Abstract GDF SUEZ, as a gas and LNG company, operates onshore and offshore facilities where accidental high pressure releases of natural gas are likely to occur. To study this hazard, experiments have been performed in uncongested areas, focusing on fires and heat flux. In parallel, trials were carried out to assess overpressures generated by a gas explosion in a congested area. Most experiments were done with a quiescent stoichiometric gas cloud and not with a real turbulent release of methane or natural gas. The same assumption of stoichiometric gas cloud is then followed for quantitative risk assessment and influence land use planning, occupational safety, safety gaps definition and design. The following review endeavours to collect existing knowledge and recent experimental and numerical works on the influence of initial turbulence on methane explosions severity and to highlight the differences between explosions with initial turbulence and explosions with quiescent stoichiometric mixtures. Small-scale experiments of methane explosions carried out by INERIS and GDF SUEZ in 2011-2013 are presented. These are 0.8 kg/s methane jets dispersion and explosion tests, in open field and within various congested areas. According to those tests, maximum overpressure is multiplied by 5 for turbulent jets ignitions compared to tests with quiescent stoichiometric mixtures of same volume. The influence of gas jet turbulence on explosion is confirmed by FLACS simulations performed by GDF SUEZ for large scale configurations. In addition, sensitivity studies on FLACS simulations show uncertainties on the CFD modelling of gas jets explosions. In order to further increase knowledge and to validate models and CFD codes, as FLACS, GDF SUEZ Research Center is currently setting up a JIP for a campaign to pursue large-scale explosion tests with pressurized natural gas releases into both congested and uncongested areas.