In this paper, a novel integrated system that combines photocatalysis, photovoltaics, thermal engine and chemical energy storage for better solar energy harvesting is assessed using energy and exergy methods. The system generates hydrogen and sulfur from sulfurous waters specific to chemical and petrochemical industries. The solar light is split into three spectra using optical surfaces covered with selected dielectric coatings: (i) the high energy spectrum, consisting of photons with wavelengths shorter than, similar to 500 nm, is used to generate hydrogen from water photolysis, (ii) the middle spectrum with wavelengths between similar to 500 nm and similar to 800 nm is used to generate electricity with photovoltaic (PV) arrays and (iii) the long wave spectrum of low energy photons with wavelengths longer than,800 nm is used to generate electricity with a thermally driven Rankine engine (RE). The electricity generated by PV and RE is employed to generate additional hydrogen by electrolysis and to drive auxiliary devices within the system. A model is developed based on conservation equations and transport equations applied for each essential component of the system. The model allows for assessment of system performance and the comparison with other solar hydrogen production systems. A case study for an oil sands exploitation area where sulfurous aqueous wastes and hydrogen demand exist Calgary (Alberta) is presented. A solar tower configuration is selected as the best choice for a large scale system with 500 MW light harvesting heliostat field. Hourly predictions of system output are obtained. The devised system requires 5526 acres of land for the solar field and produces 41.4 t hydrogen per day. If a conventional solar tower would be used instead which generates power and is coupled to a water electrolysis system the hydrogen production is lower, namely 28.7 t/day. An economic scenario is considered by assuming that the co-produced sulfur and hydrogen are both valorized on the market for 25 years with a levelized price of 1.65 $/kg out of which 10% represents operation and maintenance costs. It is shown that the system is feasible provided that the required equity investment of capital is inferior to M$ 500. (C) 2014 Published by Elsevier Ltd.