The content of this report is property of HTceramix and subject to restrictive distribution The goal of the project "Oxygen membranes for the partial oxidation of natural gas" was the pre-industrial evaluation of the concept of an electroceramic reactor, capable of converting natural gas into synthesis gas (H2+CO) by partial oxidation with pure oxygen. This project was the continuation of the pre-project on the fundamental principle (FOGA 085) accomplished with success at EPFL and which was co-financed by FOGA (50'000 CHF). In the present second Phase, FOGA supported with 250'000 CHF the spin-off company HTceramix SA, which is specialised in the development and fabrication of electroceramics for an effective conversion of natural gas. HTceramix has managed the project in close collaboration with EPFL. The goal of this second Phase was to give proof of the feasibility of a future reactor. This ambitious project was divided into 5 stages defined by milestones, all clearly fulfilled. The first milestone gave the proof of the mechanical, thermo-mechanical and chemical stability of a material selected in the 1st Phase, stability required for an operational reactor. During the next stage, ceramic tubes of this material were fabricated in collaboration with external partners. Both compaction and extrusion were successfully applied for the product processing; extrusion was preferred because it lead to a thinner wall thickness, a better mechanical stability and a reduced raw material consumption. In parallel with the tube fabrication, a catalyst compatible with the ceramic membrane was developed for the promotion of partial oxidation (POX, stage 3). A noble-metal-free oxide satisfied the requirements of 90% conversion of methane and 98% CO-selectivity. For the proof of concept, the natural gas should be converted into synthesis gas by the combination of ceramic tubes and the POX-catalyst (stage 4). The tubular reactor operated over 1'400 hours with a methane conversion of 95% and a CO-selectivity of 92%. The oxygen permeation flux across the tubular membrane was enhanced by a factor 10 compared to Phase 1 project. In stage 5, the concept of a heat integrated tubular reactor was studied. It was calculated that 1 kWth natural gas can be converted in a 0.15 litre volume reactor (consisting of 12 tubes 20cm long and of 5mm diameter each). The required surface for integrated heat exchanger is 70 cm2/kW. All formulated goals were fulfilled, respecting both the financial and time limits.