Evaluation of Cyclic Solvent Injection (‘Huff-n-Puff’) in Artificially-Fractured Shale Core Samples: Experiments & Modeling
Abstract
Multi-fractured horizontal wells (MFHWs) have enabled commercial production from low-permeability oil reservoirs but oil recoveries remain exceedingly small using the primary recovery scheme. As a result, operators have investigated the use of solvent (gas) injection schemes, such as huff-n-puff (HNP), to improve oil recovery. Laboratory experiments simulating the HNP process have been proposed to allow the investigation of recovery mechanisms, and for use in simulating HNP pilots; however, these experiments typically 1) fail to represent field conditions properly and 2) require long test times when performed on intact core plug samples. The primary objectives of this proof-of-concept study are to 1) design and implement a new experimental procedure that better reproduces HNP schemes in MFHWs and 2) use the results to explore the controls on enhanced hydrocarbon recovery in tight reservoirs.
A liquid/relative permeameter, previously developed for measuring single/two-phase gas (hydrocarbons/non-hydrocarbons) and liquid (oil, brine) flow in tight rocks, has been modified to perform core-based HNP experiments. The experimental procedure involves: 1) artificially fracturing a reservoir core plug sample under differential biaxial stress conditions to simulate an induced hydraulic fracture, 2) saturating the fractured core plug sample with oil, 3) measuring stress-dependent fracture permeability with oil under loading and unloading conditions, 4) implementing multiple HNP cycles (gas injection, soaking and production). For 4), oil and gas production is measured for each step, and oil/gas compositions are measured after selected steps. Compositional numerical simulation is used to 1) optimize the experimental design prior to the experiments and 2) history-match each cycle after the experiments, enabling fundamental controlling mechanisms to be explored.
A low-porosity (2.8%), low-permeability [slip-corrected (N2) gas permeability: 1.25·10-4 md; 900 psi effective stress] Duvernay shale core plug sample was chosen for analysis, after fracturing and saturation with de-waxed (dead) formation oil. Fractured core plug (oil) permeability exhibited significant hysteresis between loading and unloading cycles. Six HNP cycles were performed with CO2 using an injection pressure of about 1300 psi, and typical cycle lengths of 1 hour for injection and soaking, and 4 hours for production. The six HNP cycles were implemented in only 28 hours (2 cycles were abbreviated), with maximum oil recovery < 50%. As expected, the magnitude of incremental oil recovery decreased after the first 2 cycles. The data were successfully history-matched with a numerical simulator.
Previous HNP experimental approaches may be classified as either ‘flow-through-matrix’ (gas injection into an intact core plug sample) or ‘flow-around-matrix’ (gas flow around an intact core plug sample) – the former tests require exceedingly long test times and are usually performed on higher permeability samples, while the latter tests result in small test times but exceedingly high oil recoveries due to unrealistic experimental conditions. The experimental method proposed herein endeavors to reproduce field conditions by conveying the gas through a fracture within the core plug sample subjected to confining stress, and fracture area controlling gas exposure to the sample. The experimental conditions are therefore more realistic, yielding more meaningful/reasonable test results in a shorter time frame.