AbstractsPhysics

Gas-injection enhanced oil recovery can recover nearly all residual oil where the gas sweeps. Sweep efficiency in these processes is often poor, in large part because of gravity override of gas. Stone and Jenkins presented a model for gravity override in homogeneous reservoirs, showing that the distance gas and water travel before segregation depends directly on injection rate. In cases where injection pressure is limiting, injectivity is key to overcoming gravity override. Stone assumed continuous injection of gas and water as a model for WAG (Water Alternating Gas), contending that this is valid as long as slugs mix near the well. This model for co-injection can be extended to relate segregation distance for co-injection processes directly to injection pressure. Injectivity depends on saturations very near the well, however. Therefore, where injection pressure is limiting, this model is pessimistic because injectivity in WAG is greater than in co-injection. We investigate the increase in injectivity possible with WAG compared to co-injection in 1D and 2D, and the implications for gravity override in 2D, using a range of models for gas and water relative permeabilities. We confirm that the greater injectivity of WAG improves vertical sweep compared to Stone's model when injection pressure is limiting. The greatest improvements occur when slugs violate Stone's assumption: that is, when they are too large to mix fully near the well. The increase in injectivity over co-injection is greater for foam than for WAG without foam, because foam has much lower mobility when gas and water flow together.