A net-zero emissions strategy for China's power sector using carbon-capture utilization and storage

The study addresses how China’s future power system can achieve deep decarbonization while ensuring high reliability and resilience. With wind and solar expected to dominate future generation but exhibiting intermittency and weather sensitivity, the research investigates whether integrating abated fossil fuel generation using carbon capture, utilization, and storage (CCUS) can complement high shares of renewables to reduce storage and transmission needs, lower costs, and improve reliability and resilience. The work is motivated by China’s 2060 carbon neutrality pledge, the uneven spatial distribution of renewables versus load centers, and China’s significant geological CO2 storage potential. The central research questions are: (1) What combinations of storage duration, interprovincial transmission capacity, and shares of CCUS-enabled fossil generation produce reliable, low-cost near-zero systems by 2050? (2) How does including abated fossil generation affect system resilience to extreme weather? (3) Under what conditions can the power system reach net-zero emissions by pairing CCUS with biomass co-firing?

Prior studies have separately emphasized the feasibility of 100% (or near-100%) renewable systems and the economic role of CCUS as firm low-carbon generation. Research highlights the need for long-duration storage and transmission to achieve reliable high-renewable systems, and integrated assessment models and bottom-up optimization studies often include CCUS for cost-effective deep decarbonization. However, a consistent comparison between a fully renewable system and a high-renewable system complemented by CCUS within the same modeling framework has been rare. Earlier power system models commonly lacked detailed, high-resolution CCUS facility and geological constraints and often aggregated time into limited representative periods, potentially biasing results. This study fills these gaps by integrating hourly resolution across an entire year, provincial-level modeling, and explicit source–sink CCUS matching constrained by storage potential, injection capacity, and transport distance.

The authors develop a high-resolution integrated power system assessment and simulation framework for China’s 31 provinces in 2050 comprising six modules: (1) Hourly nonfossil generation potential estimation: Downscales onshore wind, offshore wind, and solar PV using historical hourly climate data; hydropower and nuclear potentials are adjusted by month. (2) Hourly electricity demand projection: An econometric fixed-effects model projects provincial annual per-capita electricity consumption to 2050 using GDP per capita, heating/cooling degree days, industrial structure, and electricity price index, then downscales to hourly using 2019 hourly load shapes (with Anhui as reference) and monthly workday/non-workday profiles. The selected scenario yields national demand of 14.53 PWh in 2050. (3) CCUS source–sink optimal matching: A multiobjective optimization matches candidate coal- and gas-fired plants to onshore/offshore storage sites constrained by geological storage potential, injection rates, transport distance ≤500 km, unit size ≥300 MW, and remaining life ≥15 years. 718 of 944 coal plants (907.8 GW) and 58 of 165 gas plants (53.6 GW) are selected, matched to 5,471 storage sites (4,926 deep saline aquifers; 545 oil fields for EOR). (4) Optimal near-zero power system simulation: Hourly supply–demand balancing across provinces with four supply modes (local real-time generation, dispatched generation via transmission, local storage discharge, and dispatched storage discharge). Interprovincial transmission includes 85 routes (50 current + 35 new). Scenarios vary short-term storage duration (0–24 h), transmission capacity (1–10× reference, 0.5 increments), and CCUS-enabled fossil share (0–20% in 1% steps), plus a zero-fossil system with long-term hydrogen storage, totaling 10,450 scenarios. Priority rules reflect grid preferences: nuclear > hydro > abated fossil > variable renewables; short-term storage discharges before long-term. Supply is dispatched to provinces with the largest unmet demand; storage charges from residual VRE potential. (5) Cost-competitiveness analysis: Computes total system cost as the sum of levelized costs for each generation type, storage (short-term and hydrogen), and transmission CAPEX (route length × unit cost × capacity), then derives system LCOE. The optimal configuration satisfies 99.9% reliability (≤0.1% national shortage rate). (6) Extreme weather resilience assessment: Simulates historical events—2008 snowstorms (7 provinces), 2021 sandstorms (8 provinces), 2022 droughts (6 provinces), and 2022 heat waves (14 provinces)—by adjusting hourly climate-dependent generation and loads, comparing zero-fossil and high-renewable+CCUS systems. Reliability and resilience metrics are based on power shortage rates at national/provincial/hourly levels. Emissions accounting includes only direct stack emissions from fossil units with 90% capture; indirect lifecycle emissions from nonfossil sources are excluded. To achieve net-zero, an optimal biomass co-firing matching model links CCUS-ready coal plants to nearby biomass (≤50 km collection radius; max 40% cofiring by heat), finding average 13% cofiring sufficient for net-zero under the same boundary.

