Rapid cost decrease of renewables and storage accelerates the decarbonization of China's power system

China’s power sector produces roughly half of the nation’s energy-related CO2 emissions (~14% of global energy-related CO2), making power-sector decarbonization central to national and global climate goals. Although China’s Paris commitments (emissions peak and 20% non-fossil energy by 2030) are important, they are insufficient to limit warming to 2 or 1.5 °C. Recent, dramatic cost declines in solar PV, wind, and battery storage have been underestimated in mainstream outlooks, potentially understating feasible decarbonization pathways. This study integrates updated cost trajectories into a detailed power-system model to explore how China’s electricity system could evolve by 2030 under stricter CO2 targets. The research questions are: (1) How would China’s power system change given rapid decreases in renewable and storage costs under more stringent CO2 targets? (2) What are the system costs to achieve those changes? (3) How would these changes reshape regional development and transmission needs?

Prior studies outline routes to high levels of non-emitting generation in China by mid-century, but many have not captured the recent rapid declines in renewable and storage costs. Major outlooks (IEA’s World Energy Outlook and EIA’s International Energy Outlook) have historically under-estimated renewable growth. Empirical cost trends—global weighted-average LCOE reductions since 2010 of ~77% (utility-scale PV), 35% (onshore wind), and 85% (battery storage)—suggest greater potential for renewable penetration than previously modeled. The literature gap addressed here is the incorporation of these updated cost trajectories into system planning models to reassess feasible and economical decarbonization by 2030, including regional development and transmission implications.

The authors updated the SWITCH-China capacity expansion and operations optimization model to represent China’s power system at provincial spatial resolution with hourly temporal resolution. SWITCH minimizes total system costs (capital, fixed and variable O&M, fuel, and transmission) while meeting projected demand and subject to reliability, operational constraints, resource availability, and policy constraints. Long-term investments (generation, storage, transmission build/retirements) and short-term dispatch are co-optimized. Costs for solar, wind, and storage follow aggressive cost-decline trajectories consistent with NREL’s Annual Technology Baseline (ATB) to 2030; other technology costs follow original SWITCH-China assumptions. The analysis defines four 2030 scenarios: (1) BAU: continuation of existing policies, conventional renewable cost projections, and no new coal plants after 2020 due to air-pollution and carbon-mitigation regulations; (2) R: low-cost renewables with rapid cost declines for solar, wind, and storage; (3) C50: R plus a power-sector CO2 cap 50% below 2015 by 2030; (4) C80: R plus an 80% CO2 reduction from 2015 by 2030. Operational simulations examine hourly dispatch feasibility, including storage charge/discharge and gas peaking. Regional transmission expansion is endogenously optimized, identifying new interprovincial transfers and corridor capacities. Sensitivity analyses test (a) higher demand (D+20%: linear 20% increase by 2030) and (b) higher renewable and storage capital costs (C+20%: 20% above R). CO2 accounting and demand projections follow documented methods in supplementary materials.

