Decreasing fracture effectiveness due to conductivity decay is a strong contributor to the steep production decline commonly observed in shale plays. The conductivity of a fracture is determined experimentally by measuring the pressure drop of a fluid flowing through a uniformly distributed proppant bed in a core with fixed length and height. Fracture conductivity degradation results from damage mechanisms and fluid interactions that occur during hydraulic fracturing operations. Rock softening and proppant embedment are some of these damage mechanisms. The impact of these interactions can be observed by measuring fracture conductivity in the laboratory under stress states similar to field conditions. This study is based on experiments performed on fractured and propped Niobrara core plugs. The samples were characterized using X-ray Diffraction (XRD), and X-ray Fluorescence (XRF), and helical CT-scans. The experiments were performed on a triaxial stress test assembly to monitor the chemical and mechanical alterations in the formation, proppant, and fluid under reservoir conditions. To achieve this, fluid chemical composition, dynamic and static moduli, and conductivity were obtained. The setup was used for the simultaneous acquisition of stress, ultrasonic compressional and shear wave velocities, flow data and fluid sampling. The results from this study indicate that stress-dependent, long-term fracture conductivity shows the sharpest decline in the early stages of the experiment. The associated fluid sample analysis indicates that the highest physicochemical dissolution of most of the elements is happening at the early contact of the fluid with the rock and is later enhanced by the pressure increase in the system. A comparison with the conductivity measurements performed on Vaca Muerta samples shows a similar behavior, yet a steeper initial decay than that observed in the Niobrara samples. The difference observed between the two samples is related to the mineralogy of the formation and the high proppant embedment observed in the Vaca Muerta samples. Although higher softening occurred in the Niobrara samples, larger embedment was observed in the Vaca Muerta sample. This experimental observation is an indication that the conductivity damage varies not only with the mineralogical content of the formation, but also with the distribution of minerals along the fracture face. Geomechanical, geochemical, and flow data integration provided a better understanding of proppant embedment and mineral distribution of the rock. It is the conclusion of this study that even if the intact core sample contains an average mineralogical composition, the heterogeneity caused by variations in the mineralogy at where the fracture is induced has the biggest impact on embedment.

A water hammer signal is generated at the end of a hydraulic fracturing treatment from the sudden change in fluid velocity in the system. Analysis of water hammer data has been proposed to be a simple, inexpensive diagnostic technique to assess the hydraulic connection between the wellbore, fracture network, and reservoir. Previous studies have reached differing conclusions, suggesting that the water hammer signal can be dominated either by the hydraulic fracture dimensions or by wellbore effects. This study evaluates the utility of water hammer diagnostics in a large dataset to determine correlations between the observed water hammer signal and treatment size, fracture fluid type, completion type, and resulting well productivity. Optimization of hydraulic fractures requires the integration of data from a variety of sources, and water hammer data may prove to be a useful tool in a multivariate diagnostic process. This study is based on the analyses of high-frequency treatment data, post-job reports, geological data, chemical tracer results, and production from more than 100 wells completed in the Niobrara, Codell, and Wolfcamp plays. All evaluated wells were horizontal, and included cemented and uncemented wells completed with plug and perf or ball activated frac sleeves in each stage. Treatment data were sourced, compiled, and analyzed using a new frac data management software which allowed rapid identification and evaluation of water hammer signals. For each stage the water hammer signal was characterized by taking the wave period, amplitude, duration, and decay rate. Twenty-nine wells exhibiting specific water hammer characteristics were selected for more detailed analyses to evaluate the competing influence of formation parameters, completion types, and stimulation strategies on well productivity and chemical tracer recovery. Using this dataset, we reviewed the relationship between the number of perforation clusters per stage, ball-seat size, stage and cluster spacing, reservoir quality and the observed water hammer effect and productivity. The water hammer signal was found to be affected by the entire well-fracture-reservoir system. The most substantial effects on the signal are associated with the completion type and fluid system. For multistage sleeve completions, water hammer signatures were found to be less pronounced on stages near the toe of the well. Friction and other signal restrictions in the system strongly influence the amplitude of the water hammer, which is useful to assess near wellbore connection quality, fracture initiation, and fracture placement. Chemical tracer recovery in neighboring wells identified longer or more interconnected fracture networks, which also showed higher signal decay. A high signal decay rate and duration also correlate with increased well production, which we attribute to formations properties and stimulation parameters. Analyses of water hammer data are subject to limitations including signal interference introduced by pumping rate variations, fracturing fluid composition, and low sampling frequency. It is also important to consider the large frictional pressures developed in the system during pumping, especially between the wellbore and the formation, and signal attenuation in long wellbores. Therefore, it is possible that this diagnostic approach would be more applicable in wells completed with a single fracture per stage; wherein water hammer can potentially yield a more detailed understanding of fracture dynamics. This paper provides several insights and examples that will be helpful to gather and analyze water hammer data.

No abstract

No abstract