- Zero-fossil baseline: With current interprovincial transmission and no storage, national total power shortage rate in 2050 could be 28.1% (21.8% under an enhanced reference with 35 added UHV lines). Even at 5× reference transmission and 12 h storage, the shortage rate only falls to 0.07%, with high associated costs (levelized supply cost 47.24 USD/MWh; transmission cost ~$64.7B; storage cost ~$71.0B). - Reliability benefits of CCUS-enabled abated fossil generation: At reference transmission and no storage, allowing 5%, 10%, and 20% CCUS shares reduces national shortage rates from 21.8% (0% CCUS) to 19.4%, 16.9%, and 12.8%, respectively (improvements of 2.4, 4.9, and 9.0 percentage points relative to the 21.8% figure noted in the text). With 5× transmission and 12 h storage, shortage rates drop to 0.06%, 0.03%, and 0.03% for 5%, 10%, and 20% CCUS shares, respectively (versus 0.07% with zero-fossil). To approximate the zero-fossil best reliability (0.07% shortage at 5× transmission, 12 h storage), required infrastructure is reduced to: 4.5× and 10 h (5% CCUS), 3.5× and 8 h (10% CCUS), and 3.5× and 7 h (20% CCUS), reducing system LCOE by 1.1%, 2.5%, and 2.8%, respectively. Estimated infrastructure savings for storage/transmission: ~$9.5–$36.7B; avoiding stranded assets from CCUS retrofits: ~$4.2–$16.8B. - Optimal 2050 power mix (min-cost at 99.9% reliability without long-term storage): A system with 16% abated fossil generation (14.9% coal+CCUS; 1.1% gas+CCUS), 30.0% wind (on+offshore), 30.5% PV, 12.6% hydro, 11.1% nuclear, and 8.6% storage-backed generation achieves the lowest cost of ~$662B and LCOE 45.64 USD/MWh, which is 2.5% lower than a zero-fossil system ($679.2B; 46.78 USD/MWh). This optimal case uses 3× transmission and 8 h storage. CCUS plants are concentrated in Jiangsu, Henan, and Hebei, with typical source–sink distances between ~10 and 147 km; 13 coastal plants connect to offshore storage. - Long-term storage comparison: A zero-fossil system with long-term hydrogen storage meeting 0.1% shortage requires 6 h short-term storage and 4× transmission, costing ~$820B (LCOE 47.15 USD/MWh), which is 19.3% costlier than the high-renewable + 16% CCUS case. Under stricter reliability (e.g., 0.01% shortage), long-term storage can outperform CCUS on LCOE (e.g., 47.19 vs 48.21 USD/MWh at 0.01%). - Resilience to extreme events (high-renewable + 16% CCUS vs zero-fossil): • Snowstorms: Hours with >10% shortage reduce from 396 to 101; max hourly shortage from 44% to 16%; regional shortages cut by 54% (−22.9 TWh). • Sandstorms: Hours with >10% shortage reduced by 5 h; regional shortages cut by 56% (0.9 to 0.4 TWh); onset of severe shortages delayed by ~7 h. • Droughts: Hours with >10% shortage reduced by 13 h; regional shortages cut by 57% (2.1 to 0.9 TWh). • Heat waves: Least impact on both systems. - Path to net-zero: Given 90% capture at fossil units, net-zero can be reached by cofiring biomass with CCUS-ready plants. Of 196 candidate CCUS fossil plants, 166 can source biomass within 50 km; an average 13% biomass cofiring ratio achieves net-zero under the study’s accounting boundary, with low additional cost for cofiring ratios <20%.

