--> Optimization of Horizontal Wells Utilizing Multivariate Analytics of Seismic Inversion in the Wolfcamp Formation of the Midland Basin

AAPG Annual Convention and Exhibition

Datapages, Inc.Print this page

Optimization of Horizontal Wells Utilizing Multivariate Analytics of Seismic Inversion in the Wolfcamp Formation of the Midland Basin

Abstract

The Midland Basin Wolfcamp play is a continuous resource play with a minimum of six independent productive zones of varying geological complexity that total in excess of 4,500 feet requires careful consideration in order to maximize value of the asset. The play is in the early stages of the development lifecycle, and provides a valuable opportunity to leverage production data from early delineation testing into a statistically valid model for predicting future horizontal production. In 2012 Laredo began testing methods of integrating geophysical, geological, petrophysical, completion, drilling, and production data to regionally high-grade drilling targets across their Midland Basin asset. Bi-variate analysis revealed that higher resolution multi-variate statistics were necessary in order to develop a local scale model that could forecast production. Utilizing multivariate statistical analytics on 82 seismic attributes (pre-stack, post-stack, and inversion), Laredo has created an “Earth Model” based on the combination of 5 seismic attributes in order to predict cumulative oil production volumes of horizontals at a set point in time. Historical well performance was then used for calibration and initial validation. The model is normalized relative to completion length, completion testing, and spacing from the calibration data to understand the variations of the drivers on production. The results of the model have undergone further refinement using microseismic, petrophysical, and empirical data, all of which ground truth the results to primary geological drivers of production. The model was confirmed utilizing a population of blind test wells which resulted in a correlation 0.85 of actual versus predicted cumulative 90-day oil production. The goal of Laredo's multivariate analytics is to improve well performance by maximizing lateral length in the most productive portions of the formation with an optimized well design and completion methodology. This technique, based on high quality 3D seismic, can be utilized in any unconventional play where good sample sizes of production and high quality geological information are available in order to high grade drilling inventories and improve individual well performance.