Sustainable Smart Factory Energy Management Across Eight Industrial Campaigns: A Retrospective Scenario Assessment of Aggregate Demand Flexibility

Sustainable smart factories require energy management that conserves industrial demand and respects receiver capacity across production campaigns. This retrospective study assessed whether a scenario framework could reduce energy above a fixed daily threshold, exceedance days, maximum daily grid energy, and purchased grid energy without deleting demand. The dataset contained 435 daily records from eight campaigns, including 423 finite positive observations used for scenario evaluation, three products, six metered energy sources, and 10,440 hourly weather records. Weather-derived photovoltaic availability was coupled with daily battery-grid accounting. Data from 2019–2024 supported development, data from 2025 supported temporal validation, and data from 2026 supported chronological testing. Five scenarios compared the baseline (S1) with deterministic tuning (S2), demand-side management (S3), a genetic algorithm (GA; S4), and particle swarm optimization (PSO; S5) under identical objectives and constraints. Across all data, S2, S4, and S5 reduced above-threshold energy by 6.9%, left exceedance days and maximum daily grid energy unchanged, and increased purchased grid energy by 0.1%. Their 2026 reduction was 1.0%. Accepted transfers represented 0.50–0.68% of demand, retained unplaced requests at source, met the three-day limit, and caused no receiver violations. GA and PSO returned identical objectives and outcomes across 30 seeded runs each, providing scenario-based evidence under aggregate daily assumptions.