The results demonstrate that a high-renewable power system complemented by CCUS-enabled abated fossil generation can achieve the required 99.9% reliability at lower overall system cost and with reduced dependence on very high interprovincial transmission capacities and extensive short-term storage compared with a zero-fossil system. CCUS provides firm, dispatchable low-carbon capacity that mitigates renewable intermittency and spatial mismatch between resource-rich and load-centered provinces. This firm capacity also enhances resilience during extreme weather, substantially cutting the severity and duration of shortages during snowstorms, sandstorms, and droughts. While a zero-fossil system augmented with long-term storage can meet even stricter reliability standards at competitive LCOE, for the 0.1% shortage criterion the combined renewables+CCUS strategy is more economical. Regional analysis indicates strong export roles for Northwest and Southwest China, and high import dependence for Beijing–Tianjin and East Coast regions, implying differentiated strategies: expand long-distance transmission where external dependence is high; deploy storage and backup in exporter regions that rely heavily on variable renewables and are more vulnerable to extreme events. Finally, to reach net-zero under direct emissions accounting, modest biomass cofiring at CCUS-ready fossil plants can offset residual emissions, with infrastructure and resource matching indicating practical feasibility.

This study provides an integrated, hourly, province-level evaluation of China’s 2050 near-zero power system pathways, jointly optimizing renewables, storage, transmission, and geologically constrained CCUS retrofits. It finds that including up to a 20% share of abated fossil generation markedly improves reliability and resilience while reducing the need for extreme storage and transmission builds. The lowest-cost 2050 system achieves 99.9% reliability with 16% CCUS-enabled fossil generation, 3× transmission, and 8 h storage, at an LCOE of 45.64 USD/MWh—2.5% below a zero-fossil alternative. For stricter reliability requirements, long-term storage becomes increasingly important and can outperform CCUS on LCOE at very low shortage thresholds. A net-zero emissions power sector is achievable by adding modest biomass cofiring (average 13%) to CCUS-equipped fossil plants. Future research should refine intra-provincial network modeling, incorporate operational constraints and ancillary services, expand uncertainty analysis of technology costs and performance, assess lifecycle emissions, and evaluate policy designs (e.g., incentives for CCUS and biomass cofiring) that align investment with reliability and resilience goals.

- Network representation: Provinces are modeled as single nodes with only interprovincial transmission; intra-provincial grid constraints and congestion are not represented. - Operational simplifications: Priority dispatch rules and simplified storage charge/discharge logic are used; ramping constraints, minimum up/down times, reserves, and ancillary services are not explicitly modeled. - Storage modeling: Short-term storage is capped by duration but excludes round-trip efficiency details in the summary; long-term storage is represented via hydrogen with assumed costs and efficiencies; technology readiness and deployment constraints are simplified. - CCUS constraints and performance: Assumes 90% capture; site characterization, permitting, and injection dynamics are abstracted via capacity and distance constraints; pipeline network optimization beyond simple distance constraints is not modeled. - Emissions boundary: Only direct stack emissions are counted; lifecycle and upstream emissions for all technologies are excluded. - Demand modeling: Hourly 2050 load shapes rely on 2019 profiles (Anhui-based scaling) and econometric projections; demand response and electrification patterns may differ by 2050. - Cost uncertainty: Future LCOEs, transmission costs, and storage costs are scenario-based and subject to uncertainty; sensitivity analyses are summarized but not exhaustive. - Extreme events: Analyses are based on selected historical events; compound risks, infrastructure damage states, and restoration logistics are simplified.