A theory to model pseudo steady state water production and predict long-term water recovery from fractured reservoirs

The environmental and economic impact of hydraulic fracturing technology in the energy industry remains a growing concern due to large water consumption and poor water recovery in return. In this study, we propose a novel theory to describe and model the drive mechanisms (fracture closure and fluid expansion) responsible for long-term water recovery from hydraulically fractured reservoirs. In the literature, pseudo steady state (PSS) flow of water has been explained and modelled under the assumption of negligible fluid influx from matrix, referred to as "supercharge conditions". The preliminary flowback data analysis of 8 different unconventional reservoirs from North America in this study shows 1) water PSS flow signature is also observed after fluid breakthrough into fractures, and 2) it can be modelled mathematically, verified with long-term field data, and validated with numerical simulation results. These observations suggest that oil and water develop different networks or flow paths through fractures. With time, oil network expands and water network shrinks. Therefore, water continues to be produced from the fractures as oil saturation increases in the fractures with expansion of oil network due to fracture pressure decline. Based on this theory, an integrated semi-analytical workflow is developed by accounting for the changes in the volumes and compressibilities of both fracture and fluid. The workflow was applied on early production data (150-200 days) from six multi-fractured horizontal wells in Eagle Ford. The results show that the workflow can use the early-time data to accurately estimate long-term water recovery (325-475 days) from fractured reservoirs (within 6 % of measured field data). However, frequent long shut-ins during the course of long-term production increases the uncertainty in water recovery prediction.