In radio access network planning and optimisation, the estimation and evaluation of cell coverage is crucial. For it, operators use several methodologies, ranging from theoretical propagation models for planning cell coverage, to drive test measurements for evaluating the performance within a cell. A novel propagation model is proposed which combines drive test measurements, cell reach statistics, antenna radiation patterns, terrain morphologies, and classical theoretical propagation models. Considering a realistic scenario of an LTE cell, the use of a classical propagation model such as Walfish- Ikegami results in an average error of -27 dB of estimations with respect to measurements, evidencing a pessimistic estimation of signal. By opposition, the proposed model fits with DT measurements.