Why demand needs to change
The energy crisis is a longstanding issue. Today's crisis, however, is different because it coincides with structural changes. Expanding renewable energy makes supply fluctuate, while extreme heat and the AI industry push up demand. These changes are occurring simultaneously and are interconnected.
At 7 p.m. on August 7, electricity demand surged to 95.3 GW, equivalent to the output of 68 new nuclear reactors. KPX estimates that demand would have exceeded 100 GW without the reduction in industrial operations during the vacation season. By contrast, there were 82 instances on the mainland alone last year of curtailment that forcibly halted solar and wind generation during daytime hours in spring and autumn. Electricity is tight in summer, while surplus electricity goes unused during spring and autumn days.
The response so far has been straightforward: build more power plants to address electricity shortages, and expand the grid and large battery energy storage systems, or ESS, to store surplus electricity. But building power plants and transmission networks takes years, and ESS is expensive. Those costs ultimately feed into electricity bills. Treating demand as fixed and increasing supply alone will not solve the problem. We now need to consider supply, demand and institutions as a single system and accelerate demand-side solutions.
This is not simply a call to conserve energy. Demand management has so far focused on how much less electricity is consumed, but the value of the same savings differs by time of day. Saving electricity in the evening can reduce the need to build power plants, whereas saving it during the day only leads to more solar generation being discarded. An alternative is therefore load shifting: reducing consumption in the evening and increasing it during the day.
Load shifting is already a practical option. Google has signed agreements totaling 1 GW with US utilities under which it will reduce or shift some AI workloads when electricity supplies are tight. In modeling by my research team, charging demand shifted to daytime when electric vehicle charging prices were set lower during the day and higher in the evening, and more workplace chargers were installed. The share of renewable energy going unused fell from 5% to 2%, and ESS worth 14 trillion won that had been needed to meet the curtailment target was no longer necessary.
Changing tariffs alone increased inconvenience for drivers and resulted in a net loss for society. Adding chargers at the same time, however, produced a net benefit even after installation costs. This means price signals and access conditions must change together.
Planning and compensation
The demand forecast proposal for the 12th Basic Plan on Long-term Electricity Supply and Demand, released in August, also included a target to reduce peak electricity demand in 2040 by 20.8 GW through demand management. That is equivalent to 15 new nuclear reactors. Alongside efficiency improvements, it included load shifting as an implementation measure. But setting a target is different from delivering it. Making load shifting a usable resource requires concrete action.
First, demand management must be evaluated on the same basis as power plants. This means comparing the cost of meeting 1 GW of evening peak demand with power plants against the cost of shifting charging or industrial processes, or drawing power from electric vehicle batteries. The EU has legislated for flexibility needs to be assessed every two years and national targets to be set. The UK also included consumer flexibility alongside generation and storage in its planning last year. Demand management targets in our Basic Plan on Long-term Electricity Supply and Demand should likewise be set by comparison with supply alternatives, and annual delivery should be reviewed just as it is for generation facilities.
Next, the value provided by shifting demand must be fairly compensated. Factories receive compensation in the demand resource trading market simply for being ready to reduce consumption. Household and commercial demand response, however, is activated only in exceptional circumstances such as supply emergencies, so it is rarely used in normal conditions. In the UK, Octopus Energy provides effectively free charging electricity to drivers who entrust it with vehicle-to-grid charging and discharging control, or V2G, and leave their cars plugged in for extended periods. It creates value through everyday charging and discharging rather than emergency activation. We, too, should enable providers that aggregate small resources from households and electric vehicles to receive market compensation commensurate with their value on a regular basis, not just during emergencies.
Enabling changes in consumption
Finally, infrastructure is needed to turn tariff signals into action. Tariffs and charging services must be improved so that time-of-use price signals reach households and electric vehicle drivers. Even if daytime charging is cheaper, demand will not shift without workplace chargers. Factories and data centers need equipment and contracts that allow them to adjust working hours. Households need automated controls that operate devices once users specify their departure time, charging amount and indoor temperature. Tariffs and access conditions must change together for load shifting to benefit society as a whole.
Future electricity planning will be judged as much by how much demand it can shift as by how many additional power plants it builds. Resources that can provide a cheaper and faster alternative to power plants are already present in our parking spaces, living rooms and factories. Reliability and carbon neutrality in the 100 GW era depend on putting in place now the rules and mechanisms needed to make that 20.8 GW available.
The views expressed in this column may differ from the publication's editorial position.
Original Korean article: ZDNet Korea
English brief by Korea Wind Intelligence, koreawindintel.com
