A Multi-Dimensional and Multi-Level Based Collaborative Evaluation of Carbon Reduction in the Power Sector

Abstract

Achieving deep carbon reduction in the power sector is central to China's green energy transition. This study develops a novel multidimensional and multi-level collaborative evaluation framework designed specifically to assess carbon reduction pathways in electricity systems. Methodologically, we begin by forecasting power demand and the evolution of the generation structure—including thermal, wind, hydro, and photovoltaic sources—based on multi-source power big data, which provides a high-resolution foundation for model parameterisation and calibration. An improved Analytic Hierarchy Process-Entropy Weight Method (AHP-EWM) integrated with a Particle Swarm Optimization (PSO) algorithm is proposed, where the availability of big data facilitates a more robust training process for the PSO, leading to scientifically robust weights for these indices. The model subsequently evaluates the synergistic efficiency of carbon reduction strategies, incorporating constraints on both carbon and air pollutant emissions to quantify co-benefits. Applied to a representative Chinese province from 2030 to 2050, the model demonstrates that under the AHP-EWM-PSO optimized scenario, carbon reduction efficiency is maximized. By 2050, carbon dioxide emissions are projected to decrease significantly, accompanied by substantial reductions in air pollutants, confirming the strong synergistic outcomes of the proposed pathway. The coupling degree of synergy approaches 1, indicating optimal alignment between carbon mitigation and auxiliary environmental benefits. This study provides a scientifically-grounded and practical tool for formulating and monitoring carbon reduction strategies, supporting policy-making aimed at accelerating the transition to a high-renewable power system with significant environmental co-benefits.