Dynamic Energy Optimization and Lighting Flexibility Classification for Sustainable Vertical Farming: A...

Vertical farming offers a promising solution to food production challenges in urban and climate-constrained regions, yet its high energy demand raises concerns about sustainability. Most existing studies assess energy demand and CO ₂ emissions under static operational assumptions and lack a comprehensive framework linking seasonal renewable availability, crop cycle timing, and operational flexibility to system resilience and grid dependency.

This study evaluates the performance and carbon footprint of a fully enclosed pilot vertical farming unit in Northern Greece using TRNSYS 18 simulations and high-resolution environmental data. Forty-eight cultivation scenarios were generated by varying photoperiods, humidity levels, and HVAC setpoints to reflect seasonal Mediterranean conditions. Each scenario was analyzed for energy consumption, grid reliance, photovoltaic sufficiency, and life cycle CO ₂ emissions. A dynamic crop cycle estimation model was applied to capture seasonal variability and align planting windows with solar energy availability. Performance benchmarking was implemented through a novel resilience-based multi-criteria framework combining energy-per-cycle metrics with two composite indices, the Grid Independence Index (GII) and the Seasonal Resilience Score (SRS), formulated in this study as a key methodological innovation. A Lighting Flexibility Classification System was further developed to determine the maximum safe reduction in artificial lighting without increasing total energy demand.

The median carbon footprint across the 48 simulated scenarios was 3.67 kg CO ₂ kg ^-1 of lettuce (Lactuca Sativa), with optimized configurations achieving substantially lower emissions compared to conventional greenhouse cultivation. Crop cycle duration varied between 34 and 52 days depending on photoperiod and temperature setpoints. Aligning cultivation cycles with high solar availability, combined with dynamic lighting adjustments, PV contribution to annual electricity coverage was significantly enhanced. In such configurations, lighting inputs were reduced by up to 10% without causing an increase in total system energy requirements.

The findings highlight the potential of simulation-based design for optimizing energy use and minimizing environmental impacts in control-environment agriculture. The proposed metrics and classification provide practical tools for improving the resilience and sustainability of vertical farming under Mediterranean conditions.