sampaio, Guilherme Bodin, Joaquim Dias Garcia, Alexandre Street.

Recent Academic Highlights

International Conference on Stochastic Programming (ICSP 2025)

PSR presented five papers at the XVII International Conference on Stochastic Programming (ICSP 2025), held from July 28 to August 1 in Paris, France. Together, these works reflect the diversity of PSR’s current research agenda.

1. “Multicut Benders Decomposition for Generation Expansion with Dynamic Probabilistic Reserves”, by Tiago Andrade, Alessandro Soares, Fernanda Thomé, Lucas Guerreiro, Luiz Carlos da Costa Junior. This study applies and compares single-cut and multicut Benders decomposition to a generation expansion problem with high renewable penetration and Dynamic Probabilistic Reserve (DPR).
2. “Incorporation of Risk Metrics Electric System Expansion Planning and Portfolio Efficiency Analysis”, by Lucas Guerreiro, Alessandro Soares, Tiago Andrade. This study proposes a methodology to incorporate risk measures into energy system expansion planning through a decomposed investment and operation model. By constructing an efficient frontier, it evaluates trade-offs between cost and risk, supporting more robust long-term decision-making.
3. “ApplicationDrivenLearning.jl: A High-Performance Library for Training Predictive Models in Decision-Making Contexts”, by Joaquim Dias Garcia, Giovanni Amorim, Alexandre Street. The ApplicationDrivenLearning.jl package in Julia provides a flexible interface and high-performance methods – including exact bilevel formulations and heuristic approaches – to make this technology more accessible.
4. “Comparison between different parallelization schemes in SDDP policy training”, by Guilherme Bodin, Joaquim Dias Garcia. This study evaluates different parallelization strategies to improve the computational performance of the SDDP algorithm in large-scale hydrothermal dispatch problems. The results indicate that hybrid approaches combining scenario-based parallelization and subproblem-based parallelization provide the most efficient and robust solution for complex energy systems.
5. “Multicyclic multistage stochastic programming using daisy chains in the electricity market”, by Gabriel Cunha, Bernardo Costa, Guilherme Bodin, Joaquim Dias Garcia, Gabriel Vidigal, Cleyton Santos. This paper analyzes different time discretization strategies for hydropower scheduling problems modeled with stochastic dynamic programming. Using the open-source IARA tool, it explores how discrete representations approximate the underlying continuous and seasonal dynamics of the system.

Brazilian Operations Research Society Conference (SBPO 2025)

PSR took part in the LVII SBPO (October 5-9), in Gramado, Brazil, with a presentation delivered by its CEO, Luiz Barroso, on “The Good, the Bad and the Ugly” of Renewable Energy in Brazil and the Challenges They Pose for Operation and Planning, in addition to three other papers with PSR’s participation that were presented at the symposium:

1. “The IARA Model: A Market Design Simulator for Short-Term Electricity Price Formation”, by Gabriel Vidigal, Gabriel Cunha, Rafael Benchimol Klausner, Guilherme Bodin, Joaquim Dias Garcia, Carolina Monteiro, Pedro Ripper, Nina Hubner and Bruno Peixoto, showcasing the IARA software developed through a PSR-led project, coordinated by CCEE and funded by the World Bank. (Further information: https://iara.psr-inc.com/)
2. “ApplicationDrivenLearning.jl: A High-Performance Julia Package for Training Predictive Models in Decision-Making Contexts”, by Giovanni Amorim, Alexandre Street and Joaquim Dias Garcia. The ApplicationDrivenLearning.jl package in Julia provides a flexible interface and high-performance methods—including exact bilevel formulations and heuristic approaches—to make this technology more accessible.
3. “Internalizing the Opportunity Costs of Integrated Operation and the Attributes of Energy Sources in Energy Auctions”, by Rafael Benchimol Klausner, Alexandre Street, Bernardo Costa, Luiz Barroso and Joaquim Dias Garcia. The Optimized Expansion Auction (LEO) modernizes Brazil’s generation planning by replacing traditional auctions with an integrated model that minimizes total costs, ensures reliability, and accounts for interactions among technologies.

Brazilian Seminar on Electrical Energy Generation and Transmission (XXVIII SNPTEE 2025)

PSR was strongly represented at the XXVIII SNPTEE (October 19-22), held in Recife, Brazil, with ten papers presented by team members and participation as co-authors in two additional papers. More information at: https://snptee.com.br/.

