Currently, the power plants fueled by natural gas (NG) are responsible for 11 percent of the electric power generation in Brazil, consuming 40 percent of gas demand in the country. The gas consumption of power plants is seasonal and volatile, since it is directly related to the level of rainfall in locations where there are hydroelectric plants. Due to this peculiarity of the Brazilian energy matrix and the country s dependence on imported gas, there is a need to establish floating stocks of Liquefied Natural Gas (LNG) in LNG carriers. The main objective of this stock is that, if it is necessary to dispatch the generation of the power plants fueled by Natural Gas, the country has a sufficient stock of NG to allow the production of the thermoelectric generation during 30 days of operation. However, this model of contingency stock of Natural Gas is inefficient and has a very high cost. This dissertation demonstrates the technical and economic feasibility of replacing the floating LNG carriers stocks by the use of open underground caves by dissolving rock salt domes onshore, presenting a prediction of the annual stock size and the necessary quantity of caves to meet the growing need for NG stock over time. As a secondary objective, the dissertation presents an analysis of the use of these stocks in underground offshore environment, as a contingency lung for the maintenance of the pre-salt oil reservoirs production in the event of an occurrence that causes the interruption of the regular flow of the runoff from the gas produced in the platforms.