Wall-bounded turbulent flow involves the developmentof multi-scale turbulent eddies, as well as a sharplyvarying boundary layer. Its theoretical descriptions are yetphenomenological. We present here a new framework calledstructural ensemble dynamics (SED), which aims at usingsystematically all relevant statistical properties of turbulentstructures for a quantitative description of ensemble means.A new set of closure equations based on the SED approachfor a turbulent channel flow is presented. SED order functionsare defined, and numerically determined from data ofdirect numerical simulations (DNS). Computational resultsshow that the new closure model reproduces accurately thesolution of the original Navier-Stokes simulation, includingthe mean velocity profile, the kinetic energy of the stream-wisevelocity component, and every term in the energy budgetequation. It is suggested that the SED-based studies of turbulentstructure builds a bridge between the studies of physicalmechanisms of turbulence and the development of accuratemodel equations for engineering predictions. Wall-bounded turbulent flow involves the development of multi-scale turbulent eddies, as well as a sharply warping boundary layer. Its present in a new framework called structural ensemble dynamics (SED), which aims at using systematically all relevant statistical properties of A new set of closure formulas based on the SED approach for a turbulent channel flow is presented. SED order functions defined by, and numerically determined from data of direct numerical simulations (DNS). Computational resultsshow that the new closure model reproduces accurately the solution of the original Navier-Stokes simulation, including the mean velocity profile, the kinetic energy of the stream-wisevelocity component, and every term in the energy budgeteation. It is suggested that the SED-based studies of turbulentstructure builds a bridge between the studies of physical mechanisms of tu rbulence and the development of accuratemodel equations for engineering predictions.