🌆 What happens to a city’s power system in the evening?

Electricity demand usually rises in the evening as people return home, turn on lights and household appliances, and charge devices and electric vehicles. At the same time, transport systems, shopping centres, businesses, and municipal infrastructure remain in operation.

⚡ For the power system, this means a rapid shift to a different operating regime. Generation, grids, and substations must adjust to increased consumption while maintaining the reliability and quality of electricity supply.

📊 This is where data and forecasting become especially important. Modern power systems analyze historical consumption, weather conditions, user behaviour, grid loading, and other parameters to anticipate when and where a peak in demand will occur. China illustrates the scale of the challenge. In 2026, amid hot weather and rising consumption, the national electricity load set new records several times, reaching 1.553 billion kW. Interregional power transfers, pumped-storage hydropower plants, energy storage systems, and virtual power plants are used to meet peak demand.

🔋 By the end of 2025, China’s cumulative installed capacity of new energy storage systems had reached approximately 145 GW. Storage systems, however, are only part of the solution. To use them effectively, operators need to forecast when demand will begin to rise and when a storage system should be dispatched.

🚗 Electric vehicles add another layer of complexity: charging peaks may occur late in the evening and overnight. As a result, demand profiles are becoming less predictable and require more accurate analytical models.

🤖 This is where machine learning, time-series forecasting, and intelligent control systems come to the fore. The energy sector increasingly relies on anticipating changes rather than responding only after they occur.