In this article, a dynamic model describing the growth of the green microalgae Chlamydomonas reinhardtii, under light attenuation and sulphur-deprived conditions leading to hydrogen production in a photobioreactor is presented. The strong interactions between biological and physical phenomena require complex mathematical expressions with an important number of parameters. This article presents a global identification procedure in three steps using data from batch experiments. First, it includes the application of a sensitivity function analysis, which allows one to determine the parameters having the greatest influence on model outputs. Secondly, the most influential parameters were identified by using the classical least-squares cost function. This stage is applied to the experimental data collected from a lab-scale batch photobioreactor. Finally, the implementation of an Extended Kalman Filter estimating the biomass concentration, extracellular and intracellular sulphur concentrations is presented. Thereby, the observer uses on-line measurements provided by a mass spectrometer measuring the outlet gas composition (O-2, CO2). Software sensor performances and limits are illustrated in simulation and with experimental data.