1. “Addressing curtailment: Regulatory Proposals to Improve Efficiency in The Power Sector”, by Thais Sobrosa, Maria Clara Valente, Maynara Aredes, Jairo Terra.
2. “Smart Meters in Brazil: Regulatory Approaches and Implementation Strategies”, by Maria Clara Valente, Thais Sobrosa, Angela Gomes.
3. “Regulatory and Economic Incentives for Pumped Storage Plants: An International Overview for Application in Brazil”, by Fernanda Thompson, Thais Sobrosa, Luiz Rodolpho Albuquerque.
4. “Regulatory Improvements for The Proper Valuation of Hydropower Plant Attributes”, by Thiago Cesar, Alessandra Daniel, Grazziano Motteran, Ivan Carneiro, Henrique Castro, Celso Dall’Orto, Gisella Siciliano, Fernanda Thompson.
5. “Optimization of Energy Procurement for Hybrid Wind-Photovoltaic Plant Portfolios in Brazil”, by Kristian Nolte, Fernanda Thompson, Bruno Fânzeres dos Santos.
6. “Implementation of a Multiterminal HVDC System in Brazil: Benefits and Results Analysis”, by Maynara Aredes, André Tiago Queiroz, Rodrigo Martins, Robson Dias.
7. “Physical Guarantee Calculation for Interconnected Systems”, by Rafael Benchimol Klausner, Bernardo Costa, Joaquim Dias Garcia.
8. “Spatiotemporal Life Cycle Assessment of Hydropower Systems: The Case of Sinop Hydropower Plant”, by Vanessa Albuquerque, Flora Calili, Luiz Rodolpho Albuquerque, Rafael Kelman.
9. “The IARA Model: A Mechanism Simulator for Short-Term Electricity Price Formation”, by Gabriel Cunha, Nina Hubner, Joaquim Dias Garcia, Guilherme Moraes, Gabriel Vidigal, Rafael Benchimol Klausner, Carolina Monteiro, Pedro Ripper.
10. “Case Study for Methodological Validation of the Valuation of Flexibility, Reliability and Resilience Attributes in a Small-Scale System”, by Ana Carolina Cunha, Celso Dall’Orto, Renan Pinho.
11. “Effects of Optimism Bias in Natural Inflow Energy Projections on the Operational Planning of the Brazilian Power System”, by Arthur Brigatto, Joaquim Dias Garcia.
12. “ApplicationDrivenLearning.jl: A High-Performance Library for Training Predictive Models in Decision-Making Contexts”, by Giovanni Amorim, Joaquim Dias Garcia, Alexandre Street.

2025 INFORMS Annual Meeting

PSR presented two papers at the 2025 INFORMS Annual Meeting (October 26-29), in Georgia, United States.

1. “A Post Processing on Locational Marginal Costs Considering Fixed Losses on DC Optimal Power Flow”, by Bruno Amaral, Tiago Andrade, Luiz Carlos da Costa Junior, Sérgio Granville.
2. “Penalty-Free SDDP: Feasibility Cuts for Robust Multi-Stage Stochastic Optimization in Energy Planning”, by Guilherme Freitas, Luiz Carlos da Costa Junior, Tiago Andrade, Alexandre Street.

Publications and preprints

Beyond conference participation, PSR’s research activity is also reflected in journal publications and preprints addressing topics such as inflow forecasting, life-cycle assessment, expansion planning, and stochastic optimization.

Recent Publications (2025 & 2026)

PSR’s research frequently appears in leading academic journals. A few recent examples include:

• “Assessing the Optimistic Bias in the Natural Inflow Forecasts: A Call for Model Monitoring in Brazil”. Arthur Brigatto, Alexandre Street, Cristiano Fernandes, Davi Valladão, Guilherme Bodin, Joaquim Dias Garcia. IEEE Transactions on Energy Markets, Policy and Regulation. 2025. This joint work from PSR and PUC-Rio provides empirical evidence of a systematic optimistic bias in Brazil’s official inflow forecasting methodology, based on the PARp-A model, using 14 years of out-of-sample analysis. https://ieeexplore.ieee.org/document/11079762
• “Integrated Spatiotemporal Life Cycle Assessment Framework for Hydroelectric Power Generation in Brazil”. Vanessa Cardoso Albuquerque, Rodrigo Flora Calili, Maria Fatima Ludovico de Almeida, Rodolpho Albuquerque, Tarcisio Castro, Rafael Kelman. Energies. 2025. This work, part of Vanessa Cardoso Albuquerque’s thesis, validates a spatiotemporal life cycle assessment of the Sinop hydropower plant in Brazil, combining temporal detail and regionalized impacts. https://www.mdpi.com/1996-1073/18/21/5606
• “Trade-off between computation time and solution quality for integrated generation and transmission expansion planning with N-1 security criterion”. Lucas Y. Okamura, Carmen L.T. Borges, Gianfranco Chicco. International Journal of Electrical Power & Energy Systems. 2025. The paper proposes alternative approaches for integrated generation and transmission planning under the N-1 criterion, balancing solution quality and computational effort in the Chilean system. https://www.sciencedirect.com/science/article/pii/S0142061524006690?via%3Dihub

Preprints

• “A General and Streamlined Differentiable Optimization Framework”. W. Rosemberg, Joaquim Dias Garcia, François Pacaud, Robert B. Parker, Benoît Legat, Kaarthik Sundar, Russell Bent, Pascal Van Hentenryck. https://arxiv.org/abs/2510.25986
• “An Efficient Hybrid Heuristic for the Transmission Expansion Planning under Uncertainties”. Yure Rocha, Teobaldo Bulhões, Anand Subramanian, Joaquim Dias Garcia. https://arxiv.org/abs/2602.11490
• “On the tightness of linear relaxations of alternative mixed integer programming formulations for the generator maintenance scheduling problem”. Tiago Andrade. https://arxiv.org/abs/2502.08855
• “Penalty-Free SDDP: Feasibility Cuts for Robust Multi-Stage Stochastic Optimization in Energy Planning”. Guilherme Freitas, Luiz Carlos da Costa Junior, Tiago Andrade, Alexandre Street. https://arxiv.org/abs/2512.03739