How long will perfluoroalkyl substances persist in sludge-amended soils?: Century-scale regional accumulation-decay modeling for North America, Europe, and Asia

초록

Perfluoroalkyl substances (PFAS) continue to be present in sludge and sludge-amended soil, despite decades of bans and phase-outs, yet their long-term persistence and regional disparities remain poorly understood. This study examined the occurrence of 14 PFAS in sewage sludge and market-balance production from 2000 to 2023 across North America, Europe, and Asia. Regional accumulation–decay models were developed to evaluate PFAS trends in sludge-amended soils and project long-term concentrations through 2300. Long-chain PFOS, PFOA, PFDA, and PFNA account for >80% of total PFAS in sludge, reflecting their high production and hydrophobicity. Repeated applications led to increased cumulative PFAS, reaching 20 ng g⁻¹ in Europe, 14 ng g⁻¹ in North America, and 1.6 ng g⁻¹ in Asia until 2023. Even if land-spreading ceased after 2023, only 60–70% of this burden would dissipate by 2300, leaving long-chain congeners dominant. PFOS is identified as a significant long-term risk, with North American soils predicted to exceed the GW/SW protection level (0.03 ng g⁻¹) until the year 4400, followed by Europe at 4297 CE and Asia at 3500 CE. PFDA, showing the longest persistence, remains above its 0.92 ng g⁻¹ criterion in European soils until 4100 CE, while PFNA exceeds its 0.042 ng g⁻¹ criterion until 3300 CE in Europe, followed by North America and Asia. The first compliance years are influenced by long-chain PFAS and the stringent of screening values. Findings emphasize natural attenuation alone is insufficient for risk reduction, especially for long-chain PFAS, underscoring the need for class-based regulation, better sludge treatment, and targeted remediation for contaminated farmland. Capsule Abstract: Past and future concentrations of PFAS in sludge and sludge-amended soils are predicted for North America, Europe, and Asia. © 2026 Elsevier B.V. All rights are reserved, including those for text and data mining, AI training, and similar technologies.

키워드