Machine learning on field data for hydraulic fracturing design optimization

Abstract

Growing amount of fracturing stimulation jobs in the recent two decades resulted in a significant amount of measured data available for construction of predictive models via machine learning (ML). Simultaneous evolution of machine learning has made it possible to apply algorithms on the hydraulic fracture database. A typical multistage fracturing job on a near-horizontal well today involves a significant number of stages. The post-fracturing production analysis (e.g., from production logging tools) reveals evidence that different stages produce very non-uniformly, and up to 30% may not be producing at all due to a combination of geomechanics and fracturing design factors. Hence, there is a significant room for fracturing design optimization, and the wealthy of field data combined with ML techniques opens a new road for solving this optimization problem. However, ML algorithms are only applicable when there is a comprehensive, well structured digital database. This paper summarizes the efforts into the creation of a digital database of field data from several thousands of multistage hydraulic fracturing jobs on near-horizontal wells from several different oilfields in Western Siberia, Russia. In terms of the number of points (fracturing jobs), the present database is a rare case of an outstandingly representative dataset of thousands of cases, compared to typical databases available in the literature, comprising tens or hundreds of pints at best. The focus is made on data gathering from various sources, data preprocessing and development of the architecture of a database as well as solving fracture design optimization via ML. We work with the database composed from reservoir properties, fracturing designs and production data. Both forward problem (prediction of production rate from fracturing design parameters) and inverse problem (selecting an optimum set of frac design parameters to maximize production) are considered. A recommendation system is designed for advising a DESC engineer on an optimized fracturing design.

Tight and Shale Oil Exploration: A Review of the Global Experience and a Case of West Siberia

“…The application of data-driven approaches for the study of West Siberian shales was dedicated mostly to the characterization of geological and petrophysical properties of tight oil reservoirs. With respect to improved and enhanced oil recovery methods, most of the research was dedicated to optimization of hydraulic fracturing techniques [145][146][147][148]. A summary of the pore-scale modeling techniques is presented in Table 5.…”

Fracture Modeling of Shale Oil and Gas Reservoirs in Texas

“…Chaikine et al reported that the developed model could predict individual well production with a 14.9% average error which can be decreased exponentially with multiple well aggregations. Other references for hydraulic fracturing design optimization are Mutalova et al [140], Wang et al [141], and Hryb et al [142]. Fracture-interaction studies have been combined with machine learning techniques, such as fracture interference [143] and automated hydraulic fracturing [144].…”