INTEGRATED WATER-ENERGY-FOOD MANAGEMENT IN A CHANGING CLIMATE: A HYBRID STOCHASTIC-EVOLUTIONARY APPROACH FOR ADAPTIVE INFRASTRUCTURE PLANNING

Código

XXVI-SBRH0155

Autores

Julieta Augusto Nhampossa, Guilherme Fernandes Marques, Ana Paula Dalcin, AMAURY TILMANT, FILIPE SAMPAIO CASULARI PINHATI, LINEU NEIVA RODRIGUES

Tema

A - Planejamento e Gestão de Recursos Hídricos e Segurança Hídrica

Resumo

Water demands for food and energy production require integrated resource management, especially under climate change. Water storage infrastructure, both small and large, is essential to address temporal imbalances and improve water access. However, if storage expansion and operation are not aligned with agricultural and energy planning, there is a risk of conflicts and missed opportunities for enhancing water security. This study analyzes how infrastructure and water use planning can be adapted to meet irrigation and hydropower demands under different climate scenarios. Rather than treating irrigation expansion as fixed, it is modeled as a decision variable, enabling exploration of trade-offs between water storage investment, crop area expansion, water allocation, and hydropower production. A hybrid hydro-economic model combining Stochastic Dual Dynamic Programming (SDDP) with a Multi-Objective Evolutionary Algorithm (MOEA) is applied to the São Marcos watershed in Brazil. Results show that irrigation can double, as projected by ANA, but this requires a 50% increase in small-scale storage and would reduce hydropower by 21 GWh/year. In a drier climate future, this trade-off increases to -37.9 GWh/year. If irrigation expands 118%, storage will need an additional 100 hm³ in small storage, but at the expense of 20% less in hydropower. The region holds potential to support both agriculture and hydropower, but this must be accompanied by significant efficiency gains (e.g., crop yield per m³) to adapt to future climate conditions. This study provides strategic guidance for decision-makers in optimizing integrated water system operations.