Each microplastic fragment passes through three real treatment stages in series; the model computes a physically-grounded removal probability at each one from the particle's size and density and the plant's flow rate.
Stage 1 — Primary sedimentation (Stokes' law)
v_s = (2/9) · |ρp − ρw| · g · r² / μ
e1 = clamp( v_s / (Q / As) , 0, 1 )
Stage 2 — Secondary bio-flocculation
e2 = 1 − exp( −k · d · doseFactor )
(larger particles are entrained by activated-sludge flocs more readily)
Stage 3 — Tertiary filtration (pore sieving)
e3 = 1 / (1 + exp(−(d − Dpore)/s))
(a logistic size-cutoff around the filter's pore diameter)
- Particle diameter — sets the Stokes settling speed (∝ r²) and the filter sieving outcome.
- Plastic type — sets density ρp; density far from water's 1000 kg/m³ (PE/PP float, PET/PVC sink) settles or skims out fastest in Stage 1.
- Flow rate — sets the hydraulic retention time and surface-overflow rate; a faster flow gives particles less time to settle or be captured.
- Filter pore size — the Stage-3 cutoff diameter; particles well above it are sieved out almost completely, well below it pass through.
Each spawned sphere draws an independent pass/fail at every stage from these probabilities — over hundreds of particles the visible split converges on the same numbers as the readouts, the way a real treatment plant's removal efficiency is a population average, not a guarantee for any one fragment. Particles that survive all three stages exit through the effluent pipe into the receiving river, exactly the pathway that lets microplastics reach open water even from a functioning treatment plant.