Voltage-induced closure of bacterial \(\beta\)-barrel channels is observed as stepwise attenuation of single-channel current, yet the physical origin of the residual low-conductance levels remains uncertain. Concentration-dependent recordings of the trimeric porin OmpF show three linked features: non-monotonic first-monomer closure across electrolyte concentration, measurable current after all three monomers occupy low-conductance configurations, and selectivity reversal from cation preference in high-conductance states to anion preference in the fully low-conductance state. These observations require analysis of complete current trajectories rather than averages of first-closure dwell time alone. This study develops a continuous-time state-space model for single-channel OmpF records with observable conductance levels \(L_{0}\), \(L_{1}\), \(L_{2}\), and \(L_{3}\), while allowing \(L_{3}\) to contain multiple latent substates. The independent-monomer reference model imposes the rate constraint \(q_{01}:q_{12}:q_{23}=3:2:1\); deviations from this ratio provide a direct estimate of inter-monomer coupling. Robust current-emission and joint conductance–selectivity components distinguish instrumental noise from reproducible low-state heterogeneity. Transition intensities are linked to electrolyte concentration, voltage polarity, and cation identity, separating the concentration dependence of first closure from the conditional kinetics of later closures. The resulting analysis converts electrochemical gating into explicit statistical tests for independence, cooperativity, hidden substates, censoring, filtering sensitivity, and state recovery. The same formulation is applicable to OmpF, VDAC, OmpG, FhuA, OccK, and biomimetic nanopores in which functional closure may occur without complete steric occlusion.