Beyond light scheduling: Harvest-constrained co-optimization of lighting and climate in vertical farms under volatile electricity prices

Indoor vertical farms (IVFs) enable sustainable, year-round food production, but high energy demands limit profitability. Demand response (DR) can reduce electricity costs by shifting energy use over time. However, two key constraints shape DR’s potential. First, crops must reach a target weight by harvest. Second, lighting and climate control share a single energy budget, which previous studies often overlook. We present a three-level hierarchical control framework that exploits both long- and short-term price fluctuations while strictly enforcing harvest weight and date. It builds on a mechanistic model of lettuce growth and indoor climate. Across 36 simulated scenarios at a harvest weight of 250 ± 1.5 g per head, the framework reduced energy costs by 13.7% on average (7.3-19.7%) and energy use by 8.4% compared to a static baseline under realistic prices. About 97% of this saving came from climate control, which ran with the LEDs during the cheapest hours and needed less dehumidification in more humid, crop-safe air. The energy-cost saving rose with the daily range of the electricity price, from 10.5% to 17.9%. Allowing the daily light integral to vary provided a 3-percentage-point benefit over a fixed schedule, while a shrinking-horizon predictive controller under uncertain intraday prices added 0.6-1.4 percentage points over day-ahead planning. Optimizing lighting alone cost about 10% more than co-optimizing lighting and climate. Thus, dynamic lighting paid off mainly together with climate control.