No abstract

A horizontal well landed in a single formation rarely encounters homogeneous rock from the heel to the toe of the wellbore. When analyzing treatment responses that occur during hydraulic fracturing, a decreasing trend in surface treating pressure in sequential stages is typically attributed to reduced friction within the casing or frac string. However, there are several variances in treating pressure that are not readily explained by examining the surface pressures and pipe friction in isolation. These variances are also apparent when looking at bottom hole injectivity. Combining surface data and geomechanical data quickly reveals the degree of variability in rock properties along a lateral and the impact that variability can have on a completion, leading to a more optimal design. This paper demonstrates how engineers can take advantage of their most detailed completions and geomechanical data by looking for trends arising from past detailed treatment analyses and applying that gained knowledge to future completions. This study relies on the analysis of proprietary high-resolution geomechanical data derived from the processing of accelerations measured at the drillbit and high-frequency fracture treatment data recorded at one-second intervals. The data were standardized to a common format, screened for quality control, normalized, and analyzed using a data management platform. The methodology combines critical mechanical rock properties such as Young's Modulus, and Poisson's ratio with high-frequency fracture treatment data, including treating pressures, rates, and fluid and proppant volumes. Further application of the geomechanical data to derive brittleness allows for construction of a more predictive petromechanical model to optimize completion approaches. A brief analysis of past completions indicated virtually no correlation between gamma ray measurements along the stage and fracture treating conditions. However, when evaluating high-resolution mechanical rock properties along the lateral, a much more useful correlation exists between minimum horizontal stress variations (calculated from Poisson's Ratio) and eventual treating pressure and proppant placement difficulties. Calculated brittleness and bottomhole injectivity (which accounts for changes in slurry rate and pipe friction) also show a relationship, especially when cluster efficiency factors are included. This study of six Eagle Ford wells suggests that rock properties are the dominant variables affecting fracture treatment pressure and bottomhole injectivity. This method can be used to predict trouble stages, improve operational efficiencies, and optimize proppant placement. This paper proposes a process to improve completion efficiency while demonstrating the value of information contained in high-resolution and high-frequency datasets. Historically underutilized, these datasets are playing an increasingly prevalent role in advanced analytics due to improved and novel technologies for data management and interpretation. This process is useful to ask better questions and to improve critical decision making with real data.

No abstract

Unexpected problems during completion create costs that can cause a well to be outside its planned AFE and even uneconomic. These problems range from merely experiencing abnormally high pressures during treatment to casing failures. The authors of this paper use machine learning methods combined with geomechanical, wellbore trajectory, and completion datasets to develop models that predict which stages will experience difficulties during completion. The operator collected geomechanical data for 26 lateral wells. Several of these wells experienced significant difficulties during completion, including casing failures. By examining the completion data for the study wells, engineers developed two objective proxies for trouble stages: a high fracture gradient cut-off and a sand-to-water ratio cut off. The geomechanical data were combined with wellbore trajectory and stage-level completion data. Machine learning techniques were then used to predict the fracture gradient at the end of treatment and the sand-to-water ratio of a stage. The data were subdivided by target horizon to ensure the model results were not skewed towards a specific horizon. The models were cross validated, hyper-parameter tuned, and then tested against a subset of the data withheld from the training algorithm (the test data subset). Multiple metrics were applied to measure the effectiveness of the models. For the fracture gradient regression, the model that minimized the root mean squared error (RMSE) was chosen. This model was compared to a prediction that each stage would have a fracture gradient equal to the average fracture gradient. The correlation between predicted and measured frac gradients was also calculated. These metrics were analyzed for both the entire data set and the test data subset. By all these measures the model performs significantly better than the control prediction. For the sand-to-water ratio metric, the model that maximized the receiver operating characteristic area under the curve (ROC-AUC or AUC) score was chosen. We chose to measure the success of the model with the precision of its prediction of outlier (trouble) stages on the test data subset. The model performs with a precision of at least 73%. Both models identify stages where casing failures occurred as trouble stages. With these initial results, the authors are confident they can predict the probability of any stage causing a significant operational disruption in these horizons. By predicting the probability of a trouble stage, mitigation plans can be implemented to reduce costs ahead of time. By further analyzing the models, engineers can identify the major drivers for the trouble stages and work to reduce them as early as the well planning stage.

No abstract