Development, test, and evaluation of exploitation technologies for the application of gas production from natural gas hydrate reservoirs and their potential application in the Danube Delta, Black Sea

One important scientific objective of the national research project SUGAR – Submarine Gas Hydrate Reservoirs was the development, improvement, and test of innovative concepts for the production of methane from natural gas hydrate reservoirs. Therefore, different production methods, such as the thermal stimulation using in situ combustion, the chemical stimulation via injection of CO2 as a gaseous, liquid or supercritical phase and depressurization were tested alone or in combination at different scales. In the laboratory experiments these ranged from pore and hydrate grain scale to 425-L reactor volume, whereas numerical models were applied to describe the related processes from molecular to reservoir scale. In addition, the numerical simulations also evaluated the feasibility and efficiency of the application of these methods in selected areas, such as the Danube Paleodelta in the Black Sea, addressing the two dominant methane hydrate reservoir settings, buried channel-levee and turbidite systems. It turned out, that the injection of CO2 or a CO2–N2 gas mixture is not applicable for the Danube Paleodelta in the Black Sea, because the local pressure and temperature conditions are too close to the equilibrium conditions of both, the CO2 hydrate and a CO2–N2 mixed hydrate stability fields. Experiments using thermal stimulation and depressurization showed promising results but also some issues, such as sufficient heat transfer. In summary it can be said that the applicability and efficiency of each method has to be proven for each specific hydrate reservoir conditions. Based on the results obtained by numerical simulations the most promising and safe method for the production of CH4 from hydrate bearing sediments in the Danube Paleodelta would be the depressurization technique. This study summarizes the main experimental and modeling results.