A low-field NMR framework for monitoring solid-liquid separation of digestate

Mechanical dewatering of anaerobic digestate is commonly limited by variable feed composition, empirical flocculant dosing, and the lack of rapid indicators linking flocculated-feed structure to downstream cake formation. This study develops a low-field nuclear magnetic resonance (LF-NMR) relaxometry framework for monitoring solid-liquid separation of digestate performance. Industrial anaerobic digestate from a red meat processing facility was conditioned with a polymeric flocculant (0–5.25 wt%) and characterised using one-dimensional T 2 relaxation, two-dimensional T 1 - T 2 and T 2 - T 2 relaxation measurements, and multivariate exploratory data analysis (MEDA). Increasing flocculant dosage produced a significant redistribution of water populations in the flocculated feed ( T 2 from 1.15 s to 1.70 s; signal fraction from 0.16 to 0.54, comparing ≤3.0 wt% and ≥3.75 wt% groups), coinciding with a significant increase in recovered cake mass after mechanical pressing (71.5 g vs. 253.2 g). After dewatering, the dominant residual-water peak remained constant from 0–3.75 wt% before decreasing sharply at ≥ 4.5 wt%, indicating a transition from heterogeneous pore collapse to homogeneous pore drainage within a mechanically stabilised floc network. A dual-parameter NMR metric, T 2,floc , defined by T 2 > 1 s and signal fraction > 0.4 in the flocculated feed, distinguished samples that subsequently formed mechanically stable, well-drained cakes. Two-dimensional relaxometry corroborated the underlying water-population assignments via T 1 / T 2 ratio trends and T 2 - T 2 exchange behaviour over a ≈ 35 μm length scale, and MEDA confirmed that the predictive descriptors are measurable in the flocculated feed prior to pressing. The underlying dual-parameter approach, combining a mobility descriptor and an abundance descriptor, provides a transferable calibration framework for digestate dewatering process monitoring.