During hydraulic-fracturing operations in low-permeability formations, spontaneous imbibition of fracturing fluid into the rock matrix is believed to have a significant impact on the retention of water-based fracturing fluids in the neighborhood of the induced fracture. This may affect the post-fracturing productivity of the well. However, there is lack of direct experimental and visual evidence of the extent of fluid retention, evolution of the resulting imbibing-fluid front, and how they relate to potential productivity hindrance. In this paper, laboratory experiments have been carefully designed to represent the vicinity of a hydraulic fracture. The evolution of fracturing fluid leakoff is monitored as a function of space and time by use of X-ray computed tomography (CT). The X-ray CT imaging technique allows us to map saturations at controlled time intervals to monitor the migration of fracturing fluid into the reservoir formation. It is generally expected for low-permeability formations (5 to 10 md) to show strong capillary forces because of their small characteristic pore radii, but this driving mechanism is in competition with the low permeability and spatial heterogeneities found in low-permeability sands. The relevance of capillarity as a driver of fluid migration and retention in a low-permeability sand sample is interpreted visually and quantified and compared with high-permeability Berea sandstone in our experiments. It is seen that although low-permeability sands are subject to strong capillary forces, the effect can be suppressed by the low permeability of the formation and the heterogeneous nature of the sample. Nevertheless, saturation values attained as a result of spontaneous imbibition are comparable with those obtained for high-permeability samples. Leakoff of fracturing fluids during the shut-in period of a well can result in delayed gas flowback and can hinder gas production. Results from this investigation are expected to provide fundamental insight regarding critical variables affecting the retention and migration of water-based fracturing fluids in the neighborhood of hydraulic fractures, and consequently affecting the post-fracturing productivity of the well.

This paper presents an innovative integrated methodology and working procedure for characterizing and simulating the strong non-linearity and non-stationariness caused by changes in confined pressure-volume temperature (PVT) properties over time related to pore-throat size, the pressure-dependent permeability, and the intervened multiple porous media created by multi-stage fracture stimulation. The complicated physics behind the observed phenomena are explored. More specifically, this paper demonstrates and discusses the following: 1) a new rate-transient analysis (RTA) procedure to infer the stimulated reservoir volume (SRV) and fracture parameters; 2) the impact of the non-stationary feature, compaction effect, and pore-throat related PVT properties on the flow regime and well performance; 3) how to incorporate the nonstationary and non-linear features into the reservoir model; 4) the integrated procedure for history matching, performance forecast, and recovery assessment; 5) several field examples in the Bakken to illustrate the procedure.The proposed procedure has been successfully applied for the following: 1) constructing the non-stationary, nonequilibrium, and highly non-linear simulation models; 2) facilitating the history matching by addressing permeability reduction and PVT property variations caused by compaction and capillary pressure; 3) and ensuring more reliable performance forecasts and recovery assessments.The study shows that the reduction of the bubblepoint pressure could be several hundred psi in the typical Bakken rock; moreover, such reduction continues following the depletion via the compaction effect. The compaction effect could impair the matrix permeability by up to one order of magnitude.The study reveals the following: 1) the confined PVT properties could widen the favored operation window, whereas the compaction effect could significantly impair the ultimate recovery of the wells; 2) the RTA-inferred SRV-related parameters are the key input for capturing the non-stationary features; 3) the impact on recovery could be over 50% without addressing the aforementioned non-stationary and non-linear issues.This paper explores several unique phenomena in unconventional oil reservoirs which have not previously been published. The proposed analysis and assessment procedure greatly enhances the understanding of the unconventional assets and we feel will improve the accuracy of long-term rate and recovery forecasts.