- Capacity mix (2030): Under the R scenario vs BAU, wind rises from 660 to 850 GW and solar from 350 to 1260 GW; storage expands from 34 to 290 GW; gas capacity declines from 300 to 170 GW; coal capacity falls from 750 to 700 GW (~7%). Under C50, coal capacity drops to 520 GW; under C80, coal falls to ~200 GW (~4% of total capacity), offset by ~1920 GW solar and ~2000 GW wind. - Generation mix (2030): R reduces coal generation from 4900 TWh (BAU) to 3000 TWh (-30%). Wind+solar provide 39% of electricity; total non-fossil share reaches 62%. C50 lowers coal to 2400 TWh and lifts non-fossil to 77%. C80 cuts coal to ~960 TWh (~10% of generation), with non-fossil approaching 90%. - Emissions and costs: BAU emissions in 2030 are 3980 MtCO2 (≈5% above 2015). R lowers emissions to 2970 MtCO2 (≈22% below 2015) while reducing average power cost from 73.52 to 65.08 $/MWh (-11%). C50 halves 2015 emissions by 2030 with a cost of 69.47 $/MWh (7% above R, 6% below BAU). C80 achieves an 80% reduction with cost 89.08 $/MWh (+21% vs BAU). Cost of conserved CO2: R = -$36/tCO2; C50 = -$9/tCO2; C80 = $21/tCO2. - System cost structure (2030): Overall system cost is $310B (BAU), $280B (R; -11%), $285B (C50), $390B (C80). Coal fuel costs drop from ~$100B (BAU) to ~$65B (R). New capex for solar/wind/storage rises modestly from ~$55B (BAU) to ~$65B (R). - Operations feasibility: Up to ~300 GW of storage is required under R to balance high VRE days; gas generation covers low VRE periods and peaks. Hourly dispatch is operationally manageable with storage and gas under R and C50. - Transmission and regional shifts: New interprovincial transmission corridors up to ~35 GW (doubling current max cross-provincial capacity) are needed to move NW wind/solar to eastern demand centers. In 2030 under R, the Northwest exports 672 TWh (to Central), 287 TWh (to North), and 90 TWh (to East). The Eastern grid imports 287 (NW), 125 (North), 111 (Central), 57 (South), and 22 TWh (Northeast). Under C80, Northwest generation exceeds its own demand by >300%, while the Eastern grid produces ~50% of its demand. - Storage scale: R requires ~307 GW storage providing ~250 TWh annual charge/discharge (~2.2 h/day); C80 requires ~525 GW and ~388 TWh (~2 h/day). With 2030 pumped hydro potentially ~130 GW, batteries would need to reach ~177 GW to meet R’s storage requirement. - Sensitivities: With demand +20% (D+20%), 2030 solar/wind capacities rise to ~1890/1040 GW. With capital costs +20% (C+20%), solar/wind capacities fall to ~980/650 GW. Generation patterns track capacity outcomes but preserve overall structural trends toward high VRE and lower coal.

The findings show that incorporating recent cost declines for solar, wind, and storage fundamentally reshapes least-cost pathways for China’s power system by 2030. Rapid VRE deployment, complemented by storage and flexible gas, can deliver large emissions reductions at equal or lower system cost than BAU, addressing the study’s questions on system evolution and costs. Regional analysis indicates a reorientation of generation to resource-rich northwest provinces and increased reliance on transmission to supply major eastern demand centers, highlighting the importance of coordinated grid planning and market reforms. Operational simulations demonstrate that high VRE shares are manageable with adequate storage and gas capacity. The results align with China’s ETS trajectory (current price ~$3–14.5/tCO2, expected ~$4–20/tCO2 by 2030) and suggest that moderate carbon prices coupled with declining VRE/storage costs can accelerate decarbonization. Policy levers—renewable targets, competitive auctions, and power-market reforms to reduce curtailment and improve interregional trading—can further enable the transition.

Rapid declines in renewable and storage costs enable China to reach substantially higher non-fossil electricity shares by 2030 at lower or modestly higher costs than BAU, depending on the stringency of CO2 caps. The study’s main contributions are the integration of updated cost trajectories into an hourly, provincial-scale optimization, evaluation of multiple carbon-constrained scenarios, and detailed regional transmission implications. By 2030, non-fossil generation could reach 62% at 11% lower cost (R) and 77% at 6% lower cost than BAU (C50); an 80% emissions cut (C80) is technically feasible with a 21% cost premium. Future research could examine deeper regional market integration, offshore wind and demand response cost trajectories, expanded flexibility options, social and equity impacts, and detailed implementation strategies for scaling storage and transmission.

Results depend on assumed cost trajectories for solar, wind, and storage, demand growth, and policy constraints. Large-scale deployment of storage (hundreds of GW) and transmission expansion faces practical constraints: supply chain and lifecycle management for batteries, financing and investment mobilization, siting/permitting, and potential stakeholder opposition. Operational modeling assumes available flexibility from gas and storage; real-world institutional barriers (e.g., interregional trading rules, curtailment) could impede uptake. Some technologies and responses (e.g., offshore wind cost declines, demand response) are not explicitly varied in core scenarios. Uncertainties in future costs and demand were partially explored via sensitivities but remain a source of risk to outcomes.