S. Hu, H. Qi, Z. Wang et al.
Environmental Science and Ecotechnology 30 (2026) 100682
Table 3 Comparison between ESG-reported and surveyed carbon emissions for representative PPPs. Plant type Surveyed mean values (t CO 2 per t paper)
ESG-reported mean values (t CO 2 per t paper)
PFPP RFPP
1.09 0.71 0.52 0.93
1.03 0.78 0.58 0.92
HPPMP LPPMP
SPPMP 0.8 Note: PFPP, primary fiber pulp plant; RFPP, recovered fiber pulp plant; HPPMP, heavyweight paper product manufacturing plant; SPPMP, specialty paper product manufacturing plant; LPPMP, lightweight paper product manufacturing plant; ESG, environmental, social, and governance. 0.81
plants, in which thermal power plants often occupy a large pro- portion of the total site area and therefore exert a stronger influ- ence on overall carbon emissions. In contrast, for larger plants, such as PFPPs and RFPPs, although thermal power plants remained an important source of direct carbon emissions, their spatial footprints typically accounted for smaller proportions of the total site areas. Consequently, their contributions to carbon emissions were small within the area-based model. Overall, these results highlight pronounced heterogeneity across functional zones shaping plant-level carbon emissions across PPP categories.
Table 4 Carbon emission accounting model evaluation indicators. Plant type R 2
MAPE
PFPP RFPP
0.90 0.96 0.88 0.75
9.95% 8.10%
LPPMP SPPMP
12.68% 19.21%
the predicted emissions agreed well with the observed values for most of the samples. Thus, we subsequently applied the confirmed models to conduct plant-level carbon accounting for 720 classified PPPs. In future work, we plan to further improve model perfor- mance by using more balanced training samples and advanced feature engineering to better capture this variability. Using the confirmed models, we analyzed the contributions of different functional zones to plant-level carbon emissions (Table 5). Across all PPP categories, wastewater treatment areas consistently contributed to higher carbon emissions. This pattern showed that increases in production capacity were accompanied by substantial growth in wastewater generation, necessitating larger wastewater treatment systems and, consequently, gener- ating higher carbon emissions. This relationship was particularly pronounced for PFPPs, reflecting their high water consumption and pollution intensity during pulping, which, in turn, led to greater reliance on wastewater treatment infrastructure. Beyond wastewater treatment areas, the influence of different functional zones on carbon emissions varied markedly across plant types. The thermal power plant areas, for instance, were positively associated with carbon emissions only for some plant categories, with the effect being particularly pronounced for SPPMPs. This pattern could be attributed to the relatively small scale of such Table 5 Driving effects of different functional area sizes on plant carbon emissions. Plant type a b c d e f ε PFPP − 0.54 0.47 9.70 14.93 − 0.36 − 0.58 50.40 RFPP - 0.24 0.00 1.86 0.00 1.75 − 13.93 LPPMP - - 0.20 5.53 0.00 0.61 7.91 SPPMP - - − 3.76 7.61 14.740 0.83 − 0.41 HPPMP - - 0.24 1.86 0.00 1.74 − 1.74 HPPMP 19.49% Note: R 2 , coefficient of determination with values closer to 1 indicates better model performance; MAPE , mean absolute percentage error; PFPP, primary fiber pulp plant; RFPP, recovered fiber pulp plant; HPPMP, heavyweight paper product manufacturing plant; SPPMP, specialty paper product manufacturing plant; LPPMP, lightweight paper product manufacturing plant. 0.76 Note: a , b , c , d , e , and f denote the regression coefficients corresponding to the primary fiber stacking area, recovered fiber stacking area, other stacking areas, wastewater treatment area, thermal power plant area, and other built-up areas, respectively, indicating their contributions to overall carbon emissions. ε denotes the error term. PFPP, primary fiber pulp plant; RFPP, recovered fiber pulp plant; HPPMP, heavyweight paper product manufacturing plant; SPPMP, specialty paper product manufacturing plant; LPPMP, lightweight paper product manufacturing plant.
3.3. PPP carbon emission inventory
In 2022, total carbon emissions from China’s PPI were approxi- mately 163.6 million tons of CO 2 , with a heterogeneous spatial distribution across 720 PPPs (Supplementary Fig. S4). Overall, 720 PPPs were distributed across 28 Chinese provinces and municipal- ities, with pronounced regional disparities in carbon emissions generated by the PPI (Fig. 6a). From a regional perspective, China’s PPPs are mainly concentrated in coastal provinces. The 10 coastal regions (Guangdong, Shandong, Jiangsu, Zhejiang, Hainan, Fujian, Shanghai, Guangxi, Liaoning, and Tianjin) collectively contributed 102.3 million tons of CO 2 to PPI carbon emissions, accounting for 62.3% of the national total (Fig. 6b). Only Guangdong and Shandong reported carbon emissions exceeding 20 million tons of CO 2 (Fig. 6a). Combined, they generated more than one-quarter of China’s total PPI carbon emissions. In contrast, most inland prov- inces reported emissions below 5 million tons of CO 2 . Total emissions from PFPPs, RFPPs, and HPPMPs amounted to 80.4, 30.9, and 37.4 million tons of CO 2 , respectively, accounting for 90.4% of national PPI emissions (Fig. 6c). Carbon emissions from PFPPs and RFPPs displayed strongly right-skewed distributions (Fig. 7), with maximum values far exceeding those for the other categories. This indicates the presence of extremely high-emission individual plants, which significantly elevates the overall carbon emission levels of the sector. In contrast, carbon emissions from LPPMPs and SPPMPs exhibited more concentrated distributions, with lower mean values and reduced extremes. This suggests relatively smaller disparities in carbon emissions among enter- prises within these plant categories. There was also considerable variation in carbon emissions among plants that produced similar products. Taking PFPPs as an example, individual plant emissions ranged from 43,600 to 450,700 tons of CO 2 , representing a difference of more than two orders of magnitude. This pronounced skewness indicates that a small number of large, high-emission plants contribute dispro- portionately to total sectoral emissions, resulting in a strongly nonlinear emissions structure. For instance, the Jinhai Pulping and Papermaking Plant in Hainan emitted 450,700 tons of CO 2 — approximately five times the PFPP category average (87,350 tons of CO 2 ) — and accounted for 5.6% of total PFPP emissions. Further statistical analysis revealed that approximately 5% of high-emission PPPs accounted for approximately 43% of total PPI
8
Made with FlippingBook interactive PDF creator