Broadband shock-associated noise (BBSAN) is a component of jet noise generated by supersonic jets operating off-design. It is characterized by multiple broad band peaks and dominates the total noise at large angles to the jet downstream axis. A new model is introduced for the prediction of broadband shock-associated noise that uses the solution of the Reynolds averaged Navier-Stokes equations. The noise model is an acoustic analogy based on the linearized Euler equations. The equivalent source terms depend on the product of the fluctuations associated with the jet's shock cell structure and the turbulent velocity fluctuations in the jet shear layer. The former are deterministic and are obtained from the RANS solution. A statistical model is introduced to describe the properties of the turbulence. Only the geometry and operating conditions of the nozzle need to be known to make noise predictions. This overcomes the limitations and empiricism present in previous BBSAN models. Results for various axisymmetric nozzles and operating conditions are compared with experimental data to validate the model.