An experimental procedure was developed to incorporate the mechanisms of Gas-Cycling Enhanced Oil Recovery (GCEOR) phase behavior, IFT change, swelling, viscosity reduction and residual-liquid shrinkage in the presence of porous media possessing propped hydraulic fractures and matrix. Relatively large hydrocarbon pore volumes are possible using this technique whereby effluent compositions, densities and volumes are measured. The importance of rock and fluid properties is investigated along with operating pressure, injection gas composition and levels of primary depletion. Experimental oil flux is measured and scaled to forecast field production rates based on field frac data.
The experimental design for core-flow testing has provided insight into GCEOR. Two dominant flow regimes are incorporated: matrix mass transfer into and from the fracture(s) and flow within the fracture(s). Reservoirs tested exhibited pressures from 3000 to over 5000 psi and temperatures from 140 to 220 F. Huff pressures were as high as 6000 psi. Over fortyseven separate primary depletions were accomplished followed by Huff and Puff (HnP) GCEOR on multiple rock and fluid types. Design parameters were changed from run to run allowing for insight into GCEOR operation and design. Simulation of experimental results with subsequent scale-up for field forecasting was performed.
From the many primary depletion tests followed by GCEOR, using a variety of injection gases and reservoir fluids the effects of cycling pressure, injection gas composition, soak time, level of primary depletion before GCEOR, and other parameters were investigated. Measured results indicate that recovery of OOIP can be more than doubled compared to primary production, in some cases, by implementing GCEOR. From the broad, accumulated data base, the following have been observed:
Cycling pressure should be optimized (highest pressure does not necessarily perform the best). Gas quality can, in some cases, play a major role but should be considered and quantified in GCEOR applications.
oak time/Huff time in many cases may be optimized to maximize production cycles and minimize injection cycles.
Gas utilization values, for well-designed GCEOR systems, are low compared to conventional continuous gas injection projects causing Huff and Puff GCEOR to approach gas storage performance. Gas utilization appears to be sensitive to the mechanisms at work in GCEOR.
Less depletion before GCEOR initiation can accelerate recovery and can, in some cases, access residual oil that was not produced at higher levels of primary depletion.
Contrary to expectation, the rock character may dominate GCEOR performance. In a subset of this testing, the rock heterogeneity had a more dominant role than fluid properties including miscibility.
Simulation can be an important tool to upscale the experimental measurements to field design. It is shown that simulation tuned to measured, experimental Primary Depletion and GCEOR performance including phase behavior and rock-fluid interaction, provides more optimistic forecasts of field GCEOR performance than a simulator based on field-scale Primary Depletion data alone.