History Matching of Vertical Gas Injection Displacement Experiments
Master thesis
Permanent lenke
http://hdl.handle.net/11250/2559709Utgivelsesdato
2018Metadata
Vis full innførselSamlinger
Sammendrag
This master thesis presents simulations of two core flooding experiments. The two experimentsconsisted of injecting gas in a gravity stablemanner into a vertical core, afterthe core had beenwaterflooded. The main difference between the two experiments wasthe use of equilibrium gas in the first experiment, while dry separator gas was used inthe second. Using these different injection gases would expose the difference in the oilproduction due to mass transfer effects.The second experiment revealed that as oil was vaporized, water production fromthe core increased aswell. Other investigatorswere unsuccessful of simulating this phenomenonand thus obtaining a proper history match of the water phase. The problemwas identified as a problem with how the capillary pressure was managed throughoutthe simulations. Traditional three-phase models does not address the issue.To properly simulate the two experiments, the ODD3P three-phase model is used.Thismodel is utilizing saturation functions between all phases, also gas and water. Themodel combines hysteresis and miscibility to both capillary pressure and relative permeabilityin a continuous manner. Themodel also applies live scaling of capillary pressurewith respect to changing interfacial tension. By using the ODD3P model, it waspossible to obtain a history match of all three phases in both experiments. The increasedwater production due to vaporization of oil is accurately simulated by the useof a transition from water-oil capillary pressure into water-gas capillary pressure. Byapplying this transition at very low oil saturation,more water is allowed to flow.The oil and gas phases are also accurately simulated to history match experimentaldata. A surface equation of state is made to perform the flash calculations from reservoirto standard conditions. Injection of gas created an oil bank in both experiments.After the oil bank was produced a rather small flow of oil was seen in the first experiment.In the second experiment oil was produced in fairly high rate after the oil bankwas produced. Analysis of the oil flowmechanisms revealed that thiswas a results of vaporizationof oil. The vaporization effect proved to be significant. Approximately twiceas much oil were recovered by using a dry separator gas instead of an equilibrium gas.