Revisiting rule-based external carbon dosing for low-cost digital transformation of full-scale WWTPs

Digital transformation is reshaping municipal wastewater treatment, yet comprehensive digital upgrading of existing WWTPs remains constrained by the cost of additional online instrumentation and automation. Consequently, external carbon dosing in many full-scale plants still relies on plant-specific operational knowledge embedded in historical operating records. This study investigated whether historical operating records could be reconstructed into transparent, site-specific dosing references for low-cost digital decision support. Using 1078 retained operating days (2023-2025) from a full-scale cyclic activated sludge system (CASS) WWTP in South China, C/N-regime-specific compliance-informed conditional operating references were reconstructed as historical dosage medians within TN-compliant subsets. A site-specific rainfall safeguard based on NH₄⁺-N + NO₃^--N was calibrated using a separate 41-observation low-C/N dataset, yielding a retained threshold of 35.04 mg L^-1. Retrospective reference substitution estimated a net carbon-input difference of 192.87 t (21.5%; 895.02 vs. 702.15 t). Same-date paired comparisons, heteroskedasticity-autocorrelation-consistent (HAC) inference, moving-block bootstrap, and C/N-stratified analyses quantified the retrospective contrast. Leakage-controlled random forest screening achieved RMSE of 1.324 mg L^-1 and an R² of 0.576 on 108 unseen operating days. Under reference-substituted inputs, the maximum split-conformal one-sided 95% upper prediction bound was 13.946 mg L^-1. An exploratory 32-day pilot (160 CASS cycles) included nine triggered dosing events (366.12 kg COD), during which no measured TN exceeded 15 mg L^-1. External evaluation in an AAO plant confirmed that the reconstructed numerical references were process-specific. Overall, historical operating records can be transformed into transparent, auditable, and plant-specific digital operational knowledge without additional online instrumentation or hardware investment. The proposed framework provides a practical digital decision-support layer for retrospective assessment and prospective commissioning, offering a scalable and low-cost pathway for the incremental digital transformation of existing and aging WWTPs.