Rising electricity prices have increased the importance of energy-aware scheduling due to higher production costs. In environments with fluctuating power demand, shifting energy-intensive operations to periods of lower electricity prices can reduce electricity costs but may lead to increased tardiness. This trade-off is addressed in this thesis using constraint programming to evaluate schedules based on a weighted objective combining electricity and tardiness costs. To reflect intra-job variability in energy use, each job was decomposed into three non-preemptive steps following a low–high–low consumption pattern. To assess the impact of different time-of-use (TOU) pricing designs, 20 configurations were analyzed, varying interval duration, number of price levels, objective weightings, and problem size. In a second step, a real-world scenario was examined by comparing forecast-based scheduling using an hourly forecast by an autoregressive integrated moving average (ARIMA) model as well as TOU prices under perfect foresight, both benchmarked against the schedule resulting from an Oracle using actual electricity prices. Results show that classical TOU schedules with two price levels and longer intervals often yield lower overall costs, reflecting the interaction between interval length and job duration. In the real-world comparison, such TOU schedules under perfect foresight outperformed those based on ARIMA forecasts in 75% of cases. Overall, the results provide insights into how TOU tariff design influences scheduling performance and justify the use of TOU pricing relative to ARIMA-based forecasting.
«
Rising electricity prices have increased the importance of energy-aware scheduling due to higher production costs. In environments with fluctuating power demand, shifting energy-intensive operations to periods of lower electricity prices can reduce electricity costs but may lead to increased tardiness. This trade-off is addressed in this thesis using constraint programming to evaluate schedules based on a weighted objective combining electricity and tardiness costs. To reflect intra-job variabil...
»