Can industrial overcapacity enable seasonal flexibility in electricity use? A case study of aluminum smelting in China

TL;DR

This study investigates whether industrial overcapacity can provide seasonal flexibility to electricity use in decarbonized grids, using China's aluminum smelting as a case study. It develops a co-optimization framework that couples detailed smelter operations with a provincial power-system expansion model, aided by D3R dimensionality reduction, across hundreds of scenarios toward 2050 net-zero. The findings show that retaining about 30–36% overcapacity can yield substantial electricity-system cost savings (15–72 billion CNY/year) and lower aluminum production costs, while also stabilizing employment through complementary seasonal patterns in industry and power generation. These results suggest overcapacity can be reframed from a problem into an asset for grid integration and industrial resilience, with implications for other energy-intensive industries and policy design to monetize such flexibility.

Abstract

In many countries, declining demand in energy-intensive industries (EIIs) such as cement, steel, and aluminum is leading to industrial overcapacity. Although overcapacity is traditionally seen as problematic, it could unlock EIIs' flexibility in electricity use. Using China's aluminum smelting sector as a case, we evaluate the system-level cost-benefit of retaining EII overcapacity for flexible electricity use in decarbonized systems. We find that overcapacity enables smelters to adopt a seasonal operation paradigm, ceasing production during winter load peaks driven by heating electrification and renewable seasonality. In a 2050-net-zero scenario, this paradigm reduces China's electricity-system investment and operating costs by 15-72 billion CNY per year (8-34% of the industry's product value), enough to offset the costs of maintaining overcapacity and product storage. Seasonal operation also cuts workforce fluctuations across aluminum smelting and thermal-power sectors by up to 62%, potentially mitigating socio-economic disruptions from industrial restructuring and the energy transition.

Figures (7)

• Figure 1: Conceptual framework for overcapacity-enabled seasonal flexibility in energy-intensive industries. The left panel illustrates the spring to autumn time period when renewable energy generation is abundant, with aluminum smelters operating at full capacity, producing aluminum for both meeting immediate demand and storing in inventory. The right panel illustrates winter scenarios where renewable generation is limited and heating electrification exacerbates peak demand, compelling coal and natural gas power plants to operate. During this period, aluminum smelters shut down and meet demand by drawing from product inventory. This seasonal operational paradigm leverages excess capacity to create complementary patterns between industrial production and energy system dynamics, enabling cost-effective decarbonization.
• Figure 2: China’s projected aluminum demand and smelting overcapacity under different scenarios. The top solid line represents total aluminum demand, the green area shows recycled aluminum production (RA), and the yellow area indicates primary aluminum production (PA) up to 2050. China’s aluminum smelting capacity at the end of 2024 (45 million tonnes (Mt) per year) is represented by the red dashed line. The area between the red dashed line (smelter capacity at 2024 level) and the yellow area represents the excess capacity of aluminum smelters unless decommissioned. The overcapacity rate is represented by the purple line, and is given by one minus the utilization rate (1 - demand/capacity). Different scenarios are constructed based on various assumptions regarding economic development and aluminum product import/export levels, with the variable settings outlined in Table \ref{['tab:scenario_settings']}. Even under the high-demand scenario, China’s aluminum industry would retain 46% excess capacity — and as much as 86% under the low-demand scenario — posing a significant challenge to the sector.
• Figure 3: Reduction in electricity system costs due to overcapacity-enabled flexibility in aluminum smelting, compared to the no-overcapacity case. Colors represent different sources of changes in system cost: renewable investment costs (green) and non-renewable operational costs (orange) account for the largest system cost reductions. The reduction in renewable investment costs reflects improved renewable energy utilization efficiency through concentrating smelting during periods of high renewable generation. The reduction in non-renewable operational costs reflects a decreased reliance on firm generation resources (coal, gas, biomass, and nuclear) to ensure power supply reliability. Cost reductions range from 15 to 72 billion CNY (8-34% of the aluminum smelting industry’s product value in 2050) across scenarios with different aluminum demand, smelter flexibility, and technology costs. The only increased investment is in heating electrification. By mitigating winter peak loads, flexible smelter operation reduces the need for costly clean firm generation and long-duration storage, thereby lowering the grid integration barriers for heating electrification.
• Figure 4: Trade-off analysis and cost-effective strategy for retaining overcapacity. In (a), electricity system cost savings (blue), smelter operational cost increase (orange), and net system benefit (black line) are shown for 2050 in the core scenario (Mid-flexibility, Mid-demand, and Mid-technology costs). The maximal net benefit (electricity system cost reduction minus smelter operational costs) is achieved with 36% overcapacity in 2050. In (b), the mean optimal overcapacity rate increases from 6% in 2030 to 30% in 2050, due to higher flexibility requirements as the electricity system approaches net-zero emissions. However, it should be noted that the cost-effective capacity of aluminum smelting still decreases across the years due to declining primary aluminum demand. The net benefit across different scenarios first decreases as primary aluminum demand declines to 2040 levels, then increases in the net-zero emissions scenario in 2050, where load flexibility is more valuable.
• Figure 5: Seasonal operation of aluminum smelters is complementary to energy system seasonality. In (a), smelter production largely ceases from mid-November to mid-March with demand met by stored inventory, but operates at full capacity from late March to early November. In (b), the electricity consumption of aluminum smelting is largely complementary to China’s heating demand, avoiding the severe winter electricity peaks driven by heating electrification.